diff --git a/.gitattributes b/.gitattributes
index a6344aac8c09253b3b630fb776ae94478aa0275b..6349c3d305b7d64258177f7eb26d4f8f350590ac 100644
--- a/.gitattributes
+++ b/.gitattributes
@@ -33,3 +33,6 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
*.zip filter=lfs diff=lfs merge=lfs -text
*.zst filter=lfs diff=lfs merge=lfs -text
*tfevents* filter=lfs diff=lfs merge=lfs -text
+checkpoint-14852/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoint-29704/tokenizer.json filter=lfs diff=lfs merge=lfs -text
diff --git a/README.md b/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..c70540d6985ab5075714ffc9546a1d2b410e4d39
--- /dev/null
+++ b/README.md
@@ -0,0 +1,59 @@
+---
+library_name: transformers
+license: other
+base_model: Qwen/Qwen3-8B
+tags:
+- llama-factory
+- full
+- generated_from_trainer
+model-index:
+- name: Terminal-complete_12k
+ results: []
+---
+
+
+
+# Terminal-complete_12k
+
+This model is a fine-tuned version of [Qwen/Qwen3-8B](https://huggingface.co/Qwen/Qwen3-8B) on the qwen_data_complete dataset.
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+- learning_rate: 1e-05
+- train_batch_size: 1
+- eval_batch_size: 8
+- seed: 42
+- distributed_type: multi-GPU
+- num_devices: 4
+- total_train_batch_size: 4
+- total_eval_batch_size: 32
+- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
+- lr_scheduler_type: cosine
+- num_epochs: 2.0
+
+### Training results
+
+
+
+### Framework versions
+
+- Transformers 5.6.0
+- Pytorch 2.11.0+cu130
+- Datasets 4.0.0
+- Tokenizers 0.22.2
diff --git a/all_results.json b/all_results.json
new file mode 100644
index 0000000000000000000000000000000000000000..0fc254ffc9041351334a41c6a4b9ba8a72490a55
--- /dev/null
+++ b/all_results.json
@@ -0,0 +1,9 @@
+{
+ "epoch": 2.0,
+ "num_input_tokens_seen": 1299153424,
+ "total_flos": 2402581590032384.0,
+ "train_loss": 0.34685841939968143,
+ "train_runtime": 229818.4762,
+ "train_samples_per_second": 0.517,
+ "train_steps_per_second": 0.129
+}
\ No newline at end of file
diff --git a/chat_template.jinja b/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..01be9b307daa2d425f7c168c9fb145a286e0afb4
--- /dev/null
+++ b/chat_template.jinja
@@ -0,0 +1,89 @@
+{%- if tools %}
+ {{- '<|im_start|>system\n' }}
+ {%- if messages[0].role == 'system' %}
+ {{- messages[0].content + '\n\n' }}
+ {%- endif %}
+ {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within XML tags:\n" }}
+ {%- for tool in tools %}
+ {{- "\n" }}
+ {{- tool | tojson }}
+ {%- endfor %}
+ {{- "\n\n\nFor each function call, return a json object with function name and arguments within XML tags:\n\n{\"name\": , \"arguments\": }\n<|im_end|>\n" }}
+{%- else %}
+ {%- if messages[0].role == 'system' %}
+ {{- '<|im_start|>system\n' + messages[0].content + '<|im_end|>\n' }}
+ {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+ {%- set index = (messages|length - 1) - loop.index0 %}
+ {%- if ns.multi_step_tool and message.role == "user" and message.content is string and not(message.content.startswith('') and message.content.endswith('')) %}
+ {%- set ns.multi_step_tool = false %}
+ {%- set ns.last_query_index = index %}
+ {%- endif %}
+{%- endfor %}
+{%- for message in messages %}
+ {%- if message.content is string %}
+ {%- set content = message.content %}
+ {%- else %}
+ {%- set content = '' %}
+ {%- endif %}
+ {%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
+ {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+ {%- elif message.role == "assistant" %}
+ {%- set reasoning_content = '' %}
+ {%- if message.reasoning_content is string %}
+ {%- set reasoning_content = message.reasoning_content %}
+ {%- else %}
+ {%- if '' in content %}
+ {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %}
+ {%- set content = content.split('')[-1].lstrip('\n') %}
+ {%- endif %}
+ {%- endif %}
+ {%- if loop.index0 > ns.last_query_index %}
+ {%- if loop.last or (not loop.last and reasoning_content) %}
+ {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content.strip('\n') + '\n\n\n' + content.lstrip('\n') }}
+ {%- else %}
+ {{- '<|im_start|>' + message.role + '\n' + content }}
+ {%- endif %}
+ {%- else %}
+ {{- '<|im_start|>' + message.role + '\n' + content }}
+ {%- endif %}
+ {%- if message.tool_calls %}
+ {%- for tool_call in message.tool_calls %}
+ {%- if (loop.first and content) or (not loop.first) %}
+ {{- '\n' }}
+ {%- endif %}
+ {%- if tool_call.function %}
+ {%- set tool_call = tool_call.function %}
+ {%- endif %}
+ {{- '\n{"name": "' }}
+ {{- tool_call.name }}
+ {{- '", "arguments": ' }}
+ {%- if tool_call.arguments is string %}
+ {{- tool_call.arguments }}
+ {%- else %}
+ {{- tool_call.arguments | tojson }}
+ {%- endif %}
+ {{- '}\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+ {%- elif message.role == "tool" %}
+ {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
+ {{- '<|im_start|>user' }}
+ {%- endif %}
+ {{- '\n\n' }}
+ {{- content }}
+ {{- '\n' }}
+ {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
+ {{- '<|im_end|>\n' }}
+ {%- endif %}
+ {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+ {{- '<|im_start|>assistant\n' }}
+ {%- if enable_thinking is defined and enable_thinking is false %}
+ {{- '\n\n\n\n' }}
+ {%- endif %}
+{%- endif %}
\ No newline at end of file
diff --git a/checkpoint-14852/chat_template.jinja b/checkpoint-14852/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..01be9b307daa2d425f7c168c9fb145a286e0afb4
--- /dev/null
+++ b/checkpoint-14852/chat_template.jinja
@@ -0,0 +1,89 @@
+{%- if tools %}
+ {{- '<|im_start|>system\n' }}
+ {%- if messages[0].role == 'system' %}
+ {{- messages[0].content + '\n\n' }}
+ {%- endif %}
+ {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within XML tags:\n" }}
+ {%- for tool in tools %}
+ {{- "\n" }}
+ {{- tool | tojson }}
+ {%- endfor %}
+ {{- "\n\n\nFor each function call, return a json object with function name and arguments within XML tags:\n\n{\"name\": , \"arguments\": }\n<|im_end|>\n" }}
+{%- else %}
+ {%- if messages[0].role == 'system' %}
+ {{- '<|im_start|>system\n' + messages[0].content + '<|im_end|>\n' }}
+ {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+ {%- set index = (messages|length - 1) - loop.index0 %}
+ {%- if ns.multi_step_tool and message.role == "user" and message.content is string and not(message.content.startswith('') and message.content.endswith('')) %}
+ {%- set ns.multi_step_tool = false %}
+ {%- set ns.last_query_index = index %}
+ {%- endif %}
+{%- endfor %}
+{%- for message in messages %}
+ {%- if message.content is string %}
+ {%- set content = message.content %}
+ {%- else %}
+ {%- set content = '' %}
+ {%- endif %}
+ {%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
+ {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+ {%- elif message.role == "assistant" %}
+ {%- set reasoning_content = '' %}
+ {%- if message.reasoning_content is string %}
+ {%- set reasoning_content = message.reasoning_content %}
+ {%- else %}
+ {%- if '' in content %}
+ {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %}
+ {%- set content = content.split('')[-1].lstrip('\n') %}
+ {%- endif %}
+ {%- endif %}
+ {%- if loop.index0 > ns.last_query_index %}
+ {%- if loop.last or (not loop.last and reasoning_content) %}
+ {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content.strip('\n') + '\n\n\n' + content.lstrip('\n') }}
+ {%- else %}
+ {{- '<|im_start|>' + message.role + '\n' + content }}
+ {%- endif %}
+ {%- else %}
+ {{- '<|im_start|>' + message.role + '\n' + content }}
+ {%- endif %}
+ {%- if message.tool_calls %}
+ {%- for tool_call in message.tool_calls %}
+ {%- if (loop.first and content) or (not loop.first) %}
+ {{- '\n' }}
+ {%- endif %}
+ {%- if tool_call.function %}
+ {%- set tool_call = tool_call.function %}
+ {%- endif %}
+ {{- '\n{"name": "' }}
+ {{- tool_call.name }}
+ {{- '", "arguments": ' }}
+ {%- if tool_call.arguments is string %}
+ {{- tool_call.arguments }}
+ {%- else %}
+ {{- tool_call.arguments | tojson }}
+ {%- endif %}
+ {{- '}\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+ {%- elif message.role == "tool" %}
+ {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
+ {{- '<|im_start|>user' }}
+ {%- endif %}
+ {{- '\n\n' }}
+ {{- content }}
+ {{- '\n' }}
+ {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
+ {{- '<|im_end|>\n' }}
+ {%- endif %}
+ {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+ {{- '<|im_start|>assistant\n' }}
+ {%- if enable_thinking is defined and enable_thinking is false %}
+ {{- '\n\n\n\n' }}
+ {%- endif %}
+{%- endif %}
\ No newline at end of file
diff --git a/checkpoint-14852/config.json b/checkpoint-14852/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..746f56efdea311c4c5e737f2e31814c30cc1e1bf
--- /dev/null
+++ b/checkpoint-14852/config.json
@@ -0,0 +1,71 @@
+{
+ "architectures": [
+ "Qwen3ForCausalLM"
+ ],
+ "attention_bias": false,
+ "attention_dropout": 0.0,
+ "bos_token_id": null,
+ "dtype": "bfloat16",
+ "eos_token_id": 151645,
+ "head_dim": 128,
+ "hidden_act": "silu",
+ "hidden_size": 4096,
+ "initializer_range": 0.02,
+ "intermediate_size": 12288,
+ "layer_types": [
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention"
+ ],
+ "max_position_embeddings": 40960,
+ "max_window_layers": 36,
+ "model_type": "qwen3",
+ "num_attention_heads": 32,
+ "num_hidden_layers": 36,
+ "num_key_value_heads": 8,
+ "pad_token_id": 151643,
+ "rms_norm_eps": 1e-06,
+ "rope_parameters": {
+ "rope_theta": 1000000,
+ "rope_type": "default"
+ },
+ "sliding_window": null,
+ "tie_word_embeddings": false,
+ "transformers_version": "5.6.0",
+ "use_cache": false,
+ "use_sliding_window": false,
+ "vocab_size": 151936
+}
diff --git a/checkpoint-14852/generation_config.json b/checkpoint-14852/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..cf5bf49b6941f1673361724230bb69080cd3f1e2
--- /dev/null
+++ b/checkpoint-14852/generation_config.json
@@ -0,0 +1,12 @@
+{
+ "do_sample": true,
+ "eos_token_id": [
+ 151645,
+ 151643
+ ],
+ "pad_token_id": 151643,
+ "temperature": 0.6,
+ "top_k": 20,
+ "top_p": 0.95,
+ "transformers_version": "5.6.0"
+}
diff --git a/checkpoint-14852/global_step14852/zero_pp_rank_0_mp_rank_00_model_states.pt b/checkpoint-14852/global_step14852/zero_pp_rank_0_mp_rank_00_model_states.pt
new file mode 100644
index 0000000000000000000000000000000000000000..a05f55c749c6c15509bfe58303617d4288ef6000
--- /dev/null
+++ b/checkpoint-14852/global_step14852/zero_pp_rank_0_mp_rank_00_model_states.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:39c79688b0d151d9262ae7b80f5a513983fadb9ca1d104424143862c551c7196
+size 202796
diff --git a/checkpoint-14852/global_step14852/zero_pp_rank_1_mp_rank_00_model_states.pt b/checkpoint-14852/global_step14852/zero_pp_rank_1_mp_rank_00_model_states.pt
new file mode 100644
index 0000000000000000000000000000000000000000..172a58d356b3065633cbf1666c8a2ecff812e5fd
--- /dev/null
+++ b/checkpoint-14852/global_step14852/zero_pp_rank_1_mp_rank_00_model_states.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3905ed44486f1f2c3a02e8219bd3875969f7c093c5a26dad0b06c6907d3f3e49
+size 202796
diff --git a/checkpoint-14852/global_step14852/zero_pp_rank_2_mp_rank_00_model_states.pt b/checkpoint-14852/global_step14852/zero_pp_rank_2_mp_rank_00_model_states.pt
new file mode 100644
index 0000000000000000000000000000000000000000..a53a6b8fbda1c0ea2d35c080a776427542c20201
--- /dev/null
+++ b/checkpoint-14852/global_step14852/zero_pp_rank_2_mp_rank_00_model_states.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1f4e53ba4e11ee63d61a62c878144353c2af1b4067d146a04f96934584e9c6f2
+size 202796
diff --git a/checkpoint-14852/global_step14852/zero_pp_rank_3_mp_rank_00_model_states.pt b/checkpoint-14852/global_step14852/zero_pp_rank_3_mp_rank_00_model_states.pt
new file mode 100644
index 0000000000000000000000000000000000000000..8cf69af6f9ce78f79b11d7e1ed069de49b48bd22
--- /dev/null
+++ b/checkpoint-14852/global_step14852/zero_pp_rank_3_mp_rank_00_model_states.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:43af40d07516f63fe9e29bed0da072a3e999c5f5ebeb9649b9d563297872304c
+size 202796
diff --git a/checkpoint-14852/latest b/checkpoint-14852/latest
new file mode 100644
index 0000000000000000000000000000000000000000..f3f629cd7e796c0d2bc5e8bc260c3451a3607ab7
--- /dev/null
+++ b/checkpoint-14852/latest
@@ -0,0 +1 @@
+global_step14852
\ No newline at end of file
diff --git a/checkpoint-14852/rng_state_0.pth b/checkpoint-14852/rng_state_0.pth
new file mode 100644
index 0000000000000000000000000000000000000000..26082348a981197df0510cae3337c958e26fcb6b
--- /dev/null
+++ b/checkpoint-14852/rng_state_0.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:659b1cdee2219458dd84ce6a632a595465680b8080e5c44bd600ff97eca8d752
+size 15429
diff --git a/checkpoint-14852/rng_state_1.pth b/checkpoint-14852/rng_state_1.pth
new file mode 100644
index 0000000000000000000000000000000000000000..d46ce0452945f0c32849661b844e8c0bb9a24957
--- /dev/null
+++ b/checkpoint-14852/rng_state_1.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:86accf27064cdd503053e90476a6bd10de333d4ff0594535ad55ea13a473c91d
+size 15429
diff --git a/checkpoint-14852/rng_state_2.pth b/checkpoint-14852/rng_state_2.pth
new file mode 100644
index 0000000000000000000000000000000000000000..558429d43da9b21fe074ba126896adb41ea605cb
--- /dev/null
+++ b/checkpoint-14852/rng_state_2.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:18ca8d714ef40be035404c1957b5a4dee84e1f43980408393f8aa710552ee6f6
+size 15429
diff --git a/checkpoint-14852/rng_state_3.pth b/checkpoint-14852/rng_state_3.pth
new file mode 100644
index 0000000000000000000000000000000000000000..4e54cbec0a57bff8415c48ea238e3281454ea9b4
--- /dev/null
+++ b/checkpoint-14852/rng_state_3.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2cfdebe99e40accc9c9d8f09c63136a14abda997d9b501969ec8e16e9d183179
+size 15429
diff --git a/checkpoint-14852/scheduler.pt b/checkpoint-14852/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..4ea6b2c166a0893f3cf6cb194853fe8afe4c0ead
--- /dev/null
+++ b/checkpoint-14852/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dacfed8f106382a6b458226636cb6eedcc74a56308e93b5dc42b418ad9209395
+size 1465
diff --git a/checkpoint-14852/tokenizer.json b/checkpoint-14852/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..c7afbed2efcdf019f88ab0572ec29d3bf595dfe2
--- /dev/null
+++ b/checkpoint-14852/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be75606093db2094d7cd20f3c2f385c212750648bd6ea4fb2bf507a6a4c55506
+size 11422650
diff --git a/checkpoint-14852/tokenizer_config.json b/checkpoint-14852/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..6623da03457d79d2cd4af9557e9f31407cdb835c
--- /dev/null
+++ b/checkpoint-14852/tokenizer_config.json
@@ -0,0 +1,31 @@
+{
+ "add_prefix_space": false,
+ "backend": "tokenizers",
+ "bos_token": null,
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "<|im_end|>",
+ "errors": "replace",
+ "extra_special_tokens": [
+ "<|im_start|>",
+ "<|im_end|>",
+ "<|object_ref_start|>",
+ "<|object_ref_end|>",
+ "<|box_start|>",
+ "<|box_end|>",
+ "<|quad_start|>",
+ "<|quad_end|>",
+ "<|vision_start|>",
+ "<|vision_end|>",
+ "<|vision_pad|>",
+ "<|image_pad|>",
+ "<|video_pad|>"
+ ],
+ "is_local": false,
+ "local_files_only": false,
+ "model_max_length": 131072,
+ "pad_token": "<|endoftext|>",
+ "padding_side": "right",
+ "split_special_tokens": false,
+ "tokenizer_class": "Qwen2Tokenizer",
+ "unk_token": null
+}
diff --git a/checkpoint-14852/trainer_state.json b/checkpoint-14852/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..90f1d3458fdd0ceedf07aab49cbde8cd4ad58fea
--- /dev/null
+++ b/checkpoint-14852/trainer_state.json
@@ -0,0 +1,29734 @@
+{
+ "best_global_step": null,
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 1.0,
+ "eval_steps": 500,
+ "global_step": 14852,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.00033665499596014,
+ "grad_norm": 1.7158506568336664,
+ "learning_rate": 9.999999552565099e-06,
+ "loss": 0.8321187973022461,
+ "num_input_tokens_seen": 213976,
+ "step": 5,
+ "train_runtime": 48.5386,
+ "train_tokens_per_second": 4408.369
+ },
+ {
+ "epoch": 0.00067330999192028,
+ "grad_norm": 1.4989508496023307,
+ "learning_rate": 9.999997734860948e-06,
+ "loss": 0.6714894294738769,
+ "num_input_tokens_seen": 418192,
+ "step": 10,
+ "train_runtime": 79.1759,
+ "train_tokens_per_second": 5281.811
+ },
+ {
+ "epoch": 0.0010099649878804202,
+ "grad_norm": 1.3987943897146944,
+ "learning_rate": 9.999994518923374e-06,
+ "loss": 0.6414656162261962,
+ "num_input_tokens_seen": 631760,
+ "step": 15,
+ "train_runtime": 116.8887,
+ "train_tokens_per_second": 5404.799
+ },
+ {
+ "epoch": 0.00134661998384056,
+ "grad_norm": 1.3594234171992137,
+ "learning_rate": 9.999989904753278e-06,
+ "loss": 0.6119038105010987,
+ "num_input_tokens_seen": 847240,
+ "step": 20,
+ "train_runtime": 160.5101,
+ "train_tokens_per_second": 5278.422
+ },
+ {
+ "epoch": 0.0016832749798007002,
+ "grad_norm": 1.1837151081821848,
+ "learning_rate": 9.999983892351947e-06,
+ "loss": 0.5918544292449951,
+ "num_input_tokens_seen": 1067560,
+ "step": 25,
+ "train_runtime": 197.4905,
+ "train_tokens_per_second": 5405.628
+ },
+ {
+ "epoch": 0.0020199299757608404,
+ "grad_norm": 1.1504621450496095,
+ "learning_rate": 9.999976481721066e-06,
+ "loss": 0.5148830890655518,
+ "num_input_tokens_seen": 1288760,
+ "step": 30,
+ "train_runtime": 231.3357,
+ "train_tokens_per_second": 5570.951
+ },
+ {
+ "epoch": 0.0023565849717209803,
+ "grad_norm": 1.0663060499576293,
+ "learning_rate": 9.999967672862704e-06,
+ "loss": 0.554470157623291,
+ "num_input_tokens_seen": 1501896,
+ "step": 35,
+ "train_runtime": 271.9258,
+ "train_tokens_per_second": 5523.183
+ },
+ {
+ "epoch": 0.00269323996768112,
+ "grad_norm": 1.1179858029742544,
+ "learning_rate": 9.999957465779326e-06,
+ "loss": 0.5239121437072753,
+ "num_input_tokens_seen": 1716792,
+ "step": 40,
+ "train_runtime": 319.9491,
+ "train_tokens_per_second": 5365.829
+ },
+ {
+ "epoch": 0.0030298949636412605,
+ "grad_norm": 1.1555971967505698,
+ "learning_rate": 9.999945860473788e-06,
+ "loss": 0.5400435924530029,
+ "num_input_tokens_seen": 1920760,
+ "step": 45,
+ "train_runtime": 364.1591,
+ "train_tokens_per_second": 5274.507
+ },
+ {
+ "epoch": 0.0033665499596014004,
+ "grad_norm": 0.9321994843580264,
+ "learning_rate": 9.999932856949333e-06,
+ "loss": 0.5238595008850098,
+ "num_input_tokens_seen": 2140344,
+ "step": 50,
+ "train_runtime": 402.577,
+ "train_tokens_per_second": 5316.608
+ },
+ {
+ "epoch": 0.0037032049555615404,
+ "grad_norm": 1.1572766450172172,
+ "learning_rate": 9.999918455209599e-06,
+ "loss": 0.5059286117553711,
+ "num_input_tokens_seen": 2368296,
+ "step": 55,
+ "train_runtime": 438.0431,
+ "train_tokens_per_second": 5406.536
+ },
+ {
+ "epoch": 0.004039859951521681,
+ "grad_norm": 1.1277411399750283,
+ "learning_rate": 9.99990265525861e-06,
+ "loss": 0.5235630989074707,
+ "num_input_tokens_seen": 2589544,
+ "step": 60,
+ "train_runtime": 475.1915,
+ "train_tokens_per_second": 5449.474
+ },
+ {
+ "epoch": 0.004376514947481821,
+ "grad_norm": 1.233348290191874,
+ "learning_rate": 9.999885457100789e-06,
+ "loss": 0.5595571517944335,
+ "num_input_tokens_seen": 2809432,
+ "step": 65,
+ "train_runtime": 503.8866,
+ "train_tokens_per_second": 5575.524
+ },
+ {
+ "epoch": 0.0047131699434419605,
+ "grad_norm": 1.2298135105742523,
+ "learning_rate": 9.999866860740943e-06,
+ "loss": 0.5092310428619384,
+ "num_input_tokens_seen": 3020248,
+ "step": 70,
+ "train_runtime": 543.5456,
+ "train_tokens_per_second": 5556.568
+ },
+ {
+ "epoch": 0.005049824939402101,
+ "grad_norm": 0.9679470345624956,
+ "learning_rate": 9.999846866184274e-06,
+ "loss": 0.5114652633666992,
+ "num_input_tokens_seen": 3249600,
+ "step": 75,
+ "train_runtime": 585.283,
+ "train_tokens_per_second": 5552.186
+ },
+ {
+ "epoch": 0.00538647993536224,
+ "grad_norm": 0.979469640417179,
+ "learning_rate": 9.99982547343637e-06,
+ "loss": 0.5040130615234375,
+ "num_input_tokens_seen": 3443864,
+ "step": 80,
+ "train_runtime": 626.1684,
+ "train_tokens_per_second": 5499.901
+ },
+ {
+ "epoch": 0.005723134931322381,
+ "grad_norm": 1.078786163056917,
+ "learning_rate": 9.999802682503217e-06,
+ "loss": 0.4742586612701416,
+ "num_input_tokens_seen": 3661032,
+ "step": 85,
+ "train_runtime": 663.7311,
+ "train_tokens_per_second": 5515.836
+ },
+ {
+ "epoch": 0.006059789927282521,
+ "grad_norm": 0.9113383787124433,
+ "learning_rate": 9.999778493391186e-06,
+ "loss": 0.5044189453125,
+ "num_input_tokens_seen": 3888680,
+ "step": 90,
+ "train_runtime": 707.589,
+ "train_tokens_per_second": 5495.676
+ },
+ {
+ "epoch": 0.0063964449232426605,
+ "grad_norm": 1.0192977781689194,
+ "learning_rate": 9.999752906107043e-06,
+ "loss": 0.5348274230957031,
+ "num_input_tokens_seen": 4114248,
+ "step": 95,
+ "train_runtime": 744.4001,
+ "train_tokens_per_second": 5526.931
+ },
+ {
+ "epoch": 0.006733099919202801,
+ "grad_norm": 1.046539367979976,
+ "learning_rate": 9.999725920657944e-06,
+ "loss": 0.5229468822479248,
+ "num_input_tokens_seen": 4329256,
+ "step": 100,
+ "train_runtime": 790.5278,
+ "train_tokens_per_second": 5476.412
+ },
+ {
+ "epoch": 0.007069754915162941,
+ "grad_norm": 1.0823101637799084,
+ "learning_rate": 9.999697537051432e-06,
+ "loss": 0.4990852355957031,
+ "num_input_tokens_seen": 4533832,
+ "step": 105,
+ "train_runtime": 828.8033,
+ "train_tokens_per_second": 5470.335
+ },
+ {
+ "epoch": 0.007406409911123081,
+ "grad_norm": 1.0579674272002448,
+ "learning_rate": 9.999667755295449e-06,
+ "loss": 0.48107209205627444,
+ "num_input_tokens_seen": 4749088,
+ "step": 110,
+ "train_runtime": 864.874,
+ "train_tokens_per_second": 5491.075
+ },
+ {
+ "epoch": 0.007743064907083221,
+ "grad_norm": 1.2773224732968154,
+ "learning_rate": 9.999636575398319e-06,
+ "loss": 0.4919854164123535,
+ "num_input_tokens_seen": 4960752,
+ "step": 115,
+ "train_runtime": 904.8595,
+ "train_tokens_per_second": 5482.345
+ },
+ {
+ "epoch": 0.008079719903043361,
+ "grad_norm": 1.0488778079229073,
+ "learning_rate": 9.999603997368766e-06,
+ "loss": 0.4375304222106934,
+ "num_input_tokens_seen": 5167864,
+ "step": 120,
+ "train_runtime": 939.3678,
+ "train_tokens_per_second": 5501.428
+ },
+ {
+ "epoch": 0.008416374899003501,
+ "grad_norm": 1.1229102076855217,
+ "learning_rate": 9.999570021215895e-06,
+ "loss": 0.4794501304626465,
+ "num_input_tokens_seen": 5384648,
+ "step": 125,
+ "train_runtime": 973.7305,
+ "train_tokens_per_second": 5529.916
+ },
+ {
+ "epoch": 0.008753029894963642,
+ "grad_norm": 1.0387093548265027,
+ "learning_rate": 9.99953464694921e-06,
+ "loss": 0.5004319667816162,
+ "num_input_tokens_seen": 5590392,
+ "step": 130,
+ "train_runtime": 1013.6767,
+ "train_tokens_per_second": 5514.965
+ },
+ {
+ "epoch": 0.009089684890923782,
+ "grad_norm": 1.0508132347808834,
+ "learning_rate": 9.999497874578606e-06,
+ "loss": 0.495189905166626,
+ "num_input_tokens_seen": 5819224,
+ "step": 135,
+ "train_runtime": 1052.6891,
+ "train_tokens_per_second": 5527.961
+ },
+ {
+ "epoch": 0.009426339886883921,
+ "grad_norm": 1.0896002910221911,
+ "learning_rate": 9.999459704114361e-06,
+ "loss": 0.5080117225646973,
+ "num_input_tokens_seen": 6037808,
+ "step": 140,
+ "train_runtime": 1094.5671,
+ "train_tokens_per_second": 5516.161
+ },
+ {
+ "epoch": 0.00976299488284406,
+ "grad_norm": 1.034914657869065,
+ "learning_rate": 9.999420135567153e-06,
+ "loss": 0.526894474029541,
+ "num_input_tokens_seen": 6252752,
+ "step": 145,
+ "train_runtime": 1135.4311,
+ "train_tokens_per_second": 5506.941
+ },
+ {
+ "epoch": 0.010099649878804202,
+ "grad_norm": 1.164626155534281,
+ "learning_rate": 9.999379168948047e-06,
+ "loss": 0.5105450630187989,
+ "num_input_tokens_seen": 6476720,
+ "step": 150,
+ "train_runtime": 1172.7989,
+ "train_tokens_per_second": 5522.447
+ },
+ {
+ "epoch": 0.010436304874764341,
+ "grad_norm": 1.2400950788317526,
+ "learning_rate": 9.999336804268496e-06,
+ "loss": 0.47139749526977537,
+ "num_input_tokens_seen": 6698024,
+ "step": 155,
+ "train_runtime": 1217.419,
+ "train_tokens_per_second": 5501.823
+ },
+ {
+ "epoch": 0.01077295987072448,
+ "grad_norm": 1.0587023212526414,
+ "learning_rate": 9.999293041540351e-06,
+ "loss": 0.5010284900665283,
+ "num_input_tokens_seen": 6910784,
+ "step": 160,
+ "train_runtime": 1258.4852,
+ "train_tokens_per_second": 5491.351
+ },
+ {
+ "epoch": 0.011109614866684622,
+ "grad_norm": 1.0126894215000632,
+ "learning_rate": 9.999247880775849e-06,
+ "loss": 0.49982900619506837,
+ "num_input_tokens_seen": 7115568,
+ "step": 165,
+ "train_runtime": 1295.6174,
+ "train_tokens_per_second": 5492.029
+ },
+ {
+ "epoch": 0.011446269862644761,
+ "grad_norm": 1.198929638798561,
+ "learning_rate": 9.999201321987617e-06,
+ "loss": 0.46059427261352537,
+ "num_input_tokens_seen": 7329488,
+ "step": 170,
+ "train_runtime": 1337.1094,
+ "train_tokens_per_second": 5481.592
+ },
+ {
+ "epoch": 0.011782924858604901,
+ "grad_norm": 1.0114658976479987,
+ "learning_rate": 9.999153365188676e-06,
+ "loss": 0.5039016723632812,
+ "num_input_tokens_seen": 7547880,
+ "step": 175,
+ "train_runtime": 1383.6734,
+ "train_tokens_per_second": 5454.958
+ },
+ {
+ "epoch": 0.012119579854565042,
+ "grad_norm": 1.0579154349493107,
+ "learning_rate": 9.99910401039244e-06,
+ "loss": 0.4546662330627441,
+ "num_input_tokens_seen": 7757728,
+ "step": 180,
+ "train_runtime": 1424.6608,
+ "train_tokens_per_second": 5445.316
+ },
+ {
+ "epoch": 0.012456234850525182,
+ "grad_norm": 1.1327328565153654,
+ "learning_rate": 9.999053257612708e-06,
+ "loss": 0.464589786529541,
+ "num_input_tokens_seen": 7984416,
+ "step": 185,
+ "train_runtime": 1468.0361,
+ "train_tokens_per_second": 5438.842
+ },
+ {
+ "epoch": 0.012792889846485321,
+ "grad_norm": 1.0695885705791155,
+ "learning_rate": 9.999001106863672e-06,
+ "loss": 0.5309792518615722,
+ "num_input_tokens_seen": 8204976,
+ "step": 190,
+ "train_runtime": 1506.2271,
+ "train_tokens_per_second": 5447.37
+ },
+ {
+ "epoch": 0.013129544842445462,
+ "grad_norm": 1.129268767285916,
+ "learning_rate": 9.998947558159916e-06,
+ "loss": 0.4964266300201416,
+ "num_input_tokens_seen": 8428800,
+ "step": 195,
+ "train_runtime": 1546.2799,
+ "train_tokens_per_second": 5451.018
+ },
+ {
+ "epoch": 0.013466199838405602,
+ "grad_norm": 1.0645176443218387,
+ "learning_rate": 9.99889261151642e-06,
+ "loss": 0.44077634811401367,
+ "num_input_tokens_seen": 8642192,
+ "step": 200,
+ "train_runtime": 1586.0871,
+ "train_tokens_per_second": 5448.75
+ },
+ {
+ "epoch": 0.013802854834365741,
+ "grad_norm": 1.0273158071004191,
+ "learning_rate": 9.998836266948542e-06,
+ "loss": 0.47046546936035155,
+ "num_input_tokens_seen": 8852840,
+ "step": 205,
+ "train_runtime": 1618.9934,
+ "train_tokens_per_second": 5468.114
+ },
+ {
+ "epoch": 0.014139509830325882,
+ "grad_norm": 1.2260591863245542,
+ "learning_rate": 9.998778524472043e-06,
+ "loss": 0.4429631233215332,
+ "num_input_tokens_seen": 9059368,
+ "step": 210,
+ "train_runtime": 1658.2941,
+ "train_tokens_per_second": 5463.065
+ },
+ {
+ "epoch": 0.014476164826286022,
+ "grad_norm": 1.1035215533216065,
+ "learning_rate": 9.99871938410307e-06,
+ "loss": 0.46187715530395507,
+ "num_input_tokens_seen": 9273936,
+ "step": 215,
+ "train_runtime": 1690.8398,
+ "train_tokens_per_second": 5484.811
+ },
+ {
+ "epoch": 0.014812819822246161,
+ "grad_norm": 1.500582938758766,
+ "learning_rate": 9.99865884585816e-06,
+ "loss": 0.5123334407806397,
+ "num_input_tokens_seen": 9489816,
+ "step": 220,
+ "train_runtime": 1724.7509,
+ "train_tokens_per_second": 5502.137
+ },
+ {
+ "epoch": 0.015149474818206303,
+ "grad_norm": 1.1104716846300207,
+ "learning_rate": 9.998596909754245e-06,
+ "loss": 0.4839674949645996,
+ "num_input_tokens_seen": 9710720,
+ "step": 225,
+ "train_runtime": 1769.0601,
+ "train_tokens_per_second": 5489.197
+ },
+ {
+ "epoch": 0.015486129814166442,
+ "grad_norm": 1.1763018457719354,
+ "learning_rate": 9.998533575808642e-06,
+ "loss": 0.455305290222168,
+ "num_input_tokens_seen": 9922160,
+ "step": 230,
+ "train_runtime": 1804.4173,
+ "train_tokens_per_second": 5498.817
+ },
+ {
+ "epoch": 0.015822784810126583,
+ "grad_norm": 0.9508268973242634,
+ "learning_rate": 9.998468844039064e-06,
+ "loss": 0.4588268280029297,
+ "num_input_tokens_seen": 10132480,
+ "step": 235,
+ "train_runtime": 1843.1618,
+ "train_tokens_per_second": 5497.336
+ },
+ {
+ "epoch": 0.016159439806086723,
+ "grad_norm": 0.982580320370019,
+ "learning_rate": 9.998402714463613e-06,
+ "loss": 0.49381322860717775,
+ "num_input_tokens_seen": 10356032,
+ "step": 240,
+ "train_runtime": 1886.9781,
+ "train_tokens_per_second": 5488.157
+ },
+ {
+ "epoch": 0.016496094802046862,
+ "grad_norm": 0.9967583322638394,
+ "learning_rate": 9.998335187100782e-06,
+ "loss": 0.4929532051086426,
+ "num_input_tokens_seen": 10571688,
+ "step": 245,
+ "train_runtime": 1931.0773,
+ "train_tokens_per_second": 5474.503
+ },
+ {
+ "epoch": 0.016832749798007002,
+ "grad_norm": 1.0005940352447382,
+ "learning_rate": 9.998266261969453e-06,
+ "loss": 0.488722038269043,
+ "num_input_tokens_seen": 10787736,
+ "step": 250,
+ "train_runtime": 1967.3657,
+ "train_tokens_per_second": 5483.341
+ },
+ {
+ "epoch": 0.01716940479396714,
+ "grad_norm": 0.9740927827374989,
+ "learning_rate": 9.998195939088905e-06,
+ "loss": 0.46301860809326173,
+ "num_input_tokens_seen": 11014240,
+ "step": 255,
+ "train_runtime": 1995.3214,
+ "train_tokens_per_second": 5520.033
+ },
+ {
+ "epoch": 0.017506059789927284,
+ "grad_norm": 1.0005631289772583,
+ "learning_rate": 9.998124218478797e-06,
+ "loss": 0.4852250099182129,
+ "num_input_tokens_seen": 11227928,
+ "step": 260,
+ "train_runtime": 2040.8305,
+ "train_tokens_per_second": 5501.646
+ },
+ {
+ "epoch": 0.017842714785887424,
+ "grad_norm": 3.2898314540824733,
+ "learning_rate": 9.998051100159192e-06,
+ "loss": 0.4806126594543457,
+ "num_input_tokens_seen": 11457336,
+ "step": 265,
+ "train_runtime": 2073.3223,
+ "train_tokens_per_second": 5526.076
+ },
+ {
+ "epoch": 0.018179369781847563,
+ "grad_norm": 1.0137684438609407,
+ "learning_rate": 9.997976584150533e-06,
+ "loss": 0.4908778667449951,
+ "num_input_tokens_seen": 11669408,
+ "step": 270,
+ "train_runtime": 2105.6176,
+ "train_tokens_per_second": 5542.036
+ },
+ {
+ "epoch": 0.018516024777807703,
+ "grad_norm": 0.882254713488144,
+ "learning_rate": 9.997900670473662e-06,
+ "loss": 0.4400303840637207,
+ "num_input_tokens_seen": 11900376,
+ "step": 275,
+ "train_runtime": 2145.8648,
+ "train_tokens_per_second": 5545.725
+ },
+ {
+ "epoch": 0.018852679773767842,
+ "grad_norm": 1.0103112069849245,
+ "learning_rate": 9.997823359149804e-06,
+ "loss": 0.47128896713256835,
+ "num_input_tokens_seen": 12115296,
+ "step": 280,
+ "train_runtime": 2180.0754,
+ "train_tokens_per_second": 5557.283
+ },
+ {
+ "epoch": 0.01918933476972798,
+ "grad_norm": 2.383562470887594,
+ "learning_rate": 9.997744650200579e-06,
+ "loss": 0.5039300441741943,
+ "num_input_tokens_seen": 12335296,
+ "step": 285,
+ "train_runtime": 2224.7672,
+ "train_tokens_per_second": 5544.533
+ },
+ {
+ "epoch": 0.01952598976568812,
+ "grad_norm": 1.012511096525962,
+ "learning_rate": 9.997664543648e-06,
+ "loss": 0.47290778160095215,
+ "num_input_tokens_seen": 12550016,
+ "step": 290,
+ "train_runtime": 2269.2912,
+ "train_tokens_per_second": 5530.368
+ },
+ {
+ "epoch": 0.019862644761648264,
+ "grad_norm": 0.9200150844694661,
+ "learning_rate": 9.997583039514469e-06,
+ "loss": 0.47549943923950194,
+ "num_input_tokens_seen": 12769944,
+ "step": 295,
+ "train_runtime": 2310.0213,
+ "train_tokens_per_second": 5528.063
+ },
+ {
+ "epoch": 0.020199299757608404,
+ "grad_norm": 1.0603122997620704,
+ "learning_rate": 9.997500137822776e-06,
+ "loss": 0.48865842819213867,
+ "num_input_tokens_seen": 12994152,
+ "step": 300,
+ "train_runtime": 2347.3612,
+ "train_tokens_per_second": 5535.642
+ },
+ {
+ "epoch": 0.020535954753568543,
+ "grad_norm": 1.1560224095031963,
+ "learning_rate": 9.997415838596106e-06,
+ "loss": 0.46552267074584963,
+ "num_input_tokens_seen": 13206104,
+ "step": 305,
+ "train_runtime": 2390.8704,
+ "train_tokens_per_second": 5523.555
+ },
+ {
+ "epoch": 0.020872609749528682,
+ "grad_norm": 0.9496962272740035,
+ "learning_rate": 9.997330141858032e-06,
+ "loss": 0.4945693492889404,
+ "num_input_tokens_seen": 13423544,
+ "step": 310,
+ "train_runtime": 2429.9223,
+ "train_tokens_per_second": 5524.269
+ },
+ {
+ "epoch": 0.021209264745488822,
+ "grad_norm": 1.1164280869632712,
+ "learning_rate": 9.997243047632519e-06,
+ "loss": 0.47294368743896487,
+ "num_input_tokens_seen": 13640584,
+ "step": 315,
+ "train_runtime": 2463.856,
+ "train_tokens_per_second": 5536.275
+ },
+ {
+ "epoch": 0.02154591974144896,
+ "grad_norm": 0.9211260067991657,
+ "learning_rate": 9.997154555943923e-06,
+ "loss": 0.43665332794189454,
+ "num_input_tokens_seen": 13858888,
+ "step": 320,
+ "train_runtime": 2506.1,
+ "train_tokens_per_second": 5530.062
+ },
+ {
+ "epoch": 0.021882574737409104,
+ "grad_norm": 1.0719268322873283,
+ "learning_rate": 9.997064666816992e-06,
+ "loss": 0.4625236034393311,
+ "num_input_tokens_seen": 14076648,
+ "step": 325,
+ "train_runtime": 2544.8417,
+ "train_tokens_per_second": 5531.443
+ },
+ {
+ "epoch": 0.022219229733369244,
+ "grad_norm": 1.146001730574971,
+ "learning_rate": 9.996973380276858e-06,
+ "loss": 0.5440494060516358,
+ "num_input_tokens_seen": 14299864,
+ "step": 330,
+ "train_runtime": 2575.9234,
+ "train_tokens_per_second": 5551.355
+ },
+ {
+ "epoch": 0.022555884729329383,
+ "grad_norm": 1.1489261273369993,
+ "learning_rate": 9.996880696349053e-06,
+ "loss": 0.45033721923828124,
+ "num_input_tokens_seen": 14522560,
+ "step": 335,
+ "train_runtime": 2614.1655,
+ "train_tokens_per_second": 5555.333
+ },
+ {
+ "epoch": 0.022892539725289523,
+ "grad_norm": 1.0606322220744733,
+ "learning_rate": 9.996786615059497e-06,
+ "loss": 0.5329171180725097,
+ "num_input_tokens_seen": 14746648,
+ "step": 340,
+ "train_runtime": 2655.6018,
+ "train_tokens_per_second": 5553.034
+ },
+ {
+ "epoch": 0.023229194721249662,
+ "grad_norm": 0.9692502297767218,
+ "learning_rate": 9.996691136434498e-06,
+ "loss": 0.4342483997344971,
+ "num_input_tokens_seen": 14969080,
+ "step": 345,
+ "train_runtime": 2690.8338,
+ "train_tokens_per_second": 5562.989
+ },
+ {
+ "epoch": 0.023565849717209802,
+ "grad_norm": 0.8994912857959015,
+ "learning_rate": 9.996594260500753e-06,
+ "loss": 0.507387924194336,
+ "num_input_tokens_seen": 15193912,
+ "step": 350,
+ "train_runtime": 2729.4748,
+ "train_tokens_per_second": 5566.606
+ },
+ {
+ "epoch": 0.023902504713169945,
+ "grad_norm": 0.9930018613669586,
+ "learning_rate": 9.996495987285358e-06,
+ "loss": 0.4612261772155762,
+ "num_input_tokens_seen": 15411408,
+ "step": 355,
+ "train_runtime": 2773.0354,
+ "train_tokens_per_second": 5557.595
+ },
+ {
+ "epoch": 0.024239159709130084,
+ "grad_norm": 1.0882060582433957,
+ "learning_rate": 9.996396316815793e-06,
+ "loss": 0.4673755645751953,
+ "num_input_tokens_seen": 15630480,
+ "step": 360,
+ "train_runtime": 2815.6019,
+ "train_tokens_per_second": 5551.381
+ },
+ {
+ "epoch": 0.024575814705090224,
+ "grad_norm": 1.1223066651934839,
+ "learning_rate": 9.99629524911993e-06,
+ "loss": 0.4566653251647949,
+ "num_input_tokens_seen": 15847552,
+ "step": 365,
+ "train_runtime": 2852.9783,
+ "train_tokens_per_second": 5554.74
+ },
+ {
+ "epoch": 0.024912469701050363,
+ "grad_norm": 1.0964642711464192,
+ "learning_rate": 9.996192784226033e-06,
+ "loss": 0.440608549118042,
+ "num_input_tokens_seen": 16049488,
+ "step": 370,
+ "train_runtime": 2887.3093,
+ "train_tokens_per_second": 5558.631
+ },
+ {
+ "epoch": 0.025249124697010503,
+ "grad_norm": 0.9458333109972462,
+ "learning_rate": 9.996088922162755e-06,
+ "loss": 0.4601402759552002,
+ "num_input_tokens_seen": 16283544,
+ "step": 375,
+ "train_runtime": 2920.1965,
+ "train_tokens_per_second": 5576.181
+ },
+ {
+ "epoch": 0.025585779692970642,
+ "grad_norm": 0.9639187009881571,
+ "learning_rate": 9.995983662959141e-06,
+ "loss": 0.4678368091583252,
+ "num_input_tokens_seen": 16507176,
+ "step": 380,
+ "train_runtime": 2959.819,
+ "train_tokens_per_second": 5577.09
+ },
+ {
+ "epoch": 0.025922434688930785,
+ "grad_norm": 0.9786802993443171,
+ "learning_rate": 9.995877006644627e-06,
+ "loss": 0.4893208026885986,
+ "num_input_tokens_seen": 16739664,
+ "step": 385,
+ "train_runtime": 3003.5655,
+ "train_tokens_per_second": 5573.264
+ },
+ {
+ "epoch": 0.026259089684890925,
+ "grad_norm": 0.9983372688388842,
+ "learning_rate": 9.995768953249038e-06,
+ "loss": 0.4782050132751465,
+ "num_input_tokens_seen": 16966112,
+ "step": 390,
+ "train_runtime": 3042.5667,
+ "train_tokens_per_second": 5576.25
+ },
+ {
+ "epoch": 0.026595744680851064,
+ "grad_norm": 0.9763900936703368,
+ "learning_rate": 9.995659502802594e-06,
+ "loss": 0.4591067314147949,
+ "num_input_tokens_seen": 17184144,
+ "step": 395,
+ "train_runtime": 3081.608,
+ "train_tokens_per_second": 5576.356
+ },
+ {
+ "epoch": 0.026932399676811204,
+ "grad_norm": 1.1889179128899698,
+ "learning_rate": 9.995548655335899e-06,
+ "loss": 0.4847611427307129,
+ "num_input_tokens_seen": 17393504,
+ "step": 400,
+ "train_runtime": 3119.919,
+ "train_tokens_per_second": 5574.986
+ },
+ {
+ "epoch": 0.027269054672771343,
+ "grad_norm": 1.0385231933952215,
+ "learning_rate": 9.995436410879951e-06,
+ "loss": 0.5011202812194824,
+ "num_input_tokens_seen": 17618352,
+ "step": 405,
+ "train_runtime": 3158.5186,
+ "train_tokens_per_second": 5578.043
+ },
+ {
+ "epoch": 0.027605709668731482,
+ "grad_norm": 1.0315411320040826,
+ "learning_rate": 9.99532276946614e-06,
+ "loss": 0.4390664577484131,
+ "num_input_tokens_seen": 17833504,
+ "step": 410,
+ "train_runtime": 3197.1139,
+ "train_tokens_per_second": 5578.001
+ },
+ {
+ "epoch": 0.027942364664691625,
+ "grad_norm": 0.9798271051589648,
+ "learning_rate": 9.995207731126246e-06,
+ "loss": 0.48116722106933596,
+ "num_input_tokens_seen": 18060040,
+ "step": 415,
+ "train_runtime": 3233.4074,
+ "train_tokens_per_second": 5585.451
+ },
+ {
+ "epoch": 0.028279019660651765,
+ "grad_norm": 0.9349852018887401,
+ "learning_rate": 9.99509129589244e-06,
+ "loss": 0.4614968776702881,
+ "num_input_tokens_seen": 18273536,
+ "step": 420,
+ "train_runtime": 3276.2797,
+ "train_tokens_per_second": 5577.526
+ },
+ {
+ "epoch": 0.028615674656611904,
+ "grad_norm": 1.1608131463568248,
+ "learning_rate": 9.994973463797279e-06,
+ "loss": 0.4571842193603516,
+ "num_input_tokens_seen": 18501288,
+ "step": 425,
+ "train_runtime": 3312.0947,
+ "train_tokens_per_second": 5585.978
+ },
+ {
+ "epoch": 0.028952329652572044,
+ "grad_norm": 0.9117559518264567,
+ "learning_rate": 9.99485423487372e-06,
+ "loss": 0.42873506546020507,
+ "num_input_tokens_seen": 18713400,
+ "step": 430,
+ "train_runtime": 3354.8456,
+ "train_tokens_per_second": 5578.021
+ },
+ {
+ "epoch": 0.029288984648532183,
+ "grad_norm": 1.0351008634262573,
+ "learning_rate": 9.994733609155097e-06,
+ "loss": 0.46686439514160155,
+ "num_input_tokens_seen": 18927424,
+ "step": 435,
+ "train_runtime": 3390.8404,
+ "train_tokens_per_second": 5581.927
+ },
+ {
+ "epoch": 0.029625639644492323,
+ "grad_norm": 0.9515109413699648,
+ "learning_rate": 9.994611586675149e-06,
+ "loss": 0.4384657382965088,
+ "num_input_tokens_seen": 19153848,
+ "step": 440,
+ "train_runtime": 3427.1383,
+ "train_tokens_per_second": 5588.875
+ },
+ {
+ "epoch": 0.029962294640452466,
+ "grad_norm": 2.170654371130382,
+ "learning_rate": 9.994488167468e-06,
+ "loss": 0.457914924621582,
+ "num_input_tokens_seen": 19367752,
+ "step": 445,
+ "train_runtime": 3463.4205,
+ "train_tokens_per_second": 5592.088
+ },
+ {
+ "epoch": 0.030298949636412605,
+ "grad_norm": 0.9213402764770521,
+ "learning_rate": 9.994363351568158e-06,
+ "loss": 0.49051642417907715,
+ "num_input_tokens_seen": 19581376,
+ "step": 450,
+ "train_runtime": 3507.0672,
+ "train_tokens_per_second": 5583.405
+ },
+ {
+ "epoch": 0.030635604632372745,
+ "grad_norm": 1.0129019967889514,
+ "learning_rate": 9.994237139010532e-06,
+ "loss": 0.4277812480926514,
+ "num_input_tokens_seen": 19787728,
+ "step": 455,
+ "train_runtime": 3546.6942,
+ "train_tokens_per_second": 5579.203
+ },
+ {
+ "epoch": 0.030972259628332884,
+ "grad_norm": 0.9251976717430972,
+ "learning_rate": 9.994109529830414e-06,
+ "loss": 0.4721639633178711,
+ "num_input_tokens_seen": 20024672,
+ "step": 460,
+ "train_runtime": 3585.3338,
+ "train_tokens_per_second": 5585.163
+ },
+ {
+ "epoch": 0.03130891462429303,
+ "grad_norm": 0.8419132637197803,
+ "learning_rate": 9.993980524063494e-06,
+ "loss": 0.44678316116333006,
+ "num_input_tokens_seen": 20247816,
+ "step": 465,
+ "train_runtime": 3622.8878,
+ "train_tokens_per_second": 5588.861
+ },
+ {
+ "epoch": 0.03164556962025317,
+ "grad_norm": 0.8727373261991417,
+ "learning_rate": 9.993850121745842e-06,
+ "loss": 0.448548698425293,
+ "num_input_tokens_seen": 20469008,
+ "step": 470,
+ "train_runtime": 3663.8929,
+ "train_tokens_per_second": 5586.683
+ },
+ {
+ "epoch": 0.031982224616213306,
+ "grad_norm": 0.9661829335825901,
+ "learning_rate": 9.99371832291393e-06,
+ "loss": 0.44652185440063474,
+ "num_input_tokens_seen": 20684368,
+ "step": 475,
+ "train_runtime": 3702.3594,
+ "train_tokens_per_second": 5586.807
+ },
+ {
+ "epoch": 0.032318879612173446,
+ "grad_norm": 1.052144245514416,
+ "learning_rate": 9.993585127604612e-06,
+ "loss": 0.4458156108856201,
+ "num_input_tokens_seen": 20888824,
+ "step": 480,
+ "train_runtime": 3737.8839,
+ "train_tokens_per_second": 5588.409
+ },
+ {
+ "epoch": 0.032655534608133585,
+ "grad_norm": 0.8727464499212149,
+ "learning_rate": 9.993450535855136e-06,
+ "loss": 0.4528946876525879,
+ "num_input_tokens_seen": 21093776,
+ "step": 485,
+ "train_runtime": 3771.0576,
+ "train_tokens_per_second": 5593.597
+ },
+ {
+ "epoch": 0.032992189604093725,
+ "grad_norm": 1.4490470715769006,
+ "learning_rate": 9.993314547703142e-06,
+ "loss": 0.43402805328369143,
+ "num_input_tokens_seen": 21314528,
+ "step": 490,
+ "train_runtime": 3810.8049,
+ "train_tokens_per_second": 5593.183
+ },
+ {
+ "epoch": 0.033328844600053864,
+ "grad_norm": 0.913157610231942,
+ "learning_rate": 9.993177163186655e-06,
+ "loss": 0.44025378227233886,
+ "num_input_tokens_seen": 21529648,
+ "step": 495,
+ "train_runtime": 3846.3087,
+ "train_tokens_per_second": 5597.483
+ },
+ {
+ "epoch": 0.033665499596014004,
+ "grad_norm": 1.0732207036056458,
+ "learning_rate": 9.993038382344097e-06,
+ "loss": 0.43718814849853516,
+ "num_input_tokens_seen": 21744008,
+ "step": 500,
+ "train_runtime": 3886.9027,
+ "train_tokens_per_second": 5594.174
+ },
+ {
+ "epoch": 0.03400215459197414,
+ "grad_norm": 0.9385843989678325,
+ "learning_rate": 9.992898205214279e-06,
+ "loss": 0.44336519241333006,
+ "num_input_tokens_seen": 21963168,
+ "step": 505,
+ "train_runtime": 3919.9586,
+ "train_tokens_per_second": 5602.908
+ },
+ {
+ "epoch": 0.03433880958793428,
+ "grad_norm": 0.9966215411858433,
+ "learning_rate": 9.992756631836397e-06,
+ "loss": 0.4398432731628418,
+ "num_input_tokens_seen": 22181088,
+ "step": 510,
+ "train_runtime": 3967.3085,
+ "train_tokens_per_second": 5590.966
+ },
+ {
+ "epoch": 0.03467546458389442,
+ "grad_norm": 0.8920648260369749,
+ "learning_rate": 9.992613662250044e-06,
+ "loss": 0.4598443031311035,
+ "num_input_tokens_seen": 22402688,
+ "step": 515,
+ "train_runtime": 4007.4675,
+ "train_tokens_per_second": 5590.236
+ },
+ {
+ "epoch": 0.03501211957985457,
+ "grad_norm": 0.9290820640509749,
+ "learning_rate": 9.992469296495201e-06,
+ "loss": 0.4575651168823242,
+ "num_input_tokens_seen": 22628840,
+ "step": 520,
+ "train_runtime": 4042.7584,
+ "train_tokens_per_second": 5597.376
+ },
+ {
+ "epoch": 0.03534877457581471,
+ "grad_norm": 0.9109745499097042,
+ "learning_rate": 9.992323534612238e-06,
+ "loss": 0.414502477645874,
+ "num_input_tokens_seen": 22848504,
+ "step": 525,
+ "train_runtime": 4082.2649,
+ "train_tokens_per_second": 5597.017
+ },
+ {
+ "epoch": 0.03568542957177485,
+ "grad_norm": 1.1344769076017303,
+ "learning_rate": 9.99217637664192e-06,
+ "loss": 0.45860743522644043,
+ "num_input_tokens_seen": 23072192,
+ "step": 530,
+ "train_runtime": 4119.8107,
+ "train_tokens_per_second": 5600.304
+ },
+ {
+ "epoch": 0.03602208456773499,
+ "grad_norm": 0.940449394205852,
+ "learning_rate": 9.992027822625397e-06,
+ "loss": 0.45654983520507814,
+ "num_input_tokens_seen": 23281952,
+ "step": 535,
+ "train_runtime": 4161.0794,
+ "train_tokens_per_second": 5595.171
+ },
+ {
+ "epoch": 0.036358739563695126,
+ "grad_norm": 1.0036660879417987,
+ "learning_rate": 9.99187787260421e-06,
+ "loss": 0.46324939727783204,
+ "num_input_tokens_seen": 23492592,
+ "step": 540,
+ "train_runtime": 4199.7934,
+ "train_tokens_per_second": 5593.749
+ },
+ {
+ "epoch": 0.036695394559655266,
+ "grad_norm": 0.9529512682682689,
+ "learning_rate": 9.991726526620294e-06,
+ "loss": 0.4310856819152832,
+ "num_input_tokens_seen": 23724784,
+ "step": 545,
+ "train_runtime": 4235.651,
+ "train_tokens_per_second": 5601.213
+ },
+ {
+ "epoch": 0.037032049555615405,
+ "grad_norm": 1.0092720255970657,
+ "learning_rate": 9.991573784715973e-06,
+ "loss": 0.4721970081329346,
+ "num_input_tokens_seen": 23941816,
+ "step": 550,
+ "train_runtime": 4278.956,
+ "train_tokens_per_second": 5595.247
+ },
+ {
+ "epoch": 0.037368704551575545,
+ "grad_norm": 1.0108753586712804,
+ "learning_rate": 9.99141964693396e-06,
+ "loss": 0.48006014823913573,
+ "num_input_tokens_seen": 24165232,
+ "step": 555,
+ "train_runtime": 4315.3047,
+ "train_tokens_per_second": 5599.89
+ },
+ {
+ "epoch": 0.037705359547535684,
+ "grad_norm": 1.0343909095504493,
+ "learning_rate": 9.99126411331736e-06,
+ "loss": 0.4417973518371582,
+ "num_input_tokens_seen": 24389896,
+ "step": 560,
+ "train_runtime": 4359.1735,
+ "train_tokens_per_second": 5595.074
+ },
+ {
+ "epoch": 0.038042014543495824,
+ "grad_norm": 1.0113397276422895,
+ "learning_rate": 9.991107183909665e-06,
+ "loss": 0.4680752754211426,
+ "num_input_tokens_seen": 24614992,
+ "step": 565,
+ "train_runtime": 4391.4505,
+ "train_tokens_per_second": 5605.208
+ },
+ {
+ "epoch": 0.03837866953945596,
+ "grad_norm": 0.9954842488566088,
+ "learning_rate": 9.990948858754763e-06,
+ "loss": 0.4640792369842529,
+ "num_input_tokens_seen": 24823080,
+ "step": 570,
+ "train_runtime": 4434.0745,
+ "train_tokens_per_second": 5598.255
+ },
+ {
+ "epoch": 0.0387153245354161,
+ "grad_norm": 0.9153320849806976,
+ "learning_rate": 9.990789137896927e-06,
+ "loss": 0.4609560966491699,
+ "num_input_tokens_seen": 25044784,
+ "step": 575,
+ "train_runtime": 4471.0247,
+ "train_tokens_per_second": 5601.576
+ },
+ {
+ "epoch": 0.03905197953137624,
+ "grad_norm": 0.9718791869837801,
+ "learning_rate": 9.990628021380823e-06,
+ "loss": 0.47663059234619143,
+ "num_input_tokens_seen": 25262200,
+ "step": 580,
+ "train_runtime": 4509.6496,
+ "train_tokens_per_second": 5601.81
+ },
+ {
+ "epoch": 0.03938863452733639,
+ "grad_norm": 1.218415508866825,
+ "learning_rate": 9.990465509251508e-06,
+ "loss": 0.46056556701660156,
+ "num_input_tokens_seen": 25473136,
+ "step": 585,
+ "train_runtime": 4541.7073,
+ "train_tokens_per_second": 5608.714
+ },
+ {
+ "epoch": 0.03972528952329653,
+ "grad_norm": 1.0200294193913204,
+ "learning_rate": 9.990301601554425e-06,
+ "loss": 0.48001799583435056,
+ "num_input_tokens_seen": 25671088,
+ "step": 590,
+ "train_runtime": 4581.4528,
+ "train_tokens_per_second": 5603.264
+ },
+ {
+ "epoch": 0.04006194451925667,
+ "grad_norm": 1.0450169479009095,
+ "learning_rate": 9.990136298335414e-06,
+ "loss": 0.40839109420776365,
+ "num_input_tokens_seen": 25887472,
+ "step": 595,
+ "train_runtime": 4618.2668,
+ "train_tokens_per_second": 5605.452
+ },
+ {
+ "epoch": 0.04039859951521681,
+ "grad_norm": 0.9533941163271269,
+ "learning_rate": 9.989969599640698e-06,
+ "loss": 0.4223615646362305,
+ "num_input_tokens_seen": 26117456,
+ "step": 600,
+ "train_runtime": 4650.9664,
+ "train_tokens_per_second": 5615.49
+ },
+ {
+ "epoch": 0.040735254511176947,
+ "grad_norm": 0.9258754840147914,
+ "learning_rate": 9.989801505516895e-06,
+ "loss": 0.505678129196167,
+ "num_input_tokens_seen": 26348056,
+ "step": 605,
+ "train_runtime": 4690.0532,
+ "train_tokens_per_second": 5617.859
+ },
+ {
+ "epoch": 0.041071909507137086,
+ "grad_norm": 0.9866589002080994,
+ "learning_rate": 9.989632016011012e-06,
+ "loss": 0.47707581520080566,
+ "num_input_tokens_seen": 26572040,
+ "step": 610,
+ "train_runtime": 4721.6631,
+ "train_tokens_per_second": 5627.687
+ },
+ {
+ "epoch": 0.041408564503097225,
+ "grad_norm": 1.0374837134071888,
+ "learning_rate": 9.989461131170448e-06,
+ "loss": 0.4833499908447266,
+ "num_input_tokens_seen": 26769320,
+ "step": 615,
+ "train_runtime": 4762.2337,
+ "train_tokens_per_second": 5621.169
+ },
+ {
+ "epoch": 0.041745219499057365,
+ "grad_norm": 0.8823212064118572,
+ "learning_rate": 9.98928885104299e-06,
+ "loss": 0.4514059543609619,
+ "num_input_tokens_seen": 26988024,
+ "step": 620,
+ "train_runtime": 4795.3541,
+ "train_tokens_per_second": 5627.952
+ },
+ {
+ "epoch": 0.042081874495017504,
+ "grad_norm": 0.8780469782436839,
+ "learning_rate": 9.989115175676812e-06,
+ "loss": 0.4226998329162598,
+ "num_input_tokens_seen": 27212624,
+ "step": 625,
+ "train_runtime": 4834.4544,
+ "train_tokens_per_second": 5628.892
+ },
+ {
+ "epoch": 0.042418529490977644,
+ "grad_norm": 0.9554506078310793,
+ "learning_rate": 9.988940105120483e-06,
+ "loss": 0.47803831100463867,
+ "num_input_tokens_seen": 27422208,
+ "step": 630,
+ "train_runtime": 4877.5946,
+ "train_tokens_per_second": 5622.076
+ },
+ {
+ "epoch": 0.04275518448693778,
+ "grad_norm": 0.93348166885423,
+ "learning_rate": 9.988763639422965e-06,
+ "loss": 0.4802694797515869,
+ "num_input_tokens_seen": 27650384,
+ "step": 635,
+ "train_runtime": 4915.966,
+ "train_tokens_per_second": 5624.608
+ },
+ {
+ "epoch": 0.04309183948289792,
+ "grad_norm": 0.9769454728471132,
+ "learning_rate": 9.988585778633604e-06,
+ "loss": 0.4367190361022949,
+ "num_input_tokens_seen": 27851424,
+ "step": 640,
+ "train_runtime": 4948.8643,
+ "train_tokens_per_second": 5627.841
+ },
+ {
+ "epoch": 0.04342849447885807,
+ "grad_norm": 0.9848655689806975,
+ "learning_rate": 9.988406522802134e-06,
+ "loss": 0.45001683235168455,
+ "num_input_tokens_seen": 28060088,
+ "step": 645,
+ "train_runtime": 4991.2246,
+ "train_tokens_per_second": 5621.884
+ },
+ {
+ "epoch": 0.04376514947481821,
+ "grad_norm": 0.9626370029590127,
+ "learning_rate": 9.98822587197869e-06,
+ "loss": 0.41652770042419435,
+ "num_input_tokens_seen": 28274912,
+ "step": 650,
+ "train_runtime": 5025.7373,
+ "train_tokens_per_second": 5626.023
+ },
+ {
+ "epoch": 0.04410180447077835,
+ "grad_norm": 0.9983674144400133,
+ "learning_rate": 9.988043826213785e-06,
+ "loss": 0.4282545566558838,
+ "num_input_tokens_seen": 28481968,
+ "step": 655,
+ "train_runtime": 5062.1603,
+ "train_tokens_per_second": 5626.445
+ },
+ {
+ "epoch": 0.04443845946673849,
+ "grad_norm": 2.4559661675719884,
+ "learning_rate": 9.98786038555833e-06,
+ "loss": 0.4085527420043945,
+ "num_input_tokens_seen": 28702472,
+ "step": 660,
+ "train_runtime": 5099.8424,
+ "train_tokens_per_second": 5628.11
+ },
+ {
+ "epoch": 0.04477511446269863,
+ "grad_norm": 0.8332845163143295,
+ "learning_rate": 9.987675550063625e-06,
+ "loss": 0.44841432571411133,
+ "num_input_tokens_seen": 28932512,
+ "step": 665,
+ "train_runtime": 5135.2428,
+ "train_tokens_per_second": 5634.108
+ },
+ {
+ "epoch": 0.04511176945865877,
+ "grad_norm": 1.0017265357539382,
+ "learning_rate": 9.987489319781355e-06,
+ "loss": 0.4431906223297119,
+ "num_input_tokens_seen": 29153640,
+ "step": 670,
+ "train_runtime": 5165.908,
+ "train_tokens_per_second": 5643.469
+ },
+ {
+ "epoch": 0.045448424454618906,
+ "grad_norm": 0.8381101119904498,
+ "learning_rate": 9.987301694763602e-06,
+ "loss": 0.4798222541809082,
+ "num_input_tokens_seen": 29368992,
+ "step": 675,
+ "train_runtime": 5201.341,
+ "train_tokens_per_second": 5646.427
+ },
+ {
+ "epoch": 0.045785079450579046,
+ "grad_norm": 0.9739030010713555,
+ "learning_rate": 9.987112675062834e-06,
+ "loss": 0.4200302600860596,
+ "num_input_tokens_seen": 29582872,
+ "step": 680,
+ "train_runtime": 5240.8223,
+ "train_tokens_per_second": 5644.7
+ },
+ {
+ "epoch": 0.046121734446539185,
+ "grad_norm": 0.9923641076223235,
+ "learning_rate": 9.986922260731908e-06,
+ "loss": 0.42931652069091797,
+ "num_input_tokens_seen": 29791056,
+ "step": 685,
+ "train_runtime": 5280.43,
+ "train_tokens_per_second": 5641.786
+ },
+ {
+ "epoch": 0.046458389442499325,
+ "grad_norm": 0.8786561915980274,
+ "learning_rate": 9.986730451824073e-06,
+ "loss": 0.4656373977661133,
+ "num_input_tokens_seen": 30001648,
+ "step": 690,
+ "train_runtime": 5325.5768,
+ "train_tokens_per_second": 5633.502
+ },
+ {
+ "epoch": 0.046795044438459464,
+ "grad_norm": 0.9155252116619295,
+ "learning_rate": 9.986537248392969e-06,
+ "loss": 0.4756736755371094,
+ "num_input_tokens_seen": 30223896,
+ "step": 695,
+ "train_runtime": 5365.8859,
+ "train_tokens_per_second": 5632.601
+ },
+ {
+ "epoch": 0.047131699434419604,
+ "grad_norm": 1.0345486643678798,
+ "learning_rate": 9.986342650492626e-06,
+ "loss": 0.4506352424621582,
+ "num_input_tokens_seen": 30445352,
+ "step": 700,
+ "train_runtime": 5399.4485,
+ "train_tokens_per_second": 5638.604
+ },
+ {
+ "epoch": 0.04746835443037975,
+ "grad_norm": 0.919553088389443,
+ "learning_rate": 9.98614665817746e-06,
+ "loss": 0.42203187942504883,
+ "num_input_tokens_seen": 30642128,
+ "step": 705,
+ "train_runtime": 5439.3583,
+ "train_tokens_per_second": 5633.409
+ },
+ {
+ "epoch": 0.04780500942633989,
+ "grad_norm": 0.9271703985798777,
+ "learning_rate": 9.98594927150228e-06,
+ "loss": 0.427701473236084,
+ "num_input_tokens_seen": 30873224,
+ "step": 710,
+ "train_runtime": 5478.7451,
+ "train_tokens_per_second": 5635.09
+ },
+ {
+ "epoch": 0.04814166442230003,
+ "grad_norm": 0.9508538400970585,
+ "learning_rate": 9.985750490522286e-06,
+ "loss": 0.39564857482910154,
+ "num_input_tokens_seen": 31087512,
+ "step": 715,
+ "train_runtime": 5521.2724,
+ "train_tokens_per_second": 5630.498
+ },
+ {
+ "epoch": 0.04847831941826017,
+ "grad_norm": 0.9372110505017411,
+ "learning_rate": 9.985550315293066e-06,
+ "loss": 0.4636754035949707,
+ "num_input_tokens_seen": 31318912,
+ "step": 720,
+ "train_runtime": 5557.4337,
+ "train_tokens_per_second": 5635.499
+ },
+ {
+ "epoch": 0.04881497441422031,
+ "grad_norm": 0.9970520932843633,
+ "learning_rate": 9.985348745870598e-06,
+ "loss": 0.4597306728363037,
+ "num_input_tokens_seen": 31546096,
+ "step": 725,
+ "train_runtime": 5592.8911,
+ "train_tokens_per_second": 5640.392
+ },
+ {
+ "epoch": 0.04915162941018045,
+ "grad_norm": 0.8271247348731543,
+ "learning_rate": 9.98514578231125e-06,
+ "loss": 0.4194610595703125,
+ "num_input_tokens_seen": 31774968,
+ "step": 730,
+ "train_runtime": 5630.8722,
+ "train_tokens_per_second": 5642.992
+ },
+ {
+ "epoch": 0.04948828440614059,
+ "grad_norm": 0.9452005671897916,
+ "learning_rate": 9.984941424671781e-06,
+ "loss": 0.4705995559692383,
+ "num_input_tokens_seen": 31991384,
+ "step": 735,
+ "train_runtime": 5672.8388,
+ "train_tokens_per_second": 5639.396
+ },
+ {
+ "epoch": 0.049824939402100726,
+ "grad_norm": 0.9711026791829106,
+ "learning_rate": 9.984735673009337e-06,
+ "loss": 0.43364248275756834,
+ "num_input_tokens_seen": 32211232,
+ "step": 740,
+ "train_runtime": 5717.5318,
+ "train_tokens_per_second": 5633.765
+ },
+ {
+ "epoch": 0.050161594398060866,
+ "grad_norm": 0.8583208108971163,
+ "learning_rate": 9.984528527381458e-06,
+ "loss": 0.4496613502502441,
+ "num_input_tokens_seen": 32432504,
+ "step": 745,
+ "train_runtime": 5758.1136,
+ "train_tokens_per_second": 5632.488
+ },
+ {
+ "epoch": 0.050498249394021005,
+ "grad_norm": 1.0983008551303333,
+ "learning_rate": 9.984319987846072e-06,
+ "loss": 0.4476910591125488,
+ "num_input_tokens_seen": 32644184,
+ "step": 750,
+ "train_runtime": 5795.8989,
+ "train_tokens_per_second": 5632.29
+ },
+ {
+ "epoch": 0.050834904389981145,
+ "grad_norm": 0.9341698177556362,
+ "learning_rate": 9.984110054461495e-06,
+ "loss": 0.4494589328765869,
+ "num_input_tokens_seen": 32883080,
+ "step": 755,
+ "train_runtime": 5831.2135,
+ "train_tokens_per_second": 5639.149
+ },
+ {
+ "epoch": 0.051171559385941284,
+ "grad_norm": 2.9705690058280205,
+ "learning_rate": 9.983898727286435e-06,
+ "loss": 0.44231281280517576,
+ "num_input_tokens_seen": 33102688,
+ "step": 760,
+ "train_runtime": 5864.3725,
+ "train_tokens_per_second": 5644.711
+ },
+ {
+ "epoch": 0.05150821438190143,
+ "grad_norm": 0.9183883314753698,
+ "learning_rate": 9.983686006379985e-06,
+ "loss": 0.45505475997924805,
+ "num_input_tokens_seen": 33319336,
+ "step": 765,
+ "train_runtime": 5901.491,
+ "train_tokens_per_second": 5645.918
+ },
+ {
+ "epoch": 0.05184486937786157,
+ "grad_norm": 0.9313957301525887,
+ "learning_rate": 9.983471891801638e-06,
+ "loss": 0.437741756439209,
+ "num_input_tokens_seen": 33545192,
+ "step": 770,
+ "train_runtime": 5938.7864,
+ "train_tokens_per_second": 5648.493
+ },
+ {
+ "epoch": 0.05218152437382171,
+ "grad_norm": 0.8973015582826919,
+ "learning_rate": 9.983256383611267e-06,
+ "loss": 0.441961669921875,
+ "num_input_tokens_seen": 33767136,
+ "step": 775,
+ "train_runtime": 5974.9184,
+ "train_tokens_per_second": 5651.481
+ },
+ {
+ "epoch": 0.05251817936978185,
+ "grad_norm": 1.380873298582274,
+ "learning_rate": 9.983039481869138e-06,
+ "loss": 0.42367143630981446,
+ "num_input_tokens_seen": 33985712,
+ "step": 780,
+ "train_runtime": 6009.8556,
+ "train_tokens_per_second": 5654.996
+ },
+ {
+ "epoch": 0.05285483436574199,
+ "grad_norm": 0.9605495456782687,
+ "learning_rate": 9.98282118663591e-06,
+ "loss": 0.4347693920135498,
+ "num_input_tokens_seen": 34208000,
+ "step": 785,
+ "train_runtime": 6046.7667,
+ "train_tokens_per_second": 5657.238
+ },
+ {
+ "epoch": 0.05319148936170213,
+ "grad_norm": 0.9570184798319451,
+ "learning_rate": 9.982601497972624e-06,
+ "loss": 0.4524986267089844,
+ "num_input_tokens_seen": 34425040,
+ "step": 790,
+ "train_runtime": 6085.7762,
+ "train_tokens_per_second": 5656.639
+ },
+ {
+ "epoch": 0.05352814435766227,
+ "grad_norm": 1.293632659414026,
+ "learning_rate": 9.982380415940717e-06,
+ "loss": 0.4259688377380371,
+ "num_input_tokens_seen": 34634712,
+ "step": 795,
+ "train_runtime": 6120.1944,
+ "train_tokens_per_second": 5659.087
+ },
+ {
+ "epoch": 0.05386479935362241,
+ "grad_norm": 0.9262120217395047,
+ "learning_rate": 9.982157940602015e-06,
+ "loss": 0.46526622772216797,
+ "num_input_tokens_seen": 34865376,
+ "step": 800,
+ "train_runtime": 6156.1798,
+ "train_tokens_per_second": 5663.476
+ },
+ {
+ "epoch": 0.05420145434958255,
+ "grad_norm": 0.892992639712076,
+ "learning_rate": 9.981934072018732e-06,
+ "loss": 0.44102907180786133,
+ "num_input_tokens_seen": 35098448,
+ "step": 805,
+ "train_runtime": 6189.6138,
+ "train_tokens_per_second": 5670.539
+ },
+ {
+ "epoch": 0.054538109345542686,
+ "grad_norm": 0.8797013169732244,
+ "learning_rate": 9.981708810253472e-06,
+ "loss": 0.4654733657836914,
+ "num_input_tokens_seen": 35317184,
+ "step": 810,
+ "train_runtime": 6223.9712,
+ "train_tokens_per_second": 5674.381
+ },
+ {
+ "epoch": 0.054874764341502826,
+ "grad_norm": 1.1385550766773516,
+ "learning_rate": 9.981482155369229e-06,
+ "loss": 0.4581299781799316,
+ "num_input_tokens_seen": 35543024,
+ "step": 815,
+ "train_runtime": 6264.1777,
+ "train_tokens_per_second": 5674.013
+ },
+ {
+ "epoch": 0.055211419337462965,
+ "grad_norm": 1.0108160361368144,
+ "learning_rate": 9.981254107429385e-06,
+ "loss": 0.4073037147521973,
+ "num_input_tokens_seen": 35752976,
+ "step": 820,
+ "train_runtime": 6301.9605,
+ "train_tokens_per_second": 5673.31
+ },
+ {
+ "epoch": 0.05554807433342311,
+ "grad_norm": 0.8271492373593813,
+ "learning_rate": 9.981024666497716e-06,
+ "loss": 0.4087496757507324,
+ "num_input_tokens_seen": 35970808,
+ "step": 825,
+ "train_runtime": 6336.9136,
+ "train_tokens_per_second": 5676.392
+ },
+ {
+ "epoch": 0.05588472932938325,
+ "grad_norm": 1.2448406201928153,
+ "learning_rate": 9.98079383263838e-06,
+ "loss": 0.451265811920166,
+ "num_input_tokens_seen": 36175264,
+ "step": 830,
+ "train_runtime": 6372.1661,
+ "train_tokens_per_second": 5677.075
+ },
+ {
+ "epoch": 0.05622138432534339,
+ "grad_norm": 1.0960090803895477,
+ "learning_rate": 9.980561605915932e-06,
+ "loss": 0.43454971313476565,
+ "num_input_tokens_seen": 36399968,
+ "step": 835,
+ "train_runtime": 6409.522,
+ "train_tokens_per_second": 5679.046
+ },
+ {
+ "epoch": 0.05655803932130353,
+ "grad_norm": 0.865350004624438,
+ "learning_rate": 9.980327986395316e-06,
+ "loss": 0.44866161346435546,
+ "num_input_tokens_seen": 36614352,
+ "step": 840,
+ "train_runtime": 6450.8502,
+ "train_tokens_per_second": 5675.896
+ },
+ {
+ "epoch": 0.05689469431726367,
+ "grad_norm": 0.9132401043618537,
+ "learning_rate": 9.980092974141855e-06,
+ "loss": 0.43543548583984376,
+ "num_input_tokens_seen": 36842440,
+ "step": 845,
+ "train_runtime": 6489.8829,
+ "train_tokens_per_second": 5676.904
+ },
+ {
+ "epoch": 0.05723134931322381,
+ "grad_norm": 0.9515826536482608,
+ "learning_rate": 9.979856569221276e-06,
+ "loss": 0.434114933013916,
+ "num_input_tokens_seen": 37064928,
+ "step": 850,
+ "train_runtime": 6535.3585,
+ "train_tokens_per_second": 5671.445
+ },
+ {
+ "epoch": 0.05756800430918395,
+ "grad_norm": 0.905243476379667,
+ "learning_rate": 9.979618771699688e-06,
+ "loss": 0.47480268478393556,
+ "num_input_tokens_seen": 37286696,
+ "step": 855,
+ "train_runtime": 6571.3163,
+ "train_tokens_per_second": 5674.159
+ },
+ {
+ "epoch": 0.05790465930514409,
+ "grad_norm": 0.9973188371129253,
+ "learning_rate": 9.979379581643588e-06,
+ "loss": 0.4271403312683105,
+ "num_input_tokens_seen": 37505104,
+ "step": 860,
+ "train_runtime": 6612.0758,
+ "train_tokens_per_second": 5672.213
+ },
+ {
+ "epoch": 0.05824131430110423,
+ "grad_norm": 0.891663822548588,
+ "learning_rate": 9.979138999119866e-06,
+ "loss": 0.46260747909545896,
+ "num_input_tokens_seen": 37721104,
+ "step": 865,
+ "train_runtime": 6657.2054,
+ "train_tokens_per_second": 5666.207
+ },
+ {
+ "epoch": 0.05857796929706437,
+ "grad_norm": 0.8193884764682856,
+ "learning_rate": 9.978897024195801e-06,
+ "loss": 0.43666720390319824,
+ "num_input_tokens_seen": 37959848,
+ "step": 870,
+ "train_runtime": 6692.9435,
+ "train_tokens_per_second": 5671.622
+ },
+ {
+ "epoch": 0.058914624293024506,
+ "grad_norm": 1.1669037394876847,
+ "learning_rate": 9.978653656939058e-06,
+ "loss": 0.4272760391235352,
+ "num_input_tokens_seen": 38179560,
+ "step": 875,
+ "train_runtime": 6736.655,
+ "train_tokens_per_second": 5667.436
+ },
+ {
+ "epoch": 0.059251279288984646,
+ "grad_norm": 1.0468766117182693,
+ "learning_rate": 9.978408897417699e-06,
+ "loss": 0.4613694190979004,
+ "num_input_tokens_seen": 38405040,
+ "step": 880,
+ "train_runtime": 6776.3735,
+ "train_tokens_per_second": 5667.492
+ },
+ {
+ "epoch": 0.059587934284944785,
+ "grad_norm": 0.8352176899590911,
+ "learning_rate": 9.978162745700165e-06,
+ "loss": 0.4417067527770996,
+ "num_input_tokens_seen": 38631232,
+ "step": 885,
+ "train_runtime": 6814.8051,
+ "train_tokens_per_second": 5668.721
+ },
+ {
+ "epoch": 0.05992458928090493,
+ "grad_norm": 1.0696188249598721,
+ "learning_rate": 9.977915201855293e-06,
+ "loss": 0.42413697242736814,
+ "num_input_tokens_seen": 38849712,
+ "step": 890,
+ "train_runtime": 6850.9387,
+ "train_tokens_per_second": 5670.714
+ },
+ {
+ "epoch": 0.06026124427686507,
+ "grad_norm": 0.9281445460422576,
+ "learning_rate": 9.977666265952309e-06,
+ "loss": 0.4329718589782715,
+ "num_input_tokens_seen": 39067416,
+ "step": 895,
+ "train_runtime": 6890.6802,
+ "train_tokens_per_second": 5669.602
+ },
+ {
+ "epoch": 0.06059789927282521,
+ "grad_norm": 0.8685965904886482,
+ "learning_rate": 9.977415938060823e-06,
+ "loss": 0.4434166431427002,
+ "num_input_tokens_seen": 39288624,
+ "step": 900,
+ "train_runtime": 6926.1271,
+ "train_tokens_per_second": 5672.524
+ },
+ {
+ "epoch": 0.06093455426878535,
+ "grad_norm": 0.9417641978342199,
+ "learning_rate": 9.977164218250846e-06,
+ "loss": 0.444110107421875,
+ "num_input_tokens_seen": 39505328,
+ "step": 905,
+ "train_runtime": 6968.8615,
+ "train_tokens_per_second": 5668.835
+ },
+ {
+ "epoch": 0.06127120926474549,
+ "grad_norm": 0.8557175345597763,
+ "learning_rate": 9.976911106592765e-06,
+ "loss": 0.4071072578430176,
+ "num_input_tokens_seen": 39717144,
+ "step": 910,
+ "train_runtime": 7009.1269,
+ "train_tokens_per_second": 5666.49
+ },
+ {
+ "epoch": 0.06160786426070563,
+ "grad_norm": 1.1399112064858985,
+ "learning_rate": 9.976656603157363e-06,
+ "loss": 0.4384052276611328,
+ "num_input_tokens_seen": 39933000,
+ "step": 915,
+ "train_runtime": 7045.7591,
+ "train_tokens_per_second": 5667.665
+ },
+ {
+ "epoch": 0.06194451925666577,
+ "grad_norm": 0.9115769460735773,
+ "learning_rate": 9.976400708015811e-06,
+ "loss": 0.42393884658813474,
+ "num_input_tokens_seen": 40145168,
+ "step": 920,
+ "train_runtime": 7079.4173,
+ "train_tokens_per_second": 5670.688
+ },
+ {
+ "epoch": 0.06228117425262591,
+ "grad_norm": 0.8414795692849011,
+ "learning_rate": 9.97614342123967e-06,
+ "loss": 0.4269254684448242,
+ "num_input_tokens_seen": 40379264,
+ "step": 925,
+ "train_runtime": 7111.1373,
+ "train_tokens_per_second": 5678.313
+ },
+ {
+ "epoch": 0.06261782924858605,
+ "grad_norm": 0.9352365482641929,
+ "learning_rate": 9.97588474290089e-06,
+ "loss": 0.4331661224365234,
+ "num_input_tokens_seen": 40594944,
+ "step": 930,
+ "train_runtime": 7152.76,
+ "train_tokens_per_second": 5675.424
+ },
+ {
+ "epoch": 0.0629544842445462,
+ "grad_norm": 1.0614310774414917,
+ "learning_rate": 9.975624673071806e-06,
+ "loss": 0.4371979236602783,
+ "num_input_tokens_seen": 40808856,
+ "step": 935,
+ "train_runtime": 7197.718,
+ "train_tokens_per_second": 5669.694
+ },
+ {
+ "epoch": 0.06329113924050633,
+ "grad_norm": 1.0079121416060646,
+ "learning_rate": 9.97536321182515e-06,
+ "loss": 0.4044677734375,
+ "num_input_tokens_seen": 41018224,
+ "step": 940,
+ "train_runtime": 7227.9115,
+ "train_tokens_per_second": 5674.976
+ },
+ {
+ "epoch": 0.06362779423646647,
+ "grad_norm": 0.9987688379119654,
+ "learning_rate": 9.975100359234037e-06,
+ "loss": 0.4564094066619873,
+ "num_input_tokens_seen": 41239392,
+ "step": 945,
+ "train_runtime": 7265.9862,
+ "train_tokens_per_second": 5675.677
+ },
+ {
+ "epoch": 0.06396444923242661,
+ "grad_norm": 0.9165255721069369,
+ "learning_rate": 9.974836115371972e-06,
+ "loss": 0.46511135101318357,
+ "num_input_tokens_seen": 41454352,
+ "step": 950,
+ "train_runtime": 7304.5238,
+ "train_tokens_per_second": 5675.161
+ },
+ {
+ "epoch": 0.06430110422838675,
+ "grad_norm": 0.9087923038014859,
+ "learning_rate": 9.974570480312852e-06,
+ "loss": 0.40421581268310547,
+ "num_input_tokens_seen": 41680576,
+ "step": 955,
+ "train_runtime": 7346.7849,
+ "train_tokens_per_second": 5673.308
+ },
+ {
+ "epoch": 0.06463775922434689,
+ "grad_norm": 1.0231903244202105,
+ "learning_rate": 9.974303454130962e-06,
+ "loss": 0.4464391231536865,
+ "num_input_tokens_seen": 41882360,
+ "step": 960,
+ "train_runtime": 7380.8428,
+ "train_tokens_per_second": 5674.469
+ },
+ {
+ "epoch": 0.06497441422030703,
+ "grad_norm": 0.8595399990273458,
+ "learning_rate": 9.97403503690097e-06,
+ "loss": 0.42637953758239744,
+ "num_input_tokens_seen": 42103128,
+ "step": 965,
+ "train_runtime": 7415.439,
+ "train_tokens_per_second": 5677.766
+ },
+ {
+ "epoch": 0.06531106921626717,
+ "grad_norm": 0.8422231085957664,
+ "learning_rate": 9.973765228697942e-06,
+ "loss": 0.39764556884765623,
+ "num_input_tokens_seen": 42328192,
+ "step": 970,
+ "train_runtime": 7453.2347,
+ "train_tokens_per_second": 5679.171
+ },
+ {
+ "epoch": 0.06564772421222731,
+ "grad_norm": 0.9247166172844082,
+ "learning_rate": 9.973494029597327e-06,
+ "loss": 0.43375320434570314,
+ "num_input_tokens_seen": 42531664,
+ "step": 975,
+ "train_runtime": 7487.6836,
+ "train_tokens_per_second": 5680.217
+ },
+ {
+ "epoch": 0.06598437920818745,
+ "grad_norm": 1.0151970555059149,
+ "learning_rate": 9.973221439674968e-06,
+ "loss": 0.4309089660644531,
+ "num_input_tokens_seen": 42740856,
+ "step": 980,
+ "train_runtime": 7527.1188,
+ "train_tokens_per_second": 5678.249
+ },
+ {
+ "epoch": 0.06632103420414759,
+ "grad_norm": 0.882649641197766,
+ "learning_rate": 9.972947459007089e-06,
+ "loss": 0.45548033714294434,
+ "num_input_tokens_seen": 42956288,
+ "step": 985,
+ "train_runtime": 7562.7131,
+ "train_tokens_per_second": 5680.01
+ },
+ {
+ "epoch": 0.06665768920010773,
+ "grad_norm": 0.9104741890803707,
+ "learning_rate": 9.972672087670313e-06,
+ "loss": 0.4385095119476318,
+ "num_input_tokens_seen": 43177272,
+ "step": 990,
+ "train_runtime": 7599.4941,
+ "train_tokens_per_second": 5681.598
+ },
+ {
+ "epoch": 0.06699434419606787,
+ "grad_norm": 0.8108451361340878,
+ "learning_rate": 9.972395325741644e-06,
+ "loss": 0.413985538482666,
+ "num_input_tokens_seen": 43400152,
+ "step": 995,
+ "train_runtime": 7640.1047,
+ "train_tokens_per_second": 5680.57
+ },
+ {
+ "epoch": 0.06733099919202801,
+ "grad_norm": 0.9025973522517067,
+ "learning_rate": 9.972117173298476e-06,
+ "loss": 0.38621523380279543,
+ "num_input_tokens_seen": 43614968,
+ "step": 1000,
+ "train_runtime": 7674.2806,
+ "train_tokens_per_second": 5683.265
+ },
+ {
+ "epoch": 0.06766765418798815,
+ "grad_norm": 0.971243318554936,
+ "learning_rate": 9.971837630418597e-06,
+ "loss": 0.42676420211791993,
+ "num_input_tokens_seen": 43835904,
+ "step": 1005,
+ "train_runtime": 7717.1168,
+ "train_tokens_per_second": 5680.347
+ },
+ {
+ "epoch": 0.06800430918394829,
+ "grad_norm": 0.9278937330533324,
+ "learning_rate": 9.971556697180179e-06,
+ "loss": 0.44001970291137693,
+ "num_input_tokens_seen": 44048512,
+ "step": 1010,
+ "train_runtime": 7756.4813,
+ "train_tokens_per_second": 5678.93
+ },
+ {
+ "epoch": 0.06834096417990843,
+ "grad_norm": 0.9429692040194982,
+ "learning_rate": 9.971274373661784e-06,
+ "loss": 0.4194046974182129,
+ "num_input_tokens_seen": 44271000,
+ "step": 1015,
+ "train_runtime": 7787.4236,
+ "train_tokens_per_second": 5684.935
+ },
+ {
+ "epoch": 0.06867761917586857,
+ "grad_norm": 1.111228553028809,
+ "learning_rate": 9.97099065994236e-06,
+ "loss": 0.39097912311553956,
+ "num_input_tokens_seen": 44499576,
+ "step": 1020,
+ "train_runtime": 7822.0558,
+ "train_tokens_per_second": 5688.987
+ },
+ {
+ "epoch": 0.0690142741718287,
+ "grad_norm": 1.0049309947436895,
+ "learning_rate": 9.970705556101253e-06,
+ "loss": 0.43691391944885255,
+ "num_input_tokens_seen": 44702000,
+ "step": 1025,
+ "train_runtime": 7857.7887,
+ "train_tokens_per_second": 5688.878
+ },
+ {
+ "epoch": 0.06935092916778884,
+ "grad_norm": 0.8982051488791603,
+ "learning_rate": 9.970419062218186e-06,
+ "loss": 0.45786590576171876,
+ "num_input_tokens_seen": 44936832,
+ "step": 1030,
+ "train_runtime": 7899.821,
+ "train_tokens_per_second": 5688.336
+ },
+ {
+ "epoch": 0.06968758416374898,
+ "grad_norm": 0.9779591960140872,
+ "learning_rate": 9.970131178373276e-06,
+ "loss": 0.4594112396240234,
+ "num_input_tokens_seen": 45166720,
+ "step": 1035,
+ "train_runtime": 7937.6344,
+ "train_tokens_per_second": 5690.199
+ },
+ {
+ "epoch": 0.07002423915970914,
+ "grad_norm": 0.9403137936019832,
+ "learning_rate": 9.969841904647033e-06,
+ "loss": 0.48036918640136717,
+ "num_input_tokens_seen": 45386112,
+ "step": 1040,
+ "train_runtime": 7975.3873,
+ "train_tokens_per_second": 5690.772
+ },
+ {
+ "epoch": 0.07036089415566928,
+ "grad_norm": 0.9489869208797291,
+ "learning_rate": 9.969551241120349e-06,
+ "loss": 0.4358417510986328,
+ "num_input_tokens_seen": 45615968,
+ "step": 1045,
+ "train_runtime": 8008.3857,
+ "train_tokens_per_second": 5696.025
+ },
+ {
+ "epoch": 0.07069754915162942,
+ "grad_norm": 0.8565455428368292,
+ "learning_rate": 9.969259187874507e-06,
+ "loss": 0.40404400825500486,
+ "num_input_tokens_seen": 45846704,
+ "step": 1050,
+ "train_runtime": 8044.3345,
+ "train_tokens_per_second": 5699.254
+ },
+ {
+ "epoch": 0.07103420414758956,
+ "grad_norm": 0.9631616045121916,
+ "learning_rate": 9.968965744991178e-06,
+ "loss": 0.43367910385131836,
+ "num_input_tokens_seen": 46069384,
+ "step": 1055,
+ "train_runtime": 8075.3735,
+ "train_tokens_per_second": 5704.923
+ },
+ {
+ "epoch": 0.0713708591435497,
+ "grad_norm": 0.923891258768515,
+ "learning_rate": 9.968670912552422e-06,
+ "loss": 0.41605100631713865,
+ "num_input_tokens_seen": 46289280,
+ "step": 1060,
+ "train_runtime": 8110.2287,
+ "train_tokens_per_second": 5707.518
+ },
+ {
+ "epoch": 0.07170751413950983,
+ "grad_norm": 0.9199449654930428,
+ "learning_rate": 9.968374690640692e-06,
+ "loss": 0.4259674072265625,
+ "num_input_tokens_seen": 46513832,
+ "step": 1065,
+ "train_runtime": 8149.1273,
+ "train_tokens_per_second": 5707.83
+ },
+ {
+ "epoch": 0.07204416913546997,
+ "grad_norm": 0.9349037827416551,
+ "learning_rate": 9.968077079338822e-06,
+ "loss": 0.44624719619750974,
+ "num_input_tokens_seen": 46732528,
+ "step": 1070,
+ "train_runtime": 8183.763,
+ "train_tokens_per_second": 5710.396
+ },
+ {
+ "epoch": 0.07238082413143011,
+ "grad_norm": 0.9411351171686876,
+ "learning_rate": 9.967778078730038e-06,
+ "loss": 0.4426886558532715,
+ "num_input_tokens_seen": 46948464,
+ "step": 1075,
+ "train_runtime": 8220.2593,
+ "train_tokens_per_second": 5711.312
+ },
+ {
+ "epoch": 0.07271747912739025,
+ "grad_norm": 0.9433124357981822,
+ "learning_rate": 9.967477688897957e-06,
+ "loss": 0.44336910247802735,
+ "num_input_tokens_seen": 47171032,
+ "step": 1080,
+ "train_runtime": 8259.7389,
+ "train_tokens_per_second": 5710.959
+ },
+ {
+ "epoch": 0.07305413412335039,
+ "grad_norm": 0.8570101786695654,
+ "learning_rate": 9.967175909926578e-06,
+ "loss": 0.46260600090026854,
+ "num_input_tokens_seen": 47402968,
+ "step": 1085,
+ "train_runtime": 8293.6887,
+ "train_tokens_per_second": 5715.547
+ },
+ {
+ "epoch": 0.07339078911931053,
+ "grad_norm": 0.967581695257143,
+ "learning_rate": 9.966872741900296e-06,
+ "loss": 0.4760030746459961,
+ "num_input_tokens_seen": 47634040,
+ "step": 1090,
+ "train_runtime": 8342.26,
+ "train_tokens_per_second": 5709.968
+ },
+ {
+ "epoch": 0.07372744411527067,
+ "grad_norm": 1.3602823694752813,
+ "learning_rate": 9.96656818490389e-06,
+ "loss": 0.4523357391357422,
+ "num_input_tokens_seen": 47849424,
+ "step": 1095,
+ "train_runtime": 8378.4116,
+ "train_tokens_per_second": 5711.038
+ },
+ {
+ "epoch": 0.07406409911123081,
+ "grad_norm": 0.8798163617081339,
+ "learning_rate": 9.966262239022529e-06,
+ "loss": 0.48172898292541505,
+ "num_input_tokens_seen": 48074544,
+ "step": 1100,
+ "train_runtime": 8414.2774,
+ "train_tokens_per_second": 5713.449
+ },
+ {
+ "epoch": 0.07440075410719095,
+ "grad_norm": 0.8689729621566785,
+ "learning_rate": 9.965954904341768e-06,
+ "loss": 0.4594748497009277,
+ "num_input_tokens_seen": 48294808,
+ "step": 1105,
+ "train_runtime": 8453.2631,
+ "train_tokens_per_second": 5713.156
+ },
+ {
+ "epoch": 0.07473740910315109,
+ "grad_norm": 0.9298500312875317,
+ "learning_rate": 9.965646180947555e-06,
+ "loss": 0.44074625968933107,
+ "num_input_tokens_seen": 48509312,
+ "step": 1110,
+ "train_runtime": 8496.7093,
+ "train_tokens_per_second": 5709.188
+ },
+ {
+ "epoch": 0.07507406409911123,
+ "grad_norm": 0.9049700587066084,
+ "learning_rate": 9.96533606892622e-06,
+ "loss": 0.42275848388671877,
+ "num_input_tokens_seen": 48715504,
+ "step": 1115,
+ "train_runtime": 8533.5132,
+ "train_tokens_per_second": 5708.728
+ },
+ {
+ "epoch": 0.07541071909507137,
+ "grad_norm": 0.882919522521347,
+ "learning_rate": 9.965024568364487e-06,
+ "loss": 0.4448400497436523,
+ "num_input_tokens_seen": 48944392,
+ "step": 1120,
+ "train_runtime": 8571.3237,
+ "train_tokens_per_second": 5710.249
+ },
+ {
+ "epoch": 0.07574737409103151,
+ "grad_norm": 0.9374688066893563,
+ "learning_rate": 9.964711679349467e-06,
+ "loss": 0.41461782455444335,
+ "num_input_tokens_seen": 49155824,
+ "step": 1125,
+ "train_runtime": 8611.0892,
+ "train_tokens_per_second": 5708.433
+ },
+ {
+ "epoch": 0.07608402908699165,
+ "grad_norm": 0.7995128542211772,
+ "learning_rate": 9.964397401968656e-06,
+ "loss": 0.46135854721069336,
+ "num_input_tokens_seen": 49380912,
+ "step": 1130,
+ "train_runtime": 8650.8346,
+ "train_tokens_per_second": 5708.225
+ },
+ {
+ "epoch": 0.07642068408295179,
+ "grad_norm": 0.8607214899503258,
+ "learning_rate": 9.964081736309942e-06,
+ "loss": 0.41400885581970215,
+ "num_input_tokens_seen": 49617456,
+ "step": 1135,
+ "train_runtime": 8688.4749,
+ "train_tokens_per_second": 5710.721
+ },
+ {
+ "epoch": 0.07675733907891193,
+ "grad_norm": 0.8715957366665356,
+ "learning_rate": 9.963764682461599e-06,
+ "loss": 0.4346445083618164,
+ "num_input_tokens_seen": 49848584,
+ "step": 1140,
+ "train_runtime": 8722.1234,
+ "train_tokens_per_second": 5715.189
+ },
+ {
+ "epoch": 0.07709399407487207,
+ "grad_norm": 0.8561235152822287,
+ "learning_rate": 9.96344624051229e-06,
+ "loss": 0.4340538501739502,
+ "num_input_tokens_seen": 50065696,
+ "step": 1145,
+ "train_runtime": 8762.0281,
+ "train_tokens_per_second": 5713.939
+ },
+ {
+ "epoch": 0.0774306490708322,
+ "grad_norm": 0.929804620150381,
+ "learning_rate": 9.96312641055107e-06,
+ "loss": 0.4169576644897461,
+ "num_input_tokens_seen": 50283968,
+ "step": 1150,
+ "train_runtime": 8801.5092,
+ "train_tokens_per_second": 5713.107
+ },
+ {
+ "epoch": 0.07776730406679234,
+ "grad_norm": 0.752419655730966,
+ "learning_rate": 9.962805192667375e-06,
+ "loss": 0.41555070877075195,
+ "num_input_tokens_seen": 50500904,
+ "step": 1155,
+ "train_runtime": 8848.2982,
+ "train_tokens_per_second": 5707.414
+ },
+ {
+ "epoch": 0.07810395906275248,
+ "grad_norm": 1.0612801529163063,
+ "learning_rate": 9.962482586951033e-06,
+ "loss": 0.4202773094177246,
+ "num_input_tokens_seen": 50733408,
+ "step": 1160,
+ "train_runtime": 8889.0849,
+ "train_tokens_per_second": 5707.383
+ },
+ {
+ "epoch": 0.07844061405871264,
+ "grad_norm": 0.9990837210097588,
+ "learning_rate": 9.962158593492258e-06,
+ "loss": 0.442533016204834,
+ "num_input_tokens_seen": 50939416,
+ "step": 1165,
+ "train_runtime": 8930.0327,
+ "train_tokens_per_second": 5704.281
+ },
+ {
+ "epoch": 0.07877726905467278,
+ "grad_norm": 0.8491329656110541,
+ "learning_rate": 9.961833212381657e-06,
+ "loss": 0.4386699676513672,
+ "num_input_tokens_seen": 51160880,
+ "step": 1170,
+ "train_runtime": 8967.7862,
+ "train_tokens_per_second": 5704.962
+ },
+ {
+ "epoch": 0.07911392405063292,
+ "grad_norm": 0.8233317233024117,
+ "learning_rate": 9.961506443710221e-06,
+ "loss": 0.4202608108520508,
+ "num_input_tokens_seen": 51382192,
+ "step": 1175,
+ "train_runtime": 9009.4779,
+ "train_tokens_per_second": 5703.126
+ },
+ {
+ "epoch": 0.07945057904659306,
+ "grad_norm": 1.1740908151711091,
+ "learning_rate": 9.961178287569326e-06,
+ "loss": 0.4113266944885254,
+ "num_input_tokens_seen": 51610400,
+ "step": 1180,
+ "train_runtime": 9043.5038,
+ "train_tokens_per_second": 5706.903
+ },
+ {
+ "epoch": 0.0797872340425532,
+ "grad_norm": 0.9683671789990694,
+ "learning_rate": 9.960848744050747e-06,
+ "loss": 0.44871106147766116,
+ "num_input_tokens_seen": 51828672,
+ "step": 1185,
+ "train_runtime": 9085.7121,
+ "train_tokens_per_second": 5704.415
+ },
+ {
+ "epoch": 0.08012388903851334,
+ "grad_norm": 0.7384402440175095,
+ "learning_rate": 9.960517813246633e-06,
+ "loss": 0.44126176834106445,
+ "num_input_tokens_seen": 52038752,
+ "step": 1190,
+ "train_runtime": 9126.3974,
+ "train_tokens_per_second": 5702.004
+ },
+ {
+ "epoch": 0.08046054403447347,
+ "grad_norm": 0.9518446142368773,
+ "learning_rate": 9.96018549524953e-06,
+ "loss": 0.4168846607208252,
+ "num_input_tokens_seen": 52257272,
+ "step": 1195,
+ "train_runtime": 9165.705,
+ "train_tokens_per_second": 5701.391
+ },
+ {
+ "epoch": 0.08079719903043361,
+ "grad_norm": 0.9426346640072512,
+ "learning_rate": 9.959851790152371e-06,
+ "loss": 0.41556696891784667,
+ "num_input_tokens_seen": 52478032,
+ "step": 1200,
+ "train_runtime": 9204.8312,
+ "train_tokens_per_second": 5701.14
+ },
+ {
+ "epoch": 0.08113385402639375,
+ "grad_norm": 0.8462737285766736,
+ "learning_rate": 9.959516698048475e-06,
+ "loss": 0.3835282325744629,
+ "num_input_tokens_seen": 52684592,
+ "step": 1205,
+ "train_runtime": 9245.4602,
+ "train_tokens_per_second": 5698.428
+ },
+ {
+ "epoch": 0.08147050902235389,
+ "grad_norm": 0.8765077742279136,
+ "learning_rate": 9.959180219031551e-06,
+ "loss": 0.4106024742126465,
+ "num_input_tokens_seen": 52895048,
+ "step": 1210,
+ "train_runtime": 9284.6342,
+ "train_tokens_per_second": 5697.052
+ },
+ {
+ "epoch": 0.08180716401831403,
+ "grad_norm": 0.8799539327811132,
+ "learning_rate": 9.958842353195688e-06,
+ "loss": 0.4241660118103027,
+ "num_input_tokens_seen": 53122496,
+ "step": 1215,
+ "train_runtime": 9323.4499,
+ "train_tokens_per_second": 5697.73
+ },
+ {
+ "epoch": 0.08214381901427417,
+ "grad_norm": 0.8851017529748658,
+ "learning_rate": 9.958503100635377e-06,
+ "loss": 0.43586392402648927,
+ "num_input_tokens_seen": 53331344,
+ "step": 1220,
+ "train_runtime": 9359.4211,
+ "train_tokens_per_second": 5698.146
+ },
+ {
+ "epoch": 0.08248047401023431,
+ "grad_norm": 0.9611985454606802,
+ "learning_rate": 9.958162461445484e-06,
+ "loss": 0.4607269287109375,
+ "num_input_tokens_seen": 53560160,
+ "step": 1225,
+ "train_runtime": 9398.2675,
+ "train_tokens_per_second": 5698.94
+ },
+ {
+ "epoch": 0.08281712900619445,
+ "grad_norm": 0.9067497311826802,
+ "learning_rate": 9.957820435721271e-06,
+ "loss": 0.4349203586578369,
+ "num_input_tokens_seen": 53775560,
+ "step": 1230,
+ "train_runtime": 9435.4003,
+ "train_tokens_per_second": 5699.341
+ },
+ {
+ "epoch": 0.08315378400215459,
+ "grad_norm": 0.963108721688269,
+ "learning_rate": 9.95747702355838e-06,
+ "loss": 0.45670351982116697,
+ "num_input_tokens_seen": 54000808,
+ "step": 1235,
+ "train_runtime": 9470.0321,
+ "train_tokens_per_second": 5702.284
+ },
+ {
+ "epoch": 0.08349043899811473,
+ "grad_norm": 0.8513390607173822,
+ "learning_rate": 9.957132225052848e-06,
+ "loss": 0.4656820297241211,
+ "num_input_tokens_seen": 54218336,
+ "step": 1240,
+ "train_runtime": 9509.4298,
+ "train_tokens_per_second": 5701.534
+ },
+ {
+ "epoch": 0.08382709399407487,
+ "grad_norm": 0.8033639657062732,
+ "learning_rate": 9.956786040301098e-06,
+ "loss": 0.407243537902832,
+ "num_input_tokens_seen": 54435576,
+ "step": 1245,
+ "train_runtime": 9549.0218,
+ "train_tokens_per_second": 5700.644
+ },
+ {
+ "epoch": 0.08416374899003501,
+ "grad_norm": 0.9112338258739832,
+ "learning_rate": 9.956438469399936e-06,
+ "loss": 0.42780628204345705,
+ "num_input_tokens_seen": 54646448,
+ "step": 1250,
+ "train_runtime": 9584.5449,
+ "train_tokens_per_second": 5701.517
+ },
+ {
+ "epoch": 0.08450040398599515,
+ "grad_norm": 0.898524764750954,
+ "learning_rate": 9.956089512446562e-06,
+ "loss": 0.4281005859375,
+ "num_input_tokens_seen": 54849384,
+ "step": 1255,
+ "train_runtime": 9624.8869,
+ "train_tokens_per_second": 5698.704
+ },
+ {
+ "epoch": 0.08483705898195529,
+ "grad_norm": 0.9393978257221618,
+ "learning_rate": 9.955739169538559e-06,
+ "loss": 0.43431816101074217,
+ "num_input_tokens_seen": 55062944,
+ "step": 1260,
+ "train_runtime": 9664.3501,
+ "train_tokens_per_second": 5697.532
+ },
+ {
+ "epoch": 0.08517371397791543,
+ "grad_norm": 0.9086736133443781,
+ "learning_rate": 9.955387440773902e-06,
+ "loss": 0.4328322887420654,
+ "num_input_tokens_seen": 55267032,
+ "step": 1265,
+ "train_runtime": 9704.5196,
+ "train_tokens_per_second": 5694.979
+ },
+ {
+ "epoch": 0.08551036897387557,
+ "grad_norm": 1.1824810596928264,
+ "learning_rate": 9.955034326250947e-06,
+ "loss": 0.42726478576660154,
+ "num_input_tokens_seen": 55485176,
+ "step": 1270,
+ "train_runtime": 9740.2982,
+ "train_tokens_per_second": 5696.456
+ },
+ {
+ "epoch": 0.0858470239698357,
+ "grad_norm": 0.981019269531651,
+ "learning_rate": 9.954679826068444e-06,
+ "loss": 0.431905460357666,
+ "num_input_tokens_seen": 55697472,
+ "step": 1275,
+ "train_runtime": 9778.2861,
+ "train_tokens_per_second": 5696.036
+ },
+ {
+ "epoch": 0.08618367896579585,
+ "grad_norm": 0.9167947068742166,
+ "learning_rate": 9.954323940325526e-06,
+ "loss": 0.4367355823516846,
+ "num_input_tokens_seen": 55909208,
+ "step": 1280,
+ "train_runtime": 9815.8422,
+ "train_tokens_per_second": 5695.814
+ },
+ {
+ "epoch": 0.086520333961756,
+ "grad_norm": 0.8284962122347568,
+ "learning_rate": 9.953966669121715e-06,
+ "loss": 0.4562966346740723,
+ "num_input_tokens_seen": 56144336,
+ "step": 1285,
+ "train_runtime": 9855.4694,
+ "train_tokens_per_second": 5696.769
+ },
+ {
+ "epoch": 0.08685698895771614,
+ "grad_norm": 0.9908373153090002,
+ "learning_rate": 9.953608012556924e-06,
+ "loss": 0.4428553581237793,
+ "num_input_tokens_seen": 56370632,
+ "step": 1290,
+ "train_runtime": 9900.5066,
+ "train_tokens_per_second": 5693.712
+ },
+ {
+ "epoch": 0.08719364395367628,
+ "grad_norm": 0.8603920936193332,
+ "learning_rate": 9.953247970731449e-06,
+ "loss": 0.4518705368041992,
+ "num_input_tokens_seen": 56581824,
+ "step": 1295,
+ "train_runtime": 9941.9268,
+ "train_tokens_per_second": 5691.233
+ },
+ {
+ "epoch": 0.08753029894963642,
+ "grad_norm": 1.0065353834353872,
+ "learning_rate": 9.952886543745973e-06,
+ "loss": 0.45711684226989746,
+ "num_input_tokens_seen": 56797912,
+ "step": 1300,
+ "train_runtime": 9983.8473,
+ "train_tokens_per_second": 5688.98
+ },
+ {
+ "epoch": 0.08786695394559656,
+ "grad_norm": 0.9704989049669684,
+ "learning_rate": 9.952523731701566e-06,
+ "loss": 0.43997964859008787,
+ "num_input_tokens_seen": 57028400,
+ "step": 1305,
+ "train_runtime": 10023.0235,
+ "train_tokens_per_second": 5689.74
+ },
+ {
+ "epoch": 0.0882036089415567,
+ "grad_norm": 0.7746848672115061,
+ "learning_rate": 9.952159534699693e-06,
+ "loss": 0.39839365482330324,
+ "num_input_tokens_seen": 57239112,
+ "step": 1310,
+ "train_runtime": 10060.0427,
+ "train_tokens_per_second": 5689.748
+ },
+ {
+ "epoch": 0.08854026393751684,
+ "grad_norm": 0.8703667437003134,
+ "learning_rate": 9.951793952842197e-06,
+ "loss": 0.46162776947021483,
+ "num_input_tokens_seen": 57451840,
+ "step": 1315,
+ "train_runtime": 10090.9799,
+ "train_tokens_per_second": 5693.386
+ },
+ {
+ "epoch": 0.08887691893347698,
+ "grad_norm": 0.9970523155610451,
+ "learning_rate": 9.95142698623131e-06,
+ "loss": 0.4125502109527588,
+ "num_input_tokens_seen": 57649616,
+ "step": 1320,
+ "train_runtime": 10134.3943,
+ "train_tokens_per_second": 5688.511
+ },
+ {
+ "epoch": 0.08921357392943711,
+ "grad_norm": 0.8982754367798736,
+ "learning_rate": 9.951058634969657e-06,
+ "loss": 0.4594625473022461,
+ "num_input_tokens_seen": 57859504,
+ "step": 1325,
+ "train_runtime": 10173.3422,
+ "train_tokens_per_second": 5687.364
+ },
+ {
+ "epoch": 0.08955022892539725,
+ "grad_norm": 0.8441984007212928,
+ "learning_rate": 9.950688899160244e-06,
+ "loss": 0.42917823791503906,
+ "num_input_tokens_seen": 58092104,
+ "step": 1330,
+ "train_runtime": 10208.5539,
+ "train_tokens_per_second": 5690.532
+ },
+ {
+ "epoch": 0.0898868839213574,
+ "grad_norm": 0.8326739767212783,
+ "learning_rate": 9.950317778906469e-06,
+ "loss": 0.4345374584197998,
+ "num_input_tokens_seen": 58319256,
+ "step": 1335,
+ "train_runtime": 10249.3358,
+ "train_tokens_per_second": 5690.052
+ },
+ {
+ "epoch": 0.09022353891731753,
+ "grad_norm": 0.8294890352349942,
+ "learning_rate": 9.949945274312109e-06,
+ "loss": 0.40816512107849123,
+ "num_input_tokens_seen": 58540424,
+ "step": 1340,
+ "train_runtime": 10287.6498,
+ "train_tokens_per_second": 5690.359
+ },
+ {
+ "epoch": 0.09056019391327767,
+ "grad_norm": 0.968521031703479,
+ "learning_rate": 9.94957138548134e-06,
+ "loss": 0.4445356369018555,
+ "num_input_tokens_seen": 58736800,
+ "step": 1345,
+ "train_runtime": 10326.024,
+ "train_tokens_per_second": 5688.23
+ },
+ {
+ "epoch": 0.09089684890923781,
+ "grad_norm": 1.007040150938243,
+ "learning_rate": 9.949196112518715e-06,
+ "loss": 0.41266183853149413,
+ "num_input_tokens_seen": 58963840,
+ "step": 1350,
+ "train_runtime": 10361.3193,
+ "train_tokens_per_second": 5690.766
+ },
+ {
+ "epoch": 0.09123350390519795,
+ "grad_norm": 0.8518178617284553,
+ "learning_rate": 9.94881945552918e-06,
+ "loss": 0.43543500900268556,
+ "num_input_tokens_seen": 59186072,
+ "step": 1355,
+ "train_runtime": 10406.6964,
+ "train_tokens_per_second": 5687.307
+ },
+ {
+ "epoch": 0.09157015890115809,
+ "grad_norm": 0.9351761092832128,
+ "learning_rate": 9.948441414618064e-06,
+ "loss": 0.45530238151550295,
+ "num_input_tokens_seen": 59396096,
+ "step": 1360,
+ "train_runtime": 10448.5625,
+ "train_tokens_per_second": 5684.619
+ },
+ {
+ "epoch": 0.09190681389711823,
+ "grad_norm": 0.976556409327906,
+ "learning_rate": 9.948061989891085e-06,
+ "loss": 0.41725835800170896,
+ "num_input_tokens_seen": 59621960,
+ "step": 1365,
+ "train_runtime": 10484.4296,
+ "train_tokens_per_second": 5686.715
+ },
+ {
+ "epoch": 0.09224346889307837,
+ "grad_norm": 0.9413072947091584,
+ "learning_rate": 9.94768118145435e-06,
+ "loss": 0.4238771438598633,
+ "num_input_tokens_seen": 59834688,
+ "step": 1370,
+ "train_runtime": 10521.7898,
+ "train_tokens_per_second": 5686.74
+ },
+ {
+ "epoch": 0.09258012388903851,
+ "grad_norm": 0.8072753824687519,
+ "learning_rate": 9.947298989414349e-06,
+ "loss": 0.4197520732879639,
+ "num_input_tokens_seen": 60061512,
+ "step": 1375,
+ "train_runtime": 10562.6863,
+ "train_tokens_per_second": 5686.197
+ },
+ {
+ "epoch": 0.09291677888499865,
+ "grad_norm": 0.8243851964827824,
+ "learning_rate": 9.946915413877963e-06,
+ "loss": 0.4294243812561035,
+ "num_input_tokens_seen": 60289152,
+ "step": 1380,
+ "train_runtime": 10608.7449,
+ "train_tokens_per_second": 5682.967
+ },
+ {
+ "epoch": 0.09325343388095879,
+ "grad_norm": 0.8639433441786879,
+ "learning_rate": 9.946530454952454e-06,
+ "loss": 0.44629898071289065,
+ "num_input_tokens_seen": 60516576,
+ "step": 1385,
+ "train_runtime": 10649.4064,
+ "train_tokens_per_second": 5682.624
+ },
+ {
+ "epoch": 0.09359008887691893,
+ "grad_norm": 0.9381874141159465,
+ "learning_rate": 9.946144112745477e-06,
+ "loss": 0.4549723625183105,
+ "num_input_tokens_seen": 60744096,
+ "step": 1390,
+ "train_runtime": 10690.6621,
+ "train_tokens_per_second": 5681.977
+ },
+ {
+ "epoch": 0.09392674387287907,
+ "grad_norm": 0.9598970905259178,
+ "learning_rate": 9.94575638736507e-06,
+ "loss": 0.4265571594238281,
+ "num_input_tokens_seen": 60970464,
+ "step": 1395,
+ "train_runtime": 10730.6636,
+ "train_tokens_per_second": 5681.891
+ },
+ {
+ "epoch": 0.09426339886883921,
+ "grad_norm": 0.8882564184072146,
+ "learning_rate": 9.945367278919662e-06,
+ "loss": 0.469175910949707,
+ "num_input_tokens_seen": 61195536,
+ "step": 1400,
+ "train_runtime": 10771.6796,
+ "train_tokens_per_second": 5681.151
+ },
+ {
+ "epoch": 0.09460005386479936,
+ "grad_norm": 0.8345645071811238,
+ "learning_rate": 9.944976787518064e-06,
+ "loss": 0.43855986595153806,
+ "num_input_tokens_seen": 61415400,
+ "step": 1405,
+ "train_runtime": 10804.9607,
+ "train_tokens_per_second": 5684.0
+ },
+ {
+ "epoch": 0.0949367088607595,
+ "grad_norm": 1.0139184695279284,
+ "learning_rate": 9.944584913269474e-06,
+ "loss": 0.43834495544433594,
+ "num_input_tokens_seen": 61645440,
+ "step": 1410,
+ "train_runtime": 10846.1692,
+ "train_tokens_per_second": 5683.614
+ },
+ {
+ "epoch": 0.09527336385671964,
+ "grad_norm": 0.8171057323619708,
+ "learning_rate": 9.944191656283481e-06,
+ "loss": 0.41381168365478516,
+ "num_input_tokens_seen": 61868952,
+ "step": 1415,
+ "train_runtime": 10885.6156,
+ "train_tokens_per_second": 5683.551
+ },
+ {
+ "epoch": 0.09561001885267978,
+ "grad_norm": 0.8671727475302144,
+ "learning_rate": 9.943797016670059e-06,
+ "loss": 0.41778225898742677,
+ "num_input_tokens_seen": 62081744,
+ "step": 1420,
+ "train_runtime": 10924.1913,
+ "train_tokens_per_second": 5682.96
+ },
+ {
+ "epoch": 0.09594667384863992,
+ "grad_norm": 0.9139643587899275,
+ "learning_rate": 9.943400994539565e-06,
+ "loss": 0.41177783012390134,
+ "num_input_tokens_seen": 62299552,
+ "step": 1425,
+ "train_runtime": 10963.2034,
+ "train_tokens_per_second": 5682.605
+ },
+ {
+ "epoch": 0.09628332884460006,
+ "grad_norm": 0.8942790678032806,
+ "learning_rate": 9.943003590002744e-06,
+ "loss": 0.4324612617492676,
+ "num_input_tokens_seen": 62524424,
+ "step": 1430,
+ "train_runtime": 11003.5976,
+ "train_tokens_per_second": 5682.18
+ },
+ {
+ "epoch": 0.0966199838405602,
+ "grad_norm": 0.9134672447344226,
+ "learning_rate": 9.942604803170733e-06,
+ "loss": 0.4207320690155029,
+ "num_input_tokens_seen": 62736736,
+ "step": 1435,
+ "train_runtime": 11041.0071,
+ "train_tokens_per_second": 5682.157
+ },
+ {
+ "epoch": 0.09695663883652034,
+ "grad_norm": 1.078284335277287,
+ "learning_rate": 9.94220463415505e-06,
+ "loss": 0.4587702751159668,
+ "num_input_tokens_seen": 62950552,
+ "step": 1440,
+ "train_runtime": 11082.5134,
+ "train_tokens_per_second": 5680.169
+ },
+ {
+ "epoch": 0.09729329383248048,
+ "grad_norm": 0.8211105532018887,
+ "learning_rate": 9.9418030830676e-06,
+ "loss": 0.4369199752807617,
+ "num_input_tokens_seen": 63173304,
+ "step": 1445,
+ "train_runtime": 11119.7396,
+ "train_tokens_per_second": 5681.186
+ },
+ {
+ "epoch": 0.09762994882844062,
+ "grad_norm": 0.9975227482376734,
+ "learning_rate": 9.941400150020676e-06,
+ "loss": 0.45250940322875977,
+ "num_input_tokens_seen": 63376456,
+ "step": 1450,
+ "train_runtime": 11153.3434,
+ "train_tokens_per_second": 5682.283
+ },
+ {
+ "epoch": 0.09796660382440076,
+ "grad_norm": 0.875023853319168,
+ "learning_rate": 9.940995835126957e-06,
+ "loss": 0.4510367393493652,
+ "num_input_tokens_seen": 63598704,
+ "step": 1455,
+ "train_runtime": 11190.4484,
+ "train_tokens_per_second": 5683.303
+ },
+ {
+ "epoch": 0.0983032588203609,
+ "grad_norm": 0.7795257917882069,
+ "learning_rate": 9.940590138499508e-06,
+ "loss": 0.4070897102355957,
+ "num_input_tokens_seen": 63824984,
+ "step": 1460,
+ "train_runtime": 11228.0323,
+ "train_tokens_per_second": 5684.432
+ },
+ {
+ "epoch": 0.09863991381632103,
+ "grad_norm": 0.9802634486624268,
+ "learning_rate": 9.940183060251783e-06,
+ "loss": 0.424533748626709,
+ "num_input_tokens_seen": 64020440,
+ "step": 1465,
+ "train_runtime": 11269.0538,
+ "train_tokens_per_second": 5681.084
+ },
+ {
+ "epoch": 0.09897656881228117,
+ "grad_norm": 0.8129225031494536,
+ "learning_rate": 9.939774600497616e-06,
+ "loss": 0.3772727489471436,
+ "num_input_tokens_seen": 64237912,
+ "step": 1470,
+ "train_runtime": 11305.1846,
+ "train_tokens_per_second": 5682.164
+ },
+ {
+ "epoch": 0.09931322380824131,
+ "grad_norm": 0.876095085095231,
+ "learning_rate": 9.939364759351236e-06,
+ "loss": 0.43614888191223145,
+ "num_input_tokens_seen": 64452808,
+ "step": 1475,
+ "train_runtime": 11342.3271,
+ "train_tokens_per_second": 5682.503
+ },
+ {
+ "epoch": 0.09964987880420145,
+ "grad_norm": 0.8249502349163025,
+ "learning_rate": 9.938953536927249e-06,
+ "loss": 0.4052250862121582,
+ "num_input_tokens_seen": 64679536,
+ "step": 1480,
+ "train_runtime": 11380.8832,
+ "train_tokens_per_second": 5683.174
+ },
+ {
+ "epoch": 0.09998653380016159,
+ "grad_norm": 0.8004037055954487,
+ "learning_rate": 9.938540933340657e-06,
+ "loss": 0.43981304168701174,
+ "num_input_tokens_seen": 64907672,
+ "step": 1485,
+ "train_runtime": 11414.3496,
+ "train_tokens_per_second": 5686.498
+ },
+ {
+ "epoch": 0.10032318879612173,
+ "grad_norm": 0.7556539983829699,
+ "learning_rate": 9.938126948706839e-06,
+ "loss": 0.4235081195831299,
+ "num_input_tokens_seen": 65134656,
+ "step": 1490,
+ "train_runtime": 11451.1493,
+ "train_tokens_per_second": 5688.045
+ },
+ {
+ "epoch": 0.10065984379208187,
+ "grad_norm": 0.7974202081471553,
+ "learning_rate": 9.937711583141567e-06,
+ "loss": 0.41340045928955077,
+ "num_input_tokens_seen": 65353344,
+ "step": 1495,
+ "train_runtime": 11488.8753,
+ "train_tokens_per_second": 5688.402
+ },
+ {
+ "epoch": 0.10099649878804201,
+ "grad_norm": 0.9113055103223531,
+ "learning_rate": 9.937294836760995e-06,
+ "loss": 0.43211612701416013,
+ "num_input_tokens_seen": 65570904,
+ "step": 1500,
+ "train_runtime": 11525.754,
+ "train_tokens_per_second": 5689.077
+ },
+ {
+ "epoch": 0.10133315378400215,
+ "grad_norm": 0.9362482868425025,
+ "learning_rate": 9.936876709681668e-06,
+ "loss": 0.4765140533447266,
+ "num_input_tokens_seen": 65784200,
+ "step": 1505,
+ "train_runtime": 11567.0824,
+ "train_tokens_per_second": 5687.19
+ },
+ {
+ "epoch": 0.10166980877996229,
+ "grad_norm": 0.7345279591609306,
+ "learning_rate": 9.93645720202051e-06,
+ "loss": 0.3693789005279541,
+ "num_input_tokens_seen": 66002240,
+ "step": 1510,
+ "train_runtime": 11603.3588,
+ "train_tokens_per_second": 5688.201
+ },
+ {
+ "epoch": 0.10200646377592243,
+ "grad_norm": 0.8440448421494396,
+ "learning_rate": 9.936036313894836e-06,
+ "loss": 0.46529722213745117,
+ "num_input_tokens_seen": 66217160,
+ "step": 1515,
+ "train_runtime": 11637.8846,
+ "train_tokens_per_second": 5689.793
+ },
+ {
+ "epoch": 0.10234311877188257,
+ "grad_norm": 0.9288537021329782,
+ "learning_rate": 9.935614045422349e-06,
+ "loss": 0.4474299907684326,
+ "num_input_tokens_seen": 66437216,
+ "step": 1520,
+ "train_runtime": 11674.3858,
+ "train_tokens_per_second": 5690.853
+ },
+ {
+ "epoch": 0.10267977376784271,
+ "grad_norm": 0.7816250239377797,
+ "learning_rate": 9.935190396721132e-06,
+ "loss": 0.4093039035797119,
+ "num_input_tokens_seen": 66659632,
+ "step": 1525,
+ "train_runtime": 11714.7321,
+ "train_tokens_per_second": 5690.24
+ },
+ {
+ "epoch": 0.10301642876380286,
+ "grad_norm": 0.868350723790852,
+ "learning_rate": 9.934765367909658e-06,
+ "loss": 0.442030143737793,
+ "num_input_tokens_seen": 66874200,
+ "step": 1530,
+ "train_runtime": 11751.0775,
+ "train_tokens_per_second": 5690.899
+ },
+ {
+ "epoch": 0.103353083759763,
+ "grad_norm": 0.9654553160320544,
+ "learning_rate": 9.934338959106783e-06,
+ "loss": 0.4125958442687988,
+ "num_input_tokens_seen": 67096584,
+ "step": 1535,
+ "train_runtime": 11787.6988,
+ "train_tokens_per_second": 5692.085
+ },
+ {
+ "epoch": 0.10368973875572314,
+ "grad_norm": 1.0001250180683936,
+ "learning_rate": 9.933911170431753e-06,
+ "loss": 0.45245895385742185,
+ "num_input_tokens_seen": 67300344,
+ "step": 1540,
+ "train_runtime": 11825.444,
+ "train_tokens_per_second": 5691.147
+ },
+ {
+ "epoch": 0.10402639375168328,
+ "grad_norm": 0.9480950052674862,
+ "learning_rate": 9.933482002004198e-06,
+ "loss": 0.4369757652282715,
+ "num_input_tokens_seen": 67528208,
+ "step": 1545,
+ "train_runtime": 11868.3124,
+ "train_tokens_per_second": 5689.79
+ },
+ {
+ "epoch": 0.10436304874764342,
+ "grad_norm": 1.049191818835591,
+ "learning_rate": 9.933051453944134e-06,
+ "loss": 0.47658519744873046,
+ "num_input_tokens_seen": 67752368,
+ "step": 1550,
+ "train_runtime": 11899.3433,
+ "train_tokens_per_second": 5693.791
+ },
+ {
+ "epoch": 0.10469970374360356,
+ "grad_norm": 0.8519840991796084,
+ "learning_rate": 9.932619526371958e-06,
+ "loss": 0.4100327014923096,
+ "num_input_tokens_seen": 67972272,
+ "step": 1555,
+ "train_runtime": 11933.0537,
+ "train_tokens_per_second": 5696.134
+ },
+ {
+ "epoch": 0.1050363587395637,
+ "grad_norm": 0.8587262694316211,
+ "learning_rate": 9.932186219408463e-06,
+ "loss": 0.40094866752624514,
+ "num_input_tokens_seen": 68182312,
+ "step": 1560,
+ "train_runtime": 11973.5123,
+ "train_tokens_per_second": 5694.429
+ },
+ {
+ "epoch": 0.10537301373552384,
+ "grad_norm": 0.9397299864907294,
+ "learning_rate": 9.931751533174819e-06,
+ "loss": 0.44454145431518555,
+ "num_input_tokens_seen": 68389480,
+ "step": 1565,
+ "train_runtime": 12014.0801,
+ "train_tokens_per_second": 5692.444
+ },
+ {
+ "epoch": 0.10570966873148398,
+ "grad_norm": 0.9745376383539546,
+ "learning_rate": 9.931315467792584e-06,
+ "loss": 0.45273237228393554,
+ "num_input_tokens_seen": 68625472,
+ "step": 1570,
+ "train_runtime": 12045.9213,
+ "train_tokens_per_second": 5696.988
+ },
+ {
+ "epoch": 0.10604632372744412,
+ "grad_norm": 0.796989168878711,
+ "learning_rate": 9.930878023383705e-06,
+ "loss": 0.4475082874298096,
+ "num_input_tokens_seen": 68842256,
+ "step": 1575,
+ "train_runtime": 12081.8379,
+ "train_tokens_per_second": 5697.995
+ },
+ {
+ "epoch": 0.10638297872340426,
+ "grad_norm": 0.7777899346165759,
+ "learning_rate": 9.93043920007051e-06,
+ "loss": 0.39358346462249755,
+ "num_input_tokens_seen": 69063336,
+ "step": 1580,
+ "train_runtime": 12120.9279,
+ "train_tokens_per_second": 5697.859
+ },
+ {
+ "epoch": 0.1067196337193644,
+ "grad_norm": 0.856296656865995,
+ "learning_rate": 9.929998997975712e-06,
+ "loss": 0.40912957191467286,
+ "num_input_tokens_seen": 69297392,
+ "step": 1585,
+ "train_runtime": 12157.5274,
+ "train_tokens_per_second": 5699.958
+ },
+ {
+ "epoch": 0.10705628871532454,
+ "grad_norm": 0.8406798760366443,
+ "learning_rate": 9.929557417222417e-06,
+ "loss": 0.4154947280883789,
+ "num_input_tokens_seen": 69502272,
+ "step": 1590,
+ "train_runtime": 12195.0467,
+ "train_tokens_per_second": 5699.221
+ },
+ {
+ "epoch": 0.10739294371128467,
+ "grad_norm": 0.8506547801105618,
+ "learning_rate": 9.929114457934109e-06,
+ "loss": 0.433414888381958,
+ "num_input_tokens_seen": 69718712,
+ "step": 1595,
+ "train_runtime": 12235.8524,
+ "train_tokens_per_second": 5697.904
+ },
+ {
+ "epoch": 0.10772959870724481,
+ "grad_norm": 0.8306444786085215,
+ "learning_rate": 9.92867012023466e-06,
+ "loss": 0.429976749420166,
+ "num_input_tokens_seen": 69942920,
+ "step": 1600,
+ "train_runtime": 12275.1053,
+ "train_tokens_per_second": 5697.949
+ },
+ {
+ "epoch": 0.10806625370320495,
+ "grad_norm": 0.8083632874129185,
+ "learning_rate": 9.92822440424833e-06,
+ "loss": 0.465771484375,
+ "num_input_tokens_seen": 70167408,
+ "step": 1605,
+ "train_runtime": 12311.856,
+ "train_tokens_per_second": 5699.174
+ },
+ {
+ "epoch": 0.1084029086991651,
+ "grad_norm": 0.8623025104703945,
+ "learning_rate": 9.927777310099759e-06,
+ "loss": 0.4795114040374756,
+ "num_input_tokens_seen": 70398096,
+ "step": 1610,
+ "train_runtime": 12355.4012,
+ "train_tokens_per_second": 5697.759
+ },
+ {
+ "epoch": 0.10873956369512523,
+ "grad_norm": 0.8734280942811361,
+ "learning_rate": 9.927328837913976e-06,
+ "loss": 0.42298336029052735,
+ "num_input_tokens_seen": 70621368,
+ "step": 1615,
+ "train_runtime": 12392.0717,
+ "train_tokens_per_second": 5698.915
+ },
+ {
+ "epoch": 0.10907621869108537,
+ "grad_norm": 0.8755533500883431,
+ "learning_rate": 9.926878987816397e-06,
+ "loss": 0.4128897666931152,
+ "num_input_tokens_seen": 70850496,
+ "step": 1620,
+ "train_runtime": 12432.0003,
+ "train_tokens_per_second": 5699.042
+ },
+ {
+ "epoch": 0.10941287368704551,
+ "grad_norm": 0.7849562344468639,
+ "learning_rate": 9.92642775993282e-06,
+ "loss": 0.40384368896484374,
+ "num_input_tokens_seen": 71078912,
+ "step": 1625,
+ "train_runtime": 12472.6839,
+ "train_tokens_per_second": 5698.766
+ },
+ {
+ "epoch": 0.10974952868300565,
+ "grad_norm": 0.8314524460612794,
+ "learning_rate": 9.925975154389428e-06,
+ "loss": 0.39646482467651367,
+ "num_input_tokens_seen": 71291944,
+ "step": 1630,
+ "train_runtime": 12512.7857,
+ "train_tokens_per_second": 5697.528
+ },
+ {
+ "epoch": 0.11008618367896579,
+ "grad_norm": 0.8397438217714046,
+ "learning_rate": 9.925521171312792e-06,
+ "loss": 0.41344375610351564,
+ "num_input_tokens_seen": 71513056,
+ "step": 1635,
+ "train_runtime": 12552.1031,
+ "train_tokens_per_second": 5697.297
+ },
+ {
+ "epoch": 0.11042283867492593,
+ "grad_norm": 0.8717786503893655,
+ "learning_rate": 9.925065810829868e-06,
+ "loss": 0.43964757919311526,
+ "num_input_tokens_seen": 71712776,
+ "step": 1640,
+ "train_runtime": 12592.8232,
+ "train_tokens_per_second": 5694.734
+ },
+ {
+ "epoch": 0.11075949367088607,
+ "grad_norm": 0.8466265387114118,
+ "learning_rate": 9.924609073067994e-06,
+ "loss": 0.45009913444519045,
+ "num_input_tokens_seen": 71918888,
+ "step": 1645,
+ "train_runtime": 12637.095,
+ "train_tokens_per_second": 5691.093
+ },
+ {
+ "epoch": 0.11109614866684622,
+ "grad_norm": 0.8202088060176734,
+ "learning_rate": 9.924150958154897e-06,
+ "loss": 0.43289871215820314,
+ "num_input_tokens_seen": 72144880,
+ "step": 1650,
+ "train_runtime": 12679.4137,
+ "train_tokens_per_second": 5689.922
+ },
+ {
+ "epoch": 0.11143280366280636,
+ "grad_norm": 0.8241473767552461,
+ "learning_rate": 9.923691466218686e-06,
+ "loss": 0.4033914566040039,
+ "num_input_tokens_seen": 72368424,
+ "step": 1655,
+ "train_runtime": 12722.761,
+ "train_tokens_per_second": 5688.107
+ },
+ {
+ "epoch": 0.1117694586587665,
+ "grad_norm": 0.9179601239629808,
+ "learning_rate": 9.923230597387856e-06,
+ "loss": 0.4340330123901367,
+ "num_input_tokens_seen": 72604440,
+ "step": 1660,
+ "train_runtime": 12760.3543,
+ "train_tokens_per_second": 5689.845
+ },
+ {
+ "epoch": 0.11210611365472664,
+ "grad_norm": 0.9667829858136964,
+ "learning_rate": 9.922768351791289e-06,
+ "loss": 0.45328373908996583,
+ "num_input_tokens_seen": 72822440,
+ "step": 1665,
+ "train_runtime": 12804.6293,
+ "train_tokens_per_second": 5687.196
+ },
+ {
+ "epoch": 0.11244276865068678,
+ "grad_norm": 0.9239911563296863,
+ "learning_rate": 9.92230472955825e-06,
+ "loss": 0.41826558113098145,
+ "num_input_tokens_seen": 73042800,
+ "step": 1670,
+ "train_runtime": 12838.5402,
+ "train_tokens_per_second": 5689.338
+ },
+ {
+ "epoch": 0.11277942364664692,
+ "grad_norm": 0.8636254640310012,
+ "learning_rate": 9.92183973081839e-06,
+ "loss": 0.3946269750595093,
+ "num_input_tokens_seen": 73268056,
+ "step": 1675,
+ "train_runtime": 12882.6482,
+ "train_tokens_per_second": 5687.344
+ },
+ {
+ "epoch": 0.11311607864260706,
+ "grad_norm": 0.7861261803530913,
+ "learning_rate": 9.921373355701743e-06,
+ "loss": 0.46369400024414065,
+ "num_input_tokens_seen": 73499832,
+ "step": 1680,
+ "train_runtime": 12924.32,
+ "train_tokens_per_second": 5686.94
+ },
+ {
+ "epoch": 0.1134527336385672,
+ "grad_norm": 0.8384955151965564,
+ "learning_rate": 9.920905604338732e-06,
+ "loss": 0.45543785095214845,
+ "num_input_tokens_seen": 73720264,
+ "step": 1685,
+ "train_runtime": 12963.3961,
+ "train_tokens_per_second": 5686.802
+ },
+ {
+ "epoch": 0.11378938863452734,
+ "grad_norm": 1.0833650625320947,
+ "learning_rate": 9.920436476860158e-06,
+ "loss": 0.38962454795837403,
+ "num_input_tokens_seen": 73931824,
+ "step": 1690,
+ "train_runtime": 13004.1451,
+ "train_tokens_per_second": 5685.251
+ },
+ {
+ "epoch": 0.11412604363048748,
+ "grad_norm": 0.8250025055357804,
+ "learning_rate": 9.919965973397214e-06,
+ "loss": 0.40491409301757814,
+ "num_input_tokens_seen": 74150736,
+ "step": 1695,
+ "train_runtime": 13037.0352,
+ "train_tokens_per_second": 5687.699
+ },
+ {
+ "epoch": 0.11446269862644762,
+ "grad_norm": 0.841239081826391,
+ "learning_rate": 9.919494094081475e-06,
+ "loss": 0.45804147720336913,
+ "num_input_tokens_seen": 74352136,
+ "step": 1700,
+ "train_runtime": 13073.6778,
+ "train_tokens_per_second": 5687.163
+ },
+ {
+ "epoch": 0.11479935362240776,
+ "grad_norm": 0.9629476764947537,
+ "learning_rate": 9.919020839044899e-06,
+ "loss": 0.4247718334197998,
+ "num_input_tokens_seen": 74588328,
+ "step": 1705,
+ "train_runtime": 13108.3031,
+ "train_tokens_per_second": 5690.159
+ },
+ {
+ "epoch": 0.1151360086183679,
+ "grad_norm": 0.8166202703347624,
+ "learning_rate": 9.918546208419832e-06,
+ "loss": 0.4040073394775391,
+ "num_input_tokens_seen": 74810232,
+ "step": 1710,
+ "train_runtime": 13152.6429,
+ "train_tokens_per_second": 5687.848
+ },
+ {
+ "epoch": 0.11547266361432804,
+ "grad_norm": 0.9143911353222378,
+ "learning_rate": 9.918070202339e-06,
+ "loss": 0.41643967628479006,
+ "num_input_tokens_seen": 75038248,
+ "step": 1715,
+ "train_runtime": 13188.009,
+ "train_tokens_per_second": 5689.885
+ },
+ {
+ "epoch": 0.11580931861028818,
+ "grad_norm": 0.854163675009985,
+ "learning_rate": 9.91759282093552e-06,
+ "loss": 0.4160469532012939,
+ "num_input_tokens_seen": 75259984,
+ "step": 1720,
+ "train_runtime": 13221.5917,
+ "train_tokens_per_second": 5692.203
+ },
+ {
+ "epoch": 0.11614597360624832,
+ "grad_norm": 0.9111231961657891,
+ "learning_rate": 9.917114064342888e-06,
+ "loss": 0.38855648040771484,
+ "num_input_tokens_seen": 75468520,
+ "step": 1725,
+ "train_runtime": 13264.901,
+ "train_tokens_per_second": 5689.339
+ },
+ {
+ "epoch": 0.11648262860220845,
+ "grad_norm": 0.8667773958295081,
+ "learning_rate": 9.916633932694988e-06,
+ "loss": 0.3962873458862305,
+ "num_input_tokens_seen": 75690392,
+ "step": 1730,
+ "train_runtime": 13301.6754,
+ "train_tokens_per_second": 5690.29
+ },
+ {
+ "epoch": 0.1168192835981686,
+ "grad_norm": 0.8572351568987933,
+ "learning_rate": 9.916152426126086e-06,
+ "loss": 0.3940844774246216,
+ "num_input_tokens_seen": 75908632,
+ "step": 1735,
+ "train_runtime": 13340.3356,
+ "train_tokens_per_second": 5690.159
+ },
+ {
+ "epoch": 0.11715593859412873,
+ "grad_norm": 0.8920380227031728,
+ "learning_rate": 9.915669544770837e-06,
+ "loss": 0.41963586807250974,
+ "num_input_tokens_seen": 76143760,
+ "step": 1740,
+ "train_runtime": 13373.7492,
+ "train_tokens_per_second": 5693.524
+ },
+ {
+ "epoch": 0.11749259359008887,
+ "grad_norm": 0.8151573693579499,
+ "learning_rate": 9.915185288764272e-06,
+ "loss": 0.4006998062133789,
+ "num_input_tokens_seen": 76381464,
+ "step": 1745,
+ "train_runtime": 13411.6709,
+ "train_tokens_per_second": 5695.149
+ },
+ {
+ "epoch": 0.11782924858604901,
+ "grad_norm": 0.7823689731999574,
+ "learning_rate": 9.914699658241815e-06,
+ "loss": 0.39071149826049806,
+ "num_input_tokens_seen": 76607736,
+ "step": 1750,
+ "train_runtime": 13455.8214,
+ "train_tokens_per_second": 5693.278
+ },
+ {
+ "epoch": 0.11816590358200915,
+ "grad_norm": 0.9650032542802038,
+ "learning_rate": 9.91421265333927e-06,
+ "loss": 0.44945535659790037,
+ "num_input_tokens_seen": 76836152,
+ "step": 1755,
+ "train_runtime": 13495.7655,
+ "train_tokens_per_second": 5693.353
+ },
+ {
+ "epoch": 0.11850255857796929,
+ "grad_norm": 0.8293434376681832,
+ "learning_rate": 9.913724274192829e-06,
+ "loss": 0.4147852897644043,
+ "num_input_tokens_seen": 77069784,
+ "step": 1760,
+ "train_runtime": 13537.8959,
+ "train_tokens_per_second": 5692.892
+ },
+ {
+ "epoch": 0.11883921357392943,
+ "grad_norm": 0.9258862128909312,
+ "learning_rate": 9.91323452093906e-06,
+ "loss": 0.41230435371398927,
+ "num_input_tokens_seen": 77288624,
+ "step": 1765,
+ "train_runtime": 13574.917,
+ "train_tokens_per_second": 5693.488
+ },
+ {
+ "epoch": 0.11917586856988957,
+ "grad_norm": 0.848930776896082,
+ "learning_rate": 9.912743393714927e-06,
+ "loss": 0.3945260286331177,
+ "num_input_tokens_seen": 77519384,
+ "step": 1770,
+ "train_runtime": 13605.814,
+ "train_tokens_per_second": 5697.519
+ },
+ {
+ "epoch": 0.11951252356584972,
+ "grad_norm": 0.8278888004348279,
+ "learning_rate": 9.912250892657769e-06,
+ "loss": 0.40961732864379885,
+ "num_input_tokens_seen": 77735336,
+ "step": 1775,
+ "train_runtime": 13644.1096,
+ "train_tokens_per_second": 5697.355
+ },
+ {
+ "epoch": 0.11984917856180986,
+ "grad_norm": 0.9501953857186206,
+ "learning_rate": 9.911757017905312e-06,
+ "loss": 0.41979494094848635,
+ "num_input_tokens_seen": 77943944,
+ "step": 1780,
+ "train_runtime": 13675.7254,
+ "train_tokens_per_second": 5699.438
+ },
+ {
+ "epoch": 0.12018583355777,
+ "grad_norm": 0.8175400797161728,
+ "learning_rate": 9.911261769595668e-06,
+ "loss": 0.4165688991546631,
+ "num_input_tokens_seen": 78176688,
+ "step": 1785,
+ "train_runtime": 13720.1102,
+ "train_tokens_per_second": 5697.964
+ },
+ {
+ "epoch": 0.12052248855373014,
+ "grad_norm": 0.9695481492778111,
+ "learning_rate": 9.910765147867332e-06,
+ "loss": 0.4288275718688965,
+ "num_input_tokens_seen": 78398920,
+ "step": 1790,
+ "train_runtime": 13754.606,
+ "train_tokens_per_second": 5699.83
+ },
+ {
+ "epoch": 0.12085914354969028,
+ "grad_norm": 0.9043853969674001,
+ "learning_rate": 9.91026715285918e-06,
+ "loss": 0.42678542137145997,
+ "num_input_tokens_seen": 78617400,
+ "step": 1795,
+ "train_runtime": 13795.7712,
+ "train_tokens_per_second": 5698.659
+ },
+ {
+ "epoch": 0.12119579854565042,
+ "grad_norm": 0.9643745725188065,
+ "learning_rate": 9.909767784710476e-06,
+ "loss": 0.4547886371612549,
+ "num_input_tokens_seen": 78858720,
+ "step": 1800,
+ "train_runtime": 13832.1459,
+ "train_tokens_per_second": 5701.12
+ },
+ {
+ "epoch": 0.12153245354161056,
+ "grad_norm": 0.8832616111060988,
+ "learning_rate": 9.909267043560868e-06,
+ "loss": 0.4295345783233643,
+ "num_input_tokens_seen": 79081264,
+ "step": 1805,
+ "train_runtime": 13867.5931,
+ "train_tokens_per_second": 5702.595
+ },
+ {
+ "epoch": 0.1218691085375707,
+ "grad_norm": 0.8001666467072275,
+ "learning_rate": 9.908764929550388e-06,
+ "loss": 0.4215728759765625,
+ "num_input_tokens_seen": 79301792,
+ "step": 1810,
+ "train_runtime": 13907.1584,
+ "train_tokens_per_second": 5702.228
+ },
+ {
+ "epoch": 0.12220576353353084,
+ "grad_norm": 0.9182690866647383,
+ "learning_rate": 9.908261442819444e-06,
+ "loss": 0.4215451717376709,
+ "num_input_tokens_seen": 79503840,
+ "step": 1815,
+ "train_runtime": 13955.1411,
+ "train_tokens_per_second": 5697.1
+ },
+ {
+ "epoch": 0.12254241852949098,
+ "grad_norm": 0.7856797264096972,
+ "learning_rate": 9.907756583508842e-06,
+ "loss": 0.3922400951385498,
+ "num_input_tokens_seen": 79740056,
+ "step": 1820,
+ "train_runtime": 13995.629,
+ "train_tokens_per_second": 5697.497
+ },
+ {
+ "epoch": 0.12287907352545112,
+ "grad_norm": 0.791802883681199,
+ "learning_rate": 9.907250351759761e-06,
+ "loss": 0.40973472595214844,
+ "num_input_tokens_seen": 79961848,
+ "step": 1825,
+ "train_runtime": 14035.6538,
+ "train_tokens_per_second": 5697.052
+ },
+ {
+ "epoch": 0.12321572852141126,
+ "grad_norm": 1.0344260725436623,
+ "learning_rate": 9.906742747713766e-06,
+ "loss": 0.4293796539306641,
+ "num_input_tokens_seen": 80180448,
+ "step": 1830,
+ "train_runtime": 14071.4195,
+ "train_tokens_per_second": 5698.107
+ },
+ {
+ "epoch": 0.1235523835173714,
+ "grad_norm": 1.0115683734845813,
+ "learning_rate": 9.906233771512808e-06,
+ "loss": 0.4379068374633789,
+ "num_input_tokens_seen": 80391288,
+ "step": 1835,
+ "train_runtime": 14108.8045,
+ "train_tokens_per_second": 5697.952
+ },
+ {
+ "epoch": 0.12388903851333154,
+ "grad_norm": 0.9113985281993394,
+ "learning_rate": 9.905723423299222e-06,
+ "loss": 0.44842100143432617,
+ "num_input_tokens_seen": 80619704,
+ "step": 1840,
+ "train_runtime": 14148.1364,
+ "train_tokens_per_second": 5698.256
+ },
+ {
+ "epoch": 0.12422569350929168,
+ "grad_norm": 0.8655689498703674,
+ "learning_rate": 9.905211703215723e-06,
+ "loss": 0.402877140045166,
+ "num_input_tokens_seen": 80832568,
+ "step": 1845,
+ "train_runtime": 14183.917,
+ "train_tokens_per_second": 5698.889
+ },
+ {
+ "epoch": 0.12456234850525182,
+ "grad_norm": 0.8753972943680609,
+ "learning_rate": 9.904698611405412e-06,
+ "loss": 0.4264510631561279,
+ "num_input_tokens_seen": 81058296,
+ "step": 1850,
+ "train_runtime": 14223.3173,
+ "train_tokens_per_second": 5698.973
+ },
+ {
+ "epoch": 0.12489900350121196,
+ "grad_norm": 0.7688694545812488,
+ "learning_rate": 9.904184148011774e-06,
+ "loss": 0.438901424407959,
+ "num_input_tokens_seen": 81274840,
+ "step": 1855,
+ "train_runtime": 14263.055,
+ "train_tokens_per_second": 5698.277
+ },
+ {
+ "epoch": 0.1252356584971721,
+ "grad_norm": 0.9233391292188109,
+ "learning_rate": 9.903668313178681e-06,
+ "loss": 0.44609222412109373,
+ "num_input_tokens_seen": 81500576,
+ "step": 1860,
+ "train_runtime": 14299.671,
+ "train_tokens_per_second": 5699.472
+ },
+ {
+ "epoch": 0.12557231349313225,
+ "grad_norm": 1.0943906900012805,
+ "learning_rate": 9.903151107050377e-06,
+ "loss": 0.396915864944458,
+ "num_input_tokens_seen": 81707648,
+ "step": 1865,
+ "train_runtime": 14337.8178,
+ "train_tokens_per_second": 5698.751
+ },
+ {
+ "epoch": 0.1259089684890924,
+ "grad_norm": 0.7168785500619042,
+ "learning_rate": 9.902632529771502e-06,
+ "loss": 0.42331228256225584,
+ "num_input_tokens_seen": 81926600,
+ "step": 1870,
+ "train_runtime": 14376.4633,
+ "train_tokens_per_second": 5698.662
+ },
+ {
+ "epoch": 0.12624562348505253,
+ "grad_norm": 1.0835704108533986,
+ "learning_rate": 9.902112581487074e-06,
+ "loss": 0.42675061225891114,
+ "num_input_tokens_seen": 82148216,
+ "step": 1875,
+ "train_runtime": 14416.4946,
+ "train_tokens_per_second": 5698.21
+ },
+ {
+ "epoch": 0.12658227848101267,
+ "grad_norm": 0.9867817197311566,
+ "learning_rate": 9.901591262342494e-06,
+ "loss": 0.4384635925292969,
+ "num_input_tokens_seen": 82358848,
+ "step": 1880,
+ "train_runtime": 14459.8702,
+ "train_tokens_per_second": 5695.684
+ },
+ {
+ "epoch": 0.1269189334769728,
+ "grad_norm": 0.9106268020914177,
+ "learning_rate": 9.901068572483548e-06,
+ "loss": 0.4060489654541016,
+ "num_input_tokens_seen": 82576440,
+ "step": 1885,
+ "train_runtime": 14497.7161,
+ "train_tokens_per_second": 5695.824
+ },
+ {
+ "epoch": 0.12725558847293295,
+ "grad_norm": 0.9009980190191491,
+ "learning_rate": 9.900544512056402e-06,
+ "loss": 0.47153472900390625,
+ "num_input_tokens_seen": 82807688,
+ "step": 1890,
+ "train_runtime": 14534.4393,
+ "train_tokens_per_second": 5697.343
+ },
+ {
+ "epoch": 0.12759224346889309,
+ "grad_norm": 0.8549638938689581,
+ "learning_rate": 9.900019081207612e-06,
+ "loss": 0.4209029197692871,
+ "num_input_tokens_seen": 83038168,
+ "step": 1895,
+ "train_runtime": 14569.2093,
+ "train_tokens_per_second": 5699.566
+ },
+ {
+ "epoch": 0.12792889846485322,
+ "grad_norm": 0.8968781974119465,
+ "learning_rate": 9.89949228008411e-06,
+ "loss": 0.42281227111816405,
+ "num_input_tokens_seen": 83267720,
+ "step": 1900,
+ "train_runtime": 14606.2305,
+ "train_tokens_per_second": 5700.836
+ },
+ {
+ "epoch": 0.12826555346081336,
+ "grad_norm": 0.8595709799025085,
+ "learning_rate": 9.898964108833216e-06,
+ "loss": 0.42650370597839354,
+ "num_input_tokens_seen": 83492616,
+ "step": 1905,
+ "train_runtime": 14646.7617,
+ "train_tokens_per_second": 5700.415
+ },
+ {
+ "epoch": 0.1286022084567735,
+ "grad_norm": 0.8815455063429253,
+ "learning_rate": 9.898434567602629e-06,
+ "loss": 0.4016720771789551,
+ "num_input_tokens_seen": 83715840,
+ "step": 1910,
+ "train_runtime": 14686.2575,
+ "train_tokens_per_second": 5700.284
+ },
+ {
+ "epoch": 0.12893886345273364,
+ "grad_norm": 0.9694766961124608,
+ "learning_rate": 9.897903656540437e-06,
+ "loss": 0.4289149284362793,
+ "num_input_tokens_seen": 83934184,
+ "step": 1915,
+ "train_runtime": 14724.9891,
+ "train_tokens_per_second": 5700.119
+ },
+ {
+ "epoch": 0.12927551844869378,
+ "grad_norm": 0.9823894967580632,
+ "learning_rate": 9.897371375795103e-06,
+ "loss": 0.43784418106079104,
+ "num_input_tokens_seen": 84150312,
+ "step": 1920,
+ "train_runtime": 14759.5087,
+ "train_tokens_per_second": 5701.43
+ },
+ {
+ "epoch": 0.12961217344465392,
+ "grad_norm": 0.8194539046535433,
+ "learning_rate": 9.896837725515484e-06,
+ "loss": 0.44328908920288085,
+ "num_input_tokens_seen": 84361600,
+ "step": 1925,
+ "train_runtime": 14805.8899,
+ "train_tokens_per_second": 5697.841
+ },
+ {
+ "epoch": 0.12994882844061406,
+ "grad_norm": 0.7081955649991436,
+ "learning_rate": 9.89630270585081e-06,
+ "loss": 0.39231128692626954,
+ "num_input_tokens_seen": 84577168,
+ "step": 1930,
+ "train_runtime": 14842.2012,
+ "train_tokens_per_second": 5698.425
+ },
+ {
+ "epoch": 0.1302854834365742,
+ "grad_norm": 0.8873554761311145,
+ "learning_rate": 9.895766316950694e-06,
+ "loss": 0.4394209384918213,
+ "num_input_tokens_seen": 84796464,
+ "step": 1935,
+ "train_runtime": 14876.8277,
+ "train_tokens_per_second": 5699.902
+ },
+ {
+ "epoch": 0.13062213843253434,
+ "grad_norm": 0.8262969710012915,
+ "learning_rate": 9.89522855896514e-06,
+ "loss": 0.41849603652954104,
+ "num_input_tokens_seen": 85025304,
+ "step": 1940,
+ "train_runtime": 14919.7496,
+ "train_tokens_per_second": 5698.843
+ },
+ {
+ "epoch": 0.13095879342849448,
+ "grad_norm": 0.9120986383703955,
+ "learning_rate": 9.89468943204453e-06,
+ "loss": 0.4164949893951416,
+ "num_input_tokens_seen": 85239728,
+ "step": 1945,
+ "train_runtime": 14964.5159,
+ "train_tokens_per_second": 5696.123
+ },
+ {
+ "epoch": 0.13129544842445462,
+ "grad_norm": 1.0478669869240167,
+ "learning_rate": 9.89414893633963e-06,
+ "loss": 0.41168971061706544,
+ "num_input_tokens_seen": 85451488,
+ "step": 1950,
+ "train_runtime": 14998.4779,
+ "train_tokens_per_second": 5697.344
+ },
+ {
+ "epoch": 0.13163210342041476,
+ "grad_norm": 0.8072074904847002,
+ "learning_rate": 9.893607072001584e-06,
+ "loss": 0.42524333000183107,
+ "num_input_tokens_seen": 85681416,
+ "step": 1955,
+ "train_runtime": 15045.5899,
+ "train_tokens_per_second": 5694.786
+ },
+ {
+ "epoch": 0.1319687584163749,
+ "grad_norm": 0.7884630260384243,
+ "learning_rate": 9.893063839181925e-06,
+ "loss": 0.41834263801574706,
+ "num_input_tokens_seen": 85897864,
+ "step": 1960,
+ "train_runtime": 15086.3562,
+ "train_tokens_per_second": 5693.745
+ },
+ {
+ "epoch": 0.13230541341233504,
+ "grad_norm": 0.8242479851716092,
+ "learning_rate": 9.892519238032566e-06,
+ "loss": 0.4272817611694336,
+ "num_input_tokens_seen": 86126184,
+ "step": 1965,
+ "train_runtime": 15130.1067,
+ "train_tokens_per_second": 5692.371
+ },
+ {
+ "epoch": 0.13264206840829518,
+ "grad_norm": 0.8883896038795247,
+ "learning_rate": 9.891973268705804e-06,
+ "loss": 0.45201525688171384,
+ "num_input_tokens_seen": 86340160,
+ "step": 1970,
+ "train_runtime": 15166.855,
+ "train_tokens_per_second": 5692.687
+ },
+ {
+ "epoch": 0.13297872340425532,
+ "grad_norm": 0.7899276150283895,
+ "learning_rate": 9.891425931354316e-06,
+ "loss": 0.4250628471374512,
+ "num_input_tokens_seen": 86565168,
+ "step": 1975,
+ "train_runtime": 15203.4839,
+ "train_tokens_per_second": 5693.772
+ },
+ {
+ "epoch": 0.13331537840021546,
+ "grad_norm": 0.8879719340880936,
+ "learning_rate": 9.890877226131165e-06,
+ "loss": 0.41506524085998536,
+ "num_input_tokens_seen": 86796552,
+ "step": 1980,
+ "train_runtime": 15247.3097,
+ "train_tokens_per_second": 5692.581
+ },
+ {
+ "epoch": 0.1336520333961756,
+ "grad_norm": 0.916769222493085,
+ "learning_rate": 9.890327153189793e-06,
+ "loss": 0.4300036907196045,
+ "num_input_tokens_seen": 87016968,
+ "step": 1985,
+ "train_runtime": 15292.8026,
+ "train_tokens_per_second": 5690.06
+ },
+ {
+ "epoch": 0.13398868839213574,
+ "grad_norm": 0.9050174324296295,
+ "learning_rate": 9.889775712684026e-06,
+ "loss": 0.4278897285461426,
+ "num_input_tokens_seen": 87240744,
+ "step": 1990,
+ "train_runtime": 15335.4097,
+ "train_tokens_per_second": 5688.843
+ },
+ {
+ "epoch": 0.13432534338809587,
+ "grad_norm": 0.8780717724520911,
+ "learning_rate": 9.889222904768074e-06,
+ "loss": 0.42691898345947266,
+ "num_input_tokens_seen": 87458888,
+ "step": 1995,
+ "train_runtime": 15375.168,
+ "train_tokens_per_second": 5688.321
+ },
+ {
+ "epoch": 0.13466199838405601,
+ "grad_norm": 0.8749943321506966,
+ "learning_rate": 9.888668729596528e-06,
+ "loss": 0.41673784255981444,
+ "num_input_tokens_seen": 87661224,
+ "step": 2000,
+ "train_runtime": 15417.6729,
+ "train_tokens_per_second": 5685.762
+ },
+ {
+ "epoch": 0.13499865338001615,
+ "grad_norm": 0.9429761954235121,
+ "learning_rate": 9.88811318732436e-06,
+ "loss": 0.4196599006652832,
+ "num_input_tokens_seen": 87871224,
+ "step": 2005,
+ "train_runtime": 15454.7159,
+ "train_tokens_per_second": 5685.722
+ },
+ {
+ "epoch": 0.1353353083759763,
+ "grad_norm": 0.8303836661224822,
+ "learning_rate": 9.887556278106927e-06,
+ "loss": 0.43319454193115237,
+ "num_input_tokens_seen": 88101424,
+ "step": 2010,
+ "train_runtime": 15492.7733,
+ "train_tokens_per_second": 5686.614
+ },
+ {
+ "epoch": 0.13567196337193643,
+ "grad_norm": 1.2988017888943026,
+ "learning_rate": 9.886998002099964e-06,
+ "loss": 0.42733011245727537,
+ "num_input_tokens_seen": 88331320,
+ "step": 2015,
+ "train_runtime": 15533.3637,
+ "train_tokens_per_second": 5686.555
+ },
+ {
+ "epoch": 0.13600861836789657,
+ "grad_norm": 0.926333904941965,
+ "learning_rate": 9.886438359459595e-06,
+ "loss": 0.4501063346862793,
+ "num_input_tokens_seen": 88546592,
+ "step": 2020,
+ "train_runtime": 15572.5905,
+ "train_tokens_per_second": 5686.054
+ },
+ {
+ "epoch": 0.1363452733638567,
+ "grad_norm": 0.8812830943715737,
+ "learning_rate": 9.885877350342319e-06,
+ "loss": 0.3861879587173462,
+ "num_input_tokens_seen": 88774480,
+ "step": 2025,
+ "train_runtime": 15611.197,
+ "train_tokens_per_second": 5686.59
+ },
+ {
+ "epoch": 0.13668192835981685,
+ "grad_norm": 0.8007094863514034,
+ "learning_rate": 9.885314974905024e-06,
+ "loss": 0.3875585556030273,
+ "num_input_tokens_seen": 89011576,
+ "step": 2030,
+ "train_runtime": 15652.3238,
+ "train_tokens_per_second": 5686.796
+ },
+ {
+ "epoch": 0.137018583355777,
+ "grad_norm": 0.9125014304117678,
+ "learning_rate": 9.884751233304973e-06,
+ "loss": 0.39730267524719237,
+ "num_input_tokens_seen": 89217520,
+ "step": 2035,
+ "train_runtime": 15686.075,
+ "train_tokens_per_second": 5687.689
+ },
+ {
+ "epoch": 0.13735523835173713,
+ "grad_norm": 0.8474671445937998,
+ "learning_rate": 9.884186125699816e-06,
+ "loss": 0.42992243766784666,
+ "num_input_tokens_seen": 89431816,
+ "step": 2040,
+ "train_runtime": 15726.6605,
+ "train_tokens_per_second": 5686.637
+ },
+ {
+ "epoch": 0.13769189334769727,
+ "grad_norm": 0.9102001150203423,
+ "learning_rate": 9.883619652247584e-06,
+ "loss": 0.4123133659362793,
+ "num_input_tokens_seen": 89651968,
+ "step": 2045,
+ "train_runtime": 15768.4222,
+ "train_tokens_per_second": 5685.538
+ },
+ {
+ "epoch": 0.1380285483436574,
+ "grad_norm": 0.6804193522827889,
+ "learning_rate": 9.883051813106687e-06,
+ "loss": 0.3782381296157837,
+ "num_input_tokens_seen": 89874624,
+ "step": 2050,
+ "train_runtime": 15813.3958,
+ "train_tokens_per_second": 5683.449
+ },
+ {
+ "epoch": 0.13836520333961755,
+ "grad_norm": 0.8989541162037309,
+ "learning_rate": 9.882482608435924e-06,
+ "loss": 0.43795013427734375,
+ "num_input_tokens_seen": 90097216,
+ "step": 2055,
+ "train_runtime": 15851.8211,
+ "train_tokens_per_second": 5683.714
+ },
+ {
+ "epoch": 0.1387018583355777,
+ "grad_norm": 0.8434117541493056,
+ "learning_rate": 9.881912038394468e-06,
+ "loss": 0.42223529815673827,
+ "num_input_tokens_seen": 90317784,
+ "step": 2060,
+ "train_runtime": 15897.2071,
+ "train_tokens_per_second": 5681.362
+ },
+ {
+ "epoch": 0.13903851333153783,
+ "grad_norm": 0.7960282493279416,
+ "learning_rate": 9.881340103141878e-06,
+ "loss": 0.4242293357849121,
+ "num_input_tokens_seen": 90546968,
+ "step": 2065,
+ "train_runtime": 15940.8703,
+ "train_tokens_per_second": 5680.177
+ },
+ {
+ "epoch": 0.13937516832749797,
+ "grad_norm": 0.8413026941754382,
+ "learning_rate": 9.880766802838092e-06,
+ "loss": 0.4057013511657715,
+ "num_input_tokens_seen": 90762816,
+ "step": 2070,
+ "train_runtime": 15983.7133,
+ "train_tokens_per_second": 5678.456
+ },
+ {
+ "epoch": 0.1397118233234581,
+ "grad_norm": 0.7350898878720474,
+ "learning_rate": 9.880192137643434e-06,
+ "loss": 0.4075471878051758,
+ "num_input_tokens_seen": 90979408,
+ "step": 2075,
+ "train_runtime": 16017.6972,
+ "train_tokens_per_second": 5679.931
+ },
+ {
+ "epoch": 0.14004847831941827,
+ "grad_norm": 0.9141607370838677,
+ "learning_rate": 9.879616107718608e-06,
+ "loss": 0.4198787689208984,
+ "num_input_tokens_seen": 91197560,
+ "step": 2080,
+ "train_runtime": 16051.0779,
+ "train_tokens_per_second": 5681.709
+ },
+ {
+ "epoch": 0.1403851333153784,
+ "grad_norm": 0.7897438400002121,
+ "learning_rate": 9.879038713224695e-06,
+ "loss": 0.4208370685577393,
+ "num_input_tokens_seen": 91421872,
+ "step": 2085,
+ "train_runtime": 16093.9488,
+ "train_tokens_per_second": 5680.512
+ },
+ {
+ "epoch": 0.14072178831133855,
+ "grad_norm": 0.8205504842451986,
+ "learning_rate": 9.878459954323165e-06,
+ "loss": 0.4030628204345703,
+ "num_input_tokens_seen": 91647112,
+ "step": 2090,
+ "train_runtime": 16126.4411,
+ "train_tokens_per_second": 5683.034
+ },
+ {
+ "epoch": 0.1410584433072987,
+ "grad_norm": 0.8325253458879384,
+ "learning_rate": 9.877879831175865e-06,
+ "loss": 0.39506003856658933,
+ "num_input_tokens_seen": 91871896,
+ "step": 2095,
+ "train_runtime": 16160.6707,
+ "train_tokens_per_second": 5684.906
+ },
+ {
+ "epoch": 0.14139509830325883,
+ "grad_norm": 0.821317865316409,
+ "learning_rate": 9.877298343945025e-06,
+ "loss": 0.41777982711791994,
+ "num_input_tokens_seen": 92094456,
+ "step": 2100,
+ "train_runtime": 16203.3038,
+ "train_tokens_per_second": 5683.684
+ },
+ {
+ "epoch": 0.14173175329921897,
+ "grad_norm": 0.7872012950426368,
+ "learning_rate": 9.876715492793256e-06,
+ "loss": 0.41480426788330077,
+ "num_input_tokens_seen": 92315440,
+ "step": 2105,
+ "train_runtime": 16242.0831,
+ "train_tokens_per_second": 5683.719
+ },
+ {
+ "epoch": 0.1420684082951791,
+ "grad_norm": 0.8092561101034788,
+ "learning_rate": 9.87613127788355e-06,
+ "loss": 0.389608097076416,
+ "num_input_tokens_seen": 92527368,
+ "step": 2110,
+ "train_runtime": 16278.4874,
+ "train_tokens_per_second": 5684.027
+ },
+ {
+ "epoch": 0.14240506329113925,
+ "grad_norm": 0.8850860821410723,
+ "learning_rate": 9.87554569937928e-06,
+ "loss": 0.4558421611785889,
+ "num_input_tokens_seen": 92759672,
+ "step": 2115,
+ "train_runtime": 16314.9617,
+ "train_tokens_per_second": 5685.559
+ },
+ {
+ "epoch": 0.1427417182870994,
+ "grad_norm": 0.8891822985539864,
+ "learning_rate": 9.874958757444203e-06,
+ "loss": 0.45552263259887693,
+ "num_input_tokens_seen": 92995096,
+ "step": 2120,
+ "train_runtime": 16350.6171,
+ "train_tokens_per_second": 5687.559
+ },
+ {
+ "epoch": 0.14307837328305953,
+ "grad_norm": 0.8780645983704002,
+ "learning_rate": 9.874370452242454e-06,
+ "loss": 0.43328495025634767,
+ "num_input_tokens_seen": 93229944,
+ "step": 2125,
+ "train_runtime": 16384.0995,
+ "train_tokens_per_second": 5690.27
+ },
+ {
+ "epoch": 0.14341502827901967,
+ "grad_norm": 0.9402959761478824,
+ "learning_rate": 9.873780783938553e-06,
+ "loss": 0.4462221622467041,
+ "num_input_tokens_seen": 93450480,
+ "step": 2130,
+ "train_runtime": 16417.1629,
+ "train_tokens_per_second": 5692.243
+ },
+ {
+ "epoch": 0.1437516832749798,
+ "grad_norm": 0.8574770635812221,
+ "learning_rate": 9.873189752697396e-06,
+ "loss": 0.3947939395904541,
+ "num_input_tokens_seen": 93670472,
+ "step": 2135,
+ "train_runtime": 16459.3253,
+ "train_tokens_per_second": 5691.027
+ },
+ {
+ "epoch": 0.14408833827093995,
+ "grad_norm": 0.8947319210921507,
+ "learning_rate": 9.872597358684265e-06,
+ "loss": 0.458558464050293,
+ "num_input_tokens_seen": 93883544,
+ "step": 2140,
+ "train_runtime": 16489.5195,
+ "train_tokens_per_second": 5693.528
+ },
+ {
+ "epoch": 0.1444249932669001,
+ "grad_norm": 0.8538177526814695,
+ "learning_rate": 9.872003602064816e-06,
+ "loss": 0.4299613475799561,
+ "num_input_tokens_seen": 94116824,
+ "step": 2145,
+ "train_runtime": 16523.1274,
+ "train_tokens_per_second": 5696.066
+ },
+ {
+ "epoch": 0.14476164826286023,
+ "grad_norm": 0.7471306795366713,
+ "learning_rate": 9.8714084830051e-06,
+ "loss": 0.4079564571380615,
+ "num_input_tokens_seen": 94349480,
+ "step": 2150,
+ "train_runtime": 16564.7452,
+ "train_tokens_per_second": 5695.8
+ },
+ {
+ "epoch": 0.14509830325882037,
+ "grad_norm": 0.8218664635099376,
+ "learning_rate": 9.870812001671533e-06,
+ "loss": 0.40776524543762205,
+ "num_input_tokens_seen": 94577624,
+ "step": 2155,
+ "train_runtime": 16600.9249,
+ "train_tokens_per_second": 5697.13
+ },
+ {
+ "epoch": 0.1454349582547805,
+ "grad_norm": 1.0737295934493194,
+ "learning_rate": 9.870214158230922e-06,
+ "loss": 0.4177376747131348,
+ "num_input_tokens_seen": 94804128,
+ "step": 2160,
+ "train_runtime": 16635.1905,
+ "train_tokens_per_second": 5699.011
+ },
+ {
+ "epoch": 0.14577161325074064,
+ "grad_norm": 0.786781014793927,
+ "learning_rate": 9.869614952850454e-06,
+ "loss": 0.40900435447692873,
+ "num_input_tokens_seen": 95026192,
+ "step": 2165,
+ "train_runtime": 16669.2594,
+ "train_tokens_per_second": 5700.685
+ },
+ {
+ "epoch": 0.14610826824670078,
+ "grad_norm": 0.8360055927785602,
+ "learning_rate": 9.86901438569769e-06,
+ "loss": 0.42821083068847654,
+ "num_input_tokens_seen": 95254672,
+ "step": 2170,
+ "train_runtime": 16706.7948,
+ "train_tokens_per_second": 5701.553
+ },
+ {
+ "epoch": 0.14644492324266092,
+ "grad_norm": 0.8933164074109287,
+ "learning_rate": 9.86841245694058e-06,
+ "loss": 0.425125789642334,
+ "num_input_tokens_seen": 95450280,
+ "step": 2175,
+ "train_runtime": 16746.2086,
+ "train_tokens_per_second": 5699.814
+ },
+ {
+ "epoch": 0.14678157823862106,
+ "grad_norm": 0.8906147805897214,
+ "learning_rate": 9.86780916674745e-06,
+ "loss": 0.45731253623962403,
+ "num_input_tokens_seen": 95651304,
+ "step": 2180,
+ "train_runtime": 16783.8632,
+ "train_tokens_per_second": 5699.004
+ },
+ {
+ "epoch": 0.1471182332345812,
+ "grad_norm": 1.0525327714019657,
+ "learning_rate": 9.86720451528701e-06,
+ "loss": 0.40560040473937986,
+ "num_input_tokens_seen": 95859952,
+ "step": 2185,
+ "train_runtime": 16822.6242,
+ "train_tokens_per_second": 5698.276
+ },
+ {
+ "epoch": 0.14745488823054134,
+ "grad_norm": 0.816498968853468,
+ "learning_rate": 9.866598502728348e-06,
+ "loss": 0.43126611709594725,
+ "num_input_tokens_seen": 96069168,
+ "step": 2190,
+ "train_runtime": 16862.6901,
+ "train_tokens_per_second": 5697.144
+ },
+ {
+ "epoch": 0.14779154322650148,
+ "grad_norm": 0.841739118951704,
+ "learning_rate": 9.865991129240931e-06,
+ "loss": 0.41145782470703124,
+ "num_input_tokens_seen": 96283672,
+ "step": 2195,
+ "train_runtime": 16896.2709,
+ "train_tokens_per_second": 5698.516
+ },
+ {
+ "epoch": 0.14812819822246162,
+ "grad_norm": 0.878868960575425,
+ "learning_rate": 9.865382394994614e-06,
+ "loss": 0.40999040603637693,
+ "num_input_tokens_seen": 96502528,
+ "step": 2200,
+ "train_runtime": 16939.3291,
+ "train_tokens_per_second": 5696.951
+ },
+ {
+ "epoch": 0.14846485321842176,
+ "grad_norm": 0.6967776248464405,
+ "learning_rate": 9.864772300159622e-06,
+ "loss": 0.3990957021713257,
+ "num_input_tokens_seen": 96720064,
+ "step": 2205,
+ "train_runtime": 16973.4551,
+ "train_tokens_per_second": 5698.313
+ },
+ {
+ "epoch": 0.1488015082143819,
+ "grad_norm": 0.9032740614218077,
+ "learning_rate": 9.864160844906571e-06,
+ "loss": 0.43683781623840334,
+ "num_input_tokens_seen": 96941160,
+ "step": 2210,
+ "train_runtime": 17009.3974,
+ "train_tokens_per_second": 5699.271
+ },
+ {
+ "epoch": 0.14913816321034204,
+ "grad_norm": 0.8444434220131386,
+ "learning_rate": 9.863548029406449e-06,
+ "loss": 0.419160795211792,
+ "num_input_tokens_seen": 97161584,
+ "step": 2215,
+ "train_runtime": 17050.6002,
+ "train_tokens_per_second": 5698.426
+ },
+ {
+ "epoch": 0.14947481820630218,
+ "grad_norm": 0.7872642768120427,
+ "learning_rate": 9.86293385383063e-06,
+ "loss": 0.3916934490203857,
+ "num_input_tokens_seen": 97372312,
+ "step": 2220,
+ "train_runtime": 17095.7206,
+ "train_tokens_per_second": 5695.713
+ },
+ {
+ "epoch": 0.14981147320226232,
+ "grad_norm": 0.9092720440628393,
+ "learning_rate": 9.862318318350865e-06,
+ "loss": 0.4371610641479492,
+ "num_input_tokens_seen": 97594144,
+ "step": 2225,
+ "train_runtime": 17132.445,
+ "train_tokens_per_second": 5696.452
+ },
+ {
+ "epoch": 0.15014812819822246,
+ "grad_norm": 0.7914448222714119,
+ "learning_rate": 9.861701423139287e-06,
+ "loss": 0.39911181926727296,
+ "num_input_tokens_seen": 97820888,
+ "step": 2230,
+ "train_runtime": 17165.9269,
+ "train_tokens_per_second": 5698.55
+ },
+ {
+ "epoch": 0.1504847831941826,
+ "grad_norm": 0.8316284130963478,
+ "learning_rate": 9.861083168368411e-06,
+ "loss": 0.41806850433349607,
+ "num_input_tokens_seen": 98045776,
+ "step": 2235,
+ "train_runtime": 17204.7719,
+ "train_tokens_per_second": 5698.755
+ },
+ {
+ "epoch": 0.15082143819014274,
+ "grad_norm": 0.9088287061584689,
+ "learning_rate": 9.860463554211125e-06,
+ "loss": 0.3891399383544922,
+ "num_input_tokens_seen": 98268344,
+ "step": 2240,
+ "train_runtime": 17246.0796,
+ "train_tokens_per_second": 5698.011
+ },
+ {
+ "epoch": 0.15115809318610288,
+ "grad_norm": 0.9381951784290052,
+ "learning_rate": 9.859842580840706e-06,
+ "loss": 0.4640520095825195,
+ "num_input_tokens_seen": 98493000,
+ "step": 2245,
+ "train_runtime": 17289.4391,
+ "train_tokens_per_second": 5696.715
+ },
+ {
+ "epoch": 0.15149474818206302,
+ "grad_norm": 0.7432535531903158,
+ "learning_rate": 9.859220248430806e-06,
+ "loss": 0.41865386962890627,
+ "num_input_tokens_seen": 98731240,
+ "step": 2250,
+ "train_runtime": 17329.2789,
+ "train_tokens_per_second": 5697.366
+ },
+ {
+ "epoch": 0.15183140317802316,
+ "grad_norm": 0.7752690728179633,
+ "learning_rate": 9.85859655715546e-06,
+ "loss": 0.39298291206359864,
+ "num_input_tokens_seen": 98953224,
+ "step": 2255,
+ "train_runtime": 17375.0272,
+ "train_tokens_per_second": 5695.141
+ },
+ {
+ "epoch": 0.1521680581739833,
+ "grad_norm": 0.8610212435397341,
+ "learning_rate": 9.857971507189077e-06,
+ "loss": 0.41016359329223634,
+ "num_input_tokens_seen": 99171512,
+ "step": 2260,
+ "train_runtime": 17414.0463,
+ "train_tokens_per_second": 5694.915
+ },
+ {
+ "epoch": 0.15250471316994343,
+ "grad_norm": 0.8902354817859172,
+ "learning_rate": 9.857345098706455e-06,
+ "loss": 0.4255502700805664,
+ "num_input_tokens_seen": 99402080,
+ "step": 2265,
+ "train_runtime": 17451.574,
+ "train_tokens_per_second": 5695.88
+ },
+ {
+ "epoch": 0.15284136816590357,
+ "grad_norm": 0.9086587862733272,
+ "learning_rate": 9.856717331882765e-06,
+ "loss": 0.4265317440032959,
+ "num_input_tokens_seen": 99607944,
+ "step": 2270,
+ "train_runtime": 17486.7082,
+ "train_tokens_per_second": 5696.209
+ },
+ {
+ "epoch": 0.1531780231618637,
+ "grad_norm": 0.9432280136780563,
+ "learning_rate": 9.85608820689356e-06,
+ "loss": 0.39440102577209474,
+ "num_input_tokens_seen": 99830656,
+ "step": 2275,
+ "train_runtime": 17528.0918,
+ "train_tokens_per_second": 5695.466
+ },
+ {
+ "epoch": 0.15351467815782385,
+ "grad_norm": 0.84487279707662,
+ "learning_rate": 9.855457723914772e-06,
+ "loss": 0.4084943771362305,
+ "num_input_tokens_seen": 100050880,
+ "step": 2280,
+ "train_runtime": 17562.2543,
+ "train_tokens_per_second": 5696.927
+ },
+ {
+ "epoch": 0.153851333153784,
+ "grad_norm": 0.9062843980819959,
+ "learning_rate": 9.854825883122713e-06,
+ "loss": 0.4020846366882324,
+ "num_input_tokens_seen": 100260608,
+ "step": 2285,
+ "train_runtime": 17602.1432,
+ "train_tokens_per_second": 5695.932
+ },
+ {
+ "epoch": 0.15418798814974413,
+ "grad_norm": 0.765345818981944,
+ "learning_rate": 9.85419268469408e-06,
+ "loss": 0.39870567321777345,
+ "num_input_tokens_seen": 100466752,
+ "step": 2290,
+ "train_runtime": 17638.1952,
+ "train_tokens_per_second": 5695.977
+ },
+ {
+ "epoch": 0.15452464314570427,
+ "grad_norm": 0.8051620604928534,
+ "learning_rate": 9.85355812880594e-06,
+ "loss": 0.3938899517059326,
+ "num_input_tokens_seen": 100686544,
+ "step": 2295,
+ "train_runtime": 17677.5734,
+ "train_tokens_per_second": 5695.722
+ },
+ {
+ "epoch": 0.1548612981416644,
+ "grad_norm": 0.7457579703786071,
+ "learning_rate": 9.852922215635747e-06,
+ "loss": 0.37209017276763917,
+ "num_input_tokens_seen": 100908792,
+ "step": 2300,
+ "train_runtime": 17719.4452,
+ "train_tokens_per_second": 5694.805
+ },
+ {
+ "epoch": 0.15519795313762455,
+ "grad_norm": 0.8179184151627187,
+ "learning_rate": 9.85228494536133e-06,
+ "loss": 0.4363281726837158,
+ "num_input_tokens_seen": 101132472,
+ "step": 2305,
+ "train_runtime": 17753.0155,
+ "train_tokens_per_second": 5696.636
+ },
+ {
+ "epoch": 0.1555346081335847,
+ "grad_norm": 0.8389619253246731,
+ "learning_rate": 9.851646318160901e-06,
+ "loss": 0.41831111907958984,
+ "num_input_tokens_seen": 101349024,
+ "step": 2310,
+ "train_runtime": 17793.8083,
+ "train_tokens_per_second": 5695.747
+ },
+ {
+ "epoch": 0.15587126312954483,
+ "grad_norm": 0.7507564820666337,
+ "learning_rate": 9.85100633421305e-06,
+ "loss": 0.39124910831451415,
+ "num_input_tokens_seen": 101558936,
+ "step": 2315,
+ "train_runtime": 17835.905,
+ "train_tokens_per_second": 5694.072
+ },
+ {
+ "epoch": 0.15620791812550497,
+ "grad_norm": 0.8895097039801884,
+ "learning_rate": 9.850364993696749e-06,
+ "loss": 0.42600722312927247,
+ "num_input_tokens_seen": 101772184,
+ "step": 2320,
+ "train_runtime": 17875.446,
+ "train_tokens_per_second": 5693.407
+ },
+ {
+ "epoch": 0.15654457312146514,
+ "grad_norm": 0.9520687878700897,
+ "learning_rate": 9.849722296791339e-06,
+ "loss": 0.4083821773529053,
+ "num_input_tokens_seen": 101982520,
+ "step": 2325,
+ "train_runtime": 17909.3087,
+ "train_tokens_per_second": 5694.386
+ },
+ {
+ "epoch": 0.15688122811742528,
+ "grad_norm": 0.9282345518369051,
+ "learning_rate": 9.849078243676558e-06,
+ "loss": 0.4323273658752441,
+ "num_input_tokens_seen": 102196704,
+ "step": 2330,
+ "train_runtime": 17950.3047,
+ "train_tokens_per_second": 5693.313
+ },
+ {
+ "epoch": 0.15721788311338541,
+ "grad_norm": 0.7571706215220008,
+ "learning_rate": 9.848432834532506e-06,
+ "loss": 0.3904409885406494,
+ "num_input_tokens_seen": 102413872,
+ "step": 2335,
+ "train_runtime": 17990.1951,
+ "train_tokens_per_second": 5692.76
+ },
+ {
+ "epoch": 0.15755453810934555,
+ "grad_norm": 0.7812435917007698,
+ "learning_rate": 9.847786069539673e-06,
+ "loss": 0.42157392501831054,
+ "num_input_tokens_seen": 102636848,
+ "step": 2340,
+ "train_runtime": 18027.5652,
+ "train_tokens_per_second": 5693.328
+ },
+ {
+ "epoch": 0.1578911931053057,
+ "grad_norm": 0.8268032160233915,
+ "learning_rate": 9.847137948878926e-06,
+ "loss": 0.4385071277618408,
+ "num_input_tokens_seen": 102861544,
+ "step": 2345,
+ "train_runtime": 18065.377,
+ "train_tokens_per_second": 5693.85
+ },
+ {
+ "epoch": 0.15822784810126583,
+ "grad_norm": 0.9872902684937591,
+ "learning_rate": 9.846488472731505e-06,
+ "loss": 0.46036734580993655,
+ "num_input_tokens_seen": 103092600,
+ "step": 2350,
+ "train_runtime": 18100.6003,
+ "train_tokens_per_second": 5695.535
+ },
+ {
+ "epoch": 0.15856450309722597,
+ "grad_norm": 0.8069383596278824,
+ "learning_rate": 9.845837641279038e-06,
+ "loss": 0.4151791572570801,
+ "num_input_tokens_seen": 103323896,
+ "step": 2355,
+ "train_runtime": 18131.0861,
+ "train_tokens_per_second": 5698.715
+ },
+ {
+ "epoch": 0.1589011580931861,
+ "grad_norm": 0.918784982161899,
+ "learning_rate": 9.845185454703528e-06,
+ "loss": 0.42508544921875,
+ "num_input_tokens_seen": 103541848,
+ "step": 2360,
+ "train_runtime": 18171.0294,
+ "train_tokens_per_second": 5698.183
+ },
+ {
+ "epoch": 0.15923781308914625,
+ "grad_norm": 0.817441107070197,
+ "learning_rate": 9.844531913187356e-06,
+ "loss": 0.3977517604827881,
+ "num_input_tokens_seen": 103759048,
+ "step": 2365,
+ "train_runtime": 18209.2359,
+ "train_tokens_per_second": 5698.155
+ },
+ {
+ "epoch": 0.1595744680851064,
+ "grad_norm": 0.8849599843025514,
+ "learning_rate": 9.84387701691328e-06,
+ "loss": 0.41399388313293456,
+ "num_input_tokens_seen": 103958648,
+ "step": 2370,
+ "train_runtime": 18252.5289,
+ "train_tokens_per_second": 5695.575
+ },
+ {
+ "epoch": 0.15991112308106653,
+ "grad_norm": 0.9775050084644975,
+ "learning_rate": 9.843220766064443e-06,
+ "loss": 0.4396827697753906,
+ "num_input_tokens_seen": 104164280,
+ "step": 2375,
+ "train_runtime": 18287.1992,
+ "train_tokens_per_second": 5696.022
+ },
+ {
+ "epoch": 0.16024777807702667,
+ "grad_norm": 1.026818097492493,
+ "learning_rate": 9.842563160824364e-06,
+ "loss": 0.4241004943847656,
+ "num_input_tokens_seen": 104376504,
+ "step": 2380,
+ "train_runtime": 18322.4835,
+ "train_tokens_per_second": 5696.635
+ },
+ {
+ "epoch": 0.1605844330729868,
+ "grad_norm": 0.8310388237023464,
+ "learning_rate": 9.841904201376939e-06,
+ "loss": 0.44525814056396484,
+ "num_input_tokens_seen": 104607616,
+ "step": 2385,
+ "train_runtime": 18361.4521,
+ "train_tokens_per_second": 5697.132
+ },
+ {
+ "epoch": 0.16092108806894695,
+ "grad_norm": 0.8107352807912601,
+ "learning_rate": 9.841243887906443e-06,
+ "loss": 0.3924037218093872,
+ "num_input_tokens_seen": 104827648,
+ "step": 2390,
+ "train_runtime": 18397.6205,
+ "train_tokens_per_second": 5697.892
+ },
+ {
+ "epoch": 0.1612577430649071,
+ "grad_norm": 0.8255737690801053,
+ "learning_rate": 9.840582220597532e-06,
+ "loss": 0.38776795864105223,
+ "num_input_tokens_seen": 105048616,
+ "step": 2395,
+ "train_runtime": 18437.9099,
+ "train_tokens_per_second": 5697.425
+ },
+ {
+ "epoch": 0.16159439806086723,
+ "grad_norm": 1.0393091287356393,
+ "learning_rate": 9.839919199635238e-06,
+ "loss": 0.44150123596191404,
+ "num_input_tokens_seen": 105271480,
+ "step": 2400,
+ "train_runtime": 18481.0333,
+ "train_tokens_per_second": 5696.19
+ },
+ {
+ "epoch": 0.16193105305682737,
+ "grad_norm": 1.778936138437422,
+ "learning_rate": 9.839254825204973e-06,
+ "loss": 0.4216805934906006,
+ "num_input_tokens_seen": 105497528,
+ "step": 2405,
+ "train_runtime": 18522.7039,
+ "train_tokens_per_second": 5695.579
+ },
+ {
+ "epoch": 0.1622677080527875,
+ "grad_norm": 0.8730955239731051,
+ "learning_rate": 9.838589097492527e-06,
+ "loss": 0.4195371627807617,
+ "num_input_tokens_seen": 105713952,
+ "step": 2410,
+ "train_runtime": 18562.2885,
+ "train_tokens_per_second": 5695.093
+ },
+ {
+ "epoch": 0.16260436304874765,
+ "grad_norm": 1.003318291974337,
+ "learning_rate": 9.83792201668407e-06,
+ "loss": 0.4166804313659668,
+ "num_input_tokens_seen": 105940976,
+ "step": 2415,
+ "train_runtime": 18598.4307,
+ "train_tokens_per_second": 5696.232
+ },
+ {
+ "epoch": 0.16294101804470779,
+ "grad_norm": 0.9882562378903975,
+ "learning_rate": 9.837253582966146e-06,
+ "loss": 0.4339381217956543,
+ "num_input_tokens_seen": 106165352,
+ "step": 2420,
+ "train_runtime": 18633.1302,
+ "train_tokens_per_second": 5697.666
+ },
+ {
+ "epoch": 0.16327767304066793,
+ "grad_norm": 0.8141828153275021,
+ "learning_rate": 9.836583796525686e-06,
+ "loss": 0.3936171531677246,
+ "num_input_tokens_seen": 106364536,
+ "step": 2425,
+ "train_runtime": 18675.1796,
+ "train_tokens_per_second": 5695.503
+ },
+ {
+ "epoch": 0.16361432803662807,
+ "grad_norm": 0.861811860528164,
+ "learning_rate": 9.835912657549986e-06,
+ "loss": 0.4514013290405273,
+ "num_input_tokens_seen": 106587560,
+ "step": 2430,
+ "train_runtime": 18717.8406,
+ "train_tokens_per_second": 5694.437
+ },
+ {
+ "epoch": 0.1639509830325882,
+ "grad_norm": 0.8811857541516364,
+ "learning_rate": 9.835240166226734e-06,
+ "loss": 0.39427995681762695,
+ "num_input_tokens_seen": 106799528,
+ "step": 2435,
+ "train_runtime": 18761.1279,
+ "train_tokens_per_second": 5692.596
+ },
+ {
+ "epoch": 0.16428763802854834,
+ "grad_norm": 0.8015505121470864,
+ "learning_rate": 9.834566322743987e-06,
+ "loss": 0.43557186126708985,
+ "num_input_tokens_seen": 107012480,
+ "step": 2440,
+ "train_runtime": 18795.981,
+ "train_tokens_per_second": 5693.37
+ },
+ {
+ "epoch": 0.16462429302450848,
+ "grad_norm": 0.8088311122015132,
+ "learning_rate": 9.833891127290186e-06,
+ "loss": 0.4289062976837158,
+ "num_input_tokens_seen": 107250496,
+ "step": 2445,
+ "train_runtime": 18827.4248,
+ "train_tokens_per_second": 5696.504
+ },
+ {
+ "epoch": 0.16496094802046862,
+ "grad_norm": 0.7483438381257065,
+ "learning_rate": 9.833214580054145e-06,
+ "loss": 0.3959768295288086,
+ "num_input_tokens_seen": 107469136,
+ "step": 2450,
+ "train_runtime": 18863.1469,
+ "train_tokens_per_second": 5697.307
+ },
+ {
+ "epoch": 0.16529760301642876,
+ "grad_norm": 0.8277673951140269,
+ "learning_rate": 9.832536681225058e-06,
+ "loss": 0.44450016021728517,
+ "num_input_tokens_seen": 107691864,
+ "step": 2455,
+ "train_runtime": 18898.3354,
+ "train_tokens_per_second": 5698.484
+ },
+ {
+ "epoch": 0.1656342580123889,
+ "grad_norm": 0.8054080794016197,
+ "learning_rate": 9.831857430992497e-06,
+ "loss": 0.3906950235366821,
+ "num_input_tokens_seen": 107905928,
+ "step": 2460,
+ "train_runtime": 18933.5537,
+ "train_tokens_per_second": 5699.19
+ },
+ {
+ "epoch": 0.16597091300834904,
+ "grad_norm": 0.8380957144167158,
+ "learning_rate": 9.831176829546416e-06,
+ "loss": 0.41658635139465333,
+ "num_input_tokens_seen": 108126176,
+ "step": 2465,
+ "train_runtime": 18970.3816,
+ "train_tokens_per_second": 5699.736
+ },
+ {
+ "epoch": 0.16630756800430918,
+ "grad_norm": 0.9452365824285418,
+ "learning_rate": 9.830494877077137e-06,
+ "loss": 0.46334171295166016,
+ "num_input_tokens_seen": 108358872,
+ "step": 2470,
+ "train_runtime": 19004.9571,
+ "train_tokens_per_second": 5701.611
+ },
+ {
+ "epoch": 0.16664422300026932,
+ "grad_norm": 1.1889597050506375,
+ "learning_rate": 9.82981157377537e-06,
+ "loss": 0.4094019889831543,
+ "num_input_tokens_seen": 108572432,
+ "step": 2475,
+ "train_runtime": 19047.273,
+ "train_tokens_per_second": 5700.156
+ },
+ {
+ "epoch": 0.16698087799622946,
+ "grad_norm": 0.8308872657286545,
+ "learning_rate": 9.829126919832198e-06,
+ "loss": 0.3822948455810547,
+ "num_input_tokens_seen": 108789616,
+ "step": 2480,
+ "train_runtime": 19086.2476,
+ "train_tokens_per_second": 5699.895
+ },
+ {
+ "epoch": 0.1673175329921896,
+ "grad_norm": 0.7290133116509023,
+ "learning_rate": 9.82844091543908e-06,
+ "loss": 0.4369968414306641,
+ "num_input_tokens_seen": 109023616,
+ "step": 2485,
+ "train_runtime": 19125.718,
+ "train_tokens_per_second": 5700.367
+ },
+ {
+ "epoch": 0.16765418798814974,
+ "grad_norm": 0.7990719470039959,
+ "learning_rate": 9.82775356078786e-06,
+ "loss": 0.4084890842437744,
+ "num_input_tokens_seen": 109238536,
+ "step": 2490,
+ "train_runtime": 19167.2805,
+ "train_tokens_per_second": 5699.219
+ },
+ {
+ "epoch": 0.16799084298410988,
+ "grad_norm": 0.7393931696314984,
+ "learning_rate": 9.82706485607075e-06,
+ "loss": 0.3479023456573486,
+ "num_input_tokens_seen": 109448064,
+ "step": 2495,
+ "train_runtime": 19202.9034,
+ "train_tokens_per_second": 5699.558
+ },
+ {
+ "epoch": 0.16832749798007002,
+ "grad_norm": 0.8223012662270877,
+ "learning_rate": 9.826374801480346e-06,
+ "loss": 0.4044170379638672,
+ "num_input_tokens_seen": 109654600,
+ "step": 2500,
+ "train_runtime": 19240.4274,
+ "train_tokens_per_second": 5699.177
+ },
+ {
+ "epoch": 0.16866415297603016,
+ "grad_norm": 0.7578035444329189,
+ "learning_rate": 9.825683397209617e-06,
+ "loss": 0.4128159999847412,
+ "num_input_tokens_seen": 109844880,
+ "step": 2505,
+ "train_runtime": 19277.0081,
+ "train_tokens_per_second": 5698.233
+ },
+ {
+ "epoch": 0.1690008079719903,
+ "grad_norm": 0.7754496384426975,
+ "learning_rate": 9.824990643451915e-06,
+ "loss": 0.40271682739257814,
+ "num_input_tokens_seen": 110072368,
+ "step": 2510,
+ "train_runtime": 19317.0275,
+ "train_tokens_per_second": 5698.204
+ },
+ {
+ "epoch": 0.16933746296795044,
+ "grad_norm": 0.7630225519369642,
+ "learning_rate": 9.824296540400965e-06,
+ "loss": 0.4032273769378662,
+ "num_input_tokens_seen": 110306168,
+ "step": 2515,
+ "train_runtime": 19359.4719,
+ "train_tokens_per_second": 5697.788
+ },
+ {
+ "epoch": 0.16967411796391058,
+ "grad_norm": 0.9445770110284926,
+ "learning_rate": 9.823601088250872e-06,
+ "loss": 0.43148112297058105,
+ "num_input_tokens_seen": 110517392,
+ "step": 2520,
+ "train_runtime": 19399.6091,
+ "train_tokens_per_second": 5696.888
+ },
+ {
+ "epoch": 0.17001077295987072,
+ "grad_norm": 0.8318843666554462,
+ "learning_rate": 9.822904287196114e-06,
+ "loss": 0.4280102729797363,
+ "num_input_tokens_seen": 110729000,
+ "step": 2525,
+ "train_runtime": 19437.1917,
+ "train_tokens_per_second": 5696.759
+ },
+ {
+ "epoch": 0.17034742795583085,
+ "grad_norm": 0.8164720266895754,
+ "learning_rate": 9.822206137431553e-06,
+ "loss": 0.4008312702178955,
+ "num_input_tokens_seen": 110951576,
+ "step": 2530,
+ "train_runtime": 19473.1465,
+ "train_tokens_per_second": 5697.671
+ },
+ {
+ "epoch": 0.170684082951791,
+ "grad_norm": 0.7906437056018215,
+ "learning_rate": 9.821506639152421e-06,
+ "loss": 0.42267236709594724,
+ "num_input_tokens_seen": 111165912,
+ "step": 2535,
+ "train_runtime": 19511.9498,
+ "train_tokens_per_second": 5697.325
+ },
+ {
+ "epoch": 0.17102073794775113,
+ "grad_norm": 0.8027232200064034,
+ "learning_rate": 9.820805792554334e-06,
+ "loss": 0.3978217124938965,
+ "num_input_tokens_seen": 111390648,
+ "step": 2540,
+ "train_runtime": 19548.9277,
+ "train_tokens_per_second": 5698.044
+ },
+ {
+ "epoch": 0.17135739294371127,
+ "grad_norm": 0.9732933783691021,
+ "learning_rate": 9.820103597833278e-06,
+ "loss": 0.40553808212280273,
+ "num_input_tokens_seen": 111613680,
+ "step": 2545,
+ "train_runtime": 19588.8813,
+ "train_tokens_per_second": 5697.808
+ },
+ {
+ "epoch": 0.1716940479396714,
+ "grad_norm": 0.82729851843596,
+ "learning_rate": 9.819400055185622e-06,
+ "loss": 0.3968106031417847,
+ "num_input_tokens_seen": 111830672,
+ "step": 2550,
+ "train_runtime": 19624.1494,
+ "train_tokens_per_second": 5698.625
+ },
+ {
+ "epoch": 0.17203070293563155,
+ "grad_norm": 0.8468905353541687,
+ "learning_rate": 9.818695164808108e-06,
+ "loss": 0.40885286331176757,
+ "num_input_tokens_seen": 112047832,
+ "step": 2555,
+ "train_runtime": 19668.7374,
+ "train_tokens_per_second": 5696.748
+ },
+ {
+ "epoch": 0.1723673579315917,
+ "grad_norm": 0.7939112227025558,
+ "learning_rate": 9.817988926897856e-06,
+ "loss": 0.39391307830810546,
+ "num_input_tokens_seen": 112267160,
+ "step": 2560,
+ "train_runtime": 19710.9584,
+ "train_tokens_per_second": 5695.672
+ },
+ {
+ "epoch": 0.17270401292755186,
+ "grad_norm": 0.8639441951747879,
+ "learning_rate": 9.817281341652367e-06,
+ "loss": 0.4082347869873047,
+ "num_input_tokens_seen": 112491800,
+ "step": 2565,
+ "train_runtime": 19749.2522,
+ "train_tokens_per_second": 5696.003
+ },
+ {
+ "epoch": 0.173040667923512,
+ "grad_norm": 0.716613288966692,
+ "learning_rate": 9.816572409269508e-06,
+ "loss": 0.3824393033981323,
+ "num_input_tokens_seen": 112719560,
+ "step": 2570,
+ "train_runtime": 19784.3744,
+ "train_tokens_per_second": 5697.403
+ },
+ {
+ "epoch": 0.17337732291947214,
+ "grad_norm": 0.7563163756690298,
+ "learning_rate": 9.815862129947537e-06,
+ "loss": 0.4002216339111328,
+ "num_input_tokens_seen": 112940512,
+ "step": 2575,
+ "train_runtime": 19821.7603,
+ "train_tokens_per_second": 5697.804
+ },
+ {
+ "epoch": 0.17371397791543228,
+ "grad_norm": 0.8087466973345561,
+ "learning_rate": 9.815150503885078e-06,
+ "loss": 0.37178959846496584,
+ "num_input_tokens_seen": 113165624,
+ "step": 2580,
+ "train_runtime": 19858.1999,
+ "train_tokens_per_second": 5698.685
+ },
+ {
+ "epoch": 0.17405063291139242,
+ "grad_norm": 0.8994912714777122,
+ "learning_rate": 9.814437531281133e-06,
+ "loss": 0.40114603042602537,
+ "num_input_tokens_seen": 113386744,
+ "step": 2585,
+ "train_runtime": 19903.4827,
+ "train_tokens_per_second": 5696.829
+ },
+ {
+ "epoch": 0.17438728790735256,
+ "grad_norm": 0.9269225867551159,
+ "learning_rate": 9.813723212335085e-06,
+ "loss": 0.43936843872070314,
+ "num_input_tokens_seen": 113593688,
+ "step": 2590,
+ "train_runtime": 19940.3903,
+ "train_tokens_per_second": 5696.663
+ },
+ {
+ "epoch": 0.1747239429033127,
+ "grad_norm": 0.8457918670255721,
+ "learning_rate": 9.813007547246691e-06,
+ "loss": 0.39149506092071534,
+ "num_input_tokens_seen": 113810536,
+ "step": 2595,
+ "train_runtime": 19976.29,
+ "train_tokens_per_second": 5697.281
+ },
+ {
+ "epoch": 0.17506059789927284,
+ "grad_norm": 0.8422559964504273,
+ "learning_rate": 9.812290536216083e-06,
+ "loss": 0.4519649982452393,
+ "num_input_tokens_seen": 114012584,
+ "step": 2600,
+ "train_runtime": 20014.4208,
+ "train_tokens_per_second": 5696.522
+ },
+ {
+ "epoch": 0.17539725289523297,
+ "grad_norm": 0.7614963966994434,
+ "learning_rate": 9.811572179443773e-06,
+ "loss": 0.3978102445602417,
+ "num_input_tokens_seen": 114240040,
+ "step": 2605,
+ "train_runtime": 20049.8879,
+ "train_tokens_per_second": 5697.789
+ },
+ {
+ "epoch": 0.17573390789119311,
+ "grad_norm": 0.7655048641611386,
+ "learning_rate": 9.810852477130645e-06,
+ "loss": 0.40557031631469725,
+ "num_input_tokens_seen": 114449696,
+ "step": 2610,
+ "train_runtime": 20089.5257,
+ "train_tokens_per_second": 5696.983
+ },
+ {
+ "epoch": 0.17607056288715325,
+ "grad_norm": 0.8093163549967093,
+ "learning_rate": 9.810131429477964e-06,
+ "loss": 0.38374016284942625,
+ "num_input_tokens_seen": 114677776,
+ "step": 2615,
+ "train_runtime": 20121.5227,
+ "train_tokens_per_second": 5699.259
+ },
+ {
+ "epoch": 0.1764072178831134,
+ "grad_norm": 0.8258310946925135,
+ "learning_rate": 9.809409036687364e-06,
+ "loss": 0.396789026260376,
+ "num_input_tokens_seen": 114896336,
+ "step": 2620,
+ "train_runtime": 20158.9581,
+ "train_tokens_per_second": 5699.518
+ },
+ {
+ "epoch": 0.17674387287907353,
+ "grad_norm": 0.7382001307664883,
+ "learning_rate": 9.808685298960867e-06,
+ "loss": 0.4137075424194336,
+ "num_input_tokens_seen": 115125264,
+ "step": 2625,
+ "train_runtime": 20201.5972,
+ "train_tokens_per_second": 5698.82
+ },
+ {
+ "epoch": 0.17708052787503367,
+ "grad_norm": 0.7663556283529158,
+ "learning_rate": 9.807960216500858e-06,
+ "loss": 0.4165202617645264,
+ "num_input_tokens_seen": 115343304,
+ "step": 2630,
+ "train_runtime": 20236.3656,
+ "train_tokens_per_second": 5699.803
+ },
+ {
+ "epoch": 0.1774171828709938,
+ "grad_norm": 0.7884797087616878,
+ "learning_rate": 9.807233789510106e-06,
+ "loss": 0.38329482078552246,
+ "num_input_tokens_seen": 115555208,
+ "step": 2635,
+ "train_runtime": 20279.8657,
+ "train_tokens_per_second": 5698.026
+ },
+ {
+ "epoch": 0.17775383786695395,
+ "grad_norm": 0.7323486746827084,
+ "learning_rate": 9.806506018191754e-06,
+ "loss": 0.39183108806610106,
+ "num_input_tokens_seen": 115777312,
+ "step": 2640,
+ "train_runtime": 20317.1207,
+ "train_tokens_per_second": 5698.51
+ },
+ {
+ "epoch": 0.1780904928629141,
+ "grad_norm": 0.8712184035601627,
+ "learning_rate": 9.80577690274932e-06,
+ "loss": 0.39883904457092284,
+ "num_input_tokens_seen": 116000320,
+ "step": 2645,
+ "train_runtime": 20353.9248,
+ "train_tokens_per_second": 5699.162
+ },
+ {
+ "epoch": 0.17842714785887423,
+ "grad_norm": 0.6517931345598202,
+ "learning_rate": 9.8050464433867e-06,
+ "loss": 0.3788264751434326,
+ "num_input_tokens_seen": 116228264,
+ "step": 2650,
+ "train_runtime": 20388.2032,
+ "train_tokens_per_second": 5700.761
+ },
+ {
+ "epoch": 0.17876380285483437,
+ "grad_norm": 0.8947080972086597,
+ "learning_rate": 9.804314640308165e-06,
+ "loss": 0.43610692024230957,
+ "num_input_tokens_seen": 116459784,
+ "step": 2655,
+ "train_runtime": 20433.6777,
+ "train_tokens_per_second": 5699.404
+ },
+ {
+ "epoch": 0.1791004578507945,
+ "grad_norm": 0.8316768119557644,
+ "learning_rate": 9.803581493718361e-06,
+ "loss": 0.4152191162109375,
+ "num_input_tokens_seen": 116677976,
+ "step": 2660,
+ "train_runtime": 20471.066,
+ "train_tokens_per_second": 5699.653
+ },
+ {
+ "epoch": 0.17943711284675465,
+ "grad_norm": 0.7694745863442054,
+ "learning_rate": 9.80284700382231e-06,
+ "loss": 0.4158869743347168,
+ "num_input_tokens_seen": 116913096,
+ "step": 2665,
+ "train_runtime": 20513.9974,
+ "train_tokens_per_second": 5699.186
+ },
+ {
+ "epoch": 0.1797737678427148,
+ "grad_norm": 0.7759018142960451,
+ "learning_rate": 9.802111170825408e-06,
+ "loss": 0.40905094146728516,
+ "num_input_tokens_seen": 117139040,
+ "step": 2670,
+ "train_runtime": 20555.3803,
+ "train_tokens_per_second": 5698.705
+ },
+ {
+ "epoch": 0.18011042283867493,
+ "grad_norm": 0.7209579007108823,
+ "learning_rate": 9.80137399493343e-06,
+ "loss": 0.4070711135864258,
+ "num_input_tokens_seen": 117354016,
+ "step": 2675,
+ "train_runtime": 20594.1226,
+ "train_tokens_per_second": 5698.423
+ },
+ {
+ "epoch": 0.18044707783463507,
+ "grad_norm": 0.7855863516525835,
+ "learning_rate": 9.800635476352525e-06,
+ "loss": 0.39305026531219484,
+ "num_input_tokens_seen": 117552920,
+ "step": 2680,
+ "train_runtime": 20636.4809,
+ "train_tokens_per_second": 5696.365
+ },
+ {
+ "epoch": 0.1807837328305952,
+ "grad_norm": 0.8011882919088659,
+ "learning_rate": 9.799895615289218e-06,
+ "loss": 0.38626153469085694,
+ "num_input_tokens_seen": 117762232,
+ "step": 2685,
+ "train_runtime": 20677.4885,
+ "train_tokens_per_second": 5695.19
+ },
+ {
+ "epoch": 0.18112038782655535,
+ "grad_norm": 0.7959442838200258,
+ "learning_rate": 9.799154411950407e-06,
+ "loss": 0.44461402893066404,
+ "num_input_tokens_seen": 117998240,
+ "step": 2690,
+ "train_runtime": 20716.7034,
+ "train_tokens_per_second": 5695.802
+ },
+ {
+ "epoch": 0.18145704282251549,
+ "grad_norm": 0.7750011575792003,
+ "learning_rate": 9.79841186654337e-06,
+ "loss": 0.4056252479553223,
+ "num_input_tokens_seen": 118211832,
+ "step": 2695,
+ "train_runtime": 20752.9555,
+ "train_tokens_per_second": 5696.144
+ },
+ {
+ "epoch": 0.18179369781847562,
+ "grad_norm": 0.8899930605283304,
+ "learning_rate": 9.797667979275752e-06,
+ "loss": 0.43903465270996095,
+ "num_input_tokens_seen": 118421416,
+ "step": 2700,
+ "train_runtime": 20785.8453,
+ "train_tokens_per_second": 5697.214
+ },
+ {
+ "epoch": 0.18213035281443576,
+ "grad_norm": 0.7964676861353408,
+ "learning_rate": 9.796922750355584e-06,
+ "loss": 0.3816461801528931,
+ "num_input_tokens_seen": 118635408,
+ "step": 2705,
+ "train_runtime": 20823.8564,
+ "train_tokens_per_second": 5697.091
+ },
+ {
+ "epoch": 0.1824670078103959,
+ "grad_norm": 0.8627424628290897,
+ "learning_rate": 9.796176179991266e-06,
+ "loss": 0.4243772029876709,
+ "num_input_tokens_seen": 118853520,
+ "step": 2710,
+ "train_runtime": 20861.0922,
+ "train_tokens_per_second": 5697.378
+ },
+ {
+ "epoch": 0.18280366280635604,
+ "grad_norm": 0.8313804598527397,
+ "learning_rate": 9.795428268391572e-06,
+ "loss": 0.39929049015045165,
+ "num_input_tokens_seen": 119057464,
+ "step": 2715,
+ "train_runtime": 20899.2219,
+ "train_tokens_per_second": 5696.741
+ },
+ {
+ "epoch": 0.18314031780231618,
+ "grad_norm": 0.9224656362131587,
+ "learning_rate": 9.794679015765655e-06,
+ "loss": 0.4025382995605469,
+ "num_input_tokens_seen": 119270504,
+ "step": 2720,
+ "train_runtime": 20936.4359,
+ "train_tokens_per_second": 5696.791
+ },
+ {
+ "epoch": 0.18347697279827632,
+ "grad_norm": 0.8830106782970203,
+ "learning_rate": 9.793928422323041e-06,
+ "loss": 0.40293426513671876,
+ "num_input_tokens_seen": 119494864,
+ "step": 2725,
+ "train_runtime": 20982.2238,
+ "train_tokens_per_second": 5695.052
+ },
+ {
+ "epoch": 0.18381362779423646,
+ "grad_norm": 0.9585116874337107,
+ "learning_rate": 9.793176488273629e-06,
+ "loss": 0.40497751235961915,
+ "num_input_tokens_seen": 119700000,
+ "step": 2730,
+ "train_runtime": 21018.3132,
+ "train_tokens_per_second": 5695.034
+ },
+ {
+ "epoch": 0.1841502827901966,
+ "grad_norm": 0.9047740005031505,
+ "learning_rate": 9.792423213827697e-06,
+ "loss": 0.3891148567199707,
+ "num_input_tokens_seen": 119916696,
+ "step": 2735,
+ "train_runtime": 21057.8627,
+ "train_tokens_per_second": 5694.628
+ },
+ {
+ "epoch": 0.18448693778615674,
+ "grad_norm": 0.8310766014320071,
+ "learning_rate": 9.791668599195897e-06,
+ "loss": 0.3831428289413452,
+ "num_input_tokens_seen": 120142080,
+ "step": 2740,
+ "train_runtime": 21101.6665,
+ "train_tokens_per_second": 5693.488
+ },
+ {
+ "epoch": 0.18482359278211688,
+ "grad_norm": 0.7855666850049595,
+ "learning_rate": 9.790912644589248e-06,
+ "loss": 0.42138309478759767,
+ "num_input_tokens_seen": 120348496,
+ "step": 2745,
+ "train_runtime": 21146.0201,
+ "train_tokens_per_second": 5691.307
+ },
+ {
+ "epoch": 0.18516024777807702,
+ "grad_norm": 0.7620864321495552,
+ "learning_rate": 9.790155350219159e-06,
+ "loss": 0.3965245246887207,
+ "num_input_tokens_seen": 120570160,
+ "step": 2750,
+ "train_runtime": 21186.6351,
+ "train_tokens_per_second": 5690.859
+ },
+ {
+ "epoch": 0.18549690277403716,
+ "grad_norm": 0.7543478062618101,
+ "learning_rate": 9.789396716297399e-06,
+ "loss": 0.4091955661773682,
+ "num_input_tokens_seen": 120800688,
+ "step": 2755,
+ "train_runtime": 21220.0669,
+ "train_tokens_per_second": 5692.757
+ },
+ {
+ "epoch": 0.1858335577699973,
+ "grad_norm": 0.8640772187026432,
+ "learning_rate": 9.788636743036122e-06,
+ "loss": 0.38339416980743407,
+ "num_input_tokens_seen": 121012328,
+ "step": 2760,
+ "train_runtime": 21257.8451,
+ "train_tokens_per_second": 5692.596
+ },
+ {
+ "epoch": 0.18617021276595744,
+ "grad_norm": 0.7868554451548372,
+ "learning_rate": 9.787875430647847e-06,
+ "loss": 0.3950460195541382,
+ "num_input_tokens_seen": 121232032,
+ "step": 2765,
+ "train_runtime": 21294.5091,
+ "train_tokens_per_second": 5693.112
+ },
+ {
+ "epoch": 0.18650686776191758,
+ "grad_norm": 0.8537247075271003,
+ "learning_rate": 9.787112779345477e-06,
+ "loss": 0.39361560344696045,
+ "num_input_tokens_seen": 121436200,
+ "step": 2770,
+ "train_runtime": 21331.5416,
+ "train_tokens_per_second": 5692.8
+ },
+ {
+ "epoch": 0.18684352275787772,
+ "grad_norm": 0.8358704371476329,
+ "learning_rate": 9.786348789342283e-06,
+ "loss": 0.4345618724822998,
+ "num_input_tokens_seen": 121656544,
+ "step": 2775,
+ "train_runtime": 21370.3261,
+ "train_tokens_per_second": 5692.779
+ },
+ {
+ "epoch": 0.18718017775383786,
+ "grad_norm": 0.7754752799232651,
+ "learning_rate": 9.785583460851912e-06,
+ "loss": 0.39746851921081544,
+ "num_input_tokens_seen": 121881368,
+ "step": 2780,
+ "train_runtime": 21412.528,
+ "train_tokens_per_second": 5692.059
+ },
+ {
+ "epoch": 0.187516832749798,
+ "grad_norm": 0.7537619511207324,
+ "learning_rate": 9.784816794088386e-06,
+ "loss": 0.4182771682739258,
+ "num_input_tokens_seen": 122116272,
+ "step": 2785,
+ "train_runtime": 21450.3054,
+ "train_tokens_per_second": 5692.985
+ },
+ {
+ "epoch": 0.18785348774575814,
+ "grad_norm": 0.86793765719075,
+ "learning_rate": 9.784048789266101e-06,
+ "loss": 0.4000124931335449,
+ "num_input_tokens_seen": 122324024,
+ "step": 2790,
+ "train_runtime": 21493.8907,
+ "train_tokens_per_second": 5691.107
+ },
+ {
+ "epoch": 0.18819014274171827,
+ "grad_norm": 0.7757014341183018,
+ "learning_rate": 9.78327944659983e-06,
+ "loss": 0.411564302444458,
+ "num_input_tokens_seen": 122539808,
+ "step": 2795,
+ "train_runtime": 21541.5867,
+ "train_tokens_per_second": 5688.523
+ },
+ {
+ "epoch": 0.18852679773767841,
+ "grad_norm": 0.8345683649256179,
+ "learning_rate": 9.78250876630471e-06,
+ "loss": 0.40423192977905276,
+ "num_input_tokens_seen": 122747320,
+ "step": 2800,
+ "train_runtime": 21579.8161,
+ "train_tokens_per_second": 5688.061
+ },
+ {
+ "epoch": 0.18886345273363855,
+ "grad_norm": 0.8512506107699013,
+ "learning_rate": 9.781736748596267e-06,
+ "loss": 0.39595718383789064,
+ "num_input_tokens_seen": 122957984,
+ "step": 2805,
+ "train_runtime": 21616.5661,
+ "train_tokens_per_second": 5688.137
+ },
+ {
+ "epoch": 0.18920010772959872,
+ "grad_norm": 0.8124112022835439,
+ "learning_rate": 9.78096339369039e-06,
+ "loss": 0.3889493942260742,
+ "num_input_tokens_seen": 123167712,
+ "step": 2810,
+ "train_runtime": 21651.6405,
+ "train_tokens_per_second": 5688.609
+ },
+ {
+ "epoch": 0.18953676272555886,
+ "grad_norm": 0.7367215294516509,
+ "learning_rate": 9.780188701803345e-06,
+ "loss": 0.42559328079223635,
+ "num_input_tokens_seen": 123386632,
+ "step": 2815,
+ "train_runtime": 21694.1793,
+ "train_tokens_per_second": 5687.546
+ },
+ {
+ "epoch": 0.189873417721519,
+ "grad_norm": 0.7954578075141111,
+ "learning_rate": 9.779412673151772e-06,
+ "loss": 0.40504965782165525,
+ "num_input_tokens_seen": 123614712,
+ "step": 2820,
+ "train_runtime": 21735.5555,
+ "train_tokens_per_second": 5687.212
+ },
+ {
+ "epoch": 0.19021007271747914,
+ "grad_norm": 0.7248412230311774,
+ "learning_rate": 9.778635307952686e-06,
+ "loss": 0.4043741703033447,
+ "num_input_tokens_seen": 123839728,
+ "step": 2825,
+ "train_runtime": 21777.9441,
+ "train_tokens_per_second": 5686.475
+ },
+ {
+ "epoch": 0.19054672771343928,
+ "grad_norm": 0.9358958553909015,
+ "learning_rate": 9.777856606423472e-06,
+ "loss": 0.403910493850708,
+ "num_input_tokens_seen": 124065288,
+ "step": 2830,
+ "train_runtime": 21817.8593,
+ "train_tokens_per_second": 5686.41
+ },
+ {
+ "epoch": 0.19088338270939942,
+ "grad_norm": 0.8463489849936312,
+ "learning_rate": 9.777076568781897e-06,
+ "loss": 0.43606057167053225,
+ "num_input_tokens_seen": 124264360,
+ "step": 2835,
+ "train_runtime": 21855.3285,
+ "train_tokens_per_second": 5685.769
+ },
+ {
+ "epoch": 0.19122003770535956,
+ "grad_norm": 0.9165712627873297,
+ "learning_rate": 9.77629519524609e-06,
+ "loss": 0.3660429954528809,
+ "num_input_tokens_seen": 124479288,
+ "step": 2840,
+ "train_runtime": 21894.9089,
+ "train_tokens_per_second": 5685.307
+ },
+ {
+ "epoch": 0.1915566927013197,
+ "grad_norm": 0.7656937544219163,
+ "learning_rate": 9.775512486034564e-06,
+ "loss": 0.4081143379211426,
+ "num_input_tokens_seen": 124692848,
+ "step": 2845,
+ "train_runtime": 21938.4076,
+ "train_tokens_per_second": 5683.769
+ },
+ {
+ "epoch": 0.19189334769727984,
+ "grad_norm": 0.867416455867072,
+ "learning_rate": 9.774728441366197e-06,
+ "loss": 0.4261101245880127,
+ "num_input_tokens_seen": 124899520,
+ "step": 2850,
+ "train_runtime": 21975.3012,
+ "train_tokens_per_second": 5683.632
+ },
+ {
+ "epoch": 0.19223000269323998,
+ "grad_norm": 0.760427761575941,
+ "learning_rate": 9.773943061460249e-06,
+ "loss": 0.39615979194641116,
+ "num_input_tokens_seen": 125109776,
+ "step": 2855,
+ "train_runtime": 22015.122,
+ "train_tokens_per_second": 5682.902
+ },
+ {
+ "epoch": 0.19256665768920012,
+ "grad_norm": 0.8768768754679016,
+ "learning_rate": 9.773156346536345e-06,
+ "loss": 0.42858476638793946,
+ "num_input_tokens_seen": 125326624,
+ "step": 2860,
+ "train_runtime": 22054.6435,
+ "train_tokens_per_second": 5682.55
+ },
+ {
+ "epoch": 0.19290331268516026,
+ "grad_norm": 0.7656261692386461,
+ "learning_rate": 9.772368296814489e-06,
+ "loss": 0.3920739650726318,
+ "num_input_tokens_seen": 125545392,
+ "step": 2865,
+ "train_runtime": 22088.3487,
+ "train_tokens_per_second": 5683.784
+ },
+ {
+ "epoch": 0.1932399676811204,
+ "grad_norm": 0.7687451750642236,
+ "learning_rate": 9.771578912515058e-06,
+ "loss": 0.439731502532959,
+ "num_input_tokens_seen": 125763512,
+ "step": 2870,
+ "train_runtime": 22123.8551,
+ "train_tokens_per_second": 5684.521
+ },
+ {
+ "epoch": 0.19357662267708053,
+ "grad_norm": 3.7881124274468654,
+ "learning_rate": 9.7707881938588e-06,
+ "loss": 0.3899986743927002,
+ "num_input_tokens_seen": 125983312,
+ "step": 2875,
+ "train_runtime": 22166.3825,
+ "train_tokens_per_second": 5683.531
+ },
+ {
+ "epoch": 0.19391327767304067,
+ "grad_norm": 0.7637738370885714,
+ "learning_rate": 9.769996141066836e-06,
+ "loss": 0.4218841552734375,
+ "num_input_tokens_seen": 126221208,
+ "step": 2880,
+ "train_runtime": 22196.5764,
+ "train_tokens_per_second": 5686.517
+ },
+ {
+ "epoch": 0.1942499326690008,
+ "grad_norm": 0.7196684362616566,
+ "learning_rate": 9.76920275436066e-06,
+ "loss": 0.41857566833496096,
+ "num_input_tokens_seen": 126460024,
+ "step": 2885,
+ "train_runtime": 22232.6411,
+ "train_tokens_per_second": 5688.034
+ },
+ {
+ "epoch": 0.19458658766496095,
+ "grad_norm": 0.7681501572608938,
+ "learning_rate": 9.768408033962145e-06,
+ "loss": 0.4293510437011719,
+ "num_input_tokens_seen": 126682488,
+ "step": 2890,
+ "train_runtime": 22274.9037,
+ "train_tokens_per_second": 5687.229
+ },
+ {
+ "epoch": 0.1949232426609211,
+ "grad_norm": 0.7776932982886595,
+ "learning_rate": 9.767611980093525e-06,
+ "loss": 0.40836372375488283,
+ "num_input_tokens_seen": 126903552,
+ "step": 2895,
+ "train_runtime": 22315.0672,
+ "train_tokens_per_second": 5686.9
+ },
+ {
+ "epoch": 0.19525989765688123,
+ "grad_norm": 0.7725164291944147,
+ "learning_rate": 9.76681459297742e-06,
+ "loss": 0.41483469009399415,
+ "num_input_tokens_seen": 127124488,
+ "step": 2900,
+ "train_runtime": 22361.1218,
+ "train_tokens_per_second": 5685.068
+ },
+ {
+ "epoch": 0.19559655265284137,
+ "grad_norm": 1.0523171028632952,
+ "learning_rate": 9.766015872836814e-06,
+ "loss": 0.4586020469665527,
+ "num_input_tokens_seen": 127351272,
+ "step": 2905,
+ "train_runtime": 22403.7587,
+ "train_tokens_per_second": 5684.371
+ },
+ {
+ "epoch": 0.1959332076488015,
+ "grad_norm": 0.981785930598382,
+ "learning_rate": 9.765215819895066e-06,
+ "loss": 0.4265925407409668,
+ "num_input_tokens_seen": 127584848,
+ "step": 2910,
+ "train_runtime": 22435.5709,
+ "train_tokens_per_second": 5686.722
+ },
+ {
+ "epoch": 0.19626986264476165,
+ "grad_norm": 0.7598810830140263,
+ "learning_rate": 9.76441443437591e-06,
+ "loss": 0.3775201320648193,
+ "num_input_tokens_seen": 127815392,
+ "step": 2915,
+ "train_runtime": 22468.3308,
+ "train_tokens_per_second": 5688.691
+ },
+ {
+ "epoch": 0.1966065176407218,
+ "grad_norm": 0.8036869153351719,
+ "learning_rate": 9.76361171650345e-06,
+ "loss": 0.4249995708465576,
+ "num_input_tokens_seen": 128025072,
+ "step": 2920,
+ "train_runtime": 22508.1392,
+ "train_tokens_per_second": 5687.946
+ },
+ {
+ "epoch": 0.19694317263668193,
+ "grad_norm": 0.8854228803465163,
+ "learning_rate": 9.762807666502164e-06,
+ "loss": 0.41265301704406737,
+ "num_input_tokens_seen": 128236720,
+ "step": 2925,
+ "train_runtime": 22548.1044,
+ "train_tokens_per_second": 5687.251
+ },
+ {
+ "epoch": 0.19727982763264207,
+ "grad_norm": 1.302088487080232,
+ "learning_rate": 9.762002284596898e-06,
+ "loss": 0.4183157444000244,
+ "num_input_tokens_seen": 128448432,
+ "step": 2930,
+ "train_runtime": 22591.5125,
+ "train_tokens_per_second": 5685.694
+ },
+ {
+ "epoch": 0.1976164826286022,
+ "grad_norm": 0.8232254317537885,
+ "learning_rate": 9.761195571012882e-06,
+ "loss": 0.4257832050323486,
+ "num_input_tokens_seen": 128657896,
+ "step": 2935,
+ "train_runtime": 22629.0733,
+ "train_tokens_per_second": 5685.513
+ },
+ {
+ "epoch": 0.19795313762456235,
+ "grad_norm": 0.8172096711574582,
+ "learning_rate": 9.760387525975705e-06,
+ "loss": 0.395475172996521,
+ "num_input_tokens_seen": 128883360,
+ "step": 2940,
+ "train_runtime": 22662.7742,
+ "train_tokens_per_second": 5687.007
+ },
+ {
+ "epoch": 0.1982897926205225,
+ "grad_norm": 0.7517384205877706,
+ "learning_rate": 9.759578149711337e-06,
+ "loss": 0.35533039569854735,
+ "num_input_tokens_seen": 129112664,
+ "step": 2945,
+ "train_runtime": 22703.1966,
+ "train_tokens_per_second": 5686.982
+ },
+ {
+ "epoch": 0.19862644761648263,
+ "grad_norm": 0.7164265548546924,
+ "learning_rate": 9.758767442446117e-06,
+ "loss": 0.38588006496429444,
+ "num_input_tokens_seen": 129342864,
+ "step": 2950,
+ "train_runtime": 22748.3032,
+ "train_tokens_per_second": 5685.825
+ },
+ {
+ "epoch": 0.19896310261244277,
+ "grad_norm": 0.733779366224309,
+ "learning_rate": 9.757955404406755e-06,
+ "loss": 0.38382346630096437,
+ "num_input_tokens_seen": 129566032,
+ "step": 2955,
+ "train_runtime": 22786.3627,
+ "train_tokens_per_second": 5686.122
+ },
+ {
+ "epoch": 0.1992997576084029,
+ "grad_norm": 0.9448244412990339,
+ "learning_rate": 9.757142035820337e-06,
+ "loss": 0.3917106628417969,
+ "num_input_tokens_seen": 129766056,
+ "step": 2960,
+ "train_runtime": 22824.9574,
+ "train_tokens_per_second": 5685.27
+ },
+ {
+ "epoch": 0.19963641260436304,
+ "grad_norm": 0.9552752355322248,
+ "learning_rate": 9.756327336914318e-06,
+ "loss": 0.4134823322296143,
+ "num_input_tokens_seen": 129977048,
+ "step": 2965,
+ "train_runtime": 22868.2844,
+ "train_tokens_per_second": 5683.725
+ },
+ {
+ "epoch": 0.19997306760032318,
+ "grad_norm": 1.0910385569978054,
+ "learning_rate": 9.755511307916526e-06,
+ "loss": 0.4055294990539551,
+ "num_input_tokens_seen": 130194224,
+ "step": 2970,
+ "train_runtime": 22901.3701,
+ "train_tokens_per_second": 5684.997
+ },
+ {
+ "epoch": 0.20030972259628332,
+ "grad_norm": 0.8363361731098393,
+ "learning_rate": 9.754693949055161e-06,
+ "loss": 0.4119466781616211,
+ "num_input_tokens_seen": 130423144,
+ "step": 2975,
+ "train_runtime": 22941.2287,
+ "train_tokens_per_second": 5685.098
+ },
+ {
+ "epoch": 0.20064637759224346,
+ "grad_norm": 0.8509098752743555,
+ "learning_rate": 9.753875260558795e-06,
+ "loss": 0.43303780555725097,
+ "num_input_tokens_seen": 130623168,
+ "step": 2980,
+ "train_runtime": 22977.1333,
+ "train_tokens_per_second": 5684.92
+ },
+ {
+ "epoch": 0.2009830325882036,
+ "grad_norm": 0.7851629802721982,
+ "learning_rate": 9.75305524265637e-06,
+ "loss": 0.39323015213012696,
+ "num_input_tokens_seen": 130853536,
+ "step": 2985,
+ "train_runtime": 23010.865,
+ "train_tokens_per_second": 5686.598
+ },
+ {
+ "epoch": 0.20131968758416374,
+ "grad_norm": 0.8423251979034995,
+ "learning_rate": 9.752233895577205e-06,
+ "loss": 0.4042644023895264,
+ "num_input_tokens_seen": 131059592,
+ "step": 2990,
+ "train_runtime": 23047.2444,
+ "train_tokens_per_second": 5686.562
+ },
+ {
+ "epoch": 0.20165634258012388,
+ "grad_norm": 0.7311264203424918,
+ "learning_rate": 9.751411219550984e-06,
+ "loss": 0.40361328125,
+ "num_input_tokens_seen": 131276760,
+ "step": 2995,
+ "train_runtime": 23082.7777,
+ "train_tokens_per_second": 5687.217
+ },
+ {
+ "epoch": 0.20199299757608402,
+ "grad_norm": 0.9307523548267392,
+ "learning_rate": 9.750587214807766e-06,
+ "loss": 0.39671123027801514,
+ "num_input_tokens_seen": 131486248,
+ "step": 3000,
+ "train_runtime": 23124.9978,
+ "train_tokens_per_second": 5685.892
+ },
+ {
+ "epoch": 0.20232965257204416,
+ "grad_norm": 0.8339744148987415,
+ "learning_rate": 9.749761881577981e-06,
+ "loss": 0.42069306373596194,
+ "num_input_tokens_seen": 131716280,
+ "step": 3005,
+ "train_runtime": 23166.3136,
+ "train_tokens_per_second": 5685.681
+ },
+ {
+ "epoch": 0.2026663075680043,
+ "grad_norm": 0.7849226835237555,
+ "learning_rate": 9.748935220092434e-06,
+ "loss": 0.39777138233184817,
+ "num_input_tokens_seen": 131935832,
+ "step": 3010,
+ "train_runtime": 23202.4925,
+ "train_tokens_per_second": 5686.278
+ },
+ {
+ "epoch": 0.20300296256396444,
+ "grad_norm": 0.8473889903914309,
+ "learning_rate": 9.748107230582295e-06,
+ "loss": 0.39002056121826173,
+ "num_input_tokens_seen": 132140024,
+ "step": 3015,
+ "train_runtime": 23239.8205,
+ "train_tokens_per_second": 5685.931
+ },
+ {
+ "epoch": 0.20333961755992458,
+ "grad_norm": 0.7565501165020906,
+ "learning_rate": 9.747277913279109e-06,
+ "loss": 0.4608659267425537,
+ "num_input_tokens_seen": 132358504,
+ "step": 3020,
+ "train_runtime": 23285.5396,
+ "train_tokens_per_second": 5684.15
+ },
+ {
+ "epoch": 0.20367627255588472,
+ "grad_norm": 0.71921389289753,
+ "learning_rate": 9.746447268414794e-06,
+ "loss": 0.44081912040710447,
+ "num_input_tokens_seen": 132589576,
+ "step": 3025,
+ "train_runtime": 23324.3416,
+ "train_tokens_per_second": 5684.601
+ },
+ {
+ "epoch": 0.20401292755184486,
+ "grad_norm": 0.8592739422898477,
+ "learning_rate": 9.745615296221635e-06,
+ "loss": 0.40747461318969724,
+ "num_input_tokens_seen": 132819496,
+ "step": 3030,
+ "train_runtime": 23358.8471,
+ "train_tokens_per_second": 5686.047
+ },
+ {
+ "epoch": 0.204349582547805,
+ "grad_norm": 0.7375243033673482,
+ "learning_rate": 9.74478199693229e-06,
+ "loss": 0.42429242134094236,
+ "num_input_tokens_seen": 133045880,
+ "step": 3035,
+ "train_runtime": 23401.9716,
+ "train_tokens_per_second": 5685.242
+ },
+ {
+ "epoch": 0.20468623754376514,
+ "grad_norm": 0.7984195828599883,
+ "learning_rate": 9.743947370779791e-06,
+ "loss": 0.4262086391448975,
+ "num_input_tokens_seen": 133273848,
+ "step": 3040,
+ "train_runtime": 23442.6366,
+ "train_tokens_per_second": 5685.105
+ },
+ {
+ "epoch": 0.20502289253972528,
+ "grad_norm": 0.7817233028800991,
+ "learning_rate": 9.743111417997536e-06,
+ "loss": 0.41955132484436036,
+ "num_input_tokens_seen": 133500640,
+ "step": 3045,
+ "train_runtime": 23481.7864,
+ "train_tokens_per_second": 5685.285
+ },
+ {
+ "epoch": 0.20535954753568542,
+ "grad_norm": 0.8413661304916547,
+ "learning_rate": 9.742274138819298e-06,
+ "loss": 0.416639518737793,
+ "num_input_tokens_seen": 133719848,
+ "step": 3050,
+ "train_runtime": 23518.5528,
+ "train_tokens_per_second": 5685.718
+ },
+ {
+ "epoch": 0.20569620253164558,
+ "grad_norm": 0.8835193289624108,
+ "learning_rate": 9.74143553347922e-06,
+ "loss": 0.45115389823913576,
+ "num_input_tokens_seen": 133944240,
+ "step": 3055,
+ "train_runtime": 23558.0231,
+ "train_tokens_per_second": 5685.716
+ },
+ {
+ "epoch": 0.20603285752760572,
+ "grad_norm": 0.7584811920041505,
+ "learning_rate": 9.740595602211814e-06,
+ "loss": 0.395220947265625,
+ "num_input_tokens_seen": 134166880,
+ "step": 3060,
+ "train_runtime": 23594.0834,
+ "train_tokens_per_second": 5686.463
+ },
+ {
+ "epoch": 0.20636951252356586,
+ "grad_norm": 0.779533118470965,
+ "learning_rate": 9.739754345251963e-06,
+ "loss": 0.3915586471557617,
+ "num_input_tokens_seen": 134373184,
+ "step": 3065,
+ "train_runtime": 23633.4149,
+ "train_tokens_per_second": 5685.729
+ },
+ {
+ "epoch": 0.206706167519526,
+ "grad_norm": 0.752085146799199,
+ "learning_rate": 9.738911762834924e-06,
+ "loss": 0.36250555515289307,
+ "num_input_tokens_seen": 134594840,
+ "step": 3070,
+ "train_runtime": 23669.854,
+ "train_tokens_per_second": 5686.34
+ },
+ {
+ "epoch": 0.20704282251548614,
+ "grad_norm": 0.9101615257113173,
+ "learning_rate": 9.738067855196323e-06,
+ "loss": 0.40388903617858884,
+ "num_input_tokens_seen": 134814144,
+ "step": 3075,
+ "train_runtime": 23702.4871,
+ "train_tokens_per_second": 5687.764
+ },
+ {
+ "epoch": 0.20737947751144628,
+ "grad_norm": 0.937886524311701,
+ "learning_rate": 9.737222622572153e-06,
+ "loss": 0.4489736557006836,
+ "num_input_tokens_seen": 135025088,
+ "step": 3080,
+ "train_runtime": 23747.5708,
+ "train_tokens_per_second": 5685.848
+ },
+ {
+ "epoch": 0.20771613250740642,
+ "grad_norm": 0.939422935382809,
+ "learning_rate": 9.736376065198784e-06,
+ "loss": 0.4101541519165039,
+ "num_input_tokens_seen": 135253064,
+ "step": 3085,
+ "train_runtime": 23786.6187,
+ "train_tokens_per_second": 5686.099
+ },
+ {
+ "epoch": 0.20805278750336656,
+ "grad_norm": 0.8449440163538374,
+ "learning_rate": 9.735528183312951e-06,
+ "loss": 0.39399616718292235,
+ "num_input_tokens_seen": 135486752,
+ "step": 3090,
+ "train_runtime": 23827.8882,
+ "train_tokens_per_second": 5686.058
+ },
+ {
+ "epoch": 0.2083894424993267,
+ "grad_norm": 0.7846090848755455,
+ "learning_rate": 9.734678977151762e-06,
+ "loss": 0.3820789098739624,
+ "num_input_tokens_seen": 135702632,
+ "step": 3095,
+ "train_runtime": 23868.8149,
+ "train_tokens_per_second": 5685.353
+ },
+ {
+ "epoch": 0.20872609749528684,
+ "grad_norm": 0.7998350921308711,
+ "learning_rate": 9.733828446952696e-06,
+ "loss": 0.40059328079223633,
+ "num_input_tokens_seen": 135925976,
+ "step": 3100,
+ "train_runtime": 23904.8211,
+ "train_tokens_per_second": 5686.132
+ },
+ {
+ "epoch": 0.20906275249124698,
+ "grad_norm": 0.7794995917456119,
+ "learning_rate": 9.732976592953598e-06,
+ "loss": 0.3992907524108887,
+ "num_input_tokens_seen": 136155256,
+ "step": 3105,
+ "train_runtime": 23947.0473,
+ "train_tokens_per_second": 5685.68
+ },
+ {
+ "epoch": 0.20939940748720712,
+ "grad_norm": 0.8028611006413868,
+ "learning_rate": 9.73212341539269e-06,
+ "loss": 0.40617761611938474,
+ "num_input_tokens_seen": 136377968,
+ "step": 3110,
+ "train_runtime": 23995.829,
+ "train_tokens_per_second": 5683.403
+ },
+ {
+ "epoch": 0.20973606248316726,
+ "grad_norm": 0.7679301554806387,
+ "learning_rate": 9.731268914508556e-06,
+ "loss": 0.39726805686950684,
+ "num_input_tokens_seen": 136595144,
+ "step": 3115,
+ "train_runtime": 24037.8732,
+ "train_tokens_per_second": 5682.497
+ },
+ {
+ "epoch": 0.2100727174791274,
+ "grad_norm": 0.7757940100269232,
+ "learning_rate": 9.730413090540158e-06,
+ "loss": 0.3819112777709961,
+ "num_input_tokens_seen": 136818256,
+ "step": 3120,
+ "train_runtime": 24075.9596,
+ "train_tokens_per_second": 5682.775
+ },
+ {
+ "epoch": 0.21040937247508754,
+ "grad_norm": 0.7507007143030089,
+ "learning_rate": 9.729555943726825e-06,
+ "loss": 0.40000243186950685,
+ "num_input_tokens_seen": 137031872,
+ "step": 3125,
+ "train_runtime": 24120.3223,
+ "train_tokens_per_second": 5681.179
+ },
+ {
+ "epoch": 0.21074602747104768,
+ "grad_norm": 0.8053852599963457,
+ "learning_rate": 9.728697474308253e-06,
+ "loss": 0.40857748985290526,
+ "num_input_tokens_seen": 137246200,
+ "step": 3130,
+ "train_runtime": 24159.6984,
+ "train_tokens_per_second": 5680.791
+ },
+ {
+ "epoch": 0.21108268246700782,
+ "grad_norm": 0.8178855876896101,
+ "learning_rate": 9.72783768252451e-06,
+ "loss": 0.38510382175445557,
+ "num_input_tokens_seen": 137454056,
+ "step": 3135,
+ "train_runtime": 24197.6049,
+ "train_tokens_per_second": 5680.482
+ },
+ {
+ "epoch": 0.21141933746296795,
+ "grad_norm": 0.7840878331434453,
+ "learning_rate": 9.726976568616036e-06,
+ "loss": 0.40147695541381834,
+ "num_input_tokens_seen": 137663648,
+ "step": 3140,
+ "train_runtime": 24241.393,
+ "train_tokens_per_second": 5678.867
+ },
+ {
+ "epoch": 0.2117559924589281,
+ "grad_norm": 0.9228616942450966,
+ "learning_rate": 9.726114132823638e-06,
+ "loss": 0.389466381072998,
+ "num_input_tokens_seen": 137882432,
+ "step": 3145,
+ "train_runtime": 24280.2706,
+ "train_tokens_per_second": 5678.785
+ },
+ {
+ "epoch": 0.21209264745488823,
+ "grad_norm": 0.6934203083016008,
+ "learning_rate": 9.725250375388493e-06,
+ "loss": 0.4213107109069824,
+ "num_input_tokens_seen": 138093344,
+ "step": 3150,
+ "train_runtime": 24320.3928,
+ "train_tokens_per_second": 5678.089
+ },
+ {
+ "epoch": 0.21242930245084837,
+ "grad_norm": 0.8669084820101542,
+ "learning_rate": 9.724385296552147e-06,
+ "loss": 0.43596372604370115,
+ "num_input_tokens_seen": 138316336,
+ "step": 3155,
+ "train_runtime": 24362.0648,
+ "train_tokens_per_second": 5677.529
+ },
+ {
+ "epoch": 0.2127659574468085,
+ "grad_norm": 0.9337337457982863,
+ "learning_rate": 9.72351889655652e-06,
+ "loss": 0.4109373092651367,
+ "num_input_tokens_seen": 138541888,
+ "step": 3160,
+ "train_runtime": 24397.2747,
+ "train_tokens_per_second": 5678.58
+ },
+ {
+ "epoch": 0.21310261244276865,
+ "grad_norm": 0.8342554979170712,
+ "learning_rate": 9.722651175643895e-06,
+ "loss": 0.40399460792541503,
+ "num_input_tokens_seen": 138760928,
+ "step": 3165,
+ "train_runtime": 24436.674,
+ "train_tokens_per_second": 5678.388
+ },
+ {
+ "epoch": 0.2134392674387288,
+ "grad_norm": 0.7886653586812639,
+ "learning_rate": 9.721782134056928e-06,
+ "loss": 0.36252574920654296,
+ "num_input_tokens_seen": 138982088,
+ "step": 3170,
+ "train_runtime": 24480.3769,
+ "train_tokens_per_second": 5677.285
+ },
+ {
+ "epoch": 0.21377592243468893,
+ "grad_norm": 0.7172325919553721,
+ "learning_rate": 9.720911772038645e-06,
+ "loss": 0.4204305648803711,
+ "num_input_tokens_seen": 139208968,
+ "step": 3175,
+ "train_runtime": 24514.7644,
+ "train_tokens_per_second": 5678.577
+ },
+ {
+ "epoch": 0.21411257743064907,
+ "grad_norm": 0.8184330706813875,
+ "learning_rate": 9.720040089832436e-06,
+ "loss": 0.4118339538574219,
+ "num_input_tokens_seen": 139404296,
+ "step": 3180,
+ "train_runtime": 24555.2547,
+ "train_tokens_per_second": 5677.168
+ },
+ {
+ "epoch": 0.2144492324266092,
+ "grad_norm": 0.7810238480356294,
+ "learning_rate": 9.719167087682069e-06,
+ "loss": 0.38251235485076907,
+ "num_input_tokens_seen": 139626528,
+ "step": 3185,
+ "train_runtime": 24596.095,
+ "train_tokens_per_second": 5676.776
+ },
+ {
+ "epoch": 0.21478588742256935,
+ "grad_norm": 0.8237949114328076,
+ "learning_rate": 9.718292765831673e-06,
+ "loss": 0.39778890609741213,
+ "num_input_tokens_seen": 139852040,
+ "step": 3190,
+ "train_runtime": 24636.5885,
+ "train_tokens_per_second": 5676.599
+ },
+ {
+ "epoch": 0.2151225424185295,
+ "grad_norm": 0.823181949661376,
+ "learning_rate": 9.71741712452575e-06,
+ "loss": 0.4149151802062988,
+ "num_input_tokens_seen": 140074344,
+ "step": 3195,
+ "train_runtime": 24669.1536,
+ "train_tokens_per_second": 5678.117
+ },
+ {
+ "epoch": 0.21545919741448963,
+ "grad_norm": 0.7987011374780209,
+ "learning_rate": 9.716540164009174e-06,
+ "loss": 0.3865767478942871,
+ "num_input_tokens_seen": 140277680,
+ "step": 3200,
+ "train_runtime": 24707.2868,
+ "train_tokens_per_second": 5677.583
+ },
+ {
+ "epoch": 0.21579585241044977,
+ "grad_norm": 0.9131858261391301,
+ "learning_rate": 9.715661884527178e-06,
+ "loss": 0.4173252582550049,
+ "num_input_tokens_seen": 140503936,
+ "step": 3205,
+ "train_runtime": 24750.0835,
+ "train_tokens_per_second": 5676.908
+ },
+ {
+ "epoch": 0.2161325074064099,
+ "grad_norm": 0.9283199650974979,
+ "learning_rate": 9.714782286325374e-06,
+ "loss": 0.4102613925933838,
+ "num_input_tokens_seen": 140723000,
+ "step": 3210,
+ "train_runtime": 24793.1175,
+ "train_tokens_per_second": 5675.89
+ },
+ {
+ "epoch": 0.21646916240237005,
+ "grad_norm": 0.9449783149453059,
+ "learning_rate": 9.713901369649738e-06,
+ "loss": 0.3924571990966797,
+ "num_input_tokens_seen": 140943200,
+ "step": 3215,
+ "train_runtime": 24831.5013,
+ "train_tokens_per_second": 5675.984
+ },
+ {
+ "epoch": 0.2168058173983302,
+ "grad_norm": 0.7571957619819966,
+ "learning_rate": 9.713019134746618e-06,
+ "loss": 0.38917341232299807,
+ "num_input_tokens_seen": 141178808,
+ "step": 3220,
+ "train_runtime": 24867.52,
+ "train_tokens_per_second": 5677.237
+ },
+ {
+ "epoch": 0.21714247239429033,
+ "grad_norm": 0.8569813128796843,
+ "learning_rate": 9.712135581862722e-06,
+ "loss": 0.39778847694396974,
+ "num_input_tokens_seen": 141410600,
+ "step": 3225,
+ "train_runtime": 24909.1487,
+ "train_tokens_per_second": 5677.055
+ },
+ {
+ "epoch": 0.21747912739025047,
+ "grad_norm": 0.7957771696921552,
+ "learning_rate": 9.711250711245139e-06,
+ "loss": 0.41150903701782227,
+ "num_input_tokens_seen": 141622744,
+ "step": 3230,
+ "train_runtime": 24947.3479,
+ "train_tokens_per_second": 5676.866
+ },
+ {
+ "epoch": 0.2178157823862106,
+ "grad_norm": 0.7674615378265653,
+ "learning_rate": 9.710364523141317e-06,
+ "loss": 0.4173532009124756,
+ "num_input_tokens_seen": 141845232,
+ "step": 3235,
+ "train_runtime": 24990.9712,
+ "train_tokens_per_second": 5675.859
+ },
+ {
+ "epoch": 0.21815243738217074,
+ "grad_norm": 0.881520768069259,
+ "learning_rate": 9.709477017799076e-06,
+ "loss": 0.4302227973937988,
+ "num_input_tokens_seen": 142068920,
+ "step": 3240,
+ "train_runtime": 25022.7411,
+ "train_tokens_per_second": 5677.592
+ },
+ {
+ "epoch": 0.21848909237813088,
+ "grad_norm": 0.9412217646326109,
+ "learning_rate": 9.708588195466604e-06,
+ "loss": 0.4126433849334717,
+ "num_input_tokens_seen": 142289600,
+ "step": 3245,
+ "train_runtime": 25054.9859,
+ "train_tokens_per_second": 5679.093
+ },
+ {
+ "epoch": 0.21882574737409102,
+ "grad_norm": 0.7430440131191046,
+ "learning_rate": 9.707698056392458e-06,
+ "loss": 0.4167231559753418,
+ "num_input_tokens_seen": 142506920,
+ "step": 3250,
+ "train_runtime": 25090.3796,
+ "train_tokens_per_second": 5679.743
+ },
+ {
+ "epoch": 0.21916240237005116,
+ "grad_norm": 0.8971811831202605,
+ "learning_rate": 9.706806600825562e-06,
+ "loss": 0.4131289005279541,
+ "num_input_tokens_seen": 142719992,
+ "step": 3255,
+ "train_runtime": 25133.8043,
+ "train_tokens_per_second": 5678.408
+ },
+ {
+ "epoch": 0.2194990573660113,
+ "grad_norm": 0.9964244321532095,
+ "learning_rate": 9.70591382901521e-06,
+ "loss": 0.42434329986572267,
+ "num_input_tokens_seen": 142935376,
+ "step": 3260,
+ "train_runtime": 25170.1091,
+ "train_tokens_per_second": 5678.775
+ },
+ {
+ "epoch": 0.21983571236197144,
+ "grad_norm": 0.7540863225432197,
+ "learning_rate": 9.705019741211061e-06,
+ "loss": 0.39461374282836914,
+ "num_input_tokens_seen": 143155712,
+ "step": 3265,
+ "train_runtime": 25208.8229,
+ "train_tokens_per_second": 5678.794
+ },
+ {
+ "epoch": 0.22017236735793158,
+ "grad_norm": 0.8596938023198728,
+ "learning_rate": 9.704124337663145e-06,
+ "loss": 0.38963048458099364,
+ "num_input_tokens_seen": 143381832,
+ "step": 3270,
+ "train_runtime": 25243.421,
+ "train_tokens_per_second": 5679.968
+ },
+ {
+ "epoch": 0.22050902235389172,
+ "grad_norm": 0.8540309395774371,
+ "learning_rate": 9.703227618621859e-06,
+ "loss": 0.4044978618621826,
+ "num_input_tokens_seen": 143607328,
+ "step": 3275,
+ "train_runtime": 25284.0238,
+ "train_tokens_per_second": 5679.766
+ },
+ {
+ "epoch": 0.22084567734985186,
+ "grad_norm": 0.8072539756195105,
+ "learning_rate": 9.702329584337965e-06,
+ "loss": 0.41426868438720704,
+ "num_input_tokens_seen": 143809000,
+ "step": 3280,
+ "train_runtime": 25326.637,
+ "train_tokens_per_second": 5678.172
+ },
+ {
+ "epoch": 0.221182332345812,
+ "grad_norm": 0.9295650810498607,
+ "learning_rate": 9.701430235062599e-06,
+ "loss": 0.43207788467407227,
+ "num_input_tokens_seen": 144023696,
+ "step": 3285,
+ "train_runtime": 25366.0897,
+ "train_tokens_per_second": 5677.804
+ },
+ {
+ "epoch": 0.22151898734177214,
+ "grad_norm": 0.9183012776415354,
+ "learning_rate": 9.700529571047259e-06,
+ "loss": 0.4370114803314209,
+ "num_input_tokens_seen": 144236616,
+ "step": 3290,
+ "train_runtime": 25403.6973,
+ "train_tokens_per_second": 5677.78
+ },
+ {
+ "epoch": 0.22185564233773228,
+ "grad_norm": 0.8676275172270271,
+ "learning_rate": 9.699627592543814e-06,
+ "loss": 0.3935758829116821,
+ "num_input_tokens_seen": 144451272,
+ "step": 3295,
+ "train_runtime": 25449.4742,
+ "train_tokens_per_second": 5676.002
+ },
+ {
+ "epoch": 0.22219229733369245,
+ "grad_norm": 0.7791903636055438,
+ "learning_rate": 9.698724299804498e-06,
+ "loss": 0.41451497077941896,
+ "num_input_tokens_seen": 144675240,
+ "step": 3300,
+ "train_runtime": 25489.8713,
+ "train_tokens_per_second": 5675.793
+ },
+ {
+ "epoch": 0.22252895232965259,
+ "grad_norm": 0.8026534486815711,
+ "learning_rate": 9.697819693081915e-06,
+ "loss": 0.40555124282836913,
+ "num_input_tokens_seen": 144894992,
+ "step": 3305,
+ "train_runtime": 25526.7574,
+ "train_tokens_per_second": 5676.2
+ },
+ {
+ "epoch": 0.22286560732561272,
+ "grad_norm": 0.7189128013140877,
+ "learning_rate": 9.696913772629034e-06,
+ "loss": 0.38943185806274416,
+ "num_input_tokens_seen": 145119152,
+ "step": 3310,
+ "train_runtime": 25563.6911,
+ "train_tokens_per_second": 5676.768
+ },
+ {
+ "epoch": 0.22320226232157286,
+ "grad_norm": 0.788476880680751,
+ "learning_rate": 9.696006538699195e-06,
+ "loss": 0.42933268547058107,
+ "num_input_tokens_seen": 145340576,
+ "step": 3315,
+ "train_runtime": 25595.5361,
+ "train_tokens_per_second": 5678.356
+ },
+ {
+ "epoch": 0.223538917317533,
+ "grad_norm": 0.9173858981265539,
+ "learning_rate": 9.695097991546102e-06,
+ "loss": 0.41726884841918943,
+ "num_input_tokens_seen": 145536392,
+ "step": 3320,
+ "train_runtime": 25635.2268,
+ "train_tokens_per_second": 5677.203
+ },
+ {
+ "epoch": 0.22387557231349314,
+ "grad_norm": 0.85644319812075,
+ "learning_rate": 9.694188131423828e-06,
+ "loss": 0.38661761283874513,
+ "num_input_tokens_seen": 145752592,
+ "step": 3325,
+ "train_runtime": 25675.7087,
+ "train_tokens_per_second": 5676.673
+ },
+ {
+ "epoch": 0.22421222730945328,
+ "grad_norm": 0.8283061394820891,
+ "learning_rate": 9.693276958586809e-06,
+ "loss": 0.4395936965942383,
+ "num_input_tokens_seen": 145975632,
+ "step": 3330,
+ "train_runtime": 25717.6903,
+ "train_tokens_per_second": 5676.079
+ },
+ {
+ "epoch": 0.22454888230541342,
+ "grad_norm": 1.107175434088717,
+ "learning_rate": 9.692364473289856e-06,
+ "loss": 0.37718255519866944,
+ "num_input_tokens_seen": 146190016,
+ "step": 3335,
+ "train_runtime": 25758.607,
+ "train_tokens_per_second": 5675.385
+ },
+ {
+ "epoch": 0.22488553730137356,
+ "grad_norm": 0.8144408251561523,
+ "learning_rate": 9.69145067578814e-06,
+ "loss": 0.38504073619842527,
+ "num_input_tokens_seen": 146415880,
+ "step": 3340,
+ "train_runtime": 25799.3872,
+ "train_tokens_per_second": 5675.169
+ },
+ {
+ "epoch": 0.2252221922973337,
+ "grad_norm": 0.7888334595920984,
+ "learning_rate": 9.690535566337205e-06,
+ "loss": 0.44390153884887695,
+ "num_input_tokens_seen": 146637056,
+ "step": 3345,
+ "train_runtime": 25841.4558,
+ "train_tokens_per_second": 5674.489
+ },
+ {
+ "epoch": 0.22555884729329384,
+ "grad_norm": 0.8121918469969017,
+ "learning_rate": 9.689619145192955e-06,
+ "loss": 0.4186522960662842,
+ "num_input_tokens_seen": 146858960,
+ "step": 3350,
+ "train_runtime": 25884.465,
+ "train_tokens_per_second": 5673.633
+ },
+ {
+ "epoch": 0.22589550228925398,
+ "grad_norm": 0.7357528337250359,
+ "learning_rate": 9.688701412611663e-06,
+ "loss": 0.39981067180633545,
+ "num_input_tokens_seen": 147088760,
+ "step": 3355,
+ "train_runtime": 25924.1782,
+ "train_tokens_per_second": 5673.806
+ },
+ {
+ "epoch": 0.22623215728521412,
+ "grad_norm": 0.8039258104050244,
+ "learning_rate": 9.687782368849973e-06,
+ "loss": 0.40915794372558595,
+ "num_input_tokens_seen": 147306920,
+ "step": 3360,
+ "train_runtime": 25967.0787,
+ "train_tokens_per_second": 5672.834
+ },
+ {
+ "epoch": 0.22656881228117426,
+ "grad_norm": 0.8366828009872761,
+ "learning_rate": 9.686862014164893e-06,
+ "loss": 0.3879998207092285,
+ "num_input_tokens_seen": 147534144,
+ "step": 3365,
+ "train_runtime": 26004.3358,
+ "train_tokens_per_second": 5673.444
+ },
+ {
+ "epoch": 0.2269054672771344,
+ "grad_norm": 0.8583728877321222,
+ "learning_rate": 9.685940348813795e-06,
+ "loss": 0.3802268266677856,
+ "num_input_tokens_seen": 147751544,
+ "step": 3370,
+ "train_runtime": 26042.4838,
+ "train_tokens_per_second": 5673.481
+ },
+ {
+ "epoch": 0.22724212227309454,
+ "grad_norm": 0.8211994293131717,
+ "learning_rate": 9.685017373054421e-06,
+ "loss": 0.39569883346557616,
+ "num_input_tokens_seen": 147971208,
+ "step": 3375,
+ "train_runtime": 26081.1089,
+ "train_tokens_per_second": 5673.501
+ },
+ {
+ "epoch": 0.22757877726905468,
+ "grad_norm": 0.7908688808511072,
+ "learning_rate": 9.68409308714488e-06,
+ "loss": 0.3963681697845459,
+ "num_input_tokens_seen": 148186648,
+ "step": 3380,
+ "train_runtime": 26119.5816,
+ "train_tokens_per_second": 5673.393
+ },
+ {
+ "epoch": 0.22791543226501482,
+ "grad_norm": 0.7696519906828798,
+ "learning_rate": 9.68316749134364e-06,
+ "loss": 0.4038090705871582,
+ "num_input_tokens_seen": 148417752,
+ "step": 3385,
+ "train_runtime": 26162.0325,
+ "train_tokens_per_second": 5673.021
+ },
+ {
+ "epoch": 0.22825208726097496,
+ "grad_norm": 0.7784468571247383,
+ "learning_rate": 9.682240585909545e-06,
+ "loss": 0.38401403427124026,
+ "num_input_tokens_seen": 148628520,
+ "step": 3390,
+ "train_runtime": 26201.9074,
+ "train_tokens_per_second": 5672.431
+ },
+ {
+ "epoch": 0.2285887422569351,
+ "grad_norm": 0.7682945918583601,
+ "learning_rate": 9.681312371101803e-06,
+ "loss": 0.36235072612762453,
+ "num_input_tokens_seen": 148860120,
+ "step": 3395,
+ "train_runtime": 26235.4949,
+ "train_tokens_per_second": 5673.997
+ },
+ {
+ "epoch": 0.22892539725289524,
+ "grad_norm": 0.7562252932302131,
+ "learning_rate": 9.680382847179984e-06,
+ "loss": 0.4153308868408203,
+ "num_input_tokens_seen": 149075360,
+ "step": 3400,
+ "train_runtime": 26279.0336,
+ "train_tokens_per_second": 5672.787
+ },
+ {
+ "epoch": 0.22926205224885537,
+ "grad_norm": 0.8026900227545827,
+ "learning_rate": 9.679452014404025e-06,
+ "loss": 0.39875361919403074,
+ "num_input_tokens_seen": 149283904,
+ "step": 3405,
+ "train_runtime": 26318.8542,
+ "train_tokens_per_second": 5672.128
+ },
+ {
+ "epoch": 0.22959870724481551,
+ "grad_norm": 0.8571181299293755,
+ "learning_rate": 9.678519873034234e-06,
+ "loss": 0.435106086730957,
+ "num_input_tokens_seen": 149511344,
+ "step": 3410,
+ "train_runtime": 26357.0843,
+ "train_tokens_per_second": 5672.53
+ },
+ {
+ "epoch": 0.22993536224077565,
+ "grad_norm": 0.9132428672371381,
+ "learning_rate": 9.677586423331277e-06,
+ "loss": 0.42625818252563474,
+ "num_input_tokens_seen": 149734672,
+ "step": 3415,
+ "train_runtime": 26396.1327,
+ "train_tokens_per_second": 5672.599
+ },
+ {
+ "epoch": 0.2302720172367358,
+ "grad_norm": 0.7227848772197648,
+ "learning_rate": 9.676651665556194e-06,
+ "loss": 0.3853285312652588,
+ "num_input_tokens_seen": 149966984,
+ "step": 3420,
+ "train_runtime": 26437.1674,
+ "train_tokens_per_second": 5672.581
+ },
+ {
+ "epoch": 0.23060867223269593,
+ "grad_norm": 0.9758464763691787,
+ "learning_rate": 9.675715599970383e-06,
+ "loss": 0.39388408660888674,
+ "num_input_tokens_seen": 150173312,
+ "step": 3425,
+ "train_runtime": 26476.8347,
+ "train_tokens_per_second": 5671.876
+ },
+ {
+ "epoch": 0.23094532722865607,
+ "grad_norm": 0.8789640499008409,
+ "learning_rate": 9.674778226835617e-06,
+ "loss": 0.43226184844970705,
+ "num_input_tokens_seen": 150378728,
+ "step": 3430,
+ "train_runtime": 26514.5982,
+ "train_tokens_per_second": 5671.545
+ },
+ {
+ "epoch": 0.2312819822246162,
+ "grad_norm": 0.8223781330799458,
+ "learning_rate": 9.673839546414026e-06,
+ "loss": 0.4151242733001709,
+ "num_input_tokens_seen": 150589928,
+ "step": 3435,
+ "train_runtime": 26555.5388,
+ "train_tokens_per_second": 5670.754
+ },
+ {
+ "epoch": 0.23161863722057635,
+ "grad_norm": 0.6819707841325117,
+ "learning_rate": 9.672899558968107e-06,
+ "loss": 0.39006340503692627,
+ "num_input_tokens_seen": 150802664,
+ "step": 3440,
+ "train_runtime": 26599.7743,
+ "train_tokens_per_second": 5669.321
+ },
+ {
+ "epoch": 0.2319552922165365,
+ "grad_norm": 0.8192017234463131,
+ "learning_rate": 9.67195826476073e-06,
+ "loss": 0.44136791229248046,
+ "num_input_tokens_seen": 151027880,
+ "step": 3445,
+ "train_runtime": 26637.6563,
+ "train_tokens_per_second": 5669.714
+ },
+ {
+ "epoch": 0.23229194721249663,
+ "grad_norm": 0.8137299353827566,
+ "learning_rate": 9.671015664055122e-06,
+ "loss": 0.3967195749282837,
+ "num_input_tokens_seen": 151239928,
+ "step": 3450,
+ "train_runtime": 26671.0692,
+ "train_tokens_per_second": 5670.561
+ },
+ {
+ "epoch": 0.23262860220845677,
+ "grad_norm": 0.819274273484707,
+ "learning_rate": 9.670071757114877e-06,
+ "loss": 0.4030034065246582,
+ "num_input_tokens_seen": 151459832,
+ "step": 3455,
+ "train_runtime": 26710.598,
+ "train_tokens_per_second": 5670.402
+ },
+ {
+ "epoch": 0.2329652572044169,
+ "grad_norm": 0.8052967688256787,
+ "learning_rate": 9.669126544203957e-06,
+ "loss": 0.38636322021484376,
+ "num_input_tokens_seen": 151667896,
+ "step": 3460,
+ "train_runtime": 26749.1234,
+ "train_tokens_per_second": 5670.014
+ },
+ {
+ "epoch": 0.23330191220037705,
+ "grad_norm": 0.8630240510385624,
+ "learning_rate": 9.668180025586689e-06,
+ "loss": 0.4077507972717285,
+ "num_input_tokens_seen": 151902352,
+ "step": 3465,
+ "train_runtime": 26791.3212,
+ "train_tokens_per_second": 5669.834
+ },
+ {
+ "epoch": 0.2336385671963372,
+ "grad_norm": 0.7786921906276983,
+ "learning_rate": 9.667232201527761e-06,
+ "loss": 0.3916687250137329,
+ "num_input_tokens_seen": 152102680,
+ "step": 3470,
+ "train_runtime": 26827.5433,
+ "train_tokens_per_second": 5669.646
+ },
+ {
+ "epoch": 0.23397522219229733,
+ "grad_norm": 0.8593939179148706,
+ "learning_rate": 9.666283072292234e-06,
+ "loss": 0.41549954414367674,
+ "num_input_tokens_seen": 152303592,
+ "step": 3475,
+ "train_runtime": 26864.7416,
+ "train_tokens_per_second": 5669.274
+ },
+ {
+ "epoch": 0.23431187718825747,
+ "grad_norm": 0.7173015967840941,
+ "learning_rate": 9.665332638145522e-06,
+ "loss": 0.40798091888427734,
+ "num_input_tokens_seen": 152526320,
+ "step": 3480,
+ "train_runtime": 26904.6608,
+ "train_tokens_per_second": 5669.141
+ },
+ {
+ "epoch": 0.2346485321842176,
+ "grad_norm": 0.7787780767474121,
+ "learning_rate": 9.664380899353414e-06,
+ "loss": 0.44217610359191895,
+ "num_input_tokens_seen": 152747800,
+ "step": 3485,
+ "train_runtime": 26938.7633,
+ "train_tokens_per_second": 5670.186
+ },
+ {
+ "epoch": 0.23498518718017775,
+ "grad_norm": 0.875181117161752,
+ "learning_rate": 9.663427856182062e-06,
+ "loss": 0.418949556350708,
+ "num_input_tokens_seen": 152978904,
+ "step": 3490,
+ "train_runtime": 26976.7683,
+ "train_tokens_per_second": 5670.765
+ },
+ {
+ "epoch": 0.23532184217613789,
+ "grad_norm": 0.8234151432189463,
+ "learning_rate": 9.662473508897981e-06,
+ "loss": 0.41005191802978513,
+ "num_input_tokens_seen": 153200600,
+ "step": 3495,
+ "train_runtime": 27021.1553,
+ "train_tokens_per_second": 5669.654
+ },
+ {
+ "epoch": 0.23565849717209802,
+ "grad_norm": 0.9381121440419455,
+ "learning_rate": 9.661517857768052e-06,
+ "loss": 0.43746633529663087,
+ "num_input_tokens_seen": 153413320,
+ "step": 3500,
+ "train_runtime": 27058.2447,
+ "train_tokens_per_second": 5669.744
+ },
+ {
+ "epoch": 0.23599515216805816,
+ "grad_norm": 0.9159040524887876,
+ "learning_rate": 9.660560903059517e-06,
+ "loss": 0.4074262619018555,
+ "num_input_tokens_seen": 153619928,
+ "step": 3505,
+ "train_runtime": 27099.4693,
+ "train_tokens_per_second": 5668.743
+ },
+ {
+ "epoch": 0.2363318071640183,
+ "grad_norm": 0.7088057637221193,
+ "learning_rate": 9.659602645039988e-06,
+ "loss": 0.4169961929321289,
+ "num_input_tokens_seen": 153844808,
+ "step": 3510,
+ "train_runtime": 27137.4232,
+ "train_tokens_per_second": 5669.102
+ },
+ {
+ "epoch": 0.23666846215997844,
+ "grad_norm": 0.7425326417567343,
+ "learning_rate": 9.658643083977435e-06,
+ "loss": 0.37222681045532224,
+ "num_input_tokens_seen": 154057008,
+ "step": 3515,
+ "train_runtime": 27172.217,
+ "train_tokens_per_second": 5669.652
+ },
+ {
+ "epoch": 0.23700511715593858,
+ "grad_norm": 0.7835915209475461,
+ "learning_rate": 9.657682220140202e-06,
+ "loss": 0.3870618104934692,
+ "num_input_tokens_seen": 154257952,
+ "step": 3520,
+ "train_runtime": 27216.0159,
+ "train_tokens_per_second": 5667.911
+ },
+ {
+ "epoch": 0.23734177215189872,
+ "grad_norm": 0.944898207115128,
+ "learning_rate": 9.656720053796987e-06,
+ "loss": 0.4269230365753174,
+ "num_input_tokens_seen": 154468968,
+ "step": 3525,
+ "train_runtime": 27253.774,
+ "train_tokens_per_second": 5667.801
+ },
+ {
+ "epoch": 0.23767842714785886,
+ "grad_norm": 0.964924059220392,
+ "learning_rate": 9.655756585216859e-06,
+ "loss": 0.39426770210266116,
+ "num_input_tokens_seen": 154679920,
+ "step": 3530,
+ "train_runtime": 27297.8899,
+ "train_tokens_per_second": 5666.369
+ },
+ {
+ "epoch": 0.238015082143819,
+ "grad_norm": 0.8389859916982031,
+ "learning_rate": 9.654791814669246e-06,
+ "loss": 0.41727919578552247,
+ "num_input_tokens_seen": 154894088,
+ "step": 3535,
+ "train_runtime": 27338.9009,
+ "train_tokens_per_second": 5665.703
+ },
+ {
+ "epoch": 0.23835173713977914,
+ "grad_norm": 0.7799448759874524,
+ "learning_rate": 9.653825742423949e-06,
+ "loss": 0.36776251792907716,
+ "num_input_tokens_seen": 155118416,
+ "step": 3540,
+ "train_runtime": 27380.1523,
+ "train_tokens_per_second": 5665.36
+ },
+ {
+ "epoch": 0.2386883921357393,
+ "grad_norm": 0.7611172988563449,
+ "learning_rate": 9.652858368751118e-06,
+ "loss": 0.38267481327056885,
+ "num_input_tokens_seen": 155325944,
+ "step": 3545,
+ "train_runtime": 27419.2328,
+ "train_tokens_per_second": 5664.854
+ },
+ {
+ "epoch": 0.23902504713169945,
+ "grad_norm": 0.849616732171608,
+ "learning_rate": 9.651889693921284e-06,
+ "loss": 0.39499473571777344,
+ "num_input_tokens_seen": 155553936,
+ "step": 3550,
+ "train_runtime": 27453.0014,
+ "train_tokens_per_second": 5666.191
+ },
+ {
+ "epoch": 0.2393617021276596,
+ "grad_norm": 0.8290842696777342,
+ "learning_rate": 9.650919718205331e-06,
+ "loss": 0.46364564895629884,
+ "num_input_tokens_seen": 155767440,
+ "step": 3555,
+ "train_runtime": 27489.7425,
+ "train_tokens_per_second": 5666.384
+ },
+ {
+ "epoch": 0.23969835712361973,
+ "grad_norm": 0.6972330732061462,
+ "learning_rate": 9.649948441874509e-06,
+ "loss": 0.41176652908325195,
+ "num_input_tokens_seen": 155991432,
+ "step": 3560,
+ "train_runtime": 27533.7281,
+ "train_tokens_per_second": 5665.467
+ },
+ {
+ "epoch": 0.24003501211957987,
+ "grad_norm": 0.8232987599322095,
+ "learning_rate": 9.648975865200434e-06,
+ "loss": 0.3719303607940674,
+ "num_input_tokens_seen": 156200840,
+ "step": 3565,
+ "train_runtime": 27572.2717,
+ "train_tokens_per_second": 5665.142
+ },
+ {
+ "epoch": 0.24037166711554,
+ "grad_norm": 0.7784058710951964,
+ "learning_rate": 9.648001988455082e-06,
+ "loss": 0.42696056365966795,
+ "num_input_tokens_seen": 156423232,
+ "step": 3570,
+ "train_runtime": 27611.0095,
+ "train_tokens_per_second": 5665.249
+ },
+ {
+ "epoch": 0.24070832211150014,
+ "grad_norm": 0.8101799377558996,
+ "learning_rate": 9.647026811910794e-06,
+ "loss": 0.42109375,
+ "num_input_tokens_seen": 156638144,
+ "step": 3575,
+ "train_runtime": 27652.042,
+ "train_tokens_per_second": 5664.614
+ },
+ {
+ "epoch": 0.24104497710746028,
+ "grad_norm": 0.7776435744711748,
+ "learning_rate": 9.646050335840279e-06,
+ "loss": 0.440277624130249,
+ "num_input_tokens_seen": 156867440,
+ "step": 3580,
+ "train_runtime": 27693.6123,
+ "train_tokens_per_second": 5664.391
+ },
+ {
+ "epoch": 0.24138163210342042,
+ "grad_norm": 0.8251722449916711,
+ "learning_rate": 9.6450725605166e-06,
+ "loss": 0.4013186454772949,
+ "num_input_tokens_seen": 157079728,
+ "step": 3585,
+ "train_runtime": 27726.0789,
+ "train_tokens_per_second": 5665.414
+ },
+ {
+ "epoch": 0.24171828709938056,
+ "grad_norm": 0.8433224749813183,
+ "learning_rate": 9.644093486213192e-06,
+ "loss": 0.4329217910766602,
+ "num_input_tokens_seen": 157312856,
+ "step": 3590,
+ "train_runtime": 27767.0627,
+ "train_tokens_per_second": 5665.448
+ },
+ {
+ "epoch": 0.2420549420953407,
+ "grad_norm": 0.8662917363975451,
+ "learning_rate": 9.64311311320385e-06,
+ "loss": 0.4125063896179199,
+ "num_input_tokens_seen": 157536824,
+ "step": 3595,
+ "train_runtime": 27801.6779,
+ "train_tokens_per_second": 5666.45
+ },
+ {
+ "epoch": 0.24239159709130084,
+ "grad_norm": 0.8971134461583352,
+ "learning_rate": 9.642131441762732e-06,
+ "loss": 0.39645483493804934,
+ "num_input_tokens_seen": 157744672,
+ "step": 3600,
+ "train_runtime": 27838.8811,
+ "train_tokens_per_second": 5666.344
+ },
+ {
+ "epoch": 0.24272825208726098,
+ "grad_norm": 0.8431994851395936,
+ "learning_rate": 9.641148472164359e-06,
+ "loss": 0.3727452516555786,
+ "num_input_tokens_seen": 157957176,
+ "step": 3605,
+ "train_runtime": 27875.6937,
+ "train_tokens_per_second": 5666.484
+ },
+ {
+ "epoch": 0.24306490708322112,
+ "grad_norm": 0.8088195809393041,
+ "learning_rate": 9.640164204683616e-06,
+ "loss": 0.39656991958618165,
+ "num_input_tokens_seen": 158185920,
+ "step": 3610,
+ "train_runtime": 27916.062,
+ "train_tokens_per_second": 5666.484
+ },
+ {
+ "epoch": 0.24340156207918126,
+ "grad_norm": 0.8326446982063453,
+ "learning_rate": 9.639178639595748e-06,
+ "loss": 0.43434457778930663,
+ "num_input_tokens_seen": 158410048,
+ "step": 3615,
+ "train_runtime": 27944.4625,
+ "train_tokens_per_second": 5668.746
+ },
+ {
+ "epoch": 0.2437382170751414,
+ "grad_norm": 0.7829852555366933,
+ "learning_rate": 9.638191777176367e-06,
+ "loss": 0.40036301612854003,
+ "num_input_tokens_seen": 158632248,
+ "step": 3620,
+ "train_runtime": 27978.7949,
+ "train_tokens_per_second": 5669.731
+ },
+ {
+ "epoch": 0.24407487207110154,
+ "grad_norm": 0.7554602216351339,
+ "learning_rate": 9.637203617701446e-06,
+ "loss": 0.41867575645446775,
+ "num_input_tokens_seen": 158861832,
+ "step": 3625,
+ "train_runtime": 28021.0376,
+ "train_tokens_per_second": 5669.377
+ },
+ {
+ "epoch": 0.24441152706706168,
+ "grad_norm": 0.7700724836949714,
+ "learning_rate": 9.63621416144732e-06,
+ "loss": 0.3974162578582764,
+ "num_input_tokens_seen": 159076104,
+ "step": 3630,
+ "train_runtime": 28057.1542,
+ "train_tokens_per_second": 5669.716
+ },
+ {
+ "epoch": 0.24474818206302182,
+ "grad_norm": 0.819901947145382,
+ "learning_rate": 9.635223408690688e-06,
+ "loss": 0.39289240837097167,
+ "num_input_tokens_seen": 159275688,
+ "step": 3635,
+ "train_runtime": 28092.5481,
+ "train_tokens_per_second": 5669.678
+ },
+ {
+ "epoch": 0.24508483705898196,
+ "grad_norm": 0.74792422760683,
+ "learning_rate": 9.63423135970861e-06,
+ "loss": 0.4178638935089111,
+ "num_input_tokens_seen": 159506232,
+ "step": 3640,
+ "train_runtime": 28131.9808,
+ "train_tokens_per_second": 5669.925
+ },
+ {
+ "epoch": 0.2454214920549421,
+ "grad_norm": 0.8690926777352274,
+ "learning_rate": 9.633238014778511e-06,
+ "loss": 0.4228395462036133,
+ "num_input_tokens_seen": 159735712,
+ "step": 3645,
+ "train_runtime": 28164.6263,
+ "train_tokens_per_second": 5671.501
+ },
+ {
+ "epoch": 0.24575814705090224,
+ "grad_norm": 0.8408876506004788,
+ "learning_rate": 9.632243374178173e-06,
+ "loss": 0.3977930784225464,
+ "num_input_tokens_seen": 159959000,
+ "step": 3650,
+ "train_runtime": 28206.3804,
+ "train_tokens_per_second": 5671.022
+ },
+ {
+ "epoch": 0.24609480204686238,
+ "grad_norm": 0.8178577594154023,
+ "learning_rate": 9.63124743818575e-06,
+ "loss": 0.42597079277038574,
+ "num_input_tokens_seen": 160183304,
+ "step": 3655,
+ "train_runtime": 28242.8523,
+ "train_tokens_per_second": 5671.64
+ },
+ {
+ "epoch": 0.24643145704282252,
+ "grad_norm": 0.7174819842350753,
+ "learning_rate": 9.630250207079745e-06,
+ "loss": 0.385437536239624,
+ "num_input_tokens_seen": 160396208,
+ "step": 3660,
+ "train_runtime": 28279.1927,
+ "train_tokens_per_second": 5671.881
+ },
+ {
+ "epoch": 0.24676811203878266,
+ "grad_norm": 0.9558291181876568,
+ "learning_rate": 9.629251681139038e-06,
+ "loss": 0.41692662239074707,
+ "num_input_tokens_seen": 160619648,
+ "step": 3665,
+ "train_runtime": 28315.8408,
+ "train_tokens_per_second": 5672.431
+ },
+ {
+ "epoch": 0.2471047670347428,
+ "grad_norm": 0.7393493036785912,
+ "learning_rate": 9.628251860642857e-06,
+ "loss": 0.40501389503479,
+ "num_input_tokens_seen": 160832936,
+ "step": 3670,
+ "train_runtime": 28358.3536,
+ "train_tokens_per_second": 5671.448
+ },
+ {
+ "epoch": 0.24744142203070293,
+ "grad_norm": 1.3913304565406266,
+ "learning_rate": 9.627250745870805e-06,
+ "loss": 0.4110020637512207,
+ "num_input_tokens_seen": 161045560,
+ "step": 3675,
+ "train_runtime": 28395.0175,
+ "train_tokens_per_second": 5671.613
+ },
+ {
+ "epoch": 0.24777807702666307,
+ "grad_norm": 0.7989443375204172,
+ "learning_rate": 9.626248337102834e-06,
+ "loss": 0.4464590072631836,
+ "num_input_tokens_seen": 161267264,
+ "step": 3680,
+ "train_runtime": 28433.9863,
+ "train_tokens_per_second": 5671.638
+ },
+ {
+ "epoch": 0.2481147320226232,
+ "grad_norm": 0.8100743623465356,
+ "learning_rate": 9.625244634619269e-06,
+ "loss": 0.42216014862060547,
+ "num_input_tokens_seen": 161484928,
+ "step": 3685,
+ "train_runtime": 28470.6875,
+ "train_tokens_per_second": 5671.971
+ },
+ {
+ "epoch": 0.24845138701858335,
+ "grad_norm": 0.8140092686493631,
+ "learning_rate": 9.62423963870079e-06,
+ "loss": 0.4016411781311035,
+ "num_input_tokens_seen": 161695760,
+ "step": 3690,
+ "train_runtime": 28513.9311,
+ "train_tokens_per_second": 5670.764
+ },
+ {
+ "epoch": 0.2487880420145435,
+ "grad_norm": 0.7547948701222302,
+ "learning_rate": 9.623233349628441e-06,
+ "loss": 0.36588501930236816,
+ "num_input_tokens_seen": 161906344,
+ "step": 3695,
+ "train_runtime": 28554.4813,
+ "train_tokens_per_second": 5670.085
+ },
+ {
+ "epoch": 0.24912469701050363,
+ "grad_norm": 0.7855444375692547,
+ "learning_rate": 9.62222576768363e-06,
+ "loss": 0.37874155044555663,
+ "num_input_tokens_seen": 162121952,
+ "step": 3700,
+ "train_runtime": 28593.7113,
+ "train_tokens_per_second": 5669.846
+ },
+ {
+ "epoch": 0.24946135200646377,
+ "grad_norm": 0.8013508326751752,
+ "learning_rate": 9.62121689314812e-06,
+ "loss": 0.4508801460266113,
+ "num_input_tokens_seen": 162339592,
+ "step": 3705,
+ "train_runtime": 28637.9282,
+ "train_tokens_per_second": 5668.692
+ },
+ {
+ "epoch": 0.2497980070024239,
+ "grad_norm": 0.8341179639748463,
+ "learning_rate": 9.620206726304045e-06,
+ "loss": 0.4121876239776611,
+ "num_input_tokens_seen": 162551488,
+ "step": 3710,
+ "train_runtime": 28677.3311,
+ "train_tokens_per_second": 5668.292
+ },
+ {
+ "epoch": 0.2501346619983841,
+ "grad_norm": 0.7606955484077755,
+ "learning_rate": 9.61919526743389e-06,
+ "loss": 0.4007285118103027,
+ "num_input_tokens_seen": 162775296,
+ "step": 3715,
+ "train_runtime": 28717.1122,
+ "train_tokens_per_second": 5668.233
+ },
+ {
+ "epoch": 0.2504713169943442,
+ "grad_norm": 0.7586767868766064,
+ "learning_rate": 9.618182516820507e-06,
+ "loss": 0.4169614315032959,
+ "num_input_tokens_seen": 163004008,
+ "step": 3720,
+ "train_runtime": 28752.7905,
+ "train_tokens_per_second": 5669.154
+ },
+ {
+ "epoch": 0.25080797199030436,
+ "grad_norm": 0.7958372407148936,
+ "learning_rate": 9.61716847474711e-06,
+ "loss": 0.38545975685119627,
+ "num_input_tokens_seen": 163216408,
+ "step": 3725,
+ "train_runtime": 28788.1056,
+ "train_tokens_per_second": 5669.578
+ },
+ {
+ "epoch": 0.2511446269862645,
+ "grad_norm": 0.7821165797023588,
+ "learning_rate": 9.616153141497273e-06,
+ "loss": 0.41805257797241213,
+ "num_input_tokens_seen": 163430088,
+ "step": 3730,
+ "train_runtime": 28827.4553,
+ "train_tokens_per_second": 5669.251
+ },
+ {
+ "epoch": 0.25148128198222464,
+ "grad_norm": 0.8340007896342425,
+ "learning_rate": 9.61513651735493e-06,
+ "loss": 0.37319297790527345,
+ "num_input_tokens_seen": 163642328,
+ "step": 3735,
+ "train_runtime": 28865.8432,
+ "train_tokens_per_second": 5669.065
+ },
+ {
+ "epoch": 0.2518179369781848,
+ "grad_norm": 0.8830424913719485,
+ "learning_rate": 9.614118602604376e-06,
+ "loss": 0.3945745944976807,
+ "num_input_tokens_seen": 163838104,
+ "step": 3740,
+ "train_runtime": 28903.991,
+ "train_tokens_per_second": 5668.356
+ },
+ {
+ "epoch": 0.2521545919741449,
+ "grad_norm": 0.8052297510787393,
+ "learning_rate": 9.613099397530267e-06,
+ "loss": 0.3722787380218506,
+ "num_input_tokens_seen": 164064200,
+ "step": 3745,
+ "train_runtime": 28947.9133,
+ "train_tokens_per_second": 5667.566
+ },
+ {
+ "epoch": 0.25249124697010505,
+ "grad_norm": 0.7862742185087508,
+ "learning_rate": 9.612078902417623e-06,
+ "loss": 0.4092113018035889,
+ "num_input_tokens_seen": 164293832,
+ "step": 3750,
+ "train_runtime": 28984.6022,
+ "train_tokens_per_second": 5668.314
+ },
+ {
+ "epoch": 0.2528279019660652,
+ "grad_norm": 0.917777159281807,
+ "learning_rate": 9.611057117551821e-06,
+ "loss": 0.4049520492553711,
+ "num_input_tokens_seen": 164494120,
+ "step": 3755,
+ "train_runtime": 29022.09,
+ "train_tokens_per_second": 5667.894
+ },
+ {
+ "epoch": 0.25316455696202533,
+ "grad_norm": 0.7487164986896401,
+ "learning_rate": 9.610034043218598e-06,
+ "loss": 0.41929073333740235,
+ "num_input_tokens_seen": 164721320,
+ "step": 3760,
+ "train_runtime": 29064.6791,
+ "train_tokens_per_second": 5667.405
+ },
+ {
+ "epoch": 0.2535012119579855,
+ "grad_norm": 0.8821854609765899,
+ "learning_rate": 9.609009679704057e-06,
+ "loss": 0.4089076995849609,
+ "num_input_tokens_seen": 164941328,
+ "step": 3765,
+ "train_runtime": 29100.0265,
+ "train_tokens_per_second": 5668.082
+ },
+ {
+ "epoch": 0.2538378669539456,
+ "grad_norm": 0.8451569510904378,
+ "learning_rate": 9.607984027294654e-06,
+ "loss": 0.4017925262451172,
+ "num_input_tokens_seen": 165148144,
+ "step": 3770,
+ "train_runtime": 29142.825,
+ "train_tokens_per_second": 5666.854
+ },
+ {
+ "epoch": 0.25417452194990575,
+ "grad_norm": 0.7139702054990582,
+ "learning_rate": 9.606957086277213e-06,
+ "loss": 0.4037941455841064,
+ "num_input_tokens_seen": 165372064,
+ "step": 3775,
+ "train_runtime": 29183.2104,
+ "train_tokens_per_second": 5666.685
+ },
+ {
+ "epoch": 0.2545111769458659,
+ "grad_norm": 0.8399965732594856,
+ "learning_rate": 9.605928856938913e-06,
+ "loss": 0.4064283847808838,
+ "num_input_tokens_seen": 165589000,
+ "step": 3780,
+ "train_runtime": 29221.4526,
+ "train_tokens_per_second": 5666.693
+ },
+ {
+ "epoch": 0.25484783194182603,
+ "grad_norm": 0.7840999876703054,
+ "learning_rate": 9.604899339567298e-06,
+ "loss": 0.3694040060043335,
+ "num_input_tokens_seen": 165806496,
+ "step": 3785,
+ "train_runtime": 29262.2531,
+ "train_tokens_per_second": 5666.225
+ },
+ {
+ "epoch": 0.25518448693778617,
+ "grad_norm": 0.8254996088572744,
+ "learning_rate": 9.603868534450263e-06,
+ "loss": 0.370687198638916,
+ "num_input_tokens_seen": 166040152,
+ "step": 3790,
+ "train_runtime": 29297.3995,
+ "train_tokens_per_second": 5667.402
+ },
+ {
+ "epoch": 0.2555211419337463,
+ "grad_norm": 0.8205302355074221,
+ "learning_rate": 9.602836441876074e-06,
+ "loss": 0.3572968006134033,
+ "num_input_tokens_seen": 166264712,
+ "step": 3795,
+ "train_runtime": 29337.7818,
+ "train_tokens_per_second": 5667.256
+ },
+ {
+ "epoch": 0.25585779692970645,
+ "grad_norm": 0.9811797993807015,
+ "learning_rate": 9.601803062133353e-06,
+ "loss": 0.3995164394378662,
+ "num_input_tokens_seen": 166486544,
+ "step": 3800,
+ "train_runtime": 29381.5758,
+ "train_tokens_per_second": 5666.359
+ },
+ {
+ "epoch": 0.2561944519256666,
+ "grad_norm": 0.7702336235429141,
+ "learning_rate": 9.600768395511078e-06,
+ "loss": 0.3709728717803955,
+ "num_input_tokens_seen": 166705432,
+ "step": 3805,
+ "train_runtime": 29417.1853,
+ "train_tokens_per_second": 5666.94
+ },
+ {
+ "epoch": 0.25653110692162673,
+ "grad_norm": 0.8067121262918427,
+ "learning_rate": 9.599732442298592e-06,
+ "loss": 0.39358603954315186,
+ "num_input_tokens_seen": 166928072,
+ "step": 3810,
+ "train_runtime": 29453.7851,
+ "train_tokens_per_second": 5667.457
+ },
+ {
+ "epoch": 0.25686776191758687,
+ "grad_norm": 0.8388091751069785,
+ "learning_rate": 9.598695202785596e-06,
+ "loss": 0.3980567216873169,
+ "num_input_tokens_seen": 167135336,
+ "step": 3815,
+ "train_runtime": 29489.9025,
+ "train_tokens_per_second": 5667.545
+ },
+ {
+ "epoch": 0.257204416913547,
+ "grad_norm": 0.7655332462464155,
+ "learning_rate": 9.59765667726215e-06,
+ "loss": 0.3841102600097656,
+ "num_input_tokens_seen": 167356664,
+ "step": 3820,
+ "train_runtime": 29526.6684,
+ "train_tokens_per_second": 5667.983
+ },
+ {
+ "epoch": 0.25754107190950715,
+ "grad_norm": 0.8418059528589851,
+ "learning_rate": 9.596616866018677e-06,
+ "loss": 0.46039524078369143,
+ "num_input_tokens_seen": 167568600,
+ "step": 3825,
+ "train_runtime": 29568.9722,
+ "train_tokens_per_second": 5667.042
+ },
+ {
+ "epoch": 0.2578777269054673,
+ "grad_norm": 0.87917393714771,
+ "learning_rate": 9.595575769345955e-06,
+ "loss": 0.3880587100982666,
+ "num_input_tokens_seen": 167766736,
+ "step": 3830,
+ "train_runtime": 29608.4677,
+ "train_tokens_per_second": 5666.174
+ },
+ {
+ "epoch": 0.2582143819014274,
+ "grad_norm": 0.7766137760871757,
+ "learning_rate": 9.594533387535122e-06,
+ "loss": 0.41129217147827146,
+ "num_input_tokens_seen": 167995600,
+ "step": 3835,
+ "train_runtime": 29647.1244,
+ "train_tokens_per_second": 5666.506
+ },
+ {
+ "epoch": 0.25855103689738757,
+ "grad_norm": 0.7405296564338093,
+ "learning_rate": 9.593489720877678e-06,
+ "loss": 0.4008028984069824,
+ "num_input_tokens_seen": 168214488,
+ "step": 3840,
+ "train_runtime": 29683.6625,
+ "train_tokens_per_second": 5666.905
+ },
+ {
+ "epoch": 0.2588876918933477,
+ "grad_norm": 0.8634330150776178,
+ "learning_rate": 9.592444769665482e-06,
+ "loss": 0.37677698135375975,
+ "num_input_tokens_seen": 168446672,
+ "step": 3845,
+ "train_runtime": 29726.006,
+ "train_tokens_per_second": 5666.643
+ },
+ {
+ "epoch": 0.25922434688930784,
+ "grad_norm": 0.967218365398896,
+ "learning_rate": 9.59139853419075e-06,
+ "loss": 0.3776779413223267,
+ "num_input_tokens_seen": 168649368,
+ "step": 3850,
+ "train_runtime": 29767.8639,
+ "train_tokens_per_second": 5665.484
+ },
+ {
+ "epoch": 0.259561001885268,
+ "grad_norm": 0.7776571735726057,
+ "learning_rate": 9.590351014746059e-06,
+ "loss": 0.40406084060668945,
+ "num_input_tokens_seen": 168875024,
+ "step": 3855,
+ "train_runtime": 29804.1223,
+ "train_tokens_per_second": 5666.163
+ },
+ {
+ "epoch": 0.2598976568812281,
+ "grad_norm": 1.1756300159580646,
+ "learning_rate": 9.589302211624343e-06,
+ "loss": 0.39687304496765136,
+ "num_input_tokens_seen": 169090984,
+ "step": 3860,
+ "train_runtime": 29840.2711,
+ "train_tokens_per_second": 5666.536
+ },
+ {
+ "epoch": 0.26023431187718826,
+ "grad_norm": 0.819456740850995,
+ "learning_rate": 9.5882521251189e-06,
+ "loss": 0.3968550682067871,
+ "num_input_tokens_seen": 169310120,
+ "step": 3865,
+ "train_runtime": 29877.9357,
+ "train_tokens_per_second": 5666.728
+ },
+ {
+ "epoch": 0.2605709668731484,
+ "grad_norm": 0.8515658291218985,
+ "learning_rate": 9.58720075552338e-06,
+ "loss": 0.42056074142456057,
+ "num_input_tokens_seen": 169524504,
+ "step": 3870,
+ "train_runtime": 29909.434,
+ "train_tokens_per_second": 5667.928
+ },
+ {
+ "epoch": 0.26090762186910854,
+ "grad_norm": 0.8783050270072679,
+ "learning_rate": 9.586148103131797e-06,
+ "loss": 0.40169625282287597,
+ "num_input_tokens_seen": 169733248,
+ "step": 3875,
+ "train_runtime": 29944.6725,
+ "train_tokens_per_second": 5668.229
+ },
+ {
+ "epoch": 0.2612442768650687,
+ "grad_norm": 0.857473518999431,
+ "learning_rate": 9.585094168238518e-06,
+ "loss": 0.4272635459899902,
+ "num_input_tokens_seen": 169943480,
+ "step": 3880,
+ "train_runtime": 29983.9133,
+ "train_tokens_per_second": 5667.822
+ },
+ {
+ "epoch": 0.2615809318610288,
+ "grad_norm": 0.8162084475991911,
+ "learning_rate": 9.58403895113828e-06,
+ "loss": 0.39615912437438966,
+ "num_input_tokens_seen": 170164936,
+ "step": 3885,
+ "train_runtime": 30023.2923,
+ "train_tokens_per_second": 5667.764
+ },
+ {
+ "epoch": 0.26191758685698896,
+ "grad_norm": 0.773095299777449,
+ "learning_rate": 9.582982452126164e-06,
+ "loss": 0.380422306060791,
+ "num_input_tokens_seen": 170380728,
+ "step": 3890,
+ "train_runtime": 30061.2728,
+ "train_tokens_per_second": 5667.782
+ },
+ {
+ "epoch": 0.2622542418529491,
+ "grad_norm": 0.7250445142657057,
+ "learning_rate": 9.58192467149762e-06,
+ "loss": 0.3720268726348877,
+ "num_input_tokens_seen": 170591568,
+ "step": 3895,
+ "train_runtime": 30102.8751,
+ "train_tokens_per_second": 5666.953
+ },
+ {
+ "epoch": 0.26259089684890924,
+ "grad_norm": 0.8455822434092976,
+ "learning_rate": 9.580865609548452e-06,
+ "loss": 0.4109041213989258,
+ "num_input_tokens_seen": 170793416,
+ "step": 3900,
+ "train_runtime": 30139.1143,
+ "train_tokens_per_second": 5666.836
+ },
+ {
+ "epoch": 0.2629275518448694,
+ "grad_norm": 0.669931714570705,
+ "learning_rate": 9.579805266574824e-06,
+ "loss": 0.3842399835586548,
+ "num_input_tokens_seen": 171006328,
+ "step": 3905,
+ "train_runtime": 30178.5838,
+ "train_tokens_per_second": 5666.48
+ },
+ {
+ "epoch": 0.2632642068408295,
+ "grad_norm": 0.6870277390309487,
+ "learning_rate": 9.578743642873258e-06,
+ "loss": 0.40047216415405273,
+ "num_input_tokens_seen": 171209880,
+ "step": 3910,
+ "train_runtime": 30217.8046,
+ "train_tokens_per_second": 5665.861
+ },
+ {
+ "epoch": 0.26360086183678966,
+ "grad_norm": 0.7479204866431858,
+ "learning_rate": 9.57768073874063e-06,
+ "loss": 0.40533103942871096,
+ "num_input_tokens_seen": 171434584,
+ "step": 3915,
+ "train_runtime": 30250.4257,
+ "train_tokens_per_second": 5667.179
+ },
+ {
+ "epoch": 0.2639375168327498,
+ "grad_norm": 0.822857586143697,
+ "learning_rate": 9.576616554474182e-06,
+ "loss": 0.41611337661743164,
+ "num_input_tokens_seen": 171637768,
+ "step": 3920,
+ "train_runtime": 30287.808,
+ "train_tokens_per_second": 5666.893
+ },
+ {
+ "epoch": 0.26427417182870994,
+ "grad_norm": 0.7335208763069258,
+ "learning_rate": 9.575551090371512e-06,
+ "loss": 0.39514660835266113,
+ "num_input_tokens_seen": 171845072,
+ "step": 3925,
+ "train_runtime": 30328.6771,
+ "train_tokens_per_second": 5666.092
+ },
+ {
+ "epoch": 0.2646108268246701,
+ "grad_norm": 0.7927281370222846,
+ "learning_rate": 9.574484346730564e-06,
+ "loss": 0.4031528472900391,
+ "num_input_tokens_seen": 172066528,
+ "step": 3930,
+ "train_runtime": 30370.1573,
+ "train_tokens_per_second": 5665.645
+ },
+ {
+ "epoch": 0.2649474818206302,
+ "grad_norm": 0.7830404478638062,
+ "learning_rate": 9.573416323849658e-06,
+ "loss": 0.4126886367797852,
+ "num_input_tokens_seen": 172285504,
+ "step": 3935,
+ "train_runtime": 30411.2446,
+ "train_tokens_per_second": 5665.191
+ },
+ {
+ "epoch": 0.26528413681659035,
+ "grad_norm": 0.7639494210600385,
+ "learning_rate": 9.572347022027461e-06,
+ "loss": 0.3947685718536377,
+ "num_input_tokens_seen": 172512448,
+ "step": 3940,
+ "train_runtime": 30452.9824,
+ "train_tokens_per_second": 5664.879
+ },
+ {
+ "epoch": 0.2656207918125505,
+ "grad_norm": 0.7465504294757535,
+ "learning_rate": 9.571276441563e-06,
+ "loss": 0.40018043518066404,
+ "num_input_tokens_seen": 172712640,
+ "step": 3945,
+ "train_runtime": 30484.642,
+ "train_tokens_per_second": 5665.562
+ },
+ {
+ "epoch": 0.26595744680851063,
+ "grad_norm": 0.79021748664721,
+ "learning_rate": 9.570204582755659e-06,
+ "loss": 0.38366117477416994,
+ "num_input_tokens_seen": 172934512,
+ "step": 3950,
+ "train_runtime": 30524.2614,
+ "train_tokens_per_second": 5665.477
+ },
+ {
+ "epoch": 0.2662941018044708,
+ "grad_norm": 0.9006277156621565,
+ "learning_rate": 9.569131445905178e-06,
+ "loss": 0.4079860210418701,
+ "num_input_tokens_seen": 173148536,
+ "step": 3955,
+ "train_runtime": 30557.4244,
+ "train_tokens_per_second": 5666.333
+ },
+ {
+ "epoch": 0.2666307568004309,
+ "grad_norm": 0.8299161339057768,
+ "learning_rate": 9.568057031311659e-06,
+ "loss": 0.41454520225524905,
+ "num_input_tokens_seen": 173357456,
+ "step": 3960,
+ "train_runtime": 30592.1621,
+ "train_tokens_per_second": 5666.728
+ },
+ {
+ "epoch": 0.26696741179639105,
+ "grad_norm": 0.9401299888719756,
+ "learning_rate": 9.566981339275557e-06,
+ "loss": 0.43401298522949217,
+ "num_input_tokens_seen": 173565960,
+ "step": 3965,
+ "train_runtime": 30636.8629,
+ "train_tokens_per_second": 5665.265
+ },
+ {
+ "epoch": 0.2673040667923512,
+ "grad_norm": 0.7744089765372347,
+ "learning_rate": 9.565904370097687e-06,
+ "loss": 0.4027921199798584,
+ "num_input_tokens_seen": 173765184,
+ "step": 3970,
+ "train_runtime": 30675.2756,
+ "train_tokens_per_second": 5664.666
+ },
+ {
+ "epoch": 0.26764072178831133,
+ "grad_norm": 1.1387875256699147,
+ "learning_rate": 9.564826124079219e-06,
+ "loss": 0.3860454082489014,
+ "num_input_tokens_seen": 173982664,
+ "step": 3975,
+ "train_runtime": 30722.6571,
+ "train_tokens_per_second": 5663.008
+ },
+ {
+ "epoch": 0.26797737678427147,
+ "grad_norm": 0.7717245266267511,
+ "learning_rate": 9.563746601521681e-06,
+ "loss": 0.37142066955566405,
+ "num_input_tokens_seen": 174207720,
+ "step": 3980,
+ "train_runtime": 30764.4205,
+ "train_tokens_per_second": 5662.636
+ },
+ {
+ "epoch": 0.2683140317802316,
+ "grad_norm": 0.8633313009939835,
+ "learning_rate": 9.562665802726961e-06,
+ "loss": 0.39812779426574707,
+ "num_input_tokens_seen": 174437200,
+ "step": 3985,
+ "train_runtime": 30809.6747,
+ "train_tokens_per_second": 5661.767
+ },
+ {
+ "epoch": 0.26865068677619175,
+ "grad_norm": 0.79235769535237,
+ "learning_rate": 9.561583727997297e-06,
+ "loss": 0.3973848819732666,
+ "num_input_tokens_seen": 174649000,
+ "step": 3990,
+ "train_runtime": 30852.8543,
+ "train_tokens_per_second": 5660.708
+ },
+ {
+ "epoch": 0.2689873417721519,
+ "grad_norm": 0.866887420036326,
+ "learning_rate": 9.560500377635288e-06,
+ "loss": 0.4132582664489746,
+ "num_input_tokens_seen": 174869968,
+ "step": 3995,
+ "train_runtime": 30894.8991,
+ "train_tokens_per_second": 5660.157
+ },
+ {
+ "epoch": 0.26932399676811203,
+ "grad_norm": 0.8490880534158571,
+ "learning_rate": 9.559415751943893e-06,
+ "loss": 0.4121055603027344,
+ "num_input_tokens_seen": 175087088,
+ "step": 4000,
+ "train_runtime": 30928.2656,
+ "train_tokens_per_second": 5661.07
+ },
+ {
+ "epoch": 0.26966065176407217,
+ "grad_norm": 0.7247750937341004,
+ "learning_rate": 9.558329851226422e-06,
+ "loss": 0.3906848430633545,
+ "num_input_tokens_seen": 175314216,
+ "step": 4005,
+ "train_runtime": 30967.3219,
+ "train_tokens_per_second": 5661.265
+ },
+ {
+ "epoch": 0.2699973067600323,
+ "grad_norm": 0.8575764814098764,
+ "learning_rate": 9.557242675786542e-06,
+ "loss": 0.4193889141082764,
+ "num_input_tokens_seen": 175541064,
+ "step": 4010,
+ "train_runtime": 31013.366,
+ "train_tokens_per_second": 5660.175
+ },
+ {
+ "epoch": 0.27033396175599245,
+ "grad_norm": 0.7331252032970815,
+ "learning_rate": 9.55615422592828e-06,
+ "loss": 0.4430998802185059,
+ "num_input_tokens_seen": 175770592,
+ "step": 4015,
+ "train_runtime": 31053.1239,
+ "train_tokens_per_second": 5660.319
+ },
+ {
+ "epoch": 0.2706706167519526,
+ "grad_norm": 0.8233208505066181,
+ "learning_rate": 9.555064501956018e-06,
+ "loss": 0.4185193061828613,
+ "num_input_tokens_seen": 175988096,
+ "step": 4020,
+ "train_runtime": 31090.813,
+ "train_tokens_per_second": 5660.453
+ },
+ {
+ "epoch": 0.2710072717479127,
+ "grad_norm": 0.7929784850811431,
+ "learning_rate": 9.553973504174493e-06,
+ "loss": 0.3700920581817627,
+ "num_input_tokens_seen": 176213488,
+ "step": 4025,
+ "train_runtime": 31131.4846,
+ "train_tokens_per_second": 5660.298
+ },
+ {
+ "epoch": 0.27134392674387287,
+ "grad_norm": 0.8802925347262501,
+ "learning_rate": 9.5528812328888e-06,
+ "loss": 0.41269850730895996,
+ "num_input_tokens_seen": 176420144,
+ "step": 4030,
+ "train_runtime": 31173.5245,
+ "train_tokens_per_second": 5659.294
+ },
+ {
+ "epoch": 0.271680581739833,
+ "grad_norm": 0.7894741861959437,
+ "learning_rate": 9.551787688404386e-06,
+ "loss": 0.3877572536468506,
+ "num_input_tokens_seen": 176635200,
+ "step": 4035,
+ "train_runtime": 31218.475,
+ "train_tokens_per_second": 5658.034
+ },
+ {
+ "epoch": 0.27201723673579314,
+ "grad_norm": 0.7609981119661086,
+ "learning_rate": 9.550692871027061e-06,
+ "loss": 0.3883471727371216,
+ "num_input_tokens_seen": 176856552,
+ "step": 4040,
+ "train_runtime": 31261.3151,
+ "train_tokens_per_second": 5657.361
+ },
+ {
+ "epoch": 0.2723538917317533,
+ "grad_norm": 0.7842118778601046,
+ "learning_rate": 9.549596781062984e-06,
+ "loss": 0.37834415435791013,
+ "num_input_tokens_seen": 177085864,
+ "step": 4045,
+ "train_runtime": 31299.8686,
+ "train_tokens_per_second": 5657.719
+ },
+ {
+ "epoch": 0.2726905467277134,
+ "grad_norm": 0.8324923888304009,
+ "learning_rate": 9.548499418818677e-06,
+ "loss": 0.3912312984466553,
+ "num_input_tokens_seen": 177292800,
+ "step": 4050,
+ "train_runtime": 31343.6672,
+ "train_tokens_per_second": 5656.415
+ },
+ {
+ "epoch": 0.27302720172367356,
+ "grad_norm": 0.886343399880937,
+ "learning_rate": 9.547400784601011e-06,
+ "loss": 0.448984956741333,
+ "num_input_tokens_seen": 177509120,
+ "step": 4055,
+ "train_runtime": 31388.1483,
+ "train_tokens_per_second": 5655.291
+ },
+ {
+ "epoch": 0.2733638567196337,
+ "grad_norm": 0.7611731029617864,
+ "learning_rate": 9.546300878717216e-06,
+ "loss": 0.41724767684936526,
+ "num_input_tokens_seen": 177737192,
+ "step": 4060,
+ "train_runtime": 31419.4811,
+ "train_tokens_per_second": 5656.91
+ },
+ {
+ "epoch": 0.27370051171559384,
+ "grad_norm": 0.7391294054271582,
+ "learning_rate": 9.545199701474877e-06,
+ "loss": 0.3876931667327881,
+ "num_input_tokens_seen": 177955528,
+ "step": 4065,
+ "train_runtime": 31460.3252,
+ "train_tokens_per_second": 5656.506
+ },
+ {
+ "epoch": 0.274037166711554,
+ "grad_norm": 0.8122534322473342,
+ "learning_rate": 9.544097253181935e-06,
+ "loss": 0.37918825149536134,
+ "num_input_tokens_seen": 178168688,
+ "step": 4070,
+ "train_runtime": 31501.2931,
+ "train_tokens_per_second": 5655.917
+ },
+ {
+ "epoch": 0.2743738217075141,
+ "grad_norm": 0.7594531409775707,
+ "learning_rate": 9.542993534146687e-06,
+ "loss": 0.41271028518676756,
+ "num_input_tokens_seen": 178382216,
+ "step": 4075,
+ "train_runtime": 31548.6623,
+ "train_tokens_per_second": 5654.193
+ },
+ {
+ "epoch": 0.27471047670347426,
+ "grad_norm": 0.813877169259893,
+ "learning_rate": 9.541888544677784e-06,
+ "loss": 0.4018410682678223,
+ "num_input_tokens_seen": 178601400,
+ "step": 4080,
+ "train_runtime": 31583.5769,
+ "train_tokens_per_second": 5654.882
+ },
+ {
+ "epoch": 0.2750471316994344,
+ "grad_norm": 0.8060293108187504,
+ "learning_rate": 9.540782285084232e-06,
+ "loss": 0.36272602081298827,
+ "num_input_tokens_seen": 178815704,
+ "step": 4085,
+ "train_runtime": 31624.6421,
+ "train_tokens_per_second": 5654.316
+ },
+ {
+ "epoch": 0.27538378669539454,
+ "grad_norm": 0.7676915231843366,
+ "learning_rate": 9.539674755675392e-06,
+ "loss": 0.39566807746887206,
+ "num_input_tokens_seen": 179033752,
+ "step": 4090,
+ "train_runtime": 31667.4816,
+ "train_tokens_per_second": 5653.552
+ },
+ {
+ "epoch": 0.2757204416913547,
+ "grad_norm": 0.6913331471393089,
+ "learning_rate": 9.538565956760983e-06,
+ "loss": 0.35824832916259763,
+ "num_input_tokens_seen": 179269312,
+ "step": 4095,
+ "train_runtime": 31701.6169,
+ "train_tokens_per_second": 5654.895
+ },
+ {
+ "epoch": 0.2760570966873148,
+ "grad_norm": 0.7630923484514573,
+ "learning_rate": 9.537455888651078e-06,
+ "loss": 0.3908623933792114,
+ "num_input_tokens_seen": 179500456,
+ "step": 4100,
+ "train_runtime": 31737.6882,
+ "train_tokens_per_second": 5655.751
+ },
+ {
+ "epoch": 0.27639375168327496,
+ "grad_norm": 0.8057751017404143,
+ "learning_rate": 9.536344551656103e-06,
+ "loss": 0.4113442420959473,
+ "num_input_tokens_seen": 179727568,
+ "step": 4105,
+ "train_runtime": 31778.3965,
+ "train_tokens_per_second": 5655.653
+ },
+ {
+ "epoch": 0.2767304066792351,
+ "grad_norm": 0.7452483379248684,
+ "learning_rate": 9.535231946086838e-06,
+ "loss": 0.392077112197876,
+ "num_input_tokens_seen": 179947920,
+ "step": 4110,
+ "train_runtime": 31810.5777,
+ "train_tokens_per_second": 5656.858
+ },
+ {
+ "epoch": 0.27706706167519524,
+ "grad_norm": 0.7935004896667632,
+ "learning_rate": 9.534118072254421e-06,
+ "loss": 0.41229848861694335,
+ "num_input_tokens_seen": 180172368,
+ "step": 4115,
+ "train_runtime": 31849.5051,
+ "train_tokens_per_second": 5656.991
+ },
+ {
+ "epoch": 0.2774037166711554,
+ "grad_norm": 0.782426266760514,
+ "learning_rate": 9.533002930470345e-06,
+ "loss": 0.38343167304992676,
+ "num_input_tokens_seen": 180392712,
+ "step": 4120,
+ "train_runtime": 31893.6268,
+ "train_tokens_per_second": 5656.074
+ },
+ {
+ "epoch": 0.2777403716671155,
+ "grad_norm": 0.6585746266135624,
+ "learning_rate": 9.531886521046452e-06,
+ "loss": 0.38598432540893557,
+ "num_input_tokens_seen": 180594944,
+ "step": 4125,
+ "train_runtime": 31934.9141,
+ "train_tokens_per_second": 5655.094
+ },
+ {
+ "epoch": 0.27807702666307565,
+ "grad_norm": 0.8150349479481921,
+ "learning_rate": 9.530768844294947e-06,
+ "loss": 0.41448087692260743,
+ "num_input_tokens_seen": 180816024,
+ "step": 4130,
+ "train_runtime": 31975.4857,
+ "train_tokens_per_second": 5654.833
+ },
+ {
+ "epoch": 0.2784136816590358,
+ "grad_norm": 0.8985901076929287,
+ "learning_rate": 9.52964990052838e-06,
+ "loss": 0.41909160614013674,
+ "num_input_tokens_seen": 181035024,
+ "step": 4135,
+ "train_runtime": 32014.2767,
+ "train_tokens_per_second": 5654.822
+ },
+ {
+ "epoch": 0.27875033665499593,
+ "grad_norm": 0.7912879897563827,
+ "learning_rate": 9.528529690059663e-06,
+ "loss": 0.3741894245147705,
+ "num_input_tokens_seen": 181258672,
+ "step": 4140,
+ "train_runtime": 32055.2042,
+ "train_tokens_per_second": 5654.579
+ },
+ {
+ "epoch": 0.2790869916509561,
+ "grad_norm": 0.9153777980141117,
+ "learning_rate": 9.52740821320206e-06,
+ "loss": 0.4044219970703125,
+ "num_input_tokens_seen": 181475056,
+ "step": 4145,
+ "train_runtime": 32086.7668,
+ "train_tokens_per_second": 5655.76
+ },
+ {
+ "epoch": 0.2794236466469162,
+ "grad_norm": 0.7181310809160721,
+ "learning_rate": 9.526285470269185e-06,
+ "loss": 0.4395021915435791,
+ "num_input_tokens_seen": 181701504,
+ "step": 4150,
+ "train_runtime": 32121.2174,
+ "train_tokens_per_second": 5656.744
+ },
+ {
+ "epoch": 0.2797603016428764,
+ "grad_norm": 0.833451101611962,
+ "learning_rate": 9.525161461575012e-06,
+ "loss": 0.4143718719482422,
+ "num_input_tokens_seen": 181905448,
+ "step": 4155,
+ "train_runtime": 32162.0174,
+ "train_tokens_per_second": 5655.909
+ },
+ {
+ "epoch": 0.28009695663883655,
+ "grad_norm": 0.9347376730252263,
+ "learning_rate": 9.524036187433867e-06,
+ "loss": 0.38724517822265625,
+ "num_input_tokens_seen": 182111832,
+ "step": 4160,
+ "train_runtime": 32199.5756,
+ "train_tokens_per_second": 5655.722
+ },
+ {
+ "epoch": 0.2804336116347967,
+ "grad_norm": 0.7395436682357247,
+ "learning_rate": 9.522909648160427e-06,
+ "loss": 0.38469388484954836,
+ "num_input_tokens_seen": 182331456,
+ "step": 4165,
+ "train_runtime": 32237.7321,
+ "train_tokens_per_second": 5655.84
+ },
+ {
+ "epoch": 0.2807702666307568,
+ "grad_norm": 0.9082810498252065,
+ "learning_rate": 9.521781844069727e-06,
+ "loss": 0.4097011566162109,
+ "num_input_tokens_seen": 182555536,
+ "step": 4170,
+ "train_runtime": 32282.4632,
+ "train_tokens_per_second": 5654.944
+ },
+ {
+ "epoch": 0.28110692162671697,
+ "grad_norm": 0.8211424225834406,
+ "learning_rate": 9.520652775477154e-06,
+ "loss": 0.38606297969818115,
+ "num_input_tokens_seen": 182776104,
+ "step": 4175,
+ "train_runtime": 32324.1374,
+ "train_tokens_per_second": 5654.477
+ },
+ {
+ "epoch": 0.2814435766226771,
+ "grad_norm": 0.6834380753616917,
+ "learning_rate": 9.519522442698447e-06,
+ "loss": 0.36697864532470703,
+ "num_input_tokens_seen": 182986640,
+ "step": 4180,
+ "train_runtime": 32354.2522,
+ "train_tokens_per_second": 5655.722
+ },
+ {
+ "epoch": 0.28178023161863724,
+ "grad_norm": 0.8351889375982321,
+ "learning_rate": 9.518390846049699e-06,
+ "loss": 0.4045860290527344,
+ "num_input_tokens_seen": 183200160,
+ "step": 4185,
+ "train_runtime": 32398.1959,
+ "train_tokens_per_second": 5654.641
+ },
+ {
+ "epoch": 0.2821168866145974,
+ "grad_norm": 0.7607561247725104,
+ "learning_rate": 9.517257985847361e-06,
+ "loss": 0.42084197998046874,
+ "num_input_tokens_seen": 183411672,
+ "step": 4190,
+ "train_runtime": 32441.1387,
+ "train_tokens_per_second": 5653.676
+ },
+ {
+ "epoch": 0.2824535416105575,
+ "grad_norm": 0.7318149860850908,
+ "learning_rate": 9.516123862408232e-06,
+ "loss": 0.4293764591217041,
+ "num_input_tokens_seen": 183636784,
+ "step": 4195,
+ "train_runtime": 32485.1985,
+ "train_tokens_per_second": 5652.937
+ },
+ {
+ "epoch": 0.28279019660651766,
+ "grad_norm": 1.0902481858366875,
+ "learning_rate": 9.514988476049466e-06,
+ "loss": 0.3744691848754883,
+ "num_input_tokens_seen": 183850024,
+ "step": 4200,
+ "train_runtime": 32521.0388,
+ "train_tokens_per_second": 5653.264
+ },
+ {
+ "epoch": 0.2831268516024778,
+ "grad_norm": 0.7196863893637454,
+ "learning_rate": 9.513851827088567e-06,
+ "loss": 0.39441556930541993,
+ "num_input_tokens_seen": 184055312,
+ "step": 4205,
+ "train_runtime": 32558.0875,
+ "train_tokens_per_second": 5653.136
+ },
+ {
+ "epoch": 0.28346350659843794,
+ "grad_norm": 0.901667167722795,
+ "learning_rate": 9.512713915843402e-06,
+ "loss": 0.399395227432251,
+ "num_input_tokens_seen": 184280840,
+ "step": 4210,
+ "train_runtime": 32591.1094,
+ "train_tokens_per_second": 5654.329
+ },
+ {
+ "epoch": 0.2838001615943981,
+ "grad_norm": 0.7871649191207917,
+ "learning_rate": 9.511574742632177e-06,
+ "loss": 0.38000011444091797,
+ "num_input_tokens_seen": 184499800,
+ "step": 4215,
+ "train_runtime": 32635.0953,
+ "train_tokens_per_second": 5653.417
+ },
+ {
+ "epoch": 0.2841368165903582,
+ "grad_norm": 0.7805206391000944,
+ "learning_rate": 9.510434307773464e-06,
+ "loss": 0.45735921859741213,
+ "num_input_tokens_seen": 184725080,
+ "step": 4220,
+ "train_runtime": 32673.8214,
+ "train_tokens_per_second": 5653.611
+ },
+ {
+ "epoch": 0.28447347158631836,
+ "grad_norm": 0.6332223820184592,
+ "learning_rate": 9.509292611586179e-06,
+ "loss": 0.35365097522735595,
+ "num_input_tokens_seen": 184912768,
+ "step": 4225,
+ "train_runtime": 32714.0896,
+ "train_tokens_per_second": 5652.389
+ },
+ {
+ "epoch": 0.2848101265822785,
+ "grad_norm": 0.8305072132980625,
+ "learning_rate": 9.508149654389592e-06,
+ "loss": 0.40491819381713867,
+ "num_input_tokens_seen": 185136408,
+ "step": 4230,
+ "train_runtime": 32755.4509,
+ "train_tokens_per_second": 5652.079
+ },
+ {
+ "epoch": 0.28514678157823864,
+ "grad_norm": 0.6650135309711892,
+ "learning_rate": 9.507005436503331e-06,
+ "loss": 0.3746334552764893,
+ "num_input_tokens_seen": 185367152,
+ "step": 4235,
+ "train_runtime": 32798.7805,
+ "train_tokens_per_second": 5651.648
+ },
+ {
+ "epoch": 0.2854834365741988,
+ "grad_norm": 0.8834317004719436,
+ "learning_rate": 9.505859958247373e-06,
+ "loss": 0.4024499893188477,
+ "num_input_tokens_seen": 185580528,
+ "step": 4240,
+ "train_runtime": 32837.3509,
+ "train_tokens_per_second": 5651.507
+ },
+ {
+ "epoch": 0.2858200915701589,
+ "grad_norm": 0.703931424320413,
+ "learning_rate": 9.504713219942045e-06,
+ "loss": 0.3693699359893799,
+ "num_input_tokens_seen": 185800152,
+ "step": 4245,
+ "train_runtime": 32873.2494,
+ "train_tokens_per_second": 5652.017
+ },
+ {
+ "epoch": 0.28615674656611906,
+ "grad_norm": 0.7673763574951983,
+ "learning_rate": 9.503565221908029e-06,
+ "loss": 0.3989689350128174,
+ "num_input_tokens_seen": 186023072,
+ "step": 4250,
+ "train_runtime": 32908.5929,
+ "train_tokens_per_second": 5652.72
+ },
+ {
+ "epoch": 0.2864934015620792,
+ "grad_norm": 0.7942180767418593,
+ "learning_rate": 9.50241596446636e-06,
+ "loss": 0.41822104454040526,
+ "num_input_tokens_seen": 186251344,
+ "step": 4255,
+ "train_runtime": 32948.6784,
+ "train_tokens_per_second": 5652.771
+ },
+ {
+ "epoch": 0.28683005655803934,
+ "grad_norm": 0.8790228905730333,
+ "learning_rate": 9.501265447938423e-06,
+ "loss": 0.40343165397644043,
+ "num_input_tokens_seen": 186459392,
+ "step": 4260,
+ "train_runtime": 32985.9245,
+ "train_tokens_per_second": 5652.696
+ },
+ {
+ "epoch": 0.2871667115539995,
+ "grad_norm": 0.8360538773951802,
+ "learning_rate": 9.500113672645958e-06,
+ "loss": 0.41559224128723143,
+ "num_input_tokens_seen": 186674536,
+ "step": 4265,
+ "train_runtime": 33020.78,
+ "train_tokens_per_second": 5653.244
+ },
+ {
+ "epoch": 0.2875033665499596,
+ "grad_norm": 0.8611168073988301,
+ "learning_rate": 9.498960638911054e-06,
+ "loss": 0.3698106288909912,
+ "num_input_tokens_seen": 186883568,
+ "step": 4270,
+ "train_runtime": 33066.6343,
+ "train_tokens_per_second": 5651.726
+ },
+ {
+ "epoch": 0.28784002154591976,
+ "grad_norm": 0.7947411629350094,
+ "learning_rate": 9.497806347056153e-06,
+ "loss": 0.394811749458313,
+ "num_input_tokens_seen": 187101448,
+ "step": 4275,
+ "train_runtime": 33108.0971,
+ "train_tokens_per_second": 5651.229
+ },
+ {
+ "epoch": 0.2881766765418799,
+ "grad_norm": 0.8271288744866215,
+ "learning_rate": 9.49665079740405e-06,
+ "loss": 0.41505231857299807,
+ "num_input_tokens_seen": 187323568,
+ "step": 4280,
+ "train_runtime": 33144.1133,
+ "train_tokens_per_second": 5651.79
+ },
+ {
+ "epoch": 0.28851333153784003,
+ "grad_norm": 0.8870087883370243,
+ "learning_rate": 9.495493990277889e-06,
+ "loss": 0.41215314865112307,
+ "num_input_tokens_seen": 187536768,
+ "step": 4285,
+ "train_runtime": 33177.0201,
+ "train_tokens_per_second": 5652.61
+ },
+ {
+ "epoch": 0.2888499865338002,
+ "grad_norm": 0.7416815526876334,
+ "learning_rate": 9.49433592600117e-06,
+ "loss": 0.39543790817260743,
+ "num_input_tokens_seen": 187761936,
+ "step": 4290,
+ "train_runtime": 33218.3652,
+ "train_tokens_per_second": 5652.353
+ },
+ {
+ "epoch": 0.2891866415297603,
+ "grad_norm": 0.8456820374140082,
+ "learning_rate": 9.49317660489774e-06,
+ "loss": 0.38694300651550295,
+ "num_input_tokens_seen": 187989768,
+ "step": 4295,
+ "train_runtime": 33254.7613,
+ "train_tokens_per_second": 5653.018
+ },
+ {
+ "epoch": 0.28952329652572045,
+ "grad_norm": 0.7340597270803788,
+ "learning_rate": 9.4920160272918e-06,
+ "loss": 0.39570274353027346,
+ "num_input_tokens_seen": 188221328,
+ "step": 4300,
+ "train_runtime": 33292.1232,
+ "train_tokens_per_second": 5653.629
+ },
+ {
+ "epoch": 0.2898599515216806,
+ "grad_norm": 0.9417045729052645,
+ "learning_rate": 9.490854193507904e-06,
+ "loss": 0.40100975036621095,
+ "num_input_tokens_seen": 188436984,
+ "step": 4305,
+ "train_runtime": 33333.4329,
+ "train_tokens_per_second": 5653.093
+ },
+ {
+ "epoch": 0.29019660651764073,
+ "grad_norm": 0.8307091778797395,
+ "learning_rate": 9.489691103870951e-06,
+ "loss": 0.4165069580078125,
+ "num_input_tokens_seen": 188666800,
+ "step": 4310,
+ "train_runtime": 33370.0506,
+ "train_tokens_per_second": 5653.776
+ },
+ {
+ "epoch": 0.29053326151360087,
+ "grad_norm": 0.6898420962550031,
+ "learning_rate": 9.488526758706198e-06,
+ "loss": 0.3968646764755249,
+ "num_input_tokens_seen": 188890864,
+ "step": 4315,
+ "train_runtime": 33408.9132,
+ "train_tokens_per_second": 5653.906
+ },
+ {
+ "epoch": 0.290869916509561,
+ "grad_norm": 0.9744166501910104,
+ "learning_rate": 9.48736115833925e-06,
+ "loss": 0.42871532440185545,
+ "num_input_tokens_seen": 189097152,
+ "step": 4320,
+ "train_runtime": 33444.6308,
+ "train_tokens_per_second": 5654.036
+ },
+ {
+ "epoch": 0.29120657150552115,
+ "grad_norm": 0.7709225593851357,
+ "learning_rate": 9.486194303096062e-06,
+ "loss": 0.388767147064209,
+ "num_input_tokens_seen": 189319032,
+ "step": 4325,
+ "train_runtime": 33480.301,
+ "train_tokens_per_second": 5654.639
+ },
+ {
+ "epoch": 0.2915432265014813,
+ "grad_norm": 0.6810949642678763,
+ "learning_rate": 9.485026193302945e-06,
+ "loss": 0.37795631885528563,
+ "num_input_tokens_seen": 189540056,
+ "step": 4330,
+ "train_runtime": 33515.0635,
+ "train_tokens_per_second": 5655.369
+ },
+ {
+ "epoch": 0.29187988149744143,
+ "grad_norm": 0.6374851744198567,
+ "learning_rate": 9.483856829286552e-06,
+ "loss": 0.43641042709350586,
+ "num_input_tokens_seen": 189765816,
+ "step": 4335,
+ "train_runtime": 33556.7326,
+ "train_tokens_per_second": 5655.074
+ },
+ {
+ "epoch": 0.29221653649340157,
+ "grad_norm": 0.8345515619506504,
+ "learning_rate": 9.482686211373896e-06,
+ "loss": 0.4038511276245117,
+ "num_input_tokens_seen": 189992032,
+ "step": 4340,
+ "train_runtime": 33597.0529,
+ "train_tokens_per_second": 5655.021
+ },
+ {
+ "epoch": 0.2925531914893617,
+ "grad_norm": 0.8775359668232541,
+ "learning_rate": 9.481514339892335e-06,
+ "loss": 0.4172847747802734,
+ "num_input_tokens_seen": 190224704,
+ "step": 4345,
+ "train_runtime": 33633.3907,
+ "train_tokens_per_second": 5655.829
+ },
+ {
+ "epoch": 0.29288984648532185,
+ "grad_norm": 0.7728830127459773,
+ "learning_rate": 9.48034121516958e-06,
+ "loss": 0.41391868591308595,
+ "num_input_tokens_seen": 190437840,
+ "step": 4350,
+ "train_runtime": 33671.0816,
+ "train_tokens_per_second": 5655.828
+ },
+ {
+ "epoch": 0.293226501481282,
+ "grad_norm": 0.7347575506312092,
+ "learning_rate": 9.47916683753369e-06,
+ "loss": 0.38980252742767335,
+ "num_input_tokens_seen": 190656112,
+ "step": 4355,
+ "train_runtime": 33713.543,
+ "train_tokens_per_second": 5655.179
+ },
+ {
+ "epoch": 0.2935631564772421,
+ "grad_norm": 0.7225462237696316,
+ "learning_rate": 9.477991207313077e-06,
+ "loss": 0.4024055480957031,
+ "num_input_tokens_seen": 190883776,
+ "step": 4360,
+ "train_runtime": 33754.4757,
+ "train_tokens_per_second": 5655.066
+ },
+ {
+ "epoch": 0.29389981147320227,
+ "grad_norm": 0.7312109916350326,
+ "learning_rate": 9.476814324836504e-06,
+ "loss": 0.38675341606140134,
+ "num_input_tokens_seen": 191111960,
+ "step": 4365,
+ "train_runtime": 33789.2531,
+ "train_tokens_per_second": 5655.998
+ },
+ {
+ "epoch": 0.2942364664691624,
+ "grad_norm": 0.7752639421468126,
+ "learning_rate": 9.475636190433078e-06,
+ "loss": 0.44519591331481934,
+ "num_input_tokens_seen": 191343016,
+ "step": 4370,
+ "train_runtime": 33827.4824,
+ "train_tokens_per_second": 5656.437
+ },
+ {
+ "epoch": 0.29457312146512254,
+ "grad_norm": 0.7480674780381356,
+ "learning_rate": 9.474456804432264e-06,
+ "loss": 0.3989989995956421,
+ "num_input_tokens_seen": 191572600,
+ "step": 4375,
+ "train_runtime": 33864.2914,
+ "train_tokens_per_second": 5657.068
+ },
+ {
+ "epoch": 0.2949097764610827,
+ "grad_norm": 0.7492609939786371,
+ "learning_rate": 9.473276167163872e-06,
+ "loss": 0.41669330596923826,
+ "num_input_tokens_seen": 191802056,
+ "step": 4380,
+ "train_runtime": 33903.6008,
+ "train_tokens_per_second": 5657.277
+ },
+ {
+ "epoch": 0.2952464314570428,
+ "grad_norm": 0.7254851965917017,
+ "learning_rate": 9.472094278958066e-06,
+ "loss": 0.41675920486450196,
+ "num_input_tokens_seen": 192033856,
+ "step": 4385,
+ "train_runtime": 33947.504,
+ "train_tokens_per_second": 5656.789
+ },
+ {
+ "epoch": 0.29558308645300296,
+ "grad_norm": 0.7301224106455536,
+ "learning_rate": 9.470911140145353e-06,
+ "loss": 0.3885148763656616,
+ "num_input_tokens_seen": 192244376,
+ "step": 4390,
+ "train_runtime": 33990.9485,
+ "train_tokens_per_second": 5655.752
+ },
+ {
+ "epoch": 0.2959197414489631,
+ "grad_norm": 0.8599934571941987,
+ "learning_rate": 9.469726751056599e-06,
+ "loss": 0.3994889736175537,
+ "num_input_tokens_seen": 192466432,
+ "step": 4395,
+ "train_runtime": 34031.0021,
+ "train_tokens_per_second": 5655.62
+ },
+ {
+ "epoch": 0.29625639644492324,
+ "grad_norm": 0.8672385958944913,
+ "learning_rate": 9.468541112023009e-06,
+ "loss": 0.42194247245788574,
+ "num_input_tokens_seen": 192676720,
+ "step": 4400,
+ "train_runtime": 34068.9438,
+ "train_tokens_per_second": 5655.494
+ },
+ {
+ "epoch": 0.2965930514408834,
+ "grad_norm": 0.8654785396115573,
+ "learning_rate": 9.467354223376149e-06,
+ "loss": 0.4138471603393555,
+ "num_input_tokens_seen": 192867824,
+ "step": 4405,
+ "train_runtime": 34101.5758,
+ "train_tokens_per_second": 5655.687
+ },
+ {
+ "epoch": 0.2969297064368435,
+ "grad_norm": 0.7586880855501626,
+ "learning_rate": 9.466166085447923e-06,
+ "loss": 0.4040337562561035,
+ "num_input_tokens_seen": 193079552,
+ "step": 4410,
+ "train_runtime": 34144.2523,
+ "train_tokens_per_second": 5654.819
+ },
+ {
+ "epoch": 0.29726636143280366,
+ "grad_norm": 0.8413204922562729,
+ "learning_rate": 9.464976698570595e-06,
+ "loss": 0.3852673053741455,
+ "num_input_tokens_seen": 193274704,
+ "step": 4415,
+ "train_runtime": 34179.0678,
+ "train_tokens_per_second": 5654.768
+ },
+ {
+ "epoch": 0.2976030164287638,
+ "grad_norm": 0.8227530497102407,
+ "learning_rate": 9.463786063076769e-06,
+ "loss": 0.42119736671447755,
+ "num_input_tokens_seen": 193494016,
+ "step": 4420,
+ "train_runtime": 34217.4917,
+ "train_tokens_per_second": 5654.828
+ },
+ {
+ "epoch": 0.29793967142472394,
+ "grad_norm": 0.8498857055875949,
+ "learning_rate": 9.462594179299408e-06,
+ "loss": 0.34813380241394043,
+ "num_input_tokens_seen": 193708944,
+ "step": 4425,
+ "train_runtime": 34255.6112,
+ "train_tokens_per_second": 5654.809
+ },
+ {
+ "epoch": 0.2982763264206841,
+ "grad_norm": 0.7591384035910306,
+ "learning_rate": 9.461401047571811e-06,
+ "loss": 0.39094252586364747,
+ "num_input_tokens_seen": 193940712,
+ "step": 4430,
+ "train_runtime": 34291.0594,
+ "train_tokens_per_second": 5655.722
+ },
+ {
+ "epoch": 0.2986129814166442,
+ "grad_norm": 0.8057409790041082,
+ "learning_rate": 9.460206668227639e-06,
+ "loss": 0.4155745983123779,
+ "num_input_tokens_seen": 194168936,
+ "step": 4435,
+ "train_runtime": 34330.974,
+ "train_tokens_per_second": 5655.795
+ },
+ {
+ "epoch": 0.29894963641260436,
+ "grad_norm": 0.7522439279097726,
+ "learning_rate": 9.459011041600895e-06,
+ "loss": 0.39428043365478516,
+ "num_input_tokens_seen": 194397872,
+ "step": 4440,
+ "train_runtime": 34371.2051,
+ "train_tokens_per_second": 5655.835
+ },
+ {
+ "epoch": 0.2992862914085645,
+ "grad_norm": 0.8885125640048409,
+ "learning_rate": 9.457814168025932e-06,
+ "loss": 0.3819078207015991,
+ "num_input_tokens_seen": 194618856,
+ "step": 4445,
+ "train_runtime": 34409.2545,
+ "train_tokens_per_second": 5656.003
+ },
+ {
+ "epoch": 0.29962294640452464,
+ "grad_norm": 0.8798594554422206,
+ "learning_rate": 9.456616047837452e-06,
+ "loss": 0.4244089603424072,
+ "num_input_tokens_seen": 194815632,
+ "step": 4450,
+ "train_runtime": 34451.3067,
+ "train_tokens_per_second": 5654.811
+ },
+ {
+ "epoch": 0.2999596014004848,
+ "grad_norm": 0.7140830042785486,
+ "learning_rate": 9.455416681370506e-06,
+ "loss": 0.43105688095092776,
+ "num_input_tokens_seen": 195043320,
+ "step": 4455,
+ "train_runtime": 34491.908,
+ "train_tokens_per_second": 5654.756
+ },
+ {
+ "epoch": 0.3002962563964449,
+ "grad_norm": 0.7933555904953391,
+ "learning_rate": 9.45421606896049e-06,
+ "loss": 0.4107870101928711,
+ "num_input_tokens_seen": 195273800,
+ "step": 4460,
+ "train_runtime": 34532.6899,
+ "train_tokens_per_second": 5654.752
+ },
+ {
+ "epoch": 0.30063291139240506,
+ "grad_norm": 0.7656195298280324,
+ "learning_rate": 9.453014210943155e-06,
+ "loss": 0.39834353923797605,
+ "num_input_tokens_seen": 195492912,
+ "step": 4465,
+ "train_runtime": 34574.8621,
+ "train_tokens_per_second": 5654.192
+ },
+ {
+ "epoch": 0.3009695663883652,
+ "grad_norm": 0.7817448952040295,
+ "learning_rate": 9.451811107654596e-06,
+ "loss": 0.39757986068725587,
+ "num_input_tokens_seen": 195716896,
+ "step": 4470,
+ "train_runtime": 34621.7737,
+ "train_tokens_per_second": 5653.0
+ },
+ {
+ "epoch": 0.30130622138432533,
+ "grad_norm": 0.7841726368249392,
+ "learning_rate": 9.450606759431255e-06,
+ "loss": 0.40091800689697266,
+ "num_input_tokens_seen": 195945064,
+ "step": 4475,
+ "train_runtime": 34656.1692,
+ "train_tokens_per_second": 5653.974
+ },
+ {
+ "epoch": 0.3016428763802855,
+ "grad_norm": 0.8817669510853672,
+ "learning_rate": 9.449401166609928e-06,
+ "loss": 0.42542014122009275,
+ "num_input_tokens_seen": 196155296,
+ "step": 4480,
+ "train_runtime": 34691.0919,
+ "train_tokens_per_second": 5654.342
+ },
+ {
+ "epoch": 0.3019795313762456,
+ "grad_norm": 0.958339871121062,
+ "learning_rate": 9.448194329527752e-06,
+ "loss": 0.40885372161865235,
+ "num_input_tokens_seen": 196367168,
+ "step": 4485,
+ "train_runtime": 34725.1749,
+ "train_tokens_per_second": 5654.894
+ },
+ {
+ "epoch": 0.30231618637220575,
+ "grad_norm": 0.7219539133285781,
+ "learning_rate": 9.446986248522216e-06,
+ "loss": 0.38100175857543944,
+ "num_input_tokens_seen": 196585760,
+ "step": 4490,
+ "train_runtime": 34763.5979,
+ "train_tokens_per_second": 5654.931
+ },
+ {
+ "epoch": 0.3026528413681659,
+ "grad_norm": 0.7623511745984738,
+ "learning_rate": 9.445776923931157e-06,
+ "loss": 0.3696560859680176,
+ "num_input_tokens_seen": 196799296,
+ "step": 4495,
+ "train_runtime": 34801.3851,
+ "train_tokens_per_second": 5654.927
+ },
+ {
+ "epoch": 0.30298949636412603,
+ "grad_norm": 0.7900772624336759,
+ "learning_rate": 9.444566356092754e-06,
+ "loss": 0.3985780715942383,
+ "num_input_tokens_seen": 197027464,
+ "step": 4500,
+ "train_runtime": 34835.2173,
+ "train_tokens_per_second": 5655.985
+ },
+ {
+ "epoch": 0.30332615136008617,
+ "grad_norm": 0.8352145795658186,
+ "learning_rate": 9.443354545345545e-06,
+ "loss": 0.36351020336151124,
+ "num_input_tokens_seen": 197246488,
+ "step": 4505,
+ "train_runtime": 34872.1679,
+ "train_tokens_per_second": 5656.273
+ },
+ {
+ "epoch": 0.3036628063560463,
+ "grad_norm": 0.7235873259829829,
+ "learning_rate": 9.442141492028406e-06,
+ "loss": 0.3931256294250488,
+ "num_input_tokens_seen": 197464240,
+ "step": 4510,
+ "train_runtime": 34916.8698,
+ "train_tokens_per_second": 5655.268
+ },
+ {
+ "epoch": 0.30399946135200645,
+ "grad_norm": 0.7287455155410504,
+ "learning_rate": 9.440927196480563e-06,
+ "loss": 0.3859192609786987,
+ "num_input_tokens_seen": 197693416,
+ "step": 4515,
+ "train_runtime": 34952.6701,
+ "train_tokens_per_second": 5656.032
+ },
+ {
+ "epoch": 0.3043361163479666,
+ "grad_norm": 0.7124742821058738,
+ "learning_rate": 9.439711659041593e-06,
+ "loss": 0.3971871376037598,
+ "num_input_tokens_seen": 197908880,
+ "step": 4520,
+ "train_runtime": 34995.5331,
+ "train_tokens_per_second": 5655.261
+ },
+ {
+ "epoch": 0.30467277134392673,
+ "grad_norm": 0.8049967305897996,
+ "learning_rate": 9.438494880051413e-06,
+ "loss": 0.4050449371337891,
+ "num_input_tokens_seen": 198130240,
+ "step": 4525,
+ "train_runtime": 35031.404,
+ "train_tokens_per_second": 5655.789
+ },
+ {
+ "epoch": 0.30500942633988687,
+ "grad_norm": 0.8167232313014808,
+ "learning_rate": 9.437276859850293e-06,
+ "loss": 0.39906351566314696,
+ "num_input_tokens_seen": 198354624,
+ "step": 4530,
+ "train_runtime": 35070.2405,
+ "train_tokens_per_second": 5655.924
+ },
+ {
+ "epoch": 0.305346081335847,
+ "grad_norm": 0.8219257479433354,
+ "learning_rate": 9.436057598778848e-06,
+ "loss": 0.41982173919677734,
+ "num_input_tokens_seen": 198569592,
+ "step": 4535,
+ "train_runtime": 35103.488,
+ "train_tokens_per_second": 5656.691
+ },
+ {
+ "epoch": 0.30568273633180715,
+ "grad_norm": 0.8373417043715061,
+ "learning_rate": 9.434837097178043e-06,
+ "loss": 0.44964447021484377,
+ "num_input_tokens_seen": 198790848,
+ "step": 4540,
+ "train_runtime": 35136.0232,
+ "train_tokens_per_second": 5657.75
+ },
+ {
+ "epoch": 0.3060193913277673,
+ "grad_norm": 0.8551654564560078,
+ "learning_rate": 9.433615355389183e-06,
+ "loss": 0.4145000457763672,
+ "num_input_tokens_seen": 198995864,
+ "step": 4545,
+ "train_runtime": 35171.0896,
+ "train_tokens_per_second": 5657.939
+ },
+ {
+ "epoch": 0.3063560463237274,
+ "grad_norm": 0.8539775390580207,
+ "learning_rate": 9.432392373753926e-06,
+ "loss": 0.38867299556732177,
+ "num_input_tokens_seen": 199228112,
+ "step": 4550,
+ "train_runtime": 35204.7853,
+ "train_tokens_per_second": 5659.12
+ },
+ {
+ "epoch": 0.30669270131968757,
+ "grad_norm": 0.7417346548045687,
+ "learning_rate": 9.431168152614278e-06,
+ "loss": 0.398314380645752,
+ "num_input_tokens_seen": 199444448,
+ "step": 4555,
+ "train_runtime": 35246.8496,
+ "train_tokens_per_second": 5658.504
+ },
+ {
+ "epoch": 0.3070293563156477,
+ "grad_norm": 0.7867984496679544,
+ "learning_rate": 9.429942692312585e-06,
+ "loss": 0.36063923835754397,
+ "num_input_tokens_seen": 199665776,
+ "step": 4560,
+ "train_runtime": 35277.6142,
+ "train_tokens_per_second": 5659.844
+ },
+ {
+ "epoch": 0.30736601131160785,
+ "grad_norm": 0.6064626251289161,
+ "learning_rate": 9.428715993191546e-06,
+ "loss": 0.38531041145324707,
+ "num_input_tokens_seen": 199888952,
+ "step": 4565,
+ "train_runtime": 35318.7352,
+ "train_tokens_per_second": 5659.573
+ },
+ {
+ "epoch": 0.307702666307568,
+ "grad_norm": 0.7823320659003977,
+ "learning_rate": 9.427488055594199e-06,
+ "loss": 0.4015589714050293,
+ "num_input_tokens_seen": 200094448,
+ "step": 4570,
+ "train_runtime": 35356.6905,
+ "train_tokens_per_second": 5659.309
+ },
+ {
+ "epoch": 0.3080393213035281,
+ "grad_norm": 0.7797326812244397,
+ "learning_rate": 9.426258879863937e-06,
+ "loss": 0.40995278358459475,
+ "num_input_tokens_seen": 200330936,
+ "step": 4575,
+ "train_runtime": 35388.9695,
+ "train_tokens_per_second": 5660.83
+ },
+ {
+ "epoch": 0.30837597629948826,
+ "grad_norm": 0.6558180029153803,
+ "learning_rate": 9.425028466344494e-06,
+ "loss": 0.35400965213775637,
+ "num_input_tokens_seen": 200550336,
+ "step": 4580,
+ "train_runtime": 35423.4598,
+ "train_tokens_per_second": 5661.512
+ },
+ {
+ "epoch": 0.3087126312954484,
+ "grad_norm": 0.9357449426999332,
+ "learning_rate": 9.42379681537995e-06,
+ "loss": 0.40514564514160156,
+ "num_input_tokens_seen": 200767584,
+ "step": 4585,
+ "train_runtime": 35466.8621,
+ "train_tokens_per_second": 5660.709
+ },
+ {
+ "epoch": 0.30904928629140854,
+ "grad_norm": 1.1029634658189478,
+ "learning_rate": 9.422563927314732e-06,
+ "loss": 0.39676432609558104,
+ "num_input_tokens_seen": 200990368,
+ "step": 4590,
+ "train_runtime": 35503.267,
+ "train_tokens_per_second": 5661.18
+ },
+ {
+ "epoch": 0.3093859412873687,
+ "grad_norm": 0.7960058470606449,
+ "learning_rate": 9.421329802493615e-06,
+ "loss": 0.3798659801483154,
+ "num_input_tokens_seen": 201213136,
+ "step": 4595,
+ "train_runtime": 35545.6953,
+ "train_tokens_per_second": 5660.689
+ },
+ {
+ "epoch": 0.3097225962833288,
+ "grad_norm": 0.7764466461774601,
+ "learning_rate": 9.420094441261717e-06,
+ "loss": 0.4128886222839355,
+ "num_input_tokens_seen": 201436008,
+ "step": 4600,
+ "train_runtime": 35591.0272,
+ "train_tokens_per_second": 5659.741
+ },
+ {
+ "epoch": 0.31005925127928896,
+ "grad_norm": 0.8514809314658448,
+ "learning_rate": 9.418857843964506e-06,
+ "loss": 0.3874027252197266,
+ "num_input_tokens_seen": 201651024,
+ "step": 4605,
+ "train_runtime": 35633.1312,
+ "train_tokens_per_second": 5659.088
+ },
+ {
+ "epoch": 0.3103959062752491,
+ "grad_norm": 0.791242525351627,
+ "learning_rate": 9.417620010947786e-06,
+ "loss": 0.41231350898742675,
+ "num_input_tokens_seen": 201862568,
+ "step": 4610,
+ "train_runtime": 35672.8097,
+ "train_tokens_per_second": 5658.724
+ },
+ {
+ "epoch": 0.31073256127120924,
+ "grad_norm": 0.8196831043483923,
+ "learning_rate": 9.416380942557719e-06,
+ "loss": 0.3759624719619751,
+ "num_input_tokens_seen": 202081848,
+ "step": 4615,
+ "train_runtime": 35711.7461,
+ "train_tokens_per_second": 5658.694
+ },
+ {
+ "epoch": 0.3110692162671694,
+ "grad_norm": 0.814113401127112,
+ "learning_rate": 9.415140639140803e-06,
+ "loss": 0.40959873199462893,
+ "num_input_tokens_seen": 202288416,
+ "step": 4620,
+ "train_runtime": 35748.2065,
+ "train_tokens_per_second": 5658.701
+ },
+ {
+ "epoch": 0.3114058712631295,
+ "grad_norm": 0.6867352095609621,
+ "learning_rate": 9.413899101043887e-06,
+ "loss": 0.4299426555633545,
+ "num_input_tokens_seen": 202507280,
+ "step": 4625,
+ "train_runtime": 35787.2021,
+ "train_tokens_per_second": 5658.651
+ },
+ {
+ "epoch": 0.31174252625908966,
+ "grad_norm": 0.9127540749426457,
+ "learning_rate": 9.412656328614162e-06,
+ "loss": 0.4051331043243408,
+ "num_input_tokens_seen": 202708520,
+ "step": 4630,
+ "train_runtime": 35826.4689,
+ "train_tokens_per_second": 5658.066
+ },
+ {
+ "epoch": 0.3120791812550498,
+ "grad_norm": 0.7769819809945303,
+ "learning_rate": 9.411412322199165e-06,
+ "loss": 0.40700321197509765,
+ "num_input_tokens_seen": 202938288,
+ "step": 4635,
+ "train_runtime": 35860.2769,
+ "train_tokens_per_second": 5659.139
+ },
+ {
+ "epoch": 0.31241583625100994,
+ "grad_norm": 0.7763840440333488,
+ "learning_rate": 9.41016708214678e-06,
+ "loss": 0.397929310798645,
+ "num_input_tokens_seen": 203140768,
+ "step": 4640,
+ "train_runtime": 35900.8969,
+ "train_tokens_per_second": 5658.376
+ },
+ {
+ "epoch": 0.31275249124697013,
+ "grad_norm": 0.838793433476314,
+ "learning_rate": 9.408920608805233e-06,
+ "loss": 0.3729093074798584,
+ "num_input_tokens_seen": 203353200,
+ "step": 4645,
+ "train_runtime": 35943.6665,
+ "train_tokens_per_second": 5657.553
+ },
+ {
+ "epoch": 0.31308914624293027,
+ "grad_norm": 0.759403553618091,
+ "learning_rate": 9.407672902523097e-06,
+ "loss": 0.42061967849731446,
+ "num_input_tokens_seen": 203576960,
+ "step": 4650,
+ "train_runtime": 35989.9311,
+ "train_tokens_per_second": 5656.498
+ },
+ {
+ "epoch": 0.3134258012388904,
+ "grad_norm": 0.8516208752301061,
+ "learning_rate": 9.406423963649287e-06,
+ "loss": 0.39535982608795167,
+ "num_input_tokens_seen": 203780664,
+ "step": 4655,
+ "train_runtime": 36033.4005,
+ "train_tokens_per_second": 5655.327
+ },
+ {
+ "epoch": 0.31376245623485055,
+ "grad_norm": 0.6661646334234551,
+ "learning_rate": 9.405173792533069e-06,
+ "loss": 0.3896958827972412,
+ "num_input_tokens_seen": 204002504,
+ "step": 4660,
+ "train_runtime": 36071.9272,
+ "train_tokens_per_second": 5655.437
+ },
+ {
+ "epoch": 0.3140991112308107,
+ "grad_norm": 0.802128758388854,
+ "learning_rate": 9.403922389524045e-06,
+ "loss": 0.383258056640625,
+ "num_input_tokens_seen": 204219216,
+ "step": 4665,
+ "train_runtime": 36107.3083,
+ "train_tokens_per_second": 5655.897
+ },
+ {
+ "epoch": 0.31443576622677083,
+ "grad_norm": 0.8311848290104593,
+ "learning_rate": 9.40266975497217e-06,
+ "loss": 0.41370835304260256,
+ "num_input_tokens_seen": 204435080,
+ "step": 4670,
+ "train_runtime": 36146.1312,
+ "train_tokens_per_second": 5655.794
+ },
+ {
+ "epoch": 0.31477242122273097,
+ "grad_norm": 0.7524391894573029,
+ "learning_rate": 9.401415889227736e-06,
+ "loss": 0.4065183162689209,
+ "num_input_tokens_seen": 204665984,
+ "step": 4675,
+ "train_runtime": 36189.9551,
+ "train_tokens_per_second": 5655.326
+ },
+ {
+ "epoch": 0.3151090762186911,
+ "grad_norm": 0.6516968499884368,
+ "learning_rate": 9.400160792641385e-06,
+ "loss": 0.40553603172302244,
+ "num_input_tokens_seen": 204900584,
+ "step": 4680,
+ "train_runtime": 36230.3148,
+ "train_tokens_per_second": 5655.501
+ },
+ {
+ "epoch": 0.31544573121465125,
+ "grad_norm": 0.713149562707507,
+ "learning_rate": 9.398904465564097e-06,
+ "loss": 0.3963156700134277,
+ "num_input_tokens_seen": 205121896,
+ "step": 4685,
+ "train_runtime": 36261.5038,
+ "train_tokens_per_second": 5656.74
+ },
+ {
+ "epoch": 0.3157823862106114,
+ "grad_norm": 0.7445101762458055,
+ "learning_rate": 9.397646908347203e-06,
+ "loss": 0.3701439142227173,
+ "num_input_tokens_seen": 205337424,
+ "step": 4690,
+ "train_runtime": 36299.7979,
+ "train_tokens_per_second": 5656.71
+ },
+ {
+ "epoch": 0.3161190412065715,
+ "grad_norm": 0.7876242336850148,
+ "learning_rate": 9.396388121342375e-06,
+ "loss": 0.37079999446868894,
+ "num_input_tokens_seen": 205566880,
+ "step": 4695,
+ "train_runtime": 36337.2628,
+ "train_tokens_per_second": 5657.192
+ },
+ {
+ "epoch": 0.31645569620253167,
+ "grad_norm": 0.8752489859691854,
+ "learning_rate": 9.395128104901628e-06,
+ "loss": 0.382697606086731,
+ "num_input_tokens_seen": 205780336,
+ "step": 4700,
+ "train_runtime": 36373.0101,
+ "train_tokens_per_second": 5657.501
+ },
+ {
+ "epoch": 0.3167923511984918,
+ "grad_norm": 0.6981702689866264,
+ "learning_rate": 9.393866859377321e-06,
+ "loss": 0.4276883125305176,
+ "num_input_tokens_seen": 206004992,
+ "step": 4705,
+ "train_runtime": 36409.7995,
+ "train_tokens_per_second": 5657.955
+ },
+ {
+ "epoch": 0.31712900619445195,
+ "grad_norm": 0.7289408984257015,
+ "learning_rate": 9.392604385122157e-06,
+ "loss": 0.39430766105651854,
+ "num_input_tokens_seen": 206237344,
+ "step": 4710,
+ "train_runtime": 36441.1815,
+ "train_tokens_per_second": 5659.458
+ },
+ {
+ "epoch": 0.3174656611904121,
+ "grad_norm": 0.7707325345782555,
+ "learning_rate": 9.391340682489185e-06,
+ "loss": 0.3808866024017334,
+ "num_input_tokens_seen": 206449496,
+ "step": 4715,
+ "train_runtime": 36480.1405,
+ "train_tokens_per_second": 5659.23
+ },
+ {
+ "epoch": 0.3178023161863722,
+ "grad_norm": 0.8379775589672267,
+ "learning_rate": 9.390075751831794e-06,
+ "loss": 0.39603471755981445,
+ "num_input_tokens_seen": 206677304,
+ "step": 4720,
+ "train_runtime": 36513.5133,
+ "train_tokens_per_second": 5660.296
+ },
+ {
+ "epoch": 0.31813897118233236,
+ "grad_norm": 1.1929541297616924,
+ "learning_rate": 9.388809593503718e-06,
+ "loss": 0.41028714179992676,
+ "num_input_tokens_seen": 206901904,
+ "step": 4725,
+ "train_runtime": 36554.4289,
+ "train_tokens_per_second": 5660.105
+ },
+ {
+ "epoch": 0.3184756261782925,
+ "grad_norm": 0.8413335424516502,
+ "learning_rate": 9.387542207859034e-06,
+ "loss": 0.428537654876709,
+ "num_input_tokens_seen": 207122320,
+ "step": 4730,
+ "train_runtime": 36584.6009,
+ "train_tokens_per_second": 5661.462
+ },
+ {
+ "epoch": 0.31881228117425264,
+ "grad_norm": 0.7898247112897376,
+ "learning_rate": 9.386273595252161e-06,
+ "loss": 0.41155662536621096,
+ "num_input_tokens_seen": 207352296,
+ "step": 4735,
+ "train_runtime": 36620.4486,
+ "train_tokens_per_second": 5662.2
+ },
+ {
+ "epoch": 0.3191489361702128,
+ "grad_norm": 0.8389868484209729,
+ "learning_rate": 9.385003756037865e-06,
+ "loss": 0.4223154067993164,
+ "num_input_tokens_seen": 207575072,
+ "step": 4740,
+ "train_runtime": 36662.9004,
+ "train_tokens_per_second": 5661.72
+ },
+ {
+ "epoch": 0.3194855911661729,
+ "grad_norm": 0.827214827358242,
+ "learning_rate": 9.383732690571253e-06,
+ "loss": 0.40798358917236327,
+ "num_input_tokens_seen": 207817360,
+ "step": 4745,
+ "train_runtime": 36697.2613,
+ "train_tokens_per_second": 5663.021
+ },
+ {
+ "epoch": 0.31982224616213306,
+ "grad_norm": 0.8033752581626638,
+ "learning_rate": 9.382460399207769e-06,
+ "loss": 0.40032100677490234,
+ "num_input_tokens_seen": 208045464,
+ "step": 4750,
+ "train_runtime": 36739.6396,
+ "train_tokens_per_second": 5662.697
+ },
+ {
+ "epoch": 0.3201589011580932,
+ "grad_norm": 0.7745249893682031,
+ "learning_rate": 9.381186882303212e-06,
+ "loss": 0.3821758985519409,
+ "num_input_tokens_seen": 208245512,
+ "step": 4755,
+ "train_runtime": 36775.3408,
+ "train_tokens_per_second": 5662.64
+ },
+ {
+ "epoch": 0.32049555615405334,
+ "grad_norm": 0.7835752828676862,
+ "learning_rate": 9.379912140213713e-06,
+ "loss": 0.38662168979644773,
+ "num_input_tokens_seen": 208458376,
+ "step": 4760,
+ "train_runtime": 36810.474,
+ "train_tokens_per_second": 5663.018
+ },
+ {
+ "epoch": 0.3208322111500135,
+ "grad_norm": 0.7427789671372642,
+ "learning_rate": 9.37863617329575e-06,
+ "loss": 0.37845721244812014,
+ "num_input_tokens_seen": 208662464,
+ "step": 4765,
+ "train_runtime": 36848.709,
+ "train_tokens_per_second": 5662.68
+ },
+ {
+ "epoch": 0.3211688661459736,
+ "grad_norm": 0.7906549867893844,
+ "learning_rate": 9.377358981906145e-06,
+ "loss": 0.3855098485946655,
+ "num_input_tokens_seen": 208883752,
+ "step": 4770,
+ "train_runtime": 36885.8469,
+ "train_tokens_per_second": 5662.978
+ },
+ {
+ "epoch": 0.32150552114193376,
+ "grad_norm": 0.7000304206526682,
+ "learning_rate": 9.376080566402059e-06,
+ "loss": 0.41453986167907714,
+ "num_input_tokens_seen": 209114544,
+ "step": 4775,
+ "train_runtime": 36923.9104,
+ "train_tokens_per_second": 5663.391
+ },
+ {
+ "epoch": 0.3218421761378939,
+ "grad_norm": 0.7569304220647727,
+ "learning_rate": 9.374800927140994e-06,
+ "loss": 0.33689126968383787,
+ "num_input_tokens_seen": 209342880,
+ "step": 4780,
+ "train_runtime": 36959.0708,
+ "train_tokens_per_second": 5664.181
+ },
+ {
+ "epoch": 0.32217883113385404,
+ "grad_norm": 0.7693618085921944,
+ "learning_rate": 9.373520064480804e-06,
+ "loss": 0.3944025278091431,
+ "num_input_tokens_seen": 209567016,
+ "step": 4785,
+ "train_runtime": 36999.5953,
+ "train_tokens_per_second": 5664.035
+ },
+ {
+ "epoch": 0.3225154861298142,
+ "grad_norm": 0.7545108220793344,
+ "learning_rate": 9.372237978779672e-06,
+ "loss": 0.3935252666473389,
+ "num_input_tokens_seen": 209798224,
+ "step": 4790,
+ "train_runtime": 37035.8347,
+ "train_tokens_per_second": 5664.736
+ },
+ {
+ "epoch": 0.3228521411257743,
+ "grad_norm": 0.7487128271214056,
+ "learning_rate": 9.37095467039613e-06,
+ "loss": 0.3892039775848389,
+ "num_input_tokens_seen": 209998416,
+ "step": 4795,
+ "train_runtime": 37074.0037,
+ "train_tokens_per_second": 5664.304
+ },
+ {
+ "epoch": 0.32318879612173446,
+ "grad_norm": 0.7480661049890698,
+ "learning_rate": 9.369670139689056e-06,
+ "loss": 0.3860762119293213,
+ "num_input_tokens_seen": 210212136,
+ "step": 4800,
+ "train_runtime": 37113.4829,
+ "train_tokens_per_second": 5664.037
+ },
+ {
+ "epoch": 0.3235254511176946,
+ "grad_norm": 0.7846044101372698,
+ "learning_rate": 9.368384387017659e-06,
+ "loss": 0.377323055267334,
+ "num_input_tokens_seen": 210412912,
+ "step": 4805,
+ "train_runtime": 37150.0136,
+ "train_tokens_per_second": 5663.872
+ },
+ {
+ "epoch": 0.32386210611365474,
+ "grad_norm": 0.77183510086721,
+ "learning_rate": 9.367097412741497e-06,
+ "loss": 0.3832876443862915,
+ "num_input_tokens_seen": 210626632,
+ "step": 4810,
+ "train_runtime": 37185.5052,
+ "train_tokens_per_second": 5664.213
+ },
+ {
+ "epoch": 0.3241987611096149,
+ "grad_norm": 0.9148498149260199,
+ "learning_rate": 9.36580921722047e-06,
+ "loss": 0.41832733154296875,
+ "num_input_tokens_seen": 210847304,
+ "step": 4815,
+ "train_runtime": 37226.3372,
+ "train_tokens_per_second": 5663.928
+ },
+ {
+ "epoch": 0.324535416105575,
+ "grad_norm": 0.6839164180369833,
+ "learning_rate": 9.364519800814818e-06,
+ "loss": 0.3626012563705444,
+ "num_input_tokens_seen": 211071840,
+ "step": 4820,
+ "train_runtime": 37269.7466,
+ "train_tokens_per_second": 5663.356
+ },
+ {
+ "epoch": 0.32487207110153515,
+ "grad_norm": 0.6658908418064057,
+ "learning_rate": 9.36322916388512e-06,
+ "loss": 0.3882687330245972,
+ "num_input_tokens_seen": 211297288,
+ "step": 4825,
+ "train_runtime": 37311.4159,
+ "train_tokens_per_second": 5663.073
+ },
+ {
+ "epoch": 0.3252087260974953,
+ "grad_norm": 0.7420928499905488,
+ "learning_rate": 9.3619373067923e-06,
+ "loss": 0.3612262487411499,
+ "num_input_tokens_seen": 211529896,
+ "step": 4830,
+ "train_runtime": 37348.4533,
+ "train_tokens_per_second": 5663.686
+ },
+ {
+ "epoch": 0.32554538109345543,
+ "grad_norm": 0.7491528413015758,
+ "learning_rate": 9.360644229897622e-06,
+ "loss": 0.39133119583129883,
+ "num_input_tokens_seen": 211744104,
+ "step": 4835,
+ "train_runtime": 37390.6137,
+ "train_tokens_per_second": 5663.028
+ },
+ {
+ "epoch": 0.32588203608941557,
+ "grad_norm": 0.8362580597115173,
+ "learning_rate": 9.35934993356269e-06,
+ "loss": 0.4070117950439453,
+ "num_input_tokens_seen": 211960664,
+ "step": 4840,
+ "train_runtime": 37434.3008,
+ "train_tokens_per_second": 5662.204
+ },
+ {
+ "epoch": 0.3262186910853757,
+ "grad_norm": 0.8872020316370295,
+ "learning_rate": 9.35805441814945e-06,
+ "loss": 0.37361793518066405,
+ "num_input_tokens_seen": 212184888,
+ "step": 4845,
+ "train_runtime": 37481.2897,
+ "train_tokens_per_second": 5661.088
+ },
+ {
+ "epoch": 0.32655534608133585,
+ "grad_norm": 0.7655564350170675,
+ "learning_rate": 9.356757684020188e-06,
+ "loss": 0.40865292549133303,
+ "num_input_tokens_seen": 212387000,
+ "step": 4850,
+ "train_runtime": 37521.1842,
+ "train_tokens_per_second": 5660.456
+ },
+ {
+ "epoch": 0.326892001077296,
+ "grad_norm": 0.9840119081755966,
+ "learning_rate": 9.355459731537533e-06,
+ "loss": 0.37503221035003664,
+ "num_input_tokens_seen": 212602240,
+ "step": 4855,
+ "train_runtime": 37563.6776,
+ "train_tokens_per_second": 5659.782
+ },
+ {
+ "epoch": 0.32722865607325613,
+ "grad_norm": 0.8416963207567933,
+ "learning_rate": 9.354160561064451e-06,
+ "loss": 0.4150728225708008,
+ "num_input_tokens_seen": 212829424,
+ "step": 4860,
+ "train_runtime": 37600.8083,
+ "train_tokens_per_second": 5660.235
+ },
+ {
+ "epoch": 0.32756531106921627,
+ "grad_norm": 0.6549002669311522,
+ "learning_rate": 9.352860172964254e-06,
+ "loss": 0.41104774475097655,
+ "num_input_tokens_seen": 213059616,
+ "step": 4865,
+ "train_runtime": 37639.635,
+ "train_tokens_per_second": 5660.512
+ },
+ {
+ "epoch": 0.3279019660651764,
+ "grad_norm": 0.7662647212420475,
+ "learning_rate": 9.35155856760059e-06,
+ "loss": 0.37230610847473145,
+ "num_input_tokens_seen": 213291344,
+ "step": 4870,
+ "train_runtime": 37683.6107,
+ "train_tokens_per_second": 5660.056
+ },
+ {
+ "epoch": 0.32823862106113655,
+ "grad_norm": 0.8032964428930659,
+ "learning_rate": 9.35025574533745e-06,
+ "loss": 0.4028165817260742,
+ "num_input_tokens_seen": 213518768,
+ "step": 4875,
+ "train_runtime": 37716.795,
+ "train_tokens_per_second": 5661.106
+ },
+ {
+ "epoch": 0.3285752760570967,
+ "grad_norm": 0.709809886638744,
+ "learning_rate": 9.34895170653916e-06,
+ "loss": 0.37210307121276853,
+ "num_input_tokens_seen": 213742008,
+ "step": 4880,
+ "train_runtime": 37761.13,
+ "train_tokens_per_second": 5660.371
+ },
+ {
+ "epoch": 0.3289119310530568,
+ "grad_norm": 1.0245077977220034,
+ "learning_rate": 9.347646451570394e-06,
+ "loss": 0.3780977725982666,
+ "num_input_tokens_seen": 213967120,
+ "step": 4885,
+ "train_runtime": 37795.7217,
+ "train_tokens_per_second": 5661.147
+ },
+ {
+ "epoch": 0.32924858604901697,
+ "grad_norm": 0.7783625701111361,
+ "learning_rate": 9.346339980796162e-06,
+ "loss": 0.44457039833068845,
+ "num_input_tokens_seen": 214204152,
+ "step": 4890,
+ "train_runtime": 37830.001,
+ "train_tokens_per_second": 5662.282
+ },
+ {
+ "epoch": 0.3295852410449771,
+ "grad_norm": 0.7619788379384177,
+ "learning_rate": 9.345032294581813e-06,
+ "loss": 0.3991098403930664,
+ "num_input_tokens_seen": 214426328,
+ "step": 4895,
+ "train_runtime": 37861.0229,
+ "train_tokens_per_second": 5663.511
+ },
+ {
+ "epoch": 0.32992189604093725,
+ "grad_norm": 0.7063552896906414,
+ "learning_rate": 9.343723393293038e-06,
+ "loss": 0.37960023880004884,
+ "num_input_tokens_seen": 214641400,
+ "step": 4900,
+ "train_runtime": 37900.5478,
+ "train_tokens_per_second": 5663.28
+ },
+ {
+ "epoch": 0.3302585510368974,
+ "grad_norm": 0.9300418965561537,
+ "learning_rate": 9.342413277295869e-06,
+ "loss": 0.3947901725769043,
+ "num_input_tokens_seen": 214865368,
+ "step": 4905,
+ "train_runtime": 37941.8194,
+ "train_tokens_per_second": 5663.022
+ },
+ {
+ "epoch": 0.3305952060328575,
+ "grad_norm": 0.660978808262332,
+ "learning_rate": 9.341101946956672e-06,
+ "loss": 0.39415407180786133,
+ "num_input_tokens_seen": 215077016,
+ "step": 4910,
+ "train_runtime": 37976.1785,
+ "train_tokens_per_second": 5663.472
+ },
+ {
+ "epoch": 0.33093186102881766,
+ "grad_norm": 0.7480448807625937,
+ "learning_rate": 9.33978940264216e-06,
+ "loss": 0.4069984436035156,
+ "num_input_tokens_seen": 215281168,
+ "step": 4915,
+ "train_runtime": 38020.9215,
+ "train_tokens_per_second": 5662.176
+ },
+ {
+ "epoch": 0.3312685160247778,
+ "grad_norm": 0.9363661586418124,
+ "learning_rate": 9.338475644719377e-06,
+ "loss": 0.396682071685791,
+ "num_input_tokens_seen": 215508928,
+ "step": 4920,
+ "train_runtime": 38060.6534,
+ "train_tokens_per_second": 5662.25
+ },
+ {
+ "epoch": 0.33160517102073794,
+ "grad_norm": 0.7522573238030015,
+ "learning_rate": 9.337160673555719e-06,
+ "loss": 0.3950789451599121,
+ "num_input_tokens_seen": 215730176,
+ "step": 4925,
+ "train_runtime": 38096.6721,
+ "train_tokens_per_second": 5662.704
+ },
+ {
+ "epoch": 0.3319418260166981,
+ "grad_norm": 0.7866963830157505,
+ "learning_rate": 9.335844489518905e-06,
+ "loss": 0.3960768222808838,
+ "num_input_tokens_seen": 215941544,
+ "step": 4930,
+ "train_runtime": 38135.6021,
+ "train_tokens_per_second": 5662.466
+ },
+ {
+ "epoch": 0.3322784810126582,
+ "grad_norm": 0.7771252275599534,
+ "learning_rate": 9.334527092977008e-06,
+ "loss": 0.3927053928375244,
+ "num_input_tokens_seen": 216139784,
+ "step": 4935,
+ "train_runtime": 38172.7334,
+ "train_tokens_per_second": 5662.151
+ },
+ {
+ "epoch": 0.33261513600861836,
+ "grad_norm": 0.7031312284914388,
+ "learning_rate": 9.33320848429843e-06,
+ "loss": 0.3905383825302124,
+ "num_input_tokens_seen": 216376872,
+ "step": 4940,
+ "train_runtime": 38213.8052,
+ "train_tokens_per_second": 5662.27
+ },
+ {
+ "epoch": 0.3329517910045785,
+ "grad_norm": 0.7992397342780773,
+ "learning_rate": 9.331888663851917e-06,
+ "loss": 0.41385631561279296,
+ "num_input_tokens_seen": 216602432,
+ "step": 4945,
+ "train_runtime": 38247.7905,
+ "train_tokens_per_second": 5663.136
+ },
+ {
+ "epoch": 0.33328844600053864,
+ "grad_norm": 0.8325843513269255,
+ "learning_rate": 9.330567632006552e-06,
+ "loss": 0.402661657333374,
+ "num_input_tokens_seen": 216820040,
+ "step": 4950,
+ "train_runtime": 38282.5886,
+ "train_tokens_per_second": 5663.672
+ },
+ {
+ "epoch": 0.3336251009964988,
+ "grad_norm": 0.8157610005939018,
+ "learning_rate": 9.329245389131759e-06,
+ "loss": 0.4041293144226074,
+ "num_input_tokens_seen": 217040648,
+ "step": 4955,
+ "train_runtime": 38321.0859,
+ "train_tokens_per_second": 5663.74
+ },
+ {
+ "epoch": 0.3339617559924589,
+ "grad_norm": 0.7514981142846812,
+ "learning_rate": 9.327921935597298e-06,
+ "loss": 0.373291015625,
+ "num_input_tokens_seen": 217253544,
+ "step": 4960,
+ "train_runtime": 38362.9141,
+ "train_tokens_per_second": 5663.114
+ },
+ {
+ "epoch": 0.33429841098841906,
+ "grad_norm": 0.9584412932125773,
+ "learning_rate": 9.326597271773267e-06,
+ "loss": 0.39936108589172364,
+ "num_input_tokens_seen": 217452896,
+ "step": 4965,
+ "train_runtime": 38404.9607,
+ "train_tokens_per_second": 5662.104
+ },
+ {
+ "epoch": 0.3346350659843792,
+ "grad_norm": 0.7914463925442367,
+ "learning_rate": 9.325271398030107e-06,
+ "loss": 0.3928532600402832,
+ "num_input_tokens_seen": 217666152,
+ "step": 4970,
+ "train_runtime": 38441.97,
+ "train_tokens_per_second": 5662.201
+ },
+ {
+ "epoch": 0.33497172098033934,
+ "grad_norm": 0.8843279935185989,
+ "learning_rate": 9.323944314738591e-06,
+ "loss": 0.40926666259765626,
+ "num_input_tokens_seen": 217886336,
+ "step": 4975,
+ "train_runtime": 38486.6001,
+ "train_tokens_per_second": 5661.356
+ },
+ {
+ "epoch": 0.3353083759762995,
+ "grad_norm": 0.7413319586045699,
+ "learning_rate": 9.322616022269838e-06,
+ "loss": 0.3675239562988281,
+ "num_input_tokens_seen": 218092776,
+ "step": 4980,
+ "train_runtime": 38523.0427,
+ "train_tokens_per_second": 5661.359
+ },
+ {
+ "epoch": 0.3356450309722596,
+ "grad_norm": 0.8050536335877444,
+ "learning_rate": 9.321286520995297e-06,
+ "loss": 0.40448784828186035,
+ "num_input_tokens_seen": 218303136,
+ "step": 4985,
+ "train_runtime": 38563.0671,
+ "train_tokens_per_second": 5660.938
+ },
+ {
+ "epoch": 0.33598168596821976,
+ "grad_norm": 0.8220117119004713,
+ "learning_rate": 9.319955811286758e-06,
+ "loss": 0.38973236083984375,
+ "num_input_tokens_seen": 218521848,
+ "step": 4990,
+ "train_runtime": 38606.842,
+ "train_tokens_per_second": 5660.184
+ },
+ {
+ "epoch": 0.3363183409641799,
+ "grad_norm": 0.724656690749584,
+ "learning_rate": 9.318623893516354e-06,
+ "loss": 0.38318960666656493,
+ "num_input_tokens_seen": 218733712,
+ "step": 4995,
+ "train_runtime": 38643.6596,
+ "train_tokens_per_second": 5660.274
+ },
+ {
+ "epoch": 0.33665499596014004,
+ "grad_norm": 0.777455286936359,
+ "learning_rate": 9.317290768056548e-06,
+ "loss": 0.3933337450027466,
+ "num_input_tokens_seen": 218942384,
+ "step": 5000,
+ "train_runtime": 38685.6918,
+ "train_tokens_per_second": 5659.518
+ },
+ {
+ "epoch": 0.3369916509561002,
+ "grad_norm": 0.7933540714498255,
+ "learning_rate": 9.315956435280147e-06,
+ "loss": 0.40426931381225584,
+ "num_input_tokens_seen": 219148584,
+ "step": 5005,
+ "train_runtime": 38726.8749,
+ "train_tokens_per_second": 5658.824
+ },
+ {
+ "epoch": 0.3373283059520603,
+ "grad_norm": 0.7310653600558,
+ "learning_rate": 9.314620895560292e-06,
+ "loss": 0.3823359251022339,
+ "num_input_tokens_seen": 219366640,
+ "step": 5010,
+ "train_runtime": 38759.9606,
+ "train_tokens_per_second": 5659.62
+ },
+ {
+ "epoch": 0.33766496094802045,
+ "grad_norm": 0.8607865029821252,
+ "learning_rate": 9.313284149270459e-06,
+ "loss": 0.4273562431335449,
+ "num_input_tokens_seen": 219580136,
+ "step": 5015,
+ "train_runtime": 38797.085,
+ "train_tokens_per_second": 5659.707
+ },
+ {
+ "epoch": 0.3380016159439806,
+ "grad_norm": 0.8291229922855889,
+ "learning_rate": 9.311946196784469e-06,
+ "loss": 0.396860933303833,
+ "num_input_tokens_seen": 219795256,
+ "step": 5020,
+ "train_runtime": 38838.6299,
+ "train_tokens_per_second": 5659.192
+ },
+ {
+ "epoch": 0.33833827093994073,
+ "grad_norm": 0.746756264356882,
+ "learning_rate": 9.310607038476476e-06,
+ "loss": 0.40080647468566893,
+ "num_input_tokens_seen": 220007056,
+ "step": 5025,
+ "train_runtime": 38882.3681,
+ "train_tokens_per_second": 5658.273
+ },
+ {
+ "epoch": 0.3386749259359009,
+ "grad_norm": 0.7715704037440936,
+ "learning_rate": 9.30926667472097e-06,
+ "loss": 0.3821887016296387,
+ "num_input_tokens_seen": 220230040,
+ "step": 5030,
+ "train_runtime": 38929.962,
+ "train_tokens_per_second": 5657.083
+ },
+ {
+ "epoch": 0.339011580931861,
+ "grad_norm": 0.695121253798547,
+ "learning_rate": 9.307925105892779e-06,
+ "loss": 0.4106111526489258,
+ "num_input_tokens_seen": 220463304,
+ "step": 5035,
+ "train_runtime": 38964.7555,
+ "train_tokens_per_second": 5658.018
+ },
+ {
+ "epoch": 0.33934823592782115,
+ "grad_norm": 0.7852274341067195,
+ "learning_rate": 9.30658233236707e-06,
+ "loss": 0.38834619522094727,
+ "num_input_tokens_seen": 220695456,
+ "step": 5040,
+ "train_runtime": 39007.1954,
+ "train_tokens_per_second": 5657.814
+ },
+ {
+ "epoch": 0.3396848909237813,
+ "grad_norm": 0.759484699190671,
+ "learning_rate": 9.30523835451934e-06,
+ "loss": 0.3923861265182495,
+ "num_input_tokens_seen": 220915344,
+ "step": 5045,
+ "train_runtime": 39053.4813,
+ "train_tokens_per_second": 5656.739
+ },
+ {
+ "epoch": 0.34002154591974143,
+ "grad_norm": 0.7964278768577083,
+ "learning_rate": 9.303893172725437e-06,
+ "loss": 0.4306797981262207,
+ "num_input_tokens_seen": 221146072,
+ "step": 5050,
+ "train_runtime": 39091.2343,
+ "train_tokens_per_second": 5657.178
+ },
+ {
+ "epoch": 0.34035820091570157,
+ "grad_norm": 0.7340454424901695,
+ "learning_rate": 9.302546787361533e-06,
+ "loss": 0.38288087844848634,
+ "num_input_tokens_seen": 221371344,
+ "step": 5055,
+ "train_runtime": 39133.7258,
+ "train_tokens_per_second": 5656.792
+ },
+ {
+ "epoch": 0.3406948559116617,
+ "grad_norm": 0.7383658204672374,
+ "learning_rate": 9.301199198804139e-06,
+ "loss": 0.37805628776550293,
+ "num_input_tokens_seen": 221583280,
+ "step": 5060,
+ "train_runtime": 39172.7195,
+ "train_tokens_per_second": 5656.571
+ },
+ {
+ "epoch": 0.34103151090762185,
+ "grad_norm": 0.8155446108957558,
+ "learning_rate": 9.299850407430102e-06,
+ "loss": 0.4092425346374512,
+ "num_input_tokens_seen": 221808808,
+ "step": 5065,
+ "train_runtime": 39215.053,
+ "train_tokens_per_second": 5656.216
+ },
+ {
+ "epoch": 0.341368165903582,
+ "grad_norm": 0.8253751293294798,
+ "learning_rate": 9.298500413616613e-06,
+ "loss": 0.3920405864715576,
+ "num_input_tokens_seen": 222029984,
+ "step": 5070,
+ "train_runtime": 39251.0729,
+ "train_tokens_per_second": 5656.66
+ },
+ {
+ "epoch": 0.3417048208995421,
+ "grad_norm": 0.7618503900471122,
+ "learning_rate": 9.297149217741188e-06,
+ "loss": 0.3707792043685913,
+ "num_input_tokens_seen": 222251080,
+ "step": 5075,
+ "train_runtime": 39288.8528,
+ "train_tokens_per_second": 5656.848
+ },
+ {
+ "epoch": 0.34204147589550227,
+ "grad_norm": 0.7834580864284257,
+ "learning_rate": 9.295796820181688e-06,
+ "loss": 0.4238715171813965,
+ "num_input_tokens_seen": 222475224,
+ "step": 5080,
+ "train_runtime": 39326.0211,
+ "train_tokens_per_second": 5657.201
+ },
+ {
+ "epoch": 0.3423781308914624,
+ "grad_norm": 0.7875688230457064,
+ "learning_rate": 9.294443221316305e-06,
+ "loss": 0.3979609727859497,
+ "num_input_tokens_seen": 222689712,
+ "step": 5085,
+ "train_runtime": 39362.9107,
+ "train_tokens_per_second": 5657.349
+ },
+ {
+ "epoch": 0.34271478588742255,
+ "grad_norm": 0.6945475678075447,
+ "learning_rate": 9.293088421523569e-06,
+ "loss": 0.41632776260375975,
+ "num_input_tokens_seen": 222903112,
+ "step": 5090,
+ "train_runtime": 39399.1126,
+ "train_tokens_per_second": 5657.567
+ },
+ {
+ "epoch": 0.3430514408833827,
+ "grad_norm": 0.8727362030270556,
+ "learning_rate": 9.291732421182346e-06,
+ "loss": 0.3804980754852295,
+ "num_input_tokens_seen": 223116536,
+ "step": 5095,
+ "train_runtime": 39444.3485,
+ "train_tokens_per_second": 5656.489
+ },
+ {
+ "epoch": 0.3433880958793428,
+ "grad_norm": 0.7868108132642002,
+ "learning_rate": 9.290375220671837e-06,
+ "loss": 0.3959723949432373,
+ "num_input_tokens_seen": 223331328,
+ "step": 5100,
+ "train_runtime": 39483.2714,
+ "train_tokens_per_second": 5656.353
+ },
+ {
+ "epoch": 0.34372475087530296,
+ "grad_norm": 0.8977037421010565,
+ "learning_rate": 9.28901682037158e-06,
+ "loss": 0.3575240135192871,
+ "num_input_tokens_seen": 223545784,
+ "step": 5105,
+ "train_runtime": 39520.4864,
+ "train_tokens_per_second": 5656.453
+ },
+ {
+ "epoch": 0.3440614058712631,
+ "grad_norm": 0.724529969116944,
+ "learning_rate": 9.287657220661442e-06,
+ "loss": 0.41797566413879395,
+ "num_input_tokens_seen": 223772168,
+ "step": 5110,
+ "train_runtime": 39556.4135,
+ "train_tokens_per_second": 5657.039
+ },
+ {
+ "epoch": 0.34439806086722324,
+ "grad_norm": 0.8314015078496961,
+ "learning_rate": 9.286296421921636e-06,
+ "loss": 0.3681803703308105,
+ "num_input_tokens_seen": 223987880,
+ "step": 5115,
+ "train_runtime": 39597.3882,
+ "train_tokens_per_second": 5656.633
+ },
+ {
+ "epoch": 0.3447347158631834,
+ "grad_norm": 0.8879862781401929,
+ "learning_rate": 9.284934424532706e-06,
+ "loss": 0.4501188278198242,
+ "num_input_tokens_seen": 224197912,
+ "step": 5120,
+ "train_runtime": 39637.597,
+ "train_tokens_per_second": 5656.193
+ },
+ {
+ "epoch": 0.3450713708591435,
+ "grad_norm": 0.7067262341475377,
+ "learning_rate": 9.283571228875525e-06,
+ "loss": 0.35322113037109376,
+ "num_input_tokens_seen": 224398888,
+ "step": 5125,
+ "train_runtime": 39676.3313,
+ "train_tokens_per_second": 5655.737
+ },
+ {
+ "epoch": 0.3454080258551037,
+ "grad_norm": 0.7651046532838097,
+ "learning_rate": 9.282206835331311e-06,
+ "loss": 0.4238561153411865,
+ "num_input_tokens_seen": 224624840,
+ "step": 5130,
+ "train_runtime": 39714.5021,
+ "train_tokens_per_second": 5655.99
+ },
+ {
+ "epoch": 0.34574468085106386,
+ "grad_norm": 1.0540025361763135,
+ "learning_rate": 9.28084124428161e-06,
+ "loss": 0.3581863880157471,
+ "num_input_tokens_seen": 224840616,
+ "step": 5135,
+ "train_runtime": 39760.1229,
+ "train_tokens_per_second": 5654.928
+ },
+ {
+ "epoch": 0.346081335847024,
+ "grad_norm": 0.712972521802771,
+ "learning_rate": 9.279474456108305e-06,
+ "loss": 0.37324137687683107,
+ "num_input_tokens_seen": 225076296,
+ "step": 5140,
+ "train_runtime": 39795.5671,
+ "train_tokens_per_second": 5655.813
+ },
+ {
+ "epoch": 0.34641799084298414,
+ "grad_norm": 0.7436296711193636,
+ "learning_rate": 9.278106471193615e-06,
+ "loss": 0.36896820068359376,
+ "num_input_tokens_seen": 225294528,
+ "step": 5145,
+ "train_runtime": 39833.6798,
+ "train_tokens_per_second": 5655.88
+ },
+ {
+ "epoch": 0.3467546458389443,
+ "grad_norm": 0.7808819136322714,
+ "learning_rate": 9.276737289920093e-06,
+ "loss": 0.4088866710662842,
+ "num_input_tokens_seen": 225512648,
+ "step": 5150,
+ "train_runtime": 39875.84,
+ "train_tokens_per_second": 5655.37
+ },
+ {
+ "epoch": 0.3470913008349044,
+ "grad_norm": 0.74354681758287,
+ "learning_rate": 9.275366912670624e-06,
+ "loss": 0.4070155143737793,
+ "num_input_tokens_seen": 225724000,
+ "step": 5155,
+ "train_runtime": 39910.7695,
+ "train_tokens_per_second": 5655.717
+ },
+ {
+ "epoch": 0.34742795583086455,
+ "grad_norm": 0.7443700356039168,
+ "learning_rate": 9.273995339828432e-06,
+ "loss": 0.3892085075378418,
+ "num_input_tokens_seen": 225939192,
+ "step": 5160,
+ "train_runtime": 39950.0434,
+ "train_tokens_per_second": 5655.543
+ },
+ {
+ "epoch": 0.3477646108268247,
+ "grad_norm": 0.8699331338781421,
+ "learning_rate": 9.272622571777072e-06,
+ "loss": 0.42869272232055666,
+ "num_input_tokens_seen": 226161472,
+ "step": 5165,
+ "train_runtime": 39986.0253,
+ "train_tokens_per_second": 5656.013
+ },
+ {
+ "epoch": 0.34810126582278483,
+ "grad_norm": 0.7946248476997148,
+ "learning_rate": 9.271248608900434e-06,
+ "loss": 0.3663766860961914,
+ "num_input_tokens_seen": 226388672,
+ "step": 5170,
+ "train_runtime": 40022.8625,
+ "train_tokens_per_second": 5656.484
+ },
+ {
+ "epoch": 0.348437920818745,
+ "grad_norm": 0.820181936155836,
+ "learning_rate": 9.269873451582741e-06,
+ "loss": 0.3990020275115967,
+ "num_input_tokens_seen": 226604448,
+ "step": 5175,
+ "train_runtime": 40065.8484,
+ "train_tokens_per_second": 5655.801
+ },
+ {
+ "epoch": 0.3487745758147051,
+ "grad_norm": 0.7865052705818104,
+ "learning_rate": 9.268497100208556e-06,
+ "loss": 0.40553035736083987,
+ "num_input_tokens_seen": 226814192,
+ "step": 5180,
+ "train_runtime": 40108.7169,
+ "train_tokens_per_second": 5654.985
+ },
+ {
+ "epoch": 0.34911123081066525,
+ "grad_norm": 0.8597790154802004,
+ "learning_rate": 9.267119555162765e-06,
+ "loss": 0.40463762283325194,
+ "num_input_tokens_seen": 227044824,
+ "step": 5185,
+ "train_runtime": 40148.6974,
+ "train_tokens_per_second": 5655.098
+ },
+ {
+ "epoch": 0.3494478858066254,
+ "grad_norm": 0.7383395894951612,
+ "learning_rate": 9.2657408168306e-06,
+ "loss": 0.3876547336578369,
+ "num_input_tokens_seen": 227264048,
+ "step": 5190,
+ "train_runtime": 40181.6708,
+ "train_tokens_per_second": 5655.913
+ },
+ {
+ "epoch": 0.34978454080258553,
+ "grad_norm": 0.8478101930950076,
+ "learning_rate": 9.264360885597617e-06,
+ "loss": 0.3948791027069092,
+ "num_input_tokens_seen": 227482568,
+ "step": 5195,
+ "train_runtime": 40224.9641,
+ "train_tokens_per_second": 5655.258
+ },
+ {
+ "epoch": 0.35012119579854567,
+ "grad_norm": 0.8312712862431509,
+ "learning_rate": 9.262979761849709e-06,
+ "loss": 0.3999017238616943,
+ "num_input_tokens_seen": 227701672,
+ "step": 5200,
+ "train_runtime": 40265.2495,
+ "train_tokens_per_second": 5655.042
+ },
+ {
+ "epoch": 0.3504578507945058,
+ "grad_norm": 0.8047649086754547,
+ "learning_rate": 9.261597445973106e-06,
+ "loss": 0.40303988456726075,
+ "num_input_tokens_seen": 227930736,
+ "step": 5205,
+ "train_runtime": 40300.5889,
+ "train_tokens_per_second": 5655.767
+ },
+ {
+ "epoch": 0.35079450579046595,
+ "grad_norm": 0.7279256779386029,
+ "learning_rate": 9.260213938354365e-06,
+ "loss": 0.38801677227020265,
+ "num_input_tokens_seen": 228142248,
+ "step": 5210,
+ "train_runtime": 40339.465,
+ "train_tokens_per_second": 5655.56
+ },
+ {
+ "epoch": 0.3511311607864261,
+ "grad_norm": 0.703365188592268,
+ "learning_rate": 9.25882923938038e-06,
+ "loss": 0.40598435401916505,
+ "num_input_tokens_seen": 228355000,
+ "step": 5215,
+ "train_runtime": 40382.8232,
+ "train_tokens_per_second": 5654.756
+ },
+ {
+ "epoch": 0.35146781578238623,
+ "grad_norm": 0.8641270145143162,
+ "learning_rate": 9.257443349438382e-06,
+ "loss": 0.42745413780212405,
+ "num_input_tokens_seen": 228579256,
+ "step": 5220,
+ "train_runtime": 40419.0219,
+ "train_tokens_per_second": 5655.24
+ },
+ {
+ "epoch": 0.35180447077834637,
+ "grad_norm": 0.8147972074351507,
+ "learning_rate": 9.256056268915924e-06,
+ "loss": 0.40657625198364256,
+ "num_input_tokens_seen": 228801840,
+ "step": 5225,
+ "train_runtime": 40461.4517,
+ "train_tokens_per_second": 5654.81
+ },
+ {
+ "epoch": 0.3521411257743065,
+ "grad_norm": 0.7053549338628146,
+ "learning_rate": 9.254667998200906e-06,
+ "loss": 0.4108288764953613,
+ "num_input_tokens_seen": 229023432,
+ "step": 5230,
+ "train_runtime": 40502.8889,
+ "train_tokens_per_second": 5654.496
+ },
+ {
+ "epoch": 0.35247778077026665,
+ "grad_norm": 0.8352493252466918,
+ "learning_rate": 9.253278537681548e-06,
+ "loss": 0.4314168930053711,
+ "num_input_tokens_seen": 229259792,
+ "step": 5235,
+ "train_runtime": 40538.046,
+ "train_tokens_per_second": 5655.423
+ },
+ {
+ "epoch": 0.3528144357662268,
+ "grad_norm": 0.7807713230559138,
+ "learning_rate": 9.251887887746407e-06,
+ "loss": 0.3931559085845947,
+ "num_input_tokens_seen": 229485656,
+ "step": 5240,
+ "train_runtime": 40577.5224,
+ "train_tokens_per_second": 5655.487
+ },
+ {
+ "epoch": 0.3531510907621869,
+ "grad_norm": 0.8636190729399873,
+ "learning_rate": 9.25049604878438e-06,
+ "loss": 0.38311750888824464,
+ "num_input_tokens_seen": 229704840,
+ "step": 5245,
+ "train_runtime": 40615.5331,
+ "train_tokens_per_second": 5655.591
+ },
+ {
+ "epoch": 0.35348774575814707,
+ "grad_norm": 0.8592849372799231,
+ "learning_rate": 9.249103021184684e-06,
+ "loss": 0.3960615634918213,
+ "num_input_tokens_seen": 229920256,
+ "step": 5250,
+ "train_runtime": 40653.9984,
+ "train_tokens_per_second": 5655.539
+ },
+ {
+ "epoch": 0.3538244007541072,
+ "grad_norm": 0.7502281989758054,
+ "learning_rate": 9.24770880533688e-06,
+ "loss": 0.4205930709838867,
+ "num_input_tokens_seen": 230150768,
+ "step": 5255,
+ "train_runtime": 40691.9738,
+ "train_tokens_per_second": 5655.925
+ },
+ {
+ "epoch": 0.35416105575006734,
+ "grad_norm": 0.9736374450794365,
+ "learning_rate": 9.24631340163085e-06,
+ "loss": 0.41542859077453614,
+ "num_input_tokens_seen": 230338192,
+ "step": 5260,
+ "train_runtime": 40727.6217,
+ "train_tokens_per_second": 5655.577
+ },
+ {
+ "epoch": 0.3544977107460275,
+ "grad_norm": 0.7629428550659803,
+ "learning_rate": 9.244916810456822e-06,
+ "loss": 0.3574668884277344,
+ "num_input_tokens_seen": 230568472,
+ "step": 5265,
+ "train_runtime": 40758.3906,
+ "train_tokens_per_second": 5656.957
+ },
+ {
+ "epoch": 0.3548343657419876,
+ "grad_norm": 0.7107248914856642,
+ "learning_rate": 9.24351903220534e-06,
+ "loss": 0.40234909057617185,
+ "num_input_tokens_seen": 230793504,
+ "step": 5270,
+ "train_runtime": 40798.996,
+ "train_tokens_per_second": 5656.843
+ },
+ {
+ "epoch": 0.35517102073794776,
+ "grad_norm": 0.8468122239159248,
+ "learning_rate": 9.242120067267295e-06,
+ "loss": 0.3589898347854614,
+ "num_input_tokens_seen": 231011256,
+ "step": 5275,
+ "train_runtime": 40837.4422,
+ "train_tokens_per_second": 5656.849
+ },
+ {
+ "epoch": 0.3555076757339079,
+ "grad_norm": 0.8384350564515061,
+ "learning_rate": 9.2407199160339e-06,
+ "loss": 0.3916470050811768,
+ "num_input_tokens_seen": 231224304,
+ "step": 5280,
+ "train_runtime": 40879.0573,
+ "train_tokens_per_second": 5656.302
+ },
+ {
+ "epoch": 0.35584433072986804,
+ "grad_norm": 0.7878128731634584,
+ "learning_rate": 9.239318578896701e-06,
+ "loss": 0.3869880199432373,
+ "num_input_tokens_seen": 231440576,
+ "step": 5285,
+ "train_runtime": 40917.8128,
+ "train_tokens_per_second": 5656.23
+ },
+ {
+ "epoch": 0.3561809857258282,
+ "grad_norm": 0.9404923069120379,
+ "learning_rate": 9.237916056247583e-06,
+ "loss": 0.40935673713684084,
+ "num_input_tokens_seen": 231653280,
+ "step": 5290,
+ "train_runtime": 40955.4085,
+ "train_tokens_per_second": 5656.232
+ },
+ {
+ "epoch": 0.3565176407217883,
+ "grad_norm": 0.790253055764951,
+ "learning_rate": 9.23651234847875e-06,
+ "loss": 0.4153427600860596,
+ "num_input_tokens_seen": 231867296,
+ "step": 5295,
+ "train_runtime": 40993.5671,
+ "train_tokens_per_second": 5656.187
+ },
+ {
+ "epoch": 0.35685429571774846,
+ "grad_norm": 0.8209369098367891,
+ "learning_rate": 9.23510745598275e-06,
+ "loss": 0.3877981185913086,
+ "num_input_tokens_seen": 232073536,
+ "step": 5300,
+ "train_runtime": 41032.3162,
+ "train_tokens_per_second": 5655.872
+ },
+ {
+ "epoch": 0.3571909507137086,
+ "grad_norm": 0.6660402351658175,
+ "learning_rate": 9.233701379152456e-06,
+ "loss": 0.3829216957092285,
+ "num_input_tokens_seen": 232295728,
+ "step": 5305,
+ "train_runtime": 41070.6587,
+ "train_tokens_per_second": 5656.002
+ },
+ {
+ "epoch": 0.35752760570966874,
+ "grad_norm": 0.7635535443595722,
+ "learning_rate": 9.232294118381069e-06,
+ "loss": 0.41249933242797854,
+ "num_input_tokens_seen": 232520456,
+ "step": 5310,
+ "train_runtime": 41107.4158,
+ "train_tokens_per_second": 5656.411
+ },
+ {
+ "epoch": 0.3578642607056289,
+ "grad_norm": 0.7665309192538949,
+ "learning_rate": 9.230885674062129e-06,
+ "loss": 0.3704015493392944,
+ "num_input_tokens_seen": 232740384,
+ "step": 5315,
+ "train_runtime": 41146.5227,
+ "train_tokens_per_second": 5656.38
+ },
+ {
+ "epoch": 0.358200915701589,
+ "grad_norm": 0.805325429526549,
+ "learning_rate": 9.2294760465895e-06,
+ "loss": 0.4049508094787598,
+ "num_input_tokens_seen": 232969144,
+ "step": 5320,
+ "train_runtime": 41180.0542,
+ "train_tokens_per_second": 5657.33
+ },
+ {
+ "epoch": 0.35853757069754916,
+ "grad_norm": 0.8309484529457033,
+ "learning_rate": 9.228065236357383e-06,
+ "loss": 0.39788720607757566,
+ "num_input_tokens_seen": 233196464,
+ "step": 5325,
+ "train_runtime": 41217.3194,
+ "train_tokens_per_second": 5657.73
+ },
+ {
+ "epoch": 0.3588742256935093,
+ "grad_norm": 0.7008865285408351,
+ "learning_rate": 9.226653243760304e-06,
+ "loss": 0.3763267993927002,
+ "num_input_tokens_seen": 233413024,
+ "step": 5330,
+ "train_runtime": 41251.2257,
+ "train_tokens_per_second": 5658.329
+ },
+ {
+ "epoch": 0.35921088068946944,
+ "grad_norm": 0.7229398353747315,
+ "learning_rate": 9.225240069193124e-06,
+ "loss": 0.4120461463928223,
+ "num_input_tokens_seen": 233629584,
+ "step": 5335,
+ "train_runtime": 41295.9805,
+ "train_tokens_per_second": 5657.441
+ },
+ {
+ "epoch": 0.3595475356854296,
+ "grad_norm": 0.854235630422482,
+ "learning_rate": 9.223825713051034e-06,
+ "loss": 0.4018904685974121,
+ "num_input_tokens_seen": 233844536,
+ "step": 5340,
+ "train_runtime": 41326.7978,
+ "train_tokens_per_second": 5658.424
+ },
+ {
+ "epoch": 0.3598841906813897,
+ "grad_norm": 0.8244749095839433,
+ "learning_rate": 9.222410175729549e-06,
+ "loss": 0.38492045402526853,
+ "num_input_tokens_seen": 234063096,
+ "step": 5345,
+ "train_runtime": 41367.4954,
+ "train_tokens_per_second": 5658.14
+ },
+ {
+ "epoch": 0.36022084567734985,
+ "grad_norm": 0.7424522527415252,
+ "learning_rate": 9.220993457624523e-06,
+ "loss": 0.3723012924194336,
+ "num_input_tokens_seen": 234279256,
+ "step": 5350,
+ "train_runtime": 41414.7783,
+ "train_tokens_per_second": 5656.9
+ },
+ {
+ "epoch": 0.36055750067331,
+ "grad_norm": 0.8296218729091308,
+ "learning_rate": 9.219575559132136e-06,
+ "loss": 0.4064267635345459,
+ "num_input_tokens_seen": 234496160,
+ "step": 5355,
+ "train_runtime": 41454.7573,
+ "train_tokens_per_second": 5656.677
+ },
+ {
+ "epoch": 0.36089415566927013,
+ "grad_norm": 0.8270954179905022,
+ "learning_rate": 9.2181564806489e-06,
+ "loss": 0.3711550235748291,
+ "num_input_tokens_seen": 234705872,
+ "step": 5360,
+ "train_runtime": 41488.3956,
+ "train_tokens_per_second": 5657.145
+ },
+ {
+ "epoch": 0.3612308106652303,
+ "grad_norm": 0.7010804881835582,
+ "learning_rate": 9.216736222571655e-06,
+ "loss": 0.36968867778778075,
+ "num_input_tokens_seen": 234928056,
+ "step": 5365,
+ "train_runtime": 41528.5306,
+ "train_tokens_per_second": 5657.028
+ },
+ {
+ "epoch": 0.3615674656611904,
+ "grad_norm": 0.7909884318309988,
+ "learning_rate": 9.215314785297572e-06,
+ "loss": 0.4384300708770752,
+ "num_input_tokens_seen": 235127224,
+ "step": 5370,
+ "train_runtime": 41566.5449,
+ "train_tokens_per_second": 5656.646
+ },
+ {
+ "epoch": 0.36190412065715055,
+ "grad_norm": 0.7730294782111428,
+ "learning_rate": 9.21389216922415e-06,
+ "loss": 0.3896491527557373,
+ "num_input_tokens_seen": 235350232,
+ "step": 5375,
+ "train_runtime": 41604.4719,
+ "train_tokens_per_second": 5656.849
+ },
+ {
+ "epoch": 0.3622407756531107,
+ "grad_norm": 1.5247558034207205,
+ "learning_rate": 9.212468374749222e-06,
+ "loss": 0.4335320949554443,
+ "num_input_tokens_seen": 235570704,
+ "step": 5380,
+ "train_runtime": 41642.8804,
+ "train_tokens_per_second": 5656.926
+ },
+ {
+ "epoch": 0.36257743064907083,
+ "grad_norm": 0.7761987245970543,
+ "learning_rate": 9.211043402270942e-06,
+ "loss": 0.39116473197937013,
+ "num_input_tokens_seen": 235785280,
+ "step": 5385,
+ "train_runtime": 41679.7538,
+ "train_tokens_per_second": 5657.07
+ },
+ {
+ "epoch": 0.36291408564503097,
+ "grad_norm": 0.7417904936648765,
+ "learning_rate": 9.209617252187806e-06,
+ "loss": 0.3792844533920288,
+ "num_input_tokens_seen": 235999696,
+ "step": 5390,
+ "train_runtime": 41717.5624,
+ "train_tokens_per_second": 5657.083
+ },
+ {
+ "epoch": 0.3632507406409911,
+ "grad_norm": 0.8168071916778651,
+ "learning_rate": 9.208189924898627e-06,
+ "loss": 0.3464975833892822,
+ "num_input_tokens_seen": 236210480,
+ "step": 5395,
+ "train_runtime": 41760.5338,
+ "train_tokens_per_second": 5656.309
+ },
+ {
+ "epoch": 0.36358739563695125,
+ "grad_norm": 0.7857550769816507,
+ "learning_rate": 9.206761420802554e-06,
+ "loss": 0.4210227966308594,
+ "num_input_tokens_seen": 236432304,
+ "step": 5400,
+ "train_runtime": 41802.7339,
+ "train_tokens_per_second": 5655.905
+ },
+ {
+ "epoch": 0.3639240506329114,
+ "grad_norm": 0.724781136378123,
+ "learning_rate": 9.205331740299065e-06,
+ "loss": 0.3803566932678223,
+ "num_input_tokens_seen": 236642552,
+ "step": 5405,
+ "train_runtime": 41844.665,
+ "train_tokens_per_second": 5655.262
+ },
+ {
+ "epoch": 0.36426070562887153,
+ "grad_norm": 0.7382233662482024,
+ "learning_rate": 9.203900883787967e-06,
+ "loss": 0.4222862243652344,
+ "num_input_tokens_seen": 236866080,
+ "step": 5410,
+ "train_runtime": 41880.1302,
+ "train_tokens_per_second": 5655.811
+ },
+ {
+ "epoch": 0.36459736062483167,
+ "grad_norm": 0.7707623946075276,
+ "learning_rate": 9.202468851669388e-06,
+ "loss": 0.42064590454101564,
+ "num_input_tokens_seen": 237084656,
+ "step": 5415,
+ "train_runtime": 41922.4674,
+ "train_tokens_per_second": 5655.313
+ },
+ {
+ "epoch": 0.3649340156207918,
+ "grad_norm": 0.7505014032426987,
+ "learning_rate": 9.201035644343797e-06,
+ "loss": 0.40491809844970705,
+ "num_input_tokens_seen": 237299088,
+ "step": 5420,
+ "train_runtime": 41963.0407,
+ "train_tokens_per_second": 5654.955
+ },
+ {
+ "epoch": 0.36527067061675195,
+ "grad_norm": 0.8419562349359976,
+ "learning_rate": 9.199601262211985e-06,
+ "loss": 0.3938578128814697,
+ "num_input_tokens_seen": 237505024,
+ "step": 5425,
+ "train_runtime": 41999.6624,
+ "train_tokens_per_second": 5654.927
+ },
+ {
+ "epoch": 0.3656073256127121,
+ "grad_norm": 0.7844116167185397,
+ "learning_rate": 9.19816570567507e-06,
+ "loss": 0.4014750480651855,
+ "num_input_tokens_seen": 237729312,
+ "step": 5430,
+ "train_runtime": 42038.7844,
+ "train_tokens_per_second": 5655.0
+ },
+ {
+ "epoch": 0.3659439806086722,
+ "grad_norm": 0.8709777369310681,
+ "learning_rate": 9.196728975134503e-06,
+ "loss": 0.41324663162231445,
+ "num_input_tokens_seen": 237937544,
+ "step": 5435,
+ "train_runtime": 42082.2656,
+ "train_tokens_per_second": 5654.105
+ },
+ {
+ "epoch": 0.36628063560463237,
+ "grad_norm": 0.943067660878324,
+ "learning_rate": 9.195291070992061e-06,
+ "loss": 0.39972093105316164,
+ "num_input_tokens_seen": 238144328,
+ "step": 5440,
+ "train_runtime": 42119.7987,
+ "train_tokens_per_second": 5653.976
+ },
+ {
+ "epoch": 0.3666172906005925,
+ "grad_norm": 0.7866408242613544,
+ "learning_rate": 9.193851993649849e-06,
+ "loss": 0.36598644256591795,
+ "num_input_tokens_seen": 238371752,
+ "step": 5445,
+ "train_runtime": 42157.0955,
+ "train_tokens_per_second": 5654.368
+ },
+ {
+ "epoch": 0.36695394559655264,
+ "grad_norm": 0.7572508667451172,
+ "learning_rate": 9.1924117435103e-06,
+ "loss": 0.3939794063568115,
+ "num_input_tokens_seen": 238586128,
+ "step": 5450,
+ "train_runtime": 42198.1981,
+ "train_tokens_per_second": 5653.941
+ },
+ {
+ "epoch": 0.3672906005925128,
+ "grad_norm": 0.885004223430614,
+ "learning_rate": 9.190970320976176e-06,
+ "loss": 0.3719717264175415,
+ "num_input_tokens_seen": 238811344,
+ "step": 5455,
+ "train_runtime": 42232.691,
+ "train_tokens_per_second": 5654.656
+ },
+ {
+ "epoch": 0.3676272555884729,
+ "grad_norm": 0.8940362227103387,
+ "learning_rate": 9.189527726450565e-06,
+ "loss": 0.4350133895874023,
+ "num_input_tokens_seen": 239003208,
+ "step": 5460,
+ "train_runtime": 42266.5178,
+ "train_tokens_per_second": 5654.67
+ },
+ {
+ "epoch": 0.36796391058443306,
+ "grad_norm": 0.6837019733666716,
+ "learning_rate": 9.188083960336885e-06,
+ "loss": 0.3788605213165283,
+ "num_input_tokens_seen": 239225040,
+ "step": 5465,
+ "train_runtime": 42305.7631,
+ "train_tokens_per_second": 5654.668
+ },
+ {
+ "epoch": 0.3683005655803932,
+ "grad_norm": 0.7278725929539929,
+ "learning_rate": 9.186639023038879e-06,
+ "loss": 0.4101285934448242,
+ "num_input_tokens_seen": 239445688,
+ "step": 5470,
+ "train_runtime": 42345.9353,
+ "train_tokens_per_second": 5654.514
+ },
+ {
+ "epoch": 0.36863722057635334,
+ "grad_norm": 0.8240745816268156,
+ "learning_rate": 9.185192914960622e-06,
+ "loss": 0.43002777099609374,
+ "num_input_tokens_seen": 239660856,
+ "step": 5475,
+ "train_runtime": 42383.8589,
+ "train_tokens_per_second": 5654.531
+ },
+ {
+ "epoch": 0.3689738755723135,
+ "grad_norm": 0.7561111927294512,
+ "learning_rate": 9.183745636506511e-06,
+ "loss": 0.400468635559082,
+ "num_input_tokens_seen": 239863160,
+ "step": 5480,
+ "train_runtime": 42415.6347,
+ "train_tokens_per_second": 5655.065
+ },
+ {
+ "epoch": 0.3693105305682736,
+ "grad_norm": 0.6657976994723397,
+ "learning_rate": 9.182297188081274e-06,
+ "loss": 0.39147305488586426,
+ "num_input_tokens_seen": 240088504,
+ "step": 5485,
+ "train_runtime": 42445.5983,
+ "train_tokens_per_second": 5656.382
+ },
+ {
+ "epoch": 0.36964718556423376,
+ "grad_norm": 0.7438674827381799,
+ "learning_rate": 9.180847570089964e-06,
+ "loss": 0.38025264739990233,
+ "num_input_tokens_seen": 240285080,
+ "step": 5490,
+ "train_runtime": 42487.4443,
+ "train_tokens_per_second": 5655.437
+ },
+ {
+ "epoch": 0.3699838405601939,
+ "grad_norm": 0.7687571314606845,
+ "learning_rate": 9.179396782937965e-06,
+ "loss": 0.4285304546356201,
+ "num_input_tokens_seen": 240518464,
+ "step": 5495,
+ "train_runtime": 42522.1167,
+ "train_tokens_per_second": 5656.314
+ },
+ {
+ "epoch": 0.37032049555615404,
+ "grad_norm": 0.8258932109796826,
+ "learning_rate": 9.17794482703098e-06,
+ "loss": 0.41852726936340334,
+ "num_input_tokens_seen": 240737112,
+ "step": 5500,
+ "train_runtime": 42561.0578,
+ "train_tokens_per_second": 5656.277
+ },
+ {
+ "epoch": 0.3706571505521142,
+ "grad_norm": 0.7712097331858169,
+ "learning_rate": 9.176491702775049e-06,
+ "loss": 0.4076284408569336,
+ "num_input_tokens_seen": 240940632,
+ "step": 5505,
+ "train_runtime": 42605.8265,
+ "train_tokens_per_second": 5655.11
+ },
+ {
+ "epoch": 0.3709938055480743,
+ "grad_norm": 0.7293507629162065,
+ "learning_rate": 9.17503741057653e-06,
+ "loss": 0.42336077690124513,
+ "num_input_tokens_seen": 241163512,
+ "step": 5510,
+ "train_runtime": 42641.1976,
+ "train_tokens_per_second": 5655.646
+ },
+ {
+ "epoch": 0.37133046054403446,
+ "grad_norm": 0.7873802897373652,
+ "learning_rate": 9.173581950842111e-06,
+ "loss": 0.3858926773071289,
+ "num_input_tokens_seen": 241380280,
+ "step": 5515,
+ "train_runtime": 42677.1161,
+ "train_tokens_per_second": 5655.965
+ },
+ {
+ "epoch": 0.3716671155399946,
+ "grad_norm": 0.833898462602849,
+ "learning_rate": 9.172125323978811e-06,
+ "loss": 0.39653089046478274,
+ "num_input_tokens_seen": 241601112,
+ "step": 5520,
+ "train_runtime": 42713.21,
+ "train_tokens_per_second": 5656.356
+ },
+ {
+ "epoch": 0.37200377053595474,
+ "grad_norm": 0.7933645694043111,
+ "learning_rate": 9.170667530393966e-06,
+ "loss": 0.4098936080932617,
+ "num_input_tokens_seen": 241828432,
+ "step": 5525,
+ "train_runtime": 42747.8553,
+ "train_tokens_per_second": 5657.089
+ },
+ {
+ "epoch": 0.3723404255319149,
+ "grad_norm": 0.724581396473804,
+ "learning_rate": 9.169208570495247e-06,
+ "loss": 0.3672953605651855,
+ "num_input_tokens_seen": 242051344,
+ "step": 5530,
+ "train_runtime": 42787.8177,
+ "train_tokens_per_second": 5657.015
+ },
+ {
+ "epoch": 0.372677080527875,
+ "grad_norm": 0.720329801926469,
+ "learning_rate": 9.167748444690646e-06,
+ "loss": 0.3897575855255127,
+ "num_input_tokens_seen": 242250584,
+ "step": 5535,
+ "train_runtime": 42826.4052,
+ "train_tokens_per_second": 5656.571
+ },
+ {
+ "epoch": 0.37301373552383515,
+ "grad_norm": 0.7934647008536321,
+ "learning_rate": 9.16628715338848e-06,
+ "loss": 0.3871308326721191,
+ "num_input_tokens_seen": 242459936,
+ "step": 5540,
+ "train_runtime": 42859.1014,
+ "train_tokens_per_second": 5657.14
+ },
+ {
+ "epoch": 0.3733503905197953,
+ "grad_norm": 0.7931279904448103,
+ "learning_rate": 9.164824696997398e-06,
+ "loss": 0.37985944747924805,
+ "num_input_tokens_seen": 242665952,
+ "step": 5545,
+ "train_runtime": 42896.1293,
+ "train_tokens_per_second": 5657.059
+ },
+ {
+ "epoch": 0.37368704551575543,
+ "grad_norm": 0.7548635507515563,
+ "learning_rate": 9.163361075926371e-06,
+ "loss": 0.4154071807861328,
+ "num_input_tokens_seen": 242885960,
+ "step": 5550,
+ "train_runtime": 42932.7348,
+ "train_tokens_per_second": 5657.361
+ },
+ {
+ "epoch": 0.3740237005117156,
+ "grad_norm": 0.824128032145257,
+ "learning_rate": 9.161896290584695e-06,
+ "loss": 0.37793383598327634,
+ "num_input_tokens_seen": 243108432,
+ "step": 5555,
+ "train_runtime": 42974.4088,
+ "train_tokens_per_second": 5657.051
+ },
+ {
+ "epoch": 0.3743603555076757,
+ "grad_norm": 0.755788883617862,
+ "learning_rate": 9.160430341381991e-06,
+ "loss": 0.3929650545120239,
+ "num_input_tokens_seen": 243321200,
+ "step": 5560,
+ "train_runtime": 43020.7871,
+ "train_tokens_per_second": 5655.898
+ },
+ {
+ "epoch": 0.37469701050363585,
+ "grad_norm": 0.8046945349779168,
+ "learning_rate": 9.15896322872821e-06,
+ "loss": 0.39426612854003906,
+ "num_input_tokens_seen": 243541160,
+ "step": 5565,
+ "train_runtime": 43065.6274,
+ "train_tokens_per_second": 5655.117
+ },
+ {
+ "epoch": 0.375033665499596,
+ "grad_norm": 0.8408220299669169,
+ "learning_rate": 9.157494953033625e-06,
+ "loss": 0.4226834297180176,
+ "num_input_tokens_seen": 243766296,
+ "step": 5570,
+ "train_runtime": 43100.4977,
+ "train_tokens_per_second": 5655.765
+ },
+ {
+ "epoch": 0.37537032049555613,
+ "grad_norm": 0.7701423462069609,
+ "learning_rate": 9.156025514708832e-06,
+ "loss": 0.40192432403564454,
+ "num_input_tokens_seen": 244002616,
+ "step": 5575,
+ "train_runtime": 43146.2968,
+ "train_tokens_per_second": 5655.239
+ },
+ {
+ "epoch": 0.37570697549151627,
+ "grad_norm": 0.7773289349591549,
+ "learning_rate": 9.154554914164757e-06,
+ "loss": 0.38442695140838623,
+ "num_input_tokens_seen": 244237240,
+ "step": 5580,
+ "train_runtime": 43188.3958,
+ "train_tokens_per_second": 5655.159
+ },
+ {
+ "epoch": 0.3760436304874764,
+ "grad_norm": 0.6847338403952183,
+ "learning_rate": 9.153083151812649e-06,
+ "loss": 0.377176308631897,
+ "num_input_tokens_seen": 244463576,
+ "step": 5585,
+ "train_runtime": 43225.3626,
+ "train_tokens_per_second": 5655.559
+ },
+ {
+ "epoch": 0.37638028548343655,
+ "grad_norm": 0.780657947894897,
+ "learning_rate": 9.151610228064078e-06,
+ "loss": 0.40382232666015627,
+ "num_input_tokens_seen": 244685816,
+ "step": 5590,
+ "train_runtime": 43264.9573,
+ "train_tokens_per_second": 5655.52
+ },
+ {
+ "epoch": 0.3767169404793967,
+ "grad_norm": 0.7617598851673253,
+ "learning_rate": 9.150136143330949e-06,
+ "loss": 0.42247867584228516,
+ "num_input_tokens_seen": 244901992,
+ "step": 5595,
+ "train_runtime": 43305.9839,
+ "train_tokens_per_second": 5655.154
+ },
+ {
+ "epoch": 0.37705359547535683,
+ "grad_norm": 0.6965545191686713,
+ "learning_rate": 9.14866089802548e-06,
+ "loss": 0.3760505676269531,
+ "num_input_tokens_seen": 245127560,
+ "step": 5600,
+ "train_runtime": 43348.427,
+ "train_tokens_per_second": 5654.82
+ },
+ {
+ "epoch": 0.37739025047131697,
+ "grad_norm": 0.8578077401564244,
+ "learning_rate": 9.147184492560219e-06,
+ "loss": 0.3698588848114014,
+ "num_input_tokens_seen": 245339952,
+ "step": 5605,
+ "train_runtime": 43388.8794,
+ "train_tokens_per_second": 5654.443
+ },
+ {
+ "epoch": 0.3777269054672771,
+ "grad_norm": 1.0507318076347205,
+ "learning_rate": 9.14570692734804e-06,
+ "loss": 0.3787434816360474,
+ "num_input_tokens_seen": 245565672,
+ "step": 5610,
+ "train_runtime": 43425.1962,
+ "train_tokens_per_second": 5654.912
+ },
+ {
+ "epoch": 0.37806356046323725,
+ "grad_norm": 0.8234417174719115,
+ "learning_rate": 9.144228202802137e-06,
+ "loss": 0.41345486640930174,
+ "num_input_tokens_seen": 245793416,
+ "step": 5615,
+ "train_runtime": 43457.0254,
+ "train_tokens_per_second": 5656.011
+ },
+ {
+ "epoch": 0.37840021545919744,
+ "grad_norm": 0.8147806646148233,
+ "learning_rate": 9.142748319336033e-06,
+ "loss": 0.39602549076080323,
+ "num_input_tokens_seen": 246014600,
+ "step": 5620,
+ "train_runtime": 43495.3144,
+ "train_tokens_per_second": 5656.117
+ },
+ {
+ "epoch": 0.3787368704551576,
+ "grad_norm": 0.8294007935657565,
+ "learning_rate": 9.141267277363573e-06,
+ "loss": 0.3865189552307129,
+ "num_input_tokens_seen": 246209040,
+ "step": 5625,
+ "train_runtime": 43530.2862,
+ "train_tokens_per_second": 5656.04
+ },
+ {
+ "epoch": 0.3790735254511177,
+ "grad_norm": 0.8057103649691175,
+ "learning_rate": 9.139785077298921e-06,
+ "loss": 0.3669655084609985,
+ "num_input_tokens_seen": 246435304,
+ "step": 5630,
+ "train_runtime": 43570.2665,
+ "train_tokens_per_second": 5656.043
+ },
+ {
+ "epoch": 0.37941018044707786,
+ "grad_norm": 0.7570731119383132,
+ "learning_rate": 9.138301719556577e-06,
+ "loss": 0.3826982736587524,
+ "num_input_tokens_seen": 246652712,
+ "step": 5635,
+ "train_runtime": 43609.2254,
+ "train_tokens_per_second": 5655.976
+ },
+ {
+ "epoch": 0.379746835443038,
+ "grad_norm": 0.7474769926348979,
+ "learning_rate": 9.136817204551352e-06,
+ "loss": 0.386492919921875,
+ "num_input_tokens_seen": 246882568,
+ "step": 5640,
+ "train_runtime": 43646.3585,
+ "train_tokens_per_second": 5656.43
+ },
+ {
+ "epoch": 0.38008349043899814,
+ "grad_norm": 0.7350249063601988,
+ "learning_rate": 9.135331532698385e-06,
+ "loss": 0.3880929470062256,
+ "num_input_tokens_seen": 247094472,
+ "step": 5645,
+ "train_runtime": 43685.4708,
+ "train_tokens_per_second": 5656.216
+ },
+ {
+ "epoch": 0.3804201454349583,
+ "grad_norm": 0.7440003142918237,
+ "learning_rate": 9.133844704413144e-06,
+ "loss": 0.37475876808166503,
+ "num_input_tokens_seen": 247305056,
+ "step": 5650,
+ "train_runtime": 43724.0377,
+ "train_tokens_per_second": 5656.043
+ },
+ {
+ "epoch": 0.3807568004309184,
+ "grad_norm": 0.829580642455565,
+ "learning_rate": 9.132356720111413e-06,
+ "loss": 0.40270652770996096,
+ "num_input_tokens_seen": 247521856,
+ "step": 5655,
+ "train_runtime": 43766.9087,
+ "train_tokens_per_second": 5655.457
+ },
+ {
+ "epoch": 0.38109345542687856,
+ "grad_norm": 0.7890155830011499,
+ "learning_rate": 9.1308675802093e-06,
+ "loss": 0.3874132394790649,
+ "num_input_tokens_seen": 247735120,
+ "step": 5660,
+ "train_runtime": 43801.888,
+ "train_tokens_per_second": 5655.809
+ },
+ {
+ "epoch": 0.3814301104228387,
+ "grad_norm": 0.7482568242628747,
+ "learning_rate": 9.129377285123241e-06,
+ "loss": 0.3984661102294922,
+ "num_input_tokens_seen": 247949328,
+ "step": 5665,
+ "train_runtime": 43845.4472,
+ "train_tokens_per_second": 5655.076
+ },
+ {
+ "epoch": 0.38176676541879884,
+ "grad_norm": 0.7947458109924397,
+ "learning_rate": 9.127885835269996e-06,
+ "loss": 0.4090867519378662,
+ "num_input_tokens_seen": 248174032,
+ "step": 5670,
+ "train_runtime": 43879.3187,
+ "train_tokens_per_second": 5655.831
+ },
+ {
+ "epoch": 0.382103420414759,
+ "grad_norm": 0.726446478217637,
+ "learning_rate": 9.126393231066636e-06,
+ "loss": 0.3848613977432251,
+ "num_input_tokens_seen": 248384368,
+ "step": 5675,
+ "train_runtime": 43917.9285,
+ "train_tokens_per_second": 5655.649
+ },
+ {
+ "epoch": 0.3824400754107191,
+ "grad_norm": 0.7750970581531168,
+ "learning_rate": 9.124899472930568e-06,
+ "loss": 0.4047283172607422,
+ "num_input_tokens_seen": 248607600,
+ "step": 5680,
+ "train_runtime": 43961.6528,
+ "train_tokens_per_second": 5655.101
+ },
+ {
+ "epoch": 0.38277673040667926,
+ "grad_norm": 1.9536806017675672,
+ "learning_rate": 9.123404561279517e-06,
+ "loss": 0.41053128242492676,
+ "num_input_tokens_seen": 248811928,
+ "step": 5685,
+ "train_runtime": 44006.8302,
+ "train_tokens_per_second": 5653.939
+ },
+ {
+ "epoch": 0.3831133854026394,
+ "grad_norm": 0.7123153967912388,
+ "learning_rate": 9.121908496531527e-06,
+ "loss": 0.3802008867263794,
+ "num_input_tokens_seen": 249035424,
+ "step": 5690,
+ "train_runtime": 44040.1743,
+ "train_tokens_per_second": 5654.733
+ },
+ {
+ "epoch": 0.38345004039859953,
+ "grad_norm": 0.7190654994858774,
+ "learning_rate": 9.120411279104971e-06,
+ "loss": 0.3904706954956055,
+ "num_input_tokens_seen": 249254456,
+ "step": 5695,
+ "train_runtime": 44083.466,
+ "train_tokens_per_second": 5654.148
+ },
+ {
+ "epoch": 0.3837866953945597,
+ "grad_norm": 0.9187897813904784,
+ "learning_rate": 9.118912909418538e-06,
+ "loss": 0.39504170417785645,
+ "num_input_tokens_seen": 249477272,
+ "step": 5700,
+ "train_runtime": 44127.086,
+ "train_tokens_per_second": 5653.609
+ },
+ {
+ "epoch": 0.3841233503905198,
+ "grad_norm": 0.7651587963325609,
+ "learning_rate": 9.117413387891246e-06,
+ "loss": 0.4095571041107178,
+ "num_input_tokens_seen": 249693520,
+ "step": 5705,
+ "train_runtime": 44170.6884,
+ "train_tokens_per_second": 5652.923
+ },
+ {
+ "epoch": 0.38446000538647995,
+ "grad_norm": 0.7879046146044965,
+ "learning_rate": 9.11591271494243e-06,
+ "loss": 0.39027183055877684,
+ "num_input_tokens_seen": 249909984,
+ "step": 5710,
+ "train_runtime": 44209.3015,
+ "train_tokens_per_second": 5652.882
+ },
+ {
+ "epoch": 0.3847966603824401,
+ "grad_norm": 0.7861690179999627,
+ "learning_rate": 9.114410890991748e-06,
+ "loss": 0.40212602615356446,
+ "num_input_tokens_seen": 250124664,
+ "step": 5715,
+ "train_runtime": 44250.9806,
+ "train_tokens_per_second": 5652.41
+ },
+ {
+ "epoch": 0.38513331537840023,
+ "grad_norm": 0.7798159280525913,
+ "learning_rate": 9.112907916459177e-06,
+ "loss": 0.37251665592193606,
+ "num_input_tokens_seen": 250317264,
+ "step": 5720,
+ "train_runtime": 44293.6615,
+ "train_tokens_per_second": 5651.311
+ },
+ {
+ "epoch": 0.38546997037436037,
+ "grad_norm": 0.7886435715633883,
+ "learning_rate": 9.111403791765025e-06,
+ "loss": 0.38798794746398924,
+ "num_input_tokens_seen": 250544576,
+ "step": 5725,
+ "train_runtime": 44329.9756,
+ "train_tokens_per_second": 5651.809
+ },
+ {
+ "epoch": 0.3858066253703205,
+ "grad_norm": 0.7210916045331931,
+ "learning_rate": 9.109898517329914e-06,
+ "loss": 0.3711621046066284,
+ "num_input_tokens_seen": 250774080,
+ "step": 5730,
+ "train_runtime": 44362.0555,
+ "train_tokens_per_second": 5652.896
+ },
+ {
+ "epoch": 0.38614328036628065,
+ "grad_norm": 0.757278528605975,
+ "learning_rate": 9.108392093574785e-06,
+ "loss": 0.3934319972991943,
+ "num_input_tokens_seen": 250992112,
+ "step": 5735,
+ "train_runtime": 44396.0233,
+ "train_tokens_per_second": 5653.482
+ },
+ {
+ "epoch": 0.3864799353622408,
+ "grad_norm": 0.7725316475525575,
+ "learning_rate": 9.10688452092091e-06,
+ "loss": 0.36261420249938964,
+ "num_input_tokens_seen": 251210912,
+ "step": 5740,
+ "train_runtime": 44426.4194,
+ "train_tokens_per_second": 5654.539
+ },
+ {
+ "epoch": 0.38681659035820093,
+ "grad_norm": 0.8668597404685281,
+ "learning_rate": 9.105375799789874e-06,
+ "loss": 0.39951817989349364,
+ "num_input_tokens_seen": 251426584,
+ "step": 5745,
+ "train_runtime": 44475.8422,
+ "train_tokens_per_second": 5653.105
+ },
+ {
+ "epoch": 0.38715324535416107,
+ "grad_norm": 0.8168129882131703,
+ "learning_rate": 9.103865930603586e-06,
+ "loss": 0.3850820064544678,
+ "num_input_tokens_seen": 251633720,
+ "step": 5750,
+ "train_runtime": 44516.691,
+ "train_tokens_per_second": 5652.57
+ },
+ {
+ "epoch": 0.3874899003501212,
+ "grad_norm": 0.7659015613854286,
+ "learning_rate": 9.102354913784279e-06,
+ "loss": 0.3853774070739746,
+ "num_input_tokens_seen": 251855296,
+ "step": 5755,
+ "train_runtime": 44559.5986,
+ "train_tokens_per_second": 5652.1
+ },
+ {
+ "epoch": 0.38782655534608135,
+ "grad_norm": 0.7966891158323898,
+ "learning_rate": 9.100842749754499e-06,
+ "loss": 0.3578346252441406,
+ "num_input_tokens_seen": 252080400,
+ "step": 5760,
+ "train_runtime": 44605.3404,
+ "train_tokens_per_second": 5651.35
+ },
+ {
+ "epoch": 0.3881632103420415,
+ "grad_norm": 1.3427031718401212,
+ "learning_rate": 9.099329438937122e-06,
+ "loss": 0.40854997634887696,
+ "num_input_tokens_seen": 252296312,
+ "step": 5765,
+ "train_runtime": 44649.458,
+ "train_tokens_per_second": 5650.602
+ },
+ {
+ "epoch": 0.3884998653380016,
+ "grad_norm": 0.6230050104908084,
+ "learning_rate": 9.09781498175534e-06,
+ "loss": 0.3587602138519287,
+ "num_input_tokens_seen": 252509656,
+ "step": 5770,
+ "train_runtime": 44685.4879,
+ "train_tokens_per_second": 5650.82
+ },
+ {
+ "epoch": 0.38883652033396177,
+ "grad_norm": 0.6750080511665485,
+ "learning_rate": 9.096299378632666e-06,
+ "loss": 0.37611684799194334,
+ "num_input_tokens_seen": 252740072,
+ "step": 5775,
+ "train_runtime": 44727.054,
+ "train_tokens_per_second": 5650.72
+ },
+ {
+ "epoch": 0.3891731753299219,
+ "grad_norm": 0.7734243231158888,
+ "learning_rate": 9.09478262999293e-06,
+ "loss": 0.4280585289001465,
+ "num_input_tokens_seen": 252973392,
+ "step": 5780,
+ "train_runtime": 44771.0866,
+ "train_tokens_per_second": 5650.374
+ },
+ {
+ "epoch": 0.38950983032588204,
+ "grad_norm": 0.8173587277273618,
+ "learning_rate": 9.09326473626029e-06,
+ "loss": 0.38141813278198244,
+ "num_input_tokens_seen": 253184320,
+ "step": 5785,
+ "train_runtime": 44809.6538,
+ "train_tokens_per_second": 5650.218
+ },
+ {
+ "epoch": 0.3898464853218422,
+ "grad_norm": 0.7487752402423106,
+ "learning_rate": 9.091745697859218e-06,
+ "loss": 0.41326370239257815,
+ "num_input_tokens_seen": 253411776,
+ "step": 5790,
+ "train_runtime": 44845.5045,
+ "train_tokens_per_second": 5650.773
+ },
+ {
+ "epoch": 0.3901831403178023,
+ "grad_norm": 0.8197406274404395,
+ "learning_rate": 9.090225515214512e-06,
+ "loss": 0.4117950439453125,
+ "num_input_tokens_seen": 253627512,
+ "step": 5795,
+ "train_runtime": 44885.8047,
+ "train_tokens_per_second": 5650.506
+ },
+ {
+ "epoch": 0.39051979531376246,
+ "grad_norm": 0.8857575262398083,
+ "learning_rate": 9.088704188751281e-06,
+ "loss": 0.4129201889038086,
+ "num_input_tokens_seen": 253840400,
+ "step": 5800,
+ "train_runtime": 44924.1283,
+ "train_tokens_per_second": 5650.425
+ },
+ {
+ "epoch": 0.3908564503097226,
+ "grad_norm": 0.85730454063038,
+ "learning_rate": 9.087181718894963e-06,
+ "loss": 0.39529869556427,
+ "num_input_tokens_seen": 254064096,
+ "step": 5805,
+ "train_runtime": 44968.4154,
+ "train_tokens_per_second": 5649.834
+ },
+ {
+ "epoch": 0.39119310530568274,
+ "grad_norm": 0.7118238052198951,
+ "learning_rate": 9.085658106071309e-06,
+ "loss": 0.39313058853149413,
+ "num_input_tokens_seen": 254281632,
+ "step": 5810,
+ "train_runtime": 45005.0615,
+ "train_tokens_per_second": 5650.067
+ },
+ {
+ "epoch": 0.3915297603016429,
+ "grad_norm": 0.7349746949660144,
+ "learning_rate": 9.084133350706394e-06,
+ "loss": 0.3712653398513794,
+ "num_input_tokens_seen": 254493944,
+ "step": 5815,
+ "train_runtime": 45042.1578,
+ "train_tokens_per_second": 5650.128
+ },
+ {
+ "epoch": 0.391866415297603,
+ "grad_norm": 0.7166943198737387,
+ "learning_rate": 9.082607453226613e-06,
+ "loss": 0.38478312492370603,
+ "num_input_tokens_seen": 254715712,
+ "step": 5820,
+ "train_runtime": 45077.3472,
+ "train_tokens_per_second": 5650.637
+ },
+ {
+ "epoch": 0.39220307029356316,
+ "grad_norm": 0.7543487043231831,
+ "learning_rate": 9.081080414058674e-06,
+ "loss": 0.3994553565979004,
+ "num_input_tokens_seen": 254927592,
+ "step": 5825,
+ "train_runtime": 45113.9999,
+ "train_tokens_per_second": 5650.742
+ },
+ {
+ "epoch": 0.3925397252895233,
+ "grad_norm": 0.7896679447300116,
+ "learning_rate": 9.079552233629612e-06,
+ "loss": 0.3710631847381592,
+ "num_input_tokens_seen": 255138464,
+ "step": 5830,
+ "train_runtime": 45152.0089,
+ "train_tokens_per_second": 5650.656
+ },
+ {
+ "epoch": 0.39287638028548344,
+ "grad_norm": 0.815361722369787,
+ "learning_rate": 9.078022912366776e-06,
+ "loss": 0.3914163827896118,
+ "num_input_tokens_seen": 255367048,
+ "step": 5835,
+ "train_runtime": 45192.1775,
+ "train_tokens_per_second": 5650.691
+ },
+ {
+ "epoch": 0.3932130352814436,
+ "grad_norm": 0.8005773109782048,
+ "learning_rate": 9.076492450697835e-06,
+ "loss": 0.40029544830322267,
+ "num_input_tokens_seen": 255598184,
+ "step": 5840,
+ "train_runtime": 45232.9765,
+ "train_tokens_per_second": 5650.704
+ },
+ {
+ "epoch": 0.3935496902774037,
+ "grad_norm": 0.849726633762188,
+ "learning_rate": 9.074960849050782e-06,
+ "loss": 0.3852508068084717,
+ "num_input_tokens_seen": 255819944,
+ "step": 5845,
+ "train_runtime": 45272.8634,
+ "train_tokens_per_second": 5650.624
+ },
+ {
+ "epoch": 0.39388634527336386,
+ "grad_norm": 0.7762826196998844,
+ "learning_rate": 9.07342810785392e-06,
+ "loss": 0.37957100868225097,
+ "num_input_tokens_seen": 256019616,
+ "step": 5850,
+ "train_runtime": 45311.4802,
+ "train_tokens_per_second": 5650.215
+ },
+ {
+ "epoch": 0.394223000269324,
+ "grad_norm": 0.7731972533551551,
+ "learning_rate": 9.071894227535878e-06,
+ "loss": 0.38977823257446287,
+ "num_input_tokens_seen": 256237256,
+ "step": 5855,
+ "train_runtime": 45352.5452,
+ "train_tokens_per_second": 5649.898
+ },
+ {
+ "epoch": 0.39455965526528414,
+ "grad_norm": 0.8423525268519388,
+ "learning_rate": 9.070359208525598e-06,
+ "loss": 0.406312084197998,
+ "num_input_tokens_seen": 256464368,
+ "step": 5860,
+ "train_runtime": 45388.423,
+ "train_tokens_per_second": 5650.436
+ },
+ {
+ "epoch": 0.3948963102612443,
+ "grad_norm": 0.8527879610056384,
+ "learning_rate": 9.068823051252346e-06,
+ "loss": 0.3981436491012573,
+ "num_input_tokens_seen": 256688032,
+ "step": 5865,
+ "train_runtime": 45429.2415,
+ "train_tokens_per_second": 5650.282
+ },
+ {
+ "epoch": 0.3952329652572044,
+ "grad_norm": 0.8385538303650153,
+ "learning_rate": 9.067285756145702e-06,
+ "loss": 0.3921481132507324,
+ "num_input_tokens_seen": 256917840,
+ "step": 5870,
+ "train_runtime": 45462.4848,
+ "train_tokens_per_second": 5651.205
+ },
+ {
+ "epoch": 0.39556962025316456,
+ "grad_norm": 0.8390456887457934,
+ "learning_rate": 9.065747323635565e-06,
+ "loss": 0.3327157020568848,
+ "num_input_tokens_seen": 257101736,
+ "step": 5875,
+ "train_runtime": 45501.4164,
+ "train_tokens_per_second": 5650.412
+ },
+ {
+ "epoch": 0.3959062752491247,
+ "grad_norm": 0.7851013810213687,
+ "learning_rate": 9.064207754152154e-06,
+ "loss": 0.3997276782989502,
+ "num_input_tokens_seen": 257317696,
+ "step": 5880,
+ "train_runtime": 45545.2236,
+ "train_tokens_per_second": 5649.719
+ },
+ {
+ "epoch": 0.39624293024508483,
+ "grad_norm": 0.7365128564544189,
+ "learning_rate": 9.062667048126005e-06,
+ "loss": 0.3864887714385986,
+ "num_input_tokens_seen": 257547392,
+ "step": 5885,
+ "train_runtime": 45589.9837,
+ "train_tokens_per_second": 5649.21
+ },
+ {
+ "epoch": 0.396579585241045,
+ "grad_norm": 0.8935927015959244,
+ "learning_rate": 9.061125205987971e-06,
+ "loss": 0.4157553672790527,
+ "num_input_tokens_seen": 257761408,
+ "step": 5890,
+ "train_runtime": 45630.8437,
+ "train_tokens_per_second": 5648.842
+ },
+ {
+ "epoch": 0.3969162402370051,
+ "grad_norm": 0.7876564044642139,
+ "learning_rate": 9.059582228169222e-06,
+ "loss": 0.37337889671325686,
+ "num_input_tokens_seen": 257963800,
+ "step": 5895,
+ "train_runtime": 45664.7,
+ "train_tokens_per_second": 5649.086
+ },
+ {
+ "epoch": 0.39725289523296525,
+ "grad_norm": 0.7288959690931758,
+ "learning_rate": 9.058038115101248e-06,
+ "loss": 0.4037358283996582,
+ "num_input_tokens_seen": 258195648,
+ "step": 5900,
+ "train_runtime": 45702.5616,
+ "train_tokens_per_second": 5649.479
+ },
+ {
+ "epoch": 0.3975895502289254,
+ "grad_norm": 0.7904683688779008,
+ "learning_rate": 9.056492867215857e-06,
+ "loss": 0.3476010084152222,
+ "num_input_tokens_seen": 258400784,
+ "step": 5905,
+ "train_runtime": 45748.63,
+ "train_tokens_per_second": 5648.274
+ },
+ {
+ "epoch": 0.39792620522488553,
+ "grad_norm": 0.6759275337501953,
+ "learning_rate": 9.05494648494517e-06,
+ "loss": 0.3933530330657959,
+ "num_input_tokens_seen": 258606968,
+ "step": 5910,
+ "train_runtime": 45786.7698,
+ "train_tokens_per_second": 5648.072
+ },
+ {
+ "epoch": 0.39826286022084567,
+ "grad_norm": 0.6825204404349337,
+ "learning_rate": 9.053398968721629e-06,
+ "loss": 0.37357821464538576,
+ "num_input_tokens_seen": 258832840,
+ "step": 5915,
+ "train_runtime": 45827.0425,
+ "train_tokens_per_second": 5648.037
+ },
+ {
+ "epoch": 0.3985995152168058,
+ "grad_norm": 0.7549266235464368,
+ "learning_rate": 9.05185031897799e-06,
+ "loss": 0.4264666080474854,
+ "num_input_tokens_seen": 259049920,
+ "step": 5920,
+ "train_runtime": 45860.4564,
+ "train_tokens_per_second": 5648.656
+ },
+ {
+ "epoch": 0.39893617021276595,
+ "grad_norm": 0.8046151442540065,
+ "learning_rate": 9.050300536147331e-06,
+ "loss": 0.42224712371826173,
+ "num_input_tokens_seen": 259285872,
+ "step": 5925,
+ "train_runtime": 45898.8286,
+ "train_tokens_per_second": 5649.074
+ },
+ {
+ "epoch": 0.3992728252087261,
+ "grad_norm": 0.8013279899857307,
+ "learning_rate": 9.048749620663044e-06,
+ "loss": 0.36397323608398435,
+ "num_input_tokens_seen": 259520536,
+ "step": 5930,
+ "train_runtime": 45935.2903,
+ "train_tokens_per_second": 5649.698
+ },
+ {
+ "epoch": 0.39960948020468623,
+ "grad_norm": 0.7188029194424288,
+ "learning_rate": 9.047197572958834e-06,
+ "loss": 0.37349615097045896,
+ "num_input_tokens_seen": 259742248,
+ "step": 5935,
+ "train_runtime": 45973.4138,
+ "train_tokens_per_second": 5649.836
+ },
+ {
+ "epoch": 0.39994613520064637,
+ "grad_norm": 0.8120677037958034,
+ "learning_rate": 9.04564439346873e-06,
+ "loss": 0.3917673110961914,
+ "num_input_tokens_seen": 259952080,
+ "step": 5940,
+ "train_runtime": 46008.964,
+ "train_tokens_per_second": 5650.031
+ },
+ {
+ "epoch": 0.4002827901966065,
+ "grad_norm": 0.8737926960149425,
+ "learning_rate": 9.044090082627071e-06,
+ "loss": 0.39895708560943605,
+ "num_input_tokens_seen": 260160320,
+ "step": 5945,
+ "train_runtime": 46045.9436,
+ "train_tokens_per_second": 5650.016
+ },
+ {
+ "epoch": 0.40061944519256665,
+ "grad_norm": 0.7302283729471113,
+ "learning_rate": 9.042534640868514e-06,
+ "loss": 0.4021873950958252,
+ "num_input_tokens_seen": 260368008,
+ "step": 5950,
+ "train_runtime": 46081.9071,
+ "train_tokens_per_second": 5650.114
+ },
+ {
+ "epoch": 0.4009561001885268,
+ "grad_norm": 0.749168823407156,
+ "learning_rate": 9.04097806862804e-06,
+ "loss": 0.40214033126831056,
+ "num_input_tokens_seen": 260592824,
+ "step": 5955,
+ "train_runtime": 46117.1368,
+ "train_tokens_per_second": 5650.672
+ },
+ {
+ "epoch": 0.4012927551844869,
+ "grad_norm": 0.723288812952783,
+ "learning_rate": 9.03942036634093e-06,
+ "loss": 0.4055626392364502,
+ "num_input_tokens_seen": 260810288,
+ "step": 5960,
+ "train_runtime": 46156.4921,
+ "train_tokens_per_second": 5650.566
+ },
+ {
+ "epoch": 0.40162941018044707,
+ "grad_norm": 0.7518699289987396,
+ "learning_rate": 9.037861534442797e-06,
+ "loss": 0.3734895706176758,
+ "num_input_tokens_seen": 261038312,
+ "step": 5965,
+ "train_runtime": 46198.2126,
+ "train_tokens_per_second": 5650.399
+ },
+ {
+ "epoch": 0.4019660651764072,
+ "grad_norm": 1.0026827049870985,
+ "learning_rate": 9.036301573369563e-06,
+ "loss": 0.40189208984375,
+ "num_input_tokens_seen": 261266152,
+ "step": 5970,
+ "train_runtime": 46239.4807,
+ "train_tokens_per_second": 5650.283
+ },
+ {
+ "epoch": 0.40230272017236735,
+ "grad_norm": 0.8605436747902855,
+ "learning_rate": 9.034740483557465e-06,
+ "loss": 0.3900024175643921,
+ "num_input_tokens_seen": 261483152,
+ "step": 5975,
+ "train_runtime": 46274.8434,
+ "train_tokens_per_second": 5650.655
+ },
+ {
+ "epoch": 0.4026393751683275,
+ "grad_norm": 0.7213971067668634,
+ "learning_rate": 9.033178265443055e-06,
+ "loss": 0.38559999465942385,
+ "num_input_tokens_seen": 261703416,
+ "step": 5980,
+ "train_runtime": 46306.9931,
+ "train_tokens_per_second": 5651.488
+ },
+ {
+ "epoch": 0.4029760301642876,
+ "grad_norm": 0.7467717729146578,
+ "learning_rate": 9.031614919463206e-06,
+ "loss": 0.3772167682647705,
+ "num_input_tokens_seen": 261926896,
+ "step": 5985,
+ "train_runtime": 46342.877,
+ "train_tokens_per_second": 5651.934
+ },
+ {
+ "epoch": 0.40331268516024776,
+ "grad_norm": 0.8052956550974576,
+ "learning_rate": 9.030050446055099e-06,
+ "loss": 0.3684276342391968,
+ "num_input_tokens_seen": 262143080,
+ "step": 5990,
+ "train_runtime": 46380.2479,
+ "train_tokens_per_second": 5652.041
+ },
+ {
+ "epoch": 0.4036493401562079,
+ "grad_norm": 0.6725874955047775,
+ "learning_rate": 9.028484845656235e-06,
+ "loss": 0.3965585231781006,
+ "num_input_tokens_seen": 262378952,
+ "step": 5995,
+ "train_runtime": 46413.7998,
+ "train_tokens_per_second": 5653.038
+ },
+ {
+ "epoch": 0.40398599515216804,
+ "grad_norm": 0.8065835959924061,
+ "learning_rate": 9.026918118704431e-06,
+ "loss": 0.42124166488647463,
+ "num_input_tokens_seen": 262599696,
+ "step": 6000,
+ "train_runtime": 46454.2053,
+ "train_tokens_per_second": 5652.872
+ },
+ {
+ "epoch": 0.4043226501481282,
+ "grad_norm": 0.6374762319564959,
+ "learning_rate": 9.025350265637816e-06,
+ "loss": 0.3700655221939087,
+ "num_input_tokens_seen": 262830992,
+ "step": 6005,
+ "train_runtime": 46490.7965,
+ "train_tokens_per_second": 5653.398
+ },
+ {
+ "epoch": 0.4046593051440883,
+ "grad_norm": 0.7573194160968504,
+ "learning_rate": 9.023781286894834e-06,
+ "loss": 0.38486456871032715,
+ "num_input_tokens_seen": 263044288,
+ "step": 6010,
+ "train_runtime": 46526.3581,
+ "train_tokens_per_second": 5653.662
+ },
+ {
+ "epoch": 0.40499596014004846,
+ "grad_norm": 0.8562717514351039,
+ "learning_rate": 9.022211182914247e-06,
+ "loss": 0.3843253135681152,
+ "num_input_tokens_seen": 263264800,
+ "step": 6015,
+ "train_runtime": 46564.9415,
+ "train_tokens_per_second": 5653.713
+ },
+ {
+ "epoch": 0.4053326151360086,
+ "grad_norm": 0.7322028797978756,
+ "learning_rate": 9.020639954135128e-06,
+ "loss": 0.3804231405258179,
+ "num_input_tokens_seen": 263476880,
+ "step": 6020,
+ "train_runtime": 46608.1754,
+ "train_tokens_per_second": 5653.019
+ },
+ {
+ "epoch": 0.40566927013196874,
+ "grad_norm": 0.7924756161102612,
+ "learning_rate": 9.019067600996867e-06,
+ "loss": 0.32609758377075193,
+ "num_input_tokens_seen": 263676440,
+ "step": 6025,
+ "train_runtime": 46647.9449,
+ "train_tokens_per_second": 5652.477
+ },
+ {
+ "epoch": 0.4060059251279289,
+ "grad_norm": 0.7406207884260587,
+ "learning_rate": 9.017494123939169e-06,
+ "loss": 0.3878431558609009,
+ "num_input_tokens_seen": 263891832,
+ "step": 6030,
+ "train_runtime": 46686.4579,
+ "train_tokens_per_second": 5652.428
+ },
+ {
+ "epoch": 0.406342580123889,
+ "grad_norm": 0.6859816815928503,
+ "learning_rate": 9.015919523402048e-06,
+ "loss": 0.38746066093444825,
+ "num_input_tokens_seen": 264108024,
+ "step": 6035,
+ "train_runtime": 46723.9959,
+ "train_tokens_per_second": 5652.514
+ },
+ {
+ "epoch": 0.40667923511984916,
+ "grad_norm": 1.0398516753802722,
+ "learning_rate": 9.014343799825838e-06,
+ "loss": 0.3903219699859619,
+ "num_input_tokens_seen": 264311560,
+ "step": 6040,
+ "train_runtime": 46758.4333,
+ "train_tokens_per_second": 5652.704
+ },
+ {
+ "epoch": 0.4070158901158093,
+ "grad_norm": 0.6954397754211817,
+ "learning_rate": 9.012766953651186e-06,
+ "loss": 0.42699203491210935,
+ "num_input_tokens_seen": 264536928,
+ "step": 6045,
+ "train_runtime": 46800.6106,
+ "train_tokens_per_second": 5652.425
+ },
+ {
+ "epoch": 0.40735254511176944,
+ "grad_norm": 0.6848650266078808,
+ "learning_rate": 9.01118898531905e-06,
+ "loss": 0.4040683746337891,
+ "num_input_tokens_seen": 264744560,
+ "step": 6050,
+ "train_runtime": 46841.7035,
+ "train_tokens_per_second": 5651.899
+ },
+ {
+ "epoch": 0.4076892001077296,
+ "grad_norm": 0.7157357093690987,
+ "learning_rate": 9.009609895270708e-06,
+ "loss": 0.3655755043029785,
+ "num_input_tokens_seen": 264973096,
+ "step": 6055,
+ "train_runtime": 46881.3659,
+ "train_tokens_per_second": 5651.992
+ },
+ {
+ "epoch": 0.4080258551036897,
+ "grad_norm": 0.8098322622781278,
+ "learning_rate": 9.008029683947743e-06,
+ "loss": 0.42575840950012206,
+ "num_input_tokens_seen": 265197544,
+ "step": 6060,
+ "train_runtime": 46917.7458,
+ "train_tokens_per_second": 5652.393
+ },
+ {
+ "epoch": 0.40836251009964986,
+ "grad_norm": 0.7104130729995994,
+ "learning_rate": 9.006448351792057e-06,
+ "loss": 0.3974879741668701,
+ "num_input_tokens_seen": 265418944,
+ "step": 6065,
+ "train_runtime": 46962.89,
+ "train_tokens_per_second": 5651.674
+ },
+ {
+ "epoch": 0.40869916509561,
+ "grad_norm": 0.6632525596263023,
+ "learning_rate": 9.004865899245864e-06,
+ "loss": 0.38395133018493655,
+ "num_input_tokens_seen": 265636776,
+ "step": 6070,
+ "train_runtime": 46999.967,
+ "train_tokens_per_second": 5651.85
+ },
+ {
+ "epoch": 0.40903582009157013,
+ "grad_norm": 0.7043912136666165,
+ "learning_rate": 9.003282326751694e-06,
+ "loss": 0.4008152961730957,
+ "num_input_tokens_seen": 265857800,
+ "step": 6075,
+ "train_runtime": 47039.3368,
+ "train_tokens_per_second": 5651.819
+ },
+ {
+ "epoch": 0.4093724750875303,
+ "grad_norm": 0.9655541200247993,
+ "learning_rate": 9.001697634752387e-06,
+ "loss": 0.40933728218078613,
+ "num_input_tokens_seen": 266070448,
+ "step": 6080,
+ "train_runtime": 47074.6096,
+ "train_tokens_per_second": 5652.101
+ },
+ {
+ "epoch": 0.4097091300834904,
+ "grad_norm": 0.683185307022826,
+ "learning_rate": 9.000111823691097e-06,
+ "loss": 0.37639498710632324,
+ "num_input_tokens_seen": 266291816,
+ "step": 6085,
+ "train_runtime": 47112.3669,
+ "train_tokens_per_second": 5652.27
+ },
+ {
+ "epoch": 0.41004578507945055,
+ "grad_norm": 0.8219233379550768,
+ "learning_rate": 8.99852489401129e-06,
+ "loss": 0.40433249473571775,
+ "num_input_tokens_seen": 266517904,
+ "step": 6090,
+ "train_runtime": 47148.2501,
+ "train_tokens_per_second": 5652.763
+ },
+ {
+ "epoch": 0.4103824400754107,
+ "grad_norm": 0.7423675573693415,
+ "learning_rate": 8.996936846156748e-06,
+ "loss": 0.3635870456695557,
+ "num_input_tokens_seen": 266736992,
+ "step": 6095,
+ "train_runtime": 47183.3361,
+ "train_tokens_per_second": 5653.203
+ },
+ {
+ "epoch": 0.41071909507137083,
+ "grad_norm": 0.8008363109798308,
+ "learning_rate": 8.995347680571563e-06,
+ "loss": 0.41033496856689455,
+ "num_input_tokens_seen": 266959272,
+ "step": 6100,
+ "train_runtime": 47217.625,
+ "train_tokens_per_second": 5653.806
+ },
+ {
+ "epoch": 0.41105575006733097,
+ "grad_norm": 0.6633014625775752,
+ "learning_rate": 8.993757397700137e-06,
+ "loss": 0.38593282699584963,
+ "num_input_tokens_seen": 267178696,
+ "step": 6105,
+ "train_runtime": 47254.1693,
+ "train_tokens_per_second": 5654.077
+ },
+ {
+ "epoch": 0.41139240506329117,
+ "grad_norm": 0.7915805562051228,
+ "learning_rate": 8.99216599798719e-06,
+ "loss": 0.42214293479919435,
+ "num_input_tokens_seen": 267415152,
+ "step": 6110,
+ "train_runtime": 47290.6934,
+ "train_tokens_per_second": 5654.71
+ },
+ {
+ "epoch": 0.4117290600592513,
+ "grad_norm": 0.7917765167657784,
+ "learning_rate": 8.990573481877753e-06,
+ "loss": 0.3639373779296875,
+ "num_input_tokens_seen": 267618552,
+ "step": 6115,
+ "train_runtime": 47325.6204,
+ "train_tokens_per_second": 5654.835
+ },
+ {
+ "epoch": 0.41206571505521145,
+ "grad_norm": 0.694124505400035,
+ "learning_rate": 8.988979849817167e-06,
+ "loss": 0.36659536361694334,
+ "num_input_tokens_seen": 267814904,
+ "step": 6120,
+ "train_runtime": 47366.4119,
+ "train_tokens_per_second": 5654.11
+ },
+ {
+ "epoch": 0.4124023700511716,
+ "grad_norm": 0.8642974554518021,
+ "learning_rate": 8.987385102251085e-06,
+ "loss": 0.3810924291610718,
+ "num_input_tokens_seen": 268015432,
+ "step": 6125,
+ "train_runtime": 47405.0635,
+ "train_tokens_per_second": 5653.73
+ },
+ {
+ "epoch": 0.4127390250471317,
+ "grad_norm": 0.7416982400357927,
+ "learning_rate": 8.985789239625475e-06,
+ "loss": 0.3897581100463867,
+ "num_input_tokens_seen": 268235696,
+ "step": 6130,
+ "train_runtime": 47449.5896,
+ "train_tokens_per_second": 5653.067
+ },
+ {
+ "epoch": 0.41307568004309186,
+ "grad_norm": 0.7586989342588731,
+ "learning_rate": 8.984192262386613e-06,
+ "loss": 0.36377105712890623,
+ "num_input_tokens_seen": 268454176,
+ "step": 6135,
+ "train_runtime": 47483.2101,
+ "train_tokens_per_second": 5653.665
+ },
+ {
+ "epoch": 0.413412335039052,
+ "grad_norm": 0.7553176125857815,
+ "learning_rate": 8.98259417098109e-06,
+ "loss": 0.3994285583496094,
+ "num_input_tokens_seen": 268685424,
+ "step": 6140,
+ "train_runtime": 47523.5954,
+ "train_tokens_per_second": 5653.727
+ },
+ {
+ "epoch": 0.41374899003501214,
+ "grad_norm": 0.7619380388493862,
+ "learning_rate": 8.980994965855808e-06,
+ "loss": 0.39436161518096924,
+ "num_input_tokens_seen": 268906160,
+ "step": 6145,
+ "train_runtime": 47557.2625,
+ "train_tokens_per_second": 5654.366
+ },
+ {
+ "epoch": 0.4140856450309723,
+ "grad_norm": 0.7194190068295008,
+ "learning_rate": 8.979394647457976e-06,
+ "loss": 0.3927819490432739,
+ "num_input_tokens_seen": 269128016,
+ "step": 6150,
+ "train_runtime": 47600.5297,
+ "train_tokens_per_second": 5653.887
+ },
+ {
+ "epoch": 0.4144223000269324,
+ "grad_norm": 0.741178107175373,
+ "learning_rate": 8.97779321623512e-06,
+ "loss": 0.35940072536468504,
+ "num_input_tokens_seen": 269333048,
+ "step": 6155,
+ "train_runtime": 47638.7497,
+ "train_tokens_per_second": 5653.655
+ },
+ {
+ "epoch": 0.41475895502289256,
+ "grad_norm": 0.7334172734354966,
+ "learning_rate": 8.976190672635077e-06,
+ "loss": 0.400145435333252,
+ "num_input_tokens_seen": 269557248,
+ "step": 6160,
+ "train_runtime": 47679.304,
+ "train_tokens_per_second": 5653.548
+ },
+ {
+ "epoch": 0.4150956100188527,
+ "grad_norm": 0.7612088304227586,
+ "learning_rate": 8.974587017105991e-06,
+ "loss": 0.41222319602966306,
+ "num_input_tokens_seen": 269768408,
+ "step": 6165,
+ "train_runtime": 47713.1048,
+ "train_tokens_per_second": 5653.969
+ },
+ {
+ "epoch": 0.41543226501481284,
+ "grad_norm": 0.7344591815245108,
+ "learning_rate": 8.97298225009632e-06,
+ "loss": 0.3931986570358276,
+ "num_input_tokens_seen": 269990368,
+ "step": 6170,
+ "train_runtime": 47754.3241,
+ "train_tokens_per_second": 5653.737
+ },
+ {
+ "epoch": 0.415768920010773,
+ "grad_norm": 0.7575071698285096,
+ "learning_rate": 8.971376372054829e-06,
+ "loss": 0.3943799495697021,
+ "num_input_tokens_seen": 270208056,
+ "step": 6175,
+ "train_runtime": 47794.7334,
+ "train_tokens_per_second": 5653.511
+ },
+ {
+ "epoch": 0.4161055750067331,
+ "grad_norm": 0.6823047182780082,
+ "learning_rate": 8.969769383430602e-06,
+ "loss": 0.3858570337295532,
+ "num_input_tokens_seen": 270432152,
+ "step": 6180,
+ "train_runtime": 47829.7321,
+ "train_tokens_per_second": 5654.06
+ },
+ {
+ "epoch": 0.41644223000269326,
+ "grad_norm": 0.7476934311068052,
+ "learning_rate": 8.968161284673027e-06,
+ "loss": 0.37152729034423826,
+ "num_input_tokens_seen": 270661216,
+ "step": 6185,
+ "train_runtime": 47868.2088,
+ "train_tokens_per_second": 5654.3
+ },
+ {
+ "epoch": 0.4167788849986534,
+ "grad_norm": 0.7528325409155248,
+ "learning_rate": 8.9665520762318e-06,
+ "loss": 0.37863969802856445,
+ "num_input_tokens_seen": 270875256,
+ "step": 6190,
+ "train_runtime": 47904.4792,
+ "train_tokens_per_second": 5654.487
+ },
+ {
+ "epoch": 0.41711553999461354,
+ "grad_norm": 0.7282239751820792,
+ "learning_rate": 8.964941758556934e-06,
+ "loss": 0.3976884126663208,
+ "num_input_tokens_seen": 271088288,
+ "step": 6195,
+ "train_runtime": 47948.7488,
+ "train_tokens_per_second": 5653.709
+ },
+ {
+ "epoch": 0.4174521949905737,
+ "grad_norm": 0.755222562109598,
+ "learning_rate": 8.963330332098747e-06,
+ "loss": 0.3814378261566162,
+ "num_input_tokens_seen": 271296528,
+ "step": 6200,
+ "train_runtime": 47985.6256,
+ "train_tokens_per_second": 5653.704
+ },
+ {
+ "epoch": 0.4177888499865338,
+ "grad_norm": 0.7830273045211564,
+ "learning_rate": 8.961717797307872e-06,
+ "loss": 0.37258970737457275,
+ "num_input_tokens_seen": 271515568,
+ "step": 6205,
+ "train_runtime": 48019.4934,
+ "train_tokens_per_second": 5654.278
+ },
+ {
+ "epoch": 0.41812550498249396,
+ "grad_norm": 0.7683395292659414,
+ "learning_rate": 8.960104154635248e-06,
+ "loss": 0.4039943695068359,
+ "num_input_tokens_seen": 271733704,
+ "step": 6210,
+ "train_runtime": 48057.6768,
+ "train_tokens_per_second": 5654.325
+ },
+ {
+ "epoch": 0.4184621599784541,
+ "grad_norm": 0.7964266950242961,
+ "learning_rate": 8.958489404532125e-06,
+ "loss": 0.4006314277648926,
+ "num_input_tokens_seen": 271941040,
+ "step": 6215,
+ "train_runtime": 48094.9785,
+ "train_tokens_per_second": 5654.25
+ },
+ {
+ "epoch": 0.41879881497441424,
+ "grad_norm": 0.7229961706218653,
+ "learning_rate": 8.956873547450062e-06,
+ "loss": 0.3996126651763916,
+ "num_input_tokens_seen": 272155560,
+ "step": 6220,
+ "train_runtime": 48135.1561,
+ "train_tokens_per_second": 5653.987
+ },
+ {
+ "epoch": 0.4191354699703744,
+ "grad_norm": 0.8692942989729212,
+ "learning_rate": 8.95525658384093e-06,
+ "loss": 0.4290035247802734,
+ "num_input_tokens_seen": 272369192,
+ "step": 6225,
+ "train_runtime": 48171.8576,
+ "train_tokens_per_second": 5654.114
+ },
+ {
+ "epoch": 0.4194721249663345,
+ "grad_norm": 0.7600990815023999,
+ "learning_rate": 8.953638514156908e-06,
+ "loss": 0.3810266494750977,
+ "num_input_tokens_seen": 272591248,
+ "step": 6230,
+ "train_runtime": 48208.9238,
+ "train_tokens_per_second": 5654.373
+ },
+ {
+ "epoch": 0.41980877996229465,
+ "grad_norm": 0.7355727181513495,
+ "learning_rate": 8.952019338850481e-06,
+ "loss": 0.3950758695602417,
+ "num_input_tokens_seen": 272810616,
+ "step": 6235,
+ "train_runtime": 48248.4869,
+ "train_tokens_per_second": 5654.283
+ },
+ {
+ "epoch": 0.4201454349582548,
+ "grad_norm": 0.7644200111226909,
+ "learning_rate": 8.950399058374447e-06,
+ "loss": 0.377358603477478,
+ "num_input_tokens_seen": 273036976,
+ "step": 6240,
+ "train_runtime": 48285.7901,
+ "train_tokens_per_second": 5654.603
+ },
+ {
+ "epoch": 0.42048208995421493,
+ "grad_norm": 0.7918087336685585,
+ "learning_rate": 8.948777673181913e-06,
+ "loss": 0.389499831199646,
+ "num_input_tokens_seen": 273260832,
+ "step": 6245,
+ "train_runtime": 48322.4015,
+ "train_tokens_per_second": 5654.951
+ },
+ {
+ "epoch": 0.42081874495017507,
+ "grad_norm": 0.7765030452628453,
+ "learning_rate": 8.947155183726296e-06,
+ "loss": 0.35937402248382566,
+ "num_input_tokens_seen": 273488360,
+ "step": 6250,
+ "train_runtime": 48361.4271,
+ "train_tokens_per_second": 5655.093
+ },
+ {
+ "epoch": 0.4211553999461352,
+ "grad_norm": 0.7014097499994082,
+ "learning_rate": 8.94553159046132e-06,
+ "loss": 0.3715068817138672,
+ "num_input_tokens_seen": 273707472,
+ "step": 6255,
+ "train_runtime": 48401.1614,
+ "train_tokens_per_second": 5654.977
+ },
+ {
+ "epoch": 0.42149205494209535,
+ "grad_norm": 0.6408398788945804,
+ "learning_rate": 8.943906893841015e-06,
+ "loss": 0.36748714447021485,
+ "num_input_tokens_seen": 273928544,
+ "step": 6260,
+ "train_runtime": 48433.3616,
+ "train_tokens_per_second": 5655.782
+ },
+ {
+ "epoch": 0.4218287099380555,
+ "grad_norm": 0.754509294899034,
+ "learning_rate": 8.942281094319721e-06,
+ "loss": 0.4157464027404785,
+ "num_input_tokens_seen": 274141592,
+ "step": 6265,
+ "train_runtime": 48468.7848,
+ "train_tokens_per_second": 5656.044
+ },
+ {
+ "epoch": 0.42216536493401563,
+ "grad_norm": 0.7840590763768849,
+ "learning_rate": 8.940654192352092e-06,
+ "loss": 0.40111403465270995,
+ "num_input_tokens_seen": 274349792,
+ "step": 6270,
+ "train_runtime": 48510.3168,
+ "train_tokens_per_second": 5655.494
+ },
+ {
+ "epoch": 0.42250201992997577,
+ "grad_norm": 0.8213882192132025,
+ "learning_rate": 8.939026188393083e-06,
+ "loss": 0.3853888034820557,
+ "num_input_tokens_seen": 274569104,
+ "step": 6275,
+ "train_runtime": 48556.531,
+ "train_tokens_per_second": 5654.628
+ },
+ {
+ "epoch": 0.4228386749259359,
+ "grad_norm": 0.8791839426260795,
+ "learning_rate": 8.937397082897963e-06,
+ "loss": 0.40306596755981444,
+ "num_input_tokens_seen": 274796616,
+ "step": 6280,
+ "train_runtime": 48596.7026,
+ "train_tokens_per_second": 5654.635
+ },
+ {
+ "epoch": 0.42317532992189605,
+ "grad_norm": 0.7843983011163462,
+ "learning_rate": 8.935766876322303e-06,
+ "loss": 0.4140591621398926,
+ "num_input_tokens_seen": 275015624,
+ "step": 6285,
+ "train_runtime": 48635.2618,
+ "train_tokens_per_second": 5654.655
+ },
+ {
+ "epoch": 0.4235119849178562,
+ "grad_norm": 0.7007974546940894,
+ "learning_rate": 8.934135569121985e-06,
+ "loss": 0.3959988832473755,
+ "num_input_tokens_seen": 275242840,
+ "step": 6290,
+ "train_runtime": 48673.0052,
+ "train_tokens_per_second": 5654.938
+ },
+ {
+ "epoch": 0.4238486399138163,
+ "grad_norm": 0.7078146309363205,
+ "learning_rate": 8.932503161753202e-06,
+ "loss": 0.38456072807312014,
+ "num_input_tokens_seen": 275462504,
+ "step": 6295,
+ "train_runtime": 48709.4026,
+ "train_tokens_per_second": 5655.222
+ },
+ {
+ "epoch": 0.42418529490977647,
+ "grad_norm": 1.0351275568049705,
+ "learning_rate": 8.930869654672449e-06,
+ "loss": 0.38416409492492676,
+ "num_input_tokens_seen": 275681808,
+ "step": 6300,
+ "train_runtime": 48747.6893,
+ "train_tokens_per_second": 5655.279
+ },
+ {
+ "epoch": 0.4245219499057366,
+ "grad_norm": 0.8236989821133165,
+ "learning_rate": 8.929235048336531e-06,
+ "loss": 0.37823896408081054,
+ "num_input_tokens_seen": 275903576,
+ "step": 6305,
+ "train_runtime": 48788.5888,
+ "train_tokens_per_second": 5655.084
+ },
+ {
+ "epoch": 0.42485860490169675,
+ "grad_norm": 0.8519274837856372,
+ "learning_rate": 8.927599343202563e-06,
+ "loss": 0.4197996139526367,
+ "num_input_tokens_seen": 276132792,
+ "step": 6310,
+ "train_runtime": 48821.7451,
+ "train_tokens_per_second": 5655.939
+ },
+ {
+ "epoch": 0.4251952598976569,
+ "grad_norm": 0.7601906981584227,
+ "learning_rate": 8.92596253972796e-06,
+ "loss": 0.39647321701049804,
+ "num_input_tokens_seen": 276373576,
+ "step": 6315,
+ "train_runtime": 48856.0141,
+ "train_tokens_per_second": 5656.9
+ },
+ {
+ "epoch": 0.425531914893617,
+ "grad_norm": 0.7436129276391935,
+ "learning_rate": 8.924324638370454e-06,
+ "loss": 0.4159084320068359,
+ "num_input_tokens_seen": 276601344,
+ "step": 6320,
+ "train_runtime": 48892.2285,
+ "train_tokens_per_second": 5657.368
+ },
+ {
+ "epoch": 0.42586856988957716,
+ "grad_norm": 0.619825666948129,
+ "learning_rate": 8.922685639588076e-06,
+ "loss": 0.3394161701202393,
+ "num_input_tokens_seen": 276801216,
+ "step": 6325,
+ "train_runtime": 48929.4586,
+ "train_tokens_per_second": 5657.149
+ },
+ {
+ "epoch": 0.4262052248855373,
+ "grad_norm": 0.7566990752584487,
+ "learning_rate": 8.921045543839167e-06,
+ "loss": 0.412354850769043,
+ "num_input_tokens_seen": 277025984,
+ "step": 6330,
+ "train_runtime": 48964.4429,
+ "train_tokens_per_second": 5657.697
+ },
+ {
+ "epoch": 0.42654187988149744,
+ "grad_norm": 0.7706529692481164,
+ "learning_rate": 8.919404351582376e-06,
+ "loss": 0.37722196578979494,
+ "num_input_tokens_seen": 277234256,
+ "step": 6335,
+ "train_runtime": 49000.4277,
+ "train_tokens_per_second": 5657.793
+ },
+ {
+ "epoch": 0.4268785348774576,
+ "grad_norm": 0.7056105235159689,
+ "learning_rate": 8.917762063276657e-06,
+ "loss": 0.38721909523010256,
+ "num_input_tokens_seen": 277448936,
+ "step": 6340,
+ "train_runtime": 49039.3589,
+ "train_tokens_per_second": 5657.679
+ },
+ {
+ "epoch": 0.4272151898734177,
+ "grad_norm": 0.7187047540510013,
+ "learning_rate": 8.916118679381268e-06,
+ "loss": 0.3728447198867798,
+ "num_input_tokens_seen": 277667920,
+ "step": 6345,
+ "train_runtime": 49076.9224,
+ "train_tokens_per_second": 5657.81
+ },
+ {
+ "epoch": 0.42755184486937786,
+ "grad_norm": 0.8958354766388671,
+ "learning_rate": 8.914474200355777e-06,
+ "loss": 0.39159350395202636,
+ "num_input_tokens_seen": 277885288,
+ "step": 6350,
+ "train_runtime": 49121.5781,
+ "train_tokens_per_second": 5657.092
+ },
+ {
+ "epoch": 0.427888499865338,
+ "grad_norm": 0.8411282788940809,
+ "learning_rate": 8.912828626660059e-06,
+ "loss": 0.4010343074798584,
+ "num_input_tokens_seen": 278112192,
+ "step": 6355,
+ "train_runtime": 49161.2354,
+ "train_tokens_per_second": 5657.144
+ },
+ {
+ "epoch": 0.42822515486129814,
+ "grad_norm": 0.7895834565221768,
+ "learning_rate": 8.911181958754294e-06,
+ "loss": 0.4207916259765625,
+ "num_input_tokens_seen": 278332088,
+ "step": 6360,
+ "train_runtime": 49198.4506,
+ "train_tokens_per_second": 5657.334
+ },
+ {
+ "epoch": 0.4285618098572583,
+ "grad_norm": 0.6974429090218124,
+ "learning_rate": 8.909534197098966e-06,
+ "loss": 0.4089661121368408,
+ "num_input_tokens_seen": 278560320,
+ "step": 6365,
+ "train_runtime": 49240.3447,
+ "train_tokens_per_second": 5657.156
+ },
+ {
+ "epoch": 0.4288984648532184,
+ "grad_norm": 0.7879065723514492,
+ "learning_rate": 8.907885342154865e-06,
+ "loss": 0.3871757507324219,
+ "num_input_tokens_seen": 278780408,
+ "step": 6370,
+ "train_runtime": 49273.5524,
+ "train_tokens_per_second": 5657.81
+ },
+ {
+ "epoch": 0.42923511984917856,
+ "grad_norm": 0.809098238394157,
+ "learning_rate": 8.906235394383089e-06,
+ "loss": 0.3723259925842285,
+ "num_input_tokens_seen": 278977576,
+ "step": 6375,
+ "train_runtime": 49314.5399,
+ "train_tokens_per_second": 5657.106
+ },
+ {
+ "epoch": 0.4295717748451387,
+ "grad_norm": 0.7788656469697295,
+ "learning_rate": 8.904584354245042e-06,
+ "loss": 0.38217949867248535,
+ "num_input_tokens_seen": 279198056,
+ "step": 6380,
+ "train_runtime": 49349.1158,
+ "train_tokens_per_second": 5657.61
+ },
+ {
+ "epoch": 0.42990842984109884,
+ "grad_norm": 0.7480156005511179,
+ "learning_rate": 8.902932222202433e-06,
+ "loss": 0.39810030460357665,
+ "num_input_tokens_seen": 279431128,
+ "step": 6385,
+ "train_runtime": 49390.7865,
+ "train_tokens_per_second": 5657.556
+ },
+ {
+ "epoch": 0.430245084837059,
+ "grad_norm": 0.7111926102410158,
+ "learning_rate": 8.901278998717272e-06,
+ "loss": 0.37800168991088867,
+ "num_input_tokens_seen": 279655568,
+ "step": 6390,
+ "train_runtime": 49420.9322,
+ "train_tokens_per_second": 5658.646
+ },
+ {
+ "epoch": 0.4305817398330191,
+ "grad_norm": 0.8127379256774018,
+ "learning_rate": 8.899624684251878e-06,
+ "loss": 0.38466432094573977,
+ "num_input_tokens_seen": 279872920,
+ "step": 6395,
+ "train_runtime": 49461.0295,
+ "train_tokens_per_second": 5658.453
+ },
+ {
+ "epoch": 0.43091839482897926,
+ "grad_norm": 0.7440397221517976,
+ "learning_rate": 8.897969279268877e-06,
+ "loss": 0.37961535453796386,
+ "num_input_tokens_seen": 280093688,
+ "step": 6400,
+ "train_runtime": 49494.6401,
+ "train_tokens_per_second": 5659.071
+ },
+ {
+ "epoch": 0.4312550498249394,
+ "grad_norm": 0.7095617451331785,
+ "learning_rate": 8.896312784231196e-06,
+ "loss": 0.3712193012237549,
+ "num_input_tokens_seen": 280305104,
+ "step": 6405,
+ "train_runtime": 49529.3102,
+ "train_tokens_per_second": 5659.378
+ },
+ {
+ "epoch": 0.43159170482089954,
+ "grad_norm": 0.8249055107614544,
+ "learning_rate": 8.89465519960207e-06,
+ "loss": 0.42307000160217284,
+ "num_input_tokens_seen": 280534896,
+ "step": 6410,
+ "train_runtime": 49562.7367,
+ "train_tokens_per_second": 5660.198
+ },
+ {
+ "epoch": 0.4319283598168597,
+ "grad_norm": 0.7044078667504021,
+ "learning_rate": 8.892996525845037e-06,
+ "loss": 0.39340283870697024,
+ "num_input_tokens_seen": 280760536,
+ "step": 6415,
+ "train_runtime": 49609.124,
+ "train_tokens_per_second": 5659.454
+ },
+ {
+ "epoch": 0.4322650148128198,
+ "grad_norm": 0.8758617954616612,
+ "learning_rate": 8.89133676342394e-06,
+ "loss": 0.4034512996673584,
+ "num_input_tokens_seen": 280959600,
+ "step": 6420,
+ "train_runtime": 49647.7511,
+ "train_tokens_per_second": 5659.06
+ },
+ {
+ "epoch": 0.43260166980877995,
+ "grad_norm": 0.7612824097484856,
+ "learning_rate": 8.889675912802923e-06,
+ "loss": 0.38754520416259763,
+ "num_input_tokens_seen": 281173760,
+ "step": 6425,
+ "train_runtime": 49684.4958,
+ "train_tokens_per_second": 5659.185
+ },
+ {
+ "epoch": 0.4329383248047401,
+ "grad_norm": 0.8243557630186539,
+ "learning_rate": 8.888013974446443e-06,
+ "loss": 0.4189424514770508,
+ "num_input_tokens_seen": 281395808,
+ "step": 6430,
+ "train_runtime": 49725.5827,
+ "train_tokens_per_second": 5658.975
+ },
+ {
+ "epoch": 0.43327497980070023,
+ "grad_norm": 0.8439524750777101,
+ "learning_rate": 8.886350948819253e-06,
+ "loss": 0.38129727840423583,
+ "num_input_tokens_seen": 281606072,
+ "step": 6435,
+ "train_runtime": 49758.4427,
+ "train_tokens_per_second": 5659.463
+ },
+ {
+ "epoch": 0.4336116347966604,
+ "grad_norm": 0.8007298009088516,
+ "learning_rate": 8.884686836386412e-06,
+ "loss": 0.38291594982147215,
+ "num_input_tokens_seen": 281822424,
+ "step": 6440,
+ "train_runtime": 49799.792,
+ "train_tokens_per_second": 5659.108
+ },
+ {
+ "epoch": 0.4339482897926205,
+ "grad_norm": 0.7881388654599728,
+ "learning_rate": 8.883021637613286e-06,
+ "loss": 0.4050453662872314,
+ "num_input_tokens_seen": 282049648,
+ "step": 6445,
+ "train_runtime": 49835.0256,
+ "train_tokens_per_second": 5659.667
+ },
+ {
+ "epoch": 0.43428494478858065,
+ "grad_norm": 0.6887341883009596,
+ "learning_rate": 8.881355352965537e-06,
+ "loss": 0.38397510051727296,
+ "num_input_tokens_seen": 282264512,
+ "step": 6450,
+ "train_runtime": 49867.6041,
+ "train_tokens_per_second": 5660.278
+ },
+ {
+ "epoch": 0.4346215997845408,
+ "grad_norm": 0.672101432469643,
+ "learning_rate": 8.879687982909145e-06,
+ "loss": 0.3574419736862183,
+ "num_input_tokens_seen": 282465912,
+ "step": 6455,
+ "train_runtime": 49905.6117,
+ "train_tokens_per_second": 5660.003
+ },
+ {
+ "epoch": 0.43495825478050093,
+ "grad_norm": 0.8916903644937464,
+ "learning_rate": 8.878019527910378e-06,
+ "loss": 0.41092991828918457,
+ "num_input_tokens_seen": 282670240,
+ "step": 6460,
+ "train_runtime": 49945.0146,
+ "train_tokens_per_second": 5659.629
+ },
+ {
+ "epoch": 0.43529490977646107,
+ "grad_norm": 0.8298316229675538,
+ "learning_rate": 8.876349988435815e-06,
+ "loss": 0.44412593841552733,
+ "num_input_tokens_seen": 282887856,
+ "step": 6465,
+ "train_runtime": 49983.0629,
+ "train_tokens_per_second": 5659.674
+ },
+ {
+ "epoch": 0.4356315647724212,
+ "grad_norm": 0.6431274180779993,
+ "learning_rate": 8.874679364952339e-06,
+ "loss": 0.38706722259521487,
+ "num_input_tokens_seen": 283127040,
+ "step": 6470,
+ "train_runtime": 50021.3951,
+ "train_tokens_per_second": 5660.119
+ },
+ {
+ "epoch": 0.43596821976838135,
+ "grad_norm": 0.7656387607672229,
+ "learning_rate": 8.873007657927135e-06,
+ "loss": 0.3893885135650635,
+ "num_input_tokens_seen": 283357232,
+ "step": 6475,
+ "train_runtime": 50062.5333,
+ "train_tokens_per_second": 5660.066
+ },
+ {
+ "epoch": 0.4363048747643415,
+ "grad_norm": 0.7096570879969786,
+ "learning_rate": 8.87133486782769e-06,
+ "loss": 0.35058715343475344,
+ "num_input_tokens_seen": 283558232,
+ "step": 6480,
+ "train_runtime": 50096.5148,
+ "train_tokens_per_second": 5660.239
+ },
+ {
+ "epoch": 0.4366415297603016,
+ "grad_norm": 0.8761505694419173,
+ "learning_rate": 8.869660995121792e-06,
+ "loss": 0.3999923229217529,
+ "num_input_tokens_seen": 283777800,
+ "step": 6485,
+ "train_runtime": 50131.8591,
+ "train_tokens_per_second": 5660.628
+ },
+ {
+ "epoch": 0.43697818475626177,
+ "grad_norm": 0.7555798070151492,
+ "learning_rate": 8.867986040277537e-06,
+ "loss": 0.4016421318054199,
+ "num_input_tokens_seen": 283985824,
+ "step": 6490,
+ "train_runtime": 50172.9,
+ "train_tokens_per_second": 5660.144
+ },
+ {
+ "epoch": 0.4373148397522219,
+ "grad_norm": 0.7001554230524034,
+ "learning_rate": 8.86631000376332e-06,
+ "loss": 0.4068891525268555,
+ "num_input_tokens_seen": 284193496,
+ "step": 6495,
+ "train_runtime": 50210.1755,
+ "train_tokens_per_second": 5660.078
+ },
+ {
+ "epoch": 0.43765149474818205,
+ "grad_norm": 0.809001796546686,
+ "learning_rate": 8.864632886047841e-06,
+ "loss": 0.3719168186187744,
+ "num_input_tokens_seen": 284406672,
+ "step": 6500,
+ "train_runtime": 50246.2209,
+ "train_tokens_per_second": 5660.26
+ },
+ {
+ "epoch": 0.4379881497441422,
+ "grad_norm": 0.8525357252115519,
+ "learning_rate": 8.862954687600097e-06,
+ "loss": 0.4161073684692383,
+ "num_input_tokens_seen": 284620064,
+ "step": 6505,
+ "train_runtime": 50283.9357,
+ "train_tokens_per_second": 5660.258
+ },
+ {
+ "epoch": 0.4383248047401023,
+ "grad_norm": 0.818348330515836,
+ "learning_rate": 8.861275408889393e-06,
+ "loss": 0.4044035911560059,
+ "num_input_tokens_seen": 284827680,
+ "step": 6510,
+ "train_runtime": 50319.756,
+ "train_tokens_per_second": 5660.355
+ },
+ {
+ "epoch": 0.43866145973606246,
+ "grad_norm": 0.8194841529764723,
+ "learning_rate": 8.859595050385332e-06,
+ "loss": 0.40093269348144533,
+ "num_input_tokens_seen": 285050288,
+ "step": 6515,
+ "train_runtime": 50356.7378,
+ "train_tokens_per_second": 5660.619
+ },
+ {
+ "epoch": 0.4389981147320226,
+ "grad_norm": 0.7603950111901223,
+ "learning_rate": 8.857913612557825e-06,
+ "loss": 0.39177913665771485,
+ "num_input_tokens_seen": 285275776,
+ "step": 6520,
+ "train_runtime": 50399.864,
+ "train_tokens_per_second": 5660.249
+ },
+ {
+ "epoch": 0.43933476972798274,
+ "grad_norm": 0.9171110192985658,
+ "learning_rate": 8.856231095877077e-06,
+ "loss": 0.37299246788024903,
+ "num_input_tokens_seen": 285483840,
+ "step": 6525,
+ "train_runtime": 50437.6759,
+ "train_tokens_per_second": 5660.131
+ },
+ {
+ "epoch": 0.4396714247239429,
+ "grad_norm": 0.6722265713323919,
+ "learning_rate": 8.8545475008136e-06,
+ "loss": 0.3728148937225342,
+ "num_input_tokens_seen": 285708616,
+ "step": 6530,
+ "train_runtime": 50478.218,
+ "train_tokens_per_second": 5660.038
+ },
+ {
+ "epoch": 0.440008079719903,
+ "grad_norm": 0.8225035496811262,
+ "learning_rate": 8.852862827838207e-06,
+ "loss": 0.37645254135131834,
+ "num_input_tokens_seen": 285914952,
+ "step": 6535,
+ "train_runtime": 50513.9654,
+ "train_tokens_per_second": 5660.117
+ },
+ {
+ "epoch": 0.44034473471586316,
+ "grad_norm": 0.672139181444301,
+ "learning_rate": 8.851177077422008e-06,
+ "loss": 0.4013407230377197,
+ "num_input_tokens_seen": 286146336,
+ "step": 6540,
+ "train_runtime": 50553.4611,
+ "train_tokens_per_second": 5660.272
+ },
+ {
+ "epoch": 0.4406813897118233,
+ "grad_norm": 0.7644327948595142,
+ "learning_rate": 8.849490250036421e-06,
+ "loss": 0.4249086380004883,
+ "num_input_tokens_seen": 286362088,
+ "step": 6545,
+ "train_runtime": 50590.1862,
+ "train_tokens_per_second": 5660.428
+ },
+ {
+ "epoch": 0.44101804470778344,
+ "grad_norm": 0.8087430911202192,
+ "learning_rate": 8.847802346153159e-06,
+ "loss": 0.3778714895248413,
+ "num_input_tokens_seen": 286576504,
+ "step": 6550,
+ "train_runtime": 50623.1732,
+ "train_tokens_per_second": 5660.975
+ },
+ {
+ "epoch": 0.4413546997037436,
+ "grad_norm": 0.822277412759685,
+ "learning_rate": 8.846113366244241e-06,
+ "loss": 0.40006532669067385,
+ "num_input_tokens_seen": 286802512,
+ "step": 6555,
+ "train_runtime": 50653.7568,
+ "train_tokens_per_second": 5662.019
+ },
+ {
+ "epoch": 0.4416913546997037,
+ "grad_norm": 0.7788883116407057,
+ "learning_rate": 8.844423310781986e-06,
+ "loss": 0.3659277677536011,
+ "num_input_tokens_seen": 287019968,
+ "step": 6560,
+ "train_runtime": 50693.4968,
+ "train_tokens_per_second": 5661.87
+ },
+ {
+ "epoch": 0.44202800969566386,
+ "grad_norm": 0.8189856612851919,
+ "learning_rate": 8.842732180239011e-06,
+ "loss": 0.417068338394165,
+ "num_input_tokens_seen": 287233880,
+ "step": 6565,
+ "train_runtime": 50732.9626,
+ "train_tokens_per_second": 5661.682
+ },
+ {
+ "epoch": 0.442364664691624,
+ "grad_norm": 0.7955988955008321,
+ "learning_rate": 8.841039975088234e-06,
+ "loss": 0.37861127853393556,
+ "num_input_tokens_seen": 287450920,
+ "step": 6570,
+ "train_runtime": 50770.9895,
+ "train_tokens_per_second": 5661.716
+ },
+ {
+ "epoch": 0.44270131968758414,
+ "grad_norm": 0.7673850184930923,
+ "learning_rate": 8.839346695802878e-06,
+ "loss": 0.44512476921081545,
+ "num_input_tokens_seen": 287680008,
+ "step": 6575,
+ "train_runtime": 50808.7713,
+ "train_tokens_per_second": 5662.015
+ },
+ {
+ "epoch": 0.4430379746835443,
+ "grad_norm": 0.707236454889657,
+ "learning_rate": 8.837652342856459e-06,
+ "loss": 0.4026740550994873,
+ "num_input_tokens_seen": 287904440,
+ "step": 6580,
+ "train_runtime": 50842.5258,
+ "train_tokens_per_second": 5662.67
+ },
+ {
+ "epoch": 0.4433746296795044,
+ "grad_norm": 0.7131491003211293,
+ "learning_rate": 8.835956916722803e-06,
+ "loss": 0.37171001434326173,
+ "num_input_tokens_seen": 288123072,
+ "step": 6585,
+ "train_runtime": 50884.3886,
+ "train_tokens_per_second": 5662.308
+ },
+ {
+ "epoch": 0.44371128467546456,
+ "grad_norm": 0.7279859643678576,
+ "learning_rate": 8.834260417876024e-06,
+ "loss": 0.39625725746154783,
+ "num_input_tokens_seen": 288337792,
+ "step": 6590,
+ "train_runtime": 50922.6708,
+ "train_tokens_per_second": 5662.268
+ },
+ {
+ "epoch": 0.4440479396714247,
+ "grad_norm": 0.6822187993121582,
+ "learning_rate": 8.832562846790549e-06,
+ "loss": 0.4032885074615479,
+ "num_input_tokens_seen": 288562504,
+ "step": 6595,
+ "train_runtime": 50958.16,
+ "train_tokens_per_second": 5662.734
+ },
+ {
+ "epoch": 0.4443845946673849,
+ "grad_norm": 0.7696796062355198,
+ "learning_rate": 8.830864203941092e-06,
+ "loss": 0.4127056121826172,
+ "num_input_tokens_seen": 288783304,
+ "step": 6600,
+ "train_runtime": 50994.9294,
+ "train_tokens_per_second": 5662.981
+ },
+ {
+ "epoch": 0.44472124966334503,
+ "grad_norm": 0.796046357490389,
+ "learning_rate": 8.829164489802677e-06,
+ "loss": 0.37214198112487795,
+ "num_input_tokens_seen": 288995032,
+ "step": 6605,
+ "train_runtime": 51030.6262,
+ "train_tokens_per_second": 5663.168
+ },
+ {
+ "epoch": 0.44505790465930517,
+ "grad_norm": 0.8194054840462127,
+ "learning_rate": 8.827463704850622e-06,
+ "loss": 0.3779730796813965,
+ "num_input_tokens_seen": 289211784,
+ "step": 6610,
+ "train_runtime": 51065.1791,
+ "train_tokens_per_second": 5663.581
+ },
+ {
+ "epoch": 0.4453945596552653,
+ "grad_norm": 0.853917923414557,
+ "learning_rate": 8.825761849560547e-06,
+ "loss": 0.3779446125030518,
+ "num_input_tokens_seen": 289426496,
+ "step": 6615,
+ "train_runtime": 51104.5336,
+ "train_tokens_per_second": 5663.421
+ },
+ {
+ "epoch": 0.44573121465122545,
+ "grad_norm": 0.8243805829529494,
+ "learning_rate": 8.824058924408371e-06,
+ "loss": 0.3913204431533813,
+ "num_input_tokens_seen": 289642104,
+ "step": 6620,
+ "train_runtime": 51148.3439,
+ "train_tokens_per_second": 5662.786
+ },
+ {
+ "epoch": 0.4460678696471856,
+ "grad_norm": 0.6977791985037581,
+ "learning_rate": 8.822354929870311e-06,
+ "loss": 0.3713568687438965,
+ "num_input_tokens_seen": 289855080,
+ "step": 6625,
+ "train_runtime": 51186.4829,
+ "train_tokens_per_second": 5662.727
+ },
+ {
+ "epoch": 0.44640452464314573,
+ "grad_norm": 0.6544439461875585,
+ "learning_rate": 8.820649866422884e-06,
+ "loss": 0.3891900539398193,
+ "num_input_tokens_seen": 290080272,
+ "step": 6630,
+ "train_runtime": 51221.0131,
+ "train_tokens_per_second": 5663.306
+ },
+ {
+ "epoch": 0.44674117963910587,
+ "grad_norm": 0.7628155627258327,
+ "learning_rate": 8.818943734542905e-06,
+ "loss": 0.38543388843536375,
+ "num_input_tokens_seen": 290310296,
+ "step": 6635,
+ "train_runtime": 51260.636,
+ "train_tokens_per_second": 5663.416
+ },
+ {
+ "epoch": 0.447077834635066,
+ "grad_norm": 0.7154748015782274,
+ "learning_rate": 8.817236534707486e-06,
+ "loss": 0.38082070350646974,
+ "num_input_tokens_seen": 290530408,
+ "step": 6640,
+ "train_runtime": 51298.209,
+ "train_tokens_per_second": 5663.559
+ },
+ {
+ "epoch": 0.44741448963102615,
+ "grad_norm": 0.719435209044012,
+ "learning_rate": 8.815528267394045e-06,
+ "loss": 0.38055622577667236,
+ "num_input_tokens_seen": 290752096,
+ "step": 6645,
+ "train_runtime": 51339.4277,
+ "train_tokens_per_second": 5663.33
+ },
+ {
+ "epoch": 0.4477511446269863,
+ "grad_norm": 0.8130810621777962,
+ "learning_rate": 8.81381893308029e-06,
+ "loss": 0.42238750457763674,
+ "num_input_tokens_seen": 290976696,
+ "step": 6650,
+ "train_runtime": 51376.6803,
+ "train_tokens_per_second": 5663.595
+ },
+ {
+ "epoch": 0.4480877996229464,
+ "grad_norm": 0.6802969800260558,
+ "learning_rate": 8.81210853224423e-06,
+ "loss": 0.397933292388916,
+ "num_input_tokens_seen": 291195128,
+ "step": 6655,
+ "train_runtime": 51416.0644,
+ "train_tokens_per_second": 5663.505
+ },
+ {
+ "epoch": 0.44842445461890656,
+ "grad_norm": 0.7683337052075077,
+ "learning_rate": 8.810397065364177e-06,
+ "loss": 0.3660165309906006,
+ "num_input_tokens_seen": 291397432,
+ "step": 6660,
+ "train_runtime": 51460.1895,
+ "train_tokens_per_second": 5662.58
+ },
+ {
+ "epoch": 0.4487611096148667,
+ "grad_norm": 0.7385595517387719,
+ "learning_rate": 8.808684532918735e-06,
+ "loss": 0.4035042762756348,
+ "num_input_tokens_seen": 291611984,
+ "step": 6665,
+ "train_runtime": 51497.2825,
+ "train_tokens_per_second": 5662.667
+ },
+ {
+ "epoch": 0.44909776461082684,
+ "grad_norm": 0.8670076691450054,
+ "learning_rate": 8.806970935386807e-06,
+ "loss": 0.37046058177948,
+ "num_input_tokens_seen": 291817952,
+ "step": 6670,
+ "train_runtime": 51539.7964,
+ "train_tokens_per_second": 5661.993
+ },
+ {
+ "epoch": 0.449434419606787,
+ "grad_norm": 0.845351499495225,
+ "learning_rate": 8.805256273247597e-06,
+ "loss": 0.3828955411911011,
+ "num_input_tokens_seen": 292030296,
+ "step": 6675,
+ "train_runtime": 51585.7162,
+ "train_tokens_per_second": 5661.069
+ },
+ {
+ "epoch": 0.4497710746027471,
+ "grad_norm": 0.7558486825710268,
+ "learning_rate": 8.803540546980605e-06,
+ "loss": 0.3911006450653076,
+ "num_input_tokens_seen": 292240392,
+ "step": 6680,
+ "train_runtime": 51625.5292,
+ "train_tokens_per_second": 5660.773
+ },
+ {
+ "epoch": 0.45010772959870726,
+ "grad_norm": 0.7143473072799023,
+ "learning_rate": 8.801823757065628e-06,
+ "loss": 0.39244418144226073,
+ "num_input_tokens_seen": 292456784,
+ "step": 6685,
+ "train_runtime": 51661.3439,
+ "train_tokens_per_second": 5661.037
+ },
+ {
+ "epoch": 0.4504443845946674,
+ "grad_norm": 0.7046306873714272,
+ "learning_rate": 8.800105903982758e-06,
+ "loss": 0.37194869518280027,
+ "num_input_tokens_seen": 292669672,
+ "step": 6690,
+ "train_runtime": 51695.7982,
+ "train_tokens_per_second": 5661.382
+ },
+ {
+ "epoch": 0.45078103959062754,
+ "grad_norm": 0.9038260012464076,
+ "learning_rate": 8.798386988212393e-06,
+ "loss": 0.36289196014404296,
+ "num_input_tokens_seen": 292866896,
+ "step": 6695,
+ "train_runtime": 51737.8021,
+ "train_tokens_per_second": 5660.598
+ },
+ {
+ "epoch": 0.4511176945865877,
+ "grad_norm": 0.7097725989673881,
+ "learning_rate": 8.796667010235216e-06,
+ "loss": 0.4039313793182373,
+ "num_input_tokens_seen": 293088104,
+ "step": 6700,
+ "train_runtime": 51775.0194,
+ "train_tokens_per_second": 5660.801
+ },
+ {
+ "epoch": 0.4514543495825478,
+ "grad_norm": 0.6983233354591538,
+ "learning_rate": 8.794945970532219e-06,
+ "loss": 0.38548543453216555,
+ "num_input_tokens_seen": 293312720,
+ "step": 6705,
+ "train_runtime": 51819.136,
+ "train_tokens_per_second": 5660.317
+ },
+ {
+ "epoch": 0.45179100457850796,
+ "grad_norm": 0.7715442099717464,
+ "learning_rate": 8.79322386958468e-06,
+ "loss": 0.42191410064697266,
+ "num_input_tokens_seen": 293540040,
+ "step": 6710,
+ "train_runtime": 51850.4773,
+ "train_tokens_per_second": 5661.279
+ },
+ {
+ "epoch": 0.4521276595744681,
+ "grad_norm": 0.8667322084375894,
+ "learning_rate": 8.791500707874183e-06,
+ "loss": 0.3590597152709961,
+ "num_input_tokens_seen": 293753504,
+ "step": 6715,
+ "train_runtime": 51887.9851,
+ "train_tokens_per_second": 5661.301
+ },
+ {
+ "epoch": 0.45246431457042824,
+ "grad_norm": 0.7001275719009605,
+ "learning_rate": 8.789776485882601e-06,
+ "loss": 0.40479679107666017,
+ "num_input_tokens_seen": 293980328,
+ "step": 6720,
+ "train_runtime": 51930.6229,
+ "train_tokens_per_second": 5661.021
+ },
+ {
+ "epoch": 0.4528009695663884,
+ "grad_norm": 0.7486619231659534,
+ "learning_rate": 8.788051204092112e-06,
+ "loss": 0.380597448348999,
+ "num_input_tokens_seen": 294201288,
+ "step": 6725,
+ "train_runtime": 51969.1487,
+ "train_tokens_per_second": 5661.076
+ },
+ {
+ "epoch": 0.4531376245623485,
+ "grad_norm": 0.837803412390275,
+ "learning_rate": 8.786324862985183e-06,
+ "loss": 0.39116497039794923,
+ "num_input_tokens_seen": 294414792,
+ "step": 6730,
+ "train_runtime": 52007.4528,
+ "train_tokens_per_second": 5661.012
+ },
+ {
+ "epoch": 0.45347427955830866,
+ "grad_norm": 0.7989569302827977,
+ "learning_rate": 8.78459746304458e-06,
+ "loss": 0.37044734954833985,
+ "num_input_tokens_seen": 294593136,
+ "step": 6735,
+ "train_runtime": 52048.2566,
+ "train_tokens_per_second": 5660.0
+ },
+ {
+ "epoch": 0.4538109345542688,
+ "grad_norm": 0.7910699073925987,
+ "learning_rate": 8.782869004753363e-06,
+ "loss": 0.40142025947570803,
+ "num_input_tokens_seen": 294808392,
+ "step": 6740,
+ "train_runtime": 52084.1003,
+ "train_tokens_per_second": 5660.238
+ },
+ {
+ "epoch": 0.45414758955022894,
+ "grad_norm": 0.7578062648097129,
+ "learning_rate": 8.781139488594892e-06,
+ "loss": 0.39638261795043944,
+ "num_input_tokens_seen": 295032712,
+ "step": 6745,
+ "train_runtime": 52116.6196,
+ "train_tokens_per_second": 5661.01
+ },
+ {
+ "epoch": 0.4544842445461891,
+ "grad_norm": 0.7479401557774803,
+ "learning_rate": 8.779408915052821e-06,
+ "loss": 0.3619338274002075,
+ "num_input_tokens_seen": 295250848,
+ "step": 6750,
+ "train_runtime": 52158.5344,
+ "train_tokens_per_second": 5660.643
+ },
+ {
+ "epoch": 0.4548208995421492,
+ "grad_norm": 0.7929634941210536,
+ "learning_rate": 8.777677284611096e-06,
+ "loss": 0.36865906715393065,
+ "num_input_tokens_seen": 295444808,
+ "step": 6755,
+ "train_runtime": 52194.881,
+ "train_tokens_per_second": 5660.417
+ },
+ {
+ "epoch": 0.45515755453810935,
+ "grad_norm": 0.7977612087147963,
+ "learning_rate": 8.775944597753969e-06,
+ "loss": 0.3817254304885864,
+ "num_input_tokens_seen": 295651640,
+ "step": 6760,
+ "train_runtime": 52235.5188,
+ "train_tokens_per_second": 5659.973
+ },
+ {
+ "epoch": 0.4554942095340695,
+ "grad_norm": 0.7276333392461809,
+ "learning_rate": 8.774210854965972e-06,
+ "loss": 0.4110917568206787,
+ "num_input_tokens_seen": 295872600,
+ "step": 6765,
+ "train_runtime": 52282.5363,
+ "train_tokens_per_second": 5659.11
+ },
+ {
+ "epoch": 0.45583086453002963,
+ "grad_norm": 0.7743347789370404,
+ "learning_rate": 8.772476056731946e-06,
+ "loss": 0.3540276050567627,
+ "num_input_tokens_seen": 296100760,
+ "step": 6770,
+ "train_runtime": 52321.5826,
+ "train_tokens_per_second": 5659.247
+ },
+ {
+ "epoch": 0.4561675195259898,
+ "grad_norm": 0.8417565391365261,
+ "learning_rate": 8.77074020353702e-06,
+ "loss": 0.37112271785736084,
+ "num_input_tokens_seen": 296310112,
+ "step": 6775,
+ "train_runtime": 52364.5089,
+ "train_tokens_per_second": 5658.606
+ },
+ {
+ "epoch": 0.4565041745219499,
+ "grad_norm": 0.8971045549426276,
+ "learning_rate": 8.76900329586662e-06,
+ "loss": 0.416597318649292,
+ "num_input_tokens_seen": 296529072,
+ "step": 6780,
+ "train_runtime": 52408.4219,
+ "train_tokens_per_second": 5658.042
+ },
+ {
+ "epoch": 0.45684082951791005,
+ "grad_norm": 0.7289171592528997,
+ "learning_rate": 8.767265334206467e-06,
+ "loss": 0.38317620754241943,
+ "num_input_tokens_seen": 296750896,
+ "step": 6785,
+ "train_runtime": 52453.3014,
+ "train_tokens_per_second": 5657.43
+ },
+ {
+ "epoch": 0.4571774845138702,
+ "grad_norm": 0.742990584892427,
+ "learning_rate": 8.765526319042577e-06,
+ "loss": 0.38066811561584474,
+ "num_input_tokens_seen": 296973576,
+ "step": 6790,
+ "train_runtime": 52484.833,
+ "train_tokens_per_second": 5658.274
+ },
+ {
+ "epoch": 0.45751413950983033,
+ "grad_norm": 0.7941387596625845,
+ "learning_rate": 8.763786250861258e-06,
+ "loss": 0.35711822509765623,
+ "num_input_tokens_seen": 297198488,
+ "step": 6795,
+ "train_runtime": 52525.9579,
+ "train_tokens_per_second": 5658.126
+ },
+ {
+ "epoch": 0.45785079450579047,
+ "grad_norm": 0.7637071617244607,
+ "learning_rate": 8.762045130149114e-06,
+ "loss": 0.37547504901885986,
+ "num_input_tokens_seen": 297424408,
+ "step": 6800,
+ "train_runtime": 52568.1627,
+ "train_tokens_per_second": 5657.881
+ },
+ {
+ "epoch": 0.4581874495017506,
+ "grad_norm": 0.7114371063573226,
+ "learning_rate": 8.76030295739305e-06,
+ "loss": 0.391666579246521,
+ "num_input_tokens_seen": 297652664,
+ "step": 6805,
+ "train_runtime": 52605.904,
+ "train_tokens_per_second": 5658.161
+ },
+ {
+ "epoch": 0.45852410449771075,
+ "grad_norm": 0.7263315072433297,
+ "learning_rate": 8.758559733080252e-06,
+ "loss": 0.4111741065979004,
+ "num_input_tokens_seen": 297874936,
+ "step": 6810,
+ "train_runtime": 52637.0942,
+ "train_tokens_per_second": 5659.031
+ },
+ {
+ "epoch": 0.4588607594936709,
+ "grad_norm": 0.7359316103796298,
+ "learning_rate": 8.75681545769821e-06,
+ "loss": 0.38690164089202883,
+ "num_input_tokens_seen": 298107584,
+ "step": 6815,
+ "train_runtime": 52669.678,
+ "train_tokens_per_second": 5659.947
+ },
+ {
+ "epoch": 0.45919741448963103,
+ "grad_norm": 0.8032263081480874,
+ "learning_rate": 8.755070131734706e-06,
+ "loss": 0.3867367744445801,
+ "num_input_tokens_seen": 298320208,
+ "step": 6820,
+ "train_runtime": 52707.8474,
+ "train_tokens_per_second": 5659.882
+ },
+ {
+ "epoch": 0.45953406948559117,
+ "grad_norm": 0.7833333891541101,
+ "learning_rate": 8.753323755677812e-06,
+ "loss": 0.4013802528381348,
+ "num_input_tokens_seen": 298533552,
+ "step": 6825,
+ "train_runtime": 52749.6034,
+ "train_tokens_per_second": 5659.446
+ },
+ {
+ "epoch": 0.4598707244815513,
+ "grad_norm": 0.7609318114113853,
+ "learning_rate": 8.7515763300159e-06,
+ "loss": 0.3463205575942993,
+ "num_input_tokens_seen": 298746696,
+ "step": 6830,
+ "train_runtime": 52789.3591,
+ "train_tokens_per_second": 5659.222
+ },
+ {
+ "epoch": 0.46020737947751145,
+ "grad_norm": 0.6930204184723102,
+ "learning_rate": 8.74982785523763e-06,
+ "loss": 0.40210418701171874,
+ "num_input_tokens_seen": 298972384,
+ "step": 6835,
+ "train_runtime": 52827.2223,
+ "train_tokens_per_second": 5659.438
+ },
+ {
+ "epoch": 0.4605440344734716,
+ "grad_norm": 0.7026299121313064,
+ "learning_rate": 8.748078331831957e-06,
+ "loss": 0.3934351682662964,
+ "num_input_tokens_seen": 299200112,
+ "step": 6840,
+ "train_runtime": 52862.9542,
+ "train_tokens_per_second": 5659.92
+ },
+ {
+ "epoch": 0.4608806894694317,
+ "grad_norm": 0.7019416941915204,
+ "learning_rate": 8.746327760288129e-06,
+ "loss": 0.3836989164352417,
+ "num_input_tokens_seen": 299437024,
+ "step": 6845,
+ "train_runtime": 52894.4779,
+ "train_tokens_per_second": 5661.026
+ },
+ {
+ "epoch": 0.46121734446539187,
+ "grad_norm": 0.8234006493355744,
+ "learning_rate": 8.744576141095691e-06,
+ "loss": 0.39553108215332033,
+ "num_input_tokens_seen": 299641752,
+ "step": 6850,
+ "train_runtime": 52932.8412,
+ "train_tokens_per_second": 5660.791
+ },
+ {
+ "epoch": 0.461553999461352,
+ "grad_norm": 0.846531534553998,
+ "learning_rate": 8.742823474744476e-06,
+ "loss": 0.39920735359191895,
+ "num_input_tokens_seen": 299852504,
+ "step": 6855,
+ "train_runtime": 52970.3403,
+ "train_tokens_per_second": 5660.762
+ },
+ {
+ "epoch": 0.46189065445731214,
+ "grad_norm": 0.7339872597188516,
+ "learning_rate": 8.74106976172461e-06,
+ "loss": 0.40232462882995607,
+ "num_input_tokens_seen": 300073192,
+ "step": 6860,
+ "train_runtime": 53013.6318,
+ "train_tokens_per_second": 5660.302
+ },
+ {
+ "epoch": 0.4622273094532723,
+ "grad_norm": 0.8137834490765111,
+ "learning_rate": 8.739315002526513e-06,
+ "loss": 0.40238103866577146,
+ "num_input_tokens_seen": 300287552,
+ "step": 6865,
+ "train_runtime": 53055.019,
+ "train_tokens_per_second": 5659.927
+ },
+ {
+ "epoch": 0.4625639644492324,
+ "grad_norm": 0.7042015061046794,
+ "learning_rate": 8.737559197640902e-06,
+ "loss": 0.3717548131942749,
+ "num_input_tokens_seen": 300496376,
+ "step": 6870,
+ "train_runtime": 53091.6564,
+ "train_tokens_per_second": 5659.955
+ },
+ {
+ "epoch": 0.46290061944519256,
+ "grad_norm": 0.7325334644495457,
+ "learning_rate": 8.735802347558778e-06,
+ "loss": 0.4027840614318848,
+ "num_input_tokens_seen": 300709416,
+ "step": 6875,
+ "train_runtime": 53135.1809,
+ "train_tokens_per_second": 5659.328
+ },
+ {
+ "epoch": 0.4632372744411527,
+ "grad_norm": 0.7655214239764673,
+ "learning_rate": 8.734044452771442e-06,
+ "loss": 0.38804852962493896,
+ "num_input_tokens_seen": 300918752,
+ "step": 6880,
+ "train_runtime": 53179.1079,
+ "train_tokens_per_second": 5658.59
+ },
+ {
+ "epoch": 0.46357392943711284,
+ "grad_norm": 0.8904799503928702,
+ "learning_rate": 8.73228551377048e-06,
+ "loss": 0.40845584869384766,
+ "num_input_tokens_seen": 301139720,
+ "step": 6885,
+ "train_runtime": 53208.9509,
+ "train_tokens_per_second": 5659.569
+ },
+ {
+ "epoch": 0.463910584433073,
+ "grad_norm": 0.6748289632604044,
+ "learning_rate": 8.730525531047776e-06,
+ "loss": 0.4137433052062988,
+ "num_input_tokens_seen": 301362752,
+ "step": 6890,
+ "train_runtime": 53254.1898,
+ "train_tokens_per_second": 5658.949
+ },
+ {
+ "epoch": 0.4642472394290331,
+ "grad_norm": 0.7035224245591345,
+ "learning_rate": 8.728764505095503e-06,
+ "loss": 0.4120682716369629,
+ "num_input_tokens_seen": 301587944,
+ "step": 6895,
+ "train_runtime": 53291.4984,
+ "train_tokens_per_second": 5659.213
+ },
+ {
+ "epoch": 0.46458389442499326,
+ "grad_norm": 0.6435345300249111,
+ "learning_rate": 8.727002436406126e-06,
+ "loss": 0.3724491834640503,
+ "num_input_tokens_seen": 301823392,
+ "step": 6900,
+ "train_runtime": 53331.8202,
+ "train_tokens_per_second": 5659.349
+ },
+ {
+ "epoch": 0.4649205494209534,
+ "grad_norm": 0.7100192481355228,
+ "learning_rate": 8.725239325472404e-06,
+ "loss": 0.4057363510131836,
+ "num_input_tokens_seen": 302046392,
+ "step": 6905,
+ "train_runtime": 53370.503,
+ "train_tokens_per_second": 5659.426
+ },
+ {
+ "epoch": 0.46525720441691354,
+ "grad_norm": 0.7998034880411997,
+ "learning_rate": 8.723475172787381e-06,
+ "loss": 0.3984755277633667,
+ "num_input_tokens_seen": 302270848,
+ "step": 6910,
+ "train_runtime": 53413.2206,
+ "train_tokens_per_second": 5659.102
+ },
+ {
+ "epoch": 0.4655938594128737,
+ "grad_norm": 0.9282661770223051,
+ "learning_rate": 8.7217099788444e-06,
+ "loss": 0.38761351108551023,
+ "num_input_tokens_seen": 302484424,
+ "step": 6915,
+ "train_runtime": 53452.6504,
+ "train_tokens_per_second": 5658.923
+ },
+ {
+ "epoch": 0.4659305144088338,
+ "grad_norm": 0.6829252086088465,
+ "learning_rate": 8.719943744137091e-06,
+ "loss": 0.3843079566955566,
+ "num_input_tokens_seen": 302706608,
+ "step": 6920,
+ "train_runtime": 53494.2752,
+ "train_tokens_per_second": 5658.673
+ },
+ {
+ "epoch": 0.46626716940479396,
+ "grad_norm": 0.7370197158639645,
+ "learning_rate": 8.718176469159378e-06,
+ "loss": 0.40040907859802244,
+ "num_input_tokens_seen": 302946080,
+ "step": 6925,
+ "train_runtime": 53527.8006,
+ "train_tokens_per_second": 5659.603
+ },
+ {
+ "epoch": 0.4666038244007541,
+ "grad_norm": 0.7744303086322805,
+ "learning_rate": 8.716408154405469e-06,
+ "loss": 0.370440411567688,
+ "num_input_tokens_seen": 303172472,
+ "step": 6930,
+ "train_runtime": 53558.5862,
+ "train_tokens_per_second": 5660.576
+ },
+ {
+ "epoch": 0.46694047939671424,
+ "grad_norm": 0.9149869901290688,
+ "learning_rate": 8.714638800369872e-06,
+ "loss": 0.4163699150085449,
+ "num_input_tokens_seen": 303374536,
+ "step": 6935,
+ "train_runtime": 53596.9127,
+ "train_tokens_per_second": 5660.299
+ },
+ {
+ "epoch": 0.4672771343926744,
+ "grad_norm": 0.6478644031670531,
+ "learning_rate": 8.712868407547378e-06,
+ "loss": 0.41262283325195315,
+ "num_input_tokens_seen": 303604704,
+ "step": 6940,
+ "train_runtime": 53634.4097,
+ "train_tokens_per_second": 5660.633
+ },
+ {
+ "epoch": 0.4676137893886345,
+ "grad_norm": 0.8540394960614904,
+ "learning_rate": 8.711096976433076e-06,
+ "loss": 0.40007991790771485,
+ "num_input_tokens_seen": 303830584,
+ "step": 6945,
+ "train_runtime": 53675.8336,
+ "train_tokens_per_second": 5660.473
+ },
+ {
+ "epoch": 0.46795044438459465,
+ "grad_norm": 0.6854573080771246,
+ "learning_rate": 8.709324507522337e-06,
+ "loss": 0.3609914779663086,
+ "num_input_tokens_seen": 304051832,
+ "step": 6950,
+ "train_runtime": 53710.0276,
+ "train_tokens_per_second": 5660.988
+ },
+ {
+ "epoch": 0.4682870993805548,
+ "grad_norm": 0.74529061138361,
+ "learning_rate": 8.707551001310828e-06,
+ "loss": 0.3842365264892578,
+ "num_input_tokens_seen": 304279352,
+ "step": 6955,
+ "train_runtime": 53749.8544,
+ "train_tokens_per_second": 5661.027
+ },
+ {
+ "epoch": 0.46862375437651493,
+ "grad_norm": 0.7026982064246363,
+ "learning_rate": 8.705776458294503e-06,
+ "loss": 0.387627911567688,
+ "num_input_tokens_seen": 304487200,
+ "step": 6960,
+ "train_runtime": 53785.1129,
+ "train_tokens_per_second": 5661.18
+ },
+ {
+ "epoch": 0.4689604093724751,
+ "grad_norm": 0.7214247021583069,
+ "learning_rate": 8.704000878969608e-06,
+ "loss": 0.37026023864746094,
+ "num_input_tokens_seen": 304693840,
+ "step": 6965,
+ "train_runtime": 53822.9634,
+ "train_tokens_per_second": 5661.038
+ },
+ {
+ "epoch": 0.4692970643684352,
+ "grad_norm": 0.6979359178232242,
+ "learning_rate": 8.702224263832679e-06,
+ "loss": 0.3628101825714111,
+ "num_input_tokens_seen": 304921392,
+ "step": 6970,
+ "train_runtime": 53861.4766,
+ "train_tokens_per_second": 5661.215
+ },
+ {
+ "epoch": 0.46963371936439535,
+ "grad_norm": 0.7813572135644429,
+ "learning_rate": 8.700446613380542e-06,
+ "loss": 0.4275547504425049,
+ "num_input_tokens_seen": 305149136,
+ "step": 6975,
+ "train_runtime": 53899.1275,
+ "train_tokens_per_second": 5661.486
+ },
+ {
+ "epoch": 0.4699703743603555,
+ "grad_norm": 0.800833875206026,
+ "learning_rate": 8.698667928110307e-06,
+ "loss": 0.373789119720459,
+ "num_input_tokens_seen": 305369336,
+ "step": 6980,
+ "train_runtime": 53936.8451,
+ "train_tokens_per_second": 5661.609
+ },
+ {
+ "epoch": 0.47030702935631563,
+ "grad_norm": 0.8980829202572613,
+ "learning_rate": 8.696888208519381e-06,
+ "loss": 0.3973226070404053,
+ "num_input_tokens_seen": 305581504,
+ "step": 6985,
+ "train_runtime": 53973.4435,
+ "train_tokens_per_second": 5661.701
+ },
+ {
+ "epoch": 0.47064368435227577,
+ "grad_norm": 0.7351189903914839,
+ "learning_rate": 8.695107455105454e-06,
+ "loss": 0.34683785438537595,
+ "num_input_tokens_seen": 305790632,
+ "step": 6990,
+ "train_runtime": 54006.1459,
+ "train_tokens_per_second": 5662.145
+ },
+ {
+ "epoch": 0.4709803393482359,
+ "grad_norm": 0.7928821233878864,
+ "learning_rate": 8.693325668366513e-06,
+ "loss": 0.39075660705566406,
+ "num_input_tokens_seen": 306004232,
+ "step": 6995,
+ "train_runtime": 54042.3536,
+ "train_tokens_per_second": 5662.304
+ },
+ {
+ "epoch": 0.47131699434419605,
+ "grad_norm": 0.8527776847132168,
+ "learning_rate": 8.691542848800825e-06,
+ "loss": 0.39600386619567873,
+ "num_input_tokens_seen": 306212208,
+ "step": 7000,
+ "train_runtime": 54084.5742,
+ "train_tokens_per_second": 5661.729
+ },
+ {
+ "epoch": 0.4716536493401562,
+ "grad_norm": 0.8808374134194763,
+ "learning_rate": 8.68975899690695e-06,
+ "loss": 0.38110928535461425,
+ "num_input_tokens_seen": 306437744,
+ "step": 7005,
+ "train_runtime": 54125.0428,
+ "train_tokens_per_second": 5661.663
+ },
+ {
+ "epoch": 0.47199030433611633,
+ "grad_norm": 0.7423709319863485,
+ "learning_rate": 8.687974113183738e-06,
+ "loss": 0.37776660919189453,
+ "num_input_tokens_seen": 306667400,
+ "step": 7010,
+ "train_runtime": 54160.5555,
+ "train_tokens_per_second": 5662.191
+ },
+ {
+ "epoch": 0.47232695933207647,
+ "grad_norm": 0.8031207862343954,
+ "learning_rate": 8.686188198130326e-06,
+ "loss": 0.3893289089202881,
+ "num_input_tokens_seen": 306891232,
+ "step": 7015,
+ "train_runtime": 54195.218,
+ "train_tokens_per_second": 5662.699
+ },
+ {
+ "epoch": 0.4726636143280366,
+ "grad_norm": 0.716418279114708,
+ "learning_rate": 8.684401252246138e-06,
+ "loss": 0.3991250038146973,
+ "num_input_tokens_seen": 307116336,
+ "step": 7020,
+ "train_runtime": 54233.7853,
+ "train_tokens_per_second": 5662.823
+ },
+ {
+ "epoch": 0.47300026932399675,
+ "grad_norm": 0.7731462314856246,
+ "learning_rate": 8.68261327603089e-06,
+ "loss": 0.42037010192871094,
+ "num_input_tokens_seen": 307328112,
+ "step": 7025,
+ "train_runtime": 54273.3793,
+ "train_tokens_per_second": 5662.594
+ },
+ {
+ "epoch": 0.4733369243199569,
+ "grad_norm": 0.7619151144650879,
+ "learning_rate": 8.68082426998458e-06,
+ "loss": 0.42117109298706057,
+ "num_input_tokens_seen": 307558384,
+ "step": 7030,
+ "train_runtime": 54315.1352,
+ "train_tokens_per_second": 5662.48
+ },
+ {
+ "epoch": 0.473673579315917,
+ "grad_norm": 0.8366328207039446,
+ "learning_rate": 8.6790342346075e-06,
+ "loss": 0.355437707901001,
+ "num_input_tokens_seen": 307757832,
+ "step": 7035,
+ "train_runtime": 54355.0453,
+ "train_tokens_per_second": 5661.992
+ },
+ {
+ "epoch": 0.47401023431187717,
+ "grad_norm": 0.694627660462694,
+ "learning_rate": 8.677243170400228e-06,
+ "loss": 0.3744319438934326,
+ "num_input_tokens_seen": 307964848,
+ "step": 7040,
+ "train_runtime": 54392.4804,
+ "train_tokens_per_second": 5661.901
+ },
+ {
+ "epoch": 0.4743468893078373,
+ "grad_norm": 0.7593548972261533,
+ "learning_rate": 8.675451077863632e-06,
+ "loss": 0.40117859840393066,
+ "num_input_tokens_seen": 308180176,
+ "step": 7045,
+ "train_runtime": 54429.867,
+ "train_tokens_per_second": 5661.968
+ },
+ {
+ "epoch": 0.47468354430379744,
+ "grad_norm": 0.7788878005901869,
+ "learning_rate": 8.67365795749886e-06,
+ "loss": 0.37619283199310305,
+ "num_input_tokens_seen": 308393256,
+ "step": 7050,
+ "train_runtime": 54475.6142,
+ "train_tokens_per_second": 5661.125
+ },
+ {
+ "epoch": 0.4750201992997576,
+ "grad_norm": 0.6935999303413757,
+ "learning_rate": 8.671863809807355e-06,
+ "loss": 0.36246466636657715,
+ "num_input_tokens_seen": 308598120,
+ "step": 7055,
+ "train_runtime": 54512.6696,
+ "train_tokens_per_second": 5661.035
+ },
+ {
+ "epoch": 0.4753568542957177,
+ "grad_norm": 0.7796482197914519,
+ "learning_rate": 8.670068635290844e-06,
+ "loss": 0.3815620899200439,
+ "num_input_tokens_seen": 308814312,
+ "step": 7060,
+ "train_runtime": 54553.6102,
+ "train_tokens_per_second": 5660.749
+ },
+ {
+ "epoch": 0.47569350929167786,
+ "grad_norm": 0.6394547128504924,
+ "learning_rate": 8.668272434451343e-06,
+ "loss": 0.3547971725463867,
+ "num_input_tokens_seen": 309036648,
+ "step": 7065,
+ "train_runtime": 54589.1907,
+ "train_tokens_per_second": 5661.133
+ },
+ {
+ "epoch": 0.476030164287638,
+ "grad_norm": 0.8424232555426785,
+ "learning_rate": 8.666475207791154e-06,
+ "loss": 0.3938549518585205,
+ "num_input_tokens_seen": 309237312,
+ "step": 7070,
+ "train_runtime": 54626.0249,
+ "train_tokens_per_second": 5660.989
+ },
+ {
+ "epoch": 0.47636681928359814,
+ "grad_norm": 0.7156676868042459,
+ "learning_rate": 8.664676955812863e-06,
+ "loss": 0.3805175065994263,
+ "num_input_tokens_seen": 309461072,
+ "step": 7075,
+ "train_runtime": 54665.5373,
+ "train_tokens_per_second": 5660.99
+ },
+ {
+ "epoch": 0.4767034742795583,
+ "grad_norm": 0.7803230987384645,
+ "learning_rate": 8.662877679019348e-06,
+ "loss": 0.38736543655395506,
+ "num_input_tokens_seen": 309676264,
+ "step": 7080,
+ "train_runtime": 54702.5852,
+ "train_tokens_per_second": 5661.09
+ },
+ {
+ "epoch": 0.4770401292755185,
+ "grad_norm": 0.811046251111573,
+ "learning_rate": 8.66107737791377e-06,
+ "loss": 0.40035252571105956,
+ "num_input_tokens_seen": 309908584,
+ "step": 7085,
+ "train_runtime": 54739.8422,
+ "train_tokens_per_second": 5661.481
+ },
+ {
+ "epoch": 0.4773767842714786,
+ "grad_norm": 0.8088100128547995,
+ "learning_rate": 8.659276052999574e-06,
+ "loss": 0.37158200740814207,
+ "num_input_tokens_seen": 310109536,
+ "step": 7090,
+ "train_runtime": 54784.3231,
+ "train_tokens_per_second": 5660.552
+ },
+ {
+ "epoch": 0.47771343926743876,
+ "grad_norm": 0.6607865351139958,
+ "learning_rate": 8.657473704780504e-06,
+ "loss": 0.3958035707473755,
+ "num_input_tokens_seen": 310316416,
+ "step": 7095,
+ "train_runtime": 54823.0111,
+ "train_tokens_per_second": 5660.331
+ },
+ {
+ "epoch": 0.4780500942633989,
+ "grad_norm": 0.7508776157927014,
+ "learning_rate": 8.65567033376057e-06,
+ "loss": 0.3861806392669678,
+ "num_input_tokens_seen": 310543280,
+ "step": 7100,
+ "train_runtime": 54858.1842,
+ "train_tokens_per_second": 5660.838
+ },
+ {
+ "epoch": 0.47838674925935903,
+ "grad_norm": 0.6796527894531441,
+ "learning_rate": 8.65386594044409e-06,
+ "loss": 0.38321571350097655,
+ "num_input_tokens_seen": 310767656,
+ "step": 7105,
+ "train_runtime": 54899.3833,
+ "train_tokens_per_second": 5660.677
+ },
+ {
+ "epoch": 0.4787234042553192,
+ "grad_norm": 0.7839875292722721,
+ "learning_rate": 8.652060525335647e-06,
+ "loss": 0.35958447456359866,
+ "num_input_tokens_seen": 310977648,
+ "step": 7110,
+ "train_runtime": 54935.0727,
+ "train_tokens_per_second": 5660.822
+ },
+ {
+ "epoch": 0.4790600592512793,
+ "grad_norm": 0.6805039530532168,
+ "learning_rate": 8.650254088940125e-06,
+ "loss": 0.37933621406555174,
+ "num_input_tokens_seen": 311184472,
+ "step": 7115,
+ "train_runtime": 54970.8187,
+ "train_tokens_per_second": 5660.903
+ },
+ {
+ "epoch": 0.47939671424723945,
+ "grad_norm": 0.8690524861477827,
+ "learning_rate": 8.648446631762686e-06,
+ "loss": 0.3870699882507324,
+ "num_input_tokens_seen": 311386608,
+ "step": 7120,
+ "train_runtime": 55007.3848,
+ "train_tokens_per_second": 5660.815
+ },
+ {
+ "epoch": 0.4797333692431996,
+ "grad_norm": 0.7834933561721031,
+ "learning_rate": 8.646638154308781e-06,
+ "loss": 0.3672234296798706,
+ "num_input_tokens_seen": 311600392,
+ "step": 7125,
+ "train_runtime": 55045.8437,
+ "train_tokens_per_second": 5660.743
+ },
+ {
+ "epoch": 0.48007002423915973,
+ "grad_norm": 0.6821280974528945,
+ "learning_rate": 8.644828657084144e-06,
+ "loss": 0.3536520004272461,
+ "num_input_tokens_seen": 311826176,
+ "step": 7130,
+ "train_runtime": 55086.3186,
+ "train_tokens_per_second": 5660.683
+ },
+ {
+ "epoch": 0.48040667923511987,
+ "grad_norm": 0.7090523007119766,
+ "learning_rate": 8.643018140594797e-06,
+ "loss": 0.35062260627746583,
+ "num_input_tokens_seen": 312040616,
+ "step": 7135,
+ "train_runtime": 55118.7948,
+ "train_tokens_per_second": 5661.238
+ },
+ {
+ "epoch": 0.48074333423108,
+ "grad_norm": 0.7694829828758608,
+ "learning_rate": 8.641206605347042e-06,
+ "loss": 0.3680727958679199,
+ "num_input_tokens_seen": 312250776,
+ "step": 7140,
+ "train_runtime": 55154.654,
+ "train_tokens_per_second": 5661.368
+ },
+ {
+ "epoch": 0.48107998922704015,
+ "grad_norm": 0.7834141310420747,
+ "learning_rate": 8.639394051847472e-06,
+ "loss": 0.3820206642150879,
+ "num_input_tokens_seen": 312469768,
+ "step": 7145,
+ "train_runtime": 55194.213,
+ "train_tokens_per_second": 5661.278
+ },
+ {
+ "epoch": 0.4814166442230003,
+ "grad_norm": 0.8086494042960466,
+ "learning_rate": 8.63758048060296e-06,
+ "loss": 0.3749295473098755,
+ "num_input_tokens_seen": 312686816,
+ "step": 7150,
+ "train_runtime": 55234.7643,
+ "train_tokens_per_second": 5661.051
+ },
+ {
+ "epoch": 0.48175329921896043,
+ "grad_norm": 1.146252827120329,
+ "learning_rate": 8.635765892120666e-06,
+ "loss": 0.35863730907440183,
+ "num_input_tokens_seen": 312928240,
+ "step": 7155,
+ "train_runtime": 55276.3406,
+ "train_tokens_per_second": 5661.161
+ },
+ {
+ "epoch": 0.48208995421492057,
+ "grad_norm": 0.6322723599846541,
+ "learning_rate": 8.633950286908032e-06,
+ "loss": 0.3957623243331909,
+ "num_input_tokens_seen": 313146400,
+ "step": 7160,
+ "train_runtime": 55318.9273,
+ "train_tokens_per_second": 5660.746
+ },
+ {
+ "epoch": 0.4824266092108807,
+ "grad_norm": 0.8242447392294052,
+ "learning_rate": 8.63213366547279e-06,
+ "loss": 0.41290960311889646,
+ "num_input_tokens_seen": 313353080,
+ "step": 7165,
+ "train_runtime": 55354.3316,
+ "train_tokens_per_second": 5660.859
+ },
+ {
+ "epoch": 0.48276326420684085,
+ "grad_norm": 0.8189218668862145,
+ "learning_rate": 8.630316028322952e-06,
+ "loss": 0.3599478960037231,
+ "num_input_tokens_seen": 313556520,
+ "step": 7170,
+ "train_runtime": 55393.1558,
+ "train_tokens_per_second": 5660.564
+ },
+ {
+ "epoch": 0.483099919202801,
+ "grad_norm": 0.8049613617878884,
+ "learning_rate": 8.628497375966811e-06,
+ "loss": 0.3688513278961182,
+ "num_input_tokens_seen": 313791304,
+ "step": 7175,
+ "train_runtime": 55433.9631,
+ "train_tokens_per_second": 5660.633
+ },
+ {
+ "epoch": 0.4834365741987611,
+ "grad_norm": 0.7403145334116261,
+ "learning_rate": 8.62667770891295e-06,
+ "loss": 0.4179924488067627,
+ "num_input_tokens_seen": 314025432,
+ "step": 7180,
+ "train_runtime": 55479.2927,
+ "train_tokens_per_second": 5660.228
+ },
+ {
+ "epoch": 0.48377322919472127,
+ "grad_norm": 0.6161204381365467,
+ "learning_rate": 8.624857027670232e-06,
+ "loss": 0.3718327522277832,
+ "num_input_tokens_seen": 314253744,
+ "step": 7185,
+ "train_runtime": 55518.1876,
+ "train_tokens_per_second": 5660.375
+ },
+ {
+ "epoch": 0.4841098841906814,
+ "grad_norm": 0.6673541170021607,
+ "learning_rate": 8.623035332747804e-06,
+ "loss": 0.4232837677001953,
+ "num_input_tokens_seen": 314457360,
+ "step": 7190,
+ "train_runtime": 55555.7684,
+ "train_tokens_per_second": 5660.211
+ },
+ {
+ "epoch": 0.48444653918664154,
+ "grad_norm": 0.8639989632411573,
+ "learning_rate": 8.6212126246551e-06,
+ "loss": 0.35295858383178713,
+ "num_input_tokens_seen": 314659296,
+ "step": 7195,
+ "train_runtime": 55590.9252,
+ "train_tokens_per_second": 5660.264
+ },
+ {
+ "epoch": 0.4847831941826017,
+ "grad_norm": 0.7194371682580046,
+ "learning_rate": 8.619388903901833e-06,
+ "loss": 0.397071385383606,
+ "num_input_tokens_seen": 314886400,
+ "step": 7200,
+ "train_runtime": 55632.7504,
+ "train_tokens_per_second": 5660.09
+ },
+ {
+ "epoch": 0.4851198491785618,
+ "grad_norm": 0.7895653948671645,
+ "learning_rate": 8.617564170997998e-06,
+ "loss": 0.393186354637146,
+ "num_input_tokens_seen": 315113560,
+ "step": 7205,
+ "train_runtime": 55665.6693,
+ "train_tokens_per_second": 5660.824
+ },
+ {
+ "epoch": 0.48545650417452196,
+ "grad_norm": 0.6916492568995722,
+ "learning_rate": 8.61573842645388e-06,
+ "loss": 0.38975393772125244,
+ "num_input_tokens_seen": 315339760,
+ "step": 7210,
+ "train_runtime": 55701.6461,
+ "train_tokens_per_second": 5661.229
+ },
+ {
+ "epoch": 0.4857931591704821,
+ "grad_norm": 0.7125114232933183,
+ "learning_rate": 8.61391167078004e-06,
+ "loss": 0.3773904085159302,
+ "num_input_tokens_seen": 315564560,
+ "step": 7215,
+ "train_runtime": 55739.932,
+ "train_tokens_per_second": 5661.373
+ },
+ {
+ "epoch": 0.48612981416644224,
+ "grad_norm": 0.7816578333674177,
+ "learning_rate": 8.612083904487324e-06,
+ "loss": 0.4223905563354492,
+ "num_input_tokens_seen": 315780664,
+ "step": 7220,
+ "train_runtime": 55782.1816,
+ "train_tokens_per_second": 5660.959
+ },
+ {
+ "epoch": 0.4864664691624024,
+ "grad_norm": 0.8225392117617899,
+ "learning_rate": 8.610255128086867e-06,
+ "loss": 0.39981255531311033,
+ "num_input_tokens_seen": 315985224,
+ "step": 7225,
+ "train_runtime": 55817.5809,
+ "train_tokens_per_second": 5661.034
+ },
+ {
+ "epoch": 0.4868031241583625,
+ "grad_norm": 0.7192041520493799,
+ "learning_rate": 8.60842534209007e-06,
+ "loss": 0.35159389972686766,
+ "num_input_tokens_seen": 316206120,
+ "step": 7230,
+ "train_runtime": 55856.1823,
+ "train_tokens_per_second": 5661.076
+ },
+ {
+ "epoch": 0.48713977915432266,
+ "grad_norm": 0.7538381499748801,
+ "learning_rate": 8.606594547008636e-06,
+ "loss": 0.33089404106140136,
+ "num_input_tokens_seen": 316428112,
+ "step": 7235,
+ "train_runtime": 55889.528,
+ "train_tokens_per_second": 5661.671
+ },
+ {
+ "epoch": 0.4874764341502828,
+ "grad_norm": 0.7734857420248509,
+ "learning_rate": 8.604762743354536e-06,
+ "loss": 0.39783239364624023,
+ "num_input_tokens_seen": 316666376,
+ "step": 7240,
+ "train_runtime": 55922.8766,
+ "train_tokens_per_second": 5662.555
+ },
+ {
+ "epoch": 0.48781308914624294,
+ "grad_norm": 0.7475280321551805,
+ "learning_rate": 8.602929931640029e-06,
+ "loss": 0.3995728254318237,
+ "num_input_tokens_seen": 316868832,
+ "step": 7245,
+ "train_runtime": 55959.5907,
+ "train_tokens_per_second": 5662.458
+ },
+ {
+ "epoch": 0.4881497441422031,
+ "grad_norm": 0.8349680082803242,
+ "learning_rate": 8.601096112377659e-06,
+ "loss": 0.4125559329986572,
+ "num_input_tokens_seen": 317082248,
+ "step": 7250,
+ "train_runtime": 56002.6917,
+ "train_tokens_per_second": 5661.911
+ },
+ {
+ "epoch": 0.4884863991381632,
+ "grad_norm": 0.6943776304621931,
+ "learning_rate": 8.59926128608024e-06,
+ "loss": 0.3738250255584717,
+ "num_input_tokens_seen": 317303120,
+ "step": 7255,
+ "train_runtime": 56039.6874,
+ "train_tokens_per_second": 5662.114
+ },
+ {
+ "epoch": 0.48882305413412336,
+ "grad_norm": 0.7789443419004642,
+ "learning_rate": 8.597425453260881e-06,
+ "loss": 0.3914709806442261,
+ "num_input_tokens_seen": 317515672,
+ "step": 7260,
+ "train_runtime": 56082.3268,
+ "train_tokens_per_second": 5661.599
+ },
+ {
+ "epoch": 0.4891597091300835,
+ "grad_norm": 0.8422191416131971,
+ "learning_rate": 8.595588614432967e-06,
+ "loss": 0.3669420719146729,
+ "num_input_tokens_seen": 317720352,
+ "step": 7265,
+ "train_runtime": 56128.1181,
+ "train_tokens_per_second": 5660.627
+ },
+ {
+ "epoch": 0.48949636412604364,
+ "grad_norm": 0.7163712412405564,
+ "learning_rate": 8.59375077011016e-06,
+ "loss": 0.36806163787841795,
+ "num_input_tokens_seen": 317948384,
+ "step": 7270,
+ "train_runtime": 56167.6356,
+ "train_tokens_per_second": 5660.704
+ },
+ {
+ "epoch": 0.4898330191220038,
+ "grad_norm": 0.8174970324207671,
+ "learning_rate": 8.591911920806412e-06,
+ "loss": 0.39014444351196287,
+ "num_input_tokens_seen": 318179072,
+ "step": 7275,
+ "train_runtime": 56213.9579,
+ "train_tokens_per_second": 5660.144
+ },
+ {
+ "epoch": 0.4901696741179639,
+ "grad_norm": 0.6974694369606549,
+ "learning_rate": 8.590072067035947e-06,
+ "loss": 0.38969347476959226,
+ "num_input_tokens_seen": 318410480,
+ "step": 7280,
+ "train_runtime": 56255.4338,
+ "train_tokens_per_second": 5660.084
+ },
+ {
+ "epoch": 0.49050632911392406,
+ "grad_norm": 0.6692328837890084,
+ "learning_rate": 8.588231209313278e-06,
+ "loss": 0.37391541004180906,
+ "num_input_tokens_seen": 318654632,
+ "step": 7285,
+ "train_runtime": 56291.6831,
+ "train_tokens_per_second": 5660.776
+ },
+ {
+ "epoch": 0.4908429841098842,
+ "grad_norm": 0.7608072412958533,
+ "learning_rate": 8.586389348153192e-06,
+ "loss": 0.3896606206893921,
+ "num_input_tokens_seen": 318874824,
+ "step": 7290,
+ "train_runtime": 56326.4539,
+ "train_tokens_per_second": 5661.191
+ },
+ {
+ "epoch": 0.49117963910584433,
+ "grad_norm": 0.7736109340706484,
+ "learning_rate": 8.584546484070762e-06,
+ "loss": 0.40265254974365233,
+ "num_input_tokens_seen": 319084200,
+ "step": 7295,
+ "train_runtime": 56365.3177,
+ "train_tokens_per_second": 5661.002
+ },
+ {
+ "epoch": 0.4915162941018045,
+ "grad_norm": 0.7875873984707813,
+ "learning_rate": 8.582702617581338e-06,
+ "loss": 0.37858309745788576,
+ "num_input_tokens_seen": 319307784,
+ "step": 7300,
+ "train_runtime": 56403.5288,
+ "train_tokens_per_second": 5661.131
+ },
+ {
+ "epoch": 0.4918529490977646,
+ "grad_norm": 0.7183058860492505,
+ "learning_rate": 8.58085774920055e-06,
+ "loss": 0.38918490409851075,
+ "num_input_tokens_seen": 319513304,
+ "step": 7305,
+ "train_runtime": 56438.5945,
+ "train_tokens_per_second": 5661.256
+ },
+ {
+ "epoch": 0.49218960409372475,
+ "grad_norm": 0.7356534041270457,
+ "learning_rate": 8.579011879444313e-06,
+ "loss": 0.414321231842041,
+ "num_input_tokens_seen": 319743560,
+ "step": 7310,
+ "train_runtime": 56483.4424,
+ "train_tokens_per_second": 5660.837
+ },
+ {
+ "epoch": 0.4925262590896849,
+ "grad_norm": 0.6584093897897996,
+ "learning_rate": 8.577165008828817e-06,
+ "loss": 0.34825422763824465,
+ "num_input_tokens_seen": 319947112,
+ "step": 7315,
+ "train_runtime": 56521.6053,
+ "train_tokens_per_second": 5660.616
+ },
+ {
+ "epoch": 0.49286291408564503,
+ "grad_norm": 0.7722746099858304,
+ "learning_rate": 8.57531713787053e-06,
+ "loss": 0.3858429193496704,
+ "num_input_tokens_seen": 320155920,
+ "step": 7320,
+ "train_runtime": 56558.1587,
+ "train_tokens_per_second": 5660.65
+ },
+ {
+ "epoch": 0.49319956908160517,
+ "grad_norm": 0.8389191394295502,
+ "learning_rate": 8.573468267086208e-06,
+ "loss": 0.40245862007141114,
+ "num_input_tokens_seen": 320372792,
+ "step": 7325,
+ "train_runtime": 56592.9849,
+ "train_tokens_per_second": 5660.998
+ },
+ {
+ "epoch": 0.4935362240775653,
+ "grad_norm": 0.7633923302039957,
+ "learning_rate": 8.57161839699288e-06,
+ "loss": 0.36266283988952636,
+ "num_input_tokens_seen": 320592160,
+ "step": 7330,
+ "train_runtime": 56631.6072,
+ "train_tokens_per_second": 5661.011
+ },
+ {
+ "epoch": 0.49387287907352545,
+ "grad_norm": 0.8780389979845686,
+ "learning_rate": 8.569767528107857e-06,
+ "loss": 0.3856183052062988,
+ "num_input_tokens_seen": 320807512,
+ "step": 7335,
+ "train_runtime": 56669.1815,
+ "train_tokens_per_second": 5661.058
+ },
+ {
+ "epoch": 0.4942095340694856,
+ "grad_norm": 0.6800471334043537,
+ "learning_rate": 8.567915660948727e-06,
+ "loss": 0.3747032880783081,
+ "num_input_tokens_seen": 321031440,
+ "step": 7340,
+ "train_runtime": 56709.5179,
+ "train_tokens_per_second": 5660.98
+ },
+ {
+ "epoch": 0.49454618906544573,
+ "grad_norm": 0.8334581237618652,
+ "learning_rate": 8.566062796033358e-06,
+ "loss": 0.4063011646270752,
+ "num_input_tokens_seen": 321244008,
+ "step": 7345,
+ "train_runtime": 56746.898,
+ "train_tokens_per_second": 5660.997
+ },
+ {
+ "epoch": 0.49488284406140587,
+ "grad_norm": 0.7667199760691221,
+ "learning_rate": 8.564208933879903e-06,
+ "loss": 0.37364883422851564,
+ "num_input_tokens_seen": 321456992,
+ "step": 7350,
+ "train_runtime": 56787.18,
+ "train_tokens_per_second": 5660.732
+ },
+ {
+ "epoch": 0.495219499057366,
+ "grad_norm": 0.7580149996088524,
+ "learning_rate": 8.562354075006782e-06,
+ "loss": 0.3439052104949951,
+ "num_input_tokens_seen": 321675512,
+ "step": 7355,
+ "train_runtime": 56825.1931,
+ "train_tokens_per_second": 5660.79
+ },
+ {
+ "epoch": 0.49555615405332615,
+ "grad_norm": 0.6724287776480686,
+ "learning_rate": 8.560498219932706e-06,
+ "loss": 0.4257340431213379,
+ "num_input_tokens_seen": 321898448,
+ "step": 7360,
+ "train_runtime": 56859.2761,
+ "train_tokens_per_second": 5661.318
+ },
+ {
+ "epoch": 0.4958928090492863,
+ "grad_norm": 0.735779732523222,
+ "learning_rate": 8.558641369176654e-06,
+ "loss": 0.33791351318359375,
+ "num_input_tokens_seen": 322099904,
+ "step": 7365,
+ "train_runtime": 56892.2085,
+ "train_tokens_per_second": 5661.582
+ },
+ {
+ "epoch": 0.4962294640452464,
+ "grad_norm": 0.7979574200557189,
+ "learning_rate": 8.556783523257892e-06,
+ "loss": 0.3972716569900513,
+ "num_input_tokens_seen": 322310456,
+ "step": 7370,
+ "train_runtime": 56935.1565,
+ "train_tokens_per_second": 5661.009
+ },
+ {
+ "epoch": 0.49656611904120657,
+ "grad_norm": 0.7637866611420476,
+ "learning_rate": 8.554924682695959e-06,
+ "loss": 0.3530308246612549,
+ "num_input_tokens_seen": 322518744,
+ "step": 7375,
+ "train_runtime": 56971.9645,
+ "train_tokens_per_second": 5661.008
+ },
+ {
+ "epoch": 0.4969027740371667,
+ "grad_norm": 0.7640709256586813,
+ "learning_rate": 8.553064848010677e-06,
+ "loss": 0.3955545902252197,
+ "num_input_tokens_seen": 322725520,
+ "step": 7380,
+ "train_runtime": 57007.6953,
+ "train_tokens_per_second": 5661.087
+ },
+ {
+ "epoch": 0.49723942903312685,
+ "grad_norm": 0.8082539381866035,
+ "learning_rate": 8.551204019722137e-06,
+ "loss": 0.37126519680023196,
+ "num_input_tokens_seen": 322938184,
+ "step": 7385,
+ "train_runtime": 57046.2643,
+ "train_tokens_per_second": 5660.987
+ },
+ {
+ "epoch": 0.497576084029087,
+ "grad_norm": 0.7610367285948725,
+ "learning_rate": 8.549342198350718e-06,
+ "loss": 0.40738329887390134,
+ "num_input_tokens_seen": 323160704,
+ "step": 7390,
+ "train_runtime": 57086.0056,
+ "train_tokens_per_second": 5660.944
+ },
+ {
+ "epoch": 0.4979127390250471,
+ "grad_norm": 0.7004854714973048,
+ "learning_rate": 8.54747938441707e-06,
+ "loss": 0.36683225631713867,
+ "num_input_tokens_seen": 323382024,
+ "step": 7395,
+ "train_runtime": 57128.4646,
+ "train_tokens_per_second": 5660.611
+ },
+ {
+ "epoch": 0.49824939402100726,
+ "grad_norm": 0.6981945654130406,
+ "learning_rate": 8.545615578442126e-06,
+ "loss": 0.35772018432617186,
+ "num_input_tokens_seen": 323601056,
+ "step": 7400,
+ "train_runtime": 57166.4275,
+ "train_tokens_per_second": 5660.684
+ },
+ {
+ "epoch": 0.4985860490169674,
+ "grad_norm": 0.8538725530295028,
+ "learning_rate": 8.54375078094709e-06,
+ "loss": 0.3746786117553711,
+ "num_input_tokens_seen": 323824464,
+ "step": 7405,
+ "train_runtime": 57216.1649,
+ "train_tokens_per_second": 5659.667
+ },
+ {
+ "epoch": 0.49892270401292754,
+ "grad_norm": 0.6772326544260556,
+ "learning_rate": 8.541884992453449e-06,
+ "loss": 0.3711742401123047,
+ "num_input_tokens_seen": 324047472,
+ "step": 7410,
+ "train_runtime": 57254.2655,
+ "train_tokens_per_second": 5659.796
+ },
+ {
+ "epoch": 0.4992593590088877,
+ "grad_norm": 0.6789887270206421,
+ "learning_rate": 8.540018213482966e-06,
+ "loss": 0.37660138607025145,
+ "num_input_tokens_seen": 324280688,
+ "step": 7415,
+ "train_runtime": 57294.4005,
+ "train_tokens_per_second": 5659.902
+ },
+ {
+ "epoch": 0.4995960140048478,
+ "grad_norm": 0.7605591952463279,
+ "learning_rate": 8.538150444557676e-06,
+ "loss": 0.3920082330703735,
+ "num_input_tokens_seen": 324488104,
+ "step": 7420,
+ "train_runtime": 57333.8655,
+ "train_tokens_per_second": 5659.624
+ },
+ {
+ "epoch": 0.49993266900080796,
+ "grad_norm": 0.826315107413246,
+ "learning_rate": 8.536281686199896e-06,
+ "loss": 0.385677433013916,
+ "num_input_tokens_seen": 324711216,
+ "step": 7425,
+ "train_runtime": 57366.6278,
+ "train_tokens_per_second": 5660.281
+ },
+ {
+ "epoch": 0.5002693239967682,
+ "grad_norm": 0.6877893318576644,
+ "learning_rate": 8.534411938932218e-06,
+ "loss": 0.3787369728088379,
+ "num_input_tokens_seen": 324939096,
+ "step": 7430,
+ "train_runtime": 57400.8374,
+ "train_tokens_per_second": 5660.877
+ },
+ {
+ "epoch": 0.5006059789927283,
+ "grad_norm": 0.7206196203304496,
+ "learning_rate": 8.532541203277515e-06,
+ "loss": 0.3697648525238037,
+ "num_input_tokens_seen": 325163640,
+ "step": 7435,
+ "train_runtime": 57437.2769,
+ "train_tokens_per_second": 5661.195
+ },
+ {
+ "epoch": 0.5009426339886884,
+ "grad_norm": 0.8219400190681228,
+ "learning_rate": 8.530669479758926e-06,
+ "loss": 0.37598912715911864,
+ "num_input_tokens_seen": 325383192,
+ "step": 7440,
+ "train_runtime": 57475.956,
+ "train_tokens_per_second": 5661.205
+ },
+ {
+ "epoch": 0.5012792889846486,
+ "grad_norm": 0.8185284831231338,
+ "learning_rate": 8.528796768899878e-06,
+ "loss": 0.39255619049072266,
+ "num_input_tokens_seen": 325585216,
+ "step": 7445,
+ "train_runtime": 57516.4354,
+ "train_tokens_per_second": 5660.734
+ },
+ {
+ "epoch": 0.5016159439806087,
+ "grad_norm": 0.748875770223375,
+ "learning_rate": 8.526923071224064e-06,
+ "loss": 0.39842636585235597,
+ "num_input_tokens_seen": 325810472,
+ "step": 7450,
+ "train_runtime": 57549.4748,
+ "train_tokens_per_second": 5661.398
+ },
+ {
+ "epoch": 0.5019525989765689,
+ "grad_norm": 0.7148185230181737,
+ "learning_rate": 8.525048387255461e-06,
+ "loss": 0.3887434482574463,
+ "num_input_tokens_seen": 326034504,
+ "step": 7455,
+ "train_runtime": 57590.0671,
+ "train_tokens_per_second": 5661.298
+ },
+ {
+ "epoch": 0.502289253972529,
+ "grad_norm": 0.7229968516896057,
+ "learning_rate": 8.523172717518315e-06,
+ "loss": 0.3696585178375244,
+ "num_input_tokens_seen": 326245696,
+ "step": 7460,
+ "train_runtime": 57627.1988,
+ "train_tokens_per_second": 5661.314
+ },
+ {
+ "epoch": 0.5026259089684891,
+ "grad_norm": 0.7305037275102111,
+ "learning_rate": 8.521296062537156e-06,
+ "loss": 0.37876601219177247,
+ "num_input_tokens_seen": 326460104,
+ "step": 7465,
+ "train_runtime": 57664.9318,
+ "train_tokens_per_second": 5661.328
+ },
+ {
+ "epoch": 0.5029625639644493,
+ "grad_norm": 0.7847466467811594,
+ "learning_rate": 8.51941842283678e-06,
+ "loss": 0.36045174598693847,
+ "num_input_tokens_seen": 326673104,
+ "step": 7470,
+ "train_runtime": 57704.8246,
+ "train_tokens_per_second": 5661.106
+ },
+ {
+ "epoch": 0.5032992189604094,
+ "grad_norm": 0.991284631688998,
+ "learning_rate": 8.517539798942265e-06,
+ "loss": 0.3735339641571045,
+ "num_input_tokens_seen": 326906960,
+ "step": 7475,
+ "train_runtime": 57737.7344,
+ "train_tokens_per_second": 5661.929
+ },
+ {
+ "epoch": 0.5036358739563696,
+ "grad_norm": 0.6872665714663292,
+ "learning_rate": 8.515660191378962e-06,
+ "loss": 0.39676947593688966,
+ "num_input_tokens_seen": 327128416,
+ "step": 7480,
+ "train_runtime": 57776.877,
+ "train_tokens_per_second": 5661.926
+ },
+ {
+ "epoch": 0.5039725289523297,
+ "grad_norm": 0.6751937767879514,
+ "learning_rate": 8.513779600672495e-06,
+ "loss": 0.3535287380218506,
+ "num_input_tokens_seen": 327364152,
+ "step": 7485,
+ "train_runtime": 57811.2991,
+ "train_tokens_per_second": 5662.633
+ },
+ {
+ "epoch": 0.5043091839482898,
+ "grad_norm": 0.7000671921171517,
+ "learning_rate": 8.511898027348771e-06,
+ "loss": 0.37773268222808837,
+ "num_input_tokens_seen": 327568048,
+ "step": 7490,
+ "train_runtime": 57853.1955,
+ "train_tokens_per_second": 5662.056
+ },
+ {
+ "epoch": 0.50464583894425,
+ "grad_norm": 0.6936107754040254,
+ "learning_rate": 8.51001547193396e-06,
+ "loss": 0.3704256296157837,
+ "num_input_tokens_seen": 327792152,
+ "step": 7495,
+ "train_runtime": 57893.3414,
+ "train_tokens_per_second": 5662.001
+ },
+ {
+ "epoch": 0.5049824939402101,
+ "grad_norm": 0.7889020414532445,
+ "learning_rate": 8.508131934954515e-06,
+ "loss": 0.36852757930755614,
+ "num_input_tokens_seen": 327994400,
+ "step": 7500,
+ "train_runtime": 57932.8569,
+ "train_tokens_per_second": 5661.63
+ },
+ {
+ "epoch": 0.5053191489361702,
+ "grad_norm": 0.715078307452266,
+ "learning_rate": 8.50624741693716e-06,
+ "loss": 0.3951726913452148,
+ "num_input_tokens_seen": 328200448,
+ "step": 7505,
+ "train_runtime": 57971.7531,
+ "train_tokens_per_second": 5661.386
+ },
+ {
+ "epoch": 0.5056558039321304,
+ "grad_norm": 1.068005291058121,
+ "learning_rate": 8.5043619184089e-06,
+ "loss": 0.4119250297546387,
+ "num_input_tokens_seen": 328421224,
+ "step": 7510,
+ "train_runtime": 58005.869,
+ "train_tokens_per_second": 5661.862
+ },
+ {
+ "epoch": 0.5059924589280905,
+ "grad_norm": 1.638455940254133,
+ "learning_rate": 8.502475439897e-06,
+ "loss": 0.4078369140625,
+ "num_input_tokens_seen": 328646192,
+ "step": 7515,
+ "train_runtime": 58043.9094,
+ "train_tokens_per_second": 5662.027
+ },
+ {
+ "epoch": 0.5063291139240507,
+ "grad_norm": 0.7358398144623932,
+ "learning_rate": 8.500587981929013e-06,
+ "loss": 0.3898489475250244,
+ "num_input_tokens_seen": 328867064,
+ "step": 7520,
+ "train_runtime": 58088.0612,
+ "train_tokens_per_second": 5661.526
+ },
+ {
+ "epoch": 0.5066657689200108,
+ "grad_norm": 0.7487957564819914,
+ "learning_rate": 8.498699545032758e-06,
+ "loss": 0.38454551696777345,
+ "num_input_tokens_seen": 329103168,
+ "step": 7525,
+ "train_runtime": 58124.1106,
+ "train_tokens_per_second": 5662.077
+ },
+ {
+ "epoch": 0.507002423915971,
+ "grad_norm": 0.7172633060502269,
+ "learning_rate": 8.496810129736334e-06,
+ "loss": 0.3850215435028076,
+ "num_input_tokens_seen": 329322920,
+ "step": 7530,
+ "train_runtime": 58166.9838,
+ "train_tokens_per_second": 5661.681
+ },
+ {
+ "epoch": 0.5073390789119311,
+ "grad_norm": 0.7747557646585903,
+ "learning_rate": 8.494919736568105e-06,
+ "loss": 0.3788848638534546,
+ "num_input_tokens_seen": 329533896,
+ "step": 7535,
+ "train_runtime": 58210.0232,
+ "train_tokens_per_second": 5661.119
+ },
+ {
+ "epoch": 0.5076757339078912,
+ "grad_norm": 0.7933496638959117,
+ "learning_rate": 8.49302836605672e-06,
+ "loss": 0.37194099426269533,
+ "num_input_tokens_seen": 329760624,
+ "step": 7540,
+ "train_runtime": 58246.3752,
+ "train_tokens_per_second": 5661.479
+ },
+ {
+ "epoch": 0.5080123889038514,
+ "grad_norm": 0.7365707403857205,
+ "learning_rate": 8.491136018731088e-06,
+ "loss": 0.3646608114242554,
+ "num_input_tokens_seen": 329984728,
+ "step": 7545,
+ "train_runtime": 58287.6488,
+ "train_tokens_per_second": 5661.315
+ },
+ {
+ "epoch": 0.5083490438998115,
+ "grad_norm": 0.8956737633683735,
+ "learning_rate": 8.4892426951204e-06,
+ "loss": 0.3987517118453979,
+ "num_input_tokens_seen": 330206096,
+ "step": 7550,
+ "train_runtime": 58316.2366,
+ "train_tokens_per_second": 5662.335
+ },
+ {
+ "epoch": 0.5086856988957716,
+ "grad_norm": 0.7170163461314129,
+ "learning_rate": 8.48734839575412e-06,
+ "loss": 0.3717711687088013,
+ "num_input_tokens_seen": 330437336,
+ "step": 7555,
+ "train_runtime": 58351.3409,
+ "train_tokens_per_second": 5662.892
+ },
+ {
+ "epoch": 0.5090223538917318,
+ "grad_norm": 0.6630558999546488,
+ "learning_rate": 8.485453121161983e-06,
+ "loss": 0.37380030155181887,
+ "num_input_tokens_seen": 330653616,
+ "step": 7560,
+ "train_runtime": 58392.093,
+ "train_tokens_per_second": 5662.644
+ },
+ {
+ "epoch": 0.5093590088876919,
+ "grad_norm": 0.7845896930720607,
+ "learning_rate": 8.483556871873993e-06,
+ "loss": 0.42113819122314455,
+ "num_input_tokens_seen": 330875672,
+ "step": 7565,
+ "train_runtime": 58433.2419,
+ "train_tokens_per_second": 5662.456
+ },
+ {
+ "epoch": 0.5096956638836521,
+ "grad_norm": 0.760051472590646,
+ "learning_rate": 8.481659648420433e-06,
+ "loss": 0.3786670207977295,
+ "num_input_tokens_seen": 331096488,
+ "step": 7570,
+ "train_runtime": 58475.0649,
+ "train_tokens_per_second": 5662.182
+ },
+ {
+ "epoch": 0.5100323188796122,
+ "grad_norm": 0.8835596263525625,
+ "learning_rate": 8.479761451331855e-06,
+ "loss": 0.3809337615966797,
+ "num_input_tokens_seen": 331285184,
+ "step": 7575,
+ "train_runtime": 58515.9798,
+ "train_tokens_per_second": 5661.448
+ },
+ {
+ "epoch": 0.5103689738755723,
+ "grad_norm": 0.9301429501501213,
+ "learning_rate": 8.477862281139082e-06,
+ "loss": 0.3540458917617798,
+ "num_input_tokens_seen": 331501760,
+ "step": 7580,
+ "train_runtime": 58558.4957,
+ "train_tokens_per_second": 5661.036
+ },
+ {
+ "epoch": 0.5107056288715325,
+ "grad_norm": 0.7609757852463125,
+ "learning_rate": 8.475962138373212e-06,
+ "loss": 0.3986123323440552,
+ "num_input_tokens_seen": 331702152,
+ "step": 7585,
+ "train_runtime": 58591.7614,
+ "train_tokens_per_second": 5661.242
+ },
+ {
+ "epoch": 0.5110422838674926,
+ "grad_norm": 0.6930333171792868,
+ "learning_rate": 8.474061023565614e-06,
+ "loss": 0.3940688371658325,
+ "num_input_tokens_seen": 331929992,
+ "step": 7590,
+ "train_runtime": 58631.4664,
+ "train_tokens_per_second": 5661.294
+ },
+ {
+ "epoch": 0.5113789388634528,
+ "grad_norm": 0.7801001592101691,
+ "learning_rate": 8.472158937247931e-06,
+ "loss": 0.40605816841125486,
+ "num_input_tokens_seen": 332148472,
+ "step": 7595,
+ "train_runtime": 58666.3698,
+ "train_tokens_per_second": 5661.65
+ },
+ {
+ "epoch": 0.5117155938594129,
+ "grad_norm": 0.7185541728168203,
+ "learning_rate": 8.47025587995207e-06,
+ "loss": 0.38570313453674315,
+ "num_input_tokens_seen": 332362728,
+ "step": 7600,
+ "train_runtime": 58704.4102,
+ "train_tokens_per_second": 5661.631
+ },
+ {
+ "epoch": 0.512052248855373,
+ "grad_norm": 0.7887783994842774,
+ "learning_rate": 8.46835185221022e-06,
+ "loss": 0.39829018115997317,
+ "num_input_tokens_seen": 332569544,
+ "step": 7605,
+ "train_runtime": 58743.9844,
+ "train_tokens_per_second": 5661.338
+ },
+ {
+ "epoch": 0.5123889038513332,
+ "grad_norm": 0.8602264236155862,
+ "learning_rate": 8.466446854554835e-06,
+ "loss": 0.3549684762954712,
+ "num_input_tokens_seen": 332778920,
+ "step": 7610,
+ "train_runtime": 58780.8818,
+ "train_tokens_per_second": 5661.346
+ },
+ {
+ "epoch": 0.5127255588472933,
+ "grad_norm": 0.7228150824538518,
+ "learning_rate": 8.464540887518638e-06,
+ "loss": 0.37906813621520996,
+ "num_input_tokens_seen": 332984080,
+ "step": 7615,
+ "train_runtime": 58823.2164,
+ "train_tokens_per_second": 5660.759
+ },
+ {
+ "epoch": 0.5130622138432535,
+ "grad_norm": 0.7996970703617401,
+ "learning_rate": 8.462633951634631e-06,
+ "loss": 0.3685762405395508,
+ "num_input_tokens_seen": 333202640,
+ "step": 7620,
+ "train_runtime": 58868.3048,
+ "train_tokens_per_second": 5660.136
+ },
+ {
+ "epoch": 0.5133988688392136,
+ "grad_norm": 0.6901634902422549,
+ "learning_rate": 8.46072604743608e-06,
+ "loss": 0.35547614097595215,
+ "num_input_tokens_seen": 333413904,
+ "step": 7625,
+ "train_runtime": 58905.0124,
+ "train_tokens_per_second": 5660.196
+ },
+ {
+ "epoch": 0.5137355238351737,
+ "grad_norm": 0.9033857052689097,
+ "learning_rate": 8.458817175456528e-06,
+ "loss": 0.3930670261383057,
+ "num_input_tokens_seen": 333630616,
+ "step": 7630,
+ "train_runtime": 58947.1829,
+ "train_tokens_per_second": 5659.823
+ },
+ {
+ "epoch": 0.5140721788311339,
+ "grad_norm": 0.7781771666334231,
+ "learning_rate": 8.45690733622978e-06,
+ "loss": 0.35927329063415525,
+ "num_input_tokens_seen": 333854600,
+ "step": 7635,
+ "train_runtime": 58986.6913,
+ "train_tokens_per_second": 5659.829
+ },
+ {
+ "epoch": 0.514408833827094,
+ "grad_norm": 0.8455368762157274,
+ "learning_rate": 8.45499653028992e-06,
+ "loss": 0.36787919998168944,
+ "num_input_tokens_seen": 334071192,
+ "step": 7640,
+ "train_runtime": 59022.2041,
+ "train_tokens_per_second": 5660.093
+ },
+ {
+ "epoch": 0.5147454888230542,
+ "grad_norm": 0.7586644939540588,
+ "learning_rate": 8.453084758171297e-06,
+ "loss": 0.37957746982574464,
+ "num_input_tokens_seen": 334291888,
+ "step": 7645,
+ "train_runtime": 59067.1738,
+ "train_tokens_per_second": 5659.521
+ },
+ {
+ "epoch": 0.5150821438190143,
+ "grad_norm": 0.7524816129003911,
+ "learning_rate": 8.451172020408532e-06,
+ "loss": 0.38945965766906737,
+ "num_input_tokens_seen": 334516424,
+ "step": 7650,
+ "train_runtime": 59101.459,
+ "train_tokens_per_second": 5660.037
+ },
+ {
+ "epoch": 0.5154187988149744,
+ "grad_norm": 0.7137520182282876,
+ "learning_rate": 8.449258317536517e-06,
+ "loss": 0.35926315784454343,
+ "num_input_tokens_seen": 334732808,
+ "step": 7655,
+ "train_runtime": 59134.7468,
+ "train_tokens_per_second": 5660.51
+ },
+ {
+ "epoch": 0.5157554538109346,
+ "grad_norm": 0.7129464795573726,
+ "learning_rate": 8.447343650090412e-06,
+ "loss": 0.35725347995758056,
+ "num_input_tokens_seen": 334954280,
+ "step": 7660,
+ "train_runtime": 59172.9555,
+ "train_tokens_per_second": 5660.597
+ },
+ {
+ "epoch": 0.5160921088068947,
+ "grad_norm": 0.7272448686794939,
+ "learning_rate": 8.445428018605647e-06,
+ "loss": 0.406490421295166,
+ "num_input_tokens_seen": 335178568,
+ "step": 7665,
+ "train_runtime": 59210.2513,
+ "train_tokens_per_second": 5660.82
+ },
+ {
+ "epoch": 0.5164287638028549,
+ "grad_norm": 0.6593329295977397,
+ "learning_rate": 8.443511423617924e-06,
+ "loss": 0.34410724639892576,
+ "num_input_tokens_seen": 335398608,
+ "step": 7670,
+ "train_runtime": 59246.6136,
+ "train_tokens_per_second": 5661.06
+ },
+ {
+ "epoch": 0.516765418798815,
+ "grad_norm": 0.7054753575271266,
+ "learning_rate": 8.441593865663213e-06,
+ "loss": 0.39874303340911865,
+ "num_input_tokens_seen": 335621696,
+ "step": 7675,
+ "train_runtime": 59285.0684,
+ "train_tokens_per_second": 5661.151
+ },
+ {
+ "epoch": 0.5171020737947751,
+ "grad_norm": 0.7429064696844442,
+ "learning_rate": 8.439675345277751e-06,
+ "loss": 0.40140485763549805,
+ "num_input_tokens_seen": 335839504,
+ "step": 7680,
+ "train_runtime": 59327.3379,
+ "train_tokens_per_second": 5660.788
+ },
+ {
+ "epoch": 0.5174387287907353,
+ "grad_norm": 0.6392657750321254,
+ "learning_rate": 8.437755862998047e-06,
+ "loss": 0.3619039297103882,
+ "num_input_tokens_seen": 336062080,
+ "step": 7685,
+ "train_runtime": 59362.2516,
+ "train_tokens_per_second": 5661.208
+ },
+ {
+ "epoch": 0.5177753837866954,
+ "grad_norm": 0.6333778064119364,
+ "learning_rate": 8.435835419360878e-06,
+ "loss": 0.3882591247558594,
+ "num_input_tokens_seen": 336299944,
+ "step": 7690,
+ "train_runtime": 59398.7224,
+ "train_tokens_per_second": 5661.737
+ },
+ {
+ "epoch": 0.5181120387826555,
+ "grad_norm": 0.7397126752113882,
+ "learning_rate": 8.433914014903291e-06,
+ "loss": 0.42388019561767576,
+ "num_input_tokens_seen": 336514968,
+ "step": 7695,
+ "train_runtime": 59437.773,
+ "train_tokens_per_second": 5661.635
+ },
+ {
+ "epoch": 0.5184486937786157,
+ "grad_norm": 0.6888095905405301,
+ "learning_rate": 8.431991650162599e-06,
+ "loss": 0.37606000900268555,
+ "num_input_tokens_seen": 336732600,
+ "step": 7700,
+ "train_runtime": 59483.4411,
+ "train_tokens_per_second": 5660.947
+ },
+ {
+ "epoch": 0.5187853487745758,
+ "grad_norm": 0.6980617402046243,
+ "learning_rate": 8.430068325676386e-06,
+ "loss": 0.39625885486602785,
+ "num_input_tokens_seen": 336948776,
+ "step": 7705,
+ "train_runtime": 59521.0447,
+ "train_tokens_per_second": 5661.002
+ },
+ {
+ "epoch": 0.519122003770536,
+ "grad_norm": 0.732968462897192,
+ "learning_rate": 8.428144041982502e-06,
+ "loss": 0.39479496479034426,
+ "num_input_tokens_seen": 337163520,
+ "step": 7710,
+ "train_runtime": 59556.4,
+ "train_tokens_per_second": 5661.247
+ },
+ {
+ "epoch": 0.5194586587664961,
+ "grad_norm": 0.6401198443952045,
+ "learning_rate": 8.42621879961907e-06,
+ "loss": 0.38160171508789065,
+ "num_input_tokens_seen": 337382992,
+ "step": 7715,
+ "train_runtime": 59597.2808,
+ "train_tokens_per_second": 5661.047
+ },
+ {
+ "epoch": 0.5197953137624562,
+ "grad_norm": 0.7011673418789361,
+ "learning_rate": 8.424292599124476e-06,
+ "loss": 0.3532627820968628,
+ "num_input_tokens_seen": 337609304,
+ "step": 7720,
+ "train_runtime": 59636.4965,
+ "train_tokens_per_second": 5661.119
+ },
+ {
+ "epoch": 0.5201319687584164,
+ "grad_norm": 0.7609717187854771,
+ "learning_rate": 8.422365441037374e-06,
+ "loss": 0.39478967189788816,
+ "num_input_tokens_seen": 337838072,
+ "step": 7725,
+ "train_runtime": 59669.1056,
+ "train_tokens_per_second": 5661.859
+ },
+ {
+ "epoch": 0.5204686237543765,
+ "grad_norm": 0.743645872276741,
+ "learning_rate": 8.420437325896692e-06,
+ "loss": 0.3906223297119141,
+ "num_input_tokens_seen": 338067512,
+ "step": 7730,
+ "train_runtime": 59705.1498,
+ "train_tokens_per_second": 5662.284
+ },
+ {
+ "epoch": 0.5208052787503367,
+ "grad_norm": 0.7586775310938808,
+ "learning_rate": 8.418508254241617e-06,
+ "loss": 0.3925325870513916,
+ "num_input_tokens_seen": 338284504,
+ "step": 7735,
+ "train_runtime": 59748.0729,
+ "train_tokens_per_second": 5661.848
+ },
+ {
+ "epoch": 0.5211419337462968,
+ "grad_norm": 0.8223238925961407,
+ "learning_rate": 8.416578226611612e-06,
+ "loss": 0.35642185211181643,
+ "num_input_tokens_seen": 338495440,
+ "step": 7740,
+ "train_runtime": 59783.2802,
+ "train_tokens_per_second": 5662.042
+ },
+ {
+ "epoch": 0.5214785887422569,
+ "grad_norm": 0.8532576047124709,
+ "learning_rate": 8.414647243546398e-06,
+ "loss": 0.38446238040924074,
+ "num_input_tokens_seen": 338717192,
+ "step": 7745,
+ "train_runtime": 59821.3869,
+ "train_tokens_per_second": 5662.142
+ },
+ {
+ "epoch": 0.5218152437382171,
+ "grad_norm": 0.6899241307861492,
+ "learning_rate": 8.412715305585973e-06,
+ "loss": 0.367059588432312,
+ "num_input_tokens_seen": 338946888,
+ "step": 7750,
+ "train_runtime": 59859.9792,
+ "train_tokens_per_second": 5662.329
+ },
+ {
+ "epoch": 0.5221518987341772,
+ "grad_norm": 0.7311441327867719,
+ "learning_rate": 8.410782413270594e-06,
+ "loss": 0.3799006462097168,
+ "num_input_tokens_seen": 339172224,
+ "step": 7755,
+ "train_runtime": 59900.1807,
+ "train_tokens_per_second": 5662.29
+ },
+ {
+ "epoch": 0.5224885537301374,
+ "grad_norm": 0.9420308727064326,
+ "learning_rate": 8.408848567140787e-06,
+ "loss": 0.37287948131561277,
+ "num_input_tokens_seen": 339364672,
+ "step": 7760,
+ "train_runtime": 59941.0763,
+ "train_tokens_per_second": 5661.638
+ },
+ {
+ "epoch": 0.5228252087260975,
+ "grad_norm": 0.7245874800766633,
+ "learning_rate": 8.406913767737353e-06,
+ "loss": 0.37642457485198977,
+ "num_input_tokens_seen": 339579064,
+ "step": 7765,
+ "train_runtime": 59982.2891,
+ "train_tokens_per_second": 5661.322
+ },
+ {
+ "epoch": 0.5231618637220576,
+ "grad_norm": 0.7199997285512201,
+ "learning_rate": 8.404978015601344e-06,
+ "loss": 0.3964075565338135,
+ "num_input_tokens_seen": 339801840,
+ "step": 7770,
+ "train_runtime": 60029.6628,
+ "train_tokens_per_second": 5660.566
+ },
+ {
+ "epoch": 0.5234985187180178,
+ "grad_norm": 0.697918705663631,
+ "learning_rate": 8.403041311274091e-06,
+ "loss": 0.39160215854644775,
+ "num_input_tokens_seen": 340025832,
+ "step": 7775,
+ "train_runtime": 60071.2747,
+ "train_tokens_per_second": 5660.373
+ },
+ {
+ "epoch": 0.5238351737139779,
+ "grad_norm": 0.8157700767931442,
+ "learning_rate": 8.401103655297188e-06,
+ "loss": 0.3895624876022339,
+ "num_input_tokens_seen": 340239616,
+ "step": 7780,
+ "train_runtime": 60105.9113,
+ "train_tokens_per_second": 5660.668
+ },
+ {
+ "epoch": 0.5241718287099381,
+ "grad_norm": 0.838923642880136,
+ "learning_rate": 8.399165048212494e-06,
+ "loss": 0.3589164257049561,
+ "num_input_tokens_seen": 340454984,
+ "step": 7785,
+ "train_runtime": 60145.3789,
+ "train_tokens_per_second": 5660.534
+ },
+ {
+ "epoch": 0.5245084837058982,
+ "grad_norm": 0.7246920315479125,
+ "learning_rate": 8.397225490562131e-06,
+ "loss": 0.4032485008239746,
+ "num_input_tokens_seen": 340684288,
+ "step": 7790,
+ "train_runtime": 60180.0663,
+ "train_tokens_per_second": 5661.082
+ },
+ {
+ "epoch": 0.5248451387018583,
+ "grad_norm": 0.7826576263190983,
+ "learning_rate": 8.395284982888494e-06,
+ "loss": 0.3588679313659668,
+ "num_input_tokens_seen": 340904512,
+ "step": 7795,
+ "train_runtime": 60216.628,
+ "train_tokens_per_second": 5661.302
+ },
+ {
+ "epoch": 0.5251817936978185,
+ "grad_norm": 0.7381573756482819,
+ "learning_rate": 8.393343525734239e-06,
+ "loss": 0.39098711013793946,
+ "num_input_tokens_seen": 341128688,
+ "step": 7800,
+ "train_runtime": 60254.8697,
+ "train_tokens_per_second": 5661.429
+ },
+ {
+ "epoch": 0.5255184486937786,
+ "grad_norm": 0.9510438503863269,
+ "learning_rate": 8.391401119642286e-06,
+ "loss": 0.39932818412780763,
+ "num_input_tokens_seen": 341326256,
+ "step": 7805,
+ "train_runtime": 60302.5843,
+ "train_tokens_per_second": 5660.226
+ },
+ {
+ "epoch": 0.5258551036897388,
+ "grad_norm": 0.7551888581389101,
+ "learning_rate": 8.389457765155824e-06,
+ "loss": 0.40273380279541016,
+ "num_input_tokens_seen": 341526840,
+ "step": 7810,
+ "train_runtime": 60347.108,
+ "train_tokens_per_second": 5659.374
+ },
+ {
+ "epoch": 0.5261917586856989,
+ "grad_norm": 0.7817537228902953,
+ "learning_rate": 8.387513462818309e-06,
+ "loss": 0.36842942237854004,
+ "num_input_tokens_seen": 341747720,
+ "step": 7815,
+ "train_runtime": 60388.0509,
+ "train_tokens_per_second": 5659.194
+ },
+ {
+ "epoch": 0.526528413681659,
+ "grad_norm": 0.76396158135225,
+ "learning_rate": 8.385568213173453e-06,
+ "loss": 0.3687218427658081,
+ "num_input_tokens_seen": 341961736,
+ "step": 7820,
+ "train_runtime": 60430.695,
+ "train_tokens_per_second": 5658.742
+ },
+ {
+ "epoch": 0.5268650686776192,
+ "grad_norm": 1.0538784566401151,
+ "learning_rate": 8.383622016765242e-06,
+ "loss": 0.3595340013504028,
+ "num_input_tokens_seen": 342180688,
+ "step": 7825,
+ "train_runtime": 60474.2774,
+ "train_tokens_per_second": 5658.285
+ },
+ {
+ "epoch": 0.5272017236735793,
+ "grad_norm": 0.7683842178996437,
+ "learning_rate": 8.381674874137926e-06,
+ "loss": 0.4014129638671875,
+ "num_input_tokens_seen": 342390504,
+ "step": 7830,
+ "train_runtime": 60507.7742,
+ "train_tokens_per_second": 5658.62
+ },
+ {
+ "epoch": 0.5275383786695395,
+ "grad_norm": 0.7767541075637882,
+ "learning_rate": 8.379726785836013e-06,
+ "loss": 0.3654167652130127,
+ "num_input_tokens_seen": 342602080,
+ "step": 7835,
+ "train_runtime": 60550.4658,
+ "train_tokens_per_second": 5658.125
+ },
+ {
+ "epoch": 0.5278750336654996,
+ "grad_norm": 0.6734458895156976,
+ "learning_rate": 8.377777752404283e-06,
+ "loss": 0.36977086067199705,
+ "num_input_tokens_seen": 342831800,
+ "step": 7840,
+ "train_runtime": 60591.1256,
+ "train_tokens_per_second": 5658.119
+ },
+ {
+ "epoch": 0.5282116886614597,
+ "grad_norm": 0.8168160522078569,
+ "learning_rate": 8.375827774387774e-06,
+ "loss": 0.34728164672851564,
+ "num_input_tokens_seen": 343029568,
+ "step": 7845,
+ "train_runtime": 60633.6042,
+ "train_tokens_per_second": 5657.417
+ },
+ {
+ "epoch": 0.5285483436574199,
+ "grad_norm": 0.6622532156086101,
+ "learning_rate": 8.373876852331793e-06,
+ "loss": 0.3438518762588501,
+ "num_input_tokens_seen": 343254472,
+ "step": 7850,
+ "train_runtime": 60671.9643,
+ "train_tokens_per_second": 5657.547
+ },
+ {
+ "epoch": 0.52888499865338,
+ "grad_norm": 0.7502437558347258,
+ "learning_rate": 8.371924986781909e-06,
+ "loss": 0.38301682472229004,
+ "num_input_tokens_seen": 343470008,
+ "step": 7855,
+ "train_runtime": 60709.3414,
+ "train_tokens_per_second": 5657.614
+ },
+ {
+ "epoch": 0.5292216536493402,
+ "grad_norm": 0.7587114327145728,
+ "learning_rate": 8.369972178283955e-06,
+ "loss": 0.36738762855529783,
+ "num_input_tokens_seen": 343698072,
+ "step": 7860,
+ "train_runtime": 60747.59,
+ "train_tokens_per_second": 5657.806
+ },
+ {
+ "epoch": 0.5295583086453003,
+ "grad_norm": 0.6881092460250557,
+ "learning_rate": 8.368018427384027e-06,
+ "loss": 0.38115482330322265,
+ "num_input_tokens_seen": 343919960,
+ "step": 7865,
+ "train_runtime": 60791.0097,
+ "train_tokens_per_second": 5657.415
+ },
+ {
+ "epoch": 0.5298949636412604,
+ "grad_norm": 0.8355237541845086,
+ "learning_rate": 8.366063734628485e-06,
+ "loss": 0.38156135082244874,
+ "num_input_tokens_seen": 344123568,
+ "step": 7870,
+ "train_runtime": 60827.6415,
+ "train_tokens_per_second": 5657.355
+ },
+ {
+ "epoch": 0.5302316186372206,
+ "grad_norm": 0.8303838221096392,
+ "learning_rate": 8.364108100563955e-06,
+ "loss": 0.4007874011993408,
+ "num_input_tokens_seen": 344341784,
+ "step": 7875,
+ "train_runtime": 60864.4296,
+ "train_tokens_per_second": 5657.521
+ },
+ {
+ "epoch": 0.5305682736331807,
+ "grad_norm": 0.7925795609225155,
+ "learning_rate": 8.362151525737321e-06,
+ "loss": 0.4212652206420898,
+ "num_input_tokens_seen": 344562552,
+ "step": 7880,
+ "train_runtime": 60901.6398,
+ "train_tokens_per_second": 5657.689
+ },
+ {
+ "epoch": 0.5309049286291408,
+ "grad_norm": 0.7468959848481416,
+ "learning_rate": 8.360194010695733e-06,
+ "loss": 0.3570136547088623,
+ "num_input_tokens_seen": 344782336,
+ "step": 7885,
+ "train_runtime": 60942.4135,
+ "train_tokens_per_second": 5657.51
+ },
+ {
+ "epoch": 0.531241583625101,
+ "grad_norm": 0.7105479533553207,
+ "learning_rate": 8.358235555986607e-06,
+ "loss": 0.39661598205566406,
+ "num_input_tokens_seen": 345008672,
+ "step": 7890,
+ "train_runtime": 60976.8653,
+ "train_tokens_per_second": 5658.026
+ },
+ {
+ "epoch": 0.5315782386210611,
+ "grad_norm": 0.768825896681359,
+ "learning_rate": 8.356276162157613e-06,
+ "loss": 0.3597188234329224,
+ "num_input_tokens_seen": 345216360,
+ "step": 7895,
+ "train_runtime": 61021.9355,
+ "train_tokens_per_second": 5657.25
+ },
+ {
+ "epoch": 0.5319148936170213,
+ "grad_norm": 0.711887791776799,
+ "learning_rate": 8.354315829756696e-06,
+ "loss": 0.37782256603240966,
+ "num_input_tokens_seen": 345442968,
+ "step": 7900,
+ "train_runtime": 61056.926,
+ "train_tokens_per_second": 5657.72
+ },
+ {
+ "epoch": 0.5322515486129814,
+ "grad_norm": 0.7271488313267995,
+ "learning_rate": 8.352354559332053e-06,
+ "loss": 0.36695542335510256,
+ "num_input_tokens_seen": 345663248,
+ "step": 7905,
+ "train_runtime": 61093.5998,
+ "train_tokens_per_second": 5657.929
+ },
+ {
+ "epoch": 0.5325882036089415,
+ "grad_norm": 0.7790624119649752,
+ "learning_rate": 8.350392351432146e-06,
+ "loss": 0.3877138376235962,
+ "num_input_tokens_seen": 345877640,
+ "step": 7910,
+ "train_runtime": 61125.7433,
+ "train_tokens_per_second": 5658.461
+ },
+ {
+ "epoch": 0.5329248586049017,
+ "grad_norm": 0.652480198172122,
+ "learning_rate": 8.348429206605702e-06,
+ "loss": 0.3623414754867554,
+ "num_input_tokens_seen": 346113080,
+ "step": 7915,
+ "train_runtime": 61170.1897,
+ "train_tokens_per_second": 5658.199
+ },
+ {
+ "epoch": 0.5332615136008618,
+ "grad_norm": 0.8166602948279773,
+ "learning_rate": 8.346465125401706e-06,
+ "loss": 0.41824941635131835,
+ "num_input_tokens_seen": 346331632,
+ "step": 7920,
+ "train_runtime": 61213.7632,
+ "train_tokens_per_second": 5657.741
+ },
+ {
+ "epoch": 0.533598168596822,
+ "grad_norm": 0.7338720699869766,
+ "learning_rate": 8.344500108369412e-06,
+ "loss": 0.3689335107803345,
+ "num_input_tokens_seen": 346540104,
+ "step": 7925,
+ "train_runtime": 61254.7218,
+ "train_tokens_per_second": 5657.361
+ },
+ {
+ "epoch": 0.5339348235927821,
+ "grad_norm": 0.7241659753104186,
+ "learning_rate": 8.342534156058326e-06,
+ "loss": 0.3848552942276001,
+ "num_input_tokens_seen": 346759048,
+ "step": 7930,
+ "train_runtime": 61293.0183,
+ "train_tokens_per_second": 5657.399
+ },
+ {
+ "epoch": 0.5342714785887422,
+ "grad_norm": 0.7284900345851684,
+ "learning_rate": 8.34056726901822e-06,
+ "loss": 0.39329664707183837,
+ "num_input_tokens_seen": 346971992,
+ "step": 7935,
+ "train_runtime": 61333.9961,
+ "train_tokens_per_second": 5657.091
+ },
+ {
+ "epoch": 0.5346081335847024,
+ "grad_norm": 0.9343840115025148,
+ "learning_rate": 8.338599447799132e-06,
+ "loss": 0.37961349487304685,
+ "num_input_tokens_seen": 347176784,
+ "step": 7940,
+ "train_runtime": 61367.1004,
+ "train_tokens_per_second": 5657.376
+ },
+ {
+ "epoch": 0.5349447885806625,
+ "grad_norm": 0.7650504516817562,
+ "learning_rate": 8.336630692951353e-06,
+ "loss": 0.3706023931503296,
+ "num_input_tokens_seen": 347396336,
+ "step": 7945,
+ "train_runtime": 61405.4086,
+ "train_tokens_per_second": 5657.422
+ },
+ {
+ "epoch": 0.5352814435766227,
+ "grad_norm": 0.7809182592661493,
+ "learning_rate": 8.334661005025441e-06,
+ "loss": 0.3996375560760498,
+ "num_input_tokens_seen": 347612392,
+ "step": 7950,
+ "train_runtime": 61443.353,
+ "train_tokens_per_second": 5657.445
+ },
+ {
+ "epoch": 0.5356180985725828,
+ "grad_norm": 1.0173782664960012,
+ "learning_rate": 8.332690384572213e-06,
+ "loss": 0.4039102077484131,
+ "num_input_tokens_seen": 347827352,
+ "step": 7955,
+ "train_runtime": 61481.7721,
+ "train_tokens_per_second": 5657.406
+ },
+ {
+ "epoch": 0.5359547535685429,
+ "grad_norm": 0.7319934278286511,
+ "learning_rate": 8.330718832142746e-06,
+ "loss": 0.3981315612792969,
+ "num_input_tokens_seen": 348060432,
+ "step": 7960,
+ "train_runtime": 61520.2514,
+ "train_tokens_per_second": 5657.656
+ },
+ {
+ "epoch": 0.5362914085645031,
+ "grad_norm": 0.6935185521555359,
+ "learning_rate": 8.328746348288379e-06,
+ "loss": 0.3804043769836426,
+ "num_input_tokens_seen": 348266616,
+ "step": 7965,
+ "train_runtime": 61562.5177,
+ "train_tokens_per_second": 5657.121
+ },
+ {
+ "epoch": 0.5366280635604632,
+ "grad_norm": 0.7663090177174399,
+ "learning_rate": 8.32677293356071e-06,
+ "loss": 0.42568235397338866,
+ "num_input_tokens_seen": 348498272,
+ "step": 7970,
+ "train_runtime": 61598.1572,
+ "train_tokens_per_second": 5657.609
+ },
+ {
+ "epoch": 0.5369647185564234,
+ "grad_norm": 0.8375018048980476,
+ "learning_rate": 8.324798588511598e-06,
+ "loss": 0.3702836275100708,
+ "num_input_tokens_seen": 348696416,
+ "step": 7975,
+ "train_runtime": 61635.4226,
+ "train_tokens_per_second": 5657.403
+ },
+ {
+ "epoch": 0.5373013735523835,
+ "grad_norm": 0.8187637529093583,
+ "learning_rate": 8.322823313693162e-06,
+ "loss": 0.38841800689697265,
+ "num_input_tokens_seen": 348910592,
+ "step": 7980,
+ "train_runtime": 61671.4891,
+ "train_tokens_per_second": 5657.567
+ },
+ {
+ "epoch": 0.5376380285483436,
+ "grad_norm": 0.7918621668094143,
+ "learning_rate": 8.320847109657782e-06,
+ "loss": 0.3783905506134033,
+ "num_input_tokens_seen": 349120256,
+ "step": 7985,
+ "train_runtime": 61712.9559,
+ "train_tokens_per_second": 5657.163
+ },
+ {
+ "epoch": 0.5379746835443038,
+ "grad_norm": 0.7362861473446518,
+ "learning_rate": 8.318869976958099e-06,
+ "loss": 0.3449576377868652,
+ "num_input_tokens_seen": 349340688,
+ "step": 7990,
+ "train_runtime": 61755.9567,
+ "train_tokens_per_second": 5656.793
+ },
+ {
+ "epoch": 0.5383113385402639,
+ "grad_norm": 0.7351144206496397,
+ "learning_rate": 8.316891916147007e-06,
+ "loss": 0.40518803596496583,
+ "num_input_tokens_seen": 349557408,
+ "step": 7995,
+ "train_runtime": 61795.1163,
+ "train_tokens_per_second": 5656.716
+ },
+ {
+ "epoch": 0.5386479935362241,
+ "grad_norm": 0.8155814621488839,
+ "learning_rate": 8.314912927777671e-06,
+ "loss": 0.3914914608001709,
+ "num_input_tokens_seen": 349776936,
+ "step": 8000,
+ "train_runtime": 61829.5854,
+ "train_tokens_per_second": 5657.113
+ },
+ {
+ "epoch": 0.5389846485321842,
+ "grad_norm": 0.7399640627157749,
+ "learning_rate": 8.312933012403503e-06,
+ "loss": 0.3830781698226929,
+ "num_input_tokens_seen": 349981944,
+ "step": 8005,
+ "train_runtime": 61870.1471,
+ "train_tokens_per_second": 5656.718
+ },
+ {
+ "epoch": 0.5393213035281443,
+ "grad_norm": 0.6837705266722391,
+ "learning_rate": 8.310952170578184e-06,
+ "loss": 0.3917949676513672,
+ "num_input_tokens_seen": 350196560,
+ "step": 8010,
+ "train_runtime": 61910.9608,
+ "train_tokens_per_second": 5656.455
+ },
+ {
+ "epoch": 0.5396579585241045,
+ "grad_norm": 0.8108035687999815,
+ "learning_rate": 8.308970402855647e-06,
+ "loss": 0.35423126220703127,
+ "num_input_tokens_seen": 350407320,
+ "step": 8015,
+ "train_runtime": 61942.9362,
+ "train_tokens_per_second": 5656.938
+ },
+ {
+ "epoch": 0.5399946135200646,
+ "grad_norm": 0.7391585249859578,
+ "learning_rate": 8.306987709790087e-06,
+ "loss": 0.36304192543029784,
+ "num_input_tokens_seen": 350621920,
+ "step": 8020,
+ "train_runtime": 61974.5184,
+ "train_tokens_per_second": 5657.517
+ },
+ {
+ "epoch": 0.5403312685160248,
+ "grad_norm": 0.8297400381885977,
+ "learning_rate": 8.305004091935962e-06,
+ "loss": 0.3827538013458252,
+ "num_input_tokens_seen": 350840936,
+ "step": 8025,
+ "train_runtime": 62010.6476,
+ "train_tokens_per_second": 5657.753
+ },
+ {
+ "epoch": 0.5406679235119849,
+ "grad_norm": 0.8790917583658322,
+ "learning_rate": 8.303019549847979e-06,
+ "loss": 0.40552120208740233,
+ "num_input_tokens_seen": 351057864,
+ "step": 8030,
+ "train_runtime": 62051.8481,
+ "train_tokens_per_second": 5657.492
+ },
+ {
+ "epoch": 0.541004578507945,
+ "grad_norm": 0.7794541942972262,
+ "learning_rate": 8.301034084081114e-06,
+ "loss": 0.38002500534057615,
+ "num_input_tokens_seen": 351285720,
+ "step": 8035,
+ "train_runtime": 62090.6336,
+ "train_tokens_per_second": 5657.628
+ },
+ {
+ "epoch": 0.5413412335039052,
+ "grad_norm": 0.8524170296510677,
+ "learning_rate": 8.299047695190592e-06,
+ "loss": 0.408983850479126,
+ "num_input_tokens_seen": 351491576,
+ "step": 8040,
+ "train_runtime": 62127.0664,
+ "train_tokens_per_second": 5657.624
+ },
+ {
+ "epoch": 0.5416778884998653,
+ "grad_norm": 0.8835316549178089,
+ "learning_rate": 8.297060383731903e-06,
+ "loss": 0.38953261375427245,
+ "num_input_tokens_seen": 351706376,
+ "step": 8045,
+ "train_runtime": 62162.9599,
+ "train_tokens_per_second": 5657.813
+ },
+ {
+ "epoch": 0.5420145434958255,
+ "grad_norm": 0.8130215721187897,
+ "learning_rate": 8.29507215026079e-06,
+ "loss": 0.3989473819732666,
+ "num_input_tokens_seen": 351925296,
+ "step": 8050,
+ "train_runtime": 62194.3013,
+ "train_tokens_per_second": 5658.481
+ },
+ {
+ "epoch": 0.5423511984917856,
+ "grad_norm": 0.7698865553206388,
+ "learning_rate": 8.293082995333257e-06,
+ "loss": 0.3805055618286133,
+ "num_input_tokens_seen": 352152728,
+ "step": 8055,
+ "train_runtime": 62235.2959,
+ "train_tokens_per_second": 5658.409
+ },
+ {
+ "epoch": 0.5426878534877457,
+ "grad_norm": 0.7925089143300325,
+ "learning_rate": 8.291092919505568e-06,
+ "loss": 0.359727144241333,
+ "num_input_tokens_seen": 352368040,
+ "step": 8060,
+ "train_runtime": 62272.5239,
+ "train_tokens_per_second": 5658.483
+ },
+ {
+ "epoch": 0.5430245084837059,
+ "grad_norm": 0.7473619474168591,
+ "learning_rate": 8.289101923334237e-06,
+ "loss": 0.3927903175354004,
+ "num_input_tokens_seen": 352592456,
+ "step": 8065,
+ "train_runtime": 62314.1115,
+ "train_tokens_per_second": 5658.308
+ },
+ {
+ "epoch": 0.543361163479666,
+ "grad_norm": 0.8359023777365873,
+ "learning_rate": 8.28711000737604e-06,
+ "loss": 0.37551441192626955,
+ "num_input_tokens_seen": 352801680,
+ "step": 8070,
+ "train_runtime": 62352.2218,
+ "train_tokens_per_second": 5658.205
+ },
+ {
+ "epoch": 0.5436978184756261,
+ "grad_norm": 0.8077808908921092,
+ "learning_rate": 8.285117172188013e-06,
+ "loss": 0.3938951253890991,
+ "num_input_tokens_seen": 353025760,
+ "step": 8075,
+ "train_runtime": 62386.0037,
+ "train_tokens_per_second": 5658.733
+ },
+ {
+ "epoch": 0.5440344734715863,
+ "grad_norm": 0.8228182913825491,
+ "learning_rate": 8.283123418327443e-06,
+ "loss": 0.4004195213317871,
+ "num_input_tokens_seen": 353244352,
+ "step": 8080,
+ "train_runtime": 62419.1441,
+ "train_tokens_per_second": 5659.231
+ },
+ {
+ "epoch": 0.5443711284675464,
+ "grad_norm": 0.693682105218099,
+ "learning_rate": 8.281128746351878e-06,
+ "loss": 0.3595161437988281,
+ "num_input_tokens_seen": 353475168,
+ "step": 8085,
+ "train_runtime": 62453.3189,
+ "train_tokens_per_second": 5659.83
+ },
+ {
+ "epoch": 0.5447077834635066,
+ "grad_norm": 0.7187659246186631,
+ "learning_rate": 8.279133156819123e-06,
+ "loss": 0.40300402641296384,
+ "num_input_tokens_seen": 353691088,
+ "step": 8090,
+ "train_runtime": 62496.7209,
+ "train_tokens_per_second": 5659.354
+ },
+ {
+ "epoch": 0.5450444384594667,
+ "grad_norm": 0.725512834705303,
+ "learning_rate": 8.277136650287237e-06,
+ "loss": 0.3694662570953369,
+ "num_input_tokens_seen": 353917984,
+ "step": 8095,
+ "train_runtime": 62536.3526,
+ "train_tokens_per_second": 5659.396
+ },
+ {
+ "epoch": 0.5453810934554268,
+ "grad_norm": 0.8301428582075541,
+ "learning_rate": 8.275139227314537e-06,
+ "loss": 0.3757363557815552,
+ "num_input_tokens_seen": 354125952,
+ "step": 8100,
+ "train_runtime": 62575.4142,
+ "train_tokens_per_second": 5659.187
+ },
+ {
+ "epoch": 0.545717748451387,
+ "grad_norm": 0.7999915594881919,
+ "learning_rate": 8.273140888459594e-06,
+ "loss": 0.4242837905883789,
+ "num_input_tokens_seen": 354358280,
+ "step": 8105,
+ "train_runtime": 62606.4679,
+ "train_tokens_per_second": 5660.091
+ },
+ {
+ "epoch": 0.5460544034473471,
+ "grad_norm": 0.9175431453565677,
+ "learning_rate": 8.27114163428124e-06,
+ "loss": 0.3934792518615723,
+ "num_input_tokens_seen": 354575664,
+ "step": 8110,
+ "train_runtime": 62642.1464,
+ "train_tokens_per_second": 5660.337
+ },
+ {
+ "epoch": 0.5463910584433073,
+ "grad_norm": 0.7408300214764466,
+ "learning_rate": 8.26914146533856e-06,
+ "loss": 0.38090806007385253,
+ "num_input_tokens_seen": 354808384,
+ "step": 8115,
+ "train_runtime": 62681.8186,
+ "train_tokens_per_second": 5660.467
+ },
+ {
+ "epoch": 0.5467277134392674,
+ "grad_norm": 0.7033405910036602,
+ "learning_rate": 8.267140382190892e-06,
+ "loss": 0.3828258037567139,
+ "num_input_tokens_seen": 355026976,
+ "step": 8120,
+ "train_runtime": 62720.5013,
+ "train_tokens_per_second": 5660.461
+ },
+ {
+ "epoch": 0.5470643684352275,
+ "grad_norm": 0.6635248689946966,
+ "learning_rate": 8.265138385397835e-06,
+ "loss": 0.3777566432952881,
+ "num_input_tokens_seen": 355244064,
+ "step": 8125,
+ "train_runtime": 62758.0924,
+ "train_tokens_per_second": 5660.53
+ },
+ {
+ "epoch": 0.5474010234311877,
+ "grad_norm": 0.6184317548675163,
+ "learning_rate": 8.263135475519241e-06,
+ "loss": 0.34886605739593507,
+ "num_input_tokens_seen": 355483272,
+ "step": 8130,
+ "train_runtime": 62794.3565,
+ "train_tokens_per_second": 5661.07
+ },
+ {
+ "epoch": 0.5477376784271478,
+ "grad_norm": 0.7287895081916465,
+ "learning_rate": 8.261131653115216e-06,
+ "loss": 0.42696261405944824,
+ "num_input_tokens_seen": 355711240,
+ "step": 8135,
+ "train_runtime": 62834.333,
+ "train_tokens_per_second": 5661.097
+ },
+ {
+ "epoch": 0.548074333423108,
+ "grad_norm": 0.7794435545394348,
+ "learning_rate": 8.259126918746122e-06,
+ "loss": 0.36356501579284667,
+ "num_input_tokens_seen": 355930576,
+ "step": 8140,
+ "train_runtime": 62868.698,
+ "train_tokens_per_second": 5661.491
+ },
+ {
+ "epoch": 0.5484109884190681,
+ "grad_norm": 0.6736849599162793,
+ "learning_rate": 8.25712127297258e-06,
+ "loss": 0.3778836250305176,
+ "num_input_tokens_seen": 356150816,
+ "step": 8145,
+ "train_runtime": 62904.5449,
+ "train_tokens_per_second": 5661.766
+ },
+ {
+ "epoch": 0.5487476434150282,
+ "grad_norm": 0.7082862793695406,
+ "learning_rate": 8.255114716355459e-06,
+ "loss": 0.4094264507293701,
+ "num_input_tokens_seen": 356382088,
+ "step": 8150,
+ "train_runtime": 62942.0397,
+ "train_tokens_per_second": 5662.068
+ },
+ {
+ "epoch": 0.5490842984109884,
+ "grad_norm": 0.6051005207909932,
+ "learning_rate": 8.253107249455887e-06,
+ "loss": 0.3917287826538086,
+ "num_input_tokens_seen": 356602352,
+ "step": 8155,
+ "train_runtime": 62975.3252,
+ "train_tokens_per_second": 5662.573
+ },
+ {
+ "epoch": 0.5494209534069485,
+ "grad_norm": 0.7713641151309204,
+ "learning_rate": 8.251098872835247e-06,
+ "loss": 0.39372997283935546,
+ "num_input_tokens_seen": 356825280,
+ "step": 8160,
+ "train_runtime": 63017.9451,
+ "train_tokens_per_second": 5662.28
+ },
+ {
+ "epoch": 0.5497576084029087,
+ "grad_norm": 0.6787537110984104,
+ "learning_rate": 8.249089587055175e-06,
+ "loss": 0.3526461601257324,
+ "num_input_tokens_seen": 357037664,
+ "step": 8165,
+ "train_runtime": 63059.5973,
+ "train_tokens_per_second": 5661.908
+ },
+ {
+ "epoch": 0.5500942633988688,
+ "grad_norm": 0.6748882087262432,
+ "learning_rate": 8.247079392677558e-06,
+ "loss": 0.3740200996398926,
+ "num_input_tokens_seen": 357257424,
+ "step": 8170,
+ "train_runtime": 63101.2168,
+ "train_tokens_per_second": 5661.657
+ },
+ {
+ "epoch": 0.5504309183948289,
+ "grad_norm": 0.7274044025423265,
+ "learning_rate": 8.245068290264544e-06,
+ "loss": 0.3697060108184814,
+ "num_input_tokens_seen": 357480144,
+ "step": 8175,
+ "train_runtime": 63140.8867,
+ "train_tokens_per_second": 5661.627
+ },
+ {
+ "epoch": 0.5507675733907891,
+ "grad_norm": 0.7084601443387462,
+ "learning_rate": 8.243056280378531e-06,
+ "loss": 0.37436251640319823,
+ "num_input_tokens_seen": 357699472,
+ "step": 8180,
+ "train_runtime": 63181.8065,
+ "train_tokens_per_second": 5661.432
+ },
+ {
+ "epoch": 0.5511042283867492,
+ "grad_norm": 0.7481357333059493,
+ "learning_rate": 8.241043363582171e-06,
+ "loss": 0.36437788009643557,
+ "num_input_tokens_seen": 357926808,
+ "step": 8185,
+ "train_runtime": 63218.2393,
+ "train_tokens_per_second": 5661.765
+ },
+ {
+ "epoch": 0.5514408833827094,
+ "grad_norm": 1.0989196799559033,
+ "learning_rate": 8.239029540438369e-06,
+ "loss": 0.3505790948867798,
+ "num_input_tokens_seen": 358145064,
+ "step": 8190,
+ "train_runtime": 63257.3639,
+ "train_tokens_per_second": 5661.713
+ },
+ {
+ "epoch": 0.5517775383786695,
+ "grad_norm": 0.7305407817056846,
+ "learning_rate": 8.237014811510284e-06,
+ "loss": 0.3656752109527588,
+ "num_input_tokens_seen": 358369824,
+ "step": 8195,
+ "train_runtime": 63304.7711,
+ "train_tokens_per_second": 5661.024
+ },
+ {
+ "epoch": 0.5521141933746296,
+ "grad_norm": 0.6472079948654392,
+ "learning_rate": 8.234999177361329e-06,
+ "loss": 0.3714550018310547,
+ "num_input_tokens_seen": 358596784,
+ "step": 8200,
+ "train_runtime": 63340.7687,
+ "train_tokens_per_second": 5661.39
+ },
+ {
+ "epoch": 0.5524508483705898,
+ "grad_norm": 0.7398046092203671,
+ "learning_rate": 8.232982638555169e-06,
+ "loss": 0.40174274444580077,
+ "num_input_tokens_seen": 358820312,
+ "step": 8205,
+ "train_runtime": 63372.9424,
+ "train_tokens_per_second": 5662.043
+ },
+ {
+ "epoch": 0.5527875033665499,
+ "grad_norm": 0.7062238114257713,
+ "learning_rate": 8.230965195655724e-06,
+ "loss": 0.39043726921081545,
+ "num_input_tokens_seen": 359055496,
+ "step": 8210,
+ "train_runtime": 63412.4919,
+ "train_tokens_per_second": 5662.22
+ },
+ {
+ "epoch": 0.55312415836251,
+ "grad_norm": 0.8164987537109517,
+ "learning_rate": 8.228946849227167e-06,
+ "loss": 0.3990142345428467,
+ "num_input_tokens_seen": 359287376,
+ "step": 8215,
+ "train_runtime": 63449.3085,
+ "train_tokens_per_second": 5662.589
+ },
+ {
+ "epoch": 0.5534608133584702,
+ "grad_norm": 0.6938755136640083,
+ "learning_rate": 8.226927599833916e-06,
+ "loss": 0.386507511138916,
+ "num_input_tokens_seen": 359514576,
+ "step": 8220,
+ "train_runtime": 63481.4052,
+ "train_tokens_per_second": 5663.305
+ },
+ {
+ "epoch": 0.5537974683544303,
+ "grad_norm": 0.6819170611796698,
+ "learning_rate": 8.224907448040655e-06,
+ "loss": 0.37993340492248534,
+ "num_input_tokens_seen": 359745160,
+ "step": 8225,
+ "train_runtime": 63513.0016,
+ "train_tokens_per_second": 5664.118
+ },
+ {
+ "epoch": 0.5541341233503905,
+ "grad_norm": 0.7464665475450891,
+ "learning_rate": 8.222886394412309e-06,
+ "loss": 0.4058523178100586,
+ "num_input_tokens_seen": 359965008,
+ "step": 8230,
+ "train_runtime": 63553.6566,
+ "train_tokens_per_second": 5663.954
+ },
+ {
+ "epoch": 0.5544707783463506,
+ "grad_norm": 0.6722302273522195,
+ "learning_rate": 8.220864439514057e-06,
+ "loss": 0.384306526184082,
+ "num_input_tokens_seen": 360194040,
+ "step": 8235,
+ "train_runtime": 63592.1988,
+ "train_tokens_per_second": 5664.123
+ },
+ {
+ "epoch": 0.5548074333423108,
+ "grad_norm": 0.6641141478245969,
+ "learning_rate": 8.218841583911338e-06,
+ "loss": 0.3680997848510742,
+ "num_input_tokens_seen": 360409248,
+ "step": 8240,
+ "train_runtime": 63634.667,
+ "train_tokens_per_second": 5663.725
+ },
+ {
+ "epoch": 0.5551440883382709,
+ "grad_norm": 0.7528778151335748,
+ "learning_rate": 8.216817828169834e-06,
+ "loss": 0.413088321685791,
+ "num_input_tokens_seen": 360624480,
+ "step": 8245,
+ "train_runtime": 63670.5285,
+ "train_tokens_per_second": 5663.915
+ },
+ {
+ "epoch": 0.555480743334231,
+ "grad_norm": 0.8364309542706772,
+ "learning_rate": 8.214793172855481e-06,
+ "loss": 0.36465370655059814,
+ "num_input_tokens_seen": 360837048,
+ "step": 8250,
+ "train_runtime": 63711.1159,
+ "train_tokens_per_second": 5663.644
+ },
+ {
+ "epoch": 0.5558173983301912,
+ "grad_norm": 0.6793083765205596,
+ "learning_rate": 8.21276761853447e-06,
+ "loss": 0.36249728202819825,
+ "num_input_tokens_seen": 361060272,
+ "step": 8255,
+ "train_runtime": 63754.6776,
+ "train_tokens_per_second": 5663.275
+ },
+ {
+ "epoch": 0.5561540533261513,
+ "grad_norm": 0.8280046723618791,
+ "learning_rate": 8.210741165773238e-06,
+ "loss": 0.39040985107421877,
+ "num_input_tokens_seen": 361294192,
+ "step": 8260,
+ "train_runtime": 63785.4799,
+ "train_tokens_per_second": 5664.207
+ },
+ {
+ "epoch": 0.5564907083221114,
+ "grad_norm": 0.7577410939409384,
+ "learning_rate": 8.208713815138476e-06,
+ "loss": 0.36567642688751223,
+ "num_input_tokens_seen": 361524240,
+ "step": 8265,
+ "train_runtime": 63826.3741,
+ "train_tokens_per_second": 5664.183
+ },
+ {
+ "epoch": 0.5568273633180716,
+ "grad_norm": 0.6700074247940968,
+ "learning_rate": 8.20668556719713e-06,
+ "loss": 0.35714569091796877,
+ "num_input_tokens_seen": 361732392,
+ "step": 8270,
+ "train_runtime": 63860.1654,
+ "train_tokens_per_second": 5664.445
+ },
+ {
+ "epoch": 0.5571640183140317,
+ "grad_norm": 0.7602580503776288,
+ "learning_rate": 8.204656422516388e-06,
+ "loss": 0.3372737646102905,
+ "num_input_tokens_seen": 361946008,
+ "step": 8275,
+ "train_runtime": 63901.2504,
+ "train_tokens_per_second": 5664.146
+ },
+ {
+ "epoch": 0.5575006733099919,
+ "grad_norm": 0.8510184000147942,
+ "learning_rate": 8.2026263816637e-06,
+ "loss": 0.38366384506225587,
+ "num_input_tokens_seen": 362152248,
+ "step": 8280,
+ "train_runtime": 63938.3717,
+ "train_tokens_per_second": 5664.083
+ },
+ {
+ "epoch": 0.557837328305952,
+ "grad_norm": 0.7613111925807169,
+ "learning_rate": 8.200595445206753e-06,
+ "loss": 0.37342875003814696,
+ "num_input_tokens_seen": 362371096,
+ "step": 8285,
+ "train_runtime": 63979.792,
+ "train_tokens_per_second": 5663.837
+ },
+ {
+ "epoch": 0.5581739833019121,
+ "grad_norm": 1.256734854809266,
+ "learning_rate": 8.198563613713497e-06,
+ "loss": 0.37401297092437746,
+ "num_input_tokens_seen": 362579576,
+ "step": 8290,
+ "train_runtime": 64020.2056,
+ "train_tokens_per_second": 5663.518
+ },
+ {
+ "epoch": 0.5585106382978723,
+ "grad_norm": 0.7385210193529426,
+ "learning_rate": 8.196530887752127e-06,
+ "loss": 0.3654287338256836,
+ "num_input_tokens_seen": 362803720,
+ "step": 8295,
+ "train_runtime": 64054.0322,
+ "train_tokens_per_second": 5664.026
+ },
+ {
+ "epoch": 0.5588472932938324,
+ "grad_norm": 0.8102882370091832,
+ "learning_rate": 8.194497267891086e-06,
+ "loss": 0.37125024795532224,
+ "num_input_tokens_seen": 363027432,
+ "step": 8300,
+ "train_runtime": 64092.3585,
+ "train_tokens_per_second": 5664.13
+ },
+ {
+ "epoch": 0.5591839482897926,
+ "grad_norm": 0.7432402409233543,
+ "learning_rate": 8.19246275469907e-06,
+ "loss": 0.3796513795852661,
+ "num_input_tokens_seen": 363254432,
+ "step": 8305,
+ "train_runtime": 64136.3098,
+ "train_tokens_per_second": 5663.788
+ },
+ {
+ "epoch": 0.5595206032857528,
+ "grad_norm": 0.6849455308836975,
+ "learning_rate": 8.190427348745027e-06,
+ "loss": 0.36992511749267576,
+ "num_input_tokens_seen": 363479976,
+ "step": 8310,
+ "train_runtime": 64171.4185,
+ "train_tokens_per_second": 5664.204
+ },
+ {
+ "epoch": 0.559857258281713,
+ "grad_norm": 0.6998379838638995,
+ "learning_rate": 8.188391050598146e-06,
+ "loss": 0.3632178544998169,
+ "num_input_tokens_seen": 363712504,
+ "step": 8315,
+ "train_runtime": 64216.5044,
+ "train_tokens_per_second": 5663.848
+ },
+ {
+ "epoch": 0.5601939132776731,
+ "grad_norm": 0.7737605990707936,
+ "learning_rate": 8.186353860827878e-06,
+ "loss": 0.37043955326080324,
+ "num_input_tokens_seen": 363919784,
+ "step": 8320,
+ "train_runtime": 64256.872,
+ "train_tokens_per_second": 5663.515
+ },
+ {
+ "epoch": 0.5605305682736332,
+ "grad_norm": 0.6957700417795362,
+ "learning_rate": 8.184315780003911e-06,
+ "loss": 0.3709959745407104,
+ "num_input_tokens_seen": 364140736,
+ "step": 8325,
+ "train_runtime": 64302.6454,
+ "train_tokens_per_second": 5662.92
+ },
+ {
+ "epoch": 0.5608672232695934,
+ "grad_norm": 0.64141155335251,
+ "learning_rate": 8.182276808696193e-06,
+ "loss": 0.40601282119750975,
+ "num_input_tokens_seen": 364348280,
+ "step": 8330,
+ "train_runtime": 64347.1042,
+ "train_tokens_per_second": 5662.233
+ },
+ {
+ "epoch": 0.5612038782655535,
+ "grad_norm": 0.6158998839242338,
+ "learning_rate": 8.18023694747491e-06,
+ "loss": 0.3827899694442749,
+ "num_input_tokens_seen": 364574968,
+ "step": 8335,
+ "train_runtime": 64381.872,
+ "train_tokens_per_second": 5662.696
+ },
+ {
+ "epoch": 0.5615405332615137,
+ "grad_norm": 0.7883261214385181,
+ "learning_rate": 8.178196196910508e-06,
+ "loss": 0.3656409740447998,
+ "num_input_tokens_seen": 364803736,
+ "step": 8340,
+ "train_runtime": 64421.7397,
+ "train_tokens_per_second": 5662.743
+ },
+ {
+ "epoch": 0.5618771882574738,
+ "grad_norm": 0.7366395620437122,
+ "learning_rate": 8.176154557573672e-06,
+ "loss": 0.4000543594360352,
+ "num_input_tokens_seen": 365037240,
+ "step": 8345,
+ "train_runtime": 64459.3416,
+ "train_tokens_per_second": 5663.062
+ },
+ {
+ "epoch": 0.5622138432534339,
+ "grad_norm": 0.6985009938577207,
+ "learning_rate": 8.174112030035344e-06,
+ "loss": 0.3629467010498047,
+ "num_input_tokens_seen": 365268200,
+ "step": 8350,
+ "train_runtime": 64493.078,
+ "train_tokens_per_second": 5663.681
+ },
+ {
+ "epoch": 0.5625504982493941,
+ "grad_norm": 0.7274783538197038,
+ "learning_rate": 8.172068614866707e-06,
+ "loss": 0.36340205669403075,
+ "num_input_tokens_seen": 365477184,
+ "step": 8355,
+ "train_runtime": 64531.8006,
+ "train_tokens_per_second": 5663.521
+ },
+ {
+ "epoch": 0.5628871532453542,
+ "grad_norm": 0.7368665647506253,
+ "learning_rate": 8.170024312639197e-06,
+ "loss": 0.39545679092407227,
+ "num_input_tokens_seen": 365707680,
+ "step": 8360,
+ "train_runtime": 64570.9604,
+ "train_tokens_per_second": 5663.656
+ },
+ {
+ "epoch": 0.5632238082413143,
+ "grad_norm": 0.7340645583889084,
+ "learning_rate": 8.167979123924496e-06,
+ "loss": 0.3859426498413086,
+ "num_input_tokens_seen": 365930048,
+ "step": 8365,
+ "train_runtime": 64604.6699,
+ "train_tokens_per_second": 5664.142
+ },
+ {
+ "epoch": 0.5635604632372745,
+ "grad_norm": 0.7462104536621238,
+ "learning_rate": 8.165933049294537e-06,
+ "loss": 0.4087996482849121,
+ "num_input_tokens_seen": 366147352,
+ "step": 8370,
+ "train_runtime": 64646.3443,
+ "train_tokens_per_second": 5663.852
+ },
+ {
+ "epoch": 0.5638971182332346,
+ "grad_norm": 0.7582565637830365,
+ "learning_rate": 8.163886089321493e-06,
+ "loss": 0.37345223426818847,
+ "num_input_tokens_seen": 366377160,
+ "step": 8375,
+ "train_runtime": 64688.6158,
+ "train_tokens_per_second": 5663.704
+ },
+ {
+ "epoch": 0.5642337732291948,
+ "grad_norm": 0.6581884199846296,
+ "learning_rate": 8.161838244577795e-06,
+ "loss": 0.3573953151702881,
+ "num_input_tokens_seen": 366596264,
+ "step": 8380,
+ "train_runtime": 64726.1749,
+ "train_tokens_per_second": 5663.802
+ },
+ {
+ "epoch": 0.5645704282251549,
+ "grad_norm": 0.7540331450420041,
+ "learning_rate": 8.159789515636114e-06,
+ "loss": 0.3621669769287109,
+ "num_input_tokens_seen": 366814672,
+ "step": 8385,
+ "train_runtime": 64768.6142,
+ "train_tokens_per_second": 5663.463
+ },
+ {
+ "epoch": 0.564907083221115,
+ "grad_norm": 0.7846236965599341,
+ "learning_rate": 8.15773990306937e-06,
+ "loss": 0.39387807846069334,
+ "num_input_tokens_seen": 367010840,
+ "step": 8390,
+ "train_runtime": 64803.6605,
+ "train_tokens_per_second": 5663.428
+ },
+ {
+ "epoch": 0.5652437382170752,
+ "grad_norm": 0.8105957793723372,
+ "learning_rate": 8.155689407450732e-06,
+ "loss": 0.37447366714477537,
+ "num_input_tokens_seen": 367235664,
+ "step": 8395,
+ "train_runtime": 64840.3814,
+ "train_tokens_per_second": 5663.688
+ },
+ {
+ "epoch": 0.5655803932130353,
+ "grad_norm": 0.7593251086886306,
+ "learning_rate": 8.153638029353615e-06,
+ "loss": 0.3721097707748413,
+ "num_input_tokens_seen": 367439472,
+ "step": 8400,
+ "train_runtime": 64880.3333,
+ "train_tokens_per_second": 5663.341
+ },
+ {
+ "epoch": 0.5659170482089955,
+ "grad_norm": 0.7080065499934168,
+ "learning_rate": 8.151585769351678e-06,
+ "loss": 0.416213321685791,
+ "num_input_tokens_seen": 367665000,
+ "step": 8405,
+ "train_runtime": 64915.8566,
+ "train_tokens_per_second": 5663.716
+ },
+ {
+ "epoch": 0.5662537032049556,
+ "grad_norm": 0.8290526825274906,
+ "learning_rate": 8.149532628018833e-06,
+ "loss": 0.3906120538711548,
+ "num_input_tokens_seen": 367873544,
+ "step": 8410,
+ "train_runtime": 64954.7801,
+ "train_tokens_per_second": 5663.533
+ },
+ {
+ "epoch": 0.5665903582009157,
+ "grad_norm": 0.7887226131302195,
+ "learning_rate": 8.147478605929228e-06,
+ "loss": 0.36734464168548586,
+ "num_input_tokens_seen": 368106424,
+ "step": 8415,
+ "train_runtime": 64988.7811,
+ "train_tokens_per_second": 5664.153
+ },
+ {
+ "epoch": 0.5669270131968759,
+ "grad_norm": 0.7105615187211392,
+ "learning_rate": 8.145423703657269e-06,
+ "loss": 0.3696259021759033,
+ "num_input_tokens_seen": 368329816,
+ "step": 8420,
+ "train_runtime": 65035.2437,
+ "train_tokens_per_second": 5663.542
+ },
+ {
+ "epoch": 0.567263668192836,
+ "grad_norm": 0.7106675383557719,
+ "learning_rate": 8.143367921777601e-06,
+ "loss": 0.3901944637298584,
+ "num_input_tokens_seen": 368556080,
+ "step": 8425,
+ "train_runtime": 65073.1284,
+ "train_tokens_per_second": 5663.722
+ },
+ {
+ "epoch": 0.5676003231887962,
+ "grad_norm": 0.7158037922920018,
+ "learning_rate": 8.141311260865117e-06,
+ "loss": 0.42763843536376955,
+ "num_input_tokens_seen": 368786216,
+ "step": 8430,
+ "train_runtime": 65107.0411,
+ "train_tokens_per_second": 5664.306
+ },
+ {
+ "epoch": 0.5679369781847563,
+ "grad_norm": 0.7147505710629092,
+ "learning_rate": 8.139253721494956e-06,
+ "loss": 0.4124480724334717,
+ "num_input_tokens_seen": 368998320,
+ "step": 8435,
+ "train_runtime": 65143.6774,
+ "train_tokens_per_second": 5664.377
+ },
+ {
+ "epoch": 0.5682736331807164,
+ "grad_norm": 0.690454346302624,
+ "learning_rate": 8.137195304242502e-06,
+ "loss": 0.3694611549377441,
+ "num_input_tokens_seen": 369221808,
+ "step": 8440,
+ "train_runtime": 65180.1258,
+ "train_tokens_per_second": 5664.638
+ },
+ {
+ "epoch": 0.5686102881766766,
+ "grad_norm": 0.6647768706585024,
+ "learning_rate": 8.135136009683387e-06,
+ "loss": 0.3659370899200439,
+ "num_input_tokens_seen": 369439952,
+ "step": 8445,
+ "train_runtime": 65214.1207,
+ "train_tokens_per_second": 5665.03
+ },
+ {
+ "epoch": 0.5689469431726367,
+ "grad_norm": 0.7123577020124194,
+ "learning_rate": 8.13307583839348e-06,
+ "loss": 0.37821600437164304,
+ "num_input_tokens_seen": 369665384,
+ "step": 8450,
+ "train_runtime": 65253.9634,
+ "train_tokens_per_second": 5665.026
+ },
+ {
+ "epoch": 0.5692835981685969,
+ "grad_norm": 0.7080847091486382,
+ "learning_rate": 8.131014790948908e-06,
+ "loss": 0.3932258367538452,
+ "num_input_tokens_seen": 369885352,
+ "step": 8455,
+ "train_runtime": 65292.5756,
+ "train_tokens_per_second": 5665.045
+ },
+ {
+ "epoch": 0.569620253164557,
+ "grad_norm": 0.8175401785926597,
+ "learning_rate": 8.128952867926035e-06,
+ "loss": 0.42728986740112307,
+ "num_input_tokens_seen": 370102168,
+ "step": 8460,
+ "train_runtime": 65326.5467,
+ "train_tokens_per_second": 5665.418
+ },
+ {
+ "epoch": 0.5699569081605171,
+ "grad_norm": 0.7636022480938197,
+ "learning_rate": 8.126890069901468e-06,
+ "loss": 0.3966690540313721,
+ "num_input_tokens_seen": 370323328,
+ "step": 8465,
+ "train_runtime": 65365.5452,
+ "train_tokens_per_second": 5665.421
+ },
+ {
+ "epoch": 0.5702935631564773,
+ "grad_norm": 0.8306888968108936,
+ "learning_rate": 8.124826397452064e-06,
+ "loss": 0.3729779005050659,
+ "num_input_tokens_seen": 370551792,
+ "step": 8470,
+ "train_runtime": 65403.3941,
+ "train_tokens_per_second": 5665.636
+ },
+ {
+ "epoch": 0.5706302181524374,
+ "grad_norm": 0.6975418197600278,
+ "learning_rate": 8.122761851154923e-06,
+ "loss": 0.37224202156066893,
+ "num_input_tokens_seen": 370769584,
+ "step": 8475,
+ "train_runtime": 65437.1957,
+ "train_tokens_per_second": 5666.037
+ },
+ {
+ "epoch": 0.5709668731483976,
+ "grad_norm": 0.7691847068239698,
+ "learning_rate": 8.120696431587387e-06,
+ "loss": 0.3793643474578857,
+ "num_input_tokens_seen": 370986856,
+ "step": 8480,
+ "train_runtime": 65477.4237,
+ "train_tokens_per_second": 5665.874
+ },
+ {
+ "epoch": 0.5713035281443577,
+ "grad_norm": 0.653136501814857,
+ "learning_rate": 8.118630139327046e-06,
+ "loss": 0.36144418716430665,
+ "num_input_tokens_seen": 371203736,
+ "step": 8485,
+ "train_runtime": 65520.6447,
+ "train_tokens_per_second": 5665.447
+ },
+ {
+ "epoch": 0.5716401831403178,
+ "grad_norm": 0.6688068417437787,
+ "learning_rate": 8.116562974951731e-06,
+ "loss": 0.3969851016998291,
+ "num_input_tokens_seen": 371433168,
+ "step": 8490,
+ "train_runtime": 65562.0835,
+ "train_tokens_per_second": 5665.366
+ },
+ {
+ "epoch": 0.571976838136278,
+ "grad_norm": 0.6693856135821081,
+ "learning_rate": 8.11449493903952e-06,
+ "loss": 0.37904407978057864,
+ "num_input_tokens_seen": 371661928,
+ "step": 8495,
+ "train_runtime": 65600.9804,
+ "train_tokens_per_second": 5665.494
+ },
+ {
+ "epoch": 0.5723134931322381,
+ "grad_norm": 0.6989787293355911,
+ "learning_rate": 8.112426032168727e-06,
+ "loss": 0.39436898231506345,
+ "num_input_tokens_seen": 371877904,
+ "step": 8500,
+ "train_runtime": 65639.9628,
+ "train_tokens_per_second": 5665.419
+ },
+ {
+ "epoch": 0.5726501481281983,
+ "grad_norm": 0.6921966447452655,
+ "learning_rate": 8.11035625491792e-06,
+ "loss": 0.35653374195098875,
+ "num_input_tokens_seen": 372083344,
+ "step": 8505,
+ "train_runtime": 65673.4132,
+ "train_tokens_per_second": 5665.662
+ },
+ {
+ "epoch": 0.5729868031241584,
+ "grad_norm": 0.7863535213377899,
+ "learning_rate": 8.108285607865907e-06,
+ "loss": 0.3748607158660889,
+ "num_input_tokens_seen": 372296008,
+ "step": 8510,
+ "train_runtime": 65709.8413,
+ "train_tokens_per_second": 5665.757
+ },
+ {
+ "epoch": 0.5733234581201185,
+ "grad_norm": 0.7103469377120155,
+ "learning_rate": 8.106214091591731e-06,
+ "loss": 0.36768136024475095,
+ "num_input_tokens_seen": 372517376,
+ "step": 8515,
+ "train_runtime": 65747.0301,
+ "train_tokens_per_second": 5665.919
+ },
+ {
+ "epoch": 0.5736601131160787,
+ "grad_norm": 0.8043097859632866,
+ "learning_rate": 8.104141706674693e-06,
+ "loss": 0.3505872249603271,
+ "num_input_tokens_seen": 372744072,
+ "step": 8520,
+ "train_runtime": 65786.4844,
+ "train_tokens_per_second": 5665.967
+ },
+ {
+ "epoch": 0.5739967681120388,
+ "grad_norm": 0.7899542715564848,
+ "learning_rate": 8.102068453694324e-06,
+ "loss": 0.39815869331359866,
+ "num_input_tokens_seen": 372965496,
+ "step": 8525,
+ "train_runtime": 65827.1156,
+ "train_tokens_per_second": 5665.834
+ },
+ {
+ "epoch": 0.574333423107999,
+ "grad_norm": 0.7007321773126729,
+ "learning_rate": 8.099994333230404e-06,
+ "loss": 0.39165263175964354,
+ "num_input_tokens_seen": 373176000,
+ "step": 8530,
+ "train_runtime": 65869.8446,
+ "train_tokens_per_second": 5665.354
+ },
+ {
+ "epoch": 0.5746700781039591,
+ "grad_norm": 0.8059453918533225,
+ "learning_rate": 8.097919345862952e-06,
+ "loss": 0.387595534324646,
+ "num_input_tokens_seen": 373375616,
+ "step": 8535,
+ "train_runtime": 65901.6062,
+ "train_tokens_per_second": 5665.653
+ },
+ {
+ "epoch": 0.5750067330999192,
+ "grad_norm": 0.7748964359890478,
+ "learning_rate": 8.095843492172235e-06,
+ "loss": 0.3876967906951904,
+ "num_input_tokens_seen": 373591328,
+ "step": 8540,
+ "train_runtime": 65941.7008,
+ "train_tokens_per_second": 5665.479
+ },
+ {
+ "epoch": 0.5753433880958794,
+ "grad_norm": 0.773534060871464,
+ "learning_rate": 8.093766772738758e-06,
+ "loss": 0.3301082611083984,
+ "num_input_tokens_seen": 373805712,
+ "step": 8545,
+ "train_runtime": 65984.3171,
+ "train_tokens_per_second": 5665.069
+ },
+ {
+ "epoch": 0.5756800430918395,
+ "grad_norm": 0.7746351168910522,
+ "learning_rate": 8.091689188143267e-06,
+ "loss": 0.3910914659500122,
+ "num_input_tokens_seen": 374017824,
+ "step": 8550,
+ "train_runtime": 66030.3105,
+ "train_tokens_per_second": 5664.335
+ },
+ {
+ "epoch": 0.5760166980877997,
+ "grad_norm": 0.7771670247365485,
+ "learning_rate": 8.089610738966754e-06,
+ "loss": 0.3839992046356201,
+ "num_input_tokens_seen": 374221088,
+ "step": 8555,
+ "train_runtime": 66067.2485,
+ "train_tokens_per_second": 5664.245
+ },
+ {
+ "epoch": 0.5763533530837598,
+ "grad_norm": 0.8314883209793272,
+ "learning_rate": 8.08753142579045e-06,
+ "loss": 0.41312179565429685,
+ "num_input_tokens_seen": 374438784,
+ "step": 8560,
+ "train_runtime": 66103.0863,
+ "train_tokens_per_second": 5664.468
+ },
+ {
+ "epoch": 0.5766900080797199,
+ "grad_norm": 0.7070329659969831,
+ "learning_rate": 8.085451249195828e-06,
+ "loss": 0.369646692276001,
+ "num_input_tokens_seen": 374674400,
+ "step": 8565,
+ "train_runtime": 66144.5198,
+ "train_tokens_per_second": 5664.481
+ },
+ {
+ "epoch": 0.5770266630756801,
+ "grad_norm": 0.728101829102978,
+ "learning_rate": 8.083370209764603e-06,
+ "loss": 0.38304858207702636,
+ "num_input_tokens_seen": 374895048,
+ "step": 8570,
+ "train_runtime": 66187.4516,
+ "train_tokens_per_second": 5664.141
+ },
+ {
+ "epoch": 0.5773633180716402,
+ "grad_norm": 0.7613212980235913,
+ "learning_rate": 8.081288308078731e-06,
+ "loss": 0.39395363330841066,
+ "num_input_tokens_seen": 375130432,
+ "step": 8575,
+ "train_runtime": 66220.5986,
+ "train_tokens_per_second": 5664.86
+ },
+ {
+ "epoch": 0.5776999730676003,
+ "grad_norm": 0.6846318822730251,
+ "learning_rate": 8.07920554472041e-06,
+ "loss": 0.35710582733154295,
+ "num_input_tokens_seen": 375356416,
+ "step": 8580,
+ "train_runtime": 66266.8733,
+ "train_tokens_per_second": 5664.315
+ },
+ {
+ "epoch": 0.5780366280635605,
+ "grad_norm": 0.791144672174043,
+ "learning_rate": 8.077121920272075e-06,
+ "loss": 0.3807799816131592,
+ "num_input_tokens_seen": 375586768,
+ "step": 8585,
+ "train_runtime": 66303.4383,
+ "train_tokens_per_second": 5664.665
+ },
+ {
+ "epoch": 0.5783732830595206,
+ "grad_norm": 0.6749521518269885,
+ "learning_rate": 8.07503743531641e-06,
+ "loss": 0.3493949890136719,
+ "num_input_tokens_seen": 375796816,
+ "step": 8590,
+ "train_runtime": 66347.6746,
+ "train_tokens_per_second": 5664.054
+ },
+ {
+ "epoch": 0.5787099380554808,
+ "grad_norm": 0.7233256642537537,
+ "learning_rate": 8.072952090436328e-06,
+ "loss": 0.34188032150268555,
+ "num_input_tokens_seen": 376025856,
+ "step": 8595,
+ "train_runtime": 66379.0077,
+ "train_tokens_per_second": 5664.831
+ },
+ {
+ "epoch": 0.5790465930514409,
+ "grad_norm": 0.8183957573318358,
+ "learning_rate": 8.070865886214996e-06,
+ "loss": 0.378603458404541,
+ "num_input_tokens_seen": 376248416,
+ "step": 8600,
+ "train_runtime": 66412.9945,
+ "train_tokens_per_second": 5665.283
+ },
+ {
+ "epoch": 0.579383248047401,
+ "grad_norm": 0.8303736324901154,
+ "learning_rate": 8.06877882323581e-06,
+ "loss": 0.3776103496551514,
+ "num_input_tokens_seen": 376468952,
+ "step": 8605,
+ "train_runtime": 66445.0637,
+ "train_tokens_per_second": 5665.868
+ },
+ {
+ "epoch": 0.5797199030433612,
+ "grad_norm": 0.6628257697173621,
+ "learning_rate": 8.066690902082413e-06,
+ "loss": 0.35461578369140623,
+ "num_input_tokens_seen": 376690720,
+ "step": 8610,
+ "train_runtime": 66482.3424,
+ "train_tokens_per_second": 5666.027
+ },
+ {
+ "epoch": 0.5800565580393213,
+ "grad_norm": 0.6914119662414026,
+ "learning_rate": 8.064602123338683e-06,
+ "loss": 0.4189821720123291,
+ "num_input_tokens_seen": 376926952,
+ "step": 8615,
+ "train_runtime": 66523.5883,
+ "train_tokens_per_second": 5666.065
+ },
+ {
+ "epoch": 0.5803932130352815,
+ "grad_norm": 0.7255952843828343,
+ "learning_rate": 8.06251248758874e-06,
+ "loss": 0.3706305980682373,
+ "num_input_tokens_seen": 377137424,
+ "step": 8620,
+ "train_runtime": 66560.9464,
+ "train_tokens_per_second": 5666.047
+ },
+ {
+ "epoch": 0.5807298680312416,
+ "grad_norm": 0.7888860114164491,
+ "learning_rate": 8.060421995416949e-06,
+ "loss": 0.4162487030029297,
+ "num_input_tokens_seen": 377341792,
+ "step": 8625,
+ "train_runtime": 66601.9228,
+ "train_tokens_per_second": 5665.629
+ },
+ {
+ "epoch": 0.5810665230272017,
+ "grad_norm": 0.8097744521376653,
+ "learning_rate": 8.058330647407904e-06,
+ "loss": 0.3742199420928955,
+ "num_input_tokens_seen": 377553480,
+ "step": 8630,
+ "train_runtime": 66640.7232,
+ "train_tokens_per_second": 5665.507
+ },
+ {
+ "epoch": 0.5814031780231619,
+ "grad_norm": 0.9311380768738048,
+ "learning_rate": 8.056238444146445e-06,
+ "loss": 0.4032620429992676,
+ "num_input_tokens_seen": 377768576,
+ "step": 8635,
+ "train_runtime": 66676.5105,
+ "train_tokens_per_second": 5665.692
+ },
+ {
+ "epoch": 0.581739833019122,
+ "grad_norm": 0.7908376307807878,
+ "learning_rate": 8.054145386217652e-06,
+ "loss": 0.35233349800109864,
+ "num_input_tokens_seen": 377990912,
+ "step": 8640,
+ "train_runtime": 66709.357,
+ "train_tokens_per_second": 5666.235
+ },
+ {
+ "epoch": 0.5820764880150822,
+ "grad_norm": 0.7544609324314604,
+ "learning_rate": 8.05205147420684e-06,
+ "loss": 0.36639404296875,
+ "num_input_tokens_seen": 378205832,
+ "step": 8645,
+ "train_runtime": 66742.4566,
+ "train_tokens_per_second": 5666.645
+ },
+ {
+ "epoch": 0.5824131430110423,
+ "grad_norm": 0.8852523330609511,
+ "learning_rate": 8.049956708699567e-06,
+ "loss": 0.41215500831604,
+ "num_input_tokens_seen": 378422632,
+ "step": 8650,
+ "train_runtime": 66779.3643,
+ "train_tokens_per_second": 5666.76
+ },
+ {
+ "epoch": 0.5827497980070024,
+ "grad_norm": 0.7748134898838014,
+ "learning_rate": 8.047861090281623e-06,
+ "loss": 0.4017993450164795,
+ "num_input_tokens_seen": 378636312,
+ "step": 8655,
+ "train_runtime": 66817.0207,
+ "train_tokens_per_second": 5666.764
+ },
+ {
+ "epoch": 0.5830864530029626,
+ "grad_norm": 0.7174936151093502,
+ "learning_rate": 8.045764619539047e-06,
+ "loss": 0.36281633377075195,
+ "num_input_tokens_seen": 378862480,
+ "step": 8660,
+ "train_runtime": 66859.5287,
+ "train_tokens_per_second": 5666.544
+ },
+ {
+ "epoch": 0.5834231079989227,
+ "grad_norm": 0.7177129776779663,
+ "learning_rate": 8.043667297058106e-06,
+ "loss": 0.3568310022354126,
+ "num_input_tokens_seen": 379073648,
+ "step": 8665,
+ "train_runtime": 66894.7683,
+ "train_tokens_per_second": 5666.716
+ },
+ {
+ "epoch": 0.5837597629948829,
+ "grad_norm": 0.671557492383721,
+ "learning_rate": 8.04156912342531e-06,
+ "loss": 0.3924489259719849,
+ "num_input_tokens_seen": 379296048,
+ "step": 8670,
+ "train_runtime": 66930.6354,
+ "train_tokens_per_second": 5667.002
+ },
+ {
+ "epoch": 0.584096417990843,
+ "grad_norm": 0.8338696138817213,
+ "learning_rate": 8.03947009922741e-06,
+ "loss": 0.3730125904083252,
+ "num_input_tokens_seen": 379514216,
+ "step": 8675,
+ "train_runtime": 66971.0506,
+ "train_tokens_per_second": 5666.84
+ },
+ {
+ "epoch": 0.5844330729868031,
+ "grad_norm": 0.8046977070640711,
+ "learning_rate": 8.037370225051387e-06,
+ "loss": 0.4212483406066895,
+ "num_input_tokens_seen": 379726496,
+ "step": 8680,
+ "train_runtime": 67011.5775,
+ "train_tokens_per_second": 5666.58
+ },
+ {
+ "epoch": 0.5847697279827633,
+ "grad_norm": 0.747137316814288,
+ "learning_rate": 8.035269501484467e-06,
+ "loss": 0.3820772886276245,
+ "num_input_tokens_seen": 379952664,
+ "step": 8685,
+ "train_runtime": 67045.976,
+ "train_tokens_per_second": 5667.047
+ },
+ {
+ "epoch": 0.5851063829787234,
+ "grad_norm": 0.7779439978184742,
+ "learning_rate": 8.03316792911411e-06,
+ "loss": 0.41135897636413576,
+ "num_input_tokens_seen": 380168720,
+ "step": 8690,
+ "train_runtime": 67086.3352,
+ "train_tokens_per_second": 5666.858
+ },
+ {
+ "epoch": 0.5854430379746836,
+ "grad_norm": 0.6815474179189787,
+ "learning_rate": 8.031065508528011e-06,
+ "loss": 0.35835254192352295,
+ "num_input_tokens_seen": 380396160,
+ "step": 8695,
+ "train_runtime": 67124.515,
+ "train_tokens_per_second": 5667.023
+ },
+ {
+ "epoch": 0.5857796929706437,
+ "grad_norm": 0.7133086557979279,
+ "learning_rate": 8.028962240314111e-06,
+ "loss": 0.3686984539031982,
+ "num_input_tokens_seen": 380620472,
+ "step": 8700,
+ "train_runtime": 67161.6095,
+ "train_tokens_per_second": 5667.233
+ },
+ {
+ "epoch": 0.5861163479666038,
+ "grad_norm": 0.7088226284029326,
+ "learning_rate": 8.026858125060577e-06,
+ "loss": 0.3879531383514404,
+ "num_input_tokens_seen": 380837592,
+ "step": 8705,
+ "train_runtime": 67201.0534,
+ "train_tokens_per_second": 5667.137
+ },
+ {
+ "epoch": 0.586453002962564,
+ "grad_norm": 0.7612804968151702,
+ "learning_rate": 8.024753163355821e-06,
+ "loss": 0.4247321128845215,
+ "num_input_tokens_seen": 381061216,
+ "step": 8710,
+ "train_runtime": 67242.9084,
+ "train_tokens_per_second": 5666.935
+ },
+ {
+ "epoch": 0.5867896579585241,
+ "grad_norm": 0.7005358220776335,
+ "learning_rate": 8.022647355788489e-06,
+ "loss": 0.3679481029510498,
+ "num_input_tokens_seen": 381280288,
+ "step": 8715,
+ "train_runtime": 67281.4067,
+ "train_tokens_per_second": 5666.949
+ },
+ {
+ "epoch": 0.5871263129544843,
+ "grad_norm": 0.728044784671465,
+ "learning_rate": 8.02054070294746e-06,
+ "loss": 0.40247626304626466,
+ "num_input_tokens_seen": 381496992,
+ "step": 8720,
+ "train_runtime": 67319.4589,
+ "train_tokens_per_second": 5666.965
+ },
+ {
+ "epoch": 0.5874629679504444,
+ "grad_norm": 0.7052792472604399,
+ "learning_rate": 8.018433205421859e-06,
+ "loss": 0.36869523525238035,
+ "num_input_tokens_seen": 381727008,
+ "step": 8725,
+ "train_runtime": 67359.8794,
+ "train_tokens_per_second": 5666.979
+ },
+ {
+ "epoch": 0.5877996229464045,
+ "grad_norm": 0.8148613184034657,
+ "learning_rate": 8.016324863801036e-06,
+ "loss": 0.3896385669708252,
+ "num_input_tokens_seen": 381950608,
+ "step": 8730,
+ "train_runtime": 67395.2795,
+ "train_tokens_per_second": 5667.32
+ },
+ {
+ "epoch": 0.5881362779423647,
+ "grad_norm": 0.7008873936428064,
+ "learning_rate": 8.014215678674581e-06,
+ "loss": 0.397304368019104,
+ "num_input_tokens_seen": 382173664,
+ "step": 8735,
+ "train_runtime": 67436.6612,
+ "train_tokens_per_second": 5667.15
+ },
+ {
+ "epoch": 0.5884729329383248,
+ "grad_norm": 0.6667026328000061,
+ "learning_rate": 8.012105650632325e-06,
+ "loss": 0.37210235595703123,
+ "num_input_tokens_seen": 382388568,
+ "step": 8740,
+ "train_runtime": 67476.723,
+ "train_tokens_per_second": 5666.97
+ },
+ {
+ "epoch": 0.588809587934285,
+ "grad_norm": 0.7893753618716253,
+ "learning_rate": 8.009994780264328e-06,
+ "loss": 0.38750429153442384,
+ "num_input_tokens_seen": 382584944,
+ "step": 8745,
+ "train_runtime": 67517.8366,
+ "train_tokens_per_second": 5666.428
+ },
+ {
+ "epoch": 0.5891462429302451,
+ "grad_norm": 0.7520095638811525,
+ "learning_rate": 8.00788306816089e-06,
+ "loss": 0.38128883838653566,
+ "num_input_tokens_seen": 382801808,
+ "step": 8750,
+ "train_runtime": 67554.4917,
+ "train_tokens_per_second": 5666.563
+ },
+ {
+ "epoch": 0.5894828979262052,
+ "grad_norm": 0.7841152997217926,
+ "learning_rate": 8.005770514912543e-06,
+ "loss": 0.39379191398620605,
+ "num_input_tokens_seen": 383025832,
+ "step": 8755,
+ "train_runtime": 67593.0084,
+ "train_tokens_per_second": 5666.649
+ },
+ {
+ "epoch": 0.5898195529221654,
+ "grad_norm": 0.728445499817985,
+ "learning_rate": 8.003657121110055e-06,
+ "loss": 0.3845071792602539,
+ "num_input_tokens_seen": 383260728,
+ "step": 8760,
+ "train_runtime": 67628.415,
+ "train_tokens_per_second": 5667.155
+ },
+ {
+ "epoch": 0.5901562079181255,
+ "grad_norm": 0.7109388567699519,
+ "learning_rate": 8.001542887344432e-06,
+ "loss": 0.36870722770690917,
+ "num_input_tokens_seen": 383481016,
+ "step": 8765,
+ "train_runtime": 67663.424,
+ "train_tokens_per_second": 5667.479
+ },
+ {
+ "epoch": 0.5904928629140856,
+ "grad_norm": 0.7379119456141738,
+ "learning_rate": 7.999427814206911e-06,
+ "loss": 0.38972065448760984,
+ "num_input_tokens_seen": 383699928,
+ "step": 8770,
+ "train_runtime": 67698.1314,
+ "train_tokens_per_second": 5667.807
+ },
+ {
+ "epoch": 0.5908295179100458,
+ "grad_norm": 0.7977299068456677,
+ "learning_rate": 7.997311902288968e-06,
+ "loss": 0.3479362964630127,
+ "num_input_tokens_seen": 383923208,
+ "step": 8775,
+ "train_runtime": 67735.622,
+ "train_tokens_per_second": 5667.966
+ },
+ {
+ "epoch": 0.5911661729060059,
+ "grad_norm": 0.6927052461533478,
+ "learning_rate": 7.995195152182307e-06,
+ "loss": 0.35241219997406004,
+ "num_input_tokens_seen": 384124912,
+ "step": 8780,
+ "train_runtime": 67775.1001,
+ "train_tokens_per_second": 5667.641
+ },
+ {
+ "epoch": 0.5915028279019661,
+ "grad_norm": 1.8520827992951523,
+ "learning_rate": 7.993077564478875e-06,
+ "loss": 0.367362380027771,
+ "num_input_tokens_seen": 384352936,
+ "step": 8785,
+ "train_runtime": 67813.9355,
+ "train_tokens_per_second": 5667.757
+ },
+ {
+ "epoch": 0.5918394828979262,
+ "grad_norm": 0.8136091755899332,
+ "learning_rate": 7.990959139770844e-06,
+ "loss": 0.3980868339538574,
+ "num_input_tokens_seen": 384586952,
+ "step": 8790,
+ "train_runtime": 67855.5637,
+ "train_tokens_per_second": 5667.729
+ },
+ {
+ "epoch": 0.5921761378938863,
+ "grad_norm": 0.7689488310801835,
+ "learning_rate": 7.988839878650628e-06,
+ "loss": 0.3829834461212158,
+ "num_input_tokens_seen": 384813912,
+ "step": 8795,
+ "train_runtime": 67899.407,
+ "train_tokens_per_second": 5667.412
+ },
+ {
+ "epoch": 0.5925127928898465,
+ "grad_norm": 0.7981067639524714,
+ "learning_rate": 7.986719781710872e-06,
+ "loss": 0.3936153411865234,
+ "num_input_tokens_seen": 385040600,
+ "step": 8800,
+ "train_runtime": 67941.2356,
+ "train_tokens_per_second": 5667.259
+ },
+ {
+ "epoch": 0.5928494478858066,
+ "grad_norm": 0.7888638773984543,
+ "learning_rate": 7.984598849544452e-06,
+ "loss": 0.37601919174194337,
+ "num_input_tokens_seen": 385256736,
+ "step": 8805,
+ "train_runtime": 67980.1555,
+ "train_tokens_per_second": 5667.194
+ },
+ {
+ "epoch": 0.5931861028817668,
+ "grad_norm": 0.8141275660860262,
+ "learning_rate": 7.98247708274448e-06,
+ "loss": 0.3897037744522095,
+ "num_input_tokens_seen": 385471192,
+ "step": 8810,
+ "train_runtime": 68015.2704,
+ "train_tokens_per_second": 5667.421
+ },
+ {
+ "epoch": 0.5935227578777269,
+ "grad_norm": 0.7312305067797448,
+ "learning_rate": 7.980354481904303e-06,
+ "loss": 0.4111155033111572,
+ "num_input_tokens_seen": 385695968,
+ "step": 8815,
+ "train_runtime": 68053.7353,
+ "train_tokens_per_second": 5667.521
+ },
+ {
+ "epoch": 0.593859412873687,
+ "grad_norm": 0.8597141063820571,
+ "learning_rate": 7.978231047617498e-06,
+ "loss": 0.41021156311035156,
+ "num_input_tokens_seen": 385910576,
+ "step": 8820,
+ "train_runtime": 68091.813,
+ "train_tokens_per_second": 5667.503
+ },
+ {
+ "epoch": 0.5941960678696472,
+ "grad_norm": 0.7977576371807428,
+ "learning_rate": 7.976106780477878e-06,
+ "loss": 0.4051943778991699,
+ "num_input_tokens_seen": 386135568,
+ "step": 8825,
+ "train_runtime": 68134.8361,
+ "train_tokens_per_second": 5667.227
+ },
+ {
+ "epoch": 0.5945327228656073,
+ "grad_norm": 0.7369358873224232,
+ "learning_rate": 7.973981681079487e-06,
+ "loss": 0.3718585729598999,
+ "num_input_tokens_seen": 386344944,
+ "step": 8830,
+ "train_runtime": 68175.9892,
+ "train_tokens_per_second": 5666.877
+ },
+ {
+ "epoch": 0.5948693778615675,
+ "grad_norm": 0.7361018680183479,
+ "learning_rate": 7.971855750016603e-06,
+ "loss": 0.37234504222869874,
+ "num_input_tokens_seen": 386561056,
+ "step": 8835,
+ "train_runtime": 68210.534,
+ "train_tokens_per_second": 5667.175
+ },
+ {
+ "epoch": 0.5952060328575276,
+ "grad_norm": 0.7646438523333068,
+ "learning_rate": 7.969728987883734e-06,
+ "loss": 0.386487340927124,
+ "num_input_tokens_seen": 386782376,
+ "step": 8840,
+ "train_runtime": 68250.0947,
+ "train_tokens_per_second": 5667.133
+ },
+ {
+ "epoch": 0.5955426878534877,
+ "grad_norm": 0.8191747882763178,
+ "learning_rate": 7.967601395275624e-06,
+ "loss": 0.38187594413757325,
+ "num_input_tokens_seen": 386999696,
+ "step": 8845,
+ "train_runtime": 68291.3193,
+ "train_tokens_per_second": 5666.894
+ },
+ {
+ "epoch": 0.5958793428494479,
+ "grad_norm": 0.6459146656871505,
+ "learning_rate": 7.965472972787247e-06,
+ "loss": 0.35875325202941893,
+ "num_input_tokens_seen": 387220408,
+ "step": 8850,
+ "train_runtime": 68328.0647,
+ "train_tokens_per_second": 5667.077
+ },
+ {
+ "epoch": 0.596215997845408,
+ "grad_norm": 0.7549464702278598,
+ "learning_rate": 7.963343721013808e-06,
+ "loss": 0.3718303680419922,
+ "num_input_tokens_seen": 387455840,
+ "step": 8855,
+ "train_runtime": 68366.913,
+ "train_tokens_per_second": 5667.3
+ },
+ {
+ "epoch": 0.5965526528413682,
+ "grad_norm": 0.7558527061932778,
+ "learning_rate": 7.961213640550748e-06,
+ "loss": 0.3902780771255493,
+ "num_input_tokens_seen": 387678720,
+ "step": 8860,
+ "train_runtime": 68410.3443,
+ "train_tokens_per_second": 5666.961
+ },
+ {
+ "epoch": 0.5968893078373283,
+ "grad_norm": 0.7082248157089,
+ "learning_rate": 7.959082731993736e-06,
+ "loss": 0.3638461589813232,
+ "num_input_tokens_seen": 387897464,
+ "step": 8865,
+ "train_runtime": 68443.0142,
+ "train_tokens_per_second": 5667.452
+ },
+ {
+ "epoch": 0.5972259628332884,
+ "grad_norm": 0.9079660049109616,
+ "learning_rate": 7.956950995938673e-06,
+ "loss": 0.3705005168914795,
+ "num_input_tokens_seen": 388107624,
+ "step": 8870,
+ "train_runtime": 68486.4151,
+ "train_tokens_per_second": 5666.929
+ },
+ {
+ "epoch": 0.5975626178292486,
+ "grad_norm": 0.7984526639513112,
+ "learning_rate": 7.954818432981694e-06,
+ "loss": 0.3910086631774902,
+ "num_input_tokens_seen": 388310680,
+ "step": 8875,
+ "train_runtime": 68525.481,
+ "train_tokens_per_second": 5666.661
+ },
+ {
+ "epoch": 0.5978992728252087,
+ "grad_norm": 0.7147181194308313,
+ "learning_rate": 7.952685043719162e-06,
+ "loss": 0.3656719923019409,
+ "num_input_tokens_seen": 388533832,
+ "step": 8880,
+ "train_runtime": 68568.5855,
+ "train_tokens_per_second": 5666.353
+ },
+ {
+ "epoch": 0.5982359278211689,
+ "grad_norm": 0.7782537183487828,
+ "learning_rate": 7.950550828747672e-06,
+ "loss": 0.37486002445220945,
+ "num_input_tokens_seen": 388754728,
+ "step": 8885,
+ "train_runtime": 68608.906,
+ "train_tokens_per_second": 5666.243
+ },
+ {
+ "epoch": 0.598572582817129,
+ "grad_norm": 0.6958894420319796,
+ "learning_rate": 7.948415788664052e-06,
+ "loss": 0.40519165992736816,
+ "num_input_tokens_seen": 388977056,
+ "step": 8890,
+ "train_runtime": 68641.9616,
+ "train_tokens_per_second": 5666.753
+ },
+ {
+ "epoch": 0.5989092378130891,
+ "grad_norm": 0.8244159660705994,
+ "learning_rate": 7.946279924065359e-06,
+ "loss": 0.3696116924285889,
+ "num_input_tokens_seen": 389195368,
+ "step": 8895,
+ "train_runtime": 68676.2104,
+ "train_tokens_per_second": 5667.106
+ },
+ {
+ "epoch": 0.5992458928090493,
+ "grad_norm": 0.7624772559058,
+ "learning_rate": 7.94414323554888e-06,
+ "loss": 0.3721637725830078,
+ "num_input_tokens_seen": 389424256,
+ "step": 8900,
+ "train_runtime": 68707.9383,
+ "train_tokens_per_second": 5667.82
+ },
+ {
+ "epoch": 0.5995825478050094,
+ "grad_norm": 0.7421871285525378,
+ "learning_rate": 7.942005723712135e-06,
+ "loss": 0.3689994812011719,
+ "num_input_tokens_seen": 389659456,
+ "step": 8905,
+ "train_runtime": 68742.7631,
+ "train_tokens_per_second": 5668.371
+ },
+ {
+ "epoch": 0.5999192028009696,
+ "grad_norm": 0.7598715011111157,
+ "learning_rate": 7.939867389152868e-06,
+ "loss": 0.36181044578552246,
+ "num_input_tokens_seen": 389878008,
+ "step": 8910,
+ "train_runtime": 68782.9286,
+ "train_tokens_per_second": 5668.238
+ },
+ {
+ "epoch": 0.6002558577969297,
+ "grad_norm": 0.7642969305957489,
+ "learning_rate": 7.937728232469062e-06,
+ "loss": 0.3904815435409546,
+ "num_input_tokens_seen": 390090680,
+ "step": 8915,
+ "train_runtime": 68823.5271,
+ "train_tokens_per_second": 5667.984
+ },
+ {
+ "epoch": 0.6005925127928898,
+ "grad_norm": 0.7743252489033806,
+ "learning_rate": 7.935588254258921e-06,
+ "loss": 0.407269811630249,
+ "num_input_tokens_seen": 390298200,
+ "step": 8920,
+ "train_runtime": 68859.6954,
+ "train_tokens_per_second": 5668.021
+ },
+ {
+ "epoch": 0.60092916778885,
+ "grad_norm": 0.7626396676106256,
+ "learning_rate": 7.933447455120889e-06,
+ "loss": 0.3707296848297119,
+ "num_input_tokens_seen": 390513112,
+ "step": 8925,
+ "train_runtime": 68906.8163,
+ "train_tokens_per_second": 5667.264
+ },
+ {
+ "epoch": 0.6012658227848101,
+ "grad_norm": 0.9270450940086636,
+ "learning_rate": 7.931305835653629e-06,
+ "loss": 0.34468765258789064,
+ "num_input_tokens_seen": 390741944,
+ "step": 8930,
+ "train_runtime": 68945.5046,
+ "train_tokens_per_second": 5667.403
+ },
+ {
+ "epoch": 0.6016024777807703,
+ "grad_norm": 0.7523711200271979,
+ "learning_rate": 7.929163396456038e-06,
+ "loss": 0.39664325714111326,
+ "num_input_tokens_seen": 390950400,
+ "step": 8935,
+ "train_runtime": 68982.333,
+ "train_tokens_per_second": 5667.399
+ },
+ {
+ "epoch": 0.6019391327767304,
+ "grad_norm": 0.7743378155770866,
+ "learning_rate": 7.927020138127245e-06,
+ "loss": 0.369353199005127,
+ "num_input_tokens_seen": 391172832,
+ "step": 8940,
+ "train_runtime": 69022.2251,
+ "train_tokens_per_second": 5667.346
+ },
+ {
+ "epoch": 0.6022757877726905,
+ "grad_norm": 0.6649737691389936,
+ "learning_rate": 7.924876061266603e-06,
+ "loss": 0.3477231502532959,
+ "num_input_tokens_seen": 391403392,
+ "step": 8945,
+ "train_runtime": 69060.7271,
+ "train_tokens_per_second": 5667.525
+ },
+ {
+ "epoch": 0.6026124427686507,
+ "grad_norm": 0.8428633382608517,
+ "learning_rate": 7.922731166473698e-06,
+ "loss": 0.40044050216674804,
+ "num_input_tokens_seen": 391619536,
+ "step": 8950,
+ "train_runtime": 69095.3564,
+ "train_tokens_per_second": 5667.813
+ },
+ {
+ "epoch": 0.6029490977646108,
+ "grad_norm": 0.7797139187172423,
+ "learning_rate": 7.920585454348341e-06,
+ "loss": 0.35779447555541993,
+ "num_input_tokens_seen": 391832424,
+ "step": 8955,
+ "train_runtime": 69130.8816,
+ "train_tokens_per_second": 5667.98
+ },
+ {
+ "epoch": 0.603285752760571,
+ "grad_norm": 0.6454426330600102,
+ "learning_rate": 7.918438925490578e-06,
+ "loss": 0.35099799633026124,
+ "num_input_tokens_seen": 392057736,
+ "step": 8960,
+ "train_runtime": 69168.5952,
+ "train_tokens_per_second": 5668.147
+ },
+ {
+ "epoch": 0.6036224077565311,
+ "grad_norm": 0.9240196513971288,
+ "learning_rate": 7.916291580500672e-06,
+ "loss": 0.37104010581970215,
+ "num_input_tokens_seen": 392274152,
+ "step": 8965,
+ "train_runtime": 69209.1039,
+ "train_tokens_per_second": 5667.956
+ },
+ {
+ "epoch": 0.6039590627524912,
+ "grad_norm": 0.753245828305525,
+ "learning_rate": 7.914143419979127e-06,
+ "loss": 0.36745285987854004,
+ "num_input_tokens_seen": 392499120,
+ "step": 8970,
+ "train_runtime": 69247.1372,
+ "train_tokens_per_second": 5668.092
+ },
+ {
+ "epoch": 0.6042957177484514,
+ "grad_norm": 0.6569111554819693,
+ "learning_rate": 7.911994444526667e-06,
+ "loss": 0.34250531196594236,
+ "num_input_tokens_seen": 392720016,
+ "step": 8975,
+ "train_runtime": 69281.6006,
+ "train_tokens_per_second": 5668.46
+ },
+ {
+ "epoch": 0.6046323727444115,
+ "grad_norm": 0.6710386941472357,
+ "learning_rate": 7.909844654744246e-06,
+ "loss": 0.3526078939437866,
+ "num_input_tokens_seen": 392945160,
+ "step": 8980,
+ "train_runtime": 69318.5011,
+ "train_tokens_per_second": 5668.691
+ },
+ {
+ "epoch": 0.6049690277403716,
+ "grad_norm": 0.7627659966441698,
+ "learning_rate": 7.907694051233044e-06,
+ "loss": 0.38707795143127444,
+ "num_input_tokens_seen": 393164952,
+ "step": 8985,
+ "train_runtime": 69355.9491,
+ "train_tokens_per_second": 5668.799
+ },
+ {
+ "epoch": 0.6053056827363318,
+ "grad_norm": 0.6996994069888867,
+ "learning_rate": 7.905542634594476e-06,
+ "loss": 0.3759605884552002,
+ "num_input_tokens_seen": 393392000,
+ "step": 8990,
+ "train_runtime": 69403.0227,
+ "train_tokens_per_second": 5668.226
+ },
+ {
+ "epoch": 0.6056423377322919,
+ "grad_norm": 0.7505686616004086,
+ "learning_rate": 7.903390405430174e-06,
+ "loss": 0.36977381706237794,
+ "num_input_tokens_seen": 393611632,
+ "step": 8995,
+ "train_runtime": 69443.7718,
+ "train_tokens_per_second": 5668.062
+ },
+ {
+ "epoch": 0.6059789927282521,
+ "grad_norm": 0.7461690774426525,
+ "learning_rate": 7.901237364342003e-06,
+ "loss": 0.3545074939727783,
+ "num_input_tokens_seen": 393813184,
+ "step": 9000,
+ "train_runtime": 69478.2997,
+ "train_tokens_per_second": 5668.147
+ },
+ {
+ "epoch": 0.6063156477242122,
+ "grad_norm": 0.692733389183626,
+ "learning_rate": 7.899083511932053e-06,
+ "loss": 0.3470455646514893,
+ "num_input_tokens_seen": 394027272,
+ "step": 9005,
+ "train_runtime": 69518.7288,
+ "train_tokens_per_second": 5667.93
+ },
+ {
+ "epoch": 0.6066523027201723,
+ "grad_norm": 0.7055838759339507,
+ "learning_rate": 7.896928848802645e-06,
+ "loss": 0.372318959236145,
+ "num_input_tokens_seen": 394252480,
+ "step": 9010,
+ "train_runtime": 69553.6455,
+ "train_tokens_per_second": 5668.322
+ },
+ {
+ "epoch": 0.6069889577161325,
+ "grad_norm": 0.7576301760274441,
+ "learning_rate": 7.894773375556322e-06,
+ "loss": 0.39665966033935546,
+ "num_input_tokens_seen": 394449024,
+ "step": 9015,
+ "train_runtime": 69593.8001,
+ "train_tokens_per_second": 5667.876
+ },
+ {
+ "epoch": 0.6073256127120926,
+ "grad_norm": 0.8095166727188953,
+ "learning_rate": 7.892617092795855e-06,
+ "loss": 0.38990225791931155,
+ "num_input_tokens_seen": 394672680,
+ "step": 9020,
+ "train_runtime": 69632.6302,
+ "train_tokens_per_second": 5667.927
+ },
+ {
+ "epoch": 0.6076622677080528,
+ "grad_norm": 0.6966710594837849,
+ "learning_rate": 7.890460001124242e-06,
+ "loss": 0.3693479061126709,
+ "num_input_tokens_seen": 394882720,
+ "step": 9025,
+ "train_runtime": 69668.2129,
+ "train_tokens_per_second": 5668.047
+ },
+ {
+ "epoch": 0.6079989227040129,
+ "grad_norm": 0.7467966463857733,
+ "learning_rate": 7.888302101144707e-06,
+ "loss": 0.3727553367614746,
+ "num_input_tokens_seen": 395096336,
+ "step": 9030,
+ "train_runtime": 69710.5053,
+ "train_tokens_per_second": 5667.673
+ },
+ {
+ "epoch": 0.608335577699973,
+ "grad_norm": 0.7033388740227385,
+ "learning_rate": 7.886143393460699e-06,
+ "loss": 0.38645596504211427,
+ "num_input_tokens_seen": 395327104,
+ "step": 9035,
+ "train_runtime": 69746.6807,
+ "train_tokens_per_second": 5668.042
+ },
+ {
+ "epoch": 0.6086722326959332,
+ "grad_norm": 0.717725417977876,
+ "learning_rate": 7.883983878675895e-06,
+ "loss": 0.3641960620880127,
+ "num_input_tokens_seen": 395561000,
+ "step": 9040,
+ "train_runtime": 69794.3801,
+ "train_tokens_per_second": 5667.519
+ },
+ {
+ "epoch": 0.6090088876918933,
+ "grad_norm": 0.737338173324883,
+ "learning_rate": 7.881823557394194e-06,
+ "loss": 0.3701021194458008,
+ "num_input_tokens_seen": 395790544,
+ "step": 9045,
+ "train_runtime": 69832.741,
+ "train_tokens_per_second": 5667.693
+ },
+ {
+ "epoch": 0.6093455426878535,
+ "grad_norm": 0.7527693128608011,
+ "learning_rate": 7.879662430219728e-06,
+ "loss": 0.35156404972076416,
+ "num_input_tokens_seen": 395987680,
+ "step": 9050,
+ "train_runtime": 69873.2033,
+ "train_tokens_per_second": 5667.232
+ },
+ {
+ "epoch": 0.6096821976838136,
+ "grad_norm": 0.8175940814840023,
+ "learning_rate": 7.877500497756843e-06,
+ "loss": 0.4006368637084961,
+ "num_input_tokens_seen": 396207224,
+ "step": 9055,
+ "train_runtime": 69914.8014,
+ "train_tokens_per_second": 5667.001
+ },
+ {
+ "epoch": 0.6100188526797737,
+ "grad_norm": 1.0466330679614098,
+ "learning_rate": 7.87533776061012e-06,
+ "loss": 0.3919075012207031,
+ "num_input_tokens_seen": 396431144,
+ "step": 9060,
+ "train_runtime": 69954.8331,
+ "train_tokens_per_second": 5666.959
+ },
+ {
+ "epoch": 0.6103555076757339,
+ "grad_norm": 0.7304596174173837,
+ "learning_rate": 7.873174219384362e-06,
+ "loss": 0.3918348789215088,
+ "num_input_tokens_seen": 396647760,
+ "step": 9065,
+ "train_runtime": 69996.3796,
+ "train_tokens_per_second": 5666.69
+ },
+ {
+ "epoch": 0.610692162671694,
+ "grad_norm": 0.7556787705072924,
+ "learning_rate": 7.871009874684595e-06,
+ "loss": 0.3941289186477661,
+ "num_input_tokens_seen": 396880640,
+ "step": 9070,
+ "train_runtime": 70028.8034,
+ "train_tokens_per_second": 5667.391
+ },
+ {
+ "epoch": 0.6110288176676542,
+ "grad_norm": 0.8397284206623034,
+ "learning_rate": 7.868844727116072e-06,
+ "loss": 0.3926485300064087,
+ "num_input_tokens_seen": 397103976,
+ "step": 9075,
+ "train_runtime": 70068.2945,
+ "train_tokens_per_second": 5667.385
+ },
+ {
+ "epoch": 0.6113654726636143,
+ "grad_norm": 0.6869244210168536,
+ "learning_rate": 7.866678777284267e-06,
+ "loss": 0.3840303897857666,
+ "num_input_tokens_seen": 397331160,
+ "step": 9080,
+ "train_runtime": 70102.9706,
+ "train_tokens_per_second": 5667.822
+ },
+ {
+ "epoch": 0.6117021276595744,
+ "grad_norm": 0.6366270970878026,
+ "learning_rate": 7.864512025794886e-06,
+ "loss": 0.3659891843795776,
+ "num_input_tokens_seen": 397555992,
+ "step": 9085,
+ "train_runtime": 70142.9573,
+ "train_tokens_per_second": 5667.796
+ },
+ {
+ "epoch": 0.6120387826555346,
+ "grad_norm": 0.7702173454778595,
+ "learning_rate": 7.86234447325385e-06,
+ "loss": 0.36196537017822267,
+ "num_input_tokens_seen": 397769720,
+ "step": 9090,
+ "train_runtime": 70185.3981,
+ "train_tokens_per_second": 5667.414
+ },
+ {
+ "epoch": 0.6123754376514947,
+ "grad_norm": 0.6638241416823153,
+ "learning_rate": 7.860176120267309e-06,
+ "loss": 0.4144433498382568,
+ "num_input_tokens_seen": 398006352,
+ "step": 9095,
+ "train_runtime": 70227.0746,
+ "train_tokens_per_second": 5667.42
+ },
+ {
+ "epoch": 0.6127120926474549,
+ "grad_norm": 0.7106604037911711,
+ "learning_rate": 7.858006967441637e-06,
+ "loss": 0.3602044105529785,
+ "num_input_tokens_seen": 398216992,
+ "step": 9100,
+ "train_runtime": 70268.2952,
+ "train_tokens_per_second": 5667.093
+ },
+ {
+ "epoch": 0.613048747643415,
+ "grad_norm": 0.8013079858059877,
+ "learning_rate": 7.85583701538343e-06,
+ "loss": 0.40687074661254885,
+ "num_input_tokens_seen": 398447240,
+ "step": 9105,
+ "train_runtime": 70313.64,
+ "train_tokens_per_second": 5666.713
+ },
+ {
+ "epoch": 0.6133854026393751,
+ "grad_norm": 0.8363243678158666,
+ "learning_rate": 7.853666264699506e-06,
+ "loss": 0.3638491153717041,
+ "num_input_tokens_seen": 398656928,
+ "step": 9110,
+ "train_runtime": 70355.8426,
+ "train_tokens_per_second": 5666.295
+ },
+ {
+ "epoch": 0.6137220576353353,
+ "grad_norm": 0.6855947504133412,
+ "learning_rate": 7.851494715996914e-06,
+ "loss": 0.41129069328308104,
+ "num_input_tokens_seen": 398870368,
+ "step": 9115,
+ "train_runtime": 70395.581,
+ "train_tokens_per_second": 5666.128
+ },
+ {
+ "epoch": 0.6140587126312954,
+ "grad_norm": 0.7613022857869545,
+ "learning_rate": 7.849322369882915e-06,
+ "loss": 0.3685478687286377,
+ "num_input_tokens_seen": 399088536,
+ "step": 9120,
+ "train_runtime": 70444.8636,
+ "train_tokens_per_second": 5665.261
+ },
+ {
+ "epoch": 0.6143953676272556,
+ "grad_norm": 0.6988300008519205,
+ "learning_rate": 7.847149226965003e-06,
+ "loss": 0.34872984886169434,
+ "num_input_tokens_seen": 399301232,
+ "step": 9125,
+ "train_runtime": 70485.0506,
+ "train_tokens_per_second": 5665.049
+ },
+ {
+ "epoch": 0.6147320226232157,
+ "grad_norm": 0.8477724491109673,
+ "learning_rate": 7.844975287850886e-06,
+ "loss": 0.35025968551635744,
+ "num_input_tokens_seen": 399523504,
+ "step": 9130,
+ "train_runtime": 70522.5255,
+ "train_tokens_per_second": 5665.19
+ },
+ {
+ "epoch": 0.6150686776191758,
+ "grad_norm": 0.7692634619323344,
+ "learning_rate": 7.842800553148506e-06,
+ "loss": 0.3998849630355835,
+ "num_input_tokens_seen": 399735256,
+ "step": 9135,
+ "train_runtime": 70558.7444,
+ "train_tokens_per_second": 5665.283
+ },
+ {
+ "epoch": 0.615405332615136,
+ "grad_norm": 0.7313920740237436,
+ "learning_rate": 7.840625023466014e-06,
+ "loss": 0.3812243938446045,
+ "num_input_tokens_seen": 399958824,
+ "step": 9140,
+ "train_runtime": 70598.2086,
+ "train_tokens_per_second": 5665.283
+ },
+ {
+ "epoch": 0.6157419876110961,
+ "grad_norm": 0.7274387978674163,
+ "learning_rate": 7.838448699411792e-06,
+ "loss": 0.3413673400878906,
+ "num_input_tokens_seen": 400181528,
+ "step": 9145,
+ "train_runtime": 70641.2437,
+ "train_tokens_per_second": 5664.984
+ },
+ {
+ "epoch": 0.6160786426070562,
+ "grad_norm": 0.7006449570449917,
+ "learning_rate": 7.836271581594442e-06,
+ "loss": 0.4210306167602539,
+ "num_input_tokens_seen": 400404728,
+ "step": 9150,
+ "train_runtime": 70677.6291,
+ "train_tokens_per_second": 5665.226
+ },
+ {
+ "epoch": 0.6164152976030164,
+ "grad_norm": 0.8552267430665972,
+ "learning_rate": 7.834093670622789e-06,
+ "loss": 0.3944708824157715,
+ "num_input_tokens_seen": 400620128,
+ "step": 9155,
+ "train_runtime": 70718.8549,
+ "train_tokens_per_second": 5664.969
+ },
+ {
+ "epoch": 0.6167519525989765,
+ "grad_norm": 0.7631479691553084,
+ "learning_rate": 7.831914967105878e-06,
+ "loss": 0.4071969985961914,
+ "num_input_tokens_seen": 400851832,
+ "step": 9160,
+ "train_runtime": 70759.3063,
+ "train_tokens_per_second": 5665.005
+ },
+ {
+ "epoch": 0.6170886075949367,
+ "grad_norm": 0.6980894707116475,
+ "learning_rate": 7.829735471652978e-06,
+ "loss": 0.3579000234603882,
+ "num_input_tokens_seen": 401070744,
+ "step": 9165,
+ "train_runtime": 70799.3413,
+ "train_tokens_per_second": 5664.894
+ },
+ {
+ "epoch": 0.6174252625908968,
+ "grad_norm": 0.8778577602385942,
+ "learning_rate": 7.827555184873575e-06,
+ "loss": 0.38968710899353026,
+ "num_input_tokens_seen": 401294816,
+ "step": 9170,
+ "train_runtime": 70831.4898,
+ "train_tokens_per_second": 5665.486
+ },
+ {
+ "epoch": 0.617761917586857,
+ "grad_norm": 0.6780563561113029,
+ "learning_rate": 7.825374107377381e-06,
+ "loss": 0.35830237865448,
+ "num_input_tokens_seen": 401509200,
+ "step": 9175,
+ "train_runtime": 70873.5987,
+ "train_tokens_per_second": 5665.145
+ },
+ {
+ "epoch": 0.6180985725828171,
+ "grad_norm": 0.7956649065370722,
+ "learning_rate": 7.823192239774327e-06,
+ "loss": 0.3784201145172119,
+ "num_input_tokens_seen": 401720584,
+ "step": 9180,
+ "train_runtime": 70913.365,
+ "train_tokens_per_second": 5664.949
+ },
+ {
+ "epoch": 0.6184352275787772,
+ "grad_norm": 0.689595538429294,
+ "learning_rate": 7.821009582674567e-06,
+ "loss": 0.3932128667831421,
+ "num_input_tokens_seen": 401937000,
+ "step": 9185,
+ "train_runtime": 70950.5951,
+ "train_tokens_per_second": 5665.026
+ },
+ {
+ "epoch": 0.6187718825747374,
+ "grad_norm": 0.7661957206566802,
+ "learning_rate": 7.818826136688473e-06,
+ "loss": 0.41967573165893557,
+ "num_input_tokens_seen": 402163016,
+ "step": 9190,
+ "train_runtime": 70988.3051,
+ "train_tokens_per_second": 5665.201
+ },
+ {
+ "epoch": 0.6191085375706975,
+ "grad_norm": 1.24632241828128,
+ "learning_rate": 7.816641902426638e-06,
+ "loss": 0.3871551513671875,
+ "num_input_tokens_seen": 402390608,
+ "step": 9195,
+ "train_runtime": 71026.2689,
+ "train_tokens_per_second": 5665.377
+ },
+ {
+ "epoch": 0.6194451925666576,
+ "grad_norm": 0.7754009003596685,
+ "learning_rate": 7.814456880499877e-06,
+ "loss": 0.39989705085754396,
+ "num_input_tokens_seen": 402608440,
+ "step": 9200,
+ "train_runtime": 71059.0686,
+ "train_tokens_per_second": 5665.828
+ },
+ {
+ "epoch": 0.6197818475626178,
+ "grad_norm": 0.8050073203615641,
+ "learning_rate": 7.812271071519222e-06,
+ "loss": 0.40456953048706057,
+ "num_input_tokens_seen": 402830072,
+ "step": 9205,
+ "train_runtime": 71092.5653,
+ "train_tokens_per_second": 5666.276
+ },
+ {
+ "epoch": 0.6201185025585779,
+ "grad_norm": 0.636871155540424,
+ "learning_rate": 7.810084476095932e-06,
+ "loss": 0.3537304401397705,
+ "num_input_tokens_seen": 403051544,
+ "step": 9210,
+ "train_runtime": 71132.4539,
+ "train_tokens_per_second": 5666.212
+ },
+ {
+ "epoch": 0.6204551575545381,
+ "grad_norm": 0.7771337802144807,
+ "learning_rate": 7.807897094841476e-06,
+ "loss": 0.3726407527923584,
+ "num_input_tokens_seen": 403260152,
+ "step": 9215,
+ "train_runtime": 71171.5092,
+ "train_tokens_per_second": 5666.033
+ },
+ {
+ "epoch": 0.6207918125504982,
+ "grad_norm": 0.8018367108579978,
+ "learning_rate": 7.805708928367555e-06,
+ "loss": 0.3734896183013916,
+ "num_input_tokens_seen": 403468824,
+ "step": 9220,
+ "train_runtime": 71211.2179,
+ "train_tokens_per_second": 5665.804
+ },
+ {
+ "epoch": 0.6211284675464583,
+ "grad_norm": 0.7409189256900955,
+ "learning_rate": 7.803519977286075e-06,
+ "loss": 0.4021895885467529,
+ "num_input_tokens_seen": 403677704,
+ "step": 9225,
+ "train_runtime": 71249.4495,
+ "train_tokens_per_second": 5665.696
+ },
+ {
+ "epoch": 0.6214651225424185,
+ "grad_norm": 0.7435590926911755,
+ "learning_rate": 7.801330242209176e-06,
+ "loss": 0.38091113567352297,
+ "num_input_tokens_seen": 403883200,
+ "step": 9230,
+ "train_runtime": 71290.1392,
+ "train_tokens_per_second": 5665.345
+ },
+ {
+ "epoch": 0.6218017775383786,
+ "grad_norm": 0.7623979983718003,
+ "learning_rate": 7.799139723749207e-06,
+ "loss": 0.3860699415206909,
+ "num_input_tokens_seen": 404093128,
+ "step": 9235,
+ "train_runtime": 71328.4839,
+ "train_tokens_per_second": 5665.242
+ },
+ {
+ "epoch": 0.6221384325343388,
+ "grad_norm": 0.752221691602039,
+ "learning_rate": 7.79694842251874e-06,
+ "loss": 0.3864139080047607,
+ "num_input_tokens_seen": 404319728,
+ "step": 9240,
+ "train_runtime": 71368.1022,
+ "train_tokens_per_second": 5665.272
+ },
+ {
+ "epoch": 0.6224750875302989,
+ "grad_norm": 0.669956786900923,
+ "learning_rate": 7.794756339130565e-06,
+ "loss": 0.3734030485153198,
+ "num_input_tokens_seen": 404541968,
+ "step": 9245,
+ "train_runtime": 71411.2121,
+ "train_tokens_per_second": 5664.964
+ },
+ {
+ "epoch": 0.622811742526259,
+ "grad_norm": 0.6983661416716165,
+ "learning_rate": 7.792563474197692e-06,
+ "loss": 0.3814718008041382,
+ "num_input_tokens_seen": 404754312,
+ "step": 9250,
+ "train_runtime": 71448.6745,
+ "train_tokens_per_second": 5664.966
+ },
+ {
+ "epoch": 0.6231483975222192,
+ "grad_norm": 0.7124131828610318,
+ "learning_rate": 7.79036982833335e-06,
+ "loss": 0.4030300617218018,
+ "num_input_tokens_seen": 404985704,
+ "step": 9255,
+ "train_runtime": 71479.1768,
+ "train_tokens_per_second": 5665.786
+ },
+ {
+ "epoch": 0.6234850525181793,
+ "grad_norm": 0.7886397068317147,
+ "learning_rate": 7.78817540215098e-06,
+ "loss": 0.3688653469085693,
+ "num_input_tokens_seen": 405208848,
+ "step": 9260,
+ "train_runtime": 71523.5445,
+ "train_tokens_per_second": 5665.391
+ },
+ {
+ "epoch": 0.6238217075141395,
+ "grad_norm": 0.7219385997208987,
+ "learning_rate": 7.78598019626425e-06,
+ "loss": 0.37583611011505125,
+ "num_input_tokens_seen": 405447552,
+ "step": 9265,
+ "train_runtime": 71555.7936,
+ "train_tokens_per_second": 5666.174
+ },
+ {
+ "epoch": 0.6241583625100996,
+ "grad_norm": 0.7695265603997767,
+ "learning_rate": 7.783784211287044e-06,
+ "loss": 0.4006516456604004,
+ "num_input_tokens_seen": 405651112,
+ "step": 9270,
+ "train_runtime": 71597.6416,
+ "train_tokens_per_second": 5665.705
+ },
+ {
+ "epoch": 0.6244950175060597,
+ "grad_norm": 0.9194045285642082,
+ "learning_rate": 7.78158744783346e-06,
+ "loss": 0.408467960357666,
+ "num_input_tokens_seen": 405873152,
+ "step": 9275,
+ "train_runtime": 71636.7797,
+ "train_tokens_per_second": 5665.709
+ },
+ {
+ "epoch": 0.6248316725020199,
+ "grad_norm": 0.7515447419413074,
+ "learning_rate": 7.779389906517813e-06,
+ "loss": 0.35638930797576907,
+ "num_input_tokens_seen": 406086632,
+ "step": 9280,
+ "train_runtime": 71670.2143,
+ "train_tokens_per_second": 5666.045
+ },
+ {
+ "epoch": 0.6251683274979801,
+ "grad_norm": 0.7184945703842209,
+ "learning_rate": 7.777191587954645e-06,
+ "loss": 0.3749531745910645,
+ "num_input_tokens_seen": 406298704,
+ "step": 9285,
+ "train_runtime": 71712.7407,
+ "train_tokens_per_second": 5665.642
+ },
+ {
+ "epoch": 0.6255049824939403,
+ "grad_norm": 0.8027326683853181,
+ "learning_rate": 7.774992492758704e-06,
+ "loss": 0.36041698455810545,
+ "num_input_tokens_seen": 406502360,
+ "step": 9290,
+ "train_runtime": 71750.3732,
+ "train_tokens_per_second": 5665.509
+ },
+ {
+ "epoch": 0.6258416374899004,
+ "grad_norm": 0.7186651340067036,
+ "learning_rate": 7.77279262154496e-06,
+ "loss": 0.38366947174072263,
+ "num_input_tokens_seen": 406727224,
+ "step": 9295,
+ "train_runtime": 71789.6701,
+ "train_tokens_per_second": 5665.54
+ },
+ {
+ "epoch": 0.6261782924858605,
+ "grad_norm": 0.8295991455328212,
+ "learning_rate": 7.770591974928604e-06,
+ "loss": 0.3997042655944824,
+ "num_input_tokens_seen": 406958480,
+ "step": 9300,
+ "train_runtime": 71824.8324,
+ "train_tokens_per_second": 5665.986
+ },
+ {
+ "epoch": 0.6265149474818207,
+ "grad_norm": 0.7658676057217327,
+ "learning_rate": 7.768390553525034e-06,
+ "loss": 0.3705620527267456,
+ "num_input_tokens_seen": 407186440,
+ "step": 9305,
+ "train_runtime": 71865.0099,
+ "train_tokens_per_second": 5665.99
+ },
+ {
+ "epoch": 0.6268516024777808,
+ "grad_norm": 0.7085619378845062,
+ "learning_rate": 7.766188357949875e-06,
+ "loss": 0.3714077711105347,
+ "num_input_tokens_seen": 407407560,
+ "step": 9310,
+ "train_runtime": 71909.2798,
+ "train_tokens_per_second": 5665.577
+ },
+ {
+ "epoch": 0.627188257473741,
+ "grad_norm": 0.8341480094252892,
+ "learning_rate": 7.763985388818963e-06,
+ "loss": 0.36980538368225097,
+ "num_input_tokens_seen": 407628232,
+ "step": 9315,
+ "train_runtime": 71949.141,
+ "train_tokens_per_second": 5665.505
+ },
+ {
+ "epoch": 0.6275249124697011,
+ "grad_norm": 0.8636262815852889,
+ "learning_rate": 7.76178164674835e-06,
+ "loss": 0.3548977613449097,
+ "num_input_tokens_seen": 407838184,
+ "step": 9320,
+ "train_runtime": 71979.0743,
+ "train_tokens_per_second": 5666.066
+ },
+ {
+ "epoch": 0.6278615674656612,
+ "grad_norm": 0.7192127713238392,
+ "learning_rate": 7.759577132354305e-06,
+ "loss": 0.3763836622238159,
+ "num_input_tokens_seen": 408065528,
+ "step": 9325,
+ "train_runtime": 72013.1228,
+ "train_tokens_per_second": 5666.544
+ },
+ {
+ "epoch": 0.6281982224616214,
+ "grad_norm": 0.7987348676964026,
+ "learning_rate": 7.757371846253315e-06,
+ "loss": 0.3501241683959961,
+ "num_input_tokens_seen": 408275816,
+ "step": 9330,
+ "train_runtime": 72052.1024,
+ "train_tokens_per_second": 5666.397
+ },
+ {
+ "epoch": 0.6285348774575815,
+ "grad_norm": 0.7470954183186648,
+ "learning_rate": 7.755165789062083e-06,
+ "loss": 0.4022618293762207,
+ "num_input_tokens_seen": 408498408,
+ "step": 9335,
+ "train_runtime": 72090.7302,
+ "train_tokens_per_second": 5666.448
+ },
+ {
+ "epoch": 0.6288715324535417,
+ "grad_norm": 0.7151603552727446,
+ "learning_rate": 7.752958961397522e-06,
+ "loss": 0.3877957105636597,
+ "num_input_tokens_seen": 408733352,
+ "step": 9340,
+ "train_runtime": 72120.5104,
+ "train_tokens_per_second": 5667.366
+ },
+ {
+ "epoch": 0.6292081874495018,
+ "grad_norm": 0.7351213604942984,
+ "learning_rate": 7.750751363876765e-06,
+ "loss": 0.39035401344299314,
+ "num_input_tokens_seen": 408941920,
+ "step": 9345,
+ "train_runtime": 72164.8249,
+ "train_tokens_per_second": 5666.776
+ },
+ {
+ "epoch": 0.6295448424454619,
+ "grad_norm": 0.8453874832584365,
+ "learning_rate": 7.748542997117162e-06,
+ "loss": 0.3737621307373047,
+ "num_input_tokens_seen": 409162768,
+ "step": 9350,
+ "train_runtime": 72196.2605,
+ "train_tokens_per_second": 5667.368
+ },
+ {
+ "epoch": 0.6298814974414221,
+ "grad_norm": 0.7429854652224369,
+ "learning_rate": 7.746333861736275e-06,
+ "loss": 0.37030327320098877,
+ "num_input_tokens_seen": 409372800,
+ "step": 9355,
+ "train_runtime": 72230.736,
+ "train_tokens_per_second": 5667.571
+ },
+ {
+ "epoch": 0.6302181524373822,
+ "grad_norm": 0.8081848245845143,
+ "learning_rate": 7.74412395835188e-06,
+ "loss": 0.3631394386291504,
+ "num_input_tokens_seen": 409583128,
+ "step": 9360,
+ "train_runtime": 72272.2265,
+ "train_tokens_per_second": 5667.227
+ },
+ {
+ "epoch": 0.6305548074333424,
+ "grad_norm": 0.6576701373107874,
+ "learning_rate": 7.74191328758197e-06,
+ "loss": 0.37232241630554197,
+ "num_input_tokens_seen": 409794192,
+ "step": 9365,
+ "train_runtime": 72313.6306,
+ "train_tokens_per_second": 5666.901
+ },
+ {
+ "epoch": 0.6308914624293025,
+ "grad_norm": 0.735743623178803,
+ "learning_rate": 7.739701850044752e-06,
+ "loss": 0.4075340270996094,
+ "num_input_tokens_seen": 410020536,
+ "step": 9370,
+ "train_runtime": 72350.0803,
+ "train_tokens_per_second": 5667.175
+ },
+ {
+ "epoch": 0.6312281174252626,
+ "grad_norm": 0.8241305453975977,
+ "learning_rate": 7.73748964635865e-06,
+ "loss": 0.3718372583389282,
+ "num_input_tokens_seen": 410241072,
+ "step": 9375,
+ "train_runtime": 72387.3064,
+ "train_tokens_per_second": 5667.307
+ },
+ {
+ "epoch": 0.6315647724212228,
+ "grad_norm": 0.6824612940355289,
+ "learning_rate": 7.735276677142297e-06,
+ "loss": 0.35228054523468016,
+ "num_input_tokens_seen": 410443760,
+ "step": 9380,
+ "train_runtime": 72423.3738,
+ "train_tokens_per_second": 5667.283
+ },
+ {
+ "epoch": 0.6319014274171829,
+ "grad_norm": 0.7420817513182338,
+ "learning_rate": 7.733062943014542e-06,
+ "loss": 0.3992206573486328,
+ "num_input_tokens_seen": 410670656,
+ "step": 9385,
+ "train_runtime": 72463.1943,
+ "train_tokens_per_second": 5667.3
+ },
+ {
+ "epoch": 0.632238082413143,
+ "grad_norm": 0.7090881615954795,
+ "learning_rate": 7.730848444594452e-06,
+ "loss": 0.3573267698287964,
+ "num_input_tokens_seen": 410903448,
+ "step": 9390,
+ "train_runtime": 72497.2497,
+ "train_tokens_per_second": 5667.849
+ },
+ {
+ "epoch": 0.6325747374091032,
+ "grad_norm": 0.7269251761731511,
+ "learning_rate": 7.7286331825013e-06,
+ "loss": 0.37146763801574706,
+ "num_input_tokens_seen": 411126128,
+ "step": 9395,
+ "train_runtime": 72531.5005,
+ "train_tokens_per_second": 5668.242
+ },
+ {
+ "epoch": 0.6329113924050633,
+ "grad_norm": 0.8641784035147556,
+ "learning_rate": 7.72641715735458e-06,
+ "loss": 0.3655902624130249,
+ "num_input_tokens_seen": 411336720,
+ "step": 9400,
+ "train_runtime": 72578.3797,
+ "train_tokens_per_second": 5667.483
+ },
+ {
+ "epoch": 0.6332480474010235,
+ "grad_norm": 0.7283249670371048,
+ "learning_rate": 7.724200369773998e-06,
+ "loss": 0.38719449043273924,
+ "num_input_tokens_seen": 411563872,
+ "step": 9405,
+ "train_runtime": 72618.0679,
+ "train_tokens_per_second": 5667.513
+ },
+ {
+ "epoch": 0.6335847023969836,
+ "grad_norm": 0.6863677356579423,
+ "learning_rate": 7.721982820379467e-06,
+ "loss": 0.37484476566314695,
+ "num_input_tokens_seen": 411787952,
+ "step": 9410,
+ "train_runtime": 72652.1342,
+ "train_tokens_per_second": 5667.94
+ },
+ {
+ "epoch": 0.6339213573929438,
+ "grad_norm": 0.7975234970060876,
+ "learning_rate": 7.719764509791123e-06,
+ "loss": 0.3741923332214355,
+ "num_input_tokens_seen": 411975928,
+ "step": 9415,
+ "train_runtime": 72685.292,
+ "train_tokens_per_second": 5667.941
+ },
+ {
+ "epoch": 0.6342580123889039,
+ "grad_norm": 0.735648228811728,
+ "learning_rate": 7.717545438629303e-06,
+ "loss": 0.43533172607421877,
+ "num_input_tokens_seen": 412214368,
+ "step": 9420,
+ "train_runtime": 72721.9474,
+ "train_tokens_per_second": 5668.363
+ },
+ {
+ "epoch": 0.634594667384864,
+ "grad_norm": 0.6852408582111305,
+ "learning_rate": 7.71532560751457e-06,
+ "loss": 0.3732184410095215,
+ "num_input_tokens_seen": 412434712,
+ "step": 9425,
+ "train_runtime": 72757.9756,
+ "train_tokens_per_second": 5668.584
+ },
+ {
+ "epoch": 0.6349313223808242,
+ "grad_norm": 0.7315681955128508,
+ "learning_rate": 7.713105017067686e-06,
+ "loss": 0.38740081787109376,
+ "num_input_tokens_seen": 412668752,
+ "step": 9430,
+ "train_runtime": 72791.6587,
+ "train_tokens_per_second": 5669.176
+ },
+ {
+ "epoch": 0.6352679773767843,
+ "grad_norm": 0.840365118006204,
+ "learning_rate": 7.710883667909639e-06,
+ "loss": 0.37178239822387693,
+ "num_input_tokens_seen": 412886888,
+ "step": 9435,
+ "train_runtime": 72827.5502,
+ "train_tokens_per_second": 5669.378
+ },
+ {
+ "epoch": 0.6356046323727444,
+ "grad_norm": 0.7096620735032497,
+ "learning_rate": 7.708661560661617e-06,
+ "loss": 0.3967950105667114,
+ "num_input_tokens_seen": 413111704,
+ "step": 9440,
+ "train_runtime": 72863.4631,
+ "train_tokens_per_second": 5669.669
+ },
+ {
+ "epoch": 0.6359412873687046,
+ "grad_norm": 0.7646005775520524,
+ "learning_rate": 7.706438695945026e-06,
+ "loss": 0.39545307159423826,
+ "num_input_tokens_seen": 413330656,
+ "step": 9445,
+ "train_runtime": 72901.529,
+ "train_tokens_per_second": 5669.712
+ },
+ {
+ "epoch": 0.6362779423646647,
+ "grad_norm": 0.7774443776922321,
+ "learning_rate": 7.704215074381484e-06,
+ "loss": 0.3966789960861206,
+ "num_input_tokens_seen": 413537792,
+ "step": 9450,
+ "train_runtime": 72937.8522,
+ "train_tokens_per_second": 5669.728
+ },
+ {
+ "epoch": 0.6366145973606249,
+ "grad_norm": 0.7509437079381308,
+ "learning_rate": 7.70199069659282e-06,
+ "loss": 0.38933987617492677,
+ "num_input_tokens_seen": 413751760,
+ "step": 9455,
+ "train_runtime": 72973.609,
+ "train_tokens_per_second": 5669.882
+ },
+ {
+ "epoch": 0.636951252356585,
+ "grad_norm": 0.7325824670692431,
+ "learning_rate": 7.699765563201075e-06,
+ "loss": 0.3942718982696533,
+ "num_input_tokens_seen": 413978016,
+ "step": 9460,
+ "train_runtime": 73003.9087,
+ "train_tokens_per_second": 5670.628
+ },
+ {
+ "epoch": 0.6372879073525451,
+ "grad_norm": 1.211650286474132,
+ "learning_rate": 7.697539674828498e-06,
+ "loss": 0.36391506195068357,
+ "num_input_tokens_seen": 414211120,
+ "step": 9465,
+ "train_runtime": 73043.0392,
+ "train_tokens_per_second": 5670.782
+ },
+ {
+ "epoch": 0.6376245623485053,
+ "grad_norm": 0.7046659464816453,
+ "learning_rate": 7.695313032097554e-06,
+ "loss": 0.35442147254943845,
+ "num_input_tokens_seen": 414430048,
+ "step": 9470,
+ "train_runtime": 73079.8991,
+ "train_tokens_per_second": 5670.917
+ },
+ {
+ "epoch": 0.6379612173444654,
+ "grad_norm": 0.6947182191976057,
+ "learning_rate": 7.693085635630913e-06,
+ "loss": 0.37603862285614015,
+ "num_input_tokens_seen": 414632320,
+ "step": 9475,
+ "train_runtime": 73126.3077,
+ "train_tokens_per_second": 5670.084
+ },
+ {
+ "epoch": 0.6382978723404256,
+ "grad_norm": 0.7520528416471214,
+ "learning_rate": 7.690857486051463e-06,
+ "loss": 0.3694300174713135,
+ "num_input_tokens_seen": 414859816,
+ "step": 9480,
+ "train_runtime": 73166.4344,
+ "train_tokens_per_second": 5670.084
+ },
+ {
+ "epoch": 0.6386345273363857,
+ "grad_norm": 0.7317534743877758,
+ "learning_rate": 7.688628583982298e-06,
+ "loss": 0.3734642267227173,
+ "num_input_tokens_seen": 415071856,
+ "step": 9485,
+ "train_runtime": 73208.2652,
+ "train_tokens_per_second": 5669.74
+ },
+ {
+ "epoch": 0.6389711823323458,
+ "grad_norm": 0.7042485073751668,
+ "learning_rate": 7.68639893004672e-06,
+ "loss": 0.3597568035125732,
+ "num_input_tokens_seen": 415294888,
+ "step": 9490,
+ "train_runtime": 73249.0074,
+ "train_tokens_per_second": 5669.632
+ },
+ {
+ "epoch": 0.639307837328306,
+ "grad_norm": 0.7092737777153273,
+ "learning_rate": 7.684168524868253e-06,
+ "loss": 0.3566474437713623,
+ "num_input_tokens_seen": 415526408,
+ "step": 9495,
+ "train_runtime": 73288.7331,
+ "train_tokens_per_second": 5669.717
+ },
+ {
+ "epoch": 0.6396444923242661,
+ "grad_norm": 0.788876930336864,
+ "learning_rate": 7.681937369070612e-06,
+ "loss": 0.37108328342437746,
+ "num_input_tokens_seen": 415741912,
+ "step": 9500,
+ "train_runtime": 73321.3834,
+ "train_tokens_per_second": 5670.132
+ },
+ {
+ "epoch": 0.6399811473202263,
+ "grad_norm": 0.7482496645801756,
+ "learning_rate": 7.67970546327774e-06,
+ "loss": 0.3800150156021118,
+ "num_input_tokens_seen": 415962440,
+ "step": 9505,
+ "train_runtime": 73359.2213,
+ "train_tokens_per_second": 5670.213
+ },
+ {
+ "epoch": 0.6403178023161864,
+ "grad_norm": 0.7003781550580643,
+ "learning_rate": 7.67747280811378e-06,
+ "loss": 0.3659864902496338,
+ "num_input_tokens_seen": 416179208,
+ "step": 9510,
+ "train_runtime": 73399.2881,
+ "train_tokens_per_second": 5670.071
+ },
+ {
+ "epoch": 0.6406544573121465,
+ "grad_norm": 0.6823118163683989,
+ "learning_rate": 7.675239404203088e-06,
+ "loss": 0.35662126541137695,
+ "num_input_tokens_seen": 416394640,
+ "step": 9515,
+ "train_runtime": 73439.5055,
+ "train_tokens_per_second": 5669.9
+ },
+ {
+ "epoch": 0.6409911123081067,
+ "grad_norm": 0.7167030632794295,
+ "learning_rate": 7.673005252170226e-06,
+ "loss": 0.3727959156036377,
+ "num_input_tokens_seen": 416620784,
+ "step": 9520,
+ "train_runtime": 73484.0265,
+ "train_tokens_per_second": 5669.542
+ },
+ {
+ "epoch": 0.6413277673040668,
+ "grad_norm": 0.7593758397546905,
+ "learning_rate": 7.670770352639968e-06,
+ "loss": 0.3673388957977295,
+ "num_input_tokens_seen": 416826168,
+ "step": 9525,
+ "train_runtime": 73525.2289,
+ "train_tokens_per_second": 5669.158
+ },
+ {
+ "epoch": 0.641664422300027,
+ "grad_norm": 0.7362340230809692,
+ "learning_rate": 7.668534706237299e-06,
+ "loss": 0.37580122947692873,
+ "num_input_tokens_seen": 417047480,
+ "step": 9530,
+ "train_runtime": 73556.9783,
+ "train_tokens_per_second": 5669.72
+ },
+ {
+ "epoch": 0.6420010772959871,
+ "grad_norm": 0.7543938330712561,
+ "learning_rate": 7.666298313587409e-06,
+ "loss": 0.4170052528381348,
+ "num_input_tokens_seen": 417268504,
+ "step": 9535,
+ "train_runtime": 73597.5716,
+ "train_tokens_per_second": 5669.596
+ },
+ {
+ "epoch": 0.6423377322919472,
+ "grad_norm": 0.8008472981307811,
+ "learning_rate": 7.664061175315696e-06,
+ "loss": 0.3715662956237793,
+ "num_input_tokens_seen": 417464000,
+ "step": 9540,
+ "train_runtime": 73631.8047,
+ "train_tokens_per_second": 5669.615
+ },
+ {
+ "epoch": 0.6426743872879074,
+ "grad_norm": 0.7188017157103265,
+ "learning_rate": 7.661823292047771e-06,
+ "loss": 0.3764479160308838,
+ "num_input_tokens_seen": 417683920,
+ "step": 9545,
+ "train_runtime": 73672.7623,
+ "train_tokens_per_second": 5669.448
+ },
+ {
+ "epoch": 0.6430110422838675,
+ "grad_norm": 0.8895228408009356,
+ "learning_rate": 7.659584664409452e-06,
+ "loss": 0.3806633949279785,
+ "num_input_tokens_seen": 417910080,
+ "step": 9550,
+ "train_runtime": 73714.3818,
+ "train_tokens_per_second": 5669.315
+ },
+ {
+ "epoch": 0.6433476972798277,
+ "grad_norm": 0.7813418710468062,
+ "learning_rate": 7.65734529302676e-06,
+ "loss": 0.33327679634094237,
+ "num_input_tokens_seen": 418114864,
+ "step": 9555,
+ "train_runtime": 73751.4559,
+ "train_tokens_per_second": 5669.242
+ },
+ {
+ "epoch": 0.6436843522757878,
+ "grad_norm": 0.7440032604844485,
+ "learning_rate": 7.655105178525932e-06,
+ "loss": 0.36725008487701416,
+ "num_input_tokens_seen": 418331608,
+ "step": 9560,
+ "train_runtime": 73789.4942,
+ "train_tokens_per_second": 5669.257
+ },
+ {
+ "epoch": 0.6440210072717479,
+ "grad_norm": 0.7428051543591697,
+ "learning_rate": 7.652864321533406e-06,
+ "loss": 0.37874999046325686,
+ "num_input_tokens_seen": 418542736,
+ "step": 9565,
+ "train_runtime": 73827.1886,
+ "train_tokens_per_second": 5669.222
+ },
+ {
+ "epoch": 0.6443576622677081,
+ "grad_norm": 0.6906490664289745,
+ "learning_rate": 7.650622722675833e-06,
+ "loss": 0.39217414855957033,
+ "num_input_tokens_seen": 418769104,
+ "step": 9570,
+ "train_runtime": 73869.6555,
+ "train_tokens_per_second": 5669.027
+ },
+ {
+ "epoch": 0.6446943172636682,
+ "grad_norm": 0.7836457733135137,
+ "learning_rate": 7.648380382580067e-06,
+ "loss": 0.3911130428314209,
+ "num_input_tokens_seen": 418983272,
+ "step": 9575,
+ "train_runtime": 73911.9877,
+ "train_tokens_per_second": 5668.678
+ },
+ {
+ "epoch": 0.6450309722596284,
+ "grad_norm": 0.7191438751807192,
+ "learning_rate": 7.646137301873173e-06,
+ "loss": 0.3877583026885986,
+ "num_input_tokens_seen": 419193152,
+ "step": 9580,
+ "train_runtime": 73952.718,
+ "train_tokens_per_second": 5668.394
+ },
+ {
+ "epoch": 0.6453676272555885,
+ "grad_norm": 0.8722963532886264,
+ "learning_rate": 7.643893481182421e-06,
+ "loss": 0.35512332916259765,
+ "num_input_tokens_seen": 419415144,
+ "step": 9585,
+ "train_runtime": 73990.0513,
+ "train_tokens_per_second": 5668.534
+ },
+ {
+ "epoch": 0.6457042822515486,
+ "grad_norm": 0.7893095386452317,
+ "learning_rate": 7.641648921135287e-06,
+ "loss": 0.4030734062194824,
+ "num_input_tokens_seen": 419642840,
+ "step": 9590,
+ "train_runtime": 74032.6519,
+ "train_tokens_per_second": 5668.348
+ },
+ {
+ "epoch": 0.6460409372475088,
+ "grad_norm": 0.8691991766270082,
+ "learning_rate": 7.639403622359455e-06,
+ "loss": 0.38401873111724855,
+ "num_input_tokens_seen": 419851960,
+ "step": 9595,
+ "train_runtime": 74068.9663,
+ "train_tokens_per_second": 5668.392
+ },
+ {
+ "epoch": 0.6463775922434689,
+ "grad_norm": 0.7669041161649485,
+ "learning_rate": 7.637157585482817e-06,
+ "loss": 0.3698650598526001,
+ "num_input_tokens_seen": 420069368,
+ "step": 9600,
+ "train_runtime": 74113.2241,
+ "train_tokens_per_second": 5667.941
+ },
+ {
+ "epoch": 0.646714247239429,
+ "grad_norm": 0.6764016385660067,
+ "learning_rate": 7.63491081113347e-06,
+ "loss": 0.3979948043823242,
+ "num_input_tokens_seen": 420301264,
+ "step": 9605,
+ "train_runtime": 74149.7053,
+ "train_tokens_per_second": 5668.28
+ },
+ {
+ "epoch": 0.6470509022353892,
+ "grad_norm": 0.7632828915951216,
+ "learning_rate": 7.632663299939718e-06,
+ "loss": 0.4009084701538086,
+ "num_input_tokens_seen": 420528480,
+ "step": 9610,
+ "train_runtime": 74191.7204,
+ "train_tokens_per_second": 5668.132
+ },
+ {
+ "epoch": 0.6473875572313493,
+ "grad_norm": 0.7759849429505897,
+ "learning_rate": 7.630415052530066e-06,
+ "loss": 0.37017059326171875,
+ "num_input_tokens_seen": 420744352,
+ "step": 9615,
+ "train_runtime": 74228.9204,
+ "train_tokens_per_second": 5668.2
+ },
+ {
+ "epoch": 0.6477242122273095,
+ "grad_norm": 0.7521586068543743,
+ "learning_rate": 7.628166069533235e-06,
+ "loss": 0.3800983905792236,
+ "num_input_tokens_seen": 420951552,
+ "step": 9620,
+ "train_runtime": 74264.9231,
+ "train_tokens_per_second": 5668.242
+ },
+ {
+ "epoch": 0.6480608672232696,
+ "grad_norm": 0.8746614520646498,
+ "learning_rate": 7.625916351578141e-06,
+ "loss": 0.4101721286773682,
+ "num_input_tokens_seen": 421168904,
+ "step": 9625,
+ "train_runtime": 74302.0251,
+ "train_tokens_per_second": 5668.337
+ },
+ {
+ "epoch": 0.6483975222192297,
+ "grad_norm": 0.645310594782184,
+ "learning_rate": 7.623665899293914e-06,
+ "loss": 0.35078365802764894,
+ "num_input_tokens_seen": 421386640,
+ "step": 9630,
+ "train_runtime": 74343.6656,
+ "train_tokens_per_second": 5668.091
+ },
+ {
+ "epoch": 0.6487341772151899,
+ "grad_norm": 0.709824008157462,
+ "learning_rate": 7.621414713309885e-06,
+ "loss": 0.3464890241622925,
+ "num_input_tokens_seen": 421604504,
+ "step": 9635,
+ "train_runtime": 74380.7993,
+ "train_tokens_per_second": 5668.19
+ },
+ {
+ "epoch": 0.64907083221115,
+ "grad_norm": 0.6987475287759364,
+ "learning_rate": 7.61916279425559e-06,
+ "loss": 0.38260612487792967,
+ "num_input_tokens_seen": 421833672,
+ "step": 9640,
+ "train_runtime": 74422.4312,
+ "train_tokens_per_second": 5668.099
+ },
+ {
+ "epoch": 0.6494074872071102,
+ "grad_norm": 0.727875668145333,
+ "learning_rate": 7.6169101427607716e-06,
+ "loss": 0.38509573936462405,
+ "num_input_tokens_seen": 422067496,
+ "step": 9645,
+ "train_runtime": 74456.6368,
+ "train_tokens_per_second": 5668.635
+ },
+ {
+ "epoch": 0.6497441422030703,
+ "grad_norm": 0.7549567194100304,
+ "learning_rate": 7.614656759455375e-06,
+ "loss": 0.38222179412841795,
+ "num_input_tokens_seen": 422289144,
+ "step": 9650,
+ "train_runtime": 74495.8893,
+ "train_tokens_per_second": 5668.623
+ },
+ {
+ "epoch": 0.6500807971990304,
+ "grad_norm": 0.8363753913347584,
+ "learning_rate": 7.612402644969555e-06,
+ "loss": 0.38720729351043703,
+ "num_input_tokens_seen": 422500848,
+ "step": 9655,
+ "train_runtime": 74537.09,
+ "train_tokens_per_second": 5668.33
+ },
+ {
+ "epoch": 0.6504174521949906,
+ "grad_norm": 0.718216642211442,
+ "learning_rate": 7.6101477999336646e-06,
+ "loss": 0.37707133293151857,
+ "num_input_tokens_seen": 422718672,
+ "step": 9660,
+ "train_runtime": 74584.7849,
+ "train_tokens_per_second": 5667.626
+ },
+ {
+ "epoch": 0.6507541071909507,
+ "grad_norm": 0.8872821374211818,
+ "learning_rate": 7.607892224978266e-06,
+ "loss": 0.3739673376083374,
+ "num_input_tokens_seen": 422929480,
+ "step": 9665,
+ "train_runtime": 74620.552,
+ "train_tokens_per_second": 5667.735
+ },
+ {
+ "epoch": 0.6510907621869109,
+ "grad_norm": 0.691277480453269,
+ "learning_rate": 7.605635920734122e-06,
+ "loss": 0.3418617010116577,
+ "num_input_tokens_seen": 423142128,
+ "step": 9670,
+ "train_runtime": 74655.9636,
+ "train_tokens_per_second": 5667.895
+ },
+ {
+ "epoch": 0.651427417182871,
+ "grad_norm": 0.9025153211455825,
+ "learning_rate": 7.603378887832202e-06,
+ "loss": 0.3855189561843872,
+ "num_input_tokens_seen": 423349056,
+ "step": 9675,
+ "train_runtime": 74695.4667,
+ "train_tokens_per_second": 5667.667
+ },
+ {
+ "epoch": 0.6517640721788311,
+ "grad_norm": 0.7073570329177757,
+ "learning_rate": 7.601121126903676e-06,
+ "loss": 0.3803261280059814,
+ "num_input_tokens_seen": 423573240,
+ "step": 9680,
+ "train_runtime": 74738.2061,
+ "train_tokens_per_second": 5667.426
+ },
+ {
+ "epoch": 0.6521007271747913,
+ "grad_norm": 0.7609618594052814,
+ "learning_rate": 7.598862638579922e-06,
+ "loss": 0.3738381385803223,
+ "num_input_tokens_seen": 423809440,
+ "step": 9685,
+ "train_runtime": 74781.8146,
+ "train_tokens_per_second": 5667.279
+ },
+ {
+ "epoch": 0.6524373821707514,
+ "grad_norm": 0.6698743309851123,
+ "learning_rate": 7.5966034234925205e-06,
+ "loss": 0.3761345624923706,
+ "num_input_tokens_seen": 424041528,
+ "step": 9690,
+ "train_runtime": 74821.3132,
+ "train_tokens_per_second": 5667.39
+ },
+ {
+ "epoch": 0.6527740371667116,
+ "grad_norm": 0.7414693193437045,
+ "learning_rate": 7.59434348227325e-06,
+ "loss": 0.43105735778808596,
+ "num_input_tokens_seen": 424268184,
+ "step": 9695,
+ "train_runtime": 74855.3245,
+ "train_tokens_per_second": 5667.842
+ },
+ {
+ "epoch": 0.6531106921626717,
+ "grad_norm": 0.8237582570303243,
+ "learning_rate": 7.592082815554098e-06,
+ "loss": 0.40675368309021,
+ "num_input_tokens_seen": 424492616,
+ "step": 9700,
+ "train_runtime": 74894.094,
+ "train_tokens_per_second": 5667.905
+ },
+ {
+ "epoch": 0.6534473471586318,
+ "grad_norm": 0.9212376319324829,
+ "learning_rate": 7.58982142396725e-06,
+ "loss": 0.419390869140625,
+ "num_input_tokens_seen": 424686848,
+ "step": 9705,
+ "train_runtime": 74939.3759,
+ "train_tokens_per_second": 5667.072
+ },
+ {
+ "epoch": 0.653784002154592,
+ "grad_norm": 0.7691414977900423,
+ "learning_rate": 7.5875593081451e-06,
+ "loss": 0.3695266008377075,
+ "num_input_tokens_seen": 424913632,
+ "step": 9710,
+ "train_runtime": 74985.8205,
+ "train_tokens_per_second": 5666.586
+ },
+ {
+ "epoch": 0.6541206571505521,
+ "grad_norm": 0.7263975698689714,
+ "learning_rate": 7.585296468720243e-06,
+ "loss": 0.3619475126266479,
+ "num_input_tokens_seen": 425139040,
+ "step": 9715,
+ "train_runtime": 75026.5575,
+ "train_tokens_per_second": 5666.514
+ },
+ {
+ "epoch": 0.6544573121465123,
+ "grad_norm": 0.712706963298683,
+ "learning_rate": 7.5830329063254705e-06,
+ "loss": 0.38277277946472166,
+ "num_input_tokens_seen": 425352824,
+ "step": 9720,
+ "train_runtime": 75065.2226,
+ "train_tokens_per_second": 5666.443
+ },
+ {
+ "epoch": 0.6547939671424724,
+ "grad_norm": 0.6764513937336157,
+ "learning_rate": 7.5807686215937835e-06,
+ "loss": 0.38765199184417726,
+ "num_input_tokens_seen": 425565784,
+ "step": 9725,
+ "train_runtime": 75105.3744,
+ "train_tokens_per_second": 5666.249
+ },
+ {
+ "epoch": 0.6551306221384325,
+ "grad_norm": 0.7196173883927505,
+ "learning_rate": 7.578503615158379e-06,
+ "loss": 0.3398409366607666,
+ "num_input_tokens_seen": 425778128,
+ "step": 9730,
+ "train_runtime": 75149.2065,
+ "train_tokens_per_second": 5665.77
+ },
+ {
+ "epoch": 0.6554672771343927,
+ "grad_norm": 0.646960134820036,
+ "learning_rate": 7.576237887652662e-06,
+ "loss": 0.38698830604553225,
+ "num_input_tokens_seen": 426007040,
+ "step": 9735,
+ "train_runtime": 75188.2698,
+ "train_tokens_per_second": 5665.871
+ },
+ {
+ "epoch": 0.6558039321303528,
+ "grad_norm": 0.8063318146524463,
+ "learning_rate": 7.573971439710235e-06,
+ "loss": 0.34668645858764646,
+ "num_input_tokens_seen": 426223488,
+ "step": 9740,
+ "train_runtime": 75225.1526,
+ "train_tokens_per_second": 5665.97
+ },
+ {
+ "epoch": 0.656140587126313,
+ "grad_norm": 0.8258091202754753,
+ "learning_rate": 7.571704271964902e-06,
+ "loss": 0.3675506591796875,
+ "num_input_tokens_seen": 426433528,
+ "step": 9745,
+ "train_runtime": 75260.6499,
+ "train_tokens_per_second": 5666.089
+ },
+ {
+ "epoch": 0.6564772421222731,
+ "grad_norm": 0.8027675726296949,
+ "learning_rate": 7.56943638505067e-06,
+ "loss": 0.3636334419250488,
+ "num_input_tokens_seen": 426658696,
+ "step": 9750,
+ "train_runtime": 75296.5611,
+ "train_tokens_per_second": 5666.377
+ },
+ {
+ "epoch": 0.6568138971182332,
+ "grad_norm": 0.7190477185688757,
+ "learning_rate": 7.567167779601746e-06,
+ "loss": 0.39123508930206297,
+ "num_input_tokens_seen": 426886792,
+ "step": 9755,
+ "train_runtime": 75338.0495,
+ "train_tokens_per_second": 5666.284
+ },
+ {
+ "epoch": 0.6571505521141934,
+ "grad_norm": 0.7906692823017727,
+ "learning_rate": 7.56489845625254e-06,
+ "loss": 0.413257360458374,
+ "num_input_tokens_seen": 427104192,
+ "step": 9760,
+ "train_runtime": 75386.2254,
+ "train_tokens_per_second": 5665.547
+ },
+ {
+ "epoch": 0.6574872071101535,
+ "grad_norm": 0.7301115269654299,
+ "learning_rate": 7.562628415637659e-06,
+ "loss": 0.369996976852417,
+ "num_input_tokens_seen": 427312504,
+ "step": 9765,
+ "train_runtime": 75426.5577,
+ "train_tokens_per_second": 5665.279
+ },
+ {
+ "epoch": 0.6578238621061137,
+ "grad_norm": 0.7224806256488935,
+ "learning_rate": 7.560357658391913e-06,
+ "loss": 0.4071043491363525,
+ "num_input_tokens_seen": 427533792,
+ "step": 9770,
+ "train_runtime": 75465.4869,
+ "train_tokens_per_second": 5665.289
+ },
+ {
+ "epoch": 0.6581605171020738,
+ "grad_norm": 0.7268098674561758,
+ "learning_rate": 7.558086185150311e-06,
+ "loss": 0.3804465770721436,
+ "num_input_tokens_seen": 427747608,
+ "step": 9775,
+ "train_runtime": 75507.0805,
+ "train_tokens_per_second": 5665.0
+ },
+ {
+ "epoch": 0.6584971720980339,
+ "grad_norm": 0.7850729912855515,
+ "learning_rate": 7.555813996548065e-06,
+ "loss": 0.40397090911865235,
+ "num_input_tokens_seen": 427972008,
+ "step": 9780,
+ "train_runtime": 75543.708,
+ "train_tokens_per_second": 5665.224
+ },
+ {
+ "epoch": 0.6588338270939941,
+ "grad_norm": 0.7662329566093665,
+ "learning_rate": 7.553541093220586e-06,
+ "loss": 0.3842537641525269,
+ "num_input_tokens_seen": 428186312,
+ "step": 9785,
+ "train_runtime": 75579.4178,
+ "train_tokens_per_second": 5665.383
+ },
+ {
+ "epoch": 0.6591704820899542,
+ "grad_norm": 0.70627363493871,
+ "learning_rate": 7.551267475803481e-06,
+ "loss": 0.37559702396392824,
+ "num_input_tokens_seen": 428413952,
+ "step": 9790,
+ "train_runtime": 75619.0124,
+ "train_tokens_per_second": 5665.426
+ },
+ {
+ "epoch": 0.6595071370859144,
+ "grad_norm": 0.7354101668176468,
+ "learning_rate": 7.5489931449325635e-06,
+ "loss": 0.3736457347869873,
+ "num_input_tokens_seen": 428630728,
+ "step": 9795,
+ "train_runtime": 75658.6215,
+ "train_tokens_per_second": 5665.326
+ },
+ {
+ "epoch": 0.6598437920818745,
+ "grad_norm": 0.7892531883838505,
+ "learning_rate": 7.546718101243842e-06,
+ "loss": 0.39138045310974123,
+ "num_input_tokens_seen": 428842760,
+ "step": 9800,
+ "train_runtime": 75693.8787,
+ "train_tokens_per_second": 5665.488
+ },
+ {
+ "epoch": 0.6601804470778346,
+ "grad_norm": 0.8775677410278766,
+ "learning_rate": 7.544442345373522e-06,
+ "loss": 0.3698113918304443,
+ "num_input_tokens_seen": 429051072,
+ "step": 9805,
+ "train_runtime": 75734.5916,
+ "train_tokens_per_second": 5665.193
+ },
+ {
+ "epoch": 0.6605171020737948,
+ "grad_norm": 0.6944135512445987,
+ "learning_rate": 7.542165877958014e-06,
+ "loss": 0.3717064380645752,
+ "num_input_tokens_seen": 429266232,
+ "step": 9810,
+ "train_runtime": 75771.7628,
+ "train_tokens_per_second": 5665.253
+ },
+ {
+ "epoch": 0.6608537570697549,
+ "grad_norm": 0.7361334690779209,
+ "learning_rate": 7.539888699633926e-06,
+ "loss": 0.38160529136657717,
+ "num_input_tokens_seen": 429495032,
+ "step": 9815,
+ "train_runtime": 75808.8176,
+ "train_tokens_per_second": 5665.502
+ },
+ {
+ "epoch": 0.661190412065715,
+ "grad_norm": 0.8126000202476551,
+ "learning_rate": 7.5376108110380606e-06,
+ "loss": 0.34899187088012695,
+ "num_input_tokens_seen": 429723384,
+ "step": 9820,
+ "train_runtime": 75847.4225,
+ "train_tokens_per_second": 5665.629
+ },
+ {
+ "epoch": 0.6615270670616752,
+ "grad_norm": 0.6957748548678448,
+ "learning_rate": 7.535332212807426e-06,
+ "loss": 0.3428369998931885,
+ "num_input_tokens_seen": 429924744,
+ "step": 9825,
+ "train_runtime": 75886.3897,
+ "train_tokens_per_second": 5665.374
+ },
+ {
+ "epoch": 0.6618637220576353,
+ "grad_norm": 0.7233083648231398,
+ "learning_rate": 7.53305290557922e-06,
+ "loss": 0.39725618362426757,
+ "num_input_tokens_seen": 430154928,
+ "step": 9830,
+ "train_runtime": 75925.3068,
+ "train_tokens_per_second": 5665.501
+ },
+ {
+ "epoch": 0.6622003770535955,
+ "grad_norm": 0.7049996354735961,
+ "learning_rate": 7.530772889990848e-06,
+ "loss": 0.3740149974822998,
+ "num_input_tokens_seen": 430386800,
+ "step": 9835,
+ "train_runtime": 75959.5503,
+ "train_tokens_per_second": 5666.0
+ },
+ {
+ "epoch": 0.6625370320495556,
+ "grad_norm": 0.6720383481048683,
+ "learning_rate": 7.528492166679907e-06,
+ "loss": 0.35007429122924805,
+ "num_input_tokens_seen": 430616592,
+ "step": 9840,
+ "train_runtime": 76001.567,
+ "train_tokens_per_second": 5665.891
+ },
+ {
+ "epoch": 0.6628736870455157,
+ "grad_norm": 0.6809307882652357,
+ "learning_rate": 7.526210736284194e-06,
+ "loss": 0.38547792434692385,
+ "num_input_tokens_seen": 430820368,
+ "step": 9845,
+ "train_runtime": 76036.0252,
+ "train_tokens_per_second": 5666.003
+ },
+ {
+ "epoch": 0.6632103420414759,
+ "grad_norm": 0.7241468747622457,
+ "learning_rate": 7.523928599441706e-06,
+ "loss": 0.3851293087005615,
+ "num_input_tokens_seen": 431045232,
+ "step": 9850,
+ "train_runtime": 76076.642,
+ "train_tokens_per_second": 5665.934
+ },
+ {
+ "epoch": 0.663546997037436,
+ "grad_norm": 0.8065697031009399,
+ "learning_rate": 7.5216457567906295e-06,
+ "loss": 0.3921393632888794,
+ "num_input_tokens_seen": 431272728,
+ "step": 9855,
+ "train_runtime": 76119.1303,
+ "train_tokens_per_second": 5665.76
+ },
+ {
+ "epoch": 0.6638836520333962,
+ "grad_norm": 0.7363581812525671,
+ "learning_rate": 7.519362208969362e-06,
+ "loss": 0.3589025974273682,
+ "num_input_tokens_seen": 431497576,
+ "step": 9860,
+ "train_runtime": 76164.9303,
+ "train_tokens_per_second": 5665.305
+ },
+ {
+ "epoch": 0.6642203070293563,
+ "grad_norm": 0.7028175355387544,
+ "learning_rate": 7.5170779566164855e-06,
+ "loss": 0.39649462699890137,
+ "num_input_tokens_seen": 431716768,
+ "step": 9865,
+ "train_runtime": 76207.7038,
+ "train_tokens_per_second": 5665.002
+ },
+ {
+ "epoch": 0.6645569620253164,
+ "grad_norm": 0.8083377822452007,
+ "learning_rate": 7.514793000370783e-06,
+ "loss": 0.39783761501312254,
+ "num_input_tokens_seen": 431927680,
+ "step": 9870,
+ "train_runtime": 76252.3393,
+ "train_tokens_per_second": 5664.452
+ },
+ {
+ "epoch": 0.6648936170212766,
+ "grad_norm": 0.7361462900334614,
+ "learning_rate": 7.512507340871239e-06,
+ "loss": 0.3664602756500244,
+ "num_input_tokens_seen": 432144784,
+ "step": 9875,
+ "train_runtime": 76293.2197,
+ "train_tokens_per_second": 5664.262
+ },
+ {
+ "epoch": 0.6652302720172367,
+ "grad_norm": 0.7771322596278541,
+ "learning_rate": 7.510220978757028e-06,
+ "loss": 0.39647390842437746,
+ "num_input_tokens_seen": 432361504,
+ "step": 9880,
+ "train_runtime": 76335.1474,
+ "train_tokens_per_second": 5663.99
+ },
+ {
+ "epoch": 0.6655669270131969,
+ "grad_norm": 0.9351257512839886,
+ "learning_rate": 7.507933914667524e-06,
+ "loss": 0.4201164722442627,
+ "num_input_tokens_seen": 432586992,
+ "step": 9885,
+ "train_runtime": 76375.8314,
+ "train_tokens_per_second": 5663.925
+ },
+ {
+ "epoch": 0.665903582009157,
+ "grad_norm": 0.7531976870702812,
+ "learning_rate": 7.505646149242299e-06,
+ "loss": 0.37460851669311523,
+ "num_input_tokens_seen": 432809080,
+ "step": 9890,
+ "train_runtime": 76415.9118,
+ "train_tokens_per_second": 5663.861
+ },
+ {
+ "epoch": 0.6662402370051171,
+ "grad_norm": 0.719734237163997,
+ "learning_rate": 7.503357683121119e-06,
+ "loss": 0.34871246814727785,
+ "num_input_tokens_seen": 433035120,
+ "step": 9895,
+ "train_runtime": 76462.7297,
+ "train_tokens_per_second": 5663.349
+ },
+ {
+ "epoch": 0.6665768920010773,
+ "grad_norm": 0.7130561120719486,
+ "learning_rate": 7.501068516943944e-06,
+ "loss": 0.3231820583343506,
+ "num_input_tokens_seen": 433248272,
+ "step": 9900,
+ "train_runtime": 76502.2615,
+ "train_tokens_per_second": 5663.209
+ },
+ {
+ "epoch": 0.6669135469970374,
+ "grad_norm": 0.6874890275442226,
+ "learning_rate": 7.498778651350934e-06,
+ "loss": 0.3865694284439087,
+ "num_input_tokens_seen": 433472920,
+ "step": 9905,
+ "train_runtime": 76539.0929,
+ "train_tokens_per_second": 5663.419
+ },
+ {
+ "epoch": 0.6672502019929976,
+ "grad_norm": 0.6837593508656135,
+ "learning_rate": 7.496488086982443e-06,
+ "loss": 0.3751338720321655,
+ "num_input_tokens_seen": 433694440,
+ "step": 9910,
+ "train_runtime": 76582.0454,
+ "train_tokens_per_second": 5663.135
+ },
+ {
+ "epoch": 0.6675868569889577,
+ "grad_norm": 0.834960347653905,
+ "learning_rate": 7.494196824479019e-06,
+ "loss": 0.37354316711425783,
+ "num_input_tokens_seen": 433909064,
+ "step": 9915,
+ "train_runtime": 76619.6521,
+ "train_tokens_per_second": 5663.156
+ },
+ {
+ "epoch": 0.6679235119849178,
+ "grad_norm": 0.6905943083082271,
+ "learning_rate": 7.491904864481406e-06,
+ "loss": 0.3818557024002075,
+ "num_input_tokens_seen": 434122768,
+ "step": 9920,
+ "train_runtime": 76661.7307,
+ "train_tokens_per_second": 5662.835
+ },
+ {
+ "epoch": 0.668260166980878,
+ "grad_norm": 0.6608540738794552,
+ "learning_rate": 7.489612207630543e-06,
+ "loss": 0.34175124168396,
+ "num_input_tokens_seen": 434354528,
+ "step": 9925,
+ "train_runtime": 76695.5171,
+ "train_tokens_per_second": 5663.363
+ },
+ {
+ "epoch": 0.6685968219768381,
+ "grad_norm": 0.747448764956391,
+ "learning_rate": 7.487318854567566e-06,
+ "loss": 0.37988905906677245,
+ "num_input_tokens_seen": 434564096,
+ "step": 9930,
+ "train_runtime": 76729.4177,
+ "train_tokens_per_second": 5663.592
+ },
+ {
+ "epoch": 0.6689334769727983,
+ "grad_norm": 0.7551042289550628,
+ "learning_rate": 7.485024805933803e-06,
+ "loss": 0.3599559307098389,
+ "num_input_tokens_seen": 434767584,
+ "step": 9935,
+ "train_runtime": 76765.8819,
+ "train_tokens_per_second": 5663.552
+ },
+ {
+ "epoch": 0.6692701319687584,
+ "grad_norm": 0.6631253047912142,
+ "learning_rate": 7.482730062370776e-06,
+ "loss": 0.3699958086013794,
+ "num_input_tokens_seen": 434981392,
+ "step": 9940,
+ "train_runtime": 76810.3453,
+ "train_tokens_per_second": 5663.057
+ },
+ {
+ "epoch": 0.6696067869647185,
+ "grad_norm": 0.7419282684235264,
+ "learning_rate": 7.480434624520205e-06,
+ "loss": 0.3593663215637207,
+ "num_input_tokens_seen": 435217648,
+ "step": 9945,
+ "train_runtime": 76843.0552,
+ "train_tokens_per_second": 5663.721
+ },
+ {
+ "epoch": 0.6699434419606787,
+ "grad_norm": 0.7057608866633803,
+ "learning_rate": 7.478138493024001e-06,
+ "loss": 0.3879536151885986,
+ "num_input_tokens_seen": 435444608,
+ "step": 9950,
+ "train_runtime": 76882.5546,
+ "train_tokens_per_second": 5663.764
+ },
+ {
+ "epoch": 0.6702800969566388,
+ "grad_norm": 0.740329239724491,
+ "learning_rate": 7.475841668524268e-06,
+ "loss": 0.34809188842773436,
+ "num_input_tokens_seen": 435656280,
+ "step": 9955,
+ "train_runtime": 76923.9627,
+ "train_tokens_per_second": 5663.466
+ },
+ {
+ "epoch": 0.670616751952599,
+ "grad_norm": 0.8121045332443452,
+ "learning_rate": 7.473544151663309e-06,
+ "loss": 0.4048673152923584,
+ "num_input_tokens_seen": 435875456,
+ "step": 9960,
+ "train_runtime": 76958.8069,
+ "train_tokens_per_second": 5663.75
+ },
+ {
+ "epoch": 0.6709534069485591,
+ "grad_norm": 0.8149107981141157,
+ "learning_rate": 7.471245943083615e-06,
+ "loss": 0.3677277326583862,
+ "num_input_tokens_seen": 436097568,
+ "step": 9965,
+ "train_runtime": 76998.2656,
+ "train_tokens_per_second": 5663.732
+ },
+ {
+ "epoch": 0.6712900619445192,
+ "grad_norm": 0.7104981842358721,
+ "learning_rate": 7.468947043427873e-06,
+ "loss": 0.3423744201660156,
+ "num_input_tokens_seen": 436314872,
+ "step": 9970,
+ "train_runtime": 77032.0267,
+ "train_tokens_per_second": 5664.071
+ },
+ {
+ "epoch": 0.6716267169404794,
+ "grad_norm": 0.8060621663622469,
+ "learning_rate": 7.466647453338965e-06,
+ "loss": 0.38165833950042727,
+ "num_input_tokens_seen": 436524088,
+ "step": 9975,
+ "train_runtime": 77070.1105,
+ "train_tokens_per_second": 5663.987
+ },
+ {
+ "epoch": 0.6719633719364395,
+ "grad_norm": 0.6817255880743095,
+ "learning_rate": 7.464347173459958e-06,
+ "loss": 0.3900636911392212,
+ "num_input_tokens_seen": 436763792,
+ "step": 9980,
+ "train_runtime": 77108.4939,
+ "train_tokens_per_second": 5664.276
+ },
+ {
+ "epoch": 0.6723000269323997,
+ "grad_norm": 0.5849966120883113,
+ "learning_rate": 7.462046204434126e-06,
+ "loss": 0.35174126625061036,
+ "num_input_tokens_seen": 437001592,
+ "step": 9985,
+ "train_runtime": 77143.495,
+ "train_tokens_per_second": 5664.789
+ },
+ {
+ "epoch": 0.6726366819283598,
+ "grad_norm": 0.819711833914903,
+ "learning_rate": 7.459744546904922e-06,
+ "loss": 0.37253947257995607,
+ "num_input_tokens_seen": 437217472,
+ "step": 9990,
+ "train_runtime": 77177.4484,
+ "train_tokens_per_second": 5665.094
+ },
+ {
+ "epoch": 0.6729733369243199,
+ "grad_norm": 0.7103670160362342,
+ "learning_rate": 7.457442201516001e-06,
+ "loss": 0.34635467529296876,
+ "num_input_tokens_seen": 437432256,
+ "step": 9995,
+ "train_runtime": 77222.7561,
+ "train_tokens_per_second": 5664.551
+ },
+ {
+ "epoch": 0.6733099919202801,
+ "grad_norm": 0.7772791959365979,
+ "learning_rate": 7.455139168911203e-06,
+ "loss": 0.3714813232421875,
+ "num_input_tokens_seen": 437632184,
+ "step": 10000,
+ "train_runtime": 77259.692,
+ "train_tokens_per_second": 5664.431
+ },
+ {
+ "epoch": 0.6736466469162402,
+ "grad_norm": 0.8663644820452563,
+ "learning_rate": 7.452835449734564e-06,
+ "loss": 0.37085671424865724,
+ "num_input_tokens_seen": 437852448,
+ "step": 10005,
+ "train_runtime": 77297.557,
+ "train_tokens_per_second": 5664.506
+ },
+ {
+ "epoch": 0.6739833019122003,
+ "grad_norm": 0.6628154115799231,
+ "learning_rate": 7.450531044630315e-06,
+ "loss": 0.35476658344268797,
+ "num_input_tokens_seen": 438072720,
+ "step": 10010,
+ "train_runtime": 77330.8077,
+ "train_tokens_per_second": 5664.918
+ },
+ {
+ "epoch": 0.6743199569081605,
+ "grad_norm": 0.7435766802454172,
+ "learning_rate": 7.448225954242874e-06,
+ "loss": 0.3909281253814697,
+ "num_input_tokens_seen": 438300752,
+ "step": 10015,
+ "train_runtime": 77372.9883,
+ "train_tokens_per_second": 5664.777
+ },
+ {
+ "epoch": 0.6746566119041206,
+ "grad_norm": 0.7076966807661021,
+ "learning_rate": 7.44592017921685e-06,
+ "loss": 0.37700977325439455,
+ "num_input_tokens_seen": 438521952,
+ "step": 10020,
+ "train_runtime": 77408.1985,
+ "train_tokens_per_second": 5665.058
+ },
+ {
+ "epoch": 0.6749932669000808,
+ "grad_norm": 0.7283120289124543,
+ "learning_rate": 7.443613720197048e-06,
+ "loss": 0.39572868347167967,
+ "num_input_tokens_seen": 438742144,
+ "step": 10025,
+ "train_runtime": 77451.6366,
+ "train_tokens_per_second": 5664.724
+ },
+ {
+ "epoch": 0.6753299218960409,
+ "grad_norm": 0.8184114176248545,
+ "learning_rate": 7.44130657782846e-06,
+ "loss": 0.35796802043914794,
+ "num_input_tokens_seen": 438946944,
+ "step": 10030,
+ "train_runtime": 77487.5196,
+ "train_tokens_per_second": 5664.744
+ },
+ {
+ "epoch": 0.675666576892001,
+ "grad_norm": 0.7168914852518904,
+ "learning_rate": 7.438998752756274e-06,
+ "loss": 0.3807486057281494,
+ "num_input_tokens_seen": 439164968,
+ "step": 10035,
+ "train_runtime": 77530.1548,
+ "train_tokens_per_second": 5664.441
+ },
+ {
+ "epoch": 0.6760032318879612,
+ "grad_norm": 0.7003642669524813,
+ "learning_rate": 7.436690245625864e-06,
+ "loss": 0.36953206062316896,
+ "num_input_tokens_seen": 439382032,
+ "step": 10040,
+ "train_runtime": 77568.57,
+ "train_tokens_per_second": 5664.434
+ },
+ {
+ "epoch": 0.6763398868839213,
+ "grad_norm": 0.8332693301417188,
+ "learning_rate": 7.434381057082797e-06,
+ "loss": 0.369028639793396,
+ "num_input_tokens_seen": 439590072,
+ "step": 10045,
+ "train_runtime": 77608.9784,
+ "train_tokens_per_second": 5664.165
+ },
+ {
+ "epoch": 0.6766765418798815,
+ "grad_norm": 0.7659413901414965,
+ "learning_rate": 7.4320711877728305e-06,
+ "loss": 0.3970766067504883,
+ "num_input_tokens_seen": 439814224,
+ "step": 10050,
+ "train_runtime": 77655.5893,
+ "train_tokens_per_second": 5663.652
+ },
+ {
+ "epoch": 0.6770131968758416,
+ "grad_norm": 0.7616766803709787,
+ "learning_rate": 7.429760638341909e-06,
+ "loss": 0.34726481437683104,
+ "num_input_tokens_seen": 440013360,
+ "step": 10055,
+ "train_runtime": 77694.3971,
+ "train_tokens_per_second": 5663.386
+ },
+ {
+ "epoch": 0.6773498518718017,
+ "grad_norm": 0.7214303745868783,
+ "learning_rate": 7.427449409436176e-06,
+ "loss": 0.42118000984191895,
+ "num_input_tokens_seen": 440232216,
+ "step": 10060,
+ "train_runtime": 77735.8693,
+ "train_tokens_per_second": 5663.18
+ },
+ {
+ "epoch": 0.6776865068677619,
+ "grad_norm": 0.6480309395843215,
+ "learning_rate": 7.425137501701955e-06,
+ "loss": 0.35333929061889646,
+ "num_input_tokens_seen": 440465120,
+ "step": 10065,
+ "train_runtime": 77778.1439,
+ "train_tokens_per_second": 5663.096
+ },
+ {
+ "epoch": 0.678023161863722,
+ "grad_norm": 0.8425975984479226,
+ "learning_rate": 7.422824915785765e-06,
+ "loss": 0.3732592582702637,
+ "num_input_tokens_seen": 440684520,
+ "step": 10070,
+ "train_runtime": 77813.6788,
+ "train_tokens_per_second": 5663.33
+ },
+ {
+ "epoch": 0.6783598168596822,
+ "grad_norm": 0.7751732244738783,
+ "learning_rate": 7.420511652334313e-06,
+ "loss": 0.3922509431838989,
+ "num_input_tokens_seen": 440888096,
+ "step": 10075,
+ "train_runtime": 77856.5574,
+ "train_tokens_per_second": 5662.825
+ },
+ {
+ "epoch": 0.6786964718556423,
+ "grad_norm": 0.746917656110605,
+ "learning_rate": 7.418197711994495e-06,
+ "loss": 0.3609792947769165,
+ "num_input_tokens_seen": 441099168,
+ "step": 10080,
+ "train_runtime": 77896.5389,
+ "train_tokens_per_second": 5662.629
+ },
+ {
+ "epoch": 0.6790331268516024,
+ "grad_norm": 0.7979148491415948,
+ "learning_rate": 7.4158830954133984e-06,
+ "loss": 0.37356295585632326,
+ "num_input_tokens_seen": 441329664,
+ "step": 10085,
+ "train_runtime": 77936.3536,
+ "train_tokens_per_second": 5662.693
+ },
+ {
+ "epoch": 0.6793697818475626,
+ "grad_norm": 0.7047169214626602,
+ "learning_rate": 7.413567803238298e-06,
+ "loss": 0.36446294784545896,
+ "num_input_tokens_seen": 441545160,
+ "step": 10090,
+ "train_runtime": 77973.8265,
+ "train_tokens_per_second": 5662.736
+ },
+ {
+ "epoch": 0.6797064368435227,
+ "grad_norm": 0.6385696916431561,
+ "learning_rate": 7.411251836116658e-06,
+ "loss": 0.3654582977294922,
+ "num_input_tokens_seen": 441770600,
+ "step": 10095,
+ "train_runtime": 78021.2974,
+ "train_tokens_per_second": 5662.18
+ },
+ {
+ "epoch": 0.6800430918394829,
+ "grad_norm": 0.675501708680357,
+ "learning_rate": 7.408935194696132e-06,
+ "loss": 0.38489809036254885,
+ "num_input_tokens_seen": 441989608,
+ "step": 10100,
+ "train_runtime": 78062.7748,
+ "train_tokens_per_second": 5661.977
+ },
+ {
+ "epoch": 0.680379746835443,
+ "grad_norm": 0.6976086448861777,
+ "learning_rate": 7.406617879624558e-06,
+ "loss": 0.39519965648651123,
+ "num_input_tokens_seen": 442219464,
+ "step": 10105,
+ "train_runtime": 78104.0438,
+ "train_tokens_per_second": 5661.928
+ },
+ {
+ "epoch": 0.6807164018314031,
+ "grad_norm": 0.6737767286617782,
+ "learning_rate": 7.40429989154997e-06,
+ "loss": 0.35456171035766604,
+ "num_input_tokens_seen": 442434504,
+ "step": 10110,
+ "train_runtime": 78146.9907,
+ "train_tokens_per_second": 5661.568
+ },
+ {
+ "epoch": 0.6810530568273633,
+ "grad_norm": 0.6824592254845565,
+ "learning_rate": 7.4019812311205826e-06,
+ "loss": 0.396740984916687,
+ "num_input_tokens_seen": 442663960,
+ "step": 10115,
+ "train_runtime": 78186.9172,
+ "train_tokens_per_second": 5661.612
+ },
+ {
+ "epoch": 0.6813897118233234,
+ "grad_norm": 0.6995329567504307,
+ "learning_rate": 7.399661898984805e-06,
+ "loss": 0.39820547103881837,
+ "num_input_tokens_seen": 442904544,
+ "step": 10120,
+ "train_runtime": 78228.9374,
+ "train_tokens_per_second": 5661.646
+ },
+ {
+ "epoch": 0.6817263668192836,
+ "grad_norm": 0.7146974067721376,
+ "learning_rate": 7.397341895791228e-06,
+ "loss": 0.3785873889923096,
+ "num_input_tokens_seen": 443119656,
+ "step": 10125,
+ "train_runtime": 78271.2656,
+ "train_tokens_per_second": 5661.332
+ },
+ {
+ "epoch": 0.6820630218152437,
+ "grad_norm": 0.7979315787216683,
+ "learning_rate": 7.395021222188634e-06,
+ "loss": 0.37588326930999755,
+ "num_input_tokens_seen": 443345048,
+ "step": 10130,
+ "train_runtime": 78307.7048,
+ "train_tokens_per_second": 5661.576
+ },
+ {
+ "epoch": 0.6823996768112038,
+ "grad_norm": 0.8569372783831029,
+ "learning_rate": 7.392699878825993e-06,
+ "loss": 0.4026289939880371,
+ "num_input_tokens_seen": 443550416,
+ "step": 10135,
+ "train_runtime": 78348.8182,
+ "train_tokens_per_second": 5661.227
+ },
+ {
+ "epoch": 0.682736331807164,
+ "grad_norm": 0.6799351048003278,
+ "learning_rate": 7.390377866352462e-06,
+ "loss": 0.3739679574966431,
+ "num_input_tokens_seen": 443760728,
+ "step": 10140,
+ "train_runtime": 78391.1203,
+ "train_tokens_per_second": 5660.855
+ },
+ {
+ "epoch": 0.6830729868031241,
+ "grad_norm": 0.7679162845866374,
+ "learning_rate": 7.388055185417382e-06,
+ "loss": 0.36473128795623777,
+ "num_input_tokens_seen": 443986040,
+ "step": 10145,
+ "train_runtime": 78429.781,
+ "train_tokens_per_second": 5660.937
+ },
+ {
+ "epoch": 0.6834096417990843,
+ "grad_norm": 0.7045812738235366,
+ "learning_rate": 7.3857318366702844e-06,
+ "loss": 0.35809712409973143,
+ "num_input_tokens_seen": 444206480,
+ "step": 10150,
+ "train_runtime": 78472.8507,
+ "train_tokens_per_second": 5660.639
+ },
+ {
+ "epoch": 0.6837462967950444,
+ "grad_norm": 0.7004993125342208,
+ "learning_rate": 7.383407820760885e-06,
+ "loss": 0.3713981151580811,
+ "num_input_tokens_seen": 444432096,
+ "step": 10155,
+ "train_runtime": 78515.6453,
+ "train_tokens_per_second": 5660.427
+ },
+ {
+ "epoch": 0.6840829517910045,
+ "grad_norm": 0.7774580479484972,
+ "learning_rate": 7.381083138339089e-06,
+ "loss": 0.38779325485229493,
+ "num_input_tokens_seen": 444643008,
+ "step": 10160,
+ "train_runtime": 78555.8979,
+ "train_tokens_per_second": 5660.212
+ },
+ {
+ "epoch": 0.6844196067869647,
+ "grad_norm": 0.7163976333149118,
+ "learning_rate": 7.378757790054988e-06,
+ "loss": 0.39658608436584475,
+ "num_input_tokens_seen": 444875528,
+ "step": 10165,
+ "train_runtime": 78592.4434,
+ "train_tokens_per_second": 5660.538
+ },
+ {
+ "epoch": 0.6847562617829248,
+ "grad_norm": 0.7470536381083663,
+ "learning_rate": 7.376431776558855e-06,
+ "loss": 0.3956130504608154,
+ "num_input_tokens_seen": 445097312,
+ "step": 10170,
+ "train_runtime": 78630.7483,
+ "train_tokens_per_second": 5660.601
+ },
+ {
+ "epoch": 0.685092916778885,
+ "grad_norm": 0.7408806601027974,
+ "learning_rate": 7.3741050985011546e-06,
+ "loss": 0.3585235118865967,
+ "num_input_tokens_seen": 445319152,
+ "step": 10175,
+ "train_runtime": 78670.4341,
+ "train_tokens_per_second": 5660.566
+ },
+ {
+ "epoch": 0.6854295717748451,
+ "grad_norm": 0.7086975222942481,
+ "learning_rate": 7.371777756532531e-06,
+ "loss": 0.34885153770446775,
+ "num_input_tokens_seen": 445544848,
+ "step": 10180,
+ "train_runtime": 78704.6359,
+ "train_tokens_per_second": 5660.973
+ },
+ {
+ "epoch": 0.6857662267708052,
+ "grad_norm": 0.7956956482990667,
+ "learning_rate": 7.369449751303823e-06,
+ "loss": 0.3933322191238403,
+ "num_input_tokens_seen": 445764768,
+ "step": 10185,
+ "train_runtime": 78741.4834,
+ "train_tokens_per_second": 5661.117
+ },
+ {
+ "epoch": 0.6861028817667654,
+ "grad_norm": 0.5998016401864463,
+ "learning_rate": 7.367121083466048e-06,
+ "loss": 0.3657109498977661,
+ "num_input_tokens_seen": 445988536,
+ "step": 10190,
+ "train_runtime": 78776.8049,
+ "train_tokens_per_second": 5661.419
+ },
+ {
+ "epoch": 0.6864395367627255,
+ "grad_norm": 0.8579022323586973,
+ "learning_rate": 7.364791753670409e-06,
+ "loss": 0.3847865104675293,
+ "num_input_tokens_seen": 446213600,
+ "step": 10195,
+ "train_runtime": 78810.0378,
+ "train_tokens_per_second": 5661.888
+ },
+ {
+ "epoch": 0.6867761917586857,
+ "grad_norm": 0.729478615131132,
+ "learning_rate": 7.362461762568298e-06,
+ "loss": 0.372448205947876,
+ "num_input_tokens_seen": 446432912,
+ "step": 10200,
+ "train_runtime": 78848.3501,
+ "train_tokens_per_second": 5661.918
+ },
+ {
+ "epoch": 0.6871128467546458,
+ "grad_norm": 0.67715047966365,
+ "learning_rate": 7.360131110811287e-06,
+ "loss": 0.3532183408737183,
+ "num_input_tokens_seen": 446656416,
+ "step": 10205,
+ "train_runtime": 78888.3323,
+ "train_tokens_per_second": 5661.882
+ },
+ {
+ "epoch": 0.6874495017506059,
+ "grad_norm": 0.7095724500319497,
+ "learning_rate": 7.357799799051138e-06,
+ "loss": 0.3632460355758667,
+ "num_input_tokens_seen": 446871712,
+ "step": 10210,
+ "train_runtime": 78927.4603,
+ "train_tokens_per_second": 5661.803
+ },
+ {
+ "epoch": 0.6877861567465661,
+ "grad_norm": 0.9476488222825441,
+ "learning_rate": 7.355467827939794e-06,
+ "loss": 0.38349783420562744,
+ "num_input_tokens_seen": 447082192,
+ "step": 10215,
+ "train_runtime": 78967.9316,
+ "train_tokens_per_second": 5661.566
+ },
+ {
+ "epoch": 0.6881228117425262,
+ "grad_norm": 0.7980347922355381,
+ "learning_rate": 7.353135198129382e-06,
+ "loss": 0.3912900686264038,
+ "num_input_tokens_seen": 447291168,
+ "step": 10220,
+ "train_runtime": 79002.7871,
+ "train_tokens_per_second": 5661.714
+ },
+ {
+ "epoch": 0.6884594667384863,
+ "grad_norm": 0.796413758238857,
+ "learning_rate": 7.350801910272216e-06,
+ "loss": 0.38165950775146484,
+ "num_input_tokens_seen": 447512392,
+ "step": 10225,
+ "train_runtime": 79046.6142,
+ "train_tokens_per_second": 5661.373
+ },
+ {
+ "epoch": 0.6887961217344465,
+ "grad_norm": 0.7746118196960639,
+ "learning_rate": 7.348467965020791e-06,
+ "loss": 0.34427103996276853,
+ "num_input_tokens_seen": 447731304,
+ "step": 10230,
+ "train_runtime": 79082.2762,
+ "train_tokens_per_second": 5661.588
+ },
+ {
+ "epoch": 0.6891327767304066,
+ "grad_norm": 0.8166556060856294,
+ "learning_rate": 7.3461333630277885e-06,
+ "loss": 0.3613192081451416,
+ "num_input_tokens_seen": 447922416,
+ "step": 10235,
+ "train_runtime": 79117.4476,
+ "train_tokens_per_second": 5661.487
+ },
+ {
+ "epoch": 0.6894694317263668,
+ "grad_norm": 0.6976640506234788,
+ "learning_rate": 7.343798104946074e-06,
+ "loss": 0.3512859344482422,
+ "num_input_tokens_seen": 448148520,
+ "step": 10240,
+ "train_runtime": 79156.1315,
+ "train_tokens_per_second": 5661.577
+ },
+ {
+ "epoch": 0.6898060867223269,
+ "grad_norm": 0.74195160383717,
+ "learning_rate": 7.341462191428692e-06,
+ "loss": 0.381636905670166,
+ "num_input_tokens_seen": 448364664,
+ "step": 10245,
+ "train_runtime": 79199.1829,
+ "train_tokens_per_second": 5661.228
+ },
+ {
+ "epoch": 0.690142741718287,
+ "grad_norm": 0.7527010867181296,
+ "learning_rate": 7.339125623128874e-06,
+ "loss": 0.3913971662521362,
+ "num_input_tokens_seen": 448587016,
+ "step": 10250,
+ "train_runtime": 79248.9634,
+ "train_tokens_per_second": 5660.478
+ },
+ {
+ "epoch": 0.6904793967142472,
+ "grad_norm": 0.7210165301469443,
+ "learning_rate": 7.336788400700035e-06,
+ "loss": 0.36610395908355714,
+ "num_input_tokens_seen": 448798728,
+ "step": 10255,
+ "train_runtime": 79292.6113,
+ "train_tokens_per_second": 5660.032
+ },
+ {
+ "epoch": 0.6908160517102074,
+ "grad_norm": 0.9096784607094398,
+ "learning_rate": 7.334450524795771e-06,
+ "loss": 0.408566951751709,
+ "num_input_tokens_seen": 449023728,
+ "step": 10260,
+ "train_runtime": 79333.8153,
+ "train_tokens_per_second": 5659.929
+ },
+ {
+ "epoch": 0.6911527067061676,
+ "grad_norm": 0.6784070074683265,
+ "learning_rate": 7.332111996069863e-06,
+ "loss": 0.3746421575546265,
+ "num_input_tokens_seen": 449245560,
+ "step": 10265,
+ "train_runtime": 79370.4334,
+ "train_tokens_per_second": 5660.112
+ },
+ {
+ "epoch": 0.6914893617021277,
+ "grad_norm": 0.8780985630278836,
+ "learning_rate": 7.32977281517627e-06,
+ "loss": 0.39997851848602295,
+ "num_input_tokens_seen": 449446632,
+ "step": 10270,
+ "train_runtime": 79406.7923,
+ "train_tokens_per_second": 5660.053
+ },
+ {
+ "epoch": 0.6918260166980879,
+ "grad_norm": 0.7238138438393178,
+ "learning_rate": 7.3274329827691406e-06,
+ "loss": 0.3859497308731079,
+ "num_input_tokens_seen": 449659944,
+ "step": 10275,
+ "train_runtime": 79449.7577,
+ "train_tokens_per_second": 5659.677
+ },
+ {
+ "epoch": 0.692162671694048,
+ "grad_norm": 0.6417831874350313,
+ "learning_rate": 7.325092499502796e-06,
+ "loss": 0.35598082542419435,
+ "num_input_tokens_seen": 449892656,
+ "step": 10280,
+ "train_runtime": 79489.514,
+ "train_tokens_per_second": 5659.774
+ },
+ {
+ "epoch": 0.6924993266900081,
+ "grad_norm": 0.6886975793380838,
+ "learning_rate": 7.322751366031751e-06,
+ "loss": 0.3739547967910767,
+ "num_input_tokens_seen": 450115160,
+ "step": 10285,
+ "train_runtime": 79528.7582,
+ "train_tokens_per_second": 5659.779
+ },
+ {
+ "epoch": 0.6928359816859683,
+ "grad_norm": 0.8067129855431404,
+ "learning_rate": 7.320409583010691e-06,
+ "loss": 0.3593981027603149,
+ "num_input_tokens_seen": 450339952,
+ "step": 10290,
+ "train_runtime": 79567.7249,
+ "train_tokens_per_second": 5659.832
+ },
+ {
+ "epoch": 0.6931726366819284,
+ "grad_norm": 0.720222949795357,
+ "learning_rate": 7.318067151094493e-06,
+ "loss": 0.4208518981933594,
+ "num_input_tokens_seen": 450562568,
+ "step": 10295,
+ "train_runtime": 79603.3438,
+ "train_tokens_per_second": 5660.096
+ },
+ {
+ "epoch": 0.6935092916778886,
+ "grad_norm": 0.6662609649687479,
+ "learning_rate": 7.3157240709382046e-06,
+ "loss": 0.3863611936569214,
+ "num_input_tokens_seen": 450794024,
+ "step": 10300,
+ "train_runtime": 79634.484,
+ "train_tokens_per_second": 5660.789
+ },
+ {
+ "epoch": 0.6938459466738487,
+ "grad_norm": 0.7343868155686462,
+ "learning_rate": 7.313380343197066e-06,
+ "loss": 0.3639280080795288,
+ "num_input_tokens_seen": 451027856,
+ "step": 10305,
+ "train_runtime": 79674.0903,
+ "train_tokens_per_second": 5660.91
+ },
+ {
+ "epoch": 0.6941826016698088,
+ "grad_norm": 0.7926474138155267,
+ "learning_rate": 7.311035968526492e-06,
+ "loss": 0.3794968605041504,
+ "num_input_tokens_seen": 451247432,
+ "step": 10310,
+ "train_runtime": 79715.3108,
+ "train_tokens_per_second": 5660.737
+ },
+ {
+ "epoch": 0.694519256665769,
+ "grad_norm": 0.8428213530948624,
+ "learning_rate": 7.308690947582078e-06,
+ "loss": 0.38078649044036866,
+ "num_input_tokens_seen": 451456696,
+ "step": 10315,
+ "train_runtime": 79750.434,
+ "train_tokens_per_second": 5660.868
+ },
+ {
+ "epoch": 0.6948559116617291,
+ "grad_norm": 0.7185545573840266,
+ "learning_rate": 7.306345281019602e-06,
+ "loss": 0.37413511276245115,
+ "num_input_tokens_seen": 451677704,
+ "step": 10320,
+ "train_runtime": 79790.1325,
+ "train_tokens_per_second": 5660.822
+ },
+ {
+ "epoch": 0.6951925666576892,
+ "grad_norm": 0.7678020136994413,
+ "learning_rate": 7.303998969495023e-06,
+ "loss": 0.40517454147338866,
+ "num_input_tokens_seen": 451892384,
+ "step": 10325,
+ "train_runtime": 79832.0377,
+ "train_tokens_per_second": 5660.539
+ },
+ {
+ "epoch": 0.6955292216536494,
+ "grad_norm": 0.7116247415151229,
+ "learning_rate": 7.301652013664479e-06,
+ "loss": 0.39863314628601076,
+ "num_input_tokens_seen": 452118752,
+ "step": 10330,
+ "train_runtime": 79868.7262,
+ "train_tokens_per_second": 5660.773
+ },
+ {
+ "epoch": 0.6958658766496095,
+ "grad_norm": 0.7656084452706404,
+ "learning_rate": 7.299304414184288e-06,
+ "loss": 0.40618147850036623,
+ "num_input_tokens_seen": 452338552,
+ "step": 10335,
+ "train_runtime": 79909.7465,
+ "train_tokens_per_second": 5660.618
+ },
+ {
+ "epoch": 0.6962025316455697,
+ "grad_norm": 0.6668364749984544,
+ "learning_rate": 7.296956171710951e-06,
+ "loss": 0.3879892587661743,
+ "num_input_tokens_seen": 452573592,
+ "step": 10340,
+ "train_runtime": 79946.2615,
+ "train_tokens_per_second": 5660.973
+ },
+ {
+ "epoch": 0.6965391866415298,
+ "grad_norm": 0.8038854323996706,
+ "learning_rate": 7.294607286901146e-06,
+ "loss": 0.3926121711730957,
+ "num_input_tokens_seen": 452795696,
+ "step": 10345,
+ "train_runtime": 79987.6707,
+ "train_tokens_per_second": 5660.819
+ },
+ {
+ "epoch": 0.69687584163749,
+ "grad_norm": 0.7769622893093939,
+ "learning_rate": 7.292257760411726e-06,
+ "loss": 0.3886493444442749,
+ "num_input_tokens_seen": 453009328,
+ "step": 10350,
+ "train_runtime": 80023.6895,
+ "train_tokens_per_second": 5660.94
+ },
+ {
+ "epoch": 0.6972124966334501,
+ "grad_norm": 0.7478012896643832,
+ "learning_rate": 7.289907592899737e-06,
+ "loss": 0.37845487594604493,
+ "num_input_tokens_seen": 453214464,
+ "step": 10355,
+ "train_runtime": 80062.9601,
+ "train_tokens_per_second": 5660.726
+ },
+ {
+ "epoch": 0.6975491516294102,
+ "grad_norm": 0.7047299342802962,
+ "learning_rate": 7.287556785022389e-06,
+ "loss": 0.3894767761230469,
+ "num_input_tokens_seen": 453437368,
+ "step": 10360,
+ "train_runtime": 80102.9256,
+ "train_tokens_per_second": 5660.684
+ },
+ {
+ "epoch": 0.6978858066253704,
+ "grad_norm": 0.7604979984461925,
+ "learning_rate": 7.285205337437082e-06,
+ "loss": 0.3892651557922363,
+ "num_input_tokens_seen": 453647680,
+ "step": 10365,
+ "train_runtime": 80141.1595,
+ "train_tokens_per_second": 5660.608
+ },
+ {
+ "epoch": 0.6982224616213305,
+ "grad_norm": 0.7970440490683166,
+ "learning_rate": 7.2828532508013895e-06,
+ "loss": 0.399712347984314,
+ "num_input_tokens_seen": 453867592,
+ "step": 10370,
+ "train_runtime": 80178.1608,
+ "train_tokens_per_second": 5660.738
+ },
+ {
+ "epoch": 0.6985591166172906,
+ "grad_norm": 0.7584286596844286,
+ "learning_rate": 7.2805005257730625e-06,
+ "loss": 0.3772284030914307,
+ "num_input_tokens_seen": 454087008,
+ "step": 10375,
+ "train_runtime": 80211.2544,
+ "train_tokens_per_second": 5661.138
+ },
+ {
+ "epoch": 0.6988957716132508,
+ "grad_norm": 0.6810825697097003,
+ "learning_rate": 7.278147163010037e-06,
+ "loss": 0.36341052055358886,
+ "num_input_tokens_seen": 454313224,
+ "step": 10380,
+ "train_runtime": 80250.7125,
+ "train_tokens_per_second": 5661.174
+ },
+ {
+ "epoch": 0.6992324266092109,
+ "grad_norm": 0.7849426670217893,
+ "learning_rate": 7.275793163170423e-06,
+ "loss": 0.36755235195159913,
+ "num_input_tokens_seen": 454539992,
+ "step": 10385,
+ "train_runtime": 80288.0816,
+ "train_tokens_per_second": 5661.363
+ },
+ {
+ "epoch": 0.6995690816051711,
+ "grad_norm": 0.7022327990344984,
+ "learning_rate": 7.273438526912506e-06,
+ "loss": 0.3956979990005493,
+ "num_input_tokens_seen": 454748824,
+ "step": 10390,
+ "train_runtime": 80325.0195,
+ "train_tokens_per_second": 5661.36
+ },
+ {
+ "epoch": 0.6999057366011312,
+ "grad_norm": 0.7244301627454164,
+ "learning_rate": 7.271083254894757e-06,
+ "loss": 0.36125593185424804,
+ "num_input_tokens_seen": 454962456,
+ "step": 10395,
+ "train_runtime": 80364.6433,
+ "train_tokens_per_second": 5661.227
+ },
+ {
+ "epoch": 0.7002423915970913,
+ "grad_norm": 0.7856991901608305,
+ "learning_rate": 7.268727347775815e-06,
+ "loss": 0.3766735076904297,
+ "num_input_tokens_seen": 455189544,
+ "step": 10400,
+ "train_runtime": 80398.3718,
+ "train_tokens_per_second": 5661.676
+ },
+ {
+ "epoch": 0.7005790465930515,
+ "grad_norm": 0.7076340782949039,
+ "learning_rate": 7.266370806214506e-06,
+ "loss": 0.32742795944213865,
+ "num_input_tokens_seen": 455409928,
+ "step": 10405,
+ "train_runtime": 80446.5155,
+ "train_tokens_per_second": 5661.027
+ },
+ {
+ "epoch": 0.7009157015890116,
+ "grad_norm": 0.6900523877236565,
+ "learning_rate": 7.264013630869827e-06,
+ "loss": 0.37685844898223875,
+ "num_input_tokens_seen": 455626104,
+ "step": 10410,
+ "train_runtime": 80491.2258,
+ "train_tokens_per_second": 5660.569
+ },
+ {
+ "epoch": 0.7012523565849718,
+ "grad_norm": 0.6710950318779684,
+ "learning_rate": 7.261655822400957e-06,
+ "loss": 0.39275636672973635,
+ "num_input_tokens_seen": 455857232,
+ "step": 10415,
+ "train_runtime": 80526.04,
+ "train_tokens_per_second": 5660.992
+ },
+ {
+ "epoch": 0.7015890115809319,
+ "grad_norm": 0.7198657361358307,
+ "learning_rate": 7.259297381467248e-06,
+ "loss": 0.38013458251953125,
+ "num_input_tokens_seen": 456077832,
+ "step": 10420,
+ "train_runtime": 80569.4365,
+ "train_tokens_per_second": 5660.68
+ },
+ {
+ "epoch": 0.701925666576892,
+ "grad_norm": 0.7005819646574629,
+ "learning_rate": 7.2569383087282285e-06,
+ "loss": 0.34854826927185056,
+ "num_input_tokens_seen": 456272960,
+ "step": 10425,
+ "train_runtime": 80611.1139,
+ "train_tokens_per_second": 5660.174
+ },
+ {
+ "epoch": 0.7022623215728522,
+ "grad_norm": 0.7072800232396435,
+ "learning_rate": 7.25457860484361e-06,
+ "loss": 0.39771556854248047,
+ "num_input_tokens_seen": 456493600,
+ "step": 10430,
+ "train_runtime": 80651.8402,
+ "train_tokens_per_second": 5660.052
+ },
+ {
+ "epoch": 0.7025989765688123,
+ "grad_norm": 0.7445696785976343,
+ "learning_rate": 7.252218270473274e-06,
+ "loss": 0.3760495662689209,
+ "num_input_tokens_seen": 456704912,
+ "step": 10435,
+ "train_runtime": 80692.0569,
+ "train_tokens_per_second": 5659.85
+ },
+ {
+ "epoch": 0.7029356315647725,
+ "grad_norm": 0.7457222056984448,
+ "learning_rate": 7.249857306277279e-06,
+ "loss": 0.369310188293457,
+ "num_input_tokens_seen": 456921864,
+ "step": 10440,
+ "train_runtime": 80726.6828,
+ "train_tokens_per_second": 5660.109
+ },
+ {
+ "epoch": 0.7032722865607326,
+ "grad_norm": 0.9221604408221112,
+ "learning_rate": 7.247495712915864e-06,
+ "loss": 0.35996944904327394,
+ "num_input_tokens_seen": 457143032,
+ "step": 10445,
+ "train_runtime": 80760.6102,
+ "train_tokens_per_second": 5660.47
+ },
+ {
+ "epoch": 0.7036089415566927,
+ "grad_norm": 0.6862985079879047,
+ "learning_rate": 7.245133491049439e-06,
+ "loss": 0.3736717939376831,
+ "num_input_tokens_seen": 457363400,
+ "step": 10450,
+ "train_runtime": 80799.8061,
+ "train_tokens_per_second": 5660.452
+ },
+ {
+ "epoch": 0.7039455965526529,
+ "grad_norm": 0.6935497313758425,
+ "learning_rate": 7.242770641338592e-06,
+ "loss": 0.4129630088806152,
+ "num_input_tokens_seen": 457584008,
+ "step": 10455,
+ "train_runtime": 80841.1453,
+ "train_tokens_per_second": 5660.286
+ },
+ {
+ "epoch": 0.704282251548613,
+ "grad_norm": 0.7202390162158359,
+ "learning_rate": 7.240407164444088e-06,
+ "loss": 0.36982312202453616,
+ "num_input_tokens_seen": 457798528,
+ "step": 10460,
+ "train_runtime": 80879.602,
+ "train_tokens_per_second": 5660.247
+ },
+ {
+ "epoch": 0.7046189065445732,
+ "grad_norm": 0.7863736874260931,
+ "learning_rate": 7.238043061026862e-06,
+ "loss": 0.34938392639160154,
+ "num_input_tokens_seen": 458006080,
+ "step": 10465,
+ "train_runtime": 80921.1996,
+ "train_tokens_per_second": 5659.902
+ },
+ {
+ "epoch": 0.7049555615405333,
+ "grad_norm": 0.7013473236004093,
+ "learning_rate": 7.235678331748034e-06,
+ "loss": 0.3546873092651367,
+ "num_input_tokens_seen": 458203208,
+ "step": 10470,
+ "train_runtime": 80956.6805,
+ "train_tokens_per_second": 5659.857
+ },
+ {
+ "epoch": 0.7052922165364934,
+ "grad_norm": 0.7707435582118967,
+ "learning_rate": 7.233312977268887e-06,
+ "loss": 0.40183773040771487,
+ "num_input_tokens_seen": 458427952,
+ "step": 10475,
+ "train_runtime": 80993.4895,
+ "train_tokens_per_second": 5660.059
+ },
+ {
+ "epoch": 0.7056288715324536,
+ "grad_norm": 0.7366337541289394,
+ "learning_rate": 7.230946998250889e-06,
+ "loss": 0.36978516578674314,
+ "num_input_tokens_seen": 458631504,
+ "step": 10480,
+ "train_runtime": 81028.6939,
+ "train_tokens_per_second": 5660.112
+ },
+ {
+ "epoch": 0.7059655265284137,
+ "grad_norm": 0.7121052684917011,
+ "learning_rate": 7.228580395355676e-06,
+ "loss": 0.34577772617340086,
+ "num_input_tokens_seen": 458845056,
+ "step": 10485,
+ "train_runtime": 81072.1146,
+ "train_tokens_per_second": 5659.715
+ },
+ {
+ "epoch": 0.7063021815243739,
+ "grad_norm": 0.7741334918584597,
+ "learning_rate": 7.226213169245062e-06,
+ "loss": 0.38037862777709963,
+ "num_input_tokens_seen": 459063872,
+ "step": 10490,
+ "train_runtime": 81113.0274,
+ "train_tokens_per_second": 5659.558
+ },
+ {
+ "epoch": 0.706638836520334,
+ "grad_norm": 0.7871663193272351,
+ "learning_rate": 7.223845320581035e-06,
+ "loss": 0.3649828672409058,
+ "num_input_tokens_seen": 459276768,
+ "step": 10495,
+ "train_runtime": 81158.0461,
+ "train_tokens_per_second": 5659.042
+ },
+ {
+ "epoch": 0.7069754915162941,
+ "grad_norm": 0.8550643751341503,
+ "learning_rate": 7.221476850025754e-06,
+ "loss": 0.3691579818725586,
+ "num_input_tokens_seen": 459504320,
+ "step": 10500,
+ "train_runtime": 81197.4298,
+ "train_tokens_per_second": 5659.099
+ },
+ {
+ "epoch": 0.7073121465122543,
+ "grad_norm": 0.6802216728098077,
+ "learning_rate": 7.219107758241557e-06,
+ "loss": 0.381423807144165,
+ "num_input_tokens_seen": 459726544,
+ "step": 10505,
+ "train_runtime": 81239.3131,
+ "train_tokens_per_second": 5658.917
+ },
+ {
+ "epoch": 0.7076488015082144,
+ "grad_norm": 0.7476015422415041,
+ "learning_rate": 7.2167380458909505e-06,
+ "loss": 0.3745572090148926,
+ "num_input_tokens_seen": 459949656,
+ "step": 10510,
+ "train_runtime": 81274.379,
+ "train_tokens_per_second": 5659.221
+ },
+ {
+ "epoch": 0.7079854565041745,
+ "grad_norm": 0.6742333237089346,
+ "learning_rate": 7.214367713636619e-06,
+ "loss": 0.375703763961792,
+ "num_input_tokens_seen": 460172128,
+ "step": 10515,
+ "train_runtime": 81312.8461,
+ "train_tokens_per_second": 5659.28
+ },
+ {
+ "epoch": 0.7083221115001347,
+ "grad_norm": 0.6954522791161296,
+ "learning_rate": 7.211996762141418e-06,
+ "loss": 0.37557039260864256,
+ "num_input_tokens_seen": 460390848,
+ "step": 10520,
+ "train_runtime": 81345.9866,
+ "train_tokens_per_second": 5659.663
+ },
+ {
+ "epoch": 0.7086587664960948,
+ "grad_norm": 0.7339967212433919,
+ "learning_rate": 7.209625192068374e-06,
+ "loss": 0.4028931617736816,
+ "num_input_tokens_seen": 460611832,
+ "step": 10525,
+ "train_runtime": 81379.8389,
+ "train_tokens_per_second": 5660.024
+ },
+ {
+ "epoch": 0.708995421492055,
+ "grad_norm": 0.701987120656059,
+ "learning_rate": 7.207253004080692e-06,
+ "loss": 0.37696285247802735,
+ "num_input_tokens_seen": 460844344,
+ "step": 10530,
+ "train_runtime": 81418.5196,
+ "train_tokens_per_second": 5660.191
+ },
+ {
+ "epoch": 0.7093320764880151,
+ "grad_norm": 0.7755594445584074,
+ "learning_rate": 7.204880198841746e-06,
+ "loss": 0.34694790840148926,
+ "num_input_tokens_seen": 461058504,
+ "step": 10535,
+ "train_runtime": 81453.7678,
+ "train_tokens_per_second": 5660.37
+ },
+ {
+ "epoch": 0.7096687314839752,
+ "grad_norm": 0.8426741622679441,
+ "learning_rate": 7.2025067770150835e-06,
+ "loss": 0.4015054225921631,
+ "num_input_tokens_seen": 461286472,
+ "step": 10540,
+ "train_runtime": 81493.9088,
+ "train_tokens_per_second": 5660.38
+ },
+ {
+ "epoch": 0.7100053864799354,
+ "grad_norm": 0.7557428415974053,
+ "learning_rate": 7.2001327392644235e-06,
+ "loss": 0.3497929573059082,
+ "num_input_tokens_seen": 461515496,
+ "step": 10545,
+ "train_runtime": 81528.2463,
+ "train_tokens_per_second": 5660.805
+ },
+ {
+ "epoch": 0.7103420414758955,
+ "grad_norm": 0.7845759323690452,
+ "learning_rate": 7.197758086253659e-06,
+ "loss": 0.39971365928649905,
+ "num_input_tokens_seen": 461738504,
+ "step": 10550,
+ "train_runtime": 81568.9369,
+ "train_tokens_per_second": 5660.715
+ },
+ {
+ "epoch": 0.7106786964718557,
+ "grad_norm": 0.7112358056324952,
+ "learning_rate": 7.195382818646854e-06,
+ "loss": 0.39182677268981936,
+ "num_input_tokens_seen": 461947016,
+ "step": 10555,
+ "train_runtime": 81608.3983,
+ "train_tokens_per_second": 5660.533
+ },
+ {
+ "epoch": 0.7110153514678158,
+ "grad_norm": 0.772603293296723,
+ "learning_rate": 7.1930069371082445e-06,
+ "loss": 0.36522927284240725,
+ "num_input_tokens_seen": 462152784,
+ "step": 10560,
+ "train_runtime": 81652.6367,
+ "train_tokens_per_second": 5659.986
+ },
+ {
+ "epoch": 0.7113520064637759,
+ "grad_norm": 0.7156871566744746,
+ "learning_rate": 7.190630442302238e-06,
+ "loss": 0.3833275079727173,
+ "num_input_tokens_seen": 462379128,
+ "step": 10565,
+ "train_runtime": 81694.2563,
+ "train_tokens_per_second": 5659.873
+ },
+ {
+ "epoch": 0.7116886614597361,
+ "grad_norm": 0.7033879162554953,
+ "learning_rate": 7.188253334893414e-06,
+ "loss": 0.3999313831329346,
+ "num_input_tokens_seen": 462598648,
+ "step": 10570,
+ "train_runtime": 81730.3714,
+ "train_tokens_per_second": 5660.058
+ },
+ {
+ "epoch": 0.7120253164556962,
+ "grad_norm": 0.8834420627247958,
+ "learning_rate": 7.18587561554652e-06,
+ "loss": 0.3910887956619263,
+ "num_input_tokens_seen": 462807208,
+ "step": 10575,
+ "train_runtime": 81771.4349,
+ "train_tokens_per_second": 5659.766
+ },
+ {
+ "epoch": 0.7123619714516564,
+ "grad_norm": 0.6918016219416666,
+ "learning_rate": 7.183497284926483e-06,
+ "loss": 0.36777448654174805,
+ "num_input_tokens_seen": 463042928,
+ "step": 10580,
+ "train_runtime": 81807.4141,
+ "train_tokens_per_second": 5660.158
+ },
+ {
+ "epoch": 0.7126986264476165,
+ "grad_norm": 0.7599277759479904,
+ "learning_rate": 7.181118343698392e-06,
+ "loss": 0.36260690689086916,
+ "num_input_tokens_seen": 463252424,
+ "step": 10585,
+ "train_runtime": 81851.2205,
+ "train_tokens_per_second": 5659.689
+ },
+ {
+ "epoch": 0.7130352814435766,
+ "grad_norm": 0.7224431307545425,
+ "learning_rate": 7.178738792527512e-06,
+ "loss": 0.39423255920410155,
+ "num_input_tokens_seen": 463469984,
+ "step": 10590,
+ "train_runtime": 81891.039,
+ "train_tokens_per_second": 5659.593
+ },
+ {
+ "epoch": 0.7133719364395368,
+ "grad_norm": 0.7536526351492208,
+ "learning_rate": 7.176358632079277e-06,
+ "loss": 0.37251434326171873,
+ "num_input_tokens_seen": 463682208,
+ "step": 10595,
+ "train_runtime": 81929.5391,
+ "train_tokens_per_second": 5659.524
+ },
+ {
+ "epoch": 0.7137085914354969,
+ "grad_norm": 0.6793493050362683,
+ "learning_rate": 7.173977863019288e-06,
+ "loss": 0.37181549072265624,
+ "num_input_tokens_seen": 463907464,
+ "step": 10600,
+ "train_runtime": 81970.1484,
+ "train_tokens_per_second": 5659.468
+ },
+ {
+ "epoch": 0.7140452464314571,
+ "grad_norm": 0.6822436534253042,
+ "learning_rate": 7.171596486013323e-06,
+ "loss": 0.34790878295898436,
+ "num_input_tokens_seen": 464138256,
+ "step": 10605,
+ "train_runtime": 82007.4481,
+ "train_tokens_per_second": 5659.709
+ },
+ {
+ "epoch": 0.7143819014274172,
+ "grad_norm": 0.8051213804159209,
+ "learning_rate": 7.169214501727326e-06,
+ "loss": 0.37795867919921877,
+ "num_input_tokens_seen": 464338936,
+ "step": 10610,
+ "train_runtime": 82043.2488,
+ "train_tokens_per_second": 5659.685
+ },
+ {
+ "epoch": 0.7147185564233773,
+ "grad_norm": 0.686749483511892,
+ "learning_rate": 7.16683191082741e-06,
+ "loss": 0.36943345069885253,
+ "num_input_tokens_seen": 464560144,
+ "step": 10615,
+ "train_runtime": 82079.0859,
+ "train_tokens_per_second": 5659.909
+ },
+ {
+ "epoch": 0.7150552114193375,
+ "grad_norm": 0.8746419009334871,
+ "learning_rate": 7.16444871397986e-06,
+ "loss": 0.368468713760376,
+ "num_input_tokens_seen": 464765496,
+ "step": 10620,
+ "train_runtime": 82117.3295,
+ "train_tokens_per_second": 5659.774
+ },
+ {
+ "epoch": 0.7153918664152976,
+ "grad_norm": 1.042013296164164,
+ "learning_rate": 7.162064911851129e-06,
+ "loss": 0.36743860244750975,
+ "num_input_tokens_seen": 464978408,
+ "step": 10625,
+ "train_runtime": 82160.6887,
+ "train_tokens_per_second": 5659.378
+ },
+ {
+ "epoch": 0.7157285214112578,
+ "grad_norm": 0.7321276232893942,
+ "learning_rate": 7.15968050510784e-06,
+ "loss": 0.40234060287475587,
+ "num_input_tokens_seen": 465192704,
+ "step": 10630,
+ "train_runtime": 82198.1606,
+ "train_tokens_per_second": 5659.405
+ },
+ {
+ "epoch": 0.7160651764072179,
+ "grad_norm": 0.8428920187826066,
+ "learning_rate": 7.1572954944167834e-06,
+ "loss": 0.3755232810974121,
+ "num_input_tokens_seen": 465408616,
+ "step": 10635,
+ "train_runtime": 82239.586,
+ "train_tokens_per_second": 5659.18
+ },
+ {
+ "epoch": 0.716401831403178,
+ "grad_norm": 0.679996578354662,
+ "learning_rate": 7.154909880444922e-06,
+ "loss": 0.3969583988189697,
+ "num_input_tokens_seen": 465633232,
+ "step": 10640,
+ "train_runtime": 82282.7769,
+ "train_tokens_per_second": 5658.939
+ },
+ {
+ "epoch": 0.7167384863991382,
+ "grad_norm": 1.5740350922344166,
+ "learning_rate": 7.152523663859384e-06,
+ "loss": 0.37885403633117676,
+ "num_input_tokens_seen": 465845256,
+ "step": 10645,
+ "train_runtime": 82324.9766,
+ "train_tokens_per_second": 5658.614
+ },
+ {
+ "epoch": 0.7170751413950983,
+ "grad_norm": 0.8328362306471117,
+ "learning_rate": 7.150136845327466e-06,
+ "loss": 0.3639181613922119,
+ "num_input_tokens_seen": 466056584,
+ "step": 10650,
+ "train_runtime": 82369.823,
+ "train_tokens_per_second": 5658.099
+ },
+ {
+ "epoch": 0.7174117963910585,
+ "grad_norm": 0.737238654335971,
+ "learning_rate": 7.147749425516637e-06,
+ "loss": 0.3720954418182373,
+ "num_input_tokens_seen": 466287992,
+ "step": 10655,
+ "train_runtime": 82407.102,
+ "train_tokens_per_second": 5658.347
+ },
+ {
+ "epoch": 0.7177484513870186,
+ "grad_norm": 0.7078523897720794,
+ "learning_rate": 7.145361405094529e-06,
+ "loss": 0.3578562021255493,
+ "num_input_tokens_seen": 466523176,
+ "step": 10660,
+ "train_runtime": 82439.2799,
+ "train_tokens_per_second": 5658.991
+ },
+ {
+ "epoch": 0.7180851063829787,
+ "grad_norm": 0.7362901383213949,
+ "learning_rate": 7.142972784728945e-06,
+ "loss": 0.3409886360168457,
+ "num_input_tokens_seen": 466734112,
+ "step": 10665,
+ "train_runtime": 82483.0478,
+ "train_tokens_per_second": 5658.546
+ },
+ {
+ "epoch": 0.7184217613789389,
+ "grad_norm": 0.8146924235270421,
+ "learning_rate": 7.140583565087856e-06,
+ "loss": 0.35494747161865237,
+ "num_input_tokens_seen": 466952312,
+ "step": 10670,
+ "train_runtime": 82519.1924,
+ "train_tokens_per_second": 5658.712
+ },
+ {
+ "epoch": 0.718758416374899,
+ "grad_norm": 0.7593849226158301,
+ "learning_rate": 7.138193746839399e-06,
+ "loss": 0.3933510065078735,
+ "num_input_tokens_seen": 467185248,
+ "step": 10675,
+ "train_runtime": 82565.1418,
+ "train_tokens_per_second": 5658.384
+ },
+ {
+ "epoch": 0.7190950713708592,
+ "grad_norm": 0.7169033125551184,
+ "learning_rate": 7.135803330651878e-06,
+ "loss": 0.37902536392211916,
+ "num_input_tokens_seen": 467421448,
+ "step": 10680,
+ "train_runtime": 82599.409,
+ "train_tokens_per_second": 5658.896
+ },
+ {
+ "epoch": 0.7194317263668193,
+ "grad_norm": 0.6927480355842544,
+ "learning_rate": 7.133412317193766e-06,
+ "loss": 0.3649748802185059,
+ "num_input_tokens_seen": 467648088,
+ "step": 10685,
+ "train_runtime": 82647.9029,
+ "train_tokens_per_second": 5658.318
+ },
+ {
+ "epoch": 0.7197683813627794,
+ "grad_norm": 0.8368737935217295,
+ "learning_rate": 7.131020707133703e-06,
+ "loss": 0.39012324810028076,
+ "num_input_tokens_seen": 467852784,
+ "step": 10690,
+ "train_runtime": 82685.2143,
+ "train_tokens_per_second": 5658.24
+ },
+ {
+ "epoch": 0.7201050363587396,
+ "grad_norm": 0.7246874489536738,
+ "learning_rate": 7.128628501140495e-06,
+ "loss": 0.3994007587432861,
+ "num_input_tokens_seen": 468062360,
+ "step": 10695,
+ "train_runtime": 82719.4298,
+ "train_tokens_per_second": 5658.433
+ },
+ {
+ "epoch": 0.7204416913546997,
+ "grad_norm": 0.7034507707765917,
+ "learning_rate": 7.126235699883115e-06,
+ "loss": 0.3543556213378906,
+ "num_input_tokens_seen": 468260752,
+ "step": 10700,
+ "train_runtime": 82762.0259,
+ "train_tokens_per_second": 5657.918
+ },
+ {
+ "epoch": 0.7207783463506598,
+ "grad_norm": 0.6713729985803661,
+ "learning_rate": 7.1238423040306995e-06,
+ "loss": 0.3659778594970703,
+ "num_input_tokens_seen": 468480296,
+ "step": 10705,
+ "train_runtime": 82807.3194,
+ "train_tokens_per_second": 5657.474
+ },
+ {
+ "epoch": 0.72111500134662,
+ "grad_norm": 0.6714547248755331,
+ "learning_rate": 7.121448314252557e-06,
+ "loss": 0.33471217155456545,
+ "num_input_tokens_seen": 468714640,
+ "step": 10710,
+ "train_runtime": 82849.0161,
+ "train_tokens_per_second": 5657.456
+ },
+ {
+ "epoch": 0.7214516563425801,
+ "grad_norm": 0.7542079623814812,
+ "learning_rate": 7.1190537312181586e-06,
+ "loss": 0.3590662479400635,
+ "num_input_tokens_seen": 468920200,
+ "step": 10715,
+ "train_runtime": 82883.6718,
+ "train_tokens_per_second": 5657.57
+ },
+ {
+ "epoch": 0.7217883113385403,
+ "grad_norm": 0.788689096156777,
+ "learning_rate": 7.116658555597141e-06,
+ "loss": 0.35546276569366453,
+ "num_input_tokens_seen": 469105088,
+ "step": 10720,
+ "train_runtime": 82920.6765,
+ "train_tokens_per_second": 5657.275
+ },
+ {
+ "epoch": 0.7221249663345004,
+ "grad_norm": 0.7720777482295464,
+ "learning_rate": 7.114262788059308e-06,
+ "loss": 0.36028232574462893,
+ "num_input_tokens_seen": 469316144,
+ "step": 10725,
+ "train_runtime": 82956.2558,
+ "train_tokens_per_second": 5657.393
+ },
+ {
+ "epoch": 0.7224616213304605,
+ "grad_norm": 0.7595426301739953,
+ "learning_rate": 7.111866429274627e-06,
+ "loss": 0.3734895706176758,
+ "num_input_tokens_seen": 469533344,
+ "step": 10730,
+ "train_runtime": 82995.9693,
+ "train_tokens_per_second": 5657.303
+ },
+ {
+ "epoch": 0.7227982763264207,
+ "grad_norm": 0.7895201856991956,
+ "learning_rate": 7.1094694799132356e-06,
+ "loss": 0.38394627571105955,
+ "num_input_tokens_seen": 469757776,
+ "step": 10735,
+ "train_runtime": 83039.5254,
+ "train_tokens_per_second": 5657.038
+ },
+ {
+ "epoch": 0.7231349313223808,
+ "grad_norm": 0.7170404350301777,
+ "learning_rate": 7.10707194064543e-06,
+ "loss": 0.355570125579834,
+ "num_input_tokens_seen": 469973048,
+ "step": 10740,
+ "train_runtime": 83082.4999,
+ "train_tokens_per_second": 5656.703
+ },
+ {
+ "epoch": 0.723471586318341,
+ "grad_norm": 0.751527765717788,
+ "learning_rate": 7.104673812141676e-06,
+ "loss": 0.4048286437988281,
+ "num_input_tokens_seen": 470200352,
+ "step": 10745,
+ "train_runtime": 83123.5825,
+ "train_tokens_per_second": 5656.642
+ },
+ {
+ "epoch": 0.7238082413143011,
+ "grad_norm": 0.7209513181054711,
+ "learning_rate": 7.102275095072601e-06,
+ "loss": 0.3921640157699585,
+ "num_input_tokens_seen": 470434416,
+ "step": 10750,
+ "train_runtime": 83170.1972,
+ "train_tokens_per_second": 5656.286
+ },
+ {
+ "epoch": 0.7241448963102612,
+ "grad_norm": 0.7141189337990387,
+ "learning_rate": 7.099875790108998e-06,
+ "loss": 0.36566712856292727,
+ "num_input_tokens_seen": 470662208,
+ "step": 10755,
+ "train_runtime": 83219.7805,
+ "train_tokens_per_second": 5655.653
+ },
+ {
+ "epoch": 0.7244815513062214,
+ "grad_norm": 0.6677452072067016,
+ "learning_rate": 7.097475897921827e-06,
+ "loss": 0.37137765884399415,
+ "num_input_tokens_seen": 470876192,
+ "step": 10760,
+ "train_runtime": 83259.577,
+ "train_tokens_per_second": 5655.52
+ },
+ {
+ "epoch": 0.7248182063021815,
+ "grad_norm": 0.8654836254283663,
+ "learning_rate": 7.09507541918221e-06,
+ "loss": 0.3979093790054321,
+ "num_input_tokens_seen": 471091920,
+ "step": 10765,
+ "train_runtime": 83298.0964,
+ "train_tokens_per_second": 5655.494
+ },
+ {
+ "epoch": 0.7251548612981417,
+ "grad_norm": 0.6072122160998616,
+ "learning_rate": 7.092674354561433e-06,
+ "loss": 0.3879514217376709,
+ "num_input_tokens_seen": 471316336,
+ "step": 10770,
+ "train_runtime": 83339.0663,
+ "train_tokens_per_second": 5655.407
+ },
+ {
+ "epoch": 0.7254915162941018,
+ "grad_norm": 0.753636374019264,
+ "learning_rate": 7.090272704730945e-06,
+ "loss": 0.3511503219604492,
+ "num_input_tokens_seen": 471518800,
+ "step": 10775,
+ "train_runtime": 83375.4431,
+ "train_tokens_per_second": 5655.368
+ },
+ {
+ "epoch": 0.7258281712900619,
+ "grad_norm": 0.7575044992088539,
+ "learning_rate": 7.08787047036236e-06,
+ "loss": 0.35472354888916013,
+ "num_input_tokens_seen": 471754616,
+ "step": 10780,
+ "train_runtime": 83418.0251,
+ "train_tokens_per_second": 5655.308
+ },
+ {
+ "epoch": 0.7261648262860221,
+ "grad_norm": 0.822516802699336,
+ "learning_rate": 7.085467652127457e-06,
+ "loss": 0.3795314788818359,
+ "num_input_tokens_seen": 471981872,
+ "step": 10785,
+ "train_runtime": 83450.133,
+ "train_tokens_per_second": 5655.855
+ },
+ {
+ "epoch": 0.7265014812819822,
+ "grad_norm": 0.6385132066085284,
+ "learning_rate": 7.083064250698175e-06,
+ "loss": 0.3529500961303711,
+ "num_input_tokens_seen": 472213144,
+ "step": 10790,
+ "train_runtime": 83491.9091,
+ "train_tokens_per_second": 5655.795
+ },
+ {
+ "epoch": 0.7268381362779424,
+ "grad_norm": 0.6595130665327932,
+ "learning_rate": 7.0806602667466175e-06,
+ "loss": 0.37914767265319826,
+ "num_input_tokens_seen": 472443584,
+ "step": 10795,
+ "train_runtime": 83534.7223,
+ "train_tokens_per_second": 5655.655
+ },
+ {
+ "epoch": 0.7271747912739025,
+ "grad_norm": 0.6988131699183094,
+ "learning_rate": 7.078255700945051e-06,
+ "loss": 0.38105590343475343,
+ "num_input_tokens_seen": 472655528,
+ "step": 10800,
+ "train_runtime": 83566.9737,
+ "train_tokens_per_second": 5656.009
+ },
+ {
+ "epoch": 0.7275114462698626,
+ "grad_norm": 0.7284700984007974,
+ "learning_rate": 7.075850553965904e-06,
+ "loss": 0.366654634475708,
+ "num_input_tokens_seen": 472863848,
+ "step": 10805,
+ "train_runtime": 83605.9871,
+ "train_tokens_per_second": 5655.861
+ },
+ {
+ "epoch": 0.7278481012658228,
+ "grad_norm": 0.7093309792405998,
+ "learning_rate": 7.07344482648177e-06,
+ "loss": 0.3818976879119873,
+ "num_input_tokens_seen": 473087584,
+ "step": 10810,
+ "train_runtime": 83637.9528,
+ "train_tokens_per_second": 5656.375
+ },
+ {
+ "epoch": 0.7281847562617829,
+ "grad_norm": 0.7655242160543194,
+ "learning_rate": 7.071038519165402e-06,
+ "loss": 0.3804640293121338,
+ "num_input_tokens_seen": 473302440,
+ "step": 10815,
+ "train_runtime": 83680.6746,
+ "train_tokens_per_second": 5656.054
+ },
+ {
+ "epoch": 0.7285214112577431,
+ "grad_norm": 0.7319764295438455,
+ "learning_rate": 7.068631632689716e-06,
+ "loss": 0.3819848299026489,
+ "num_input_tokens_seen": 473528888,
+ "step": 10820,
+ "train_runtime": 83719.4607,
+ "train_tokens_per_second": 5656.139
+ },
+ {
+ "epoch": 0.7288580662537032,
+ "grad_norm": 9.329714081971986,
+ "learning_rate": 7.066224167727791e-06,
+ "loss": 0.3868781566619873,
+ "num_input_tokens_seen": 473749232,
+ "step": 10825,
+ "train_runtime": 83753.7938,
+ "train_tokens_per_second": 5656.451
+ },
+ {
+ "epoch": 0.7291947212496633,
+ "grad_norm": 0.841160080727543,
+ "learning_rate": 7.063816124952865e-06,
+ "loss": 0.36863956451416013,
+ "num_input_tokens_seen": 473966688,
+ "step": 10830,
+ "train_runtime": 83791.8425,
+ "train_tokens_per_second": 5656.478
+ },
+ {
+ "epoch": 0.7295313762456235,
+ "grad_norm": 0.6549972032651611,
+ "learning_rate": 7.061407505038341e-06,
+ "loss": 0.3608518362045288,
+ "num_input_tokens_seen": 474183856,
+ "step": 10835,
+ "train_runtime": 83830.8355,
+ "train_tokens_per_second": 5656.437
+ },
+ {
+ "epoch": 0.7298680312415836,
+ "grad_norm": 0.6949899381534901,
+ "learning_rate": 7.058998308657782e-06,
+ "loss": 0.3866151809692383,
+ "num_input_tokens_seen": 474414672,
+ "step": 10840,
+ "train_runtime": 83863.1172,
+ "train_tokens_per_second": 5657.012
+ },
+ {
+ "epoch": 0.7302046862375438,
+ "grad_norm": 0.6794084004266899,
+ "learning_rate": 7.056588536484912e-06,
+ "loss": 0.35731284618377684,
+ "num_input_tokens_seen": 474622840,
+ "step": 10845,
+ "train_runtime": 83892.8777,
+ "train_tokens_per_second": 5657.487
+ },
+ {
+ "epoch": 0.7305413412335039,
+ "grad_norm": 1.3762092654034057,
+ "learning_rate": 7.0541781891936145e-06,
+ "loss": 0.35231194496154783,
+ "num_input_tokens_seen": 474842808,
+ "step": 10850,
+ "train_runtime": 83931.0911,
+ "train_tokens_per_second": 5657.532
+ },
+ {
+ "epoch": 0.730877996229464,
+ "grad_norm": 0.8182830044909798,
+ "learning_rate": 7.0517672674579364e-06,
+ "loss": 0.3739452838897705,
+ "num_input_tokens_seen": 475060040,
+ "step": 10855,
+ "train_runtime": 83971.3377,
+ "train_tokens_per_second": 5657.407
+ },
+ {
+ "epoch": 0.7312146512254242,
+ "grad_norm": 0.7405063263799381,
+ "learning_rate": 7.049355771952087e-06,
+ "loss": 0.367911171913147,
+ "num_input_tokens_seen": 475270960,
+ "step": 10860,
+ "train_runtime": 84012.5713,
+ "train_tokens_per_second": 5657.141
+ },
+ {
+ "epoch": 0.7315513062213843,
+ "grad_norm": 0.743519981806223,
+ "learning_rate": 7.046943703350429e-06,
+ "loss": 0.3706895112991333,
+ "num_input_tokens_seen": 475496232,
+ "step": 10865,
+ "train_runtime": 84056.0111,
+ "train_tokens_per_second": 5656.897
+ },
+ {
+ "epoch": 0.7318879612173445,
+ "grad_norm": 0.6596197156889615,
+ "learning_rate": 7.044531062327492e-06,
+ "loss": 0.356654691696167,
+ "num_input_tokens_seen": 475709832,
+ "step": 10870,
+ "train_runtime": 84089.5859,
+ "train_tokens_per_second": 5657.179
+ },
+ {
+ "epoch": 0.7322246162133046,
+ "grad_norm": 0.6526147436786898,
+ "learning_rate": 7.042117849557963e-06,
+ "loss": 0.3708888053894043,
+ "num_input_tokens_seen": 475932072,
+ "step": 10875,
+ "train_runtime": 84133.6504,
+ "train_tokens_per_second": 5656.858
+ },
+ {
+ "epoch": 0.7325612712092647,
+ "grad_norm": 0.744802883666199,
+ "learning_rate": 7.0397040657166885e-06,
+ "loss": 0.36887059211730955,
+ "num_input_tokens_seen": 476162776,
+ "step": 10880,
+ "train_runtime": 84169.2025,
+ "train_tokens_per_second": 5657.209
+ },
+ {
+ "epoch": 0.7328979262052249,
+ "grad_norm": 0.7068220775813847,
+ "learning_rate": 7.037289711478676e-06,
+ "loss": 0.34908170700073243,
+ "num_input_tokens_seen": 476376384,
+ "step": 10885,
+ "train_runtime": 84204.347,
+ "train_tokens_per_second": 5657.385
+ },
+ {
+ "epoch": 0.733234581201185,
+ "grad_norm": 0.7211408903358524,
+ "learning_rate": 7.034874787519095e-06,
+ "loss": 0.3759069204330444,
+ "num_input_tokens_seen": 476599400,
+ "step": 10890,
+ "train_runtime": 84242.2013,
+ "train_tokens_per_second": 5657.49
+ },
+ {
+ "epoch": 0.7335712361971451,
+ "grad_norm": 0.6965350386219624,
+ "learning_rate": 7.0324592945132655e-06,
+ "loss": 0.4069540023803711,
+ "num_input_tokens_seen": 476821304,
+ "step": 10895,
+ "train_runtime": 84276.594,
+ "train_tokens_per_second": 5657.814
+ },
+ {
+ "epoch": 0.7339078911931053,
+ "grad_norm": 0.7259339581846704,
+ "learning_rate": 7.0300432331366764e-06,
+ "loss": 0.3802814245223999,
+ "num_input_tokens_seen": 477036424,
+ "step": 10900,
+ "train_runtime": 84309.7162,
+ "train_tokens_per_second": 5658.143
+ },
+ {
+ "epoch": 0.7342445461890654,
+ "grad_norm": 0.726630192334144,
+ "learning_rate": 7.02762660406497e-06,
+ "loss": 0.3741856813430786,
+ "num_input_tokens_seen": 477257288,
+ "step": 10905,
+ "train_runtime": 84348.0236,
+ "train_tokens_per_second": 5658.192
+ },
+ {
+ "epoch": 0.7345812011850256,
+ "grad_norm": 0.6353064783947603,
+ "learning_rate": 7.02520940797395e-06,
+ "loss": 0.3721340179443359,
+ "num_input_tokens_seen": 477464672,
+ "step": 10910,
+ "train_runtime": 84385.7451,
+ "train_tokens_per_second": 5658.12
+ },
+ {
+ "epoch": 0.7349178561809857,
+ "grad_norm": 0.6280488461396918,
+ "learning_rate": 7.022791645539576e-06,
+ "loss": 0.37077898979187013,
+ "num_input_tokens_seen": 477670600,
+ "step": 10915,
+ "train_runtime": 84423.4019,
+ "train_tokens_per_second": 5658.035
+ },
+ {
+ "epoch": 0.7352545111769458,
+ "grad_norm": 0.7582582970874048,
+ "learning_rate": 7.02037331743797e-06,
+ "loss": 0.4068927764892578,
+ "num_input_tokens_seen": 477891112,
+ "step": 10920,
+ "train_runtime": 84461.3924,
+ "train_tokens_per_second": 5658.101
+ },
+ {
+ "epoch": 0.735591166172906,
+ "grad_norm": 0.6908837097564919,
+ "learning_rate": 7.0179544243454076e-06,
+ "loss": 0.3525757551193237,
+ "num_input_tokens_seen": 478120960,
+ "step": 10925,
+ "train_runtime": 84493.3059,
+ "train_tokens_per_second": 5658.684
+ },
+ {
+ "epoch": 0.7359278211688661,
+ "grad_norm": 0.8025906131360969,
+ "learning_rate": 7.0155349669383245e-06,
+ "loss": 0.36997184753417967,
+ "num_input_tokens_seen": 478344240,
+ "step": 10930,
+ "train_runtime": 84534.5442,
+ "train_tokens_per_second": 5658.565
+ },
+ {
+ "epoch": 0.7362644761648263,
+ "grad_norm": 0.691884724659927,
+ "learning_rate": 7.013114945893316e-06,
+ "loss": 0.36500933170318606,
+ "num_input_tokens_seen": 478567880,
+ "step": 10935,
+ "train_runtime": 84579.7591,
+ "train_tokens_per_second": 5658.184
+ },
+ {
+ "epoch": 0.7366011311607864,
+ "grad_norm": 0.7278702815389843,
+ "learning_rate": 7.010694361887131e-06,
+ "loss": 0.3656963109970093,
+ "num_input_tokens_seen": 478795792,
+ "step": 10940,
+ "train_runtime": 84624.1001,
+ "train_tokens_per_second": 5657.913
+ },
+ {
+ "epoch": 0.7369377861567465,
+ "grad_norm": 0.6970616747051929,
+ "learning_rate": 7.008273215596681e-06,
+ "loss": 0.3773782253265381,
+ "num_input_tokens_seen": 479014424,
+ "step": 10945,
+ "train_runtime": 84661.672,
+ "train_tokens_per_second": 5657.984
+ },
+ {
+ "epoch": 0.7372744411527067,
+ "grad_norm": 0.7390738693319422,
+ "learning_rate": 7.005851507699029e-06,
+ "loss": 0.3759654760360718,
+ "num_input_tokens_seen": 479243632,
+ "step": 10950,
+ "train_runtime": 84694.665,
+ "train_tokens_per_second": 5658.487
+ },
+ {
+ "epoch": 0.7376110961486668,
+ "grad_norm": 0.6808946119767743,
+ "learning_rate": 7.003429238871399e-06,
+ "loss": 0.3657572031021118,
+ "num_input_tokens_seen": 479473192,
+ "step": 10955,
+ "train_runtime": 84738.238,
+ "train_tokens_per_second": 5658.286
+ },
+ {
+ "epoch": 0.737947751144627,
+ "grad_norm": 0.7178504014412501,
+ "learning_rate": 7.001006409791171e-06,
+ "loss": 0.40654420852661133,
+ "num_input_tokens_seen": 479687920,
+ "step": 10960,
+ "train_runtime": 84776.0398,
+ "train_tokens_per_second": 5658.296
+ },
+ {
+ "epoch": 0.7382844061405871,
+ "grad_norm": 0.7628213758613887,
+ "learning_rate": 6.998583021135882e-06,
+ "loss": 0.3780509948730469,
+ "num_input_tokens_seen": 479906096,
+ "step": 10965,
+ "train_runtime": 84815.2532,
+ "train_tokens_per_second": 5658.252
+ },
+ {
+ "epoch": 0.7386210611365472,
+ "grad_norm": 0.8356715268382896,
+ "learning_rate": 6.996159073583224e-06,
+ "loss": 0.36737675666809083,
+ "num_input_tokens_seen": 480125696,
+ "step": 10970,
+ "train_runtime": 84852.9815,
+ "train_tokens_per_second": 5658.324
+ },
+ {
+ "epoch": 0.7389577161325074,
+ "grad_norm": 0.8055302742937802,
+ "learning_rate": 6.993734567811046e-06,
+ "loss": 0.34843399524688723,
+ "num_input_tokens_seen": 480343296,
+ "step": 10975,
+ "train_runtime": 84896.1427,
+ "train_tokens_per_second": 5658.011
+ },
+ {
+ "epoch": 0.7392943711284675,
+ "grad_norm": 0.6594571098967035,
+ "learning_rate": 6.991309504497352e-06,
+ "loss": 0.37664797306060793,
+ "num_input_tokens_seen": 480566544,
+ "step": 10980,
+ "train_runtime": 84934.0218,
+ "train_tokens_per_second": 5658.116
+ },
+ {
+ "epoch": 0.7396310261244277,
+ "grad_norm": 0.7694992058673898,
+ "learning_rate": 6.988883884320307e-06,
+ "loss": 0.3457742214202881,
+ "num_input_tokens_seen": 480774232,
+ "step": 10985,
+ "train_runtime": 84970.4518,
+ "train_tokens_per_second": 5658.134
+ },
+ {
+ "epoch": 0.7399676811203878,
+ "grad_norm": 0.7402878857378402,
+ "learning_rate": 6.9864577079582254e-06,
+ "loss": 0.39705471992492675,
+ "num_input_tokens_seen": 481000200,
+ "step": 10990,
+ "train_runtime": 85008.5973,
+ "train_tokens_per_second": 5658.254
+ },
+ {
+ "epoch": 0.7403043361163479,
+ "grad_norm": 0.8393816654846036,
+ "learning_rate": 6.984030976089581e-06,
+ "loss": 0.3594042301177979,
+ "num_input_tokens_seen": 481219544,
+ "step": 10995,
+ "train_runtime": 85043.9927,
+ "train_tokens_per_second": 5658.478
+ },
+ {
+ "epoch": 0.7406409911123081,
+ "grad_norm": 0.7035982175377999,
+ "learning_rate": 6.981603689392999e-06,
+ "loss": 0.3729727745056152,
+ "num_input_tokens_seen": 481433184,
+ "step": 11000,
+ "train_runtime": 85091.999,
+ "train_tokens_per_second": 5657.796
+ },
+ {
+ "epoch": 0.7409776461082682,
+ "grad_norm": 0.7063885505049592,
+ "learning_rate": 6.979175848547263e-06,
+ "loss": 0.3587386131286621,
+ "num_input_tokens_seen": 481648384,
+ "step": 11005,
+ "train_runtime": 85133.4023,
+ "train_tokens_per_second": 5657.572
+ },
+ {
+ "epoch": 0.7413143011042284,
+ "grad_norm": 0.818479059925005,
+ "learning_rate": 6.976747454231316e-06,
+ "loss": 0.36256041526794436,
+ "num_input_tokens_seen": 481853600,
+ "step": 11010,
+ "train_runtime": 85175.3796,
+ "train_tokens_per_second": 5657.193
+ },
+ {
+ "epoch": 0.7416509561001885,
+ "grad_norm": 0.7213537892974411,
+ "learning_rate": 6.974318507124245e-06,
+ "loss": 0.37796127796173096,
+ "num_input_tokens_seen": 482067848,
+ "step": 11015,
+ "train_runtime": 85211.3126,
+ "train_tokens_per_second": 5657.322
+ },
+ {
+ "epoch": 0.7419876110961486,
+ "grad_norm": 0.9553094338809663,
+ "learning_rate": 6.971889007905299e-06,
+ "loss": 0.38568458557128904,
+ "num_input_tokens_seen": 482272608,
+ "step": 11020,
+ "train_runtime": 85242.7243,
+ "train_tokens_per_second": 5657.64
+ },
+ {
+ "epoch": 0.7423242660921088,
+ "grad_norm": 0.6954588658200586,
+ "learning_rate": 6.9694589572538795e-06,
+ "loss": 0.4131626605987549,
+ "num_input_tokens_seen": 482500568,
+ "step": 11025,
+ "train_runtime": 85279.0869,
+ "train_tokens_per_second": 5657.9
+ },
+ {
+ "epoch": 0.7426609210880689,
+ "grad_norm": 0.6724190701417789,
+ "learning_rate": 6.9670283558495436e-06,
+ "loss": 0.3677205085754395,
+ "num_input_tokens_seen": 482727520,
+ "step": 11030,
+ "train_runtime": 85315.8408,
+ "train_tokens_per_second": 5658.123
+ },
+ {
+ "epoch": 0.742997576084029,
+ "grad_norm": 0.7736013776309341,
+ "learning_rate": 6.964597204371999e-06,
+ "loss": 0.3734050035476685,
+ "num_input_tokens_seen": 482943600,
+ "step": 11035,
+ "train_runtime": 85349.9838,
+ "train_tokens_per_second": 5658.391
+ },
+ {
+ "epoch": 0.7433342310799892,
+ "grad_norm": 0.7557593534274823,
+ "learning_rate": 6.962165503501111e-06,
+ "loss": 0.3419958591461182,
+ "num_input_tokens_seen": 483163040,
+ "step": 11040,
+ "train_runtime": 85392.7525,
+ "train_tokens_per_second": 5658.127
+ },
+ {
+ "epoch": 0.7436708860759493,
+ "grad_norm": 0.786614283633,
+ "learning_rate": 6.959733253916899e-06,
+ "loss": 0.39629015922546384,
+ "num_input_tokens_seen": 483387400,
+ "step": 11045,
+ "train_runtime": 85430.316,
+ "train_tokens_per_second": 5658.265
+ },
+ {
+ "epoch": 0.7440075410719095,
+ "grad_norm": 0.7118679551230915,
+ "learning_rate": 6.95730045629953e-06,
+ "loss": 0.3637987613677979,
+ "num_input_tokens_seen": 483586168,
+ "step": 11050,
+ "train_runtime": 85469.2598,
+ "train_tokens_per_second": 5658.013
+ },
+ {
+ "epoch": 0.7443441960678696,
+ "grad_norm": 0.775880823500199,
+ "learning_rate": 6.954867111329329e-06,
+ "loss": 0.3997496604919434,
+ "num_input_tokens_seen": 483812040,
+ "step": 11055,
+ "train_runtime": 85501.004,
+ "train_tokens_per_second": 5658.554
+ },
+ {
+ "epoch": 0.7446808510638298,
+ "grad_norm": 0.8104111912789507,
+ "learning_rate": 6.952433219686775e-06,
+ "loss": 0.36485040187835693,
+ "num_input_tokens_seen": 484033200,
+ "step": 11060,
+ "train_runtime": 85544.4427,
+ "train_tokens_per_second": 5658.266
+ },
+ {
+ "epoch": 0.7450175060597899,
+ "grad_norm": 0.745606714965393,
+ "learning_rate": 6.949998782052497e-06,
+ "loss": 0.38755149841308595,
+ "num_input_tokens_seen": 484251832,
+ "step": 11065,
+ "train_runtime": 85584.405,
+ "train_tokens_per_second": 5658.178
+ },
+ {
+ "epoch": 0.74535416105575,
+ "grad_norm": 0.7941311038883269,
+ "learning_rate": 6.947563799107277e-06,
+ "loss": 0.36263265609741213,
+ "num_input_tokens_seen": 484460952,
+ "step": 11070,
+ "train_runtime": 85617.994,
+ "train_tokens_per_second": 5658.401
+ },
+ {
+ "epoch": 0.7456908160517102,
+ "grad_norm": 0.795109174361431,
+ "learning_rate": 6.945128271532052e-06,
+ "loss": 0.37139124870300294,
+ "num_input_tokens_seen": 484696168,
+ "step": 11075,
+ "train_runtime": 85658.1904,
+ "train_tokens_per_second": 5658.492
+ },
+ {
+ "epoch": 0.7460274710476703,
+ "grad_norm": 0.7555629636652617,
+ "learning_rate": 6.942692200007908e-06,
+ "loss": 0.37421250343322754,
+ "num_input_tokens_seen": 484914304,
+ "step": 11080,
+ "train_runtime": 85693.7663,
+ "train_tokens_per_second": 5658.688
+ },
+ {
+ "epoch": 0.7463641260436304,
+ "grad_norm": 0.7710680161258006,
+ "learning_rate": 6.9402555852160845e-06,
+ "loss": 0.3868663787841797,
+ "num_input_tokens_seen": 485140952,
+ "step": 11085,
+ "train_runtime": 85733.9493,
+ "train_tokens_per_second": 5658.68
+ },
+ {
+ "epoch": 0.7467007810395906,
+ "grad_norm": 0.6618662744692935,
+ "learning_rate": 6.937818427837975e-06,
+ "loss": 0.35773367881774903,
+ "num_input_tokens_seen": 485364088,
+ "step": 11090,
+ "train_runtime": 85775.7043,
+ "train_tokens_per_second": 5658.526
+ },
+ {
+ "epoch": 0.7470374360355507,
+ "grad_norm": 0.8138674839150098,
+ "learning_rate": 6.93538072855512e-06,
+ "loss": 0.40668773651123047,
+ "num_input_tokens_seen": 485585456,
+ "step": 11095,
+ "train_runtime": 85811.2509,
+ "train_tokens_per_second": 5658.762
+ },
+ {
+ "epoch": 0.7473740910315109,
+ "grad_norm": 0.7237738810724516,
+ "learning_rate": 6.932942488049217e-06,
+ "loss": 0.3412586212158203,
+ "num_input_tokens_seen": 485799760,
+ "step": 11100,
+ "train_runtime": 85848.9089,
+ "train_tokens_per_second": 5658.776
+ },
+ {
+ "epoch": 0.747710746027471,
+ "grad_norm": 0.7192380582338419,
+ "learning_rate": 6.930503707002108e-06,
+ "loss": 0.39055523872375486,
+ "num_input_tokens_seen": 486017000,
+ "step": 11105,
+ "train_runtime": 85893.6142,
+ "train_tokens_per_second": 5658.36
+ },
+ {
+ "epoch": 0.7480474010234311,
+ "grad_norm": 0.7129551106844463,
+ "learning_rate": 6.928064386095796e-06,
+ "loss": 0.3696824789047241,
+ "num_input_tokens_seen": 486243416,
+ "step": 11110,
+ "train_runtime": 85934.5192,
+ "train_tokens_per_second": 5658.301
+ },
+ {
+ "epoch": 0.7483840560193913,
+ "grad_norm": 0.8296106328518675,
+ "learning_rate": 6.9256245260124264e-06,
+ "loss": 0.369446325302124,
+ "num_input_tokens_seen": 486469728,
+ "step": 11115,
+ "train_runtime": 85966.8967,
+ "train_tokens_per_second": 5658.803
+ },
+ {
+ "epoch": 0.7487207110153514,
+ "grad_norm": 0.8112267099332857,
+ "learning_rate": 6.923184127434298e-06,
+ "loss": 0.3864800453186035,
+ "num_input_tokens_seen": 486670960,
+ "step": 11120,
+ "train_runtime": 86001.8182,
+ "train_tokens_per_second": 5658.845
+ },
+ {
+ "epoch": 0.7490573660113116,
+ "grad_norm": 0.8933045922508607,
+ "learning_rate": 6.920743191043861e-06,
+ "loss": 0.35923840999603274,
+ "num_input_tokens_seen": 486881136,
+ "step": 11125,
+ "train_runtime": 86039.0507,
+ "train_tokens_per_second": 5658.839
+ },
+ {
+ "epoch": 0.7493940210072717,
+ "grad_norm": 0.6534193785965624,
+ "learning_rate": 6.918301717523714e-06,
+ "loss": 0.3482653617858887,
+ "num_input_tokens_seen": 487095072,
+ "step": 11130,
+ "train_runtime": 86077.5734,
+ "train_tokens_per_second": 5658.792
+ },
+ {
+ "epoch": 0.7497306760032318,
+ "grad_norm": 0.7160660990648459,
+ "learning_rate": 6.915859707556611e-06,
+ "loss": 0.3809635639190674,
+ "num_input_tokens_seen": 487312640,
+ "step": 11135,
+ "train_runtime": 86112.0956,
+ "train_tokens_per_second": 5659.05
+ },
+ {
+ "epoch": 0.750067330999192,
+ "grad_norm": 0.7880329367953478,
+ "learning_rate": 6.913417161825449e-06,
+ "loss": 0.40050134658813474,
+ "num_input_tokens_seen": 487525824,
+ "step": 11140,
+ "train_runtime": 86145.5605,
+ "train_tokens_per_second": 5659.326
+ },
+ {
+ "epoch": 0.7504039859951521,
+ "grad_norm": 0.7445156362898442,
+ "learning_rate": 6.91097408101328e-06,
+ "loss": 0.3789609432220459,
+ "num_input_tokens_seen": 487742328,
+ "step": 11145,
+ "train_runtime": 86186.4755,
+ "train_tokens_per_second": 5659.152
+ },
+ {
+ "epoch": 0.7507406409911123,
+ "grad_norm": 0.8940165351109152,
+ "learning_rate": 6.908530465803302e-06,
+ "loss": 0.370347785949707,
+ "num_input_tokens_seen": 487949632,
+ "step": 11150,
+ "train_runtime": 86217.4676,
+ "train_tokens_per_second": 5659.522
+ },
+ {
+ "epoch": 0.7510772959870724,
+ "grad_norm": 0.8568185419619945,
+ "learning_rate": 6.906086316878865e-06,
+ "loss": 0.3664633989334106,
+ "num_input_tokens_seen": 488152136,
+ "step": 11155,
+ "train_runtime": 86258.4672,
+ "train_tokens_per_second": 5659.179
+ },
+ {
+ "epoch": 0.7514139509830325,
+ "grad_norm": 0.6561053217546615,
+ "learning_rate": 6.903641634923468e-06,
+ "loss": 0.3477164268493652,
+ "num_input_tokens_seen": 488367048,
+ "step": 11160,
+ "train_runtime": 86297.7659,
+ "train_tokens_per_second": 5659.093
+ },
+ {
+ "epoch": 0.7517506059789927,
+ "grad_norm": 0.7394466907995108,
+ "learning_rate": 6.9011964206207585e-06,
+ "loss": 0.3994622230529785,
+ "num_input_tokens_seen": 488567560,
+ "step": 11165,
+ "train_runtime": 86330.4394,
+ "train_tokens_per_second": 5659.273
+ },
+ {
+ "epoch": 0.7520872609749528,
+ "grad_norm": 0.7890852129954726,
+ "learning_rate": 6.898750674654531e-06,
+ "loss": 0.36271369457244873,
+ "num_input_tokens_seen": 488788328,
+ "step": 11170,
+ "train_runtime": 86369.9487,
+ "train_tokens_per_second": 5659.241
+ },
+ {
+ "epoch": 0.752423915970913,
+ "grad_norm": 0.6327317994824692,
+ "learning_rate": 6.896304397708733e-06,
+ "loss": 0.35087008476257325,
+ "num_input_tokens_seen": 488995192,
+ "step": 11175,
+ "train_runtime": 86408.8053,
+ "train_tokens_per_second": 5659.09
+ },
+ {
+ "epoch": 0.7527605709668731,
+ "grad_norm": 0.7895910131617881,
+ "learning_rate": 6.893857590467456e-06,
+ "loss": 0.3524162292480469,
+ "num_input_tokens_seen": 489213016,
+ "step": 11180,
+ "train_runtime": 86445.1307,
+ "train_tokens_per_second": 5659.232
+ },
+ {
+ "epoch": 0.7530972259628332,
+ "grad_norm": 0.7380392223605537,
+ "learning_rate": 6.891410253614944e-06,
+ "loss": 0.34860227108001707,
+ "num_input_tokens_seen": 489418272,
+ "step": 11185,
+ "train_runtime": 86484.3079,
+ "train_tokens_per_second": 5659.041
+ },
+ {
+ "epoch": 0.7534338809587934,
+ "grad_norm": 0.702987221857376,
+ "learning_rate": 6.888962387835586e-06,
+ "loss": 0.384684419631958,
+ "num_input_tokens_seen": 489652568,
+ "step": 11190,
+ "train_runtime": 86517.2831,
+ "train_tokens_per_second": 5659.593
+ },
+ {
+ "epoch": 0.7537705359547535,
+ "grad_norm": 0.7485124246366612,
+ "learning_rate": 6.886513993813919e-06,
+ "loss": 0.38306422233581544,
+ "num_input_tokens_seen": 489864096,
+ "step": 11195,
+ "train_runtime": 86557.4583,
+ "train_tokens_per_second": 5659.409
+ },
+ {
+ "epoch": 0.7541071909507137,
+ "grad_norm": 0.7067279867145273,
+ "learning_rate": 6.88406507223463e-06,
+ "loss": 0.36569690704345703,
+ "num_input_tokens_seen": 490092880,
+ "step": 11200,
+ "train_runtime": 86594.5272,
+ "train_tokens_per_second": 5659.629
+ },
+ {
+ "epoch": 0.7544438459466738,
+ "grad_norm": 0.7595465228382173,
+ "learning_rate": 6.881615623782551e-06,
+ "loss": 0.35335478782653806,
+ "num_input_tokens_seen": 490301440,
+ "step": 11205,
+ "train_runtime": 86631.5997,
+ "train_tokens_per_second": 5659.614
+ },
+ {
+ "epoch": 0.7547805009426339,
+ "grad_norm": 0.7480898301661179,
+ "learning_rate": 6.879165649142663e-06,
+ "loss": 0.3659096956253052,
+ "num_input_tokens_seen": 490500792,
+ "step": 11210,
+ "train_runtime": 86669.1193,
+ "train_tokens_per_second": 5659.464
+ },
+ {
+ "epoch": 0.7551171559385941,
+ "grad_norm": 0.6467936248180215,
+ "learning_rate": 6.876715149000093e-06,
+ "loss": 0.36233794689178467,
+ "num_input_tokens_seen": 490705136,
+ "step": 11215,
+ "train_runtime": 86708.2047,
+ "train_tokens_per_second": 5659.27
+ },
+ {
+ "epoch": 0.7554538109345542,
+ "grad_norm": 0.7242780017786447,
+ "learning_rate": 6.874264124040117e-06,
+ "loss": 0.3791798114776611,
+ "num_input_tokens_seen": 490930904,
+ "step": 11220,
+ "train_runtime": 86744.2352,
+ "train_tokens_per_second": 5659.522
+ },
+ {
+ "epoch": 0.7557904659305144,
+ "grad_norm": 0.7534511874997005,
+ "learning_rate": 6.871812574948154e-06,
+ "loss": 0.394960880279541,
+ "num_input_tokens_seen": 491158544,
+ "step": 11225,
+ "train_runtime": 86777.1348,
+ "train_tokens_per_second": 5660.0
+ },
+ {
+ "epoch": 0.7561271209264745,
+ "grad_norm": 0.7151778561469025,
+ "learning_rate": 6.869360502409773e-06,
+ "loss": 0.3845766544342041,
+ "num_input_tokens_seen": 491385232,
+ "step": 11230,
+ "train_runtime": 86814.8579,
+ "train_tokens_per_second": 5660.151
+ },
+ {
+ "epoch": 0.7564637759224346,
+ "grad_norm": 0.7179238868789197,
+ "learning_rate": 6.866907907110689e-06,
+ "loss": 0.351468563079834,
+ "num_input_tokens_seen": 491593976,
+ "step": 11235,
+ "train_runtime": 86852.7406,
+ "train_tokens_per_second": 5660.086
+ },
+ {
+ "epoch": 0.7568004309183949,
+ "grad_norm": 0.6534240517661798,
+ "learning_rate": 6.864454789736763e-06,
+ "loss": 0.3472340822219849,
+ "num_input_tokens_seen": 491827504,
+ "step": 11240,
+ "train_runtime": 86890.0801,
+ "train_tokens_per_second": 5660.341
+ },
+ {
+ "epoch": 0.757137085914355,
+ "grad_norm": 0.6984388211704616,
+ "learning_rate": 6.8620011509740005e-06,
+ "loss": 0.3707188844680786,
+ "num_input_tokens_seen": 492042264,
+ "step": 11245,
+ "train_runtime": 86929.2774,
+ "train_tokens_per_second": 5660.259
+ },
+ {
+ "epoch": 0.7574737409103152,
+ "grad_norm": 0.8438964758084152,
+ "learning_rate": 6.859546991508552e-06,
+ "loss": 0.3541130542755127,
+ "num_input_tokens_seen": 492261592,
+ "step": 11250,
+ "train_runtime": 86971.6043,
+ "train_tokens_per_second": 5660.027
+ },
+ {
+ "epoch": 0.7578103959062753,
+ "grad_norm": 0.6972841594933251,
+ "learning_rate": 6.857092312026716e-06,
+ "loss": 0.3816833972930908,
+ "num_input_tokens_seen": 492495240,
+ "step": 11255,
+ "train_runtime": 87006.7607,
+ "train_tokens_per_second": 5660.425
+ },
+ {
+ "epoch": 0.7581470509022354,
+ "grad_norm": 0.7161736954331712,
+ "learning_rate": 6.8546371132149395e-06,
+ "loss": 0.39476380348205564,
+ "num_input_tokens_seen": 492712328,
+ "step": 11260,
+ "train_runtime": 87046.3192,
+ "train_tokens_per_second": 5660.346
+ },
+ {
+ "epoch": 0.7584837058981956,
+ "grad_norm": 0.7348844948766989,
+ "learning_rate": 6.852181395759807e-06,
+ "loss": 0.39330875873565674,
+ "num_input_tokens_seen": 492943248,
+ "step": 11265,
+ "train_runtime": 87087.1631,
+ "train_tokens_per_second": 5660.343
+ },
+ {
+ "epoch": 0.7588203608941557,
+ "grad_norm": 0.5683273529508269,
+ "learning_rate": 6.849725160348054e-06,
+ "loss": 0.3455686330795288,
+ "num_input_tokens_seen": 493152344,
+ "step": 11270,
+ "train_runtime": 87127.3625,
+ "train_tokens_per_second": 5660.132
+ },
+ {
+ "epoch": 0.7591570158901159,
+ "grad_norm": 0.7258002012386351,
+ "learning_rate": 6.847268407666557e-06,
+ "loss": 0.35138664245605467,
+ "num_input_tokens_seen": 493375584,
+ "step": 11275,
+ "train_runtime": 87171.3314,
+ "train_tokens_per_second": 5659.838
+ },
+ {
+ "epoch": 0.759493670886076,
+ "grad_norm": 0.7379175392341268,
+ "learning_rate": 6.844811138402341e-06,
+ "loss": 0.37623722553253175,
+ "num_input_tokens_seen": 493594016,
+ "step": 11280,
+ "train_runtime": 87212.4908,
+ "train_tokens_per_second": 5659.671
+ },
+ {
+ "epoch": 0.7598303258820361,
+ "grad_norm": 0.7131462906095328,
+ "learning_rate": 6.842353353242574e-06,
+ "loss": 0.36553401947021485,
+ "num_input_tokens_seen": 493803080,
+ "step": 11285,
+ "train_runtime": 87254.8113,
+ "train_tokens_per_second": 5659.322
+ },
+ {
+ "epoch": 0.7601669808779963,
+ "grad_norm": 0.7404989121841616,
+ "learning_rate": 6.839895052874565e-06,
+ "loss": 0.3759953737258911,
+ "num_input_tokens_seen": 494027672,
+ "step": 11290,
+ "train_runtime": 87290.4517,
+ "train_tokens_per_second": 5659.584
+ },
+ {
+ "epoch": 0.7605036358739564,
+ "grad_norm": 0.6838203757609232,
+ "learning_rate": 6.837436237985772e-06,
+ "loss": 0.3641194343566895,
+ "num_input_tokens_seen": 494238208,
+ "step": 11295,
+ "train_runtime": 87333.9419,
+ "train_tokens_per_second": 5659.177
+ },
+ {
+ "epoch": 0.7608402908699166,
+ "grad_norm": 0.850634796737577,
+ "learning_rate": 6.834976909263794e-06,
+ "loss": 0.37904021739959715,
+ "num_input_tokens_seen": 494450104,
+ "step": 11300,
+ "train_runtime": 87372.8846,
+ "train_tokens_per_second": 5659.08
+ },
+ {
+ "epoch": 0.7611769458658767,
+ "grad_norm": 0.7504214621662669,
+ "learning_rate": 6.832517067396374e-06,
+ "loss": 0.3586588859558105,
+ "num_input_tokens_seen": 494683264,
+ "step": 11305,
+ "train_runtime": 87416.6251,
+ "train_tokens_per_second": 5658.915
+ },
+ {
+ "epoch": 0.7615136008618368,
+ "grad_norm": 0.7181780717405477,
+ "learning_rate": 6.8300567130714e-06,
+ "loss": 0.39732561111450193,
+ "num_input_tokens_seen": 494912080,
+ "step": 11310,
+ "train_runtime": 87455.7279,
+ "train_tokens_per_second": 5659.001
+ },
+ {
+ "epoch": 0.761850255857797,
+ "grad_norm": 0.8326466499790485,
+ "learning_rate": 6.8275958469769035e-06,
+ "loss": 0.38528354167938234,
+ "num_input_tokens_seen": 495143768,
+ "step": 11315,
+ "train_runtime": 87496.2946,
+ "train_tokens_per_second": 5659.026
+ },
+ {
+ "epoch": 0.7621869108537571,
+ "grad_norm": 0.8043440720773757,
+ "learning_rate": 6.825134469801055e-06,
+ "loss": 0.3580173969268799,
+ "num_input_tokens_seen": 495373776,
+ "step": 11320,
+ "train_runtime": 87536.7056,
+ "train_tokens_per_second": 5659.041
+ },
+ {
+ "epoch": 0.7625235658497173,
+ "grad_norm": 0.7612366109464449,
+ "learning_rate": 6.8226725822321715e-06,
+ "loss": 0.4001301765441895,
+ "num_input_tokens_seen": 495581360,
+ "step": 11325,
+ "train_runtime": 87574.6303,
+ "train_tokens_per_second": 5658.96
+ },
+ {
+ "epoch": 0.7628602208456774,
+ "grad_norm": 0.7254035816528333,
+ "learning_rate": 6.820210184958712e-06,
+ "loss": 0.38148252964019774,
+ "num_input_tokens_seen": 495805664,
+ "step": 11330,
+ "train_runtime": 87611.2812,
+ "train_tokens_per_second": 5659.153
+ },
+ {
+ "epoch": 0.7631968758416375,
+ "grad_norm": 0.7816370531432046,
+ "learning_rate": 6.81774727866928e-06,
+ "loss": 0.38228158950805663,
+ "num_input_tokens_seen": 496015888,
+ "step": 11335,
+ "train_runtime": 87647.9698,
+ "train_tokens_per_second": 5659.183
+ },
+ {
+ "epoch": 0.7635335308375977,
+ "grad_norm": 0.9719681651052307,
+ "learning_rate": 6.815283864052617e-06,
+ "loss": 0.3840935468673706,
+ "num_input_tokens_seen": 496237176,
+ "step": 11340,
+ "train_runtime": 87687.7469,
+ "train_tokens_per_second": 5659.139
+ },
+ {
+ "epoch": 0.7638701858335578,
+ "grad_norm": 0.7821628809832356,
+ "learning_rate": 6.812819941797611e-06,
+ "loss": 0.38556504249572754,
+ "num_input_tokens_seen": 496442520,
+ "step": 11345,
+ "train_runtime": 87730.1254,
+ "train_tokens_per_second": 5658.746
+ },
+ {
+ "epoch": 0.764206840829518,
+ "grad_norm": 0.7793739338464833,
+ "learning_rate": 6.810355512593288e-06,
+ "loss": 0.3861824035644531,
+ "num_input_tokens_seen": 496669864,
+ "step": 11350,
+ "train_runtime": 87767.0891,
+ "train_tokens_per_second": 5658.953
+ },
+ {
+ "epoch": 0.7645434958254781,
+ "grad_norm": 0.7769049083652588,
+ "learning_rate": 6.807890577128817e-06,
+ "loss": 0.33897886276245115,
+ "num_input_tokens_seen": 496892112,
+ "step": 11355,
+ "train_runtime": 87813.3112,
+ "train_tokens_per_second": 5658.506
+ },
+ {
+ "epoch": 0.7648801508214382,
+ "grad_norm": 0.8235786764795517,
+ "learning_rate": 6.805425136093513e-06,
+ "loss": 0.3582730531692505,
+ "num_input_tokens_seen": 497111304,
+ "step": 11360,
+ "train_runtime": 87847.6951,
+ "train_tokens_per_second": 5658.786
+ },
+ {
+ "epoch": 0.7652168058173984,
+ "grad_norm": 0.825724303695909,
+ "learning_rate": 6.802959190176827e-06,
+ "loss": 0.38888885974884035,
+ "num_input_tokens_seen": 497318752,
+ "step": 11365,
+ "train_runtime": 87886.6126,
+ "train_tokens_per_second": 5658.641
+ },
+ {
+ "epoch": 0.7655534608133585,
+ "grad_norm": 0.7692237966789585,
+ "learning_rate": 6.800492740068352e-06,
+ "loss": 0.3655484676361084,
+ "num_input_tokens_seen": 497523256,
+ "step": 11370,
+ "train_runtime": 87923.25,
+ "train_tokens_per_second": 5658.609
+ },
+ {
+ "epoch": 0.7658901158093187,
+ "grad_norm": 0.8501602672533506,
+ "learning_rate": 6.798025786457825e-06,
+ "loss": 0.35955286026000977,
+ "num_input_tokens_seen": 497749392,
+ "step": 11375,
+ "train_runtime": 87963.9631,
+ "train_tokens_per_second": 5658.56
+ },
+ {
+ "epoch": 0.7662267708052788,
+ "grad_norm": 0.6654629734284263,
+ "learning_rate": 6.795558330035118e-06,
+ "loss": 0.3689566612243652,
+ "num_input_tokens_seen": 497978632,
+ "step": 11380,
+ "train_runtime": 88010.5357,
+ "train_tokens_per_second": 5658.171
+ },
+ {
+ "epoch": 0.7665634258012389,
+ "grad_norm": 0.571714653501898,
+ "learning_rate": 6.793090371490251e-06,
+ "loss": 0.37029447555541994,
+ "num_input_tokens_seen": 498211416,
+ "step": 11385,
+ "train_runtime": 88043.3102,
+ "train_tokens_per_second": 5658.708
+ },
+ {
+ "epoch": 0.7669000807971991,
+ "grad_norm": 0.6562894791866329,
+ "learning_rate": 6.790621911513378e-06,
+ "loss": 0.36082067489624026,
+ "num_input_tokens_seen": 498430376,
+ "step": 11390,
+ "train_runtime": 88076.9318,
+ "train_tokens_per_second": 5659.034
+ },
+ {
+ "epoch": 0.7672367357931592,
+ "grad_norm": 0.6830456569016362,
+ "learning_rate": 6.788152950794799e-06,
+ "loss": 0.3912408113479614,
+ "num_input_tokens_seen": 498646776,
+ "step": 11395,
+ "train_runtime": 88113.1004,
+ "train_tokens_per_second": 5659.167
+ },
+ {
+ "epoch": 0.7675733907891193,
+ "grad_norm": 0.8063375531181857,
+ "learning_rate": 6.7856834900249495e-06,
+ "loss": 0.3919992208480835,
+ "num_input_tokens_seen": 498865144,
+ "step": 11400,
+ "train_runtime": 88157.9923,
+ "train_tokens_per_second": 5658.763
+ },
+ {
+ "epoch": 0.7679100457850795,
+ "grad_norm": 0.7488462821998634,
+ "learning_rate": 6.783213529894405e-06,
+ "loss": 0.3473628044128418,
+ "num_input_tokens_seen": 499082328,
+ "step": 11405,
+ "train_runtime": 88202.6214,
+ "train_tokens_per_second": 5658.362
+ },
+ {
+ "epoch": 0.7682467007810396,
+ "grad_norm": 0.7845757383023176,
+ "learning_rate": 6.780743071093883e-06,
+ "loss": 0.35947589874267577,
+ "num_input_tokens_seen": 499306216,
+ "step": 11410,
+ "train_runtime": 88242.1656,
+ "train_tokens_per_second": 5658.363
+ },
+ {
+ "epoch": 0.7685833557769998,
+ "grad_norm": 1.4808480239457869,
+ "learning_rate": 6.77827211431424e-06,
+ "loss": 0.36936905384063723,
+ "num_input_tokens_seen": 499532072,
+ "step": 11415,
+ "train_runtime": 88280.0767,
+ "train_tokens_per_second": 5658.492
+ },
+ {
+ "epoch": 0.7689200107729599,
+ "grad_norm": 0.7193580497484903,
+ "learning_rate": 6.775800660246471e-06,
+ "loss": 0.36505095958709716,
+ "num_input_tokens_seen": 499732272,
+ "step": 11420,
+ "train_runtime": 88316.7765,
+ "train_tokens_per_second": 5658.407
+ },
+ {
+ "epoch": 0.76925666576892,
+ "grad_norm": 0.7225028793880158,
+ "learning_rate": 6.77332870958171e-06,
+ "loss": 0.3719770908355713,
+ "num_input_tokens_seen": 499963848,
+ "step": 11425,
+ "train_runtime": 88361.2969,
+ "train_tokens_per_second": 5658.177
+ },
+ {
+ "epoch": 0.7695933207648802,
+ "grad_norm": 0.7310175751597672,
+ "learning_rate": 6.770856263011229e-06,
+ "loss": 0.3673600196838379,
+ "num_input_tokens_seen": 500175216,
+ "step": 11430,
+ "train_runtime": 88394.2396,
+ "train_tokens_per_second": 5658.459
+ },
+ {
+ "epoch": 0.7699299757608403,
+ "grad_norm": 0.8011531340232532,
+ "learning_rate": 6.768383321226442e-06,
+ "loss": 0.38931114673614503,
+ "num_input_tokens_seen": 500380304,
+ "step": 11435,
+ "train_runtime": 88432.6121,
+ "train_tokens_per_second": 5658.323
+ },
+ {
+ "epoch": 0.7702666307568005,
+ "grad_norm": 1.540406850238615,
+ "learning_rate": 6.765909884918898e-06,
+ "loss": 0.3583209753036499,
+ "num_input_tokens_seen": 500590256,
+ "step": 11440,
+ "train_runtime": 88469.3611,
+ "train_tokens_per_second": 5658.346
+ },
+ {
+ "epoch": 0.7706032857527606,
+ "grad_norm": 0.6645503979893219,
+ "learning_rate": 6.763435954780286e-06,
+ "loss": 0.3920794725418091,
+ "num_input_tokens_seen": 500810896,
+ "step": 11445,
+ "train_runtime": 88504.9842,
+ "train_tokens_per_second": 5658.561
+ },
+ {
+ "epoch": 0.7709399407487207,
+ "grad_norm": 0.7669134143397409,
+ "learning_rate": 6.7609615315024325e-06,
+ "loss": 0.37000212669372556,
+ "num_input_tokens_seen": 501027440,
+ "step": 11450,
+ "train_runtime": 88543.2243,
+ "train_tokens_per_second": 5658.563
+ },
+ {
+ "epoch": 0.7712765957446809,
+ "grad_norm": 0.7415266590401455,
+ "learning_rate": 6.758486615777301e-06,
+ "loss": 0.37416849136352537,
+ "num_input_tokens_seen": 501241576,
+ "step": 11455,
+ "train_runtime": 88583.8287,
+ "train_tokens_per_second": 5658.387
+ },
+ {
+ "epoch": 0.771613250740641,
+ "grad_norm": 0.8407214839390016,
+ "learning_rate": 6.756011208296995e-06,
+ "loss": 0.3626854419708252,
+ "num_input_tokens_seen": 501454696,
+ "step": 11460,
+ "train_runtime": 88621.4278,
+ "train_tokens_per_second": 5658.391
+ },
+ {
+ "epoch": 0.7719499057366012,
+ "grad_norm": 0.7461388799032196,
+ "learning_rate": 6.753535309753755e-06,
+ "loss": 0.35544500350952146,
+ "num_input_tokens_seen": 501658528,
+ "step": 11465,
+ "train_runtime": 88662.546,
+ "train_tokens_per_second": 5658.066
+ },
+ {
+ "epoch": 0.7722865607325613,
+ "grad_norm": 0.6919457018289741,
+ "learning_rate": 6.7510589208399565e-06,
+ "loss": 0.3747244119644165,
+ "num_input_tokens_seen": 501877728,
+ "step": 11470,
+ "train_runtime": 88703.6872,
+ "train_tokens_per_second": 5657.913
+ },
+ {
+ "epoch": 0.7726232157285214,
+ "grad_norm": 0.7585317880232398,
+ "learning_rate": 6.748582042248115e-06,
+ "loss": 0.35962686538696287,
+ "num_input_tokens_seen": 502089032,
+ "step": 11475,
+ "train_runtime": 88740.6334,
+ "train_tokens_per_second": 5657.938
+ },
+ {
+ "epoch": 0.7729598707244816,
+ "grad_norm": 0.7198688949375671,
+ "learning_rate": 6.746104674670882e-06,
+ "loss": 0.38360817432403566,
+ "num_input_tokens_seen": 502313408,
+ "step": 11480,
+ "train_runtime": 88777.3916,
+ "train_tokens_per_second": 5658.123
+ },
+ {
+ "epoch": 0.7732965257204417,
+ "grad_norm": 0.6976205058667109,
+ "learning_rate": 6.743626818801042e-06,
+ "loss": 0.39516823291778563,
+ "num_input_tokens_seen": 502534360,
+ "step": 11485,
+ "train_runtime": 88818.4462,
+ "train_tokens_per_second": 5657.995
+ },
+ {
+ "epoch": 0.7736331807164019,
+ "grad_norm": 0.6984390114365983,
+ "learning_rate": 6.741148475331525e-06,
+ "loss": 0.37048795223236086,
+ "num_input_tokens_seen": 502757200,
+ "step": 11490,
+ "train_runtime": 88854.2343,
+ "train_tokens_per_second": 5658.224
+ },
+ {
+ "epoch": 0.773969835712362,
+ "grad_norm": 0.6953225919926989,
+ "learning_rate": 6.738669644955388e-06,
+ "loss": 0.3468164443969727,
+ "num_input_tokens_seen": 502970848,
+ "step": 11495,
+ "train_runtime": 88897.1339,
+ "train_tokens_per_second": 5657.897
+ },
+ {
+ "epoch": 0.7743064907083221,
+ "grad_norm": 0.7656686554212365,
+ "learning_rate": 6.736190328365828e-06,
+ "loss": 0.3672001838684082,
+ "num_input_tokens_seen": 503195432,
+ "step": 11500,
+ "train_runtime": 88938.4411,
+ "train_tokens_per_second": 5657.795
+ },
+ {
+ "epoch": 0.7746431457042823,
+ "grad_norm": 0.743345653178439,
+ "learning_rate": 6.73371052625618e-06,
+ "loss": 0.37738971710205077,
+ "num_input_tokens_seen": 503417816,
+ "step": 11505,
+ "train_runtime": 88979.7242,
+ "train_tokens_per_second": 5657.669
+ },
+ {
+ "epoch": 0.7749798007002424,
+ "grad_norm": 0.7491341805650183,
+ "learning_rate": 6.731230239319912e-06,
+ "loss": 0.37387096881866455,
+ "num_input_tokens_seen": 503642040,
+ "step": 11510,
+ "train_runtime": 89021.4956,
+ "train_tokens_per_second": 5657.533
+ },
+ {
+ "epoch": 0.7753164556962026,
+ "grad_norm": 0.6256746258505268,
+ "learning_rate": 6.728749468250627e-06,
+ "loss": 0.38176569938659666,
+ "num_input_tokens_seen": 503870440,
+ "step": 11515,
+ "train_runtime": 89058.3538,
+ "train_tokens_per_second": 5657.756
+ },
+ {
+ "epoch": 0.7756531106921627,
+ "grad_norm": 0.8527701221713213,
+ "learning_rate": 6.726268213742066e-06,
+ "loss": 0.375903582572937,
+ "num_input_tokens_seen": 504076280,
+ "step": 11520,
+ "train_runtime": 89100.0272,
+ "train_tokens_per_second": 5657.42
+ },
+ {
+ "epoch": 0.7759897656881228,
+ "grad_norm": 0.7746670396544606,
+ "learning_rate": 6.723786476488105e-06,
+ "loss": 0.38536639213562013,
+ "num_input_tokens_seen": 504303624,
+ "step": 11525,
+ "train_runtime": 89134.096,
+ "train_tokens_per_second": 5657.808
+ },
+ {
+ "epoch": 0.776326420684083,
+ "grad_norm": 0.6906699278173198,
+ "learning_rate": 6.72130425718275e-06,
+ "loss": 0.3504000186920166,
+ "num_input_tokens_seen": 504511120,
+ "step": 11530,
+ "train_runtime": 89170.4494,
+ "train_tokens_per_second": 5657.829
+ },
+ {
+ "epoch": 0.7766630756800431,
+ "grad_norm": 0.7640134079354167,
+ "learning_rate": 6.718821556520151e-06,
+ "loss": 0.3902331829071045,
+ "num_input_tokens_seen": 504720744,
+ "step": 11535,
+ "train_runtime": 89203.783,
+ "train_tokens_per_second": 5658.064
+ },
+ {
+ "epoch": 0.7769997306760033,
+ "grad_norm": 0.676852860064436,
+ "learning_rate": 6.716338375194583e-06,
+ "loss": 0.3558895826339722,
+ "num_input_tokens_seen": 504945672,
+ "step": 11540,
+ "train_runtime": 89239.0086,
+ "train_tokens_per_second": 5658.351
+ },
+ {
+ "epoch": 0.7773363856719634,
+ "grad_norm": 0.7081316231079949,
+ "learning_rate": 6.713854713900463e-06,
+ "loss": 0.39481263160705565,
+ "num_input_tokens_seen": 505168112,
+ "step": 11545,
+ "train_runtime": 89280.0653,
+ "train_tokens_per_second": 5658.241
+ },
+ {
+ "epoch": 0.7776730406679235,
+ "grad_norm": 0.7771142918437934,
+ "learning_rate": 6.7113705733323365e-06,
+ "loss": 0.3915176153182983,
+ "num_input_tokens_seen": 505373904,
+ "step": 11550,
+ "train_runtime": 89321.9305,
+ "train_tokens_per_second": 5657.893
+ },
+ {
+ "epoch": 0.7780096956638837,
+ "grad_norm": 0.7862100764061776,
+ "learning_rate": 6.708885954184885e-06,
+ "loss": 0.36997599601745607,
+ "num_input_tokens_seen": 505592960,
+ "step": 11555,
+ "train_runtime": 89362.1395,
+ "train_tokens_per_second": 5657.798
+ },
+ {
+ "epoch": 0.7783463506598438,
+ "grad_norm": 0.8044048640519208,
+ "learning_rate": 6.7064008571529275e-06,
+ "loss": 0.3574796199798584,
+ "num_input_tokens_seen": 505815992,
+ "step": 11560,
+ "train_runtime": 89407.5364,
+ "train_tokens_per_second": 5657.42
+ },
+ {
+ "epoch": 0.778683005655804,
+ "grad_norm": 0.63451019489799,
+ "learning_rate": 6.703915282931411e-06,
+ "loss": 0.37357444763183595,
+ "num_input_tokens_seen": 506041792,
+ "step": 11565,
+ "train_runtime": 89452.1107,
+ "train_tokens_per_second": 5657.125
+ },
+ {
+ "epoch": 0.7790196606517641,
+ "grad_norm": 0.7400296332193441,
+ "learning_rate": 6.701429232215418e-06,
+ "loss": 0.3822374105453491,
+ "num_input_tokens_seen": 506244224,
+ "step": 11570,
+ "train_runtime": 89491.3476,
+ "train_tokens_per_second": 5656.907
+ },
+ {
+ "epoch": 0.7793563156477242,
+ "grad_norm": 0.8414435559233742,
+ "learning_rate": 6.6989427057001665e-06,
+ "loss": 0.3804741144180298,
+ "num_input_tokens_seen": 506470096,
+ "step": 11575,
+ "train_runtime": 89532.7906,
+ "train_tokens_per_second": 5656.811
+ },
+ {
+ "epoch": 0.7796929706436844,
+ "grad_norm": 0.6787888680159854,
+ "learning_rate": 6.696455704081006e-06,
+ "loss": 0.3787699699401855,
+ "num_input_tokens_seen": 506688376,
+ "step": 11580,
+ "train_runtime": 89574.3736,
+ "train_tokens_per_second": 5656.622
+ },
+ {
+ "epoch": 0.7800296256396445,
+ "grad_norm": 0.711221275891421,
+ "learning_rate": 6.693968228053416e-06,
+ "loss": 0.38877334594726565,
+ "num_input_tokens_seen": 506908432,
+ "step": 11585,
+ "train_runtime": 89609.8238,
+ "train_tokens_per_second": 5656.84
+ },
+ {
+ "epoch": 0.7803662806356046,
+ "grad_norm": 0.7045284132631514,
+ "learning_rate": 6.691480278313013e-06,
+ "loss": 0.37210679054260254,
+ "num_input_tokens_seen": 507123272,
+ "step": 11590,
+ "train_runtime": 89650.9072,
+ "train_tokens_per_second": 5656.644
+ },
+ {
+ "epoch": 0.7807029356315648,
+ "grad_norm": 0.718080559096589,
+ "learning_rate": 6.688991855555545e-06,
+ "loss": 0.35350847244262695,
+ "num_input_tokens_seen": 507339456,
+ "step": 11595,
+ "train_runtime": 89694.1289,
+ "train_tokens_per_second": 5656.328
+ },
+ {
+ "epoch": 0.7810395906275249,
+ "grad_norm": 0.692068019249801,
+ "learning_rate": 6.686502960476888e-06,
+ "loss": 0.3558246612548828,
+ "num_input_tokens_seen": 507577136,
+ "step": 11600,
+ "train_runtime": 89731.2991,
+ "train_tokens_per_second": 5656.634
+ },
+ {
+ "epoch": 0.7813762456234851,
+ "grad_norm": 0.7053421542399363,
+ "learning_rate": 6.684013593773058e-06,
+ "loss": 0.3307861089706421,
+ "num_input_tokens_seen": 507811288,
+ "step": 11605,
+ "train_runtime": 89766.4366,
+ "train_tokens_per_second": 5657.028
+ },
+ {
+ "epoch": 0.7817129006194452,
+ "grad_norm": 0.7619528494233906,
+ "learning_rate": 6.6815237561401955e-06,
+ "loss": 0.33051750659942625,
+ "num_input_tokens_seen": 508006304,
+ "step": 11610,
+ "train_runtime": 89802.2646,
+ "train_tokens_per_second": 5656.943
+ },
+ {
+ "epoch": 0.7820495556154053,
+ "grad_norm": 0.8056101436691785,
+ "learning_rate": 6.679033448274577e-06,
+ "loss": 0.35503997802734377,
+ "num_input_tokens_seen": 508223832,
+ "step": 11615,
+ "train_runtime": 89850.1477,
+ "train_tokens_per_second": 5656.349
+ },
+ {
+ "epoch": 0.7823862106113655,
+ "grad_norm": 0.6611022339527542,
+ "learning_rate": 6.676542670872609e-06,
+ "loss": 0.3701477527618408,
+ "num_input_tokens_seen": 508440024,
+ "step": 11620,
+ "train_runtime": 89890.5249,
+ "train_tokens_per_second": 5656.214
+ },
+ {
+ "epoch": 0.7827228656073256,
+ "grad_norm": 0.6851834640607047,
+ "learning_rate": 6.674051424630829e-06,
+ "loss": 0.36185245513916015,
+ "num_input_tokens_seen": 508666440,
+ "step": 11625,
+ "train_runtime": 89924.4436,
+ "train_tokens_per_second": 5656.598
+ },
+ {
+ "epoch": 0.7830595206032858,
+ "grad_norm": 0.7081381172789725,
+ "learning_rate": 6.671559710245905e-06,
+ "loss": 0.36898825168609617,
+ "num_input_tokens_seen": 508892304,
+ "step": 11630,
+ "train_runtime": 89970.0326,
+ "train_tokens_per_second": 5656.242
+ },
+ {
+ "epoch": 0.7833961755992459,
+ "grad_norm": 0.7864088827874502,
+ "learning_rate": 6.669067528414639e-06,
+ "loss": 0.36539783477783205,
+ "num_input_tokens_seen": 509118040,
+ "step": 11635,
+ "train_runtime": 90012.286,
+ "train_tokens_per_second": 5656.095
+ },
+ {
+ "epoch": 0.783732830595206,
+ "grad_norm": 0.7072947789302337,
+ "learning_rate": 6.6665748798339606e-06,
+ "loss": 0.39262051582336427,
+ "num_input_tokens_seen": 509346248,
+ "step": 11640,
+ "train_runtime": 90050.1585,
+ "train_tokens_per_second": 5656.25
+ },
+ {
+ "epoch": 0.7840694855911662,
+ "grad_norm": 0.7696238239442139,
+ "learning_rate": 6.664081765200932e-06,
+ "loss": 0.38922972679138185,
+ "num_input_tokens_seen": 509547192,
+ "step": 11645,
+ "train_runtime": 90082.5361,
+ "train_tokens_per_second": 5656.448
+ },
+ {
+ "epoch": 0.7844061405871263,
+ "grad_norm": 0.9564255408577432,
+ "learning_rate": 6.661588185212745e-06,
+ "loss": 0.38661224842071534,
+ "num_input_tokens_seen": 509767456,
+ "step": 11650,
+ "train_runtime": 90124.0819,
+ "train_tokens_per_second": 5656.285
+ },
+ {
+ "epoch": 0.7847427955830865,
+ "grad_norm": 0.6920498506689123,
+ "learning_rate": 6.65909414056672e-06,
+ "loss": 0.37398757934570315,
+ "num_input_tokens_seen": 509978544,
+ "step": 11655,
+ "train_runtime": 90162.88,
+ "train_tokens_per_second": 5656.192
+ },
+ {
+ "epoch": 0.7850794505790466,
+ "grad_norm": 0.7983355414544516,
+ "learning_rate": 6.6565996319603075e-06,
+ "loss": 0.36649136543273925,
+ "num_input_tokens_seen": 510194488,
+ "step": 11660,
+ "train_runtime": 90205.1793,
+ "train_tokens_per_second": 5655.933
+ },
+ {
+ "epoch": 0.7854161055750067,
+ "grad_norm": 0.705582165882806,
+ "learning_rate": 6.6541046600910915e-06,
+ "loss": 0.35945792198181153,
+ "num_input_tokens_seen": 510408368,
+ "step": 11665,
+ "train_runtime": 90242.6397,
+ "train_tokens_per_second": 5655.956
+ },
+ {
+ "epoch": 0.7857527605709669,
+ "grad_norm": 0.7659229146292734,
+ "learning_rate": 6.6516092256567824e-06,
+ "loss": 0.36875619888305666,
+ "num_input_tokens_seen": 510635536,
+ "step": 11670,
+ "train_runtime": 90279.7373,
+ "train_tokens_per_second": 5656.148
+ },
+ {
+ "epoch": 0.786089415566927,
+ "grad_norm": 0.8578375495363305,
+ "learning_rate": 6.64911332935522e-06,
+ "loss": 0.39651241302490237,
+ "num_input_tokens_seen": 510860904,
+ "step": 11675,
+ "train_runtime": 90320.5501,
+ "train_tokens_per_second": 5656.087
+ },
+ {
+ "epoch": 0.7864260705628872,
+ "grad_norm": 0.7134409162325845,
+ "learning_rate": 6.646616971884373e-06,
+ "loss": 0.3647106885910034,
+ "num_input_tokens_seen": 511085760,
+ "step": 11680,
+ "train_runtime": 90359.7652,
+ "train_tokens_per_second": 5656.121
+ },
+ {
+ "epoch": 0.7867627255588473,
+ "grad_norm": 0.6706622363930574,
+ "learning_rate": 6.64412015394234e-06,
+ "loss": 0.35194456577301025,
+ "num_input_tokens_seen": 511285624,
+ "step": 11685,
+ "train_runtime": 90397.4925,
+ "train_tokens_per_second": 5655.971
+ },
+ {
+ "epoch": 0.7870993805548074,
+ "grad_norm": 0.8444783042140578,
+ "learning_rate": 6.641622876227349e-06,
+ "loss": 0.3812446594238281,
+ "num_input_tokens_seen": 511491128,
+ "step": 11690,
+ "train_runtime": 90439.6501,
+ "train_tokens_per_second": 5655.607
+ },
+ {
+ "epoch": 0.7874360355507676,
+ "grad_norm": 0.7494745665567035,
+ "learning_rate": 6.639125139437756e-06,
+ "loss": 0.3623605012893677,
+ "num_input_tokens_seen": 511707944,
+ "step": 11695,
+ "train_runtime": 90482.4407,
+ "train_tokens_per_second": 5655.329
+ },
+ {
+ "epoch": 0.7877726905467277,
+ "grad_norm": 0.7902023749862866,
+ "learning_rate": 6.6366269442720446e-06,
+ "loss": 0.36579647064208987,
+ "num_input_tokens_seen": 511918704,
+ "step": 11700,
+ "train_runtime": 90522.6671,
+ "train_tokens_per_second": 5655.144
+ },
+ {
+ "epoch": 0.7881093455426879,
+ "grad_norm": 0.7084908330150312,
+ "learning_rate": 6.634128291428827e-06,
+ "loss": 0.4204265594482422,
+ "num_input_tokens_seen": 512130360,
+ "step": 11705,
+ "train_runtime": 90561.121,
+ "train_tokens_per_second": 5655.08
+ },
+ {
+ "epoch": 0.788446000538648,
+ "grad_norm": 0.7463863432263244,
+ "learning_rate": 6.631629181606842e-06,
+ "loss": 0.37847514152526857,
+ "num_input_tokens_seen": 512353072,
+ "step": 11710,
+ "train_runtime": 90604.8421,
+ "train_tokens_per_second": 5654.809
+ },
+ {
+ "epoch": 0.7887826555346081,
+ "grad_norm": 0.7252447688219217,
+ "learning_rate": 6.629129615504961e-06,
+ "loss": 0.3778711795806885,
+ "num_input_tokens_seen": 512578040,
+ "step": 11715,
+ "train_runtime": 90644.5653,
+ "train_tokens_per_second": 5654.813
+ },
+ {
+ "epoch": 0.7891193105305683,
+ "grad_norm": 0.6877935716049668,
+ "learning_rate": 6.626629593822178e-06,
+ "loss": 0.349456787109375,
+ "num_input_tokens_seen": 512795568,
+ "step": 11720,
+ "train_runtime": 90683.8764,
+ "train_tokens_per_second": 5654.76
+ },
+ {
+ "epoch": 0.7894559655265284,
+ "grad_norm": 0.7159875503391986,
+ "learning_rate": 6.624129117257616e-06,
+ "loss": 0.3877539157867432,
+ "num_input_tokens_seen": 513020704,
+ "step": 11725,
+ "train_runtime": 90721.6257,
+ "train_tokens_per_second": 5654.889
+ },
+ {
+ "epoch": 0.7897926205224886,
+ "grad_norm": 0.8198647095115159,
+ "learning_rate": 6.621628186510524e-06,
+ "loss": 0.37906270027160643,
+ "num_input_tokens_seen": 513238728,
+ "step": 11730,
+ "train_runtime": 90763.6807,
+ "train_tokens_per_second": 5654.671
+ },
+ {
+ "epoch": 0.7901292755184487,
+ "grad_norm": 0.6710230144644209,
+ "learning_rate": 6.619126802280281e-06,
+ "loss": 0.35970292091369627,
+ "num_input_tokens_seen": 513454240,
+ "step": 11735,
+ "train_runtime": 90806.8266,
+ "train_tokens_per_second": 5654.357
+ },
+ {
+ "epoch": 0.7904659305144088,
+ "grad_norm": 0.6839883499284557,
+ "learning_rate": 6.616624965266392e-06,
+ "loss": 0.3722025632858276,
+ "num_input_tokens_seen": 513674120,
+ "step": 11740,
+ "train_runtime": 90845.0403,
+ "train_tokens_per_second": 5654.399
+ },
+ {
+ "epoch": 0.790802585510369,
+ "grad_norm": 0.7422766392719585,
+ "learning_rate": 6.614122676168486e-06,
+ "loss": 0.39596946239471437,
+ "num_input_tokens_seen": 513889520,
+ "step": 11745,
+ "train_runtime": 90889.19,
+ "train_tokens_per_second": 5654.022
+ },
+ {
+ "epoch": 0.7911392405063291,
+ "grad_norm": 0.8574462720949246,
+ "learning_rate": 6.611619935686321e-06,
+ "loss": 0.34925665855407717,
+ "num_input_tokens_seen": 514094568,
+ "step": 11750,
+ "train_runtime": 90923.7862,
+ "train_tokens_per_second": 5654.126
+ },
+ {
+ "epoch": 0.7914758955022893,
+ "grad_norm": 0.7396446955629874,
+ "learning_rate": 6.609116744519779e-06,
+ "loss": 0.3882757663726807,
+ "num_input_tokens_seen": 514319376,
+ "step": 11755,
+ "train_runtime": 90961.7449,
+ "train_tokens_per_second": 5654.238
+ },
+ {
+ "epoch": 0.7918125504982494,
+ "grad_norm": 0.8047711473168563,
+ "learning_rate": 6.60661310336887e-06,
+ "loss": 0.3384627103805542,
+ "num_input_tokens_seen": 514527584,
+ "step": 11760,
+ "train_runtime": 91008.8104,
+ "train_tokens_per_second": 5653.602
+ },
+ {
+ "epoch": 0.7921492054942095,
+ "grad_norm": 0.7981581690317775,
+ "learning_rate": 6.604109012933732e-06,
+ "loss": 0.37465848922729494,
+ "num_input_tokens_seen": 514729464,
+ "step": 11765,
+ "train_runtime": 91040.9107,
+ "train_tokens_per_second": 5653.826
+ },
+ {
+ "epoch": 0.7924858604901697,
+ "grad_norm": 0.7333032087882904,
+ "learning_rate": 6.601604473914622e-06,
+ "loss": 0.39129223823547366,
+ "num_input_tokens_seen": 514948712,
+ "step": 11770,
+ "train_runtime": 91080.7609,
+ "train_tokens_per_second": 5653.759
+ },
+ {
+ "epoch": 0.7928225154861298,
+ "grad_norm": 0.8203529810583842,
+ "learning_rate": 6.599099487011928e-06,
+ "loss": 0.3703568458557129,
+ "num_input_tokens_seen": 515174152,
+ "step": 11775,
+ "train_runtime": 91123.7113,
+ "train_tokens_per_second": 5653.569
+ },
+ {
+ "epoch": 0.79315917048209,
+ "grad_norm": 0.7692920757506985,
+ "learning_rate": 6.596594052926163e-06,
+ "loss": 0.37824716567993166,
+ "num_input_tokens_seen": 515398616,
+ "step": 11780,
+ "train_runtime": 91161.9049,
+ "train_tokens_per_second": 5653.662
+ },
+ {
+ "epoch": 0.7934958254780501,
+ "grad_norm": 0.6620821907912957,
+ "learning_rate": 6.5940881723579595e-06,
+ "loss": 0.36647372245788573,
+ "num_input_tokens_seen": 515618208,
+ "step": 11785,
+ "train_runtime": 91199.6136,
+ "train_tokens_per_second": 5653.732
+ },
+ {
+ "epoch": 0.7938324804740102,
+ "grad_norm": 0.7308558763801775,
+ "learning_rate": 6.591581846008085e-06,
+ "loss": 0.34111249446868896,
+ "num_input_tokens_seen": 515838000,
+ "step": 11790,
+ "train_runtime": 91242.8853,
+ "train_tokens_per_second": 5653.46
+ },
+ {
+ "epoch": 0.7941691354699704,
+ "grad_norm": 0.7510405382733509,
+ "learning_rate": 6.589075074577421e-06,
+ "loss": 0.3614863634109497,
+ "num_input_tokens_seen": 516058872,
+ "step": 11795,
+ "train_runtime": 91288.1701,
+ "train_tokens_per_second": 5653.075
+ },
+ {
+ "epoch": 0.7945057904659305,
+ "grad_norm": 0.682167345785408,
+ "learning_rate": 6.5865678587669805e-06,
+ "loss": 0.35341219902038573,
+ "num_input_tokens_seen": 516288480,
+ "step": 11800,
+ "train_runtime": 91324.6105,
+ "train_tokens_per_second": 5653.334
+ },
+ {
+ "epoch": 0.7948424454618906,
+ "grad_norm": 0.8727392399581251,
+ "learning_rate": 6.584060199277897e-06,
+ "loss": 0.3922135829925537,
+ "num_input_tokens_seen": 516500456,
+ "step": 11805,
+ "train_runtime": 91356.4687,
+ "train_tokens_per_second": 5653.682
+ },
+ {
+ "epoch": 0.7951791004578508,
+ "grad_norm": 0.7263526124399556,
+ "learning_rate": 6.581552096811428e-06,
+ "loss": 0.3736285209655762,
+ "num_input_tokens_seen": 516717184,
+ "step": 11810,
+ "train_runtime": 91391.2945,
+ "train_tokens_per_second": 5653.899
+ },
+ {
+ "epoch": 0.7955157554538109,
+ "grad_norm": 0.7483107004591556,
+ "learning_rate": 6.57904355206896e-06,
+ "loss": 0.3753064632415771,
+ "num_input_tokens_seen": 516937256,
+ "step": 11815,
+ "train_runtime": 91433.3951,
+ "train_tokens_per_second": 5653.703
+ },
+ {
+ "epoch": 0.7958524104497711,
+ "grad_norm": 0.741351883349472,
+ "learning_rate": 6.576534565751997e-06,
+ "loss": 0.35328757762908936,
+ "num_input_tokens_seen": 517155784,
+ "step": 11820,
+ "train_runtime": 91472.4934,
+ "train_tokens_per_second": 5653.675
+ },
+ {
+ "epoch": 0.7961890654457312,
+ "grad_norm": 0.7392430177231608,
+ "learning_rate": 6.574025138562172e-06,
+ "loss": 0.3843714475631714,
+ "num_input_tokens_seen": 517361632,
+ "step": 11825,
+ "train_runtime": 91507.0999,
+ "train_tokens_per_second": 5653.787
+ },
+ {
+ "epoch": 0.7965257204416913,
+ "grad_norm": 0.6186336035373811,
+ "learning_rate": 6.571515271201234e-06,
+ "loss": 0.3463775634765625,
+ "num_input_tokens_seen": 517585680,
+ "step": 11830,
+ "train_runtime": 91547.5139,
+ "train_tokens_per_second": 5653.738
+ },
+ {
+ "epoch": 0.7968623754376515,
+ "grad_norm": 0.6495706347541744,
+ "learning_rate": 6.569004964371062e-06,
+ "loss": 0.3861451864242554,
+ "num_input_tokens_seen": 517810808,
+ "step": 11835,
+ "train_runtime": 91580.6678,
+ "train_tokens_per_second": 5654.15
+ },
+ {
+ "epoch": 0.7971990304336116,
+ "grad_norm": 0.7143293886304013,
+ "learning_rate": 6.566494218773656e-06,
+ "loss": 0.3636291265487671,
+ "num_input_tokens_seen": 518036944,
+ "step": 11840,
+ "train_runtime": 91623.7035,
+ "train_tokens_per_second": 5653.962
+ },
+ {
+ "epoch": 0.7975356854295718,
+ "grad_norm": 0.7702610503844717,
+ "learning_rate": 6.563983035111136e-06,
+ "loss": 0.376003098487854,
+ "num_input_tokens_seen": 518266216,
+ "step": 11845,
+ "train_runtime": 91660.207,
+ "train_tokens_per_second": 5654.212
+ },
+ {
+ "epoch": 0.7978723404255319,
+ "grad_norm": 0.7600176188341217,
+ "learning_rate": 6.561471414085748e-06,
+ "loss": 0.36989893913269045,
+ "num_input_tokens_seen": 518477336,
+ "step": 11850,
+ "train_runtime": 91696.4011,
+ "train_tokens_per_second": 5654.282
+ },
+ {
+ "epoch": 0.798208995421492,
+ "grad_norm": 0.7119305326521156,
+ "learning_rate": 6.558959356399859e-06,
+ "loss": 0.3683455944061279,
+ "num_input_tokens_seen": 518691120,
+ "step": 11855,
+ "train_runtime": 91735.153,
+ "train_tokens_per_second": 5654.224
+ },
+ {
+ "epoch": 0.7985456504174522,
+ "grad_norm": 0.6724381741463443,
+ "learning_rate": 6.5564468627559544e-06,
+ "loss": 0.3579908609390259,
+ "num_input_tokens_seen": 518897696,
+ "step": 11860,
+ "train_runtime": 91775.8953,
+ "train_tokens_per_second": 5653.965
+ },
+ {
+ "epoch": 0.7988823054134123,
+ "grad_norm": 0.7387321689727897,
+ "learning_rate": 6.5539339338566495e-06,
+ "loss": 0.38613746166229246,
+ "num_input_tokens_seen": 519112712,
+ "step": 11865,
+ "train_runtime": 91816.066,
+ "train_tokens_per_second": 5653.833
+ },
+ {
+ "epoch": 0.7992189604093725,
+ "grad_norm": 0.8155036594396956,
+ "learning_rate": 6.5514205704046754e-06,
+ "loss": 0.34462246894836424,
+ "num_input_tokens_seen": 519326360,
+ "step": 11870,
+ "train_runtime": 91857.9272,
+ "train_tokens_per_second": 5653.582
+ },
+ {
+ "epoch": 0.7995556154053326,
+ "grad_norm": 0.7554873227873684,
+ "learning_rate": 6.548906773102886e-06,
+ "loss": 0.34732234477996826,
+ "num_input_tokens_seen": 519538040,
+ "step": 11875,
+ "train_runtime": 91892.7088,
+ "train_tokens_per_second": 5653.746
+ },
+ {
+ "epoch": 0.7998922704012927,
+ "grad_norm": 0.6865110717768457,
+ "learning_rate": 6.546392542654257e-06,
+ "loss": 0.3555797576904297,
+ "num_input_tokens_seen": 519754440,
+ "step": 11880,
+ "train_runtime": 91931.8443,
+ "train_tokens_per_second": 5653.693
+ },
+ {
+ "epoch": 0.8002289253972529,
+ "grad_norm": 0.7775168324740175,
+ "learning_rate": 6.543877879761882e-06,
+ "loss": 0.39305408000946046,
+ "num_input_tokens_seen": 519971256,
+ "step": 11885,
+ "train_runtime": 91972.3549,
+ "train_tokens_per_second": 5653.56
+ },
+ {
+ "epoch": 0.800565580393213,
+ "grad_norm": 0.6425158999557459,
+ "learning_rate": 6.5413627851289825e-06,
+ "loss": 0.3501169681549072,
+ "num_input_tokens_seen": 520197464,
+ "step": 11890,
+ "train_runtime": 92010.0574,
+ "train_tokens_per_second": 5653.702
+ },
+ {
+ "epoch": 0.8009022353891732,
+ "grad_norm": 0.7227536117889709,
+ "learning_rate": 6.538847259458896e-06,
+ "loss": 0.3551483631134033,
+ "num_input_tokens_seen": 520422616,
+ "step": 11895,
+ "train_runtime": 92042.9843,
+ "train_tokens_per_second": 5654.126
+ },
+ {
+ "epoch": 0.8012388903851333,
+ "grad_norm": 0.7042669271875746,
+ "learning_rate": 6.5363313034550805e-06,
+ "loss": 0.34134459495544434,
+ "num_input_tokens_seen": 520633392,
+ "step": 11900,
+ "train_runtime": 92087.6293,
+ "train_tokens_per_second": 5653.674
+ },
+ {
+ "epoch": 0.8015755453810934,
+ "grad_norm": 0.739311915732445,
+ "learning_rate": 6.533814917821114e-06,
+ "loss": 0.3754891395568848,
+ "num_input_tokens_seen": 520855784,
+ "step": 11905,
+ "train_runtime": 92127.8168,
+ "train_tokens_per_second": 5653.621
+ },
+ {
+ "epoch": 0.8019122003770536,
+ "grad_norm": 0.6302684231473084,
+ "learning_rate": 6.531298103260696e-06,
+ "loss": 0.3585964202880859,
+ "num_input_tokens_seen": 521066736,
+ "step": 11910,
+ "train_runtime": 92163.5443,
+ "train_tokens_per_second": 5653.719
+ },
+ {
+ "epoch": 0.8022488553730137,
+ "grad_norm": 0.7417671721607515,
+ "learning_rate": 6.528780860477648e-06,
+ "loss": 0.3797733783721924,
+ "num_input_tokens_seen": 521300416,
+ "step": 11915,
+ "train_runtime": 92200.3318,
+ "train_tokens_per_second": 5653.997
+ },
+ {
+ "epoch": 0.8025855103689739,
+ "grad_norm": 0.6890068438400968,
+ "learning_rate": 6.526263190175907e-06,
+ "loss": 0.3375946044921875,
+ "num_input_tokens_seen": 521532280,
+ "step": 11920,
+ "train_runtime": 92234.4047,
+ "train_tokens_per_second": 5654.422
+ },
+ {
+ "epoch": 0.802922165364934,
+ "grad_norm": 0.6823445253166819,
+ "learning_rate": 6.523745093059531e-06,
+ "loss": 0.3637096643447876,
+ "num_input_tokens_seen": 521751944,
+ "step": 11925,
+ "train_runtime": 92276.588,
+ "train_tokens_per_second": 5654.218
+ },
+ {
+ "epoch": 0.8032588203608941,
+ "grad_norm": 0.8198308621507191,
+ "learning_rate": 6.521226569832699e-06,
+ "loss": 0.3677360057830811,
+ "num_input_tokens_seen": 521974016,
+ "step": 11930,
+ "train_runtime": 92311.2607,
+ "train_tokens_per_second": 5654.5
+ },
+ {
+ "epoch": 0.8035954753568543,
+ "grad_norm": 0.7478623966170844,
+ "learning_rate": 6.518707621199707e-06,
+ "loss": 0.355226993560791,
+ "num_input_tokens_seen": 522190040,
+ "step": 11935,
+ "train_runtime": 92353.1434,
+ "train_tokens_per_second": 5654.275
+ },
+ {
+ "epoch": 0.8039321303528144,
+ "grad_norm": 0.7995961908090801,
+ "learning_rate": 6.516188247864972e-06,
+ "loss": 0.3752752304077148,
+ "num_input_tokens_seen": 522418072,
+ "step": 11940,
+ "train_runtime": 92392.9504,
+ "train_tokens_per_second": 5654.307
+ },
+ {
+ "epoch": 0.8042687853487746,
+ "grad_norm": 0.8084497458000778,
+ "learning_rate": 6.513668450533028e-06,
+ "loss": 0.3633857250213623,
+ "num_input_tokens_seen": 522645296,
+ "step": 11945,
+ "train_runtime": 92428.2209,
+ "train_tokens_per_second": 5654.607
+ },
+ {
+ "epoch": 0.8046054403447347,
+ "grad_norm": 0.7480700245413793,
+ "learning_rate": 6.511148229908531e-06,
+ "loss": 0.39172661304473877,
+ "num_input_tokens_seen": 522868752,
+ "step": 11950,
+ "train_runtime": 92461.7106,
+ "train_tokens_per_second": 5654.976
+ },
+ {
+ "epoch": 0.8049420953406948,
+ "grad_norm": 0.8708977640452219,
+ "learning_rate": 6.508627586696248e-06,
+ "loss": 0.3619298696517944,
+ "num_input_tokens_seen": 523082504,
+ "step": 11955,
+ "train_runtime": 92499.8318,
+ "train_tokens_per_second": 5654.956
+ },
+ {
+ "epoch": 0.805278750336655,
+ "grad_norm": 0.7261115677502934,
+ "learning_rate": 6.5061065216010695e-06,
+ "loss": 0.3667879581451416,
+ "num_input_tokens_seen": 523299872,
+ "step": 11960,
+ "train_runtime": 92542.5023,
+ "train_tokens_per_second": 5654.698
+ },
+ {
+ "epoch": 0.8056154053326151,
+ "grad_norm": 0.7422593678216647,
+ "learning_rate": 6.5035850353280075e-06,
+ "loss": 0.3360764503479004,
+ "num_input_tokens_seen": 523516992,
+ "step": 11965,
+ "train_runtime": 92581.276,
+ "train_tokens_per_second": 5654.675
+ },
+ {
+ "epoch": 0.8059520603285752,
+ "grad_norm": 0.7763633318074254,
+ "learning_rate": 6.501063128582183e-06,
+ "loss": 0.35983996391296386,
+ "num_input_tokens_seen": 523724536,
+ "step": 11970,
+ "train_runtime": 92615.6401,
+ "train_tokens_per_second": 5654.817
+ },
+ {
+ "epoch": 0.8062887153245354,
+ "grad_norm": 0.8599327424362851,
+ "learning_rate": 6.498540802068843e-06,
+ "loss": 0.36988580226898193,
+ "num_input_tokens_seen": 523941648,
+ "step": 11975,
+ "train_runtime": 92650.751,
+ "train_tokens_per_second": 5655.018
+ },
+ {
+ "epoch": 0.8066253703204955,
+ "grad_norm": 0.7313407988142705,
+ "learning_rate": 6.4960180564933454e-06,
+ "loss": 0.3595852851867676,
+ "num_input_tokens_seen": 524162560,
+ "step": 11980,
+ "train_runtime": 92691.1408,
+ "train_tokens_per_second": 5654.937
+ },
+ {
+ "epoch": 0.8069620253164557,
+ "grad_norm": 0.7521095206194225,
+ "learning_rate": 6.493494892561167e-06,
+ "loss": 0.3594093084335327,
+ "num_input_tokens_seen": 524374312,
+ "step": 11985,
+ "train_runtime": 92728.3173,
+ "train_tokens_per_second": 5654.953
+ },
+ {
+ "epoch": 0.8072986803124158,
+ "grad_norm": 0.8430159653893533,
+ "learning_rate": 6.490971310977907e-06,
+ "loss": 0.3831813812255859,
+ "num_input_tokens_seen": 524596088,
+ "step": 11990,
+ "train_runtime": 92765.8417,
+ "train_tokens_per_second": 5655.057
+ },
+ {
+ "epoch": 0.807635335308376,
+ "grad_norm": 0.7203196408716581,
+ "learning_rate": 6.488447312449273e-06,
+ "loss": 0.384428596496582,
+ "num_input_tokens_seen": 524821680,
+ "step": 11995,
+ "train_runtime": 92796.5099,
+ "train_tokens_per_second": 5655.619
+ },
+ {
+ "epoch": 0.8079719903043361,
+ "grad_norm": 0.7501646695656614,
+ "learning_rate": 6.485922897681095e-06,
+ "loss": 0.3725404739379883,
+ "num_input_tokens_seen": 525045696,
+ "step": 12000,
+ "train_runtime": 92832.3227,
+ "train_tokens_per_second": 5655.85
+ },
+ {
+ "epoch": 0.8083086453002962,
+ "grad_norm": 0.6490994037223768,
+ "learning_rate": 6.483398067379316e-06,
+ "loss": 0.4056462287902832,
+ "num_input_tokens_seen": 525279952,
+ "step": 12005,
+ "train_runtime": 92862.4479,
+ "train_tokens_per_second": 5656.538
+ },
+ {
+ "epoch": 0.8086453002962564,
+ "grad_norm": 0.719666275525647,
+ "learning_rate": 6.480872822249998e-06,
+ "loss": 0.3556082248687744,
+ "num_input_tokens_seen": 525506864,
+ "step": 12010,
+ "train_runtime": 92904.5673,
+ "train_tokens_per_second": 5656.416
+ },
+ {
+ "epoch": 0.8089819552922165,
+ "grad_norm": 0.764344635124093,
+ "learning_rate": 6.478347162999318e-06,
+ "loss": 0.3839968919754028,
+ "num_input_tokens_seen": 525719016,
+ "step": 12015,
+ "train_runtime": 92947.2954,
+ "train_tokens_per_second": 5656.098
+ },
+ {
+ "epoch": 0.8093186102881766,
+ "grad_norm": 0.6701860780093635,
+ "learning_rate": 6.475821090333568e-06,
+ "loss": 0.3664535999298096,
+ "num_input_tokens_seen": 525939808,
+ "step": 12020,
+ "train_runtime": 92994.8165,
+ "train_tokens_per_second": 5655.582
+ },
+ {
+ "epoch": 0.8096552652841368,
+ "grad_norm": 0.7604615691324128,
+ "learning_rate": 6.473294604959156e-06,
+ "loss": 0.38412063121795653,
+ "num_input_tokens_seen": 526150200,
+ "step": 12025,
+ "train_runtime": 93031.4012,
+ "train_tokens_per_second": 5655.619
+ },
+ {
+ "epoch": 0.8099919202800969,
+ "grad_norm": 0.7314264087625333,
+ "learning_rate": 6.470767707582605e-06,
+ "loss": 0.3502906322479248,
+ "num_input_tokens_seen": 526371776,
+ "step": 12030,
+ "train_runtime": 93073.7526,
+ "train_tokens_per_second": 5655.427
+ },
+ {
+ "epoch": 0.8103285752760571,
+ "grad_norm": 0.8031066646434316,
+ "learning_rate": 6.468240398910555e-06,
+ "loss": 0.36304702758789065,
+ "num_input_tokens_seen": 526566912,
+ "step": 12035,
+ "train_runtime": 93113.568,
+ "train_tokens_per_second": 5655.104
+ },
+ {
+ "epoch": 0.8106652302720172,
+ "grad_norm": 0.7056043206827206,
+ "learning_rate": 6.4657126796497595e-06,
+ "loss": 0.3545745849609375,
+ "num_input_tokens_seen": 526798856,
+ "step": 12040,
+ "train_runtime": 93156.2331,
+ "train_tokens_per_second": 5655.004
+ },
+ {
+ "epoch": 0.8110018852679773,
+ "grad_norm": 0.8107764954705456,
+ "learning_rate": 6.463184550507087e-06,
+ "loss": 0.37404961585998536,
+ "num_input_tokens_seen": 527012296,
+ "step": 12045,
+ "train_runtime": 93201.4479,
+ "train_tokens_per_second": 5654.551
+ },
+ {
+ "epoch": 0.8113385402639375,
+ "grad_norm": 0.6987203165103377,
+ "learning_rate": 6.46065601218952e-06,
+ "loss": 0.34106624126434326,
+ "num_input_tokens_seen": 527230672,
+ "step": 12050,
+ "train_runtime": 93238.6997,
+ "train_tokens_per_second": 5654.633
+ },
+ {
+ "epoch": 0.8116751952598976,
+ "grad_norm": 0.6797008656627719,
+ "learning_rate": 6.458127065404156e-06,
+ "loss": 0.3812570095062256,
+ "num_input_tokens_seen": 527453272,
+ "step": 12055,
+ "train_runtime": 93278.2031,
+ "train_tokens_per_second": 5654.625
+ },
+ {
+ "epoch": 0.8120118502558578,
+ "grad_norm": 0.6832021023137033,
+ "learning_rate": 6.45559771085821e-06,
+ "loss": 0.35934345722198485,
+ "num_input_tokens_seen": 527679752,
+ "step": 12060,
+ "train_runtime": 93316.0381,
+ "train_tokens_per_second": 5654.759
+ },
+ {
+ "epoch": 0.8123485052518179,
+ "grad_norm": 0.7280405860607188,
+ "learning_rate": 6.453067949259005e-06,
+ "loss": 0.35833373069763186,
+ "num_input_tokens_seen": 527905280,
+ "step": 12065,
+ "train_runtime": 93350.8689,
+ "train_tokens_per_second": 5655.066
+ },
+ {
+ "epoch": 0.812685160247778,
+ "grad_norm": 0.7042450928577643,
+ "learning_rate": 6.450537781313982e-06,
+ "loss": 0.33934841156005857,
+ "num_input_tokens_seen": 528118504,
+ "step": 12070,
+ "train_runtime": 93398.5502,
+ "train_tokens_per_second": 5654.461
+ },
+ {
+ "epoch": 0.8130218152437382,
+ "grad_norm": 0.8291985738023927,
+ "learning_rate": 6.448007207730693e-06,
+ "loss": 0.3756300926208496,
+ "num_input_tokens_seen": 528340504,
+ "step": 12075,
+ "train_runtime": 93436.5515,
+ "train_tokens_per_second": 5654.538
+ },
+ {
+ "epoch": 0.8133584702396983,
+ "grad_norm": 0.7204057314942438,
+ "learning_rate": 6.445476229216805e-06,
+ "loss": 0.35174586772918703,
+ "num_input_tokens_seen": 528575776,
+ "step": 12080,
+ "train_runtime": 93474.5152,
+ "train_tokens_per_second": 5654.758
+ },
+ {
+ "epoch": 0.8136951252356585,
+ "grad_norm": 0.7511444963498415,
+ "learning_rate": 6.442944846480099e-06,
+ "loss": 0.39254226684570315,
+ "num_input_tokens_seen": 528786904,
+ "step": 12085,
+ "train_runtime": 93515.4145,
+ "train_tokens_per_second": 5654.543
+ },
+ {
+ "epoch": 0.8140317802316186,
+ "grad_norm": 0.7816657648101096,
+ "learning_rate": 6.44041306022847e-06,
+ "loss": 0.39618897438049316,
+ "num_input_tokens_seen": 529015944,
+ "step": 12090,
+ "train_runtime": 93554.4561,
+ "train_tokens_per_second": 5654.631
+ },
+ {
+ "epoch": 0.8143684352275787,
+ "grad_norm": 0.8141588683065977,
+ "learning_rate": 6.43788087116992e-06,
+ "loss": 0.39859035015106203,
+ "num_input_tokens_seen": 529246224,
+ "step": 12095,
+ "train_runtime": 93595.1858,
+ "train_tokens_per_second": 5654.631
+ },
+ {
+ "epoch": 0.8147050902235389,
+ "grad_norm": 0.7932908932160116,
+ "learning_rate": 6.43534828001257e-06,
+ "loss": 0.3658564567565918,
+ "num_input_tokens_seen": 529476248,
+ "step": 12100,
+ "train_runtime": 93626.3021,
+ "train_tokens_per_second": 5655.208
+ },
+ {
+ "epoch": 0.815041745219499,
+ "grad_norm": 0.79019755327484,
+ "learning_rate": 6.43281528746465e-06,
+ "loss": 0.372475266456604,
+ "num_input_tokens_seen": 529683080,
+ "step": 12105,
+ "train_runtime": 93660.6895,
+ "train_tokens_per_second": 5655.34
+ },
+ {
+ "epoch": 0.8153784002154592,
+ "grad_norm": 0.6339454129210152,
+ "learning_rate": 6.430281894234504e-06,
+ "loss": 0.36925411224365234,
+ "num_input_tokens_seen": 529906632,
+ "step": 12110,
+ "train_runtime": 93709.4206,
+ "train_tokens_per_second": 5654.785
+ },
+ {
+ "epoch": 0.8157150552114193,
+ "grad_norm": 0.8085146313091093,
+ "learning_rate": 6.427748101030587e-06,
+ "loss": 0.3794065237045288,
+ "num_input_tokens_seen": 530122608,
+ "step": 12115,
+ "train_runtime": 93746.5407,
+ "train_tokens_per_second": 5654.85
+ },
+ {
+ "epoch": 0.8160517102073794,
+ "grad_norm": 0.7189030374735652,
+ "learning_rate": 6.425213908561468e-06,
+ "loss": 0.3644280433654785,
+ "num_input_tokens_seen": 530344760,
+ "step": 12120,
+ "train_runtime": 93782.6096,
+ "train_tokens_per_second": 5655.044
+ },
+ {
+ "epoch": 0.8163883652033396,
+ "grad_norm": 0.7379513095206941,
+ "learning_rate": 6.422679317535823e-06,
+ "loss": 0.3801795721054077,
+ "num_input_tokens_seen": 530554320,
+ "step": 12125,
+ "train_runtime": 93818.6237,
+ "train_tokens_per_second": 5655.107
+ },
+ {
+ "epoch": 0.8167250201992997,
+ "grad_norm": 0.7361384056056562,
+ "learning_rate": 6.420144328662441e-06,
+ "loss": 0.4110711097717285,
+ "num_input_tokens_seen": 530784216,
+ "step": 12130,
+ "train_runtime": 93858.9763,
+ "train_tokens_per_second": 5655.125
+ },
+ {
+ "epoch": 0.8170616751952599,
+ "grad_norm": 0.6984222353139848,
+ "learning_rate": 6.417608942650229e-06,
+ "loss": 0.36923315525054934,
+ "num_input_tokens_seen": 530985400,
+ "step": 12135,
+ "train_runtime": 93900.3001,
+ "train_tokens_per_second": 5654.779
+ },
+ {
+ "epoch": 0.81739833019122,
+ "grad_norm": 0.7696394382831785,
+ "learning_rate": 6.415073160208196e-06,
+ "loss": 0.368359637260437,
+ "num_input_tokens_seen": 531219992,
+ "step": 12140,
+ "train_runtime": 93938.5272,
+ "train_tokens_per_second": 5654.975
+ },
+ {
+ "epoch": 0.8177349851871801,
+ "grad_norm": 0.8778188370382649,
+ "learning_rate": 6.412536982045465e-06,
+ "loss": 0.38031327724456787,
+ "num_input_tokens_seen": 531416856,
+ "step": 12145,
+ "train_runtime": 93975.0195,
+ "train_tokens_per_second": 5654.874
+ },
+ {
+ "epoch": 0.8180716401831403,
+ "grad_norm": 0.7668958438378366,
+ "learning_rate": 6.410000408871273e-06,
+ "loss": 0.3821054458618164,
+ "num_input_tokens_seen": 531631960,
+ "step": 12150,
+ "train_runtime": 94010.159,
+ "train_tokens_per_second": 5655.048
+ },
+ {
+ "epoch": 0.8184082951791004,
+ "grad_norm": 0.7888864971284683,
+ "learning_rate": 6.407463441394961e-06,
+ "loss": 0.37601168155670167,
+ "num_input_tokens_seen": 531859288,
+ "step": 12155,
+ "train_runtime": 94046.1948,
+ "train_tokens_per_second": 5655.298
+ },
+ {
+ "epoch": 0.8187449501750605,
+ "grad_norm": 0.8191097049266588,
+ "learning_rate": 6.4049260803259874e-06,
+ "loss": 0.38929071426391604,
+ "num_input_tokens_seen": 532082152,
+ "step": 12160,
+ "train_runtime": 94088.2586,
+ "train_tokens_per_second": 5655.139
+ },
+ {
+ "epoch": 0.8190816051710207,
+ "grad_norm": 0.8221152177897472,
+ "learning_rate": 6.402388326373914e-06,
+ "loss": 0.3606018304824829,
+ "num_input_tokens_seen": 532304432,
+ "step": 12165,
+ "train_runtime": 94128.8133,
+ "train_tokens_per_second": 5655.064
+ },
+ {
+ "epoch": 0.8194182601669808,
+ "grad_norm": 0.6644138882185266,
+ "learning_rate": 6.399850180248418e-06,
+ "loss": 0.4054881572723389,
+ "num_input_tokens_seen": 532538000,
+ "step": 12170,
+ "train_runtime": 94167.2148,
+ "train_tokens_per_second": 5655.238
+ },
+ {
+ "epoch": 0.819754915162941,
+ "grad_norm": 0.7674282579500541,
+ "learning_rate": 6.397311642659283e-06,
+ "loss": 0.3806436538696289,
+ "num_input_tokens_seen": 532749856,
+ "step": 12175,
+ "train_runtime": 94203.2208,
+ "train_tokens_per_second": 5655.325
+ },
+ {
+ "epoch": 0.8200915701589011,
+ "grad_norm": 0.7532047074504019,
+ "learning_rate": 6.394772714316402e-06,
+ "loss": 0.3875831842422485,
+ "num_input_tokens_seen": 532981096,
+ "step": 12180,
+ "train_runtime": 94238.5397,
+ "train_tokens_per_second": 5655.66
+ },
+ {
+ "epoch": 0.8204282251548612,
+ "grad_norm": 0.691733853853237,
+ "learning_rate": 6.392233395929779e-06,
+ "loss": 0.34835638999938967,
+ "num_input_tokens_seen": 533193808,
+ "step": 12185,
+ "train_runtime": 94267.7706,
+ "train_tokens_per_second": 5656.162
+ },
+ {
+ "epoch": 0.8207648801508214,
+ "grad_norm": 0.6752837485492823,
+ "learning_rate": 6.389693688209526e-06,
+ "loss": 0.362264609336853,
+ "num_input_tokens_seen": 533413896,
+ "step": 12190,
+ "train_runtime": 94307.5026,
+ "train_tokens_per_second": 5656.113
+ },
+ {
+ "epoch": 0.8211015351467815,
+ "grad_norm": 0.6640861040569711,
+ "learning_rate": 6.387153591865867e-06,
+ "loss": 0.3703956604003906,
+ "num_input_tokens_seen": 533642648,
+ "step": 12195,
+ "train_runtime": 94351.3757,
+ "train_tokens_per_second": 5655.907
+ },
+ {
+ "epoch": 0.8214381901427417,
+ "grad_norm": 0.8639133957294087,
+ "learning_rate": 6.384613107609127e-06,
+ "loss": 0.38781538009643557,
+ "num_input_tokens_seen": 533855672,
+ "step": 12200,
+ "train_runtime": 94388.037,
+ "train_tokens_per_second": 5655.968
+ },
+ {
+ "epoch": 0.8217748451387018,
+ "grad_norm": 0.6461164084347707,
+ "learning_rate": 6.382072236149747e-06,
+ "loss": 0.3636336326599121,
+ "num_input_tokens_seen": 534086688,
+ "step": 12205,
+ "train_runtime": 94418.101,
+ "train_tokens_per_second": 5656.613
+ },
+ {
+ "epoch": 0.8221115001346619,
+ "grad_norm": 1.259849346653607,
+ "learning_rate": 6.379530978198273e-06,
+ "loss": 0.33962430953979494,
+ "num_input_tokens_seen": 534295064,
+ "step": 12210,
+ "train_runtime": 94452.6928,
+ "train_tokens_per_second": 5656.748
+ },
+ {
+ "epoch": 0.8224481551306222,
+ "grad_norm": 0.6840892751944132,
+ "learning_rate": 6.376989334465361e-06,
+ "loss": 0.34957318305969237,
+ "num_input_tokens_seen": 534518376,
+ "step": 12215,
+ "train_runtime": 94487.4762,
+ "train_tokens_per_second": 5657.029
+ },
+ {
+ "epoch": 0.8227848101265823,
+ "grad_norm": 0.7080671395957931,
+ "learning_rate": 6.374447305661772e-06,
+ "loss": 0.359330415725708,
+ "num_input_tokens_seen": 534753352,
+ "step": 12220,
+ "train_runtime": 94526.7452,
+ "train_tokens_per_second": 5657.165
+ },
+ {
+ "epoch": 0.8231214651225425,
+ "grad_norm": 0.7000813266501178,
+ "learning_rate": 6.3719048924983776e-06,
+ "loss": 0.3536032199859619,
+ "num_input_tokens_seen": 534959200,
+ "step": 12225,
+ "train_runtime": 94565.9434,
+ "train_tokens_per_second": 5656.996
+ },
+ {
+ "epoch": 0.8234581201185026,
+ "grad_norm": 0.7684419131244044,
+ "learning_rate": 6.369362095686152e-06,
+ "loss": 0.3803597688674927,
+ "num_input_tokens_seen": 535175808,
+ "step": 12230,
+ "train_runtime": 94601.305,
+ "train_tokens_per_second": 5657.172
+ },
+ {
+ "epoch": 0.8237947751144628,
+ "grad_norm": 0.6924107525351223,
+ "learning_rate": 6.366818915936185e-06,
+ "loss": 0.34799120426177976,
+ "num_input_tokens_seen": 535394008,
+ "step": 12235,
+ "train_runtime": 94637.8962,
+ "train_tokens_per_second": 5657.29
+ },
+ {
+ "epoch": 0.8241314301104229,
+ "grad_norm": 0.7585078799576096,
+ "learning_rate": 6.364275353959667e-06,
+ "loss": 0.3779747486114502,
+ "num_input_tokens_seen": 535603560,
+ "step": 12240,
+ "train_runtime": 94679.4987,
+ "train_tokens_per_second": 5657.017
+ },
+ {
+ "epoch": 0.824468085106383,
+ "grad_norm": 0.7685591544772424,
+ "learning_rate": 6.3617314104678966e-06,
+ "loss": 0.36225066184997556,
+ "num_input_tokens_seen": 535838080,
+ "step": 12245,
+ "train_runtime": 94720.1071,
+ "train_tokens_per_second": 5657.068
+ },
+ {
+ "epoch": 0.8248047401023432,
+ "grad_norm": 0.76161666147799,
+ "learning_rate": 6.359187086172278e-06,
+ "loss": 0.3977229118347168,
+ "num_input_tokens_seen": 536041408,
+ "step": 12250,
+ "train_runtime": 94755.4024,
+ "train_tokens_per_second": 5657.107
+ },
+ {
+ "epoch": 0.8251413950983033,
+ "grad_norm": 0.6732965305210872,
+ "learning_rate": 6.356642381784326e-06,
+ "loss": 0.3828271389007568,
+ "num_input_tokens_seen": 536255712,
+ "step": 12255,
+ "train_runtime": 94789.5374,
+ "train_tokens_per_second": 5657.33
+ },
+ {
+ "epoch": 0.8254780500942634,
+ "grad_norm": 0.9216909752320883,
+ "learning_rate": 6.354097298015658e-06,
+ "loss": 0.3526939392089844,
+ "num_input_tokens_seen": 536450456,
+ "step": 12260,
+ "train_runtime": 94828.1952,
+ "train_tokens_per_second": 5657.078
+ },
+ {
+ "epoch": 0.8258147050902236,
+ "grad_norm": 0.7392662084519794,
+ "learning_rate": 6.3515518355779985e-06,
+ "loss": 0.3932209014892578,
+ "num_input_tokens_seen": 536655032,
+ "step": 12265,
+ "train_runtime": 94865.9619,
+ "train_tokens_per_second": 5656.982
+ },
+ {
+ "epoch": 0.8261513600861837,
+ "grad_norm": 0.630920453324516,
+ "learning_rate": 6.349005995183177e-06,
+ "loss": 0.3485422134399414,
+ "num_input_tokens_seen": 536883576,
+ "step": 12270,
+ "train_runtime": 94899.668,
+ "train_tokens_per_second": 5657.381
+ },
+ {
+ "epoch": 0.8264880150821439,
+ "grad_norm": 0.6584126043910694,
+ "learning_rate": 6.346459777543131e-06,
+ "loss": 0.3577850341796875,
+ "num_input_tokens_seen": 537102112,
+ "step": 12275,
+ "train_runtime": 94933.4199,
+ "train_tokens_per_second": 5657.672
+ },
+ {
+ "epoch": 0.826824670078104,
+ "grad_norm": 0.69670427508635,
+ "learning_rate": 6.343913183369902e-06,
+ "loss": 0.3586315155029297,
+ "num_input_tokens_seen": 537306360,
+ "step": 12280,
+ "train_runtime": 94970.6969,
+ "train_tokens_per_second": 5657.602
+ },
+ {
+ "epoch": 0.8271613250740641,
+ "grad_norm": 0.6795435309811332,
+ "learning_rate": 6.341366213375637e-06,
+ "loss": 0.3537776231765747,
+ "num_input_tokens_seen": 537534592,
+ "step": 12285,
+ "train_runtime": 95003.3121,
+ "train_tokens_per_second": 5658.062
+ },
+ {
+ "epoch": 0.8274979800700243,
+ "grad_norm": 0.7334092343426839,
+ "learning_rate": 6.338818868272587e-06,
+ "loss": 0.3713988780975342,
+ "num_input_tokens_seen": 537747240,
+ "step": 12290,
+ "train_runtime": 95041.1213,
+ "train_tokens_per_second": 5658.048
+ },
+ {
+ "epoch": 0.8278346350659844,
+ "grad_norm": 0.7029457386112722,
+ "learning_rate": 6.336271148773111e-06,
+ "loss": 0.3642918109893799,
+ "num_input_tokens_seen": 537982080,
+ "step": 12295,
+ "train_runtime": 95085.5253,
+ "train_tokens_per_second": 5657.876
+ },
+ {
+ "epoch": 0.8281712900619446,
+ "grad_norm": 0.664085080042037,
+ "learning_rate": 6.333723055589669e-06,
+ "loss": 0.4124462127685547,
+ "num_input_tokens_seen": 538202248,
+ "step": 12300,
+ "train_runtime": 95123.8409,
+ "train_tokens_per_second": 5657.911
+ },
+ {
+ "epoch": 0.8285079450579047,
+ "grad_norm": 0.7689236487654127,
+ "learning_rate": 6.331174589434826e-06,
+ "loss": 0.3763313293457031,
+ "num_input_tokens_seen": 538420056,
+ "step": 12305,
+ "train_runtime": 95170.7251,
+ "train_tokens_per_second": 5657.413
+ },
+ {
+ "epoch": 0.8288446000538648,
+ "grad_norm": 0.7341729953622874,
+ "learning_rate": 6.328625751021254e-06,
+ "loss": 0.35441150665283205,
+ "num_input_tokens_seen": 538627360,
+ "step": 12310,
+ "train_runtime": 95214.1667,
+ "train_tokens_per_second": 5657.009
+ },
+ {
+ "epoch": 0.829181255049825,
+ "grad_norm": 0.7056989368636907,
+ "learning_rate": 6.326076541061729e-06,
+ "loss": 0.334460711479187,
+ "num_input_tokens_seen": 538844328,
+ "step": 12315,
+ "train_runtime": 95255.2861,
+ "train_tokens_per_second": 5656.844
+ },
+ {
+ "epoch": 0.8295179100457851,
+ "grad_norm": 0.7430926488416755,
+ "learning_rate": 6.323526960269127e-06,
+ "loss": 0.3949697971343994,
+ "num_input_tokens_seen": 539054992,
+ "step": 12320,
+ "train_runtime": 95291.3412,
+ "train_tokens_per_second": 5656.915
+ },
+ {
+ "epoch": 0.8298545650417453,
+ "grad_norm": 0.7296769222033309,
+ "learning_rate": 6.3209770093564315e-06,
+ "loss": 0.3561330556869507,
+ "num_input_tokens_seen": 539280216,
+ "step": 12325,
+ "train_runtime": 95332.3002,
+ "train_tokens_per_second": 5656.847
+ },
+ {
+ "epoch": 0.8301912200377054,
+ "grad_norm": 0.8506709358501154,
+ "learning_rate": 6.318426689036727e-06,
+ "loss": 0.3534719705581665,
+ "num_input_tokens_seen": 539494040,
+ "step": 12330,
+ "train_runtime": 95376.0518,
+ "train_tokens_per_second": 5656.494
+ },
+ {
+ "epoch": 0.8305278750336655,
+ "grad_norm": 0.7341499431044767,
+ "learning_rate": 6.315876000023202e-06,
+ "loss": 0.37946810722351076,
+ "num_input_tokens_seen": 539706008,
+ "step": 12335,
+ "train_runtime": 95416.5574,
+ "train_tokens_per_second": 5656.314
+ },
+ {
+ "epoch": 0.8308645300296257,
+ "grad_norm": 3.744199811122827,
+ "learning_rate": 6.313324943029151e-06,
+ "loss": 0.3778734922409058,
+ "num_input_tokens_seen": 539917272,
+ "step": 12340,
+ "train_runtime": 95460.2877,
+ "train_tokens_per_second": 5655.936
+ },
+ {
+ "epoch": 0.8312011850255858,
+ "grad_norm": 0.742412638594704,
+ "learning_rate": 6.310773518767967e-06,
+ "loss": 0.35590362548828125,
+ "num_input_tokens_seen": 540130136,
+ "step": 12345,
+ "train_runtime": 95493.567,
+ "train_tokens_per_second": 5656.194
+ },
+ {
+ "epoch": 0.831537840021546,
+ "grad_norm": 0.702835896250019,
+ "learning_rate": 6.30822172795315e-06,
+ "loss": 0.36078743934631347,
+ "num_input_tokens_seen": 540336200,
+ "step": 12350,
+ "train_runtime": 95528.2473,
+ "train_tokens_per_second": 5656.298
+ },
+ {
+ "epoch": 0.8318744950175061,
+ "grad_norm": 0.7843017900068701,
+ "learning_rate": 6.3056695712982965e-06,
+ "loss": 0.4154792785644531,
+ "num_input_tokens_seen": 540561200,
+ "step": 12355,
+ "train_runtime": 95559.8481,
+ "train_tokens_per_second": 5656.782
+ },
+ {
+ "epoch": 0.8322111500134662,
+ "grad_norm": 0.6757106331074613,
+ "learning_rate": 6.303117049517111e-06,
+ "loss": 0.3399226427078247,
+ "num_input_tokens_seen": 540776592,
+ "step": 12360,
+ "train_runtime": 95600.3642,
+ "train_tokens_per_second": 5656.637
+ },
+ {
+ "epoch": 0.8325478050094264,
+ "grad_norm": 0.7135184795154971,
+ "learning_rate": 6.300564163323398e-06,
+ "loss": 0.35541400909423826,
+ "num_input_tokens_seen": 540989568,
+ "step": 12365,
+ "train_runtime": 95640.6371,
+ "train_tokens_per_second": 5656.482
+ },
+ {
+ "epoch": 0.8328844600053865,
+ "grad_norm": 0.7511263678823955,
+ "learning_rate": 6.298010913431065e-06,
+ "loss": 0.36992754936218264,
+ "num_input_tokens_seen": 541205456,
+ "step": 12370,
+ "train_runtime": 95679.064,
+ "train_tokens_per_second": 5656.467
+ },
+ {
+ "epoch": 0.8332211150013467,
+ "grad_norm": 0.735359416996797,
+ "learning_rate": 6.295457300554119e-06,
+ "loss": 0.3586021184921265,
+ "num_input_tokens_seen": 541434640,
+ "step": 12375,
+ "train_runtime": 95717.4462,
+ "train_tokens_per_second": 5656.593
+ },
+ {
+ "epoch": 0.8335577699973068,
+ "grad_norm": 0.8567636961499998,
+ "learning_rate": 6.292903325406668e-06,
+ "loss": 0.38977057933807374,
+ "num_input_tokens_seen": 541635304,
+ "step": 12380,
+ "train_runtime": 95751.5012,
+ "train_tokens_per_second": 5656.677
+ },
+ {
+ "epoch": 0.8338944249932669,
+ "grad_norm": 0.7502505399469903,
+ "learning_rate": 6.2903489887029255e-06,
+ "loss": 0.34544246196746825,
+ "num_input_tokens_seen": 541858656,
+ "step": 12385,
+ "train_runtime": 95786.9128,
+ "train_tokens_per_second": 5656.917
+ },
+ {
+ "epoch": 0.8342310799892271,
+ "grad_norm": 0.6987916307726929,
+ "learning_rate": 6.287794291157204e-06,
+ "loss": 0.36299178600311277,
+ "num_input_tokens_seen": 542076592,
+ "step": 12390,
+ "train_runtime": 95820.7633,
+ "train_tokens_per_second": 5657.193
+ },
+ {
+ "epoch": 0.8345677349851872,
+ "grad_norm": 0.8150952905885787,
+ "learning_rate": 6.285239233483915e-06,
+ "loss": 0.36720116138458253,
+ "num_input_tokens_seen": 542309048,
+ "step": 12395,
+ "train_runtime": 95859.9652,
+ "train_tokens_per_second": 5657.305
+ },
+ {
+ "epoch": 0.8349043899811474,
+ "grad_norm": 0.683112485827057,
+ "learning_rate": 6.2826838163975724e-06,
+ "loss": 0.3783855438232422,
+ "num_input_tokens_seen": 542526280,
+ "step": 12400,
+ "train_runtime": 95898.4045,
+ "train_tokens_per_second": 5657.302
+ },
+ {
+ "epoch": 0.8352410449771075,
+ "grad_norm": 0.8138279320478745,
+ "learning_rate": 6.2801280406127905e-06,
+ "loss": 0.35349900722503663,
+ "num_input_tokens_seen": 542750064,
+ "step": 12405,
+ "train_runtime": 95934.8168,
+ "train_tokens_per_second": 5657.488
+ },
+ {
+ "epoch": 0.8355776999730676,
+ "grad_norm": 0.7615550252521133,
+ "learning_rate": 6.2775719068442845e-06,
+ "loss": 0.36704421043395996,
+ "num_input_tokens_seen": 542968240,
+ "step": 12410,
+ "train_runtime": 95972.339,
+ "train_tokens_per_second": 5657.549
+ },
+ {
+ "epoch": 0.8359143549690278,
+ "grad_norm": 0.7219105847794507,
+ "learning_rate": 6.275015415806869e-06,
+ "loss": 0.3532416343688965,
+ "num_input_tokens_seen": 543175408,
+ "step": 12415,
+ "train_runtime": 96016.4623,
+ "train_tokens_per_second": 5657.107
+ },
+ {
+ "epoch": 0.8362510099649879,
+ "grad_norm": 0.6790545597022978,
+ "learning_rate": 6.272458568215458e-06,
+ "loss": 0.3773714303970337,
+ "num_input_tokens_seen": 543389992,
+ "step": 12420,
+ "train_runtime": 96057.0753,
+ "train_tokens_per_second": 5656.949
+ },
+ {
+ "epoch": 0.836587664960948,
+ "grad_norm": 0.758978760003948,
+ "learning_rate": 6.269901364785066e-06,
+ "loss": 0.37525081634521484,
+ "num_input_tokens_seen": 543616944,
+ "step": 12425,
+ "train_runtime": 96095.6171,
+ "train_tokens_per_second": 5657.042
+ },
+ {
+ "epoch": 0.8369243199569082,
+ "grad_norm": 0.6823393482935673,
+ "learning_rate": 6.2673438062308055e-06,
+ "loss": 0.37595961093902586,
+ "num_input_tokens_seen": 543859984,
+ "step": 12430,
+ "train_runtime": 96131.3734,
+ "train_tokens_per_second": 5657.466
+ },
+ {
+ "epoch": 0.8372609749528683,
+ "grad_norm": 0.9689604562670696,
+ "learning_rate": 6.264785893267892e-06,
+ "loss": 0.3472005367279053,
+ "num_input_tokens_seen": 544063976,
+ "step": 12435,
+ "train_runtime": 96162.9294,
+ "train_tokens_per_second": 5657.731
+ },
+ {
+ "epoch": 0.8375976299488285,
+ "grad_norm": 0.7213840812199924,
+ "learning_rate": 6.2622276266116366e-06,
+ "loss": 0.36200590133666993,
+ "num_input_tokens_seen": 544288640,
+ "step": 12440,
+ "train_runtime": 96199.7795,
+ "train_tokens_per_second": 5657.899
+ },
+ {
+ "epoch": 0.8379342849447886,
+ "grad_norm": 0.6737267934047665,
+ "learning_rate": 6.25966900697745e-06,
+ "loss": 0.37799394130706787,
+ "num_input_tokens_seen": 544514152,
+ "step": 12445,
+ "train_runtime": 96237.3914,
+ "train_tokens_per_second": 5658.031
+ },
+ {
+ "epoch": 0.8382709399407487,
+ "grad_norm": 0.6929403643911132,
+ "learning_rate": 6.257110035080843e-06,
+ "loss": 0.3491935729980469,
+ "num_input_tokens_seen": 544741824,
+ "step": 12450,
+ "train_runtime": 96274.7948,
+ "train_tokens_per_second": 5658.198
+ },
+ {
+ "epoch": 0.8386075949367089,
+ "grad_norm": 0.6942505611698238,
+ "learning_rate": 6.254550711637422e-06,
+ "loss": 0.40814781188964844,
+ "num_input_tokens_seen": 544950448,
+ "step": 12455,
+ "train_runtime": 96314.5199,
+ "train_tokens_per_second": 5658.03
+ },
+ {
+ "epoch": 0.838944249932669,
+ "grad_norm": 0.7575909904692931,
+ "learning_rate": 6.251991037362897e-06,
+ "loss": 0.38295414447784426,
+ "num_input_tokens_seen": 545159072,
+ "step": 12460,
+ "train_runtime": 96349.3253,
+ "train_tokens_per_second": 5658.151
+ },
+ {
+ "epoch": 0.8392809049286292,
+ "grad_norm": 0.7581123803912536,
+ "learning_rate": 6.249431012973068e-06,
+ "loss": 0.37713871002197263,
+ "num_input_tokens_seen": 545367120,
+ "step": 12465,
+ "train_runtime": 96388.6672,
+ "train_tokens_per_second": 5658.0
+ },
+ {
+ "epoch": 0.8396175599245893,
+ "grad_norm": 0.7683081862438146,
+ "learning_rate": 6.246870639183843e-06,
+ "loss": 0.37398107051849366,
+ "num_input_tokens_seen": 545578664,
+ "step": 12470,
+ "train_runtime": 96426.5472,
+ "train_tokens_per_second": 5657.972
+ },
+ {
+ "epoch": 0.8399542149205494,
+ "grad_norm": 0.8223266832633437,
+ "learning_rate": 6.2443099167112185e-06,
+ "loss": 0.39701476097106936,
+ "num_input_tokens_seen": 545803912,
+ "step": 12475,
+ "train_runtime": 96462.5666,
+ "train_tokens_per_second": 5658.194
+ },
+ {
+ "epoch": 0.8402908699165096,
+ "grad_norm": 0.6969953488947742,
+ "learning_rate": 6.241748846271293e-06,
+ "loss": 0.3778507709503174,
+ "num_input_tokens_seen": 546030536,
+ "step": 12480,
+ "train_runtime": 96496.5077,
+ "train_tokens_per_second": 5658.552
+ },
+ {
+ "epoch": 0.8406275249124697,
+ "grad_norm": 0.7778677325039889,
+ "learning_rate": 6.239187428580263e-06,
+ "loss": 0.37636547088623046,
+ "num_input_tokens_seen": 546236768,
+ "step": 12485,
+ "train_runtime": 96538.6051,
+ "train_tokens_per_second": 5658.221
+ },
+ {
+ "epoch": 0.8409641799084299,
+ "grad_norm": 0.8059997970084556,
+ "learning_rate": 6.236625664354421e-06,
+ "loss": 0.3784951686859131,
+ "num_input_tokens_seen": 546462096,
+ "step": 12490,
+ "train_runtime": 96574.6617,
+ "train_tokens_per_second": 5658.442
+ },
+ {
+ "epoch": 0.84130083490439,
+ "grad_norm": 0.8520532583827878,
+ "learning_rate": 6.2340635543101545e-06,
+ "loss": 0.3676713466644287,
+ "num_input_tokens_seen": 546657168,
+ "step": 12495,
+ "train_runtime": 96614.4259,
+ "train_tokens_per_second": 5658.132
+ },
+ {
+ "epoch": 0.8416374899003501,
+ "grad_norm": 0.6539078891022843,
+ "learning_rate": 6.23150109916395e-06,
+ "loss": 0.33987085819244384,
+ "num_input_tokens_seen": 546881208,
+ "step": 12500,
+ "train_runtime": 96655.3189,
+ "train_tokens_per_second": 5658.056
+ },
+ {
+ "epoch": 0.8419741448963103,
+ "grad_norm": 0.6624040098906347,
+ "learning_rate": 6.22893829963239e-06,
+ "loss": 0.3859384059906006,
+ "num_input_tokens_seen": 547092816,
+ "step": 12505,
+ "train_runtime": 96691.3494,
+ "train_tokens_per_second": 5658.136
+ },
+ {
+ "epoch": 0.8423107998922704,
+ "grad_norm": 0.7298942090178581,
+ "learning_rate": 6.226375156432153e-06,
+ "loss": 0.357137393951416,
+ "num_input_tokens_seen": 547322656,
+ "step": 12510,
+ "train_runtime": 96731.546,
+ "train_tokens_per_second": 5658.161
+ },
+ {
+ "epoch": 0.8426474548882306,
+ "grad_norm": 0.7316486599091492,
+ "learning_rate": 6.2238116702800145e-06,
+ "loss": 0.3710299491882324,
+ "num_input_tokens_seen": 547537944,
+ "step": 12515,
+ "train_runtime": 96762.4833,
+ "train_tokens_per_second": 5658.577
+ },
+ {
+ "epoch": 0.8429841098841907,
+ "grad_norm": 0.7998109937320629,
+ "learning_rate": 6.221247841892846e-06,
+ "loss": 0.3879734992980957,
+ "num_input_tokens_seen": 547746624,
+ "step": 12520,
+ "train_runtime": 96805.6441,
+ "train_tokens_per_second": 5658.21
+ },
+ {
+ "epoch": 0.8433207648801508,
+ "grad_norm": 0.7307238083679223,
+ "learning_rate": 6.218683671987611e-06,
+ "loss": 0.3691586971282959,
+ "num_input_tokens_seen": 547983928,
+ "step": 12525,
+ "train_runtime": 96840.3547,
+ "train_tokens_per_second": 5658.632
+ },
+ {
+ "epoch": 0.843657419876111,
+ "grad_norm": 0.7271187490915773,
+ "learning_rate": 6.216119161281373e-06,
+ "loss": 0.37589943408966064,
+ "num_input_tokens_seen": 548209600,
+ "step": 12530,
+ "train_runtime": 96879.0798,
+ "train_tokens_per_second": 5658.699
+ },
+ {
+ "epoch": 0.8439940748720711,
+ "grad_norm": 0.7311632772691732,
+ "learning_rate": 6.213554310491291e-06,
+ "loss": 0.3672966003417969,
+ "num_input_tokens_seen": 548421688,
+ "step": 12535,
+ "train_runtime": 96928.6015,
+ "train_tokens_per_second": 5657.996
+ },
+ {
+ "epoch": 0.8443307298680313,
+ "grad_norm": 0.7450184498891292,
+ "learning_rate": 6.210989120334616e-06,
+ "loss": 0.362609338760376,
+ "num_input_tokens_seen": 548644072,
+ "step": 12540,
+ "train_runtime": 96971.5458,
+ "train_tokens_per_second": 5657.784
+ },
+ {
+ "epoch": 0.8446673848639914,
+ "grad_norm": 0.7758427572495085,
+ "learning_rate": 6.208423591528694e-06,
+ "loss": 0.38643317222595214,
+ "num_input_tokens_seen": 548854336,
+ "step": 12545,
+ "train_runtime": 97009.7887,
+ "train_tokens_per_second": 5657.721
+ },
+ {
+ "epoch": 0.8450040398599515,
+ "grad_norm": 0.688742133220875,
+ "learning_rate": 6.205857724790969e-06,
+ "loss": 0.3612468481063843,
+ "num_input_tokens_seen": 549072600,
+ "step": 12550,
+ "train_runtime": 97048.5889,
+ "train_tokens_per_second": 5657.708
+ },
+ {
+ "epoch": 0.8453406948559117,
+ "grad_norm": 0.7926232068697935,
+ "learning_rate": 6.203291520838974e-06,
+ "loss": 0.409790563583374,
+ "num_input_tokens_seen": 549303984,
+ "step": 12555,
+ "train_runtime": 97086.2605,
+ "train_tokens_per_second": 5657.896
+ },
+ {
+ "epoch": 0.8456773498518718,
+ "grad_norm": 0.7041401739211672,
+ "learning_rate": 6.200724980390344e-06,
+ "loss": 0.34305951595306394,
+ "num_input_tokens_seen": 549497696,
+ "step": 12560,
+ "train_runtime": 97125.3875,
+ "train_tokens_per_second": 5657.611
+ },
+ {
+ "epoch": 0.846014004847832,
+ "grad_norm": 0.6959581809215943,
+ "learning_rate": 6.198158104162804e-06,
+ "loss": 0.3178701400756836,
+ "num_input_tokens_seen": 549713600,
+ "step": 12565,
+ "train_runtime": 97158.9336,
+ "train_tokens_per_second": 5657.88
+ },
+ {
+ "epoch": 0.8463506598437921,
+ "grad_norm": 0.6766745725154476,
+ "learning_rate": 6.1955908928741685e-06,
+ "loss": 0.37602977752685546,
+ "num_input_tokens_seen": 549953184,
+ "step": 12570,
+ "train_runtime": 97195.8557,
+ "train_tokens_per_second": 5658.196
+ },
+ {
+ "epoch": 0.8466873148397522,
+ "grad_norm": 0.6881093744203187,
+ "learning_rate": 6.193023347242353e-06,
+ "loss": 0.38083457946777344,
+ "num_input_tokens_seen": 550184352,
+ "step": 12575,
+ "train_runtime": 97232.314,
+ "train_tokens_per_second": 5658.452
+ },
+ {
+ "epoch": 0.8470239698357124,
+ "grad_norm": 0.741931440728507,
+ "learning_rate": 6.190455467985361e-06,
+ "loss": 0.3633177042007446,
+ "num_input_tokens_seen": 550405552,
+ "step": 12580,
+ "train_runtime": 97272.8862,
+ "train_tokens_per_second": 5658.366
+ },
+ {
+ "epoch": 0.8473606248316725,
+ "grad_norm": 0.7792859259983645,
+ "learning_rate": 6.187887255821295e-06,
+ "loss": 0.3954836368560791,
+ "num_input_tokens_seen": 550625896,
+ "step": 12585,
+ "train_runtime": 97305.8601,
+ "train_tokens_per_second": 5658.713
+ },
+ {
+ "epoch": 0.8476972798276327,
+ "grad_norm": 0.7394004400669401,
+ "learning_rate": 6.185318711468345e-06,
+ "loss": 0.3501850128173828,
+ "num_input_tokens_seen": 550827088,
+ "step": 12590,
+ "train_runtime": 97346.1248,
+ "train_tokens_per_second": 5658.439
+ },
+ {
+ "epoch": 0.8480339348235928,
+ "grad_norm": 0.7906419360212633,
+ "learning_rate": 6.182749835644799e-06,
+ "loss": 0.3587897777557373,
+ "num_input_tokens_seen": 551036712,
+ "step": 12595,
+ "train_runtime": 97378.9075,
+ "train_tokens_per_second": 5658.687
+ },
+ {
+ "epoch": 0.8483705898195529,
+ "grad_norm": 0.6720676543784547,
+ "learning_rate": 6.180180629069031e-06,
+ "loss": 0.3534966468811035,
+ "num_input_tokens_seen": 551269008,
+ "step": 12600,
+ "train_runtime": 97416.9713,
+ "train_tokens_per_second": 5658.86
+ },
+ {
+ "epoch": 0.8487072448155131,
+ "grad_norm": 1.0844161686609408,
+ "learning_rate": 6.177611092459514e-06,
+ "loss": 0.3404214859008789,
+ "num_input_tokens_seen": 551486584,
+ "step": 12605,
+ "train_runtime": 97454.9229,
+ "train_tokens_per_second": 5658.889
+ },
+ {
+ "epoch": 0.8490438998114732,
+ "grad_norm": 0.8339846261840644,
+ "learning_rate": 6.175041226534809e-06,
+ "loss": 0.3584805965423584,
+ "num_input_tokens_seen": 551707952,
+ "step": 12610,
+ "train_runtime": 97487.377,
+ "train_tokens_per_second": 5659.276
+ },
+ {
+ "epoch": 0.8493805548074334,
+ "grad_norm": 0.688269893800755,
+ "learning_rate": 6.172471032013574e-06,
+ "loss": 0.327656888961792,
+ "num_input_tokens_seen": 551932560,
+ "step": 12615,
+ "train_runtime": 97526.9729,
+ "train_tokens_per_second": 5659.281
+ },
+ {
+ "epoch": 0.8497172098033935,
+ "grad_norm": 0.6595384550469138,
+ "learning_rate": 6.169900509614553e-06,
+ "loss": 0.3601629972457886,
+ "num_input_tokens_seen": 552163800,
+ "step": 12620,
+ "train_runtime": 97577.6237,
+ "train_tokens_per_second": 5658.713
+ },
+ {
+ "epoch": 0.8500538647993536,
+ "grad_norm": 0.8030731667400421,
+ "learning_rate": 6.167329660056585e-06,
+ "loss": 0.36720144748687744,
+ "num_input_tokens_seen": 552375472,
+ "step": 12625,
+ "train_runtime": 97617.7309,
+ "train_tokens_per_second": 5658.557
+ },
+ {
+ "epoch": 0.8503905197953138,
+ "grad_norm": 0.6580730799233167,
+ "learning_rate": 6.1647584840586e-06,
+ "loss": 0.36503446102142334,
+ "num_input_tokens_seen": 552599576,
+ "step": 12630,
+ "train_runtime": 97665.6991,
+ "train_tokens_per_second": 5658.072
+ },
+ {
+ "epoch": 0.8507271747912739,
+ "grad_norm": 0.7006154386468065,
+ "learning_rate": 6.1621869823396175e-06,
+ "loss": 0.37128496170043945,
+ "num_input_tokens_seen": 552832696,
+ "step": 12635,
+ "train_runtime": 97703.9353,
+ "train_tokens_per_second": 5658.244
+ },
+ {
+ "epoch": 0.851063829787234,
+ "grad_norm": 0.7225238597915238,
+ "learning_rate": 6.159615155618752e-06,
+ "loss": 0.37408173084259033,
+ "num_input_tokens_seen": 553059920,
+ "step": 12640,
+ "train_runtime": 97737.0325,
+ "train_tokens_per_second": 5658.653
+ },
+ {
+ "epoch": 0.8514004847831942,
+ "grad_norm": 0.8136467242343185,
+ "learning_rate": 6.157043004615207e-06,
+ "loss": 0.38596835136413576,
+ "num_input_tokens_seen": 553300640,
+ "step": 12645,
+ "train_runtime": 97780.3263,
+ "train_tokens_per_second": 5658.609
+ },
+ {
+ "epoch": 0.8517371397791543,
+ "grad_norm": 0.7939080541143891,
+ "learning_rate": 6.1544705300482755e-06,
+ "loss": 0.3616677761077881,
+ "num_input_tokens_seen": 553532032,
+ "step": 12650,
+ "train_runtime": 97820.9636,
+ "train_tokens_per_second": 5658.624
+ },
+ {
+ "epoch": 0.8520737947751145,
+ "grad_norm": 0.6610977560801387,
+ "learning_rate": 6.151897732637341e-06,
+ "loss": 0.3330955982208252,
+ "num_input_tokens_seen": 553739416,
+ "step": 12655,
+ "train_runtime": 97857.6515,
+ "train_tokens_per_second": 5658.622
+ },
+ {
+ "epoch": 0.8524104497710746,
+ "grad_norm": 0.8508497854348367,
+ "learning_rate": 6.149324613101877e-06,
+ "loss": 0.37527806758880616,
+ "num_input_tokens_seen": 553959976,
+ "step": 12660,
+ "train_runtime": 97897.0387,
+ "train_tokens_per_second": 5658.598
+ },
+ {
+ "epoch": 0.8527471047670347,
+ "grad_norm": 0.6193562313644957,
+ "learning_rate": 6.146751172161453e-06,
+ "loss": 0.33863201141357424,
+ "num_input_tokens_seen": 554189776,
+ "step": 12665,
+ "train_runtime": 97940.3461,
+ "train_tokens_per_second": 5658.442
+ },
+ {
+ "epoch": 0.8530837597629949,
+ "grad_norm": 0.7796518940843955,
+ "learning_rate": 6.144177410535719e-06,
+ "loss": 0.3652313470840454,
+ "num_input_tokens_seen": 554414232,
+ "step": 12670,
+ "train_runtime": 97979.2461,
+ "train_tokens_per_second": 5658.486
+ },
+ {
+ "epoch": 0.853420414758955,
+ "grad_norm": 0.7186651171193743,
+ "learning_rate": 6.14160332894442e-06,
+ "loss": 0.3651700973510742,
+ "num_input_tokens_seen": 554625192,
+ "step": 12675,
+ "train_runtime": 98015.4768,
+ "train_tokens_per_second": 5658.547
+ },
+ {
+ "epoch": 0.8537570697549152,
+ "grad_norm": 0.724709697937217,
+ "learning_rate": 6.139028928107391e-06,
+ "loss": 0.36736459732055665,
+ "num_input_tokens_seen": 554842240,
+ "step": 12680,
+ "train_runtime": 98056.2763,
+ "train_tokens_per_second": 5658.406
+ },
+ {
+ "epoch": 0.8540937247508753,
+ "grad_norm": 0.7018554477013618,
+ "learning_rate": 6.136454208744553e-06,
+ "loss": 0.3479247808456421,
+ "num_input_tokens_seen": 555060144,
+ "step": 12685,
+ "train_runtime": 98097.7581,
+ "train_tokens_per_second": 5658.235
+ },
+ {
+ "epoch": 0.8544303797468354,
+ "grad_norm": 0.7210589341441987,
+ "learning_rate": 6.133879171575921e-06,
+ "loss": 0.3790309906005859,
+ "num_input_tokens_seen": 555276648,
+ "step": 12690,
+ "train_runtime": 98135.4374,
+ "train_tokens_per_second": 5658.268
+ },
+ {
+ "epoch": 0.8547670347427956,
+ "grad_norm": 0.8012051731436295,
+ "learning_rate": 6.131303817321595e-06,
+ "loss": 0.3976943254470825,
+ "num_input_tokens_seen": 555506600,
+ "step": 12695,
+ "train_runtime": 98173.0099,
+ "train_tokens_per_second": 5658.445
+ },
+ {
+ "epoch": 0.8551036897387557,
+ "grad_norm": 0.6686702415410488,
+ "learning_rate": 6.128728146701763e-06,
+ "loss": 0.3907043933868408,
+ "num_input_tokens_seen": 555735640,
+ "step": 12700,
+ "train_runtime": 98208.5683,
+ "train_tokens_per_second": 5658.729
+ },
+ {
+ "epoch": 0.8554403447347159,
+ "grad_norm": 0.6745127609405563,
+ "learning_rate": 6.126152160436705e-06,
+ "loss": 0.3401200294494629,
+ "num_input_tokens_seen": 555954936,
+ "step": 12705,
+ "train_runtime": 98247.4191,
+ "train_tokens_per_second": 5658.723
+ },
+ {
+ "epoch": 0.855776999730676,
+ "grad_norm": 0.9952487891058109,
+ "learning_rate": 6.123575859246784e-06,
+ "loss": 0.3820429801940918,
+ "num_input_tokens_seen": 556162568,
+ "step": 12710,
+ "train_runtime": 98285.6421,
+ "train_tokens_per_second": 5658.635
+ },
+ {
+ "epoch": 0.8561136547266361,
+ "grad_norm": 0.7343846388254841,
+ "learning_rate": 6.120999243852459e-06,
+ "loss": 0.3488877773284912,
+ "num_input_tokens_seen": 556374088,
+ "step": 12715,
+ "train_runtime": 98325.3673,
+ "train_tokens_per_second": 5658.5
+ },
+ {
+ "epoch": 0.8564503097225963,
+ "grad_norm": 0.8403978577775663,
+ "learning_rate": 6.118422314974269e-06,
+ "loss": 0.3801596164703369,
+ "num_input_tokens_seen": 556596000,
+ "step": 12720,
+ "train_runtime": 98368.0103,
+ "train_tokens_per_second": 5658.303
+ },
+ {
+ "epoch": 0.8567869647185564,
+ "grad_norm": 0.7544062671681226,
+ "learning_rate": 6.115845073332845e-06,
+ "loss": 0.36686029434204104,
+ "num_input_tokens_seen": 556828456,
+ "step": 12725,
+ "train_runtime": 98413.3452,
+ "train_tokens_per_second": 5658.058
+ },
+ {
+ "epoch": 0.8571236197145166,
+ "grad_norm": 0.8566960890477198,
+ "learning_rate": 6.113267519648905e-06,
+ "loss": 0.39199421405792234,
+ "num_input_tokens_seen": 557067040,
+ "step": 12730,
+ "train_runtime": 98457.4483,
+ "train_tokens_per_second": 5657.947
+ },
+ {
+ "epoch": 0.8574602747104767,
+ "grad_norm": 0.6753240704632734,
+ "learning_rate": 6.110689654643253e-06,
+ "loss": 0.36382412910461426,
+ "num_input_tokens_seen": 557274264,
+ "step": 12735,
+ "train_runtime": 98497.8079,
+ "train_tokens_per_second": 5657.733
+ },
+ {
+ "epoch": 0.8577969297064368,
+ "grad_norm": 0.7623447924884228,
+ "learning_rate": 6.1081114790367805e-06,
+ "loss": 0.37306723594665525,
+ "num_input_tokens_seen": 557484064,
+ "step": 12740,
+ "train_runtime": 98531.9864,
+ "train_tokens_per_second": 5657.899
+ },
+ {
+ "epoch": 0.858133584702397,
+ "grad_norm": 0.7690727585148884,
+ "learning_rate": 6.105532993550467e-06,
+ "loss": 0.37079343795776365,
+ "num_input_tokens_seen": 557707168,
+ "step": 12745,
+ "train_runtime": 98570.6779,
+ "train_tokens_per_second": 5657.942
+ },
+ {
+ "epoch": 0.8584702396983571,
+ "grad_norm": 0.7227829734774792,
+ "learning_rate": 6.1029541989053765e-06,
+ "loss": 0.3675063610076904,
+ "num_input_tokens_seen": 557923328,
+ "step": 12750,
+ "train_runtime": 98611.8723,
+ "train_tokens_per_second": 5657.77
+ },
+ {
+ "epoch": 0.8588068946943173,
+ "grad_norm": 0.6216338336006675,
+ "learning_rate": 6.100375095822661e-06,
+ "loss": 0.3660135746002197,
+ "num_input_tokens_seen": 558141912,
+ "step": 12755,
+ "train_runtime": 98653.6004,
+ "train_tokens_per_second": 5657.593
+ },
+ {
+ "epoch": 0.8591435496902774,
+ "grad_norm": 0.7752180098632147,
+ "learning_rate": 6.097795685023558e-06,
+ "loss": 0.3654344081878662,
+ "num_input_tokens_seen": 558363664,
+ "step": 12760,
+ "train_runtime": 98689.3362,
+ "train_tokens_per_second": 5657.791
+ },
+ {
+ "epoch": 0.8594802046862375,
+ "grad_norm": 0.7139656897700937,
+ "learning_rate": 6.095215967229396e-06,
+ "loss": 0.3468952655792236,
+ "num_input_tokens_seen": 558569128,
+ "step": 12765,
+ "train_runtime": 98726.8939,
+ "train_tokens_per_second": 5657.72
+ },
+ {
+ "epoch": 0.8598168596821977,
+ "grad_norm": 0.6751611340302733,
+ "learning_rate": 6.092635943161578e-06,
+ "loss": 0.37009525299072266,
+ "num_input_tokens_seen": 558798144,
+ "step": 12770,
+ "train_runtime": 98770.45,
+ "train_tokens_per_second": 5657.544
+ },
+ {
+ "epoch": 0.8601535146781578,
+ "grad_norm": 0.7202350000001299,
+ "learning_rate": 6.090055613541603e-06,
+ "loss": 0.34846246242523193,
+ "num_input_tokens_seen": 559026000,
+ "step": 12775,
+ "train_runtime": 98813.4858,
+ "train_tokens_per_second": 5657.386
+ },
+ {
+ "epoch": 0.860490169674118,
+ "grad_norm": 0.7513541698981121,
+ "learning_rate": 6.087474979091052e-06,
+ "loss": 0.3586388111114502,
+ "num_input_tokens_seen": 559233040,
+ "step": 12780,
+ "train_runtime": 98860.409,
+ "train_tokens_per_second": 5656.795
+ },
+ {
+ "epoch": 0.8608268246700781,
+ "grad_norm": 0.7207910613458428,
+ "learning_rate": 6.084894040531591e-06,
+ "loss": 0.3598465919494629,
+ "num_input_tokens_seen": 559456240,
+ "step": 12785,
+ "train_runtime": 98903.5566,
+ "train_tokens_per_second": 5656.584
+ },
+ {
+ "epoch": 0.8611634796660382,
+ "grad_norm": 0.7450858357428686,
+ "learning_rate": 6.0823127985849705e-06,
+ "loss": 0.35327887535095215,
+ "num_input_tokens_seen": 559662696,
+ "step": 12790,
+ "train_runtime": 98942.3562,
+ "train_tokens_per_second": 5656.452
+ },
+ {
+ "epoch": 0.8615001346619984,
+ "grad_norm": 0.7413157970110851,
+ "learning_rate": 6.079731253973028e-06,
+ "loss": 0.3576143741607666,
+ "num_input_tokens_seen": 559886624,
+ "step": 12795,
+ "train_runtime": 98988.6398,
+ "train_tokens_per_second": 5656.069
+ },
+ {
+ "epoch": 0.8618367896579585,
+ "grad_norm": 0.7990871684605166,
+ "learning_rate": 6.077149407417683e-06,
+ "loss": 0.36043851375579833,
+ "num_input_tokens_seen": 560101432,
+ "step": 12800,
+ "train_runtime": 99024.1718,
+ "train_tokens_per_second": 5656.209
+ },
+ {
+ "epoch": 0.8621734446539187,
+ "grad_norm": 0.7364962437233615,
+ "learning_rate": 6.07456725964094e-06,
+ "loss": 0.3793959617614746,
+ "num_input_tokens_seen": 560309472,
+ "step": 12805,
+ "train_runtime": 99059.299,
+ "train_tokens_per_second": 5656.304
+ },
+ {
+ "epoch": 0.8625100996498788,
+ "grad_norm": 0.680598341183089,
+ "learning_rate": 6.07198481136489e-06,
+ "loss": 0.3436178684234619,
+ "num_input_tokens_seen": 560520408,
+ "step": 12810,
+ "train_runtime": 99092.681,
+ "train_tokens_per_second": 5656.527
+ },
+ {
+ "epoch": 0.8628467546458389,
+ "grad_norm": 0.765274818340009,
+ "learning_rate": 6.069402063311706e-06,
+ "loss": 0.3549325704574585,
+ "num_input_tokens_seen": 560731064,
+ "step": 12815,
+ "train_runtime": 99135.7209,
+ "train_tokens_per_second": 5656.196
+ },
+ {
+ "epoch": 0.8631834096417991,
+ "grad_norm": 0.8395744171280569,
+ "learning_rate": 6.066819016203646e-06,
+ "loss": 0.36306018829345704,
+ "num_input_tokens_seen": 560944360,
+ "step": 12820,
+ "train_runtime": 99173.0123,
+ "train_tokens_per_second": 5656.22
+ },
+ {
+ "epoch": 0.8635200646377592,
+ "grad_norm": 0.7436029507337527,
+ "learning_rate": 6.064235670763048e-06,
+ "loss": 0.36905884742736816,
+ "num_input_tokens_seen": 561164928,
+ "step": 12825,
+ "train_runtime": 99209.242,
+ "train_tokens_per_second": 5656.378
+ },
+ {
+ "epoch": 0.8638567196337193,
+ "grad_norm": 0.6763369772275482,
+ "learning_rate": 6.06165202771234e-06,
+ "loss": 0.3765472412109375,
+ "num_input_tokens_seen": 561394056,
+ "step": 12830,
+ "train_runtime": 99250.4089,
+ "train_tokens_per_second": 5656.34
+ },
+ {
+ "epoch": 0.8641933746296795,
+ "grad_norm": 0.6931570065850123,
+ "learning_rate": 6.059068087774026e-06,
+ "loss": 0.34474341869354247,
+ "num_input_tokens_seen": 561593424,
+ "step": 12835,
+ "train_runtime": 99291.1127,
+ "train_tokens_per_second": 5656.029
+ },
+ {
+ "epoch": 0.8645300296256396,
+ "grad_norm": 0.7306127243942253,
+ "learning_rate": 6.0564838516707005e-06,
+ "loss": 0.38959674835205077,
+ "num_input_tokens_seen": 561821784,
+ "step": 12840,
+ "train_runtime": 99328.0768,
+ "train_tokens_per_second": 5656.223
+ },
+ {
+ "epoch": 0.8648666846215998,
+ "grad_norm": 0.6692585369253214,
+ "learning_rate": 6.053899320125033e-06,
+ "loss": 0.34748311042785646,
+ "num_input_tokens_seen": 562042168,
+ "step": 12845,
+ "train_runtime": 99368.2816,
+ "train_tokens_per_second": 5656.153
+ },
+ {
+ "epoch": 0.8652033396175599,
+ "grad_norm": 0.731387810949299,
+ "learning_rate": 6.051314493859782e-06,
+ "loss": 0.37701342105865476,
+ "num_input_tokens_seen": 562268616,
+ "step": 12850,
+ "train_runtime": 99405.5774,
+ "train_tokens_per_second": 5656.309
+ },
+ {
+ "epoch": 0.86553999461352,
+ "grad_norm": 0.744983156104233,
+ "learning_rate": 6.048729373597786e-06,
+ "loss": 0.36400351524353025,
+ "num_input_tokens_seen": 562488728,
+ "step": 12855,
+ "train_runtime": 99441.1097,
+ "train_tokens_per_second": 5656.501
+ },
+ {
+ "epoch": 0.8658766496094802,
+ "grad_norm": 0.7052968569468162,
+ "learning_rate": 6.046143960061963e-06,
+ "loss": 0.38647964000701907,
+ "num_input_tokens_seen": 562715632,
+ "step": 12860,
+ "train_runtime": 99479.679,
+ "train_tokens_per_second": 5656.589
+ },
+ {
+ "epoch": 0.8662133046054403,
+ "grad_norm": 0.8490020320396742,
+ "learning_rate": 6.043558253975319e-06,
+ "loss": 0.3893282413482666,
+ "num_input_tokens_seen": 562949544,
+ "step": 12865,
+ "train_runtime": 99522.1085,
+ "train_tokens_per_second": 5656.528
+ },
+ {
+ "epoch": 0.8665499596014005,
+ "grad_norm": 0.7443300106386931,
+ "learning_rate": 6.040972256060936e-06,
+ "loss": 0.35498225688934326,
+ "num_input_tokens_seen": 563161168,
+ "step": 12870,
+ "train_runtime": 99567.3086,
+ "train_tokens_per_second": 5656.085
+ },
+ {
+ "epoch": 0.8668866145973606,
+ "grad_norm": 0.6849806622056446,
+ "learning_rate": 6.038385967041982e-06,
+ "loss": 0.3538069248199463,
+ "num_input_tokens_seen": 563369656,
+ "step": 12875,
+ "train_runtime": 99613.8289,
+ "train_tokens_per_second": 5655.537
+ },
+ {
+ "epoch": 0.8672232695933207,
+ "grad_norm": 0.6384169452977918,
+ "learning_rate": 6.035799387641703e-06,
+ "loss": 0.36878552436828616,
+ "num_input_tokens_seen": 563594552,
+ "step": 12880,
+ "train_runtime": 99655.4177,
+ "train_tokens_per_second": 5655.433
+ },
+ {
+ "epoch": 0.8675599245892809,
+ "grad_norm": 0.7636357717419208,
+ "learning_rate": 6.033212518583427e-06,
+ "loss": 0.3786178112030029,
+ "num_input_tokens_seen": 563803512,
+ "step": 12885,
+ "train_runtime": 99695.3116,
+ "train_tokens_per_second": 5655.266
+ },
+ {
+ "epoch": 0.867896579585241,
+ "grad_norm": 0.8323239828663773,
+ "learning_rate": 6.030625360590567e-06,
+ "loss": 0.35876643657684326,
+ "num_input_tokens_seen": 564030520,
+ "step": 12890,
+ "train_runtime": 99734.3991,
+ "train_tokens_per_second": 5655.326
+ },
+ {
+ "epoch": 0.8682332345812012,
+ "grad_norm": 0.7006071245579706,
+ "learning_rate": 6.028037914386609e-06,
+ "loss": 0.3620233297348022,
+ "num_input_tokens_seen": 564252688,
+ "step": 12895,
+ "train_runtime": 99773.9473,
+ "train_tokens_per_second": 5655.311
+ },
+ {
+ "epoch": 0.8685698895771613,
+ "grad_norm": 0.7617834642572114,
+ "learning_rate": 6.025450180695128e-06,
+ "loss": 0.38210210800170896,
+ "num_input_tokens_seen": 564471472,
+ "step": 12900,
+ "train_runtime": 99808.9271,
+ "train_tokens_per_second": 5655.521
+ },
+ {
+ "epoch": 0.8689065445731214,
+ "grad_norm": 0.8560111437943349,
+ "learning_rate": 6.022862160239774e-06,
+ "loss": 0.3630260467529297,
+ "num_input_tokens_seen": 564683208,
+ "step": 12905,
+ "train_runtime": 99850.9033,
+ "train_tokens_per_second": 5655.264
+ },
+ {
+ "epoch": 0.8692431995690816,
+ "grad_norm": 0.6306746195476955,
+ "learning_rate": 6.020273853744277e-06,
+ "loss": 0.3535295009613037,
+ "num_input_tokens_seen": 564903760,
+ "step": 12910,
+ "train_runtime": 99888.6442,
+ "train_tokens_per_second": 5655.335
+ },
+ {
+ "epoch": 0.8695798545650417,
+ "grad_norm": 0.6756821029513189,
+ "learning_rate": 6.017685261932452e-06,
+ "loss": 0.35518722534179686,
+ "num_input_tokens_seen": 565105760,
+ "step": 12915,
+ "train_runtime": 99923.6252,
+ "train_tokens_per_second": 5655.377
+ },
+ {
+ "epoch": 0.8699165095610019,
+ "grad_norm": 0.75004570307814,
+ "learning_rate": 6.015096385528189e-06,
+ "loss": 0.37440195083618166,
+ "num_input_tokens_seen": 565318768,
+ "step": 12920,
+ "train_runtime": 99960.4635,
+ "train_tokens_per_second": 5655.424
+ },
+ {
+ "epoch": 0.870253164556962,
+ "grad_norm": 0.725287349099034,
+ "learning_rate": 6.012507225255457e-06,
+ "loss": 0.3945824146270752,
+ "num_input_tokens_seen": 565515560,
+ "step": 12925,
+ "train_runtime": 99995.6077,
+ "train_tokens_per_second": 5655.404
+ },
+ {
+ "epoch": 0.8705898195529221,
+ "grad_norm": 0.6923174837526764,
+ "learning_rate": 6.009917781838309e-06,
+ "loss": 0.3706216812133789,
+ "num_input_tokens_seen": 565736648,
+ "step": 12930,
+ "train_runtime": 100027.8477,
+ "train_tokens_per_second": 5655.791
+ },
+ {
+ "epoch": 0.8709264745488823,
+ "grad_norm": 0.6934584410887805,
+ "learning_rate": 6.007328056000873e-06,
+ "loss": 0.37667386531829833,
+ "num_input_tokens_seen": 565960800,
+ "step": 12935,
+ "train_runtime": 100063.8699,
+ "train_tokens_per_second": 5655.996
+ },
+ {
+ "epoch": 0.8712631295448424,
+ "grad_norm": 0.7940274414990574,
+ "learning_rate": 6.004738048467359e-06,
+ "loss": 0.379984712600708,
+ "num_input_tokens_seen": 566180672,
+ "step": 12940,
+ "train_runtime": 100101.3217,
+ "train_tokens_per_second": 5656.076
+ },
+ {
+ "epoch": 0.8715997845408026,
+ "grad_norm": 0.7344390394871104,
+ "learning_rate": 6.002147759962054e-06,
+ "loss": 0.37138075828552247,
+ "num_input_tokens_seen": 566401496,
+ "step": 12945,
+ "train_runtime": 100139.2781,
+ "train_tokens_per_second": 5656.137
+ },
+ {
+ "epoch": 0.8719364395367627,
+ "grad_norm": 0.7421475803173249,
+ "learning_rate": 5.999557191209323e-06,
+ "loss": 0.39212737083435056,
+ "num_input_tokens_seen": 566630216,
+ "step": 12950,
+ "train_runtime": 100175.2546,
+ "train_tokens_per_second": 5656.389
+ },
+ {
+ "epoch": 0.8722730945327228,
+ "grad_norm": 0.8003491943698586,
+ "learning_rate": 5.996966342933611e-06,
+ "loss": 0.33372251987457274,
+ "num_input_tokens_seen": 566848176,
+ "step": 12955,
+ "train_runtime": 100212.8634,
+ "train_tokens_per_second": 5656.441
+ },
+ {
+ "epoch": 0.872609749528683,
+ "grad_norm": 0.7450010892993638,
+ "learning_rate": 5.99437521585944e-06,
+ "loss": 0.37457475662231443,
+ "num_input_tokens_seen": 567056408,
+ "step": 12960,
+ "train_runtime": 100250.1105,
+ "train_tokens_per_second": 5656.417
+ },
+ {
+ "epoch": 0.8729464045246431,
+ "grad_norm": 0.7217432409540151,
+ "learning_rate": 5.99178381071141e-06,
+ "loss": 0.3829956531524658,
+ "num_input_tokens_seen": 567275168,
+ "step": 12965,
+ "train_runtime": 100287.1326,
+ "train_tokens_per_second": 5656.51
+ },
+ {
+ "epoch": 0.8732830595206033,
+ "grad_norm": 0.8055483175054227,
+ "learning_rate": 5.989192128214201e-06,
+ "loss": 0.36388020515441893,
+ "num_input_tokens_seen": 567493544,
+ "step": 12970,
+ "train_runtime": 100327.3129,
+ "train_tokens_per_second": 5656.421
+ },
+ {
+ "epoch": 0.8736197145165634,
+ "grad_norm": 0.7033510804995021,
+ "learning_rate": 5.986600169092568e-06,
+ "loss": 0.350065803527832,
+ "num_input_tokens_seen": 567708176,
+ "step": 12975,
+ "train_runtime": 100364.1125,
+ "train_tokens_per_second": 5656.486
+ },
+ {
+ "epoch": 0.8739563695125235,
+ "grad_norm": 0.7167335785321889,
+ "learning_rate": 5.984007934071343e-06,
+ "loss": 0.3742049217224121,
+ "num_input_tokens_seen": 567941160,
+ "step": 12980,
+ "train_runtime": 100407.0573,
+ "train_tokens_per_second": 5656.387
+ },
+ {
+ "epoch": 0.8742930245084837,
+ "grad_norm": 0.7554361554619513,
+ "learning_rate": 5.981415423875436e-06,
+ "loss": 0.36548290252685545,
+ "num_input_tokens_seen": 568166040,
+ "step": 12985,
+ "train_runtime": 100442.582,
+ "train_tokens_per_second": 5656.625
+ },
+ {
+ "epoch": 0.8746296795044438,
+ "grad_norm": 0.9815756056409435,
+ "learning_rate": 5.978822639229836e-06,
+ "loss": 0.3484132528305054,
+ "num_input_tokens_seen": 568367984,
+ "step": 12990,
+ "train_runtime": 100479.4893,
+ "train_tokens_per_second": 5656.557
+ },
+ {
+ "epoch": 0.874966334500404,
+ "grad_norm": 0.6864898867694621,
+ "learning_rate": 5.976229580859607e-06,
+ "loss": 0.36139798164367676,
+ "num_input_tokens_seen": 568596000,
+ "step": 12995,
+ "train_runtime": 100516.4986,
+ "train_tokens_per_second": 5656.743
+ },
+ {
+ "epoch": 0.8753029894963641,
+ "grad_norm": 0.7186739597846133,
+ "learning_rate": 5.9736362494898895e-06,
+ "loss": 0.3539880275726318,
+ "num_input_tokens_seen": 568814440,
+ "step": 13000,
+ "train_runtime": 100555.0904,
+ "train_tokens_per_second": 5656.744
+ },
+ {
+ "epoch": 0.8756396444923242,
+ "grad_norm": 0.6506103454572264,
+ "learning_rate": 5.971042645845897e-06,
+ "loss": 0.3666597843170166,
+ "num_input_tokens_seen": 569029824,
+ "step": 13005,
+ "train_runtime": 100595.3102,
+ "train_tokens_per_second": 5656.624
+ },
+ {
+ "epoch": 0.8759762994882844,
+ "grad_norm": 0.7706124062857171,
+ "learning_rate": 5.968448770652926e-06,
+ "loss": 0.36510615348815917,
+ "num_input_tokens_seen": 569265096,
+ "step": 13010,
+ "train_runtime": 100627.6528,
+ "train_tokens_per_second": 5657.144
+ },
+ {
+ "epoch": 0.8763129544842445,
+ "grad_norm": 0.7419816582789281,
+ "learning_rate": 5.965854624636345e-06,
+ "loss": 0.3825402736663818,
+ "num_input_tokens_seen": 569479936,
+ "step": 13015,
+ "train_runtime": 100669.5038,
+ "train_tokens_per_second": 5656.926
+ },
+ {
+ "epoch": 0.8766496094802047,
+ "grad_norm": 0.7717112571968371,
+ "learning_rate": 5.963260208521598e-06,
+ "loss": 0.3698084354400635,
+ "num_input_tokens_seen": 569691888,
+ "step": 13020,
+ "train_runtime": 100708.4391,
+ "train_tokens_per_second": 5656.844
+ },
+ {
+ "epoch": 0.8769862644761648,
+ "grad_norm": 0.6761324100161831,
+ "learning_rate": 5.960665523034205e-06,
+ "loss": 0.38090927600860597,
+ "num_input_tokens_seen": 569901576,
+ "step": 13025,
+ "train_runtime": 100749.3523,
+ "train_tokens_per_second": 5656.628
+ },
+ {
+ "epoch": 0.8773229194721249,
+ "grad_norm": 0.7359973742678347,
+ "learning_rate": 5.958070568899762e-06,
+ "loss": 0.33312511444091797,
+ "num_input_tokens_seen": 570123424,
+ "step": 13030,
+ "train_runtime": 100790.5444,
+ "train_tokens_per_second": 5656.517
+ },
+ {
+ "epoch": 0.8776595744680851,
+ "grad_norm": 1.2136155199409513,
+ "learning_rate": 5.955475346843939e-06,
+ "loss": 0.3555187225341797,
+ "num_input_tokens_seen": 570337248,
+ "step": 13035,
+ "train_runtime": 100824.0572,
+ "train_tokens_per_second": 5656.758
+ },
+ {
+ "epoch": 0.8779962294640452,
+ "grad_norm": 0.6546063737811157,
+ "learning_rate": 5.952879857592482e-06,
+ "loss": 0.3440871000289917,
+ "num_input_tokens_seen": 570549824,
+ "step": 13040,
+ "train_runtime": 100867.9537,
+ "train_tokens_per_second": 5656.403
+ },
+ {
+ "epoch": 0.8783328844600053,
+ "grad_norm": 0.6844892975959102,
+ "learning_rate": 5.950284101871212e-06,
+ "loss": 0.38202319145202634,
+ "num_input_tokens_seen": 570757152,
+ "step": 13045,
+ "train_runtime": 100906.3146,
+ "train_tokens_per_second": 5656.308
+ },
+ {
+ "epoch": 0.8786695394559655,
+ "grad_norm": 0.747172974034045,
+ "learning_rate": 5.947688080406023e-06,
+ "loss": 0.3463796615600586,
+ "num_input_tokens_seen": 570960928,
+ "step": 13050,
+ "train_runtime": 100940.5074,
+ "train_tokens_per_second": 5656.41
+ },
+ {
+ "epoch": 0.8790061944519256,
+ "grad_norm": 0.7561590864609555,
+ "learning_rate": 5.945091793922885e-06,
+ "loss": 0.3679834842681885,
+ "num_input_tokens_seen": 571177360,
+ "step": 13055,
+ "train_runtime": 100974.1104,
+ "train_tokens_per_second": 5656.671
+ },
+ {
+ "epoch": 0.8793428494478858,
+ "grad_norm": 0.6782284385723677,
+ "learning_rate": 5.942495243147839e-06,
+ "loss": 0.3929899215698242,
+ "num_input_tokens_seen": 571399192,
+ "step": 13060,
+ "train_runtime": 101005.759,
+ "train_tokens_per_second": 5657.095
+ },
+ {
+ "epoch": 0.8796795044438459,
+ "grad_norm": 0.7143338491446609,
+ "learning_rate": 5.939898428807004e-06,
+ "loss": 0.3815101146697998,
+ "num_input_tokens_seen": 571612808,
+ "step": 13065,
+ "train_runtime": 101043.3454,
+ "train_tokens_per_second": 5657.105
+ },
+ {
+ "epoch": 0.880016159439806,
+ "grad_norm": 0.8449030830666151,
+ "learning_rate": 5.937301351626571e-06,
+ "loss": 0.379590368270874,
+ "num_input_tokens_seen": 571824200,
+ "step": 13070,
+ "train_runtime": 101081.8232,
+ "train_tokens_per_second": 5657.043
+ },
+ {
+ "epoch": 0.8803528144357662,
+ "grad_norm": 0.6830378970909649,
+ "learning_rate": 5.934704012332803e-06,
+ "loss": 0.35865001678466796,
+ "num_input_tokens_seen": 572052552,
+ "step": 13075,
+ "train_runtime": 101122.8991,
+ "train_tokens_per_second": 5657.003
+ },
+ {
+ "epoch": 0.8806894694317263,
+ "grad_norm": 0.8010624909706995,
+ "learning_rate": 5.93210641165204e-06,
+ "loss": 0.4059410095214844,
+ "num_input_tokens_seen": 572274280,
+ "step": 13080,
+ "train_runtime": 101165.6483,
+ "train_tokens_per_second": 5656.804
+ },
+ {
+ "epoch": 0.8810261244276865,
+ "grad_norm": 0.7989521029331256,
+ "learning_rate": 5.92950855031069e-06,
+ "loss": 0.36967930793762205,
+ "num_input_tokens_seen": 572493808,
+ "step": 13085,
+ "train_runtime": 101202.3764,
+ "train_tokens_per_second": 5656.921
+ },
+ {
+ "epoch": 0.8813627794236466,
+ "grad_norm": 0.8258054867788488,
+ "learning_rate": 5.926910429035239e-06,
+ "loss": 0.36150131225585935,
+ "num_input_tokens_seen": 572706072,
+ "step": 13090,
+ "train_runtime": 101239.3964,
+ "train_tokens_per_second": 5656.949
+ },
+ {
+ "epoch": 0.8816994344196067,
+ "grad_norm": 0.7636365729370889,
+ "learning_rate": 5.924312048552241e-06,
+ "loss": 0.3525059223175049,
+ "num_input_tokens_seen": 572932096,
+ "step": 13095,
+ "train_runtime": 101273.4837,
+ "train_tokens_per_second": 5657.276
+ },
+ {
+ "epoch": 0.8820360894155669,
+ "grad_norm": 0.7191821908315644,
+ "learning_rate": 5.9217134095883265e-06,
+ "loss": 0.3641670227050781,
+ "num_input_tokens_seen": 573150384,
+ "step": 13100,
+ "train_runtime": 101314.4917,
+ "train_tokens_per_second": 5657.141
+ },
+ {
+ "epoch": 0.882372744411527,
+ "grad_norm": 0.7363436743411855,
+ "learning_rate": 5.919114512870197e-06,
+ "loss": 0.3603955268859863,
+ "num_input_tokens_seen": 573381144,
+ "step": 13105,
+ "train_runtime": 101359.314,
+ "train_tokens_per_second": 5656.916
+ },
+ {
+ "epoch": 0.8827093994074872,
+ "grad_norm": 0.7568233135879606,
+ "learning_rate": 5.916515359124623e-06,
+ "loss": 0.3425842046737671,
+ "num_input_tokens_seen": 573581832,
+ "step": 13110,
+ "train_runtime": 101399.4218,
+ "train_tokens_per_second": 5656.658
+ },
+ {
+ "epoch": 0.8830460544034473,
+ "grad_norm": 0.7215795683330004,
+ "learning_rate": 5.913915949078453e-06,
+ "loss": 0.36979331970214846,
+ "num_input_tokens_seen": 573801512,
+ "step": 13115,
+ "train_runtime": 101438.6834,
+ "train_tokens_per_second": 5656.634
+ },
+ {
+ "epoch": 0.8833827093994074,
+ "grad_norm": 0.7764516396496984,
+ "learning_rate": 5.911316283458601e-06,
+ "loss": 0.3616600513458252,
+ "num_input_tokens_seen": 574007544,
+ "step": 13120,
+ "train_runtime": 101478.4371,
+ "train_tokens_per_second": 5656.448
+ },
+ {
+ "epoch": 0.8837193643953676,
+ "grad_norm": 0.7391044514688772,
+ "learning_rate": 5.908716362992057e-06,
+ "loss": 0.36803760528564455,
+ "num_input_tokens_seen": 574215752,
+ "step": 13125,
+ "train_runtime": 101514.9279,
+ "train_tokens_per_second": 5656.466
+ },
+ {
+ "epoch": 0.8840560193913277,
+ "grad_norm": 0.6863603049824145,
+ "learning_rate": 5.90611618840588e-06,
+ "loss": 0.379679536819458,
+ "num_input_tokens_seen": 574434272,
+ "step": 13130,
+ "train_runtime": 101554.971,
+ "train_tokens_per_second": 5656.388
+ },
+ {
+ "epoch": 0.8843926743872879,
+ "grad_norm": 0.9353489354148398,
+ "learning_rate": 5.903515760427198e-06,
+ "loss": 0.4000962257385254,
+ "num_input_tokens_seen": 574653864,
+ "step": 13135,
+ "train_runtime": 101588.2234,
+ "train_tokens_per_second": 5656.698
+ },
+ {
+ "epoch": 0.884729329383248,
+ "grad_norm": 0.732273202002582,
+ "learning_rate": 5.900915079783218e-06,
+ "loss": 0.3558906316757202,
+ "num_input_tokens_seen": 574887384,
+ "step": 13140,
+ "train_runtime": 101622.3485,
+ "train_tokens_per_second": 5657.096
+ },
+ {
+ "epoch": 0.8850659843792081,
+ "grad_norm": 0.7460917342917756,
+ "learning_rate": 5.898314147201207e-06,
+ "loss": 0.3416097640991211,
+ "num_input_tokens_seen": 575101544,
+ "step": 13145,
+ "train_runtime": 101664.6483,
+ "train_tokens_per_second": 5656.849
+ },
+ {
+ "epoch": 0.8854026393751683,
+ "grad_norm": 1.0430878064810838,
+ "learning_rate": 5.895712963408511e-06,
+ "loss": 0.3921904325485229,
+ "num_input_tokens_seen": 575309032,
+ "step": 13150,
+ "train_runtime": 101704.7114,
+ "train_tokens_per_second": 5656.661
+ },
+ {
+ "epoch": 0.8857392943711284,
+ "grad_norm": 0.7135808350871835,
+ "learning_rate": 5.893111529132539e-06,
+ "loss": 0.38542513847351073,
+ "num_input_tokens_seen": 575518184,
+ "step": 13155,
+ "train_runtime": 101742.2738,
+ "train_tokens_per_second": 5656.628
+ },
+ {
+ "epoch": 0.8860759493670886,
+ "grad_norm": 0.7661970012022706,
+ "learning_rate": 5.8905098451007755e-06,
+ "loss": 0.37914443016052246,
+ "num_input_tokens_seen": 575721848,
+ "step": 13160,
+ "train_runtime": 101781.9439,
+ "train_tokens_per_second": 5656.424
+ },
+ {
+ "epoch": 0.8864126043630487,
+ "grad_norm": 0.7203329040983677,
+ "learning_rate": 5.887907912040774e-06,
+ "loss": 0.3751526355743408,
+ "num_input_tokens_seen": 575946104,
+ "step": 13165,
+ "train_runtime": 101814.7009,
+ "train_tokens_per_second": 5656.807
+ },
+ {
+ "epoch": 0.8867492593590088,
+ "grad_norm": 0.7424424433197104,
+ "learning_rate": 5.885305730680158e-06,
+ "loss": 0.3788419723510742,
+ "num_input_tokens_seen": 576173784,
+ "step": 13170,
+ "train_runtime": 101847.3914,
+ "train_tokens_per_second": 5657.227
+ },
+ {
+ "epoch": 0.887085914354969,
+ "grad_norm": 0.7289665105864628,
+ "learning_rate": 5.882703301746616e-06,
+ "loss": 0.36290099620819094,
+ "num_input_tokens_seen": 576389120,
+ "step": 13175,
+ "train_runtime": 101889.8836,
+ "train_tokens_per_second": 5656.981
+ },
+ {
+ "epoch": 0.8874225693509291,
+ "grad_norm": 0.7262294984371385,
+ "learning_rate": 5.88010062596791e-06,
+ "loss": 0.35980806350708006,
+ "num_input_tokens_seen": 576598016,
+ "step": 13180,
+ "train_runtime": 101930.8807,
+ "train_tokens_per_second": 5656.755
+ },
+ {
+ "epoch": 0.8877592243468893,
+ "grad_norm": 0.601051210488872,
+ "learning_rate": 5.877497704071871e-06,
+ "loss": 0.3475220203399658,
+ "num_input_tokens_seen": 576831392,
+ "step": 13185,
+ "train_runtime": 101969.0556,
+ "train_tokens_per_second": 5656.926
+ },
+ {
+ "epoch": 0.8880958793428494,
+ "grad_norm": 0.7662456138801742,
+ "learning_rate": 5.874894536786396e-06,
+ "loss": 0.3340578079223633,
+ "num_input_tokens_seen": 577044568,
+ "step": 13190,
+ "train_runtime": 102009.2868,
+ "train_tokens_per_second": 5656.785
+ },
+ {
+ "epoch": 0.8884325343388096,
+ "grad_norm": 0.7991035565617537,
+ "learning_rate": 5.8722911248394555e-06,
+ "loss": 0.33777599334716796,
+ "num_input_tokens_seen": 577243936,
+ "step": 13195,
+ "train_runtime": 102050.722,
+ "train_tokens_per_second": 5656.441
+ },
+ {
+ "epoch": 0.8887691893347698,
+ "grad_norm": 0.7706395477682242,
+ "learning_rate": 5.8696874689590824e-06,
+ "loss": 0.3437623977661133,
+ "num_input_tokens_seen": 577453888,
+ "step": 13200,
+ "train_runtime": 102083.4149,
+ "train_tokens_per_second": 5656.687
+ },
+ {
+ "epoch": 0.8891058443307299,
+ "grad_norm": 0.8134775525528564,
+ "learning_rate": 5.867083569873383e-06,
+ "loss": 0.3634350776672363,
+ "num_input_tokens_seen": 577679688,
+ "step": 13205,
+ "train_runtime": 102117.5448,
+ "train_tokens_per_second": 5657.007
+ },
+ {
+ "epoch": 0.8894424993266901,
+ "grad_norm": 0.7673260905685215,
+ "learning_rate": 5.8644794283105255e-06,
+ "loss": 0.36774277687072754,
+ "num_input_tokens_seen": 577900960,
+ "step": 13210,
+ "train_runtime": 102155.9766,
+ "train_tokens_per_second": 5657.045
+ },
+ {
+ "epoch": 0.8897791543226502,
+ "grad_norm": 0.7882719202387701,
+ "learning_rate": 5.861875044998755e-06,
+ "loss": 0.40239391326904295,
+ "num_input_tokens_seen": 578125704,
+ "step": 13215,
+ "train_runtime": 102190.4444,
+ "train_tokens_per_second": 5657.336
+ },
+ {
+ "epoch": 0.8901158093186103,
+ "grad_norm": 0.7563037927913291,
+ "learning_rate": 5.859270420666377e-06,
+ "loss": 0.36860146522521975,
+ "num_input_tokens_seen": 578343816,
+ "step": 13220,
+ "train_runtime": 102228.533,
+ "train_tokens_per_second": 5657.362
+ },
+ {
+ "epoch": 0.8904524643145705,
+ "grad_norm": 0.7756762347173053,
+ "learning_rate": 5.856665556041764e-06,
+ "loss": 0.3481595993041992,
+ "num_input_tokens_seen": 578548144,
+ "step": 13225,
+ "train_runtime": 102265.8667,
+ "train_tokens_per_second": 5657.295
+ },
+ {
+ "epoch": 0.8907891193105306,
+ "grad_norm": 0.7183886665375777,
+ "learning_rate": 5.8540604518533605e-06,
+ "loss": 0.3859154224395752,
+ "num_input_tokens_seen": 578772776,
+ "step": 13230,
+ "train_runtime": 102299.6045,
+ "train_tokens_per_second": 5657.625
+ },
+ {
+ "epoch": 0.8911257743064908,
+ "grad_norm": 1.4271805656006054,
+ "learning_rate": 5.851455108829675e-06,
+ "loss": 0.34519567489624026,
+ "num_input_tokens_seen": 578981024,
+ "step": 13235,
+ "train_runtime": 102340.8563,
+ "train_tokens_per_second": 5657.379
+ },
+ {
+ "epoch": 0.8914624293024509,
+ "grad_norm": 0.6789149460339803,
+ "learning_rate": 5.848849527699283e-06,
+ "loss": 0.3698782205581665,
+ "num_input_tokens_seen": 579201768,
+ "step": 13240,
+ "train_runtime": 102379.4065,
+ "train_tokens_per_second": 5657.405
+ },
+ {
+ "epoch": 0.891799084298411,
+ "grad_norm": 0.7639170132777864,
+ "learning_rate": 5.846243709190827e-06,
+ "loss": 0.35757324695587156,
+ "num_input_tokens_seen": 579419880,
+ "step": 13245,
+ "train_runtime": 102417.7033,
+ "train_tokens_per_second": 5657.419
+ },
+ {
+ "epoch": 0.8921357392943712,
+ "grad_norm": 0.8837562529973417,
+ "learning_rate": 5.843637654033019e-06,
+ "loss": 0.37553791999816893,
+ "num_input_tokens_seen": 579620760,
+ "step": 13250,
+ "train_runtime": 102459.8571,
+ "train_tokens_per_second": 5657.052
+ },
+ {
+ "epoch": 0.8924723942903313,
+ "grad_norm": 0.676305723344451,
+ "learning_rate": 5.841031362954629e-06,
+ "loss": 0.39575533866882323,
+ "num_input_tokens_seen": 579849048,
+ "step": 13255,
+ "train_runtime": 102495.0631,
+ "train_tokens_per_second": 5657.336
+ },
+ {
+ "epoch": 0.8928090492862915,
+ "grad_norm": 0.6628658100615732,
+ "learning_rate": 5.8384248366845e-06,
+ "loss": 0.37047085762023924,
+ "num_input_tokens_seen": 580077352,
+ "step": 13260,
+ "train_runtime": 102540.1881,
+ "train_tokens_per_second": 5657.073
+ },
+ {
+ "epoch": 0.8931457042822516,
+ "grad_norm": 0.6825809402512039,
+ "learning_rate": 5.8358180759515396e-06,
+ "loss": 0.37846667766571046,
+ "num_input_tokens_seen": 580288200,
+ "step": 13265,
+ "train_runtime": 102577.1371,
+ "train_tokens_per_second": 5657.091
+ },
+ {
+ "epoch": 0.8934823592782117,
+ "grad_norm": 0.7471824938587698,
+ "learning_rate": 5.833211081484719e-06,
+ "loss": 0.37510037422180176,
+ "num_input_tokens_seen": 580506904,
+ "step": 13270,
+ "train_runtime": 102618.782,
+ "train_tokens_per_second": 5656.926
+ },
+ {
+ "epoch": 0.8938190142741719,
+ "grad_norm": 0.641717053888179,
+ "learning_rate": 5.830603854013077e-06,
+ "loss": 0.3673901081085205,
+ "num_input_tokens_seen": 580723408,
+ "step": 13275,
+ "train_runtime": 102660.3376,
+ "train_tokens_per_second": 5656.746
+ },
+ {
+ "epoch": 0.894155669270132,
+ "grad_norm": 0.8523312033525736,
+ "learning_rate": 5.827996394265716e-06,
+ "loss": 0.35067110061645507,
+ "num_input_tokens_seen": 580935112,
+ "step": 13280,
+ "train_runtime": 102701.6639,
+ "train_tokens_per_second": 5656.531
+ },
+ {
+ "epoch": 0.8944923242660922,
+ "grad_norm": 0.7794670905441262,
+ "learning_rate": 5.825388702971802e-06,
+ "loss": 0.35003342628479006,
+ "num_input_tokens_seen": 581151304,
+ "step": 13285,
+ "train_runtime": 102733.2999,
+ "train_tokens_per_second": 5656.893
+ },
+ {
+ "epoch": 0.8948289792620523,
+ "grad_norm": 0.7697508084044369,
+ "learning_rate": 5.822780780860571e-06,
+ "loss": 0.36042060852050783,
+ "num_input_tokens_seen": 581367952,
+ "step": 13290,
+ "train_runtime": 102770.8878,
+ "train_tokens_per_second": 5656.932
+ },
+ {
+ "epoch": 0.8951656342580124,
+ "grad_norm": 0.7094962387963322,
+ "learning_rate": 5.820172628661317e-06,
+ "loss": 0.3688685178756714,
+ "num_input_tokens_seen": 581603552,
+ "step": 13295,
+ "train_runtime": 102804.4905,
+ "train_tokens_per_second": 5657.375
+ },
+ {
+ "epoch": 0.8955022892539726,
+ "grad_norm": 0.7277970714037896,
+ "learning_rate": 5.8175642471034045e-06,
+ "loss": 0.34978251457214354,
+ "num_input_tokens_seen": 581823928,
+ "step": 13300,
+ "train_runtime": 102838.1084,
+ "train_tokens_per_second": 5657.669
+ },
+ {
+ "epoch": 0.8958389442499327,
+ "grad_norm": 0.7417598590201532,
+ "learning_rate": 5.8149556369162565e-06,
+ "loss": 0.3728696346282959,
+ "num_input_tokens_seen": 582031800,
+ "step": 13305,
+ "train_runtime": 102877.0099,
+ "train_tokens_per_second": 5657.55
+ },
+ {
+ "epoch": 0.8961755992458929,
+ "grad_norm": 0.7192634087406454,
+ "learning_rate": 5.812346798829361e-06,
+ "loss": 0.37227132320404055,
+ "num_input_tokens_seen": 582257072,
+ "step": 13310,
+ "train_runtime": 102918.7234,
+ "train_tokens_per_second": 5657.446
+ },
+ {
+ "epoch": 0.896512254241853,
+ "grad_norm": 0.7396821495476227,
+ "learning_rate": 5.809737733572276e-06,
+ "loss": 0.35695157051086424,
+ "num_input_tokens_seen": 582470280,
+ "step": 13315,
+ "train_runtime": 102957.1094,
+ "train_tokens_per_second": 5657.407
+ },
+ {
+ "epoch": 0.8968489092378131,
+ "grad_norm": 0.8286698107403867,
+ "learning_rate": 5.8071284418746166e-06,
+ "loss": 0.3608556747436523,
+ "num_input_tokens_seen": 582689968,
+ "step": 13320,
+ "train_runtime": 102997.8183,
+ "train_tokens_per_second": 5657.304
+ },
+ {
+ "epoch": 0.8971855642337733,
+ "grad_norm": 0.7429942677291922,
+ "learning_rate": 5.80451892446606e-06,
+ "loss": 0.38565864562988283,
+ "num_input_tokens_seen": 582917944,
+ "step": 13325,
+ "train_runtime": 103036.6727,
+ "train_tokens_per_second": 5657.383
+ },
+ {
+ "epoch": 0.8975222192297334,
+ "grad_norm": 0.6807484219365846,
+ "learning_rate": 5.801909182076354e-06,
+ "loss": 0.3657494068145752,
+ "num_input_tokens_seen": 583143528,
+ "step": 13330,
+ "train_runtime": 103071.4117,
+ "train_tokens_per_second": 5657.665
+ },
+ {
+ "epoch": 0.8978588742256935,
+ "grad_norm": 0.7911812244430602,
+ "learning_rate": 5.7992992154353e-06,
+ "loss": 0.3965157985687256,
+ "num_input_tokens_seen": 583360120,
+ "step": 13335,
+ "train_runtime": 103106.3833,
+ "train_tokens_per_second": 5657.847
+ },
+ {
+ "epoch": 0.8981955292216537,
+ "grad_norm": 0.7134344977635636,
+ "learning_rate": 5.79668902527277e-06,
+ "loss": 0.3728632926940918,
+ "num_input_tokens_seen": 583576392,
+ "step": 13340,
+ "train_runtime": 103143.3779,
+ "train_tokens_per_second": 5657.914
+ },
+ {
+ "epoch": 0.8985321842176138,
+ "grad_norm": 0.7068010881578768,
+ "learning_rate": 5.794078612318695e-06,
+ "loss": 0.36152513027191163,
+ "num_input_tokens_seen": 583793888,
+ "step": 13345,
+ "train_runtime": 103187.6757,
+ "train_tokens_per_second": 5657.593
+ },
+ {
+ "epoch": 0.898868839213574,
+ "grad_norm": 1.018985024468395,
+ "learning_rate": 5.791467977303068e-06,
+ "loss": 0.3539185285568237,
+ "num_input_tokens_seen": 584007880,
+ "step": 13350,
+ "train_runtime": 103221.6694,
+ "train_tokens_per_second": 5657.803
+ },
+ {
+ "epoch": 0.8992054942095341,
+ "grad_norm": 0.8787085802082749,
+ "learning_rate": 5.788857120955945e-06,
+ "loss": 0.3753284692764282,
+ "num_input_tokens_seen": 584222720,
+ "step": 13355,
+ "train_runtime": 103259.3517,
+ "train_tokens_per_second": 5657.819
+ },
+ {
+ "epoch": 0.8995421492054942,
+ "grad_norm": 0.7526792175486189,
+ "learning_rate": 5.786246044007443e-06,
+ "loss": 0.3408811092376709,
+ "num_input_tokens_seen": 584430032,
+ "step": 13360,
+ "train_runtime": 103303.2518,
+ "train_tokens_per_second": 5657.421
+ },
+ {
+ "epoch": 0.8998788042014544,
+ "grad_norm": 0.7042471293534047,
+ "learning_rate": 5.783634747187743e-06,
+ "loss": 0.38433132171630857,
+ "num_input_tokens_seen": 584657272,
+ "step": 13365,
+ "train_runtime": 103336.5354,
+ "train_tokens_per_second": 5657.798
+ },
+ {
+ "epoch": 0.9002154591974145,
+ "grad_norm": 0.7283278180613847,
+ "learning_rate": 5.781023231227084e-06,
+ "loss": 0.3663713693618774,
+ "num_input_tokens_seen": 584873064,
+ "step": 13370,
+ "train_runtime": 103383.5959,
+ "train_tokens_per_second": 5657.31
+ },
+ {
+ "epoch": 0.9005521141933747,
+ "grad_norm": 0.6987374788445736,
+ "learning_rate": 5.778411496855769e-06,
+ "loss": 0.3779282093048096,
+ "num_input_tokens_seen": 585095664,
+ "step": 13375,
+ "train_runtime": 103422.1497,
+ "train_tokens_per_second": 5657.354
+ },
+ {
+ "epoch": 0.9008887691893348,
+ "grad_norm": 0.7973255027303662,
+ "learning_rate": 5.775799544804161e-06,
+ "loss": 0.3911623477935791,
+ "num_input_tokens_seen": 585323224,
+ "step": 13380,
+ "train_runtime": 103458.2471,
+ "train_tokens_per_second": 5657.579
+ },
+ {
+ "epoch": 0.9012254241852949,
+ "grad_norm": 0.7376867678943736,
+ "learning_rate": 5.7731873758026845e-06,
+ "loss": 0.3567417860031128,
+ "num_input_tokens_seen": 585539800,
+ "step": 13385,
+ "train_runtime": 103499.6202,
+ "train_tokens_per_second": 5657.41
+ },
+ {
+ "epoch": 0.9015620791812551,
+ "grad_norm": 0.8164983344754712,
+ "learning_rate": 5.770574990581823e-06,
+ "loss": 0.3617391109466553,
+ "num_input_tokens_seen": 585756160,
+ "step": 13390,
+ "train_runtime": 103538.5541,
+ "train_tokens_per_second": 5657.372
+ },
+ {
+ "epoch": 0.9018987341772152,
+ "grad_norm": 0.7386008524112068,
+ "learning_rate": 5.767962389872124e-06,
+ "loss": 0.34462101459503175,
+ "num_input_tokens_seen": 585972312,
+ "step": 13395,
+ "train_runtime": 103578.7054,
+ "train_tokens_per_second": 5657.266
+ },
+ {
+ "epoch": 0.9022353891731754,
+ "grad_norm": 0.7749458765170661,
+ "learning_rate": 5.765349574404191e-06,
+ "loss": 0.3628375053405762,
+ "num_input_tokens_seen": 586175624,
+ "step": 13400,
+ "train_runtime": 103611.757,
+ "train_tokens_per_second": 5657.424
+ },
+ {
+ "epoch": 0.9025720441691355,
+ "grad_norm": 0.7292901746189114,
+ "learning_rate": 5.7627365449086905e-06,
+ "loss": 0.364549446105957,
+ "num_input_tokens_seen": 586370072,
+ "step": 13405,
+ "train_runtime": 103645.7307,
+ "train_tokens_per_second": 5657.445
+ },
+ {
+ "epoch": 0.9029086991650956,
+ "grad_norm": 0.7620757507465177,
+ "learning_rate": 5.760123302116345e-06,
+ "loss": 0.3901118040084839,
+ "num_input_tokens_seen": 586593560,
+ "step": 13410,
+ "train_runtime": 103677.957,
+ "train_tokens_per_second": 5657.843
+ },
+ {
+ "epoch": 0.9032453541610558,
+ "grad_norm": 0.7598225601483857,
+ "learning_rate": 5.757509846757945e-06,
+ "loss": 0.36999359130859377,
+ "num_input_tokens_seen": 586819408,
+ "step": 13415,
+ "train_runtime": 103716.3963,
+ "train_tokens_per_second": 5657.923
+ },
+ {
+ "epoch": 0.9035820091570159,
+ "grad_norm": 1.147309642657896,
+ "learning_rate": 5.75489617956433e-06,
+ "loss": 0.3971453428268433,
+ "num_input_tokens_seen": 587040712,
+ "step": 13420,
+ "train_runtime": 103762.395,
+ "train_tokens_per_second": 5657.548
+ },
+ {
+ "epoch": 0.9039186641529761,
+ "grad_norm": 0.6772540219988885,
+ "learning_rate": 5.752282301266406e-06,
+ "loss": 0.37274601459503176,
+ "num_input_tokens_seen": 587260024,
+ "step": 13425,
+ "train_runtime": 103799.3896,
+ "train_tokens_per_second": 5657.644
+ },
+ {
+ "epoch": 0.9042553191489362,
+ "grad_norm": 0.6865687048122345,
+ "learning_rate": 5.749668212595136e-06,
+ "loss": 0.36958420276641846,
+ "num_input_tokens_seen": 587475152,
+ "step": 13430,
+ "train_runtime": 103838.2234,
+ "train_tokens_per_second": 5657.6
+ },
+ {
+ "epoch": 0.9045919741448963,
+ "grad_norm": 0.7480660958831125,
+ "learning_rate": 5.747053914281539e-06,
+ "loss": 0.36184403896331785,
+ "num_input_tokens_seen": 587679656,
+ "step": 13435,
+ "train_runtime": 103881.14,
+ "train_tokens_per_second": 5657.231
+ },
+ {
+ "epoch": 0.9049286291408565,
+ "grad_norm": 0.7323661070573674,
+ "learning_rate": 5.744439407056699e-06,
+ "loss": 0.3511162757873535,
+ "num_input_tokens_seen": 587888648,
+ "step": 13440,
+ "train_runtime": 103918.9245,
+ "train_tokens_per_second": 5657.186
+ },
+ {
+ "epoch": 0.9052652841368166,
+ "grad_norm": 0.6204712137626076,
+ "learning_rate": 5.741824691651752e-06,
+ "loss": 0.3408759832382202,
+ "num_input_tokens_seen": 588114264,
+ "step": 13445,
+ "train_runtime": 103954.7773,
+ "train_tokens_per_second": 5657.405
+ },
+ {
+ "epoch": 0.9056019391327768,
+ "grad_norm": 0.8116090187953918,
+ "learning_rate": 5.739209768797896e-06,
+ "loss": 0.36322770118713377,
+ "num_input_tokens_seen": 588322904,
+ "step": 13450,
+ "train_runtime": 103991.6356,
+ "train_tokens_per_second": 5657.406
+ },
+ {
+ "epoch": 0.9059385941287369,
+ "grad_norm": 0.7146086921110422,
+ "learning_rate": 5.736594639226385e-06,
+ "loss": 0.32691378593444825,
+ "num_input_tokens_seen": 588542624,
+ "step": 13455,
+ "train_runtime": 104026.9412,
+ "train_tokens_per_second": 5657.598
+ },
+ {
+ "epoch": 0.906275249124697,
+ "grad_norm": 0.6758364668603787,
+ "learning_rate": 5.7339793036685306e-06,
+ "loss": 0.3772242546081543,
+ "num_input_tokens_seen": 588766448,
+ "step": 13460,
+ "train_runtime": 104068.0394,
+ "train_tokens_per_second": 5657.515
+ },
+ {
+ "epoch": 0.9066119041206572,
+ "grad_norm": 0.7332497573577879,
+ "learning_rate": 5.731363762855706e-06,
+ "loss": 0.3471216678619385,
+ "num_input_tokens_seen": 588981896,
+ "step": 13465,
+ "train_runtime": 104109.0854,
+ "train_tokens_per_second": 5657.353
+ },
+ {
+ "epoch": 0.9069485591166173,
+ "grad_norm": 0.8052750996260324,
+ "learning_rate": 5.728748017519337e-06,
+ "loss": 0.3998070001602173,
+ "num_input_tokens_seen": 589199016,
+ "step": 13470,
+ "train_runtime": 104146.4665,
+ "train_tokens_per_second": 5657.408
+ },
+ {
+ "epoch": 0.9072852141125775,
+ "grad_norm": 0.7072206601280142,
+ "learning_rate": 5.726132068390908e-06,
+ "loss": 0.3687425136566162,
+ "num_input_tokens_seen": 589402952,
+ "step": 13475,
+ "train_runtime": 104180.5112,
+ "train_tokens_per_second": 5657.516
+ },
+ {
+ "epoch": 0.9076218691085376,
+ "grad_norm": 0.7386177069645364,
+ "learning_rate": 5.723515916201962e-06,
+ "loss": 0.4100666046142578,
+ "num_input_tokens_seen": 589630440,
+ "step": 13480,
+ "train_runtime": 104218.7952,
+ "train_tokens_per_second": 5657.621
+ },
+ {
+ "epoch": 0.9079585241044977,
+ "grad_norm": 0.6255645042824444,
+ "learning_rate": 5.720899561684098e-06,
+ "loss": 0.3434640645980835,
+ "num_input_tokens_seen": 589849368,
+ "step": 13485,
+ "train_runtime": 104262.0828,
+ "train_tokens_per_second": 5657.372
+ },
+ {
+ "epoch": 0.9082951791004579,
+ "grad_norm": 0.7488072842696437,
+ "learning_rate": 5.71828300556897e-06,
+ "loss": 0.36190190315246584,
+ "num_input_tokens_seen": 590075552,
+ "step": 13490,
+ "train_runtime": 104304.1746,
+ "train_tokens_per_second": 5657.257
+ },
+ {
+ "epoch": 0.908631834096418,
+ "grad_norm": 0.595597879983287,
+ "learning_rate": 5.715666248588289e-06,
+ "loss": 0.3661900520324707,
+ "num_input_tokens_seen": 590300848,
+ "step": 13495,
+ "train_runtime": 104350.0273,
+ "train_tokens_per_second": 5656.93
+ },
+ {
+ "epoch": 0.9089684890923782,
+ "grad_norm": 0.7843852676552852,
+ "learning_rate": 5.713049291473825e-06,
+ "loss": 0.39053096771240237,
+ "num_input_tokens_seen": 590529208,
+ "step": 13500,
+ "train_runtime": 104390.3684,
+ "train_tokens_per_second": 5656.932
+ },
+ {
+ "epoch": 0.9093051440883383,
+ "grad_norm": 0.7754635800602863,
+ "learning_rate": 5.710432134957399e-06,
+ "loss": 0.3421627998352051,
+ "num_input_tokens_seen": 590729424,
+ "step": 13505,
+ "train_runtime": 104437.8595,
+ "train_tokens_per_second": 5656.277
+ },
+ {
+ "epoch": 0.9096417990842984,
+ "grad_norm": 0.7329858480957279,
+ "learning_rate": 5.707814779770892e-06,
+ "loss": 0.34261083602905273,
+ "num_input_tokens_seen": 590940928,
+ "step": 13510,
+ "train_runtime": 104484.5885,
+ "train_tokens_per_second": 5655.771
+ },
+ {
+ "epoch": 0.9099784540802586,
+ "grad_norm": 0.7345589839622834,
+ "learning_rate": 5.7051972266462395e-06,
+ "loss": 0.35787391662597656,
+ "num_input_tokens_seen": 591158192,
+ "step": 13515,
+ "train_runtime": 104521.8258,
+ "train_tokens_per_second": 5655.835
+ },
+ {
+ "epoch": 0.9103151090762187,
+ "grad_norm": 0.8008449716401149,
+ "learning_rate": 5.702579476315431e-06,
+ "loss": 0.3310979127883911,
+ "num_input_tokens_seen": 591376656,
+ "step": 13520,
+ "train_runtime": 104566.7223,
+ "train_tokens_per_second": 5655.496
+ },
+ {
+ "epoch": 0.9106517640721788,
+ "grad_norm": 0.7423340401003228,
+ "learning_rate": 5.699961529510513e-06,
+ "loss": 0.3578958988189697,
+ "num_input_tokens_seen": 591581336,
+ "step": 13525,
+ "train_runtime": 104607.1051,
+ "train_tokens_per_second": 5655.269
+ },
+ {
+ "epoch": 0.910988419068139,
+ "grad_norm": 0.7857402111034575,
+ "learning_rate": 5.697343386963586e-06,
+ "loss": 0.3766601800918579,
+ "num_input_tokens_seen": 591797696,
+ "step": 13530,
+ "train_runtime": 104651.9901,
+ "train_tokens_per_second": 5654.911
+ },
+ {
+ "epoch": 0.9113250740640991,
+ "grad_norm": 0.6657292352842593,
+ "learning_rate": 5.694725049406801e-06,
+ "loss": 0.3710474967956543,
+ "num_input_tokens_seen": 592014888,
+ "step": 13535,
+ "train_runtime": 104697.3063,
+ "train_tokens_per_second": 5654.538
+ },
+ {
+ "epoch": 0.9116617290600593,
+ "grad_norm": 0.6966081276914712,
+ "learning_rate": 5.692106517572373e-06,
+ "loss": 0.38451037406921384,
+ "num_input_tokens_seen": 592229672,
+ "step": 13540,
+ "train_runtime": 104731.9456,
+ "train_tokens_per_second": 5654.719
+ },
+ {
+ "epoch": 0.9119983840560194,
+ "grad_norm": 0.7724805146942423,
+ "learning_rate": 5.689487792192566e-06,
+ "loss": 0.3990262746810913,
+ "num_input_tokens_seen": 592451696,
+ "step": 13545,
+ "train_runtime": 104773.5682,
+ "train_tokens_per_second": 5654.591
+ },
+ {
+ "epoch": 0.9123350390519795,
+ "grad_norm": 0.7342026376072939,
+ "learning_rate": 5.686868873999696e-06,
+ "loss": 0.3797611236572266,
+ "num_input_tokens_seen": 592668960,
+ "step": 13550,
+ "train_runtime": 104816.4545,
+ "train_tokens_per_second": 5654.35
+ },
+ {
+ "epoch": 0.9126716940479397,
+ "grad_norm": 0.666648695750785,
+ "learning_rate": 5.684249763726135e-06,
+ "loss": 0.3764026641845703,
+ "num_input_tokens_seen": 592890144,
+ "step": 13555,
+ "train_runtime": 104853.017,
+ "train_tokens_per_second": 5654.488
+ },
+ {
+ "epoch": 0.9130083490438998,
+ "grad_norm": 0.796183302090131,
+ "learning_rate": 5.681630462104308e-06,
+ "loss": 0.3508880615234375,
+ "num_input_tokens_seen": 593092200,
+ "step": 13560,
+ "train_runtime": 104890.6314,
+ "train_tokens_per_second": 5654.387
+ },
+ {
+ "epoch": 0.91334500403986,
+ "grad_norm": 0.6941135923026022,
+ "learning_rate": 5.679010969866697e-06,
+ "loss": 0.3628666877746582,
+ "num_input_tokens_seen": 593304400,
+ "step": 13565,
+ "train_runtime": 104927.8794,
+ "train_tokens_per_second": 5654.402
+ },
+ {
+ "epoch": 0.9136816590358201,
+ "grad_norm": 0.7543532537354826,
+ "learning_rate": 5.676391287745835e-06,
+ "loss": 0.35071396827697754,
+ "num_input_tokens_seen": 593537816,
+ "step": 13570,
+ "train_runtime": 104959.059,
+ "train_tokens_per_second": 5654.946
+ },
+ {
+ "epoch": 0.9140183140317802,
+ "grad_norm": 0.6731321122510332,
+ "learning_rate": 5.673771416474305e-06,
+ "loss": 0.3439845085144043,
+ "num_input_tokens_seen": 593761520,
+ "step": 13575,
+ "train_runtime": 104997.0967,
+ "train_tokens_per_second": 5655.028
+ },
+ {
+ "epoch": 0.9143549690277404,
+ "grad_norm": 0.7587431284561748,
+ "learning_rate": 5.671151356784747e-06,
+ "loss": 0.3460673809051514,
+ "num_input_tokens_seen": 593970768,
+ "step": 13580,
+ "train_runtime": 105034.4592,
+ "train_tokens_per_second": 5655.009
+ },
+ {
+ "epoch": 0.9146916240237005,
+ "grad_norm": 0.6614026194133914,
+ "learning_rate": 5.668531109409851e-06,
+ "loss": 0.3715926170349121,
+ "num_input_tokens_seen": 594193064,
+ "step": 13585,
+ "train_runtime": 105071.7331,
+ "train_tokens_per_second": 5655.118
+ },
+ {
+ "epoch": 0.9150282790196607,
+ "grad_norm": 0.7852840297073502,
+ "learning_rate": 5.665910675082362e-06,
+ "loss": 0.36374545097351074,
+ "num_input_tokens_seen": 594409288,
+ "step": 13590,
+ "train_runtime": 105109.8348,
+ "train_tokens_per_second": 5655.125
+ },
+ {
+ "epoch": 0.9153649340156208,
+ "grad_norm": 0.7103415655589641,
+ "learning_rate": 5.663290054535078e-06,
+ "loss": 0.34845700263977053,
+ "num_input_tokens_seen": 594620336,
+ "step": 13595,
+ "train_runtime": 105150.7645,
+ "train_tokens_per_second": 5654.931
+ },
+ {
+ "epoch": 0.9157015890115809,
+ "grad_norm": 0.8295807713349123,
+ "learning_rate": 5.660669248500846e-06,
+ "loss": 0.39510331153869627,
+ "num_input_tokens_seen": 594838312,
+ "step": 13600,
+ "train_runtime": 105189.0392,
+ "train_tokens_per_second": 5654.946
+ },
+ {
+ "epoch": 0.9160382440075411,
+ "grad_norm": 0.7158029671235298,
+ "learning_rate": 5.658048257712563e-06,
+ "loss": 0.3466895818710327,
+ "num_input_tokens_seen": 595060816,
+ "step": 13605,
+ "train_runtime": 105234.3634,
+ "train_tokens_per_second": 5654.625
+ },
+ {
+ "epoch": 0.9163748990035012,
+ "grad_norm": 0.7282803359887597,
+ "learning_rate": 5.655427082903184e-06,
+ "loss": 0.3633090019226074,
+ "num_input_tokens_seen": 595279584,
+ "step": 13610,
+ "train_runtime": 105277.1838,
+ "train_tokens_per_second": 5654.403
+ },
+ {
+ "epoch": 0.9167115539994614,
+ "grad_norm": 0.6738750845313296,
+ "learning_rate": 5.652805724805711e-06,
+ "loss": 0.35727887153625487,
+ "num_input_tokens_seen": 595499192,
+ "step": 13615,
+ "train_runtime": 105314.5327,
+ "train_tokens_per_second": 5654.483
+ },
+ {
+ "epoch": 0.9170482089954215,
+ "grad_norm": 0.8264756955958018,
+ "learning_rate": 5.650184184153199e-06,
+ "loss": 0.3687946319580078,
+ "num_input_tokens_seen": 595723264,
+ "step": 13620,
+ "train_runtime": 105352.1128,
+ "train_tokens_per_second": 5654.592
+ },
+ {
+ "epoch": 0.9173848639913816,
+ "grad_norm": 0.7363635027786596,
+ "learning_rate": 5.647562461678754e-06,
+ "loss": 0.39588251113891604,
+ "num_input_tokens_seen": 595947688,
+ "step": 13625,
+ "train_runtime": 105388.882,
+ "train_tokens_per_second": 5654.749
+ },
+ {
+ "epoch": 0.9177215189873418,
+ "grad_norm": 0.7671112595712233,
+ "learning_rate": 5.64494055811553e-06,
+ "loss": 0.35439510345458985,
+ "num_input_tokens_seen": 596167280,
+ "step": 13630,
+ "train_runtime": 105432.2515,
+ "train_tokens_per_second": 5654.506
+ },
+ {
+ "epoch": 0.9180581739833019,
+ "grad_norm": 0.7391097804364258,
+ "learning_rate": 5.642318474196737e-06,
+ "loss": 0.3450098991394043,
+ "num_input_tokens_seen": 596389280,
+ "step": 13635,
+ "train_runtime": 105470.2709,
+ "train_tokens_per_second": 5654.572
+ },
+ {
+ "epoch": 0.9183948289792621,
+ "grad_norm": 0.7394314350713648,
+ "learning_rate": 5.63969621065563e-06,
+ "loss": 0.36347208023071287,
+ "num_input_tokens_seen": 596603992,
+ "step": 13640,
+ "train_runtime": 105510.1258,
+ "train_tokens_per_second": 5654.471
+ },
+ {
+ "epoch": 0.9187314839752222,
+ "grad_norm": 0.6523195737553367,
+ "learning_rate": 5.637073768225517e-06,
+ "loss": 0.3404127836227417,
+ "num_input_tokens_seen": 596814560,
+ "step": 13645,
+ "train_runtime": 105555.5698,
+ "train_tokens_per_second": 5654.032
+ },
+ {
+ "epoch": 0.9190681389711823,
+ "grad_norm": 0.7286950960994242,
+ "learning_rate": 5.634451147639758e-06,
+ "loss": 0.3368164300918579,
+ "num_input_tokens_seen": 597016352,
+ "step": 13650,
+ "train_runtime": 105596.0923,
+ "train_tokens_per_second": 5653.773
+ },
+ {
+ "epoch": 0.9194047939671425,
+ "grad_norm": 0.7381840389633337,
+ "learning_rate": 5.631828349631757e-06,
+ "loss": 0.34926304817199705,
+ "num_input_tokens_seen": 597241424,
+ "step": 13655,
+ "train_runtime": 105633.1844,
+ "train_tokens_per_second": 5653.919
+ },
+ {
+ "epoch": 0.9197414489631026,
+ "grad_norm": 0.7377168394438081,
+ "learning_rate": 5.629205374934974e-06,
+ "loss": 0.3418566703796387,
+ "num_input_tokens_seen": 597449976,
+ "step": 13660,
+ "train_runtime": 105670.7222,
+ "train_tokens_per_second": 5653.884
+ },
+ {
+ "epoch": 0.9200781039590628,
+ "grad_norm": 0.7062836493823704,
+ "learning_rate": 5.626582224282913e-06,
+ "loss": 0.36661295890808104,
+ "num_input_tokens_seen": 597675264,
+ "step": 13665,
+ "train_runtime": 105714.0467,
+ "train_tokens_per_second": 5653.698
+ },
+ {
+ "epoch": 0.9204147589550229,
+ "grad_norm": 0.7520494279886026,
+ "learning_rate": 5.623958898409132e-06,
+ "loss": 0.39702858924865725,
+ "num_input_tokens_seen": 597900752,
+ "step": 13670,
+ "train_runtime": 105743.9982,
+ "train_tokens_per_second": 5654.229
+ },
+ {
+ "epoch": 0.920751413950983,
+ "grad_norm": 0.6987426458318066,
+ "learning_rate": 5.621335398047234e-06,
+ "loss": 0.3848641157150269,
+ "num_input_tokens_seen": 598115680,
+ "step": 13675,
+ "train_runtime": 105787.0252,
+ "train_tokens_per_second": 5653.961
+ },
+ {
+ "epoch": 0.9210880689469432,
+ "grad_norm": 0.7307345712178913,
+ "learning_rate": 5.618711723930874e-06,
+ "loss": 0.36865854263305664,
+ "num_input_tokens_seen": 598333664,
+ "step": 13680,
+ "train_runtime": 105817.9321,
+ "train_tokens_per_second": 5654.369
+ },
+ {
+ "epoch": 0.9214247239429033,
+ "grad_norm": 0.7963356189127992,
+ "learning_rate": 5.616087876793753e-06,
+ "loss": 0.3818777561187744,
+ "num_input_tokens_seen": 598560984,
+ "step": 13685,
+ "train_runtime": 105853.2283,
+ "train_tokens_per_second": 5654.631
+ },
+ {
+ "epoch": 0.9217613789388635,
+ "grad_norm": 0.6658404353690085,
+ "learning_rate": 5.613463857369621e-06,
+ "loss": 0.3648639678955078,
+ "num_input_tokens_seen": 598784192,
+ "step": 13690,
+ "train_runtime": 105895.3239,
+ "train_tokens_per_second": 5654.491
+ },
+ {
+ "epoch": 0.9220980339348236,
+ "grad_norm": 0.7245035749434348,
+ "learning_rate": 5.610839666392278e-06,
+ "loss": 0.35527966022491453,
+ "num_input_tokens_seen": 599005456,
+ "step": 13695,
+ "train_runtime": 105941.0406,
+ "train_tokens_per_second": 5654.14
+ },
+ {
+ "epoch": 0.9224346889307837,
+ "grad_norm": 0.7564271755937382,
+ "learning_rate": 5.608215304595571e-06,
+ "loss": 0.3818879842758179,
+ "num_input_tokens_seen": 599222856,
+ "step": 13700,
+ "train_runtime": 105987.7905,
+ "train_tokens_per_second": 5653.697
+ },
+ {
+ "epoch": 0.9227713439267439,
+ "grad_norm": 0.71469916144582,
+ "learning_rate": 5.605590772713393e-06,
+ "loss": 0.35137481689453126,
+ "num_input_tokens_seen": 599421784,
+ "step": 13705,
+ "train_runtime": 106025.2333,
+ "train_tokens_per_second": 5653.577
+ },
+ {
+ "epoch": 0.923107998922704,
+ "grad_norm": 0.8680683000673615,
+ "learning_rate": 5.602966071479687e-06,
+ "loss": 0.3719889402389526,
+ "num_input_tokens_seen": 599626912,
+ "step": 13710,
+ "train_runtime": 106057.3995,
+ "train_tokens_per_second": 5653.796
+ },
+ {
+ "epoch": 0.9234446539186641,
+ "grad_norm": 0.7763025193380707,
+ "learning_rate": 5.600341201628439e-06,
+ "loss": 0.37358889579772947,
+ "num_input_tokens_seen": 599854192,
+ "step": 13715,
+ "train_runtime": 106090.67,
+ "train_tokens_per_second": 5654.165
+ },
+ {
+ "epoch": 0.9237813089146243,
+ "grad_norm": 0.6656080663010198,
+ "learning_rate": 5.597716163893692e-06,
+ "loss": 0.31069388389587405,
+ "num_input_tokens_seen": 600085080,
+ "step": 13720,
+ "train_runtime": 106134.8208,
+ "train_tokens_per_second": 5653.989
+ },
+ {
+ "epoch": 0.9241179639105844,
+ "grad_norm": 0.7327001774115326,
+ "learning_rate": 5.595090959009525e-06,
+ "loss": 0.34327847957611085,
+ "num_input_tokens_seen": 600310096,
+ "step": 13725,
+ "train_runtime": 106180.1722,
+ "train_tokens_per_second": 5653.693
+ },
+ {
+ "epoch": 0.9244546189065446,
+ "grad_norm": 0.6125120195863898,
+ "learning_rate": 5.592465587710068e-06,
+ "loss": 0.34142630100250243,
+ "num_input_tokens_seen": 600545032,
+ "step": 13730,
+ "train_runtime": 106218.5007,
+ "train_tokens_per_second": 5653.865
+ },
+ {
+ "epoch": 0.9247912739025047,
+ "grad_norm": 0.6167159170504363,
+ "learning_rate": 5.589840050729498e-06,
+ "loss": 0.3286783218383789,
+ "num_input_tokens_seen": 600768064,
+ "step": 13735,
+ "train_runtime": 106258.0895,
+ "train_tokens_per_second": 5653.857
+ },
+ {
+ "epoch": 0.9251279288984648,
+ "grad_norm": 0.7881839614694872,
+ "learning_rate": 5.587214348802039e-06,
+ "loss": 0.3909108638763428,
+ "num_input_tokens_seen": 600980632,
+ "step": 13740,
+ "train_runtime": 106298.4773,
+ "train_tokens_per_second": 5653.709
+ },
+ {
+ "epoch": 0.925464583894425,
+ "grad_norm": 0.7022898228386897,
+ "learning_rate": 5.5845884826619615e-06,
+ "loss": 0.36343495845794677,
+ "num_input_tokens_seen": 601207448,
+ "step": 13745,
+ "train_runtime": 106333.7277,
+ "train_tokens_per_second": 5653.968
+ },
+ {
+ "epoch": 0.9258012388903851,
+ "grad_norm": 0.6378832684361027,
+ "learning_rate": 5.5819624530435775e-06,
+ "loss": 0.3585641622543335,
+ "num_input_tokens_seen": 601419704,
+ "step": 13750,
+ "train_runtime": 106373.8251,
+ "train_tokens_per_second": 5653.832
+ },
+ {
+ "epoch": 0.9261378938863453,
+ "grad_norm": 0.7242428710010252,
+ "learning_rate": 5.57933626068125e-06,
+ "loss": 0.3933718681335449,
+ "num_input_tokens_seen": 601639824,
+ "step": 13755,
+ "train_runtime": 106414.6703,
+ "train_tokens_per_second": 5653.73
+ },
+ {
+ "epoch": 0.9264745488823054,
+ "grad_norm": 0.7473031695779977,
+ "learning_rate": 5.5767099063093835e-06,
+ "loss": 0.3709197998046875,
+ "num_input_tokens_seen": 601868536,
+ "step": 13760,
+ "train_runtime": 106452.2637,
+ "train_tokens_per_second": 5653.882
+ },
+ {
+ "epoch": 0.9268112038782655,
+ "grad_norm": 0.803164002917514,
+ "learning_rate": 5.574083390662431e-06,
+ "loss": 0.373383617401123,
+ "num_input_tokens_seen": 602098696,
+ "step": 13765,
+ "train_runtime": 106489.0364,
+ "train_tokens_per_second": 5654.091
+ },
+ {
+ "epoch": 0.9271478588742257,
+ "grad_norm": 0.7108278884261701,
+ "learning_rate": 5.57145671447489e-06,
+ "loss": 0.3655261993408203,
+ "num_input_tokens_seen": 602302480,
+ "step": 13770,
+ "train_runtime": 106525.8625,
+ "train_tokens_per_second": 5654.049
+ },
+ {
+ "epoch": 0.9274845138701858,
+ "grad_norm": 0.7613900230491544,
+ "learning_rate": 5.5688298784813e-06,
+ "loss": 0.3984712600708008,
+ "num_input_tokens_seen": 602522208,
+ "step": 13775,
+ "train_runtime": 106564.2964,
+ "train_tokens_per_second": 5654.072
+ },
+ {
+ "epoch": 0.927821168866146,
+ "grad_norm": 0.6708141182451115,
+ "learning_rate": 5.566202883416249e-06,
+ "loss": 0.3970597743988037,
+ "num_input_tokens_seen": 602750424,
+ "step": 13780,
+ "train_runtime": 106606.7053,
+ "train_tokens_per_second": 5653.964
+ },
+ {
+ "epoch": 0.9281578238621061,
+ "grad_norm": 0.7821647487746413,
+ "learning_rate": 5.563575730014366e-06,
+ "loss": 0.39171276092529295,
+ "num_input_tokens_seen": 602965320,
+ "step": 13785,
+ "train_runtime": 106643.4726,
+ "train_tokens_per_second": 5654.029
+ },
+ {
+ "epoch": 0.9284944788580662,
+ "grad_norm": 0.7438040498615117,
+ "learning_rate": 5.560948419010328e-06,
+ "loss": 0.3737471342086792,
+ "num_input_tokens_seen": 603186136,
+ "step": 13790,
+ "train_runtime": 106681.291,
+ "train_tokens_per_second": 5654.095
+ },
+ {
+ "epoch": 0.9288311338540264,
+ "grad_norm": 0.7446914306967766,
+ "learning_rate": 5.558320951138853e-06,
+ "loss": 0.34050416946411133,
+ "num_input_tokens_seen": 603397288,
+ "step": 13795,
+ "train_runtime": 106722.5651,
+ "train_tokens_per_second": 5653.887
+ },
+ {
+ "epoch": 0.9291677888499865,
+ "grad_norm": 0.7759821070900785,
+ "learning_rate": 5.5556933271347046e-06,
+ "loss": 0.3723226308822632,
+ "num_input_tokens_seen": 603606088,
+ "step": 13800,
+ "train_runtime": 106763.5139,
+ "train_tokens_per_second": 5653.674
+ },
+ {
+ "epoch": 0.9295044438459467,
+ "grad_norm": 0.7098216673132232,
+ "learning_rate": 5.55306554773269e-06,
+ "loss": 0.3821061611175537,
+ "num_input_tokens_seen": 603816144,
+ "step": 13805,
+ "train_runtime": 106805.761,
+ "train_tokens_per_second": 5653.404
+ },
+ {
+ "epoch": 0.9298410988419068,
+ "grad_norm": 0.9275669407149835,
+ "learning_rate": 5.550437613667658e-06,
+ "loss": 0.3593113660812378,
+ "num_input_tokens_seen": 604004552,
+ "step": 13810,
+ "train_runtime": 106836.865,
+ "train_tokens_per_second": 5653.522
+ },
+ {
+ "epoch": 0.9301777538378669,
+ "grad_norm": 0.6896177858346889,
+ "learning_rate": 5.547809525674501e-06,
+ "loss": 0.35454933643341063,
+ "num_input_tokens_seen": 604212440,
+ "step": 13815,
+ "train_runtime": 106875.1406,
+ "train_tokens_per_second": 5653.442
+ },
+ {
+ "epoch": 0.9305144088338271,
+ "grad_norm": 0.7927279590495904,
+ "learning_rate": 5.545181284488159e-06,
+ "loss": 0.38926100730895996,
+ "num_input_tokens_seen": 604439792,
+ "step": 13820,
+ "train_runtime": 106917.4476,
+ "train_tokens_per_second": 5653.332
+ },
+ {
+ "epoch": 0.9308510638297872,
+ "grad_norm": 0.8386569241979678,
+ "learning_rate": 5.5425528908436085e-06,
+ "loss": 0.35577256679534913,
+ "num_input_tokens_seen": 604643600,
+ "step": 13825,
+ "train_runtime": 106953.2903,
+ "train_tokens_per_second": 5653.343
+ },
+ {
+ "epoch": 0.9311877188257474,
+ "grad_norm": 0.7592336539171605,
+ "learning_rate": 5.539924345475873e-06,
+ "loss": 0.3907893180847168,
+ "num_input_tokens_seen": 604839736,
+ "step": 13830,
+ "train_runtime": 106988.9857,
+ "train_tokens_per_second": 5653.29
+ },
+ {
+ "epoch": 0.9315243738217075,
+ "grad_norm": 0.8126152719075438,
+ "learning_rate": 5.5372956491200145e-06,
+ "loss": 0.37813870906829833,
+ "num_input_tokens_seen": 605051728,
+ "step": 13835,
+ "train_runtime": 107025.6706,
+ "train_tokens_per_second": 5653.333
+ },
+ {
+ "epoch": 0.9318610288176676,
+ "grad_norm": 0.7879475237888696,
+ "learning_rate": 5.534666802511143e-06,
+ "loss": 0.36364994049072263,
+ "num_input_tokens_seen": 605251488,
+ "step": 13840,
+ "train_runtime": 107062.9642,
+ "train_tokens_per_second": 5653.229
+ },
+ {
+ "epoch": 0.9321976838136278,
+ "grad_norm": 0.6977984667014386,
+ "learning_rate": 5.532037806384404e-06,
+ "loss": 0.372878885269165,
+ "num_input_tokens_seen": 605478256,
+ "step": 13845,
+ "train_runtime": 107101.7476,
+ "train_tokens_per_second": 5653.299
+ },
+ {
+ "epoch": 0.9325343388095879,
+ "grad_norm": 0.7377782721135303,
+ "learning_rate": 5.529408661474989e-06,
+ "loss": 0.33782191276550294,
+ "num_input_tokens_seen": 605678152,
+ "step": 13850,
+ "train_runtime": 107140.4741,
+ "train_tokens_per_second": 5653.122
+ },
+ {
+ "epoch": 0.932870993805548,
+ "grad_norm": 0.7120793619791788,
+ "learning_rate": 5.526779368518129e-06,
+ "loss": 0.3900485038757324,
+ "num_input_tokens_seen": 605902976,
+ "step": 13855,
+ "train_runtime": 107174.2345,
+ "train_tokens_per_second": 5653.439
+ },
+ {
+ "epoch": 0.9332076488015082,
+ "grad_norm": 0.7800631196768846,
+ "learning_rate": 5.5241499282491e-06,
+ "loss": 0.35591723918914797,
+ "num_input_tokens_seen": 606103560,
+ "step": 13860,
+ "train_runtime": 107212.7655,
+ "train_tokens_per_second": 5653.278
+ },
+ {
+ "epoch": 0.9335443037974683,
+ "grad_norm": 0.6995200522098299,
+ "learning_rate": 5.5215203414032135e-06,
+ "loss": 0.3359647274017334,
+ "num_input_tokens_seen": 606332784,
+ "step": 13865,
+ "train_runtime": 107261.2791,
+ "train_tokens_per_second": 5652.858
+ },
+ {
+ "epoch": 0.9338809587934285,
+ "grad_norm": 0.7315710239059265,
+ "learning_rate": 5.518890608715828e-06,
+ "loss": 0.3636192798614502,
+ "num_input_tokens_seen": 606549080,
+ "step": 13870,
+ "train_runtime": 107306.9882,
+ "train_tokens_per_second": 5652.466
+ },
+ {
+ "epoch": 0.9342176137893886,
+ "grad_norm": 0.8443286259971129,
+ "learning_rate": 5.516260730922337e-06,
+ "loss": 0.3732717275619507,
+ "num_input_tokens_seen": 606754664,
+ "step": 13875,
+ "train_runtime": 107344.4184,
+ "train_tokens_per_second": 5652.41
+ },
+ {
+ "epoch": 0.9345542687853488,
+ "grad_norm": 0.7163150661251769,
+ "learning_rate": 5.51363070875818e-06,
+ "loss": 0.3665364503860474,
+ "num_input_tokens_seen": 606971336,
+ "step": 13880,
+ "train_runtime": 107380.8754,
+ "train_tokens_per_second": 5652.509
+ },
+ {
+ "epoch": 0.9348909237813089,
+ "grad_norm": 0.7826404929432437,
+ "learning_rate": 5.511000542958832e-06,
+ "loss": 0.3829598903656006,
+ "num_input_tokens_seen": 607182888,
+ "step": 13885,
+ "train_runtime": 107425.4489,
+ "train_tokens_per_second": 5652.133
+ },
+ {
+ "epoch": 0.935227578777269,
+ "grad_norm": 0.7974161880488172,
+ "learning_rate": 5.508370234259812e-06,
+ "loss": 0.3617416858673096,
+ "num_input_tokens_seen": 607395376,
+ "step": 13890,
+ "train_runtime": 107467.1083,
+ "train_tokens_per_second": 5651.919
+ },
+ {
+ "epoch": 0.9355642337732292,
+ "grad_norm": 0.7468198908899866,
+ "learning_rate": 5.505739783396678e-06,
+ "loss": 0.4197434425354004,
+ "num_input_tokens_seen": 607606560,
+ "step": 13895,
+ "train_runtime": 107501.3365,
+ "train_tokens_per_second": 5652.084
+ },
+ {
+ "epoch": 0.9359008887691893,
+ "grad_norm": 0.8018465922168921,
+ "learning_rate": 5.503109191105026e-06,
+ "loss": 0.368869423866272,
+ "num_input_tokens_seen": 607818184,
+ "step": 13900,
+ "train_runtime": 107542.256,
+ "train_tokens_per_second": 5651.901
+ },
+ {
+ "epoch": 0.9362375437651494,
+ "grad_norm": 0.6896918447153622,
+ "learning_rate": 5.500478458120493e-06,
+ "loss": 0.3878501892089844,
+ "num_input_tokens_seen": 608046144,
+ "step": 13905,
+ "train_runtime": 107578.9863,
+ "train_tokens_per_second": 5652.09
+ },
+ {
+ "epoch": 0.9365741987611096,
+ "grad_norm": 0.6502629376388114,
+ "learning_rate": 5.497847585178754e-06,
+ "loss": 0.37720532417297364,
+ "num_input_tokens_seen": 608274496,
+ "step": 13910,
+ "train_runtime": 107614.2489,
+ "train_tokens_per_second": 5652.36
+ },
+ {
+ "epoch": 0.9369108537570697,
+ "grad_norm": 0.671150617120138,
+ "learning_rate": 5.495216573015526e-06,
+ "loss": 0.34202401638031005,
+ "num_input_tokens_seen": 608487072,
+ "step": 13915,
+ "train_runtime": 107649.7272,
+ "train_tokens_per_second": 5652.472
+ },
+ {
+ "epoch": 0.9372475087530299,
+ "grad_norm": 0.7302905322159697,
+ "learning_rate": 5.492585422366562e-06,
+ "loss": 0.3395219802856445,
+ "num_input_tokens_seen": 608702136,
+ "step": 13920,
+ "train_runtime": 107690.4313,
+ "train_tokens_per_second": 5652.333
+ },
+ {
+ "epoch": 0.93758416374899,
+ "grad_norm": 0.7232012493162839,
+ "learning_rate": 5.489954133967657e-06,
+ "loss": 0.3353360652923584,
+ "num_input_tokens_seen": 608907648,
+ "step": 13925,
+ "train_runtime": 107730.6462,
+ "train_tokens_per_second": 5652.13
+ },
+ {
+ "epoch": 0.9379208187449501,
+ "grad_norm": 0.6700324546008681,
+ "learning_rate": 5.48732270855464e-06,
+ "loss": 0.3447211027145386,
+ "num_input_tokens_seen": 609137704,
+ "step": 13930,
+ "train_runtime": 107770.2254,
+ "train_tokens_per_second": 5652.189
+ },
+ {
+ "epoch": 0.9382574737409103,
+ "grad_norm": 0.744553222908747,
+ "learning_rate": 5.48469114686338e-06,
+ "loss": 0.36331336498260497,
+ "num_input_tokens_seen": 609342520,
+ "step": 13935,
+ "train_runtime": 107808.8795,
+ "train_tokens_per_second": 5652.062
+ },
+ {
+ "epoch": 0.9385941287368704,
+ "grad_norm": 0.7229277790841906,
+ "learning_rate": 5.482059449629788e-06,
+ "loss": 0.3539914131164551,
+ "num_input_tokens_seen": 609571440,
+ "step": 13940,
+ "train_runtime": 107847.7103,
+ "train_tokens_per_second": 5652.15
+ },
+ {
+ "epoch": 0.9389307837328306,
+ "grad_norm": 0.7003088104586751,
+ "learning_rate": 5.479427617589808e-06,
+ "loss": 0.3414954662322998,
+ "num_input_tokens_seen": 609787136,
+ "step": 13945,
+ "train_runtime": 107885.5574,
+ "train_tokens_per_second": 5652.167
+ },
+ {
+ "epoch": 0.9392674387287907,
+ "grad_norm": 0.6891150838884817,
+ "learning_rate": 5.4767956514794226e-06,
+ "loss": 0.3689284801483154,
+ "num_input_tokens_seen": 610015912,
+ "step": 13950,
+ "train_runtime": 107918.1469,
+ "train_tokens_per_second": 5652.58
+ },
+ {
+ "epoch": 0.9396040937247508,
+ "grad_norm": 0.8843526225142938,
+ "learning_rate": 5.474163552034655e-06,
+ "loss": 0.3568282127380371,
+ "num_input_tokens_seen": 610224872,
+ "step": 13955,
+ "train_runtime": 107958.8593,
+ "train_tokens_per_second": 5652.383
+ },
+ {
+ "epoch": 0.939940748720711,
+ "grad_norm": 0.7953669523701649,
+ "learning_rate": 5.4715313199915595e-06,
+ "loss": 0.33748788833618165,
+ "num_input_tokens_seen": 610433896,
+ "step": 13960,
+ "train_runtime": 107998.8796,
+ "train_tokens_per_second": 5652.224
+ },
+ {
+ "epoch": 0.9402774037166711,
+ "grad_norm": 0.8436524922865566,
+ "learning_rate": 5.468898956086236e-06,
+ "loss": 0.36934537887573243,
+ "num_input_tokens_seen": 610657928,
+ "step": 13965,
+ "train_runtime": 108037.0093,
+ "train_tokens_per_second": 5652.303
+ },
+ {
+ "epoch": 0.9406140587126313,
+ "grad_norm": 0.7062142551000551,
+ "learning_rate": 5.466266461054815e-06,
+ "loss": 0.3693251132965088,
+ "num_input_tokens_seen": 610879072,
+ "step": 13970,
+ "train_runtime": 108078.9259,
+ "train_tokens_per_second": 5652.157
+ },
+ {
+ "epoch": 0.9409507137085914,
+ "grad_norm": 0.7405368560790437,
+ "learning_rate": 5.463633835633464e-06,
+ "loss": 0.38323063850402833,
+ "num_input_tokens_seen": 611111488,
+ "step": 13975,
+ "train_runtime": 108116.6852,
+ "train_tokens_per_second": 5652.333
+ },
+ {
+ "epoch": 0.9412873687045515,
+ "grad_norm": 0.6906471137787829,
+ "learning_rate": 5.46100108055839e-06,
+ "loss": 0.37401399612426756,
+ "num_input_tokens_seen": 611338984,
+ "step": 13980,
+ "train_runtime": 108153.7009,
+ "train_tokens_per_second": 5652.502
+ },
+ {
+ "epoch": 0.9416240237005117,
+ "grad_norm": 0.699157303658586,
+ "learning_rate": 5.458368196565833e-06,
+ "loss": 0.34747042655944826,
+ "num_input_tokens_seen": 611557384,
+ "step": 13985,
+ "train_runtime": 108187.4002,
+ "train_tokens_per_second": 5652.76
+ },
+ {
+ "epoch": 0.9419606786964718,
+ "grad_norm": 0.6950434602213443,
+ "learning_rate": 5.455735184392073e-06,
+ "loss": 0.39068779945373533,
+ "num_input_tokens_seen": 611785288,
+ "step": 13990,
+ "train_runtime": 108222.1017,
+ "train_tokens_per_second": 5653.053
+ },
+ {
+ "epoch": 0.942297333692432,
+ "grad_norm": 0.8222775495837801,
+ "learning_rate": 5.4531020447734204e-06,
+ "loss": 0.3776577949523926,
+ "num_input_tokens_seen": 611999096,
+ "step": 13995,
+ "train_runtime": 108255.2451,
+ "train_tokens_per_second": 5653.297
+ },
+ {
+ "epoch": 0.9426339886883921,
+ "grad_norm": 0.6942977966354065,
+ "learning_rate": 5.450468778446226e-06,
+ "loss": 0.38014535903930663,
+ "num_input_tokens_seen": 612229616,
+ "step": 14000,
+ "train_runtime": 108302.2562,
+ "train_tokens_per_second": 5652.972
+ },
+ {
+ "epoch": 0.9429706436843522,
+ "grad_norm": 0.8181331571770365,
+ "learning_rate": 5.447835386146875e-06,
+ "loss": 0.3620244264602661,
+ "num_input_tokens_seen": 612457288,
+ "step": 14005,
+ "train_runtime": 108339.7787,
+ "train_tokens_per_second": 5653.116
+ },
+ {
+ "epoch": 0.9433072986803124,
+ "grad_norm": 0.7249133494502058,
+ "learning_rate": 5.445201868611784e-06,
+ "loss": 0.4034558296203613,
+ "num_input_tokens_seen": 612668352,
+ "step": 14010,
+ "train_runtime": 108376.897,
+ "train_tokens_per_second": 5653.127
+ },
+ {
+ "epoch": 0.9436439536762725,
+ "grad_norm": 0.7004943272407984,
+ "learning_rate": 5.442568226577413e-06,
+ "loss": 0.37315356731414795,
+ "num_input_tokens_seen": 612885272,
+ "step": 14015,
+ "train_runtime": 108415.5424,
+ "train_tokens_per_second": 5653.113
+ },
+ {
+ "epoch": 0.9439806086722327,
+ "grad_norm": 0.6490441295997635,
+ "learning_rate": 5.439934460780247e-06,
+ "loss": 0.3591346740722656,
+ "num_input_tokens_seen": 613107944,
+ "step": 14020,
+ "train_runtime": 108453.9122,
+ "train_tokens_per_second": 5653.166
+ },
+ {
+ "epoch": 0.9443172636681928,
+ "grad_norm": 0.6918392163792407,
+ "learning_rate": 5.437300571956813e-06,
+ "loss": 0.3745762586593628,
+ "num_input_tokens_seen": 613308168,
+ "step": 14025,
+ "train_runtime": 108489.5473,
+ "train_tokens_per_second": 5653.154
+ },
+ {
+ "epoch": 0.9446539186641529,
+ "grad_norm": 0.654742109915237,
+ "learning_rate": 5.434666560843667e-06,
+ "loss": 0.3551823616027832,
+ "num_input_tokens_seen": 613516376,
+ "step": 14030,
+ "train_runtime": 108525.2798,
+ "train_tokens_per_second": 5653.212
+ },
+ {
+ "epoch": 0.9449905736601131,
+ "grad_norm": 0.7025947454416406,
+ "learning_rate": 5.432032428177405e-06,
+ "loss": 0.3733973979949951,
+ "num_input_tokens_seen": 613743992,
+ "step": 14035,
+ "train_runtime": 108565.1399,
+ "train_tokens_per_second": 5653.233
+ },
+ {
+ "epoch": 0.9453272286560732,
+ "grad_norm": 0.7446809227599923,
+ "learning_rate": 5.429398174694651e-06,
+ "loss": 0.35695390701293944,
+ "num_input_tokens_seen": 613975032,
+ "step": 14040,
+ "train_runtime": 108600.6982,
+ "train_tokens_per_second": 5653.509
+ },
+ {
+ "epoch": 0.9456638836520334,
+ "grad_norm": 0.6898581042276205,
+ "learning_rate": 5.426763801132067e-06,
+ "loss": 0.3617269039154053,
+ "num_input_tokens_seen": 614162568,
+ "step": 14045,
+ "train_runtime": 108637.4509,
+ "train_tokens_per_second": 5653.323
+ },
+ {
+ "epoch": 0.9460005386479935,
+ "grad_norm": 0.7678862695446375,
+ "learning_rate": 5.424129308226347e-06,
+ "loss": 0.3843861103057861,
+ "num_input_tokens_seen": 614370656,
+ "step": 14050,
+ "train_runtime": 108675.5655,
+ "train_tokens_per_second": 5653.255
+ },
+ {
+ "epoch": 0.9463371936439536,
+ "grad_norm": 0.7654766109095031,
+ "learning_rate": 5.421494696714219e-06,
+ "loss": 0.3902265548706055,
+ "num_input_tokens_seen": 614608864,
+ "step": 14055,
+ "train_runtime": 108715.2383,
+ "train_tokens_per_second": 5653.383
+ },
+ {
+ "epoch": 0.9466738486399138,
+ "grad_norm": 0.7609882464523344,
+ "learning_rate": 5.418859967332442e-06,
+ "loss": 0.3707080841064453,
+ "num_input_tokens_seen": 614836608,
+ "step": 14060,
+ "train_runtime": 108755.1102,
+ "train_tokens_per_second": 5653.404
+ },
+ {
+ "epoch": 0.9470105036358739,
+ "grad_norm": 0.9394544427990754,
+ "learning_rate": 5.416225120817811e-06,
+ "loss": 0.3681238412857056,
+ "num_input_tokens_seen": 615027064,
+ "step": 14065,
+ "train_runtime": 108790.9955,
+ "train_tokens_per_second": 5653.29
+ },
+ {
+ "epoch": 0.947347158631834,
+ "grad_norm": 0.6539346876126566,
+ "learning_rate": 5.413590157907153e-06,
+ "loss": 0.3346075534820557,
+ "num_input_tokens_seen": 615270632,
+ "step": 14070,
+ "train_runtime": 108826.6066,
+ "train_tokens_per_second": 5653.678
+ },
+ {
+ "epoch": 0.9476838136277942,
+ "grad_norm": 0.7338889594818504,
+ "learning_rate": 5.4109550793373255e-06,
+ "loss": 0.355800724029541,
+ "num_input_tokens_seen": 615493608,
+ "step": 14075,
+ "train_runtime": 108868.2257,
+ "train_tokens_per_second": 5653.565
+ },
+ {
+ "epoch": 0.9480204686237543,
+ "grad_norm": 0.7464991259586956,
+ "learning_rate": 5.40831988584522e-06,
+ "loss": 0.38836262226104734,
+ "num_input_tokens_seen": 615719216,
+ "step": 14080,
+ "train_runtime": 108904.0303,
+ "train_tokens_per_second": 5653.778
+ },
+ {
+ "epoch": 0.9483571236197145,
+ "grad_norm": 0.7069248942260824,
+ "learning_rate": 5.40568457816776e-06,
+ "loss": 0.37587385177612304,
+ "num_input_tokens_seen": 615951848,
+ "step": 14085,
+ "train_runtime": 108947.057,
+ "train_tokens_per_second": 5653.68
+ },
+ {
+ "epoch": 0.9486937786156746,
+ "grad_norm": 0.6655009611716466,
+ "learning_rate": 5.403049157041903e-06,
+ "loss": 0.35560946464538573,
+ "num_input_tokens_seen": 616172928,
+ "step": 14090,
+ "train_runtime": 108984.6239,
+ "train_tokens_per_second": 5653.76
+ },
+ {
+ "epoch": 0.9490304336116347,
+ "grad_norm": 0.7730572202934075,
+ "learning_rate": 5.400413623204635e-06,
+ "loss": 0.35712380409240724,
+ "num_input_tokens_seen": 616390224,
+ "step": 14095,
+ "train_runtime": 109023.2559,
+ "train_tokens_per_second": 5653.75
+ },
+ {
+ "epoch": 0.9493670886075949,
+ "grad_norm": 0.7946160574063879,
+ "learning_rate": 5.397777977392973e-06,
+ "loss": 0.39092726707458497,
+ "num_input_tokens_seen": 616612064,
+ "step": 14100,
+ "train_runtime": 109071.1955,
+ "train_tokens_per_second": 5653.299
+ },
+ {
+ "epoch": 0.949703743603555,
+ "grad_norm": 0.7261331777714058,
+ "learning_rate": 5.395142220343967e-06,
+ "loss": 0.37581348419189453,
+ "num_input_tokens_seen": 616835560,
+ "step": 14105,
+ "train_runtime": 109114.6463,
+ "train_tokens_per_second": 5653.096
+ },
+ {
+ "epoch": 0.9500403985995152,
+ "grad_norm": 0.7685617426175168,
+ "learning_rate": 5.3925063527947e-06,
+ "loss": 0.34862990379333497,
+ "num_input_tokens_seen": 617051200,
+ "step": 14110,
+ "train_runtime": 109161.8492,
+ "train_tokens_per_second": 5652.627
+ },
+ {
+ "epoch": 0.9503770535954753,
+ "grad_norm": 0.771130071661086,
+ "learning_rate": 5.389870375482283e-06,
+ "loss": 0.37904350757598876,
+ "num_input_tokens_seen": 617248976,
+ "step": 14115,
+ "train_runtime": 109194.634,
+ "train_tokens_per_second": 5652.741
+ },
+ {
+ "epoch": 0.9507137085914354,
+ "grad_norm": 0.7205683359655156,
+ "learning_rate": 5.387234289143859e-06,
+ "loss": 0.331813383102417,
+ "num_input_tokens_seen": 617469104,
+ "step": 14120,
+ "train_runtime": 109226.9679,
+ "train_tokens_per_second": 5653.083
+ },
+ {
+ "epoch": 0.9510503635873956,
+ "grad_norm": 0.7693964177160706,
+ "learning_rate": 5.384598094516601e-06,
+ "loss": 0.3729023218154907,
+ "num_input_tokens_seen": 617692328,
+ "step": 14125,
+ "train_runtime": 109271.6235,
+ "train_tokens_per_second": 5652.816
+ },
+ {
+ "epoch": 0.9513870185833557,
+ "grad_norm": 0.6640672951882179,
+ "learning_rate": 5.381961792337712e-06,
+ "loss": 0.3759523153305054,
+ "num_input_tokens_seen": 617921872,
+ "step": 14130,
+ "train_runtime": 109307.5627,
+ "train_tokens_per_second": 5653.057
+ },
+ {
+ "epoch": 0.9517236735793159,
+ "grad_norm": 0.6678437973528463,
+ "learning_rate": 5.3793253833444245e-06,
+ "loss": 0.3401124715805054,
+ "num_input_tokens_seen": 618147624,
+ "step": 14135,
+ "train_runtime": 109342.9092,
+ "train_tokens_per_second": 5653.294
+ },
+ {
+ "epoch": 0.952060328575276,
+ "grad_norm": 0.6761734645912748,
+ "learning_rate": 5.3766888682740045e-06,
+ "loss": 0.3542716264724731,
+ "num_input_tokens_seen": 618375336,
+ "step": 14140,
+ "train_runtime": 109378.6235,
+ "train_tokens_per_second": 5653.53
+ },
+ {
+ "epoch": 0.9523969835712361,
+ "grad_norm": 0.7917338745636189,
+ "learning_rate": 5.374052247863745e-06,
+ "loss": 0.3719951152801514,
+ "num_input_tokens_seen": 618588568,
+ "step": 14145,
+ "train_runtime": 109412.8524,
+ "train_tokens_per_second": 5653.71
+ },
+ {
+ "epoch": 0.9527336385671963,
+ "grad_norm": 0.7643089958554735,
+ "learning_rate": 5.371415522850966e-06,
+ "loss": 0.34739990234375,
+ "num_input_tokens_seen": 618807136,
+ "step": 14150,
+ "train_runtime": 109455.6515,
+ "train_tokens_per_second": 5653.496
+ },
+ {
+ "epoch": 0.9530702935631564,
+ "grad_norm": 0.7334076755125779,
+ "learning_rate": 5.368778693973019e-06,
+ "loss": 0.35079226493835447,
+ "num_input_tokens_seen": 619020192,
+ "step": 14155,
+ "train_runtime": 109497.2351,
+ "train_tokens_per_second": 5653.295
+ },
+ {
+ "epoch": 0.9534069485591166,
+ "grad_norm": 0.6518007663729203,
+ "learning_rate": 5.366141761967287e-06,
+ "loss": 0.3731173753738403,
+ "num_input_tokens_seen": 619234312,
+ "step": 14160,
+ "train_runtime": 109535.8084,
+ "train_tokens_per_second": 5653.259
+ },
+ {
+ "epoch": 0.9537436035550767,
+ "grad_norm": 0.884684695372941,
+ "learning_rate": 5.363504727571179e-06,
+ "loss": 0.38480224609375,
+ "num_input_tokens_seen": 619455112,
+ "step": 14165,
+ "train_runtime": 109574.3358,
+ "train_tokens_per_second": 5653.286
+ },
+ {
+ "epoch": 0.954080258551037,
+ "grad_norm": 0.7444612393985186,
+ "learning_rate": 5.360867591522133e-06,
+ "loss": 0.4258725166320801,
+ "num_input_tokens_seen": 619686368,
+ "step": 14170,
+ "train_runtime": 109614.4844,
+ "train_tokens_per_second": 5653.326
+ },
+ {
+ "epoch": 0.9544169135469971,
+ "grad_norm": 0.766088989031395,
+ "learning_rate": 5.358230354557617e-06,
+ "loss": 0.34602572917938235,
+ "num_input_tokens_seen": 619912184,
+ "step": 14175,
+ "train_runtime": 109647.5189,
+ "train_tokens_per_second": 5653.682
+ },
+ {
+ "epoch": 0.9547535685429572,
+ "grad_norm": 0.8426997667284004,
+ "learning_rate": 5.355593017415123e-06,
+ "loss": 0.39844369888305664,
+ "num_input_tokens_seen": 620138400,
+ "step": 14180,
+ "train_runtime": 109681.0251,
+ "train_tokens_per_second": 5654.017
+ },
+ {
+ "epoch": 0.9550902235389174,
+ "grad_norm": 0.6855456694998082,
+ "learning_rate": 5.352955580832176e-06,
+ "loss": 0.36531069278717043,
+ "num_input_tokens_seen": 620363336,
+ "step": 14185,
+ "train_runtime": 109720.9773,
+ "train_tokens_per_second": 5654.008
+ },
+ {
+ "epoch": 0.9554268785348775,
+ "grad_norm": 0.6489568473620863,
+ "learning_rate": 5.350318045546326e-06,
+ "loss": 0.378409218788147,
+ "num_input_tokens_seen": 620586608,
+ "step": 14190,
+ "train_runtime": 109762.864,
+ "train_tokens_per_second": 5653.885
+ },
+ {
+ "epoch": 0.9557635335308377,
+ "grad_norm": 0.690203133783407,
+ "learning_rate": 5.347680412295152e-06,
+ "loss": 0.3777376413345337,
+ "num_input_tokens_seen": 620803016,
+ "step": 14195,
+ "train_runtime": 109803.0202,
+ "train_tokens_per_second": 5653.788
+ },
+ {
+ "epoch": 0.9561001885267978,
+ "grad_norm": 0.8763898868581147,
+ "learning_rate": 5.34504268181626e-06,
+ "loss": 0.37516131401062014,
+ "num_input_tokens_seen": 621009712,
+ "step": 14200,
+ "train_runtime": 109842.1757,
+ "train_tokens_per_second": 5653.655
+ },
+ {
+ "epoch": 0.9564368435227579,
+ "grad_norm": 0.8351648882525854,
+ "learning_rate": 5.342404854847282e-06,
+ "loss": 0.36751456260681153,
+ "num_input_tokens_seen": 621209912,
+ "step": 14205,
+ "train_runtime": 109878.6528,
+ "train_tokens_per_second": 5653.6
+ },
+ {
+ "epoch": 0.9567734985187181,
+ "grad_norm": 0.695513608466477,
+ "learning_rate": 5.3397669321258776e-06,
+ "loss": 0.3533796310424805,
+ "num_input_tokens_seen": 621439136,
+ "step": 14210,
+ "train_runtime": 109913.5419,
+ "train_tokens_per_second": 5653.891
+ },
+ {
+ "epoch": 0.9571101535146782,
+ "grad_norm": 0.7212650216714175,
+ "learning_rate": 5.337128914389734e-06,
+ "loss": 0.3940083980560303,
+ "num_input_tokens_seen": 621670256,
+ "step": 14215,
+ "train_runtime": 109948.6255,
+ "train_tokens_per_second": 5654.189
+ },
+ {
+ "epoch": 0.9574468085106383,
+ "grad_norm": 0.7045386935112185,
+ "learning_rate": 5.334490802376565e-06,
+ "loss": 0.3630239009857178,
+ "num_input_tokens_seen": 621896344,
+ "step": 14220,
+ "train_runtime": 109992.4071,
+ "train_tokens_per_second": 5653.993
+ },
+ {
+ "epoch": 0.9577834635065985,
+ "grad_norm": 1.0648426671565956,
+ "learning_rate": 5.33185259682411e-06,
+ "loss": 0.38904123306274413,
+ "num_input_tokens_seen": 622111488,
+ "step": 14225,
+ "train_runtime": 110028.3421,
+ "train_tokens_per_second": 5654.102
+ },
+ {
+ "epoch": 0.9581201185025586,
+ "grad_norm": 0.6854092676166333,
+ "learning_rate": 5.329214298470134e-06,
+ "loss": 0.3952740907669067,
+ "num_input_tokens_seen": 622335752,
+ "step": 14230,
+ "train_runtime": 110067.3201,
+ "train_tokens_per_second": 5654.137
+ },
+ {
+ "epoch": 0.9584567734985188,
+ "grad_norm": 0.6504683472772627,
+ "learning_rate": 5.326575908052428e-06,
+ "loss": 0.34938879013061525,
+ "num_input_tokens_seen": 622563512,
+ "step": 14235,
+ "train_runtime": 110103.0055,
+ "train_tokens_per_second": 5654.373
+ },
+ {
+ "epoch": 0.9587934284944789,
+ "grad_norm": 0.8743571550942602,
+ "learning_rate": 5.323937426308813e-06,
+ "loss": 0.4064329147338867,
+ "num_input_tokens_seen": 622789680,
+ "step": 14240,
+ "train_runtime": 110140.2833,
+ "train_tokens_per_second": 5654.513
+ },
+ {
+ "epoch": 0.959130083490439,
+ "grad_norm": 0.7364135099684984,
+ "learning_rate": 5.321298853977128e-06,
+ "loss": 0.3482068538665771,
+ "num_input_tokens_seen": 623030288,
+ "step": 14245,
+ "train_runtime": 110172.6755,
+ "train_tokens_per_second": 5655.035
+ },
+ {
+ "epoch": 0.9594667384863992,
+ "grad_norm": 1.04514521012973,
+ "learning_rate": 5.318660191795244e-06,
+ "loss": 0.35120339393615724,
+ "num_input_tokens_seen": 623250248,
+ "step": 14250,
+ "train_runtime": 110210.2613,
+ "train_tokens_per_second": 5655.102
+ },
+ {
+ "epoch": 0.9598033934823593,
+ "grad_norm": 0.6859815669971231,
+ "learning_rate": 5.316021440501053e-06,
+ "loss": 0.32753448486328124,
+ "num_input_tokens_seen": 623468032,
+ "step": 14255,
+ "train_runtime": 110254.6316,
+ "train_tokens_per_second": 5654.801
+ },
+ {
+ "epoch": 0.9601400484783195,
+ "grad_norm": 0.6696719492054304,
+ "learning_rate": 5.313382600832474e-06,
+ "loss": 0.32216849327087405,
+ "num_input_tokens_seen": 623687536,
+ "step": 14260,
+ "train_runtime": 110295.1281,
+ "train_tokens_per_second": 5654.715
+ },
+ {
+ "epoch": 0.9604767034742796,
+ "grad_norm": 0.6796334454419847,
+ "learning_rate": 5.310743673527451e-06,
+ "loss": 0.3820059776306152,
+ "num_input_tokens_seen": 623916888,
+ "step": 14265,
+ "train_runtime": 110337.4663,
+ "train_tokens_per_second": 5654.624
+ },
+ {
+ "epoch": 0.9608133584702397,
+ "grad_norm": 0.7257174247083755,
+ "learning_rate": 5.3081046593239496e-06,
+ "loss": 0.38223018646240237,
+ "num_input_tokens_seen": 624136416,
+ "step": 14270,
+ "train_runtime": 110378.6302,
+ "train_tokens_per_second": 5654.504
+ },
+ {
+ "epoch": 0.9611500134661999,
+ "grad_norm": 0.7285256715224561,
+ "learning_rate": 5.305465558959964e-06,
+ "loss": 0.366811728477478,
+ "num_input_tokens_seen": 624349152,
+ "step": 14275,
+ "train_runtime": 110421.354,
+ "train_tokens_per_second": 5654.243
+ },
+ {
+ "epoch": 0.96148666846216,
+ "grad_norm": 0.7064345566990189,
+ "learning_rate": 5.302826373173506e-06,
+ "loss": 0.3593759536743164,
+ "num_input_tokens_seen": 624577352,
+ "step": 14280,
+ "train_runtime": 110460.1396,
+ "train_tokens_per_second": 5654.323
+ },
+ {
+ "epoch": 0.9618233234581202,
+ "grad_norm": 0.7256414932456383,
+ "learning_rate": 5.3001871027026194e-06,
+ "loss": 0.3493657112121582,
+ "num_input_tokens_seen": 624788664,
+ "step": 14285,
+ "train_runtime": 110502.0095,
+ "train_tokens_per_second": 5654.093
+ },
+ {
+ "epoch": 0.9621599784540803,
+ "grad_norm": 0.7017567146140217,
+ "learning_rate": 5.2975477482853685e-06,
+ "loss": 0.3475620746612549,
+ "num_input_tokens_seen": 625013976,
+ "step": 14290,
+ "train_runtime": 110543.2223,
+ "train_tokens_per_second": 5654.023
+ },
+ {
+ "epoch": 0.9624966334500404,
+ "grad_norm": 0.8025787225363398,
+ "learning_rate": 5.294908310659838e-06,
+ "loss": 0.3636843204498291,
+ "num_input_tokens_seen": 625243416,
+ "step": 14295,
+ "train_runtime": 110579.9936,
+ "train_tokens_per_second": 5654.218
+ },
+ {
+ "epoch": 0.9628332884460006,
+ "grad_norm": 0.796752899535943,
+ "learning_rate": 5.292268790564138e-06,
+ "loss": 0.3458188772201538,
+ "num_input_tokens_seen": 625456472,
+ "step": 14300,
+ "train_runtime": 110621.9221,
+ "train_tokens_per_second": 5654.001
+ },
+ {
+ "epoch": 0.9631699434419607,
+ "grad_norm": 0.7622734681695542,
+ "learning_rate": 5.289629188736403e-06,
+ "loss": 0.39026923179626466,
+ "num_input_tokens_seen": 625676128,
+ "step": 14305,
+ "train_runtime": 110667.8056,
+ "train_tokens_per_second": 5653.642
+ },
+ {
+ "epoch": 0.9635065984379209,
+ "grad_norm": 0.973852297067902,
+ "learning_rate": 5.286989505914788e-06,
+ "loss": 0.38088412284851075,
+ "num_input_tokens_seen": 625904384,
+ "step": 14310,
+ "train_runtime": 110702.6608,
+ "train_tokens_per_second": 5653.924
+ },
+ {
+ "epoch": 0.963843253433881,
+ "grad_norm": 0.8147025206812348,
+ "learning_rate": 5.284349742837475e-06,
+ "loss": 0.3758919954299927,
+ "num_input_tokens_seen": 626134784,
+ "step": 14315,
+ "train_runtime": 110744.3819,
+ "train_tokens_per_second": 5653.874
+ },
+ {
+ "epoch": 0.9641799084298411,
+ "grad_norm": 0.8865610211887707,
+ "learning_rate": 5.281709900242662e-06,
+ "loss": 0.35522916316986086,
+ "num_input_tokens_seen": 626357296,
+ "step": 14320,
+ "train_runtime": 110789.1337,
+ "train_tokens_per_second": 5653.599
+ },
+ {
+ "epoch": 0.9645165634258013,
+ "grad_norm": 0.7685303095133553,
+ "learning_rate": 5.279069978868574e-06,
+ "loss": 0.3454322099685669,
+ "num_input_tokens_seen": 626576680,
+ "step": 14325,
+ "train_runtime": 110823.2787,
+ "train_tokens_per_second": 5653.836
+ },
+ {
+ "epoch": 0.9648532184217614,
+ "grad_norm": 0.7582279578264484,
+ "learning_rate": 5.276429979453455e-06,
+ "loss": 0.36722531318664553,
+ "num_input_tokens_seen": 626795800,
+ "step": 14330,
+ "train_runtime": 110855.2605,
+ "train_tokens_per_second": 5654.182
+ },
+ {
+ "epoch": 0.9651898734177216,
+ "grad_norm": 0.6609643186180265,
+ "learning_rate": 5.273789902735575e-06,
+ "loss": 0.36404809951782224,
+ "num_input_tokens_seen": 627011600,
+ "step": 14335,
+ "train_runtime": 110890.1684,
+ "train_tokens_per_second": 5654.348
+ },
+ {
+ "epoch": 0.9655265284136817,
+ "grad_norm": 0.7160660322102089,
+ "learning_rate": 5.27114974945322e-06,
+ "loss": 0.33545889854431155,
+ "num_input_tokens_seen": 627226888,
+ "step": 14340,
+ "train_runtime": 110937.4224,
+ "train_tokens_per_second": 5653.88
+ },
+ {
+ "epoch": 0.9658631834096418,
+ "grad_norm": 0.7371545286431122,
+ "learning_rate": 5.2685095203447046e-06,
+ "loss": 0.3596933841705322,
+ "num_input_tokens_seen": 627453768,
+ "step": 14345,
+ "train_runtime": 110976.5578,
+ "train_tokens_per_second": 5653.931
+ },
+ {
+ "epoch": 0.966199838405602,
+ "grad_norm": 0.6944067783154944,
+ "learning_rate": 5.265869216148357e-06,
+ "loss": 0.352811861038208,
+ "num_input_tokens_seen": 627673216,
+ "step": 14350,
+ "train_runtime": 111014.9066,
+ "train_tokens_per_second": 5653.954
+ },
+ {
+ "epoch": 0.9665364934015621,
+ "grad_norm": 0.7335669224993876,
+ "learning_rate": 5.263228837602532e-06,
+ "loss": 0.42299542427062986,
+ "num_input_tokens_seen": 627887552,
+ "step": 14355,
+ "train_runtime": 111050.8572,
+ "train_tokens_per_second": 5654.054
+ },
+ {
+ "epoch": 0.9668731483975223,
+ "grad_norm": 0.83794134450925,
+ "learning_rate": 5.2605883854456e-06,
+ "loss": 0.3737421989440918,
+ "num_input_tokens_seen": 628108328,
+ "step": 14360,
+ "train_runtime": 111089.5086,
+ "train_tokens_per_second": 5654.074
+ },
+ {
+ "epoch": 0.9672098033934824,
+ "grad_norm": 0.6342873275404923,
+ "learning_rate": 5.257947860415957e-06,
+ "loss": 0.3357997894287109,
+ "num_input_tokens_seen": 628335808,
+ "step": 14365,
+ "train_runtime": 111122.6788,
+ "train_tokens_per_second": 5654.434
+ },
+ {
+ "epoch": 0.9675464583894425,
+ "grad_norm": 0.8822334071658794,
+ "learning_rate": 5.255307263252019e-06,
+ "loss": 0.3541123867034912,
+ "num_input_tokens_seen": 628566792,
+ "step": 14370,
+ "train_runtime": 111156.9119,
+ "train_tokens_per_second": 5654.77
+ },
+ {
+ "epoch": 0.9678831133854027,
+ "grad_norm": 0.674709973682241,
+ "learning_rate": 5.2526665946922185e-06,
+ "loss": 0.3823538303375244,
+ "num_input_tokens_seen": 628784336,
+ "step": 14375,
+ "train_runtime": 111196.5312,
+ "train_tokens_per_second": 5654.712
+ },
+ {
+ "epoch": 0.9682197683813628,
+ "grad_norm": 0.658823173011451,
+ "learning_rate": 5.25002585547501e-06,
+ "loss": 0.35442066192626953,
+ "num_input_tokens_seen": 629008320,
+ "step": 14380,
+ "train_runtime": 111234.2162,
+ "train_tokens_per_second": 5654.81
+ },
+ {
+ "epoch": 0.968556423377323,
+ "grad_norm": 0.723111666266317,
+ "learning_rate": 5.2473850463388675e-06,
+ "loss": 0.3943564176559448,
+ "num_input_tokens_seen": 629232256,
+ "step": 14385,
+ "train_runtime": 111274.7039,
+ "train_tokens_per_second": 5654.765
+ },
+ {
+ "epoch": 0.9688930783732831,
+ "grad_norm": 0.6809256630686321,
+ "learning_rate": 5.244744168022288e-06,
+ "loss": 0.3620326995849609,
+ "num_input_tokens_seen": 629455456,
+ "step": 14390,
+ "train_runtime": 111314.9521,
+ "train_tokens_per_second": 5654.725
+ },
+ {
+ "epoch": 0.9692297333692432,
+ "grad_norm": 0.6956920321641313,
+ "learning_rate": 5.24210322126378e-06,
+ "loss": 0.36865873336791993,
+ "num_input_tokens_seen": 629691848,
+ "step": 14395,
+ "train_runtime": 111353.3824,
+ "train_tokens_per_second": 5654.896
+ },
+ {
+ "epoch": 0.9695663883652034,
+ "grad_norm": 0.7152647788363251,
+ "learning_rate": 5.23946220680188e-06,
+ "loss": 0.3680101871490479,
+ "num_input_tokens_seen": 629928368,
+ "step": 14400,
+ "train_runtime": 111391.3207,
+ "train_tokens_per_second": 5655.094
+ },
+ {
+ "epoch": 0.9699030433611635,
+ "grad_norm": 0.7371345942960084,
+ "learning_rate": 5.236821125375136e-06,
+ "loss": 0.3584733486175537,
+ "num_input_tokens_seen": 630150304,
+ "step": 14405,
+ "train_runtime": 111429.6066,
+ "train_tokens_per_second": 5655.142
+ },
+ {
+ "epoch": 0.9702396983571236,
+ "grad_norm": 0.6525643955944114,
+ "learning_rate": 5.234179977722119e-06,
+ "loss": 0.31973896026611326,
+ "num_input_tokens_seen": 630389640,
+ "step": 14410,
+ "train_runtime": 111465.3624,
+ "train_tokens_per_second": 5655.476
+ },
+ {
+ "epoch": 0.9705763533530838,
+ "grad_norm": 0.6728032683654267,
+ "learning_rate": 5.23153876458142e-06,
+ "loss": 0.36993358135223386,
+ "num_input_tokens_seen": 630599464,
+ "step": 14415,
+ "train_runtime": 111509.2822,
+ "train_tokens_per_second": 5655.13
+ },
+ {
+ "epoch": 0.9709130083490439,
+ "grad_norm": 0.7435706382156235,
+ "learning_rate": 5.2288974866916435e-06,
+ "loss": 0.36161553859710693,
+ "num_input_tokens_seen": 630794912,
+ "step": 14420,
+ "train_runtime": 111547.8849,
+ "train_tokens_per_second": 5654.925
+ },
+ {
+ "epoch": 0.9712496633450041,
+ "grad_norm": 0.702712902975055,
+ "learning_rate": 5.226256144791414e-06,
+ "loss": 0.36026706695556643,
+ "num_input_tokens_seen": 631000992,
+ "step": 14425,
+ "train_runtime": 111582.5203,
+ "train_tokens_per_second": 5655.016
+ },
+ {
+ "epoch": 0.9715863183409642,
+ "grad_norm": 0.7340375348827854,
+ "learning_rate": 5.223614739619375e-06,
+ "loss": 0.37513973712921145,
+ "num_input_tokens_seen": 631202192,
+ "step": 14430,
+ "train_runtime": 111620.6231,
+ "train_tokens_per_second": 5654.889
+ },
+ {
+ "epoch": 0.9719229733369243,
+ "grad_norm": 0.6505020134915022,
+ "learning_rate": 5.2209732719141855e-06,
+ "loss": 0.33454012870788574,
+ "num_input_tokens_seen": 631411712,
+ "step": 14435,
+ "train_runtime": 111660.3902,
+ "train_tokens_per_second": 5654.751
+ },
+ {
+ "epoch": 0.9722596283328845,
+ "grad_norm": 0.6834948719980648,
+ "learning_rate": 5.218331742414528e-06,
+ "loss": 0.3455739736557007,
+ "num_input_tokens_seen": 631641904,
+ "step": 14440,
+ "train_runtime": 111694.5696,
+ "train_tokens_per_second": 5655.082
+ },
+ {
+ "epoch": 0.9725962833288446,
+ "grad_norm": 0.7031614396342304,
+ "learning_rate": 5.215690151859094e-06,
+ "loss": 0.3380829572677612,
+ "num_input_tokens_seen": 631848056,
+ "step": 14445,
+ "train_runtime": 111733.8193,
+ "train_tokens_per_second": 5654.94
+ },
+ {
+ "epoch": 0.9729329383248048,
+ "grad_norm": 0.737272924329504,
+ "learning_rate": 5.213048500986597e-06,
+ "loss": 0.3667490243911743,
+ "num_input_tokens_seen": 632068664,
+ "step": 14450,
+ "train_runtime": 111782.1877,
+ "train_tokens_per_second": 5654.467
+ },
+ {
+ "epoch": 0.9732695933207649,
+ "grad_norm": 0.8384823634165054,
+ "learning_rate": 5.210406790535766e-06,
+ "loss": 0.3663896083831787,
+ "num_input_tokens_seen": 632279544,
+ "step": 14455,
+ "train_runtime": 111817.1549,
+ "train_tokens_per_second": 5654.584
+ },
+ {
+ "epoch": 0.973606248316725,
+ "grad_norm": 0.7339723372461647,
+ "learning_rate": 5.207765021245346e-06,
+ "loss": 0.3660223722457886,
+ "num_input_tokens_seen": 632492984,
+ "step": 14460,
+ "train_runtime": 111858.5212,
+ "train_tokens_per_second": 5654.401
+ },
+ {
+ "epoch": 0.9739429033126852,
+ "grad_norm": 0.7329956341586107,
+ "learning_rate": 5.2051231938541015e-06,
+ "loss": 0.3708993911743164,
+ "num_input_tokens_seen": 632694568,
+ "step": 14465,
+ "train_runtime": 111894.0881,
+ "train_tokens_per_second": 5654.406
+ },
+ {
+ "epoch": 0.9742795583086453,
+ "grad_norm": 0.7751460530884079,
+ "learning_rate": 5.202481309100808e-06,
+ "loss": 0.3687734127044678,
+ "num_input_tokens_seen": 632904360,
+ "step": 14470,
+ "train_runtime": 111935.2636,
+ "train_tokens_per_second": 5654.2
+ },
+ {
+ "epoch": 0.9746162133046055,
+ "grad_norm": 0.6898579692907033,
+ "learning_rate": 5.199839367724266e-06,
+ "loss": 0.3596921920776367,
+ "num_input_tokens_seen": 633132800,
+ "step": 14475,
+ "train_runtime": 111976.0175,
+ "train_tokens_per_second": 5654.182
+ },
+ {
+ "epoch": 0.9749528683005656,
+ "grad_norm": 0.7400137611963173,
+ "learning_rate": 5.197197370463278e-06,
+ "loss": 0.343227481842041,
+ "num_input_tokens_seen": 633370648,
+ "step": 14480,
+ "train_runtime": 112017.4444,
+ "train_tokens_per_second": 5654.214
+ },
+ {
+ "epoch": 0.9752895232965257,
+ "grad_norm": 0.6958590027819616,
+ "learning_rate": 5.194555318056675e-06,
+ "loss": 0.39154577255249023,
+ "num_input_tokens_seen": 633605816,
+ "step": 14485,
+ "train_runtime": 112050.1996,
+ "train_tokens_per_second": 5654.66
+ },
+ {
+ "epoch": 0.9756261782924859,
+ "grad_norm": 0.9461003642017359,
+ "learning_rate": 5.191913211243299e-06,
+ "loss": 0.37716875076293943,
+ "num_input_tokens_seen": 633832456,
+ "step": 14490,
+ "train_runtime": 112095.5877,
+ "train_tokens_per_second": 5654.393
+ },
+ {
+ "epoch": 0.975962833288446,
+ "grad_norm": 0.7296195050735405,
+ "learning_rate": 5.1892710507620055e-06,
+ "loss": 0.36735081672668457,
+ "num_input_tokens_seen": 634052208,
+ "step": 14495,
+ "train_runtime": 112132.123,
+ "train_tokens_per_second": 5654.51
+ },
+ {
+ "epoch": 0.9762994882844062,
+ "grad_norm": 0.713237266588106,
+ "learning_rate": 5.186628837351666e-06,
+ "loss": 0.3544626235961914,
+ "num_input_tokens_seen": 634282120,
+ "step": 14500,
+ "train_runtime": 112168.6707,
+ "train_tokens_per_second": 5654.717
+ },
+ {
+ "epoch": 0.9766361432803663,
+ "grad_norm": 0.7152784729674379,
+ "learning_rate": 5.183986571751167e-06,
+ "loss": 0.3681643962860107,
+ "num_input_tokens_seen": 634503040,
+ "step": 14505,
+ "train_runtime": 112208.6433,
+ "train_tokens_per_second": 5654.672
+ },
+ {
+ "epoch": 0.9769727982763264,
+ "grad_norm": 0.7143902821508025,
+ "learning_rate": 5.181344254699408e-06,
+ "loss": 0.384121036529541,
+ "num_input_tokens_seen": 634718224,
+ "step": 14510,
+ "train_runtime": 112246.9754,
+ "train_tokens_per_second": 5654.658
+ },
+ {
+ "epoch": 0.9773094532722866,
+ "grad_norm": 0.7563338962960594,
+ "learning_rate": 5.178701886935309e-06,
+ "loss": 0.3466785907745361,
+ "num_input_tokens_seen": 634926976,
+ "step": 14515,
+ "train_runtime": 112283.3188,
+ "train_tokens_per_second": 5654.687
+ },
+ {
+ "epoch": 0.9776461082682467,
+ "grad_norm": 0.6709897911941165,
+ "learning_rate": 5.176059469197796e-06,
+ "loss": 0.3563254356384277,
+ "num_input_tokens_seen": 635147104,
+ "step": 14520,
+ "train_runtime": 112318.4084,
+ "train_tokens_per_second": 5654.88
+ },
+ {
+ "epoch": 0.9779827632642069,
+ "grad_norm": 0.84998527536875,
+ "learning_rate": 5.173417002225815e-06,
+ "loss": 0.3569037914276123,
+ "num_input_tokens_seen": 635366792,
+ "step": 14525,
+ "train_runtime": 112357.718,
+ "train_tokens_per_second": 5654.857
+ },
+ {
+ "epoch": 0.978319418260167,
+ "grad_norm": 0.7892134632795219,
+ "learning_rate": 5.170774486758322e-06,
+ "loss": 0.36590094566345216,
+ "num_input_tokens_seen": 635560568,
+ "step": 14530,
+ "train_runtime": 112393.8137,
+ "train_tokens_per_second": 5654.765
+ },
+ {
+ "epoch": 0.9786560732561271,
+ "grad_norm": 0.6842756660337252,
+ "learning_rate": 5.168131923534287e-06,
+ "loss": 0.38362579345703124,
+ "num_input_tokens_seen": 635787144,
+ "step": 14535,
+ "train_runtime": 112428.7709,
+ "train_tokens_per_second": 5655.022
+ },
+ {
+ "epoch": 0.9789927282520873,
+ "grad_norm": 0.6691518942357926,
+ "learning_rate": 5.165489313292696e-06,
+ "loss": 0.35508084297180176,
+ "num_input_tokens_seen": 636000120,
+ "step": 14540,
+ "train_runtime": 112469.9454,
+ "train_tokens_per_second": 5654.845
+ },
+ {
+ "epoch": 0.9793293832480474,
+ "grad_norm": 0.7564357234909826,
+ "learning_rate": 5.162846656772548e-06,
+ "loss": 0.35072224140167235,
+ "num_input_tokens_seen": 636214280,
+ "step": 14545,
+ "train_runtime": 112512.921,
+ "train_tokens_per_second": 5654.589
+ },
+ {
+ "epoch": 0.9796660382440076,
+ "grad_norm": 0.7450101360653049,
+ "learning_rate": 5.16020395471285e-06,
+ "loss": 0.3438509464263916,
+ "num_input_tokens_seen": 636435816,
+ "step": 14550,
+ "train_runtime": 112554.2979,
+ "train_tokens_per_second": 5654.478
+ },
+ {
+ "epoch": 0.9800026932399677,
+ "grad_norm": 0.8137120525286944,
+ "learning_rate": 5.157561207852628e-06,
+ "loss": 0.3831623077392578,
+ "num_input_tokens_seen": 636640224,
+ "step": 14555,
+ "train_runtime": 112591.2086,
+ "train_tokens_per_second": 5654.44
+ },
+ {
+ "epoch": 0.9803393482359278,
+ "grad_norm": 0.7297592438951429,
+ "learning_rate": 5.154918416930915e-06,
+ "loss": 0.3411484479904175,
+ "num_input_tokens_seen": 636858520,
+ "step": 14560,
+ "train_runtime": 112629.598,
+ "train_tokens_per_second": 5654.451
+ },
+ {
+ "epoch": 0.980676003231888,
+ "grad_norm": 0.760019959304734,
+ "learning_rate": 5.152275582686761e-06,
+ "loss": 0.3839298486709595,
+ "num_input_tokens_seen": 637073856,
+ "step": 14565,
+ "train_runtime": 112666.3482,
+ "train_tokens_per_second": 5654.518
+ },
+ {
+ "epoch": 0.9810126582278481,
+ "grad_norm": 0.6959300895402221,
+ "learning_rate": 5.149632705859227e-06,
+ "loss": 0.36935174465179443,
+ "num_input_tokens_seen": 637299440,
+ "step": 14570,
+ "train_runtime": 112707.9899,
+ "train_tokens_per_second": 5654.43
+ },
+ {
+ "epoch": 0.9813493132238083,
+ "grad_norm": 0.7586977280516466,
+ "learning_rate": 5.146989787187383e-06,
+ "loss": 0.37570605278015134,
+ "num_input_tokens_seen": 637529440,
+ "step": 14575,
+ "train_runtime": 112743.4679,
+ "train_tokens_per_second": 5654.691
+ },
+ {
+ "epoch": 0.9816859682197684,
+ "grad_norm": 0.6825264977223254,
+ "learning_rate": 5.144346827410314e-06,
+ "loss": 0.3703890562057495,
+ "num_input_tokens_seen": 637754408,
+ "step": 14580,
+ "train_runtime": 112773.1649,
+ "train_tokens_per_second": 5655.197
+ },
+ {
+ "epoch": 0.9820226232157285,
+ "grad_norm": 0.826290290377258,
+ "learning_rate": 5.141703827267117e-06,
+ "loss": 0.3702056646347046,
+ "num_input_tokens_seen": 637982336,
+ "step": 14585,
+ "train_runtime": 112804.5446,
+ "train_tokens_per_second": 5655.644
+ },
+ {
+ "epoch": 0.9823592782116887,
+ "grad_norm": 0.7212214874544945,
+ "learning_rate": 5.139060787496893e-06,
+ "loss": 0.3754692077636719,
+ "num_input_tokens_seen": 638203872,
+ "step": 14590,
+ "train_runtime": 112841.7495,
+ "train_tokens_per_second": 5655.742
+ },
+ {
+ "epoch": 0.9826959332076488,
+ "grad_norm": 0.6942239873292527,
+ "learning_rate": 5.136417708838766e-06,
+ "loss": 0.3775701284408569,
+ "num_input_tokens_seen": 638407968,
+ "step": 14595,
+ "train_runtime": 112885.1355,
+ "train_tokens_per_second": 5655.377
+ },
+ {
+ "epoch": 0.983032588203609,
+ "grad_norm": 0.7575191844718635,
+ "learning_rate": 5.13377459203186e-06,
+ "loss": 0.3628222465515137,
+ "num_input_tokens_seen": 638616040,
+ "step": 14600,
+ "train_runtime": 112923.0341,
+ "train_tokens_per_second": 5655.321
+ },
+ {
+ "epoch": 0.9833692431995691,
+ "grad_norm": 0.7100885514863359,
+ "learning_rate": 5.131131437815318e-06,
+ "loss": 0.3371574401855469,
+ "num_input_tokens_seen": 638837176,
+ "step": 14605,
+ "train_runtime": 112963.1632,
+ "train_tokens_per_second": 5655.27
+ },
+ {
+ "epoch": 0.9837058981955292,
+ "grad_norm": 0.7208738472250812,
+ "learning_rate": 5.128488246928287e-06,
+ "loss": 0.36636881828308104,
+ "num_input_tokens_seen": 639065552,
+ "step": 14610,
+ "train_runtime": 113000.1458,
+ "train_tokens_per_second": 5655.44
+ },
+ {
+ "epoch": 0.9840425531914894,
+ "grad_norm": 0.7630145727500303,
+ "learning_rate": 5.125845020109925e-06,
+ "loss": 0.37039480209350584,
+ "num_input_tokens_seen": 639285272,
+ "step": 14615,
+ "train_runtime": 113047.3523,
+ "train_tokens_per_second": 5655.022
+ },
+ {
+ "epoch": 0.9843792081874495,
+ "grad_norm": 0.7720868800544706,
+ "learning_rate": 5.123201758099406e-06,
+ "loss": 0.36545164585113527,
+ "num_input_tokens_seen": 639499208,
+ "step": 14620,
+ "train_runtime": 113085.6686,
+ "train_tokens_per_second": 5654.998
+ },
+ {
+ "epoch": 0.9847158631834096,
+ "grad_norm": 0.7035125490360992,
+ "learning_rate": 5.12055846163591e-06,
+ "loss": 0.37010979652404785,
+ "num_input_tokens_seen": 639714312,
+ "step": 14625,
+ "train_runtime": 113123.7227,
+ "train_tokens_per_second": 5654.997
+ },
+ {
+ "epoch": 0.9850525181793698,
+ "grad_norm": 0.6519589735545973,
+ "learning_rate": 5.117915131458623e-06,
+ "loss": 0.3291551828384399,
+ "num_input_tokens_seen": 639938488,
+ "step": 14630,
+ "train_runtime": 113162.111,
+ "train_tokens_per_second": 5655.06
+ },
+ {
+ "epoch": 0.9853891731753299,
+ "grad_norm": 0.8145405370454951,
+ "learning_rate": 5.115271768306745e-06,
+ "loss": 0.36716470718383787,
+ "num_input_tokens_seen": 640152840,
+ "step": 14635,
+ "train_runtime": 113199.5845,
+ "train_tokens_per_second": 5655.081
+ },
+ {
+ "epoch": 0.9857258281712901,
+ "grad_norm": 0.790933823693933,
+ "learning_rate": 5.112628372919485e-06,
+ "loss": 0.3815732717514038,
+ "num_input_tokens_seen": 640369952,
+ "step": 14640,
+ "train_runtime": 113234.1823,
+ "train_tokens_per_second": 5655.271
+ },
+ {
+ "epoch": 0.9860624831672502,
+ "grad_norm": 0.9699782510514777,
+ "learning_rate": 5.1099849460360595e-06,
+ "loss": 0.34613332748413084,
+ "num_input_tokens_seen": 640567240,
+ "step": 14645,
+ "train_runtime": 113271.7361,
+ "train_tokens_per_second": 5655.137
+ },
+ {
+ "epoch": 0.9863991381632103,
+ "grad_norm": 0.6723948378020855,
+ "learning_rate": 5.107341488395695e-06,
+ "loss": 0.3379997730255127,
+ "num_input_tokens_seen": 640788384,
+ "step": 14650,
+ "train_runtime": 113310.4257,
+ "train_tokens_per_second": 5655.158
+ },
+ {
+ "epoch": 0.9867357931591705,
+ "grad_norm": 0.7000157996530737,
+ "learning_rate": 5.104698000737626e-06,
+ "loss": 0.37780165672302246,
+ "num_input_tokens_seen": 641011792,
+ "step": 14655,
+ "train_runtime": 113351.4983,
+ "train_tokens_per_second": 5655.08
+ },
+ {
+ "epoch": 0.9870724481551306,
+ "grad_norm": 0.7543705648632995,
+ "learning_rate": 5.102054483801094e-06,
+ "loss": 0.38533506393432615,
+ "num_input_tokens_seen": 641239984,
+ "step": 14660,
+ "train_runtime": 113390.3486,
+ "train_tokens_per_second": 5655.155
+ },
+ {
+ "epoch": 0.9874091031510908,
+ "grad_norm": 0.7193117605335543,
+ "learning_rate": 5.099410938325351e-06,
+ "loss": 0.3746023178100586,
+ "num_input_tokens_seen": 641468200,
+ "step": 14665,
+ "train_runtime": 113424.7218,
+ "train_tokens_per_second": 5655.453
+ },
+ {
+ "epoch": 0.9877457581470509,
+ "grad_norm": 0.8146622438199124,
+ "learning_rate": 5.096767365049657e-06,
+ "loss": 0.3502329111099243,
+ "num_input_tokens_seen": 641672792,
+ "step": 14670,
+ "train_runtime": 113464.964,
+ "train_tokens_per_second": 5655.25
+ },
+ {
+ "epoch": 0.988082413143011,
+ "grad_norm": 0.6875126263724606,
+ "learning_rate": 5.094123764713277e-06,
+ "loss": 0.3421436309814453,
+ "num_input_tokens_seen": 641904736,
+ "step": 14675,
+ "train_runtime": 113500.904,
+ "train_tokens_per_second": 5655.503
+ },
+ {
+ "epoch": 0.9884190681389712,
+ "grad_norm": 0.7258139194348168,
+ "learning_rate": 5.091480138055487e-06,
+ "loss": 0.36416258811950686,
+ "num_input_tokens_seen": 642136592,
+ "step": 14680,
+ "train_runtime": 113531.8425,
+ "train_tokens_per_second": 5656.004
+ },
+ {
+ "epoch": 0.9887557231349313,
+ "grad_norm": 0.7250476516557031,
+ "learning_rate": 5.088836485815568e-06,
+ "loss": 0.35428576469421386,
+ "num_input_tokens_seen": 642334648,
+ "step": 14685,
+ "train_runtime": 113566.8784,
+ "train_tokens_per_second": 5656.003
+ },
+ {
+ "epoch": 0.9890923781308915,
+ "grad_norm": 0.7385563837095339,
+ "learning_rate": 5.086192808732809e-06,
+ "loss": 0.35656602382659913,
+ "num_input_tokens_seen": 642538480,
+ "step": 14690,
+ "train_runtime": 113608.5759,
+ "train_tokens_per_second": 5655.722
+ },
+ {
+ "epoch": 0.9894290331268516,
+ "grad_norm": 0.7946980704838973,
+ "learning_rate": 5.083549107546505e-06,
+ "loss": 0.36766040325164795,
+ "num_input_tokens_seen": 642744792,
+ "step": 14695,
+ "train_runtime": 113647.6019,
+ "train_tokens_per_second": 5655.595
+ },
+ {
+ "epoch": 0.9897656881228117,
+ "grad_norm": 0.771038431226611,
+ "learning_rate": 5.080905382995961e-06,
+ "loss": 0.38895292282104493,
+ "num_input_tokens_seen": 642974608,
+ "step": 14700,
+ "train_runtime": 113683.9469,
+ "train_tokens_per_second": 5655.808
+ },
+ {
+ "epoch": 0.9901023431187719,
+ "grad_norm": 0.7343676499490034,
+ "learning_rate": 5.0782616358204835e-06,
+ "loss": 0.37378993034362795,
+ "num_input_tokens_seen": 643191928,
+ "step": 14705,
+ "train_runtime": 113723.5317,
+ "train_tokens_per_second": 5655.751
+ },
+ {
+ "epoch": 0.990438998114732,
+ "grad_norm": 0.8524692289540274,
+ "learning_rate": 5.075617866759389e-06,
+ "loss": 0.33492867946624755,
+ "num_input_tokens_seen": 643403376,
+ "step": 14710,
+ "train_runtime": 113766.092,
+ "train_tokens_per_second": 5655.493
+ },
+ {
+ "epoch": 0.9907756531106922,
+ "grad_norm": 0.7901150595595263,
+ "learning_rate": 5.0729740765519995e-06,
+ "loss": 0.4053798675537109,
+ "num_input_tokens_seen": 643617760,
+ "step": 14715,
+ "train_runtime": 113802.6369,
+ "train_tokens_per_second": 5655.561
+ },
+ {
+ "epoch": 0.9911123081066523,
+ "grad_norm": 0.7850703195564729,
+ "learning_rate": 5.070330265937643e-06,
+ "loss": 0.3686873435974121,
+ "num_input_tokens_seen": 643827376,
+ "step": 14720,
+ "train_runtime": 113838.4716,
+ "train_tokens_per_second": 5655.622
+ },
+ {
+ "epoch": 0.9914489631026124,
+ "grad_norm": 0.7339072075707702,
+ "learning_rate": 5.067686435655651e-06,
+ "loss": 0.3935819625854492,
+ "num_input_tokens_seen": 644051992,
+ "step": 14725,
+ "train_runtime": 113875.0085,
+ "train_tokens_per_second": 5655.78
+ },
+ {
+ "epoch": 0.9917856180985726,
+ "grad_norm": 0.7717913369926017,
+ "learning_rate": 5.065042586445363e-06,
+ "loss": 0.39838461875915526,
+ "num_input_tokens_seen": 644255512,
+ "step": 14730,
+ "train_runtime": 113907.4094,
+ "train_tokens_per_second": 5655.958
+ },
+ {
+ "epoch": 0.9921222730945327,
+ "grad_norm": 0.7852173126276834,
+ "learning_rate": 5.0623987190461245e-06,
+ "loss": 0.30464193820953367,
+ "num_input_tokens_seen": 644460824,
+ "step": 14735,
+ "train_runtime": 113944.9392,
+ "train_tokens_per_second": 5655.897
+ },
+ {
+ "epoch": 0.9924589280904929,
+ "grad_norm": 0.7913343599249734,
+ "learning_rate": 5.059754834197281e-06,
+ "loss": 0.3790989875793457,
+ "num_input_tokens_seen": 644675152,
+ "step": 14740,
+ "train_runtime": 113988.6481,
+ "train_tokens_per_second": 5655.608
+ },
+ {
+ "epoch": 0.992795583086453,
+ "grad_norm": 0.6313018834314664,
+ "learning_rate": 5.05711093263819e-06,
+ "loss": 0.32619643211364746,
+ "num_input_tokens_seen": 644890856,
+ "step": 14745,
+ "train_runtime": 114026.2898,
+ "train_tokens_per_second": 5655.633
+ },
+ {
+ "epoch": 0.9931322380824131,
+ "grad_norm": 0.7700737076224284,
+ "learning_rate": 5.054467015108208e-06,
+ "loss": 0.37743263244628905,
+ "num_input_tokens_seen": 645124328,
+ "step": 14750,
+ "train_runtime": 114061.9307,
+ "train_tokens_per_second": 5655.913
+ },
+ {
+ "epoch": 0.9934688930783733,
+ "grad_norm": 0.8548358470170475,
+ "learning_rate": 5.051823082346701e-06,
+ "loss": 0.37684605121612547,
+ "num_input_tokens_seen": 645342896,
+ "step": 14755,
+ "train_runtime": 114095.225,
+ "train_tokens_per_second": 5656.178
+ },
+ {
+ "epoch": 0.9938055480743334,
+ "grad_norm": 0.7487000463516748,
+ "learning_rate": 5.049179135093033e-06,
+ "loss": 0.3667085409164429,
+ "num_input_tokens_seen": 645572616,
+ "step": 14760,
+ "train_runtime": 114128.9739,
+ "train_tokens_per_second": 5656.518
+ },
+ {
+ "epoch": 0.9941422030702936,
+ "grad_norm": 0.7554586359729867,
+ "learning_rate": 5.046535174086575e-06,
+ "loss": 0.37410502433776854,
+ "num_input_tokens_seen": 645784112,
+ "step": 14765,
+ "train_runtime": 114171.4105,
+ "train_tokens_per_second": 5656.268
+ },
+ {
+ "epoch": 0.9944788580662537,
+ "grad_norm": 0.806201449115856,
+ "learning_rate": 5.043891200066706e-06,
+ "loss": 0.35855536460876464,
+ "num_input_tokens_seen": 646000480,
+ "step": 14770,
+ "train_runtime": 114211.5644,
+ "train_tokens_per_second": 5656.174
+ },
+ {
+ "epoch": 0.9948155130622138,
+ "grad_norm": 0.7970070550101245,
+ "learning_rate": 5.041247213772802e-06,
+ "loss": 0.3647798538208008,
+ "num_input_tokens_seen": 646228928,
+ "step": 14775,
+ "train_runtime": 114248.8453,
+ "train_tokens_per_second": 5656.328
+ },
+ {
+ "epoch": 0.995152168058174,
+ "grad_norm": 0.9274665922453337,
+ "learning_rate": 5.038603215944247e-06,
+ "loss": 0.3697342872619629,
+ "num_input_tokens_seen": 646427096,
+ "step": 14780,
+ "train_runtime": 114283.9377,
+ "train_tokens_per_second": 5656.325
+ },
+ {
+ "epoch": 0.9954888230541341,
+ "grad_norm": 0.7200893083392556,
+ "learning_rate": 5.0359592073204254e-06,
+ "loss": 0.34227933883666994,
+ "num_input_tokens_seen": 646643272,
+ "step": 14785,
+ "train_runtime": 114327.3802,
+ "train_tokens_per_second": 5656.067
+ },
+ {
+ "epoch": 0.9958254780500942,
+ "grad_norm": 1.0261926014877596,
+ "learning_rate": 5.033315188640725e-06,
+ "loss": 0.3627659797668457,
+ "num_input_tokens_seen": 646872200,
+ "step": 14790,
+ "train_runtime": 114366.7864,
+ "train_tokens_per_second": 5656.119
+ },
+ {
+ "epoch": 0.9961621330460544,
+ "grad_norm": 0.6681278213263397,
+ "learning_rate": 5.03067116064454e-06,
+ "loss": 0.3450484752655029,
+ "num_input_tokens_seen": 647094832,
+ "step": 14795,
+ "train_runtime": 114408.9697,
+ "train_tokens_per_second": 5655.98
+ },
+ {
+ "epoch": 0.9964987880420145,
+ "grad_norm": 0.6613774299958016,
+ "learning_rate": 5.028027124071262e-06,
+ "loss": 0.36185712814331056,
+ "num_input_tokens_seen": 647301568,
+ "step": 14800,
+ "train_runtime": 114449.8831,
+ "train_tokens_per_second": 5655.764
+ },
+ {
+ "epoch": 0.9968354430379747,
+ "grad_norm": 0.7005897120089515,
+ "learning_rate": 5.025383079660289e-06,
+ "loss": 0.3468491554260254,
+ "num_input_tokens_seen": 647531264,
+ "step": 14805,
+ "train_runtime": 114488.5353,
+ "train_tokens_per_second": 5655.861
+ },
+ {
+ "epoch": 0.9971720980339348,
+ "grad_norm": 0.7061938748348358,
+ "learning_rate": 5.0227390281510166e-06,
+ "loss": 0.3663314342498779,
+ "num_input_tokens_seen": 647753976,
+ "step": 14810,
+ "train_runtime": 114526.0357,
+ "train_tokens_per_second": 5655.954
+ },
+ {
+ "epoch": 0.997508753029895,
+ "grad_norm": 0.7212945140660744,
+ "learning_rate": 5.020094970282849e-06,
+ "loss": 0.3663675308227539,
+ "num_input_tokens_seen": 647955768,
+ "step": 14815,
+ "train_runtime": 114566.6744,
+ "train_tokens_per_second": 5655.709
+ },
+ {
+ "epoch": 0.9978454080258551,
+ "grad_norm": 0.7579147457391882,
+ "learning_rate": 5.017450906795187e-06,
+ "loss": 0.36784412860870364,
+ "num_input_tokens_seen": 648181384,
+ "step": 14820,
+ "train_runtime": 114601.4001,
+ "train_tokens_per_second": 5655.964
+ },
+ {
+ "epoch": 0.9981820630218152,
+ "grad_norm": 0.7496488759479608,
+ "learning_rate": 5.014806838427434e-06,
+ "loss": 0.36796035766601565,
+ "num_input_tokens_seen": 648382672,
+ "step": 14825,
+ "train_runtime": 114646.9011,
+ "train_tokens_per_second": 5655.475
+ },
+ {
+ "epoch": 0.9985187180177754,
+ "grad_norm": 0.6851905965880742,
+ "learning_rate": 5.0121627659189964e-06,
+ "loss": 0.3591431140899658,
+ "num_input_tokens_seen": 648612168,
+ "step": 14830,
+ "train_runtime": 114682.2822,
+ "train_tokens_per_second": 5655.731
+ },
+ {
+ "epoch": 0.9988553730137355,
+ "grad_norm": 0.6696970669236173,
+ "learning_rate": 5.009518690009281e-06,
+ "loss": 0.3566739082336426,
+ "num_input_tokens_seen": 648821208,
+ "step": 14835,
+ "train_runtime": 114724.3707,
+ "train_tokens_per_second": 5655.478
+ },
+ {
+ "epoch": 0.9991920280096956,
+ "grad_norm": 0.7614065048649256,
+ "learning_rate": 5.006874611437692e-06,
+ "loss": 0.3776280879974365,
+ "num_input_tokens_seen": 649046288,
+ "step": 14840,
+ "train_runtime": 114764.8122,
+ "train_tokens_per_second": 5655.447
+ },
+ {
+ "epoch": 0.9995286830056558,
+ "grad_norm": 0.6492499944472357,
+ "learning_rate": 5.004230530943641e-06,
+ "loss": 0.35755281448364257,
+ "num_input_tokens_seen": 649262232,
+ "step": 14845,
+ "train_runtime": 114804.5614,
+ "train_tokens_per_second": 5655.37
+ },
+ {
+ "epoch": 0.9998653380016159,
+ "grad_norm": 0.7184203567667401,
+ "learning_rate": 5.001586449266536e-06,
+ "loss": 0.3260646343231201,
+ "num_input_tokens_seen": 649487024,
+ "step": 14850,
+ "train_runtime": 114845.5757,
+ "train_tokens_per_second": 5655.307
+ }
+ ],
+ "logging_steps": 5,
+ "max_steps": 29704,
+ "num_input_tokens_seen": 649573848,
+ "num_train_epochs": 2,
+ "save_steps": 500,
+ "stateful_callbacks": {
+ "TrainerControl": {
+ "args": {
+ "should_epoch_stop": false,
+ "should_evaluate": false,
+ "should_log": false,
+ "should_save": true,
+ "should_training_stop": false
+ },
+ "attributes": {}
+ }
+ },
+ "total_flos": 1201285497962496.0,
+ "train_batch_size": 1,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/checkpoint-14852/training_args.bin b/checkpoint-14852/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..9f35ce99940d0d7f4628bd149b2a2e946f1e8ec6
--- /dev/null
+++ b/checkpoint-14852/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c414b624070ab6682af9c82937e863fa17942c0203f8862756f767014d4fbaf7
+size 7761
diff --git a/checkpoint-14852/zero_to_fp32.py b/checkpoint-14852/zero_to_fp32.py
new file mode 100644
index 0000000000000000000000000000000000000000..5995d6e6f04e43b989587aa9022a3aef0c66d694
--- /dev/null
+++ b/checkpoint-14852/zero_to_fp32.py
@@ -0,0 +1,760 @@
+#!/usr/bin/env python
+
+# Copyright (c) Microsoft Corporation.
+# SPDX-License-Identifier: Apache-2.0
+
+# DeepSpeed Team
+
+# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets
+# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in
+# the future. Once extracted, the weights don't require DeepSpeed and can be used in any
+# application.
+#
+# example:
+# python zero_to_fp32.py . output_dir/
+# or
+# python zero_to_fp32.py . output_dir/ --safe_serialization
+
+import argparse
+import torch
+import glob
+import math
+import os
+import re
+import gc
+import json
+import numpy as np
+from tqdm import tqdm
+from collections import OrderedDict
+from dataclasses import dataclass
+
+# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with
+# DeepSpeed data structures it has to be available in the current python environment.
+from deepspeed.utils import logger
+from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS,
+ FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES,
+ FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS)
+
+
+@dataclass
+class zero_model_state:
+ buffers: dict()
+ param_shapes: dict()
+ shared_params: list
+ ds_version: int
+ frozen_param_shapes: dict()
+ frozen_param_fragments: dict()
+
+
+debug = 0
+
+# load to cpu
+device = torch.device('cpu')
+
+
+def atoi(text):
+ return int(text) if text.isdigit() else text
+
+
+def natural_keys(text):
+ '''
+ alist.sort(key=natural_keys) sorts in human order
+ http://nedbatchelder.com/blog/200712/human_sorting.html
+ (See Toothy's implementation in the comments)
+ '''
+ return [atoi(c) for c in re.split(r'(\d+)', text)]
+
+
+def get_model_state_file(checkpoint_dir, zero_stage):
+ if not os.path.isdir(checkpoint_dir):
+ raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist")
+
+ # there should be only one file
+ if zero_stage <= 2:
+ file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt")
+ elif zero_stage == 3:
+ file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt")
+
+ if not os.path.exists(file):
+ raise FileNotFoundError(f"can't find model states file at '{file}'")
+
+ return file
+
+
+def get_checkpoint_files(checkpoint_dir, glob_pattern):
+ # XXX: need to test that this simple glob rule works for multi-node setup too
+ ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys)
+
+ if len(ckpt_files) == 0:
+ raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'")
+
+ return ckpt_files
+
+
+def get_optim_files(checkpoint_dir):
+ return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt")
+
+
+def get_model_state_files(checkpoint_dir):
+ return get_checkpoint_files(checkpoint_dir, "*_model_states.pt")
+
+
+def parse_model_states(files):
+ zero_model_states = []
+ for file in files:
+ state_dict = torch.load(file, map_location=device, weights_only=False)
+
+ if BUFFER_NAMES not in state_dict:
+ raise ValueError(f"{file} is not a model state checkpoint")
+ buffer_names = state_dict[BUFFER_NAMES]
+ if debug:
+ print("Found buffers:", buffer_names)
+
+ # recover just the buffers while restoring them to fp32 if they were saved in fp16
+ buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names}
+ param_shapes = state_dict[PARAM_SHAPES]
+
+ # collect parameters that are included in param_shapes
+ param_names = []
+ for s in param_shapes:
+ for name in s.keys():
+ param_names.append(name)
+
+ # update with frozen parameters
+ frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None)
+ if frozen_param_shapes is not None:
+ if debug:
+ print(f"Found frozen_param_shapes: {frozen_param_shapes}")
+ param_names += list(frozen_param_shapes.keys())
+
+ # handle shared params
+ shared_params = [[k, v] for k, v in state_dict["shared_params"].items()]
+
+ ds_version = state_dict.get(DS_VERSION, None)
+
+ frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None)
+
+ z_model_state = zero_model_state(buffers=buffers,
+ param_shapes=param_shapes,
+ shared_params=shared_params,
+ ds_version=ds_version,
+ frozen_param_shapes=frozen_param_shapes,
+ frozen_param_fragments=frozen_param_fragments)
+ zero_model_states.append(z_model_state)
+
+ return zero_model_states
+
+
+def parse_optim_states(files, ds_checkpoint_dir):
+ total_files = len(files)
+ state_dicts = []
+ for f in tqdm(files, desc='Loading checkpoint shards'):
+ state_dict = torch.load(f, map_location=device, mmap=True, weights_only=False)
+ # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights
+ # and also handle the case where it was already removed by another helper script
+ state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None)
+ state_dicts.append(state_dict)
+
+ if ZERO_STAGE not in state_dicts[0][OPTIMIZER_STATE_DICT]:
+ raise ValueError(f"{files[0]} is not a zero checkpoint")
+ zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE]
+ world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT]
+
+ # For ZeRO-2 each param group can have different partition_count as data parallelism for expert
+ # parameters can be different from data parallelism for non-expert parameters. So we can just
+ # use the max of the partition_count to get the dp world_size.
+
+ if type(world_size) is list:
+ world_size = max(world_size)
+
+ if world_size != total_files:
+ raise ValueError(
+ f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. "
+ "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes."
+ )
+
+ # the groups are named differently in each stage
+ if zero_stage <= 2:
+ fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS
+ elif zero_stage == 3:
+ fp32_groups_key = FP32_FLAT_GROUPS
+ else:
+ raise ValueError(f"unknown zero stage {zero_stage}")
+
+ fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))]
+ return zero_stage, world_size, fp32_flat_groups
+
+
+def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters):
+ """
+ Returns fp32 state_dict reconstructed from ds checkpoint
+
+ Args:
+ - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are)
+
+ """
+ print(f"Processing zero checkpoint '{ds_checkpoint_dir}'")
+
+ optim_files = get_optim_files(ds_checkpoint_dir)
+ zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir)
+ print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}")
+
+ model_files = get_model_state_files(ds_checkpoint_dir)
+
+ zero_model_states = parse_model_states(model_files)
+ print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}')
+
+ if zero_stage <= 2:
+ return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+ exclude_frozen_parameters)
+ elif zero_stage == 3:
+ return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+ exclude_frozen_parameters)
+
+
+def _zero2_merge_frozen_params(state_dict, zero_model_states):
+ if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+ return
+
+ frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+ frozen_param_fragments = zero_model_states[0].frozen_param_fragments
+
+ if debug:
+ num_elem = sum(s.numel() for s in frozen_param_shapes.values())
+ print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+ wanted_params = len(frozen_param_shapes)
+ wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+ avail_numel = sum([p.numel() for p in frozen_param_fragments.values()])
+ print(f'Frozen params: Have {avail_numel} numels to process.')
+ print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+ total_params = 0
+ total_numel = 0
+ for name, shape in frozen_param_shapes.items():
+ total_params += 1
+ unpartitioned_numel = shape.numel()
+ total_numel += unpartitioned_numel
+
+ state_dict[name] = frozen_param_fragments[name]
+
+ if debug:
+ print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+
+ print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _has_callable(obj, fn):
+ attr = getattr(obj, fn, None)
+ return callable(attr)
+
+
+def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+ param_shapes = zero_model_states[0].param_shapes
+
+ # Reconstruction protocol:
+ #
+ # XXX: document this
+
+ if debug:
+ for i in range(world_size):
+ for j in range(len(fp32_flat_groups[0])):
+ print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}")
+
+ # XXX: memory usage doubles here (zero2)
+ num_param_groups = len(fp32_flat_groups[0])
+ merged_single_partition_of_fp32_groups = []
+ for i in range(num_param_groups):
+ merged_partitions = [sd[i] for sd in fp32_flat_groups]
+ full_single_fp32_vector = torch.cat(merged_partitions, 0)
+ merged_single_partition_of_fp32_groups.append(full_single_fp32_vector)
+ avail_numel = sum(
+ [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups])
+
+ if debug:
+ wanted_params = sum([len(shapes) for shapes in param_shapes])
+ wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes])
+ # not asserting if there is a mismatch due to possible padding
+ print(f"Have {avail_numel} numels to process.")
+ print(f"Need {wanted_numel} numels in {wanted_params} params.")
+
+ # params
+ # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+ # out-of-core computing solution
+ total_numel = 0
+ total_params = 0
+ for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups):
+ offset = 0
+ avail_numel = full_single_fp32_vector.numel()
+ for name, shape in shapes.items():
+
+ unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape)
+ total_numel += unpartitioned_numel
+ total_params += 1
+
+ if debug:
+ print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+ state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape)
+ offset += unpartitioned_numel
+
+ # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and
+ # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex
+ # paddings performed in the code it's almost impossible to predict the exact numbers w/o the
+ # live optimizer object, so we are checking that the numbers are within the right range
+ align_to = 2 * world_size
+
+ def zero2_align(x):
+ return align_to * math.ceil(x / align_to)
+
+ if debug:
+ print(f"original offset={offset}, avail_numel={avail_numel}")
+
+ offset = zero2_align(offset)
+ avail_numel = zero2_align(avail_numel)
+
+ if debug:
+ print(f"aligned offset={offset}, avail_numel={avail_numel}")
+
+ # Sanity check
+ if offset != avail_numel:
+ raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+ print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+ exclude_frozen_parameters):
+ state_dict = OrderedDict()
+
+ # buffers
+ buffers = zero_model_states[0].buffers
+ state_dict.update(buffers)
+ if debug:
+ print(f"added {len(buffers)} buffers")
+
+ if not exclude_frozen_parameters:
+ _zero2_merge_frozen_params(state_dict, zero_model_states)
+
+ _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+ # recover shared parameters
+ for pair in zero_model_states[0].shared_params:
+ if pair[1] in state_dict:
+ state_dict[pair[0]] = state_dict[pair[1]]
+
+ return state_dict
+
+
+def zero3_partitioned_param_info(unpartitioned_numel, world_size):
+ remainder = unpartitioned_numel % world_size
+ padding_numel = (world_size - remainder) if remainder else 0
+ partitioned_numel = math.ceil(unpartitioned_numel / world_size)
+ return partitioned_numel, padding_numel
+
+
+def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states):
+ if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+ return
+
+ if debug:
+ for i in range(world_size):
+ num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values())
+ print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+ frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+ wanted_params = len(frozen_param_shapes)
+ wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+ avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size
+ print(f'Frozen params: Have {avail_numel} numels to process.')
+ print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+ total_params = 0
+ total_numel = 0
+ for name, shape in zero_model_states[0].frozen_param_shapes.items():
+ total_params += 1
+ unpartitioned_numel = shape.numel()
+ total_numel += unpartitioned_numel
+
+ param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states)
+ state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape)
+
+ partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+ if debug:
+ print(
+ f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+ )
+
+ print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+class GatheredTensor:
+ """
+ A pseudo tensor that collects partitioned weights.
+ It is more memory efficient when there are multiple groups.
+ """
+
+ def __init__(self, flat_groups, flat_groups_offset, offset, partitioned_numel, shape):
+ self.flat_groups = flat_groups
+ self.flat_groups_offset = flat_groups_offset
+ self.offset = offset
+ self.partitioned_numel = partitioned_numel
+ self.shape = shape
+ self.dtype = self.flat_groups[0][0].dtype
+
+ def contiguous(self):
+ """
+ Merge partitioned weights from flat_groups into a single tensor.
+ """
+ end_idx = self.offset + self.partitioned_numel
+ world_size = len(self.flat_groups)
+ pad_flat_param_chunks = []
+
+ for rank_i in range(world_size):
+ # for each rank, we need to collect weights from related group/groups
+ flat_groups_at_rank_i = self.flat_groups[rank_i]
+ start_group_id = None
+ end_group_id = None
+ for group_id in range(len(self.flat_groups_offset)):
+ if self.flat_groups_offset[group_id] <= self.offset < self.flat_groups_offset[group_id + 1]:
+ start_group_id = group_id
+ if self.flat_groups_offset[group_id] < end_idx <= self.flat_groups_offset[group_id + 1]:
+ end_group_id = group_id
+ break
+ # collect weights from related group/groups
+ for group_id in range(start_group_id, end_group_id + 1):
+ flat_tensor = flat_groups_at_rank_i[group_id]
+ start_offset = self.offset - self.flat_groups_offset[group_id]
+ end_offset = min(end_idx, self.flat_groups_offset[group_id + 1]) - self.flat_groups_offset[group_id]
+ pad_flat_param_chunks.append(flat_tensor[start_offset:end_offset])
+
+ # collect weights from all ranks
+ pad_flat_param = torch.cat(pad_flat_param_chunks, dim=0)
+ param = pad_flat_param[:self.shape.numel()].view(self.shape).contiguous()
+ return param
+
+
+def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+ param_shapes = zero_model_states[0].param_shapes
+ avail_numel = sum([flat_group.numel() for flat_group in fp32_flat_groups[0]]) * world_size
+
+ # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each
+ # param, re-consolidating each param, while dealing with padding if any
+
+ # merge list of dicts, preserving order
+ param_shapes = {k: v for d in param_shapes for k, v in d.items()}
+
+ if debug:
+ for i in range(world_size):
+ print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}")
+
+ wanted_params = len(param_shapes)
+ wanted_numel = sum(shape.numel() for shape in param_shapes.values())
+ # not asserting if there is a mismatch due to possible padding
+ avail_numel = fp32_flat_groups[0].numel() * world_size
+ print(f"Trainable params: Have {avail_numel} numels to process.")
+ print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.")
+
+ # params
+ # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+ # out-of-core computing solution
+ offset = 0
+ total_numel = 0
+ total_params = 0
+ flat_groups_offset = [0] + list(np.cumsum([flat_tensor.numel() for flat_tensor in fp32_flat_groups[0]]))
+ for name, shape in tqdm(param_shapes.items(), desc='Gathering sharded weights'):
+ unpartitioned_numel = shape.numel()
+ total_numel += unpartitioned_numel
+ total_params += 1
+ partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+ if debug:
+ print(
+ f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+ )
+
+ # memory efficient tensor
+ tensor = GatheredTensor(fp32_flat_groups, flat_groups_offset, offset, partitioned_numel, shape)
+ state_dict[name] = tensor
+ offset += partitioned_numel
+
+ offset *= world_size
+
+ # Sanity check
+ if offset != avail_numel:
+ raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+ print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+ exclude_frozen_parameters):
+ state_dict = OrderedDict()
+
+ # buffers
+ buffers = zero_model_states[0].buffers
+ state_dict.update(buffers)
+ if debug:
+ print(f"added {len(buffers)} buffers")
+
+ if not exclude_frozen_parameters:
+ _zero3_merge_frozen_params(state_dict, world_size, zero_model_states)
+
+ _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+ # recover shared parameters
+ for pair in zero_model_states[0].shared_params:
+ if pair[1] in state_dict:
+ state_dict[pair[0]] = state_dict[pair[1]]
+
+ return state_dict
+
+
+def to_torch_tensor(state_dict, return_empty_tensor=False):
+ """
+ Convert state_dict of GatheredTensor to torch tensor
+ """
+ torch_state_dict = {}
+ converted_tensors = {}
+ for name, tensor in state_dict.items():
+ tensor_id = id(tensor)
+ if tensor_id in converted_tensors: # shared tensors
+ shared_tensor = torch_state_dict[converted_tensors[tensor_id]]
+ torch_state_dict[name] = shared_tensor
+ else:
+ converted_tensors[tensor_id] = name
+ if return_empty_tensor:
+ torch_state_dict[name] = torch.empty(tensor.shape, dtype=tensor.dtype)
+ else:
+ torch_state_dict[name] = tensor.contiguous()
+ return torch_state_dict
+
+
+def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir,
+ tag=None,
+ exclude_frozen_parameters=False,
+ lazy_mode=False):
+ """
+ Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with
+ ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example
+ via a model hub.
+
+ Args:
+ - ``checkpoint_dir``: path to the desired checkpoint folder
+ - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14``
+ - ``exclude_frozen_parameters``: exclude frozen parameters
+ - ``lazy_mode``: get state_dict in lazy mode. It returns a dict of pesduo tensor instead of torch tensor, which is more memory efficient.
+ Convert the pesduo tensor to torch tensor by ``.contiguous()``
+
+ Returns:
+ - pytorch ``state_dict``
+
+ A typical usage might be ::
+
+ from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+ # do the training and checkpoint saving
+ state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu
+ model = model.cpu() # move to cpu
+ model.load_state_dict(state_dict)
+ # submit to model hub or save the model to share with others
+
+ In this example the ``model`` will no longer be usable in the deepspeed context of the same
+ application. i.e. you will need to re-initialize the deepspeed engine, since
+ ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+ If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead.
+
+ Note: the above usage may not work if your application doesn't have sufficient free CPU memory.
+ You may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
+ the checkpoint. Or you can load state_dict in lazy mode ::
+
+ from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+ state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, lazy_mode=True) # not on cpu
+ for name, lazy_tensor in state_dict.item():
+ tensor = lazy_tensor.contiguous() # to cpu
+ print(name, tensor)
+ # del tensor to release memory if it no longer in use
+ """
+ if tag is None:
+ latest_path = os.path.join(checkpoint_dir, 'latest')
+ if os.path.isfile(latest_path):
+ with open(latest_path, 'r') as fd:
+ tag = fd.read().strip()
+ else:
+ raise ValueError(f"Unable to find 'latest' file at {latest_path}")
+
+ ds_checkpoint_dir = os.path.join(checkpoint_dir, tag)
+
+ if not os.path.isdir(ds_checkpoint_dir):
+ raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist")
+
+ state_dict = _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters)
+ if lazy_mode:
+ return state_dict
+ else:
+ return to_torch_tensor(state_dict)
+
+
+def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir,
+ output_dir,
+ max_shard_size="5GB",
+ safe_serialization=False,
+ tag=None,
+ exclude_frozen_parameters=False):
+ """
+ Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be
+ loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed.
+
+ Args:
+ - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+ - ``output_dir``: directory to the pytorch fp32 state_dict output files
+ - ``max_shard_size``: the maximum size for a checkpoint before being sharded, default value is 5GB
+ - ``safe_serialization``: whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).
+ - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+ - ``exclude_frozen_parameters``: exclude frozen parameters
+ """
+
+ # Dependency pre-check
+ if safe_serialization:
+ try:
+ from safetensors.torch import save_file
+ except ImportError:
+ print('If you want to use `safe_serialization`, please `pip install safetensors`')
+ raise
+ if max_shard_size is not None:
+ try:
+ from huggingface_hub import split_torch_state_dict_into_shards
+ except ImportError:
+ print('If you want to use `max_shard_size`, please `pip install huggingface_hub`')
+ raise
+
+ # Convert zero checkpoint to state_dict
+ state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir,
+ tag,
+ exclude_frozen_parameters,
+ lazy_mode=True)
+
+ # Shard the model if it is too big.
+ weights_name = "model.safetensors" if safe_serialization else "pytorch_model.bin"
+ if max_shard_size is not None:
+ filename_pattern = weights_name.replace(".bin", "{suffix}.bin").replace(".safetensors", "{suffix}.safetensors")
+ # an memory-efficient approach for sharding
+ empty_state_dict = to_torch_tensor(state_dict, return_empty_tensor=True)
+ state_dict_split = split_torch_state_dict_into_shards(empty_state_dict,
+ filename_pattern=filename_pattern,
+ max_shard_size=max_shard_size)
+ else:
+ from collections import namedtuple
+ StateDictSplit = namedtuple("StateDictSplit", ["is_sharded", "filename_to_tensors"])
+ state_dict_split = StateDictSplit(is_sharded=False,
+ filename_to_tensors={weights_name: list(state_dict.keys())})
+
+ # Save the model by shard
+ os.makedirs(output_dir, exist_ok=True)
+ filename_to_tensors = state_dict_split.filename_to_tensors.items()
+ for shard_file, tensors in tqdm(filename_to_tensors, desc="Saving checkpoint shards"):
+ shard_state_dict = {tensor_name: state_dict[tensor_name] for tensor_name in tensors}
+ shard_state_dict = to_torch_tensor(shard_state_dict)
+ output_path = os.path.join(output_dir, shard_file)
+ if safe_serialization:
+ save_file(shard_state_dict, output_path, metadata={"format": "pt"})
+ else:
+ torch.save(shard_state_dict, output_path)
+ # release the memory of current shard
+ for tensor_name in list(shard_state_dict.keys()):
+ del state_dict[tensor_name]
+ del shard_state_dict[tensor_name]
+ del shard_state_dict
+ gc.collect()
+
+ # Save index if sharded
+ if state_dict_split.is_sharded:
+ index = {
+ "metadata": state_dict_split.metadata,
+ "weight_map": state_dict_split.tensor_to_filename,
+ }
+ save_index_file = "model.safetensors.index.json" if safe_serialization else "pytorch_model.bin.index.json"
+ save_index_file = os.path.join(output_dir, save_index_file)
+ with open(save_index_file, "w", encoding="utf-8") as f:
+ content = json.dumps(index, indent=2, sort_keys=True) + "\n"
+ f.write(content)
+
+
+def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None):
+ """
+ 1. Put the provided model to cpu
+ 2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict``
+ 3. Load it into the provided model
+
+ Args:
+ - ``model``: the model object to update
+ - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+ - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+
+ Returns:
+ - ``model`: modified model
+
+ Make sure you have plenty of CPU memory available before you call this function. If you don't
+ have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it
+ conveniently placed for you in the checkpoint folder.
+
+ A typical usage might be ::
+
+ from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint
+ model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir)
+ # submit to model hub or save the model to share with others
+
+ Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context
+ of the same application. i.e. you will need to re-initialize the deepspeed engine, since
+ ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+ """
+ logger.info("Extracting fp32 weights")
+ state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
+
+ logger.info("Overwriting model with fp32 weights")
+ model = model.cpu()
+ model.load_state_dict(state_dict, strict=False)
+
+ return model
+
+
+if __name__ == "__main__":
+ parser = argparse.ArgumentParser()
+ parser.add_argument("checkpoint_dir",
+ type=str,
+ help="path to the desired checkpoint folder, e.g., path/checkpoint-12")
+ parser.add_argument("output_dir",
+ type=str,
+ help="directory to the pytorch fp32 state_dict output files"
+ "(e.g. path/checkpoint-12-output/)")
+ parser.add_argument(
+ "--max_shard_size",
+ type=str,
+ default="5GB",
+ help="The maximum size for a checkpoint before being sharded. Checkpoints shard will then be each of size"
+ "lower than this size. If expressed as a string, needs to be digits followed by a unit (like `5MB`"
+ "We default it to 5GB in order for models to be able to run easily on free-tier google colab instances"
+ "without CPU OOM issues.")
+ parser.add_argument(
+ "--safe_serialization",
+ default=False,
+ action='store_true',
+ help="Whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).")
+ parser.add_argument("-t",
+ "--tag",
+ type=str,
+ default=None,
+ help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1")
+ parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters")
+ parser.add_argument("-d", "--debug", action='store_true', help="enable debug")
+ args = parser.parse_args()
+
+ debug = args.debug
+
+ convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir,
+ args.output_dir,
+ max_shard_size=args.max_shard_size,
+ safe_serialization=args.safe_serialization,
+ tag=args.tag,
+ exclude_frozen_parameters=args.exclude_frozen_parameters)
diff --git a/checkpoint-29704/chat_template.jinja b/checkpoint-29704/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..01be9b307daa2d425f7c168c9fb145a286e0afb4
--- /dev/null
+++ b/checkpoint-29704/chat_template.jinja
@@ -0,0 +1,89 @@
+{%- if tools %}
+ {{- '<|im_start|>system\n' }}
+ {%- if messages[0].role == 'system' %}
+ {{- messages[0].content + '\n\n' }}
+ {%- endif %}
+ {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within XML tags:\n" }}
+ {%- for tool in tools %}
+ {{- "\n" }}
+ {{- tool | tojson }}
+ {%- endfor %}
+ {{- "\n\n\nFor each function call, return a json object with function name and arguments within XML tags:\n\n{\"name\": , \"arguments\": }\n<|im_end|>\n" }}
+{%- else %}
+ {%- if messages[0].role == 'system' %}
+ {{- '<|im_start|>system\n' + messages[0].content + '<|im_end|>\n' }}
+ {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+ {%- set index = (messages|length - 1) - loop.index0 %}
+ {%- if ns.multi_step_tool and message.role == "user" and message.content is string and not(message.content.startswith('') and message.content.endswith('')) %}
+ {%- set ns.multi_step_tool = false %}
+ {%- set ns.last_query_index = index %}
+ {%- endif %}
+{%- endfor %}
+{%- for message in messages %}
+ {%- if message.content is string %}
+ {%- set content = message.content %}
+ {%- else %}
+ {%- set content = '' %}
+ {%- endif %}
+ {%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
+ {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+ {%- elif message.role == "assistant" %}
+ {%- set reasoning_content = '' %}
+ {%- if message.reasoning_content is string %}
+ {%- set reasoning_content = message.reasoning_content %}
+ {%- else %}
+ {%- if '' in content %}
+ {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %}
+ {%- set content = content.split('')[-1].lstrip('\n') %}
+ {%- endif %}
+ {%- endif %}
+ {%- if loop.index0 > ns.last_query_index %}
+ {%- if loop.last or (not loop.last and reasoning_content) %}
+ {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content.strip('\n') + '\n\n\n' + content.lstrip('\n') }}
+ {%- else %}
+ {{- '<|im_start|>' + message.role + '\n' + content }}
+ {%- endif %}
+ {%- else %}
+ {{- '<|im_start|>' + message.role + '\n' + content }}
+ {%- endif %}
+ {%- if message.tool_calls %}
+ {%- for tool_call in message.tool_calls %}
+ {%- if (loop.first and content) or (not loop.first) %}
+ {{- '\n' }}
+ {%- endif %}
+ {%- if tool_call.function %}
+ {%- set tool_call = tool_call.function %}
+ {%- endif %}
+ {{- '\n{"name": "' }}
+ {{- tool_call.name }}
+ {{- '", "arguments": ' }}
+ {%- if tool_call.arguments is string %}
+ {{- tool_call.arguments }}
+ {%- else %}
+ {{- tool_call.arguments | tojson }}
+ {%- endif %}
+ {{- '}\n' }}
+ {%- endfor %}
+ {%- endif %}
+ {{- '<|im_end|>\n' }}
+ {%- elif message.role == "tool" %}
+ {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
+ {{- '<|im_start|>user' }}
+ {%- endif %}
+ {{- '\n\n' }}
+ {{- content }}
+ {{- '\n' }}
+ {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
+ {{- '<|im_end|>\n' }}
+ {%- endif %}
+ {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+ {{- '<|im_start|>assistant\n' }}
+ {%- if enable_thinking is defined and enable_thinking is false %}
+ {{- '\n\n\n\n' }}
+ {%- endif %}
+{%- endif %}
\ No newline at end of file
diff --git a/checkpoint-29704/config.json b/checkpoint-29704/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..746f56efdea311c4c5e737f2e31814c30cc1e1bf
--- /dev/null
+++ b/checkpoint-29704/config.json
@@ -0,0 +1,71 @@
+{
+ "architectures": [
+ "Qwen3ForCausalLM"
+ ],
+ "attention_bias": false,
+ "attention_dropout": 0.0,
+ "bos_token_id": null,
+ "dtype": "bfloat16",
+ "eos_token_id": 151645,
+ "head_dim": 128,
+ "hidden_act": "silu",
+ "hidden_size": 4096,
+ "initializer_range": 0.02,
+ "intermediate_size": 12288,
+ "layer_types": [
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention"
+ ],
+ "max_position_embeddings": 40960,
+ "max_window_layers": 36,
+ "model_type": "qwen3",
+ "num_attention_heads": 32,
+ "num_hidden_layers": 36,
+ "num_key_value_heads": 8,
+ "pad_token_id": 151643,
+ "rms_norm_eps": 1e-06,
+ "rope_parameters": {
+ "rope_theta": 1000000,
+ "rope_type": "default"
+ },
+ "sliding_window": null,
+ "tie_word_embeddings": false,
+ "transformers_version": "5.6.0",
+ "use_cache": false,
+ "use_sliding_window": false,
+ "vocab_size": 151936
+}
diff --git a/checkpoint-29704/generation_config.json b/checkpoint-29704/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..cf5bf49b6941f1673361724230bb69080cd3f1e2
--- /dev/null
+++ b/checkpoint-29704/generation_config.json
@@ -0,0 +1,12 @@
+{
+ "do_sample": true,
+ "eos_token_id": [
+ 151645,
+ 151643
+ ],
+ "pad_token_id": 151643,
+ "temperature": 0.6,
+ "top_k": 20,
+ "top_p": 0.95,
+ "transformers_version": "5.6.0"
+}
diff --git a/checkpoint-29704/global_step29704/zero_pp_rank_0_mp_rank_00_model_states.pt b/checkpoint-29704/global_step29704/zero_pp_rank_0_mp_rank_00_model_states.pt
new file mode 100644
index 0000000000000000000000000000000000000000..84c9e8703e2453eefc814b42b923fda62d85ee98
--- /dev/null
+++ b/checkpoint-29704/global_step29704/zero_pp_rank_0_mp_rank_00_model_states.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b5ce3cbe67867e10ba5940798e76e013bd60b038a8cb1725bf0237320f83e425
+size 202796
diff --git a/checkpoint-29704/global_step29704/zero_pp_rank_1_mp_rank_00_model_states.pt b/checkpoint-29704/global_step29704/zero_pp_rank_1_mp_rank_00_model_states.pt
new file mode 100644
index 0000000000000000000000000000000000000000..5f0bf234e93bec1c9906e244a849fcbf96f54633
--- /dev/null
+++ b/checkpoint-29704/global_step29704/zero_pp_rank_1_mp_rank_00_model_states.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0e9c5cba9532ef23495aa00e210bd9628d87eb94d9a4cd20bfcd12d58662384a
+size 202796
diff --git a/checkpoint-29704/global_step29704/zero_pp_rank_2_mp_rank_00_model_states.pt b/checkpoint-29704/global_step29704/zero_pp_rank_2_mp_rank_00_model_states.pt
new file mode 100644
index 0000000000000000000000000000000000000000..30ea59c9a3e5a7cd37f64d5df770a504dbeb42a2
--- /dev/null
+++ b/checkpoint-29704/global_step29704/zero_pp_rank_2_mp_rank_00_model_states.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aa51c9af0a2d49ec63210ec1ee429643445e70267fc72628dc36fedb6d1540da
+size 202796
diff --git a/checkpoint-29704/global_step29704/zero_pp_rank_3_mp_rank_00_model_states.pt b/checkpoint-29704/global_step29704/zero_pp_rank_3_mp_rank_00_model_states.pt
new file mode 100644
index 0000000000000000000000000000000000000000..8e323144886d9fa644ea043a91d8518603f0b529
--- /dev/null
+++ b/checkpoint-29704/global_step29704/zero_pp_rank_3_mp_rank_00_model_states.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8c19fa1f9859e7dacce4463d408833041bd8ed9f6925c27007cdde4b39e98e92
+size 202796
diff --git a/checkpoint-29704/latest b/checkpoint-29704/latest
new file mode 100644
index 0000000000000000000000000000000000000000..adb0f0ec85d351bceb097b295b1d0c4bb9f1b450
--- /dev/null
+++ b/checkpoint-29704/latest
@@ -0,0 +1 @@
+global_step29704
\ No newline at end of file
diff --git a/checkpoint-29704/model.safetensors b/checkpoint-29704/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..e2403050ca19b0edeb2a4af3a391812ca8dad69c
--- /dev/null
+++ b/checkpoint-29704/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2d372672c4b3c93543df0eb6f735f1fb9c60753d3ca0b93a30cd40b192dfadcb
+size 16381517208
diff --git a/checkpoint-29704/rng_state_0.pth b/checkpoint-29704/rng_state_0.pth
new file mode 100644
index 0000000000000000000000000000000000000000..b2e0e7337f62c934e90c141589ca185ff78fae36
--- /dev/null
+++ b/checkpoint-29704/rng_state_0.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2fe6b873f1da64bb980afe982fc1c6b6bc461ad8c33073d856804323f64b26d9
+size 15429
diff --git a/checkpoint-29704/rng_state_1.pth b/checkpoint-29704/rng_state_1.pth
new file mode 100644
index 0000000000000000000000000000000000000000..204fc909b58b8ae0fab9c8b0f41ea5cc8f3277f8
--- /dev/null
+++ b/checkpoint-29704/rng_state_1.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:13db4dd10b22b5464f5ef0b539152b46390eae6a827b355108cb54b1af2eb830
+size 15429
diff --git a/checkpoint-29704/rng_state_2.pth b/checkpoint-29704/rng_state_2.pth
new file mode 100644
index 0000000000000000000000000000000000000000..0c3095386194e2f629b38c7957abe0b9ab74aa8b
--- /dev/null
+++ b/checkpoint-29704/rng_state_2.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06e293f13fb6e873e8081636b6b07115410b397c55f1ed08b8772b059c8eb74e
+size 15429
diff --git a/checkpoint-29704/rng_state_3.pth b/checkpoint-29704/rng_state_3.pth
new file mode 100644
index 0000000000000000000000000000000000000000..fdf7f18d5d86f0d363c9ca44b9ad169d469e47f7
--- /dev/null
+++ b/checkpoint-29704/rng_state_3.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c595c077664f94b7f91088561951be20e39483e01865774a83bc46f3971a3152
+size 15429
diff --git a/checkpoint-29704/scheduler.pt b/checkpoint-29704/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..e0c1d0812609d3c8bea84bf07ea7f0c6c42daf2c
--- /dev/null
+++ b/checkpoint-29704/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9cab43bbb04071f651f3c970690935ffc1dbe98c766c61ea1e7d8be3eefc3944
+size 1465
diff --git a/checkpoint-29704/tokenizer.json b/checkpoint-29704/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..c7afbed2efcdf019f88ab0572ec29d3bf595dfe2
--- /dev/null
+++ b/checkpoint-29704/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be75606093db2094d7cd20f3c2f385c212750648bd6ea4fb2bf507a6a4c55506
+size 11422650
diff --git a/checkpoint-29704/tokenizer_config.json b/checkpoint-29704/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..6623da03457d79d2cd4af9557e9f31407cdb835c
--- /dev/null
+++ b/checkpoint-29704/tokenizer_config.json
@@ -0,0 +1,31 @@
+{
+ "add_prefix_space": false,
+ "backend": "tokenizers",
+ "bos_token": null,
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "<|im_end|>",
+ "errors": "replace",
+ "extra_special_tokens": [
+ "<|im_start|>",
+ "<|im_end|>",
+ "<|object_ref_start|>",
+ "<|object_ref_end|>",
+ "<|box_start|>",
+ "<|box_end|>",
+ "<|quad_start|>",
+ "<|quad_end|>",
+ "<|vision_start|>",
+ "<|vision_end|>",
+ "<|vision_pad|>",
+ "<|image_pad|>",
+ "<|video_pad|>"
+ ],
+ "is_local": false,
+ "local_files_only": false,
+ "model_max_length": 131072,
+ "pad_token": "<|endoftext|>",
+ "padding_side": "right",
+ "split_special_tokens": false,
+ "tokenizer_class": "Qwen2Tokenizer",
+ "unk_token": null
+}
diff --git a/checkpoint-29704/trainer_state.json b/checkpoint-29704/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..5df1854b2130eb9bfe4b557790778c191ef4de8d
--- /dev/null
+++ b/checkpoint-29704/trainer_state.json
@@ -0,0 +1,59434 @@
+{
+ "best_global_step": null,
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 2.0,
+ "eval_steps": 500,
+ "global_step": 29704,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.00033665499596014,
+ "grad_norm": 1.7158506568336664,
+ "learning_rate": 9.999999552565099e-06,
+ "loss": 0.8321187973022461,
+ "num_input_tokens_seen": 213976,
+ "step": 5,
+ "train_runtime": 48.5386,
+ "train_tokens_per_second": 4408.369
+ },
+ {
+ "epoch": 0.00067330999192028,
+ "grad_norm": 1.4989508496023307,
+ "learning_rate": 9.999997734860948e-06,
+ "loss": 0.6714894294738769,
+ "num_input_tokens_seen": 418192,
+ "step": 10,
+ "train_runtime": 79.1759,
+ "train_tokens_per_second": 5281.811
+ },
+ {
+ "epoch": 0.0010099649878804202,
+ "grad_norm": 1.3987943897146944,
+ "learning_rate": 9.999994518923374e-06,
+ "loss": 0.6414656162261962,
+ "num_input_tokens_seen": 631760,
+ "step": 15,
+ "train_runtime": 116.8887,
+ "train_tokens_per_second": 5404.799
+ },
+ {
+ "epoch": 0.00134661998384056,
+ "grad_norm": 1.3594234171992137,
+ "learning_rate": 9.999989904753278e-06,
+ "loss": 0.6119038105010987,
+ "num_input_tokens_seen": 847240,
+ "step": 20,
+ "train_runtime": 160.5101,
+ "train_tokens_per_second": 5278.422
+ },
+ {
+ "epoch": 0.0016832749798007002,
+ "grad_norm": 1.1837151081821848,
+ "learning_rate": 9.999983892351947e-06,
+ "loss": 0.5918544292449951,
+ "num_input_tokens_seen": 1067560,
+ "step": 25,
+ "train_runtime": 197.4905,
+ "train_tokens_per_second": 5405.628
+ },
+ {
+ "epoch": 0.0020199299757608404,
+ "grad_norm": 1.1504621450496095,
+ "learning_rate": 9.999976481721066e-06,
+ "loss": 0.5148830890655518,
+ "num_input_tokens_seen": 1288760,
+ "step": 30,
+ "train_runtime": 231.3357,
+ "train_tokens_per_second": 5570.951
+ },
+ {
+ "epoch": 0.0023565849717209803,
+ "grad_norm": 1.0663060499576293,
+ "learning_rate": 9.999967672862704e-06,
+ "loss": 0.554470157623291,
+ "num_input_tokens_seen": 1501896,
+ "step": 35,
+ "train_runtime": 271.9258,
+ "train_tokens_per_second": 5523.183
+ },
+ {
+ "epoch": 0.00269323996768112,
+ "grad_norm": 1.1179858029742544,
+ "learning_rate": 9.999957465779326e-06,
+ "loss": 0.5239121437072753,
+ "num_input_tokens_seen": 1716792,
+ "step": 40,
+ "train_runtime": 319.9491,
+ "train_tokens_per_second": 5365.829
+ },
+ {
+ "epoch": 0.0030298949636412605,
+ "grad_norm": 1.1555971967505698,
+ "learning_rate": 9.999945860473788e-06,
+ "loss": 0.5400435924530029,
+ "num_input_tokens_seen": 1920760,
+ "step": 45,
+ "train_runtime": 364.1591,
+ "train_tokens_per_second": 5274.507
+ },
+ {
+ "epoch": 0.0033665499596014004,
+ "grad_norm": 0.9321994843580264,
+ "learning_rate": 9.999932856949333e-06,
+ "loss": 0.5238595008850098,
+ "num_input_tokens_seen": 2140344,
+ "step": 50,
+ "train_runtime": 402.577,
+ "train_tokens_per_second": 5316.608
+ },
+ {
+ "epoch": 0.0037032049555615404,
+ "grad_norm": 1.1572766450172172,
+ "learning_rate": 9.999918455209599e-06,
+ "loss": 0.5059286117553711,
+ "num_input_tokens_seen": 2368296,
+ "step": 55,
+ "train_runtime": 438.0431,
+ "train_tokens_per_second": 5406.536
+ },
+ {
+ "epoch": 0.004039859951521681,
+ "grad_norm": 1.1277411399750283,
+ "learning_rate": 9.99990265525861e-06,
+ "loss": 0.5235630989074707,
+ "num_input_tokens_seen": 2589544,
+ "step": 60,
+ "train_runtime": 475.1915,
+ "train_tokens_per_second": 5449.474
+ },
+ {
+ "epoch": 0.004376514947481821,
+ "grad_norm": 1.233348290191874,
+ "learning_rate": 9.999885457100789e-06,
+ "loss": 0.5595571517944335,
+ "num_input_tokens_seen": 2809432,
+ "step": 65,
+ "train_runtime": 503.8866,
+ "train_tokens_per_second": 5575.524
+ },
+ {
+ "epoch": 0.0047131699434419605,
+ "grad_norm": 1.2298135105742523,
+ "learning_rate": 9.999866860740943e-06,
+ "loss": 0.5092310428619384,
+ "num_input_tokens_seen": 3020248,
+ "step": 70,
+ "train_runtime": 543.5456,
+ "train_tokens_per_second": 5556.568
+ },
+ {
+ "epoch": 0.005049824939402101,
+ "grad_norm": 0.9679470345624956,
+ "learning_rate": 9.999846866184274e-06,
+ "loss": 0.5114652633666992,
+ "num_input_tokens_seen": 3249600,
+ "step": 75,
+ "train_runtime": 585.283,
+ "train_tokens_per_second": 5552.186
+ },
+ {
+ "epoch": 0.00538647993536224,
+ "grad_norm": 0.979469640417179,
+ "learning_rate": 9.99982547343637e-06,
+ "loss": 0.5040130615234375,
+ "num_input_tokens_seen": 3443864,
+ "step": 80,
+ "train_runtime": 626.1684,
+ "train_tokens_per_second": 5499.901
+ },
+ {
+ "epoch": 0.005723134931322381,
+ "grad_norm": 1.078786163056917,
+ "learning_rate": 9.999802682503217e-06,
+ "loss": 0.4742586612701416,
+ "num_input_tokens_seen": 3661032,
+ "step": 85,
+ "train_runtime": 663.7311,
+ "train_tokens_per_second": 5515.836
+ },
+ {
+ "epoch": 0.006059789927282521,
+ "grad_norm": 0.9113383787124433,
+ "learning_rate": 9.999778493391186e-06,
+ "loss": 0.5044189453125,
+ "num_input_tokens_seen": 3888680,
+ "step": 90,
+ "train_runtime": 707.589,
+ "train_tokens_per_second": 5495.676
+ },
+ {
+ "epoch": 0.0063964449232426605,
+ "grad_norm": 1.0192977781689194,
+ "learning_rate": 9.999752906107043e-06,
+ "loss": 0.5348274230957031,
+ "num_input_tokens_seen": 4114248,
+ "step": 95,
+ "train_runtime": 744.4001,
+ "train_tokens_per_second": 5526.931
+ },
+ {
+ "epoch": 0.006733099919202801,
+ "grad_norm": 1.046539367979976,
+ "learning_rate": 9.999725920657944e-06,
+ "loss": 0.5229468822479248,
+ "num_input_tokens_seen": 4329256,
+ "step": 100,
+ "train_runtime": 790.5278,
+ "train_tokens_per_second": 5476.412
+ },
+ {
+ "epoch": 0.007069754915162941,
+ "grad_norm": 1.0823101637799084,
+ "learning_rate": 9.999697537051432e-06,
+ "loss": 0.4990852355957031,
+ "num_input_tokens_seen": 4533832,
+ "step": 105,
+ "train_runtime": 828.8033,
+ "train_tokens_per_second": 5470.335
+ },
+ {
+ "epoch": 0.007406409911123081,
+ "grad_norm": 1.0579674272002448,
+ "learning_rate": 9.999667755295449e-06,
+ "loss": 0.48107209205627444,
+ "num_input_tokens_seen": 4749088,
+ "step": 110,
+ "train_runtime": 864.874,
+ "train_tokens_per_second": 5491.075
+ },
+ {
+ "epoch": 0.007743064907083221,
+ "grad_norm": 1.2773224732968154,
+ "learning_rate": 9.999636575398319e-06,
+ "loss": 0.4919854164123535,
+ "num_input_tokens_seen": 4960752,
+ "step": 115,
+ "train_runtime": 904.8595,
+ "train_tokens_per_second": 5482.345
+ },
+ {
+ "epoch": 0.008079719903043361,
+ "grad_norm": 1.0488778079229073,
+ "learning_rate": 9.999603997368766e-06,
+ "loss": 0.4375304222106934,
+ "num_input_tokens_seen": 5167864,
+ "step": 120,
+ "train_runtime": 939.3678,
+ "train_tokens_per_second": 5501.428
+ },
+ {
+ "epoch": 0.008416374899003501,
+ "grad_norm": 1.1229102076855217,
+ "learning_rate": 9.999570021215895e-06,
+ "loss": 0.4794501304626465,
+ "num_input_tokens_seen": 5384648,
+ "step": 125,
+ "train_runtime": 973.7305,
+ "train_tokens_per_second": 5529.916
+ },
+ {
+ "epoch": 0.008753029894963642,
+ "grad_norm": 1.0387093548265027,
+ "learning_rate": 9.99953464694921e-06,
+ "loss": 0.5004319667816162,
+ "num_input_tokens_seen": 5590392,
+ "step": 130,
+ "train_runtime": 1013.6767,
+ "train_tokens_per_second": 5514.965
+ },
+ {
+ "epoch": 0.009089684890923782,
+ "grad_norm": 1.0508132347808834,
+ "learning_rate": 9.999497874578606e-06,
+ "loss": 0.495189905166626,
+ "num_input_tokens_seen": 5819224,
+ "step": 135,
+ "train_runtime": 1052.6891,
+ "train_tokens_per_second": 5527.961
+ },
+ {
+ "epoch": 0.009426339886883921,
+ "grad_norm": 1.0896002910221911,
+ "learning_rate": 9.999459704114361e-06,
+ "loss": 0.5080117225646973,
+ "num_input_tokens_seen": 6037808,
+ "step": 140,
+ "train_runtime": 1094.5671,
+ "train_tokens_per_second": 5516.161
+ },
+ {
+ "epoch": 0.00976299488284406,
+ "grad_norm": 1.034914657869065,
+ "learning_rate": 9.999420135567153e-06,
+ "loss": 0.526894474029541,
+ "num_input_tokens_seen": 6252752,
+ "step": 145,
+ "train_runtime": 1135.4311,
+ "train_tokens_per_second": 5506.941
+ },
+ {
+ "epoch": 0.010099649878804202,
+ "grad_norm": 1.164626155534281,
+ "learning_rate": 9.999379168948047e-06,
+ "loss": 0.5105450630187989,
+ "num_input_tokens_seen": 6476720,
+ "step": 150,
+ "train_runtime": 1172.7989,
+ "train_tokens_per_second": 5522.447
+ },
+ {
+ "epoch": 0.010436304874764341,
+ "grad_norm": 1.2400950788317526,
+ "learning_rate": 9.999336804268496e-06,
+ "loss": 0.47139749526977537,
+ "num_input_tokens_seen": 6698024,
+ "step": 155,
+ "train_runtime": 1217.419,
+ "train_tokens_per_second": 5501.823
+ },
+ {
+ "epoch": 0.01077295987072448,
+ "grad_norm": 1.0587023212526414,
+ "learning_rate": 9.999293041540351e-06,
+ "loss": 0.5010284900665283,
+ "num_input_tokens_seen": 6910784,
+ "step": 160,
+ "train_runtime": 1258.4852,
+ "train_tokens_per_second": 5491.351
+ },
+ {
+ "epoch": 0.011109614866684622,
+ "grad_norm": 1.0126894215000632,
+ "learning_rate": 9.999247880775849e-06,
+ "loss": 0.49982900619506837,
+ "num_input_tokens_seen": 7115568,
+ "step": 165,
+ "train_runtime": 1295.6174,
+ "train_tokens_per_second": 5492.029
+ },
+ {
+ "epoch": 0.011446269862644761,
+ "grad_norm": 1.198929638798561,
+ "learning_rate": 9.999201321987617e-06,
+ "loss": 0.46059427261352537,
+ "num_input_tokens_seen": 7329488,
+ "step": 170,
+ "train_runtime": 1337.1094,
+ "train_tokens_per_second": 5481.592
+ },
+ {
+ "epoch": 0.011782924858604901,
+ "grad_norm": 1.0114658976479987,
+ "learning_rate": 9.999153365188676e-06,
+ "loss": 0.5039016723632812,
+ "num_input_tokens_seen": 7547880,
+ "step": 175,
+ "train_runtime": 1383.6734,
+ "train_tokens_per_second": 5454.958
+ },
+ {
+ "epoch": 0.012119579854565042,
+ "grad_norm": 1.0579154349493107,
+ "learning_rate": 9.99910401039244e-06,
+ "loss": 0.4546662330627441,
+ "num_input_tokens_seen": 7757728,
+ "step": 180,
+ "train_runtime": 1424.6608,
+ "train_tokens_per_second": 5445.316
+ },
+ {
+ "epoch": 0.012456234850525182,
+ "grad_norm": 1.1327328565153654,
+ "learning_rate": 9.999053257612708e-06,
+ "loss": 0.464589786529541,
+ "num_input_tokens_seen": 7984416,
+ "step": 185,
+ "train_runtime": 1468.0361,
+ "train_tokens_per_second": 5438.842
+ },
+ {
+ "epoch": 0.012792889846485321,
+ "grad_norm": 1.0695885705791155,
+ "learning_rate": 9.999001106863672e-06,
+ "loss": 0.5309792518615722,
+ "num_input_tokens_seen": 8204976,
+ "step": 190,
+ "train_runtime": 1506.2271,
+ "train_tokens_per_second": 5447.37
+ },
+ {
+ "epoch": 0.013129544842445462,
+ "grad_norm": 1.129268767285916,
+ "learning_rate": 9.998947558159916e-06,
+ "loss": 0.4964266300201416,
+ "num_input_tokens_seen": 8428800,
+ "step": 195,
+ "train_runtime": 1546.2799,
+ "train_tokens_per_second": 5451.018
+ },
+ {
+ "epoch": 0.013466199838405602,
+ "grad_norm": 1.0645176443218387,
+ "learning_rate": 9.99889261151642e-06,
+ "loss": 0.44077634811401367,
+ "num_input_tokens_seen": 8642192,
+ "step": 200,
+ "train_runtime": 1586.0871,
+ "train_tokens_per_second": 5448.75
+ },
+ {
+ "epoch": 0.013802854834365741,
+ "grad_norm": 1.0273158071004191,
+ "learning_rate": 9.998836266948542e-06,
+ "loss": 0.47046546936035155,
+ "num_input_tokens_seen": 8852840,
+ "step": 205,
+ "train_runtime": 1618.9934,
+ "train_tokens_per_second": 5468.114
+ },
+ {
+ "epoch": 0.014139509830325882,
+ "grad_norm": 1.2260591863245542,
+ "learning_rate": 9.998778524472043e-06,
+ "loss": 0.4429631233215332,
+ "num_input_tokens_seen": 9059368,
+ "step": 210,
+ "train_runtime": 1658.2941,
+ "train_tokens_per_second": 5463.065
+ },
+ {
+ "epoch": 0.014476164826286022,
+ "grad_norm": 1.1035215533216065,
+ "learning_rate": 9.99871938410307e-06,
+ "loss": 0.46187715530395507,
+ "num_input_tokens_seen": 9273936,
+ "step": 215,
+ "train_runtime": 1690.8398,
+ "train_tokens_per_second": 5484.811
+ },
+ {
+ "epoch": 0.014812819822246161,
+ "grad_norm": 1.500582938758766,
+ "learning_rate": 9.99865884585816e-06,
+ "loss": 0.5123334407806397,
+ "num_input_tokens_seen": 9489816,
+ "step": 220,
+ "train_runtime": 1724.7509,
+ "train_tokens_per_second": 5502.137
+ },
+ {
+ "epoch": 0.015149474818206303,
+ "grad_norm": 1.1104716846300207,
+ "learning_rate": 9.998596909754245e-06,
+ "loss": 0.4839674949645996,
+ "num_input_tokens_seen": 9710720,
+ "step": 225,
+ "train_runtime": 1769.0601,
+ "train_tokens_per_second": 5489.197
+ },
+ {
+ "epoch": 0.015486129814166442,
+ "grad_norm": 1.1763018457719354,
+ "learning_rate": 9.998533575808642e-06,
+ "loss": 0.455305290222168,
+ "num_input_tokens_seen": 9922160,
+ "step": 230,
+ "train_runtime": 1804.4173,
+ "train_tokens_per_second": 5498.817
+ },
+ {
+ "epoch": 0.015822784810126583,
+ "grad_norm": 0.9508268973242634,
+ "learning_rate": 9.998468844039064e-06,
+ "loss": 0.4588268280029297,
+ "num_input_tokens_seen": 10132480,
+ "step": 235,
+ "train_runtime": 1843.1618,
+ "train_tokens_per_second": 5497.336
+ },
+ {
+ "epoch": 0.016159439806086723,
+ "grad_norm": 0.982580320370019,
+ "learning_rate": 9.998402714463613e-06,
+ "loss": 0.49381322860717775,
+ "num_input_tokens_seen": 10356032,
+ "step": 240,
+ "train_runtime": 1886.9781,
+ "train_tokens_per_second": 5488.157
+ },
+ {
+ "epoch": 0.016496094802046862,
+ "grad_norm": 0.9967583322638394,
+ "learning_rate": 9.998335187100782e-06,
+ "loss": 0.4929532051086426,
+ "num_input_tokens_seen": 10571688,
+ "step": 245,
+ "train_runtime": 1931.0773,
+ "train_tokens_per_second": 5474.503
+ },
+ {
+ "epoch": 0.016832749798007002,
+ "grad_norm": 1.0005940352447382,
+ "learning_rate": 9.998266261969453e-06,
+ "loss": 0.488722038269043,
+ "num_input_tokens_seen": 10787736,
+ "step": 250,
+ "train_runtime": 1967.3657,
+ "train_tokens_per_second": 5483.341
+ },
+ {
+ "epoch": 0.01716940479396714,
+ "grad_norm": 0.9740927827374989,
+ "learning_rate": 9.998195939088905e-06,
+ "loss": 0.46301860809326173,
+ "num_input_tokens_seen": 11014240,
+ "step": 255,
+ "train_runtime": 1995.3214,
+ "train_tokens_per_second": 5520.033
+ },
+ {
+ "epoch": 0.017506059789927284,
+ "grad_norm": 1.0005631289772583,
+ "learning_rate": 9.998124218478797e-06,
+ "loss": 0.4852250099182129,
+ "num_input_tokens_seen": 11227928,
+ "step": 260,
+ "train_runtime": 2040.8305,
+ "train_tokens_per_second": 5501.646
+ },
+ {
+ "epoch": 0.017842714785887424,
+ "grad_norm": 3.2898314540824733,
+ "learning_rate": 9.998051100159192e-06,
+ "loss": 0.4806126594543457,
+ "num_input_tokens_seen": 11457336,
+ "step": 265,
+ "train_runtime": 2073.3223,
+ "train_tokens_per_second": 5526.076
+ },
+ {
+ "epoch": 0.018179369781847563,
+ "grad_norm": 1.0137684438609407,
+ "learning_rate": 9.997976584150533e-06,
+ "loss": 0.4908778667449951,
+ "num_input_tokens_seen": 11669408,
+ "step": 270,
+ "train_runtime": 2105.6176,
+ "train_tokens_per_second": 5542.036
+ },
+ {
+ "epoch": 0.018516024777807703,
+ "grad_norm": 0.882254713488144,
+ "learning_rate": 9.997900670473662e-06,
+ "loss": 0.4400303840637207,
+ "num_input_tokens_seen": 11900376,
+ "step": 275,
+ "train_runtime": 2145.8648,
+ "train_tokens_per_second": 5545.725
+ },
+ {
+ "epoch": 0.018852679773767842,
+ "grad_norm": 1.0103112069849245,
+ "learning_rate": 9.997823359149804e-06,
+ "loss": 0.47128896713256835,
+ "num_input_tokens_seen": 12115296,
+ "step": 280,
+ "train_runtime": 2180.0754,
+ "train_tokens_per_second": 5557.283
+ },
+ {
+ "epoch": 0.01918933476972798,
+ "grad_norm": 2.383562470887594,
+ "learning_rate": 9.997744650200579e-06,
+ "loss": 0.5039300441741943,
+ "num_input_tokens_seen": 12335296,
+ "step": 285,
+ "train_runtime": 2224.7672,
+ "train_tokens_per_second": 5544.533
+ },
+ {
+ "epoch": 0.01952598976568812,
+ "grad_norm": 1.012511096525962,
+ "learning_rate": 9.997664543648e-06,
+ "loss": 0.47290778160095215,
+ "num_input_tokens_seen": 12550016,
+ "step": 290,
+ "train_runtime": 2269.2912,
+ "train_tokens_per_second": 5530.368
+ },
+ {
+ "epoch": 0.019862644761648264,
+ "grad_norm": 0.9200150844694661,
+ "learning_rate": 9.997583039514469e-06,
+ "loss": 0.47549943923950194,
+ "num_input_tokens_seen": 12769944,
+ "step": 295,
+ "train_runtime": 2310.0213,
+ "train_tokens_per_second": 5528.063
+ },
+ {
+ "epoch": 0.020199299757608404,
+ "grad_norm": 1.0603122997620704,
+ "learning_rate": 9.997500137822776e-06,
+ "loss": 0.48865842819213867,
+ "num_input_tokens_seen": 12994152,
+ "step": 300,
+ "train_runtime": 2347.3612,
+ "train_tokens_per_second": 5535.642
+ },
+ {
+ "epoch": 0.020535954753568543,
+ "grad_norm": 1.1560224095031963,
+ "learning_rate": 9.997415838596106e-06,
+ "loss": 0.46552267074584963,
+ "num_input_tokens_seen": 13206104,
+ "step": 305,
+ "train_runtime": 2390.8704,
+ "train_tokens_per_second": 5523.555
+ },
+ {
+ "epoch": 0.020872609749528682,
+ "grad_norm": 0.9496962272740035,
+ "learning_rate": 9.997330141858032e-06,
+ "loss": 0.4945693492889404,
+ "num_input_tokens_seen": 13423544,
+ "step": 310,
+ "train_runtime": 2429.9223,
+ "train_tokens_per_second": 5524.269
+ },
+ {
+ "epoch": 0.021209264745488822,
+ "grad_norm": 1.1164280869632712,
+ "learning_rate": 9.997243047632519e-06,
+ "loss": 0.47294368743896487,
+ "num_input_tokens_seen": 13640584,
+ "step": 315,
+ "train_runtime": 2463.856,
+ "train_tokens_per_second": 5536.275
+ },
+ {
+ "epoch": 0.02154591974144896,
+ "grad_norm": 0.9211260067991657,
+ "learning_rate": 9.997154555943923e-06,
+ "loss": 0.43665332794189454,
+ "num_input_tokens_seen": 13858888,
+ "step": 320,
+ "train_runtime": 2506.1,
+ "train_tokens_per_second": 5530.062
+ },
+ {
+ "epoch": 0.021882574737409104,
+ "grad_norm": 1.0719268322873283,
+ "learning_rate": 9.997064666816992e-06,
+ "loss": 0.4625236034393311,
+ "num_input_tokens_seen": 14076648,
+ "step": 325,
+ "train_runtime": 2544.8417,
+ "train_tokens_per_second": 5531.443
+ },
+ {
+ "epoch": 0.022219229733369244,
+ "grad_norm": 1.146001730574971,
+ "learning_rate": 9.996973380276858e-06,
+ "loss": 0.5440494060516358,
+ "num_input_tokens_seen": 14299864,
+ "step": 330,
+ "train_runtime": 2575.9234,
+ "train_tokens_per_second": 5551.355
+ },
+ {
+ "epoch": 0.022555884729329383,
+ "grad_norm": 1.1489261273369993,
+ "learning_rate": 9.996880696349053e-06,
+ "loss": 0.45033721923828124,
+ "num_input_tokens_seen": 14522560,
+ "step": 335,
+ "train_runtime": 2614.1655,
+ "train_tokens_per_second": 5555.333
+ },
+ {
+ "epoch": 0.022892539725289523,
+ "grad_norm": 1.0606322220744733,
+ "learning_rate": 9.996786615059497e-06,
+ "loss": 0.5329171180725097,
+ "num_input_tokens_seen": 14746648,
+ "step": 340,
+ "train_runtime": 2655.6018,
+ "train_tokens_per_second": 5553.034
+ },
+ {
+ "epoch": 0.023229194721249662,
+ "grad_norm": 0.9692502297767218,
+ "learning_rate": 9.996691136434498e-06,
+ "loss": 0.4342483997344971,
+ "num_input_tokens_seen": 14969080,
+ "step": 345,
+ "train_runtime": 2690.8338,
+ "train_tokens_per_second": 5562.989
+ },
+ {
+ "epoch": 0.023565849717209802,
+ "grad_norm": 0.8994912857959015,
+ "learning_rate": 9.996594260500753e-06,
+ "loss": 0.507387924194336,
+ "num_input_tokens_seen": 15193912,
+ "step": 350,
+ "train_runtime": 2729.4748,
+ "train_tokens_per_second": 5566.606
+ },
+ {
+ "epoch": 0.023902504713169945,
+ "grad_norm": 0.9930018613669586,
+ "learning_rate": 9.996495987285358e-06,
+ "loss": 0.4612261772155762,
+ "num_input_tokens_seen": 15411408,
+ "step": 355,
+ "train_runtime": 2773.0354,
+ "train_tokens_per_second": 5557.595
+ },
+ {
+ "epoch": 0.024239159709130084,
+ "grad_norm": 1.0882060582433957,
+ "learning_rate": 9.996396316815793e-06,
+ "loss": 0.4673755645751953,
+ "num_input_tokens_seen": 15630480,
+ "step": 360,
+ "train_runtime": 2815.6019,
+ "train_tokens_per_second": 5551.381
+ },
+ {
+ "epoch": 0.024575814705090224,
+ "grad_norm": 1.1223066651934839,
+ "learning_rate": 9.99629524911993e-06,
+ "loss": 0.4566653251647949,
+ "num_input_tokens_seen": 15847552,
+ "step": 365,
+ "train_runtime": 2852.9783,
+ "train_tokens_per_second": 5554.74
+ },
+ {
+ "epoch": 0.024912469701050363,
+ "grad_norm": 1.0964642711464192,
+ "learning_rate": 9.996192784226033e-06,
+ "loss": 0.440608549118042,
+ "num_input_tokens_seen": 16049488,
+ "step": 370,
+ "train_runtime": 2887.3093,
+ "train_tokens_per_second": 5558.631
+ },
+ {
+ "epoch": 0.025249124697010503,
+ "grad_norm": 0.9458333109972462,
+ "learning_rate": 9.996088922162755e-06,
+ "loss": 0.4601402759552002,
+ "num_input_tokens_seen": 16283544,
+ "step": 375,
+ "train_runtime": 2920.1965,
+ "train_tokens_per_second": 5576.181
+ },
+ {
+ "epoch": 0.025585779692970642,
+ "grad_norm": 0.9639187009881571,
+ "learning_rate": 9.995983662959141e-06,
+ "loss": 0.4678368091583252,
+ "num_input_tokens_seen": 16507176,
+ "step": 380,
+ "train_runtime": 2959.819,
+ "train_tokens_per_second": 5577.09
+ },
+ {
+ "epoch": 0.025922434688930785,
+ "grad_norm": 0.9786802993443171,
+ "learning_rate": 9.995877006644627e-06,
+ "loss": 0.4893208026885986,
+ "num_input_tokens_seen": 16739664,
+ "step": 385,
+ "train_runtime": 3003.5655,
+ "train_tokens_per_second": 5573.264
+ },
+ {
+ "epoch": 0.026259089684890925,
+ "grad_norm": 0.9983372688388842,
+ "learning_rate": 9.995768953249038e-06,
+ "loss": 0.4782050132751465,
+ "num_input_tokens_seen": 16966112,
+ "step": 390,
+ "train_runtime": 3042.5667,
+ "train_tokens_per_second": 5576.25
+ },
+ {
+ "epoch": 0.026595744680851064,
+ "grad_norm": 0.9763900936703368,
+ "learning_rate": 9.995659502802594e-06,
+ "loss": 0.4591067314147949,
+ "num_input_tokens_seen": 17184144,
+ "step": 395,
+ "train_runtime": 3081.608,
+ "train_tokens_per_second": 5576.356
+ },
+ {
+ "epoch": 0.026932399676811204,
+ "grad_norm": 1.1889179128899698,
+ "learning_rate": 9.995548655335899e-06,
+ "loss": 0.4847611427307129,
+ "num_input_tokens_seen": 17393504,
+ "step": 400,
+ "train_runtime": 3119.919,
+ "train_tokens_per_second": 5574.986
+ },
+ {
+ "epoch": 0.027269054672771343,
+ "grad_norm": 1.0385231933952215,
+ "learning_rate": 9.995436410879951e-06,
+ "loss": 0.5011202812194824,
+ "num_input_tokens_seen": 17618352,
+ "step": 405,
+ "train_runtime": 3158.5186,
+ "train_tokens_per_second": 5578.043
+ },
+ {
+ "epoch": 0.027605709668731482,
+ "grad_norm": 1.0315411320040826,
+ "learning_rate": 9.99532276946614e-06,
+ "loss": 0.4390664577484131,
+ "num_input_tokens_seen": 17833504,
+ "step": 410,
+ "train_runtime": 3197.1139,
+ "train_tokens_per_second": 5578.001
+ },
+ {
+ "epoch": 0.027942364664691625,
+ "grad_norm": 0.9798271051589648,
+ "learning_rate": 9.995207731126246e-06,
+ "loss": 0.48116722106933596,
+ "num_input_tokens_seen": 18060040,
+ "step": 415,
+ "train_runtime": 3233.4074,
+ "train_tokens_per_second": 5585.451
+ },
+ {
+ "epoch": 0.028279019660651765,
+ "grad_norm": 0.9349852018887401,
+ "learning_rate": 9.99509129589244e-06,
+ "loss": 0.4614968776702881,
+ "num_input_tokens_seen": 18273536,
+ "step": 420,
+ "train_runtime": 3276.2797,
+ "train_tokens_per_second": 5577.526
+ },
+ {
+ "epoch": 0.028615674656611904,
+ "grad_norm": 1.1608131463568248,
+ "learning_rate": 9.994973463797279e-06,
+ "loss": 0.4571842193603516,
+ "num_input_tokens_seen": 18501288,
+ "step": 425,
+ "train_runtime": 3312.0947,
+ "train_tokens_per_second": 5585.978
+ },
+ {
+ "epoch": 0.028952329652572044,
+ "grad_norm": 0.9117559518264567,
+ "learning_rate": 9.99485423487372e-06,
+ "loss": 0.42873506546020507,
+ "num_input_tokens_seen": 18713400,
+ "step": 430,
+ "train_runtime": 3354.8456,
+ "train_tokens_per_second": 5578.021
+ },
+ {
+ "epoch": 0.029288984648532183,
+ "grad_norm": 1.0351008634262573,
+ "learning_rate": 9.994733609155097e-06,
+ "loss": 0.46686439514160155,
+ "num_input_tokens_seen": 18927424,
+ "step": 435,
+ "train_runtime": 3390.8404,
+ "train_tokens_per_second": 5581.927
+ },
+ {
+ "epoch": 0.029625639644492323,
+ "grad_norm": 0.9515109413699648,
+ "learning_rate": 9.994611586675149e-06,
+ "loss": 0.4384657382965088,
+ "num_input_tokens_seen": 19153848,
+ "step": 440,
+ "train_runtime": 3427.1383,
+ "train_tokens_per_second": 5588.875
+ },
+ {
+ "epoch": 0.029962294640452466,
+ "grad_norm": 2.170654371130382,
+ "learning_rate": 9.994488167468e-06,
+ "loss": 0.457914924621582,
+ "num_input_tokens_seen": 19367752,
+ "step": 445,
+ "train_runtime": 3463.4205,
+ "train_tokens_per_second": 5592.088
+ },
+ {
+ "epoch": 0.030298949636412605,
+ "grad_norm": 0.9213402764770521,
+ "learning_rate": 9.994363351568158e-06,
+ "loss": 0.49051642417907715,
+ "num_input_tokens_seen": 19581376,
+ "step": 450,
+ "train_runtime": 3507.0672,
+ "train_tokens_per_second": 5583.405
+ },
+ {
+ "epoch": 0.030635604632372745,
+ "grad_norm": 1.0129019967889514,
+ "learning_rate": 9.994237139010532e-06,
+ "loss": 0.4277812480926514,
+ "num_input_tokens_seen": 19787728,
+ "step": 455,
+ "train_runtime": 3546.6942,
+ "train_tokens_per_second": 5579.203
+ },
+ {
+ "epoch": 0.030972259628332884,
+ "grad_norm": 0.9251976717430972,
+ "learning_rate": 9.994109529830414e-06,
+ "loss": 0.4721639633178711,
+ "num_input_tokens_seen": 20024672,
+ "step": 460,
+ "train_runtime": 3585.3338,
+ "train_tokens_per_second": 5585.163
+ },
+ {
+ "epoch": 0.03130891462429303,
+ "grad_norm": 0.8419132637197803,
+ "learning_rate": 9.993980524063494e-06,
+ "loss": 0.44678316116333006,
+ "num_input_tokens_seen": 20247816,
+ "step": 465,
+ "train_runtime": 3622.8878,
+ "train_tokens_per_second": 5588.861
+ },
+ {
+ "epoch": 0.03164556962025317,
+ "grad_norm": 0.8727373261991417,
+ "learning_rate": 9.993850121745842e-06,
+ "loss": 0.448548698425293,
+ "num_input_tokens_seen": 20469008,
+ "step": 470,
+ "train_runtime": 3663.8929,
+ "train_tokens_per_second": 5586.683
+ },
+ {
+ "epoch": 0.031982224616213306,
+ "grad_norm": 0.9661829335825901,
+ "learning_rate": 9.99371832291393e-06,
+ "loss": 0.44652185440063474,
+ "num_input_tokens_seen": 20684368,
+ "step": 475,
+ "train_runtime": 3702.3594,
+ "train_tokens_per_second": 5586.807
+ },
+ {
+ "epoch": 0.032318879612173446,
+ "grad_norm": 1.052144245514416,
+ "learning_rate": 9.993585127604612e-06,
+ "loss": 0.4458156108856201,
+ "num_input_tokens_seen": 20888824,
+ "step": 480,
+ "train_runtime": 3737.8839,
+ "train_tokens_per_second": 5588.409
+ },
+ {
+ "epoch": 0.032655534608133585,
+ "grad_norm": 0.8727464499212149,
+ "learning_rate": 9.993450535855136e-06,
+ "loss": 0.4528946876525879,
+ "num_input_tokens_seen": 21093776,
+ "step": 485,
+ "train_runtime": 3771.0576,
+ "train_tokens_per_second": 5593.597
+ },
+ {
+ "epoch": 0.032992189604093725,
+ "grad_norm": 1.4490470715769006,
+ "learning_rate": 9.993314547703142e-06,
+ "loss": 0.43402805328369143,
+ "num_input_tokens_seen": 21314528,
+ "step": 490,
+ "train_runtime": 3810.8049,
+ "train_tokens_per_second": 5593.183
+ },
+ {
+ "epoch": 0.033328844600053864,
+ "grad_norm": 0.913157610231942,
+ "learning_rate": 9.993177163186655e-06,
+ "loss": 0.44025378227233886,
+ "num_input_tokens_seen": 21529648,
+ "step": 495,
+ "train_runtime": 3846.3087,
+ "train_tokens_per_second": 5597.483
+ },
+ {
+ "epoch": 0.033665499596014004,
+ "grad_norm": 1.0732207036056458,
+ "learning_rate": 9.993038382344097e-06,
+ "loss": 0.43718814849853516,
+ "num_input_tokens_seen": 21744008,
+ "step": 500,
+ "train_runtime": 3886.9027,
+ "train_tokens_per_second": 5594.174
+ },
+ {
+ "epoch": 0.03400215459197414,
+ "grad_norm": 0.9385843989678325,
+ "learning_rate": 9.992898205214279e-06,
+ "loss": 0.44336519241333006,
+ "num_input_tokens_seen": 21963168,
+ "step": 505,
+ "train_runtime": 3919.9586,
+ "train_tokens_per_second": 5602.908
+ },
+ {
+ "epoch": 0.03433880958793428,
+ "grad_norm": 0.9966215411858433,
+ "learning_rate": 9.992756631836397e-06,
+ "loss": 0.4398432731628418,
+ "num_input_tokens_seen": 22181088,
+ "step": 510,
+ "train_runtime": 3967.3085,
+ "train_tokens_per_second": 5590.966
+ },
+ {
+ "epoch": 0.03467546458389442,
+ "grad_norm": 0.8920648260369749,
+ "learning_rate": 9.992613662250044e-06,
+ "loss": 0.4598443031311035,
+ "num_input_tokens_seen": 22402688,
+ "step": 515,
+ "train_runtime": 4007.4675,
+ "train_tokens_per_second": 5590.236
+ },
+ {
+ "epoch": 0.03501211957985457,
+ "grad_norm": 0.9290820640509749,
+ "learning_rate": 9.992469296495201e-06,
+ "loss": 0.4575651168823242,
+ "num_input_tokens_seen": 22628840,
+ "step": 520,
+ "train_runtime": 4042.7584,
+ "train_tokens_per_second": 5597.376
+ },
+ {
+ "epoch": 0.03534877457581471,
+ "grad_norm": 0.9109745499097042,
+ "learning_rate": 9.992323534612238e-06,
+ "loss": 0.414502477645874,
+ "num_input_tokens_seen": 22848504,
+ "step": 525,
+ "train_runtime": 4082.2649,
+ "train_tokens_per_second": 5597.017
+ },
+ {
+ "epoch": 0.03568542957177485,
+ "grad_norm": 1.1344769076017303,
+ "learning_rate": 9.99217637664192e-06,
+ "loss": 0.45860743522644043,
+ "num_input_tokens_seen": 23072192,
+ "step": 530,
+ "train_runtime": 4119.8107,
+ "train_tokens_per_second": 5600.304
+ },
+ {
+ "epoch": 0.03602208456773499,
+ "grad_norm": 0.940449394205852,
+ "learning_rate": 9.992027822625397e-06,
+ "loss": 0.45654983520507814,
+ "num_input_tokens_seen": 23281952,
+ "step": 535,
+ "train_runtime": 4161.0794,
+ "train_tokens_per_second": 5595.171
+ },
+ {
+ "epoch": 0.036358739563695126,
+ "grad_norm": 1.0036660879417987,
+ "learning_rate": 9.99187787260421e-06,
+ "loss": 0.46324939727783204,
+ "num_input_tokens_seen": 23492592,
+ "step": 540,
+ "train_runtime": 4199.7934,
+ "train_tokens_per_second": 5593.749
+ },
+ {
+ "epoch": 0.036695394559655266,
+ "grad_norm": 0.9529512682682689,
+ "learning_rate": 9.991726526620294e-06,
+ "loss": 0.4310856819152832,
+ "num_input_tokens_seen": 23724784,
+ "step": 545,
+ "train_runtime": 4235.651,
+ "train_tokens_per_second": 5601.213
+ },
+ {
+ "epoch": 0.037032049555615405,
+ "grad_norm": 1.0092720255970657,
+ "learning_rate": 9.991573784715973e-06,
+ "loss": 0.4721970081329346,
+ "num_input_tokens_seen": 23941816,
+ "step": 550,
+ "train_runtime": 4278.956,
+ "train_tokens_per_second": 5595.247
+ },
+ {
+ "epoch": 0.037368704551575545,
+ "grad_norm": 1.0108753586712804,
+ "learning_rate": 9.99141964693396e-06,
+ "loss": 0.48006014823913573,
+ "num_input_tokens_seen": 24165232,
+ "step": 555,
+ "train_runtime": 4315.3047,
+ "train_tokens_per_second": 5599.89
+ },
+ {
+ "epoch": 0.037705359547535684,
+ "grad_norm": 1.0343909095504493,
+ "learning_rate": 9.99126411331736e-06,
+ "loss": 0.4417973518371582,
+ "num_input_tokens_seen": 24389896,
+ "step": 560,
+ "train_runtime": 4359.1735,
+ "train_tokens_per_second": 5595.074
+ },
+ {
+ "epoch": 0.038042014543495824,
+ "grad_norm": 1.0113397276422895,
+ "learning_rate": 9.991107183909665e-06,
+ "loss": 0.4680752754211426,
+ "num_input_tokens_seen": 24614992,
+ "step": 565,
+ "train_runtime": 4391.4505,
+ "train_tokens_per_second": 5605.208
+ },
+ {
+ "epoch": 0.03837866953945596,
+ "grad_norm": 0.9954842488566088,
+ "learning_rate": 9.990948858754763e-06,
+ "loss": 0.4640792369842529,
+ "num_input_tokens_seen": 24823080,
+ "step": 570,
+ "train_runtime": 4434.0745,
+ "train_tokens_per_second": 5598.255
+ },
+ {
+ "epoch": 0.0387153245354161,
+ "grad_norm": 0.9153320849806976,
+ "learning_rate": 9.990789137896927e-06,
+ "loss": 0.4609560966491699,
+ "num_input_tokens_seen": 25044784,
+ "step": 575,
+ "train_runtime": 4471.0247,
+ "train_tokens_per_second": 5601.576
+ },
+ {
+ "epoch": 0.03905197953137624,
+ "grad_norm": 0.9718791869837801,
+ "learning_rate": 9.990628021380823e-06,
+ "loss": 0.47663059234619143,
+ "num_input_tokens_seen": 25262200,
+ "step": 580,
+ "train_runtime": 4509.6496,
+ "train_tokens_per_second": 5601.81
+ },
+ {
+ "epoch": 0.03938863452733639,
+ "grad_norm": 1.218415508866825,
+ "learning_rate": 9.990465509251508e-06,
+ "loss": 0.46056556701660156,
+ "num_input_tokens_seen": 25473136,
+ "step": 585,
+ "train_runtime": 4541.7073,
+ "train_tokens_per_second": 5608.714
+ },
+ {
+ "epoch": 0.03972528952329653,
+ "grad_norm": 1.0200294193913204,
+ "learning_rate": 9.990301601554425e-06,
+ "loss": 0.48001799583435056,
+ "num_input_tokens_seen": 25671088,
+ "step": 590,
+ "train_runtime": 4581.4528,
+ "train_tokens_per_second": 5603.264
+ },
+ {
+ "epoch": 0.04006194451925667,
+ "grad_norm": 1.0450169479009095,
+ "learning_rate": 9.990136298335414e-06,
+ "loss": 0.40839109420776365,
+ "num_input_tokens_seen": 25887472,
+ "step": 595,
+ "train_runtime": 4618.2668,
+ "train_tokens_per_second": 5605.452
+ },
+ {
+ "epoch": 0.04039859951521681,
+ "grad_norm": 0.9533941163271269,
+ "learning_rate": 9.989969599640698e-06,
+ "loss": 0.4223615646362305,
+ "num_input_tokens_seen": 26117456,
+ "step": 600,
+ "train_runtime": 4650.9664,
+ "train_tokens_per_second": 5615.49
+ },
+ {
+ "epoch": 0.040735254511176947,
+ "grad_norm": 0.9258754840147914,
+ "learning_rate": 9.989801505516895e-06,
+ "loss": 0.505678129196167,
+ "num_input_tokens_seen": 26348056,
+ "step": 605,
+ "train_runtime": 4690.0532,
+ "train_tokens_per_second": 5617.859
+ },
+ {
+ "epoch": 0.041071909507137086,
+ "grad_norm": 0.9866589002080994,
+ "learning_rate": 9.989632016011012e-06,
+ "loss": 0.47707581520080566,
+ "num_input_tokens_seen": 26572040,
+ "step": 610,
+ "train_runtime": 4721.6631,
+ "train_tokens_per_second": 5627.687
+ },
+ {
+ "epoch": 0.041408564503097225,
+ "grad_norm": 1.0374837134071888,
+ "learning_rate": 9.989461131170448e-06,
+ "loss": 0.4833499908447266,
+ "num_input_tokens_seen": 26769320,
+ "step": 615,
+ "train_runtime": 4762.2337,
+ "train_tokens_per_second": 5621.169
+ },
+ {
+ "epoch": 0.041745219499057365,
+ "grad_norm": 0.8823212064118572,
+ "learning_rate": 9.98928885104299e-06,
+ "loss": 0.4514059543609619,
+ "num_input_tokens_seen": 26988024,
+ "step": 620,
+ "train_runtime": 4795.3541,
+ "train_tokens_per_second": 5627.952
+ },
+ {
+ "epoch": 0.042081874495017504,
+ "grad_norm": 0.8780469782436839,
+ "learning_rate": 9.989115175676812e-06,
+ "loss": 0.4226998329162598,
+ "num_input_tokens_seen": 27212624,
+ "step": 625,
+ "train_runtime": 4834.4544,
+ "train_tokens_per_second": 5628.892
+ },
+ {
+ "epoch": 0.042418529490977644,
+ "grad_norm": 0.9554506078310793,
+ "learning_rate": 9.988940105120483e-06,
+ "loss": 0.47803831100463867,
+ "num_input_tokens_seen": 27422208,
+ "step": 630,
+ "train_runtime": 4877.5946,
+ "train_tokens_per_second": 5622.076
+ },
+ {
+ "epoch": 0.04275518448693778,
+ "grad_norm": 0.93348166885423,
+ "learning_rate": 9.988763639422965e-06,
+ "loss": 0.4802694797515869,
+ "num_input_tokens_seen": 27650384,
+ "step": 635,
+ "train_runtime": 4915.966,
+ "train_tokens_per_second": 5624.608
+ },
+ {
+ "epoch": 0.04309183948289792,
+ "grad_norm": 0.9769454728471132,
+ "learning_rate": 9.988585778633604e-06,
+ "loss": 0.4367190361022949,
+ "num_input_tokens_seen": 27851424,
+ "step": 640,
+ "train_runtime": 4948.8643,
+ "train_tokens_per_second": 5627.841
+ },
+ {
+ "epoch": 0.04342849447885807,
+ "grad_norm": 0.9848655689806975,
+ "learning_rate": 9.988406522802134e-06,
+ "loss": 0.45001683235168455,
+ "num_input_tokens_seen": 28060088,
+ "step": 645,
+ "train_runtime": 4991.2246,
+ "train_tokens_per_second": 5621.884
+ },
+ {
+ "epoch": 0.04376514947481821,
+ "grad_norm": 0.9626370029590127,
+ "learning_rate": 9.98822587197869e-06,
+ "loss": 0.41652770042419435,
+ "num_input_tokens_seen": 28274912,
+ "step": 650,
+ "train_runtime": 5025.7373,
+ "train_tokens_per_second": 5626.023
+ },
+ {
+ "epoch": 0.04410180447077835,
+ "grad_norm": 0.9983674144400133,
+ "learning_rate": 9.988043826213785e-06,
+ "loss": 0.4282545566558838,
+ "num_input_tokens_seen": 28481968,
+ "step": 655,
+ "train_runtime": 5062.1603,
+ "train_tokens_per_second": 5626.445
+ },
+ {
+ "epoch": 0.04443845946673849,
+ "grad_norm": 2.4559661675719884,
+ "learning_rate": 9.98786038555833e-06,
+ "loss": 0.4085527420043945,
+ "num_input_tokens_seen": 28702472,
+ "step": 660,
+ "train_runtime": 5099.8424,
+ "train_tokens_per_second": 5628.11
+ },
+ {
+ "epoch": 0.04477511446269863,
+ "grad_norm": 0.8332845163143295,
+ "learning_rate": 9.987675550063625e-06,
+ "loss": 0.44841432571411133,
+ "num_input_tokens_seen": 28932512,
+ "step": 665,
+ "train_runtime": 5135.2428,
+ "train_tokens_per_second": 5634.108
+ },
+ {
+ "epoch": 0.04511176945865877,
+ "grad_norm": 1.0017265357539382,
+ "learning_rate": 9.987489319781355e-06,
+ "loss": 0.4431906223297119,
+ "num_input_tokens_seen": 29153640,
+ "step": 670,
+ "train_runtime": 5165.908,
+ "train_tokens_per_second": 5643.469
+ },
+ {
+ "epoch": 0.045448424454618906,
+ "grad_norm": 0.8381101119904498,
+ "learning_rate": 9.987301694763602e-06,
+ "loss": 0.4798222541809082,
+ "num_input_tokens_seen": 29368992,
+ "step": 675,
+ "train_runtime": 5201.341,
+ "train_tokens_per_second": 5646.427
+ },
+ {
+ "epoch": 0.045785079450579046,
+ "grad_norm": 0.9739030010713555,
+ "learning_rate": 9.987112675062834e-06,
+ "loss": 0.4200302600860596,
+ "num_input_tokens_seen": 29582872,
+ "step": 680,
+ "train_runtime": 5240.8223,
+ "train_tokens_per_second": 5644.7
+ },
+ {
+ "epoch": 0.046121734446539185,
+ "grad_norm": 0.9923641076223235,
+ "learning_rate": 9.986922260731908e-06,
+ "loss": 0.42931652069091797,
+ "num_input_tokens_seen": 29791056,
+ "step": 685,
+ "train_runtime": 5280.43,
+ "train_tokens_per_second": 5641.786
+ },
+ {
+ "epoch": 0.046458389442499325,
+ "grad_norm": 0.8786561915980274,
+ "learning_rate": 9.986730451824073e-06,
+ "loss": 0.4656373977661133,
+ "num_input_tokens_seen": 30001648,
+ "step": 690,
+ "train_runtime": 5325.5768,
+ "train_tokens_per_second": 5633.502
+ },
+ {
+ "epoch": 0.046795044438459464,
+ "grad_norm": 0.9155252116619295,
+ "learning_rate": 9.986537248392969e-06,
+ "loss": 0.4756736755371094,
+ "num_input_tokens_seen": 30223896,
+ "step": 695,
+ "train_runtime": 5365.8859,
+ "train_tokens_per_second": 5632.601
+ },
+ {
+ "epoch": 0.047131699434419604,
+ "grad_norm": 1.0345486643678798,
+ "learning_rate": 9.986342650492626e-06,
+ "loss": 0.4506352424621582,
+ "num_input_tokens_seen": 30445352,
+ "step": 700,
+ "train_runtime": 5399.4485,
+ "train_tokens_per_second": 5638.604
+ },
+ {
+ "epoch": 0.04746835443037975,
+ "grad_norm": 0.919553088389443,
+ "learning_rate": 9.98614665817746e-06,
+ "loss": 0.42203187942504883,
+ "num_input_tokens_seen": 30642128,
+ "step": 705,
+ "train_runtime": 5439.3583,
+ "train_tokens_per_second": 5633.409
+ },
+ {
+ "epoch": 0.04780500942633989,
+ "grad_norm": 0.9271703985798777,
+ "learning_rate": 9.98594927150228e-06,
+ "loss": 0.427701473236084,
+ "num_input_tokens_seen": 30873224,
+ "step": 710,
+ "train_runtime": 5478.7451,
+ "train_tokens_per_second": 5635.09
+ },
+ {
+ "epoch": 0.04814166442230003,
+ "grad_norm": 0.9508538400970585,
+ "learning_rate": 9.985750490522286e-06,
+ "loss": 0.39564857482910154,
+ "num_input_tokens_seen": 31087512,
+ "step": 715,
+ "train_runtime": 5521.2724,
+ "train_tokens_per_second": 5630.498
+ },
+ {
+ "epoch": 0.04847831941826017,
+ "grad_norm": 0.9372110505017411,
+ "learning_rate": 9.985550315293066e-06,
+ "loss": 0.4636754035949707,
+ "num_input_tokens_seen": 31318912,
+ "step": 720,
+ "train_runtime": 5557.4337,
+ "train_tokens_per_second": 5635.499
+ },
+ {
+ "epoch": 0.04881497441422031,
+ "grad_norm": 0.9970520932843633,
+ "learning_rate": 9.985348745870598e-06,
+ "loss": 0.4597306728363037,
+ "num_input_tokens_seen": 31546096,
+ "step": 725,
+ "train_runtime": 5592.8911,
+ "train_tokens_per_second": 5640.392
+ },
+ {
+ "epoch": 0.04915162941018045,
+ "grad_norm": 0.8271247348731543,
+ "learning_rate": 9.98514578231125e-06,
+ "loss": 0.4194610595703125,
+ "num_input_tokens_seen": 31774968,
+ "step": 730,
+ "train_runtime": 5630.8722,
+ "train_tokens_per_second": 5642.992
+ },
+ {
+ "epoch": 0.04948828440614059,
+ "grad_norm": 0.9452005671897916,
+ "learning_rate": 9.984941424671781e-06,
+ "loss": 0.4705995559692383,
+ "num_input_tokens_seen": 31991384,
+ "step": 735,
+ "train_runtime": 5672.8388,
+ "train_tokens_per_second": 5639.396
+ },
+ {
+ "epoch": 0.049824939402100726,
+ "grad_norm": 0.9711026791829106,
+ "learning_rate": 9.984735673009337e-06,
+ "loss": 0.43364248275756834,
+ "num_input_tokens_seen": 32211232,
+ "step": 740,
+ "train_runtime": 5717.5318,
+ "train_tokens_per_second": 5633.765
+ },
+ {
+ "epoch": 0.050161594398060866,
+ "grad_norm": 0.8583208108971163,
+ "learning_rate": 9.984528527381458e-06,
+ "loss": 0.4496613502502441,
+ "num_input_tokens_seen": 32432504,
+ "step": 745,
+ "train_runtime": 5758.1136,
+ "train_tokens_per_second": 5632.488
+ },
+ {
+ "epoch": 0.050498249394021005,
+ "grad_norm": 1.0983008551303333,
+ "learning_rate": 9.984319987846072e-06,
+ "loss": 0.4476910591125488,
+ "num_input_tokens_seen": 32644184,
+ "step": 750,
+ "train_runtime": 5795.8989,
+ "train_tokens_per_second": 5632.29
+ },
+ {
+ "epoch": 0.050834904389981145,
+ "grad_norm": 0.9341698177556362,
+ "learning_rate": 9.984110054461495e-06,
+ "loss": 0.4494589328765869,
+ "num_input_tokens_seen": 32883080,
+ "step": 755,
+ "train_runtime": 5831.2135,
+ "train_tokens_per_second": 5639.149
+ },
+ {
+ "epoch": 0.051171559385941284,
+ "grad_norm": 2.9705690058280205,
+ "learning_rate": 9.983898727286435e-06,
+ "loss": 0.44231281280517576,
+ "num_input_tokens_seen": 33102688,
+ "step": 760,
+ "train_runtime": 5864.3725,
+ "train_tokens_per_second": 5644.711
+ },
+ {
+ "epoch": 0.05150821438190143,
+ "grad_norm": 0.9183883314753698,
+ "learning_rate": 9.983686006379985e-06,
+ "loss": 0.45505475997924805,
+ "num_input_tokens_seen": 33319336,
+ "step": 765,
+ "train_runtime": 5901.491,
+ "train_tokens_per_second": 5645.918
+ },
+ {
+ "epoch": 0.05184486937786157,
+ "grad_norm": 0.9313957301525887,
+ "learning_rate": 9.983471891801638e-06,
+ "loss": 0.437741756439209,
+ "num_input_tokens_seen": 33545192,
+ "step": 770,
+ "train_runtime": 5938.7864,
+ "train_tokens_per_second": 5648.493
+ },
+ {
+ "epoch": 0.05218152437382171,
+ "grad_norm": 0.8973015582826919,
+ "learning_rate": 9.983256383611267e-06,
+ "loss": 0.441961669921875,
+ "num_input_tokens_seen": 33767136,
+ "step": 775,
+ "train_runtime": 5974.9184,
+ "train_tokens_per_second": 5651.481
+ },
+ {
+ "epoch": 0.05251817936978185,
+ "grad_norm": 1.380873298582274,
+ "learning_rate": 9.983039481869138e-06,
+ "loss": 0.42367143630981446,
+ "num_input_tokens_seen": 33985712,
+ "step": 780,
+ "train_runtime": 6009.8556,
+ "train_tokens_per_second": 5654.996
+ },
+ {
+ "epoch": 0.05285483436574199,
+ "grad_norm": 0.9605495456782687,
+ "learning_rate": 9.98282118663591e-06,
+ "loss": 0.4347693920135498,
+ "num_input_tokens_seen": 34208000,
+ "step": 785,
+ "train_runtime": 6046.7667,
+ "train_tokens_per_second": 5657.238
+ },
+ {
+ "epoch": 0.05319148936170213,
+ "grad_norm": 0.9570184798319451,
+ "learning_rate": 9.982601497972624e-06,
+ "loss": 0.4524986267089844,
+ "num_input_tokens_seen": 34425040,
+ "step": 790,
+ "train_runtime": 6085.7762,
+ "train_tokens_per_second": 5656.639
+ },
+ {
+ "epoch": 0.05352814435766227,
+ "grad_norm": 1.293632659414026,
+ "learning_rate": 9.982380415940717e-06,
+ "loss": 0.4259688377380371,
+ "num_input_tokens_seen": 34634712,
+ "step": 795,
+ "train_runtime": 6120.1944,
+ "train_tokens_per_second": 5659.087
+ },
+ {
+ "epoch": 0.05386479935362241,
+ "grad_norm": 0.9262120217395047,
+ "learning_rate": 9.982157940602015e-06,
+ "loss": 0.46526622772216797,
+ "num_input_tokens_seen": 34865376,
+ "step": 800,
+ "train_runtime": 6156.1798,
+ "train_tokens_per_second": 5663.476
+ },
+ {
+ "epoch": 0.05420145434958255,
+ "grad_norm": 0.892992639712076,
+ "learning_rate": 9.981934072018732e-06,
+ "loss": 0.44102907180786133,
+ "num_input_tokens_seen": 35098448,
+ "step": 805,
+ "train_runtime": 6189.6138,
+ "train_tokens_per_second": 5670.539
+ },
+ {
+ "epoch": 0.054538109345542686,
+ "grad_norm": 0.8797013169732244,
+ "learning_rate": 9.981708810253472e-06,
+ "loss": 0.4654733657836914,
+ "num_input_tokens_seen": 35317184,
+ "step": 810,
+ "train_runtime": 6223.9712,
+ "train_tokens_per_second": 5674.381
+ },
+ {
+ "epoch": 0.054874764341502826,
+ "grad_norm": 1.1385550766773516,
+ "learning_rate": 9.981482155369229e-06,
+ "loss": 0.4581299781799316,
+ "num_input_tokens_seen": 35543024,
+ "step": 815,
+ "train_runtime": 6264.1777,
+ "train_tokens_per_second": 5674.013
+ },
+ {
+ "epoch": 0.055211419337462965,
+ "grad_norm": 1.0108160361368144,
+ "learning_rate": 9.981254107429385e-06,
+ "loss": 0.4073037147521973,
+ "num_input_tokens_seen": 35752976,
+ "step": 820,
+ "train_runtime": 6301.9605,
+ "train_tokens_per_second": 5673.31
+ },
+ {
+ "epoch": 0.05554807433342311,
+ "grad_norm": 0.8271492373593813,
+ "learning_rate": 9.981024666497716e-06,
+ "loss": 0.4087496757507324,
+ "num_input_tokens_seen": 35970808,
+ "step": 825,
+ "train_runtime": 6336.9136,
+ "train_tokens_per_second": 5676.392
+ },
+ {
+ "epoch": 0.05588472932938325,
+ "grad_norm": 1.2448406201928153,
+ "learning_rate": 9.98079383263838e-06,
+ "loss": 0.451265811920166,
+ "num_input_tokens_seen": 36175264,
+ "step": 830,
+ "train_runtime": 6372.1661,
+ "train_tokens_per_second": 5677.075
+ },
+ {
+ "epoch": 0.05622138432534339,
+ "grad_norm": 1.0960090803895477,
+ "learning_rate": 9.980561605915932e-06,
+ "loss": 0.43454971313476565,
+ "num_input_tokens_seen": 36399968,
+ "step": 835,
+ "train_runtime": 6409.522,
+ "train_tokens_per_second": 5679.046
+ },
+ {
+ "epoch": 0.05655803932130353,
+ "grad_norm": 0.865350004624438,
+ "learning_rate": 9.980327986395316e-06,
+ "loss": 0.44866161346435546,
+ "num_input_tokens_seen": 36614352,
+ "step": 840,
+ "train_runtime": 6450.8502,
+ "train_tokens_per_second": 5675.896
+ },
+ {
+ "epoch": 0.05689469431726367,
+ "grad_norm": 0.9132401043618537,
+ "learning_rate": 9.980092974141855e-06,
+ "loss": 0.43543548583984376,
+ "num_input_tokens_seen": 36842440,
+ "step": 845,
+ "train_runtime": 6489.8829,
+ "train_tokens_per_second": 5676.904
+ },
+ {
+ "epoch": 0.05723134931322381,
+ "grad_norm": 0.9515826536482608,
+ "learning_rate": 9.979856569221276e-06,
+ "loss": 0.434114933013916,
+ "num_input_tokens_seen": 37064928,
+ "step": 850,
+ "train_runtime": 6535.3585,
+ "train_tokens_per_second": 5671.445
+ },
+ {
+ "epoch": 0.05756800430918395,
+ "grad_norm": 0.905243476379667,
+ "learning_rate": 9.979618771699688e-06,
+ "loss": 0.47480268478393556,
+ "num_input_tokens_seen": 37286696,
+ "step": 855,
+ "train_runtime": 6571.3163,
+ "train_tokens_per_second": 5674.159
+ },
+ {
+ "epoch": 0.05790465930514409,
+ "grad_norm": 0.9973188371129253,
+ "learning_rate": 9.979379581643588e-06,
+ "loss": 0.4271403312683105,
+ "num_input_tokens_seen": 37505104,
+ "step": 860,
+ "train_runtime": 6612.0758,
+ "train_tokens_per_second": 5672.213
+ },
+ {
+ "epoch": 0.05824131430110423,
+ "grad_norm": 0.891663822548588,
+ "learning_rate": 9.979138999119866e-06,
+ "loss": 0.46260747909545896,
+ "num_input_tokens_seen": 37721104,
+ "step": 865,
+ "train_runtime": 6657.2054,
+ "train_tokens_per_second": 5666.207
+ },
+ {
+ "epoch": 0.05857796929706437,
+ "grad_norm": 0.8193884764682856,
+ "learning_rate": 9.978897024195801e-06,
+ "loss": 0.43666720390319824,
+ "num_input_tokens_seen": 37959848,
+ "step": 870,
+ "train_runtime": 6692.9435,
+ "train_tokens_per_second": 5671.622
+ },
+ {
+ "epoch": 0.058914624293024506,
+ "grad_norm": 1.1669037394876847,
+ "learning_rate": 9.978653656939058e-06,
+ "loss": 0.4272760391235352,
+ "num_input_tokens_seen": 38179560,
+ "step": 875,
+ "train_runtime": 6736.655,
+ "train_tokens_per_second": 5667.436
+ },
+ {
+ "epoch": 0.059251279288984646,
+ "grad_norm": 1.0468766117182693,
+ "learning_rate": 9.978408897417699e-06,
+ "loss": 0.4613694190979004,
+ "num_input_tokens_seen": 38405040,
+ "step": 880,
+ "train_runtime": 6776.3735,
+ "train_tokens_per_second": 5667.492
+ },
+ {
+ "epoch": 0.059587934284944785,
+ "grad_norm": 0.8352176899590911,
+ "learning_rate": 9.978162745700165e-06,
+ "loss": 0.4417067527770996,
+ "num_input_tokens_seen": 38631232,
+ "step": 885,
+ "train_runtime": 6814.8051,
+ "train_tokens_per_second": 5668.721
+ },
+ {
+ "epoch": 0.05992458928090493,
+ "grad_norm": 1.0696188249598721,
+ "learning_rate": 9.977915201855293e-06,
+ "loss": 0.42413697242736814,
+ "num_input_tokens_seen": 38849712,
+ "step": 890,
+ "train_runtime": 6850.9387,
+ "train_tokens_per_second": 5670.714
+ },
+ {
+ "epoch": 0.06026124427686507,
+ "grad_norm": 0.9281445460422576,
+ "learning_rate": 9.977666265952309e-06,
+ "loss": 0.4329718589782715,
+ "num_input_tokens_seen": 39067416,
+ "step": 895,
+ "train_runtime": 6890.6802,
+ "train_tokens_per_second": 5669.602
+ },
+ {
+ "epoch": 0.06059789927282521,
+ "grad_norm": 0.8685965904886482,
+ "learning_rate": 9.977415938060823e-06,
+ "loss": 0.4434166431427002,
+ "num_input_tokens_seen": 39288624,
+ "step": 900,
+ "train_runtime": 6926.1271,
+ "train_tokens_per_second": 5672.524
+ },
+ {
+ "epoch": 0.06093455426878535,
+ "grad_norm": 0.9417641978342199,
+ "learning_rate": 9.977164218250846e-06,
+ "loss": 0.444110107421875,
+ "num_input_tokens_seen": 39505328,
+ "step": 905,
+ "train_runtime": 6968.8615,
+ "train_tokens_per_second": 5668.835
+ },
+ {
+ "epoch": 0.06127120926474549,
+ "grad_norm": 0.8557175345597763,
+ "learning_rate": 9.976911106592765e-06,
+ "loss": 0.4071072578430176,
+ "num_input_tokens_seen": 39717144,
+ "step": 910,
+ "train_runtime": 7009.1269,
+ "train_tokens_per_second": 5666.49
+ },
+ {
+ "epoch": 0.06160786426070563,
+ "grad_norm": 1.1399112064858985,
+ "learning_rate": 9.976656603157363e-06,
+ "loss": 0.4384052276611328,
+ "num_input_tokens_seen": 39933000,
+ "step": 915,
+ "train_runtime": 7045.7591,
+ "train_tokens_per_second": 5667.665
+ },
+ {
+ "epoch": 0.06194451925666577,
+ "grad_norm": 0.9115769460735773,
+ "learning_rate": 9.976400708015811e-06,
+ "loss": 0.42393884658813474,
+ "num_input_tokens_seen": 40145168,
+ "step": 920,
+ "train_runtime": 7079.4173,
+ "train_tokens_per_second": 5670.688
+ },
+ {
+ "epoch": 0.06228117425262591,
+ "grad_norm": 0.8414795692849011,
+ "learning_rate": 9.97614342123967e-06,
+ "loss": 0.4269254684448242,
+ "num_input_tokens_seen": 40379264,
+ "step": 925,
+ "train_runtime": 7111.1373,
+ "train_tokens_per_second": 5678.313
+ },
+ {
+ "epoch": 0.06261782924858605,
+ "grad_norm": 0.9352365482641929,
+ "learning_rate": 9.97588474290089e-06,
+ "loss": 0.4331661224365234,
+ "num_input_tokens_seen": 40594944,
+ "step": 930,
+ "train_runtime": 7152.76,
+ "train_tokens_per_second": 5675.424
+ },
+ {
+ "epoch": 0.0629544842445462,
+ "grad_norm": 1.0614310774414917,
+ "learning_rate": 9.975624673071806e-06,
+ "loss": 0.4371979236602783,
+ "num_input_tokens_seen": 40808856,
+ "step": 935,
+ "train_runtime": 7197.718,
+ "train_tokens_per_second": 5669.694
+ },
+ {
+ "epoch": 0.06329113924050633,
+ "grad_norm": 1.0079121416060646,
+ "learning_rate": 9.97536321182515e-06,
+ "loss": 0.4044677734375,
+ "num_input_tokens_seen": 41018224,
+ "step": 940,
+ "train_runtime": 7227.9115,
+ "train_tokens_per_second": 5674.976
+ },
+ {
+ "epoch": 0.06362779423646647,
+ "grad_norm": 0.9987688379119654,
+ "learning_rate": 9.975100359234037e-06,
+ "loss": 0.4564094066619873,
+ "num_input_tokens_seen": 41239392,
+ "step": 945,
+ "train_runtime": 7265.9862,
+ "train_tokens_per_second": 5675.677
+ },
+ {
+ "epoch": 0.06396444923242661,
+ "grad_norm": 0.9165255721069369,
+ "learning_rate": 9.974836115371972e-06,
+ "loss": 0.46511135101318357,
+ "num_input_tokens_seen": 41454352,
+ "step": 950,
+ "train_runtime": 7304.5238,
+ "train_tokens_per_second": 5675.161
+ },
+ {
+ "epoch": 0.06430110422838675,
+ "grad_norm": 0.9087923038014859,
+ "learning_rate": 9.974570480312852e-06,
+ "loss": 0.40421581268310547,
+ "num_input_tokens_seen": 41680576,
+ "step": 955,
+ "train_runtime": 7346.7849,
+ "train_tokens_per_second": 5673.308
+ },
+ {
+ "epoch": 0.06463775922434689,
+ "grad_norm": 1.0231903244202105,
+ "learning_rate": 9.974303454130962e-06,
+ "loss": 0.4464391231536865,
+ "num_input_tokens_seen": 41882360,
+ "step": 960,
+ "train_runtime": 7380.8428,
+ "train_tokens_per_second": 5674.469
+ },
+ {
+ "epoch": 0.06497441422030703,
+ "grad_norm": 0.8595399990273458,
+ "learning_rate": 9.97403503690097e-06,
+ "loss": 0.42637953758239744,
+ "num_input_tokens_seen": 42103128,
+ "step": 965,
+ "train_runtime": 7415.439,
+ "train_tokens_per_second": 5677.766
+ },
+ {
+ "epoch": 0.06531106921626717,
+ "grad_norm": 0.8422231085957664,
+ "learning_rate": 9.973765228697942e-06,
+ "loss": 0.39764556884765623,
+ "num_input_tokens_seen": 42328192,
+ "step": 970,
+ "train_runtime": 7453.2347,
+ "train_tokens_per_second": 5679.171
+ },
+ {
+ "epoch": 0.06564772421222731,
+ "grad_norm": 0.9247166172844082,
+ "learning_rate": 9.973494029597327e-06,
+ "loss": 0.43375320434570314,
+ "num_input_tokens_seen": 42531664,
+ "step": 975,
+ "train_runtime": 7487.6836,
+ "train_tokens_per_second": 5680.217
+ },
+ {
+ "epoch": 0.06598437920818745,
+ "grad_norm": 1.0151970555059149,
+ "learning_rate": 9.973221439674968e-06,
+ "loss": 0.4309089660644531,
+ "num_input_tokens_seen": 42740856,
+ "step": 980,
+ "train_runtime": 7527.1188,
+ "train_tokens_per_second": 5678.249
+ },
+ {
+ "epoch": 0.06632103420414759,
+ "grad_norm": 0.882649641197766,
+ "learning_rate": 9.972947459007089e-06,
+ "loss": 0.45548033714294434,
+ "num_input_tokens_seen": 42956288,
+ "step": 985,
+ "train_runtime": 7562.7131,
+ "train_tokens_per_second": 5680.01
+ },
+ {
+ "epoch": 0.06665768920010773,
+ "grad_norm": 0.9104741890803707,
+ "learning_rate": 9.972672087670313e-06,
+ "loss": 0.4385095119476318,
+ "num_input_tokens_seen": 43177272,
+ "step": 990,
+ "train_runtime": 7599.4941,
+ "train_tokens_per_second": 5681.598
+ },
+ {
+ "epoch": 0.06699434419606787,
+ "grad_norm": 0.8108451361340878,
+ "learning_rate": 9.972395325741644e-06,
+ "loss": 0.413985538482666,
+ "num_input_tokens_seen": 43400152,
+ "step": 995,
+ "train_runtime": 7640.1047,
+ "train_tokens_per_second": 5680.57
+ },
+ {
+ "epoch": 0.06733099919202801,
+ "grad_norm": 0.9025973522517067,
+ "learning_rate": 9.972117173298476e-06,
+ "loss": 0.38621523380279543,
+ "num_input_tokens_seen": 43614968,
+ "step": 1000,
+ "train_runtime": 7674.2806,
+ "train_tokens_per_second": 5683.265
+ },
+ {
+ "epoch": 0.06766765418798815,
+ "grad_norm": 0.971243318554936,
+ "learning_rate": 9.971837630418597e-06,
+ "loss": 0.42676420211791993,
+ "num_input_tokens_seen": 43835904,
+ "step": 1005,
+ "train_runtime": 7717.1168,
+ "train_tokens_per_second": 5680.347
+ },
+ {
+ "epoch": 0.06800430918394829,
+ "grad_norm": 0.9278937330533324,
+ "learning_rate": 9.971556697180179e-06,
+ "loss": 0.44001970291137693,
+ "num_input_tokens_seen": 44048512,
+ "step": 1010,
+ "train_runtime": 7756.4813,
+ "train_tokens_per_second": 5678.93
+ },
+ {
+ "epoch": 0.06834096417990843,
+ "grad_norm": 0.9429692040194982,
+ "learning_rate": 9.971274373661784e-06,
+ "loss": 0.4194046974182129,
+ "num_input_tokens_seen": 44271000,
+ "step": 1015,
+ "train_runtime": 7787.4236,
+ "train_tokens_per_second": 5684.935
+ },
+ {
+ "epoch": 0.06867761917586857,
+ "grad_norm": 1.111228553028809,
+ "learning_rate": 9.97099065994236e-06,
+ "loss": 0.39097912311553956,
+ "num_input_tokens_seen": 44499576,
+ "step": 1020,
+ "train_runtime": 7822.0558,
+ "train_tokens_per_second": 5688.987
+ },
+ {
+ "epoch": 0.0690142741718287,
+ "grad_norm": 1.0049309947436895,
+ "learning_rate": 9.970705556101253e-06,
+ "loss": 0.43691391944885255,
+ "num_input_tokens_seen": 44702000,
+ "step": 1025,
+ "train_runtime": 7857.7887,
+ "train_tokens_per_second": 5688.878
+ },
+ {
+ "epoch": 0.06935092916778884,
+ "grad_norm": 0.8982051488791603,
+ "learning_rate": 9.970419062218186e-06,
+ "loss": 0.45786590576171876,
+ "num_input_tokens_seen": 44936832,
+ "step": 1030,
+ "train_runtime": 7899.821,
+ "train_tokens_per_second": 5688.336
+ },
+ {
+ "epoch": 0.06968758416374898,
+ "grad_norm": 0.9779591960140872,
+ "learning_rate": 9.970131178373276e-06,
+ "loss": 0.4594112396240234,
+ "num_input_tokens_seen": 45166720,
+ "step": 1035,
+ "train_runtime": 7937.6344,
+ "train_tokens_per_second": 5690.199
+ },
+ {
+ "epoch": 0.07002423915970914,
+ "grad_norm": 0.9403137936019832,
+ "learning_rate": 9.969841904647033e-06,
+ "loss": 0.48036918640136717,
+ "num_input_tokens_seen": 45386112,
+ "step": 1040,
+ "train_runtime": 7975.3873,
+ "train_tokens_per_second": 5690.772
+ },
+ {
+ "epoch": 0.07036089415566928,
+ "grad_norm": 0.9489869208797291,
+ "learning_rate": 9.969551241120349e-06,
+ "loss": 0.4358417510986328,
+ "num_input_tokens_seen": 45615968,
+ "step": 1045,
+ "train_runtime": 8008.3857,
+ "train_tokens_per_second": 5696.025
+ },
+ {
+ "epoch": 0.07069754915162942,
+ "grad_norm": 0.8565455428368292,
+ "learning_rate": 9.969259187874507e-06,
+ "loss": 0.40404400825500486,
+ "num_input_tokens_seen": 45846704,
+ "step": 1050,
+ "train_runtime": 8044.3345,
+ "train_tokens_per_second": 5699.254
+ },
+ {
+ "epoch": 0.07103420414758956,
+ "grad_norm": 0.9631616045121916,
+ "learning_rate": 9.968965744991178e-06,
+ "loss": 0.43367910385131836,
+ "num_input_tokens_seen": 46069384,
+ "step": 1055,
+ "train_runtime": 8075.3735,
+ "train_tokens_per_second": 5704.923
+ },
+ {
+ "epoch": 0.0713708591435497,
+ "grad_norm": 0.923891258768515,
+ "learning_rate": 9.968670912552422e-06,
+ "loss": 0.41605100631713865,
+ "num_input_tokens_seen": 46289280,
+ "step": 1060,
+ "train_runtime": 8110.2287,
+ "train_tokens_per_second": 5707.518
+ },
+ {
+ "epoch": 0.07170751413950983,
+ "grad_norm": 0.9199449654930428,
+ "learning_rate": 9.968374690640692e-06,
+ "loss": 0.4259674072265625,
+ "num_input_tokens_seen": 46513832,
+ "step": 1065,
+ "train_runtime": 8149.1273,
+ "train_tokens_per_second": 5707.83
+ },
+ {
+ "epoch": 0.07204416913546997,
+ "grad_norm": 0.9349037827416551,
+ "learning_rate": 9.968077079338822e-06,
+ "loss": 0.44624719619750974,
+ "num_input_tokens_seen": 46732528,
+ "step": 1070,
+ "train_runtime": 8183.763,
+ "train_tokens_per_second": 5710.396
+ },
+ {
+ "epoch": 0.07238082413143011,
+ "grad_norm": 0.9411351171686876,
+ "learning_rate": 9.967778078730038e-06,
+ "loss": 0.4426886558532715,
+ "num_input_tokens_seen": 46948464,
+ "step": 1075,
+ "train_runtime": 8220.2593,
+ "train_tokens_per_second": 5711.312
+ },
+ {
+ "epoch": 0.07271747912739025,
+ "grad_norm": 0.9433124357981822,
+ "learning_rate": 9.967477688897957e-06,
+ "loss": 0.44336910247802735,
+ "num_input_tokens_seen": 47171032,
+ "step": 1080,
+ "train_runtime": 8259.7389,
+ "train_tokens_per_second": 5710.959
+ },
+ {
+ "epoch": 0.07305413412335039,
+ "grad_norm": 0.8570101786695654,
+ "learning_rate": 9.967175909926578e-06,
+ "loss": 0.46260600090026854,
+ "num_input_tokens_seen": 47402968,
+ "step": 1085,
+ "train_runtime": 8293.6887,
+ "train_tokens_per_second": 5715.547
+ },
+ {
+ "epoch": 0.07339078911931053,
+ "grad_norm": 0.967581695257143,
+ "learning_rate": 9.966872741900296e-06,
+ "loss": 0.4760030746459961,
+ "num_input_tokens_seen": 47634040,
+ "step": 1090,
+ "train_runtime": 8342.26,
+ "train_tokens_per_second": 5709.968
+ },
+ {
+ "epoch": 0.07372744411527067,
+ "grad_norm": 1.3602823694752813,
+ "learning_rate": 9.96656818490389e-06,
+ "loss": 0.4523357391357422,
+ "num_input_tokens_seen": 47849424,
+ "step": 1095,
+ "train_runtime": 8378.4116,
+ "train_tokens_per_second": 5711.038
+ },
+ {
+ "epoch": 0.07406409911123081,
+ "grad_norm": 0.8798163617081339,
+ "learning_rate": 9.966262239022529e-06,
+ "loss": 0.48172898292541505,
+ "num_input_tokens_seen": 48074544,
+ "step": 1100,
+ "train_runtime": 8414.2774,
+ "train_tokens_per_second": 5713.449
+ },
+ {
+ "epoch": 0.07440075410719095,
+ "grad_norm": 0.8689729621566785,
+ "learning_rate": 9.965954904341768e-06,
+ "loss": 0.4594748497009277,
+ "num_input_tokens_seen": 48294808,
+ "step": 1105,
+ "train_runtime": 8453.2631,
+ "train_tokens_per_second": 5713.156
+ },
+ {
+ "epoch": 0.07473740910315109,
+ "grad_norm": 0.9298500312875317,
+ "learning_rate": 9.965646180947555e-06,
+ "loss": 0.44074625968933107,
+ "num_input_tokens_seen": 48509312,
+ "step": 1110,
+ "train_runtime": 8496.7093,
+ "train_tokens_per_second": 5709.188
+ },
+ {
+ "epoch": 0.07507406409911123,
+ "grad_norm": 0.9049700587066084,
+ "learning_rate": 9.96533606892622e-06,
+ "loss": 0.42275848388671877,
+ "num_input_tokens_seen": 48715504,
+ "step": 1115,
+ "train_runtime": 8533.5132,
+ "train_tokens_per_second": 5708.728
+ },
+ {
+ "epoch": 0.07541071909507137,
+ "grad_norm": 0.882919522521347,
+ "learning_rate": 9.965024568364487e-06,
+ "loss": 0.4448400497436523,
+ "num_input_tokens_seen": 48944392,
+ "step": 1120,
+ "train_runtime": 8571.3237,
+ "train_tokens_per_second": 5710.249
+ },
+ {
+ "epoch": 0.07574737409103151,
+ "grad_norm": 0.9374688066893563,
+ "learning_rate": 9.964711679349467e-06,
+ "loss": 0.41461782455444335,
+ "num_input_tokens_seen": 49155824,
+ "step": 1125,
+ "train_runtime": 8611.0892,
+ "train_tokens_per_second": 5708.433
+ },
+ {
+ "epoch": 0.07608402908699165,
+ "grad_norm": 0.7995128542211772,
+ "learning_rate": 9.964397401968656e-06,
+ "loss": 0.46135854721069336,
+ "num_input_tokens_seen": 49380912,
+ "step": 1130,
+ "train_runtime": 8650.8346,
+ "train_tokens_per_second": 5708.225
+ },
+ {
+ "epoch": 0.07642068408295179,
+ "grad_norm": 0.8607214899503258,
+ "learning_rate": 9.964081736309942e-06,
+ "loss": 0.41400885581970215,
+ "num_input_tokens_seen": 49617456,
+ "step": 1135,
+ "train_runtime": 8688.4749,
+ "train_tokens_per_second": 5710.721
+ },
+ {
+ "epoch": 0.07675733907891193,
+ "grad_norm": 0.8715957366665356,
+ "learning_rate": 9.963764682461599e-06,
+ "loss": 0.4346445083618164,
+ "num_input_tokens_seen": 49848584,
+ "step": 1140,
+ "train_runtime": 8722.1234,
+ "train_tokens_per_second": 5715.189
+ },
+ {
+ "epoch": 0.07709399407487207,
+ "grad_norm": 0.8561235152822287,
+ "learning_rate": 9.96344624051229e-06,
+ "loss": 0.4340538501739502,
+ "num_input_tokens_seen": 50065696,
+ "step": 1145,
+ "train_runtime": 8762.0281,
+ "train_tokens_per_second": 5713.939
+ },
+ {
+ "epoch": 0.0774306490708322,
+ "grad_norm": 0.929804620150381,
+ "learning_rate": 9.96312641055107e-06,
+ "loss": 0.4169576644897461,
+ "num_input_tokens_seen": 50283968,
+ "step": 1150,
+ "train_runtime": 8801.5092,
+ "train_tokens_per_second": 5713.107
+ },
+ {
+ "epoch": 0.07776730406679234,
+ "grad_norm": 0.752419655730966,
+ "learning_rate": 9.962805192667375e-06,
+ "loss": 0.41555070877075195,
+ "num_input_tokens_seen": 50500904,
+ "step": 1155,
+ "train_runtime": 8848.2982,
+ "train_tokens_per_second": 5707.414
+ },
+ {
+ "epoch": 0.07810395906275248,
+ "grad_norm": 1.0612801529163063,
+ "learning_rate": 9.962482586951033e-06,
+ "loss": 0.4202773094177246,
+ "num_input_tokens_seen": 50733408,
+ "step": 1160,
+ "train_runtime": 8889.0849,
+ "train_tokens_per_second": 5707.383
+ },
+ {
+ "epoch": 0.07844061405871264,
+ "grad_norm": 0.9990837210097588,
+ "learning_rate": 9.962158593492258e-06,
+ "loss": 0.442533016204834,
+ "num_input_tokens_seen": 50939416,
+ "step": 1165,
+ "train_runtime": 8930.0327,
+ "train_tokens_per_second": 5704.281
+ },
+ {
+ "epoch": 0.07877726905467278,
+ "grad_norm": 0.8491329656110541,
+ "learning_rate": 9.961833212381657e-06,
+ "loss": 0.4386699676513672,
+ "num_input_tokens_seen": 51160880,
+ "step": 1170,
+ "train_runtime": 8967.7862,
+ "train_tokens_per_second": 5704.962
+ },
+ {
+ "epoch": 0.07911392405063292,
+ "grad_norm": 0.8233317233024117,
+ "learning_rate": 9.961506443710221e-06,
+ "loss": 0.4202608108520508,
+ "num_input_tokens_seen": 51382192,
+ "step": 1175,
+ "train_runtime": 9009.4779,
+ "train_tokens_per_second": 5703.126
+ },
+ {
+ "epoch": 0.07945057904659306,
+ "grad_norm": 1.1740908151711091,
+ "learning_rate": 9.961178287569326e-06,
+ "loss": 0.4113266944885254,
+ "num_input_tokens_seen": 51610400,
+ "step": 1180,
+ "train_runtime": 9043.5038,
+ "train_tokens_per_second": 5706.903
+ },
+ {
+ "epoch": 0.0797872340425532,
+ "grad_norm": 0.9683671789990694,
+ "learning_rate": 9.960848744050747e-06,
+ "loss": 0.44871106147766116,
+ "num_input_tokens_seen": 51828672,
+ "step": 1185,
+ "train_runtime": 9085.7121,
+ "train_tokens_per_second": 5704.415
+ },
+ {
+ "epoch": 0.08012388903851334,
+ "grad_norm": 0.7384402440175095,
+ "learning_rate": 9.960517813246633e-06,
+ "loss": 0.44126176834106445,
+ "num_input_tokens_seen": 52038752,
+ "step": 1190,
+ "train_runtime": 9126.3974,
+ "train_tokens_per_second": 5702.004
+ },
+ {
+ "epoch": 0.08046054403447347,
+ "grad_norm": 0.9518446142368773,
+ "learning_rate": 9.96018549524953e-06,
+ "loss": 0.4168846607208252,
+ "num_input_tokens_seen": 52257272,
+ "step": 1195,
+ "train_runtime": 9165.705,
+ "train_tokens_per_second": 5701.391
+ },
+ {
+ "epoch": 0.08079719903043361,
+ "grad_norm": 0.9426346640072512,
+ "learning_rate": 9.959851790152371e-06,
+ "loss": 0.41556696891784667,
+ "num_input_tokens_seen": 52478032,
+ "step": 1200,
+ "train_runtime": 9204.8312,
+ "train_tokens_per_second": 5701.14
+ },
+ {
+ "epoch": 0.08113385402639375,
+ "grad_norm": 0.8462737285766736,
+ "learning_rate": 9.959516698048475e-06,
+ "loss": 0.3835282325744629,
+ "num_input_tokens_seen": 52684592,
+ "step": 1205,
+ "train_runtime": 9245.4602,
+ "train_tokens_per_second": 5698.428
+ },
+ {
+ "epoch": 0.08147050902235389,
+ "grad_norm": 0.8765077742279136,
+ "learning_rate": 9.959180219031551e-06,
+ "loss": 0.4106024742126465,
+ "num_input_tokens_seen": 52895048,
+ "step": 1210,
+ "train_runtime": 9284.6342,
+ "train_tokens_per_second": 5697.052
+ },
+ {
+ "epoch": 0.08180716401831403,
+ "grad_norm": 0.8799539327811132,
+ "learning_rate": 9.958842353195688e-06,
+ "loss": 0.4241660118103027,
+ "num_input_tokens_seen": 53122496,
+ "step": 1215,
+ "train_runtime": 9323.4499,
+ "train_tokens_per_second": 5697.73
+ },
+ {
+ "epoch": 0.08214381901427417,
+ "grad_norm": 0.8851017529748658,
+ "learning_rate": 9.958503100635377e-06,
+ "loss": 0.43586392402648927,
+ "num_input_tokens_seen": 53331344,
+ "step": 1220,
+ "train_runtime": 9359.4211,
+ "train_tokens_per_second": 5698.146
+ },
+ {
+ "epoch": 0.08248047401023431,
+ "grad_norm": 0.9611985454606802,
+ "learning_rate": 9.958162461445484e-06,
+ "loss": 0.4607269287109375,
+ "num_input_tokens_seen": 53560160,
+ "step": 1225,
+ "train_runtime": 9398.2675,
+ "train_tokens_per_second": 5698.94
+ },
+ {
+ "epoch": 0.08281712900619445,
+ "grad_norm": 0.9067497311826802,
+ "learning_rate": 9.957820435721271e-06,
+ "loss": 0.4349203586578369,
+ "num_input_tokens_seen": 53775560,
+ "step": 1230,
+ "train_runtime": 9435.4003,
+ "train_tokens_per_second": 5699.341
+ },
+ {
+ "epoch": 0.08315378400215459,
+ "grad_norm": 0.963108721688269,
+ "learning_rate": 9.95747702355838e-06,
+ "loss": 0.45670351982116697,
+ "num_input_tokens_seen": 54000808,
+ "step": 1235,
+ "train_runtime": 9470.0321,
+ "train_tokens_per_second": 5702.284
+ },
+ {
+ "epoch": 0.08349043899811473,
+ "grad_norm": 0.8513390607173822,
+ "learning_rate": 9.957132225052848e-06,
+ "loss": 0.4656820297241211,
+ "num_input_tokens_seen": 54218336,
+ "step": 1240,
+ "train_runtime": 9509.4298,
+ "train_tokens_per_second": 5701.534
+ },
+ {
+ "epoch": 0.08382709399407487,
+ "grad_norm": 0.8033639657062732,
+ "learning_rate": 9.956786040301098e-06,
+ "loss": 0.407243537902832,
+ "num_input_tokens_seen": 54435576,
+ "step": 1245,
+ "train_runtime": 9549.0218,
+ "train_tokens_per_second": 5700.644
+ },
+ {
+ "epoch": 0.08416374899003501,
+ "grad_norm": 0.9112338258739832,
+ "learning_rate": 9.956438469399936e-06,
+ "loss": 0.42780628204345705,
+ "num_input_tokens_seen": 54646448,
+ "step": 1250,
+ "train_runtime": 9584.5449,
+ "train_tokens_per_second": 5701.517
+ },
+ {
+ "epoch": 0.08450040398599515,
+ "grad_norm": 0.898524764750954,
+ "learning_rate": 9.956089512446562e-06,
+ "loss": 0.4281005859375,
+ "num_input_tokens_seen": 54849384,
+ "step": 1255,
+ "train_runtime": 9624.8869,
+ "train_tokens_per_second": 5698.704
+ },
+ {
+ "epoch": 0.08483705898195529,
+ "grad_norm": 0.9393978257221618,
+ "learning_rate": 9.955739169538559e-06,
+ "loss": 0.43431816101074217,
+ "num_input_tokens_seen": 55062944,
+ "step": 1260,
+ "train_runtime": 9664.3501,
+ "train_tokens_per_second": 5697.532
+ },
+ {
+ "epoch": 0.08517371397791543,
+ "grad_norm": 0.9086736133443781,
+ "learning_rate": 9.955387440773902e-06,
+ "loss": 0.4328322887420654,
+ "num_input_tokens_seen": 55267032,
+ "step": 1265,
+ "train_runtime": 9704.5196,
+ "train_tokens_per_second": 5694.979
+ },
+ {
+ "epoch": 0.08551036897387557,
+ "grad_norm": 1.1824810596928264,
+ "learning_rate": 9.955034326250947e-06,
+ "loss": 0.42726478576660154,
+ "num_input_tokens_seen": 55485176,
+ "step": 1270,
+ "train_runtime": 9740.2982,
+ "train_tokens_per_second": 5696.456
+ },
+ {
+ "epoch": 0.0858470239698357,
+ "grad_norm": 0.981019269531651,
+ "learning_rate": 9.954679826068444e-06,
+ "loss": 0.431905460357666,
+ "num_input_tokens_seen": 55697472,
+ "step": 1275,
+ "train_runtime": 9778.2861,
+ "train_tokens_per_second": 5696.036
+ },
+ {
+ "epoch": 0.08618367896579585,
+ "grad_norm": 0.9167947068742166,
+ "learning_rate": 9.954323940325526e-06,
+ "loss": 0.4367355823516846,
+ "num_input_tokens_seen": 55909208,
+ "step": 1280,
+ "train_runtime": 9815.8422,
+ "train_tokens_per_second": 5695.814
+ },
+ {
+ "epoch": 0.086520333961756,
+ "grad_norm": 0.8284962122347568,
+ "learning_rate": 9.953966669121715e-06,
+ "loss": 0.4562966346740723,
+ "num_input_tokens_seen": 56144336,
+ "step": 1285,
+ "train_runtime": 9855.4694,
+ "train_tokens_per_second": 5696.769
+ },
+ {
+ "epoch": 0.08685698895771614,
+ "grad_norm": 0.9908373153090002,
+ "learning_rate": 9.953608012556924e-06,
+ "loss": 0.4428553581237793,
+ "num_input_tokens_seen": 56370632,
+ "step": 1290,
+ "train_runtime": 9900.5066,
+ "train_tokens_per_second": 5693.712
+ },
+ {
+ "epoch": 0.08719364395367628,
+ "grad_norm": 0.8603920936193332,
+ "learning_rate": 9.953247970731449e-06,
+ "loss": 0.4518705368041992,
+ "num_input_tokens_seen": 56581824,
+ "step": 1295,
+ "train_runtime": 9941.9268,
+ "train_tokens_per_second": 5691.233
+ },
+ {
+ "epoch": 0.08753029894963642,
+ "grad_norm": 1.0065353834353872,
+ "learning_rate": 9.952886543745973e-06,
+ "loss": 0.45711684226989746,
+ "num_input_tokens_seen": 56797912,
+ "step": 1300,
+ "train_runtime": 9983.8473,
+ "train_tokens_per_second": 5688.98
+ },
+ {
+ "epoch": 0.08786695394559656,
+ "grad_norm": 0.9704989049669684,
+ "learning_rate": 9.952523731701566e-06,
+ "loss": 0.43997964859008787,
+ "num_input_tokens_seen": 57028400,
+ "step": 1305,
+ "train_runtime": 10023.0235,
+ "train_tokens_per_second": 5689.74
+ },
+ {
+ "epoch": 0.0882036089415567,
+ "grad_norm": 0.7746848672115061,
+ "learning_rate": 9.952159534699693e-06,
+ "loss": 0.39839365482330324,
+ "num_input_tokens_seen": 57239112,
+ "step": 1310,
+ "train_runtime": 10060.0427,
+ "train_tokens_per_second": 5689.748
+ },
+ {
+ "epoch": 0.08854026393751684,
+ "grad_norm": 0.8703667437003134,
+ "learning_rate": 9.951793952842197e-06,
+ "loss": 0.46162776947021483,
+ "num_input_tokens_seen": 57451840,
+ "step": 1315,
+ "train_runtime": 10090.9799,
+ "train_tokens_per_second": 5693.386
+ },
+ {
+ "epoch": 0.08887691893347698,
+ "grad_norm": 0.9970523155610451,
+ "learning_rate": 9.95142698623131e-06,
+ "loss": 0.4125502109527588,
+ "num_input_tokens_seen": 57649616,
+ "step": 1320,
+ "train_runtime": 10134.3943,
+ "train_tokens_per_second": 5688.511
+ },
+ {
+ "epoch": 0.08921357392943711,
+ "grad_norm": 0.8982754367798736,
+ "learning_rate": 9.951058634969657e-06,
+ "loss": 0.4594625473022461,
+ "num_input_tokens_seen": 57859504,
+ "step": 1325,
+ "train_runtime": 10173.3422,
+ "train_tokens_per_second": 5687.364
+ },
+ {
+ "epoch": 0.08955022892539725,
+ "grad_norm": 0.8441984007212928,
+ "learning_rate": 9.950688899160244e-06,
+ "loss": 0.42917823791503906,
+ "num_input_tokens_seen": 58092104,
+ "step": 1330,
+ "train_runtime": 10208.5539,
+ "train_tokens_per_second": 5690.532
+ },
+ {
+ "epoch": 0.0898868839213574,
+ "grad_norm": 0.8326739767212783,
+ "learning_rate": 9.950317778906469e-06,
+ "loss": 0.4345374584197998,
+ "num_input_tokens_seen": 58319256,
+ "step": 1335,
+ "train_runtime": 10249.3358,
+ "train_tokens_per_second": 5690.052
+ },
+ {
+ "epoch": 0.09022353891731753,
+ "grad_norm": 0.8294890352349942,
+ "learning_rate": 9.949945274312109e-06,
+ "loss": 0.40816512107849123,
+ "num_input_tokens_seen": 58540424,
+ "step": 1340,
+ "train_runtime": 10287.6498,
+ "train_tokens_per_second": 5690.359
+ },
+ {
+ "epoch": 0.09056019391327767,
+ "grad_norm": 0.968521031703479,
+ "learning_rate": 9.94957138548134e-06,
+ "loss": 0.4445356369018555,
+ "num_input_tokens_seen": 58736800,
+ "step": 1345,
+ "train_runtime": 10326.024,
+ "train_tokens_per_second": 5688.23
+ },
+ {
+ "epoch": 0.09089684890923781,
+ "grad_norm": 1.007040150938243,
+ "learning_rate": 9.949196112518715e-06,
+ "loss": 0.41266183853149413,
+ "num_input_tokens_seen": 58963840,
+ "step": 1350,
+ "train_runtime": 10361.3193,
+ "train_tokens_per_second": 5690.766
+ },
+ {
+ "epoch": 0.09123350390519795,
+ "grad_norm": 0.8518178617284553,
+ "learning_rate": 9.94881945552918e-06,
+ "loss": 0.43543500900268556,
+ "num_input_tokens_seen": 59186072,
+ "step": 1355,
+ "train_runtime": 10406.6964,
+ "train_tokens_per_second": 5687.307
+ },
+ {
+ "epoch": 0.09157015890115809,
+ "grad_norm": 0.9351761092832128,
+ "learning_rate": 9.948441414618064e-06,
+ "loss": 0.45530238151550295,
+ "num_input_tokens_seen": 59396096,
+ "step": 1360,
+ "train_runtime": 10448.5625,
+ "train_tokens_per_second": 5684.619
+ },
+ {
+ "epoch": 0.09190681389711823,
+ "grad_norm": 0.976556409327906,
+ "learning_rate": 9.948061989891085e-06,
+ "loss": 0.41725835800170896,
+ "num_input_tokens_seen": 59621960,
+ "step": 1365,
+ "train_runtime": 10484.4296,
+ "train_tokens_per_second": 5686.715
+ },
+ {
+ "epoch": 0.09224346889307837,
+ "grad_norm": 0.9413072947091584,
+ "learning_rate": 9.94768118145435e-06,
+ "loss": 0.4238771438598633,
+ "num_input_tokens_seen": 59834688,
+ "step": 1370,
+ "train_runtime": 10521.7898,
+ "train_tokens_per_second": 5686.74
+ },
+ {
+ "epoch": 0.09258012388903851,
+ "grad_norm": 0.8072753824687519,
+ "learning_rate": 9.947298989414349e-06,
+ "loss": 0.4197520732879639,
+ "num_input_tokens_seen": 60061512,
+ "step": 1375,
+ "train_runtime": 10562.6863,
+ "train_tokens_per_second": 5686.197
+ },
+ {
+ "epoch": 0.09291677888499865,
+ "grad_norm": 0.8243851964827824,
+ "learning_rate": 9.946915413877963e-06,
+ "loss": 0.4294243812561035,
+ "num_input_tokens_seen": 60289152,
+ "step": 1380,
+ "train_runtime": 10608.7449,
+ "train_tokens_per_second": 5682.967
+ },
+ {
+ "epoch": 0.09325343388095879,
+ "grad_norm": 0.8639433441786879,
+ "learning_rate": 9.946530454952454e-06,
+ "loss": 0.44629898071289065,
+ "num_input_tokens_seen": 60516576,
+ "step": 1385,
+ "train_runtime": 10649.4064,
+ "train_tokens_per_second": 5682.624
+ },
+ {
+ "epoch": 0.09359008887691893,
+ "grad_norm": 0.9381874141159465,
+ "learning_rate": 9.946144112745477e-06,
+ "loss": 0.4549723625183105,
+ "num_input_tokens_seen": 60744096,
+ "step": 1390,
+ "train_runtime": 10690.6621,
+ "train_tokens_per_second": 5681.977
+ },
+ {
+ "epoch": 0.09392674387287907,
+ "grad_norm": 0.9598970905259178,
+ "learning_rate": 9.94575638736507e-06,
+ "loss": 0.4265571594238281,
+ "num_input_tokens_seen": 60970464,
+ "step": 1395,
+ "train_runtime": 10730.6636,
+ "train_tokens_per_second": 5681.891
+ },
+ {
+ "epoch": 0.09426339886883921,
+ "grad_norm": 0.8882564184072146,
+ "learning_rate": 9.945367278919662e-06,
+ "loss": 0.469175910949707,
+ "num_input_tokens_seen": 61195536,
+ "step": 1400,
+ "train_runtime": 10771.6796,
+ "train_tokens_per_second": 5681.151
+ },
+ {
+ "epoch": 0.09460005386479936,
+ "grad_norm": 0.8345645071811238,
+ "learning_rate": 9.944976787518064e-06,
+ "loss": 0.43855986595153806,
+ "num_input_tokens_seen": 61415400,
+ "step": 1405,
+ "train_runtime": 10804.9607,
+ "train_tokens_per_second": 5684.0
+ },
+ {
+ "epoch": 0.0949367088607595,
+ "grad_norm": 1.0139184695279284,
+ "learning_rate": 9.944584913269474e-06,
+ "loss": 0.43834495544433594,
+ "num_input_tokens_seen": 61645440,
+ "step": 1410,
+ "train_runtime": 10846.1692,
+ "train_tokens_per_second": 5683.614
+ },
+ {
+ "epoch": 0.09527336385671964,
+ "grad_norm": 0.8171057323619708,
+ "learning_rate": 9.944191656283481e-06,
+ "loss": 0.41381168365478516,
+ "num_input_tokens_seen": 61868952,
+ "step": 1415,
+ "train_runtime": 10885.6156,
+ "train_tokens_per_second": 5683.551
+ },
+ {
+ "epoch": 0.09561001885267978,
+ "grad_norm": 0.8671727475302144,
+ "learning_rate": 9.943797016670059e-06,
+ "loss": 0.41778225898742677,
+ "num_input_tokens_seen": 62081744,
+ "step": 1420,
+ "train_runtime": 10924.1913,
+ "train_tokens_per_second": 5682.96
+ },
+ {
+ "epoch": 0.09594667384863992,
+ "grad_norm": 0.9139643587899275,
+ "learning_rate": 9.943400994539565e-06,
+ "loss": 0.41177783012390134,
+ "num_input_tokens_seen": 62299552,
+ "step": 1425,
+ "train_runtime": 10963.2034,
+ "train_tokens_per_second": 5682.605
+ },
+ {
+ "epoch": 0.09628332884460006,
+ "grad_norm": 0.8942790678032806,
+ "learning_rate": 9.943003590002744e-06,
+ "loss": 0.4324612617492676,
+ "num_input_tokens_seen": 62524424,
+ "step": 1430,
+ "train_runtime": 11003.5976,
+ "train_tokens_per_second": 5682.18
+ },
+ {
+ "epoch": 0.0966199838405602,
+ "grad_norm": 0.9134672447344226,
+ "learning_rate": 9.942604803170733e-06,
+ "loss": 0.4207320690155029,
+ "num_input_tokens_seen": 62736736,
+ "step": 1435,
+ "train_runtime": 11041.0071,
+ "train_tokens_per_second": 5682.157
+ },
+ {
+ "epoch": 0.09695663883652034,
+ "grad_norm": 1.078284335277287,
+ "learning_rate": 9.94220463415505e-06,
+ "loss": 0.4587702751159668,
+ "num_input_tokens_seen": 62950552,
+ "step": 1440,
+ "train_runtime": 11082.5134,
+ "train_tokens_per_second": 5680.169
+ },
+ {
+ "epoch": 0.09729329383248048,
+ "grad_norm": 0.8211105532018887,
+ "learning_rate": 9.9418030830676e-06,
+ "loss": 0.4369199752807617,
+ "num_input_tokens_seen": 63173304,
+ "step": 1445,
+ "train_runtime": 11119.7396,
+ "train_tokens_per_second": 5681.186
+ },
+ {
+ "epoch": 0.09762994882844062,
+ "grad_norm": 0.9975227482376734,
+ "learning_rate": 9.941400150020676e-06,
+ "loss": 0.45250940322875977,
+ "num_input_tokens_seen": 63376456,
+ "step": 1450,
+ "train_runtime": 11153.3434,
+ "train_tokens_per_second": 5682.283
+ },
+ {
+ "epoch": 0.09796660382440076,
+ "grad_norm": 0.875023853319168,
+ "learning_rate": 9.940995835126957e-06,
+ "loss": 0.4510367393493652,
+ "num_input_tokens_seen": 63598704,
+ "step": 1455,
+ "train_runtime": 11190.4484,
+ "train_tokens_per_second": 5683.303
+ },
+ {
+ "epoch": 0.0983032588203609,
+ "grad_norm": 0.7795257917882069,
+ "learning_rate": 9.940590138499508e-06,
+ "loss": 0.4070897102355957,
+ "num_input_tokens_seen": 63824984,
+ "step": 1460,
+ "train_runtime": 11228.0323,
+ "train_tokens_per_second": 5684.432
+ },
+ {
+ "epoch": 0.09863991381632103,
+ "grad_norm": 0.9802634486624268,
+ "learning_rate": 9.940183060251783e-06,
+ "loss": 0.424533748626709,
+ "num_input_tokens_seen": 64020440,
+ "step": 1465,
+ "train_runtime": 11269.0538,
+ "train_tokens_per_second": 5681.084
+ },
+ {
+ "epoch": 0.09897656881228117,
+ "grad_norm": 0.8129225031494536,
+ "learning_rate": 9.939774600497616e-06,
+ "loss": 0.3772727489471436,
+ "num_input_tokens_seen": 64237912,
+ "step": 1470,
+ "train_runtime": 11305.1846,
+ "train_tokens_per_second": 5682.164
+ },
+ {
+ "epoch": 0.09931322380824131,
+ "grad_norm": 0.876095085095231,
+ "learning_rate": 9.939364759351236e-06,
+ "loss": 0.43614888191223145,
+ "num_input_tokens_seen": 64452808,
+ "step": 1475,
+ "train_runtime": 11342.3271,
+ "train_tokens_per_second": 5682.503
+ },
+ {
+ "epoch": 0.09964987880420145,
+ "grad_norm": 0.8249502349163025,
+ "learning_rate": 9.938953536927249e-06,
+ "loss": 0.4052250862121582,
+ "num_input_tokens_seen": 64679536,
+ "step": 1480,
+ "train_runtime": 11380.8832,
+ "train_tokens_per_second": 5683.174
+ },
+ {
+ "epoch": 0.09998653380016159,
+ "grad_norm": 0.8004037055954487,
+ "learning_rate": 9.938540933340657e-06,
+ "loss": 0.43981304168701174,
+ "num_input_tokens_seen": 64907672,
+ "step": 1485,
+ "train_runtime": 11414.3496,
+ "train_tokens_per_second": 5686.498
+ },
+ {
+ "epoch": 0.10032318879612173,
+ "grad_norm": 0.7556539983829699,
+ "learning_rate": 9.938126948706839e-06,
+ "loss": 0.4235081195831299,
+ "num_input_tokens_seen": 65134656,
+ "step": 1490,
+ "train_runtime": 11451.1493,
+ "train_tokens_per_second": 5688.045
+ },
+ {
+ "epoch": 0.10065984379208187,
+ "grad_norm": 0.7974202081471553,
+ "learning_rate": 9.937711583141567e-06,
+ "loss": 0.41340045928955077,
+ "num_input_tokens_seen": 65353344,
+ "step": 1495,
+ "train_runtime": 11488.8753,
+ "train_tokens_per_second": 5688.402
+ },
+ {
+ "epoch": 0.10099649878804201,
+ "grad_norm": 0.9113055103223531,
+ "learning_rate": 9.937294836760995e-06,
+ "loss": 0.43211612701416013,
+ "num_input_tokens_seen": 65570904,
+ "step": 1500,
+ "train_runtime": 11525.754,
+ "train_tokens_per_second": 5689.077
+ },
+ {
+ "epoch": 0.10133315378400215,
+ "grad_norm": 0.9362482868425025,
+ "learning_rate": 9.936876709681668e-06,
+ "loss": 0.4765140533447266,
+ "num_input_tokens_seen": 65784200,
+ "step": 1505,
+ "train_runtime": 11567.0824,
+ "train_tokens_per_second": 5687.19
+ },
+ {
+ "epoch": 0.10166980877996229,
+ "grad_norm": 0.7345279591609306,
+ "learning_rate": 9.93645720202051e-06,
+ "loss": 0.3693789005279541,
+ "num_input_tokens_seen": 66002240,
+ "step": 1510,
+ "train_runtime": 11603.3588,
+ "train_tokens_per_second": 5688.201
+ },
+ {
+ "epoch": 0.10200646377592243,
+ "grad_norm": 0.8440448421494396,
+ "learning_rate": 9.936036313894836e-06,
+ "loss": 0.46529722213745117,
+ "num_input_tokens_seen": 66217160,
+ "step": 1515,
+ "train_runtime": 11637.8846,
+ "train_tokens_per_second": 5689.793
+ },
+ {
+ "epoch": 0.10234311877188257,
+ "grad_norm": 0.9288537021329782,
+ "learning_rate": 9.935614045422349e-06,
+ "loss": 0.4474299907684326,
+ "num_input_tokens_seen": 66437216,
+ "step": 1520,
+ "train_runtime": 11674.3858,
+ "train_tokens_per_second": 5690.853
+ },
+ {
+ "epoch": 0.10267977376784271,
+ "grad_norm": 0.7816250239377797,
+ "learning_rate": 9.935190396721132e-06,
+ "loss": 0.4093039035797119,
+ "num_input_tokens_seen": 66659632,
+ "step": 1525,
+ "train_runtime": 11714.7321,
+ "train_tokens_per_second": 5690.24
+ },
+ {
+ "epoch": 0.10301642876380286,
+ "grad_norm": 0.868350723790852,
+ "learning_rate": 9.934765367909658e-06,
+ "loss": 0.442030143737793,
+ "num_input_tokens_seen": 66874200,
+ "step": 1530,
+ "train_runtime": 11751.0775,
+ "train_tokens_per_second": 5690.899
+ },
+ {
+ "epoch": 0.103353083759763,
+ "grad_norm": 0.9654553160320544,
+ "learning_rate": 9.934338959106783e-06,
+ "loss": 0.4125958442687988,
+ "num_input_tokens_seen": 67096584,
+ "step": 1535,
+ "train_runtime": 11787.6988,
+ "train_tokens_per_second": 5692.085
+ },
+ {
+ "epoch": 0.10368973875572314,
+ "grad_norm": 1.0001250180683936,
+ "learning_rate": 9.933911170431753e-06,
+ "loss": 0.45245895385742185,
+ "num_input_tokens_seen": 67300344,
+ "step": 1540,
+ "train_runtime": 11825.444,
+ "train_tokens_per_second": 5691.147
+ },
+ {
+ "epoch": 0.10402639375168328,
+ "grad_norm": 0.9480950052674862,
+ "learning_rate": 9.933482002004198e-06,
+ "loss": 0.4369757652282715,
+ "num_input_tokens_seen": 67528208,
+ "step": 1545,
+ "train_runtime": 11868.3124,
+ "train_tokens_per_second": 5689.79
+ },
+ {
+ "epoch": 0.10436304874764342,
+ "grad_norm": 1.049191818835591,
+ "learning_rate": 9.933051453944134e-06,
+ "loss": 0.47658519744873046,
+ "num_input_tokens_seen": 67752368,
+ "step": 1550,
+ "train_runtime": 11899.3433,
+ "train_tokens_per_second": 5693.791
+ },
+ {
+ "epoch": 0.10469970374360356,
+ "grad_norm": 0.8519840991796084,
+ "learning_rate": 9.932619526371958e-06,
+ "loss": 0.4100327014923096,
+ "num_input_tokens_seen": 67972272,
+ "step": 1555,
+ "train_runtime": 11933.0537,
+ "train_tokens_per_second": 5696.134
+ },
+ {
+ "epoch": 0.1050363587395637,
+ "grad_norm": 0.8587262694316211,
+ "learning_rate": 9.932186219408463e-06,
+ "loss": 0.40094866752624514,
+ "num_input_tokens_seen": 68182312,
+ "step": 1560,
+ "train_runtime": 11973.5123,
+ "train_tokens_per_second": 5694.429
+ },
+ {
+ "epoch": 0.10537301373552384,
+ "grad_norm": 0.9397299864907294,
+ "learning_rate": 9.931751533174819e-06,
+ "loss": 0.44454145431518555,
+ "num_input_tokens_seen": 68389480,
+ "step": 1565,
+ "train_runtime": 12014.0801,
+ "train_tokens_per_second": 5692.444
+ },
+ {
+ "epoch": 0.10570966873148398,
+ "grad_norm": 0.9745376383539546,
+ "learning_rate": 9.931315467792584e-06,
+ "loss": 0.45273237228393554,
+ "num_input_tokens_seen": 68625472,
+ "step": 1570,
+ "train_runtime": 12045.9213,
+ "train_tokens_per_second": 5696.988
+ },
+ {
+ "epoch": 0.10604632372744412,
+ "grad_norm": 0.796989168878711,
+ "learning_rate": 9.930878023383705e-06,
+ "loss": 0.4475082874298096,
+ "num_input_tokens_seen": 68842256,
+ "step": 1575,
+ "train_runtime": 12081.8379,
+ "train_tokens_per_second": 5697.995
+ },
+ {
+ "epoch": 0.10638297872340426,
+ "grad_norm": 0.7777899346165759,
+ "learning_rate": 9.93043920007051e-06,
+ "loss": 0.39358346462249755,
+ "num_input_tokens_seen": 69063336,
+ "step": 1580,
+ "train_runtime": 12120.9279,
+ "train_tokens_per_second": 5697.859
+ },
+ {
+ "epoch": 0.1067196337193644,
+ "grad_norm": 0.856296656865995,
+ "learning_rate": 9.929998997975712e-06,
+ "loss": 0.40912957191467286,
+ "num_input_tokens_seen": 69297392,
+ "step": 1585,
+ "train_runtime": 12157.5274,
+ "train_tokens_per_second": 5699.958
+ },
+ {
+ "epoch": 0.10705628871532454,
+ "grad_norm": 0.8406798760366443,
+ "learning_rate": 9.929557417222417e-06,
+ "loss": 0.4154947280883789,
+ "num_input_tokens_seen": 69502272,
+ "step": 1590,
+ "train_runtime": 12195.0467,
+ "train_tokens_per_second": 5699.221
+ },
+ {
+ "epoch": 0.10739294371128467,
+ "grad_norm": 0.8506547801105618,
+ "learning_rate": 9.929114457934109e-06,
+ "loss": 0.433414888381958,
+ "num_input_tokens_seen": 69718712,
+ "step": 1595,
+ "train_runtime": 12235.8524,
+ "train_tokens_per_second": 5697.904
+ },
+ {
+ "epoch": 0.10772959870724481,
+ "grad_norm": 0.8306444786085215,
+ "learning_rate": 9.92867012023466e-06,
+ "loss": 0.429976749420166,
+ "num_input_tokens_seen": 69942920,
+ "step": 1600,
+ "train_runtime": 12275.1053,
+ "train_tokens_per_second": 5697.949
+ },
+ {
+ "epoch": 0.10806625370320495,
+ "grad_norm": 0.8083632874129185,
+ "learning_rate": 9.92822440424833e-06,
+ "loss": 0.465771484375,
+ "num_input_tokens_seen": 70167408,
+ "step": 1605,
+ "train_runtime": 12311.856,
+ "train_tokens_per_second": 5699.174
+ },
+ {
+ "epoch": 0.1084029086991651,
+ "grad_norm": 0.8623025104703945,
+ "learning_rate": 9.927777310099759e-06,
+ "loss": 0.4795114040374756,
+ "num_input_tokens_seen": 70398096,
+ "step": 1610,
+ "train_runtime": 12355.4012,
+ "train_tokens_per_second": 5697.759
+ },
+ {
+ "epoch": 0.10873956369512523,
+ "grad_norm": 0.8734280942811361,
+ "learning_rate": 9.927328837913976e-06,
+ "loss": 0.42298336029052735,
+ "num_input_tokens_seen": 70621368,
+ "step": 1615,
+ "train_runtime": 12392.0717,
+ "train_tokens_per_second": 5698.915
+ },
+ {
+ "epoch": 0.10907621869108537,
+ "grad_norm": 0.8755533500883431,
+ "learning_rate": 9.926878987816397e-06,
+ "loss": 0.4128897666931152,
+ "num_input_tokens_seen": 70850496,
+ "step": 1620,
+ "train_runtime": 12432.0003,
+ "train_tokens_per_second": 5699.042
+ },
+ {
+ "epoch": 0.10941287368704551,
+ "grad_norm": 0.7849562344468639,
+ "learning_rate": 9.92642775993282e-06,
+ "loss": 0.40384368896484374,
+ "num_input_tokens_seen": 71078912,
+ "step": 1625,
+ "train_runtime": 12472.6839,
+ "train_tokens_per_second": 5698.766
+ },
+ {
+ "epoch": 0.10974952868300565,
+ "grad_norm": 0.8314524460612794,
+ "learning_rate": 9.925975154389428e-06,
+ "loss": 0.39646482467651367,
+ "num_input_tokens_seen": 71291944,
+ "step": 1630,
+ "train_runtime": 12512.7857,
+ "train_tokens_per_second": 5697.528
+ },
+ {
+ "epoch": 0.11008618367896579,
+ "grad_norm": 0.8397438217714046,
+ "learning_rate": 9.925521171312792e-06,
+ "loss": 0.41344375610351564,
+ "num_input_tokens_seen": 71513056,
+ "step": 1635,
+ "train_runtime": 12552.1031,
+ "train_tokens_per_second": 5697.297
+ },
+ {
+ "epoch": 0.11042283867492593,
+ "grad_norm": 0.8717786503893655,
+ "learning_rate": 9.925065810829868e-06,
+ "loss": 0.43964757919311526,
+ "num_input_tokens_seen": 71712776,
+ "step": 1640,
+ "train_runtime": 12592.8232,
+ "train_tokens_per_second": 5694.734
+ },
+ {
+ "epoch": 0.11075949367088607,
+ "grad_norm": 0.8466265387114118,
+ "learning_rate": 9.924609073067994e-06,
+ "loss": 0.45009913444519045,
+ "num_input_tokens_seen": 71918888,
+ "step": 1645,
+ "train_runtime": 12637.095,
+ "train_tokens_per_second": 5691.093
+ },
+ {
+ "epoch": 0.11109614866684622,
+ "grad_norm": 0.8202088060176734,
+ "learning_rate": 9.924150958154897e-06,
+ "loss": 0.43289871215820314,
+ "num_input_tokens_seen": 72144880,
+ "step": 1650,
+ "train_runtime": 12679.4137,
+ "train_tokens_per_second": 5689.922
+ },
+ {
+ "epoch": 0.11143280366280636,
+ "grad_norm": 0.8241473767552461,
+ "learning_rate": 9.923691466218686e-06,
+ "loss": 0.4033914566040039,
+ "num_input_tokens_seen": 72368424,
+ "step": 1655,
+ "train_runtime": 12722.761,
+ "train_tokens_per_second": 5688.107
+ },
+ {
+ "epoch": 0.1117694586587665,
+ "grad_norm": 0.9179601239629808,
+ "learning_rate": 9.923230597387856e-06,
+ "loss": 0.4340330123901367,
+ "num_input_tokens_seen": 72604440,
+ "step": 1660,
+ "train_runtime": 12760.3543,
+ "train_tokens_per_second": 5689.845
+ },
+ {
+ "epoch": 0.11210611365472664,
+ "grad_norm": 0.9667829858136964,
+ "learning_rate": 9.922768351791289e-06,
+ "loss": 0.45328373908996583,
+ "num_input_tokens_seen": 72822440,
+ "step": 1665,
+ "train_runtime": 12804.6293,
+ "train_tokens_per_second": 5687.196
+ },
+ {
+ "epoch": 0.11244276865068678,
+ "grad_norm": 0.9239911563296863,
+ "learning_rate": 9.92230472955825e-06,
+ "loss": 0.41826558113098145,
+ "num_input_tokens_seen": 73042800,
+ "step": 1670,
+ "train_runtime": 12838.5402,
+ "train_tokens_per_second": 5689.338
+ },
+ {
+ "epoch": 0.11277942364664692,
+ "grad_norm": 0.8636254640310012,
+ "learning_rate": 9.92183973081839e-06,
+ "loss": 0.3946269750595093,
+ "num_input_tokens_seen": 73268056,
+ "step": 1675,
+ "train_runtime": 12882.6482,
+ "train_tokens_per_second": 5687.344
+ },
+ {
+ "epoch": 0.11311607864260706,
+ "grad_norm": 0.7861261803530913,
+ "learning_rate": 9.921373355701743e-06,
+ "loss": 0.46369400024414065,
+ "num_input_tokens_seen": 73499832,
+ "step": 1680,
+ "train_runtime": 12924.32,
+ "train_tokens_per_second": 5686.94
+ },
+ {
+ "epoch": 0.1134527336385672,
+ "grad_norm": 0.8384955151965564,
+ "learning_rate": 9.920905604338732e-06,
+ "loss": 0.45543785095214845,
+ "num_input_tokens_seen": 73720264,
+ "step": 1685,
+ "train_runtime": 12963.3961,
+ "train_tokens_per_second": 5686.802
+ },
+ {
+ "epoch": 0.11378938863452734,
+ "grad_norm": 1.0833650625320947,
+ "learning_rate": 9.920436476860158e-06,
+ "loss": 0.38962454795837403,
+ "num_input_tokens_seen": 73931824,
+ "step": 1690,
+ "train_runtime": 13004.1451,
+ "train_tokens_per_second": 5685.251
+ },
+ {
+ "epoch": 0.11412604363048748,
+ "grad_norm": 0.8250025055357804,
+ "learning_rate": 9.919965973397214e-06,
+ "loss": 0.40491409301757814,
+ "num_input_tokens_seen": 74150736,
+ "step": 1695,
+ "train_runtime": 13037.0352,
+ "train_tokens_per_second": 5687.699
+ },
+ {
+ "epoch": 0.11446269862644762,
+ "grad_norm": 0.841239081826391,
+ "learning_rate": 9.919494094081475e-06,
+ "loss": 0.45804147720336913,
+ "num_input_tokens_seen": 74352136,
+ "step": 1700,
+ "train_runtime": 13073.6778,
+ "train_tokens_per_second": 5687.163
+ },
+ {
+ "epoch": 0.11479935362240776,
+ "grad_norm": 0.9629476764947537,
+ "learning_rate": 9.919020839044899e-06,
+ "loss": 0.4247718334197998,
+ "num_input_tokens_seen": 74588328,
+ "step": 1705,
+ "train_runtime": 13108.3031,
+ "train_tokens_per_second": 5690.159
+ },
+ {
+ "epoch": 0.1151360086183679,
+ "grad_norm": 0.8166202703347624,
+ "learning_rate": 9.918546208419832e-06,
+ "loss": 0.4040073394775391,
+ "num_input_tokens_seen": 74810232,
+ "step": 1710,
+ "train_runtime": 13152.6429,
+ "train_tokens_per_second": 5687.848
+ },
+ {
+ "epoch": 0.11547266361432804,
+ "grad_norm": 0.9143911353222378,
+ "learning_rate": 9.918070202339e-06,
+ "loss": 0.41643967628479006,
+ "num_input_tokens_seen": 75038248,
+ "step": 1715,
+ "train_runtime": 13188.009,
+ "train_tokens_per_second": 5689.885
+ },
+ {
+ "epoch": 0.11580931861028818,
+ "grad_norm": 0.854163675009985,
+ "learning_rate": 9.91759282093552e-06,
+ "loss": 0.4160469532012939,
+ "num_input_tokens_seen": 75259984,
+ "step": 1720,
+ "train_runtime": 13221.5917,
+ "train_tokens_per_second": 5692.203
+ },
+ {
+ "epoch": 0.11614597360624832,
+ "grad_norm": 0.9111231961657891,
+ "learning_rate": 9.917114064342888e-06,
+ "loss": 0.38855648040771484,
+ "num_input_tokens_seen": 75468520,
+ "step": 1725,
+ "train_runtime": 13264.901,
+ "train_tokens_per_second": 5689.339
+ },
+ {
+ "epoch": 0.11648262860220845,
+ "grad_norm": 0.8667773958295081,
+ "learning_rate": 9.916633932694988e-06,
+ "loss": 0.3962873458862305,
+ "num_input_tokens_seen": 75690392,
+ "step": 1730,
+ "train_runtime": 13301.6754,
+ "train_tokens_per_second": 5690.29
+ },
+ {
+ "epoch": 0.1168192835981686,
+ "grad_norm": 0.8572351568987933,
+ "learning_rate": 9.916152426126086e-06,
+ "loss": 0.3940844774246216,
+ "num_input_tokens_seen": 75908632,
+ "step": 1735,
+ "train_runtime": 13340.3356,
+ "train_tokens_per_second": 5690.159
+ },
+ {
+ "epoch": 0.11715593859412873,
+ "grad_norm": 0.8920380227031728,
+ "learning_rate": 9.915669544770837e-06,
+ "loss": 0.41963586807250974,
+ "num_input_tokens_seen": 76143760,
+ "step": 1740,
+ "train_runtime": 13373.7492,
+ "train_tokens_per_second": 5693.524
+ },
+ {
+ "epoch": 0.11749259359008887,
+ "grad_norm": 0.8151573693579499,
+ "learning_rate": 9.915185288764272e-06,
+ "loss": 0.4006998062133789,
+ "num_input_tokens_seen": 76381464,
+ "step": 1745,
+ "train_runtime": 13411.6709,
+ "train_tokens_per_second": 5695.149
+ },
+ {
+ "epoch": 0.11782924858604901,
+ "grad_norm": 0.7823689731999574,
+ "learning_rate": 9.914699658241815e-06,
+ "loss": 0.39071149826049806,
+ "num_input_tokens_seen": 76607736,
+ "step": 1750,
+ "train_runtime": 13455.8214,
+ "train_tokens_per_second": 5693.278
+ },
+ {
+ "epoch": 0.11816590358200915,
+ "grad_norm": 0.9650032542802038,
+ "learning_rate": 9.91421265333927e-06,
+ "loss": 0.44945535659790037,
+ "num_input_tokens_seen": 76836152,
+ "step": 1755,
+ "train_runtime": 13495.7655,
+ "train_tokens_per_second": 5693.353
+ },
+ {
+ "epoch": 0.11850255857796929,
+ "grad_norm": 0.8293434376681832,
+ "learning_rate": 9.913724274192829e-06,
+ "loss": 0.4147852897644043,
+ "num_input_tokens_seen": 77069784,
+ "step": 1760,
+ "train_runtime": 13537.8959,
+ "train_tokens_per_second": 5692.892
+ },
+ {
+ "epoch": 0.11883921357392943,
+ "grad_norm": 0.9258862128909312,
+ "learning_rate": 9.91323452093906e-06,
+ "loss": 0.41230435371398927,
+ "num_input_tokens_seen": 77288624,
+ "step": 1765,
+ "train_runtime": 13574.917,
+ "train_tokens_per_second": 5693.488
+ },
+ {
+ "epoch": 0.11917586856988957,
+ "grad_norm": 0.848930776896082,
+ "learning_rate": 9.912743393714927e-06,
+ "loss": 0.3945260286331177,
+ "num_input_tokens_seen": 77519384,
+ "step": 1770,
+ "train_runtime": 13605.814,
+ "train_tokens_per_second": 5697.519
+ },
+ {
+ "epoch": 0.11951252356584972,
+ "grad_norm": 0.8278888004348279,
+ "learning_rate": 9.912250892657769e-06,
+ "loss": 0.40961732864379885,
+ "num_input_tokens_seen": 77735336,
+ "step": 1775,
+ "train_runtime": 13644.1096,
+ "train_tokens_per_second": 5697.355
+ },
+ {
+ "epoch": 0.11984917856180986,
+ "grad_norm": 0.9501953857186206,
+ "learning_rate": 9.911757017905312e-06,
+ "loss": 0.41979494094848635,
+ "num_input_tokens_seen": 77943944,
+ "step": 1780,
+ "train_runtime": 13675.7254,
+ "train_tokens_per_second": 5699.438
+ },
+ {
+ "epoch": 0.12018583355777,
+ "grad_norm": 0.8175400797161728,
+ "learning_rate": 9.911261769595668e-06,
+ "loss": 0.4165688991546631,
+ "num_input_tokens_seen": 78176688,
+ "step": 1785,
+ "train_runtime": 13720.1102,
+ "train_tokens_per_second": 5697.964
+ },
+ {
+ "epoch": 0.12052248855373014,
+ "grad_norm": 0.9695481492778111,
+ "learning_rate": 9.910765147867332e-06,
+ "loss": 0.4288275718688965,
+ "num_input_tokens_seen": 78398920,
+ "step": 1790,
+ "train_runtime": 13754.606,
+ "train_tokens_per_second": 5699.83
+ },
+ {
+ "epoch": 0.12085914354969028,
+ "grad_norm": 0.9043853969674001,
+ "learning_rate": 9.91026715285918e-06,
+ "loss": 0.42678542137145997,
+ "num_input_tokens_seen": 78617400,
+ "step": 1795,
+ "train_runtime": 13795.7712,
+ "train_tokens_per_second": 5698.659
+ },
+ {
+ "epoch": 0.12119579854565042,
+ "grad_norm": 0.9643745725188065,
+ "learning_rate": 9.909767784710476e-06,
+ "loss": 0.4547886371612549,
+ "num_input_tokens_seen": 78858720,
+ "step": 1800,
+ "train_runtime": 13832.1459,
+ "train_tokens_per_second": 5701.12
+ },
+ {
+ "epoch": 0.12153245354161056,
+ "grad_norm": 0.8832616111060988,
+ "learning_rate": 9.909267043560868e-06,
+ "loss": 0.4295345783233643,
+ "num_input_tokens_seen": 79081264,
+ "step": 1805,
+ "train_runtime": 13867.5931,
+ "train_tokens_per_second": 5702.595
+ },
+ {
+ "epoch": 0.1218691085375707,
+ "grad_norm": 0.8001666467072275,
+ "learning_rate": 9.908764929550388e-06,
+ "loss": 0.4215728759765625,
+ "num_input_tokens_seen": 79301792,
+ "step": 1810,
+ "train_runtime": 13907.1584,
+ "train_tokens_per_second": 5702.228
+ },
+ {
+ "epoch": 0.12220576353353084,
+ "grad_norm": 0.9182690866647383,
+ "learning_rate": 9.908261442819444e-06,
+ "loss": 0.4215451717376709,
+ "num_input_tokens_seen": 79503840,
+ "step": 1815,
+ "train_runtime": 13955.1411,
+ "train_tokens_per_second": 5697.1
+ },
+ {
+ "epoch": 0.12254241852949098,
+ "grad_norm": 0.7856797264096972,
+ "learning_rate": 9.907756583508842e-06,
+ "loss": 0.3922400951385498,
+ "num_input_tokens_seen": 79740056,
+ "step": 1820,
+ "train_runtime": 13995.629,
+ "train_tokens_per_second": 5697.497
+ },
+ {
+ "epoch": 0.12287907352545112,
+ "grad_norm": 0.791802883681199,
+ "learning_rate": 9.907250351759761e-06,
+ "loss": 0.40973472595214844,
+ "num_input_tokens_seen": 79961848,
+ "step": 1825,
+ "train_runtime": 14035.6538,
+ "train_tokens_per_second": 5697.052
+ },
+ {
+ "epoch": 0.12321572852141126,
+ "grad_norm": 1.0344260725436623,
+ "learning_rate": 9.906742747713766e-06,
+ "loss": 0.4293796539306641,
+ "num_input_tokens_seen": 80180448,
+ "step": 1830,
+ "train_runtime": 14071.4195,
+ "train_tokens_per_second": 5698.107
+ },
+ {
+ "epoch": 0.1235523835173714,
+ "grad_norm": 1.0115683734845813,
+ "learning_rate": 9.906233771512808e-06,
+ "loss": 0.4379068374633789,
+ "num_input_tokens_seen": 80391288,
+ "step": 1835,
+ "train_runtime": 14108.8045,
+ "train_tokens_per_second": 5697.952
+ },
+ {
+ "epoch": 0.12388903851333154,
+ "grad_norm": 0.9113985281993394,
+ "learning_rate": 9.905723423299222e-06,
+ "loss": 0.44842100143432617,
+ "num_input_tokens_seen": 80619704,
+ "step": 1840,
+ "train_runtime": 14148.1364,
+ "train_tokens_per_second": 5698.256
+ },
+ {
+ "epoch": 0.12422569350929168,
+ "grad_norm": 0.8655689498703674,
+ "learning_rate": 9.905211703215723e-06,
+ "loss": 0.402877140045166,
+ "num_input_tokens_seen": 80832568,
+ "step": 1845,
+ "train_runtime": 14183.917,
+ "train_tokens_per_second": 5698.889
+ },
+ {
+ "epoch": 0.12456234850525182,
+ "grad_norm": 0.8753972943680609,
+ "learning_rate": 9.904698611405412e-06,
+ "loss": 0.4264510631561279,
+ "num_input_tokens_seen": 81058296,
+ "step": 1850,
+ "train_runtime": 14223.3173,
+ "train_tokens_per_second": 5698.973
+ },
+ {
+ "epoch": 0.12489900350121196,
+ "grad_norm": 0.7688694545812488,
+ "learning_rate": 9.904184148011774e-06,
+ "loss": 0.438901424407959,
+ "num_input_tokens_seen": 81274840,
+ "step": 1855,
+ "train_runtime": 14263.055,
+ "train_tokens_per_second": 5698.277
+ },
+ {
+ "epoch": 0.1252356584971721,
+ "grad_norm": 0.9233391292188109,
+ "learning_rate": 9.903668313178681e-06,
+ "loss": 0.44609222412109373,
+ "num_input_tokens_seen": 81500576,
+ "step": 1860,
+ "train_runtime": 14299.671,
+ "train_tokens_per_second": 5699.472
+ },
+ {
+ "epoch": 0.12557231349313225,
+ "grad_norm": 1.0943906900012805,
+ "learning_rate": 9.903151107050377e-06,
+ "loss": 0.396915864944458,
+ "num_input_tokens_seen": 81707648,
+ "step": 1865,
+ "train_runtime": 14337.8178,
+ "train_tokens_per_second": 5698.751
+ },
+ {
+ "epoch": 0.1259089684890924,
+ "grad_norm": 0.7168785500619042,
+ "learning_rate": 9.902632529771502e-06,
+ "loss": 0.42331228256225584,
+ "num_input_tokens_seen": 81926600,
+ "step": 1870,
+ "train_runtime": 14376.4633,
+ "train_tokens_per_second": 5698.662
+ },
+ {
+ "epoch": 0.12624562348505253,
+ "grad_norm": 1.0835704108533986,
+ "learning_rate": 9.902112581487074e-06,
+ "loss": 0.42675061225891114,
+ "num_input_tokens_seen": 82148216,
+ "step": 1875,
+ "train_runtime": 14416.4946,
+ "train_tokens_per_second": 5698.21
+ },
+ {
+ "epoch": 0.12658227848101267,
+ "grad_norm": 0.9867817197311566,
+ "learning_rate": 9.901591262342494e-06,
+ "loss": 0.4384635925292969,
+ "num_input_tokens_seen": 82358848,
+ "step": 1880,
+ "train_runtime": 14459.8702,
+ "train_tokens_per_second": 5695.684
+ },
+ {
+ "epoch": 0.1269189334769728,
+ "grad_norm": 0.9106268020914177,
+ "learning_rate": 9.901068572483548e-06,
+ "loss": 0.4060489654541016,
+ "num_input_tokens_seen": 82576440,
+ "step": 1885,
+ "train_runtime": 14497.7161,
+ "train_tokens_per_second": 5695.824
+ },
+ {
+ "epoch": 0.12725558847293295,
+ "grad_norm": 0.9009980190191491,
+ "learning_rate": 9.900544512056402e-06,
+ "loss": 0.47153472900390625,
+ "num_input_tokens_seen": 82807688,
+ "step": 1890,
+ "train_runtime": 14534.4393,
+ "train_tokens_per_second": 5697.343
+ },
+ {
+ "epoch": 0.12759224346889309,
+ "grad_norm": 0.8549638938689581,
+ "learning_rate": 9.900019081207612e-06,
+ "loss": 0.4209029197692871,
+ "num_input_tokens_seen": 83038168,
+ "step": 1895,
+ "train_runtime": 14569.2093,
+ "train_tokens_per_second": 5699.566
+ },
+ {
+ "epoch": 0.12792889846485322,
+ "grad_norm": 0.8968781974119465,
+ "learning_rate": 9.89949228008411e-06,
+ "loss": 0.42281227111816405,
+ "num_input_tokens_seen": 83267720,
+ "step": 1900,
+ "train_runtime": 14606.2305,
+ "train_tokens_per_second": 5700.836
+ },
+ {
+ "epoch": 0.12826555346081336,
+ "grad_norm": 0.8595709799025085,
+ "learning_rate": 9.898964108833216e-06,
+ "loss": 0.42650370597839354,
+ "num_input_tokens_seen": 83492616,
+ "step": 1905,
+ "train_runtime": 14646.7617,
+ "train_tokens_per_second": 5700.415
+ },
+ {
+ "epoch": 0.1286022084567735,
+ "grad_norm": 0.8815455063429253,
+ "learning_rate": 9.898434567602629e-06,
+ "loss": 0.4016720771789551,
+ "num_input_tokens_seen": 83715840,
+ "step": 1910,
+ "train_runtime": 14686.2575,
+ "train_tokens_per_second": 5700.284
+ },
+ {
+ "epoch": 0.12893886345273364,
+ "grad_norm": 0.9694766961124608,
+ "learning_rate": 9.897903656540437e-06,
+ "loss": 0.4289149284362793,
+ "num_input_tokens_seen": 83934184,
+ "step": 1915,
+ "train_runtime": 14724.9891,
+ "train_tokens_per_second": 5700.119
+ },
+ {
+ "epoch": 0.12927551844869378,
+ "grad_norm": 0.9823894967580632,
+ "learning_rate": 9.897371375795103e-06,
+ "loss": 0.43784418106079104,
+ "num_input_tokens_seen": 84150312,
+ "step": 1920,
+ "train_runtime": 14759.5087,
+ "train_tokens_per_second": 5701.43
+ },
+ {
+ "epoch": 0.12961217344465392,
+ "grad_norm": 0.8194539046535433,
+ "learning_rate": 9.896837725515484e-06,
+ "loss": 0.44328908920288085,
+ "num_input_tokens_seen": 84361600,
+ "step": 1925,
+ "train_runtime": 14805.8899,
+ "train_tokens_per_second": 5697.841
+ },
+ {
+ "epoch": 0.12994882844061406,
+ "grad_norm": 0.7081955649991436,
+ "learning_rate": 9.89630270585081e-06,
+ "loss": 0.39231128692626954,
+ "num_input_tokens_seen": 84577168,
+ "step": 1930,
+ "train_runtime": 14842.2012,
+ "train_tokens_per_second": 5698.425
+ },
+ {
+ "epoch": 0.1302854834365742,
+ "grad_norm": 0.8873554761311145,
+ "learning_rate": 9.895766316950694e-06,
+ "loss": 0.4394209384918213,
+ "num_input_tokens_seen": 84796464,
+ "step": 1935,
+ "train_runtime": 14876.8277,
+ "train_tokens_per_second": 5699.902
+ },
+ {
+ "epoch": 0.13062213843253434,
+ "grad_norm": 0.8262969710012915,
+ "learning_rate": 9.89522855896514e-06,
+ "loss": 0.41849603652954104,
+ "num_input_tokens_seen": 85025304,
+ "step": 1940,
+ "train_runtime": 14919.7496,
+ "train_tokens_per_second": 5698.843
+ },
+ {
+ "epoch": 0.13095879342849448,
+ "grad_norm": 0.9120986383703955,
+ "learning_rate": 9.89468943204453e-06,
+ "loss": 0.4164949893951416,
+ "num_input_tokens_seen": 85239728,
+ "step": 1945,
+ "train_runtime": 14964.5159,
+ "train_tokens_per_second": 5696.123
+ },
+ {
+ "epoch": 0.13129544842445462,
+ "grad_norm": 1.0478669869240167,
+ "learning_rate": 9.89414893633963e-06,
+ "loss": 0.41168971061706544,
+ "num_input_tokens_seen": 85451488,
+ "step": 1950,
+ "train_runtime": 14998.4779,
+ "train_tokens_per_second": 5697.344
+ },
+ {
+ "epoch": 0.13163210342041476,
+ "grad_norm": 0.8072074904847002,
+ "learning_rate": 9.893607072001584e-06,
+ "loss": 0.42524333000183107,
+ "num_input_tokens_seen": 85681416,
+ "step": 1955,
+ "train_runtime": 15045.5899,
+ "train_tokens_per_second": 5694.786
+ },
+ {
+ "epoch": 0.1319687584163749,
+ "grad_norm": 0.7884630260384243,
+ "learning_rate": 9.893063839181925e-06,
+ "loss": 0.41834263801574706,
+ "num_input_tokens_seen": 85897864,
+ "step": 1960,
+ "train_runtime": 15086.3562,
+ "train_tokens_per_second": 5693.745
+ },
+ {
+ "epoch": 0.13230541341233504,
+ "grad_norm": 0.8242479851716092,
+ "learning_rate": 9.892519238032566e-06,
+ "loss": 0.4272817611694336,
+ "num_input_tokens_seen": 86126184,
+ "step": 1965,
+ "train_runtime": 15130.1067,
+ "train_tokens_per_second": 5692.371
+ },
+ {
+ "epoch": 0.13264206840829518,
+ "grad_norm": 0.8883896038795247,
+ "learning_rate": 9.891973268705804e-06,
+ "loss": 0.45201525688171384,
+ "num_input_tokens_seen": 86340160,
+ "step": 1970,
+ "train_runtime": 15166.855,
+ "train_tokens_per_second": 5692.687
+ },
+ {
+ "epoch": 0.13297872340425532,
+ "grad_norm": 0.7899276150283895,
+ "learning_rate": 9.891425931354316e-06,
+ "loss": 0.4250628471374512,
+ "num_input_tokens_seen": 86565168,
+ "step": 1975,
+ "train_runtime": 15203.4839,
+ "train_tokens_per_second": 5693.772
+ },
+ {
+ "epoch": 0.13331537840021546,
+ "grad_norm": 0.8879719340880936,
+ "learning_rate": 9.890877226131165e-06,
+ "loss": 0.41506524085998536,
+ "num_input_tokens_seen": 86796552,
+ "step": 1980,
+ "train_runtime": 15247.3097,
+ "train_tokens_per_second": 5692.581
+ },
+ {
+ "epoch": 0.1336520333961756,
+ "grad_norm": 0.916769222493085,
+ "learning_rate": 9.890327153189793e-06,
+ "loss": 0.4300036907196045,
+ "num_input_tokens_seen": 87016968,
+ "step": 1985,
+ "train_runtime": 15292.8026,
+ "train_tokens_per_second": 5690.06
+ },
+ {
+ "epoch": 0.13398868839213574,
+ "grad_norm": 0.9050174324296295,
+ "learning_rate": 9.889775712684026e-06,
+ "loss": 0.4278897285461426,
+ "num_input_tokens_seen": 87240744,
+ "step": 1990,
+ "train_runtime": 15335.4097,
+ "train_tokens_per_second": 5688.843
+ },
+ {
+ "epoch": 0.13432534338809587,
+ "grad_norm": 0.8780717724520911,
+ "learning_rate": 9.889222904768074e-06,
+ "loss": 0.42691898345947266,
+ "num_input_tokens_seen": 87458888,
+ "step": 1995,
+ "train_runtime": 15375.168,
+ "train_tokens_per_second": 5688.321
+ },
+ {
+ "epoch": 0.13466199838405601,
+ "grad_norm": 0.8749943321506966,
+ "learning_rate": 9.888668729596528e-06,
+ "loss": 0.41673784255981444,
+ "num_input_tokens_seen": 87661224,
+ "step": 2000,
+ "train_runtime": 15417.6729,
+ "train_tokens_per_second": 5685.762
+ },
+ {
+ "epoch": 0.13499865338001615,
+ "grad_norm": 0.9429761954235121,
+ "learning_rate": 9.88811318732436e-06,
+ "loss": 0.4196599006652832,
+ "num_input_tokens_seen": 87871224,
+ "step": 2005,
+ "train_runtime": 15454.7159,
+ "train_tokens_per_second": 5685.722
+ },
+ {
+ "epoch": 0.1353353083759763,
+ "grad_norm": 0.8303836661224822,
+ "learning_rate": 9.887556278106927e-06,
+ "loss": 0.43319454193115237,
+ "num_input_tokens_seen": 88101424,
+ "step": 2010,
+ "train_runtime": 15492.7733,
+ "train_tokens_per_second": 5686.614
+ },
+ {
+ "epoch": 0.13567196337193643,
+ "grad_norm": 1.2988017888943026,
+ "learning_rate": 9.886998002099964e-06,
+ "loss": 0.42733011245727537,
+ "num_input_tokens_seen": 88331320,
+ "step": 2015,
+ "train_runtime": 15533.3637,
+ "train_tokens_per_second": 5686.555
+ },
+ {
+ "epoch": 0.13600861836789657,
+ "grad_norm": 0.926333904941965,
+ "learning_rate": 9.886438359459595e-06,
+ "loss": 0.4501063346862793,
+ "num_input_tokens_seen": 88546592,
+ "step": 2020,
+ "train_runtime": 15572.5905,
+ "train_tokens_per_second": 5686.054
+ },
+ {
+ "epoch": 0.1363452733638567,
+ "grad_norm": 0.8812830943715737,
+ "learning_rate": 9.885877350342319e-06,
+ "loss": 0.3861879587173462,
+ "num_input_tokens_seen": 88774480,
+ "step": 2025,
+ "train_runtime": 15611.197,
+ "train_tokens_per_second": 5686.59
+ },
+ {
+ "epoch": 0.13668192835981685,
+ "grad_norm": 0.8007094863514034,
+ "learning_rate": 9.885314974905024e-06,
+ "loss": 0.3875585556030273,
+ "num_input_tokens_seen": 89011576,
+ "step": 2030,
+ "train_runtime": 15652.3238,
+ "train_tokens_per_second": 5686.796
+ },
+ {
+ "epoch": 0.137018583355777,
+ "grad_norm": 0.9125014304117678,
+ "learning_rate": 9.884751233304973e-06,
+ "loss": 0.39730267524719237,
+ "num_input_tokens_seen": 89217520,
+ "step": 2035,
+ "train_runtime": 15686.075,
+ "train_tokens_per_second": 5687.689
+ },
+ {
+ "epoch": 0.13735523835173713,
+ "grad_norm": 0.8474671445937998,
+ "learning_rate": 9.884186125699816e-06,
+ "loss": 0.42992243766784666,
+ "num_input_tokens_seen": 89431816,
+ "step": 2040,
+ "train_runtime": 15726.6605,
+ "train_tokens_per_second": 5686.637
+ },
+ {
+ "epoch": 0.13769189334769727,
+ "grad_norm": 0.9102001150203423,
+ "learning_rate": 9.883619652247584e-06,
+ "loss": 0.4123133659362793,
+ "num_input_tokens_seen": 89651968,
+ "step": 2045,
+ "train_runtime": 15768.4222,
+ "train_tokens_per_second": 5685.538
+ },
+ {
+ "epoch": 0.1380285483436574,
+ "grad_norm": 0.6804193522827889,
+ "learning_rate": 9.883051813106687e-06,
+ "loss": 0.3782381296157837,
+ "num_input_tokens_seen": 89874624,
+ "step": 2050,
+ "train_runtime": 15813.3958,
+ "train_tokens_per_second": 5683.449
+ },
+ {
+ "epoch": 0.13836520333961755,
+ "grad_norm": 0.8989541162037309,
+ "learning_rate": 9.882482608435924e-06,
+ "loss": 0.43795013427734375,
+ "num_input_tokens_seen": 90097216,
+ "step": 2055,
+ "train_runtime": 15851.8211,
+ "train_tokens_per_second": 5683.714
+ },
+ {
+ "epoch": 0.1387018583355777,
+ "grad_norm": 0.8434117541493056,
+ "learning_rate": 9.881912038394468e-06,
+ "loss": 0.42223529815673827,
+ "num_input_tokens_seen": 90317784,
+ "step": 2060,
+ "train_runtime": 15897.2071,
+ "train_tokens_per_second": 5681.362
+ },
+ {
+ "epoch": 0.13903851333153783,
+ "grad_norm": 0.7960282493279416,
+ "learning_rate": 9.881340103141878e-06,
+ "loss": 0.4242293357849121,
+ "num_input_tokens_seen": 90546968,
+ "step": 2065,
+ "train_runtime": 15940.8703,
+ "train_tokens_per_second": 5680.177
+ },
+ {
+ "epoch": 0.13937516832749797,
+ "grad_norm": 0.8413026941754382,
+ "learning_rate": 9.880766802838092e-06,
+ "loss": 0.4057013511657715,
+ "num_input_tokens_seen": 90762816,
+ "step": 2070,
+ "train_runtime": 15983.7133,
+ "train_tokens_per_second": 5678.456
+ },
+ {
+ "epoch": 0.1397118233234581,
+ "grad_norm": 0.7350898878720474,
+ "learning_rate": 9.880192137643434e-06,
+ "loss": 0.4075471878051758,
+ "num_input_tokens_seen": 90979408,
+ "step": 2075,
+ "train_runtime": 16017.6972,
+ "train_tokens_per_second": 5679.931
+ },
+ {
+ "epoch": 0.14004847831941827,
+ "grad_norm": 0.9141607370838677,
+ "learning_rate": 9.879616107718608e-06,
+ "loss": 0.4198787689208984,
+ "num_input_tokens_seen": 91197560,
+ "step": 2080,
+ "train_runtime": 16051.0779,
+ "train_tokens_per_second": 5681.709
+ },
+ {
+ "epoch": 0.1403851333153784,
+ "grad_norm": 0.7897438400002121,
+ "learning_rate": 9.879038713224695e-06,
+ "loss": 0.4208370685577393,
+ "num_input_tokens_seen": 91421872,
+ "step": 2085,
+ "train_runtime": 16093.9488,
+ "train_tokens_per_second": 5680.512
+ },
+ {
+ "epoch": 0.14072178831133855,
+ "grad_norm": 0.8205504842451986,
+ "learning_rate": 9.878459954323165e-06,
+ "loss": 0.4030628204345703,
+ "num_input_tokens_seen": 91647112,
+ "step": 2090,
+ "train_runtime": 16126.4411,
+ "train_tokens_per_second": 5683.034
+ },
+ {
+ "epoch": 0.1410584433072987,
+ "grad_norm": 0.8325253458879384,
+ "learning_rate": 9.877879831175865e-06,
+ "loss": 0.39506003856658933,
+ "num_input_tokens_seen": 91871896,
+ "step": 2095,
+ "train_runtime": 16160.6707,
+ "train_tokens_per_second": 5684.906
+ },
+ {
+ "epoch": 0.14139509830325883,
+ "grad_norm": 0.821317865316409,
+ "learning_rate": 9.877298343945025e-06,
+ "loss": 0.41777982711791994,
+ "num_input_tokens_seen": 92094456,
+ "step": 2100,
+ "train_runtime": 16203.3038,
+ "train_tokens_per_second": 5683.684
+ },
+ {
+ "epoch": 0.14173175329921897,
+ "grad_norm": 0.7872012950426368,
+ "learning_rate": 9.876715492793256e-06,
+ "loss": 0.41480426788330077,
+ "num_input_tokens_seen": 92315440,
+ "step": 2105,
+ "train_runtime": 16242.0831,
+ "train_tokens_per_second": 5683.719
+ },
+ {
+ "epoch": 0.1420684082951791,
+ "grad_norm": 0.8092561101034788,
+ "learning_rate": 9.87613127788355e-06,
+ "loss": 0.389608097076416,
+ "num_input_tokens_seen": 92527368,
+ "step": 2110,
+ "train_runtime": 16278.4874,
+ "train_tokens_per_second": 5684.027
+ },
+ {
+ "epoch": 0.14240506329113925,
+ "grad_norm": 0.8850860821410723,
+ "learning_rate": 9.87554569937928e-06,
+ "loss": 0.4558421611785889,
+ "num_input_tokens_seen": 92759672,
+ "step": 2115,
+ "train_runtime": 16314.9617,
+ "train_tokens_per_second": 5685.559
+ },
+ {
+ "epoch": 0.1427417182870994,
+ "grad_norm": 0.8891822985539864,
+ "learning_rate": 9.874958757444203e-06,
+ "loss": 0.45552263259887693,
+ "num_input_tokens_seen": 92995096,
+ "step": 2120,
+ "train_runtime": 16350.6171,
+ "train_tokens_per_second": 5687.559
+ },
+ {
+ "epoch": 0.14307837328305953,
+ "grad_norm": 0.8780645983704002,
+ "learning_rate": 9.874370452242454e-06,
+ "loss": 0.43328495025634767,
+ "num_input_tokens_seen": 93229944,
+ "step": 2125,
+ "train_runtime": 16384.0995,
+ "train_tokens_per_second": 5690.27
+ },
+ {
+ "epoch": 0.14341502827901967,
+ "grad_norm": 0.9402959761478824,
+ "learning_rate": 9.873780783938553e-06,
+ "loss": 0.4462221622467041,
+ "num_input_tokens_seen": 93450480,
+ "step": 2130,
+ "train_runtime": 16417.1629,
+ "train_tokens_per_second": 5692.243
+ },
+ {
+ "epoch": 0.1437516832749798,
+ "grad_norm": 0.8574770635812221,
+ "learning_rate": 9.873189752697396e-06,
+ "loss": 0.3947939395904541,
+ "num_input_tokens_seen": 93670472,
+ "step": 2135,
+ "train_runtime": 16459.3253,
+ "train_tokens_per_second": 5691.027
+ },
+ {
+ "epoch": 0.14408833827093995,
+ "grad_norm": 0.8947319210921507,
+ "learning_rate": 9.872597358684265e-06,
+ "loss": 0.458558464050293,
+ "num_input_tokens_seen": 93883544,
+ "step": 2140,
+ "train_runtime": 16489.5195,
+ "train_tokens_per_second": 5693.528
+ },
+ {
+ "epoch": 0.1444249932669001,
+ "grad_norm": 0.8538177526814695,
+ "learning_rate": 9.872003602064816e-06,
+ "loss": 0.4299613475799561,
+ "num_input_tokens_seen": 94116824,
+ "step": 2145,
+ "train_runtime": 16523.1274,
+ "train_tokens_per_second": 5696.066
+ },
+ {
+ "epoch": 0.14476164826286023,
+ "grad_norm": 0.7471306795366713,
+ "learning_rate": 9.8714084830051e-06,
+ "loss": 0.4079564571380615,
+ "num_input_tokens_seen": 94349480,
+ "step": 2150,
+ "train_runtime": 16564.7452,
+ "train_tokens_per_second": 5695.8
+ },
+ {
+ "epoch": 0.14509830325882037,
+ "grad_norm": 0.8218664635099376,
+ "learning_rate": 9.870812001671533e-06,
+ "loss": 0.40776524543762205,
+ "num_input_tokens_seen": 94577624,
+ "step": 2155,
+ "train_runtime": 16600.9249,
+ "train_tokens_per_second": 5697.13
+ },
+ {
+ "epoch": 0.1454349582547805,
+ "grad_norm": 1.0737295934493194,
+ "learning_rate": 9.870214158230922e-06,
+ "loss": 0.4177376747131348,
+ "num_input_tokens_seen": 94804128,
+ "step": 2160,
+ "train_runtime": 16635.1905,
+ "train_tokens_per_second": 5699.011
+ },
+ {
+ "epoch": 0.14577161325074064,
+ "grad_norm": 0.786781014793927,
+ "learning_rate": 9.869614952850454e-06,
+ "loss": 0.40900435447692873,
+ "num_input_tokens_seen": 95026192,
+ "step": 2165,
+ "train_runtime": 16669.2594,
+ "train_tokens_per_second": 5700.685
+ },
+ {
+ "epoch": 0.14610826824670078,
+ "grad_norm": 0.8360055927785602,
+ "learning_rate": 9.86901438569769e-06,
+ "loss": 0.42821083068847654,
+ "num_input_tokens_seen": 95254672,
+ "step": 2170,
+ "train_runtime": 16706.7948,
+ "train_tokens_per_second": 5701.553
+ },
+ {
+ "epoch": 0.14644492324266092,
+ "grad_norm": 0.8933164074109287,
+ "learning_rate": 9.86841245694058e-06,
+ "loss": 0.425125789642334,
+ "num_input_tokens_seen": 95450280,
+ "step": 2175,
+ "train_runtime": 16746.2086,
+ "train_tokens_per_second": 5699.814
+ },
+ {
+ "epoch": 0.14678157823862106,
+ "grad_norm": 0.8906147805897214,
+ "learning_rate": 9.86780916674745e-06,
+ "loss": 0.45731253623962403,
+ "num_input_tokens_seen": 95651304,
+ "step": 2180,
+ "train_runtime": 16783.8632,
+ "train_tokens_per_second": 5699.004
+ },
+ {
+ "epoch": 0.1471182332345812,
+ "grad_norm": 1.0525327714019657,
+ "learning_rate": 9.86720451528701e-06,
+ "loss": 0.40560040473937986,
+ "num_input_tokens_seen": 95859952,
+ "step": 2185,
+ "train_runtime": 16822.6242,
+ "train_tokens_per_second": 5698.276
+ },
+ {
+ "epoch": 0.14745488823054134,
+ "grad_norm": 0.816498968853468,
+ "learning_rate": 9.866598502728348e-06,
+ "loss": 0.43126611709594725,
+ "num_input_tokens_seen": 96069168,
+ "step": 2190,
+ "train_runtime": 16862.6901,
+ "train_tokens_per_second": 5697.144
+ },
+ {
+ "epoch": 0.14779154322650148,
+ "grad_norm": 0.841739118951704,
+ "learning_rate": 9.865991129240931e-06,
+ "loss": 0.41145782470703124,
+ "num_input_tokens_seen": 96283672,
+ "step": 2195,
+ "train_runtime": 16896.2709,
+ "train_tokens_per_second": 5698.516
+ },
+ {
+ "epoch": 0.14812819822246162,
+ "grad_norm": 0.878868960575425,
+ "learning_rate": 9.865382394994614e-06,
+ "loss": 0.40999040603637693,
+ "num_input_tokens_seen": 96502528,
+ "step": 2200,
+ "train_runtime": 16939.3291,
+ "train_tokens_per_second": 5696.951
+ },
+ {
+ "epoch": 0.14846485321842176,
+ "grad_norm": 0.6967776248464405,
+ "learning_rate": 9.864772300159622e-06,
+ "loss": 0.3990957021713257,
+ "num_input_tokens_seen": 96720064,
+ "step": 2205,
+ "train_runtime": 16973.4551,
+ "train_tokens_per_second": 5698.313
+ },
+ {
+ "epoch": 0.1488015082143819,
+ "grad_norm": 0.9032740614218077,
+ "learning_rate": 9.864160844906571e-06,
+ "loss": 0.43683781623840334,
+ "num_input_tokens_seen": 96941160,
+ "step": 2210,
+ "train_runtime": 17009.3974,
+ "train_tokens_per_second": 5699.271
+ },
+ {
+ "epoch": 0.14913816321034204,
+ "grad_norm": 0.8444434220131386,
+ "learning_rate": 9.863548029406449e-06,
+ "loss": 0.419160795211792,
+ "num_input_tokens_seen": 97161584,
+ "step": 2215,
+ "train_runtime": 17050.6002,
+ "train_tokens_per_second": 5698.426
+ },
+ {
+ "epoch": 0.14947481820630218,
+ "grad_norm": 0.7872642768120427,
+ "learning_rate": 9.86293385383063e-06,
+ "loss": 0.3916934490203857,
+ "num_input_tokens_seen": 97372312,
+ "step": 2220,
+ "train_runtime": 17095.7206,
+ "train_tokens_per_second": 5695.713
+ },
+ {
+ "epoch": 0.14981147320226232,
+ "grad_norm": 0.9092720440628393,
+ "learning_rate": 9.862318318350865e-06,
+ "loss": 0.4371610641479492,
+ "num_input_tokens_seen": 97594144,
+ "step": 2225,
+ "train_runtime": 17132.445,
+ "train_tokens_per_second": 5696.452
+ },
+ {
+ "epoch": 0.15014812819822246,
+ "grad_norm": 0.7914448222714119,
+ "learning_rate": 9.861701423139287e-06,
+ "loss": 0.39911181926727296,
+ "num_input_tokens_seen": 97820888,
+ "step": 2230,
+ "train_runtime": 17165.9269,
+ "train_tokens_per_second": 5698.55
+ },
+ {
+ "epoch": 0.1504847831941826,
+ "grad_norm": 0.8316284130963478,
+ "learning_rate": 9.861083168368411e-06,
+ "loss": 0.41806850433349607,
+ "num_input_tokens_seen": 98045776,
+ "step": 2235,
+ "train_runtime": 17204.7719,
+ "train_tokens_per_second": 5698.755
+ },
+ {
+ "epoch": 0.15082143819014274,
+ "grad_norm": 0.9088287061584689,
+ "learning_rate": 9.860463554211125e-06,
+ "loss": 0.3891399383544922,
+ "num_input_tokens_seen": 98268344,
+ "step": 2240,
+ "train_runtime": 17246.0796,
+ "train_tokens_per_second": 5698.011
+ },
+ {
+ "epoch": 0.15115809318610288,
+ "grad_norm": 0.9381951784290052,
+ "learning_rate": 9.859842580840706e-06,
+ "loss": 0.4640520095825195,
+ "num_input_tokens_seen": 98493000,
+ "step": 2245,
+ "train_runtime": 17289.4391,
+ "train_tokens_per_second": 5696.715
+ },
+ {
+ "epoch": 0.15149474818206302,
+ "grad_norm": 0.7432535531903158,
+ "learning_rate": 9.859220248430806e-06,
+ "loss": 0.41865386962890627,
+ "num_input_tokens_seen": 98731240,
+ "step": 2250,
+ "train_runtime": 17329.2789,
+ "train_tokens_per_second": 5697.366
+ },
+ {
+ "epoch": 0.15183140317802316,
+ "grad_norm": 0.7752690728179633,
+ "learning_rate": 9.85859655715546e-06,
+ "loss": 0.39298291206359864,
+ "num_input_tokens_seen": 98953224,
+ "step": 2255,
+ "train_runtime": 17375.0272,
+ "train_tokens_per_second": 5695.141
+ },
+ {
+ "epoch": 0.1521680581739833,
+ "grad_norm": 0.8610212435397341,
+ "learning_rate": 9.857971507189077e-06,
+ "loss": 0.41016359329223634,
+ "num_input_tokens_seen": 99171512,
+ "step": 2260,
+ "train_runtime": 17414.0463,
+ "train_tokens_per_second": 5694.915
+ },
+ {
+ "epoch": 0.15250471316994343,
+ "grad_norm": 0.8902354817859172,
+ "learning_rate": 9.857345098706455e-06,
+ "loss": 0.4255502700805664,
+ "num_input_tokens_seen": 99402080,
+ "step": 2265,
+ "train_runtime": 17451.574,
+ "train_tokens_per_second": 5695.88
+ },
+ {
+ "epoch": 0.15284136816590357,
+ "grad_norm": 0.9086587862733272,
+ "learning_rate": 9.856717331882765e-06,
+ "loss": 0.4265317440032959,
+ "num_input_tokens_seen": 99607944,
+ "step": 2270,
+ "train_runtime": 17486.7082,
+ "train_tokens_per_second": 5696.209
+ },
+ {
+ "epoch": 0.1531780231618637,
+ "grad_norm": 0.9432280136780563,
+ "learning_rate": 9.85608820689356e-06,
+ "loss": 0.39440102577209474,
+ "num_input_tokens_seen": 99830656,
+ "step": 2275,
+ "train_runtime": 17528.0918,
+ "train_tokens_per_second": 5695.466
+ },
+ {
+ "epoch": 0.15351467815782385,
+ "grad_norm": 0.84487279707662,
+ "learning_rate": 9.855457723914772e-06,
+ "loss": 0.4084943771362305,
+ "num_input_tokens_seen": 100050880,
+ "step": 2280,
+ "train_runtime": 17562.2543,
+ "train_tokens_per_second": 5696.927
+ },
+ {
+ "epoch": 0.153851333153784,
+ "grad_norm": 0.9062843980819959,
+ "learning_rate": 9.854825883122713e-06,
+ "loss": 0.4020846366882324,
+ "num_input_tokens_seen": 100260608,
+ "step": 2285,
+ "train_runtime": 17602.1432,
+ "train_tokens_per_second": 5695.932
+ },
+ {
+ "epoch": 0.15418798814974413,
+ "grad_norm": 0.765345818981944,
+ "learning_rate": 9.85419268469408e-06,
+ "loss": 0.39870567321777345,
+ "num_input_tokens_seen": 100466752,
+ "step": 2290,
+ "train_runtime": 17638.1952,
+ "train_tokens_per_second": 5695.977
+ },
+ {
+ "epoch": 0.15452464314570427,
+ "grad_norm": 0.8051620604928534,
+ "learning_rate": 9.85355812880594e-06,
+ "loss": 0.3938899517059326,
+ "num_input_tokens_seen": 100686544,
+ "step": 2295,
+ "train_runtime": 17677.5734,
+ "train_tokens_per_second": 5695.722
+ },
+ {
+ "epoch": 0.1548612981416644,
+ "grad_norm": 0.7457579703786071,
+ "learning_rate": 9.852922215635747e-06,
+ "loss": 0.37209017276763917,
+ "num_input_tokens_seen": 100908792,
+ "step": 2300,
+ "train_runtime": 17719.4452,
+ "train_tokens_per_second": 5694.805
+ },
+ {
+ "epoch": 0.15519795313762455,
+ "grad_norm": 0.8179184151627187,
+ "learning_rate": 9.85228494536133e-06,
+ "loss": 0.4363281726837158,
+ "num_input_tokens_seen": 101132472,
+ "step": 2305,
+ "train_runtime": 17753.0155,
+ "train_tokens_per_second": 5696.636
+ },
+ {
+ "epoch": 0.1555346081335847,
+ "grad_norm": 0.8389619253246731,
+ "learning_rate": 9.851646318160901e-06,
+ "loss": 0.41831111907958984,
+ "num_input_tokens_seen": 101349024,
+ "step": 2310,
+ "train_runtime": 17793.8083,
+ "train_tokens_per_second": 5695.747
+ },
+ {
+ "epoch": 0.15587126312954483,
+ "grad_norm": 0.7507564820666337,
+ "learning_rate": 9.85100633421305e-06,
+ "loss": 0.39124910831451415,
+ "num_input_tokens_seen": 101558936,
+ "step": 2315,
+ "train_runtime": 17835.905,
+ "train_tokens_per_second": 5694.072
+ },
+ {
+ "epoch": 0.15620791812550497,
+ "grad_norm": 0.8895097039801884,
+ "learning_rate": 9.850364993696749e-06,
+ "loss": 0.42600722312927247,
+ "num_input_tokens_seen": 101772184,
+ "step": 2320,
+ "train_runtime": 17875.446,
+ "train_tokens_per_second": 5693.407
+ },
+ {
+ "epoch": 0.15654457312146514,
+ "grad_norm": 0.9520687878700897,
+ "learning_rate": 9.849722296791339e-06,
+ "loss": 0.4083821773529053,
+ "num_input_tokens_seen": 101982520,
+ "step": 2325,
+ "train_runtime": 17909.3087,
+ "train_tokens_per_second": 5694.386
+ },
+ {
+ "epoch": 0.15688122811742528,
+ "grad_norm": 0.9282345518369051,
+ "learning_rate": 9.849078243676558e-06,
+ "loss": 0.4323273658752441,
+ "num_input_tokens_seen": 102196704,
+ "step": 2330,
+ "train_runtime": 17950.3047,
+ "train_tokens_per_second": 5693.313
+ },
+ {
+ "epoch": 0.15721788311338541,
+ "grad_norm": 0.7571706215220008,
+ "learning_rate": 9.848432834532506e-06,
+ "loss": 0.3904409885406494,
+ "num_input_tokens_seen": 102413872,
+ "step": 2335,
+ "train_runtime": 17990.1951,
+ "train_tokens_per_second": 5692.76
+ },
+ {
+ "epoch": 0.15755453810934555,
+ "grad_norm": 0.7812435917007698,
+ "learning_rate": 9.847786069539673e-06,
+ "loss": 0.42157392501831054,
+ "num_input_tokens_seen": 102636848,
+ "step": 2340,
+ "train_runtime": 18027.5652,
+ "train_tokens_per_second": 5693.328
+ },
+ {
+ "epoch": 0.1578911931053057,
+ "grad_norm": 0.8268032160233915,
+ "learning_rate": 9.847137948878926e-06,
+ "loss": 0.4385071277618408,
+ "num_input_tokens_seen": 102861544,
+ "step": 2345,
+ "train_runtime": 18065.377,
+ "train_tokens_per_second": 5693.85
+ },
+ {
+ "epoch": 0.15822784810126583,
+ "grad_norm": 0.9872902684937591,
+ "learning_rate": 9.846488472731505e-06,
+ "loss": 0.46036734580993655,
+ "num_input_tokens_seen": 103092600,
+ "step": 2350,
+ "train_runtime": 18100.6003,
+ "train_tokens_per_second": 5695.535
+ },
+ {
+ "epoch": 0.15856450309722597,
+ "grad_norm": 0.8069383596278824,
+ "learning_rate": 9.845837641279038e-06,
+ "loss": 0.4151791572570801,
+ "num_input_tokens_seen": 103323896,
+ "step": 2355,
+ "train_runtime": 18131.0861,
+ "train_tokens_per_second": 5698.715
+ },
+ {
+ "epoch": 0.1589011580931861,
+ "grad_norm": 0.918784982161899,
+ "learning_rate": 9.845185454703528e-06,
+ "loss": 0.42508544921875,
+ "num_input_tokens_seen": 103541848,
+ "step": 2360,
+ "train_runtime": 18171.0294,
+ "train_tokens_per_second": 5698.183
+ },
+ {
+ "epoch": 0.15923781308914625,
+ "grad_norm": 0.817441107070197,
+ "learning_rate": 9.844531913187356e-06,
+ "loss": 0.3977517604827881,
+ "num_input_tokens_seen": 103759048,
+ "step": 2365,
+ "train_runtime": 18209.2359,
+ "train_tokens_per_second": 5698.155
+ },
+ {
+ "epoch": 0.1595744680851064,
+ "grad_norm": 0.8849599843025514,
+ "learning_rate": 9.84387701691328e-06,
+ "loss": 0.41399388313293456,
+ "num_input_tokens_seen": 103958648,
+ "step": 2370,
+ "train_runtime": 18252.5289,
+ "train_tokens_per_second": 5695.575
+ },
+ {
+ "epoch": 0.15991112308106653,
+ "grad_norm": 0.9775050084644975,
+ "learning_rate": 9.843220766064443e-06,
+ "loss": 0.4396827697753906,
+ "num_input_tokens_seen": 104164280,
+ "step": 2375,
+ "train_runtime": 18287.1992,
+ "train_tokens_per_second": 5696.022
+ },
+ {
+ "epoch": 0.16024777807702667,
+ "grad_norm": 1.026818097492493,
+ "learning_rate": 9.842563160824364e-06,
+ "loss": 0.4241004943847656,
+ "num_input_tokens_seen": 104376504,
+ "step": 2380,
+ "train_runtime": 18322.4835,
+ "train_tokens_per_second": 5696.635
+ },
+ {
+ "epoch": 0.1605844330729868,
+ "grad_norm": 0.8310388237023464,
+ "learning_rate": 9.841904201376939e-06,
+ "loss": 0.44525814056396484,
+ "num_input_tokens_seen": 104607616,
+ "step": 2385,
+ "train_runtime": 18361.4521,
+ "train_tokens_per_second": 5697.132
+ },
+ {
+ "epoch": 0.16092108806894695,
+ "grad_norm": 0.8107352807912601,
+ "learning_rate": 9.841243887906443e-06,
+ "loss": 0.3924037218093872,
+ "num_input_tokens_seen": 104827648,
+ "step": 2390,
+ "train_runtime": 18397.6205,
+ "train_tokens_per_second": 5697.892
+ },
+ {
+ "epoch": 0.1612577430649071,
+ "grad_norm": 0.8255737690801053,
+ "learning_rate": 9.840582220597532e-06,
+ "loss": 0.38776795864105223,
+ "num_input_tokens_seen": 105048616,
+ "step": 2395,
+ "train_runtime": 18437.9099,
+ "train_tokens_per_second": 5697.425
+ },
+ {
+ "epoch": 0.16159439806086723,
+ "grad_norm": 1.0393091287356393,
+ "learning_rate": 9.839919199635238e-06,
+ "loss": 0.44150123596191404,
+ "num_input_tokens_seen": 105271480,
+ "step": 2400,
+ "train_runtime": 18481.0333,
+ "train_tokens_per_second": 5696.19
+ },
+ {
+ "epoch": 0.16193105305682737,
+ "grad_norm": 1.778936138437422,
+ "learning_rate": 9.839254825204973e-06,
+ "loss": 0.4216805934906006,
+ "num_input_tokens_seen": 105497528,
+ "step": 2405,
+ "train_runtime": 18522.7039,
+ "train_tokens_per_second": 5695.579
+ },
+ {
+ "epoch": 0.1622677080527875,
+ "grad_norm": 0.8730955239731051,
+ "learning_rate": 9.838589097492527e-06,
+ "loss": 0.4195371627807617,
+ "num_input_tokens_seen": 105713952,
+ "step": 2410,
+ "train_runtime": 18562.2885,
+ "train_tokens_per_second": 5695.093
+ },
+ {
+ "epoch": 0.16260436304874765,
+ "grad_norm": 1.003318291974337,
+ "learning_rate": 9.83792201668407e-06,
+ "loss": 0.4166804313659668,
+ "num_input_tokens_seen": 105940976,
+ "step": 2415,
+ "train_runtime": 18598.4307,
+ "train_tokens_per_second": 5696.232
+ },
+ {
+ "epoch": 0.16294101804470779,
+ "grad_norm": 0.9882562378903975,
+ "learning_rate": 9.837253582966146e-06,
+ "loss": 0.4339381217956543,
+ "num_input_tokens_seen": 106165352,
+ "step": 2420,
+ "train_runtime": 18633.1302,
+ "train_tokens_per_second": 5697.666
+ },
+ {
+ "epoch": 0.16327767304066793,
+ "grad_norm": 0.8141828153275021,
+ "learning_rate": 9.836583796525686e-06,
+ "loss": 0.3936171531677246,
+ "num_input_tokens_seen": 106364536,
+ "step": 2425,
+ "train_runtime": 18675.1796,
+ "train_tokens_per_second": 5695.503
+ },
+ {
+ "epoch": 0.16361432803662807,
+ "grad_norm": 0.861811860528164,
+ "learning_rate": 9.835912657549986e-06,
+ "loss": 0.4514013290405273,
+ "num_input_tokens_seen": 106587560,
+ "step": 2430,
+ "train_runtime": 18717.8406,
+ "train_tokens_per_second": 5694.437
+ },
+ {
+ "epoch": 0.1639509830325882,
+ "grad_norm": 0.8811857541516364,
+ "learning_rate": 9.835240166226734e-06,
+ "loss": 0.39427995681762695,
+ "num_input_tokens_seen": 106799528,
+ "step": 2435,
+ "train_runtime": 18761.1279,
+ "train_tokens_per_second": 5692.596
+ },
+ {
+ "epoch": 0.16428763802854834,
+ "grad_norm": 0.8015505121470864,
+ "learning_rate": 9.834566322743987e-06,
+ "loss": 0.43557186126708985,
+ "num_input_tokens_seen": 107012480,
+ "step": 2440,
+ "train_runtime": 18795.981,
+ "train_tokens_per_second": 5693.37
+ },
+ {
+ "epoch": 0.16462429302450848,
+ "grad_norm": 0.8088311122015132,
+ "learning_rate": 9.833891127290186e-06,
+ "loss": 0.4289062976837158,
+ "num_input_tokens_seen": 107250496,
+ "step": 2445,
+ "train_runtime": 18827.4248,
+ "train_tokens_per_second": 5696.504
+ },
+ {
+ "epoch": 0.16496094802046862,
+ "grad_norm": 0.7483438381257065,
+ "learning_rate": 9.833214580054145e-06,
+ "loss": 0.3959768295288086,
+ "num_input_tokens_seen": 107469136,
+ "step": 2450,
+ "train_runtime": 18863.1469,
+ "train_tokens_per_second": 5697.307
+ },
+ {
+ "epoch": 0.16529760301642876,
+ "grad_norm": 0.8277673951140269,
+ "learning_rate": 9.832536681225058e-06,
+ "loss": 0.44450016021728517,
+ "num_input_tokens_seen": 107691864,
+ "step": 2455,
+ "train_runtime": 18898.3354,
+ "train_tokens_per_second": 5698.484
+ },
+ {
+ "epoch": 0.1656342580123889,
+ "grad_norm": 0.8054080794016197,
+ "learning_rate": 9.831857430992497e-06,
+ "loss": 0.3906950235366821,
+ "num_input_tokens_seen": 107905928,
+ "step": 2460,
+ "train_runtime": 18933.5537,
+ "train_tokens_per_second": 5699.19
+ },
+ {
+ "epoch": 0.16597091300834904,
+ "grad_norm": 0.8380957144167158,
+ "learning_rate": 9.831176829546416e-06,
+ "loss": 0.41658635139465333,
+ "num_input_tokens_seen": 108126176,
+ "step": 2465,
+ "train_runtime": 18970.3816,
+ "train_tokens_per_second": 5699.736
+ },
+ {
+ "epoch": 0.16630756800430918,
+ "grad_norm": 0.9452365824285418,
+ "learning_rate": 9.830494877077137e-06,
+ "loss": 0.46334171295166016,
+ "num_input_tokens_seen": 108358872,
+ "step": 2470,
+ "train_runtime": 19004.9571,
+ "train_tokens_per_second": 5701.611
+ },
+ {
+ "epoch": 0.16664422300026932,
+ "grad_norm": 1.1889597050506375,
+ "learning_rate": 9.82981157377537e-06,
+ "loss": 0.4094019889831543,
+ "num_input_tokens_seen": 108572432,
+ "step": 2475,
+ "train_runtime": 19047.273,
+ "train_tokens_per_second": 5700.156
+ },
+ {
+ "epoch": 0.16698087799622946,
+ "grad_norm": 0.8308872657286545,
+ "learning_rate": 9.829126919832198e-06,
+ "loss": 0.3822948455810547,
+ "num_input_tokens_seen": 108789616,
+ "step": 2480,
+ "train_runtime": 19086.2476,
+ "train_tokens_per_second": 5699.895
+ },
+ {
+ "epoch": 0.1673175329921896,
+ "grad_norm": 0.7290133116509023,
+ "learning_rate": 9.82844091543908e-06,
+ "loss": 0.4369968414306641,
+ "num_input_tokens_seen": 109023616,
+ "step": 2485,
+ "train_runtime": 19125.718,
+ "train_tokens_per_second": 5700.367
+ },
+ {
+ "epoch": 0.16765418798814974,
+ "grad_norm": 0.7990719470039959,
+ "learning_rate": 9.82775356078786e-06,
+ "loss": 0.4084890842437744,
+ "num_input_tokens_seen": 109238536,
+ "step": 2490,
+ "train_runtime": 19167.2805,
+ "train_tokens_per_second": 5699.219
+ },
+ {
+ "epoch": 0.16799084298410988,
+ "grad_norm": 0.7393931696314984,
+ "learning_rate": 9.82706485607075e-06,
+ "loss": 0.3479023456573486,
+ "num_input_tokens_seen": 109448064,
+ "step": 2495,
+ "train_runtime": 19202.9034,
+ "train_tokens_per_second": 5699.558
+ },
+ {
+ "epoch": 0.16832749798007002,
+ "grad_norm": 0.8223012662270877,
+ "learning_rate": 9.826374801480346e-06,
+ "loss": 0.4044170379638672,
+ "num_input_tokens_seen": 109654600,
+ "step": 2500,
+ "train_runtime": 19240.4274,
+ "train_tokens_per_second": 5699.177
+ },
+ {
+ "epoch": 0.16866415297603016,
+ "grad_norm": 0.7578035444329189,
+ "learning_rate": 9.825683397209617e-06,
+ "loss": 0.4128159999847412,
+ "num_input_tokens_seen": 109844880,
+ "step": 2505,
+ "train_runtime": 19277.0081,
+ "train_tokens_per_second": 5698.233
+ },
+ {
+ "epoch": 0.1690008079719903,
+ "grad_norm": 0.7754496384426975,
+ "learning_rate": 9.824990643451915e-06,
+ "loss": 0.40271682739257814,
+ "num_input_tokens_seen": 110072368,
+ "step": 2510,
+ "train_runtime": 19317.0275,
+ "train_tokens_per_second": 5698.204
+ },
+ {
+ "epoch": 0.16933746296795044,
+ "grad_norm": 0.7630225519369642,
+ "learning_rate": 9.824296540400965e-06,
+ "loss": 0.4032273769378662,
+ "num_input_tokens_seen": 110306168,
+ "step": 2515,
+ "train_runtime": 19359.4719,
+ "train_tokens_per_second": 5697.788
+ },
+ {
+ "epoch": 0.16967411796391058,
+ "grad_norm": 0.9445770110284926,
+ "learning_rate": 9.823601088250872e-06,
+ "loss": 0.43148112297058105,
+ "num_input_tokens_seen": 110517392,
+ "step": 2520,
+ "train_runtime": 19399.6091,
+ "train_tokens_per_second": 5696.888
+ },
+ {
+ "epoch": 0.17001077295987072,
+ "grad_norm": 0.8318843666554462,
+ "learning_rate": 9.822904287196114e-06,
+ "loss": 0.4280102729797363,
+ "num_input_tokens_seen": 110729000,
+ "step": 2525,
+ "train_runtime": 19437.1917,
+ "train_tokens_per_second": 5696.759
+ },
+ {
+ "epoch": 0.17034742795583085,
+ "grad_norm": 0.8164720266895754,
+ "learning_rate": 9.822206137431553e-06,
+ "loss": 0.4008312702178955,
+ "num_input_tokens_seen": 110951576,
+ "step": 2530,
+ "train_runtime": 19473.1465,
+ "train_tokens_per_second": 5697.671
+ },
+ {
+ "epoch": 0.170684082951791,
+ "grad_norm": 0.7906437056018215,
+ "learning_rate": 9.821506639152421e-06,
+ "loss": 0.42267236709594724,
+ "num_input_tokens_seen": 111165912,
+ "step": 2535,
+ "train_runtime": 19511.9498,
+ "train_tokens_per_second": 5697.325
+ },
+ {
+ "epoch": 0.17102073794775113,
+ "grad_norm": 0.8027232200064034,
+ "learning_rate": 9.820805792554334e-06,
+ "loss": 0.3978217124938965,
+ "num_input_tokens_seen": 111390648,
+ "step": 2540,
+ "train_runtime": 19548.9277,
+ "train_tokens_per_second": 5698.044
+ },
+ {
+ "epoch": 0.17135739294371127,
+ "grad_norm": 0.9732933783691021,
+ "learning_rate": 9.820103597833278e-06,
+ "loss": 0.40553808212280273,
+ "num_input_tokens_seen": 111613680,
+ "step": 2545,
+ "train_runtime": 19588.8813,
+ "train_tokens_per_second": 5697.808
+ },
+ {
+ "epoch": 0.1716940479396714,
+ "grad_norm": 0.82729851843596,
+ "learning_rate": 9.819400055185622e-06,
+ "loss": 0.3968106031417847,
+ "num_input_tokens_seen": 111830672,
+ "step": 2550,
+ "train_runtime": 19624.1494,
+ "train_tokens_per_second": 5698.625
+ },
+ {
+ "epoch": 0.17203070293563155,
+ "grad_norm": 0.8468905353541687,
+ "learning_rate": 9.818695164808108e-06,
+ "loss": 0.40885286331176757,
+ "num_input_tokens_seen": 112047832,
+ "step": 2555,
+ "train_runtime": 19668.7374,
+ "train_tokens_per_second": 5696.748
+ },
+ {
+ "epoch": 0.1723673579315917,
+ "grad_norm": 0.7939112227025558,
+ "learning_rate": 9.817988926897856e-06,
+ "loss": 0.39391307830810546,
+ "num_input_tokens_seen": 112267160,
+ "step": 2560,
+ "train_runtime": 19710.9584,
+ "train_tokens_per_second": 5695.672
+ },
+ {
+ "epoch": 0.17270401292755186,
+ "grad_norm": 0.8639441951747879,
+ "learning_rate": 9.817281341652367e-06,
+ "loss": 0.4082347869873047,
+ "num_input_tokens_seen": 112491800,
+ "step": 2565,
+ "train_runtime": 19749.2522,
+ "train_tokens_per_second": 5696.003
+ },
+ {
+ "epoch": 0.173040667923512,
+ "grad_norm": 0.716613288966692,
+ "learning_rate": 9.816572409269508e-06,
+ "loss": 0.3824393033981323,
+ "num_input_tokens_seen": 112719560,
+ "step": 2570,
+ "train_runtime": 19784.3744,
+ "train_tokens_per_second": 5697.403
+ },
+ {
+ "epoch": 0.17337732291947214,
+ "grad_norm": 0.7563163756690298,
+ "learning_rate": 9.815862129947537e-06,
+ "loss": 0.4002216339111328,
+ "num_input_tokens_seen": 112940512,
+ "step": 2575,
+ "train_runtime": 19821.7603,
+ "train_tokens_per_second": 5697.804
+ },
+ {
+ "epoch": 0.17371397791543228,
+ "grad_norm": 0.8087466973345561,
+ "learning_rate": 9.815150503885078e-06,
+ "loss": 0.37178959846496584,
+ "num_input_tokens_seen": 113165624,
+ "step": 2580,
+ "train_runtime": 19858.1999,
+ "train_tokens_per_second": 5698.685
+ },
+ {
+ "epoch": 0.17405063291139242,
+ "grad_norm": 0.8994912714777122,
+ "learning_rate": 9.814437531281133e-06,
+ "loss": 0.40114603042602537,
+ "num_input_tokens_seen": 113386744,
+ "step": 2585,
+ "train_runtime": 19903.4827,
+ "train_tokens_per_second": 5696.829
+ },
+ {
+ "epoch": 0.17438728790735256,
+ "grad_norm": 0.9269225867551159,
+ "learning_rate": 9.813723212335085e-06,
+ "loss": 0.43936843872070314,
+ "num_input_tokens_seen": 113593688,
+ "step": 2590,
+ "train_runtime": 19940.3903,
+ "train_tokens_per_second": 5696.663
+ },
+ {
+ "epoch": 0.1747239429033127,
+ "grad_norm": 0.8457918670255721,
+ "learning_rate": 9.813007547246691e-06,
+ "loss": 0.39149506092071534,
+ "num_input_tokens_seen": 113810536,
+ "step": 2595,
+ "train_runtime": 19976.29,
+ "train_tokens_per_second": 5697.281
+ },
+ {
+ "epoch": 0.17506059789927284,
+ "grad_norm": 0.8422559964504273,
+ "learning_rate": 9.812290536216083e-06,
+ "loss": 0.4519649982452393,
+ "num_input_tokens_seen": 114012584,
+ "step": 2600,
+ "train_runtime": 20014.4208,
+ "train_tokens_per_second": 5696.522
+ },
+ {
+ "epoch": 0.17539725289523297,
+ "grad_norm": 0.7614963966994434,
+ "learning_rate": 9.811572179443773e-06,
+ "loss": 0.3978102445602417,
+ "num_input_tokens_seen": 114240040,
+ "step": 2605,
+ "train_runtime": 20049.8879,
+ "train_tokens_per_second": 5697.789
+ },
+ {
+ "epoch": 0.17573390789119311,
+ "grad_norm": 0.7655048641611386,
+ "learning_rate": 9.810852477130645e-06,
+ "loss": 0.40557031631469725,
+ "num_input_tokens_seen": 114449696,
+ "step": 2610,
+ "train_runtime": 20089.5257,
+ "train_tokens_per_second": 5696.983
+ },
+ {
+ "epoch": 0.17607056288715325,
+ "grad_norm": 0.8093163549967093,
+ "learning_rate": 9.810131429477964e-06,
+ "loss": 0.38374016284942625,
+ "num_input_tokens_seen": 114677776,
+ "step": 2615,
+ "train_runtime": 20121.5227,
+ "train_tokens_per_second": 5699.259
+ },
+ {
+ "epoch": 0.1764072178831134,
+ "grad_norm": 0.8258310946925135,
+ "learning_rate": 9.809409036687364e-06,
+ "loss": 0.396789026260376,
+ "num_input_tokens_seen": 114896336,
+ "step": 2620,
+ "train_runtime": 20158.9581,
+ "train_tokens_per_second": 5699.518
+ },
+ {
+ "epoch": 0.17674387287907353,
+ "grad_norm": 0.7382001307664883,
+ "learning_rate": 9.808685298960867e-06,
+ "loss": 0.4137075424194336,
+ "num_input_tokens_seen": 115125264,
+ "step": 2625,
+ "train_runtime": 20201.5972,
+ "train_tokens_per_second": 5698.82
+ },
+ {
+ "epoch": 0.17708052787503367,
+ "grad_norm": 0.7663556283529158,
+ "learning_rate": 9.807960216500858e-06,
+ "loss": 0.4165202617645264,
+ "num_input_tokens_seen": 115343304,
+ "step": 2630,
+ "train_runtime": 20236.3656,
+ "train_tokens_per_second": 5699.803
+ },
+ {
+ "epoch": 0.1774171828709938,
+ "grad_norm": 0.7884797087616878,
+ "learning_rate": 9.807233789510106e-06,
+ "loss": 0.38329482078552246,
+ "num_input_tokens_seen": 115555208,
+ "step": 2635,
+ "train_runtime": 20279.8657,
+ "train_tokens_per_second": 5698.026
+ },
+ {
+ "epoch": 0.17775383786695395,
+ "grad_norm": 0.7323486746827084,
+ "learning_rate": 9.806506018191754e-06,
+ "loss": 0.39183108806610106,
+ "num_input_tokens_seen": 115777312,
+ "step": 2640,
+ "train_runtime": 20317.1207,
+ "train_tokens_per_second": 5698.51
+ },
+ {
+ "epoch": 0.1780904928629141,
+ "grad_norm": 0.8712184035601627,
+ "learning_rate": 9.80577690274932e-06,
+ "loss": 0.39883904457092284,
+ "num_input_tokens_seen": 116000320,
+ "step": 2645,
+ "train_runtime": 20353.9248,
+ "train_tokens_per_second": 5699.162
+ },
+ {
+ "epoch": 0.17842714785887423,
+ "grad_norm": 0.6517931345598202,
+ "learning_rate": 9.8050464433867e-06,
+ "loss": 0.3788264751434326,
+ "num_input_tokens_seen": 116228264,
+ "step": 2650,
+ "train_runtime": 20388.2032,
+ "train_tokens_per_second": 5700.761
+ },
+ {
+ "epoch": 0.17876380285483437,
+ "grad_norm": 0.8947080972086597,
+ "learning_rate": 9.804314640308165e-06,
+ "loss": 0.43610692024230957,
+ "num_input_tokens_seen": 116459784,
+ "step": 2655,
+ "train_runtime": 20433.6777,
+ "train_tokens_per_second": 5699.404
+ },
+ {
+ "epoch": 0.1791004578507945,
+ "grad_norm": 0.8316768119557644,
+ "learning_rate": 9.803581493718361e-06,
+ "loss": 0.4152191162109375,
+ "num_input_tokens_seen": 116677976,
+ "step": 2660,
+ "train_runtime": 20471.066,
+ "train_tokens_per_second": 5699.653
+ },
+ {
+ "epoch": 0.17943711284675465,
+ "grad_norm": 0.7694745863442054,
+ "learning_rate": 9.80284700382231e-06,
+ "loss": 0.4158869743347168,
+ "num_input_tokens_seen": 116913096,
+ "step": 2665,
+ "train_runtime": 20513.9974,
+ "train_tokens_per_second": 5699.186
+ },
+ {
+ "epoch": 0.1797737678427148,
+ "grad_norm": 0.7759018142960451,
+ "learning_rate": 9.802111170825408e-06,
+ "loss": 0.40905094146728516,
+ "num_input_tokens_seen": 117139040,
+ "step": 2670,
+ "train_runtime": 20555.3803,
+ "train_tokens_per_second": 5698.705
+ },
+ {
+ "epoch": 0.18011042283867493,
+ "grad_norm": 0.7209579007108823,
+ "learning_rate": 9.80137399493343e-06,
+ "loss": 0.4070711135864258,
+ "num_input_tokens_seen": 117354016,
+ "step": 2675,
+ "train_runtime": 20594.1226,
+ "train_tokens_per_second": 5698.423
+ },
+ {
+ "epoch": 0.18044707783463507,
+ "grad_norm": 0.7855863516525835,
+ "learning_rate": 9.800635476352525e-06,
+ "loss": 0.39305026531219484,
+ "num_input_tokens_seen": 117552920,
+ "step": 2680,
+ "train_runtime": 20636.4809,
+ "train_tokens_per_second": 5696.365
+ },
+ {
+ "epoch": 0.1807837328305952,
+ "grad_norm": 0.8011882919088659,
+ "learning_rate": 9.799895615289218e-06,
+ "loss": 0.38626153469085694,
+ "num_input_tokens_seen": 117762232,
+ "step": 2685,
+ "train_runtime": 20677.4885,
+ "train_tokens_per_second": 5695.19
+ },
+ {
+ "epoch": 0.18112038782655535,
+ "grad_norm": 0.7959442838200258,
+ "learning_rate": 9.799154411950407e-06,
+ "loss": 0.44461402893066404,
+ "num_input_tokens_seen": 117998240,
+ "step": 2690,
+ "train_runtime": 20716.7034,
+ "train_tokens_per_second": 5695.802
+ },
+ {
+ "epoch": 0.18145704282251549,
+ "grad_norm": 0.7750011575792003,
+ "learning_rate": 9.79841186654337e-06,
+ "loss": 0.4056252479553223,
+ "num_input_tokens_seen": 118211832,
+ "step": 2695,
+ "train_runtime": 20752.9555,
+ "train_tokens_per_second": 5696.144
+ },
+ {
+ "epoch": 0.18179369781847562,
+ "grad_norm": 0.8899930605283304,
+ "learning_rate": 9.797667979275752e-06,
+ "loss": 0.43903465270996095,
+ "num_input_tokens_seen": 118421416,
+ "step": 2700,
+ "train_runtime": 20785.8453,
+ "train_tokens_per_second": 5697.214
+ },
+ {
+ "epoch": 0.18213035281443576,
+ "grad_norm": 0.7964676861353408,
+ "learning_rate": 9.796922750355584e-06,
+ "loss": 0.3816461801528931,
+ "num_input_tokens_seen": 118635408,
+ "step": 2705,
+ "train_runtime": 20823.8564,
+ "train_tokens_per_second": 5697.091
+ },
+ {
+ "epoch": 0.1824670078103959,
+ "grad_norm": 0.8627424628290897,
+ "learning_rate": 9.796176179991266e-06,
+ "loss": 0.4243772029876709,
+ "num_input_tokens_seen": 118853520,
+ "step": 2710,
+ "train_runtime": 20861.0922,
+ "train_tokens_per_second": 5697.378
+ },
+ {
+ "epoch": 0.18280366280635604,
+ "grad_norm": 0.8313804598527397,
+ "learning_rate": 9.795428268391572e-06,
+ "loss": 0.39929049015045165,
+ "num_input_tokens_seen": 119057464,
+ "step": 2715,
+ "train_runtime": 20899.2219,
+ "train_tokens_per_second": 5696.741
+ },
+ {
+ "epoch": 0.18314031780231618,
+ "grad_norm": 0.9224656362131587,
+ "learning_rate": 9.794679015765655e-06,
+ "loss": 0.4025382995605469,
+ "num_input_tokens_seen": 119270504,
+ "step": 2720,
+ "train_runtime": 20936.4359,
+ "train_tokens_per_second": 5696.791
+ },
+ {
+ "epoch": 0.18347697279827632,
+ "grad_norm": 0.8830106782970203,
+ "learning_rate": 9.793928422323041e-06,
+ "loss": 0.40293426513671876,
+ "num_input_tokens_seen": 119494864,
+ "step": 2725,
+ "train_runtime": 20982.2238,
+ "train_tokens_per_second": 5695.052
+ },
+ {
+ "epoch": 0.18381362779423646,
+ "grad_norm": 0.9585116874337107,
+ "learning_rate": 9.793176488273629e-06,
+ "loss": 0.40497751235961915,
+ "num_input_tokens_seen": 119700000,
+ "step": 2730,
+ "train_runtime": 21018.3132,
+ "train_tokens_per_second": 5695.034
+ },
+ {
+ "epoch": 0.1841502827901966,
+ "grad_norm": 0.9047740005031505,
+ "learning_rate": 9.792423213827697e-06,
+ "loss": 0.3891148567199707,
+ "num_input_tokens_seen": 119916696,
+ "step": 2735,
+ "train_runtime": 21057.8627,
+ "train_tokens_per_second": 5694.628
+ },
+ {
+ "epoch": 0.18448693778615674,
+ "grad_norm": 0.8310766014320071,
+ "learning_rate": 9.791668599195897e-06,
+ "loss": 0.3831428289413452,
+ "num_input_tokens_seen": 120142080,
+ "step": 2740,
+ "train_runtime": 21101.6665,
+ "train_tokens_per_second": 5693.488
+ },
+ {
+ "epoch": 0.18482359278211688,
+ "grad_norm": 0.7855666850049595,
+ "learning_rate": 9.790912644589248e-06,
+ "loss": 0.42138309478759767,
+ "num_input_tokens_seen": 120348496,
+ "step": 2745,
+ "train_runtime": 21146.0201,
+ "train_tokens_per_second": 5691.307
+ },
+ {
+ "epoch": 0.18516024777807702,
+ "grad_norm": 0.7620864321495552,
+ "learning_rate": 9.790155350219159e-06,
+ "loss": 0.3965245246887207,
+ "num_input_tokens_seen": 120570160,
+ "step": 2750,
+ "train_runtime": 21186.6351,
+ "train_tokens_per_second": 5690.859
+ },
+ {
+ "epoch": 0.18549690277403716,
+ "grad_norm": 0.7543478062618101,
+ "learning_rate": 9.789396716297399e-06,
+ "loss": 0.4091955661773682,
+ "num_input_tokens_seen": 120800688,
+ "step": 2755,
+ "train_runtime": 21220.0669,
+ "train_tokens_per_second": 5692.757
+ },
+ {
+ "epoch": 0.1858335577699973,
+ "grad_norm": 0.8640772187026432,
+ "learning_rate": 9.788636743036122e-06,
+ "loss": 0.38339416980743407,
+ "num_input_tokens_seen": 121012328,
+ "step": 2760,
+ "train_runtime": 21257.8451,
+ "train_tokens_per_second": 5692.596
+ },
+ {
+ "epoch": 0.18617021276595744,
+ "grad_norm": 0.7868554451548372,
+ "learning_rate": 9.787875430647847e-06,
+ "loss": 0.3950460195541382,
+ "num_input_tokens_seen": 121232032,
+ "step": 2765,
+ "train_runtime": 21294.5091,
+ "train_tokens_per_second": 5693.112
+ },
+ {
+ "epoch": 0.18650686776191758,
+ "grad_norm": 0.8537247075271003,
+ "learning_rate": 9.787112779345477e-06,
+ "loss": 0.39361560344696045,
+ "num_input_tokens_seen": 121436200,
+ "step": 2770,
+ "train_runtime": 21331.5416,
+ "train_tokens_per_second": 5692.8
+ },
+ {
+ "epoch": 0.18684352275787772,
+ "grad_norm": 0.8358704371476329,
+ "learning_rate": 9.786348789342283e-06,
+ "loss": 0.4345618724822998,
+ "num_input_tokens_seen": 121656544,
+ "step": 2775,
+ "train_runtime": 21370.3261,
+ "train_tokens_per_second": 5692.779
+ },
+ {
+ "epoch": 0.18718017775383786,
+ "grad_norm": 0.7754752799232651,
+ "learning_rate": 9.785583460851912e-06,
+ "loss": 0.39746851921081544,
+ "num_input_tokens_seen": 121881368,
+ "step": 2780,
+ "train_runtime": 21412.528,
+ "train_tokens_per_second": 5692.059
+ },
+ {
+ "epoch": 0.187516832749798,
+ "grad_norm": 0.7537619511207324,
+ "learning_rate": 9.784816794088386e-06,
+ "loss": 0.4182771682739258,
+ "num_input_tokens_seen": 122116272,
+ "step": 2785,
+ "train_runtime": 21450.3054,
+ "train_tokens_per_second": 5692.985
+ },
+ {
+ "epoch": 0.18785348774575814,
+ "grad_norm": 0.86793765719075,
+ "learning_rate": 9.784048789266101e-06,
+ "loss": 0.4000124931335449,
+ "num_input_tokens_seen": 122324024,
+ "step": 2790,
+ "train_runtime": 21493.8907,
+ "train_tokens_per_second": 5691.107
+ },
+ {
+ "epoch": 0.18819014274171827,
+ "grad_norm": 0.7757014341183018,
+ "learning_rate": 9.78327944659983e-06,
+ "loss": 0.411564302444458,
+ "num_input_tokens_seen": 122539808,
+ "step": 2795,
+ "train_runtime": 21541.5867,
+ "train_tokens_per_second": 5688.523
+ },
+ {
+ "epoch": 0.18852679773767841,
+ "grad_norm": 0.8345683649256179,
+ "learning_rate": 9.78250876630471e-06,
+ "loss": 0.40423192977905276,
+ "num_input_tokens_seen": 122747320,
+ "step": 2800,
+ "train_runtime": 21579.8161,
+ "train_tokens_per_second": 5688.061
+ },
+ {
+ "epoch": 0.18886345273363855,
+ "grad_norm": 0.8512506107699013,
+ "learning_rate": 9.781736748596267e-06,
+ "loss": 0.39595718383789064,
+ "num_input_tokens_seen": 122957984,
+ "step": 2805,
+ "train_runtime": 21616.5661,
+ "train_tokens_per_second": 5688.137
+ },
+ {
+ "epoch": 0.18920010772959872,
+ "grad_norm": 0.8124112022835439,
+ "learning_rate": 9.78096339369039e-06,
+ "loss": 0.3889493942260742,
+ "num_input_tokens_seen": 123167712,
+ "step": 2810,
+ "train_runtime": 21651.6405,
+ "train_tokens_per_second": 5688.609
+ },
+ {
+ "epoch": 0.18953676272555886,
+ "grad_norm": 0.7367215294516509,
+ "learning_rate": 9.780188701803345e-06,
+ "loss": 0.42559328079223635,
+ "num_input_tokens_seen": 123386632,
+ "step": 2815,
+ "train_runtime": 21694.1793,
+ "train_tokens_per_second": 5687.546
+ },
+ {
+ "epoch": 0.189873417721519,
+ "grad_norm": 0.7954578075141111,
+ "learning_rate": 9.779412673151772e-06,
+ "loss": 0.40504965782165525,
+ "num_input_tokens_seen": 123614712,
+ "step": 2820,
+ "train_runtime": 21735.5555,
+ "train_tokens_per_second": 5687.212
+ },
+ {
+ "epoch": 0.19021007271747914,
+ "grad_norm": 0.7248412230311774,
+ "learning_rate": 9.778635307952686e-06,
+ "loss": 0.4043741703033447,
+ "num_input_tokens_seen": 123839728,
+ "step": 2825,
+ "train_runtime": 21777.9441,
+ "train_tokens_per_second": 5686.475
+ },
+ {
+ "epoch": 0.19054672771343928,
+ "grad_norm": 0.9358958553909015,
+ "learning_rate": 9.777856606423472e-06,
+ "loss": 0.403910493850708,
+ "num_input_tokens_seen": 124065288,
+ "step": 2830,
+ "train_runtime": 21817.8593,
+ "train_tokens_per_second": 5686.41
+ },
+ {
+ "epoch": 0.19088338270939942,
+ "grad_norm": 0.8463489849936312,
+ "learning_rate": 9.777076568781897e-06,
+ "loss": 0.43606057167053225,
+ "num_input_tokens_seen": 124264360,
+ "step": 2835,
+ "train_runtime": 21855.3285,
+ "train_tokens_per_second": 5685.769
+ },
+ {
+ "epoch": 0.19122003770535956,
+ "grad_norm": 0.9165712627873297,
+ "learning_rate": 9.77629519524609e-06,
+ "loss": 0.3660429954528809,
+ "num_input_tokens_seen": 124479288,
+ "step": 2840,
+ "train_runtime": 21894.9089,
+ "train_tokens_per_second": 5685.307
+ },
+ {
+ "epoch": 0.1915566927013197,
+ "grad_norm": 0.7656937544219163,
+ "learning_rate": 9.775512486034564e-06,
+ "loss": 0.4081143379211426,
+ "num_input_tokens_seen": 124692848,
+ "step": 2845,
+ "train_runtime": 21938.4076,
+ "train_tokens_per_second": 5683.769
+ },
+ {
+ "epoch": 0.19189334769727984,
+ "grad_norm": 0.867416455867072,
+ "learning_rate": 9.774728441366197e-06,
+ "loss": 0.4261101245880127,
+ "num_input_tokens_seen": 124899520,
+ "step": 2850,
+ "train_runtime": 21975.3012,
+ "train_tokens_per_second": 5683.632
+ },
+ {
+ "epoch": 0.19223000269323998,
+ "grad_norm": 0.760427761575941,
+ "learning_rate": 9.773943061460249e-06,
+ "loss": 0.39615979194641116,
+ "num_input_tokens_seen": 125109776,
+ "step": 2855,
+ "train_runtime": 22015.122,
+ "train_tokens_per_second": 5682.902
+ },
+ {
+ "epoch": 0.19256665768920012,
+ "grad_norm": 0.8768768754679016,
+ "learning_rate": 9.773156346536345e-06,
+ "loss": 0.42858476638793946,
+ "num_input_tokens_seen": 125326624,
+ "step": 2860,
+ "train_runtime": 22054.6435,
+ "train_tokens_per_second": 5682.55
+ },
+ {
+ "epoch": 0.19290331268516026,
+ "grad_norm": 0.7656261692386461,
+ "learning_rate": 9.772368296814489e-06,
+ "loss": 0.3920739650726318,
+ "num_input_tokens_seen": 125545392,
+ "step": 2865,
+ "train_runtime": 22088.3487,
+ "train_tokens_per_second": 5683.784
+ },
+ {
+ "epoch": 0.1932399676811204,
+ "grad_norm": 0.7687451750642236,
+ "learning_rate": 9.771578912515058e-06,
+ "loss": 0.439731502532959,
+ "num_input_tokens_seen": 125763512,
+ "step": 2870,
+ "train_runtime": 22123.8551,
+ "train_tokens_per_second": 5684.521
+ },
+ {
+ "epoch": 0.19357662267708053,
+ "grad_norm": 3.7881124274468654,
+ "learning_rate": 9.7707881938588e-06,
+ "loss": 0.3899986743927002,
+ "num_input_tokens_seen": 125983312,
+ "step": 2875,
+ "train_runtime": 22166.3825,
+ "train_tokens_per_second": 5683.531
+ },
+ {
+ "epoch": 0.19391327767304067,
+ "grad_norm": 0.7637738370885714,
+ "learning_rate": 9.769996141066836e-06,
+ "loss": 0.4218841552734375,
+ "num_input_tokens_seen": 126221208,
+ "step": 2880,
+ "train_runtime": 22196.5764,
+ "train_tokens_per_second": 5686.517
+ },
+ {
+ "epoch": 0.1942499326690008,
+ "grad_norm": 0.7196684362616566,
+ "learning_rate": 9.76920275436066e-06,
+ "loss": 0.41857566833496096,
+ "num_input_tokens_seen": 126460024,
+ "step": 2885,
+ "train_runtime": 22232.6411,
+ "train_tokens_per_second": 5688.034
+ },
+ {
+ "epoch": 0.19458658766496095,
+ "grad_norm": 0.7681501572608938,
+ "learning_rate": 9.768408033962145e-06,
+ "loss": 0.4293510437011719,
+ "num_input_tokens_seen": 126682488,
+ "step": 2890,
+ "train_runtime": 22274.9037,
+ "train_tokens_per_second": 5687.229
+ },
+ {
+ "epoch": 0.1949232426609211,
+ "grad_norm": 0.7776932982886595,
+ "learning_rate": 9.767611980093525e-06,
+ "loss": 0.40836372375488283,
+ "num_input_tokens_seen": 126903552,
+ "step": 2895,
+ "train_runtime": 22315.0672,
+ "train_tokens_per_second": 5686.9
+ },
+ {
+ "epoch": 0.19525989765688123,
+ "grad_norm": 0.7725164291944147,
+ "learning_rate": 9.76681459297742e-06,
+ "loss": 0.41483469009399415,
+ "num_input_tokens_seen": 127124488,
+ "step": 2900,
+ "train_runtime": 22361.1218,
+ "train_tokens_per_second": 5685.068
+ },
+ {
+ "epoch": 0.19559655265284137,
+ "grad_norm": 1.0523171028632952,
+ "learning_rate": 9.766015872836814e-06,
+ "loss": 0.4586020469665527,
+ "num_input_tokens_seen": 127351272,
+ "step": 2905,
+ "train_runtime": 22403.7587,
+ "train_tokens_per_second": 5684.371
+ },
+ {
+ "epoch": 0.1959332076488015,
+ "grad_norm": 0.981785930598382,
+ "learning_rate": 9.765215819895066e-06,
+ "loss": 0.4265925407409668,
+ "num_input_tokens_seen": 127584848,
+ "step": 2910,
+ "train_runtime": 22435.5709,
+ "train_tokens_per_second": 5686.722
+ },
+ {
+ "epoch": 0.19626986264476165,
+ "grad_norm": 0.7598810830140263,
+ "learning_rate": 9.76441443437591e-06,
+ "loss": 0.3775201320648193,
+ "num_input_tokens_seen": 127815392,
+ "step": 2915,
+ "train_runtime": 22468.3308,
+ "train_tokens_per_second": 5688.691
+ },
+ {
+ "epoch": 0.1966065176407218,
+ "grad_norm": 0.8036869153351719,
+ "learning_rate": 9.76361171650345e-06,
+ "loss": 0.4249995708465576,
+ "num_input_tokens_seen": 128025072,
+ "step": 2920,
+ "train_runtime": 22508.1392,
+ "train_tokens_per_second": 5687.946
+ },
+ {
+ "epoch": 0.19694317263668193,
+ "grad_norm": 0.8854228803465163,
+ "learning_rate": 9.762807666502164e-06,
+ "loss": 0.41265301704406737,
+ "num_input_tokens_seen": 128236720,
+ "step": 2925,
+ "train_runtime": 22548.1044,
+ "train_tokens_per_second": 5687.251
+ },
+ {
+ "epoch": 0.19727982763264207,
+ "grad_norm": 1.302088487080232,
+ "learning_rate": 9.762002284596898e-06,
+ "loss": 0.4183157444000244,
+ "num_input_tokens_seen": 128448432,
+ "step": 2930,
+ "train_runtime": 22591.5125,
+ "train_tokens_per_second": 5685.694
+ },
+ {
+ "epoch": 0.1976164826286022,
+ "grad_norm": 0.8232254317537885,
+ "learning_rate": 9.761195571012882e-06,
+ "loss": 0.4257832050323486,
+ "num_input_tokens_seen": 128657896,
+ "step": 2935,
+ "train_runtime": 22629.0733,
+ "train_tokens_per_second": 5685.513
+ },
+ {
+ "epoch": 0.19795313762456235,
+ "grad_norm": 0.8172096711574582,
+ "learning_rate": 9.760387525975705e-06,
+ "loss": 0.395475172996521,
+ "num_input_tokens_seen": 128883360,
+ "step": 2940,
+ "train_runtime": 22662.7742,
+ "train_tokens_per_second": 5687.007
+ },
+ {
+ "epoch": 0.1982897926205225,
+ "grad_norm": 0.7517384205877706,
+ "learning_rate": 9.759578149711337e-06,
+ "loss": 0.35533039569854735,
+ "num_input_tokens_seen": 129112664,
+ "step": 2945,
+ "train_runtime": 22703.1966,
+ "train_tokens_per_second": 5686.982
+ },
+ {
+ "epoch": 0.19862644761648263,
+ "grad_norm": 0.7164265548546924,
+ "learning_rate": 9.758767442446117e-06,
+ "loss": 0.38588006496429444,
+ "num_input_tokens_seen": 129342864,
+ "step": 2950,
+ "train_runtime": 22748.3032,
+ "train_tokens_per_second": 5685.825
+ },
+ {
+ "epoch": 0.19896310261244277,
+ "grad_norm": 0.733779366224309,
+ "learning_rate": 9.757955404406755e-06,
+ "loss": 0.38382346630096437,
+ "num_input_tokens_seen": 129566032,
+ "step": 2955,
+ "train_runtime": 22786.3627,
+ "train_tokens_per_second": 5686.122
+ },
+ {
+ "epoch": 0.1992997576084029,
+ "grad_norm": 0.9448244412990339,
+ "learning_rate": 9.757142035820337e-06,
+ "loss": 0.3917106628417969,
+ "num_input_tokens_seen": 129766056,
+ "step": 2960,
+ "train_runtime": 22824.9574,
+ "train_tokens_per_second": 5685.27
+ },
+ {
+ "epoch": 0.19963641260436304,
+ "grad_norm": 0.9552752355322248,
+ "learning_rate": 9.756327336914318e-06,
+ "loss": 0.4134823322296143,
+ "num_input_tokens_seen": 129977048,
+ "step": 2965,
+ "train_runtime": 22868.2844,
+ "train_tokens_per_second": 5683.725
+ },
+ {
+ "epoch": 0.19997306760032318,
+ "grad_norm": 1.0910385569978054,
+ "learning_rate": 9.755511307916526e-06,
+ "loss": 0.4055294990539551,
+ "num_input_tokens_seen": 130194224,
+ "step": 2970,
+ "train_runtime": 22901.3701,
+ "train_tokens_per_second": 5684.997
+ },
+ {
+ "epoch": 0.20030972259628332,
+ "grad_norm": 0.8363361731098393,
+ "learning_rate": 9.754693949055161e-06,
+ "loss": 0.4119466781616211,
+ "num_input_tokens_seen": 130423144,
+ "step": 2975,
+ "train_runtime": 22941.2287,
+ "train_tokens_per_second": 5685.098
+ },
+ {
+ "epoch": 0.20064637759224346,
+ "grad_norm": 0.8509098752743555,
+ "learning_rate": 9.753875260558795e-06,
+ "loss": 0.43303780555725097,
+ "num_input_tokens_seen": 130623168,
+ "step": 2980,
+ "train_runtime": 22977.1333,
+ "train_tokens_per_second": 5684.92
+ },
+ {
+ "epoch": 0.2009830325882036,
+ "grad_norm": 0.7851629802721982,
+ "learning_rate": 9.75305524265637e-06,
+ "loss": 0.39323015213012696,
+ "num_input_tokens_seen": 130853536,
+ "step": 2985,
+ "train_runtime": 23010.865,
+ "train_tokens_per_second": 5686.598
+ },
+ {
+ "epoch": 0.20131968758416374,
+ "grad_norm": 0.8423251979034995,
+ "learning_rate": 9.752233895577205e-06,
+ "loss": 0.4042644023895264,
+ "num_input_tokens_seen": 131059592,
+ "step": 2990,
+ "train_runtime": 23047.2444,
+ "train_tokens_per_second": 5686.562
+ },
+ {
+ "epoch": 0.20165634258012388,
+ "grad_norm": 0.7311264203424918,
+ "learning_rate": 9.751411219550984e-06,
+ "loss": 0.40361328125,
+ "num_input_tokens_seen": 131276760,
+ "step": 2995,
+ "train_runtime": 23082.7777,
+ "train_tokens_per_second": 5687.217
+ },
+ {
+ "epoch": 0.20199299757608402,
+ "grad_norm": 0.9307523548267392,
+ "learning_rate": 9.750587214807766e-06,
+ "loss": 0.39671123027801514,
+ "num_input_tokens_seen": 131486248,
+ "step": 3000,
+ "train_runtime": 23124.9978,
+ "train_tokens_per_second": 5685.892
+ },
+ {
+ "epoch": 0.20232965257204416,
+ "grad_norm": 0.8339744148987415,
+ "learning_rate": 9.749761881577981e-06,
+ "loss": 0.42069306373596194,
+ "num_input_tokens_seen": 131716280,
+ "step": 3005,
+ "train_runtime": 23166.3136,
+ "train_tokens_per_second": 5685.681
+ },
+ {
+ "epoch": 0.2026663075680043,
+ "grad_norm": 0.7849226835237555,
+ "learning_rate": 9.748935220092434e-06,
+ "loss": 0.39777138233184817,
+ "num_input_tokens_seen": 131935832,
+ "step": 3010,
+ "train_runtime": 23202.4925,
+ "train_tokens_per_second": 5686.278
+ },
+ {
+ "epoch": 0.20300296256396444,
+ "grad_norm": 0.8473889903914309,
+ "learning_rate": 9.748107230582295e-06,
+ "loss": 0.39002056121826173,
+ "num_input_tokens_seen": 132140024,
+ "step": 3015,
+ "train_runtime": 23239.8205,
+ "train_tokens_per_second": 5685.931
+ },
+ {
+ "epoch": 0.20333961755992458,
+ "grad_norm": 0.7565501165020906,
+ "learning_rate": 9.747277913279109e-06,
+ "loss": 0.4608659267425537,
+ "num_input_tokens_seen": 132358504,
+ "step": 3020,
+ "train_runtime": 23285.5396,
+ "train_tokens_per_second": 5684.15
+ },
+ {
+ "epoch": 0.20367627255588472,
+ "grad_norm": 0.71921389289753,
+ "learning_rate": 9.746447268414794e-06,
+ "loss": 0.44081912040710447,
+ "num_input_tokens_seen": 132589576,
+ "step": 3025,
+ "train_runtime": 23324.3416,
+ "train_tokens_per_second": 5684.601
+ },
+ {
+ "epoch": 0.20401292755184486,
+ "grad_norm": 0.8592739422898477,
+ "learning_rate": 9.745615296221635e-06,
+ "loss": 0.40747461318969724,
+ "num_input_tokens_seen": 132819496,
+ "step": 3030,
+ "train_runtime": 23358.8471,
+ "train_tokens_per_second": 5686.047
+ },
+ {
+ "epoch": 0.204349582547805,
+ "grad_norm": 0.7375243033673482,
+ "learning_rate": 9.74478199693229e-06,
+ "loss": 0.42429242134094236,
+ "num_input_tokens_seen": 133045880,
+ "step": 3035,
+ "train_runtime": 23401.9716,
+ "train_tokens_per_second": 5685.242
+ },
+ {
+ "epoch": 0.20468623754376514,
+ "grad_norm": 0.7984195828599883,
+ "learning_rate": 9.743947370779791e-06,
+ "loss": 0.4262086391448975,
+ "num_input_tokens_seen": 133273848,
+ "step": 3040,
+ "train_runtime": 23442.6366,
+ "train_tokens_per_second": 5685.105
+ },
+ {
+ "epoch": 0.20502289253972528,
+ "grad_norm": 0.7817233028800991,
+ "learning_rate": 9.743111417997536e-06,
+ "loss": 0.41955132484436036,
+ "num_input_tokens_seen": 133500640,
+ "step": 3045,
+ "train_runtime": 23481.7864,
+ "train_tokens_per_second": 5685.285
+ },
+ {
+ "epoch": 0.20535954753568542,
+ "grad_norm": 0.8413661304916547,
+ "learning_rate": 9.742274138819298e-06,
+ "loss": 0.416639518737793,
+ "num_input_tokens_seen": 133719848,
+ "step": 3050,
+ "train_runtime": 23518.5528,
+ "train_tokens_per_second": 5685.718
+ },
+ {
+ "epoch": 0.20569620253164558,
+ "grad_norm": 0.8835193289624108,
+ "learning_rate": 9.74143553347922e-06,
+ "loss": 0.45115389823913576,
+ "num_input_tokens_seen": 133944240,
+ "step": 3055,
+ "train_runtime": 23558.0231,
+ "train_tokens_per_second": 5685.716
+ },
+ {
+ "epoch": 0.20603285752760572,
+ "grad_norm": 0.7584811920041505,
+ "learning_rate": 9.740595602211814e-06,
+ "loss": 0.395220947265625,
+ "num_input_tokens_seen": 134166880,
+ "step": 3060,
+ "train_runtime": 23594.0834,
+ "train_tokens_per_second": 5686.463
+ },
+ {
+ "epoch": 0.20636951252356586,
+ "grad_norm": 0.779533118470965,
+ "learning_rate": 9.739754345251963e-06,
+ "loss": 0.3915586471557617,
+ "num_input_tokens_seen": 134373184,
+ "step": 3065,
+ "train_runtime": 23633.4149,
+ "train_tokens_per_second": 5685.729
+ },
+ {
+ "epoch": 0.206706167519526,
+ "grad_norm": 0.752085146799199,
+ "learning_rate": 9.738911762834924e-06,
+ "loss": 0.36250555515289307,
+ "num_input_tokens_seen": 134594840,
+ "step": 3070,
+ "train_runtime": 23669.854,
+ "train_tokens_per_second": 5686.34
+ },
+ {
+ "epoch": 0.20704282251548614,
+ "grad_norm": 0.9101615257113173,
+ "learning_rate": 9.738067855196323e-06,
+ "loss": 0.40388903617858884,
+ "num_input_tokens_seen": 134814144,
+ "step": 3075,
+ "train_runtime": 23702.4871,
+ "train_tokens_per_second": 5687.764
+ },
+ {
+ "epoch": 0.20737947751144628,
+ "grad_norm": 0.937886524311701,
+ "learning_rate": 9.737222622572153e-06,
+ "loss": 0.4489736557006836,
+ "num_input_tokens_seen": 135025088,
+ "step": 3080,
+ "train_runtime": 23747.5708,
+ "train_tokens_per_second": 5685.848
+ },
+ {
+ "epoch": 0.20771613250740642,
+ "grad_norm": 0.939422935382809,
+ "learning_rate": 9.736376065198784e-06,
+ "loss": 0.4101541519165039,
+ "num_input_tokens_seen": 135253064,
+ "step": 3085,
+ "train_runtime": 23786.6187,
+ "train_tokens_per_second": 5686.099
+ },
+ {
+ "epoch": 0.20805278750336656,
+ "grad_norm": 0.8449440163538374,
+ "learning_rate": 9.735528183312951e-06,
+ "loss": 0.39399616718292235,
+ "num_input_tokens_seen": 135486752,
+ "step": 3090,
+ "train_runtime": 23827.8882,
+ "train_tokens_per_second": 5686.058
+ },
+ {
+ "epoch": 0.2083894424993267,
+ "grad_norm": 0.7846090848755455,
+ "learning_rate": 9.734678977151762e-06,
+ "loss": 0.3820789098739624,
+ "num_input_tokens_seen": 135702632,
+ "step": 3095,
+ "train_runtime": 23868.8149,
+ "train_tokens_per_second": 5685.353
+ },
+ {
+ "epoch": 0.20872609749528684,
+ "grad_norm": 0.7998350921308711,
+ "learning_rate": 9.733828446952696e-06,
+ "loss": 0.40059328079223633,
+ "num_input_tokens_seen": 135925976,
+ "step": 3100,
+ "train_runtime": 23904.8211,
+ "train_tokens_per_second": 5686.132
+ },
+ {
+ "epoch": 0.20906275249124698,
+ "grad_norm": 0.7794995917456119,
+ "learning_rate": 9.732976592953598e-06,
+ "loss": 0.3992907524108887,
+ "num_input_tokens_seen": 136155256,
+ "step": 3105,
+ "train_runtime": 23947.0473,
+ "train_tokens_per_second": 5685.68
+ },
+ {
+ "epoch": 0.20939940748720712,
+ "grad_norm": 0.8028611006413868,
+ "learning_rate": 9.73212341539269e-06,
+ "loss": 0.40617761611938474,
+ "num_input_tokens_seen": 136377968,
+ "step": 3110,
+ "train_runtime": 23995.829,
+ "train_tokens_per_second": 5683.403
+ },
+ {
+ "epoch": 0.20973606248316726,
+ "grad_norm": 0.7679301554806387,
+ "learning_rate": 9.731268914508556e-06,
+ "loss": 0.39726805686950684,
+ "num_input_tokens_seen": 136595144,
+ "step": 3115,
+ "train_runtime": 24037.8732,
+ "train_tokens_per_second": 5682.497
+ },
+ {
+ "epoch": 0.2100727174791274,
+ "grad_norm": 0.7757940100269232,
+ "learning_rate": 9.730413090540158e-06,
+ "loss": 0.3819112777709961,
+ "num_input_tokens_seen": 136818256,
+ "step": 3120,
+ "train_runtime": 24075.9596,
+ "train_tokens_per_second": 5682.775
+ },
+ {
+ "epoch": 0.21040937247508754,
+ "grad_norm": 0.7507007143030089,
+ "learning_rate": 9.729555943726825e-06,
+ "loss": 0.40000243186950685,
+ "num_input_tokens_seen": 137031872,
+ "step": 3125,
+ "train_runtime": 24120.3223,
+ "train_tokens_per_second": 5681.179
+ },
+ {
+ "epoch": 0.21074602747104768,
+ "grad_norm": 0.8053852599963457,
+ "learning_rate": 9.728697474308253e-06,
+ "loss": 0.40857748985290526,
+ "num_input_tokens_seen": 137246200,
+ "step": 3130,
+ "train_runtime": 24159.6984,
+ "train_tokens_per_second": 5680.791
+ },
+ {
+ "epoch": 0.21108268246700782,
+ "grad_norm": 0.8178855876896101,
+ "learning_rate": 9.72783768252451e-06,
+ "loss": 0.38510382175445557,
+ "num_input_tokens_seen": 137454056,
+ "step": 3135,
+ "train_runtime": 24197.6049,
+ "train_tokens_per_second": 5680.482
+ },
+ {
+ "epoch": 0.21141933746296795,
+ "grad_norm": 0.7840878331434453,
+ "learning_rate": 9.726976568616036e-06,
+ "loss": 0.40147695541381834,
+ "num_input_tokens_seen": 137663648,
+ "step": 3140,
+ "train_runtime": 24241.393,
+ "train_tokens_per_second": 5678.867
+ },
+ {
+ "epoch": 0.2117559924589281,
+ "grad_norm": 0.9228616942450966,
+ "learning_rate": 9.726114132823638e-06,
+ "loss": 0.389466381072998,
+ "num_input_tokens_seen": 137882432,
+ "step": 3145,
+ "train_runtime": 24280.2706,
+ "train_tokens_per_second": 5678.785
+ },
+ {
+ "epoch": 0.21209264745488823,
+ "grad_norm": 0.6934203083016008,
+ "learning_rate": 9.725250375388493e-06,
+ "loss": 0.4213107109069824,
+ "num_input_tokens_seen": 138093344,
+ "step": 3150,
+ "train_runtime": 24320.3928,
+ "train_tokens_per_second": 5678.089
+ },
+ {
+ "epoch": 0.21242930245084837,
+ "grad_norm": 0.8669084820101542,
+ "learning_rate": 9.724385296552147e-06,
+ "loss": 0.43596372604370115,
+ "num_input_tokens_seen": 138316336,
+ "step": 3155,
+ "train_runtime": 24362.0648,
+ "train_tokens_per_second": 5677.529
+ },
+ {
+ "epoch": 0.2127659574468085,
+ "grad_norm": 0.9337337457982863,
+ "learning_rate": 9.72351889655652e-06,
+ "loss": 0.4109373092651367,
+ "num_input_tokens_seen": 138541888,
+ "step": 3160,
+ "train_runtime": 24397.2747,
+ "train_tokens_per_second": 5678.58
+ },
+ {
+ "epoch": 0.21310261244276865,
+ "grad_norm": 0.8342554979170712,
+ "learning_rate": 9.722651175643895e-06,
+ "loss": 0.40399460792541503,
+ "num_input_tokens_seen": 138760928,
+ "step": 3165,
+ "train_runtime": 24436.674,
+ "train_tokens_per_second": 5678.388
+ },
+ {
+ "epoch": 0.2134392674387288,
+ "grad_norm": 0.7886653586812639,
+ "learning_rate": 9.721782134056928e-06,
+ "loss": 0.36252574920654296,
+ "num_input_tokens_seen": 138982088,
+ "step": 3170,
+ "train_runtime": 24480.3769,
+ "train_tokens_per_second": 5677.285
+ },
+ {
+ "epoch": 0.21377592243468893,
+ "grad_norm": 0.7172325919553721,
+ "learning_rate": 9.720911772038645e-06,
+ "loss": 0.4204305648803711,
+ "num_input_tokens_seen": 139208968,
+ "step": 3175,
+ "train_runtime": 24514.7644,
+ "train_tokens_per_second": 5678.577
+ },
+ {
+ "epoch": 0.21411257743064907,
+ "grad_norm": 0.8184330706813875,
+ "learning_rate": 9.720040089832436e-06,
+ "loss": 0.4118339538574219,
+ "num_input_tokens_seen": 139404296,
+ "step": 3180,
+ "train_runtime": 24555.2547,
+ "train_tokens_per_second": 5677.168
+ },
+ {
+ "epoch": 0.2144492324266092,
+ "grad_norm": 0.7810238480356294,
+ "learning_rate": 9.719167087682069e-06,
+ "loss": 0.38251235485076907,
+ "num_input_tokens_seen": 139626528,
+ "step": 3185,
+ "train_runtime": 24596.095,
+ "train_tokens_per_second": 5676.776
+ },
+ {
+ "epoch": 0.21478588742256935,
+ "grad_norm": 0.8237949114328076,
+ "learning_rate": 9.718292765831673e-06,
+ "loss": 0.39778890609741213,
+ "num_input_tokens_seen": 139852040,
+ "step": 3190,
+ "train_runtime": 24636.5885,
+ "train_tokens_per_second": 5676.599
+ },
+ {
+ "epoch": 0.2151225424185295,
+ "grad_norm": 0.823181949661376,
+ "learning_rate": 9.71741712452575e-06,
+ "loss": 0.4149151802062988,
+ "num_input_tokens_seen": 140074344,
+ "step": 3195,
+ "train_runtime": 24669.1536,
+ "train_tokens_per_second": 5678.117
+ },
+ {
+ "epoch": 0.21545919741448963,
+ "grad_norm": 0.7987011374780209,
+ "learning_rate": 9.716540164009174e-06,
+ "loss": 0.3865767478942871,
+ "num_input_tokens_seen": 140277680,
+ "step": 3200,
+ "train_runtime": 24707.2868,
+ "train_tokens_per_second": 5677.583
+ },
+ {
+ "epoch": 0.21579585241044977,
+ "grad_norm": 0.9131858261391301,
+ "learning_rate": 9.715661884527178e-06,
+ "loss": 0.4173252582550049,
+ "num_input_tokens_seen": 140503936,
+ "step": 3205,
+ "train_runtime": 24750.0835,
+ "train_tokens_per_second": 5676.908
+ },
+ {
+ "epoch": 0.2161325074064099,
+ "grad_norm": 0.9283199650974979,
+ "learning_rate": 9.714782286325374e-06,
+ "loss": 0.4102613925933838,
+ "num_input_tokens_seen": 140723000,
+ "step": 3210,
+ "train_runtime": 24793.1175,
+ "train_tokens_per_second": 5675.89
+ },
+ {
+ "epoch": 0.21646916240237005,
+ "grad_norm": 0.9449783149453059,
+ "learning_rate": 9.713901369649738e-06,
+ "loss": 0.3924571990966797,
+ "num_input_tokens_seen": 140943200,
+ "step": 3215,
+ "train_runtime": 24831.5013,
+ "train_tokens_per_second": 5675.984
+ },
+ {
+ "epoch": 0.2168058173983302,
+ "grad_norm": 0.7571957619819966,
+ "learning_rate": 9.713019134746618e-06,
+ "loss": 0.38917341232299807,
+ "num_input_tokens_seen": 141178808,
+ "step": 3220,
+ "train_runtime": 24867.52,
+ "train_tokens_per_second": 5677.237
+ },
+ {
+ "epoch": 0.21714247239429033,
+ "grad_norm": 0.8569813128796843,
+ "learning_rate": 9.712135581862722e-06,
+ "loss": 0.39778847694396974,
+ "num_input_tokens_seen": 141410600,
+ "step": 3225,
+ "train_runtime": 24909.1487,
+ "train_tokens_per_second": 5677.055
+ },
+ {
+ "epoch": 0.21747912739025047,
+ "grad_norm": 0.7957771696921552,
+ "learning_rate": 9.711250711245139e-06,
+ "loss": 0.41150903701782227,
+ "num_input_tokens_seen": 141622744,
+ "step": 3230,
+ "train_runtime": 24947.3479,
+ "train_tokens_per_second": 5676.866
+ },
+ {
+ "epoch": 0.2178157823862106,
+ "grad_norm": 0.7674615378265653,
+ "learning_rate": 9.710364523141317e-06,
+ "loss": 0.4173532009124756,
+ "num_input_tokens_seen": 141845232,
+ "step": 3235,
+ "train_runtime": 24990.9712,
+ "train_tokens_per_second": 5675.859
+ },
+ {
+ "epoch": 0.21815243738217074,
+ "grad_norm": 0.881520768069259,
+ "learning_rate": 9.709477017799076e-06,
+ "loss": 0.4302227973937988,
+ "num_input_tokens_seen": 142068920,
+ "step": 3240,
+ "train_runtime": 25022.7411,
+ "train_tokens_per_second": 5677.592
+ },
+ {
+ "epoch": 0.21848909237813088,
+ "grad_norm": 0.9412217646326109,
+ "learning_rate": 9.708588195466604e-06,
+ "loss": 0.4126433849334717,
+ "num_input_tokens_seen": 142289600,
+ "step": 3245,
+ "train_runtime": 25054.9859,
+ "train_tokens_per_second": 5679.093
+ },
+ {
+ "epoch": 0.21882574737409102,
+ "grad_norm": 0.7430440131191046,
+ "learning_rate": 9.707698056392458e-06,
+ "loss": 0.4167231559753418,
+ "num_input_tokens_seen": 142506920,
+ "step": 3250,
+ "train_runtime": 25090.3796,
+ "train_tokens_per_second": 5679.743
+ },
+ {
+ "epoch": 0.21916240237005116,
+ "grad_norm": 0.8971811831202605,
+ "learning_rate": 9.706806600825562e-06,
+ "loss": 0.4131289005279541,
+ "num_input_tokens_seen": 142719992,
+ "step": 3255,
+ "train_runtime": 25133.8043,
+ "train_tokens_per_second": 5678.408
+ },
+ {
+ "epoch": 0.2194990573660113,
+ "grad_norm": 0.9964244321532095,
+ "learning_rate": 9.70591382901521e-06,
+ "loss": 0.42434329986572267,
+ "num_input_tokens_seen": 142935376,
+ "step": 3260,
+ "train_runtime": 25170.1091,
+ "train_tokens_per_second": 5678.775
+ },
+ {
+ "epoch": 0.21983571236197144,
+ "grad_norm": 0.7540863225432197,
+ "learning_rate": 9.705019741211061e-06,
+ "loss": 0.39461374282836914,
+ "num_input_tokens_seen": 143155712,
+ "step": 3265,
+ "train_runtime": 25208.8229,
+ "train_tokens_per_second": 5678.794
+ },
+ {
+ "epoch": 0.22017236735793158,
+ "grad_norm": 0.8596938023198728,
+ "learning_rate": 9.704124337663145e-06,
+ "loss": 0.38963048458099364,
+ "num_input_tokens_seen": 143381832,
+ "step": 3270,
+ "train_runtime": 25243.421,
+ "train_tokens_per_second": 5679.968
+ },
+ {
+ "epoch": 0.22050902235389172,
+ "grad_norm": 0.8540309395774371,
+ "learning_rate": 9.703227618621859e-06,
+ "loss": 0.4044978618621826,
+ "num_input_tokens_seen": 143607328,
+ "step": 3275,
+ "train_runtime": 25284.0238,
+ "train_tokens_per_second": 5679.766
+ },
+ {
+ "epoch": 0.22084567734985186,
+ "grad_norm": 0.8072539756195105,
+ "learning_rate": 9.702329584337965e-06,
+ "loss": 0.41426868438720704,
+ "num_input_tokens_seen": 143809000,
+ "step": 3280,
+ "train_runtime": 25326.637,
+ "train_tokens_per_second": 5678.172
+ },
+ {
+ "epoch": 0.221182332345812,
+ "grad_norm": 0.9295650810498607,
+ "learning_rate": 9.701430235062599e-06,
+ "loss": 0.43207788467407227,
+ "num_input_tokens_seen": 144023696,
+ "step": 3285,
+ "train_runtime": 25366.0897,
+ "train_tokens_per_second": 5677.804
+ },
+ {
+ "epoch": 0.22151898734177214,
+ "grad_norm": 0.9183012776415354,
+ "learning_rate": 9.700529571047259e-06,
+ "loss": 0.4370114803314209,
+ "num_input_tokens_seen": 144236616,
+ "step": 3290,
+ "train_runtime": 25403.6973,
+ "train_tokens_per_second": 5677.78
+ },
+ {
+ "epoch": 0.22185564233773228,
+ "grad_norm": 0.8676275172270271,
+ "learning_rate": 9.699627592543814e-06,
+ "loss": 0.3935758829116821,
+ "num_input_tokens_seen": 144451272,
+ "step": 3295,
+ "train_runtime": 25449.4742,
+ "train_tokens_per_second": 5676.002
+ },
+ {
+ "epoch": 0.22219229733369245,
+ "grad_norm": 0.7791903636055438,
+ "learning_rate": 9.698724299804498e-06,
+ "loss": 0.41451497077941896,
+ "num_input_tokens_seen": 144675240,
+ "step": 3300,
+ "train_runtime": 25489.8713,
+ "train_tokens_per_second": 5675.793
+ },
+ {
+ "epoch": 0.22252895232965259,
+ "grad_norm": 0.8026534486815711,
+ "learning_rate": 9.697819693081915e-06,
+ "loss": 0.40555124282836913,
+ "num_input_tokens_seen": 144894992,
+ "step": 3305,
+ "train_runtime": 25526.7574,
+ "train_tokens_per_second": 5676.2
+ },
+ {
+ "epoch": 0.22286560732561272,
+ "grad_norm": 0.7189128013140877,
+ "learning_rate": 9.696913772629034e-06,
+ "loss": 0.38943185806274416,
+ "num_input_tokens_seen": 145119152,
+ "step": 3310,
+ "train_runtime": 25563.6911,
+ "train_tokens_per_second": 5676.768
+ },
+ {
+ "epoch": 0.22320226232157286,
+ "grad_norm": 0.788476880680751,
+ "learning_rate": 9.696006538699195e-06,
+ "loss": 0.42933268547058107,
+ "num_input_tokens_seen": 145340576,
+ "step": 3315,
+ "train_runtime": 25595.5361,
+ "train_tokens_per_second": 5678.356
+ },
+ {
+ "epoch": 0.223538917317533,
+ "grad_norm": 0.9173858981265539,
+ "learning_rate": 9.695097991546102e-06,
+ "loss": 0.41726884841918943,
+ "num_input_tokens_seen": 145536392,
+ "step": 3320,
+ "train_runtime": 25635.2268,
+ "train_tokens_per_second": 5677.203
+ },
+ {
+ "epoch": 0.22387557231349314,
+ "grad_norm": 0.85644319812075,
+ "learning_rate": 9.694188131423828e-06,
+ "loss": 0.38661761283874513,
+ "num_input_tokens_seen": 145752592,
+ "step": 3325,
+ "train_runtime": 25675.7087,
+ "train_tokens_per_second": 5676.673
+ },
+ {
+ "epoch": 0.22421222730945328,
+ "grad_norm": 0.8283061394820891,
+ "learning_rate": 9.693276958586809e-06,
+ "loss": 0.4395936965942383,
+ "num_input_tokens_seen": 145975632,
+ "step": 3330,
+ "train_runtime": 25717.6903,
+ "train_tokens_per_second": 5676.079
+ },
+ {
+ "epoch": 0.22454888230541342,
+ "grad_norm": 1.107175434088717,
+ "learning_rate": 9.692364473289856e-06,
+ "loss": 0.37718255519866944,
+ "num_input_tokens_seen": 146190016,
+ "step": 3335,
+ "train_runtime": 25758.607,
+ "train_tokens_per_second": 5675.385
+ },
+ {
+ "epoch": 0.22488553730137356,
+ "grad_norm": 0.8144408251561523,
+ "learning_rate": 9.69145067578814e-06,
+ "loss": 0.38504073619842527,
+ "num_input_tokens_seen": 146415880,
+ "step": 3340,
+ "train_runtime": 25799.3872,
+ "train_tokens_per_second": 5675.169
+ },
+ {
+ "epoch": 0.2252221922973337,
+ "grad_norm": 0.7888334595920984,
+ "learning_rate": 9.690535566337205e-06,
+ "loss": 0.44390153884887695,
+ "num_input_tokens_seen": 146637056,
+ "step": 3345,
+ "train_runtime": 25841.4558,
+ "train_tokens_per_second": 5674.489
+ },
+ {
+ "epoch": 0.22555884729329384,
+ "grad_norm": 0.8121918469969017,
+ "learning_rate": 9.689619145192955e-06,
+ "loss": 0.4186522960662842,
+ "num_input_tokens_seen": 146858960,
+ "step": 3350,
+ "train_runtime": 25884.465,
+ "train_tokens_per_second": 5673.633
+ },
+ {
+ "epoch": 0.22589550228925398,
+ "grad_norm": 0.7357528337250359,
+ "learning_rate": 9.688701412611663e-06,
+ "loss": 0.39981067180633545,
+ "num_input_tokens_seen": 147088760,
+ "step": 3355,
+ "train_runtime": 25924.1782,
+ "train_tokens_per_second": 5673.806
+ },
+ {
+ "epoch": 0.22623215728521412,
+ "grad_norm": 0.8039258104050244,
+ "learning_rate": 9.687782368849973e-06,
+ "loss": 0.40915794372558595,
+ "num_input_tokens_seen": 147306920,
+ "step": 3360,
+ "train_runtime": 25967.0787,
+ "train_tokens_per_second": 5672.834
+ },
+ {
+ "epoch": 0.22656881228117426,
+ "grad_norm": 0.8366828009872761,
+ "learning_rate": 9.686862014164893e-06,
+ "loss": 0.3879998207092285,
+ "num_input_tokens_seen": 147534144,
+ "step": 3365,
+ "train_runtime": 26004.3358,
+ "train_tokens_per_second": 5673.444
+ },
+ {
+ "epoch": 0.2269054672771344,
+ "grad_norm": 0.8583728877321222,
+ "learning_rate": 9.685940348813795e-06,
+ "loss": 0.3802268266677856,
+ "num_input_tokens_seen": 147751544,
+ "step": 3370,
+ "train_runtime": 26042.4838,
+ "train_tokens_per_second": 5673.481
+ },
+ {
+ "epoch": 0.22724212227309454,
+ "grad_norm": 0.8211994293131717,
+ "learning_rate": 9.685017373054421e-06,
+ "loss": 0.39569883346557616,
+ "num_input_tokens_seen": 147971208,
+ "step": 3375,
+ "train_runtime": 26081.1089,
+ "train_tokens_per_second": 5673.501
+ },
+ {
+ "epoch": 0.22757877726905468,
+ "grad_norm": 0.7908688808511072,
+ "learning_rate": 9.68409308714488e-06,
+ "loss": 0.3963681697845459,
+ "num_input_tokens_seen": 148186648,
+ "step": 3380,
+ "train_runtime": 26119.5816,
+ "train_tokens_per_second": 5673.393
+ },
+ {
+ "epoch": 0.22791543226501482,
+ "grad_norm": 0.7696519906828798,
+ "learning_rate": 9.68316749134364e-06,
+ "loss": 0.4038090705871582,
+ "num_input_tokens_seen": 148417752,
+ "step": 3385,
+ "train_runtime": 26162.0325,
+ "train_tokens_per_second": 5673.021
+ },
+ {
+ "epoch": 0.22825208726097496,
+ "grad_norm": 0.7784468571247383,
+ "learning_rate": 9.682240585909545e-06,
+ "loss": 0.38401403427124026,
+ "num_input_tokens_seen": 148628520,
+ "step": 3390,
+ "train_runtime": 26201.9074,
+ "train_tokens_per_second": 5672.431
+ },
+ {
+ "epoch": 0.2285887422569351,
+ "grad_norm": 0.7682945918583601,
+ "learning_rate": 9.681312371101803e-06,
+ "loss": 0.36235072612762453,
+ "num_input_tokens_seen": 148860120,
+ "step": 3395,
+ "train_runtime": 26235.4949,
+ "train_tokens_per_second": 5673.997
+ },
+ {
+ "epoch": 0.22892539725289524,
+ "grad_norm": 0.7562252932302131,
+ "learning_rate": 9.680382847179984e-06,
+ "loss": 0.4153308868408203,
+ "num_input_tokens_seen": 149075360,
+ "step": 3400,
+ "train_runtime": 26279.0336,
+ "train_tokens_per_second": 5672.787
+ },
+ {
+ "epoch": 0.22926205224885537,
+ "grad_norm": 0.8026900227545827,
+ "learning_rate": 9.679452014404025e-06,
+ "loss": 0.39875361919403074,
+ "num_input_tokens_seen": 149283904,
+ "step": 3405,
+ "train_runtime": 26318.8542,
+ "train_tokens_per_second": 5672.128
+ },
+ {
+ "epoch": 0.22959870724481551,
+ "grad_norm": 0.8571181299293755,
+ "learning_rate": 9.678519873034234e-06,
+ "loss": 0.435106086730957,
+ "num_input_tokens_seen": 149511344,
+ "step": 3410,
+ "train_runtime": 26357.0843,
+ "train_tokens_per_second": 5672.53
+ },
+ {
+ "epoch": 0.22993536224077565,
+ "grad_norm": 0.9132428672371381,
+ "learning_rate": 9.677586423331277e-06,
+ "loss": 0.42625818252563474,
+ "num_input_tokens_seen": 149734672,
+ "step": 3415,
+ "train_runtime": 26396.1327,
+ "train_tokens_per_second": 5672.599
+ },
+ {
+ "epoch": 0.2302720172367358,
+ "grad_norm": 0.7227848772197648,
+ "learning_rate": 9.676651665556194e-06,
+ "loss": 0.3853285312652588,
+ "num_input_tokens_seen": 149966984,
+ "step": 3420,
+ "train_runtime": 26437.1674,
+ "train_tokens_per_second": 5672.581
+ },
+ {
+ "epoch": 0.23060867223269593,
+ "grad_norm": 0.9758464763691787,
+ "learning_rate": 9.675715599970383e-06,
+ "loss": 0.39388408660888674,
+ "num_input_tokens_seen": 150173312,
+ "step": 3425,
+ "train_runtime": 26476.8347,
+ "train_tokens_per_second": 5671.876
+ },
+ {
+ "epoch": 0.23094532722865607,
+ "grad_norm": 0.8789640499008409,
+ "learning_rate": 9.674778226835617e-06,
+ "loss": 0.43226184844970705,
+ "num_input_tokens_seen": 150378728,
+ "step": 3430,
+ "train_runtime": 26514.5982,
+ "train_tokens_per_second": 5671.545
+ },
+ {
+ "epoch": 0.2312819822246162,
+ "grad_norm": 0.8223781330799458,
+ "learning_rate": 9.673839546414026e-06,
+ "loss": 0.4151242733001709,
+ "num_input_tokens_seen": 150589928,
+ "step": 3435,
+ "train_runtime": 26555.5388,
+ "train_tokens_per_second": 5670.754
+ },
+ {
+ "epoch": 0.23161863722057635,
+ "grad_norm": 0.6819707841325117,
+ "learning_rate": 9.672899558968107e-06,
+ "loss": 0.39006340503692627,
+ "num_input_tokens_seen": 150802664,
+ "step": 3440,
+ "train_runtime": 26599.7743,
+ "train_tokens_per_second": 5669.321
+ },
+ {
+ "epoch": 0.2319552922165365,
+ "grad_norm": 0.8192017234463131,
+ "learning_rate": 9.67195826476073e-06,
+ "loss": 0.44136791229248046,
+ "num_input_tokens_seen": 151027880,
+ "step": 3445,
+ "train_runtime": 26637.6563,
+ "train_tokens_per_second": 5669.714
+ },
+ {
+ "epoch": 0.23229194721249663,
+ "grad_norm": 0.8137299353827566,
+ "learning_rate": 9.671015664055122e-06,
+ "loss": 0.3967195749282837,
+ "num_input_tokens_seen": 151239928,
+ "step": 3450,
+ "train_runtime": 26671.0692,
+ "train_tokens_per_second": 5670.561
+ },
+ {
+ "epoch": 0.23262860220845677,
+ "grad_norm": 0.819274273484707,
+ "learning_rate": 9.670071757114877e-06,
+ "loss": 0.4030034065246582,
+ "num_input_tokens_seen": 151459832,
+ "step": 3455,
+ "train_runtime": 26710.598,
+ "train_tokens_per_second": 5670.402
+ },
+ {
+ "epoch": 0.2329652572044169,
+ "grad_norm": 0.8052967688256787,
+ "learning_rate": 9.669126544203957e-06,
+ "loss": 0.38636322021484376,
+ "num_input_tokens_seen": 151667896,
+ "step": 3460,
+ "train_runtime": 26749.1234,
+ "train_tokens_per_second": 5670.014
+ },
+ {
+ "epoch": 0.23330191220037705,
+ "grad_norm": 0.8630240510385624,
+ "learning_rate": 9.668180025586689e-06,
+ "loss": 0.4077507972717285,
+ "num_input_tokens_seen": 151902352,
+ "step": 3465,
+ "train_runtime": 26791.3212,
+ "train_tokens_per_second": 5669.834
+ },
+ {
+ "epoch": 0.2336385671963372,
+ "grad_norm": 0.7786921906276983,
+ "learning_rate": 9.667232201527761e-06,
+ "loss": 0.3916687250137329,
+ "num_input_tokens_seen": 152102680,
+ "step": 3470,
+ "train_runtime": 26827.5433,
+ "train_tokens_per_second": 5669.646
+ },
+ {
+ "epoch": 0.23397522219229733,
+ "grad_norm": 0.8593939179148706,
+ "learning_rate": 9.666283072292234e-06,
+ "loss": 0.41549954414367674,
+ "num_input_tokens_seen": 152303592,
+ "step": 3475,
+ "train_runtime": 26864.7416,
+ "train_tokens_per_second": 5669.274
+ },
+ {
+ "epoch": 0.23431187718825747,
+ "grad_norm": 0.7173015967840941,
+ "learning_rate": 9.665332638145522e-06,
+ "loss": 0.40798091888427734,
+ "num_input_tokens_seen": 152526320,
+ "step": 3480,
+ "train_runtime": 26904.6608,
+ "train_tokens_per_second": 5669.141
+ },
+ {
+ "epoch": 0.2346485321842176,
+ "grad_norm": 0.7787780767474121,
+ "learning_rate": 9.664380899353414e-06,
+ "loss": 0.44217610359191895,
+ "num_input_tokens_seen": 152747800,
+ "step": 3485,
+ "train_runtime": 26938.7633,
+ "train_tokens_per_second": 5670.186
+ },
+ {
+ "epoch": 0.23498518718017775,
+ "grad_norm": 0.875181117161752,
+ "learning_rate": 9.663427856182062e-06,
+ "loss": 0.418949556350708,
+ "num_input_tokens_seen": 152978904,
+ "step": 3490,
+ "train_runtime": 26976.7683,
+ "train_tokens_per_second": 5670.765
+ },
+ {
+ "epoch": 0.23532184217613789,
+ "grad_norm": 0.8234151432189463,
+ "learning_rate": 9.662473508897981e-06,
+ "loss": 0.41005191802978513,
+ "num_input_tokens_seen": 153200600,
+ "step": 3495,
+ "train_runtime": 27021.1553,
+ "train_tokens_per_second": 5669.654
+ },
+ {
+ "epoch": 0.23565849717209802,
+ "grad_norm": 0.9381121440419455,
+ "learning_rate": 9.661517857768052e-06,
+ "loss": 0.43746633529663087,
+ "num_input_tokens_seen": 153413320,
+ "step": 3500,
+ "train_runtime": 27058.2447,
+ "train_tokens_per_second": 5669.744
+ },
+ {
+ "epoch": 0.23599515216805816,
+ "grad_norm": 0.9159040524887876,
+ "learning_rate": 9.660560903059517e-06,
+ "loss": 0.4074262619018555,
+ "num_input_tokens_seen": 153619928,
+ "step": 3505,
+ "train_runtime": 27099.4693,
+ "train_tokens_per_second": 5668.743
+ },
+ {
+ "epoch": 0.2363318071640183,
+ "grad_norm": 0.7088057637221193,
+ "learning_rate": 9.659602645039988e-06,
+ "loss": 0.4169961929321289,
+ "num_input_tokens_seen": 153844808,
+ "step": 3510,
+ "train_runtime": 27137.4232,
+ "train_tokens_per_second": 5669.102
+ },
+ {
+ "epoch": 0.23666846215997844,
+ "grad_norm": 0.7425326417567343,
+ "learning_rate": 9.658643083977435e-06,
+ "loss": 0.37222681045532224,
+ "num_input_tokens_seen": 154057008,
+ "step": 3515,
+ "train_runtime": 27172.217,
+ "train_tokens_per_second": 5669.652
+ },
+ {
+ "epoch": 0.23700511715593858,
+ "grad_norm": 0.7835915209475461,
+ "learning_rate": 9.657682220140202e-06,
+ "loss": 0.3870618104934692,
+ "num_input_tokens_seen": 154257952,
+ "step": 3520,
+ "train_runtime": 27216.0159,
+ "train_tokens_per_second": 5667.911
+ },
+ {
+ "epoch": 0.23734177215189872,
+ "grad_norm": 0.944898207115128,
+ "learning_rate": 9.656720053796987e-06,
+ "loss": 0.4269230365753174,
+ "num_input_tokens_seen": 154468968,
+ "step": 3525,
+ "train_runtime": 27253.774,
+ "train_tokens_per_second": 5667.801
+ },
+ {
+ "epoch": 0.23767842714785886,
+ "grad_norm": 0.964924059220392,
+ "learning_rate": 9.655756585216859e-06,
+ "loss": 0.39426770210266116,
+ "num_input_tokens_seen": 154679920,
+ "step": 3530,
+ "train_runtime": 27297.8899,
+ "train_tokens_per_second": 5666.369
+ },
+ {
+ "epoch": 0.238015082143819,
+ "grad_norm": 0.8389859916982031,
+ "learning_rate": 9.654791814669246e-06,
+ "loss": 0.41727919578552247,
+ "num_input_tokens_seen": 154894088,
+ "step": 3535,
+ "train_runtime": 27338.9009,
+ "train_tokens_per_second": 5665.703
+ },
+ {
+ "epoch": 0.23835173713977914,
+ "grad_norm": 0.7799448759874524,
+ "learning_rate": 9.653825742423949e-06,
+ "loss": 0.36776251792907716,
+ "num_input_tokens_seen": 155118416,
+ "step": 3540,
+ "train_runtime": 27380.1523,
+ "train_tokens_per_second": 5665.36
+ },
+ {
+ "epoch": 0.2386883921357393,
+ "grad_norm": 0.7611172988563449,
+ "learning_rate": 9.652858368751118e-06,
+ "loss": 0.38267481327056885,
+ "num_input_tokens_seen": 155325944,
+ "step": 3545,
+ "train_runtime": 27419.2328,
+ "train_tokens_per_second": 5664.854
+ },
+ {
+ "epoch": 0.23902504713169945,
+ "grad_norm": 0.849616732171608,
+ "learning_rate": 9.651889693921284e-06,
+ "loss": 0.39499473571777344,
+ "num_input_tokens_seen": 155553936,
+ "step": 3550,
+ "train_runtime": 27453.0014,
+ "train_tokens_per_second": 5666.191
+ },
+ {
+ "epoch": 0.2393617021276596,
+ "grad_norm": 0.8290842696777342,
+ "learning_rate": 9.650919718205331e-06,
+ "loss": 0.46364564895629884,
+ "num_input_tokens_seen": 155767440,
+ "step": 3555,
+ "train_runtime": 27489.7425,
+ "train_tokens_per_second": 5666.384
+ },
+ {
+ "epoch": 0.23969835712361973,
+ "grad_norm": 0.6972330732061462,
+ "learning_rate": 9.649948441874509e-06,
+ "loss": 0.41176652908325195,
+ "num_input_tokens_seen": 155991432,
+ "step": 3560,
+ "train_runtime": 27533.7281,
+ "train_tokens_per_second": 5665.467
+ },
+ {
+ "epoch": 0.24003501211957987,
+ "grad_norm": 0.8232987599322095,
+ "learning_rate": 9.648975865200434e-06,
+ "loss": 0.3719303607940674,
+ "num_input_tokens_seen": 156200840,
+ "step": 3565,
+ "train_runtime": 27572.2717,
+ "train_tokens_per_second": 5665.142
+ },
+ {
+ "epoch": 0.24037166711554,
+ "grad_norm": 0.7784058710951964,
+ "learning_rate": 9.648001988455082e-06,
+ "loss": 0.42696056365966795,
+ "num_input_tokens_seen": 156423232,
+ "step": 3570,
+ "train_runtime": 27611.0095,
+ "train_tokens_per_second": 5665.249
+ },
+ {
+ "epoch": 0.24070832211150014,
+ "grad_norm": 0.8101799377558996,
+ "learning_rate": 9.647026811910794e-06,
+ "loss": 0.42109375,
+ "num_input_tokens_seen": 156638144,
+ "step": 3575,
+ "train_runtime": 27652.042,
+ "train_tokens_per_second": 5664.614
+ },
+ {
+ "epoch": 0.24104497710746028,
+ "grad_norm": 0.7776435744711748,
+ "learning_rate": 9.646050335840279e-06,
+ "loss": 0.440277624130249,
+ "num_input_tokens_seen": 156867440,
+ "step": 3580,
+ "train_runtime": 27693.6123,
+ "train_tokens_per_second": 5664.391
+ },
+ {
+ "epoch": 0.24138163210342042,
+ "grad_norm": 0.8251722449916711,
+ "learning_rate": 9.6450725605166e-06,
+ "loss": 0.4013186454772949,
+ "num_input_tokens_seen": 157079728,
+ "step": 3585,
+ "train_runtime": 27726.0789,
+ "train_tokens_per_second": 5665.414
+ },
+ {
+ "epoch": 0.24171828709938056,
+ "grad_norm": 0.8433224749813183,
+ "learning_rate": 9.644093486213192e-06,
+ "loss": 0.4329217910766602,
+ "num_input_tokens_seen": 157312856,
+ "step": 3590,
+ "train_runtime": 27767.0627,
+ "train_tokens_per_second": 5665.448
+ },
+ {
+ "epoch": 0.2420549420953407,
+ "grad_norm": 0.8662917363975451,
+ "learning_rate": 9.64311311320385e-06,
+ "loss": 0.4125063896179199,
+ "num_input_tokens_seen": 157536824,
+ "step": 3595,
+ "train_runtime": 27801.6779,
+ "train_tokens_per_second": 5666.45
+ },
+ {
+ "epoch": 0.24239159709130084,
+ "grad_norm": 0.8971134461583352,
+ "learning_rate": 9.642131441762732e-06,
+ "loss": 0.39645483493804934,
+ "num_input_tokens_seen": 157744672,
+ "step": 3600,
+ "train_runtime": 27838.8811,
+ "train_tokens_per_second": 5666.344
+ },
+ {
+ "epoch": 0.24272825208726098,
+ "grad_norm": 0.8431994851395936,
+ "learning_rate": 9.641148472164359e-06,
+ "loss": 0.3727452516555786,
+ "num_input_tokens_seen": 157957176,
+ "step": 3605,
+ "train_runtime": 27875.6937,
+ "train_tokens_per_second": 5666.484
+ },
+ {
+ "epoch": 0.24306490708322112,
+ "grad_norm": 0.8088195809393041,
+ "learning_rate": 9.640164204683616e-06,
+ "loss": 0.39656991958618165,
+ "num_input_tokens_seen": 158185920,
+ "step": 3610,
+ "train_runtime": 27916.062,
+ "train_tokens_per_second": 5666.484
+ },
+ {
+ "epoch": 0.24340156207918126,
+ "grad_norm": 0.8326446982063453,
+ "learning_rate": 9.639178639595748e-06,
+ "loss": 0.43434457778930663,
+ "num_input_tokens_seen": 158410048,
+ "step": 3615,
+ "train_runtime": 27944.4625,
+ "train_tokens_per_second": 5668.746
+ },
+ {
+ "epoch": 0.2437382170751414,
+ "grad_norm": 0.7829852555366933,
+ "learning_rate": 9.638191777176367e-06,
+ "loss": 0.40036301612854003,
+ "num_input_tokens_seen": 158632248,
+ "step": 3620,
+ "train_runtime": 27978.7949,
+ "train_tokens_per_second": 5669.731
+ },
+ {
+ "epoch": 0.24407487207110154,
+ "grad_norm": 0.7554602216351339,
+ "learning_rate": 9.637203617701446e-06,
+ "loss": 0.41867575645446775,
+ "num_input_tokens_seen": 158861832,
+ "step": 3625,
+ "train_runtime": 28021.0376,
+ "train_tokens_per_second": 5669.377
+ },
+ {
+ "epoch": 0.24441152706706168,
+ "grad_norm": 0.7700724836949714,
+ "learning_rate": 9.63621416144732e-06,
+ "loss": 0.3974162578582764,
+ "num_input_tokens_seen": 159076104,
+ "step": 3630,
+ "train_runtime": 28057.1542,
+ "train_tokens_per_second": 5669.716
+ },
+ {
+ "epoch": 0.24474818206302182,
+ "grad_norm": 0.819901947145382,
+ "learning_rate": 9.635223408690688e-06,
+ "loss": 0.39289240837097167,
+ "num_input_tokens_seen": 159275688,
+ "step": 3635,
+ "train_runtime": 28092.5481,
+ "train_tokens_per_second": 5669.678
+ },
+ {
+ "epoch": 0.24508483705898196,
+ "grad_norm": 0.74792422760683,
+ "learning_rate": 9.63423135970861e-06,
+ "loss": 0.4178638935089111,
+ "num_input_tokens_seen": 159506232,
+ "step": 3640,
+ "train_runtime": 28131.9808,
+ "train_tokens_per_second": 5669.925
+ },
+ {
+ "epoch": 0.2454214920549421,
+ "grad_norm": 0.8690926777352274,
+ "learning_rate": 9.633238014778511e-06,
+ "loss": 0.4228395462036133,
+ "num_input_tokens_seen": 159735712,
+ "step": 3645,
+ "train_runtime": 28164.6263,
+ "train_tokens_per_second": 5671.501
+ },
+ {
+ "epoch": 0.24575814705090224,
+ "grad_norm": 0.8408876506004788,
+ "learning_rate": 9.632243374178173e-06,
+ "loss": 0.3977930784225464,
+ "num_input_tokens_seen": 159959000,
+ "step": 3650,
+ "train_runtime": 28206.3804,
+ "train_tokens_per_second": 5671.022
+ },
+ {
+ "epoch": 0.24609480204686238,
+ "grad_norm": 0.8178577594154023,
+ "learning_rate": 9.63124743818575e-06,
+ "loss": 0.42597079277038574,
+ "num_input_tokens_seen": 160183304,
+ "step": 3655,
+ "train_runtime": 28242.8523,
+ "train_tokens_per_second": 5671.64
+ },
+ {
+ "epoch": 0.24643145704282252,
+ "grad_norm": 0.7174819842350753,
+ "learning_rate": 9.630250207079745e-06,
+ "loss": 0.385437536239624,
+ "num_input_tokens_seen": 160396208,
+ "step": 3660,
+ "train_runtime": 28279.1927,
+ "train_tokens_per_second": 5671.881
+ },
+ {
+ "epoch": 0.24676811203878266,
+ "grad_norm": 0.9558291181876568,
+ "learning_rate": 9.629251681139038e-06,
+ "loss": 0.41692662239074707,
+ "num_input_tokens_seen": 160619648,
+ "step": 3665,
+ "train_runtime": 28315.8408,
+ "train_tokens_per_second": 5672.431
+ },
+ {
+ "epoch": 0.2471047670347428,
+ "grad_norm": 0.7393493036785912,
+ "learning_rate": 9.628251860642857e-06,
+ "loss": 0.40501389503479,
+ "num_input_tokens_seen": 160832936,
+ "step": 3670,
+ "train_runtime": 28358.3536,
+ "train_tokens_per_second": 5671.448
+ },
+ {
+ "epoch": 0.24744142203070293,
+ "grad_norm": 1.3913304565406266,
+ "learning_rate": 9.627250745870805e-06,
+ "loss": 0.4110020637512207,
+ "num_input_tokens_seen": 161045560,
+ "step": 3675,
+ "train_runtime": 28395.0175,
+ "train_tokens_per_second": 5671.613
+ },
+ {
+ "epoch": 0.24777807702666307,
+ "grad_norm": 0.7989443375204172,
+ "learning_rate": 9.626248337102834e-06,
+ "loss": 0.4464590072631836,
+ "num_input_tokens_seen": 161267264,
+ "step": 3680,
+ "train_runtime": 28433.9863,
+ "train_tokens_per_second": 5671.638
+ },
+ {
+ "epoch": 0.2481147320226232,
+ "grad_norm": 0.8100743623465356,
+ "learning_rate": 9.625244634619269e-06,
+ "loss": 0.42216014862060547,
+ "num_input_tokens_seen": 161484928,
+ "step": 3685,
+ "train_runtime": 28470.6875,
+ "train_tokens_per_second": 5671.971
+ },
+ {
+ "epoch": 0.24845138701858335,
+ "grad_norm": 0.8140092686493631,
+ "learning_rate": 9.62423963870079e-06,
+ "loss": 0.4016411781311035,
+ "num_input_tokens_seen": 161695760,
+ "step": 3690,
+ "train_runtime": 28513.9311,
+ "train_tokens_per_second": 5670.764
+ },
+ {
+ "epoch": 0.2487880420145435,
+ "grad_norm": 0.7547948701222302,
+ "learning_rate": 9.623233349628441e-06,
+ "loss": 0.36588501930236816,
+ "num_input_tokens_seen": 161906344,
+ "step": 3695,
+ "train_runtime": 28554.4813,
+ "train_tokens_per_second": 5670.085
+ },
+ {
+ "epoch": 0.24912469701050363,
+ "grad_norm": 0.7855444375692547,
+ "learning_rate": 9.62222576768363e-06,
+ "loss": 0.37874155044555663,
+ "num_input_tokens_seen": 162121952,
+ "step": 3700,
+ "train_runtime": 28593.7113,
+ "train_tokens_per_second": 5669.846
+ },
+ {
+ "epoch": 0.24946135200646377,
+ "grad_norm": 0.8013508326751752,
+ "learning_rate": 9.62121689314812e-06,
+ "loss": 0.4508801460266113,
+ "num_input_tokens_seen": 162339592,
+ "step": 3705,
+ "train_runtime": 28637.9282,
+ "train_tokens_per_second": 5668.692
+ },
+ {
+ "epoch": 0.2497980070024239,
+ "grad_norm": 0.8341179639748463,
+ "learning_rate": 9.620206726304045e-06,
+ "loss": 0.4121876239776611,
+ "num_input_tokens_seen": 162551488,
+ "step": 3710,
+ "train_runtime": 28677.3311,
+ "train_tokens_per_second": 5668.292
+ },
+ {
+ "epoch": 0.2501346619983841,
+ "grad_norm": 0.7606955484077755,
+ "learning_rate": 9.61919526743389e-06,
+ "loss": 0.4007285118103027,
+ "num_input_tokens_seen": 162775296,
+ "step": 3715,
+ "train_runtime": 28717.1122,
+ "train_tokens_per_second": 5668.233
+ },
+ {
+ "epoch": 0.2504713169943442,
+ "grad_norm": 0.7586767868766064,
+ "learning_rate": 9.618182516820507e-06,
+ "loss": 0.4169614315032959,
+ "num_input_tokens_seen": 163004008,
+ "step": 3720,
+ "train_runtime": 28752.7905,
+ "train_tokens_per_second": 5669.154
+ },
+ {
+ "epoch": 0.25080797199030436,
+ "grad_norm": 0.7958372407148936,
+ "learning_rate": 9.61716847474711e-06,
+ "loss": 0.38545975685119627,
+ "num_input_tokens_seen": 163216408,
+ "step": 3725,
+ "train_runtime": 28788.1056,
+ "train_tokens_per_second": 5669.578
+ },
+ {
+ "epoch": 0.2511446269862645,
+ "grad_norm": 0.7821165797023588,
+ "learning_rate": 9.616153141497273e-06,
+ "loss": 0.41805257797241213,
+ "num_input_tokens_seen": 163430088,
+ "step": 3730,
+ "train_runtime": 28827.4553,
+ "train_tokens_per_second": 5669.251
+ },
+ {
+ "epoch": 0.25148128198222464,
+ "grad_norm": 0.8340007896342425,
+ "learning_rate": 9.61513651735493e-06,
+ "loss": 0.37319297790527345,
+ "num_input_tokens_seen": 163642328,
+ "step": 3735,
+ "train_runtime": 28865.8432,
+ "train_tokens_per_second": 5669.065
+ },
+ {
+ "epoch": 0.2518179369781848,
+ "grad_norm": 0.8830424913719485,
+ "learning_rate": 9.614118602604376e-06,
+ "loss": 0.3945745944976807,
+ "num_input_tokens_seen": 163838104,
+ "step": 3740,
+ "train_runtime": 28903.991,
+ "train_tokens_per_second": 5668.356
+ },
+ {
+ "epoch": 0.2521545919741449,
+ "grad_norm": 0.8052297510787393,
+ "learning_rate": 9.613099397530267e-06,
+ "loss": 0.3722787380218506,
+ "num_input_tokens_seen": 164064200,
+ "step": 3745,
+ "train_runtime": 28947.9133,
+ "train_tokens_per_second": 5667.566
+ },
+ {
+ "epoch": 0.25249124697010505,
+ "grad_norm": 0.7862742185087508,
+ "learning_rate": 9.612078902417623e-06,
+ "loss": 0.4092113018035889,
+ "num_input_tokens_seen": 164293832,
+ "step": 3750,
+ "train_runtime": 28984.6022,
+ "train_tokens_per_second": 5668.314
+ },
+ {
+ "epoch": 0.2528279019660652,
+ "grad_norm": 0.917777159281807,
+ "learning_rate": 9.611057117551821e-06,
+ "loss": 0.4049520492553711,
+ "num_input_tokens_seen": 164494120,
+ "step": 3755,
+ "train_runtime": 29022.09,
+ "train_tokens_per_second": 5667.894
+ },
+ {
+ "epoch": 0.25316455696202533,
+ "grad_norm": 0.7487164986896401,
+ "learning_rate": 9.610034043218598e-06,
+ "loss": 0.41929073333740235,
+ "num_input_tokens_seen": 164721320,
+ "step": 3760,
+ "train_runtime": 29064.6791,
+ "train_tokens_per_second": 5667.405
+ },
+ {
+ "epoch": 0.2535012119579855,
+ "grad_norm": 0.8821854609765899,
+ "learning_rate": 9.609009679704057e-06,
+ "loss": 0.4089076995849609,
+ "num_input_tokens_seen": 164941328,
+ "step": 3765,
+ "train_runtime": 29100.0265,
+ "train_tokens_per_second": 5668.082
+ },
+ {
+ "epoch": 0.2538378669539456,
+ "grad_norm": 0.8451569510904378,
+ "learning_rate": 9.607984027294654e-06,
+ "loss": 0.4017925262451172,
+ "num_input_tokens_seen": 165148144,
+ "step": 3770,
+ "train_runtime": 29142.825,
+ "train_tokens_per_second": 5666.854
+ },
+ {
+ "epoch": 0.25417452194990575,
+ "grad_norm": 0.7139702054990582,
+ "learning_rate": 9.606957086277213e-06,
+ "loss": 0.4037941455841064,
+ "num_input_tokens_seen": 165372064,
+ "step": 3775,
+ "train_runtime": 29183.2104,
+ "train_tokens_per_second": 5666.685
+ },
+ {
+ "epoch": 0.2545111769458659,
+ "grad_norm": 0.8399965732594856,
+ "learning_rate": 9.605928856938913e-06,
+ "loss": 0.4064283847808838,
+ "num_input_tokens_seen": 165589000,
+ "step": 3780,
+ "train_runtime": 29221.4526,
+ "train_tokens_per_second": 5666.693
+ },
+ {
+ "epoch": 0.25484783194182603,
+ "grad_norm": 0.7840999876703054,
+ "learning_rate": 9.604899339567298e-06,
+ "loss": 0.3694040060043335,
+ "num_input_tokens_seen": 165806496,
+ "step": 3785,
+ "train_runtime": 29262.2531,
+ "train_tokens_per_second": 5666.225
+ },
+ {
+ "epoch": 0.25518448693778617,
+ "grad_norm": 0.8254996088572744,
+ "learning_rate": 9.603868534450263e-06,
+ "loss": 0.370687198638916,
+ "num_input_tokens_seen": 166040152,
+ "step": 3790,
+ "train_runtime": 29297.3995,
+ "train_tokens_per_second": 5667.402
+ },
+ {
+ "epoch": 0.2555211419337463,
+ "grad_norm": 0.8205302355074221,
+ "learning_rate": 9.602836441876074e-06,
+ "loss": 0.3572968006134033,
+ "num_input_tokens_seen": 166264712,
+ "step": 3795,
+ "train_runtime": 29337.7818,
+ "train_tokens_per_second": 5667.256
+ },
+ {
+ "epoch": 0.25585779692970645,
+ "grad_norm": 0.9811797993807015,
+ "learning_rate": 9.601803062133353e-06,
+ "loss": 0.3995164394378662,
+ "num_input_tokens_seen": 166486544,
+ "step": 3800,
+ "train_runtime": 29381.5758,
+ "train_tokens_per_second": 5666.359
+ },
+ {
+ "epoch": 0.2561944519256666,
+ "grad_norm": 0.7702336235429141,
+ "learning_rate": 9.600768395511078e-06,
+ "loss": 0.3709728717803955,
+ "num_input_tokens_seen": 166705432,
+ "step": 3805,
+ "train_runtime": 29417.1853,
+ "train_tokens_per_second": 5666.94
+ },
+ {
+ "epoch": 0.25653110692162673,
+ "grad_norm": 0.8067121262918427,
+ "learning_rate": 9.599732442298592e-06,
+ "loss": 0.39358603954315186,
+ "num_input_tokens_seen": 166928072,
+ "step": 3810,
+ "train_runtime": 29453.7851,
+ "train_tokens_per_second": 5667.457
+ },
+ {
+ "epoch": 0.25686776191758687,
+ "grad_norm": 0.8388091751069785,
+ "learning_rate": 9.598695202785596e-06,
+ "loss": 0.3980567216873169,
+ "num_input_tokens_seen": 167135336,
+ "step": 3815,
+ "train_runtime": 29489.9025,
+ "train_tokens_per_second": 5667.545
+ },
+ {
+ "epoch": 0.257204416913547,
+ "grad_norm": 0.7655332462464155,
+ "learning_rate": 9.59765667726215e-06,
+ "loss": 0.3841102600097656,
+ "num_input_tokens_seen": 167356664,
+ "step": 3820,
+ "train_runtime": 29526.6684,
+ "train_tokens_per_second": 5667.983
+ },
+ {
+ "epoch": 0.25754107190950715,
+ "grad_norm": 0.8418059528589851,
+ "learning_rate": 9.596616866018677e-06,
+ "loss": 0.46039524078369143,
+ "num_input_tokens_seen": 167568600,
+ "step": 3825,
+ "train_runtime": 29568.9722,
+ "train_tokens_per_second": 5667.042
+ },
+ {
+ "epoch": 0.2578777269054673,
+ "grad_norm": 0.87917393714771,
+ "learning_rate": 9.595575769345955e-06,
+ "loss": 0.3880587100982666,
+ "num_input_tokens_seen": 167766736,
+ "step": 3830,
+ "train_runtime": 29608.4677,
+ "train_tokens_per_second": 5666.174
+ },
+ {
+ "epoch": 0.2582143819014274,
+ "grad_norm": 0.7766137760871757,
+ "learning_rate": 9.594533387535122e-06,
+ "loss": 0.41129217147827146,
+ "num_input_tokens_seen": 167995600,
+ "step": 3835,
+ "train_runtime": 29647.1244,
+ "train_tokens_per_second": 5666.506
+ },
+ {
+ "epoch": 0.25855103689738757,
+ "grad_norm": 0.7405296564338093,
+ "learning_rate": 9.593489720877678e-06,
+ "loss": 0.4008028984069824,
+ "num_input_tokens_seen": 168214488,
+ "step": 3840,
+ "train_runtime": 29683.6625,
+ "train_tokens_per_second": 5666.905
+ },
+ {
+ "epoch": 0.2588876918933477,
+ "grad_norm": 0.8634330150776178,
+ "learning_rate": 9.592444769665482e-06,
+ "loss": 0.37677698135375975,
+ "num_input_tokens_seen": 168446672,
+ "step": 3845,
+ "train_runtime": 29726.006,
+ "train_tokens_per_second": 5666.643
+ },
+ {
+ "epoch": 0.25922434688930784,
+ "grad_norm": 0.967218365398896,
+ "learning_rate": 9.59139853419075e-06,
+ "loss": 0.3776779413223267,
+ "num_input_tokens_seen": 168649368,
+ "step": 3850,
+ "train_runtime": 29767.8639,
+ "train_tokens_per_second": 5665.484
+ },
+ {
+ "epoch": 0.259561001885268,
+ "grad_norm": 0.7776571735726057,
+ "learning_rate": 9.590351014746059e-06,
+ "loss": 0.40406084060668945,
+ "num_input_tokens_seen": 168875024,
+ "step": 3855,
+ "train_runtime": 29804.1223,
+ "train_tokens_per_second": 5666.163
+ },
+ {
+ "epoch": 0.2598976568812281,
+ "grad_norm": 1.1756300159580646,
+ "learning_rate": 9.589302211624343e-06,
+ "loss": 0.39687304496765136,
+ "num_input_tokens_seen": 169090984,
+ "step": 3860,
+ "train_runtime": 29840.2711,
+ "train_tokens_per_second": 5666.536
+ },
+ {
+ "epoch": 0.26023431187718826,
+ "grad_norm": 0.819456740850995,
+ "learning_rate": 9.5882521251189e-06,
+ "loss": 0.3968550682067871,
+ "num_input_tokens_seen": 169310120,
+ "step": 3865,
+ "train_runtime": 29877.9357,
+ "train_tokens_per_second": 5666.728
+ },
+ {
+ "epoch": 0.2605709668731484,
+ "grad_norm": 0.8515658291218985,
+ "learning_rate": 9.58720075552338e-06,
+ "loss": 0.42056074142456057,
+ "num_input_tokens_seen": 169524504,
+ "step": 3870,
+ "train_runtime": 29909.434,
+ "train_tokens_per_second": 5667.928
+ },
+ {
+ "epoch": 0.26090762186910854,
+ "grad_norm": 0.8783050270072679,
+ "learning_rate": 9.586148103131797e-06,
+ "loss": 0.40169625282287597,
+ "num_input_tokens_seen": 169733248,
+ "step": 3875,
+ "train_runtime": 29944.6725,
+ "train_tokens_per_second": 5668.229
+ },
+ {
+ "epoch": 0.2612442768650687,
+ "grad_norm": 0.857473518999431,
+ "learning_rate": 9.585094168238518e-06,
+ "loss": 0.4272635459899902,
+ "num_input_tokens_seen": 169943480,
+ "step": 3880,
+ "train_runtime": 29983.9133,
+ "train_tokens_per_second": 5667.822
+ },
+ {
+ "epoch": 0.2615809318610288,
+ "grad_norm": 0.8162084475991911,
+ "learning_rate": 9.58403895113828e-06,
+ "loss": 0.39615912437438966,
+ "num_input_tokens_seen": 170164936,
+ "step": 3885,
+ "train_runtime": 30023.2923,
+ "train_tokens_per_second": 5667.764
+ },
+ {
+ "epoch": 0.26191758685698896,
+ "grad_norm": 0.773095299777449,
+ "learning_rate": 9.582982452126164e-06,
+ "loss": 0.380422306060791,
+ "num_input_tokens_seen": 170380728,
+ "step": 3890,
+ "train_runtime": 30061.2728,
+ "train_tokens_per_second": 5667.782
+ },
+ {
+ "epoch": 0.2622542418529491,
+ "grad_norm": 0.7250445142657057,
+ "learning_rate": 9.58192467149762e-06,
+ "loss": 0.3720268726348877,
+ "num_input_tokens_seen": 170591568,
+ "step": 3895,
+ "train_runtime": 30102.8751,
+ "train_tokens_per_second": 5666.953
+ },
+ {
+ "epoch": 0.26259089684890924,
+ "grad_norm": 0.8455822434092976,
+ "learning_rate": 9.580865609548452e-06,
+ "loss": 0.4109041213989258,
+ "num_input_tokens_seen": 170793416,
+ "step": 3900,
+ "train_runtime": 30139.1143,
+ "train_tokens_per_second": 5666.836
+ },
+ {
+ "epoch": 0.2629275518448694,
+ "grad_norm": 0.669931714570705,
+ "learning_rate": 9.579805266574824e-06,
+ "loss": 0.3842399835586548,
+ "num_input_tokens_seen": 171006328,
+ "step": 3905,
+ "train_runtime": 30178.5838,
+ "train_tokens_per_second": 5666.48
+ },
+ {
+ "epoch": 0.2632642068408295,
+ "grad_norm": 0.6870277390309487,
+ "learning_rate": 9.578743642873258e-06,
+ "loss": 0.40047216415405273,
+ "num_input_tokens_seen": 171209880,
+ "step": 3910,
+ "train_runtime": 30217.8046,
+ "train_tokens_per_second": 5665.861
+ },
+ {
+ "epoch": 0.26360086183678966,
+ "grad_norm": 0.7479204866431858,
+ "learning_rate": 9.57768073874063e-06,
+ "loss": 0.40533103942871096,
+ "num_input_tokens_seen": 171434584,
+ "step": 3915,
+ "train_runtime": 30250.4257,
+ "train_tokens_per_second": 5667.179
+ },
+ {
+ "epoch": 0.2639375168327498,
+ "grad_norm": 0.822857586143697,
+ "learning_rate": 9.576616554474182e-06,
+ "loss": 0.41611337661743164,
+ "num_input_tokens_seen": 171637768,
+ "step": 3920,
+ "train_runtime": 30287.808,
+ "train_tokens_per_second": 5666.893
+ },
+ {
+ "epoch": 0.26427417182870994,
+ "grad_norm": 0.7335208763069258,
+ "learning_rate": 9.575551090371512e-06,
+ "loss": 0.39514660835266113,
+ "num_input_tokens_seen": 171845072,
+ "step": 3925,
+ "train_runtime": 30328.6771,
+ "train_tokens_per_second": 5666.092
+ },
+ {
+ "epoch": 0.2646108268246701,
+ "grad_norm": 0.7927281370222846,
+ "learning_rate": 9.574484346730564e-06,
+ "loss": 0.4031528472900391,
+ "num_input_tokens_seen": 172066528,
+ "step": 3930,
+ "train_runtime": 30370.1573,
+ "train_tokens_per_second": 5665.645
+ },
+ {
+ "epoch": 0.2649474818206302,
+ "grad_norm": 0.7830404478638062,
+ "learning_rate": 9.573416323849658e-06,
+ "loss": 0.4126886367797852,
+ "num_input_tokens_seen": 172285504,
+ "step": 3935,
+ "train_runtime": 30411.2446,
+ "train_tokens_per_second": 5665.191
+ },
+ {
+ "epoch": 0.26528413681659035,
+ "grad_norm": 0.7639494210600385,
+ "learning_rate": 9.572347022027461e-06,
+ "loss": 0.3947685718536377,
+ "num_input_tokens_seen": 172512448,
+ "step": 3940,
+ "train_runtime": 30452.9824,
+ "train_tokens_per_second": 5664.879
+ },
+ {
+ "epoch": 0.2656207918125505,
+ "grad_norm": 0.7465504294757535,
+ "learning_rate": 9.571276441563e-06,
+ "loss": 0.40018043518066404,
+ "num_input_tokens_seen": 172712640,
+ "step": 3945,
+ "train_runtime": 30484.642,
+ "train_tokens_per_second": 5665.562
+ },
+ {
+ "epoch": 0.26595744680851063,
+ "grad_norm": 0.79021748664721,
+ "learning_rate": 9.570204582755659e-06,
+ "loss": 0.38366117477416994,
+ "num_input_tokens_seen": 172934512,
+ "step": 3950,
+ "train_runtime": 30524.2614,
+ "train_tokens_per_second": 5665.477
+ },
+ {
+ "epoch": 0.2662941018044708,
+ "grad_norm": 0.9006277156621565,
+ "learning_rate": 9.569131445905178e-06,
+ "loss": 0.4079860210418701,
+ "num_input_tokens_seen": 173148536,
+ "step": 3955,
+ "train_runtime": 30557.4244,
+ "train_tokens_per_second": 5666.333
+ },
+ {
+ "epoch": 0.2666307568004309,
+ "grad_norm": 0.8299161339057768,
+ "learning_rate": 9.568057031311659e-06,
+ "loss": 0.41454520225524905,
+ "num_input_tokens_seen": 173357456,
+ "step": 3960,
+ "train_runtime": 30592.1621,
+ "train_tokens_per_second": 5666.728
+ },
+ {
+ "epoch": 0.26696741179639105,
+ "grad_norm": 0.9401299888719756,
+ "learning_rate": 9.566981339275557e-06,
+ "loss": 0.43401298522949217,
+ "num_input_tokens_seen": 173565960,
+ "step": 3965,
+ "train_runtime": 30636.8629,
+ "train_tokens_per_second": 5665.265
+ },
+ {
+ "epoch": 0.2673040667923512,
+ "grad_norm": 0.7744089765372347,
+ "learning_rate": 9.565904370097687e-06,
+ "loss": 0.4027921199798584,
+ "num_input_tokens_seen": 173765184,
+ "step": 3970,
+ "train_runtime": 30675.2756,
+ "train_tokens_per_second": 5664.666
+ },
+ {
+ "epoch": 0.26764072178831133,
+ "grad_norm": 1.1387875256699147,
+ "learning_rate": 9.564826124079219e-06,
+ "loss": 0.3860454082489014,
+ "num_input_tokens_seen": 173982664,
+ "step": 3975,
+ "train_runtime": 30722.6571,
+ "train_tokens_per_second": 5663.008
+ },
+ {
+ "epoch": 0.26797737678427147,
+ "grad_norm": 0.7717245266267511,
+ "learning_rate": 9.563746601521681e-06,
+ "loss": 0.37142066955566405,
+ "num_input_tokens_seen": 174207720,
+ "step": 3980,
+ "train_runtime": 30764.4205,
+ "train_tokens_per_second": 5662.636
+ },
+ {
+ "epoch": 0.2683140317802316,
+ "grad_norm": 0.8633313009939835,
+ "learning_rate": 9.562665802726961e-06,
+ "loss": 0.39812779426574707,
+ "num_input_tokens_seen": 174437200,
+ "step": 3985,
+ "train_runtime": 30809.6747,
+ "train_tokens_per_second": 5661.767
+ },
+ {
+ "epoch": 0.26865068677619175,
+ "grad_norm": 0.79235769535237,
+ "learning_rate": 9.561583727997297e-06,
+ "loss": 0.3973848819732666,
+ "num_input_tokens_seen": 174649000,
+ "step": 3990,
+ "train_runtime": 30852.8543,
+ "train_tokens_per_second": 5660.708
+ },
+ {
+ "epoch": 0.2689873417721519,
+ "grad_norm": 0.866887420036326,
+ "learning_rate": 9.560500377635288e-06,
+ "loss": 0.4132582664489746,
+ "num_input_tokens_seen": 174869968,
+ "step": 3995,
+ "train_runtime": 30894.8991,
+ "train_tokens_per_second": 5660.157
+ },
+ {
+ "epoch": 0.26932399676811203,
+ "grad_norm": 0.8490880534158571,
+ "learning_rate": 9.559415751943893e-06,
+ "loss": 0.4121055603027344,
+ "num_input_tokens_seen": 175087088,
+ "step": 4000,
+ "train_runtime": 30928.2656,
+ "train_tokens_per_second": 5661.07
+ },
+ {
+ "epoch": 0.26966065176407217,
+ "grad_norm": 0.7247750937341004,
+ "learning_rate": 9.558329851226422e-06,
+ "loss": 0.3906848430633545,
+ "num_input_tokens_seen": 175314216,
+ "step": 4005,
+ "train_runtime": 30967.3219,
+ "train_tokens_per_second": 5661.265
+ },
+ {
+ "epoch": 0.2699973067600323,
+ "grad_norm": 0.8575764814098764,
+ "learning_rate": 9.557242675786542e-06,
+ "loss": 0.4193889141082764,
+ "num_input_tokens_seen": 175541064,
+ "step": 4010,
+ "train_runtime": 31013.366,
+ "train_tokens_per_second": 5660.175
+ },
+ {
+ "epoch": 0.27033396175599245,
+ "grad_norm": 0.7331252032970815,
+ "learning_rate": 9.55615422592828e-06,
+ "loss": 0.4430998802185059,
+ "num_input_tokens_seen": 175770592,
+ "step": 4015,
+ "train_runtime": 31053.1239,
+ "train_tokens_per_second": 5660.319
+ },
+ {
+ "epoch": 0.2706706167519526,
+ "grad_norm": 0.8233208505066181,
+ "learning_rate": 9.555064501956018e-06,
+ "loss": 0.4185193061828613,
+ "num_input_tokens_seen": 175988096,
+ "step": 4020,
+ "train_runtime": 31090.813,
+ "train_tokens_per_second": 5660.453
+ },
+ {
+ "epoch": 0.2710072717479127,
+ "grad_norm": 0.7929784850811431,
+ "learning_rate": 9.553973504174493e-06,
+ "loss": 0.3700920581817627,
+ "num_input_tokens_seen": 176213488,
+ "step": 4025,
+ "train_runtime": 31131.4846,
+ "train_tokens_per_second": 5660.298
+ },
+ {
+ "epoch": 0.27134392674387287,
+ "grad_norm": 0.8802925347262501,
+ "learning_rate": 9.5528812328888e-06,
+ "loss": 0.41269850730895996,
+ "num_input_tokens_seen": 176420144,
+ "step": 4030,
+ "train_runtime": 31173.5245,
+ "train_tokens_per_second": 5659.294
+ },
+ {
+ "epoch": 0.271680581739833,
+ "grad_norm": 0.7894741861959437,
+ "learning_rate": 9.551787688404386e-06,
+ "loss": 0.3877572536468506,
+ "num_input_tokens_seen": 176635200,
+ "step": 4035,
+ "train_runtime": 31218.475,
+ "train_tokens_per_second": 5658.034
+ },
+ {
+ "epoch": 0.27201723673579314,
+ "grad_norm": 0.7609981119661086,
+ "learning_rate": 9.550692871027061e-06,
+ "loss": 0.3883471727371216,
+ "num_input_tokens_seen": 176856552,
+ "step": 4040,
+ "train_runtime": 31261.3151,
+ "train_tokens_per_second": 5657.361
+ },
+ {
+ "epoch": 0.2723538917317533,
+ "grad_norm": 0.7842118778601046,
+ "learning_rate": 9.549596781062984e-06,
+ "loss": 0.37834415435791013,
+ "num_input_tokens_seen": 177085864,
+ "step": 4045,
+ "train_runtime": 31299.8686,
+ "train_tokens_per_second": 5657.719
+ },
+ {
+ "epoch": 0.2726905467277134,
+ "grad_norm": 0.8324923888304009,
+ "learning_rate": 9.548499418818677e-06,
+ "loss": 0.3912312984466553,
+ "num_input_tokens_seen": 177292800,
+ "step": 4050,
+ "train_runtime": 31343.6672,
+ "train_tokens_per_second": 5656.415
+ },
+ {
+ "epoch": 0.27302720172367356,
+ "grad_norm": 0.886343399880937,
+ "learning_rate": 9.547400784601011e-06,
+ "loss": 0.448984956741333,
+ "num_input_tokens_seen": 177509120,
+ "step": 4055,
+ "train_runtime": 31388.1483,
+ "train_tokens_per_second": 5655.291
+ },
+ {
+ "epoch": 0.2733638567196337,
+ "grad_norm": 0.7611731029617864,
+ "learning_rate": 9.546300878717216e-06,
+ "loss": 0.41724767684936526,
+ "num_input_tokens_seen": 177737192,
+ "step": 4060,
+ "train_runtime": 31419.4811,
+ "train_tokens_per_second": 5656.91
+ },
+ {
+ "epoch": 0.27370051171559384,
+ "grad_norm": 0.7391294054271582,
+ "learning_rate": 9.545199701474877e-06,
+ "loss": 0.3876931667327881,
+ "num_input_tokens_seen": 177955528,
+ "step": 4065,
+ "train_runtime": 31460.3252,
+ "train_tokens_per_second": 5656.506
+ },
+ {
+ "epoch": 0.274037166711554,
+ "grad_norm": 0.8122534322473342,
+ "learning_rate": 9.544097253181935e-06,
+ "loss": 0.37918825149536134,
+ "num_input_tokens_seen": 178168688,
+ "step": 4070,
+ "train_runtime": 31501.2931,
+ "train_tokens_per_second": 5655.917
+ },
+ {
+ "epoch": 0.2743738217075141,
+ "grad_norm": 0.7594531409775707,
+ "learning_rate": 9.542993534146687e-06,
+ "loss": 0.41271028518676756,
+ "num_input_tokens_seen": 178382216,
+ "step": 4075,
+ "train_runtime": 31548.6623,
+ "train_tokens_per_second": 5654.193
+ },
+ {
+ "epoch": 0.27471047670347426,
+ "grad_norm": 0.813877169259893,
+ "learning_rate": 9.541888544677784e-06,
+ "loss": 0.4018410682678223,
+ "num_input_tokens_seen": 178601400,
+ "step": 4080,
+ "train_runtime": 31583.5769,
+ "train_tokens_per_second": 5654.882
+ },
+ {
+ "epoch": 0.2750471316994344,
+ "grad_norm": 0.8060293108187504,
+ "learning_rate": 9.540782285084232e-06,
+ "loss": 0.36272602081298827,
+ "num_input_tokens_seen": 178815704,
+ "step": 4085,
+ "train_runtime": 31624.6421,
+ "train_tokens_per_second": 5654.316
+ },
+ {
+ "epoch": 0.27538378669539454,
+ "grad_norm": 0.7676915231843366,
+ "learning_rate": 9.539674755675392e-06,
+ "loss": 0.39566807746887206,
+ "num_input_tokens_seen": 179033752,
+ "step": 4090,
+ "train_runtime": 31667.4816,
+ "train_tokens_per_second": 5653.552
+ },
+ {
+ "epoch": 0.2757204416913547,
+ "grad_norm": 0.6913331471393089,
+ "learning_rate": 9.538565956760983e-06,
+ "loss": 0.35824832916259763,
+ "num_input_tokens_seen": 179269312,
+ "step": 4095,
+ "train_runtime": 31701.6169,
+ "train_tokens_per_second": 5654.895
+ },
+ {
+ "epoch": 0.2760570966873148,
+ "grad_norm": 0.7630923484514573,
+ "learning_rate": 9.537455888651078e-06,
+ "loss": 0.3908623933792114,
+ "num_input_tokens_seen": 179500456,
+ "step": 4100,
+ "train_runtime": 31737.6882,
+ "train_tokens_per_second": 5655.751
+ },
+ {
+ "epoch": 0.27639375168327496,
+ "grad_norm": 0.8057751017404143,
+ "learning_rate": 9.536344551656103e-06,
+ "loss": 0.4113442420959473,
+ "num_input_tokens_seen": 179727568,
+ "step": 4105,
+ "train_runtime": 31778.3965,
+ "train_tokens_per_second": 5655.653
+ },
+ {
+ "epoch": 0.2767304066792351,
+ "grad_norm": 0.7452483379248684,
+ "learning_rate": 9.535231946086838e-06,
+ "loss": 0.392077112197876,
+ "num_input_tokens_seen": 179947920,
+ "step": 4110,
+ "train_runtime": 31810.5777,
+ "train_tokens_per_second": 5656.858
+ },
+ {
+ "epoch": 0.27706706167519524,
+ "grad_norm": 0.7935004896667632,
+ "learning_rate": 9.534118072254421e-06,
+ "loss": 0.41229848861694335,
+ "num_input_tokens_seen": 180172368,
+ "step": 4115,
+ "train_runtime": 31849.5051,
+ "train_tokens_per_second": 5656.991
+ },
+ {
+ "epoch": 0.2774037166711554,
+ "grad_norm": 0.782426266760514,
+ "learning_rate": 9.533002930470345e-06,
+ "loss": 0.38343167304992676,
+ "num_input_tokens_seen": 180392712,
+ "step": 4120,
+ "train_runtime": 31893.6268,
+ "train_tokens_per_second": 5656.074
+ },
+ {
+ "epoch": 0.2777403716671155,
+ "grad_norm": 0.6585746266135624,
+ "learning_rate": 9.531886521046452e-06,
+ "loss": 0.38598432540893557,
+ "num_input_tokens_seen": 180594944,
+ "step": 4125,
+ "train_runtime": 31934.9141,
+ "train_tokens_per_second": 5655.094
+ },
+ {
+ "epoch": 0.27807702666307565,
+ "grad_norm": 0.8150349479481921,
+ "learning_rate": 9.530768844294947e-06,
+ "loss": 0.41448087692260743,
+ "num_input_tokens_seen": 180816024,
+ "step": 4130,
+ "train_runtime": 31975.4857,
+ "train_tokens_per_second": 5654.833
+ },
+ {
+ "epoch": 0.2784136816590358,
+ "grad_norm": 0.8985901076929287,
+ "learning_rate": 9.52964990052838e-06,
+ "loss": 0.41909160614013674,
+ "num_input_tokens_seen": 181035024,
+ "step": 4135,
+ "train_runtime": 32014.2767,
+ "train_tokens_per_second": 5654.822
+ },
+ {
+ "epoch": 0.27875033665499593,
+ "grad_norm": 0.7912879897563827,
+ "learning_rate": 9.528529690059663e-06,
+ "loss": 0.3741894245147705,
+ "num_input_tokens_seen": 181258672,
+ "step": 4140,
+ "train_runtime": 32055.2042,
+ "train_tokens_per_second": 5654.579
+ },
+ {
+ "epoch": 0.2790869916509561,
+ "grad_norm": 0.9153777980141117,
+ "learning_rate": 9.52740821320206e-06,
+ "loss": 0.4044219970703125,
+ "num_input_tokens_seen": 181475056,
+ "step": 4145,
+ "train_runtime": 32086.7668,
+ "train_tokens_per_second": 5655.76
+ },
+ {
+ "epoch": 0.2794236466469162,
+ "grad_norm": 0.7181310809160721,
+ "learning_rate": 9.526285470269185e-06,
+ "loss": 0.4395021915435791,
+ "num_input_tokens_seen": 181701504,
+ "step": 4150,
+ "train_runtime": 32121.2174,
+ "train_tokens_per_second": 5656.744
+ },
+ {
+ "epoch": 0.2797603016428764,
+ "grad_norm": 0.833451101611962,
+ "learning_rate": 9.525161461575012e-06,
+ "loss": 0.4143718719482422,
+ "num_input_tokens_seen": 181905448,
+ "step": 4155,
+ "train_runtime": 32162.0174,
+ "train_tokens_per_second": 5655.909
+ },
+ {
+ "epoch": 0.28009695663883655,
+ "grad_norm": 0.9347376730252263,
+ "learning_rate": 9.524036187433867e-06,
+ "loss": 0.38724517822265625,
+ "num_input_tokens_seen": 182111832,
+ "step": 4160,
+ "train_runtime": 32199.5756,
+ "train_tokens_per_second": 5655.722
+ },
+ {
+ "epoch": 0.2804336116347967,
+ "grad_norm": 0.7395436682357247,
+ "learning_rate": 9.522909648160427e-06,
+ "loss": 0.38469388484954836,
+ "num_input_tokens_seen": 182331456,
+ "step": 4165,
+ "train_runtime": 32237.7321,
+ "train_tokens_per_second": 5655.84
+ },
+ {
+ "epoch": 0.2807702666307568,
+ "grad_norm": 0.9082810498252065,
+ "learning_rate": 9.521781844069727e-06,
+ "loss": 0.4097011566162109,
+ "num_input_tokens_seen": 182555536,
+ "step": 4170,
+ "train_runtime": 32282.4632,
+ "train_tokens_per_second": 5654.944
+ },
+ {
+ "epoch": 0.28110692162671697,
+ "grad_norm": 0.8211424225834406,
+ "learning_rate": 9.520652775477154e-06,
+ "loss": 0.38606297969818115,
+ "num_input_tokens_seen": 182776104,
+ "step": 4175,
+ "train_runtime": 32324.1374,
+ "train_tokens_per_second": 5654.477
+ },
+ {
+ "epoch": 0.2814435766226771,
+ "grad_norm": 0.6834380753616917,
+ "learning_rate": 9.519522442698447e-06,
+ "loss": 0.36697864532470703,
+ "num_input_tokens_seen": 182986640,
+ "step": 4180,
+ "train_runtime": 32354.2522,
+ "train_tokens_per_second": 5655.722
+ },
+ {
+ "epoch": 0.28178023161863724,
+ "grad_norm": 0.8351889375982321,
+ "learning_rate": 9.518390846049699e-06,
+ "loss": 0.4045860290527344,
+ "num_input_tokens_seen": 183200160,
+ "step": 4185,
+ "train_runtime": 32398.1959,
+ "train_tokens_per_second": 5654.641
+ },
+ {
+ "epoch": 0.2821168866145974,
+ "grad_norm": 0.7607561247725104,
+ "learning_rate": 9.517257985847361e-06,
+ "loss": 0.42084197998046874,
+ "num_input_tokens_seen": 183411672,
+ "step": 4190,
+ "train_runtime": 32441.1387,
+ "train_tokens_per_second": 5653.676
+ },
+ {
+ "epoch": 0.2824535416105575,
+ "grad_norm": 0.7318149860850908,
+ "learning_rate": 9.516123862408232e-06,
+ "loss": 0.4293764591217041,
+ "num_input_tokens_seen": 183636784,
+ "step": 4195,
+ "train_runtime": 32485.1985,
+ "train_tokens_per_second": 5652.937
+ },
+ {
+ "epoch": 0.28279019660651766,
+ "grad_norm": 1.0902481858366875,
+ "learning_rate": 9.514988476049466e-06,
+ "loss": 0.3744691848754883,
+ "num_input_tokens_seen": 183850024,
+ "step": 4200,
+ "train_runtime": 32521.0388,
+ "train_tokens_per_second": 5653.264
+ },
+ {
+ "epoch": 0.2831268516024778,
+ "grad_norm": 0.7196863893637454,
+ "learning_rate": 9.513851827088567e-06,
+ "loss": 0.39441556930541993,
+ "num_input_tokens_seen": 184055312,
+ "step": 4205,
+ "train_runtime": 32558.0875,
+ "train_tokens_per_second": 5653.136
+ },
+ {
+ "epoch": 0.28346350659843794,
+ "grad_norm": 0.901667167722795,
+ "learning_rate": 9.512713915843402e-06,
+ "loss": 0.399395227432251,
+ "num_input_tokens_seen": 184280840,
+ "step": 4210,
+ "train_runtime": 32591.1094,
+ "train_tokens_per_second": 5654.329
+ },
+ {
+ "epoch": 0.2838001615943981,
+ "grad_norm": 0.7871649191207917,
+ "learning_rate": 9.511574742632177e-06,
+ "loss": 0.38000011444091797,
+ "num_input_tokens_seen": 184499800,
+ "step": 4215,
+ "train_runtime": 32635.0953,
+ "train_tokens_per_second": 5653.417
+ },
+ {
+ "epoch": 0.2841368165903582,
+ "grad_norm": 0.7805206391000944,
+ "learning_rate": 9.510434307773464e-06,
+ "loss": 0.45735921859741213,
+ "num_input_tokens_seen": 184725080,
+ "step": 4220,
+ "train_runtime": 32673.8214,
+ "train_tokens_per_second": 5653.611
+ },
+ {
+ "epoch": 0.28447347158631836,
+ "grad_norm": 0.6332223820184592,
+ "learning_rate": 9.509292611586179e-06,
+ "loss": 0.35365097522735595,
+ "num_input_tokens_seen": 184912768,
+ "step": 4225,
+ "train_runtime": 32714.0896,
+ "train_tokens_per_second": 5652.389
+ },
+ {
+ "epoch": 0.2848101265822785,
+ "grad_norm": 0.8305072132980625,
+ "learning_rate": 9.508149654389592e-06,
+ "loss": 0.40491819381713867,
+ "num_input_tokens_seen": 185136408,
+ "step": 4230,
+ "train_runtime": 32755.4509,
+ "train_tokens_per_second": 5652.079
+ },
+ {
+ "epoch": 0.28514678157823864,
+ "grad_norm": 0.6650135309711892,
+ "learning_rate": 9.507005436503331e-06,
+ "loss": 0.3746334552764893,
+ "num_input_tokens_seen": 185367152,
+ "step": 4235,
+ "train_runtime": 32798.7805,
+ "train_tokens_per_second": 5651.648
+ },
+ {
+ "epoch": 0.2854834365741988,
+ "grad_norm": 0.8834317004719436,
+ "learning_rate": 9.505859958247373e-06,
+ "loss": 0.4024499893188477,
+ "num_input_tokens_seen": 185580528,
+ "step": 4240,
+ "train_runtime": 32837.3509,
+ "train_tokens_per_second": 5651.507
+ },
+ {
+ "epoch": 0.2858200915701589,
+ "grad_norm": 0.703931424320413,
+ "learning_rate": 9.504713219942045e-06,
+ "loss": 0.3693699359893799,
+ "num_input_tokens_seen": 185800152,
+ "step": 4245,
+ "train_runtime": 32873.2494,
+ "train_tokens_per_second": 5652.017
+ },
+ {
+ "epoch": 0.28615674656611906,
+ "grad_norm": 0.7673763574951983,
+ "learning_rate": 9.503565221908029e-06,
+ "loss": 0.3989689350128174,
+ "num_input_tokens_seen": 186023072,
+ "step": 4250,
+ "train_runtime": 32908.5929,
+ "train_tokens_per_second": 5652.72
+ },
+ {
+ "epoch": 0.2864934015620792,
+ "grad_norm": 0.7942180767418593,
+ "learning_rate": 9.50241596446636e-06,
+ "loss": 0.41822104454040526,
+ "num_input_tokens_seen": 186251344,
+ "step": 4255,
+ "train_runtime": 32948.6784,
+ "train_tokens_per_second": 5652.771
+ },
+ {
+ "epoch": 0.28683005655803934,
+ "grad_norm": 0.8790228905730333,
+ "learning_rate": 9.501265447938423e-06,
+ "loss": 0.40343165397644043,
+ "num_input_tokens_seen": 186459392,
+ "step": 4260,
+ "train_runtime": 32985.9245,
+ "train_tokens_per_second": 5652.696
+ },
+ {
+ "epoch": 0.2871667115539995,
+ "grad_norm": 0.8360538773951802,
+ "learning_rate": 9.500113672645958e-06,
+ "loss": 0.41559224128723143,
+ "num_input_tokens_seen": 186674536,
+ "step": 4265,
+ "train_runtime": 33020.78,
+ "train_tokens_per_second": 5653.244
+ },
+ {
+ "epoch": 0.2875033665499596,
+ "grad_norm": 0.8611168073988301,
+ "learning_rate": 9.498960638911054e-06,
+ "loss": 0.3698106288909912,
+ "num_input_tokens_seen": 186883568,
+ "step": 4270,
+ "train_runtime": 33066.6343,
+ "train_tokens_per_second": 5651.726
+ },
+ {
+ "epoch": 0.28784002154591976,
+ "grad_norm": 0.7947411629350094,
+ "learning_rate": 9.497806347056153e-06,
+ "loss": 0.394811749458313,
+ "num_input_tokens_seen": 187101448,
+ "step": 4275,
+ "train_runtime": 33108.0971,
+ "train_tokens_per_second": 5651.229
+ },
+ {
+ "epoch": 0.2881766765418799,
+ "grad_norm": 0.8271288744866215,
+ "learning_rate": 9.49665079740405e-06,
+ "loss": 0.41505231857299807,
+ "num_input_tokens_seen": 187323568,
+ "step": 4280,
+ "train_runtime": 33144.1133,
+ "train_tokens_per_second": 5651.79
+ },
+ {
+ "epoch": 0.28851333153784003,
+ "grad_norm": 0.8870087883370243,
+ "learning_rate": 9.495493990277889e-06,
+ "loss": 0.41215314865112307,
+ "num_input_tokens_seen": 187536768,
+ "step": 4285,
+ "train_runtime": 33177.0201,
+ "train_tokens_per_second": 5652.61
+ },
+ {
+ "epoch": 0.2888499865338002,
+ "grad_norm": 0.7416815526876334,
+ "learning_rate": 9.49433592600117e-06,
+ "loss": 0.39543790817260743,
+ "num_input_tokens_seen": 187761936,
+ "step": 4290,
+ "train_runtime": 33218.3652,
+ "train_tokens_per_second": 5652.353
+ },
+ {
+ "epoch": 0.2891866415297603,
+ "grad_norm": 0.8456820374140082,
+ "learning_rate": 9.49317660489774e-06,
+ "loss": 0.38694300651550295,
+ "num_input_tokens_seen": 187989768,
+ "step": 4295,
+ "train_runtime": 33254.7613,
+ "train_tokens_per_second": 5653.018
+ },
+ {
+ "epoch": 0.28952329652572045,
+ "grad_norm": 0.7340597270803788,
+ "learning_rate": 9.4920160272918e-06,
+ "loss": 0.39570274353027346,
+ "num_input_tokens_seen": 188221328,
+ "step": 4300,
+ "train_runtime": 33292.1232,
+ "train_tokens_per_second": 5653.629
+ },
+ {
+ "epoch": 0.2898599515216806,
+ "grad_norm": 0.9417045729052645,
+ "learning_rate": 9.490854193507904e-06,
+ "loss": 0.40100975036621095,
+ "num_input_tokens_seen": 188436984,
+ "step": 4305,
+ "train_runtime": 33333.4329,
+ "train_tokens_per_second": 5653.093
+ },
+ {
+ "epoch": 0.29019660651764073,
+ "grad_norm": 0.8307091778797395,
+ "learning_rate": 9.489691103870951e-06,
+ "loss": 0.4165069580078125,
+ "num_input_tokens_seen": 188666800,
+ "step": 4310,
+ "train_runtime": 33370.0506,
+ "train_tokens_per_second": 5653.776
+ },
+ {
+ "epoch": 0.29053326151360087,
+ "grad_norm": 0.6898420962550031,
+ "learning_rate": 9.488526758706198e-06,
+ "loss": 0.3968646764755249,
+ "num_input_tokens_seen": 188890864,
+ "step": 4315,
+ "train_runtime": 33408.9132,
+ "train_tokens_per_second": 5653.906
+ },
+ {
+ "epoch": 0.290869916509561,
+ "grad_norm": 0.9744166501910104,
+ "learning_rate": 9.48736115833925e-06,
+ "loss": 0.42871532440185545,
+ "num_input_tokens_seen": 189097152,
+ "step": 4320,
+ "train_runtime": 33444.6308,
+ "train_tokens_per_second": 5654.036
+ },
+ {
+ "epoch": 0.29120657150552115,
+ "grad_norm": 0.7709225593851357,
+ "learning_rate": 9.486194303096062e-06,
+ "loss": 0.388767147064209,
+ "num_input_tokens_seen": 189319032,
+ "step": 4325,
+ "train_runtime": 33480.301,
+ "train_tokens_per_second": 5654.639
+ },
+ {
+ "epoch": 0.2915432265014813,
+ "grad_norm": 0.6810949642678763,
+ "learning_rate": 9.485026193302945e-06,
+ "loss": 0.37795631885528563,
+ "num_input_tokens_seen": 189540056,
+ "step": 4330,
+ "train_runtime": 33515.0635,
+ "train_tokens_per_second": 5655.369
+ },
+ {
+ "epoch": 0.29187988149744143,
+ "grad_norm": 0.6374851744198567,
+ "learning_rate": 9.483856829286552e-06,
+ "loss": 0.43641042709350586,
+ "num_input_tokens_seen": 189765816,
+ "step": 4335,
+ "train_runtime": 33556.7326,
+ "train_tokens_per_second": 5655.074
+ },
+ {
+ "epoch": 0.29221653649340157,
+ "grad_norm": 0.8345515619506504,
+ "learning_rate": 9.482686211373896e-06,
+ "loss": 0.4038511276245117,
+ "num_input_tokens_seen": 189992032,
+ "step": 4340,
+ "train_runtime": 33597.0529,
+ "train_tokens_per_second": 5655.021
+ },
+ {
+ "epoch": 0.2925531914893617,
+ "grad_norm": 0.8775359668232541,
+ "learning_rate": 9.481514339892335e-06,
+ "loss": 0.4172847747802734,
+ "num_input_tokens_seen": 190224704,
+ "step": 4345,
+ "train_runtime": 33633.3907,
+ "train_tokens_per_second": 5655.829
+ },
+ {
+ "epoch": 0.29288984648532185,
+ "grad_norm": 0.7728830127459773,
+ "learning_rate": 9.48034121516958e-06,
+ "loss": 0.41391868591308595,
+ "num_input_tokens_seen": 190437840,
+ "step": 4350,
+ "train_runtime": 33671.0816,
+ "train_tokens_per_second": 5655.828
+ },
+ {
+ "epoch": 0.293226501481282,
+ "grad_norm": 0.7347575506312092,
+ "learning_rate": 9.47916683753369e-06,
+ "loss": 0.38980252742767335,
+ "num_input_tokens_seen": 190656112,
+ "step": 4355,
+ "train_runtime": 33713.543,
+ "train_tokens_per_second": 5655.179
+ },
+ {
+ "epoch": 0.2935631564772421,
+ "grad_norm": 0.7225462237696316,
+ "learning_rate": 9.477991207313077e-06,
+ "loss": 0.4024055480957031,
+ "num_input_tokens_seen": 190883776,
+ "step": 4360,
+ "train_runtime": 33754.4757,
+ "train_tokens_per_second": 5655.066
+ },
+ {
+ "epoch": 0.29389981147320227,
+ "grad_norm": 0.7312109916350326,
+ "learning_rate": 9.476814324836504e-06,
+ "loss": 0.38675341606140134,
+ "num_input_tokens_seen": 191111960,
+ "step": 4365,
+ "train_runtime": 33789.2531,
+ "train_tokens_per_second": 5655.998
+ },
+ {
+ "epoch": 0.2942364664691624,
+ "grad_norm": 0.7752639421468126,
+ "learning_rate": 9.475636190433078e-06,
+ "loss": 0.44519591331481934,
+ "num_input_tokens_seen": 191343016,
+ "step": 4370,
+ "train_runtime": 33827.4824,
+ "train_tokens_per_second": 5656.437
+ },
+ {
+ "epoch": 0.29457312146512254,
+ "grad_norm": 0.7480674780381356,
+ "learning_rate": 9.474456804432264e-06,
+ "loss": 0.3989989995956421,
+ "num_input_tokens_seen": 191572600,
+ "step": 4375,
+ "train_runtime": 33864.2914,
+ "train_tokens_per_second": 5657.068
+ },
+ {
+ "epoch": 0.2949097764610827,
+ "grad_norm": 0.7492609939786371,
+ "learning_rate": 9.473276167163872e-06,
+ "loss": 0.41669330596923826,
+ "num_input_tokens_seen": 191802056,
+ "step": 4380,
+ "train_runtime": 33903.6008,
+ "train_tokens_per_second": 5657.277
+ },
+ {
+ "epoch": 0.2952464314570428,
+ "grad_norm": 0.7254851965917017,
+ "learning_rate": 9.472094278958066e-06,
+ "loss": 0.41675920486450196,
+ "num_input_tokens_seen": 192033856,
+ "step": 4385,
+ "train_runtime": 33947.504,
+ "train_tokens_per_second": 5656.789
+ },
+ {
+ "epoch": 0.29558308645300296,
+ "grad_norm": 0.7301224106455536,
+ "learning_rate": 9.470911140145353e-06,
+ "loss": 0.3885148763656616,
+ "num_input_tokens_seen": 192244376,
+ "step": 4390,
+ "train_runtime": 33990.9485,
+ "train_tokens_per_second": 5655.752
+ },
+ {
+ "epoch": 0.2959197414489631,
+ "grad_norm": 0.8599934571941987,
+ "learning_rate": 9.469726751056599e-06,
+ "loss": 0.3994889736175537,
+ "num_input_tokens_seen": 192466432,
+ "step": 4395,
+ "train_runtime": 34031.0021,
+ "train_tokens_per_second": 5655.62
+ },
+ {
+ "epoch": 0.29625639644492324,
+ "grad_norm": 0.8672385958944913,
+ "learning_rate": 9.468541112023009e-06,
+ "loss": 0.42194247245788574,
+ "num_input_tokens_seen": 192676720,
+ "step": 4400,
+ "train_runtime": 34068.9438,
+ "train_tokens_per_second": 5655.494
+ },
+ {
+ "epoch": 0.2965930514408834,
+ "grad_norm": 0.8654785396115573,
+ "learning_rate": 9.467354223376149e-06,
+ "loss": 0.4138471603393555,
+ "num_input_tokens_seen": 192867824,
+ "step": 4405,
+ "train_runtime": 34101.5758,
+ "train_tokens_per_second": 5655.687
+ },
+ {
+ "epoch": 0.2969297064368435,
+ "grad_norm": 0.7586880855501626,
+ "learning_rate": 9.466166085447923e-06,
+ "loss": 0.4040337562561035,
+ "num_input_tokens_seen": 193079552,
+ "step": 4410,
+ "train_runtime": 34144.2523,
+ "train_tokens_per_second": 5654.819
+ },
+ {
+ "epoch": 0.29726636143280366,
+ "grad_norm": 0.8413204922562729,
+ "learning_rate": 9.464976698570595e-06,
+ "loss": 0.3852673053741455,
+ "num_input_tokens_seen": 193274704,
+ "step": 4415,
+ "train_runtime": 34179.0678,
+ "train_tokens_per_second": 5654.768
+ },
+ {
+ "epoch": 0.2976030164287638,
+ "grad_norm": 0.8227530497102407,
+ "learning_rate": 9.463786063076769e-06,
+ "loss": 0.42119736671447755,
+ "num_input_tokens_seen": 193494016,
+ "step": 4420,
+ "train_runtime": 34217.4917,
+ "train_tokens_per_second": 5654.828
+ },
+ {
+ "epoch": 0.29793967142472394,
+ "grad_norm": 0.8498857055875949,
+ "learning_rate": 9.462594179299408e-06,
+ "loss": 0.34813380241394043,
+ "num_input_tokens_seen": 193708944,
+ "step": 4425,
+ "train_runtime": 34255.6112,
+ "train_tokens_per_second": 5654.809
+ },
+ {
+ "epoch": 0.2982763264206841,
+ "grad_norm": 0.7591384035910306,
+ "learning_rate": 9.461401047571811e-06,
+ "loss": 0.39094252586364747,
+ "num_input_tokens_seen": 193940712,
+ "step": 4430,
+ "train_runtime": 34291.0594,
+ "train_tokens_per_second": 5655.722
+ },
+ {
+ "epoch": 0.2986129814166442,
+ "grad_norm": 0.8057409790041082,
+ "learning_rate": 9.460206668227639e-06,
+ "loss": 0.4155745983123779,
+ "num_input_tokens_seen": 194168936,
+ "step": 4435,
+ "train_runtime": 34330.974,
+ "train_tokens_per_second": 5655.795
+ },
+ {
+ "epoch": 0.29894963641260436,
+ "grad_norm": 0.7522439279097726,
+ "learning_rate": 9.459011041600895e-06,
+ "loss": 0.39428043365478516,
+ "num_input_tokens_seen": 194397872,
+ "step": 4440,
+ "train_runtime": 34371.2051,
+ "train_tokens_per_second": 5655.835
+ },
+ {
+ "epoch": 0.2992862914085645,
+ "grad_norm": 0.8885125640048409,
+ "learning_rate": 9.457814168025932e-06,
+ "loss": 0.3819078207015991,
+ "num_input_tokens_seen": 194618856,
+ "step": 4445,
+ "train_runtime": 34409.2545,
+ "train_tokens_per_second": 5656.003
+ },
+ {
+ "epoch": 0.29962294640452464,
+ "grad_norm": 0.8798594554422206,
+ "learning_rate": 9.456616047837452e-06,
+ "loss": 0.4244089603424072,
+ "num_input_tokens_seen": 194815632,
+ "step": 4450,
+ "train_runtime": 34451.3067,
+ "train_tokens_per_second": 5654.811
+ },
+ {
+ "epoch": 0.2999596014004848,
+ "grad_norm": 0.7140830042785486,
+ "learning_rate": 9.455416681370506e-06,
+ "loss": 0.43105688095092776,
+ "num_input_tokens_seen": 195043320,
+ "step": 4455,
+ "train_runtime": 34491.908,
+ "train_tokens_per_second": 5654.756
+ },
+ {
+ "epoch": 0.3002962563964449,
+ "grad_norm": 0.7933555904953391,
+ "learning_rate": 9.45421606896049e-06,
+ "loss": 0.4107870101928711,
+ "num_input_tokens_seen": 195273800,
+ "step": 4460,
+ "train_runtime": 34532.6899,
+ "train_tokens_per_second": 5654.752
+ },
+ {
+ "epoch": 0.30063291139240506,
+ "grad_norm": 0.7656195298280324,
+ "learning_rate": 9.453014210943155e-06,
+ "loss": 0.39834353923797605,
+ "num_input_tokens_seen": 195492912,
+ "step": 4465,
+ "train_runtime": 34574.8621,
+ "train_tokens_per_second": 5654.192
+ },
+ {
+ "epoch": 0.3009695663883652,
+ "grad_norm": 0.7817448952040295,
+ "learning_rate": 9.451811107654596e-06,
+ "loss": 0.39757986068725587,
+ "num_input_tokens_seen": 195716896,
+ "step": 4470,
+ "train_runtime": 34621.7737,
+ "train_tokens_per_second": 5653.0
+ },
+ {
+ "epoch": 0.30130622138432533,
+ "grad_norm": 0.7841726368249392,
+ "learning_rate": 9.450606759431255e-06,
+ "loss": 0.40091800689697266,
+ "num_input_tokens_seen": 195945064,
+ "step": 4475,
+ "train_runtime": 34656.1692,
+ "train_tokens_per_second": 5653.974
+ },
+ {
+ "epoch": 0.3016428763802855,
+ "grad_norm": 0.8817669510853672,
+ "learning_rate": 9.449401166609928e-06,
+ "loss": 0.42542014122009275,
+ "num_input_tokens_seen": 196155296,
+ "step": 4480,
+ "train_runtime": 34691.0919,
+ "train_tokens_per_second": 5654.342
+ },
+ {
+ "epoch": 0.3019795313762456,
+ "grad_norm": 0.958339871121062,
+ "learning_rate": 9.448194329527752e-06,
+ "loss": 0.40885372161865235,
+ "num_input_tokens_seen": 196367168,
+ "step": 4485,
+ "train_runtime": 34725.1749,
+ "train_tokens_per_second": 5654.894
+ },
+ {
+ "epoch": 0.30231618637220575,
+ "grad_norm": 0.7219539133285781,
+ "learning_rate": 9.446986248522216e-06,
+ "loss": 0.38100175857543944,
+ "num_input_tokens_seen": 196585760,
+ "step": 4490,
+ "train_runtime": 34763.5979,
+ "train_tokens_per_second": 5654.931
+ },
+ {
+ "epoch": 0.3026528413681659,
+ "grad_norm": 0.7623511745984738,
+ "learning_rate": 9.445776923931157e-06,
+ "loss": 0.3696560859680176,
+ "num_input_tokens_seen": 196799296,
+ "step": 4495,
+ "train_runtime": 34801.3851,
+ "train_tokens_per_second": 5654.927
+ },
+ {
+ "epoch": 0.30298949636412603,
+ "grad_norm": 0.7900772624336759,
+ "learning_rate": 9.444566356092754e-06,
+ "loss": 0.3985780715942383,
+ "num_input_tokens_seen": 197027464,
+ "step": 4500,
+ "train_runtime": 34835.2173,
+ "train_tokens_per_second": 5655.985
+ },
+ {
+ "epoch": 0.30332615136008617,
+ "grad_norm": 0.8352145795658186,
+ "learning_rate": 9.443354545345545e-06,
+ "loss": 0.36351020336151124,
+ "num_input_tokens_seen": 197246488,
+ "step": 4505,
+ "train_runtime": 34872.1679,
+ "train_tokens_per_second": 5656.273
+ },
+ {
+ "epoch": 0.3036628063560463,
+ "grad_norm": 0.7235873259829829,
+ "learning_rate": 9.442141492028406e-06,
+ "loss": 0.3931256294250488,
+ "num_input_tokens_seen": 197464240,
+ "step": 4510,
+ "train_runtime": 34916.8698,
+ "train_tokens_per_second": 5655.268
+ },
+ {
+ "epoch": 0.30399946135200645,
+ "grad_norm": 0.7287455155410504,
+ "learning_rate": 9.440927196480563e-06,
+ "loss": 0.3859192609786987,
+ "num_input_tokens_seen": 197693416,
+ "step": 4515,
+ "train_runtime": 34952.6701,
+ "train_tokens_per_second": 5656.032
+ },
+ {
+ "epoch": 0.3043361163479666,
+ "grad_norm": 0.7124742821058738,
+ "learning_rate": 9.439711659041593e-06,
+ "loss": 0.3971871376037598,
+ "num_input_tokens_seen": 197908880,
+ "step": 4520,
+ "train_runtime": 34995.5331,
+ "train_tokens_per_second": 5655.261
+ },
+ {
+ "epoch": 0.30467277134392673,
+ "grad_norm": 0.8049967305897996,
+ "learning_rate": 9.438494880051413e-06,
+ "loss": 0.4050449371337891,
+ "num_input_tokens_seen": 198130240,
+ "step": 4525,
+ "train_runtime": 35031.404,
+ "train_tokens_per_second": 5655.789
+ },
+ {
+ "epoch": 0.30500942633988687,
+ "grad_norm": 0.8167232313014808,
+ "learning_rate": 9.437276859850293e-06,
+ "loss": 0.39906351566314696,
+ "num_input_tokens_seen": 198354624,
+ "step": 4530,
+ "train_runtime": 35070.2405,
+ "train_tokens_per_second": 5655.924
+ },
+ {
+ "epoch": 0.305346081335847,
+ "grad_norm": 0.8219257479433354,
+ "learning_rate": 9.436057598778848e-06,
+ "loss": 0.41982173919677734,
+ "num_input_tokens_seen": 198569592,
+ "step": 4535,
+ "train_runtime": 35103.488,
+ "train_tokens_per_second": 5656.691
+ },
+ {
+ "epoch": 0.30568273633180715,
+ "grad_norm": 0.8373417043715061,
+ "learning_rate": 9.434837097178043e-06,
+ "loss": 0.44964447021484377,
+ "num_input_tokens_seen": 198790848,
+ "step": 4540,
+ "train_runtime": 35136.0232,
+ "train_tokens_per_second": 5657.75
+ },
+ {
+ "epoch": 0.3060193913277673,
+ "grad_norm": 0.8551654564560078,
+ "learning_rate": 9.433615355389183e-06,
+ "loss": 0.4145000457763672,
+ "num_input_tokens_seen": 198995864,
+ "step": 4545,
+ "train_runtime": 35171.0896,
+ "train_tokens_per_second": 5657.939
+ },
+ {
+ "epoch": 0.3063560463237274,
+ "grad_norm": 0.8539775390580207,
+ "learning_rate": 9.432392373753926e-06,
+ "loss": 0.38867299556732177,
+ "num_input_tokens_seen": 199228112,
+ "step": 4550,
+ "train_runtime": 35204.7853,
+ "train_tokens_per_second": 5659.12
+ },
+ {
+ "epoch": 0.30669270131968757,
+ "grad_norm": 0.7417346548045687,
+ "learning_rate": 9.431168152614278e-06,
+ "loss": 0.398314380645752,
+ "num_input_tokens_seen": 199444448,
+ "step": 4555,
+ "train_runtime": 35246.8496,
+ "train_tokens_per_second": 5658.504
+ },
+ {
+ "epoch": 0.3070293563156477,
+ "grad_norm": 0.7867984496679544,
+ "learning_rate": 9.429942692312585e-06,
+ "loss": 0.36063923835754397,
+ "num_input_tokens_seen": 199665776,
+ "step": 4560,
+ "train_runtime": 35277.6142,
+ "train_tokens_per_second": 5659.844
+ },
+ {
+ "epoch": 0.30736601131160785,
+ "grad_norm": 0.6064626251289161,
+ "learning_rate": 9.428715993191546e-06,
+ "loss": 0.38531041145324707,
+ "num_input_tokens_seen": 199888952,
+ "step": 4565,
+ "train_runtime": 35318.7352,
+ "train_tokens_per_second": 5659.573
+ },
+ {
+ "epoch": 0.307702666307568,
+ "grad_norm": 0.7823320659003977,
+ "learning_rate": 9.427488055594199e-06,
+ "loss": 0.4015589714050293,
+ "num_input_tokens_seen": 200094448,
+ "step": 4570,
+ "train_runtime": 35356.6905,
+ "train_tokens_per_second": 5659.309
+ },
+ {
+ "epoch": 0.3080393213035281,
+ "grad_norm": 0.7797326812244397,
+ "learning_rate": 9.426258879863937e-06,
+ "loss": 0.40995278358459475,
+ "num_input_tokens_seen": 200330936,
+ "step": 4575,
+ "train_runtime": 35388.9695,
+ "train_tokens_per_second": 5660.83
+ },
+ {
+ "epoch": 0.30837597629948826,
+ "grad_norm": 0.6558180029153803,
+ "learning_rate": 9.425028466344494e-06,
+ "loss": 0.35400965213775637,
+ "num_input_tokens_seen": 200550336,
+ "step": 4580,
+ "train_runtime": 35423.4598,
+ "train_tokens_per_second": 5661.512
+ },
+ {
+ "epoch": 0.3087126312954484,
+ "grad_norm": 0.9357449426999332,
+ "learning_rate": 9.42379681537995e-06,
+ "loss": 0.40514564514160156,
+ "num_input_tokens_seen": 200767584,
+ "step": 4585,
+ "train_runtime": 35466.8621,
+ "train_tokens_per_second": 5660.709
+ },
+ {
+ "epoch": 0.30904928629140854,
+ "grad_norm": 1.1029634658189478,
+ "learning_rate": 9.422563927314732e-06,
+ "loss": 0.39676432609558104,
+ "num_input_tokens_seen": 200990368,
+ "step": 4590,
+ "train_runtime": 35503.267,
+ "train_tokens_per_second": 5661.18
+ },
+ {
+ "epoch": 0.3093859412873687,
+ "grad_norm": 0.7960058470606449,
+ "learning_rate": 9.421329802493615e-06,
+ "loss": 0.3798659801483154,
+ "num_input_tokens_seen": 201213136,
+ "step": 4595,
+ "train_runtime": 35545.6953,
+ "train_tokens_per_second": 5660.689
+ },
+ {
+ "epoch": 0.3097225962833288,
+ "grad_norm": 0.7764466461774601,
+ "learning_rate": 9.420094441261717e-06,
+ "loss": 0.4128886222839355,
+ "num_input_tokens_seen": 201436008,
+ "step": 4600,
+ "train_runtime": 35591.0272,
+ "train_tokens_per_second": 5659.741
+ },
+ {
+ "epoch": 0.31005925127928896,
+ "grad_norm": 0.8514809314658448,
+ "learning_rate": 9.418857843964506e-06,
+ "loss": 0.3874027252197266,
+ "num_input_tokens_seen": 201651024,
+ "step": 4605,
+ "train_runtime": 35633.1312,
+ "train_tokens_per_second": 5659.088
+ },
+ {
+ "epoch": 0.3103959062752491,
+ "grad_norm": 0.791242525351627,
+ "learning_rate": 9.417620010947786e-06,
+ "loss": 0.41231350898742675,
+ "num_input_tokens_seen": 201862568,
+ "step": 4610,
+ "train_runtime": 35672.8097,
+ "train_tokens_per_second": 5658.724
+ },
+ {
+ "epoch": 0.31073256127120924,
+ "grad_norm": 0.8196831043483923,
+ "learning_rate": 9.416380942557719e-06,
+ "loss": 0.3759624719619751,
+ "num_input_tokens_seen": 202081848,
+ "step": 4615,
+ "train_runtime": 35711.7461,
+ "train_tokens_per_second": 5658.694
+ },
+ {
+ "epoch": 0.3110692162671694,
+ "grad_norm": 0.814113401127112,
+ "learning_rate": 9.415140639140803e-06,
+ "loss": 0.40959873199462893,
+ "num_input_tokens_seen": 202288416,
+ "step": 4620,
+ "train_runtime": 35748.2065,
+ "train_tokens_per_second": 5658.701
+ },
+ {
+ "epoch": 0.3114058712631295,
+ "grad_norm": 0.6867352095609621,
+ "learning_rate": 9.413899101043887e-06,
+ "loss": 0.4299426555633545,
+ "num_input_tokens_seen": 202507280,
+ "step": 4625,
+ "train_runtime": 35787.2021,
+ "train_tokens_per_second": 5658.651
+ },
+ {
+ "epoch": 0.31174252625908966,
+ "grad_norm": 0.9127540749426457,
+ "learning_rate": 9.412656328614162e-06,
+ "loss": 0.4051331043243408,
+ "num_input_tokens_seen": 202708520,
+ "step": 4630,
+ "train_runtime": 35826.4689,
+ "train_tokens_per_second": 5658.066
+ },
+ {
+ "epoch": 0.3120791812550498,
+ "grad_norm": 0.7769819809945303,
+ "learning_rate": 9.411412322199165e-06,
+ "loss": 0.40700321197509765,
+ "num_input_tokens_seen": 202938288,
+ "step": 4635,
+ "train_runtime": 35860.2769,
+ "train_tokens_per_second": 5659.139
+ },
+ {
+ "epoch": 0.31241583625100994,
+ "grad_norm": 0.7763840440333488,
+ "learning_rate": 9.41016708214678e-06,
+ "loss": 0.397929310798645,
+ "num_input_tokens_seen": 203140768,
+ "step": 4640,
+ "train_runtime": 35900.8969,
+ "train_tokens_per_second": 5658.376
+ },
+ {
+ "epoch": 0.31275249124697013,
+ "grad_norm": 0.838793433476314,
+ "learning_rate": 9.408920608805233e-06,
+ "loss": 0.3729093074798584,
+ "num_input_tokens_seen": 203353200,
+ "step": 4645,
+ "train_runtime": 35943.6665,
+ "train_tokens_per_second": 5657.553
+ },
+ {
+ "epoch": 0.31308914624293027,
+ "grad_norm": 0.759403553618091,
+ "learning_rate": 9.407672902523097e-06,
+ "loss": 0.42061967849731446,
+ "num_input_tokens_seen": 203576960,
+ "step": 4650,
+ "train_runtime": 35989.9311,
+ "train_tokens_per_second": 5656.498
+ },
+ {
+ "epoch": 0.3134258012388904,
+ "grad_norm": 0.8516208752301061,
+ "learning_rate": 9.406423963649287e-06,
+ "loss": 0.39535982608795167,
+ "num_input_tokens_seen": 203780664,
+ "step": 4655,
+ "train_runtime": 36033.4005,
+ "train_tokens_per_second": 5655.327
+ },
+ {
+ "epoch": 0.31376245623485055,
+ "grad_norm": 0.6661646334234551,
+ "learning_rate": 9.405173792533069e-06,
+ "loss": 0.3896958827972412,
+ "num_input_tokens_seen": 204002504,
+ "step": 4660,
+ "train_runtime": 36071.9272,
+ "train_tokens_per_second": 5655.437
+ },
+ {
+ "epoch": 0.3140991112308107,
+ "grad_norm": 0.802128758388854,
+ "learning_rate": 9.403922389524045e-06,
+ "loss": 0.383258056640625,
+ "num_input_tokens_seen": 204219216,
+ "step": 4665,
+ "train_runtime": 36107.3083,
+ "train_tokens_per_second": 5655.897
+ },
+ {
+ "epoch": 0.31443576622677083,
+ "grad_norm": 0.8311848290104593,
+ "learning_rate": 9.40266975497217e-06,
+ "loss": 0.41370835304260256,
+ "num_input_tokens_seen": 204435080,
+ "step": 4670,
+ "train_runtime": 36146.1312,
+ "train_tokens_per_second": 5655.794
+ },
+ {
+ "epoch": 0.31477242122273097,
+ "grad_norm": 0.7524391894573029,
+ "learning_rate": 9.401415889227736e-06,
+ "loss": 0.4065183162689209,
+ "num_input_tokens_seen": 204665984,
+ "step": 4675,
+ "train_runtime": 36189.9551,
+ "train_tokens_per_second": 5655.326
+ },
+ {
+ "epoch": 0.3151090762186911,
+ "grad_norm": 0.6516968499884368,
+ "learning_rate": 9.400160792641385e-06,
+ "loss": 0.40553603172302244,
+ "num_input_tokens_seen": 204900584,
+ "step": 4680,
+ "train_runtime": 36230.3148,
+ "train_tokens_per_second": 5655.501
+ },
+ {
+ "epoch": 0.31544573121465125,
+ "grad_norm": 0.713149562707507,
+ "learning_rate": 9.398904465564097e-06,
+ "loss": 0.3963156700134277,
+ "num_input_tokens_seen": 205121896,
+ "step": 4685,
+ "train_runtime": 36261.5038,
+ "train_tokens_per_second": 5656.74
+ },
+ {
+ "epoch": 0.3157823862106114,
+ "grad_norm": 0.7445101762458055,
+ "learning_rate": 9.397646908347203e-06,
+ "loss": 0.3701439142227173,
+ "num_input_tokens_seen": 205337424,
+ "step": 4690,
+ "train_runtime": 36299.7979,
+ "train_tokens_per_second": 5656.71
+ },
+ {
+ "epoch": 0.3161190412065715,
+ "grad_norm": 0.7876242336850148,
+ "learning_rate": 9.396388121342375e-06,
+ "loss": 0.37079999446868894,
+ "num_input_tokens_seen": 205566880,
+ "step": 4695,
+ "train_runtime": 36337.2628,
+ "train_tokens_per_second": 5657.192
+ },
+ {
+ "epoch": 0.31645569620253167,
+ "grad_norm": 0.8752489859691854,
+ "learning_rate": 9.395128104901628e-06,
+ "loss": 0.382697606086731,
+ "num_input_tokens_seen": 205780336,
+ "step": 4700,
+ "train_runtime": 36373.0101,
+ "train_tokens_per_second": 5657.501
+ },
+ {
+ "epoch": 0.3167923511984918,
+ "grad_norm": 0.6981702689866264,
+ "learning_rate": 9.393866859377321e-06,
+ "loss": 0.4276883125305176,
+ "num_input_tokens_seen": 206004992,
+ "step": 4705,
+ "train_runtime": 36409.7995,
+ "train_tokens_per_second": 5657.955
+ },
+ {
+ "epoch": 0.31712900619445195,
+ "grad_norm": 0.7289408984257015,
+ "learning_rate": 9.392604385122157e-06,
+ "loss": 0.39430766105651854,
+ "num_input_tokens_seen": 206237344,
+ "step": 4710,
+ "train_runtime": 36441.1815,
+ "train_tokens_per_second": 5659.458
+ },
+ {
+ "epoch": 0.3174656611904121,
+ "grad_norm": 0.7707325345782555,
+ "learning_rate": 9.391340682489185e-06,
+ "loss": 0.3808866024017334,
+ "num_input_tokens_seen": 206449496,
+ "step": 4715,
+ "train_runtime": 36480.1405,
+ "train_tokens_per_second": 5659.23
+ },
+ {
+ "epoch": 0.3178023161863722,
+ "grad_norm": 0.8379775589672267,
+ "learning_rate": 9.390075751831794e-06,
+ "loss": 0.39603471755981445,
+ "num_input_tokens_seen": 206677304,
+ "step": 4720,
+ "train_runtime": 36513.5133,
+ "train_tokens_per_second": 5660.296
+ },
+ {
+ "epoch": 0.31813897118233236,
+ "grad_norm": 1.1929541297616924,
+ "learning_rate": 9.388809593503718e-06,
+ "loss": 0.41028714179992676,
+ "num_input_tokens_seen": 206901904,
+ "step": 4725,
+ "train_runtime": 36554.4289,
+ "train_tokens_per_second": 5660.105
+ },
+ {
+ "epoch": 0.3184756261782925,
+ "grad_norm": 0.8413335424516502,
+ "learning_rate": 9.387542207859034e-06,
+ "loss": 0.428537654876709,
+ "num_input_tokens_seen": 207122320,
+ "step": 4730,
+ "train_runtime": 36584.6009,
+ "train_tokens_per_second": 5661.462
+ },
+ {
+ "epoch": 0.31881228117425264,
+ "grad_norm": 0.7898247112897376,
+ "learning_rate": 9.386273595252161e-06,
+ "loss": 0.41155662536621096,
+ "num_input_tokens_seen": 207352296,
+ "step": 4735,
+ "train_runtime": 36620.4486,
+ "train_tokens_per_second": 5662.2
+ },
+ {
+ "epoch": 0.3191489361702128,
+ "grad_norm": 0.8389868484209729,
+ "learning_rate": 9.385003756037865e-06,
+ "loss": 0.4223154067993164,
+ "num_input_tokens_seen": 207575072,
+ "step": 4740,
+ "train_runtime": 36662.9004,
+ "train_tokens_per_second": 5661.72
+ },
+ {
+ "epoch": 0.3194855911661729,
+ "grad_norm": 0.827214827358242,
+ "learning_rate": 9.383732690571253e-06,
+ "loss": 0.40798358917236327,
+ "num_input_tokens_seen": 207817360,
+ "step": 4745,
+ "train_runtime": 36697.2613,
+ "train_tokens_per_second": 5663.021
+ },
+ {
+ "epoch": 0.31982224616213306,
+ "grad_norm": 0.8033752581626638,
+ "learning_rate": 9.382460399207769e-06,
+ "loss": 0.40032100677490234,
+ "num_input_tokens_seen": 208045464,
+ "step": 4750,
+ "train_runtime": 36739.6396,
+ "train_tokens_per_second": 5662.697
+ },
+ {
+ "epoch": 0.3201589011580932,
+ "grad_norm": 0.7745249893682031,
+ "learning_rate": 9.381186882303212e-06,
+ "loss": 0.3821758985519409,
+ "num_input_tokens_seen": 208245512,
+ "step": 4755,
+ "train_runtime": 36775.3408,
+ "train_tokens_per_second": 5662.64
+ },
+ {
+ "epoch": 0.32049555615405334,
+ "grad_norm": 0.7835752828676862,
+ "learning_rate": 9.379912140213713e-06,
+ "loss": 0.38662168979644773,
+ "num_input_tokens_seen": 208458376,
+ "step": 4760,
+ "train_runtime": 36810.474,
+ "train_tokens_per_second": 5663.018
+ },
+ {
+ "epoch": 0.3208322111500135,
+ "grad_norm": 0.7427789671372642,
+ "learning_rate": 9.37863617329575e-06,
+ "loss": 0.37845721244812014,
+ "num_input_tokens_seen": 208662464,
+ "step": 4765,
+ "train_runtime": 36848.709,
+ "train_tokens_per_second": 5662.68
+ },
+ {
+ "epoch": 0.3211688661459736,
+ "grad_norm": 0.7906549867893844,
+ "learning_rate": 9.377358981906145e-06,
+ "loss": 0.3855098485946655,
+ "num_input_tokens_seen": 208883752,
+ "step": 4770,
+ "train_runtime": 36885.8469,
+ "train_tokens_per_second": 5662.978
+ },
+ {
+ "epoch": 0.32150552114193376,
+ "grad_norm": 0.7000304206526682,
+ "learning_rate": 9.376080566402059e-06,
+ "loss": 0.41453986167907714,
+ "num_input_tokens_seen": 209114544,
+ "step": 4775,
+ "train_runtime": 36923.9104,
+ "train_tokens_per_second": 5663.391
+ },
+ {
+ "epoch": 0.3218421761378939,
+ "grad_norm": 0.7569304220647727,
+ "learning_rate": 9.374800927140994e-06,
+ "loss": 0.33689126968383787,
+ "num_input_tokens_seen": 209342880,
+ "step": 4780,
+ "train_runtime": 36959.0708,
+ "train_tokens_per_second": 5664.181
+ },
+ {
+ "epoch": 0.32217883113385404,
+ "grad_norm": 0.7693618085921944,
+ "learning_rate": 9.373520064480804e-06,
+ "loss": 0.3944025278091431,
+ "num_input_tokens_seen": 209567016,
+ "step": 4785,
+ "train_runtime": 36999.5953,
+ "train_tokens_per_second": 5664.035
+ },
+ {
+ "epoch": 0.3225154861298142,
+ "grad_norm": 0.7545108220793344,
+ "learning_rate": 9.372237978779672e-06,
+ "loss": 0.3935252666473389,
+ "num_input_tokens_seen": 209798224,
+ "step": 4790,
+ "train_runtime": 37035.8347,
+ "train_tokens_per_second": 5664.736
+ },
+ {
+ "epoch": 0.3228521411257743,
+ "grad_norm": 0.7487128271214056,
+ "learning_rate": 9.37095467039613e-06,
+ "loss": 0.3892039775848389,
+ "num_input_tokens_seen": 209998416,
+ "step": 4795,
+ "train_runtime": 37074.0037,
+ "train_tokens_per_second": 5664.304
+ },
+ {
+ "epoch": 0.32318879612173446,
+ "grad_norm": 0.7480661049890698,
+ "learning_rate": 9.369670139689056e-06,
+ "loss": 0.3860762119293213,
+ "num_input_tokens_seen": 210212136,
+ "step": 4800,
+ "train_runtime": 37113.4829,
+ "train_tokens_per_second": 5664.037
+ },
+ {
+ "epoch": 0.3235254511176946,
+ "grad_norm": 0.7846044101372698,
+ "learning_rate": 9.368384387017659e-06,
+ "loss": 0.377323055267334,
+ "num_input_tokens_seen": 210412912,
+ "step": 4805,
+ "train_runtime": 37150.0136,
+ "train_tokens_per_second": 5663.872
+ },
+ {
+ "epoch": 0.32386210611365474,
+ "grad_norm": 0.77183510086721,
+ "learning_rate": 9.367097412741497e-06,
+ "loss": 0.3832876443862915,
+ "num_input_tokens_seen": 210626632,
+ "step": 4810,
+ "train_runtime": 37185.5052,
+ "train_tokens_per_second": 5664.213
+ },
+ {
+ "epoch": 0.3241987611096149,
+ "grad_norm": 0.9148498149260199,
+ "learning_rate": 9.36580921722047e-06,
+ "loss": 0.41832733154296875,
+ "num_input_tokens_seen": 210847304,
+ "step": 4815,
+ "train_runtime": 37226.3372,
+ "train_tokens_per_second": 5663.928
+ },
+ {
+ "epoch": 0.324535416105575,
+ "grad_norm": 0.6839164180369833,
+ "learning_rate": 9.364519800814818e-06,
+ "loss": 0.3626012563705444,
+ "num_input_tokens_seen": 211071840,
+ "step": 4820,
+ "train_runtime": 37269.7466,
+ "train_tokens_per_second": 5663.356
+ },
+ {
+ "epoch": 0.32487207110153515,
+ "grad_norm": 0.6658908418064057,
+ "learning_rate": 9.36322916388512e-06,
+ "loss": 0.3882687330245972,
+ "num_input_tokens_seen": 211297288,
+ "step": 4825,
+ "train_runtime": 37311.4159,
+ "train_tokens_per_second": 5663.073
+ },
+ {
+ "epoch": 0.3252087260974953,
+ "grad_norm": 0.7420928499905488,
+ "learning_rate": 9.3619373067923e-06,
+ "loss": 0.3612262487411499,
+ "num_input_tokens_seen": 211529896,
+ "step": 4830,
+ "train_runtime": 37348.4533,
+ "train_tokens_per_second": 5663.686
+ },
+ {
+ "epoch": 0.32554538109345543,
+ "grad_norm": 0.7491528413015758,
+ "learning_rate": 9.360644229897622e-06,
+ "loss": 0.39133119583129883,
+ "num_input_tokens_seen": 211744104,
+ "step": 4835,
+ "train_runtime": 37390.6137,
+ "train_tokens_per_second": 5663.028
+ },
+ {
+ "epoch": 0.32588203608941557,
+ "grad_norm": 0.8362580597115173,
+ "learning_rate": 9.35934993356269e-06,
+ "loss": 0.4070117950439453,
+ "num_input_tokens_seen": 211960664,
+ "step": 4840,
+ "train_runtime": 37434.3008,
+ "train_tokens_per_second": 5662.204
+ },
+ {
+ "epoch": 0.3262186910853757,
+ "grad_norm": 0.8872020316370295,
+ "learning_rate": 9.35805441814945e-06,
+ "loss": 0.37361793518066405,
+ "num_input_tokens_seen": 212184888,
+ "step": 4845,
+ "train_runtime": 37481.2897,
+ "train_tokens_per_second": 5661.088
+ },
+ {
+ "epoch": 0.32655534608133585,
+ "grad_norm": 0.7655564350170675,
+ "learning_rate": 9.356757684020188e-06,
+ "loss": 0.40865292549133303,
+ "num_input_tokens_seen": 212387000,
+ "step": 4850,
+ "train_runtime": 37521.1842,
+ "train_tokens_per_second": 5660.456
+ },
+ {
+ "epoch": 0.326892001077296,
+ "grad_norm": 0.9840119081755966,
+ "learning_rate": 9.355459731537533e-06,
+ "loss": 0.37503221035003664,
+ "num_input_tokens_seen": 212602240,
+ "step": 4855,
+ "train_runtime": 37563.6776,
+ "train_tokens_per_second": 5659.782
+ },
+ {
+ "epoch": 0.32722865607325613,
+ "grad_norm": 0.8416963207567933,
+ "learning_rate": 9.354160561064451e-06,
+ "loss": 0.4150728225708008,
+ "num_input_tokens_seen": 212829424,
+ "step": 4860,
+ "train_runtime": 37600.8083,
+ "train_tokens_per_second": 5660.235
+ },
+ {
+ "epoch": 0.32756531106921627,
+ "grad_norm": 0.6549002669311522,
+ "learning_rate": 9.352860172964254e-06,
+ "loss": 0.41104774475097655,
+ "num_input_tokens_seen": 213059616,
+ "step": 4865,
+ "train_runtime": 37639.635,
+ "train_tokens_per_second": 5660.512
+ },
+ {
+ "epoch": 0.3279019660651764,
+ "grad_norm": 0.7662647212420475,
+ "learning_rate": 9.35155856760059e-06,
+ "loss": 0.37230610847473145,
+ "num_input_tokens_seen": 213291344,
+ "step": 4870,
+ "train_runtime": 37683.6107,
+ "train_tokens_per_second": 5660.056
+ },
+ {
+ "epoch": 0.32823862106113655,
+ "grad_norm": 0.8032964428930659,
+ "learning_rate": 9.35025574533745e-06,
+ "loss": 0.4028165817260742,
+ "num_input_tokens_seen": 213518768,
+ "step": 4875,
+ "train_runtime": 37716.795,
+ "train_tokens_per_second": 5661.106
+ },
+ {
+ "epoch": 0.3285752760570967,
+ "grad_norm": 0.709809886638744,
+ "learning_rate": 9.34895170653916e-06,
+ "loss": 0.37210307121276853,
+ "num_input_tokens_seen": 213742008,
+ "step": 4880,
+ "train_runtime": 37761.13,
+ "train_tokens_per_second": 5660.371
+ },
+ {
+ "epoch": 0.3289119310530568,
+ "grad_norm": 1.0245077977220034,
+ "learning_rate": 9.347646451570394e-06,
+ "loss": 0.3780977725982666,
+ "num_input_tokens_seen": 213967120,
+ "step": 4885,
+ "train_runtime": 37795.7217,
+ "train_tokens_per_second": 5661.147
+ },
+ {
+ "epoch": 0.32924858604901697,
+ "grad_norm": 0.7783625701111361,
+ "learning_rate": 9.346339980796162e-06,
+ "loss": 0.44457039833068845,
+ "num_input_tokens_seen": 214204152,
+ "step": 4890,
+ "train_runtime": 37830.001,
+ "train_tokens_per_second": 5662.282
+ },
+ {
+ "epoch": 0.3295852410449771,
+ "grad_norm": 0.7619788379384177,
+ "learning_rate": 9.345032294581813e-06,
+ "loss": 0.3991098403930664,
+ "num_input_tokens_seen": 214426328,
+ "step": 4895,
+ "train_runtime": 37861.0229,
+ "train_tokens_per_second": 5663.511
+ },
+ {
+ "epoch": 0.32992189604093725,
+ "grad_norm": 0.7063552896906414,
+ "learning_rate": 9.343723393293038e-06,
+ "loss": 0.37960023880004884,
+ "num_input_tokens_seen": 214641400,
+ "step": 4900,
+ "train_runtime": 37900.5478,
+ "train_tokens_per_second": 5663.28
+ },
+ {
+ "epoch": 0.3302585510368974,
+ "grad_norm": 0.9300418965561537,
+ "learning_rate": 9.342413277295869e-06,
+ "loss": 0.3947901725769043,
+ "num_input_tokens_seen": 214865368,
+ "step": 4905,
+ "train_runtime": 37941.8194,
+ "train_tokens_per_second": 5663.022
+ },
+ {
+ "epoch": 0.3305952060328575,
+ "grad_norm": 0.660978808262332,
+ "learning_rate": 9.341101946956672e-06,
+ "loss": 0.39415407180786133,
+ "num_input_tokens_seen": 215077016,
+ "step": 4910,
+ "train_runtime": 37976.1785,
+ "train_tokens_per_second": 5663.472
+ },
+ {
+ "epoch": 0.33093186102881766,
+ "grad_norm": 0.7480448807625937,
+ "learning_rate": 9.33978940264216e-06,
+ "loss": 0.4069984436035156,
+ "num_input_tokens_seen": 215281168,
+ "step": 4915,
+ "train_runtime": 38020.9215,
+ "train_tokens_per_second": 5662.176
+ },
+ {
+ "epoch": 0.3312685160247778,
+ "grad_norm": 0.9363661586418124,
+ "learning_rate": 9.338475644719377e-06,
+ "loss": 0.396682071685791,
+ "num_input_tokens_seen": 215508928,
+ "step": 4920,
+ "train_runtime": 38060.6534,
+ "train_tokens_per_second": 5662.25
+ },
+ {
+ "epoch": 0.33160517102073794,
+ "grad_norm": 0.7522573238030015,
+ "learning_rate": 9.337160673555719e-06,
+ "loss": 0.3950789451599121,
+ "num_input_tokens_seen": 215730176,
+ "step": 4925,
+ "train_runtime": 38096.6721,
+ "train_tokens_per_second": 5662.704
+ },
+ {
+ "epoch": 0.3319418260166981,
+ "grad_norm": 0.7866963830157505,
+ "learning_rate": 9.335844489518905e-06,
+ "loss": 0.3960768222808838,
+ "num_input_tokens_seen": 215941544,
+ "step": 4930,
+ "train_runtime": 38135.6021,
+ "train_tokens_per_second": 5662.466
+ },
+ {
+ "epoch": 0.3322784810126582,
+ "grad_norm": 0.7771252275599534,
+ "learning_rate": 9.334527092977008e-06,
+ "loss": 0.3927053928375244,
+ "num_input_tokens_seen": 216139784,
+ "step": 4935,
+ "train_runtime": 38172.7334,
+ "train_tokens_per_second": 5662.151
+ },
+ {
+ "epoch": 0.33261513600861836,
+ "grad_norm": 0.7031312284914388,
+ "learning_rate": 9.33320848429843e-06,
+ "loss": 0.3905383825302124,
+ "num_input_tokens_seen": 216376872,
+ "step": 4940,
+ "train_runtime": 38213.8052,
+ "train_tokens_per_second": 5662.27
+ },
+ {
+ "epoch": 0.3329517910045785,
+ "grad_norm": 0.7992397342780773,
+ "learning_rate": 9.331888663851917e-06,
+ "loss": 0.41385631561279296,
+ "num_input_tokens_seen": 216602432,
+ "step": 4945,
+ "train_runtime": 38247.7905,
+ "train_tokens_per_second": 5663.136
+ },
+ {
+ "epoch": 0.33328844600053864,
+ "grad_norm": 0.8325843513269255,
+ "learning_rate": 9.330567632006552e-06,
+ "loss": 0.402661657333374,
+ "num_input_tokens_seen": 216820040,
+ "step": 4950,
+ "train_runtime": 38282.5886,
+ "train_tokens_per_second": 5663.672
+ },
+ {
+ "epoch": 0.3336251009964988,
+ "grad_norm": 0.8157610005939018,
+ "learning_rate": 9.329245389131759e-06,
+ "loss": 0.4041293144226074,
+ "num_input_tokens_seen": 217040648,
+ "step": 4955,
+ "train_runtime": 38321.0859,
+ "train_tokens_per_second": 5663.74
+ },
+ {
+ "epoch": 0.3339617559924589,
+ "grad_norm": 0.7514981142846812,
+ "learning_rate": 9.327921935597298e-06,
+ "loss": 0.373291015625,
+ "num_input_tokens_seen": 217253544,
+ "step": 4960,
+ "train_runtime": 38362.9141,
+ "train_tokens_per_second": 5663.114
+ },
+ {
+ "epoch": 0.33429841098841906,
+ "grad_norm": 0.9584412932125773,
+ "learning_rate": 9.326597271773267e-06,
+ "loss": 0.39936108589172364,
+ "num_input_tokens_seen": 217452896,
+ "step": 4965,
+ "train_runtime": 38404.9607,
+ "train_tokens_per_second": 5662.104
+ },
+ {
+ "epoch": 0.3346350659843792,
+ "grad_norm": 0.7914463925442367,
+ "learning_rate": 9.325271398030107e-06,
+ "loss": 0.3928532600402832,
+ "num_input_tokens_seen": 217666152,
+ "step": 4970,
+ "train_runtime": 38441.97,
+ "train_tokens_per_second": 5662.201
+ },
+ {
+ "epoch": 0.33497172098033934,
+ "grad_norm": 0.8843279935185989,
+ "learning_rate": 9.323944314738591e-06,
+ "loss": 0.40926666259765626,
+ "num_input_tokens_seen": 217886336,
+ "step": 4975,
+ "train_runtime": 38486.6001,
+ "train_tokens_per_second": 5661.356
+ },
+ {
+ "epoch": 0.3353083759762995,
+ "grad_norm": 0.7413319586045699,
+ "learning_rate": 9.322616022269838e-06,
+ "loss": 0.3675239562988281,
+ "num_input_tokens_seen": 218092776,
+ "step": 4980,
+ "train_runtime": 38523.0427,
+ "train_tokens_per_second": 5661.359
+ },
+ {
+ "epoch": 0.3356450309722596,
+ "grad_norm": 0.8050536335877444,
+ "learning_rate": 9.321286520995297e-06,
+ "loss": 0.40448784828186035,
+ "num_input_tokens_seen": 218303136,
+ "step": 4985,
+ "train_runtime": 38563.0671,
+ "train_tokens_per_second": 5660.938
+ },
+ {
+ "epoch": 0.33598168596821976,
+ "grad_norm": 0.8220117119004713,
+ "learning_rate": 9.319955811286758e-06,
+ "loss": 0.38973236083984375,
+ "num_input_tokens_seen": 218521848,
+ "step": 4990,
+ "train_runtime": 38606.842,
+ "train_tokens_per_second": 5660.184
+ },
+ {
+ "epoch": 0.3363183409641799,
+ "grad_norm": 0.724656690749584,
+ "learning_rate": 9.318623893516354e-06,
+ "loss": 0.38318960666656493,
+ "num_input_tokens_seen": 218733712,
+ "step": 4995,
+ "train_runtime": 38643.6596,
+ "train_tokens_per_second": 5660.274
+ },
+ {
+ "epoch": 0.33665499596014004,
+ "grad_norm": 0.777455286936359,
+ "learning_rate": 9.317290768056548e-06,
+ "loss": 0.3933337450027466,
+ "num_input_tokens_seen": 218942384,
+ "step": 5000,
+ "train_runtime": 38685.6918,
+ "train_tokens_per_second": 5659.518
+ },
+ {
+ "epoch": 0.3369916509561002,
+ "grad_norm": 0.7933540714498255,
+ "learning_rate": 9.315956435280147e-06,
+ "loss": 0.40426931381225584,
+ "num_input_tokens_seen": 219148584,
+ "step": 5005,
+ "train_runtime": 38726.8749,
+ "train_tokens_per_second": 5658.824
+ },
+ {
+ "epoch": 0.3373283059520603,
+ "grad_norm": 0.7310653600558,
+ "learning_rate": 9.314620895560292e-06,
+ "loss": 0.3823359251022339,
+ "num_input_tokens_seen": 219366640,
+ "step": 5010,
+ "train_runtime": 38759.9606,
+ "train_tokens_per_second": 5659.62
+ },
+ {
+ "epoch": 0.33766496094802045,
+ "grad_norm": 0.8607865029821252,
+ "learning_rate": 9.313284149270459e-06,
+ "loss": 0.4273562431335449,
+ "num_input_tokens_seen": 219580136,
+ "step": 5015,
+ "train_runtime": 38797.085,
+ "train_tokens_per_second": 5659.707
+ },
+ {
+ "epoch": 0.3380016159439806,
+ "grad_norm": 0.8291229922855889,
+ "learning_rate": 9.311946196784469e-06,
+ "loss": 0.396860933303833,
+ "num_input_tokens_seen": 219795256,
+ "step": 5020,
+ "train_runtime": 38838.6299,
+ "train_tokens_per_second": 5659.192
+ },
+ {
+ "epoch": 0.33833827093994073,
+ "grad_norm": 0.746756264356882,
+ "learning_rate": 9.310607038476476e-06,
+ "loss": 0.40080647468566893,
+ "num_input_tokens_seen": 220007056,
+ "step": 5025,
+ "train_runtime": 38882.3681,
+ "train_tokens_per_second": 5658.273
+ },
+ {
+ "epoch": 0.3386749259359009,
+ "grad_norm": 0.7715704037440936,
+ "learning_rate": 9.30926667472097e-06,
+ "loss": 0.3821887016296387,
+ "num_input_tokens_seen": 220230040,
+ "step": 5030,
+ "train_runtime": 38929.962,
+ "train_tokens_per_second": 5657.083
+ },
+ {
+ "epoch": 0.339011580931861,
+ "grad_norm": 0.695121253798547,
+ "learning_rate": 9.307925105892779e-06,
+ "loss": 0.4106111526489258,
+ "num_input_tokens_seen": 220463304,
+ "step": 5035,
+ "train_runtime": 38964.7555,
+ "train_tokens_per_second": 5658.018
+ },
+ {
+ "epoch": 0.33934823592782115,
+ "grad_norm": 0.7852274341067195,
+ "learning_rate": 9.30658233236707e-06,
+ "loss": 0.38834619522094727,
+ "num_input_tokens_seen": 220695456,
+ "step": 5040,
+ "train_runtime": 39007.1954,
+ "train_tokens_per_second": 5657.814
+ },
+ {
+ "epoch": 0.3396848909237813,
+ "grad_norm": 0.759484699190671,
+ "learning_rate": 9.30523835451934e-06,
+ "loss": 0.3923861265182495,
+ "num_input_tokens_seen": 220915344,
+ "step": 5045,
+ "train_runtime": 39053.4813,
+ "train_tokens_per_second": 5656.739
+ },
+ {
+ "epoch": 0.34002154591974143,
+ "grad_norm": 0.7964278768577083,
+ "learning_rate": 9.303893172725437e-06,
+ "loss": 0.4306797981262207,
+ "num_input_tokens_seen": 221146072,
+ "step": 5050,
+ "train_runtime": 39091.2343,
+ "train_tokens_per_second": 5657.178
+ },
+ {
+ "epoch": 0.34035820091570157,
+ "grad_norm": 0.7340454424901695,
+ "learning_rate": 9.302546787361533e-06,
+ "loss": 0.38288087844848634,
+ "num_input_tokens_seen": 221371344,
+ "step": 5055,
+ "train_runtime": 39133.7258,
+ "train_tokens_per_second": 5656.792
+ },
+ {
+ "epoch": 0.3406948559116617,
+ "grad_norm": 0.7383658204672374,
+ "learning_rate": 9.301199198804139e-06,
+ "loss": 0.37805628776550293,
+ "num_input_tokens_seen": 221583280,
+ "step": 5060,
+ "train_runtime": 39172.7195,
+ "train_tokens_per_second": 5656.571
+ },
+ {
+ "epoch": 0.34103151090762185,
+ "grad_norm": 0.8155446108957558,
+ "learning_rate": 9.299850407430102e-06,
+ "loss": 0.4092425346374512,
+ "num_input_tokens_seen": 221808808,
+ "step": 5065,
+ "train_runtime": 39215.053,
+ "train_tokens_per_second": 5656.216
+ },
+ {
+ "epoch": 0.341368165903582,
+ "grad_norm": 0.8253751293294798,
+ "learning_rate": 9.298500413616613e-06,
+ "loss": 0.3920405864715576,
+ "num_input_tokens_seen": 222029984,
+ "step": 5070,
+ "train_runtime": 39251.0729,
+ "train_tokens_per_second": 5656.66
+ },
+ {
+ "epoch": 0.3417048208995421,
+ "grad_norm": 0.7618503900471122,
+ "learning_rate": 9.297149217741188e-06,
+ "loss": 0.3707792043685913,
+ "num_input_tokens_seen": 222251080,
+ "step": 5075,
+ "train_runtime": 39288.8528,
+ "train_tokens_per_second": 5656.848
+ },
+ {
+ "epoch": 0.34204147589550227,
+ "grad_norm": 0.7834580864284257,
+ "learning_rate": 9.295796820181688e-06,
+ "loss": 0.4238715171813965,
+ "num_input_tokens_seen": 222475224,
+ "step": 5080,
+ "train_runtime": 39326.0211,
+ "train_tokens_per_second": 5657.201
+ },
+ {
+ "epoch": 0.3423781308914624,
+ "grad_norm": 0.7875688230457064,
+ "learning_rate": 9.294443221316305e-06,
+ "loss": 0.3979609727859497,
+ "num_input_tokens_seen": 222689712,
+ "step": 5085,
+ "train_runtime": 39362.9107,
+ "train_tokens_per_second": 5657.349
+ },
+ {
+ "epoch": 0.34271478588742255,
+ "grad_norm": 0.6945475678075447,
+ "learning_rate": 9.293088421523569e-06,
+ "loss": 0.41632776260375975,
+ "num_input_tokens_seen": 222903112,
+ "step": 5090,
+ "train_runtime": 39399.1126,
+ "train_tokens_per_second": 5657.567
+ },
+ {
+ "epoch": 0.3430514408833827,
+ "grad_norm": 0.8727362030270556,
+ "learning_rate": 9.291732421182346e-06,
+ "loss": 0.3804980754852295,
+ "num_input_tokens_seen": 223116536,
+ "step": 5095,
+ "train_runtime": 39444.3485,
+ "train_tokens_per_second": 5656.489
+ },
+ {
+ "epoch": 0.3433880958793428,
+ "grad_norm": 0.7868108132642002,
+ "learning_rate": 9.290375220671837e-06,
+ "loss": 0.3959723949432373,
+ "num_input_tokens_seen": 223331328,
+ "step": 5100,
+ "train_runtime": 39483.2714,
+ "train_tokens_per_second": 5656.353
+ },
+ {
+ "epoch": 0.34372475087530296,
+ "grad_norm": 0.8977037421010565,
+ "learning_rate": 9.28901682037158e-06,
+ "loss": 0.3575240135192871,
+ "num_input_tokens_seen": 223545784,
+ "step": 5105,
+ "train_runtime": 39520.4864,
+ "train_tokens_per_second": 5656.453
+ },
+ {
+ "epoch": 0.3440614058712631,
+ "grad_norm": 0.724529969116944,
+ "learning_rate": 9.287657220661442e-06,
+ "loss": 0.41797566413879395,
+ "num_input_tokens_seen": 223772168,
+ "step": 5110,
+ "train_runtime": 39556.4135,
+ "train_tokens_per_second": 5657.039
+ },
+ {
+ "epoch": 0.34439806086722324,
+ "grad_norm": 0.8314015078496961,
+ "learning_rate": 9.286296421921636e-06,
+ "loss": 0.3681803703308105,
+ "num_input_tokens_seen": 223987880,
+ "step": 5115,
+ "train_runtime": 39597.3882,
+ "train_tokens_per_second": 5656.633
+ },
+ {
+ "epoch": 0.3447347158631834,
+ "grad_norm": 0.8879862781401929,
+ "learning_rate": 9.284934424532706e-06,
+ "loss": 0.4501188278198242,
+ "num_input_tokens_seen": 224197912,
+ "step": 5120,
+ "train_runtime": 39637.597,
+ "train_tokens_per_second": 5656.193
+ },
+ {
+ "epoch": 0.3450713708591435,
+ "grad_norm": 0.7067262341475377,
+ "learning_rate": 9.283571228875525e-06,
+ "loss": 0.35322113037109376,
+ "num_input_tokens_seen": 224398888,
+ "step": 5125,
+ "train_runtime": 39676.3313,
+ "train_tokens_per_second": 5655.737
+ },
+ {
+ "epoch": 0.3454080258551037,
+ "grad_norm": 0.7651046532838097,
+ "learning_rate": 9.282206835331311e-06,
+ "loss": 0.4238561153411865,
+ "num_input_tokens_seen": 224624840,
+ "step": 5130,
+ "train_runtime": 39714.5021,
+ "train_tokens_per_second": 5655.99
+ },
+ {
+ "epoch": 0.34574468085106386,
+ "grad_norm": 1.0540025361763135,
+ "learning_rate": 9.28084124428161e-06,
+ "loss": 0.3581863880157471,
+ "num_input_tokens_seen": 224840616,
+ "step": 5135,
+ "train_runtime": 39760.1229,
+ "train_tokens_per_second": 5654.928
+ },
+ {
+ "epoch": 0.346081335847024,
+ "grad_norm": 0.712972521802771,
+ "learning_rate": 9.279474456108305e-06,
+ "loss": 0.37324137687683107,
+ "num_input_tokens_seen": 225076296,
+ "step": 5140,
+ "train_runtime": 39795.5671,
+ "train_tokens_per_second": 5655.813
+ },
+ {
+ "epoch": 0.34641799084298414,
+ "grad_norm": 0.7436296711193636,
+ "learning_rate": 9.278106471193615e-06,
+ "loss": 0.36896820068359376,
+ "num_input_tokens_seen": 225294528,
+ "step": 5145,
+ "train_runtime": 39833.6798,
+ "train_tokens_per_second": 5655.88
+ },
+ {
+ "epoch": 0.3467546458389443,
+ "grad_norm": 0.7808819136322714,
+ "learning_rate": 9.276737289920093e-06,
+ "loss": 0.4088866710662842,
+ "num_input_tokens_seen": 225512648,
+ "step": 5150,
+ "train_runtime": 39875.84,
+ "train_tokens_per_second": 5655.37
+ },
+ {
+ "epoch": 0.3470913008349044,
+ "grad_norm": 0.74354681758287,
+ "learning_rate": 9.275366912670624e-06,
+ "loss": 0.4070155143737793,
+ "num_input_tokens_seen": 225724000,
+ "step": 5155,
+ "train_runtime": 39910.7695,
+ "train_tokens_per_second": 5655.717
+ },
+ {
+ "epoch": 0.34742795583086455,
+ "grad_norm": 0.7443700356039168,
+ "learning_rate": 9.273995339828432e-06,
+ "loss": 0.3892085075378418,
+ "num_input_tokens_seen": 225939192,
+ "step": 5160,
+ "train_runtime": 39950.0434,
+ "train_tokens_per_second": 5655.543
+ },
+ {
+ "epoch": 0.3477646108268247,
+ "grad_norm": 0.8699331338781421,
+ "learning_rate": 9.272622571777072e-06,
+ "loss": 0.42869272232055666,
+ "num_input_tokens_seen": 226161472,
+ "step": 5165,
+ "train_runtime": 39986.0253,
+ "train_tokens_per_second": 5656.013
+ },
+ {
+ "epoch": 0.34810126582278483,
+ "grad_norm": 0.7946248476997148,
+ "learning_rate": 9.271248608900434e-06,
+ "loss": 0.3663766860961914,
+ "num_input_tokens_seen": 226388672,
+ "step": 5170,
+ "train_runtime": 40022.8625,
+ "train_tokens_per_second": 5656.484
+ },
+ {
+ "epoch": 0.348437920818745,
+ "grad_norm": 0.820181936155836,
+ "learning_rate": 9.269873451582741e-06,
+ "loss": 0.3990020275115967,
+ "num_input_tokens_seen": 226604448,
+ "step": 5175,
+ "train_runtime": 40065.8484,
+ "train_tokens_per_second": 5655.801
+ },
+ {
+ "epoch": 0.3487745758147051,
+ "grad_norm": 0.7865052705818104,
+ "learning_rate": 9.268497100208556e-06,
+ "loss": 0.40553035736083987,
+ "num_input_tokens_seen": 226814192,
+ "step": 5180,
+ "train_runtime": 40108.7169,
+ "train_tokens_per_second": 5654.985
+ },
+ {
+ "epoch": 0.34911123081066525,
+ "grad_norm": 0.8597790154802004,
+ "learning_rate": 9.267119555162765e-06,
+ "loss": 0.40463762283325194,
+ "num_input_tokens_seen": 227044824,
+ "step": 5185,
+ "train_runtime": 40148.6974,
+ "train_tokens_per_second": 5655.098
+ },
+ {
+ "epoch": 0.3494478858066254,
+ "grad_norm": 0.7383395894951612,
+ "learning_rate": 9.2657408168306e-06,
+ "loss": 0.3876547336578369,
+ "num_input_tokens_seen": 227264048,
+ "step": 5190,
+ "train_runtime": 40181.6708,
+ "train_tokens_per_second": 5655.913
+ },
+ {
+ "epoch": 0.34978454080258553,
+ "grad_norm": 0.8478101930950076,
+ "learning_rate": 9.264360885597617e-06,
+ "loss": 0.3948791027069092,
+ "num_input_tokens_seen": 227482568,
+ "step": 5195,
+ "train_runtime": 40224.9641,
+ "train_tokens_per_second": 5655.258
+ },
+ {
+ "epoch": 0.35012119579854567,
+ "grad_norm": 0.8312712862431509,
+ "learning_rate": 9.262979761849709e-06,
+ "loss": 0.3999017238616943,
+ "num_input_tokens_seen": 227701672,
+ "step": 5200,
+ "train_runtime": 40265.2495,
+ "train_tokens_per_second": 5655.042
+ },
+ {
+ "epoch": 0.3504578507945058,
+ "grad_norm": 0.8047649086754547,
+ "learning_rate": 9.261597445973106e-06,
+ "loss": 0.40303988456726075,
+ "num_input_tokens_seen": 227930736,
+ "step": 5205,
+ "train_runtime": 40300.5889,
+ "train_tokens_per_second": 5655.767
+ },
+ {
+ "epoch": 0.35079450579046595,
+ "grad_norm": 0.7279256779386029,
+ "learning_rate": 9.260213938354365e-06,
+ "loss": 0.38801677227020265,
+ "num_input_tokens_seen": 228142248,
+ "step": 5210,
+ "train_runtime": 40339.465,
+ "train_tokens_per_second": 5655.56
+ },
+ {
+ "epoch": 0.3511311607864261,
+ "grad_norm": 0.703365188592268,
+ "learning_rate": 9.25882923938038e-06,
+ "loss": 0.40598435401916505,
+ "num_input_tokens_seen": 228355000,
+ "step": 5215,
+ "train_runtime": 40382.8232,
+ "train_tokens_per_second": 5654.756
+ },
+ {
+ "epoch": 0.35146781578238623,
+ "grad_norm": 0.8641270145143162,
+ "learning_rate": 9.257443349438382e-06,
+ "loss": 0.42745413780212405,
+ "num_input_tokens_seen": 228579256,
+ "step": 5220,
+ "train_runtime": 40419.0219,
+ "train_tokens_per_second": 5655.24
+ },
+ {
+ "epoch": 0.35180447077834637,
+ "grad_norm": 0.8147972074351507,
+ "learning_rate": 9.256056268915924e-06,
+ "loss": 0.40657625198364256,
+ "num_input_tokens_seen": 228801840,
+ "step": 5225,
+ "train_runtime": 40461.4517,
+ "train_tokens_per_second": 5654.81
+ },
+ {
+ "epoch": 0.3521411257743065,
+ "grad_norm": 0.7053549338628146,
+ "learning_rate": 9.254667998200906e-06,
+ "loss": 0.4108288764953613,
+ "num_input_tokens_seen": 229023432,
+ "step": 5230,
+ "train_runtime": 40502.8889,
+ "train_tokens_per_second": 5654.496
+ },
+ {
+ "epoch": 0.35247778077026665,
+ "grad_norm": 0.8352493252466918,
+ "learning_rate": 9.253278537681548e-06,
+ "loss": 0.4314168930053711,
+ "num_input_tokens_seen": 229259792,
+ "step": 5235,
+ "train_runtime": 40538.046,
+ "train_tokens_per_second": 5655.423
+ },
+ {
+ "epoch": 0.3528144357662268,
+ "grad_norm": 0.7807713230559138,
+ "learning_rate": 9.251887887746407e-06,
+ "loss": 0.3931559085845947,
+ "num_input_tokens_seen": 229485656,
+ "step": 5240,
+ "train_runtime": 40577.5224,
+ "train_tokens_per_second": 5655.487
+ },
+ {
+ "epoch": 0.3531510907621869,
+ "grad_norm": 0.8636190729399873,
+ "learning_rate": 9.25049604878438e-06,
+ "loss": 0.38311750888824464,
+ "num_input_tokens_seen": 229704840,
+ "step": 5245,
+ "train_runtime": 40615.5331,
+ "train_tokens_per_second": 5655.591
+ },
+ {
+ "epoch": 0.35348774575814707,
+ "grad_norm": 0.8592849372799231,
+ "learning_rate": 9.249103021184684e-06,
+ "loss": 0.3960615634918213,
+ "num_input_tokens_seen": 229920256,
+ "step": 5250,
+ "train_runtime": 40653.9984,
+ "train_tokens_per_second": 5655.539
+ },
+ {
+ "epoch": 0.3538244007541072,
+ "grad_norm": 0.7502281989758054,
+ "learning_rate": 9.24770880533688e-06,
+ "loss": 0.4205930709838867,
+ "num_input_tokens_seen": 230150768,
+ "step": 5255,
+ "train_runtime": 40691.9738,
+ "train_tokens_per_second": 5655.925
+ },
+ {
+ "epoch": 0.35416105575006734,
+ "grad_norm": 0.9736374450794365,
+ "learning_rate": 9.24631340163085e-06,
+ "loss": 0.41542859077453614,
+ "num_input_tokens_seen": 230338192,
+ "step": 5260,
+ "train_runtime": 40727.6217,
+ "train_tokens_per_second": 5655.577
+ },
+ {
+ "epoch": 0.3544977107460275,
+ "grad_norm": 0.7629428550659803,
+ "learning_rate": 9.244916810456822e-06,
+ "loss": 0.3574668884277344,
+ "num_input_tokens_seen": 230568472,
+ "step": 5265,
+ "train_runtime": 40758.3906,
+ "train_tokens_per_second": 5656.957
+ },
+ {
+ "epoch": 0.3548343657419876,
+ "grad_norm": 0.7107248914856642,
+ "learning_rate": 9.24351903220534e-06,
+ "loss": 0.40234909057617185,
+ "num_input_tokens_seen": 230793504,
+ "step": 5270,
+ "train_runtime": 40798.996,
+ "train_tokens_per_second": 5656.843
+ },
+ {
+ "epoch": 0.35517102073794776,
+ "grad_norm": 0.8468122239159248,
+ "learning_rate": 9.242120067267295e-06,
+ "loss": 0.3589898347854614,
+ "num_input_tokens_seen": 231011256,
+ "step": 5275,
+ "train_runtime": 40837.4422,
+ "train_tokens_per_second": 5656.849
+ },
+ {
+ "epoch": 0.3555076757339079,
+ "grad_norm": 0.8384350564515061,
+ "learning_rate": 9.2407199160339e-06,
+ "loss": 0.3916470050811768,
+ "num_input_tokens_seen": 231224304,
+ "step": 5280,
+ "train_runtime": 40879.0573,
+ "train_tokens_per_second": 5656.302
+ },
+ {
+ "epoch": 0.35584433072986804,
+ "grad_norm": 0.7878128731634584,
+ "learning_rate": 9.239318578896701e-06,
+ "loss": 0.3869880199432373,
+ "num_input_tokens_seen": 231440576,
+ "step": 5285,
+ "train_runtime": 40917.8128,
+ "train_tokens_per_second": 5656.23
+ },
+ {
+ "epoch": 0.3561809857258282,
+ "grad_norm": 0.9404923069120379,
+ "learning_rate": 9.237916056247583e-06,
+ "loss": 0.40935673713684084,
+ "num_input_tokens_seen": 231653280,
+ "step": 5290,
+ "train_runtime": 40955.4085,
+ "train_tokens_per_second": 5656.232
+ },
+ {
+ "epoch": 0.3565176407217883,
+ "grad_norm": 0.790253055764951,
+ "learning_rate": 9.23651234847875e-06,
+ "loss": 0.4153427600860596,
+ "num_input_tokens_seen": 231867296,
+ "step": 5295,
+ "train_runtime": 40993.5671,
+ "train_tokens_per_second": 5656.187
+ },
+ {
+ "epoch": 0.35685429571774846,
+ "grad_norm": 0.8209369098367891,
+ "learning_rate": 9.23510745598275e-06,
+ "loss": 0.3877981185913086,
+ "num_input_tokens_seen": 232073536,
+ "step": 5300,
+ "train_runtime": 41032.3162,
+ "train_tokens_per_second": 5655.872
+ },
+ {
+ "epoch": 0.3571909507137086,
+ "grad_norm": 0.6660402351658175,
+ "learning_rate": 9.233701379152456e-06,
+ "loss": 0.3829216957092285,
+ "num_input_tokens_seen": 232295728,
+ "step": 5305,
+ "train_runtime": 41070.6587,
+ "train_tokens_per_second": 5656.002
+ },
+ {
+ "epoch": 0.35752760570966874,
+ "grad_norm": 0.7635535443595722,
+ "learning_rate": 9.232294118381069e-06,
+ "loss": 0.41249933242797854,
+ "num_input_tokens_seen": 232520456,
+ "step": 5310,
+ "train_runtime": 41107.4158,
+ "train_tokens_per_second": 5656.411
+ },
+ {
+ "epoch": 0.3578642607056289,
+ "grad_norm": 0.7665309192538949,
+ "learning_rate": 9.230885674062129e-06,
+ "loss": 0.3704015493392944,
+ "num_input_tokens_seen": 232740384,
+ "step": 5315,
+ "train_runtime": 41146.5227,
+ "train_tokens_per_second": 5656.38
+ },
+ {
+ "epoch": 0.358200915701589,
+ "grad_norm": 0.805325429526549,
+ "learning_rate": 9.2294760465895e-06,
+ "loss": 0.4049508094787598,
+ "num_input_tokens_seen": 232969144,
+ "step": 5320,
+ "train_runtime": 41180.0542,
+ "train_tokens_per_second": 5657.33
+ },
+ {
+ "epoch": 0.35853757069754916,
+ "grad_norm": 0.8309484529457033,
+ "learning_rate": 9.228065236357383e-06,
+ "loss": 0.39788720607757566,
+ "num_input_tokens_seen": 233196464,
+ "step": 5325,
+ "train_runtime": 41217.3194,
+ "train_tokens_per_second": 5657.73
+ },
+ {
+ "epoch": 0.3588742256935093,
+ "grad_norm": 0.7008865285408351,
+ "learning_rate": 9.226653243760304e-06,
+ "loss": 0.3763267993927002,
+ "num_input_tokens_seen": 233413024,
+ "step": 5330,
+ "train_runtime": 41251.2257,
+ "train_tokens_per_second": 5658.329
+ },
+ {
+ "epoch": 0.35921088068946944,
+ "grad_norm": 0.7229398353747315,
+ "learning_rate": 9.225240069193124e-06,
+ "loss": 0.4120461463928223,
+ "num_input_tokens_seen": 233629584,
+ "step": 5335,
+ "train_runtime": 41295.9805,
+ "train_tokens_per_second": 5657.441
+ },
+ {
+ "epoch": 0.3595475356854296,
+ "grad_norm": 0.854235630422482,
+ "learning_rate": 9.223825713051034e-06,
+ "loss": 0.4018904685974121,
+ "num_input_tokens_seen": 233844536,
+ "step": 5340,
+ "train_runtime": 41326.7978,
+ "train_tokens_per_second": 5658.424
+ },
+ {
+ "epoch": 0.3598841906813897,
+ "grad_norm": 0.8244749095839433,
+ "learning_rate": 9.222410175729549e-06,
+ "loss": 0.38492045402526853,
+ "num_input_tokens_seen": 234063096,
+ "step": 5345,
+ "train_runtime": 41367.4954,
+ "train_tokens_per_second": 5658.14
+ },
+ {
+ "epoch": 0.36022084567734985,
+ "grad_norm": 0.7424522527415252,
+ "learning_rate": 9.220993457624523e-06,
+ "loss": 0.3723012924194336,
+ "num_input_tokens_seen": 234279256,
+ "step": 5350,
+ "train_runtime": 41414.7783,
+ "train_tokens_per_second": 5656.9
+ },
+ {
+ "epoch": 0.36055750067331,
+ "grad_norm": 0.8296218729091308,
+ "learning_rate": 9.219575559132136e-06,
+ "loss": 0.4064267635345459,
+ "num_input_tokens_seen": 234496160,
+ "step": 5355,
+ "train_runtime": 41454.7573,
+ "train_tokens_per_second": 5656.677
+ },
+ {
+ "epoch": 0.36089415566927013,
+ "grad_norm": 0.8270954179905022,
+ "learning_rate": 9.2181564806489e-06,
+ "loss": 0.3711550235748291,
+ "num_input_tokens_seen": 234705872,
+ "step": 5360,
+ "train_runtime": 41488.3956,
+ "train_tokens_per_second": 5657.145
+ },
+ {
+ "epoch": 0.3612308106652303,
+ "grad_norm": 0.7010804881835582,
+ "learning_rate": 9.216736222571655e-06,
+ "loss": 0.36968867778778075,
+ "num_input_tokens_seen": 234928056,
+ "step": 5365,
+ "train_runtime": 41528.5306,
+ "train_tokens_per_second": 5657.028
+ },
+ {
+ "epoch": 0.3615674656611904,
+ "grad_norm": 0.7909884318309988,
+ "learning_rate": 9.215314785297572e-06,
+ "loss": 0.4384300708770752,
+ "num_input_tokens_seen": 235127224,
+ "step": 5370,
+ "train_runtime": 41566.5449,
+ "train_tokens_per_second": 5656.646
+ },
+ {
+ "epoch": 0.36190412065715055,
+ "grad_norm": 0.7730294782111428,
+ "learning_rate": 9.21389216922415e-06,
+ "loss": 0.3896491527557373,
+ "num_input_tokens_seen": 235350232,
+ "step": 5375,
+ "train_runtime": 41604.4719,
+ "train_tokens_per_second": 5656.849
+ },
+ {
+ "epoch": 0.3622407756531107,
+ "grad_norm": 1.5247558034207205,
+ "learning_rate": 9.212468374749222e-06,
+ "loss": 0.4335320949554443,
+ "num_input_tokens_seen": 235570704,
+ "step": 5380,
+ "train_runtime": 41642.8804,
+ "train_tokens_per_second": 5656.926
+ },
+ {
+ "epoch": 0.36257743064907083,
+ "grad_norm": 0.7761987245970543,
+ "learning_rate": 9.211043402270942e-06,
+ "loss": 0.39116473197937013,
+ "num_input_tokens_seen": 235785280,
+ "step": 5385,
+ "train_runtime": 41679.7538,
+ "train_tokens_per_second": 5657.07
+ },
+ {
+ "epoch": 0.36291408564503097,
+ "grad_norm": 0.7417904936648765,
+ "learning_rate": 9.209617252187806e-06,
+ "loss": 0.3792844533920288,
+ "num_input_tokens_seen": 235999696,
+ "step": 5390,
+ "train_runtime": 41717.5624,
+ "train_tokens_per_second": 5657.083
+ },
+ {
+ "epoch": 0.3632507406409911,
+ "grad_norm": 0.8168071916778651,
+ "learning_rate": 9.208189924898627e-06,
+ "loss": 0.3464975833892822,
+ "num_input_tokens_seen": 236210480,
+ "step": 5395,
+ "train_runtime": 41760.5338,
+ "train_tokens_per_second": 5656.309
+ },
+ {
+ "epoch": 0.36358739563695125,
+ "grad_norm": 0.7857550769816507,
+ "learning_rate": 9.206761420802554e-06,
+ "loss": 0.4210227966308594,
+ "num_input_tokens_seen": 236432304,
+ "step": 5400,
+ "train_runtime": 41802.7339,
+ "train_tokens_per_second": 5655.905
+ },
+ {
+ "epoch": 0.3639240506329114,
+ "grad_norm": 0.724781136378123,
+ "learning_rate": 9.205331740299065e-06,
+ "loss": 0.3803566932678223,
+ "num_input_tokens_seen": 236642552,
+ "step": 5405,
+ "train_runtime": 41844.665,
+ "train_tokens_per_second": 5655.262
+ },
+ {
+ "epoch": 0.36426070562887153,
+ "grad_norm": 0.7382233662482024,
+ "learning_rate": 9.203900883787967e-06,
+ "loss": 0.4222862243652344,
+ "num_input_tokens_seen": 236866080,
+ "step": 5410,
+ "train_runtime": 41880.1302,
+ "train_tokens_per_second": 5655.811
+ },
+ {
+ "epoch": 0.36459736062483167,
+ "grad_norm": 0.7707623946075276,
+ "learning_rate": 9.202468851669388e-06,
+ "loss": 0.42064590454101564,
+ "num_input_tokens_seen": 237084656,
+ "step": 5415,
+ "train_runtime": 41922.4674,
+ "train_tokens_per_second": 5655.313
+ },
+ {
+ "epoch": 0.3649340156207918,
+ "grad_norm": 0.7505014032426987,
+ "learning_rate": 9.201035644343797e-06,
+ "loss": 0.40491809844970705,
+ "num_input_tokens_seen": 237299088,
+ "step": 5420,
+ "train_runtime": 41963.0407,
+ "train_tokens_per_second": 5654.955
+ },
+ {
+ "epoch": 0.36527067061675195,
+ "grad_norm": 0.8419562349359976,
+ "learning_rate": 9.199601262211985e-06,
+ "loss": 0.3938578128814697,
+ "num_input_tokens_seen": 237505024,
+ "step": 5425,
+ "train_runtime": 41999.6624,
+ "train_tokens_per_second": 5654.927
+ },
+ {
+ "epoch": 0.3656073256127121,
+ "grad_norm": 0.7844116167185397,
+ "learning_rate": 9.19816570567507e-06,
+ "loss": 0.4014750480651855,
+ "num_input_tokens_seen": 237729312,
+ "step": 5430,
+ "train_runtime": 42038.7844,
+ "train_tokens_per_second": 5655.0
+ },
+ {
+ "epoch": 0.3659439806086722,
+ "grad_norm": 0.8709777369310681,
+ "learning_rate": 9.196728975134503e-06,
+ "loss": 0.41324663162231445,
+ "num_input_tokens_seen": 237937544,
+ "step": 5435,
+ "train_runtime": 42082.2656,
+ "train_tokens_per_second": 5654.105
+ },
+ {
+ "epoch": 0.36628063560463237,
+ "grad_norm": 0.943067660878324,
+ "learning_rate": 9.195291070992061e-06,
+ "loss": 0.39972093105316164,
+ "num_input_tokens_seen": 238144328,
+ "step": 5440,
+ "train_runtime": 42119.7987,
+ "train_tokens_per_second": 5653.976
+ },
+ {
+ "epoch": 0.3666172906005925,
+ "grad_norm": 0.7866408242613544,
+ "learning_rate": 9.193851993649849e-06,
+ "loss": 0.36598644256591795,
+ "num_input_tokens_seen": 238371752,
+ "step": 5445,
+ "train_runtime": 42157.0955,
+ "train_tokens_per_second": 5654.368
+ },
+ {
+ "epoch": 0.36695394559655264,
+ "grad_norm": 0.7572508667451172,
+ "learning_rate": 9.1924117435103e-06,
+ "loss": 0.3939794063568115,
+ "num_input_tokens_seen": 238586128,
+ "step": 5450,
+ "train_runtime": 42198.1981,
+ "train_tokens_per_second": 5653.941
+ },
+ {
+ "epoch": 0.3672906005925128,
+ "grad_norm": 0.885004223430614,
+ "learning_rate": 9.190970320976176e-06,
+ "loss": 0.3719717264175415,
+ "num_input_tokens_seen": 238811344,
+ "step": 5455,
+ "train_runtime": 42232.691,
+ "train_tokens_per_second": 5654.656
+ },
+ {
+ "epoch": 0.3676272555884729,
+ "grad_norm": 0.8940362227103387,
+ "learning_rate": 9.189527726450565e-06,
+ "loss": 0.4350133895874023,
+ "num_input_tokens_seen": 239003208,
+ "step": 5460,
+ "train_runtime": 42266.5178,
+ "train_tokens_per_second": 5654.67
+ },
+ {
+ "epoch": 0.36796391058443306,
+ "grad_norm": 0.6837019733666716,
+ "learning_rate": 9.188083960336885e-06,
+ "loss": 0.3788605213165283,
+ "num_input_tokens_seen": 239225040,
+ "step": 5465,
+ "train_runtime": 42305.7631,
+ "train_tokens_per_second": 5654.668
+ },
+ {
+ "epoch": 0.3683005655803932,
+ "grad_norm": 0.7278725929539929,
+ "learning_rate": 9.186639023038879e-06,
+ "loss": 0.4101285934448242,
+ "num_input_tokens_seen": 239445688,
+ "step": 5470,
+ "train_runtime": 42345.9353,
+ "train_tokens_per_second": 5654.514
+ },
+ {
+ "epoch": 0.36863722057635334,
+ "grad_norm": 0.8240745816268156,
+ "learning_rate": 9.185192914960622e-06,
+ "loss": 0.43002777099609374,
+ "num_input_tokens_seen": 239660856,
+ "step": 5475,
+ "train_runtime": 42383.8589,
+ "train_tokens_per_second": 5654.531
+ },
+ {
+ "epoch": 0.3689738755723135,
+ "grad_norm": 0.7561111927294512,
+ "learning_rate": 9.183745636506511e-06,
+ "loss": 0.400468635559082,
+ "num_input_tokens_seen": 239863160,
+ "step": 5480,
+ "train_runtime": 42415.6347,
+ "train_tokens_per_second": 5655.065
+ },
+ {
+ "epoch": 0.3693105305682736,
+ "grad_norm": 0.6657976994723397,
+ "learning_rate": 9.182297188081274e-06,
+ "loss": 0.39147305488586426,
+ "num_input_tokens_seen": 240088504,
+ "step": 5485,
+ "train_runtime": 42445.5983,
+ "train_tokens_per_second": 5656.382
+ },
+ {
+ "epoch": 0.36964718556423376,
+ "grad_norm": 0.7438674827381799,
+ "learning_rate": 9.180847570089964e-06,
+ "loss": 0.38025264739990233,
+ "num_input_tokens_seen": 240285080,
+ "step": 5490,
+ "train_runtime": 42487.4443,
+ "train_tokens_per_second": 5655.437
+ },
+ {
+ "epoch": 0.3699838405601939,
+ "grad_norm": 0.7687571314606845,
+ "learning_rate": 9.179396782937965e-06,
+ "loss": 0.4285304546356201,
+ "num_input_tokens_seen": 240518464,
+ "step": 5495,
+ "train_runtime": 42522.1167,
+ "train_tokens_per_second": 5656.314
+ },
+ {
+ "epoch": 0.37032049555615404,
+ "grad_norm": 0.8258932109796826,
+ "learning_rate": 9.17794482703098e-06,
+ "loss": 0.41852726936340334,
+ "num_input_tokens_seen": 240737112,
+ "step": 5500,
+ "train_runtime": 42561.0578,
+ "train_tokens_per_second": 5656.277
+ },
+ {
+ "epoch": 0.3706571505521142,
+ "grad_norm": 0.7712097331858169,
+ "learning_rate": 9.176491702775049e-06,
+ "loss": 0.4076284408569336,
+ "num_input_tokens_seen": 240940632,
+ "step": 5505,
+ "train_runtime": 42605.8265,
+ "train_tokens_per_second": 5655.11
+ },
+ {
+ "epoch": 0.3709938055480743,
+ "grad_norm": 0.7293507629162065,
+ "learning_rate": 9.17503741057653e-06,
+ "loss": 0.42336077690124513,
+ "num_input_tokens_seen": 241163512,
+ "step": 5510,
+ "train_runtime": 42641.1976,
+ "train_tokens_per_second": 5655.646
+ },
+ {
+ "epoch": 0.37133046054403446,
+ "grad_norm": 0.7873802897373652,
+ "learning_rate": 9.173581950842111e-06,
+ "loss": 0.3858926773071289,
+ "num_input_tokens_seen": 241380280,
+ "step": 5515,
+ "train_runtime": 42677.1161,
+ "train_tokens_per_second": 5655.965
+ },
+ {
+ "epoch": 0.3716671155399946,
+ "grad_norm": 0.833898462602849,
+ "learning_rate": 9.172125323978811e-06,
+ "loss": 0.39653089046478274,
+ "num_input_tokens_seen": 241601112,
+ "step": 5520,
+ "train_runtime": 42713.21,
+ "train_tokens_per_second": 5656.356
+ },
+ {
+ "epoch": 0.37200377053595474,
+ "grad_norm": 0.7933645694043111,
+ "learning_rate": 9.170667530393966e-06,
+ "loss": 0.4098936080932617,
+ "num_input_tokens_seen": 241828432,
+ "step": 5525,
+ "train_runtime": 42747.8553,
+ "train_tokens_per_second": 5657.089
+ },
+ {
+ "epoch": 0.3723404255319149,
+ "grad_norm": 0.724581396473804,
+ "learning_rate": 9.169208570495247e-06,
+ "loss": 0.3672953605651855,
+ "num_input_tokens_seen": 242051344,
+ "step": 5530,
+ "train_runtime": 42787.8177,
+ "train_tokens_per_second": 5657.015
+ },
+ {
+ "epoch": 0.372677080527875,
+ "grad_norm": 0.720329801926469,
+ "learning_rate": 9.167748444690646e-06,
+ "loss": 0.3897575855255127,
+ "num_input_tokens_seen": 242250584,
+ "step": 5535,
+ "train_runtime": 42826.4052,
+ "train_tokens_per_second": 5656.571
+ },
+ {
+ "epoch": 0.37301373552383515,
+ "grad_norm": 0.7934647008536321,
+ "learning_rate": 9.16628715338848e-06,
+ "loss": 0.3871308326721191,
+ "num_input_tokens_seen": 242459936,
+ "step": 5540,
+ "train_runtime": 42859.1014,
+ "train_tokens_per_second": 5657.14
+ },
+ {
+ "epoch": 0.3733503905197953,
+ "grad_norm": 0.7931279904448103,
+ "learning_rate": 9.164824696997398e-06,
+ "loss": 0.37985944747924805,
+ "num_input_tokens_seen": 242665952,
+ "step": 5545,
+ "train_runtime": 42896.1293,
+ "train_tokens_per_second": 5657.059
+ },
+ {
+ "epoch": 0.37368704551575543,
+ "grad_norm": 0.7548635507515563,
+ "learning_rate": 9.163361075926371e-06,
+ "loss": 0.4154071807861328,
+ "num_input_tokens_seen": 242885960,
+ "step": 5550,
+ "train_runtime": 42932.7348,
+ "train_tokens_per_second": 5657.361
+ },
+ {
+ "epoch": 0.3740237005117156,
+ "grad_norm": 0.824128032145257,
+ "learning_rate": 9.161896290584695e-06,
+ "loss": 0.37793383598327634,
+ "num_input_tokens_seen": 243108432,
+ "step": 5555,
+ "train_runtime": 42974.4088,
+ "train_tokens_per_second": 5657.051
+ },
+ {
+ "epoch": 0.3743603555076757,
+ "grad_norm": 0.755788883617862,
+ "learning_rate": 9.160430341381991e-06,
+ "loss": 0.3929650545120239,
+ "num_input_tokens_seen": 243321200,
+ "step": 5560,
+ "train_runtime": 43020.7871,
+ "train_tokens_per_second": 5655.898
+ },
+ {
+ "epoch": 0.37469701050363585,
+ "grad_norm": 0.8046945349779168,
+ "learning_rate": 9.15896322872821e-06,
+ "loss": 0.39426612854003906,
+ "num_input_tokens_seen": 243541160,
+ "step": 5565,
+ "train_runtime": 43065.6274,
+ "train_tokens_per_second": 5655.117
+ },
+ {
+ "epoch": 0.375033665499596,
+ "grad_norm": 0.8408220299669169,
+ "learning_rate": 9.157494953033625e-06,
+ "loss": 0.4226834297180176,
+ "num_input_tokens_seen": 243766296,
+ "step": 5570,
+ "train_runtime": 43100.4977,
+ "train_tokens_per_second": 5655.765
+ },
+ {
+ "epoch": 0.37537032049555613,
+ "grad_norm": 0.7701423462069609,
+ "learning_rate": 9.156025514708832e-06,
+ "loss": 0.40192432403564454,
+ "num_input_tokens_seen": 244002616,
+ "step": 5575,
+ "train_runtime": 43146.2968,
+ "train_tokens_per_second": 5655.239
+ },
+ {
+ "epoch": 0.37570697549151627,
+ "grad_norm": 0.7773289349591549,
+ "learning_rate": 9.154554914164757e-06,
+ "loss": 0.38442695140838623,
+ "num_input_tokens_seen": 244237240,
+ "step": 5580,
+ "train_runtime": 43188.3958,
+ "train_tokens_per_second": 5655.159
+ },
+ {
+ "epoch": 0.3760436304874764,
+ "grad_norm": 0.6847338403952183,
+ "learning_rate": 9.153083151812649e-06,
+ "loss": 0.377176308631897,
+ "num_input_tokens_seen": 244463576,
+ "step": 5585,
+ "train_runtime": 43225.3626,
+ "train_tokens_per_second": 5655.559
+ },
+ {
+ "epoch": 0.37638028548343655,
+ "grad_norm": 0.780657947894897,
+ "learning_rate": 9.151610228064078e-06,
+ "loss": 0.40382232666015627,
+ "num_input_tokens_seen": 244685816,
+ "step": 5590,
+ "train_runtime": 43264.9573,
+ "train_tokens_per_second": 5655.52
+ },
+ {
+ "epoch": 0.3767169404793967,
+ "grad_norm": 0.7617598851673253,
+ "learning_rate": 9.150136143330949e-06,
+ "loss": 0.42247867584228516,
+ "num_input_tokens_seen": 244901992,
+ "step": 5595,
+ "train_runtime": 43305.9839,
+ "train_tokens_per_second": 5655.154
+ },
+ {
+ "epoch": 0.37705359547535683,
+ "grad_norm": 0.6965545191686713,
+ "learning_rate": 9.14866089802548e-06,
+ "loss": 0.3760505676269531,
+ "num_input_tokens_seen": 245127560,
+ "step": 5600,
+ "train_runtime": 43348.427,
+ "train_tokens_per_second": 5654.82
+ },
+ {
+ "epoch": 0.37739025047131697,
+ "grad_norm": 0.8578077401564244,
+ "learning_rate": 9.147184492560219e-06,
+ "loss": 0.3698588848114014,
+ "num_input_tokens_seen": 245339952,
+ "step": 5605,
+ "train_runtime": 43388.8794,
+ "train_tokens_per_second": 5654.443
+ },
+ {
+ "epoch": 0.3777269054672771,
+ "grad_norm": 1.0507318076347205,
+ "learning_rate": 9.14570692734804e-06,
+ "loss": 0.3787434816360474,
+ "num_input_tokens_seen": 245565672,
+ "step": 5610,
+ "train_runtime": 43425.1962,
+ "train_tokens_per_second": 5654.912
+ },
+ {
+ "epoch": 0.37806356046323725,
+ "grad_norm": 0.8234417174719115,
+ "learning_rate": 9.144228202802137e-06,
+ "loss": 0.41345486640930174,
+ "num_input_tokens_seen": 245793416,
+ "step": 5615,
+ "train_runtime": 43457.0254,
+ "train_tokens_per_second": 5656.011
+ },
+ {
+ "epoch": 0.37840021545919744,
+ "grad_norm": 0.8147806646148233,
+ "learning_rate": 9.142748319336033e-06,
+ "loss": 0.39602549076080323,
+ "num_input_tokens_seen": 246014600,
+ "step": 5620,
+ "train_runtime": 43495.3144,
+ "train_tokens_per_second": 5656.117
+ },
+ {
+ "epoch": 0.3787368704551576,
+ "grad_norm": 0.8294007935657565,
+ "learning_rate": 9.141267277363573e-06,
+ "loss": 0.3865189552307129,
+ "num_input_tokens_seen": 246209040,
+ "step": 5625,
+ "train_runtime": 43530.2862,
+ "train_tokens_per_second": 5656.04
+ },
+ {
+ "epoch": 0.3790735254511177,
+ "grad_norm": 0.8057103649691175,
+ "learning_rate": 9.139785077298921e-06,
+ "loss": 0.3669655084609985,
+ "num_input_tokens_seen": 246435304,
+ "step": 5630,
+ "train_runtime": 43570.2665,
+ "train_tokens_per_second": 5656.043
+ },
+ {
+ "epoch": 0.37941018044707786,
+ "grad_norm": 0.7570731119383132,
+ "learning_rate": 9.138301719556577e-06,
+ "loss": 0.3826982736587524,
+ "num_input_tokens_seen": 246652712,
+ "step": 5635,
+ "train_runtime": 43609.2254,
+ "train_tokens_per_second": 5655.976
+ },
+ {
+ "epoch": 0.379746835443038,
+ "grad_norm": 0.7474769926348979,
+ "learning_rate": 9.136817204551352e-06,
+ "loss": 0.386492919921875,
+ "num_input_tokens_seen": 246882568,
+ "step": 5640,
+ "train_runtime": 43646.3585,
+ "train_tokens_per_second": 5656.43
+ },
+ {
+ "epoch": 0.38008349043899814,
+ "grad_norm": 0.7350249063601988,
+ "learning_rate": 9.135331532698385e-06,
+ "loss": 0.3880929470062256,
+ "num_input_tokens_seen": 247094472,
+ "step": 5645,
+ "train_runtime": 43685.4708,
+ "train_tokens_per_second": 5656.216
+ },
+ {
+ "epoch": 0.3804201454349583,
+ "grad_norm": 0.7440003142918237,
+ "learning_rate": 9.133844704413144e-06,
+ "loss": 0.37475876808166503,
+ "num_input_tokens_seen": 247305056,
+ "step": 5650,
+ "train_runtime": 43724.0377,
+ "train_tokens_per_second": 5656.043
+ },
+ {
+ "epoch": 0.3807568004309184,
+ "grad_norm": 0.829580642455565,
+ "learning_rate": 9.132356720111413e-06,
+ "loss": 0.40270652770996096,
+ "num_input_tokens_seen": 247521856,
+ "step": 5655,
+ "train_runtime": 43766.9087,
+ "train_tokens_per_second": 5655.457
+ },
+ {
+ "epoch": 0.38109345542687856,
+ "grad_norm": 0.7890155830011499,
+ "learning_rate": 9.1308675802093e-06,
+ "loss": 0.3874132394790649,
+ "num_input_tokens_seen": 247735120,
+ "step": 5660,
+ "train_runtime": 43801.888,
+ "train_tokens_per_second": 5655.809
+ },
+ {
+ "epoch": 0.3814301104228387,
+ "grad_norm": 0.7482568242628747,
+ "learning_rate": 9.129377285123241e-06,
+ "loss": 0.3984661102294922,
+ "num_input_tokens_seen": 247949328,
+ "step": 5665,
+ "train_runtime": 43845.4472,
+ "train_tokens_per_second": 5655.076
+ },
+ {
+ "epoch": 0.38176676541879884,
+ "grad_norm": 0.7947458109924397,
+ "learning_rate": 9.127885835269996e-06,
+ "loss": 0.4090867519378662,
+ "num_input_tokens_seen": 248174032,
+ "step": 5670,
+ "train_runtime": 43879.3187,
+ "train_tokens_per_second": 5655.831
+ },
+ {
+ "epoch": 0.382103420414759,
+ "grad_norm": 0.726446478217637,
+ "learning_rate": 9.126393231066636e-06,
+ "loss": 0.3848613977432251,
+ "num_input_tokens_seen": 248384368,
+ "step": 5675,
+ "train_runtime": 43917.9285,
+ "train_tokens_per_second": 5655.649
+ },
+ {
+ "epoch": 0.3824400754107191,
+ "grad_norm": 0.7750970581531168,
+ "learning_rate": 9.124899472930568e-06,
+ "loss": 0.4047283172607422,
+ "num_input_tokens_seen": 248607600,
+ "step": 5680,
+ "train_runtime": 43961.6528,
+ "train_tokens_per_second": 5655.101
+ },
+ {
+ "epoch": 0.38277673040667926,
+ "grad_norm": 1.9536806017675672,
+ "learning_rate": 9.123404561279517e-06,
+ "loss": 0.41053128242492676,
+ "num_input_tokens_seen": 248811928,
+ "step": 5685,
+ "train_runtime": 44006.8302,
+ "train_tokens_per_second": 5653.939
+ },
+ {
+ "epoch": 0.3831133854026394,
+ "grad_norm": 0.7123153967912388,
+ "learning_rate": 9.121908496531527e-06,
+ "loss": 0.3802008867263794,
+ "num_input_tokens_seen": 249035424,
+ "step": 5690,
+ "train_runtime": 44040.1743,
+ "train_tokens_per_second": 5654.733
+ },
+ {
+ "epoch": 0.38345004039859953,
+ "grad_norm": 0.7190654994858774,
+ "learning_rate": 9.120411279104971e-06,
+ "loss": 0.3904706954956055,
+ "num_input_tokens_seen": 249254456,
+ "step": 5695,
+ "train_runtime": 44083.466,
+ "train_tokens_per_second": 5654.148
+ },
+ {
+ "epoch": 0.3837866953945597,
+ "grad_norm": 0.9187897813904784,
+ "learning_rate": 9.118912909418538e-06,
+ "loss": 0.39504170417785645,
+ "num_input_tokens_seen": 249477272,
+ "step": 5700,
+ "train_runtime": 44127.086,
+ "train_tokens_per_second": 5653.609
+ },
+ {
+ "epoch": 0.3841233503905198,
+ "grad_norm": 0.7651587963325609,
+ "learning_rate": 9.117413387891246e-06,
+ "loss": 0.4095571041107178,
+ "num_input_tokens_seen": 249693520,
+ "step": 5705,
+ "train_runtime": 44170.6884,
+ "train_tokens_per_second": 5652.923
+ },
+ {
+ "epoch": 0.38446000538647995,
+ "grad_norm": 0.7879046146044965,
+ "learning_rate": 9.11591271494243e-06,
+ "loss": 0.39027183055877684,
+ "num_input_tokens_seen": 249909984,
+ "step": 5710,
+ "train_runtime": 44209.3015,
+ "train_tokens_per_second": 5652.882
+ },
+ {
+ "epoch": 0.3847966603824401,
+ "grad_norm": 0.7861690179999627,
+ "learning_rate": 9.114410890991748e-06,
+ "loss": 0.40212602615356446,
+ "num_input_tokens_seen": 250124664,
+ "step": 5715,
+ "train_runtime": 44250.9806,
+ "train_tokens_per_second": 5652.41
+ },
+ {
+ "epoch": 0.38513331537840023,
+ "grad_norm": 0.7798159280525913,
+ "learning_rate": 9.112907916459177e-06,
+ "loss": 0.37251665592193606,
+ "num_input_tokens_seen": 250317264,
+ "step": 5720,
+ "train_runtime": 44293.6615,
+ "train_tokens_per_second": 5651.311
+ },
+ {
+ "epoch": 0.38546997037436037,
+ "grad_norm": 0.7886435715633883,
+ "learning_rate": 9.111403791765025e-06,
+ "loss": 0.38798794746398924,
+ "num_input_tokens_seen": 250544576,
+ "step": 5725,
+ "train_runtime": 44329.9756,
+ "train_tokens_per_second": 5651.809
+ },
+ {
+ "epoch": 0.3858066253703205,
+ "grad_norm": 0.7210916045331931,
+ "learning_rate": 9.109898517329914e-06,
+ "loss": 0.3711621046066284,
+ "num_input_tokens_seen": 250774080,
+ "step": 5730,
+ "train_runtime": 44362.0555,
+ "train_tokens_per_second": 5652.896
+ },
+ {
+ "epoch": 0.38614328036628065,
+ "grad_norm": 0.757278528605975,
+ "learning_rate": 9.108392093574785e-06,
+ "loss": 0.3934319972991943,
+ "num_input_tokens_seen": 250992112,
+ "step": 5735,
+ "train_runtime": 44396.0233,
+ "train_tokens_per_second": 5653.482
+ },
+ {
+ "epoch": 0.3864799353622408,
+ "grad_norm": 0.7725316475525575,
+ "learning_rate": 9.10688452092091e-06,
+ "loss": 0.36261420249938964,
+ "num_input_tokens_seen": 251210912,
+ "step": 5740,
+ "train_runtime": 44426.4194,
+ "train_tokens_per_second": 5654.539
+ },
+ {
+ "epoch": 0.38681659035820093,
+ "grad_norm": 0.8668597404685281,
+ "learning_rate": 9.105375799789874e-06,
+ "loss": 0.39951817989349364,
+ "num_input_tokens_seen": 251426584,
+ "step": 5745,
+ "train_runtime": 44475.8422,
+ "train_tokens_per_second": 5653.105
+ },
+ {
+ "epoch": 0.38715324535416107,
+ "grad_norm": 0.8168129882131703,
+ "learning_rate": 9.103865930603586e-06,
+ "loss": 0.3850820064544678,
+ "num_input_tokens_seen": 251633720,
+ "step": 5750,
+ "train_runtime": 44516.691,
+ "train_tokens_per_second": 5652.57
+ },
+ {
+ "epoch": 0.3874899003501212,
+ "grad_norm": 0.7659015613854286,
+ "learning_rate": 9.102354913784279e-06,
+ "loss": 0.3853774070739746,
+ "num_input_tokens_seen": 251855296,
+ "step": 5755,
+ "train_runtime": 44559.5986,
+ "train_tokens_per_second": 5652.1
+ },
+ {
+ "epoch": 0.38782655534608135,
+ "grad_norm": 0.7966891158323898,
+ "learning_rate": 9.100842749754499e-06,
+ "loss": 0.3578346252441406,
+ "num_input_tokens_seen": 252080400,
+ "step": 5760,
+ "train_runtime": 44605.3404,
+ "train_tokens_per_second": 5651.35
+ },
+ {
+ "epoch": 0.3881632103420415,
+ "grad_norm": 1.3427031718401212,
+ "learning_rate": 9.099329438937122e-06,
+ "loss": 0.40854997634887696,
+ "num_input_tokens_seen": 252296312,
+ "step": 5765,
+ "train_runtime": 44649.458,
+ "train_tokens_per_second": 5650.602
+ },
+ {
+ "epoch": 0.3884998653380016,
+ "grad_norm": 0.6230050104908084,
+ "learning_rate": 9.09781498175534e-06,
+ "loss": 0.3587602138519287,
+ "num_input_tokens_seen": 252509656,
+ "step": 5770,
+ "train_runtime": 44685.4879,
+ "train_tokens_per_second": 5650.82
+ },
+ {
+ "epoch": 0.38883652033396177,
+ "grad_norm": 0.6750080511665485,
+ "learning_rate": 9.096299378632666e-06,
+ "loss": 0.37611684799194334,
+ "num_input_tokens_seen": 252740072,
+ "step": 5775,
+ "train_runtime": 44727.054,
+ "train_tokens_per_second": 5650.72
+ },
+ {
+ "epoch": 0.3891731753299219,
+ "grad_norm": 0.7734243231158888,
+ "learning_rate": 9.09478262999293e-06,
+ "loss": 0.4280585289001465,
+ "num_input_tokens_seen": 252973392,
+ "step": 5780,
+ "train_runtime": 44771.0866,
+ "train_tokens_per_second": 5650.374
+ },
+ {
+ "epoch": 0.38950983032588204,
+ "grad_norm": 0.8173587277273618,
+ "learning_rate": 9.09326473626029e-06,
+ "loss": 0.38141813278198244,
+ "num_input_tokens_seen": 253184320,
+ "step": 5785,
+ "train_runtime": 44809.6538,
+ "train_tokens_per_second": 5650.218
+ },
+ {
+ "epoch": 0.3898464853218422,
+ "grad_norm": 0.7487752402423106,
+ "learning_rate": 9.091745697859218e-06,
+ "loss": 0.41326370239257815,
+ "num_input_tokens_seen": 253411776,
+ "step": 5790,
+ "train_runtime": 44845.5045,
+ "train_tokens_per_second": 5650.773
+ },
+ {
+ "epoch": 0.3901831403178023,
+ "grad_norm": 0.8197406274404395,
+ "learning_rate": 9.090225515214512e-06,
+ "loss": 0.4117950439453125,
+ "num_input_tokens_seen": 253627512,
+ "step": 5795,
+ "train_runtime": 44885.8047,
+ "train_tokens_per_second": 5650.506
+ },
+ {
+ "epoch": 0.39051979531376246,
+ "grad_norm": 0.8857575262398083,
+ "learning_rate": 9.088704188751281e-06,
+ "loss": 0.4129201889038086,
+ "num_input_tokens_seen": 253840400,
+ "step": 5800,
+ "train_runtime": 44924.1283,
+ "train_tokens_per_second": 5650.425
+ },
+ {
+ "epoch": 0.3908564503097226,
+ "grad_norm": 0.85730454063038,
+ "learning_rate": 9.087181718894963e-06,
+ "loss": 0.39529869556427,
+ "num_input_tokens_seen": 254064096,
+ "step": 5805,
+ "train_runtime": 44968.4154,
+ "train_tokens_per_second": 5649.834
+ },
+ {
+ "epoch": 0.39119310530568274,
+ "grad_norm": 0.7118238052198951,
+ "learning_rate": 9.085658106071309e-06,
+ "loss": 0.39313058853149413,
+ "num_input_tokens_seen": 254281632,
+ "step": 5810,
+ "train_runtime": 45005.0615,
+ "train_tokens_per_second": 5650.067
+ },
+ {
+ "epoch": 0.3915297603016429,
+ "grad_norm": 0.7349746949660144,
+ "learning_rate": 9.084133350706394e-06,
+ "loss": 0.3712653398513794,
+ "num_input_tokens_seen": 254493944,
+ "step": 5815,
+ "train_runtime": 45042.1578,
+ "train_tokens_per_second": 5650.128
+ },
+ {
+ "epoch": 0.391866415297603,
+ "grad_norm": 0.7166943198737387,
+ "learning_rate": 9.082607453226613e-06,
+ "loss": 0.38478312492370603,
+ "num_input_tokens_seen": 254715712,
+ "step": 5820,
+ "train_runtime": 45077.3472,
+ "train_tokens_per_second": 5650.637
+ },
+ {
+ "epoch": 0.39220307029356316,
+ "grad_norm": 0.7543487043231831,
+ "learning_rate": 9.081080414058674e-06,
+ "loss": 0.3994553565979004,
+ "num_input_tokens_seen": 254927592,
+ "step": 5825,
+ "train_runtime": 45113.9999,
+ "train_tokens_per_second": 5650.742
+ },
+ {
+ "epoch": 0.3925397252895233,
+ "grad_norm": 0.7896679447300116,
+ "learning_rate": 9.079552233629612e-06,
+ "loss": 0.3710631847381592,
+ "num_input_tokens_seen": 255138464,
+ "step": 5830,
+ "train_runtime": 45152.0089,
+ "train_tokens_per_second": 5650.656
+ },
+ {
+ "epoch": 0.39287638028548344,
+ "grad_norm": 0.815361722369787,
+ "learning_rate": 9.078022912366776e-06,
+ "loss": 0.3914163827896118,
+ "num_input_tokens_seen": 255367048,
+ "step": 5835,
+ "train_runtime": 45192.1775,
+ "train_tokens_per_second": 5650.691
+ },
+ {
+ "epoch": 0.3932130352814436,
+ "grad_norm": 0.8005773109782048,
+ "learning_rate": 9.076492450697835e-06,
+ "loss": 0.40029544830322267,
+ "num_input_tokens_seen": 255598184,
+ "step": 5840,
+ "train_runtime": 45232.9765,
+ "train_tokens_per_second": 5650.704
+ },
+ {
+ "epoch": 0.3935496902774037,
+ "grad_norm": 0.849726633762188,
+ "learning_rate": 9.074960849050782e-06,
+ "loss": 0.3852508068084717,
+ "num_input_tokens_seen": 255819944,
+ "step": 5845,
+ "train_runtime": 45272.8634,
+ "train_tokens_per_second": 5650.624
+ },
+ {
+ "epoch": 0.39388634527336386,
+ "grad_norm": 0.7762826196998844,
+ "learning_rate": 9.07342810785392e-06,
+ "loss": 0.37957100868225097,
+ "num_input_tokens_seen": 256019616,
+ "step": 5850,
+ "train_runtime": 45311.4802,
+ "train_tokens_per_second": 5650.215
+ },
+ {
+ "epoch": 0.394223000269324,
+ "grad_norm": 0.7731972533551551,
+ "learning_rate": 9.071894227535878e-06,
+ "loss": 0.38977823257446287,
+ "num_input_tokens_seen": 256237256,
+ "step": 5855,
+ "train_runtime": 45352.5452,
+ "train_tokens_per_second": 5649.898
+ },
+ {
+ "epoch": 0.39455965526528414,
+ "grad_norm": 0.8423525268519388,
+ "learning_rate": 9.070359208525598e-06,
+ "loss": 0.406312084197998,
+ "num_input_tokens_seen": 256464368,
+ "step": 5860,
+ "train_runtime": 45388.423,
+ "train_tokens_per_second": 5650.436
+ },
+ {
+ "epoch": 0.3948963102612443,
+ "grad_norm": 0.8527879610056384,
+ "learning_rate": 9.068823051252346e-06,
+ "loss": 0.3981436491012573,
+ "num_input_tokens_seen": 256688032,
+ "step": 5865,
+ "train_runtime": 45429.2415,
+ "train_tokens_per_second": 5650.282
+ },
+ {
+ "epoch": 0.3952329652572044,
+ "grad_norm": 0.8385538303650153,
+ "learning_rate": 9.067285756145702e-06,
+ "loss": 0.3921481132507324,
+ "num_input_tokens_seen": 256917840,
+ "step": 5870,
+ "train_runtime": 45462.4848,
+ "train_tokens_per_second": 5651.205
+ },
+ {
+ "epoch": 0.39556962025316456,
+ "grad_norm": 0.8390456887457934,
+ "learning_rate": 9.065747323635565e-06,
+ "loss": 0.3327157020568848,
+ "num_input_tokens_seen": 257101736,
+ "step": 5875,
+ "train_runtime": 45501.4164,
+ "train_tokens_per_second": 5650.412
+ },
+ {
+ "epoch": 0.3959062752491247,
+ "grad_norm": 0.7851013810213687,
+ "learning_rate": 9.064207754152154e-06,
+ "loss": 0.3997276782989502,
+ "num_input_tokens_seen": 257317696,
+ "step": 5880,
+ "train_runtime": 45545.2236,
+ "train_tokens_per_second": 5649.719
+ },
+ {
+ "epoch": 0.39624293024508483,
+ "grad_norm": 0.7365128564544189,
+ "learning_rate": 9.062667048126005e-06,
+ "loss": 0.3864887714385986,
+ "num_input_tokens_seen": 257547392,
+ "step": 5885,
+ "train_runtime": 45589.9837,
+ "train_tokens_per_second": 5649.21
+ },
+ {
+ "epoch": 0.396579585241045,
+ "grad_norm": 0.8935927015959244,
+ "learning_rate": 9.061125205987971e-06,
+ "loss": 0.4157553672790527,
+ "num_input_tokens_seen": 257761408,
+ "step": 5890,
+ "train_runtime": 45630.8437,
+ "train_tokens_per_second": 5648.842
+ },
+ {
+ "epoch": 0.3969162402370051,
+ "grad_norm": 0.7876564044642139,
+ "learning_rate": 9.059582228169222e-06,
+ "loss": 0.37337889671325686,
+ "num_input_tokens_seen": 257963800,
+ "step": 5895,
+ "train_runtime": 45664.7,
+ "train_tokens_per_second": 5649.086
+ },
+ {
+ "epoch": 0.39725289523296525,
+ "grad_norm": 0.7288959690931758,
+ "learning_rate": 9.058038115101248e-06,
+ "loss": 0.4037358283996582,
+ "num_input_tokens_seen": 258195648,
+ "step": 5900,
+ "train_runtime": 45702.5616,
+ "train_tokens_per_second": 5649.479
+ },
+ {
+ "epoch": 0.3975895502289254,
+ "grad_norm": 0.7904683688779008,
+ "learning_rate": 9.056492867215857e-06,
+ "loss": 0.3476010084152222,
+ "num_input_tokens_seen": 258400784,
+ "step": 5905,
+ "train_runtime": 45748.63,
+ "train_tokens_per_second": 5648.274
+ },
+ {
+ "epoch": 0.39792620522488553,
+ "grad_norm": 0.6759275337501953,
+ "learning_rate": 9.05494648494517e-06,
+ "loss": 0.3933530330657959,
+ "num_input_tokens_seen": 258606968,
+ "step": 5910,
+ "train_runtime": 45786.7698,
+ "train_tokens_per_second": 5648.072
+ },
+ {
+ "epoch": 0.39826286022084567,
+ "grad_norm": 0.6825204404349337,
+ "learning_rate": 9.053398968721629e-06,
+ "loss": 0.37357821464538576,
+ "num_input_tokens_seen": 258832840,
+ "step": 5915,
+ "train_runtime": 45827.0425,
+ "train_tokens_per_second": 5648.037
+ },
+ {
+ "epoch": 0.3985995152168058,
+ "grad_norm": 0.7549266235464368,
+ "learning_rate": 9.05185031897799e-06,
+ "loss": 0.4264666080474854,
+ "num_input_tokens_seen": 259049920,
+ "step": 5920,
+ "train_runtime": 45860.4564,
+ "train_tokens_per_second": 5648.656
+ },
+ {
+ "epoch": 0.39893617021276595,
+ "grad_norm": 0.8046151442540065,
+ "learning_rate": 9.050300536147331e-06,
+ "loss": 0.42224712371826173,
+ "num_input_tokens_seen": 259285872,
+ "step": 5925,
+ "train_runtime": 45898.8286,
+ "train_tokens_per_second": 5649.074
+ },
+ {
+ "epoch": 0.3992728252087261,
+ "grad_norm": 0.8013279899857307,
+ "learning_rate": 9.048749620663044e-06,
+ "loss": 0.36397323608398435,
+ "num_input_tokens_seen": 259520536,
+ "step": 5930,
+ "train_runtime": 45935.2903,
+ "train_tokens_per_second": 5649.698
+ },
+ {
+ "epoch": 0.39960948020468623,
+ "grad_norm": 0.7188029194424288,
+ "learning_rate": 9.047197572958834e-06,
+ "loss": 0.37349615097045896,
+ "num_input_tokens_seen": 259742248,
+ "step": 5935,
+ "train_runtime": 45973.4138,
+ "train_tokens_per_second": 5649.836
+ },
+ {
+ "epoch": 0.39994613520064637,
+ "grad_norm": 0.8120677037958034,
+ "learning_rate": 9.04564439346873e-06,
+ "loss": 0.3917673110961914,
+ "num_input_tokens_seen": 259952080,
+ "step": 5940,
+ "train_runtime": 46008.964,
+ "train_tokens_per_second": 5650.031
+ },
+ {
+ "epoch": 0.4002827901966065,
+ "grad_norm": 0.8737926960149425,
+ "learning_rate": 9.044090082627071e-06,
+ "loss": 0.39895708560943605,
+ "num_input_tokens_seen": 260160320,
+ "step": 5945,
+ "train_runtime": 46045.9436,
+ "train_tokens_per_second": 5650.016
+ },
+ {
+ "epoch": 0.40061944519256665,
+ "grad_norm": 0.7302283729471113,
+ "learning_rate": 9.042534640868514e-06,
+ "loss": 0.4021873950958252,
+ "num_input_tokens_seen": 260368008,
+ "step": 5950,
+ "train_runtime": 46081.9071,
+ "train_tokens_per_second": 5650.114
+ },
+ {
+ "epoch": 0.4009561001885268,
+ "grad_norm": 0.749168823407156,
+ "learning_rate": 9.04097806862804e-06,
+ "loss": 0.40214033126831056,
+ "num_input_tokens_seen": 260592824,
+ "step": 5955,
+ "train_runtime": 46117.1368,
+ "train_tokens_per_second": 5650.672
+ },
+ {
+ "epoch": 0.4012927551844869,
+ "grad_norm": 0.723288812952783,
+ "learning_rate": 9.03942036634093e-06,
+ "loss": 0.4055626392364502,
+ "num_input_tokens_seen": 260810288,
+ "step": 5960,
+ "train_runtime": 46156.4921,
+ "train_tokens_per_second": 5650.566
+ },
+ {
+ "epoch": 0.40162941018044707,
+ "grad_norm": 0.7518699289987396,
+ "learning_rate": 9.037861534442797e-06,
+ "loss": 0.3734895706176758,
+ "num_input_tokens_seen": 261038312,
+ "step": 5965,
+ "train_runtime": 46198.2126,
+ "train_tokens_per_second": 5650.399
+ },
+ {
+ "epoch": 0.4019660651764072,
+ "grad_norm": 1.0026827049870985,
+ "learning_rate": 9.036301573369563e-06,
+ "loss": 0.40189208984375,
+ "num_input_tokens_seen": 261266152,
+ "step": 5970,
+ "train_runtime": 46239.4807,
+ "train_tokens_per_second": 5650.283
+ },
+ {
+ "epoch": 0.40230272017236735,
+ "grad_norm": 0.8605436747902855,
+ "learning_rate": 9.034740483557465e-06,
+ "loss": 0.3900024175643921,
+ "num_input_tokens_seen": 261483152,
+ "step": 5975,
+ "train_runtime": 46274.8434,
+ "train_tokens_per_second": 5650.655
+ },
+ {
+ "epoch": 0.4026393751683275,
+ "grad_norm": 0.7213971067668634,
+ "learning_rate": 9.033178265443055e-06,
+ "loss": 0.38559999465942385,
+ "num_input_tokens_seen": 261703416,
+ "step": 5980,
+ "train_runtime": 46306.9931,
+ "train_tokens_per_second": 5651.488
+ },
+ {
+ "epoch": 0.4029760301642876,
+ "grad_norm": 0.7467717729146578,
+ "learning_rate": 9.031614919463206e-06,
+ "loss": 0.3772167682647705,
+ "num_input_tokens_seen": 261926896,
+ "step": 5985,
+ "train_runtime": 46342.877,
+ "train_tokens_per_second": 5651.934
+ },
+ {
+ "epoch": 0.40331268516024776,
+ "grad_norm": 0.8052956550974576,
+ "learning_rate": 9.030050446055099e-06,
+ "loss": 0.3684276342391968,
+ "num_input_tokens_seen": 262143080,
+ "step": 5990,
+ "train_runtime": 46380.2479,
+ "train_tokens_per_second": 5652.041
+ },
+ {
+ "epoch": 0.4036493401562079,
+ "grad_norm": 0.6725874955047775,
+ "learning_rate": 9.028484845656235e-06,
+ "loss": 0.3965585231781006,
+ "num_input_tokens_seen": 262378952,
+ "step": 5995,
+ "train_runtime": 46413.7998,
+ "train_tokens_per_second": 5653.038
+ },
+ {
+ "epoch": 0.40398599515216804,
+ "grad_norm": 0.8065835959924061,
+ "learning_rate": 9.026918118704431e-06,
+ "loss": 0.42124166488647463,
+ "num_input_tokens_seen": 262599696,
+ "step": 6000,
+ "train_runtime": 46454.2053,
+ "train_tokens_per_second": 5652.872
+ },
+ {
+ "epoch": 0.4043226501481282,
+ "grad_norm": 0.6374762319564959,
+ "learning_rate": 9.025350265637816e-06,
+ "loss": 0.3700655221939087,
+ "num_input_tokens_seen": 262830992,
+ "step": 6005,
+ "train_runtime": 46490.7965,
+ "train_tokens_per_second": 5653.398
+ },
+ {
+ "epoch": 0.4046593051440883,
+ "grad_norm": 0.7573194160968504,
+ "learning_rate": 9.023781286894834e-06,
+ "loss": 0.38486456871032715,
+ "num_input_tokens_seen": 263044288,
+ "step": 6010,
+ "train_runtime": 46526.3581,
+ "train_tokens_per_second": 5653.662
+ },
+ {
+ "epoch": 0.40499596014004846,
+ "grad_norm": 0.8562717514351039,
+ "learning_rate": 9.022211182914247e-06,
+ "loss": 0.3843253135681152,
+ "num_input_tokens_seen": 263264800,
+ "step": 6015,
+ "train_runtime": 46564.9415,
+ "train_tokens_per_second": 5653.713
+ },
+ {
+ "epoch": 0.4053326151360086,
+ "grad_norm": 0.7322028797978756,
+ "learning_rate": 9.020639954135128e-06,
+ "loss": 0.3804231405258179,
+ "num_input_tokens_seen": 263476880,
+ "step": 6020,
+ "train_runtime": 46608.1754,
+ "train_tokens_per_second": 5653.019
+ },
+ {
+ "epoch": 0.40566927013196874,
+ "grad_norm": 0.7924756161102612,
+ "learning_rate": 9.019067600996867e-06,
+ "loss": 0.32609758377075193,
+ "num_input_tokens_seen": 263676440,
+ "step": 6025,
+ "train_runtime": 46647.9449,
+ "train_tokens_per_second": 5652.477
+ },
+ {
+ "epoch": 0.4060059251279289,
+ "grad_norm": 0.7406207884260587,
+ "learning_rate": 9.017494123939169e-06,
+ "loss": 0.3878431558609009,
+ "num_input_tokens_seen": 263891832,
+ "step": 6030,
+ "train_runtime": 46686.4579,
+ "train_tokens_per_second": 5652.428
+ },
+ {
+ "epoch": 0.406342580123889,
+ "grad_norm": 0.6859816815928503,
+ "learning_rate": 9.015919523402048e-06,
+ "loss": 0.38746066093444825,
+ "num_input_tokens_seen": 264108024,
+ "step": 6035,
+ "train_runtime": 46723.9959,
+ "train_tokens_per_second": 5652.514
+ },
+ {
+ "epoch": 0.40667923511984916,
+ "grad_norm": 1.0398516753802722,
+ "learning_rate": 9.014343799825838e-06,
+ "loss": 0.3903219699859619,
+ "num_input_tokens_seen": 264311560,
+ "step": 6040,
+ "train_runtime": 46758.4333,
+ "train_tokens_per_second": 5652.704
+ },
+ {
+ "epoch": 0.4070158901158093,
+ "grad_norm": 0.6954397754211817,
+ "learning_rate": 9.012766953651186e-06,
+ "loss": 0.42699203491210935,
+ "num_input_tokens_seen": 264536928,
+ "step": 6045,
+ "train_runtime": 46800.6106,
+ "train_tokens_per_second": 5652.425
+ },
+ {
+ "epoch": 0.40735254511176944,
+ "grad_norm": 0.6848650266078808,
+ "learning_rate": 9.01118898531905e-06,
+ "loss": 0.4040683746337891,
+ "num_input_tokens_seen": 264744560,
+ "step": 6050,
+ "train_runtime": 46841.7035,
+ "train_tokens_per_second": 5651.899
+ },
+ {
+ "epoch": 0.4076892001077296,
+ "grad_norm": 0.7157357093690987,
+ "learning_rate": 9.009609895270708e-06,
+ "loss": 0.3655755043029785,
+ "num_input_tokens_seen": 264973096,
+ "step": 6055,
+ "train_runtime": 46881.3659,
+ "train_tokens_per_second": 5651.992
+ },
+ {
+ "epoch": 0.4080258551036897,
+ "grad_norm": 0.8098322622781278,
+ "learning_rate": 9.008029683947743e-06,
+ "loss": 0.42575840950012206,
+ "num_input_tokens_seen": 265197544,
+ "step": 6060,
+ "train_runtime": 46917.7458,
+ "train_tokens_per_second": 5652.393
+ },
+ {
+ "epoch": 0.40836251009964986,
+ "grad_norm": 0.7104130729995994,
+ "learning_rate": 9.006448351792057e-06,
+ "loss": 0.3974879741668701,
+ "num_input_tokens_seen": 265418944,
+ "step": 6065,
+ "train_runtime": 46962.89,
+ "train_tokens_per_second": 5651.674
+ },
+ {
+ "epoch": 0.40869916509561,
+ "grad_norm": 0.6632525596263023,
+ "learning_rate": 9.004865899245864e-06,
+ "loss": 0.38395133018493655,
+ "num_input_tokens_seen": 265636776,
+ "step": 6070,
+ "train_runtime": 46999.967,
+ "train_tokens_per_second": 5651.85
+ },
+ {
+ "epoch": 0.40903582009157013,
+ "grad_norm": 0.7043912136666165,
+ "learning_rate": 9.003282326751694e-06,
+ "loss": 0.4008152961730957,
+ "num_input_tokens_seen": 265857800,
+ "step": 6075,
+ "train_runtime": 47039.3368,
+ "train_tokens_per_second": 5651.819
+ },
+ {
+ "epoch": 0.4093724750875303,
+ "grad_norm": 0.9655541200247993,
+ "learning_rate": 9.001697634752387e-06,
+ "loss": 0.40933728218078613,
+ "num_input_tokens_seen": 266070448,
+ "step": 6080,
+ "train_runtime": 47074.6096,
+ "train_tokens_per_second": 5652.101
+ },
+ {
+ "epoch": 0.4097091300834904,
+ "grad_norm": 0.683185307022826,
+ "learning_rate": 9.000111823691097e-06,
+ "loss": 0.37639498710632324,
+ "num_input_tokens_seen": 266291816,
+ "step": 6085,
+ "train_runtime": 47112.3669,
+ "train_tokens_per_second": 5652.27
+ },
+ {
+ "epoch": 0.41004578507945055,
+ "grad_norm": 0.8219233379550768,
+ "learning_rate": 8.99852489401129e-06,
+ "loss": 0.40433249473571775,
+ "num_input_tokens_seen": 266517904,
+ "step": 6090,
+ "train_runtime": 47148.2501,
+ "train_tokens_per_second": 5652.763
+ },
+ {
+ "epoch": 0.4103824400754107,
+ "grad_norm": 0.7423675573693415,
+ "learning_rate": 8.996936846156748e-06,
+ "loss": 0.3635870456695557,
+ "num_input_tokens_seen": 266736992,
+ "step": 6095,
+ "train_runtime": 47183.3361,
+ "train_tokens_per_second": 5653.203
+ },
+ {
+ "epoch": 0.41071909507137083,
+ "grad_norm": 0.8008363109798308,
+ "learning_rate": 8.995347680571563e-06,
+ "loss": 0.41033496856689455,
+ "num_input_tokens_seen": 266959272,
+ "step": 6100,
+ "train_runtime": 47217.625,
+ "train_tokens_per_second": 5653.806
+ },
+ {
+ "epoch": 0.41105575006733097,
+ "grad_norm": 0.6633014625775752,
+ "learning_rate": 8.993757397700137e-06,
+ "loss": 0.38593282699584963,
+ "num_input_tokens_seen": 267178696,
+ "step": 6105,
+ "train_runtime": 47254.1693,
+ "train_tokens_per_second": 5654.077
+ },
+ {
+ "epoch": 0.41139240506329117,
+ "grad_norm": 0.7915805562051228,
+ "learning_rate": 8.99216599798719e-06,
+ "loss": 0.42214293479919435,
+ "num_input_tokens_seen": 267415152,
+ "step": 6110,
+ "train_runtime": 47290.6934,
+ "train_tokens_per_second": 5654.71
+ },
+ {
+ "epoch": 0.4117290600592513,
+ "grad_norm": 0.7917765167657784,
+ "learning_rate": 8.990573481877753e-06,
+ "loss": 0.3639373779296875,
+ "num_input_tokens_seen": 267618552,
+ "step": 6115,
+ "train_runtime": 47325.6204,
+ "train_tokens_per_second": 5654.835
+ },
+ {
+ "epoch": 0.41206571505521145,
+ "grad_norm": 0.694124505400035,
+ "learning_rate": 8.988979849817167e-06,
+ "loss": 0.36659536361694334,
+ "num_input_tokens_seen": 267814904,
+ "step": 6120,
+ "train_runtime": 47366.4119,
+ "train_tokens_per_second": 5654.11
+ },
+ {
+ "epoch": 0.4124023700511716,
+ "grad_norm": 0.8642974554518021,
+ "learning_rate": 8.987385102251085e-06,
+ "loss": 0.3810924291610718,
+ "num_input_tokens_seen": 268015432,
+ "step": 6125,
+ "train_runtime": 47405.0635,
+ "train_tokens_per_second": 5653.73
+ },
+ {
+ "epoch": 0.4127390250471317,
+ "grad_norm": 0.7416982400357927,
+ "learning_rate": 8.985789239625475e-06,
+ "loss": 0.3897581100463867,
+ "num_input_tokens_seen": 268235696,
+ "step": 6130,
+ "train_runtime": 47449.5896,
+ "train_tokens_per_second": 5653.067
+ },
+ {
+ "epoch": 0.41307568004309186,
+ "grad_norm": 0.7586989342588731,
+ "learning_rate": 8.984192262386613e-06,
+ "loss": 0.36377105712890623,
+ "num_input_tokens_seen": 268454176,
+ "step": 6135,
+ "train_runtime": 47483.2101,
+ "train_tokens_per_second": 5653.665
+ },
+ {
+ "epoch": 0.413412335039052,
+ "grad_norm": 0.7553176125857815,
+ "learning_rate": 8.98259417098109e-06,
+ "loss": 0.3994285583496094,
+ "num_input_tokens_seen": 268685424,
+ "step": 6140,
+ "train_runtime": 47523.5954,
+ "train_tokens_per_second": 5653.727
+ },
+ {
+ "epoch": 0.41374899003501214,
+ "grad_norm": 0.7619380388493862,
+ "learning_rate": 8.980994965855808e-06,
+ "loss": 0.39436161518096924,
+ "num_input_tokens_seen": 268906160,
+ "step": 6145,
+ "train_runtime": 47557.2625,
+ "train_tokens_per_second": 5654.366
+ },
+ {
+ "epoch": 0.4140856450309723,
+ "grad_norm": 0.7194190068295008,
+ "learning_rate": 8.979394647457976e-06,
+ "loss": 0.3927819490432739,
+ "num_input_tokens_seen": 269128016,
+ "step": 6150,
+ "train_runtime": 47600.5297,
+ "train_tokens_per_second": 5653.887
+ },
+ {
+ "epoch": 0.4144223000269324,
+ "grad_norm": 0.741178107175373,
+ "learning_rate": 8.97779321623512e-06,
+ "loss": 0.35940072536468504,
+ "num_input_tokens_seen": 269333048,
+ "step": 6155,
+ "train_runtime": 47638.7497,
+ "train_tokens_per_second": 5653.655
+ },
+ {
+ "epoch": 0.41475895502289256,
+ "grad_norm": 0.7334172734354966,
+ "learning_rate": 8.976190672635077e-06,
+ "loss": 0.400145435333252,
+ "num_input_tokens_seen": 269557248,
+ "step": 6160,
+ "train_runtime": 47679.304,
+ "train_tokens_per_second": 5653.548
+ },
+ {
+ "epoch": 0.4150956100188527,
+ "grad_norm": 0.7612088304227586,
+ "learning_rate": 8.974587017105991e-06,
+ "loss": 0.41222319602966306,
+ "num_input_tokens_seen": 269768408,
+ "step": 6165,
+ "train_runtime": 47713.1048,
+ "train_tokens_per_second": 5653.969
+ },
+ {
+ "epoch": 0.41543226501481284,
+ "grad_norm": 0.7344591815245108,
+ "learning_rate": 8.97298225009632e-06,
+ "loss": 0.3931986570358276,
+ "num_input_tokens_seen": 269990368,
+ "step": 6170,
+ "train_runtime": 47754.3241,
+ "train_tokens_per_second": 5653.737
+ },
+ {
+ "epoch": 0.415768920010773,
+ "grad_norm": 0.7575071698285096,
+ "learning_rate": 8.971376372054829e-06,
+ "loss": 0.3943799495697021,
+ "num_input_tokens_seen": 270208056,
+ "step": 6175,
+ "train_runtime": 47794.7334,
+ "train_tokens_per_second": 5653.511
+ },
+ {
+ "epoch": 0.4161055750067331,
+ "grad_norm": 0.6823047182780082,
+ "learning_rate": 8.969769383430602e-06,
+ "loss": 0.3858570337295532,
+ "num_input_tokens_seen": 270432152,
+ "step": 6180,
+ "train_runtime": 47829.7321,
+ "train_tokens_per_second": 5654.06
+ },
+ {
+ "epoch": 0.41644223000269326,
+ "grad_norm": 0.7476934311068052,
+ "learning_rate": 8.968161284673027e-06,
+ "loss": 0.37152729034423826,
+ "num_input_tokens_seen": 270661216,
+ "step": 6185,
+ "train_runtime": 47868.2088,
+ "train_tokens_per_second": 5654.3
+ },
+ {
+ "epoch": 0.4167788849986534,
+ "grad_norm": 0.7528325409155248,
+ "learning_rate": 8.9665520762318e-06,
+ "loss": 0.37863969802856445,
+ "num_input_tokens_seen": 270875256,
+ "step": 6190,
+ "train_runtime": 47904.4792,
+ "train_tokens_per_second": 5654.487
+ },
+ {
+ "epoch": 0.41711553999461354,
+ "grad_norm": 0.7282239751820792,
+ "learning_rate": 8.964941758556934e-06,
+ "loss": 0.3976884126663208,
+ "num_input_tokens_seen": 271088288,
+ "step": 6195,
+ "train_runtime": 47948.7488,
+ "train_tokens_per_second": 5653.709
+ },
+ {
+ "epoch": 0.4174521949905737,
+ "grad_norm": 0.755222562109598,
+ "learning_rate": 8.963330332098747e-06,
+ "loss": 0.3814378261566162,
+ "num_input_tokens_seen": 271296528,
+ "step": 6200,
+ "train_runtime": 47985.6256,
+ "train_tokens_per_second": 5653.704
+ },
+ {
+ "epoch": 0.4177888499865338,
+ "grad_norm": 0.7830273045211564,
+ "learning_rate": 8.961717797307872e-06,
+ "loss": 0.37258970737457275,
+ "num_input_tokens_seen": 271515568,
+ "step": 6205,
+ "train_runtime": 48019.4934,
+ "train_tokens_per_second": 5654.278
+ },
+ {
+ "epoch": 0.41812550498249396,
+ "grad_norm": 0.7683395292659414,
+ "learning_rate": 8.960104154635248e-06,
+ "loss": 0.4039943695068359,
+ "num_input_tokens_seen": 271733704,
+ "step": 6210,
+ "train_runtime": 48057.6768,
+ "train_tokens_per_second": 5654.325
+ },
+ {
+ "epoch": 0.4184621599784541,
+ "grad_norm": 0.7964266950242961,
+ "learning_rate": 8.958489404532125e-06,
+ "loss": 0.4006314277648926,
+ "num_input_tokens_seen": 271941040,
+ "step": 6215,
+ "train_runtime": 48094.9785,
+ "train_tokens_per_second": 5654.25
+ },
+ {
+ "epoch": 0.41879881497441424,
+ "grad_norm": 0.7229961706218653,
+ "learning_rate": 8.956873547450062e-06,
+ "loss": 0.3996126651763916,
+ "num_input_tokens_seen": 272155560,
+ "step": 6220,
+ "train_runtime": 48135.1561,
+ "train_tokens_per_second": 5653.987
+ },
+ {
+ "epoch": 0.4191354699703744,
+ "grad_norm": 0.8692942989729212,
+ "learning_rate": 8.95525658384093e-06,
+ "loss": 0.4290035247802734,
+ "num_input_tokens_seen": 272369192,
+ "step": 6225,
+ "train_runtime": 48171.8576,
+ "train_tokens_per_second": 5654.114
+ },
+ {
+ "epoch": 0.4194721249663345,
+ "grad_norm": 0.7600990815023999,
+ "learning_rate": 8.953638514156908e-06,
+ "loss": 0.3810266494750977,
+ "num_input_tokens_seen": 272591248,
+ "step": 6230,
+ "train_runtime": 48208.9238,
+ "train_tokens_per_second": 5654.373
+ },
+ {
+ "epoch": 0.41980877996229465,
+ "grad_norm": 0.7355727181513495,
+ "learning_rate": 8.952019338850481e-06,
+ "loss": 0.3950758695602417,
+ "num_input_tokens_seen": 272810616,
+ "step": 6235,
+ "train_runtime": 48248.4869,
+ "train_tokens_per_second": 5654.283
+ },
+ {
+ "epoch": 0.4201454349582548,
+ "grad_norm": 0.7644200111226909,
+ "learning_rate": 8.950399058374447e-06,
+ "loss": 0.377358603477478,
+ "num_input_tokens_seen": 273036976,
+ "step": 6240,
+ "train_runtime": 48285.7901,
+ "train_tokens_per_second": 5654.603
+ },
+ {
+ "epoch": 0.42048208995421493,
+ "grad_norm": 0.7918087336685585,
+ "learning_rate": 8.948777673181913e-06,
+ "loss": 0.389499831199646,
+ "num_input_tokens_seen": 273260832,
+ "step": 6245,
+ "train_runtime": 48322.4015,
+ "train_tokens_per_second": 5654.951
+ },
+ {
+ "epoch": 0.42081874495017507,
+ "grad_norm": 0.7765030452628453,
+ "learning_rate": 8.947155183726296e-06,
+ "loss": 0.35937402248382566,
+ "num_input_tokens_seen": 273488360,
+ "step": 6250,
+ "train_runtime": 48361.4271,
+ "train_tokens_per_second": 5655.093
+ },
+ {
+ "epoch": 0.4211553999461352,
+ "grad_norm": 0.7014097499994082,
+ "learning_rate": 8.94553159046132e-06,
+ "loss": 0.3715068817138672,
+ "num_input_tokens_seen": 273707472,
+ "step": 6255,
+ "train_runtime": 48401.1614,
+ "train_tokens_per_second": 5654.977
+ },
+ {
+ "epoch": 0.42149205494209535,
+ "grad_norm": 0.6408398788945804,
+ "learning_rate": 8.943906893841015e-06,
+ "loss": 0.36748714447021485,
+ "num_input_tokens_seen": 273928544,
+ "step": 6260,
+ "train_runtime": 48433.3616,
+ "train_tokens_per_second": 5655.782
+ },
+ {
+ "epoch": 0.4218287099380555,
+ "grad_norm": 0.754509294899034,
+ "learning_rate": 8.942281094319721e-06,
+ "loss": 0.4157464027404785,
+ "num_input_tokens_seen": 274141592,
+ "step": 6265,
+ "train_runtime": 48468.7848,
+ "train_tokens_per_second": 5656.044
+ },
+ {
+ "epoch": 0.42216536493401563,
+ "grad_norm": 0.7840590763768849,
+ "learning_rate": 8.940654192352092e-06,
+ "loss": 0.40111403465270995,
+ "num_input_tokens_seen": 274349792,
+ "step": 6270,
+ "train_runtime": 48510.3168,
+ "train_tokens_per_second": 5655.494
+ },
+ {
+ "epoch": 0.42250201992997577,
+ "grad_norm": 0.8213882192132025,
+ "learning_rate": 8.939026188393083e-06,
+ "loss": 0.3853888034820557,
+ "num_input_tokens_seen": 274569104,
+ "step": 6275,
+ "train_runtime": 48556.531,
+ "train_tokens_per_second": 5654.628
+ },
+ {
+ "epoch": 0.4228386749259359,
+ "grad_norm": 0.8791839426260795,
+ "learning_rate": 8.937397082897963e-06,
+ "loss": 0.40306596755981444,
+ "num_input_tokens_seen": 274796616,
+ "step": 6280,
+ "train_runtime": 48596.7026,
+ "train_tokens_per_second": 5654.635
+ },
+ {
+ "epoch": 0.42317532992189605,
+ "grad_norm": 0.7843983011163462,
+ "learning_rate": 8.935766876322303e-06,
+ "loss": 0.4140591621398926,
+ "num_input_tokens_seen": 275015624,
+ "step": 6285,
+ "train_runtime": 48635.2618,
+ "train_tokens_per_second": 5654.655
+ },
+ {
+ "epoch": 0.4235119849178562,
+ "grad_norm": 0.7007974546940894,
+ "learning_rate": 8.934135569121985e-06,
+ "loss": 0.3959988832473755,
+ "num_input_tokens_seen": 275242840,
+ "step": 6290,
+ "train_runtime": 48673.0052,
+ "train_tokens_per_second": 5654.938
+ },
+ {
+ "epoch": 0.4238486399138163,
+ "grad_norm": 0.7078146309363205,
+ "learning_rate": 8.932503161753202e-06,
+ "loss": 0.38456072807312014,
+ "num_input_tokens_seen": 275462504,
+ "step": 6295,
+ "train_runtime": 48709.4026,
+ "train_tokens_per_second": 5655.222
+ },
+ {
+ "epoch": 0.42418529490977647,
+ "grad_norm": 1.0351275568049705,
+ "learning_rate": 8.930869654672449e-06,
+ "loss": 0.38416409492492676,
+ "num_input_tokens_seen": 275681808,
+ "step": 6300,
+ "train_runtime": 48747.6893,
+ "train_tokens_per_second": 5655.279
+ },
+ {
+ "epoch": 0.4245219499057366,
+ "grad_norm": 0.8236989821133165,
+ "learning_rate": 8.929235048336531e-06,
+ "loss": 0.37823896408081054,
+ "num_input_tokens_seen": 275903576,
+ "step": 6305,
+ "train_runtime": 48788.5888,
+ "train_tokens_per_second": 5655.084
+ },
+ {
+ "epoch": 0.42485860490169675,
+ "grad_norm": 0.8519274837856372,
+ "learning_rate": 8.927599343202563e-06,
+ "loss": 0.4197996139526367,
+ "num_input_tokens_seen": 276132792,
+ "step": 6310,
+ "train_runtime": 48821.7451,
+ "train_tokens_per_second": 5655.939
+ },
+ {
+ "epoch": 0.4251952598976569,
+ "grad_norm": 0.7601906981584227,
+ "learning_rate": 8.92596253972796e-06,
+ "loss": 0.39647321701049804,
+ "num_input_tokens_seen": 276373576,
+ "step": 6315,
+ "train_runtime": 48856.0141,
+ "train_tokens_per_second": 5656.9
+ },
+ {
+ "epoch": 0.425531914893617,
+ "grad_norm": 0.7436129276391935,
+ "learning_rate": 8.924324638370454e-06,
+ "loss": 0.4159084320068359,
+ "num_input_tokens_seen": 276601344,
+ "step": 6320,
+ "train_runtime": 48892.2285,
+ "train_tokens_per_second": 5657.368
+ },
+ {
+ "epoch": 0.42586856988957716,
+ "grad_norm": 0.619825666948129,
+ "learning_rate": 8.922685639588076e-06,
+ "loss": 0.3394161701202393,
+ "num_input_tokens_seen": 276801216,
+ "step": 6325,
+ "train_runtime": 48929.4586,
+ "train_tokens_per_second": 5657.149
+ },
+ {
+ "epoch": 0.4262052248855373,
+ "grad_norm": 0.7566990752584487,
+ "learning_rate": 8.921045543839167e-06,
+ "loss": 0.412354850769043,
+ "num_input_tokens_seen": 277025984,
+ "step": 6330,
+ "train_runtime": 48964.4429,
+ "train_tokens_per_second": 5657.697
+ },
+ {
+ "epoch": 0.42654187988149744,
+ "grad_norm": 0.7706529692481164,
+ "learning_rate": 8.919404351582376e-06,
+ "loss": 0.37722196578979494,
+ "num_input_tokens_seen": 277234256,
+ "step": 6335,
+ "train_runtime": 49000.4277,
+ "train_tokens_per_second": 5657.793
+ },
+ {
+ "epoch": 0.4268785348774576,
+ "grad_norm": 0.7056105235159689,
+ "learning_rate": 8.917762063276657e-06,
+ "loss": 0.38721909523010256,
+ "num_input_tokens_seen": 277448936,
+ "step": 6340,
+ "train_runtime": 49039.3589,
+ "train_tokens_per_second": 5657.679
+ },
+ {
+ "epoch": 0.4272151898734177,
+ "grad_norm": 0.7187047540510013,
+ "learning_rate": 8.916118679381268e-06,
+ "loss": 0.3728447198867798,
+ "num_input_tokens_seen": 277667920,
+ "step": 6345,
+ "train_runtime": 49076.9224,
+ "train_tokens_per_second": 5657.81
+ },
+ {
+ "epoch": 0.42755184486937786,
+ "grad_norm": 0.8958354766388671,
+ "learning_rate": 8.914474200355777e-06,
+ "loss": 0.39159350395202636,
+ "num_input_tokens_seen": 277885288,
+ "step": 6350,
+ "train_runtime": 49121.5781,
+ "train_tokens_per_second": 5657.092
+ },
+ {
+ "epoch": 0.427888499865338,
+ "grad_norm": 0.8411282788940809,
+ "learning_rate": 8.912828626660059e-06,
+ "loss": 0.4010343074798584,
+ "num_input_tokens_seen": 278112192,
+ "step": 6355,
+ "train_runtime": 49161.2354,
+ "train_tokens_per_second": 5657.144
+ },
+ {
+ "epoch": 0.42822515486129814,
+ "grad_norm": 0.7895834565221768,
+ "learning_rate": 8.911181958754294e-06,
+ "loss": 0.4207916259765625,
+ "num_input_tokens_seen": 278332088,
+ "step": 6360,
+ "train_runtime": 49198.4506,
+ "train_tokens_per_second": 5657.334
+ },
+ {
+ "epoch": 0.4285618098572583,
+ "grad_norm": 0.6974429090218124,
+ "learning_rate": 8.909534197098966e-06,
+ "loss": 0.4089661121368408,
+ "num_input_tokens_seen": 278560320,
+ "step": 6365,
+ "train_runtime": 49240.3447,
+ "train_tokens_per_second": 5657.156
+ },
+ {
+ "epoch": 0.4288984648532184,
+ "grad_norm": 0.7879065723514492,
+ "learning_rate": 8.907885342154865e-06,
+ "loss": 0.3871757507324219,
+ "num_input_tokens_seen": 278780408,
+ "step": 6370,
+ "train_runtime": 49273.5524,
+ "train_tokens_per_second": 5657.81
+ },
+ {
+ "epoch": 0.42923511984917856,
+ "grad_norm": 0.809098238394157,
+ "learning_rate": 8.906235394383089e-06,
+ "loss": 0.3723259925842285,
+ "num_input_tokens_seen": 278977576,
+ "step": 6375,
+ "train_runtime": 49314.5399,
+ "train_tokens_per_second": 5657.106
+ },
+ {
+ "epoch": 0.4295717748451387,
+ "grad_norm": 0.7788656469697295,
+ "learning_rate": 8.904584354245042e-06,
+ "loss": 0.38217949867248535,
+ "num_input_tokens_seen": 279198056,
+ "step": 6380,
+ "train_runtime": 49349.1158,
+ "train_tokens_per_second": 5657.61
+ },
+ {
+ "epoch": 0.42990842984109884,
+ "grad_norm": 0.7480156005511179,
+ "learning_rate": 8.902932222202433e-06,
+ "loss": 0.39810030460357665,
+ "num_input_tokens_seen": 279431128,
+ "step": 6385,
+ "train_runtime": 49390.7865,
+ "train_tokens_per_second": 5657.556
+ },
+ {
+ "epoch": 0.430245084837059,
+ "grad_norm": 0.7111926102410158,
+ "learning_rate": 8.901278998717272e-06,
+ "loss": 0.37800168991088867,
+ "num_input_tokens_seen": 279655568,
+ "step": 6390,
+ "train_runtime": 49420.9322,
+ "train_tokens_per_second": 5658.646
+ },
+ {
+ "epoch": 0.4305817398330191,
+ "grad_norm": 0.8127379256774018,
+ "learning_rate": 8.899624684251878e-06,
+ "loss": 0.38466432094573977,
+ "num_input_tokens_seen": 279872920,
+ "step": 6395,
+ "train_runtime": 49461.0295,
+ "train_tokens_per_second": 5658.453
+ },
+ {
+ "epoch": 0.43091839482897926,
+ "grad_norm": 0.7440397221517976,
+ "learning_rate": 8.897969279268877e-06,
+ "loss": 0.37961535453796386,
+ "num_input_tokens_seen": 280093688,
+ "step": 6400,
+ "train_runtime": 49494.6401,
+ "train_tokens_per_second": 5659.071
+ },
+ {
+ "epoch": 0.4312550498249394,
+ "grad_norm": 0.7095617451331785,
+ "learning_rate": 8.896312784231196e-06,
+ "loss": 0.3712193012237549,
+ "num_input_tokens_seen": 280305104,
+ "step": 6405,
+ "train_runtime": 49529.3102,
+ "train_tokens_per_second": 5659.378
+ },
+ {
+ "epoch": 0.43159170482089954,
+ "grad_norm": 0.8249055107614544,
+ "learning_rate": 8.89465519960207e-06,
+ "loss": 0.42307000160217284,
+ "num_input_tokens_seen": 280534896,
+ "step": 6410,
+ "train_runtime": 49562.7367,
+ "train_tokens_per_second": 5660.198
+ },
+ {
+ "epoch": 0.4319283598168597,
+ "grad_norm": 0.7044078667504021,
+ "learning_rate": 8.892996525845037e-06,
+ "loss": 0.39340283870697024,
+ "num_input_tokens_seen": 280760536,
+ "step": 6415,
+ "train_runtime": 49609.124,
+ "train_tokens_per_second": 5659.454
+ },
+ {
+ "epoch": 0.4322650148128198,
+ "grad_norm": 0.8758617954616612,
+ "learning_rate": 8.89133676342394e-06,
+ "loss": 0.4034512996673584,
+ "num_input_tokens_seen": 280959600,
+ "step": 6420,
+ "train_runtime": 49647.7511,
+ "train_tokens_per_second": 5659.06
+ },
+ {
+ "epoch": 0.43260166980877995,
+ "grad_norm": 0.7612824097484856,
+ "learning_rate": 8.889675912802923e-06,
+ "loss": 0.38754520416259763,
+ "num_input_tokens_seen": 281173760,
+ "step": 6425,
+ "train_runtime": 49684.4958,
+ "train_tokens_per_second": 5659.185
+ },
+ {
+ "epoch": 0.4329383248047401,
+ "grad_norm": 0.8243557630186539,
+ "learning_rate": 8.888013974446443e-06,
+ "loss": 0.4189424514770508,
+ "num_input_tokens_seen": 281395808,
+ "step": 6430,
+ "train_runtime": 49725.5827,
+ "train_tokens_per_second": 5658.975
+ },
+ {
+ "epoch": 0.43327497980070023,
+ "grad_norm": 0.8439524750777101,
+ "learning_rate": 8.886350948819253e-06,
+ "loss": 0.38129727840423583,
+ "num_input_tokens_seen": 281606072,
+ "step": 6435,
+ "train_runtime": 49758.4427,
+ "train_tokens_per_second": 5659.463
+ },
+ {
+ "epoch": 0.4336116347966604,
+ "grad_norm": 0.8007298009088516,
+ "learning_rate": 8.884686836386412e-06,
+ "loss": 0.38291594982147215,
+ "num_input_tokens_seen": 281822424,
+ "step": 6440,
+ "train_runtime": 49799.792,
+ "train_tokens_per_second": 5659.108
+ },
+ {
+ "epoch": 0.4339482897926205,
+ "grad_norm": 0.7881388654599728,
+ "learning_rate": 8.883021637613286e-06,
+ "loss": 0.4050453662872314,
+ "num_input_tokens_seen": 282049648,
+ "step": 6445,
+ "train_runtime": 49835.0256,
+ "train_tokens_per_second": 5659.667
+ },
+ {
+ "epoch": 0.43428494478858065,
+ "grad_norm": 0.6887341883009596,
+ "learning_rate": 8.881355352965537e-06,
+ "loss": 0.38397510051727296,
+ "num_input_tokens_seen": 282264512,
+ "step": 6450,
+ "train_runtime": 49867.6041,
+ "train_tokens_per_second": 5660.278
+ },
+ {
+ "epoch": 0.4346215997845408,
+ "grad_norm": 0.672101432469643,
+ "learning_rate": 8.879687982909145e-06,
+ "loss": 0.3574419736862183,
+ "num_input_tokens_seen": 282465912,
+ "step": 6455,
+ "train_runtime": 49905.6117,
+ "train_tokens_per_second": 5660.003
+ },
+ {
+ "epoch": 0.43495825478050093,
+ "grad_norm": 0.8916903644937464,
+ "learning_rate": 8.878019527910378e-06,
+ "loss": 0.41092991828918457,
+ "num_input_tokens_seen": 282670240,
+ "step": 6460,
+ "train_runtime": 49945.0146,
+ "train_tokens_per_second": 5659.629
+ },
+ {
+ "epoch": 0.43529490977646107,
+ "grad_norm": 0.8298316229675538,
+ "learning_rate": 8.876349988435815e-06,
+ "loss": 0.44412593841552733,
+ "num_input_tokens_seen": 282887856,
+ "step": 6465,
+ "train_runtime": 49983.0629,
+ "train_tokens_per_second": 5659.674
+ },
+ {
+ "epoch": 0.4356315647724212,
+ "grad_norm": 0.6431274180779993,
+ "learning_rate": 8.874679364952339e-06,
+ "loss": 0.38706722259521487,
+ "num_input_tokens_seen": 283127040,
+ "step": 6470,
+ "train_runtime": 50021.3951,
+ "train_tokens_per_second": 5660.119
+ },
+ {
+ "epoch": 0.43596821976838135,
+ "grad_norm": 0.7656387607672229,
+ "learning_rate": 8.873007657927135e-06,
+ "loss": 0.3893885135650635,
+ "num_input_tokens_seen": 283357232,
+ "step": 6475,
+ "train_runtime": 50062.5333,
+ "train_tokens_per_second": 5660.066
+ },
+ {
+ "epoch": 0.4363048747643415,
+ "grad_norm": 0.7096570879969786,
+ "learning_rate": 8.87133486782769e-06,
+ "loss": 0.35058715343475344,
+ "num_input_tokens_seen": 283558232,
+ "step": 6480,
+ "train_runtime": 50096.5148,
+ "train_tokens_per_second": 5660.239
+ },
+ {
+ "epoch": 0.4366415297603016,
+ "grad_norm": 0.8761505694419173,
+ "learning_rate": 8.869660995121792e-06,
+ "loss": 0.3999923229217529,
+ "num_input_tokens_seen": 283777800,
+ "step": 6485,
+ "train_runtime": 50131.8591,
+ "train_tokens_per_second": 5660.628
+ },
+ {
+ "epoch": 0.43697818475626177,
+ "grad_norm": 0.7555798070151492,
+ "learning_rate": 8.867986040277537e-06,
+ "loss": 0.4016421318054199,
+ "num_input_tokens_seen": 283985824,
+ "step": 6490,
+ "train_runtime": 50172.9,
+ "train_tokens_per_second": 5660.144
+ },
+ {
+ "epoch": 0.4373148397522219,
+ "grad_norm": 0.7001554230524034,
+ "learning_rate": 8.86631000376332e-06,
+ "loss": 0.4068891525268555,
+ "num_input_tokens_seen": 284193496,
+ "step": 6495,
+ "train_runtime": 50210.1755,
+ "train_tokens_per_second": 5660.078
+ },
+ {
+ "epoch": 0.43765149474818205,
+ "grad_norm": 0.809001796546686,
+ "learning_rate": 8.864632886047841e-06,
+ "loss": 0.3719168186187744,
+ "num_input_tokens_seen": 284406672,
+ "step": 6500,
+ "train_runtime": 50246.2209,
+ "train_tokens_per_second": 5660.26
+ },
+ {
+ "epoch": 0.4379881497441422,
+ "grad_norm": 0.8525357252115519,
+ "learning_rate": 8.862954687600097e-06,
+ "loss": 0.4161073684692383,
+ "num_input_tokens_seen": 284620064,
+ "step": 6505,
+ "train_runtime": 50283.9357,
+ "train_tokens_per_second": 5660.258
+ },
+ {
+ "epoch": 0.4383248047401023,
+ "grad_norm": 0.818348330515836,
+ "learning_rate": 8.861275408889393e-06,
+ "loss": 0.4044035911560059,
+ "num_input_tokens_seen": 284827680,
+ "step": 6510,
+ "train_runtime": 50319.756,
+ "train_tokens_per_second": 5660.355
+ },
+ {
+ "epoch": 0.43866145973606246,
+ "grad_norm": 0.8194841529764723,
+ "learning_rate": 8.859595050385332e-06,
+ "loss": 0.40093269348144533,
+ "num_input_tokens_seen": 285050288,
+ "step": 6515,
+ "train_runtime": 50356.7378,
+ "train_tokens_per_second": 5660.619
+ },
+ {
+ "epoch": 0.4389981147320226,
+ "grad_norm": 0.7603950111901223,
+ "learning_rate": 8.857913612557825e-06,
+ "loss": 0.39177913665771485,
+ "num_input_tokens_seen": 285275776,
+ "step": 6520,
+ "train_runtime": 50399.864,
+ "train_tokens_per_second": 5660.249
+ },
+ {
+ "epoch": 0.43933476972798274,
+ "grad_norm": 0.9171110192985658,
+ "learning_rate": 8.856231095877077e-06,
+ "loss": 0.37299246788024903,
+ "num_input_tokens_seen": 285483840,
+ "step": 6525,
+ "train_runtime": 50437.6759,
+ "train_tokens_per_second": 5660.131
+ },
+ {
+ "epoch": 0.4396714247239429,
+ "grad_norm": 0.6722265713323919,
+ "learning_rate": 8.8545475008136e-06,
+ "loss": 0.3728148937225342,
+ "num_input_tokens_seen": 285708616,
+ "step": 6530,
+ "train_runtime": 50478.218,
+ "train_tokens_per_second": 5660.038
+ },
+ {
+ "epoch": 0.440008079719903,
+ "grad_norm": 0.8225035496811262,
+ "learning_rate": 8.852862827838207e-06,
+ "loss": 0.37645254135131834,
+ "num_input_tokens_seen": 285914952,
+ "step": 6535,
+ "train_runtime": 50513.9654,
+ "train_tokens_per_second": 5660.117
+ },
+ {
+ "epoch": 0.44034473471586316,
+ "grad_norm": 0.672139181444301,
+ "learning_rate": 8.851177077422008e-06,
+ "loss": 0.4013407230377197,
+ "num_input_tokens_seen": 286146336,
+ "step": 6540,
+ "train_runtime": 50553.4611,
+ "train_tokens_per_second": 5660.272
+ },
+ {
+ "epoch": 0.4406813897118233,
+ "grad_norm": 0.7644327948595142,
+ "learning_rate": 8.849490250036421e-06,
+ "loss": 0.4249086380004883,
+ "num_input_tokens_seen": 286362088,
+ "step": 6545,
+ "train_runtime": 50590.1862,
+ "train_tokens_per_second": 5660.428
+ },
+ {
+ "epoch": 0.44101804470778344,
+ "grad_norm": 0.8087430911202192,
+ "learning_rate": 8.847802346153159e-06,
+ "loss": 0.3778714895248413,
+ "num_input_tokens_seen": 286576504,
+ "step": 6550,
+ "train_runtime": 50623.1732,
+ "train_tokens_per_second": 5660.975
+ },
+ {
+ "epoch": 0.4413546997037436,
+ "grad_norm": 0.822277412759685,
+ "learning_rate": 8.846113366244241e-06,
+ "loss": 0.40006532669067385,
+ "num_input_tokens_seen": 286802512,
+ "step": 6555,
+ "train_runtime": 50653.7568,
+ "train_tokens_per_second": 5662.019
+ },
+ {
+ "epoch": 0.4416913546997037,
+ "grad_norm": 0.7788883116407057,
+ "learning_rate": 8.844423310781986e-06,
+ "loss": 0.3659277677536011,
+ "num_input_tokens_seen": 287019968,
+ "step": 6560,
+ "train_runtime": 50693.4968,
+ "train_tokens_per_second": 5661.87
+ },
+ {
+ "epoch": 0.44202800969566386,
+ "grad_norm": 0.8189856612851919,
+ "learning_rate": 8.842732180239011e-06,
+ "loss": 0.417068338394165,
+ "num_input_tokens_seen": 287233880,
+ "step": 6565,
+ "train_runtime": 50732.9626,
+ "train_tokens_per_second": 5661.682
+ },
+ {
+ "epoch": 0.442364664691624,
+ "grad_norm": 0.7955988955008321,
+ "learning_rate": 8.841039975088234e-06,
+ "loss": 0.37861127853393556,
+ "num_input_tokens_seen": 287450920,
+ "step": 6570,
+ "train_runtime": 50770.9895,
+ "train_tokens_per_second": 5661.716
+ },
+ {
+ "epoch": 0.44270131968758414,
+ "grad_norm": 0.7673850184930923,
+ "learning_rate": 8.839346695802878e-06,
+ "loss": 0.44512476921081545,
+ "num_input_tokens_seen": 287680008,
+ "step": 6575,
+ "train_runtime": 50808.7713,
+ "train_tokens_per_second": 5662.015
+ },
+ {
+ "epoch": 0.4430379746835443,
+ "grad_norm": 0.707236454889657,
+ "learning_rate": 8.837652342856459e-06,
+ "loss": 0.4026740550994873,
+ "num_input_tokens_seen": 287904440,
+ "step": 6580,
+ "train_runtime": 50842.5258,
+ "train_tokens_per_second": 5662.67
+ },
+ {
+ "epoch": 0.4433746296795044,
+ "grad_norm": 0.7131491003211293,
+ "learning_rate": 8.835956916722803e-06,
+ "loss": 0.37171001434326173,
+ "num_input_tokens_seen": 288123072,
+ "step": 6585,
+ "train_runtime": 50884.3886,
+ "train_tokens_per_second": 5662.308
+ },
+ {
+ "epoch": 0.44371128467546456,
+ "grad_norm": 0.7279859643678576,
+ "learning_rate": 8.834260417876024e-06,
+ "loss": 0.39625725746154783,
+ "num_input_tokens_seen": 288337792,
+ "step": 6590,
+ "train_runtime": 50922.6708,
+ "train_tokens_per_second": 5662.268
+ },
+ {
+ "epoch": 0.4440479396714247,
+ "grad_norm": 0.6822187993121582,
+ "learning_rate": 8.832562846790549e-06,
+ "loss": 0.4032885074615479,
+ "num_input_tokens_seen": 288562504,
+ "step": 6595,
+ "train_runtime": 50958.16,
+ "train_tokens_per_second": 5662.734
+ },
+ {
+ "epoch": 0.4443845946673849,
+ "grad_norm": 0.7696796062355198,
+ "learning_rate": 8.830864203941092e-06,
+ "loss": 0.4127056121826172,
+ "num_input_tokens_seen": 288783304,
+ "step": 6600,
+ "train_runtime": 50994.9294,
+ "train_tokens_per_second": 5662.981
+ },
+ {
+ "epoch": 0.44472124966334503,
+ "grad_norm": 0.796046357490389,
+ "learning_rate": 8.829164489802677e-06,
+ "loss": 0.37214198112487795,
+ "num_input_tokens_seen": 288995032,
+ "step": 6605,
+ "train_runtime": 51030.6262,
+ "train_tokens_per_second": 5663.168
+ },
+ {
+ "epoch": 0.44505790465930517,
+ "grad_norm": 0.8194054840462127,
+ "learning_rate": 8.827463704850622e-06,
+ "loss": 0.3779730796813965,
+ "num_input_tokens_seen": 289211784,
+ "step": 6610,
+ "train_runtime": 51065.1791,
+ "train_tokens_per_second": 5663.581
+ },
+ {
+ "epoch": 0.4453945596552653,
+ "grad_norm": 0.853917923414557,
+ "learning_rate": 8.825761849560547e-06,
+ "loss": 0.3779446125030518,
+ "num_input_tokens_seen": 289426496,
+ "step": 6615,
+ "train_runtime": 51104.5336,
+ "train_tokens_per_second": 5663.421
+ },
+ {
+ "epoch": 0.44573121465122545,
+ "grad_norm": 0.8243805829529494,
+ "learning_rate": 8.824058924408371e-06,
+ "loss": 0.3913204431533813,
+ "num_input_tokens_seen": 289642104,
+ "step": 6620,
+ "train_runtime": 51148.3439,
+ "train_tokens_per_second": 5662.786
+ },
+ {
+ "epoch": 0.4460678696471856,
+ "grad_norm": 0.6977791985037581,
+ "learning_rate": 8.822354929870311e-06,
+ "loss": 0.3713568687438965,
+ "num_input_tokens_seen": 289855080,
+ "step": 6625,
+ "train_runtime": 51186.4829,
+ "train_tokens_per_second": 5662.727
+ },
+ {
+ "epoch": 0.44640452464314573,
+ "grad_norm": 0.6544439461875585,
+ "learning_rate": 8.820649866422884e-06,
+ "loss": 0.3891900539398193,
+ "num_input_tokens_seen": 290080272,
+ "step": 6630,
+ "train_runtime": 51221.0131,
+ "train_tokens_per_second": 5663.306
+ },
+ {
+ "epoch": 0.44674117963910587,
+ "grad_norm": 0.7628155627258327,
+ "learning_rate": 8.818943734542905e-06,
+ "loss": 0.38543388843536375,
+ "num_input_tokens_seen": 290310296,
+ "step": 6635,
+ "train_runtime": 51260.636,
+ "train_tokens_per_second": 5663.416
+ },
+ {
+ "epoch": 0.447077834635066,
+ "grad_norm": 0.7154748015782274,
+ "learning_rate": 8.817236534707486e-06,
+ "loss": 0.38082070350646974,
+ "num_input_tokens_seen": 290530408,
+ "step": 6640,
+ "train_runtime": 51298.209,
+ "train_tokens_per_second": 5663.559
+ },
+ {
+ "epoch": 0.44741448963102615,
+ "grad_norm": 0.719435209044012,
+ "learning_rate": 8.815528267394045e-06,
+ "loss": 0.38055622577667236,
+ "num_input_tokens_seen": 290752096,
+ "step": 6645,
+ "train_runtime": 51339.4277,
+ "train_tokens_per_second": 5663.33
+ },
+ {
+ "epoch": 0.4477511446269863,
+ "grad_norm": 0.8130810621777962,
+ "learning_rate": 8.81381893308029e-06,
+ "loss": 0.42238750457763674,
+ "num_input_tokens_seen": 290976696,
+ "step": 6650,
+ "train_runtime": 51376.6803,
+ "train_tokens_per_second": 5663.595
+ },
+ {
+ "epoch": 0.4480877996229464,
+ "grad_norm": 0.6802969800260558,
+ "learning_rate": 8.81210853224423e-06,
+ "loss": 0.397933292388916,
+ "num_input_tokens_seen": 291195128,
+ "step": 6655,
+ "train_runtime": 51416.0644,
+ "train_tokens_per_second": 5663.505
+ },
+ {
+ "epoch": 0.44842445461890656,
+ "grad_norm": 0.7683337052075077,
+ "learning_rate": 8.810397065364177e-06,
+ "loss": 0.3660165309906006,
+ "num_input_tokens_seen": 291397432,
+ "step": 6660,
+ "train_runtime": 51460.1895,
+ "train_tokens_per_second": 5662.58
+ },
+ {
+ "epoch": 0.4487611096148667,
+ "grad_norm": 0.7385595517387719,
+ "learning_rate": 8.808684532918735e-06,
+ "loss": 0.4035042762756348,
+ "num_input_tokens_seen": 291611984,
+ "step": 6665,
+ "train_runtime": 51497.2825,
+ "train_tokens_per_second": 5662.667
+ },
+ {
+ "epoch": 0.44909776461082684,
+ "grad_norm": 0.8670076691450054,
+ "learning_rate": 8.806970935386807e-06,
+ "loss": 0.37046058177948,
+ "num_input_tokens_seen": 291817952,
+ "step": 6670,
+ "train_runtime": 51539.7964,
+ "train_tokens_per_second": 5661.993
+ },
+ {
+ "epoch": 0.449434419606787,
+ "grad_norm": 0.845351499495225,
+ "learning_rate": 8.805256273247597e-06,
+ "loss": 0.3828955411911011,
+ "num_input_tokens_seen": 292030296,
+ "step": 6675,
+ "train_runtime": 51585.7162,
+ "train_tokens_per_second": 5661.069
+ },
+ {
+ "epoch": 0.4497710746027471,
+ "grad_norm": 0.7558486825710268,
+ "learning_rate": 8.803540546980605e-06,
+ "loss": 0.3911006450653076,
+ "num_input_tokens_seen": 292240392,
+ "step": 6680,
+ "train_runtime": 51625.5292,
+ "train_tokens_per_second": 5660.773
+ },
+ {
+ "epoch": 0.45010772959870726,
+ "grad_norm": 0.7143473072799023,
+ "learning_rate": 8.801823757065628e-06,
+ "loss": 0.39244418144226073,
+ "num_input_tokens_seen": 292456784,
+ "step": 6685,
+ "train_runtime": 51661.3439,
+ "train_tokens_per_second": 5661.037
+ },
+ {
+ "epoch": 0.4504443845946674,
+ "grad_norm": 0.7046306873714272,
+ "learning_rate": 8.800105903982758e-06,
+ "loss": 0.37194869518280027,
+ "num_input_tokens_seen": 292669672,
+ "step": 6690,
+ "train_runtime": 51695.7982,
+ "train_tokens_per_second": 5661.382
+ },
+ {
+ "epoch": 0.45078103959062754,
+ "grad_norm": 0.9038260012464076,
+ "learning_rate": 8.798386988212393e-06,
+ "loss": 0.36289196014404296,
+ "num_input_tokens_seen": 292866896,
+ "step": 6695,
+ "train_runtime": 51737.8021,
+ "train_tokens_per_second": 5660.598
+ },
+ {
+ "epoch": 0.4511176945865877,
+ "grad_norm": 0.7097725989673881,
+ "learning_rate": 8.796667010235216e-06,
+ "loss": 0.4039313793182373,
+ "num_input_tokens_seen": 293088104,
+ "step": 6700,
+ "train_runtime": 51775.0194,
+ "train_tokens_per_second": 5660.801
+ },
+ {
+ "epoch": 0.4514543495825478,
+ "grad_norm": 0.6983233354591538,
+ "learning_rate": 8.794945970532219e-06,
+ "loss": 0.38548543453216555,
+ "num_input_tokens_seen": 293312720,
+ "step": 6705,
+ "train_runtime": 51819.136,
+ "train_tokens_per_second": 5660.317
+ },
+ {
+ "epoch": 0.45179100457850796,
+ "grad_norm": 0.7715442099717464,
+ "learning_rate": 8.79322386958468e-06,
+ "loss": 0.42191410064697266,
+ "num_input_tokens_seen": 293540040,
+ "step": 6710,
+ "train_runtime": 51850.4773,
+ "train_tokens_per_second": 5661.279
+ },
+ {
+ "epoch": 0.4521276595744681,
+ "grad_norm": 0.8667322084375894,
+ "learning_rate": 8.791500707874183e-06,
+ "loss": 0.3590597152709961,
+ "num_input_tokens_seen": 293753504,
+ "step": 6715,
+ "train_runtime": 51887.9851,
+ "train_tokens_per_second": 5661.301
+ },
+ {
+ "epoch": 0.45246431457042824,
+ "grad_norm": 0.7001275719009605,
+ "learning_rate": 8.789776485882601e-06,
+ "loss": 0.40479679107666017,
+ "num_input_tokens_seen": 293980328,
+ "step": 6720,
+ "train_runtime": 51930.6229,
+ "train_tokens_per_second": 5661.021
+ },
+ {
+ "epoch": 0.4528009695663884,
+ "grad_norm": 0.7486619231659534,
+ "learning_rate": 8.788051204092112e-06,
+ "loss": 0.380597448348999,
+ "num_input_tokens_seen": 294201288,
+ "step": 6725,
+ "train_runtime": 51969.1487,
+ "train_tokens_per_second": 5661.076
+ },
+ {
+ "epoch": 0.4531376245623485,
+ "grad_norm": 0.837803412390275,
+ "learning_rate": 8.786324862985183e-06,
+ "loss": 0.39116497039794923,
+ "num_input_tokens_seen": 294414792,
+ "step": 6730,
+ "train_runtime": 52007.4528,
+ "train_tokens_per_second": 5661.012
+ },
+ {
+ "epoch": 0.45347427955830866,
+ "grad_norm": 0.7989569302827977,
+ "learning_rate": 8.78459746304458e-06,
+ "loss": 0.37044734954833985,
+ "num_input_tokens_seen": 294593136,
+ "step": 6735,
+ "train_runtime": 52048.2566,
+ "train_tokens_per_second": 5660.0
+ },
+ {
+ "epoch": 0.4538109345542688,
+ "grad_norm": 0.7910699073925987,
+ "learning_rate": 8.782869004753363e-06,
+ "loss": 0.40142025947570803,
+ "num_input_tokens_seen": 294808392,
+ "step": 6740,
+ "train_runtime": 52084.1003,
+ "train_tokens_per_second": 5660.238
+ },
+ {
+ "epoch": 0.45414758955022894,
+ "grad_norm": 0.7578062648097129,
+ "learning_rate": 8.781139488594892e-06,
+ "loss": 0.39638261795043944,
+ "num_input_tokens_seen": 295032712,
+ "step": 6745,
+ "train_runtime": 52116.6196,
+ "train_tokens_per_second": 5661.01
+ },
+ {
+ "epoch": 0.4544842445461891,
+ "grad_norm": 0.7479401557774803,
+ "learning_rate": 8.779408915052821e-06,
+ "loss": 0.3619338274002075,
+ "num_input_tokens_seen": 295250848,
+ "step": 6750,
+ "train_runtime": 52158.5344,
+ "train_tokens_per_second": 5660.643
+ },
+ {
+ "epoch": 0.4548208995421492,
+ "grad_norm": 0.7929634941210536,
+ "learning_rate": 8.777677284611096e-06,
+ "loss": 0.36865906715393065,
+ "num_input_tokens_seen": 295444808,
+ "step": 6755,
+ "train_runtime": 52194.881,
+ "train_tokens_per_second": 5660.417
+ },
+ {
+ "epoch": 0.45515755453810935,
+ "grad_norm": 0.7977612087147963,
+ "learning_rate": 8.775944597753969e-06,
+ "loss": 0.3817254304885864,
+ "num_input_tokens_seen": 295651640,
+ "step": 6760,
+ "train_runtime": 52235.5188,
+ "train_tokens_per_second": 5659.973
+ },
+ {
+ "epoch": 0.4554942095340695,
+ "grad_norm": 0.7276333392461809,
+ "learning_rate": 8.774210854965972e-06,
+ "loss": 0.4110917568206787,
+ "num_input_tokens_seen": 295872600,
+ "step": 6765,
+ "train_runtime": 52282.5363,
+ "train_tokens_per_second": 5659.11
+ },
+ {
+ "epoch": 0.45583086453002963,
+ "grad_norm": 0.7743347789370404,
+ "learning_rate": 8.772476056731946e-06,
+ "loss": 0.3540276050567627,
+ "num_input_tokens_seen": 296100760,
+ "step": 6770,
+ "train_runtime": 52321.5826,
+ "train_tokens_per_second": 5659.247
+ },
+ {
+ "epoch": 0.4561675195259898,
+ "grad_norm": 0.8417565391365261,
+ "learning_rate": 8.77074020353702e-06,
+ "loss": 0.37112271785736084,
+ "num_input_tokens_seen": 296310112,
+ "step": 6775,
+ "train_runtime": 52364.5089,
+ "train_tokens_per_second": 5658.606
+ },
+ {
+ "epoch": 0.4565041745219499,
+ "grad_norm": 0.8971045549426276,
+ "learning_rate": 8.76900329586662e-06,
+ "loss": 0.416597318649292,
+ "num_input_tokens_seen": 296529072,
+ "step": 6780,
+ "train_runtime": 52408.4219,
+ "train_tokens_per_second": 5658.042
+ },
+ {
+ "epoch": 0.45684082951791005,
+ "grad_norm": 0.7289171592528997,
+ "learning_rate": 8.767265334206467e-06,
+ "loss": 0.38317620754241943,
+ "num_input_tokens_seen": 296750896,
+ "step": 6785,
+ "train_runtime": 52453.3014,
+ "train_tokens_per_second": 5657.43
+ },
+ {
+ "epoch": 0.4571774845138702,
+ "grad_norm": 0.742990584892427,
+ "learning_rate": 8.765526319042577e-06,
+ "loss": 0.38066811561584474,
+ "num_input_tokens_seen": 296973576,
+ "step": 6790,
+ "train_runtime": 52484.833,
+ "train_tokens_per_second": 5658.274
+ },
+ {
+ "epoch": 0.45751413950983033,
+ "grad_norm": 0.7941387596625845,
+ "learning_rate": 8.763786250861258e-06,
+ "loss": 0.35711822509765623,
+ "num_input_tokens_seen": 297198488,
+ "step": 6795,
+ "train_runtime": 52525.9579,
+ "train_tokens_per_second": 5658.126
+ },
+ {
+ "epoch": 0.45785079450579047,
+ "grad_norm": 0.7637071617244607,
+ "learning_rate": 8.762045130149114e-06,
+ "loss": 0.37547504901885986,
+ "num_input_tokens_seen": 297424408,
+ "step": 6800,
+ "train_runtime": 52568.1627,
+ "train_tokens_per_second": 5657.881
+ },
+ {
+ "epoch": 0.4581874495017506,
+ "grad_norm": 0.7114371063573226,
+ "learning_rate": 8.76030295739305e-06,
+ "loss": 0.391666579246521,
+ "num_input_tokens_seen": 297652664,
+ "step": 6805,
+ "train_runtime": 52605.904,
+ "train_tokens_per_second": 5658.161
+ },
+ {
+ "epoch": 0.45852410449771075,
+ "grad_norm": 0.7263315072433297,
+ "learning_rate": 8.758559733080252e-06,
+ "loss": 0.4111741065979004,
+ "num_input_tokens_seen": 297874936,
+ "step": 6810,
+ "train_runtime": 52637.0942,
+ "train_tokens_per_second": 5659.031
+ },
+ {
+ "epoch": 0.4588607594936709,
+ "grad_norm": 0.7359316103796298,
+ "learning_rate": 8.75681545769821e-06,
+ "loss": 0.38690164089202883,
+ "num_input_tokens_seen": 298107584,
+ "step": 6815,
+ "train_runtime": 52669.678,
+ "train_tokens_per_second": 5659.947
+ },
+ {
+ "epoch": 0.45919741448963103,
+ "grad_norm": 0.8032263081480874,
+ "learning_rate": 8.755070131734706e-06,
+ "loss": 0.3867367744445801,
+ "num_input_tokens_seen": 298320208,
+ "step": 6820,
+ "train_runtime": 52707.8474,
+ "train_tokens_per_second": 5659.882
+ },
+ {
+ "epoch": 0.45953406948559117,
+ "grad_norm": 0.7833333891541101,
+ "learning_rate": 8.753323755677812e-06,
+ "loss": 0.4013802528381348,
+ "num_input_tokens_seen": 298533552,
+ "step": 6825,
+ "train_runtime": 52749.6034,
+ "train_tokens_per_second": 5659.446
+ },
+ {
+ "epoch": 0.4598707244815513,
+ "grad_norm": 0.7609318114113853,
+ "learning_rate": 8.7515763300159e-06,
+ "loss": 0.3463205575942993,
+ "num_input_tokens_seen": 298746696,
+ "step": 6830,
+ "train_runtime": 52789.3591,
+ "train_tokens_per_second": 5659.222
+ },
+ {
+ "epoch": 0.46020737947751145,
+ "grad_norm": 0.6930204184723102,
+ "learning_rate": 8.74982785523763e-06,
+ "loss": 0.40210418701171874,
+ "num_input_tokens_seen": 298972384,
+ "step": 6835,
+ "train_runtime": 52827.2223,
+ "train_tokens_per_second": 5659.438
+ },
+ {
+ "epoch": 0.4605440344734716,
+ "grad_norm": 0.7026299121313064,
+ "learning_rate": 8.748078331831957e-06,
+ "loss": 0.3934351682662964,
+ "num_input_tokens_seen": 299200112,
+ "step": 6840,
+ "train_runtime": 52862.9542,
+ "train_tokens_per_second": 5659.92
+ },
+ {
+ "epoch": 0.4608806894694317,
+ "grad_norm": 0.7019416941915204,
+ "learning_rate": 8.746327760288129e-06,
+ "loss": 0.3836989164352417,
+ "num_input_tokens_seen": 299437024,
+ "step": 6845,
+ "train_runtime": 52894.4779,
+ "train_tokens_per_second": 5661.026
+ },
+ {
+ "epoch": 0.46121734446539187,
+ "grad_norm": 0.8234006493355744,
+ "learning_rate": 8.744576141095691e-06,
+ "loss": 0.39553108215332033,
+ "num_input_tokens_seen": 299641752,
+ "step": 6850,
+ "train_runtime": 52932.8412,
+ "train_tokens_per_second": 5660.791
+ },
+ {
+ "epoch": 0.461553999461352,
+ "grad_norm": 0.846531534553998,
+ "learning_rate": 8.742823474744476e-06,
+ "loss": 0.39920735359191895,
+ "num_input_tokens_seen": 299852504,
+ "step": 6855,
+ "train_runtime": 52970.3403,
+ "train_tokens_per_second": 5660.762
+ },
+ {
+ "epoch": 0.46189065445731214,
+ "grad_norm": 0.7339872597188516,
+ "learning_rate": 8.74106976172461e-06,
+ "loss": 0.40232462882995607,
+ "num_input_tokens_seen": 300073192,
+ "step": 6860,
+ "train_runtime": 53013.6318,
+ "train_tokens_per_second": 5660.302
+ },
+ {
+ "epoch": 0.4622273094532723,
+ "grad_norm": 0.8137834490765111,
+ "learning_rate": 8.739315002526513e-06,
+ "loss": 0.40238103866577146,
+ "num_input_tokens_seen": 300287552,
+ "step": 6865,
+ "train_runtime": 53055.019,
+ "train_tokens_per_second": 5659.927
+ },
+ {
+ "epoch": 0.4625639644492324,
+ "grad_norm": 0.7042015061046794,
+ "learning_rate": 8.737559197640902e-06,
+ "loss": 0.3717548131942749,
+ "num_input_tokens_seen": 300496376,
+ "step": 6870,
+ "train_runtime": 53091.6564,
+ "train_tokens_per_second": 5659.955
+ },
+ {
+ "epoch": 0.46290061944519256,
+ "grad_norm": 0.7325334644495457,
+ "learning_rate": 8.735802347558778e-06,
+ "loss": 0.4027840614318848,
+ "num_input_tokens_seen": 300709416,
+ "step": 6875,
+ "train_runtime": 53135.1809,
+ "train_tokens_per_second": 5659.328
+ },
+ {
+ "epoch": 0.4632372744411527,
+ "grad_norm": 0.7655214239764673,
+ "learning_rate": 8.734044452771442e-06,
+ "loss": 0.38804852962493896,
+ "num_input_tokens_seen": 300918752,
+ "step": 6880,
+ "train_runtime": 53179.1079,
+ "train_tokens_per_second": 5658.59
+ },
+ {
+ "epoch": 0.46357392943711284,
+ "grad_norm": 0.8904799503928702,
+ "learning_rate": 8.73228551377048e-06,
+ "loss": 0.40845584869384766,
+ "num_input_tokens_seen": 301139720,
+ "step": 6885,
+ "train_runtime": 53208.9509,
+ "train_tokens_per_second": 5659.569
+ },
+ {
+ "epoch": 0.463910584433073,
+ "grad_norm": 0.6748289632604044,
+ "learning_rate": 8.730525531047776e-06,
+ "loss": 0.4137433052062988,
+ "num_input_tokens_seen": 301362752,
+ "step": 6890,
+ "train_runtime": 53254.1898,
+ "train_tokens_per_second": 5658.949
+ },
+ {
+ "epoch": 0.4642472394290331,
+ "grad_norm": 0.7035224245591345,
+ "learning_rate": 8.728764505095503e-06,
+ "loss": 0.4120682716369629,
+ "num_input_tokens_seen": 301587944,
+ "step": 6895,
+ "train_runtime": 53291.4984,
+ "train_tokens_per_second": 5659.213
+ },
+ {
+ "epoch": 0.46458389442499326,
+ "grad_norm": 0.6435345300249111,
+ "learning_rate": 8.727002436406126e-06,
+ "loss": 0.3724491834640503,
+ "num_input_tokens_seen": 301823392,
+ "step": 6900,
+ "train_runtime": 53331.8202,
+ "train_tokens_per_second": 5659.349
+ },
+ {
+ "epoch": 0.4649205494209534,
+ "grad_norm": 0.7100192481355228,
+ "learning_rate": 8.725239325472404e-06,
+ "loss": 0.4057363510131836,
+ "num_input_tokens_seen": 302046392,
+ "step": 6905,
+ "train_runtime": 53370.503,
+ "train_tokens_per_second": 5659.426
+ },
+ {
+ "epoch": 0.46525720441691354,
+ "grad_norm": 0.7998034880411997,
+ "learning_rate": 8.723475172787381e-06,
+ "loss": 0.3984755277633667,
+ "num_input_tokens_seen": 302270848,
+ "step": 6910,
+ "train_runtime": 53413.2206,
+ "train_tokens_per_second": 5659.102
+ },
+ {
+ "epoch": 0.4655938594128737,
+ "grad_norm": 0.9282661770223051,
+ "learning_rate": 8.7217099788444e-06,
+ "loss": 0.38761351108551023,
+ "num_input_tokens_seen": 302484424,
+ "step": 6915,
+ "train_runtime": 53452.6504,
+ "train_tokens_per_second": 5658.923
+ },
+ {
+ "epoch": 0.4659305144088338,
+ "grad_norm": 0.6829252086088465,
+ "learning_rate": 8.719943744137091e-06,
+ "loss": 0.3843079566955566,
+ "num_input_tokens_seen": 302706608,
+ "step": 6920,
+ "train_runtime": 53494.2752,
+ "train_tokens_per_second": 5658.673
+ },
+ {
+ "epoch": 0.46626716940479396,
+ "grad_norm": 0.7370197158639645,
+ "learning_rate": 8.718176469159378e-06,
+ "loss": 0.40040907859802244,
+ "num_input_tokens_seen": 302946080,
+ "step": 6925,
+ "train_runtime": 53527.8006,
+ "train_tokens_per_second": 5659.603
+ },
+ {
+ "epoch": 0.4666038244007541,
+ "grad_norm": 0.7744303086322805,
+ "learning_rate": 8.716408154405469e-06,
+ "loss": 0.370440411567688,
+ "num_input_tokens_seen": 303172472,
+ "step": 6930,
+ "train_runtime": 53558.5862,
+ "train_tokens_per_second": 5660.576
+ },
+ {
+ "epoch": 0.46694047939671424,
+ "grad_norm": 0.9149869901290688,
+ "learning_rate": 8.714638800369872e-06,
+ "loss": 0.4163699150085449,
+ "num_input_tokens_seen": 303374536,
+ "step": 6935,
+ "train_runtime": 53596.9127,
+ "train_tokens_per_second": 5660.299
+ },
+ {
+ "epoch": 0.4672771343926744,
+ "grad_norm": 0.6478644031670531,
+ "learning_rate": 8.712868407547378e-06,
+ "loss": 0.41262283325195315,
+ "num_input_tokens_seen": 303604704,
+ "step": 6940,
+ "train_runtime": 53634.4097,
+ "train_tokens_per_second": 5660.633
+ },
+ {
+ "epoch": 0.4676137893886345,
+ "grad_norm": 0.8540394960614904,
+ "learning_rate": 8.711096976433076e-06,
+ "loss": 0.40007991790771485,
+ "num_input_tokens_seen": 303830584,
+ "step": 6945,
+ "train_runtime": 53675.8336,
+ "train_tokens_per_second": 5660.473
+ },
+ {
+ "epoch": 0.46795044438459465,
+ "grad_norm": 0.6854573080771246,
+ "learning_rate": 8.709324507522337e-06,
+ "loss": 0.3609914779663086,
+ "num_input_tokens_seen": 304051832,
+ "step": 6950,
+ "train_runtime": 53710.0276,
+ "train_tokens_per_second": 5660.988
+ },
+ {
+ "epoch": 0.4682870993805548,
+ "grad_norm": 0.74529061138361,
+ "learning_rate": 8.707551001310828e-06,
+ "loss": 0.3842365264892578,
+ "num_input_tokens_seen": 304279352,
+ "step": 6955,
+ "train_runtime": 53749.8544,
+ "train_tokens_per_second": 5661.027
+ },
+ {
+ "epoch": 0.46862375437651493,
+ "grad_norm": 0.7026982064246363,
+ "learning_rate": 8.705776458294503e-06,
+ "loss": 0.387627911567688,
+ "num_input_tokens_seen": 304487200,
+ "step": 6960,
+ "train_runtime": 53785.1129,
+ "train_tokens_per_second": 5661.18
+ },
+ {
+ "epoch": 0.4689604093724751,
+ "grad_norm": 0.7214247021583069,
+ "learning_rate": 8.704000878969608e-06,
+ "loss": 0.37026023864746094,
+ "num_input_tokens_seen": 304693840,
+ "step": 6965,
+ "train_runtime": 53822.9634,
+ "train_tokens_per_second": 5661.038
+ },
+ {
+ "epoch": 0.4692970643684352,
+ "grad_norm": 0.6979359178232242,
+ "learning_rate": 8.702224263832679e-06,
+ "loss": 0.3628101825714111,
+ "num_input_tokens_seen": 304921392,
+ "step": 6970,
+ "train_runtime": 53861.4766,
+ "train_tokens_per_second": 5661.215
+ },
+ {
+ "epoch": 0.46963371936439535,
+ "grad_norm": 0.7813572135644429,
+ "learning_rate": 8.700446613380542e-06,
+ "loss": 0.4275547504425049,
+ "num_input_tokens_seen": 305149136,
+ "step": 6975,
+ "train_runtime": 53899.1275,
+ "train_tokens_per_second": 5661.486
+ },
+ {
+ "epoch": 0.4699703743603555,
+ "grad_norm": 0.800833875206026,
+ "learning_rate": 8.698667928110307e-06,
+ "loss": 0.373789119720459,
+ "num_input_tokens_seen": 305369336,
+ "step": 6980,
+ "train_runtime": 53936.8451,
+ "train_tokens_per_second": 5661.609
+ },
+ {
+ "epoch": 0.47030702935631563,
+ "grad_norm": 0.8980829202572613,
+ "learning_rate": 8.696888208519381e-06,
+ "loss": 0.3973226070404053,
+ "num_input_tokens_seen": 305581504,
+ "step": 6985,
+ "train_runtime": 53973.4435,
+ "train_tokens_per_second": 5661.701
+ },
+ {
+ "epoch": 0.47064368435227577,
+ "grad_norm": 0.7351189903914839,
+ "learning_rate": 8.695107455105454e-06,
+ "loss": 0.34683785438537595,
+ "num_input_tokens_seen": 305790632,
+ "step": 6990,
+ "train_runtime": 54006.1459,
+ "train_tokens_per_second": 5662.145
+ },
+ {
+ "epoch": 0.4709803393482359,
+ "grad_norm": 0.7928821233878864,
+ "learning_rate": 8.693325668366513e-06,
+ "loss": 0.39075660705566406,
+ "num_input_tokens_seen": 306004232,
+ "step": 6995,
+ "train_runtime": 54042.3536,
+ "train_tokens_per_second": 5662.304
+ },
+ {
+ "epoch": 0.47131699434419605,
+ "grad_norm": 0.8527776847132168,
+ "learning_rate": 8.691542848800825e-06,
+ "loss": 0.39600386619567873,
+ "num_input_tokens_seen": 306212208,
+ "step": 7000,
+ "train_runtime": 54084.5742,
+ "train_tokens_per_second": 5661.729
+ },
+ {
+ "epoch": 0.4716536493401562,
+ "grad_norm": 0.8808374134194763,
+ "learning_rate": 8.68975899690695e-06,
+ "loss": 0.38110928535461425,
+ "num_input_tokens_seen": 306437744,
+ "step": 7005,
+ "train_runtime": 54125.0428,
+ "train_tokens_per_second": 5661.663
+ },
+ {
+ "epoch": 0.47199030433611633,
+ "grad_norm": 0.7423709319863485,
+ "learning_rate": 8.687974113183738e-06,
+ "loss": 0.37776660919189453,
+ "num_input_tokens_seen": 306667400,
+ "step": 7010,
+ "train_runtime": 54160.5555,
+ "train_tokens_per_second": 5662.191
+ },
+ {
+ "epoch": 0.47232695933207647,
+ "grad_norm": 0.8031207862343954,
+ "learning_rate": 8.686188198130326e-06,
+ "loss": 0.3893289089202881,
+ "num_input_tokens_seen": 306891232,
+ "step": 7015,
+ "train_runtime": 54195.218,
+ "train_tokens_per_second": 5662.699
+ },
+ {
+ "epoch": 0.4726636143280366,
+ "grad_norm": 0.716418279114708,
+ "learning_rate": 8.684401252246138e-06,
+ "loss": 0.3991250038146973,
+ "num_input_tokens_seen": 307116336,
+ "step": 7020,
+ "train_runtime": 54233.7853,
+ "train_tokens_per_second": 5662.823
+ },
+ {
+ "epoch": 0.47300026932399675,
+ "grad_norm": 0.7731462314856246,
+ "learning_rate": 8.68261327603089e-06,
+ "loss": 0.42037010192871094,
+ "num_input_tokens_seen": 307328112,
+ "step": 7025,
+ "train_runtime": 54273.3793,
+ "train_tokens_per_second": 5662.594
+ },
+ {
+ "epoch": 0.4733369243199569,
+ "grad_norm": 0.7619151144650879,
+ "learning_rate": 8.68082426998458e-06,
+ "loss": 0.42117109298706057,
+ "num_input_tokens_seen": 307558384,
+ "step": 7030,
+ "train_runtime": 54315.1352,
+ "train_tokens_per_second": 5662.48
+ },
+ {
+ "epoch": 0.473673579315917,
+ "grad_norm": 0.8366328207039446,
+ "learning_rate": 8.6790342346075e-06,
+ "loss": 0.355437707901001,
+ "num_input_tokens_seen": 307757832,
+ "step": 7035,
+ "train_runtime": 54355.0453,
+ "train_tokens_per_second": 5661.992
+ },
+ {
+ "epoch": 0.47401023431187717,
+ "grad_norm": 0.694627660462694,
+ "learning_rate": 8.677243170400228e-06,
+ "loss": 0.3744319438934326,
+ "num_input_tokens_seen": 307964848,
+ "step": 7040,
+ "train_runtime": 54392.4804,
+ "train_tokens_per_second": 5661.901
+ },
+ {
+ "epoch": 0.4743468893078373,
+ "grad_norm": 0.7593548972261533,
+ "learning_rate": 8.675451077863632e-06,
+ "loss": 0.40117859840393066,
+ "num_input_tokens_seen": 308180176,
+ "step": 7045,
+ "train_runtime": 54429.867,
+ "train_tokens_per_second": 5661.968
+ },
+ {
+ "epoch": 0.47468354430379744,
+ "grad_norm": 0.7788878005901869,
+ "learning_rate": 8.67365795749886e-06,
+ "loss": 0.37619283199310305,
+ "num_input_tokens_seen": 308393256,
+ "step": 7050,
+ "train_runtime": 54475.6142,
+ "train_tokens_per_second": 5661.125
+ },
+ {
+ "epoch": 0.4750201992997576,
+ "grad_norm": 0.6935999303413757,
+ "learning_rate": 8.671863809807355e-06,
+ "loss": 0.36246466636657715,
+ "num_input_tokens_seen": 308598120,
+ "step": 7055,
+ "train_runtime": 54512.6696,
+ "train_tokens_per_second": 5661.035
+ },
+ {
+ "epoch": 0.4753568542957177,
+ "grad_norm": 0.7796482197914519,
+ "learning_rate": 8.670068635290844e-06,
+ "loss": 0.3815620899200439,
+ "num_input_tokens_seen": 308814312,
+ "step": 7060,
+ "train_runtime": 54553.6102,
+ "train_tokens_per_second": 5660.749
+ },
+ {
+ "epoch": 0.47569350929167786,
+ "grad_norm": 0.6394547128504924,
+ "learning_rate": 8.668272434451343e-06,
+ "loss": 0.3547971725463867,
+ "num_input_tokens_seen": 309036648,
+ "step": 7065,
+ "train_runtime": 54589.1907,
+ "train_tokens_per_second": 5661.133
+ },
+ {
+ "epoch": 0.476030164287638,
+ "grad_norm": 0.8424232555426785,
+ "learning_rate": 8.666475207791154e-06,
+ "loss": 0.3938549518585205,
+ "num_input_tokens_seen": 309237312,
+ "step": 7070,
+ "train_runtime": 54626.0249,
+ "train_tokens_per_second": 5660.989
+ },
+ {
+ "epoch": 0.47636681928359814,
+ "grad_norm": 0.7156676868042459,
+ "learning_rate": 8.664676955812863e-06,
+ "loss": 0.3805175065994263,
+ "num_input_tokens_seen": 309461072,
+ "step": 7075,
+ "train_runtime": 54665.5373,
+ "train_tokens_per_second": 5660.99
+ },
+ {
+ "epoch": 0.4767034742795583,
+ "grad_norm": 0.7803230987384645,
+ "learning_rate": 8.662877679019348e-06,
+ "loss": 0.38736543655395506,
+ "num_input_tokens_seen": 309676264,
+ "step": 7080,
+ "train_runtime": 54702.5852,
+ "train_tokens_per_second": 5661.09
+ },
+ {
+ "epoch": 0.4770401292755185,
+ "grad_norm": 0.811046251111573,
+ "learning_rate": 8.66107737791377e-06,
+ "loss": 0.40035252571105956,
+ "num_input_tokens_seen": 309908584,
+ "step": 7085,
+ "train_runtime": 54739.8422,
+ "train_tokens_per_second": 5661.481
+ },
+ {
+ "epoch": 0.4773767842714786,
+ "grad_norm": 0.8088100128547995,
+ "learning_rate": 8.659276052999574e-06,
+ "loss": 0.37158200740814207,
+ "num_input_tokens_seen": 310109536,
+ "step": 7090,
+ "train_runtime": 54784.3231,
+ "train_tokens_per_second": 5660.552
+ },
+ {
+ "epoch": 0.47771343926743876,
+ "grad_norm": 0.6607865351139958,
+ "learning_rate": 8.657473704780504e-06,
+ "loss": 0.3958035707473755,
+ "num_input_tokens_seen": 310316416,
+ "step": 7095,
+ "train_runtime": 54823.0111,
+ "train_tokens_per_second": 5660.331
+ },
+ {
+ "epoch": 0.4780500942633989,
+ "grad_norm": 0.7508776157927014,
+ "learning_rate": 8.65567033376057e-06,
+ "loss": 0.3861806392669678,
+ "num_input_tokens_seen": 310543280,
+ "step": 7100,
+ "train_runtime": 54858.1842,
+ "train_tokens_per_second": 5660.838
+ },
+ {
+ "epoch": 0.47838674925935903,
+ "grad_norm": 0.6796527894531441,
+ "learning_rate": 8.65386594044409e-06,
+ "loss": 0.38321571350097655,
+ "num_input_tokens_seen": 310767656,
+ "step": 7105,
+ "train_runtime": 54899.3833,
+ "train_tokens_per_second": 5660.677
+ },
+ {
+ "epoch": 0.4787234042553192,
+ "grad_norm": 0.7839875292722721,
+ "learning_rate": 8.652060525335647e-06,
+ "loss": 0.35958447456359866,
+ "num_input_tokens_seen": 310977648,
+ "step": 7110,
+ "train_runtime": 54935.0727,
+ "train_tokens_per_second": 5660.822
+ },
+ {
+ "epoch": 0.4790600592512793,
+ "grad_norm": 0.6805039530532168,
+ "learning_rate": 8.650254088940125e-06,
+ "loss": 0.37933621406555174,
+ "num_input_tokens_seen": 311184472,
+ "step": 7115,
+ "train_runtime": 54970.8187,
+ "train_tokens_per_second": 5660.903
+ },
+ {
+ "epoch": 0.47939671424723945,
+ "grad_norm": 0.8690524861477827,
+ "learning_rate": 8.648446631762686e-06,
+ "loss": 0.3870699882507324,
+ "num_input_tokens_seen": 311386608,
+ "step": 7120,
+ "train_runtime": 55007.3848,
+ "train_tokens_per_second": 5660.815
+ },
+ {
+ "epoch": 0.4797333692431996,
+ "grad_norm": 0.7834933561721031,
+ "learning_rate": 8.646638154308781e-06,
+ "loss": 0.3672234296798706,
+ "num_input_tokens_seen": 311600392,
+ "step": 7125,
+ "train_runtime": 55045.8437,
+ "train_tokens_per_second": 5660.743
+ },
+ {
+ "epoch": 0.48007002423915973,
+ "grad_norm": 0.6821280974528945,
+ "learning_rate": 8.644828657084144e-06,
+ "loss": 0.3536520004272461,
+ "num_input_tokens_seen": 311826176,
+ "step": 7130,
+ "train_runtime": 55086.3186,
+ "train_tokens_per_second": 5660.683
+ },
+ {
+ "epoch": 0.48040667923511987,
+ "grad_norm": 0.7090523007119766,
+ "learning_rate": 8.643018140594797e-06,
+ "loss": 0.35062260627746583,
+ "num_input_tokens_seen": 312040616,
+ "step": 7135,
+ "train_runtime": 55118.7948,
+ "train_tokens_per_second": 5661.238
+ },
+ {
+ "epoch": 0.48074333423108,
+ "grad_norm": 0.7694829828758608,
+ "learning_rate": 8.641206605347042e-06,
+ "loss": 0.3680727958679199,
+ "num_input_tokens_seen": 312250776,
+ "step": 7140,
+ "train_runtime": 55154.654,
+ "train_tokens_per_second": 5661.368
+ },
+ {
+ "epoch": 0.48107998922704015,
+ "grad_norm": 0.7834141310420747,
+ "learning_rate": 8.639394051847472e-06,
+ "loss": 0.3820206642150879,
+ "num_input_tokens_seen": 312469768,
+ "step": 7145,
+ "train_runtime": 55194.213,
+ "train_tokens_per_second": 5661.278
+ },
+ {
+ "epoch": 0.4814166442230003,
+ "grad_norm": 0.8086494042960466,
+ "learning_rate": 8.63758048060296e-06,
+ "loss": 0.3749295473098755,
+ "num_input_tokens_seen": 312686816,
+ "step": 7150,
+ "train_runtime": 55234.7643,
+ "train_tokens_per_second": 5661.051
+ },
+ {
+ "epoch": 0.48175329921896043,
+ "grad_norm": 1.146252827120329,
+ "learning_rate": 8.635765892120666e-06,
+ "loss": 0.35863730907440183,
+ "num_input_tokens_seen": 312928240,
+ "step": 7155,
+ "train_runtime": 55276.3406,
+ "train_tokens_per_second": 5661.161
+ },
+ {
+ "epoch": 0.48208995421492057,
+ "grad_norm": 0.6322723599846541,
+ "learning_rate": 8.633950286908032e-06,
+ "loss": 0.3957623243331909,
+ "num_input_tokens_seen": 313146400,
+ "step": 7160,
+ "train_runtime": 55318.9273,
+ "train_tokens_per_second": 5660.746
+ },
+ {
+ "epoch": 0.4824266092108807,
+ "grad_norm": 0.8242447392294052,
+ "learning_rate": 8.63213366547279e-06,
+ "loss": 0.41290960311889646,
+ "num_input_tokens_seen": 313353080,
+ "step": 7165,
+ "train_runtime": 55354.3316,
+ "train_tokens_per_second": 5660.859
+ },
+ {
+ "epoch": 0.48276326420684085,
+ "grad_norm": 0.8189218668862145,
+ "learning_rate": 8.630316028322952e-06,
+ "loss": 0.3599478960037231,
+ "num_input_tokens_seen": 313556520,
+ "step": 7170,
+ "train_runtime": 55393.1558,
+ "train_tokens_per_second": 5660.564
+ },
+ {
+ "epoch": 0.483099919202801,
+ "grad_norm": 0.8049613617878884,
+ "learning_rate": 8.628497375966811e-06,
+ "loss": 0.3688513278961182,
+ "num_input_tokens_seen": 313791304,
+ "step": 7175,
+ "train_runtime": 55433.9631,
+ "train_tokens_per_second": 5660.633
+ },
+ {
+ "epoch": 0.4834365741987611,
+ "grad_norm": 0.7403145334116261,
+ "learning_rate": 8.62667770891295e-06,
+ "loss": 0.4179924488067627,
+ "num_input_tokens_seen": 314025432,
+ "step": 7180,
+ "train_runtime": 55479.2927,
+ "train_tokens_per_second": 5660.228
+ },
+ {
+ "epoch": 0.48377322919472127,
+ "grad_norm": 0.6161204381365467,
+ "learning_rate": 8.624857027670232e-06,
+ "loss": 0.3718327522277832,
+ "num_input_tokens_seen": 314253744,
+ "step": 7185,
+ "train_runtime": 55518.1876,
+ "train_tokens_per_second": 5660.375
+ },
+ {
+ "epoch": 0.4841098841906814,
+ "grad_norm": 0.6673541170021607,
+ "learning_rate": 8.623035332747804e-06,
+ "loss": 0.4232837677001953,
+ "num_input_tokens_seen": 314457360,
+ "step": 7190,
+ "train_runtime": 55555.7684,
+ "train_tokens_per_second": 5660.211
+ },
+ {
+ "epoch": 0.48444653918664154,
+ "grad_norm": 0.8639989632411573,
+ "learning_rate": 8.6212126246551e-06,
+ "loss": 0.35295858383178713,
+ "num_input_tokens_seen": 314659296,
+ "step": 7195,
+ "train_runtime": 55590.9252,
+ "train_tokens_per_second": 5660.264
+ },
+ {
+ "epoch": 0.4847831941826017,
+ "grad_norm": 0.7194371682580046,
+ "learning_rate": 8.619388903901833e-06,
+ "loss": 0.397071385383606,
+ "num_input_tokens_seen": 314886400,
+ "step": 7200,
+ "train_runtime": 55632.7504,
+ "train_tokens_per_second": 5660.09
+ },
+ {
+ "epoch": 0.4851198491785618,
+ "grad_norm": 0.7895653948671645,
+ "learning_rate": 8.617564170997998e-06,
+ "loss": 0.393186354637146,
+ "num_input_tokens_seen": 315113560,
+ "step": 7205,
+ "train_runtime": 55665.6693,
+ "train_tokens_per_second": 5660.824
+ },
+ {
+ "epoch": 0.48545650417452196,
+ "grad_norm": 0.6916492568995722,
+ "learning_rate": 8.61573842645388e-06,
+ "loss": 0.38975393772125244,
+ "num_input_tokens_seen": 315339760,
+ "step": 7210,
+ "train_runtime": 55701.6461,
+ "train_tokens_per_second": 5661.229
+ },
+ {
+ "epoch": 0.4857931591704821,
+ "grad_norm": 0.7125114232933183,
+ "learning_rate": 8.61391167078004e-06,
+ "loss": 0.3773904085159302,
+ "num_input_tokens_seen": 315564560,
+ "step": 7215,
+ "train_runtime": 55739.932,
+ "train_tokens_per_second": 5661.373
+ },
+ {
+ "epoch": 0.48612981416644224,
+ "grad_norm": 0.7816578333674177,
+ "learning_rate": 8.612083904487324e-06,
+ "loss": 0.4223905563354492,
+ "num_input_tokens_seen": 315780664,
+ "step": 7220,
+ "train_runtime": 55782.1816,
+ "train_tokens_per_second": 5660.959
+ },
+ {
+ "epoch": 0.4864664691624024,
+ "grad_norm": 0.8225392117617899,
+ "learning_rate": 8.610255128086867e-06,
+ "loss": 0.39981255531311033,
+ "num_input_tokens_seen": 315985224,
+ "step": 7225,
+ "train_runtime": 55817.5809,
+ "train_tokens_per_second": 5661.034
+ },
+ {
+ "epoch": 0.4868031241583625,
+ "grad_norm": 0.7192041520493799,
+ "learning_rate": 8.60842534209007e-06,
+ "loss": 0.35159389972686766,
+ "num_input_tokens_seen": 316206120,
+ "step": 7230,
+ "train_runtime": 55856.1823,
+ "train_tokens_per_second": 5661.076
+ },
+ {
+ "epoch": 0.48713977915432266,
+ "grad_norm": 0.7538381499748801,
+ "learning_rate": 8.606594547008636e-06,
+ "loss": 0.33089404106140136,
+ "num_input_tokens_seen": 316428112,
+ "step": 7235,
+ "train_runtime": 55889.528,
+ "train_tokens_per_second": 5661.671
+ },
+ {
+ "epoch": 0.4874764341502828,
+ "grad_norm": 0.7734857420248509,
+ "learning_rate": 8.604762743354536e-06,
+ "loss": 0.39783239364624023,
+ "num_input_tokens_seen": 316666376,
+ "step": 7240,
+ "train_runtime": 55922.8766,
+ "train_tokens_per_second": 5662.555
+ },
+ {
+ "epoch": 0.48781308914624294,
+ "grad_norm": 0.7475280321551805,
+ "learning_rate": 8.602929931640029e-06,
+ "loss": 0.3995728254318237,
+ "num_input_tokens_seen": 316868832,
+ "step": 7245,
+ "train_runtime": 55959.5907,
+ "train_tokens_per_second": 5662.458
+ },
+ {
+ "epoch": 0.4881497441422031,
+ "grad_norm": 0.8349680082803242,
+ "learning_rate": 8.601096112377659e-06,
+ "loss": 0.4125559329986572,
+ "num_input_tokens_seen": 317082248,
+ "step": 7250,
+ "train_runtime": 56002.6917,
+ "train_tokens_per_second": 5661.911
+ },
+ {
+ "epoch": 0.4884863991381632,
+ "grad_norm": 0.6943776304621931,
+ "learning_rate": 8.59926128608024e-06,
+ "loss": 0.3738250255584717,
+ "num_input_tokens_seen": 317303120,
+ "step": 7255,
+ "train_runtime": 56039.6874,
+ "train_tokens_per_second": 5662.114
+ },
+ {
+ "epoch": 0.48882305413412336,
+ "grad_norm": 0.7789443419004642,
+ "learning_rate": 8.597425453260881e-06,
+ "loss": 0.3914709806442261,
+ "num_input_tokens_seen": 317515672,
+ "step": 7260,
+ "train_runtime": 56082.3268,
+ "train_tokens_per_second": 5661.599
+ },
+ {
+ "epoch": 0.4891597091300835,
+ "grad_norm": 0.8422191416131971,
+ "learning_rate": 8.595588614432967e-06,
+ "loss": 0.3669420719146729,
+ "num_input_tokens_seen": 317720352,
+ "step": 7265,
+ "train_runtime": 56128.1181,
+ "train_tokens_per_second": 5660.627
+ },
+ {
+ "epoch": 0.48949636412604364,
+ "grad_norm": 0.7163712412405564,
+ "learning_rate": 8.59375077011016e-06,
+ "loss": 0.36806163787841795,
+ "num_input_tokens_seen": 317948384,
+ "step": 7270,
+ "train_runtime": 56167.6356,
+ "train_tokens_per_second": 5660.704
+ },
+ {
+ "epoch": 0.4898330191220038,
+ "grad_norm": 0.8174970324207671,
+ "learning_rate": 8.591911920806412e-06,
+ "loss": 0.39014444351196287,
+ "num_input_tokens_seen": 318179072,
+ "step": 7275,
+ "train_runtime": 56213.9579,
+ "train_tokens_per_second": 5660.144
+ },
+ {
+ "epoch": 0.4901696741179639,
+ "grad_norm": 0.6974694369606549,
+ "learning_rate": 8.590072067035947e-06,
+ "loss": 0.38969347476959226,
+ "num_input_tokens_seen": 318410480,
+ "step": 7280,
+ "train_runtime": 56255.4338,
+ "train_tokens_per_second": 5660.084
+ },
+ {
+ "epoch": 0.49050632911392406,
+ "grad_norm": 0.6692328837890084,
+ "learning_rate": 8.588231209313278e-06,
+ "loss": 0.37391541004180906,
+ "num_input_tokens_seen": 318654632,
+ "step": 7285,
+ "train_runtime": 56291.6831,
+ "train_tokens_per_second": 5660.776
+ },
+ {
+ "epoch": 0.4908429841098842,
+ "grad_norm": 0.7608072412958533,
+ "learning_rate": 8.586389348153192e-06,
+ "loss": 0.3896606206893921,
+ "num_input_tokens_seen": 318874824,
+ "step": 7290,
+ "train_runtime": 56326.4539,
+ "train_tokens_per_second": 5661.191
+ },
+ {
+ "epoch": 0.49117963910584433,
+ "grad_norm": 0.7736109340706484,
+ "learning_rate": 8.584546484070762e-06,
+ "loss": 0.40265254974365233,
+ "num_input_tokens_seen": 319084200,
+ "step": 7295,
+ "train_runtime": 56365.3177,
+ "train_tokens_per_second": 5661.002
+ },
+ {
+ "epoch": 0.4915162941018045,
+ "grad_norm": 0.7875873984707813,
+ "learning_rate": 8.582702617581338e-06,
+ "loss": 0.37858309745788576,
+ "num_input_tokens_seen": 319307784,
+ "step": 7300,
+ "train_runtime": 56403.5288,
+ "train_tokens_per_second": 5661.131
+ },
+ {
+ "epoch": 0.4918529490977646,
+ "grad_norm": 0.7183058860492505,
+ "learning_rate": 8.58085774920055e-06,
+ "loss": 0.38918490409851075,
+ "num_input_tokens_seen": 319513304,
+ "step": 7305,
+ "train_runtime": 56438.5945,
+ "train_tokens_per_second": 5661.256
+ },
+ {
+ "epoch": 0.49218960409372475,
+ "grad_norm": 0.7356534041270457,
+ "learning_rate": 8.579011879444313e-06,
+ "loss": 0.414321231842041,
+ "num_input_tokens_seen": 319743560,
+ "step": 7310,
+ "train_runtime": 56483.4424,
+ "train_tokens_per_second": 5660.837
+ },
+ {
+ "epoch": 0.4925262590896849,
+ "grad_norm": 0.6584093897897996,
+ "learning_rate": 8.577165008828817e-06,
+ "loss": 0.34825422763824465,
+ "num_input_tokens_seen": 319947112,
+ "step": 7315,
+ "train_runtime": 56521.6053,
+ "train_tokens_per_second": 5660.616
+ },
+ {
+ "epoch": 0.49286291408564503,
+ "grad_norm": 0.7722746099858304,
+ "learning_rate": 8.57531713787053e-06,
+ "loss": 0.3858429193496704,
+ "num_input_tokens_seen": 320155920,
+ "step": 7320,
+ "train_runtime": 56558.1587,
+ "train_tokens_per_second": 5660.65
+ },
+ {
+ "epoch": 0.49319956908160517,
+ "grad_norm": 0.8389191394295502,
+ "learning_rate": 8.573468267086208e-06,
+ "loss": 0.40245862007141114,
+ "num_input_tokens_seen": 320372792,
+ "step": 7325,
+ "train_runtime": 56592.9849,
+ "train_tokens_per_second": 5660.998
+ },
+ {
+ "epoch": 0.4935362240775653,
+ "grad_norm": 0.7633923302039957,
+ "learning_rate": 8.57161839699288e-06,
+ "loss": 0.36266283988952636,
+ "num_input_tokens_seen": 320592160,
+ "step": 7330,
+ "train_runtime": 56631.6072,
+ "train_tokens_per_second": 5661.011
+ },
+ {
+ "epoch": 0.49387287907352545,
+ "grad_norm": 0.8780389979845686,
+ "learning_rate": 8.569767528107857e-06,
+ "loss": 0.3856183052062988,
+ "num_input_tokens_seen": 320807512,
+ "step": 7335,
+ "train_runtime": 56669.1815,
+ "train_tokens_per_second": 5661.058
+ },
+ {
+ "epoch": 0.4942095340694856,
+ "grad_norm": 0.6800471334043537,
+ "learning_rate": 8.567915660948727e-06,
+ "loss": 0.3747032880783081,
+ "num_input_tokens_seen": 321031440,
+ "step": 7340,
+ "train_runtime": 56709.5179,
+ "train_tokens_per_second": 5660.98
+ },
+ {
+ "epoch": 0.49454618906544573,
+ "grad_norm": 0.8334581237618652,
+ "learning_rate": 8.566062796033358e-06,
+ "loss": 0.4063011646270752,
+ "num_input_tokens_seen": 321244008,
+ "step": 7345,
+ "train_runtime": 56746.898,
+ "train_tokens_per_second": 5660.997
+ },
+ {
+ "epoch": 0.49488284406140587,
+ "grad_norm": 0.7667199760691221,
+ "learning_rate": 8.564208933879903e-06,
+ "loss": 0.37364883422851564,
+ "num_input_tokens_seen": 321456992,
+ "step": 7350,
+ "train_runtime": 56787.18,
+ "train_tokens_per_second": 5660.732
+ },
+ {
+ "epoch": 0.495219499057366,
+ "grad_norm": 0.7580149996088524,
+ "learning_rate": 8.562354075006782e-06,
+ "loss": 0.3439052104949951,
+ "num_input_tokens_seen": 321675512,
+ "step": 7355,
+ "train_runtime": 56825.1931,
+ "train_tokens_per_second": 5660.79
+ },
+ {
+ "epoch": 0.49555615405332615,
+ "grad_norm": 0.6724287776480686,
+ "learning_rate": 8.560498219932706e-06,
+ "loss": 0.4257340431213379,
+ "num_input_tokens_seen": 321898448,
+ "step": 7360,
+ "train_runtime": 56859.2761,
+ "train_tokens_per_second": 5661.318
+ },
+ {
+ "epoch": 0.4958928090492863,
+ "grad_norm": 0.735779732523222,
+ "learning_rate": 8.558641369176654e-06,
+ "loss": 0.33791351318359375,
+ "num_input_tokens_seen": 322099904,
+ "step": 7365,
+ "train_runtime": 56892.2085,
+ "train_tokens_per_second": 5661.582
+ },
+ {
+ "epoch": 0.4962294640452464,
+ "grad_norm": 0.7979574200557189,
+ "learning_rate": 8.556783523257892e-06,
+ "loss": 0.3972716569900513,
+ "num_input_tokens_seen": 322310456,
+ "step": 7370,
+ "train_runtime": 56935.1565,
+ "train_tokens_per_second": 5661.009
+ },
+ {
+ "epoch": 0.49656611904120657,
+ "grad_norm": 0.7637866611420476,
+ "learning_rate": 8.554924682695959e-06,
+ "loss": 0.3530308246612549,
+ "num_input_tokens_seen": 322518744,
+ "step": 7375,
+ "train_runtime": 56971.9645,
+ "train_tokens_per_second": 5661.008
+ },
+ {
+ "epoch": 0.4969027740371667,
+ "grad_norm": 0.7640709256586813,
+ "learning_rate": 8.553064848010677e-06,
+ "loss": 0.3955545902252197,
+ "num_input_tokens_seen": 322725520,
+ "step": 7380,
+ "train_runtime": 57007.6953,
+ "train_tokens_per_second": 5661.087
+ },
+ {
+ "epoch": 0.49723942903312685,
+ "grad_norm": 0.8082539381866035,
+ "learning_rate": 8.551204019722137e-06,
+ "loss": 0.37126519680023196,
+ "num_input_tokens_seen": 322938184,
+ "step": 7385,
+ "train_runtime": 57046.2643,
+ "train_tokens_per_second": 5660.987
+ },
+ {
+ "epoch": 0.497576084029087,
+ "grad_norm": 0.7610367285948725,
+ "learning_rate": 8.549342198350718e-06,
+ "loss": 0.40738329887390134,
+ "num_input_tokens_seen": 323160704,
+ "step": 7390,
+ "train_runtime": 57086.0056,
+ "train_tokens_per_second": 5660.944
+ },
+ {
+ "epoch": 0.4979127390250471,
+ "grad_norm": 0.7004854714973048,
+ "learning_rate": 8.54747938441707e-06,
+ "loss": 0.36683225631713867,
+ "num_input_tokens_seen": 323382024,
+ "step": 7395,
+ "train_runtime": 57128.4646,
+ "train_tokens_per_second": 5660.611
+ },
+ {
+ "epoch": 0.49824939402100726,
+ "grad_norm": 0.6981945654130406,
+ "learning_rate": 8.545615578442126e-06,
+ "loss": 0.35772018432617186,
+ "num_input_tokens_seen": 323601056,
+ "step": 7400,
+ "train_runtime": 57166.4275,
+ "train_tokens_per_second": 5660.684
+ },
+ {
+ "epoch": 0.4985860490169674,
+ "grad_norm": 0.8538725530295028,
+ "learning_rate": 8.54375078094709e-06,
+ "loss": 0.3746786117553711,
+ "num_input_tokens_seen": 323824464,
+ "step": 7405,
+ "train_runtime": 57216.1649,
+ "train_tokens_per_second": 5659.667
+ },
+ {
+ "epoch": 0.49892270401292754,
+ "grad_norm": 0.6772326544260556,
+ "learning_rate": 8.541884992453449e-06,
+ "loss": 0.3711742401123047,
+ "num_input_tokens_seen": 324047472,
+ "step": 7410,
+ "train_runtime": 57254.2655,
+ "train_tokens_per_second": 5659.796
+ },
+ {
+ "epoch": 0.4992593590088877,
+ "grad_norm": 0.6789887270206421,
+ "learning_rate": 8.540018213482966e-06,
+ "loss": 0.37660138607025145,
+ "num_input_tokens_seen": 324280688,
+ "step": 7415,
+ "train_runtime": 57294.4005,
+ "train_tokens_per_second": 5659.902
+ },
+ {
+ "epoch": 0.4995960140048478,
+ "grad_norm": 0.7605591952463279,
+ "learning_rate": 8.538150444557676e-06,
+ "loss": 0.3920082330703735,
+ "num_input_tokens_seen": 324488104,
+ "step": 7420,
+ "train_runtime": 57333.8655,
+ "train_tokens_per_second": 5659.624
+ },
+ {
+ "epoch": 0.49993266900080796,
+ "grad_norm": 0.826315107413246,
+ "learning_rate": 8.536281686199896e-06,
+ "loss": 0.385677433013916,
+ "num_input_tokens_seen": 324711216,
+ "step": 7425,
+ "train_runtime": 57366.6278,
+ "train_tokens_per_second": 5660.281
+ },
+ {
+ "epoch": 0.5002693239967682,
+ "grad_norm": 0.6877893318576644,
+ "learning_rate": 8.534411938932218e-06,
+ "loss": 0.3787369728088379,
+ "num_input_tokens_seen": 324939096,
+ "step": 7430,
+ "train_runtime": 57400.8374,
+ "train_tokens_per_second": 5660.877
+ },
+ {
+ "epoch": 0.5006059789927283,
+ "grad_norm": 0.7206196203304496,
+ "learning_rate": 8.532541203277515e-06,
+ "loss": 0.3697648525238037,
+ "num_input_tokens_seen": 325163640,
+ "step": 7435,
+ "train_runtime": 57437.2769,
+ "train_tokens_per_second": 5661.195
+ },
+ {
+ "epoch": 0.5009426339886884,
+ "grad_norm": 0.8219400190681228,
+ "learning_rate": 8.530669479758926e-06,
+ "loss": 0.37598912715911864,
+ "num_input_tokens_seen": 325383192,
+ "step": 7440,
+ "train_runtime": 57475.956,
+ "train_tokens_per_second": 5661.205
+ },
+ {
+ "epoch": 0.5012792889846486,
+ "grad_norm": 0.8185284831231338,
+ "learning_rate": 8.528796768899878e-06,
+ "loss": 0.39255619049072266,
+ "num_input_tokens_seen": 325585216,
+ "step": 7445,
+ "train_runtime": 57516.4354,
+ "train_tokens_per_second": 5660.734
+ },
+ {
+ "epoch": 0.5016159439806087,
+ "grad_norm": 0.748875770223375,
+ "learning_rate": 8.526923071224064e-06,
+ "loss": 0.39842636585235597,
+ "num_input_tokens_seen": 325810472,
+ "step": 7450,
+ "train_runtime": 57549.4748,
+ "train_tokens_per_second": 5661.398
+ },
+ {
+ "epoch": 0.5019525989765689,
+ "grad_norm": 0.7148185230181737,
+ "learning_rate": 8.525048387255461e-06,
+ "loss": 0.3887434482574463,
+ "num_input_tokens_seen": 326034504,
+ "step": 7455,
+ "train_runtime": 57590.0671,
+ "train_tokens_per_second": 5661.298
+ },
+ {
+ "epoch": 0.502289253972529,
+ "grad_norm": 0.7229968516896057,
+ "learning_rate": 8.523172717518315e-06,
+ "loss": 0.3696585178375244,
+ "num_input_tokens_seen": 326245696,
+ "step": 7460,
+ "train_runtime": 57627.1988,
+ "train_tokens_per_second": 5661.314
+ },
+ {
+ "epoch": 0.5026259089684891,
+ "grad_norm": 0.7305037275102111,
+ "learning_rate": 8.521296062537156e-06,
+ "loss": 0.37876601219177247,
+ "num_input_tokens_seen": 326460104,
+ "step": 7465,
+ "train_runtime": 57664.9318,
+ "train_tokens_per_second": 5661.328
+ },
+ {
+ "epoch": 0.5029625639644493,
+ "grad_norm": 0.7847466467811594,
+ "learning_rate": 8.51941842283678e-06,
+ "loss": 0.36045174598693847,
+ "num_input_tokens_seen": 326673104,
+ "step": 7470,
+ "train_runtime": 57704.8246,
+ "train_tokens_per_second": 5661.106
+ },
+ {
+ "epoch": 0.5032992189604094,
+ "grad_norm": 0.991284631688998,
+ "learning_rate": 8.517539798942265e-06,
+ "loss": 0.3735339641571045,
+ "num_input_tokens_seen": 326906960,
+ "step": 7475,
+ "train_runtime": 57737.7344,
+ "train_tokens_per_second": 5661.929
+ },
+ {
+ "epoch": 0.5036358739563696,
+ "grad_norm": 0.6872665714663292,
+ "learning_rate": 8.515660191378962e-06,
+ "loss": 0.39676947593688966,
+ "num_input_tokens_seen": 327128416,
+ "step": 7480,
+ "train_runtime": 57776.877,
+ "train_tokens_per_second": 5661.926
+ },
+ {
+ "epoch": 0.5039725289523297,
+ "grad_norm": 0.6751937767879514,
+ "learning_rate": 8.513779600672495e-06,
+ "loss": 0.3535287380218506,
+ "num_input_tokens_seen": 327364152,
+ "step": 7485,
+ "train_runtime": 57811.2991,
+ "train_tokens_per_second": 5662.633
+ },
+ {
+ "epoch": 0.5043091839482898,
+ "grad_norm": 0.7000671921171517,
+ "learning_rate": 8.511898027348771e-06,
+ "loss": 0.37773268222808837,
+ "num_input_tokens_seen": 327568048,
+ "step": 7490,
+ "train_runtime": 57853.1955,
+ "train_tokens_per_second": 5662.056
+ },
+ {
+ "epoch": 0.50464583894425,
+ "grad_norm": 0.6936107754040254,
+ "learning_rate": 8.51001547193396e-06,
+ "loss": 0.3704256296157837,
+ "num_input_tokens_seen": 327792152,
+ "step": 7495,
+ "train_runtime": 57893.3414,
+ "train_tokens_per_second": 5662.001
+ },
+ {
+ "epoch": 0.5049824939402101,
+ "grad_norm": 0.7889020414532445,
+ "learning_rate": 8.508131934954515e-06,
+ "loss": 0.36852757930755614,
+ "num_input_tokens_seen": 327994400,
+ "step": 7500,
+ "train_runtime": 57932.8569,
+ "train_tokens_per_second": 5661.63
+ },
+ {
+ "epoch": 0.5053191489361702,
+ "grad_norm": 0.715078307452266,
+ "learning_rate": 8.50624741693716e-06,
+ "loss": 0.3951726913452148,
+ "num_input_tokens_seen": 328200448,
+ "step": 7505,
+ "train_runtime": 57971.7531,
+ "train_tokens_per_second": 5661.386
+ },
+ {
+ "epoch": 0.5056558039321304,
+ "grad_norm": 1.068005291058121,
+ "learning_rate": 8.5043619184089e-06,
+ "loss": 0.4119250297546387,
+ "num_input_tokens_seen": 328421224,
+ "step": 7510,
+ "train_runtime": 58005.869,
+ "train_tokens_per_second": 5661.862
+ },
+ {
+ "epoch": 0.5059924589280905,
+ "grad_norm": 1.638455940254133,
+ "learning_rate": 8.502475439897e-06,
+ "loss": 0.4078369140625,
+ "num_input_tokens_seen": 328646192,
+ "step": 7515,
+ "train_runtime": 58043.9094,
+ "train_tokens_per_second": 5662.027
+ },
+ {
+ "epoch": 0.5063291139240507,
+ "grad_norm": 0.7358398144623932,
+ "learning_rate": 8.500587981929013e-06,
+ "loss": 0.3898489475250244,
+ "num_input_tokens_seen": 328867064,
+ "step": 7520,
+ "train_runtime": 58088.0612,
+ "train_tokens_per_second": 5661.526
+ },
+ {
+ "epoch": 0.5066657689200108,
+ "grad_norm": 0.7487957564819914,
+ "learning_rate": 8.498699545032758e-06,
+ "loss": 0.38454551696777345,
+ "num_input_tokens_seen": 329103168,
+ "step": 7525,
+ "train_runtime": 58124.1106,
+ "train_tokens_per_second": 5662.077
+ },
+ {
+ "epoch": 0.507002423915971,
+ "grad_norm": 0.7172633060502269,
+ "learning_rate": 8.496810129736334e-06,
+ "loss": 0.3850215435028076,
+ "num_input_tokens_seen": 329322920,
+ "step": 7530,
+ "train_runtime": 58166.9838,
+ "train_tokens_per_second": 5661.681
+ },
+ {
+ "epoch": 0.5073390789119311,
+ "grad_norm": 0.7747557646585903,
+ "learning_rate": 8.494919736568105e-06,
+ "loss": 0.3788848638534546,
+ "num_input_tokens_seen": 329533896,
+ "step": 7535,
+ "train_runtime": 58210.0232,
+ "train_tokens_per_second": 5661.119
+ },
+ {
+ "epoch": 0.5076757339078912,
+ "grad_norm": 0.7933496638959117,
+ "learning_rate": 8.49302836605672e-06,
+ "loss": 0.37194099426269533,
+ "num_input_tokens_seen": 329760624,
+ "step": 7540,
+ "train_runtime": 58246.3752,
+ "train_tokens_per_second": 5661.479
+ },
+ {
+ "epoch": 0.5080123889038514,
+ "grad_norm": 0.7365707403857205,
+ "learning_rate": 8.491136018731088e-06,
+ "loss": 0.3646608114242554,
+ "num_input_tokens_seen": 329984728,
+ "step": 7545,
+ "train_runtime": 58287.6488,
+ "train_tokens_per_second": 5661.315
+ },
+ {
+ "epoch": 0.5083490438998115,
+ "grad_norm": 0.8956737633683735,
+ "learning_rate": 8.4892426951204e-06,
+ "loss": 0.3987517118453979,
+ "num_input_tokens_seen": 330206096,
+ "step": 7550,
+ "train_runtime": 58316.2366,
+ "train_tokens_per_second": 5662.335
+ },
+ {
+ "epoch": 0.5086856988957716,
+ "grad_norm": 0.7170163461314129,
+ "learning_rate": 8.48734839575412e-06,
+ "loss": 0.3717711687088013,
+ "num_input_tokens_seen": 330437336,
+ "step": 7555,
+ "train_runtime": 58351.3409,
+ "train_tokens_per_second": 5662.892
+ },
+ {
+ "epoch": 0.5090223538917318,
+ "grad_norm": 0.6630558999546488,
+ "learning_rate": 8.485453121161983e-06,
+ "loss": 0.37380030155181887,
+ "num_input_tokens_seen": 330653616,
+ "step": 7560,
+ "train_runtime": 58392.093,
+ "train_tokens_per_second": 5662.644
+ },
+ {
+ "epoch": 0.5093590088876919,
+ "grad_norm": 0.7845896930720607,
+ "learning_rate": 8.483556871873993e-06,
+ "loss": 0.42113819122314455,
+ "num_input_tokens_seen": 330875672,
+ "step": 7565,
+ "train_runtime": 58433.2419,
+ "train_tokens_per_second": 5662.456
+ },
+ {
+ "epoch": 0.5096956638836521,
+ "grad_norm": 0.760051472590646,
+ "learning_rate": 8.481659648420433e-06,
+ "loss": 0.3786670207977295,
+ "num_input_tokens_seen": 331096488,
+ "step": 7570,
+ "train_runtime": 58475.0649,
+ "train_tokens_per_second": 5662.182
+ },
+ {
+ "epoch": 0.5100323188796122,
+ "grad_norm": 0.8835596263525625,
+ "learning_rate": 8.479761451331855e-06,
+ "loss": 0.3809337615966797,
+ "num_input_tokens_seen": 331285184,
+ "step": 7575,
+ "train_runtime": 58515.9798,
+ "train_tokens_per_second": 5661.448
+ },
+ {
+ "epoch": 0.5103689738755723,
+ "grad_norm": 0.9301429501501213,
+ "learning_rate": 8.477862281139082e-06,
+ "loss": 0.3540458917617798,
+ "num_input_tokens_seen": 331501760,
+ "step": 7580,
+ "train_runtime": 58558.4957,
+ "train_tokens_per_second": 5661.036
+ },
+ {
+ "epoch": 0.5107056288715325,
+ "grad_norm": 0.7609757852463125,
+ "learning_rate": 8.475962138373212e-06,
+ "loss": 0.3986123323440552,
+ "num_input_tokens_seen": 331702152,
+ "step": 7585,
+ "train_runtime": 58591.7614,
+ "train_tokens_per_second": 5661.242
+ },
+ {
+ "epoch": 0.5110422838674926,
+ "grad_norm": 0.6930333171792868,
+ "learning_rate": 8.474061023565614e-06,
+ "loss": 0.3940688371658325,
+ "num_input_tokens_seen": 331929992,
+ "step": 7590,
+ "train_runtime": 58631.4664,
+ "train_tokens_per_second": 5661.294
+ },
+ {
+ "epoch": 0.5113789388634528,
+ "grad_norm": 0.7801001592101691,
+ "learning_rate": 8.472158937247931e-06,
+ "loss": 0.40605816841125486,
+ "num_input_tokens_seen": 332148472,
+ "step": 7595,
+ "train_runtime": 58666.3698,
+ "train_tokens_per_second": 5661.65
+ },
+ {
+ "epoch": 0.5117155938594129,
+ "grad_norm": 0.7185541728168203,
+ "learning_rate": 8.47025587995207e-06,
+ "loss": 0.38570313453674315,
+ "num_input_tokens_seen": 332362728,
+ "step": 7600,
+ "train_runtime": 58704.4102,
+ "train_tokens_per_second": 5661.631
+ },
+ {
+ "epoch": 0.512052248855373,
+ "grad_norm": 0.7887783994842774,
+ "learning_rate": 8.46835185221022e-06,
+ "loss": 0.39829018115997317,
+ "num_input_tokens_seen": 332569544,
+ "step": 7605,
+ "train_runtime": 58743.9844,
+ "train_tokens_per_second": 5661.338
+ },
+ {
+ "epoch": 0.5123889038513332,
+ "grad_norm": 0.8602264236155862,
+ "learning_rate": 8.466446854554835e-06,
+ "loss": 0.3549684762954712,
+ "num_input_tokens_seen": 332778920,
+ "step": 7610,
+ "train_runtime": 58780.8818,
+ "train_tokens_per_second": 5661.346
+ },
+ {
+ "epoch": 0.5127255588472933,
+ "grad_norm": 0.7228150824538518,
+ "learning_rate": 8.464540887518638e-06,
+ "loss": 0.37906813621520996,
+ "num_input_tokens_seen": 332984080,
+ "step": 7615,
+ "train_runtime": 58823.2164,
+ "train_tokens_per_second": 5660.759
+ },
+ {
+ "epoch": 0.5130622138432535,
+ "grad_norm": 0.7996970703617401,
+ "learning_rate": 8.462633951634631e-06,
+ "loss": 0.3685762405395508,
+ "num_input_tokens_seen": 333202640,
+ "step": 7620,
+ "train_runtime": 58868.3048,
+ "train_tokens_per_second": 5660.136
+ },
+ {
+ "epoch": 0.5133988688392136,
+ "grad_norm": 0.6901634902422549,
+ "learning_rate": 8.46072604743608e-06,
+ "loss": 0.35547614097595215,
+ "num_input_tokens_seen": 333413904,
+ "step": 7625,
+ "train_runtime": 58905.0124,
+ "train_tokens_per_second": 5660.196
+ },
+ {
+ "epoch": 0.5137355238351737,
+ "grad_norm": 0.9033857052689097,
+ "learning_rate": 8.458817175456528e-06,
+ "loss": 0.3930670261383057,
+ "num_input_tokens_seen": 333630616,
+ "step": 7630,
+ "train_runtime": 58947.1829,
+ "train_tokens_per_second": 5659.823
+ },
+ {
+ "epoch": 0.5140721788311339,
+ "grad_norm": 0.7781771666334231,
+ "learning_rate": 8.45690733622978e-06,
+ "loss": 0.35927329063415525,
+ "num_input_tokens_seen": 333854600,
+ "step": 7635,
+ "train_runtime": 58986.6913,
+ "train_tokens_per_second": 5659.829
+ },
+ {
+ "epoch": 0.514408833827094,
+ "grad_norm": 0.8455368762157274,
+ "learning_rate": 8.45499653028992e-06,
+ "loss": 0.36787919998168944,
+ "num_input_tokens_seen": 334071192,
+ "step": 7640,
+ "train_runtime": 59022.2041,
+ "train_tokens_per_second": 5660.093
+ },
+ {
+ "epoch": 0.5147454888230542,
+ "grad_norm": 0.7586644939540588,
+ "learning_rate": 8.453084758171297e-06,
+ "loss": 0.37957746982574464,
+ "num_input_tokens_seen": 334291888,
+ "step": 7645,
+ "train_runtime": 59067.1738,
+ "train_tokens_per_second": 5659.521
+ },
+ {
+ "epoch": 0.5150821438190143,
+ "grad_norm": 0.7524816129003911,
+ "learning_rate": 8.451172020408532e-06,
+ "loss": 0.38945965766906737,
+ "num_input_tokens_seen": 334516424,
+ "step": 7650,
+ "train_runtime": 59101.459,
+ "train_tokens_per_second": 5660.037
+ },
+ {
+ "epoch": 0.5154187988149744,
+ "grad_norm": 0.7137520182282876,
+ "learning_rate": 8.449258317536517e-06,
+ "loss": 0.35926315784454343,
+ "num_input_tokens_seen": 334732808,
+ "step": 7655,
+ "train_runtime": 59134.7468,
+ "train_tokens_per_second": 5660.51
+ },
+ {
+ "epoch": 0.5157554538109346,
+ "grad_norm": 0.7129464795573726,
+ "learning_rate": 8.447343650090412e-06,
+ "loss": 0.35725347995758056,
+ "num_input_tokens_seen": 334954280,
+ "step": 7660,
+ "train_runtime": 59172.9555,
+ "train_tokens_per_second": 5660.597
+ },
+ {
+ "epoch": 0.5160921088068947,
+ "grad_norm": 0.7272448686794939,
+ "learning_rate": 8.445428018605647e-06,
+ "loss": 0.406490421295166,
+ "num_input_tokens_seen": 335178568,
+ "step": 7665,
+ "train_runtime": 59210.2513,
+ "train_tokens_per_second": 5660.82
+ },
+ {
+ "epoch": 0.5164287638028549,
+ "grad_norm": 0.6593329295977397,
+ "learning_rate": 8.443511423617924e-06,
+ "loss": 0.34410724639892576,
+ "num_input_tokens_seen": 335398608,
+ "step": 7670,
+ "train_runtime": 59246.6136,
+ "train_tokens_per_second": 5661.06
+ },
+ {
+ "epoch": 0.516765418798815,
+ "grad_norm": 0.7054753575271266,
+ "learning_rate": 8.441593865663213e-06,
+ "loss": 0.39874303340911865,
+ "num_input_tokens_seen": 335621696,
+ "step": 7675,
+ "train_runtime": 59285.0684,
+ "train_tokens_per_second": 5661.151
+ },
+ {
+ "epoch": 0.5171020737947751,
+ "grad_norm": 0.7429064696844442,
+ "learning_rate": 8.439675345277751e-06,
+ "loss": 0.40140485763549805,
+ "num_input_tokens_seen": 335839504,
+ "step": 7680,
+ "train_runtime": 59327.3379,
+ "train_tokens_per_second": 5660.788
+ },
+ {
+ "epoch": 0.5174387287907353,
+ "grad_norm": 0.6392657750321254,
+ "learning_rate": 8.437755862998047e-06,
+ "loss": 0.3619039297103882,
+ "num_input_tokens_seen": 336062080,
+ "step": 7685,
+ "train_runtime": 59362.2516,
+ "train_tokens_per_second": 5661.208
+ },
+ {
+ "epoch": 0.5177753837866954,
+ "grad_norm": 0.6333778064119364,
+ "learning_rate": 8.435835419360878e-06,
+ "loss": 0.3882591247558594,
+ "num_input_tokens_seen": 336299944,
+ "step": 7690,
+ "train_runtime": 59398.7224,
+ "train_tokens_per_second": 5661.737
+ },
+ {
+ "epoch": 0.5181120387826555,
+ "grad_norm": 0.7397126752113882,
+ "learning_rate": 8.433914014903291e-06,
+ "loss": 0.42388019561767576,
+ "num_input_tokens_seen": 336514968,
+ "step": 7695,
+ "train_runtime": 59437.773,
+ "train_tokens_per_second": 5661.635
+ },
+ {
+ "epoch": 0.5184486937786157,
+ "grad_norm": 0.6888095905405301,
+ "learning_rate": 8.431991650162599e-06,
+ "loss": 0.37606000900268555,
+ "num_input_tokens_seen": 336732600,
+ "step": 7700,
+ "train_runtime": 59483.4411,
+ "train_tokens_per_second": 5660.947
+ },
+ {
+ "epoch": 0.5187853487745758,
+ "grad_norm": 0.6980617402046243,
+ "learning_rate": 8.430068325676386e-06,
+ "loss": 0.39625885486602785,
+ "num_input_tokens_seen": 336948776,
+ "step": 7705,
+ "train_runtime": 59521.0447,
+ "train_tokens_per_second": 5661.002
+ },
+ {
+ "epoch": 0.519122003770536,
+ "grad_norm": 0.732968462897192,
+ "learning_rate": 8.428144041982502e-06,
+ "loss": 0.39479496479034426,
+ "num_input_tokens_seen": 337163520,
+ "step": 7710,
+ "train_runtime": 59556.4,
+ "train_tokens_per_second": 5661.247
+ },
+ {
+ "epoch": 0.5194586587664961,
+ "grad_norm": 0.6401198443952045,
+ "learning_rate": 8.42621879961907e-06,
+ "loss": 0.38160171508789065,
+ "num_input_tokens_seen": 337382992,
+ "step": 7715,
+ "train_runtime": 59597.2808,
+ "train_tokens_per_second": 5661.047
+ },
+ {
+ "epoch": 0.5197953137624562,
+ "grad_norm": 0.7011673418789361,
+ "learning_rate": 8.424292599124476e-06,
+ "loss": 0.3532627820968628,
+ "num_input_tokens_seen": 337609304,
+ "step": 7720,
+ "train_runtime": 59636.4965,
+ "train_tokens_per_second": 5661.119
+ },
+ {
+ "epoch": 0.5201319687584164,
+ "grad_norm": 0.7609717187854771,
+ "learning_rate": 8.422365441037374e-06,
+ "loss": 0.39478967189788816,
+ "num_input_tokens_seen": 337838072,
+ "step": 7725,
+ "train_runtime": 59669.1056,
+ "train_tokens_per_second": 5661.859
+ },
+ {
+ "epoch": 0.5204686237543765,
+ "grad_norm": 0.743645872276741,
+ "learning_rate": 8.420437325896692e-06,
+ "loss": 0.3906223297119141,
+ "num_input_tokens_seen": 338067512,
+ "step": 7730,
+ "train_runtime": 59705.1498,
+ "train_tokens_per_second": 5662.284
+ },
+ {
+ "epoch": 0.5208052787503367,
+ "grad_norm": 0.7586775310938808,
+ "learning_rate": 8.418508254241617e-06,
+ "loss": 0.3925325870513916,
+ "num_input_tokens_seen": 338284504,
+ "step": 7735,
+ "train_runtime": 59748.0729,
+ "train_tokens_per_second": 5661.848
+ },
+ {
+ "epoch": 0.5211419337462968,
+ "grad_norm": 0.8223238925961407,
+ "learning_rate": 8.416578226611612e-06,
+ "loss": 0.35642185211181643,
+ "num_input_tokens_seen": 338495440,
+ "step": 7740,
+ "train_runtime": 59783.2802,
+ "train_tokens_per_second": 5662.042
+ },
+ {
+ "epoch": 0.5214785887422569,
+ "grad_norm": 0.8532576047124709,
+ "learning_rate": 8.414647243546398e-06,
+ "loss": 0.38446238040924074,
+ "num_input_tokens_seen": 338717192,
+ "step": 7745,
+ "train_runtime": 59821.3869,
+ "train_tokens_per_second": 5662.142
+ },
+ {
+ "epoch": 0.5218152437382171,
+ "grad_norm": 0.6899241307861492,
+ "learning_rate": 8.412715305585973e-06,
+ "loss": 0.367059588432312,
+ "num_input_tokens_seen": 338946888,
+ "step": 7750,
+ "train_runtime": 59859.9792,
+ "train_tokens_per_second": 5662.329
+ },
+ {
+ "epoch": 0.5221518987341772,
+ "grad_norm": 0.7311441327867719,
+ "learning_rate": 8.410782413270594e-06,
+ "loss": 0.3799006462097168,
+ "num_input_tokens_seen": 339172224,
+ "step": 7755,
+ "train_runtime": 59900.1807,
+ "train_tokens_per_second": 5662.29
+ },
+ {
+ "epoch": 0.5224885537301374,
+ "grad_norm": 0.9420308727064326,
+ "learning_rate": 8.408848567140787e-06,
+ "loss": 0.37287948131561277,
+ "num_input_tokens_seen": 339364672,
+ "step": 7760,
+ "train_runtime": 59941.0763,
+ "train_tokens_per_second": 5661.638
+ },
+ {
+ "epoch": 0.5228252087260975,
+ "grad_norm": 0.7245874800766633,
+ "learning_rate": 8.406913767737353e-06,
+ "loss": 0.37642457485198977,
+ "num_input_tokens_seen": 339579064,
+ "step": 7765,
+ "train_runtime": 59982.2891,
+ "train_tokens_per_second": 5661.322
+ },
+ {
+ "epoch": 0.5231618637220576,
+ "grad_norm": 0.7199997285512201,
+ "learning_rate": 8.404978015601344e-06,
+ "loss": 0.3964075565338135,
+ "num_input_tokens_seen": 339801840,
+ "step": 7770,
+ "train_runtime": 60029.6628,
+ "train_tokens_per_second": 5660.566
+ },
+ {
+ "epoch": 0.5234985187180178,
+ "grad_norm": 0.697918705663631,
+ "learning_rate": 8.403041311274091e-06,
+ "loss": 0.39160215854644775,
+ "num_input_tokens_seen": 340025832,
+ "step": 7775,
+ "train_runtime": 60071.2747,
+ "train_tokens_per_second": 5660.373
+ },
+ {
+ "epoch": 0.5238351737139779,
+ "grad_norm": 0.8157700767931442,
+ "learning_rate": 8.401103655297188e-06,
+ "loss": 0.3895624876022339,
+ "num_input_tokens_seen": 340239616,
+ "step": 7780,
+ "train_runtime": 60105.9113,
+ "train_tokens_per_second": 5660.668
+ },
+ {
+ "epoch": 0.5241718287099381,
+ "grad_norm": 0.838923642880136,
+ "learning_rate": 8.399165048212494e-06,
+ "loss": 0.3589164257049561,
+ "num_input_tokens_seen": 340454984,
+ "step": 7785,
+ "train_runtime": 60145.3789,
+ "train_tokens_per_second": 5660.534
+ },
+ {
+ "epoch": 0.5245084837058982,
+ "grad_norm": 0.7246920315479125,
+ "learning_rate": 8.397225490562131e-06,
+ "loss": 0.4032485008239746,
+ "num_input_tokens_seen": 340684288,
+ "step": 7790,
+ "train_runtime": 60180.0663,
+ "train_tokens_per_second": 5661.082
+ },
+ {
+ "epoch": 0.5248451387018583,
+ "grad_norm": 0.7826576263190983,
+ "learning_rate": 8.395284982888494e-06,
+ "loss": 0.3588679313659668,
+ "num_input_tokens_seen": 340904512,
+ "step": 7795,
+ "train_runtime": 60216.628,
+ "train_tokens_per_second": 5661.302
+ },
+ {
+ "epoch": 0.5251817936978185,
+ "grad_norm": 0.7381573756482819,
+ "learning_rate": 8.393343525734239e-06,
+ "loss": 0.39098711013793946,
+ "num_input_tokens_seen": 341128688,
+ "step": 7800,
+ "train_runtime": 60254.8697,
+ "train_tokens_per_second": 5661.429
+ },
+ {
+ "epoch": 0.5255184486937786,
+ "grad_norm": 0.9510438503863269,
+ "learning_rate": 8.391401119642286e-06,
+ "loss": 0.39932818412780763,
+ "num_input_tokens_seen": 341326256,
+ "step": 7805,
+ "train_runtime": 60302.5843,
+ "train_tokens_per_second": 5660.226
+ },
+ {
+ "epoch": 0.5258551036897388,
+ "grad_norm": 0.7551888581389101,
+ "learning_rate": 8.389457765155824e-06,
+ "loss": 0.40273380279541016,
+ "num_input_tokens_seen": 341526840,
+ "step": 7810,
+ "train_runtime": 60347.108,
+ "train_tokens_per_second": 5659.374
+ },
+ {
+ "epoch": 0.5261917586856989,
+ "grad_norm": 0.7817537228902953,
+ "learning_rate": 8.387513462818309e-06,
+ "loss": 0.36842942237854004,
+ "num_input_tokens_seen": 341747720,
+ "step": 7815,
+ "train_runtime": 60388.0509,
+ "train_tokens_per_second": 5659.194
+ },
+ {
+ "epoch": 0.526528413681659,
+ "grad_norm": 0.76396158135225,
+ "learning_rate": 8.385568213173453e-06,
+ "loss": 0.3687218427658081,
+ "num_input_tokens_seen": 341961736,
+ "step": 7820,
+ "train_runtime": 60430.695,
+ "train_tokens_per_second": 5658.742
+ },
+ {
+ "epoch": 0.5268650686776192,
+ "grad_norm": 1.0538784566401151,
+ "learning_rate": 8.383622016765242e-06,
+ "loss": 0.3595340013504028,
+ "num_input_tokens_seen": 342180688,
+ "step": 7825,
+ "train_runtime": 60474.2774,
+ "train_tokens_per_second": 5658.285
+ },
+ {
+ "epoch": 0.5272017236735793,
+ "grad_norm": 0.7683842178996437,
+ "learning_rate": 8.381674874137926e-06,
+ "loss": 0.4014129638671875,
+ "num_input_tokens_seen": 342390504,
+ "step": 7830,
+ "train_runtime": 60507.7742,
+ "train_tokens_per_second": 5658.62
+ },
+ {
+ "epoch": 0.5275383786695395,
+ "grad_norm": 0.7767541075637882,
+ "learning_rate": 8.379726785836013e-06,
+ "loss": 0.3654167652130127,
+ "num_input_tokens_seen": 342602080,
+ "step": 7835,
+ "train_runtime": 60550.4658,
+ "train_tokens_per_second": 5658.125
+ },
+ {
+ "epoch": 0.5278750336654996,
+ "grad_norm": 0.6734458895156976,
+ "learning_rate": 8.377777752404283e-06,
+ "loss": 0.36977086067199705,
+ "num_input_tokens_seen": 342831800,
+ "step": 7840,
+ "train_runtime": 60591.1256,
+ "train_tokens_per_second": 5658.119
+ },
+ {
+ "epoch": 0.5282116886614597,
+ "grad_norm": 0.8168160522078569,
+ "learning_rate": 8.375827774387774e-06,
+ "loss": 0.34728164672851564,
+ "num_input_tokens_seen": 343029568,
+ "step": 7845,
+ "train_runtime": 60633.6042,
+ "train_tokens_per_second": 5657.417
+ },
+ {
+ "epoch": 0.5285483436574199,
+ "grad_norm": 0.6622532156086101,
+ "learning_rate": 8.373876852331793e-06,
+ "loss": 0.3438518762588501,
+ "num_input_tokens_seen": 343254472,
+ "step": 7850,
+ "train_runtime": 60671.9643,
+ "train_tokens_per_second": 5657.547
+ },
+ {
+ "epoch": 0.52888499865338,
+ "grad_norm": 0.7502437558347258,
+ "learning_rate": 8.371924986781909e-06,
+ "loss": 0.38301682472229004,
+ "num_input_tokens_seen": 343470008,
+ "step": 7855,
+ "train_runtime": 60709.3414,
+ "train_tokens_per_second": 5657.614
+ },
+ {
+ "epoch": 0.5292216536493402,
+ "grad_norm": 0.7587114327145728,
+ "learning_rate": 8.369972178283955e-06,
+ "loss": 0.36738762855529783,
+ "num_input_tokens_seen": 343698072,
+ "step": 7860,
+ "train_runtime": 60747.59,
+ "train_tokens_per_second": 5657.806
+ },
+ {
+ "epoch": 0.5295583086453003,
+ "grad_norm": 0.6881092460250557,
+ "learning_rate": 8.368018427384027e-06,
+ "loss": 0.38115482330322265,
+ "num_input_tokens_seen": 343919960,
+ "step": 7865,
+ "train_runtime": 60791.0097,
+ "train_tokens_per_second": 5657.415
+ },
+ {
+ "epoch": 0.5298949636412604,
+ "grad_norm": 0.8355237541845086,
+ "learning_rate": 8.366063734628485e-06,
+ "loss": 0.38156135082244874,
+ "num_input_tokens_seen": 344123568,
+ "step": 7870,
+ "train_runtime": 60827.6415,
+ "train_tokens_per_second": 5657.355
+ },
+ {
+ "epoch": 0.5302316186372206,
+ "grad_norm": 0.8303838221096392,
+ "learning_rate": 8.364108100563955e-06,
+ "loss": 0.4007874011993408,
+ "num_input_tokens_seen": 344341784,
+ "step": 7875,
+ "train_runtime": 60864.4296,
+ "train_tokens_per_second": 5657.521
+ },
+ {
+ "epoch": 0.5305682736331807,
+ "grad_norm": 0.7925795609225155,
+ "learning_rate": 8.362151525737321e-06,
+ "loss": 0.4212652206420898,
+ "num_input_tokens_seen": 344562552,
+ "step": 7880,
+ "train_runtime": 60901.6398,
+ "train_tokens_per_second": 5657.689
+ },
+ {
+ "epoch": 0.5309049286291408,
+ "grad_norm": 0.7468959848481416,
+ "learning_rate": 8.360194010695733e-06,
+ "loss": 0.3570136547088623,
+ "num_input_tokens_seen": 344782336,
+ "step": 7885,
+ "train_runtime": 60942.4135,
+ "train_tokens_per_second": 5657.51
+ },
+ {
+ "epoch": 0.531241583625101,
+ "grad_norm": 0.7105479533553207,
+ "learning_rate": 8.358235555986607e-06,
+ "loss": 0.39661598205566406,
+ "num_input_tokens_seen": 345008672,
+ "step": 7890,
+ "train_runtime": 60976.8653,
+ "train_tokens_per_second": 5658.026
+ },
+ {
+ "epoch": 0.5315782386210611,
+ "grad_norm": 0.768825896681359,
+ "learning_rate": 8.356276162157613e-06,
+ "loss": 0.3597188234329224,
+ "num_input_tokens_seen": 345216360,
+ "step": 7895,
+ "train_runtime": 61021.9355,
+ "train_tokens_per_second": 5657.25
+ },
+ {
+ "epoch": 0.5319148936170213,
+ "grad_norm": 0.711887791776799,
+ "learning_rate": 8.354315829756696e-06,
+ "loss": 0.37782256603240966,
+ "num_input_tokens_seen": 345442968,
+ "step": 7900,
+ "train_runtime": 61056.926,
+ "train_tokens_per_second": 5657.72
+ },
+ {
+ "epoch": 0.5322515486129814,
+ "grad_norm": 0.7271488313267995,
+ "learning_rate": 8.352354559332053e-06,
+ "loss": 0.36695542335510256,
+ "num_input_tokens_seen": 345663248,
+ "step": 7905,
+ "train_runtime": 61093.5998,
+ "train_tokens_per_second": 5657.929
+ },
+ {
+ "epoch": 0.5325882036089415,
+ "grad_norm": 0.7790624119649752,
+ "learning_rate": 8.350392351432146e-06,
+ "loss": 0.3877138376235962,
+ "num_input_tokens_seen": 345877640,
+ "step": 7910,
+ "train_runtime": 61125.7433,
+ "train_tokens_per_second": 5658.461
+ },
+ {
+ "epoch": 0.5329248586049017,
+ "grad_norm": 0.652480198172122,
+ "learning_rate": 8.348429206605702e-06,
+ "loss": 0.3623414754867554,
+ "num_input_tokens_seen": 346113080,
+ "step": 7915,
+ "train_runtime": 61170.1897,
+ "train_tokens_per_second": 5658.199
+ },
+ {
+ "epoch": 0.5332615136008618,
+ "grad_norm": 0.8166602948279773,
+ "learning_rate": 8.346465125401706e-06,
+ "loss": 0.41824941635131835,
+ "num_input_tokens_seen": 346331632,
+ "step": 7920,
+ "train_runtime": 61213.7632,
+ "train_tokens_per_second": 5657.741
+ },
+ {
+ "epoch": 0.533598168596822,
+ "grad_norm": 0.7338720699869766,
+ "learning_rate": 8.344500108369412e-06,
+ "loss": 0.3689335107803345,
+ "num_input_tokens_seen": 346540104,
+ "step": 7925,
+ "train_runtime": 61254.7218,
+ "train_tokens_per_second": 5657.361
+ },
+ {
+ "epoch": 0.5339348235927821,
+ "grad_norm": 0.7241659753104186,
+ "learning_rate": 8.342534156058326e-06,
+ "loss": 0.3848552942276001,
+ "num_input_tokens_seen": 346759048,
+ "step": 7930,
+ "train_runtime": 61293.0183,
+ "train_tokens_per_second": 5657.399
+ },
+ {
+ "epoch": 0.5342714785887422,
+ "grad_norm": 0.7284900345851684,
+ "learning_rate": 8.34056726901822e-06,
+ "loss": 0.39329664707183837,
+ "num_input_tokens_seen": 346971992,
+ "step": 7935,
+ "train_runtime": 61333.9961,
+ "train_tokens_per_second": 5657.091
+ },
+ {
+ "epoch": 0.5346081335847024,
+ "grad_norm": 0.9343840115025148,
+ "learning_rate": 8.338599447799132e-06,
+ "loss": 0.37961349487304685,
+ "num_input_tokens_seen": 347176784,
+ "step": 7940,
+ "train_runtime": 61367.1004,
+ "train_tokens_per_second": 5657.376
+ },
+ {
+ "epoch": 0.5349447885806625,
+ "grad_norm": 0.7650504516817562,
+ "learning_rate": 8.336630692951353e-06,
+ "loss": 0.3706023931503296,
+ "num_input_tokens_seen": 347396336,
+ "step": 7945,
+ "train_runtime": 61405.4086,
+ "train_tokens_per_second": 5657.422
+ },
+ {
+ "epoch": 0.5352814435766227,
+ "grad_norm": 0.7809182592661493,
+ "learning_rate": 8.334661005025441e-06,
+ "loss": 0.3996375560760498,
+ "num_input_tokens_seen": 347612392,
+ "step": 7950,
+ "train_runtime": 61443.353,
+ "train_tokens_per_second": 5657.445
+ },
+ {
+ "epoch": 0.5356180985725828,
+ "grad_norm": 1.0173782664960012,
+ "learning_rate": 8.332690384572213e-06,
+ "loss": 0.4039102077484131,
+ "num_input_tokens_seen": 347827352,
+ "step": 7955,
+ "train_runtime": 61481.7721,
+ "train_tokens_per_second": 5657.406
+ },
+ {
+ "epoch": 0.5359547535685429,
+ "grad_norm": 0.7319934278286511,
+ "learning_rate": 8.330718832142746e-06,
+ "loss": 0.3981315612792969,
+ "num_input_tokens_seen": 348060432,
+ "step": 7960,
+ "train_runtime": 61520.2514,
+ "train_tokens_per_second": 5657.656
+ },
+ {
+ "epoch": 0.5362914085645031,
+ "grad_norm": 0.6935185521555359,
+ "learning_rate": 8.328746348288379e-06,
+ "loss": 0.3804043769836426,
+ "num_input_tokens_seen": 348266616,
+ "step": 7965,
+ "train_runtime": 61562.5177,
+ "train_tokens_per_second": 5657.121
+ },
+ {
+ "epoch": 0.5366280635604632,
+ "grad_norm": 0.7663090177174399,
+ "learning_rate": 8.32677293356071e-06,
+ "loss": 0.42568235397338866,
+ "num_input_tokens_seen": 348498272,
+ "step": 7970,
+ "train_runtime": 61598.1572,
+ "train_tokens_per_second": 5657.609
+ },
+ {
+ "epoch": 0.5369647185564234,
+ "grad_norm": 0.8375018048980476,
+ "learning_rate": 8.324798588511598e-06,
+ "loss": 0.3702836275100708,
+ "num_input_tokens_seen": 348696416,
+ "step": 7975,
+ "train_runtime": 61635.4226,
+ "train_tokens_per_second": 5657.403
+ },
+ {
+ "epoch": 0.5373013735523835,
+ "grad_norm": 0.8187637529093583,
+ "learning_rate": 8.322823313693162e-06,
+ "loss": 0.38841800689697265,
+ "num_input_tokens_seen": 348910592,
+ "step": 7980,
+ "train_runtime": 61671.4891,
+ "train_tokens_per_second": 5657.567
+ },
+ {
+ "epoch": 0.5376380285483436,
+ "grad_norm": 0.7918621668094143,
+ "learning_rate": 8.320847109657782e-06,
+ "loss": 0.3783905506134033,
+ "num_input_tokens_seen": 349120256,
+ "step": 7985,
+ "train_runtime": 61712.9559,
+ "train_tokens_per_second": 5657.163
+ },
+ {
+ "epoch": 0.5379746835443038,
+ "grad_norm": 0.7362861473446518,
+ "learning_rate": 8.318869976958099e-06,
+ "loss": 0.3449576377868652,
+ "num_input_tokens_seen": 349340688,
+ "step": 7990,
+ "train_runtime": 61755.9567,
+ "train_tokens_per_second": 5656.793
+ },
+ {
+ "epoch": 0.5383113385402639,
+ "grad_norm": 0.7351144206496397,
+ "learning_rate": 8.316891916147007e-06,
+ "loss": 0.40518803596496583,
+ "num_input_tokens_seen": 349557408,
+ "step": 7995,
+ "train_runtime": 61795.1163,
+ "train_tokens_per_second": 5656.716
+ },
+ {
+ "epoch": 0.5386479935362241,
+ "grad_norm": 0.8155814621488839,
+ "learning_rate": 8.314912927777671e-06,
+ "loss": 0.3914914608001709,
+ "num_input_tokens_seen": 349776936,
+ "step": 8000,
+ "train_runtime": 61829.5854,
+ "train_tokens_per_second": 5657.113
+ },
+ {
+ "epoch": 0.5389846485321842,
+ "grad_norm": 0.7399640627157749,
+ "learning_rate": 8.312933012403503e-06,
+ "loss": 0.3830781698226929,
+ "num_input_tokens_seen": 349981944,
+ "step": 8005,
+ "train_runtime": 61870.1471,
+ "train_tokens_per_second": 5656.718
+ },
+ {
+ "epoch": 0.5393213035281443,
+ "grad_norm": 0.6837705266722391,
+ "learning_rate": 8.310952170578184e-06,
+ "loss": 0.3917949676513672,
+ "num_input_tokens_seen": 350196560,
+ "step": 8010,
+ "train_runtime": 61910.9608,
+ "train_tokens_per_second": 5656.455
+ },
+ {
+ "epoch": 0.5396579585241045,
+ "grad_norm": 0.8108035687999815,
+ "learning_rate": 8.308970402855647e-06,
+ "loss": 0.35423126220703127,
+ "num_input_tokens_seen": 350407320,
+ "step": 8015,
+ "train_runtime": 61942.9362,
+ "train_tokens_per_second": 5656.938
+ },
+ {
+ "epoch": 0.5399946135200646,
+ "grad_norm": 0.7391585249859578,
+ "learning_rate": 8.306987709790087e-06,
+ "loss": 0.36304192543029784,
+ "num_input_tokens_seen": 350621920,
+ "step": 8020,
+ "train_runtime": 61974.5184,
+ "train_tokens_per_second": 5657.517
+ },
+ {
+ "epoch": 0.5403312685160248,
+ "grad_norm": 0.8297400381885977,
+ "learning_rate": 8.305004091935962e-06,
+ "loss": 0.3827538013458252,
+ "num_input_tokens_seen": 350840936,
+ "step": 8025,
+ "train_runtime": 62010.6476,
+ "train_tokens_per_second": 5657.753
+ },
+ {
+ "epoch": 0.5406679235119849,
+ "grad_norm": 0.8790917583658322,
+ "learning_rate": 8.303019549847979e-06,
+ "loss": 0.40552120208740233,
+ "num_input_tokens_seen": 351057864,
+ "step": 8030,
+ "train_runtime": 62051.8481,
+ "train_tokens_per_second": 5657.492
+ },
+ {
+ "epoch": 0.541004578507945,
+ "grad_norm": 0.7794541942972262,
+ "learning_rate": 8.301034084081114e-06,
+ "loss": 0.38002500534057615,
+ "num_input_tokens_seen": 351285720,
+ "step": 8035,
+ "train_runtime": 62090.6336,
+ "train_tokens_per_second": 5657.628
+ },
+ {
+ "epoch": 0.5413412335039052,
+ "grad_norm": 0.8524170296510677,
+ "learning_rate": 8.299047695190592e-06,
+ "loss": 0.408983850479126,
+ "num_input_tokens_seen": 351491576,
+ "step": 8040,
+ "train_runtime": 62127.0664,
+ "train_tokens_per_second": 5657.624
+ },
+ {
+ "epoch": 0.5416778884998653,
+ "grad_norm": 0.8835316549178089,
+ "learning_rate": 8.297060383731903e-06,
+ "loss": 0.38953261375427245,
+ "num_input_tokens_seen": 351706376,
+ "step": 8045,
+ "train_runtime": 62162.9599,
+ "train_tokens_per_second": 5657.813
+ },
+ {
+ "epoch": 0.5420145434958255,
+ "grad_norm": 0.8130215721187897,
+ "learning_rate": 8.29507215026079e-06,
+ "loss": 0.3989473819732666,
+ "num_input_tokens_seen": 351925296,
+ "step": 8050,
+ "train_runtime": 62194.3013,
+ "train_tokens_per_second": 5658.481
+ },
+ {
+ "epoch": 0.5423511984917856,
+ "grad_norm": 0.7698865553206388,
+ "learning_rate": 8.293082995333257e-06,
+ "loss": 0.3805055618286133,
+ "num_input_tokens_seen": 352152728,
+ "step": 8055,
+ "train_runtime": 62235.2959,
+ "train_tokens_per_second": 5658.409
+ },
+ {
+ "epoch": 0.5426878534877457,
+ "grad_norm": 0.7925089143300325,
+ "learning_rate": 8.291092919505568e-06,
+ "loss": 0.359727144241333,
+ "num_input_tokens_seen": 352368040,
+ "step": 8060,
+ "train_runtime": 62272.5239,
+ "train_tokens_per_second": 5658.483
+ },
+ {
+ "epoch": 0.5430245084837059,
+ "grad_norm": 0.7473619474168591,
+ "learning_rate": 8.289101923334237e-06,
+ "loss": 0.3927903175354004,
+ "num_input_tokens_seen": 352592456,
+ "step": 8065,
+ "train_runtime": 62314.1115,
+ "train_tokens_per_second": 5658.308
+ },
+ {
+ "epoch": 0.543361163479666,
+ "grad_norm": 0.8359023777365873,
+ "learning_rate": 8.28711000737604e-06,
+ "loss": 0.37551441192626955,
+ "num_input_tokens_seen": 352801680,
+ "step": 8070,
+ "train_runtime": 62352.2218,
+ "train_tokens_per_second": 5658.205
+ },
+ {
+ "epoch": 0.5436978184756261,
+ "grad_norm": 0.8077808908921092,
+ "learning_rate": 8.285117172188013e-06,
+ "loss": 0.3938951253890991,
+ "num_input_tokens_seen": 353025760,
+ "step": 8075,
+ "train_runtime": 62386.0037,
+ "train_tokens_per_second": 5658.733
+ },
+ {
+ "epoch": 0.5440344734715863,
+ "grad_norm": 0.8228182913825491,
+ "learning_rate": 8.283123418327443e-06,
+ "loss": 0.4004195213317871,
+ "num_input_tokens_seen": 353244352,
+ "step": 8080,
+ "train_runtime": 62419.1441,
+ "train_tokens_per_second": 5659.231
+ },
+ {
+ "epoch": 0.5443711284675464,
+ "grad_norm": 0.693682105218099,
+ "learning_rate": 8.281128746351878e-06,
+ "loss": 0.3595161437988281,
+ "num_input_tokens_seen": 353475168,
+ "step": 8085,
+ "train_runtime": 62453.3189,
+ "train_tokens_per_second": 5659.83
+ },
+ {
+ "epoch": 0.5447077834635066,
+ "grad_norm": 0.7187659246186631,
+ "learning_rate": 8.279133156819123e-06,
+ "loss": 0.40300402641296384,
+ "num_input_tokens_seen": 353691088,
+ "step": 8090,
+ "train_runtime": 62496.7209,
+ "train_tokens_per_second": 5659.354
+ },
+ {
+ "epoch": 0.5450444384594667,
+ "grad_norm": 0.725512834705303,
+ "learning_rate": 8.277136650287237e-06,
+ "loss": 0.3694662570953369,
+ "num_input_tokens_seen": 353917984,
+ "step": 8095,
+ "train_runtime": 62536.3526,
+ "train_tokens_per_second": 5659.396
+ },
+ {
+ "epoch": 0.5453810934554268,
+ "grad_norm": 0.8301428582075541,
+ "learning_rate": 8.275139227314537e-06,
+ "loss": 0.3757363557815552,
+ "num_input_tokens_seen": 354125952,
+ "step": 8100,
+ "train_runtime": 62575.4142,
+ "train_tokens_per_second": 5659.187
+ },
+ {
+ "epoch": 0.545717748451387,
+ "grad_norm": 0.7999915594881919,
+ "learning_rate": 8.273140888459594e-06,
+ "loss": 0.4242837905883789,
+ "num_input_tokens_seen": 354358280,
+ "step": 8105,
+ "train_runtime": 62606.4679,
+ "train_tokens_per_second": 5660.091
+ },
+ {
+ "epoch": 0.5460544034473471,
+ "grad_norm": 0.9175431453565677,
+ "learning_rate": 8.27114163428124e-06,
+ "loss": 0.3934792518615723,
+ "num_input_tokens_seen": 354575664,
+ "step": 8110,
+ "train_runtime": 62642.1464,
+ "train_tokens_per_second": 5660.337
+ },
+ {
+ "epoch": 0.5463910584433073,
+ "grad_norm": 0.7408300214764466,
+ "learning_rate": 8.26914146533856e-06,
+ "loss": 0.38090806007385253,
+ "num_input_tokens_seen": 354808384,
+ "step": 8115,
+ "train_runtime": 62681.8186,
+ "train_tokens_per_second": 5660.467
+ },
+ {
+ "epoch": 0.5467277134392674,
+ "grad_norm": 0.7033405910036602,
+ "learning_rate": 8.267140382190892e-06,
+ "loss": 0.3828258037567139,
+ "num_input_tokens_seen": 355026976,
+ "step": 8120,
+ "train_runtime": 62720.5013,
+ "train_tokens_per_second": 5660.461
+ },
+ {
+ "epoch": 0.5470643684352275,
+ "grad_norm": 0.6635248689946966,
+ "learning_rate": 8.265138385397835e-06,
+ "loss": 0.3777566432952881,
+ "num_input_tokens_seen": 355244064,
+ "step": 8125,
+ "train_runtime": 62758.0924,
+ "train_tokens_per_second": 5660.53
+ },
+ {
+ "epoch": 0.5474010234311877,
+ "grad_norm": 0.6184317548675163,
+ "learning_rate": 8.263135475519241e-06,
+ "loss": 0.34886605739593507,
+ "num_input_tokens_seen": 355483272,
+ "step": 8130,
+ "train_runtime": 62794.3565,
+ "train_tokens_per_second": 5661.07
+ },
+ {
+ "epoch": 0.5477376784271478,
+ "grad_norm": 0.7287895081916465,
+ "learning_rate": 8.261131653115216e-06,
+ "loss": 0.42696261405944824,
+ "num_input_tokens_seen": 355711240,
+ "step": 8135,
+ "train_runtime": 62834.333,
+ "train_tokens_per_second": 5661.097
+ },
+ {
+ "epoch": 0.548074333423108,
+ "grad_norm": 0.7794435545394348,
+ "learning_rate": 8.259126918746122e-06,
+ "loss": 0.36356501579284667,
+ "num_input_tokens_seen": 355930576,
+ "step": 8140,
+ "train_runtime": 62868.698,
+ "train_tokens_per_second": 5661.491
+ },
+ {
+ "epoch": 0.5484109884190681,
+ "grad_norm": 0.6736849599162793,
+ "learning_rate": 8.25712127297258e-06,
+ "loss": 0.3778836250305176,
+ "num_input_tokens_seen": 356150816,
+ "step": 8145,
+ "train_runtime": 62904.5449,
+ "train_tokens_per_second": 5661.766
+ },
+ {
+ "epoch": 0.5487476434150282,
+ "grad_norm": 0.7082862793695406,
+ "learning_rate": 8.255114716355459e-06,
+ "loss": 0.4094264507293701,
+ "num_input_tokens_seen": 356382088,
+ "step": 8150,
+ "train_runtime": 62942.0397,
+ "train_tokens_per_second": 5662.068
+ },
+ {
+ "epoch": 0.5490842984109884,
+ "grad_norm": 0.6051005207909932,
+ "learning_rate": 8.253107249455887e-06,
+ "loss": 0.3917287826538086,
+ "num_input_tokens_seen": 356602352,
+ "step": 8155,
+ "train_runtime": 62975.3252,
+ "train_tokens_per_second": 5662.573
+ },
+ {
+ "epoch": 0.5494209534069485,
+ "grad_norm": 0.7713641151309204,
+ "learning_rate": 8.251098872835247e-06,
+ "loss": 0.39372997283935546,
+ "num_input_tokens_seen": 356825280,
+ "step": 8160,
+ "train_runtime": 63017.9451,
+ "train_tokens_per_second": 5662.28
+ },
+ {
+ "epoch": 0.5497576084029087,
+ "grad_norm": 0.6787537110984104,
+ "learning_rate": 8.249089587055175e-06,
+ "loss": 0.3526461601257324,
+ "num_input_tokens_seen": 357037664,
+ "step": 8165,
+ "train_runtime": 63059.5973,
+ "train_tokens_per_second": 5661.908
+ },
+ {
+ "epoch": 0.5500942633988688,
+ "grad_norm": 0.6748882087262432,
+ "learning_rate": 8.247079392677558e-06,
+ "loss": 0.3740200996398926,
+ "num_input_tokens_seen": 357257424,
+ "step": 8170,
+ "train_runtime": 63101.2168,
+ "train_tokens_per_second": 5661.657
+ },
+ {
+ "epoch": 0.5504309183948289,
+ "grad_norm": 0.7274044025423265,
+ "learning_rate": 8.245068290264544e-06,
+ "loss": 0.3697060108184814,
+ "num_input_tokens_seen": 357480144,
+ "step": 8175,
+ "train_runtime": 63140.8867,
+ "train_tokens_per_second": 5661.627
+ },
+ {
+ "epoch": 0.5507675733907891,
+ "grad_norm": 0.7084601443387462,
+ "learning_rate": 8.243056280378531e-06,
+ "loss": 0.37436251640319823,
+ "num_input_tokens_seen": 357699472,
+ "step": 8180,
+ "train_runtime": 63181.8065,
+ "train_tokens_per_second": 5661.432
+ },
+ {
+ "epoch": 0.5511042283867492,
+ "grad_norm": 0.7481357333059493,
+ "learning_rate": 8.241043363582171e-06,
+ "loss": 0.36437788009643557,
+ "num_input_tokens_seen": 357926808,
+ "step": 8185,
+ "train_runtime": 63218.2393,
+ "train_tokens_per_second": 5661.765
+ },
+ {
+ "epoch": 0.5514408833827094,
+ "grad_norm": 1.0989196799559033,
+ "learning_rate": 8.239029540438369e-06,
+ "loss": 0.3505790948867798,
+ "num_input_tokens_seen": 358145064,
+ "step": 8190,
+ "train_runtime": 63257.3639,
+ "train_tokens_per_second": 5661.713
+ },
+ {
+ "epoch": 0.5517775383786695,
+ "grad_norm": 0.7305407817056846,
+ "learning_rate": 8.237014811510284e-06,
+ "loss": 0.3656752109527588,
+ "num_input_tokens_seen": 358369824,
+ "step": 8195,
+ "train_runtime": 63304.7711,
+ "train_tokens_per_second": 5661.024
+ },
+ {
+ "epoch": 0.5521141933746296,
+ "grad_norm": 0.6472079948654392,
+ "learning_rate": 8.234999177361329e-06,
+ "loss": 0.3714550018310547,
+ "num_input_tokens_seen": 358596784,
+ "step": 8200,
+ "train_runtime": 63340.7687,
+ "train_tokens_per_second": 5661.39
+ },
+ {
+ "epoch": 0.5524508483705898,
+ "grad_norm": 0.7398046092203671,
+ "learning_rate": 8.232982638555169e-06,
+ "loss": 0.40174274444580077,
+ "num_input_tokens_seen": 358820312,
+ "step": 8205,
+ "train_runtime": 63372.9424,
+ "train_tokens_per_second": 5662.043
+ },
+ {
+ "epoch": 0.5527875033665499,
+ "grad_norm": 0.7062238114257713,
+ "learning_rate": 8.230965195655724e-06,
+ "loss": 0.39043726921081545,
+ "num_input_tokens_seen": 359055496,
+ "step": 8210,
+ "train_runtime": 63412.4919,
+ "train_tokens_per_second": 5662.22
+ },
+ {
+ "epoch": 0.55312415836251,
+ "grad_norm": 0.8164987537109517,
+ "learning_rate": 8.228946849227167e-06,
+ "loss": 0.3990142345428467,
+ "num_input_tokens_seen": 359287376,
+ "step": 8215,
+ "train_runtime": 63449.3085,
+ "train_tokens_per_second": 5662.589
+ },
+ {
+ "epoch": 0.5534608133584702,
+ "grad_norm": 0.6938755136640083,
+ "learning_rate": 8.226927599833916e-06,
+ "loss": 0.386507511138916,
+ "num_input_tokens_seen": 359514576,
+ "step": 8220,
+ "train_runtime": 63481.4052,
+ "train_tokens_per_second": 5663.305
+ },
+ {
+ "epoch": 0.5537974683544303,
+ "grad_norm": 0.6819170611796698,
+ "learning_rate": 8.224907448040655e-06,
+ "loss": 0.37993340492248534,
+ "num_input_tokens_seen": 359745160,
+ "step": 8225,
+ "train_runtime": 63513.0016,
+ "train_tokens_per_second": 5664.118
+ },
+ {
+ "epoch": 0.5541341233503905,
+ "grad_norm": 0.7464665475450891,
+ "learning_rate": 8.222886394412309e-06,
+ "loss": 0.4058523178100586,
+ "num_input_tokens_seen": 359965008,
+ "step": 8230,
+ "train_runtime": 63553.6566,
+ "train_tokens_per_second": 5663.954
+ },
+ {
+ "epoch": 0.5544707783463506,
+ "grad_norm": 0.6722302273522195,
+ "learning_rate": 8.220864439514057e-06,
+ "loss": 0.384306526184082,
+ "num_input_tokens_seen": 360194040,
+ "step": 8235,
+ "train_runtime": 63592.1988,
+ "train_tokens_per_second": 5664.123
+ },
+ {
+ "epoch": 0.5548074333423108,
+ "grad_norm": 0.6641141478245969,
+ "learning_rate": 8.218841583911338e-06,
+ "loss": 0.3680997848510742,
+ "num_input_tokens_seen": 360409248,
+ "step": 8240,
+ "train_runtime": 63634.667,
+ "train_tokens_per_second": 5663.725
+ },
+ {
+ "epoch": 0.5551440883382709,
+ "grad_norm": 0.7528778151335748,
+ "learning_rate": 8.216817828169834e-06,
+ "loss": 0.413088321685791,
+ "num_input_tokens_seen": 360624480,
+ "step": 8245,
+ "train_runtime": 63670.5285,
+ "train_tokens_per_second": 5663.915
+ },
+ {
+ "epoch": 0.555480743334231,
+ "grad_norm": 0.8364309542706772,
+ "learning_rate": 8.214793172855481e-06,
+ "loss": 0.36465370655059814,
+ "num_input_tokens_seen": 360837048,
+ "step": 8250,
+ "train_runtime": 63711.1159,
+ "train_tokens_per_second": 5663.644
+ },
+ {
+ "epoch": 0.5558173983301912,
+ "grad_norm": 0.6793083765205596,
+ "learning_rate": 8.21276761853447e-06,
+ "loss": 0.36249728202819825,
+ "num_input_tokens_seen": 361060272,
+ "step": 8255,
+ "train_runtime": 63754.6776,
+ "train_tokens_per_second": 5663.275
+ },
+ {
+ "epoch": 0.5561540533261513,
+ "grad_norm": 0.8280046723618791,
+ "learning_rate": 8.210741165773238e-06,
+ "loss": 0.39040985107421877,
+ "num_input_tokens_seen": 361294192,
+ "step": 8260,
+ "train_runtime": 63785.4799,
+ "train_tokens_per_second": 5664.207
+ },
+ {
+ "epoch": 0.5564907083221114,
+ "grad_norm": 0.7577410939409384,
+ "learning_rate": 8.208713815138476e-06,
+ "loss": 0.36567642688751223,
+ "num_input_tokens_seen": 361524240,
+ "step": 8265,
+ "train_runtime": 63826.3741,
+ "train_tokens_per_second": 5664.183
+ },
+ {
+ "epoch": 0.5568273633180716,
+ "grad_norm": 0.6700074247940968,
+ "learning_rate": 8.20668556719713e-06,
+ "loss": 0.35714569091796877,
+ "num_input_tokens_seen": 361732392,
+ "step": 8270,
+ "train_runtime": 63860.1654,
+ "train_tokens_per_second": 5664.445
+ },
+ {
+ "epoch": 0.5571640183140317,
+ "grad_norm": 0.7602580503776288,
+ "learning_rate": 8.204656422516388e-06,
+ "loss": 0.3372737646102905,
+ "num_input_tokens_seen": 361946008,
+ "step": 8275,
+ "train_runtime": 63901.2504,
+ "train_tokens_per_second": 5664.146
+ },
+ {
+ "epoch": 0.5575006733099919,
+ "grad_norm": 0.8510184000147942,
+ "learning_rate": 8.2026263816637e-06,
+ "loss": 0.38366384506225587,
+ "num_input_tokens_seen": 362152248,
+ "step": 8280,
+ "train_runtime": 63938.3717,
+ "train_tokens_per_second": 5664.083
+ },
+ {
+ "epoch": 0.557837328305952,
+ "grad_norm": 0.7613111925807169,
+ "learning_rate": 8.200595445206753e-06,
+ "loss": 0.37342875003814696,
+ "num_input_tokens_seen": 362371096,
+ "step": 8285,
+ "train_runtime": 63979.792,
+ "train_tokens_per_second": 5663.837
+ },
+ {
+ "epoch": 0.5581739833019121,
+ "grad_norm": 1.256734854809266,
+ "learning_rate": 8.198563613713497e-06,
+ "loss": 0.37401297092437746,
+ "num_input_tokens_seen": 362579576,
+ "step": 8290,
+ "train_runtime": 64020.2056,
+ "train_tokens_per_second": 5663.518
+ },
+ {
+ "epoch": 0.5585106382978723,
+ "grad_norm": 0.7385210193529426,
+ "learning_rate": 8.196530887752127e-06,
+ "loss": 0.3654287338256836,
+ "num_input_tokens_seen": 362803720,
+ "step": 8295,
+ "train_runtime": 64054.0322,
+ "train_tokens_per_second": 5664.026
+ },
+ {
+ "epoch": 0.5588472932938324,
+ "grad_norm": 0.8102882370091832,
+ "learning_rate": 8.194497267891086e-06,
+ "loss": 0.37125024795532224,
+ "num_input_tokens_seen": 363027432,
+ "step": 8300,
+ "train_runtime": 64092.3585,
+ "train_tokens_per_second": 5664.13
+ },
+ {
+ "epoch": 0.5591839482897926,
+ "grad_norm": 0.7432402409233543,
+ "learning_rate": 8.19246275469907e-06,
+ "loss": 0.3796513795852661,
+ "num_input_tokens_seen": 363254432,
+ "step": 8305,
+ "train_runtime": 64136.3098,
+ "train_tokens_per_second": 5663.788
+ },
+ {
+ "epoch": 0.5595206032857528,
+ "grad_norm": 0.6849455308836975,
+ "learning_rate": 8.190427348745027e-06,
+ "loss": 0.36992511749267576,
+ "num_input_tokens_seen": 363479976,
+ "step": 8310,
+ "train_runtime": 64171.4185,
+ "train_tokens_per_second": 5664.204
+ },
+ {
+ "epoch": 0.559857258281713,
+ "grad_norm": 0.6998379838638995,
+ "learning_rate": 8.188391050598146e-06,
+ "loss": 0.3632178544998169,
+ "num_input_tokens_seen": 363712504,
+ "step": 8315,
+ "train_runtime": 64216.5044,
+ "train_tokens_per_second": 5663.848
+ },
+ {
+ "epoch": 0.5601939132776731,
+ "grad_norm": 0.7737605990707936,
+ "learning_rate": 8.186353860827878e-06,
+ "loss": 0.37043955326080324,
+ "num_input_tokens_seen": 363919784,
+ "step": 8320,
+ "train_runtime": 64256.872,
+ "train_tokens_per_second": 5663.515
+ },
+ {
+ "epoch": 0.5605305682736332,
+ "grad_norm": 0.6957700417795362,
+ "learning_rate": 8.184315780003911e-06,
+ "loss": 0.3709959745407104,
+ "num_input_tokens_seen": 364140736,
+ "step": 8325,
+ "train_runtime": 64302.6454,
+ "train_tokens_per_second": 5662.92
+ },
+ {
+ "epoch": 0.5608672232695934,
+ "grad_norm": 0.64141155335251,
+ "learning_rate": 8.182276808696193e-06,
+ "loss": 0.40601282119750975,
+ "num_input_tokens_seen": 364348280,
+ "step": 8330,
+ "train_runtime": 64347.1042,
+ "train_tokens_per_second": 5662.233
+ },
+ {
+ "epoch": 0.5612038782655535,
+ "grad_norm": 0.6158998839242338,
+ "learning_rate": 8.18023694747491e-06,
+ "loss": 0.3827899694442749,
+ "num_input_tokens_seen": 364574968,
+ "step": 8335,
+ "train_runtime": 64381.872,
+ "train_tokens_per_second": 5662.696
+ },
+ {
+ "epoch": 0.5615405332615137,
+ "grad_norm": 0.7883261214385181,
+ "learning_rate": 8.178196196910508e-06,
+ "loss": 0.3656409740447998,
+ "num_input_tokens_seen": 364803736,
+ "step": 8340,
+ "train_runtime": 64421.7397,
+ "train_tokens_per_second": 5662.743
+ },
+ {
+ "epoch": 0.5618771882574738,
+ "grad_norm": 0.7366395620437122,
+ "learning_rate": 8.176154557573672e-06,
+ "loss": 0.4000543594360352,
+ "num_input_tokens_seen": 365037240,
+ "step": 8345,
+ "train_runtime": 64459.3416,
+ "train_tokens_per_second": 5663.062
+ },
+ {
+ "epoch": 0.5622138432534339,
+ "grad_norm": 0.6985009938577207,
+ "learning_rate": 8.174112030035344e-06,
+ "loss": 0.3629467010498047,
+ "num_input_tokens_seen": 365268200,
+ "step": 8350,
+ "train_runtime": 64493.078,
+ "train_tokens_per_second": 5663.681
+ },
+ {
+ "epoch": 0.5625504982493941,
+ "grad_norm": 0.7274783538197038,
+ "learning_rate": 8.172068614866707e-06,
+ "loss": 0.36340205669403075,
+ "num_input_tokens_seen": 365477184,
+ "step": 8355,
+ "train_runtime": 64531.8006,
+ "train_tokens_per_second": 5663.521
+ },
+ {
+ "epoch": 0.5628871532453542,
+ "grad_norm": 0.7368665647506253,
+ "learning_rate": 8.170024312639197e-06,
+ "loss": 0.39545679092407227,
+ "num_input_tokens_seen": 365707680,
+ "step": 8360,
+ "train_runtime": 64570.9604,
+ "train_tokens_per_second": 5663.656
+ },
+ {
+ "epoch": 0.5632238082413143,
+ "grad_norm": 0.7340645583889084,
+ "learning_rate": 8.167979123924496e-06,
+ "loss": 0.3859426498413086,
+ "num_input_tokens_seen": 365930048,
+ "step": 8365,
+ "train_runtime": 64604.6699,
+ "train_tokens_per_second": 5664.142
+ },
+ {
+ "epoch": 0.5635604632372745,
+ "grad_norm": 0.7462104536621238,
+ "learning_rate": 8.165933049294537e-06,
+ "loss": 0.4087996482849121,
+ "num_input_tokens_seen": 366147352,
+ "step": 8370,
+ "train_runtime": 64646.3443,
+ "train_tokens_per_second": 5663.852
+ },
+ {
+ "epoch": 0.5638971182332346,
+ "grad_norm": 0.7582565637830365,
+ "learning_rate": 8.163886089321493e-06,
+ "loss": 0.37345223426818847,
+ "num_input_tokens_seen": 366377160,
+ "step": 8375,
+ "train_runtime": 64688.6158,
+ "train_tokens_per_second": 5663.704
+ },
+ {
+ "epoch": 0.5642337732291948,
+ "grad_norm": 0.6581884199846296,
+ "learning_rate": 8.161838244577795e-06,
+ "loss": 0.3573953151702881,
+ "num_input_tokens_seen": 366596264,
+ "step": 8380,
+ "train_runtime": 64726.1749,
+ "train_tokens_per_second": 5663.802
+ },
+ {
+ "epoch": 0.5645704282251549,
+ "grad_norm": 0.7540331450420041,
+ "learning_rate": 8.159789515636114e-06,
+ "loss": 0.3621669769287109,
+ "num_input_tokens_seen": 366814672,
+ "step": 8385,
+ "train_runtime": 64768.6142,
+ "train_tokens_per_second": 5663.463
+ },
+ {
+ "epoch": 0.564907083221115,
+ "grad_norm": 0.7846236965599341,
+ "learning_rate": 8.15773990306937e-06,
+ "loss": 0.39387807846069334,
+ "num_input_tokens_seen": 367010840,
+ "step": 8390,
+ "train_runtime": 64803.6605,
+ "train_tokens_per_second": 5663.428
+ },
+ {
+ "epoch": 0.5652437382170752,
+ "grad_norm": 0.8105957793723372,
+ "learning_rate": 8.155689407450732e-06,
+ "loss": 0.37447366714477537,
+ "num_input_tokens_seen": 367235664,
+ "step": 8395,
+ "train_runtime": 64840.3814,
+ "train_tokens_per_second": 5663.688
+ },
+ {
+ "epoch": 0.5655803932130353,
+ "grad_norm": 0.7593251086886306,
+ "learning_rate": 8.153638029353615e-06,
+ "loss": 0.3721097707748413,
+ "num_input_tokens_seen": 367439472,
+ "step": 8400,
+ "train_runtime": 64880.3333,
+ "train_tokens_per_second": 5663.341
+ },
+ {
+ "epoch": 0.5659170482089955,
+ "grad_norm": 0.7080065499934168,
+ "learning_rate": 8.151585769351678e-06,
+ "loss": 0.416213321685791,
+ "num_input_tokens_seen": 367665000,
+ "step": 8405,
+ "train_runtime": 64915.8566,
+ "train_tokens_per_second": 5663.716
+ },
+ {
+ "epoch": 0.5662537032049556,
+ "grad_norm": 0.8290526825274906,
+ "learning_rate": 8.149532628018833e-06,
+ "loss": 0.3906120538711548,
+ "num_input_tokens_seen": 367873544,
+ "step": 8410,
+ "train_runtime": 64954.7801,
+ "train_tokens_per_second": 5663.533
+ },
+ {
+ "epoch": 0.5665903582009157,
+ "grad_norm": 0.7887226131302195,
+ "learning_rate": 8.147478605929228e-06,
+ "loss": 0.36734464168548586,
+ "num_input_tokens_seen": 368106424,
+ "step": 8415,
+ "train_runtime": 64988.7811,
+ "train_tokens_per_second": 5664.153
+ },
+ {
+ "epoch": 0.5669270131968759,
+ "grad_norm": 0.7105615187211392,
+ "learning_rate": 8.145423703657269e-06,
+ "loss": 0.3696259021759033,
+ "num_input_tokens_seen": 368329816,
+ "step": 8420,
+ "train_runtime": 65035.2437,
+ "train_tokens_per_second": 5663.542
+ },
+ {
+ "epoch": 0.567263668192836,
+ "grad_norm": 0.7106675383557719,
+ "learning_rate": 8.143367921777601e-06,
+ "loss": 0.3901944637298584,
+ "num_input_tokens_seen": 368556080,
+ "step": 8425,
+ "train_runtime": 65073.1284,
+ "train_tokens_per_second": 5663.722
+ },
+ {
+ "epoch": 0.5676003231887962,
+ "grad_norm": 0.7158037922920018,
+ "learning_rate": 8.141311260865117e-06,
+ "loss": 0.42763843536376955,
+ "num_input_tokens_seen": 368786216,
+ "step": 8430,
+ "train_runtime": 65107.0411,
+ "train_tokens_per_second": 5664.306
+ },
+ {
+ "epoch": 0.5679369781847563,
+ "grad_norm": 0.7147505710629092,
+ "learning_rate": 8.139253721494956e-06,
+ "loss": 0.4124480724334717,
+ "num_input_tokens_seen": 368998320,
+ "step": 8435,
+ "train_runtime": 65143.6774,
+ "train_tokens_per_second": 5664.377
+ },
+ {
+ "epoch": 0.5682736331807164,
+ "grad_norm": 0.690454346302624,
+ "learning_rate": 8.137195304242502e-06,
+ "loss": 0.3694611549377441,
+ "num_input_tokens_seen": 369221808,
+ "step": 8440,
+ "train_runtime": 65180.1258,
+ "train_tokens_per_second": 5664.638
+ },
+ {
+ "epoch": 0.5686102881766766,
+ "grad_norm": 0.6647768706585024,
+ "learning_rate": 8.135136009683387e-06,
+ "loss": 0.3659370899200439,
+ "num_input_tokens_seen": 369439952,
+ "step": 8445,
+ "train_runtime": 65214.1207,
+ "train_tokens_per_second": 5665.03
+ },
+ {
+ "epoch": 0.5689469431726367,
+ "grad_norm": 0.7123577020124194,
+ "learning_rate": 8.13307583839348e-06,
+ "loss": 0.37821600437164304,
+ "num_input_tokens_seen": 369665384,
+ "step": 8450,
+ "train_runtime": 65253.9634,
+ "train_tokens_per_second": 5665.026
+ },
+ {
+ "epoch": 0.5692835981685969,
+ "grad_norm": 0.7080847091486382,
+ "learning_rate": 8.131014790948908e-06,
+ "loss": 0.3932258367538452,
+ "num_input_tokens_seen": 369885352,
+ "step": 8455,
+ "train_runtime": 65292.5756,
+ "train_tokens_per_second": 5665.045
+ },
+ {
+ "epoch": 0.569620253164557,
+ "grad_norm": 0.8175401785926597,
+ "learning_rate": 8.128952867926035e-06,
+ "loss": 0.42728986740112307,
+ "num_input_tokens_seen": 370102168,
+ "step": 8460,
+ "train_runtime": 65326.5467,
+ "train_tokens_per_second": 5665.418
+ },
+ {
+ "epoch": 0.5699569081605171,
+ "grad_norm": 0.7636022480938197,
+ "learning_rate": 8.126890069901468e-06,
+ "loss": 0.3966690540313721,
+ "num_input_tokens_seen": 370323328,
+ "step": 8465,
+ "train_runtime": 65365.5452,
+ "train_tokens_per_second": 5665.421
+ },
+ {
+ "epoch": 0.5702935631564773,
+ "grad_norm": 0.8306888968108936,
+ "learning_rate": 8.124826397452064e-06,
+ "loss": 0.3729779005050659,
+ "num_input_tokens_seen": 370551792,
+ "step": 8470,
+ "train_runtime": 65403.3941,
+ "train_tokens_per_second": 5665.636
+ },
+ {
+ "epoch": 0.5706302181524374,
+ "grad_norm": 0.6975418197600278,
+ "learning_rate": 8.122761851154923e-06,
+ "loss": 0.37224202156066893,
+ "num_input_tokens_seen": 370769584,
+ "step": 8475,
+ "train_runtime": 65437.1957,
+ "train_tokens_per_second": 5666.037
+ },
+ {
+ "epoch": 0.5709668731483976,
+ "grad_norm": 0.7691847068239698,
+ "learning_rate": 8.120696431587387e-06,
+ "loss": 0.3793643474578857,
+ "num_input_tokens_seen": 370986856,
+ "step": 8480,
+ "train_runtime": 65477.4237,
+ "train_tokens_per_second": 5665.874
+ },
+ {
+ "epoch": 0.5713035281443577,
+ "grad_norm": 0.653136501814857,
+ "learning_rate": 8.118630139327046e-06,
+ "loss": 0.36144418716430665,
+ "num_input_tokens_seen": 371203736,
+ "step": 8485,
+ "train_runtime": 65520.6447,
+ "train_tokens_per_second": 5665.447
+ },
+ {
+ "epoch": 0.5716401831403178,
+ "grad_norm": 0.6688068417437787,
+ "learning_rate": 8.116562974951731e-06,
+ "loss": 0.3969851016998291,
+ "num_input_tokens_seen": 371433168,
+ "step": 8490,
+ "train_runtime": 65562.0835,
+ "train_tokens_per_second": 5665.366
+ },
+ {
+ "epoch": 0.571976838136278,
+ "grad_norm": 0.6693856135821081,
+ "learning_rate": 8.11449493903952e-06,
+ "loss": 0.37904407978057864,
+ "num_input_tokens_seen": 371661928,
+ "step": 8495,
+ "train_runtime": 65600.9804,
+ "train_tokens_per_second": 5665.494
+ },
+ {
+ "epoch": 0.5723134931322381,
+ "grad_norm": 0.6989787293355911,
+ "learning_rate": 8.112426032168727e-06,
+ "loss": 0.39436898231506345,
+ "num_input_tokens_seen": 371877904,
+ "step": 8500,
+ "train_runtime": 65639.9628,
+ "train_tokens_per_second": 5665.419
+ },
+ {
+ "epoch": 0.5726501481281983,
+ "grad_norm": 0.6921966447452655,
+ "learning_rate": 8.11035625491792e-06,
+ "loss": 0.35653374195098875,
+ "num_input_tokens_seen": 372083344,
+ "step": 8505,
+ "train_runtime": 65673.4132,
+ "train_tokens_per_second": 5665.662
+ },
+ {
+ "epoch": 0.5729868031241584,
+ "grad_norm": 0.7863535213377899,
+ "learning_rate": 8.108285607865907e-06,
+ "loss": 0.3748607158660889,
+ "num_input_tokens_seen": 372296008,
+ "step": 8510,
+ "train_runtime": 65709.8413,
+ "train_tokens_per_second": 5665.757
+ },
+ {
+ "epoch": 0.5733234581201185,
+ "grad_norm": 0.7103469377120155,
+ "learning_rate": 8.106214091591731e-06,
+ "loss": 0.36768136024475095,
+ "num_input_tokens_seen": 372517376,
+ "step": 8515,
+ "train_runtime": 65747.0301,
+ "train_tokens_per_second": 5665.919
+ },
+ {
+ "epoch": 0.5736601131160787,
+ "grad_norm": 0.8043097859632866,
+ "learning_rate": 8.104141706674693e-06,
+ "loss": 0.3505872249603271,
+ "num_input_tokens_seen": 372744072,
+ "step": 8520,
+ "train_runtime": 65786.4844,
+ "train_tokens_per_second": 5665.967
+ },
+ {
+ "epoch": 0.5739967681120388,
+ "grad_norm": 0.7899542715564848,
+ "learning_rate": 8.102068453694324e-06,
+ "loss": 0.39815869331359866,
+ "num_input_tokens_seen": 372965496,
+ "step": 8525,
+ "train_runtime": 65827.1156,
+ "train_tokens_per_second": 5665.834
+ },
+ {
+ "epoch": 0.574333423107999,
+ "grad_norm": 0.7007321773126729,
+ "learning_rate": 8.099994333230404e-06,
+ "loss": 0.39165263175964354,
+ "num_input_tokens_seen": 373176000,
+ "step": 8530,
+ "train_runtime": 65869.8446,
+ "train_tokens_per_second": 5665.354
+ },
+ {
+ "epoch": 0.5746700781039591,
+ "grad_norm": 0.8059453918533225,
+ "learning_rate": 8.097919345862952e-06,
+ "loss": 0.387595534324646,
+ "num_input_tokens_seen": 373375616,
+ "step": 8535,
+ "train_runtime": 65901.6062,
+ "train_tokens_per_second": 5665.653
+ },
+ {
+ "epoch": 0.5750067330999192,
+ "grad_norm": 0.7748964359890478,
+ "learning_rate": 8.095843492172235e-06,
+ "loss": 0.3876967906951904,
+ "num_input_tokens_seen": 373591328,
+ "step": 8540,
+ "train_runtime": 65941.7008,
+ "train_tokens_per_second": 5665.479
+ },
+ {
+ "epoch": 0.5753433880958794,
+ "grad_norm": 0.773534060871464,
+ "learning_rate": 8.093766772738758e-06,
+ "loss": 0.3301082611083984,
+ "num_input_tokens_seen": 373805712,
+ "step": 8545,
+ "train_runtime": 65984.3171,
+ "train_tokens_per_second": 5665.069
+ },
+ {
+ "epoch": 0.5756800430918395,
+ "grad_norm": 0.7746351168910522,
+ "learning_rate": 8.091689188143267e-06,
+ "loss": 0.3910914659500122,
+ "num_input_tokens_seen": 374017824,
+ "step": 8550,
+ "train_runtime": 66030.3105,
+ "train_tokens_per_second": 5664.335
+ },
+ {
+ "epoch": 0.5760166980877997,
+ "grad_norm": 0.7771670247365485,
+ "learning_rate": 8.089610738966754e-06,
+ "loss": 0.3839992046356201,
+ "num_input_tokens_seen": 374221088,
+ "step": 8555,
+ "train_runtime": 66067.2485,
+ "train_tokens_per_second": 5664.245
+ },
+ {
+ "epoch": 0.5763533530837598,
+ "grad_norm": 0.8314883209793272,
+ "learning_rate": 8.08753142579045e-06,
+ "loss": 0.41312179565429685,
+ "num_input_tokens_seen": 374438784,
+ "step": 8560,
+ "train_runtime": 66103.0863,
+ "train_tokens_per_second": 5664.468
+ },
+ {
+ "epoch": 0.5766900080797199,
+ "grad_norm": 0.7070329659969831,
+ "learning_rate": 8.085451249195828e-06,
+ "loss": 0.369646692276001,
+ "num_input_tokens_seen": 374674400,
+ "step": 8565,
+ "train_runtime": 66144.5198,
+ "train_tokens_per_second": 5664.481
+ },
+ {
+ "epoch": 0.5770266630756801,
+ "grad_norm": 0.728101829102978,
+ "learning_rate": 8.083370209764603e-06,
+ "loss": 0.38304858207702636,
+ "num_input_tokens_seen": 374895048,
+ "step": 8570,
+ "train_runtime": 66187.4516,
+ "train_tokens_per_second": 5664.141
+ },
+ {
+ "epoch": 0.5773633180716402,
+ "grad_norm": 0.7613212980235913,
+ "learning_rate": 8.081288308078731e-06,
+ "loss": 0.39395363330841066,
+ "num_input_tokens_seen": 375130432,
+ "step": 8575,
+ "train_runtime": 66220.5986,
+ "train_tokens_per_second": 5664.86
+ },
+ {
+ "epoch": 0.5776999730676003,
+ "grad_norm": 0.6846318822730251,
+ "learning_rate": 8.07920554472041e-06,
+ "loss": 0.35710582733154295,
+ "num_input_tokens_seen": 375356416,
+ "step": 8580,
+ "train_runtime": 66266.8733,
+ "train_tokens_per_second": 5664.315
+ },
+ {
+ "epoch": 0.5780366280635605,
+ "grad_norm": 0.791144672174043,
+ "learning_rate": 8.077121920272075e-06,
+ "loss": 0.3807799816131592,
+ "num_input_tokens_seen": 375586768,
+ "step": 8585,
+ "train_runtime": 66303.4383,
+ "train_tokens_per_second": 5664.665
+ },
+ {
+ "epoch": 0.5783732830595206,
+ "grad_norm": 0.6749521518269885,
+ "learning_rate": 8.07503743531641e-06,
+ "loss": 0.3493949890136719,
+ "num_input_tokens_seen": 375796816,
+ "step": 8590,
+ "train_runtime": 66347.6746,
+ "train_tokens_per_second": 5664.054
+ },
+ {
+ "epoch": 0.5787099380554808,
+ "grad_norm": 0.7233256642537537,
+ "learning_rate": 8.072952090436328e-06,
+ "loss": 0.34188032150268555,
+ "num_input_tokens_seen": 376025856,
+ "step": 8595,
+ "train_runtime": 66379.0077,
+ "train_tokens_per_second": 5664.831
+ },
+ {
+ "epoch": 0.5790465930514409,
+ "grad_norm": 0.8183957573318358,
+ "learning_rate": 8.070865886214996e-06,
+ "loss": 0.378603458404541,
+ "num_input_tokens_seen": 376248416,
+ "step": 8600,
+ "train_runtime": 66412.9945,
+ "train_tokens_per_second": 5665.283
+ },
+ {
+ "epoch": 0.579383248047401,
+ "grad_norm": 0.8303736324901154,
+ "learning_rate": 8.06877882323581e-06,
+ "loss": 0.3776103496551514,
+ "num_input_tokens_seen": 376468952,
+ "step": 8605,
+ "train_runtime": 66445.0637,
+ "train_tokens_per_second": 5665.868
+ },
+ {
+ "epoch": 0.5797199030433612,
+ "grad_norm": 0.6628257697173621,
+ "learning_rate": 8.066690902082413e-06,
+ "loss": 0.35461578369140623,
+ "num_input_tokens_seen": 376690720,
+ "step": 8610,
+ "train_runtime": 66482.3424,
+ "train_tokens_per_second": 5666.027
+ },
+ {
+ "epoch": 0.5800565580393213,
+ "grad_norm": 0.6914119662414026,
+ "learning_rate": 8.064602123338683e-06,
+ "loss": 0.4189821720123291,
+ "num_input_tokens_seen": 376926952,
+ "step": 8615,
+ "train_runtime": 66523.5883,
+ "train_tokens_per_second": 5666.065
+ },
+ {
+ "epoch": 0.5803932130352815,
+ "grad_norm": 0.7255952843828343,
+ "learning_rate": 8.06251248758874e-06,
+ "loss": 0.3706305980682373,
+ "num_input_tokens_seen": 377137424,
+ "step": 8620,
+ "train_runtime": 66560.9464,
+ "train_tokens_per_second": 5666.047
+ },
+ {
+ "epoch": 0.5807298680312416,
+ "grad_norm": 0.7888860114164491,
+ "learning_rate": 8.060421995416949e-06,
+ "loss": 0.4162487030029297,
+ "num_input_tokens_seen": 377341792,
+ "step": 8625,
+ "train_runtime": 66601.9228,
+ "train_tokens_per_second": 5665.629
+ },
+ {
+ "epoch": 0.5810665230272017,
+ "grad_norm": 0.8097744521376653,
+ "learning_rate": 8.058330647407904e-06,
+ "loss": 0.3742199420928955,
+ "num_input_tokens_seen": 377553480,
+ "step": 8630,
+ "train_runtime": 66640.7232,
+ "train_tokens_per_second": 5665.507
+ },
+ {
+ "epoch": 0.5814031780231619,
+ "grad_norm": 0.9311380768738048,
+ "learning_rate": 8.056238444146445e-06,
+ "loss": 0.4032620429992676,
+ "num_input_tokens_seen": 377768576,
+ "step": 8635,
+ "train_runtime": 66676.5105,
+ "train_tokens_per_second": 5665.692
+ },
+ {
+ "epoch": 0.581739833019122,
+ "grad_norm": 0.7908376307807878,
+ "learning_rate": 8.054145386217652e-06,
+ "loss": 0.35233349800109864,
+ "num_input_tokens_seen": 377990912,
+ "step": 8640,
+ "train_runtime": 66709.357,
+ "train_tokens_per_second": 5666.235
+ },
+ {
+ "epoch": 0.5820764880150822,
+ "grad_norm": 0.7544609324314604,
+ "learning_rate": 8.05205147420684e-06,
+ "loss": 0.36639404296875,
+ "num_input_tokens_seen": 378205832,
+ "step": 8645,
+ "train_runtime": 66742.4566,
+ "train_tokens_per_second": 5666.645
+ },
+ {
+ "epoch": 0.5824131430110423,
+ "grad_norm": 0.8852523330609511,
+ "learning_rate": 8.049956708699567e-06,
+ "loss": 0.41215500831604,
+ "num_input_tokens_seen": 378422632,
+ "step": 8650,
+ "train_runtime": 66779.3643,
+ "train_tokens_per_second": 5666.76
+ },
+ {
+ "epoch": 0.5827497980070024,
+ "grad_norm": 0.7748134898838014,
+ "learning_rate": 8.047861090281623e-06,
+ "loss": 0.4017993450164795,
+ "num_input_tokens_seen": 378636312,
+ "step": 8655,
+ "train_runtime": 66817.0207,
+ "train_tokens_per_second": 5666.764
+ },
+ {
+ "epoch": 0.5830864530029626,
+ "grad_norm": 0.7174936151093502,
+ "learning_rate": 8.045764619539047e-06,
+ "loss": 0.36281633377075195,
+ "num_input_tokens_seen": 378862480,
+ "step": 8660,
+ "train_runtime": 66859.5287,
+ "train_tokens_per_second": 5666.544
+ },
+ {
+ "epoch": 0.5834231079989227,
+ "grad_norm": 0.7177129776779663,
+ "learning_rate": 8.043667297058106e-06,
+ "loss": 0.3568310022354126,
+ "num_input_tokens_seen": 379073648,
+ "step": 8665,
+ "train_runtime": 66894.7683,
+ "train_tokens_per_second": 5666.716
+ },
+ {
+ "epoch": 0.5837597629948829,
+ "grad_norm": 0.671557492383721,
+ "learning_rate": 8.04156912342531e-06,
+ "loss": 0.3924489259719849,
+ "num_input_tokens_seen": 379296048,
+ "step": 8670,
+ "train_runtime": 66930.6354,
+ "train_tokens_per_second": 5667.002
+ },
+ {
+ "epoch": 0.584096417990843,
+ "grad_norm": 0.8338696138817213,
+ "learning_rate": 8.03947009922741e-06,
+ "loss": 0.3730125904083252,
+ "num_input_tokens_seen": 379514216,
+ "step": 8675,
+ "train_runtime": 66971.0506,
+ "train_tokens_per_second": 5666.84
+ },
+ {
+ "epoch": 0.5844330729868031,
+ "grad_norm": 0.8046977070640711,
+ "learning_rate": 8.037370225051387e-06,
+ "loss": 0.4212483406066895,
+ "num_input_tokens_seen": 379726496,
+ "step": 8680,
+ "train_runtime": 67011.5775,
+ "train_tokens_per_second": 5666.58
+ },
+ {
+ "epoch": 0.5847697279827633,
+ "grad_norm": 0.747137316814288,
+ "learning_rate": 8.035269501484467e-06,
+ "loss": 0.3820772886276245,
+ "num_input_tokens_seen": 379952664,
+ "step": 8685,
+ "train_runtime": 67045.976,
+ "train_tokens_per_second": 5667.047
+ },
+ {
+ "epoch": 0.5851063829787234,
+ "grad_norm": 0.7779439978184742,
+ "learning_rate": 8.03316792911411e-06,
+ "loss": 0.41135897636413576,
+ "num_input_tokens_seen": 380168720,
+ "step": 8690,
+ "train_runtime": 67086.3352,
+ "train_tokens_per_second": 5666.858
+ },
+ {
+ "epoch": 0.5854430379746836,
+ "grad_norm": 0.6815474179189787,
+ "learning_rate": 8.031065508528011e-06,
+ "loss": 0.35835254192352295,
+ "num_input_tokens_seen": 380396160,
+ "step": 8695,
+ "train_runtime": 67124.515,
+ "train_tokens_per_second": 5667.023
+ },
+ {
+ "epoch": 0.5857796929706437,
+ "grad_norm": 0.7133086557979279,
+ "learning_rate": 8.028962240314111e-06,
+ "loss": 0.3686984539031982,
+ "num_input_tokens_seen": 380620472,
+ "step": 8700,
+ "train_runtime": 67161.6095,
+ "train_tokens_per_second": 5667.233
+ },
+ {
+ "epoch": 0.5861163479666038,
+ "grad_norm": 0.7088226284029326,
+ "learning_rate": 8.026858125060577e-06,
+ "loss": 0.3879531383514404,
+ "num_input_tokens_seen": 380837592,
+ "step": 8705,
+ "train_runtime": 67201.0534,
+ "train_tokens_per_second": 5667.137
+ },
+ {
+ "epoch": 0.586453002962564,
+ "grad_norm": 0.7612804968151702,
+ "learning_rate": 8.024753163355821e-06,
+ "loss": 0.4247321128845215,
+ "num_input_tokens_seen": 381061216,
+ "step": 8710,
+ "train_runtime": 67242.9084,
+ "train_tokens_per_second": 5666.935
+ },
+ {
+ "epoch": 0.5867896579585241,
+ "grad_norm": 0.7005358220776335,
+ "learning_rate": 8.022647355788489e-06,
+ "loss": 0.3679481029510498,
+ "num_input_tokens_seen": 381280288,
+ "step": 8715,
+ "train_runtime": 67281.4067,
+ "train_tokens_per_second": 5666.949
+ },
+ {
+ "epoch": 0.5871263129544843,
+ "grad_norm": 0.728044784671465,
+ "learning_rate": 8.02054070294746e-06,
+ "loss": 0.40247626304626466,
+ "num_input_tokens_seen": 381496992,
+ "step": 8720,
+ "train_runtime": 67319.4589,
+ "train_tokens_per_second": 5666.965
+ },
+ {
+ "epoch": 0.5874629679504444,
+ "grad_norm": 0.7052792472604399,
+ "learning_rate": 8.018433205421859e-06,
+ "loss": 0.36869523525238035,
+ "num_input_tokens_seen": 381727008,
+ "step": 8725,
+ "train_runtime": 67359.8794,
+ "train_tokens_per_second": 5666.979
+ },
+ {
+ "epoch": 0.5877996229464045,
+ "grad_norm": 0.8148613184034657,
+ "learning_rate": 8.016324863801036e-06,
+ "loss": 0.3896385669708252,
+ "num_input_tokens_seen": 381950608,
+ "step": 8730,
+ "train_runtime": 67395.2795,
+ "train_tokens_per_second": 5667.32
+ },
+ {
+ "epoch": 0.5881362779423647,
+ "grad_norm": 0.7008873936428064,
+ "learning_rate": 8.014215678674581e-06,
+ "loss": 0.397304368019104,
+ "num_input_tokens_seen": 382173664,
+ "step": 8735,
+ "train_runtime": 67436.6612,
+ "train_tokens_per_second": 5667.15
+ },
+ {
+ "epoch": 0.5884729329383248,
+ "grad_norm": 0.6667026328000061,
+ "learning_rate": 8.012105650632325e-06,
+ "loss": 0.37210235595703123,
+ "num_input_tokens_seen": 382388568,
+ "step": 8740,
+ "train_runtime": 67476.723,
+ "train_tokens_per_second": 5666.97
+ },
+ {
+ "epoch": 0.588809587934285,
+ "grad_norm": 0.7893753618716253,
+ "learning_rate": 8.009994780264328e-06,
+ "loss": 0.38750429153442384,
+ "num_input_tokens_seen": 382584944,
+ "step": 8745,
+ "train_runtime": 67517.8366,
+ "train_tokens_per_second": 5666.428
+ },
+ {
+ "epoch": 0.5891462429302451,
+ "grad_norm": 0.7520095638811525,
+ "learning_rate": 8.00788306816089e-06,
+ "loss": 0.38128883838653566,
+ "num_input_tokens_seen": 382801808,
+ "step": 8750,
+ "train_runtime": 67554.4917,
+ "train_tokens_per_second": 5666.563
+ },
+ {
+ "epoch": 0.5894828979262052,
+ "grad_norm": 0.7841152997217926,
+ "learning_rate": 8.005770514912543e-06,
+ "loss": 0.39379191398620605,
+ "num_input_tokens_seen": 383025832,
+ "step": 8755,
+ "train_runtime": 67593.0084,
+ "train_tokens_per_second": 5666.649
+ },
+ {
+ "epoch": 0.5898195529221654,
+ "grad_norm": 0.728445499817985,
+ "learning_rate": 8.003657121110055e-06,
+ "loss": 0.3845071792602539,
+ "num_input_tokens_seen": 383260728,
+ "step": 8760,
+ "train_runtime": 67628.415,
+ "train_tokens_per_second": 5667.155
+ },
+ {
+ "epoch": 0.5901562079181255,
+ "grad_norm": 0.7109388567699519,
+ "learning_rate": 8.001542887344432e-06,
+ "loss": 0.36870722770690917,
+ "num_input_tokens_seen": 383481016,
+ "step": 8765,
+ "train_runtime": 67663.424,
+ "train_tokens_per_second": 5667.479
+ },
+ {
+ "epoch": 0.5904928629140856,
+ "grad_norm": 0.7379119456141738,
+ "learning_rate": 7.999427814206911e-06,
+ "loss": 0.38972065448760984,
+ "num_input_tokens_seen": 383699928,
+ "step": 8770,
+ "train_runtime": 67698.1314,
+ "train_tokens_per_second": 5667.807
+ },
+ {
+ "epoch": 0.5908295179100458,
+ "grad_norm": 0.7977299068456677,
+ "learning_rate": 7.997311902288968e-06,
+ "loss": 0.3479362964630127,
+ "num_input_tokens_seen": 383923208,
+ "step": 8775,
+ "train_runtime": 67735.622,
+ "train_tokens_per_second": 5667.966
+ },
+ {
+ "epoch": 0.5911661729060059,
+ "grad_norm": 0.6927052461533478,
+ "learning_rate": 7.995195152182307e-06,
+ "loss": 0.35241219997406004,
+ "num_input_tokens_seen": 384124912,
+ "step": 8780,
+ "train_runtime": 67775.1001,
+ "train_tokens_per_second": 5667.641
+ },
+ {
+ "epoch": 0.5915028279019661,
+ "grad_norm": 1.8520827992951523,
+ "learning_rate": 7.993077564478875e-06,
+ "loss": 0.367362380027771,
+ "num_input_tokens_seen": 384352936,
+ "step": 8785,
+ "train_runtime": 67813.9355,
+ "train_tokens_per_second": 5667.757
+ },
+ {
+ "epoch": 0.5918394828979262,
+ "grad_norm": 0.8136091755899332,
+ "learning_rate": 7.990959139770844e-06,
+ "loss": 0.3980868339538574,
+ "num_input_tokens_seen": 384586952,
+ "step": 8790,
+ "train_runtime": 67855.5637,
+ "train_tokens_per_second": 5667.729
+ },
+ {
+ "epoch": 0.5921761378938863,
+ "grad_norm": 0.7689488310801835,
+ "learning_rate": 7.988839878650628e-06,
+ "loss": 0.3829834461212158,
+ "num_input_tokens_seen": 384813912,
+ "step": 8795,
+ "train_runtime": 67899.407,
+ "train_tokens_per_second": 5667.412
+ },
+ {
+ "epoch": 0.5925127928898465,
+ "grad_norm": 0.7981067639524714,
+ "learning_rate": 7.986719781710872e-06,
+ "loss": 0.3936153411865234,
+ "num_input_tokens_seen": 385040600,
+ "step": 8800,
+ "train_runtime": 67941.2356,
+ "train_tokens_per_second": 5667.259
+ },
+ {
+ "epoch": 0.5928494478858066,
+ "grad_norm": 0.7888638773984543,
+ "learning_rate": 7.984598849544452e-06,
+ "loss": 0.37601919174194337,
+ "num_input_tokens_seen": 385256736,
+ "step": 8805,
+ "train_runtime": 67980.1555,
+ "train_tokens_per_second": 5667.194
+ },
+ {
+ "epoch": 0.5931861028817668,
+ "grad_norm": 0.8141275660860262,
+ "learning_rate": 7.98247708274448e-06,
+ "loss": 0.3897037744522095,
+ "num_input_tokens_seen": 385471192,
+ "step": 8810,
+ "train_runtime": 68015.2704,
+ "train_tokens_per_second": 5667.421
+ },
+ {
+ "epoch": 0.5935227578777269,
+ "grad_norm": 0.7312305067797448,
+ "learning_rate": 7.980354481904303e-06,
+ "loss": 0.4111155033111572,
+ "num_input_tokens_seen": 385695968,
+ "step": 8815,
+ "train_runtime": 68053.7353,
+ "train_tokens_per_second": 5667.521
+ },
+ {
+ "epoch": 0.593859412873687,
+ "grad_norm": 0.8597141063820571,
+ "learning_rate": 7.978231047617498e-06,
+ "loss": 0.41021156311035156,
+ "num_input_tokens_seen": 385910576,
+ "step": 8820,
+ "train_runtime": 68091.813,
+ "train_tokens_per_second": 5667.503
+ },
+ {
+ "epoch": 0.5941960678696472,
+ "grad_norm": 0.7977576371807428,
+ "learning_rate": 7.976106780477878e-06,
+ "loss": 0.4051943778991699,
+ "num_input_tokens_seen": 386135568,
+ "step": 8825,
+ "train_runtime": 68134.8361,
+ "train_tokens_per_second": 5667.227
+ },
+ {
+ "epoch": 0.5945327228656073,
+ "grad_norm": 0.7369358873224232,
+ "learning_rate": 7.973981681079487e-06,
+ "loss": 0.3718585729598999,
+ "num_input_tokens_seen": 386344944,
+ "step": 8830,
+ "train_runtime": 68175.9892,
+ "train_tokens_per_second": 5666.877
+ },
+ {
+ "epoch": 0.5948693778615675,
+ "grad_norm": 0.7361018680183479,
+ "learning_rate": 7.971855750016603e-06,
+ "loss": 0.37234504222869874,
+ "num_input_tokens_seen": 386561056,
+ "step": 8835,
+ "train_runtime": 68210.534,
+ "train_tokens_per_second": 5667.175
+ },
+ {
+ "epoch": 0.5952060328575276,
+ "grad_norm": 0.7646438523333068,
+ "learning_rate": 7.969728987883734e-06,
+ "loss": 0.386487340927124,
+ "num_input_tokens_seen": 386782376,
+ "step": 8840,
+ "train_runtime": 68250.0947,
+ "train_tokens_per_second": 5667.133
+ },
+ {
+ "epoch": 0.5955426878534877,
+ "grad_norm": 0.8191747882763178,
+ "learning_rate": 7.967601395275624e-06,
+ "loss": 0.38187594413757325,
+ "num_input_tokens_seen": 386999696,
+ "step": 8845,
+ "train_runtime": 68291.3193,
+ "train_tokens_per_second": 5666.894
+ },
+ {
+ "epoch": 0.5958793428494479,
+ "grad_norm": 0.6459146656871505,
+ "learning_rate": 7.965472972787247e-06,
+ "loss": 0.35875325202941893,
+ "num_input_tokens_seen": 387220408,
+ "step": 8850,
+ "train_runtime": 68328.0647,
+ "train_tokens_per_second": 5667.077
+ },
+ {
+ "epoch": 0.596215997845408,
+ "grad_norm": 0.7549464702278598,
+ "learning_rate": 7.963343721013808e-06,
+ "loss": 0.3718303680419922,
+ "num_input_tokens_seen": 387455840,
+ "step": 8855,
+ "train_runtime": 68366.913,
+ "train_tokens_per_second": 5667.3
+ },
+ {
+ "epoch": 0.5965526528413682,
+ "grad_norm": 0.7558527061932778,
+ "learning_rate": 7.961213640550748e-06,
+ "loss": 0.3902780771255493,
+ "num_input_tokens_seen": 387678720,
+ "step": 8860,
+ "train_runtime": 68410.3443,
+ "train_tokens_per_second": 5666.961
+ },
+ {
+ "epoch": 0.5968893078373283,
+ "grad_norm": 0.7082248157089,
+ "learning_rate": 7.959082731993736e-06,
+ "loss": 0.3638461589813232,
+ "num_input_tokens_seen": 387897464,
+ "step": 8865,
+ "train_runtime": 68443.0142,
+ "train_tokens_per_second": 5667.452
+ },
+ {
+ "epoch": 0.5972259628332884,
+ "grad_norm": 0.9079660049109616,
+ "learning_rate": 7.956950995938673e-06,
+ "loss": 0.3705005168914795,
+ "num_input_tokens_seen": 388107624,
+ "step": 8870,
+ "train_runtime": 68486.4151,
+ "train_tokens_per_second": 5666.929
+ },
+ {
+ "epoch": 0.5975626178292486,
+ "grad_norm": 0.7984526639513112,
+ "learning_rate": 7.954818432981694e-06,
+ "loss": 0.3910086631774902,
+ "num_input_tokens_seen": 388310680,
+ "step": 8875,
+ "train_runtime": 68525.481,
+ "train_tokens_per_second": 5666.661
+ },
+ {
+ "epoch": 0.5978992728252087,
+ "grad_norm": 0.7147181194308313,
+ "learning_rate": 7.952685043719162e-06,
+ "loss": 0.3656719923019409,
+ "num_input_tokens_seen": 388533832,
+ "step": 8880,
+ "train_runtime": 68568.5855,
+ "train_tokens_per_second": 5666.353
+ },
+ {
+ "epoch": 0.5982359278211689,
+ "grad_norm": 0.7782537183487828,
+ "learning_rate": 7.950550828747672e-06,
+ "loss": 0.37486002445220945,
+ "num_input_tokens_seen": 388754728,
+ "step": 8885,
+ "train_runtime": 68608.906,
+ "train_tokens_per_second": 5666.243
+ },
+ {
+ "epoch": 0.598572582817129,
+ "grad_norm": 0.6958894420319796,
+ "learning_rate": 7.948415788664052e-06,
+ "loss": 0.40519165992736816,
+ "num_input_tokens_seen": 388977056,
+ "step": 8890,
+ "train_runtime": 68641.9616,
+ "train_tokens_per_second": 5666.753
+ },
+ {
+ "epoch": 0.5989092378130891,
+ "grad_norm": 0.8244159660705994,
+ "learning_rate": 7.946279924065359e-06,
+ "loss": 0.3696116924285889,
+ "num_input_tokens_seen": 389195368,
+ "step": 8895,
+ "train_runtime": 68676.2104,
+ "train_tokens_per_second": 5667.106
+ },
+ {
+ "epoch": 0.5992458928090493,
+ "grad_norm": 0.7624772559058,
+ "learning_rate": 7.94414323554888e-06,
+ "loss": 0.3721637725830078,
+ "num_input_tokens_seen": 389424256,
+ "step": 8900,
+ "train_runtime": 68707.9383,
+ "train_tokens_per_second": 5667.82
+ },
+ {
+ "epoch": 0.5995825478050094,
+ "grad_norm": 0.7421871285525378,
+ "learning_rate": 7.942005723712135e-06,
+ "loss": 0.3689994812011719,
+ "num_input_tokens_seen": 389659456,
+ "step": 8905,
+ "train_runtime": 68742.7631,
+ "train_tokens_per_second": 5668.371
+ },
+ {
+ "epoch": 0.5999192028009696,
+ "grad_norm": 0.7598715011111157,
+ "learning_rate": 7.939867389152868e-06,
+ "loss": 0.36181044578552246,
+ "num_input_tokens_seen": 389878008,
+ "step": 8910,
+ "train_runtime": 68782.9286,
+ "train_tokens_per_second": 5668.238
+ },
+ {
+ "epoch": 0.6002558577969297,
+ "grad_norm": 0.7642969305957489,
+ "learning_rate": 7.937728232469062e-06,
+ "loss": 0.3904815435409546,
+ "num_input_tokens_seen": 390090680,
+ "step": 8915,
+ "train_runtime": 68823.5271,
+ "train_tokens_per_second": 5667.984
+ },
+ {
+ "epoch": 0.6005925127928898,
+ "grad_norm": 0.7743252489033806,
+ "learning_rate": 7.935588254258921e-06,
+ "loss": 0.407269811630249,
+ "num_input_tokens_seen": 390298200,
+ "step": 8920,
+ "train_runtime": 68859.6954,
+ "train_tokens_per_second": 5668.021
+ },
+ {
+ "epoch": 0.60092916778885,
+ "grad_norm": 0.7626396676106256,
+ "learning_rate": 7.933447455120889e-06,
+ "loss": 0.3707296848297119,
+ "num_input_tokens_seen": 390513112,
+ "step": 8925,
+ "train_runtime": 68906.8163,
+ "train_tokens_per_second": 5667.264
+ },
+ {
+ "epoch": 0.6012658227848101,
+ "grad_norm": 0.9270450940086636,
+ "learning_rate": 7.931305835653629e-06,
+ "loss": 0.34468765258789064,
+ "num_input_tokens_seen": 390741944,
+ "step": 8930,
+ "train_runtime": 68945.5046,
+ "train_tokens_per_second": 5667.403
+ },
+ {
+ "epoch": 0.6016024777807703,
+ "grad_norm": 0.7523711200271979,
+ "learning_rate": 7.929163396456038e-06,
+ "loss": 0.39664325714111326,
+ "num_input_tokens_seen": 390950400,
+ "step": 8935,
+ "train_runtime": 68982.333,
+ "train_tokens_per_second": 5667.399
+ },
+ {
+ "epoch": 0.6019391327767304,
+ "grad_norm": 0.7743378155770866,
+ "learning_rate": 7.927020138127245e-06,
+ "loss": 0.369353199005127,
+ "num_input_tokens_seen": 391172832,
+ "step": 8940,
+ "train_runtime": 69022.2251,
+ "train_tokens_per_second": 5667.346
+ },
+ {
+ "epoch": 0.6022757877726905,
+ "grad_norm": 0.6649737691389936,
+ "learning_rate": 7.924876061266603e-06,
+ "loss": 0.3477231502532959,
+ "num_input_tokens_seen": 391403392,
+ "step": 8945,
+ "train_runtime": 69060.7271,
+ "train_tokens_per_second": 5667.525
+ },
+ {
+ "epoch": 0.6026124427686507,
+ "grad_norm": 0.8428633382608517,
+ "learning_rate": 7.922731166473698e-06,
+ "loss": 0.40044050216674804,
+ "num_input_tokens_seen": 391619536,
+ "step": 8950,
+ "train_runtime": 69095.3564,
+ "train_tokens_per_second": 5667.813
+ },
+ {
+ "epoch": 0.6029490977646108,
+ "grad_norm": 0.7797139187172423,
+ "learning_rate": 7.920585454348341e-06,
+ "loss": 0.35779447555541993,
+ "num_input_tokens_seen": 391832424,
+ "step": 8955,
+ "train_runtime": 69130.8816,
+ "train_tokens_per_second": 5667.98
+ },
+ {
+ "epoch": 0.603285752760571,
+ "grad_norm": 0.6454426330600102,
+ "learning_rate": 7.918438925490578e-06,
+ "loss": 0.35099799633026124,
+ "num_input_tokens_seen": 392057736,
+ "step": 8960,
+ "train_runtime": 69168.5952,
+ "train_tokens_per_second": 5668.147
+ },
+ {
+ "epoch": 0.6036224077565311,
+ "grad_norm": 0.9240196513971288,
+ "learning_rate": 7.916291580500672e-06,
+ "loss": 0.37104010581970215,
+ "num_input_tokens_seen": 392274152,
+ "step": 8965,
+ "train_runtime": 69209.1039,
+ "train_tokens_per_second": 5667.956
+ },
+ {
+ "epoch": 0.6039590627524912,
+ "grad_norm": 0.753245828305525,
+ "learning_rate": 7.914143419979127e-06,
+ "loss": 0.36745285987854004,
+ "num_input_tokens_seen": 392499120,
+ "step": 8970,
+ "train_runtime": 69247.1372,
+ "train_tokens_per_second": 5668.092
+ },
+ {
+ "epoch": 0.6042957177484514,
+ "grad_norm": 0.6569111554819693,
+ "learning_rate": 7.911994444526667e-06,
+ "loss": 0.34250531196594236,
+ "num_input_tokens_seen": 392720016,
+ "step": 8975,
+ "train_runtime": 69281.6006,
+ "train_tokens_per_second": 5668.46
+ },
+ {
+ "epoch": 0.6046323727444115,
+ "grad_norm": 0.6710386941472357,
+ "learning_rate": 7.909844654744246e-06,
+ "loss": 0.3526078939437866,
+ "num_input_tokens_seen": 392945160,
+ "step": 8980,
+ "train_runtime": 69318.5011,
+ "train_tokens_per_second": 5668.691
+ },
+ {
+ "epoch": 0.6049690277403716,
+ "grad_norm": 0.7627659966441698,
+ "learning_rate": 7.907694051233044e-06,
+ "loss": 0.38707795143127444,
+ "num_input_tokens_seen": 393164952,
+ "step": 8985,
+ "train_runtime": 69355.9491,
+ "train_tokens_per_second": 5668.799
+ },
+ {
+ "epoch": 0.6053056827363318,
+ "grad_norm": 0.6996994069888867,
+ "learning_rate": 7.905542634594476e-06,
+ "loss": 0.3759605884552002,
+ "num_input_tokens_seen": 393392000,
+ "step": 8990,
+ "train_runtime": 69403.0227,
+ "train_tokens_per_second": 5668.226
+ },
+ {
+ "epoch": 0.6056423377322919,
+ "grad_norm": 0.7505686616004086,
+ "learning_rate": 7.903390405430174e-06,
+ "loss": 0.36977381706237794,
+ "num_input_tokens_seen": 393611632,
+ "step": 8995,
+ "train_runtime": 69443.7718,
+ "train_tokens_per_second": 5668.062
+ },
+ {
+ "epoch": 0.6059789927282521,
+ "grad_norm": 0.7461690774426525,
+ "learning_rate": 7.901237364342003e-06,
+ "loss": 0.3545074939727783,
+ "num_input_tokens_seen": 393813184,
+ "step": 9000,
+ "train_runtime": 69478.2997,
+ "train_tokens_per_second": 5668.147
+ },
+ {
+ "epoch": 0.6063156477242122,
+ "grad_norm": 0.692733389183626,
+ "learning_rate": 7.899083511932053e-06,
+ "loss": 0.3470455646514893,
+ "num_input_tokens_seen": 394027272,
+ "step": 9005,
+ "train_runtime": 69518.7288,
+ "train_tokens_per_second": 5667.93
+ },
+ {
+ "epoch": 0.6066523027201723,
+ "grad_norm": 0.7055838759339507,
+ "learning_rate": 7.896928848802645e-06,
+ "loss": 0.372318959236145,
+ "num_input_tokens_seen": 394252480,
+ "step": 9010,
+ "train_runtime": 69553.6455,
+ "train_tokens_per_second": 5668.322
+ },
+ {
+ "epoch": 0.6069889577161325,
+ "grad_norm": 0.7576301760274441,
+ "learning_rate": 7.894773375556322e-06,
+ "loss": 0.39665966033935546,
+ "num_input_tokens_seen": 394449024,
+ "step": 9015,
+ "train_runtime": 69593.8001,
+ "train_tokens_per_second": 5667.876
+ },
+ {
+ "epoch": 0.6073256127120926,
+ "grad_norm": 0.8095166727188953,
+ "learning_rate": 7.892617092795855e-06,
+ "loss": 0.38990225791931155,
+ "num_input_tokens_seen": 394672680,
+ "step": 9020,
+ "train_runtime": 69632.6302,
+ "train_tokens_per_second": 5667.927
+ },
+ {
+ "epoch": 0.6076622677080528,
+ "grad_norm": 0.6966710594837849,
+ "learning_rate": 7.890460001124242e-06,
+ "loss": 0.3693479061126709,
+ "num_input_tokens_seen": 394882720,
+ "step": 9025,
+ "train_runtime": 69668.2129,
+ "train_tokens_per_second": 5668.047
+ },
+ {
+ "epoch": 0.6079989227040129,
+ "grad_norm": 0.7467966463857733,
+ "learning_rate": 7.888302101144707e-06,
+ "loss": 0.3727553367614746,
+ "num_input_tokens_seen": 395096336,
+ "step": 9030,
+ "train_runtime": 69710.5053,
+ "train_tokens_per_second": 5667.673
+ },
+ {
+ "epoch": 0.608335577699973,
+ "grad_norm": 0.7033388740227385,
+ "learning_rate": 7.886143393460699e-06,
+ "loss": 0.38645596504211427,
+ "num_input_tokens_seen": 395327104,
+ "step": 9035,
+ "train_runtime": 69746.6807,
+ "train_tokens_per_second": 5668.042
+ },
+ {
+ "epoch": 0.6086722326959332,
+ "grad_norm": 0.717725417977876,
+ "learning_rate": 7.883983878675895e-06,
+ "loss": 0.3641960620880127,
+ "num_input_tokens_seen": 395561000,
+ "step": 9040,
+ "train_runtime": 69794.3801,
+ "train_tokens_per_second": 5667.519
+ },
+ {
+ "epoch": 0.6090088876918933,
+ "grad_norm": 0.737338173324883,
+ "learning_rate": 7.881823557394194e-06,
+ "loss": 0.3701021194458008,
+ "num_input_tokens_seen": 395790544,
+ "step": 9045,
+ "train_runtime": 69832.741,
+ "train_tokens_per_second": 5667.693
+ },
+ {
+ "epoch": 0.6093455426878535,
+ "grad_norm": 0.7527693128608011,
+ "learning_rate": 7.879662430219728e-06,
+ "loss": 0.35156404972076416,
+ "num_input_tokens_seen": 395987680,
+ "step": 9050,
+ "train_runtime": 69873.2033,
+ "train_tokens_per_second": 5667.232
+ },
+ {
+ "epoch": 0.6096821976838136,
+ "grad_norm": 0.8175940814840023,
+ "learning_rate": 7.877500497756843e-06,
+ "loss": 0.4006368637084961,
+ "num_input_tokens_seen": 396207224,
+ "step": 9055,
+ "train_runtime": 69914.8014,
+ "train_tokens_per_second": 5667.001
+ },
+ {
+ "epoch": 0.6100188526797737,
+ "grad_norm": 1.0466330679614098,
+ "learning_rate": 7.87533776061012e-06,
+ "loss": 0.3919075012207031,
+ "num_input_tokens_seen": 396431144,
+ "step": 9060,
+ "train_runtime": 69954.8331,
+ "train_tokens_per_second": 5666.959
+ },
+ {
+ "epoch": 0.6103555076757339,
+ "grad_norm": 0.7304596174173837,
+ "learning_rate": 7.873174219384362e-06,
+ "loss": 0.3918348789215088,
+ "num_input_tokens_seen": 396647760,
+ "step": 9065,
+ "train_runtime": 69996.3796,
+ "train_tokens_per_second": 5666.69
+ },
+ {
+ "epoch": 0.610692162671694,
+ "grad_norm": 0.7556787705072924,
+ "learning_rate": 7.871009874684595e-06,
+ "loss": 0.3941289186477661,
+ "num_input_tokens_seen": 396880640,
+ "step": 9070,
+ "train_runtime": 70028.8034,
+ "train_tokens_per_second": 5667.391
+ },
+ {
+ "epoch": 0.6110288176676542,
+ "grad_norm": 0.8397284206623034,
+ "learning_rate": 7.868844727116072e-06,
+ "loss": 0.3926485300064087,
+ "num_input_tokens_seen": 397103976,
+ "step": 9075,
+ "train_runtime": 70068.2945,
+ "train_tokens_per_second": 5667.385
+ },
+ {
+ "epoch": 0.6113654726636143,
+ "grad_norm": 0.6869244210168536,
+ "learning_rate": 7.866678777284267e-06,
+ "loss": 0.3840303897857666,
+ "num_input_tokens_seen": 397331160,
+ "step": 9080,
+ "train_runtime": 70102.9706,
+ "train_tokens_per_second": 5667.822
+ },
+ {
+ "epoch": 0.6117021276595744,
+ "grad_norm": 0.6366270970878026,
+ "learning_rate": 7.864512025794886e-06,
+ "loss": 0.3659891843795776,
+ "num_input_tokens_seen": 397555992,
+ "step": 9085,
+ "train_runtime": 70142.9573,
+ "train_tokens_per_second": 5667.796
+ },
+ {
+ "epoch": 0.6120387826555346,
+ "grad_norm": 0.7702173454778595,
+ "learning_rate": 7.86234447325385e-06,
+ "loss": 0.36196537017822267,
+ "num_input_tokens_seen": 397769720,
+ "step": 9090,
+ "train_runtime": 70185.3981,
+ "train_tokens_per_second": 5667.414
+ },
+ {
+ "epoch": 0.6123754376514947,
+ "grad_norm": 0.6638241416823153,
+ "learning_rate": 7.860176120267309e-06,
+ "loss": 0.4144433498382568,
+ "num_input_tokens_seen": 398006352,
+ "step": 9095,
+ "train_runtime": 70227.0746,
+ "train_tokens_per_second": 5667.42
+ },
+ {
+ "epoch": 0.6127120926474549,
+ "grad_norm": 0.7106604037911711,
+ "learning_rate": 7.858006967441637e-06,
+ "loss": 0.3602044105529785,
+ "num_input_tokens_seen": 398216992,
+ "step": 9100,
+ "train_runtime": 70268.2952,
+ "train_tokens_per_second": 5667.093
+ },
+ {
+ "epoch": 0.613048747643415,
+ "grad_norm": 0.8013079858059877,
+ "learning_rate": 7.85583701538343e-06,
+ "loss": 0.40687074661254885,
+ "num_input_tokens_seen": 398447240,
+ "step": 9105,
+ "train_runtime": 70313.64,
+ "train_tokens_per_second": 5666.713
+ },
+ {
+ "epoch": 0.6133854026393751,
+ "grad_norm": 0.8363243678158666,
+ "learning_rate": 7.853666264699506e-06,
+ "loss": 0.3638491153717041,
+ "num_input_tokens_seen": 398656928,
+ "step": 9110,
+ "train_runtime": 70355.8426,
+ "train_tokens_per_second": 5666.295
+ },
+ {
+ "epoch": 0.6137220576353353,
+ "grad_norm": 0.6855947504133412,
+ "learning_rate": 7.851494715996914e-06,
+ "loss": 0.41129069328308104,
+ "num_input_tokens_seen": 398870368,
+ "step": 9115,
+ "train_runtime": 70395.581,
+ "train_tokens_per_second": 5666.128
+ },
+ {
+ "epoch": 0.6140587126312954,
+ "grad_norm": 0.7613022857869545,
+ "learning_rate": 7.849322369882915e-06,
+ "loss": 0.3685478687286377,
+ "num_input_tokens_seen": 399088536,
+ "step": 9120,
+ "train_runtime": 70444.8636,
+ "train_tokens_per_second": 5665.261
+ },
+ {
+ "epoch": 0.6143953676272556,
+ "grad_norm": 0.6988300008519205,
+ "learning_rate": 7.847149226965003e-06,
+ "loss": 0.34872984886169434,
+ "num_input_tokens_seen": 399301232,
+ "step": 9125,
+ "train_runtime": 70485.0506,
+ "train_tokens_per_second": 5665.049
+ },
+ {
+ "epoch": 0.6147320226232157,
+ "grad_norm": 0.8477724491109673,
+ "learning_rate": 7.844975287850886e-06,
+ "loss": 0.35025968551635744,
+ "num_input_tokens_seen": 399523504,
+ "step": 9130,
+ "train_runtime": 70522.5255,
+ "train_tokens_per_second": 5665.19
+ },
+ {
+ "epoch": 0.6150686776191758,
+ "grad_norm": 0.7692634619323344,
+ "learning_rate": 7.842800553148506e-06,
+ "loss": 0.3998849630355835,
+ "num_input_tokens_seen": 399735256,
+ "step": 9135,
+ "train_runtime": 70558.7444,
+ "train_tokens_per_second": 5665.283
+ },
+ {
+ "epoch": 0.615405332615136,
+ "grad_norm": 0.7313920740237436,
+ "learning_rate": 7.840625023466014e-06,
+ "loss": 0.3812243938446045,
+ "num_input_tokens_seen": 399958824,
+ "step": 9140,
+ "train_runtime": 70598.2086,
+ "train_tokens_per_second": 5665.283
+ },
+ {
+ "epoch": 0.6157419876110961,
+ "grad_norm": 0.7274387978674163,
+ "learning_rate": 7.838448699411792e-06,
+ "loss": 0.3413673400878906,
+ "num_input_tokens_seen": 400181528,
+ "step": 9145,
+ "train_runtime": 70641.2437,
+ "train_tokens_per_second": 5664.984
+ },
+ {
+ "epoch": 0.6160786426070562,
+ "grad_norm": 0.7006449570449917,
+ "learning_rate": 7.836271581594442e-06,
+ "loss": 0.4210306167602539,
+ "num_input_tokens_seen": 400404728,
+ "step": 9150,
+ "train_runtime": 70677.6291,
+ "train_tokens_per_second": 5665.226
+ },
+ {
+ "epoch": 0.6164152976030164,
+ "grad_norm": 0.8552267430665972,
+ "learning_rate": 7.834093670622789e-06,
+ "loss": 0.3944708824157715,
+ "num_input_tokens_seen": 400620128,
+ "step": 9155,
+ "train_runtime": 70718.8549,
+ "train_tokens_per_second": 5664.969
+ },
+ {
+ "epoch": 0.6167519525989765,
+ "grad_norm": 0.7631479691553084,
+ "learning_rate": 7.831914967105878e-06,
+ "loss": 0.4071969985961914,
+ "num_input_tokens_seen": 400851832,
+ "step": 9160,
+ "train_runtime": 70759.3063,
+ "train_tokens_per_second": 5665.005
+ },
+ {
+ "epoch": 0.6170886075949367,
+ "grad_norm": 0.6980894707116475,
+ "learning_rate": 7.829735471652978e-06,
+ "loss": 0.3579000234603882,
+ "num_input_tokens_seen": 401070744,
+ "step": 9165,
+ "train_runtime": 70799.3413,
+ "train_tokens_per_second": 5664.894
+ },
+ {
+ "epoch": 0.6174252625908968,
+ "grad_norm": 0.8778577602385942,
+ "learning_rate": 7.827555184873575e-06,
+ "loss": 0.38968710899353026,
+ "num_input_tokens_seen": 401294816,
+ "step": 9170,
+ "train_runtime": 70831.4898,
+ "train_tokens_per_second": 5665.486
+ },
+ {
+ "epoch": 0.617761917586857,
+ "grad_norm": 0.6780563561113029,
+ "learning_rate": 7.825374107377381e-06,
+ "loss": 0.35830237865448,
+ "num_input_tokens_seen": 401509200,
+ "step": 9175,
+ "train_runtime": 70873.5987,
+ "train_tokens_per_second": 5665.145
+ },
+ {
+ "epoch": 0.6180985725828171,
+ "grad_norm": 0.7956649065370722,
+ "learning_rate": 7.823192239774327e-06,
+ "loss": 0.3784201145172119,
+ "num_input_tokens_seen": 401720584,
+ "step": 9180,
+ "train_runtime": 70913.365,
+ "train_tokens_per_second": 5664.949
+ },
+ {
+ "epoch": 0.6184352275787772,
+ "grad_norm": 0.689595538429294,
+ "learning_rate": 7.821009582674567e-06,
+ "loss": 0.3932128667831421,
+ "num_input_tokens_seen": 401937000,
+ "step": 9185,
+ "train_runtime": 70950.5951,
+ "train_tokens_per_second": 5665.026
+ },
+ {
+ "epoch": 0.6187718825747374,
+ "grad_norm": 0.7661957206566802,
+ "learning_rate": 7.818826136688473e-06,
+ "loss": 0.41967573165893557,
+ "num_input_tokens_seen": 402163016,
+ "step": 9190,
+ "train_runtime": 70988.3051,
+ "train_tokens_per_second": 5665.201
+ },
+ {
+ "epoch": 0.6191085375706975,
+ "grad_norm": 1.24632241828128,
+ "learning_rate": 7.816641902426638e-06,
+ "loss": 0.3871551513671875,
+ "num_input_tokens_seen": 402390608,
+ "step": 9195,
+ "train_runtime": 71026.2689,
+ "train_tokens_per_second": 5665.377
+ },
+ {
+ "epoch": 0.6194451925666576,
+ "grad_norm": 0.7754009003596685,
+ "learning_rate": 7.814456880499877e-06,
+ "loss": 0.39989705085754396,
+ "num_input_tokens_seen": 402608440,
+ "step": 9200,
+ "train_runtime": 71059.0686,
+ "train_tokens_per_second": 5665.828
+ },
+ {
+ "epoch": 0.6197818475626178,
+ "grad_norm": 0.8050073203615641,
+ "learning_rate": 7.812271071519222e-06,
+ "loss": 0.40456953048706057,
+ "num_input_tokens_seen": 402830072,
+ "step": 9205,
+ "train_runtime": 71092.5653,
+ "train_tokens_per_second": 5666.276
+ },
+ {
+ "epoch": 0.6201185025585779,
+ "grad_norm": 0.636871155540424,
+ "learning_rate": 7.810084476095932e-06,
+ "loss": 0.3537304401397705,
+ "num_input_tokens_seen": 403051544,
+ "step": 9210,
+ "train_runtime": 71132.4539,
+ "train_tokens_per_second": 5666.212
+ },
+ {
+ "epoch": 0.6204551575545381,
+ "grad_norm": 0.7771337802144807,
+ "learning_rate": 7.807897094841476e-06,
+ "loss": 0.3726407527923584,
+ "num_input_tokens_seen": 403260152,
+ "step": 9215,
+ "train_runtime": 71171.5092,
+ "train_tokens_per_second": 5666.033
+ },
+ {
+ "epoch": 0.6207918125504982,
+ "grad_norm": 0.8018367108579978,
+ "learning_rate": 7.805708928367555e-06,
+ "loss": 0.3734896183013916,
+ "num_input_tokens_seen": 403468824,
+ "step": 9220,
+ "train_runtime": 71211.2179,
+ "train_tokens_per_second": 5665.804
+ },
+ {
+ "epoch": 0.6211284675464583,
+ "grad_norm": 0.7409189256900955,
+ "learning_rate": 7.803519977286075e-06,
+ "loss": 0.4021895885467529,
+ "num_input_tokens_seen": 403677704,
+ "step": 9225,
+ "train_runtime": 71249.4495,
+ "train_tokens_per_second": 5665.696
+ },
+ {
+ "epoch": 0.6214651225424185,
+ "grad_norm": 0.7435590926911755,
+ "learning_rate": 7.801330242209176e-06,
+ "loss": 0.38091113567352297,
+ "num_input_tokens_seen": 403883200,
+ "step": 9230,
+ "train_runtime": 71290.1392,
+ "train_tokens_per_second": 5665.345
+ },
+ {
+ "epoch": 0.6218017775383786,
+ "grad_norm": 0.7623979983718003,
+ "learning_rate": 7.799139723749207e-06,
+ "loss": 0.3860699415206909,
+ "num_input_tokens_seen": 404093128,
+ "step": 9235,
+ "train_runtime": 71328.4839,
+ "train_tokens_per_second": 5665.242
+ },
+ {
+ "epoch": 0.6221384325343388,
+ "grad_norm": 0.752221691602039,
+ "learning_rate": 7.79694842251874e-06,
+ "loss": 0.3864139080047607,
+ "num_input_tokens_seen": 404319728,
+ "step": 9240,
+ "train_runtime": 71368.1022,
+ "train_tokens_per_second": 5665.272
+ },
+ {
+ "epoch": 0.6224750875302989,
+ "grad_norm": 0.669956786900923,
+ "learning_rate": 7.794756339130565e-06,
+ "loss": 0.3734030485153198,
+ "num_input_tokens_seen": 404541968,
+ "step": 9245,
+ "train_runtime": 71411.2121,
+ "train_tokens_per_second": 5664.964
+ },
+ {
+ "epoch": 0.622811742526259,
+ "grad_norm": 0.6983661416716165,
+ "learning_rate": 7.792563474197692e-06,
+ "loss": 0.3814718008041382,
+ "num_input_tokens_seen": 404754312,
+ "step": 9250,
+ "train_runtime": 71448.6745,
+ "train_tokens_per_second": 5664.966
+ },
+ {
+ "epoch": 0.6231483975222192,
+ "grad_norm": 0.7124131828610318,
+ "learning_rate": 7.79036982833335e-06,
+ "loss": 0.4030300617218018,
+ "num_input_tokens_seen": 404985704,
+ "step": 9255,
+ "train_runtime": 71479.1768,
+ "train_tokens_per_second": 5665.786
+ },
+ {
+ "epoch": 0.6234850525181793,
+ "grad_norm": 0.7886397068317147,
+ "learning_rate": 7.78817540215098e-06,
+ "loss": 0.3688653469085693,
+ "num_input_tokens_seen": 405208848,
+ "step": 9260,
+ "train_runtime": 71523.5445,
+ "train_tokens_per_second": 5665.391
+ },
+ {
+ "epoch": 0.6238217075141395,
+ "grad_norm": 0.7219385997208987,
+ "learning_rate": 7.78598019626425e-06,
+ "loss": 0.37583611011505125,
+ "num_input_tokens_seen": 405447552,
+ "step": 9265,
+ "train_runtime": 71555.7936,
+ "train_tokens_per_second": 5666.174
+ },
+ {
+ "epoch": 0.6241583625100996,
+ "grad_norm": 0.7695265603997767,
+ "learning_rate": 7.783784211287044e-06,
+ "loss": 0.4006516456604004,
+ "num_input_tokens_seen": 405651112,
+ "step": 9270,
+ "train_runtime": 71597.6416,
+ "train_tokens_per_second": 5665.705
+ },
+ {
+ "epoch": 0.6244950175060597,
+ "grad_norm": 0.9194045285642082,
+ "learning_rate": 7.78158744783346e-06,
+ "loss": 0.408467960357666,
+ "num_input_tokens_seen": 405873152,
+ "step": 9275,
+ "train_runtime": 71636.7797,
+ "train_tokens_per_second": 5665.709
+ },
+ {
+ "epoch": 0.6248316725020199,
+ "grad_norm": 0.7515447419413074,
+ "learning_rate": 7.779389906517813e-06,
+ "loss": 0.35638930797576907,
+ "num_input_tokens_seen": 406086632,
+ "step": 9280,
+ "train_runtime": 71670.2143,
+ "train_tokens_per_second": 5666.045
+ },
+ {
+ "epoch": 0.6251683274979801,
+ "grad_norm": 0.7184945703842209,
+ "learning_rate": 7.777191587954645e-06,
+ "loss": 0.3749531745910645,
+ "num_input_tokens_seen": 406298704,
+ "step": 9285,
+ "train_runtime": 71712.7407,
+ "train_tokens_per_second": 5665.642
+ },
+ {
+ "epoch": 0.6255049824939403,
+ "grad_norm": 0.8027326683853181,
+ "learning_rate": 7.774992492758704e-06,
+ "loss": 0.36041698455810545,
+ "num_input_tokens_seen": 406502360,
+ "step": 9290,
+ "train_runtime": 71750.3732,
+ "train_tokens_per_second": 5665.509
+ },
+ {
+ "epoch": 0.6258416374899004,
+ "grad_norm": 0.7186651340067036,
+ "learning_rate": 7.77279262154496e-06,
+ "loss": 0.38366947174072263,
+ "num_input_tokens_seen": 406727224,
+ "step": 9295,
+ "train_runtime": 71789.6701,
+ "train_tokens_per_second": 5665.54
+ },
+ {
+ "epoch": 0.6261782924858605,
+ "grad_norm": 0.8295991455328212,
+ "learning_rate": 7.770591974928604e-06,
+ "loss": 0.3997042655944824,
+ "num_input_tokens_seen": 406958480,
+ "step": 9300,
+ "train_runtime": 71824.8324,
+ "train_tokens_per_second": 5665.986
+ },
+ {
+ "epoch": 0.6265149474818207,
+ "grad_norm": 0.7658676057217327,
+ "learning_rate": 7.768390553525034e-06,
+ "loss": 0.3705620527267456,
+ "num_input_tokens_seen": 407186440,
+ "step": 9305,
+ "train_runtime": 71865.0099,
+ "train_tokens_per_second": 5665.99
+ },
+ {
+ "epoch": 0.6268516024777808,
+ "grad_norm": 0.7085619378845062,
+ "learning_rate": 7.766188357949875e-06,
+ "loss": 0.3714077711105347,
+ "num_input_tokens_seen": 407407560,
+ "step": 9310,
+ "train_runtime": 71909.2798,
+ "train_tokens_per_second": 5665.577
+ },
+ {
+ "epoch": 0.627188257473741,
+ "grad_norm": 0.8341480094252892,
+ "learning_rate": 7.763985388818963e-06,
+ "loss": 0.36980538368225097,
+ "num_input_tokens_seen": 407628232,
+ "step": 9315,
+ "train_runtime": 71949.141,
+ "train_tokens_per_second": 5665.505
+ },
+ {
+ "epoch": 0.6275249124697011,
+ "grad_norm": 0.8636262815852889,
+ "learning_rate": 7.76178164674835e-06,
+ "loss": 0.3548977613449097,
+ "num_input_tokens_seen": 407838184,
+ "step": 9320,
+ "train_runtime": 71979.0743,
+ "train_tokens_per_second": 5666.066
+ },
+ {
+ "epoch": 0.6278615674656612,
+ "grad_norm": 0.7192127713238392,
+ "learning_rate": 7.759577132354305e-06,
+ "loss": 0.3763836622238159,
+ "num_input_tokens_seen": 408065528,
+ "step": 9325,
+ "train_runtime": 72013.1228,
+ "train_tokens_per_second": 5666.544
+ },
+ {
+ "epoch": 0.6281982224616214,
+ "grad_norm": 0.7987348676964026,
+ "learning_rate": 7.757371846253315e-06,
+ "loss": 0.3501241683959961,
+ "num_input_tokens_seen": 408275816,
+ "step": 9330,
+ "train_runtime": 72052.1024,
+ "train_tokens_per_second": 5666.397
+ },
+ {
+ "epoch": 0.6285348774575815,
+ "grad_norm": 0.7470954183186648,
+ "learning_rate": 7.755165789062083e-06,
+ "loss": 0.4022618293762207,
+ "num_input_tokens_seen": 408498408,
+ "step": 9335,
+ "train_runtime": 72090.7302,
+ "train_tokens_per_second": 5666.448
+ },
+ {
+ "epoch": 0.6288715324535417,
+ "grad_norm": 0.7151603552727446,
+ "learning_rate": 7.752958961397522e-06,
+ "loss": 0.3877957105636597,
+ "num_input_tokens_seen": 408733352,
+ "step": 9340,
+ "train_runtime": 72120.5104,
+ "train_tokens_per_second": 5667.366
+ },
+ {
+ "epoch": 0.6292081874495018,
+ "grad_norm": 0.7351213604942984,
+ "learning_rate": 7.750751363876765e-06,
+ "loss": 0.39035401344299314,
+ "num_input_tokens_seen": 408941920,
+ "step": 9345,
+ "train_runtime": 72164.8249,
+ "train_tokens_per_second": 5666.776
+ },
+ {
+ "epoch": 0.6295448424454619,
+ "grad_norm": 0.8453874832584365,
+ "learning_rate": 7.748542997117162e-06,
+ "loss": 0.3737621307373047,
+ "num_input_tokens_seen": 409162768,
+ "step": 9350,
+ "train_runtime": 72196.2605,
+ "train_tokens_per_second": 5667.368
+ },
+ {
+ "epoch": 0.6298814974414221,
+ "grad_norm": 0.7429854652224369,
+ "learning_rate": 7.746333861736275e-06,
+ "loss": 0.37030327320098877,
+ "num_input_tokens_seen": 409372800,
+ "step": 9355,
+ "train_runtime": 72230.736,
+ "train_tokens_per_second": 5667.571
+ },
+ {
+ "epoch": 0.6302181524373822,
+ "grad_norm": 0.8081848245845143,
+ "learning_rate": 7.74412395835188e-06,
+ "loss": 0.3631394386291504,
+ "num_input_tokens_seen": 409583128,
+ "step": 9360,
+ "train_runtime": 72272.2265,
+ "train_tokens_per_second": 5667.227
+ },
+ {
+ "epoch": 0.6305548074333424,
+ "grad_norm": 0.6576701373107874,
+ "learning_rate": 7.74191328758197e-06,
+ "loss": 0.37232241630554197,
+ "num_input_tokens_seen": 409794192,
+ "step": 9365,
+ "train_runtime": 72313.6306,
+ "train_tokens_per_second": 5666.901
+ },
+ {
+ "epoch": 0.6308914624293025,
+ "grad_norm": 0.735743623178803,
+ "learning_rate": 7.739701850044752e-06,
+ "loss": 0.4075340270996094,
+ "num_input_tokens_seen": 410020536,
+ "step": 9370,
+ "train_runtime": 72350.0803,
+ "train_tokens_per_second": 5667.175
+ },
+ {
+ "epoch": 0.6312281174252626,
+ "grad_norm": 0.8241305453975977,
+ "learning_rate": 7.73748964635865e-06,
+ "loss": 0.3718372583389282,
+ "num_input_tokens_seen": 410241072,
+ "step": 9375,
+ "train_runtime": 72387.3064,
+ "train_tokens_per_second": 5667.307
+ },
+ {
+ "epoch": 0.6315647724212228,
+ "grad_norm": 0.6824612940355289,
+ "learning_rate": 7.735276677142297e-06,
+ "loss": 0.35228054523468016,
+ "num_input_tokens_seen": 410443760,
+ "step": 9380,
+ "train_runtime": 72423.3738,
+ "train_tokens_per_second": 5667.283
+ },
+ {
+ "epoch": 0.6319014274171829,
+ "grad_norm": 0.7420817513182338,
+ "learning_rate": 7.733062943014542e-06,
+ "loss": 0.3992206573486328,
+ "num_input_tokens_seen": 410670656,
+ "step": 9385,
+ "train_runtime": 72463.1943,
+ "train_tokens_per_second": 5667.3
+ },
+ {
+ "epoch": 0.632238082413143,
+ "grad_norm": 0.7090881615954795,
+ "learning_rate": 7.730848444594452e-06,
+ "loss": 0.3573267698287964,
+ "num_input_tokens_seen": 410903448,
+ "step": 9390,
+ "train_runtime": 72497.2497,
+ "train_tokens_per_second": 5667.849
+ },
+ {
+ "epoch": 0.6325747374091032,
+ "grad_norm": 0.7269251761731511,
+ "learning_rate": 7.7286331825013e-06,
+ "loss": 0.37146763801574706,
+ "num_input_tokens_seen": 411126128,
+ "step": 9395,
+ "train_runtime": 72531.5005,
+ "train_tokens_per_second": 5668.242
+ },
+ {
+ "epoch": 0.6329113924050633,
+ "grad_norm": 0.8641784035147556,
+ "learning_rate": 7.72641715735458e-06,
+ "loss": 0.3655902624130249,
+ "num_input_tokens_seen": 411336720,
+ "step": 9400,
+ "train_runtime": 72578.3797,
+ "train_tokens_per_second": 5667.483
+ },
+ {
+ "epoch": 0.6332480474010235,
+ "grad_norm": 0.7283249670371048,
+ "learning_rate": 7.724200369773998e-06,
+ "loss": 0.38719449043273924,
+ "num_input_tokens_seen": 411563872,
+ "step": 9405,
+ "train_runtime": 72618.0679,
+ "train_tokens_per_second": 5667.513
+ },
+ {
+ "epoch": 0.6335847023969836,
+ "grad_norm": 0.6863677356579423,
+ "learning_rate": 7.721982820379467e-06,
+ "loss": 0.37484476566314695,
+ "num_input_tokens_seen": 411787952,
+ "step": 9410,
+ "train_runtime": 72652.1342,
+ "train_tokens_per_second": 5667.94
+ },
+ {
+ "epoch": 0.6339213573929438,
+ "grad_norm": 0.7975234970060876,
+ "learning_rate": 7.719764509791123e-06,
+ "loss": 0.3741923332214355,
+ "num_input_tokens_seen": 411975928,
+ "step": 9415,
+ "train_runtime": 72685.292,
+ "train_tokens_per_second": 5667.941
+ },
+ {
+ "epoch": 0.6342580123889039,
+ "grad_norm": 0.735648228811728,
+ "learning_rate": 7.717545438629303e-06,
+ "loss": 0.43533172607421877,
+ "num_input_tokens_seen": 412214368,
+ "step": 9420,
+ "train_runtime": 72721.9474,
+ "train_tokens_per_second": 5668.363
+ },
+ {
+ "epoch": 0.634594667384864,
+ "grad_norm": 0.6852408582111305,
+ "learning_rate": 7.71532560751457e-06,
+ "loss": 0.3732184410095215,
+ "num_input_tokens_seen": 412434712,
+ "step": 9425,
+ "train_runtime": 72757.9756,
+ "train_tokens_per_second": 5668.584
+ },
+ {
+ "epoch": 0.6349313223808242,
+ "grad_norm": 0.7315681955128508,
+ "learning_rate": 7.713105017067686e-06,
+ "loss": 0.38740081787109376,
+ "num_input_tokens_seen": 412668752,
+ "step": 9430,
+ "train_runtime": 72791.6587,
+ "train_tokens_per_second": 5669.176
+ },
+ {
+ "epoch": 0.6352679773767843,
+ "grad_norm": 0.840365118006204,
+ "learning_rate": 7.710883667909639e-06,
+ "loss": 0.37178239822387693,
+ "num_input_tokens_seen": 412886888,
+ "step": 9435,
+ "train_runtime": 72827.5502,
+ "train_tokens_per_second": 5669.378
+ },
+ {
+ "epoch": 0.6356046323727444,
+ "grad_norm": 0.7096620735032497,
+ "learning_rate": 7.708661560661617e-06,
+ "loss": 0.3967950105667114,
+ "num_input_tokens_seen": 413111704,
+ "step": 9440,
+ "train_runtime": 72863.4631,
+ "train_tokens_per_second": 5669.669
+ },
+ {
+ "epoch": 0.6359412873687046,
+ "grad_norm": 0.7646005775520524,
+ "learning_rate": 7.706438695945026e-06,
+ "loss": 0.39545307159423826,
+ "num_input_tokens_seen": 413330656,
+ "step": 9445,
+ "train_runtime": 72901.529,
+ "train_tokens_per_second": 5669.712
+ },
+ {
+ "epoch": 0.6362779423646647,
+ "grad_norm": 0.7774443776922321,
+ "learning_rate": 7.704215074381484e-06,
+ "loss": 0.3966789960861206,
+ "num_input_tokens_seen": 413537792,
+ "step": 9450,
+ "train_runtime": 72937.8522,
+ "train_tokens_per_second": 5669.728
+ },
+ {
+ "epoch": 0.6366145973606249,
+ "grad_norm": 0.7509437079381308,
+ "learning_rate": 7.70199069659282e-06,
+ "loss": 0.38933987617492677,
+ "num_input_tokens_seen": 413751760,
+ "step": 9455,
+ "train_runtime": 72973.609,
+ "train_tokens_per_second": 5669.882
+ },
+ {
+ "epoch": 0.636951252356585,
+ "grad_norm": 0.7325824670692431,
+ "learning_rate": 7.699765563201075e-06,
+ "loss": 0.3942718982696533,
+ "num_input_tokens_seen": 413978016,
+ "step": 9460,
+ "train_runtime": 73003.9087,
+ "train_tokens_per_second": 5670.628
+ },
+ {
+ "epoch": 0.6372879073525451,
+ "grad_norm": 1.211650286474132,
+ "learning_rate": 7.697539674828498e-06,
+ "loss": 0.36391506195068357,
+ "num_input_tokens_seen": 414211120,
+ "step": 9465,
+ "train_runtime": 73043.0392,
+ "train_tokens_per_second": 5670.782
+ },
+ {
+ "epoch": 0.6376245623485053,
+ "grad_norm": 0.7046659464816453,
+ "learning_rate": 7.695313032097554e-06,
+ "loss": 0.35442147254943845,
+ "num_input_tokens_seen": 414430048,
+ "step": 9470,
+ "train_runtime": 73079.8991,
+ "train_tokens_per_second": 5670.917
+ },
+ {
+ "epoch": 0.6379612173444654,
+ "grad_norm": 0.6947182191976057,
+ "learning_rate": 7.693085635630913e-06,
+ "loss": 0.37603862285614015,
+ "num_input_tokens_seen": 414632320,
+ "step": 9475,
+ "train_runtime": 73126.3077,
+ "train_tokens_per_second": 5670.084
+ },
+ {
+ "epoch": 0.6382978723404256,
+ "grad_norm": 0.7520528416471214,
+ "learning_rate": 7.690857486051463e-06,
+ "loss": 0.3694300174713135,
+ "num_input_tokens_seen": 414859816,
+ "step": 9480,
+ "train_runtime": 73166.4344,
+ "train_tokens_per_second": 5670.084
+ },
+ {
+ "epoch": 0.6386345273363857,
+ "grad_norm": 0.7317534743877758,
+ "learning_rate": 7.688628583982298e-06,
+ "loss": 0.3734642267227173,
+ "num_input_tokens_seen": 415071856,
+ "step": 9485,
+ "train_runtime": 73208.2652,
+ "train_tokens_per_second": 5669.74
+ },
+ {
+ "epoch": 0.6389711823323458,
+ "grad_norm": 0.7042485073751668,
+ "learning_rate": 7.68639893004672e-06,
+ "loss": 0.3597568035125732,
+ "num_input_tokens_seen": 415294888,
+ "step": 9490,
+ "train_runtime": 73249.0074,
+ "train_tokens_per_second": 5669.632
+ },
+ {
+ "epoch": 0.639307837328306,
+ "grad_norm": 0.7092737777153273,
+ "learning_rate": 7.684168524868253e-06,
+ "loss": 0.3566474437713623,
+ "num_input_tokens_seen": 415526408,
+ "step": 9495,
+ "train_runtime": 73288.7331,
+ "train_tokens_per_second": 5669.717
+ },
+ {
+ "epoch": 0.6396444923242661,
+ "grad_norm": 0.788876930336864,
+ "learning_rate": 7.681937369070612e-06,
+ "loss": 0.37108328342437746,
+ "num_input_tokens_seen": 415741912,
+ "step": 9500,
+ "train_runtime": 73321.3834,
+ "train_tokens_per_second": 5670.132
+ },
+ {
+ "epoch": 0.6399811473202263,
+ "grad_norm": 0.7482496645801756,
+ "learning_rate": 7.67970546327774e-06,
+ "loss": 0.3800150156021118,
+ "num_input_tokens_seen": 415962440,
+ "step": 9505,
+ "train_runtime": 73359.2213,
+ "train_tokens_per_second": 5670.213
+ },
+ {
+ "epoch": 0.6403178023161864,
+ "grad_norm": 0.7003781550580643,
+ "learning_rate": 7.67747280811378e-06,
+ "loss": 0.3659864902496338,
+ "num_input_tokens_seen": 416179208,
+ "step": 9510,
+ "train_runtime": 73399.2881,
+ "train_tokens_per_second": 5670.071
+ },
+ {
+ "epoch": 0.6406544573121465,
+ "grad_norm": 0.6823118163683989,
+ "learning_rate": 7.675239404203088e-06,
+ "loss": 0.35662126541137695,
+ "num_input_tokens_seen": 416394640,
+ "step": 9515,
+ "train_runtime": 73439.5055,
+ "train_tokens_per_second": 5669.9
+ },
+ {
+ "epoch": 0.6409911123081067,
+ "grad_norm": 0.7167030632794295,
+ "learning_rate": 7.673005252170226e-06,
+ "loss": 0.3727959156036377,
+ "num_input_tokens_seen": 416620784,
+ "step": 9520,
+ "train_runtime": 73484.0265,
+ "train_tokens_per_second": 5669.542
+ },
+ {
+ "epoch": 0.6413277673040668,
+ "grad_norm": 0.7593758397546905,
+ "learning_rate": 7.670770352639968e-06,
+ "loss": 0.3673388957977295,
+ "num_input_tokens_seen": 416826168,
+ "step": 9525,
+ "train_runtime": 73525.2289,
+ "train_tokens_per_second": 5669.158
+ },
+ {
+ "epoch": 0.641664422300027,
+ "grad_norm": 0.7362340230809692,
+ "learning_rate": 7.668534706237299e-06,
+ "loss": 0.37580122947692873,
+ "num_input_tokens_seen": 417047480,
+ "step": 9530,
+ "train_runtime": 73556.9783,
+ "train_tokens_per_second": 5669.72
+ },
+ {
+ "epoch": 0.6420010772959871,
+ "grad_norm": 0.7543938330712561,
+ "learning_rate": 7.666298313587409e-06,
+ "loss": 0.4170052528381348,
+ "num_input_tokens_seen": 417268504,
+ "step": 9535,
+ "train_runtime": 73597.5716,
+ "train_tokens_per_second": 5669.596
+ },
+ {
+ "epoch": 0.6423377322919472,
+ "grad_norm": 0.8008472981307811,
+ "learning_rate": 7.664061175315696e-06,
+ "loss": 0.3715662956237793,
+ "num_input_tokens_seen": 417464000,
+ "step": 9540,
+ "train_runtime": 73631.8047,
+ "train_tokens_per_second": 5669.615
+ },
+ {
+ "epoch": 0.6426743872879074,
+ "grad_norm": 0.7188017157103265,
+ "learning_rate": 7.661823292047771e-06,
+ "loss": 0.3764479160308838,
+ "num_input_tokens_seen": 417683920,
+ "step": 9545,
+ "train_runtime": 73672.7623,
+ "train_tokens_per_second": 5669.448
+ },
+ {
+ "epoch": 0.6430110422838675,
+ "grad_norm": 0.8895228408009356,
+ "learning_rate": 7.659584664409452e-06,
+ "loss": 0.3806633949279785,
+ "num_input_tokens_seen": 417910080,
+ "step": 9550,
+ "train_runtime": 73714.3818,
+ "train_tokens_per_second": 5669.315
+ },
+ {
+ "epoch": 0.6433476972798277,
+ "grad_norm": 0.7813418710468062,
+ "learning_rate": 7.65734529302676e-06,
+ "loss": 0.33327679634094237,
+ "num_input_tokens_seen": 418114864,
+ "step": 9555,
+ "train_runtime": 73751.4559,
+ "train_tokens_per_second": 5669.242
+ },
+ {
+ "epoch": 0.6436843522757878,
+ "grad_norm": 0.7440032604844485,
+ "learning_rate": 7.655105178525932e-06,
+ "loss": 0.36725008487701416,
+ "num_input_tokens_seen": 418331608,
+ "step": 9560,
+ "train_runtime": 73789.4942,
+ "train_tokens_per_second": 5669.257
+ },
+ {
+ "epoch": 0.6440210072717479,
+ "grad_norm": 0.7428051543591697,
+ "learning_rate": 7.652864321533406e-06,
+ "loss": 0.37874999046325686,
+ "num_input_tokens_seen": 418542736,
+ "step": 9565,
+ "train_runtime": 73827.1886,
+ "train_tokens_per_second": 5669.222
+ },
+ {
+ "epoch": 0.6443576622677081,
+ "grad_norm": 0.6906490664289745,
+ "learning_rate": 7.650622722675833e-06,
+ "loss": 0.39217414855957033,
+ "num_input_tokens_seen": 418769104,
+ "step": 9570,
+ "train_runtime": 73869.6555,
+ "train_tokens_per_second": 5669.027
+ },
+ {
+ "epoch": 0.6446943172636682,
+ "grad_norm": 0.7836457733135137,
+ "learning_rate": 7.648380382580067e-06,
+ "loss": 0.3911130428314209,
+ "num_input_tokens_seen": 418983272,
+ "step": 9575,
+ "train_runtime": 73911.9877,
+ "train_tokens_per_second": 5668.678
+ },
+ {
+ "epoch": 0.6450309722596284,
+ "grad_norm": 0.7191438751807192,
+ "learning_rate": 7.646137301873173e-06,
+ "loss": 0.3877583026885986,
+ "num_input_tokens_seen": 419193152,
+ "step": 9580,
+ "train_runtime": 73952.718,
+ "train_tokens_per_second": 5668.394
+ },
+ {
+ "epoch": 0.6453676272555885,
+ "grad_norm": 0.8722963532886264,
+ "learning_rate": 7.643893481182421e-06,
+ "loss": 0.35512332916259765,
+ "num_input_tokens_seen": 419415144,
+ "step": 9585,
+ "train_runtime": 73990.0513,
+ "train_tokens_per_second": 5668.534
+ },
+ {
+ "epoch": 0.6457042822515486,
+ "grad_norm": 0.7893095386452317,
+ "learning_rate": 7.641648921135287e-06,
+ "loss": 0.4030734062194824,
+ "num_input_tokens_seen": 419642840,
+ "step": 9590,
+ "train_runtime": 74032.6519,
+ "train_tokens_per_second": 5668.348
+ },
+ {
+ "epoch": 0.6460409372475088,
+ "grad_norm": 0.8691991766270082,
+ "learning_rate": 7.639403622359455e-06,
+ "loss": 0.38401873111724855,
+ "num_input_tokens_seen": 419851960,
+ "step": 9595,
+ "train_runtime": 74068.9663,
+ "train_tokens_per_second": 5668.392
+ },
+ {
+ "epoch": 0.6463775922434689,
+ "grad_norm": 0.7669041161649485,
+ "learning_rate": 7.637157585482817e-06,
+ "loss": 0.3698650598526001,
+ "num_input_tokens_seen": 420069368,
+ "step": 9600,
+ "train_runtime": 74113.2241,
+ "train_tokens_per_second": 5667.941
+ },
+ {
+ "epoch": 0.646714247239429,
+ "grad_norm": 0.6764016385660067,
+ "learning_rate": 7.63491081113347e-06,
+ "loss": 0.3979948043823242,
+ "num_input_tokens_seen": 420301264,
+ "step": 9605,
+ "train_runtime": 74149.7053,
+ "train_tokens_per_second": 5668.28
+ },
+ {
+ "epoch": 0.6470509022353892,
+ "grad_norm": 0.7632828915951216,
+ "learning_rate": 7.632663299939718e-06,
+ "loss": 0.4009084701538086,
+ "num_input_tokens_seen": 420528480,
+ "step": 9610,
+ "train_runtime": 74191.7204,
+ "train_tokens_per_second": 5668.132
+ },
+ {
+ "epoch": 0.6473875572313493,
+ "grad_norm": 0.7759849429505897,
+ "learning_rate": 7.630415052530066e-06,
+ "loss": 0.37017059326171875,
+ "num_input_tokens_seen": 420744352,
+ "step": 9615,
+ "train_runtime": 74228.9204,
+ "train_tokens_per_second": 5668.2
+ },
+ {
+ "epoch": 0.6477242122273095,
+ "grad_norm": 0.7521586068543743,
+ "learning_rate": 7.628166069533235e-06,
+ "loss": 0.3800983905792236,
+ "num_input_tokens_seen": 420951552,
+ "step": 9620,
+ "train_runtime": 74264.9231,
+ "train_tokens_per_second": 5668.242
+ },
+ {
+ "epoch": 0.6480608672232696,
+ "grad_norm": 0.8746614520646498,
+ "learning_rate": 7.625916351578141e-06,
+ "loss": 0.4101721286773682,
+ "num_input_tokens_seen": 421168904,
+ "step": 9625,
+ "train_runtime": 74302.0251,
+ "train_tokens_per_second": 5668.337
+ },
+ {
+ "epoch": 0.6483975222192297,
+ "grad_norm": 0.645310594782184,
+ "learning_rate": 7.623665899293914e-06,
+ "loss": 0.35078365802764894,
+ "num_input_tokens_seen": 421386640,
+ "step": 9630,
+ "train_runtime": 74343.6656,
+ "train_tokens_per_second": 5668.091
+ },
+ {
+ "epoch": 0.6487341772151899,
+ "grad_norm": 0.709824008157462,
+ "learning_rate": 7.621414713309885e-06,
+ "loss": 0.3464890241622925,
+ "num_input_tokens_seen": 421604504,
+ "step": 9635,
+ "train_runtime": 74380.7993,
+ "train_tokens_per_second": 5668.19
+ },
+ {
+ "epoch": 0.64907083221115,
+ "grad_norm": 0.6987475287759364,
+ "learning_rate": 7.61916279425559e-06,
+ "loss": 0.38260612487792967,
+ "num_input_tokens_seen": 421833672,
+ "step": 9640,
+ "train_runtime": 74422.4312,
+ "train_tokens_per_second": 5668.099
+ },
+ {
+ "epoch": 0.6494074872071102,
+ "grad_norm": 0.727875668145333,
+ "learning_rate": 7.6169101427607716e-06,
+ "loss": 0.38509573936462405,
+ "num_input_tokens_seen": 422067496,
+ "step": 9645,
+ "train_runtime": 74456.6368,
+ "train_tokens_per_second": 5668.635
+ },
+ {
+ "epoch": 0.6497441422030703,
+ "grad_norm": 0.7549567194100304,
+ "learning_rate": 7.614656759455375e-06,
+ "loss": 0.38222179412841795,
+ "num_input_tokens_seen": 422289144,
+ "step": 9650,
+ "train_runtime": 74495.8893,
+ "train_tokens_per_second": 5668.623
+ },
+ {
+ "epoch": 0.6500807971990304,
+ "grad_norm": 0.8363753913347584,
+ "learning_rate": 7.612402644969555e-06,
+ "loss": 0.38720729351043703,
+ "num_input_tokens_seen": 422500848,
+ "step": 9655,
+ "train_runtime": 74537.09,
+ "train_tokens_per_second": 5668.33
+ },
+ {
+ "epoch": 0.6504174521949906,
+ "grad_norm": 0.718216642211442,
+ "learning_rate": 7.6101477999336646e-06,
+ "loss": 0.37707133293151857,
+ "num_input_tokens_seen": 422718672,
+ "step": 9660,
+ "train_runtime": 74584.7849,
+ "train_tokens_per_second": 5667.626
+ },
+ {
+ "epoch": 0.6507541071909507,
+ "grad_norm": 0.8872821374211818,
+ "learning_rate": 7.607892224978266e-06,
+ "loss": 0.3739673376083374,
+ "num_input_tokens_seen": 422929480,
+ "step": 9665,
+ "train_runtime": 74620.552,
+ "train_tokens_per_second": 5667.735
+ },
+ {
+ "epoch": 0.6510907621869109,
+ "grad_norm": 0.691277480453269,
+ "learning_rate": 7.605635920734122e-06,
+ "loss": 0.3418617010116577,
+ "num_input_tokens_seen": 423142128,
+ "step": 9670,
+ "train_runtime": 74655.9636,
+ "train_tokens_per_second": 5667.895
+ },
+ {
+ "epoch": 0.651427417182871,
+ "grad_norm": 0.9025153211455825,
+ "learning_rate": 7.603378887832202e-06,
+ "loss": 0.3855189561843872,
+ "num_input_tokens_seen": 423349056,
+ "step": 9675,
+ "train_runtime": 74695.4667,
+ "train_tokens_per_second": 5667.667
+ },
+ {
+ "epoch": 0.6517640721788311,
+ "grad_norm": 0.7073570329177757,
+ "learning_rate": 7.601121126903676e-06,
+ "loss": 0.3803261280059814,
+ "num_input_tokens_seen": 423573240,
+ "step": 9680,
+ "train_runtime": 74738.2061,
+ "train_tokens_per_second": 5667.426
+ },
+ {
+ "epoch": 0.6521007271747913,
+ "grad_norm": 0.7609618594052814,
+ "learning_rate": 7.598862638579922e-06,
+ "loss": 0.3738381385803223,
+ "num_input_tokens_seen": 423809440,
+ "step": 9685,
+ "train_runtime": 74781.8146,
+ "train_tokens_per_second": 5667.279
+ },
+ {
+ "epoch": 0.6524373821707514,
+ "grad_norm": 0.6698743309851123,
+ "learning_rate": 7.5966034234925205e-06,
+ "loss": 0.3761345624923706,
+ "num_input_tokens_seen": 424041528,
+ "step": 9690,
+ "train_runtime": 74821.3132,
+ "train_tokens_per_second": 5667.39
+ },
+ {
+ "epoch": 0.6527740371667116,
+ "grad_norm": 0.7414693193437045,
+ "learning_rate": 7.59434348227325e-06,
+ "loss": 0.43105735778808596,
+ "num_input_tokens_seen": 424268184,
+ "step": 9695,
+ "train_runtime": 74855.3245,
+ "train_tokens_per_second": 5667.842
+ },
+ {
+ "epoch": 0.6531106921626717,
+ "grad_norm": 0.8237582570303243,
+ "learning_rate": 7.592082815554098e-06,
+ "loss": 0.40675368309021,
+ "num_input_tokens_seen": 424492616,
+ "step": 9700,
+ "train_runtime": 74894.094,
+ "train_tokens_per_second": 5667.905
+ },
+ {
+ "epoch": 0.6534473471586318,
+ "grad_norm": 0.9212376319324829,
+ "learning_rate": 7.58982142396725e-06,
+ "loss": 0.419390869140625,
+ "num_input_tokens_seen": 424686848,
+ "step": 9705,
+ "train_runtime": 74939.3759,
+ "train_tokens_per_second": 5667.072
+ },
+ {
+ "epoch": 0.653784002154592,
+ "grad_norm": 0.7691414977900423,
+ "learning_rate": 7.5875593081451e-06,
+ "loss": 0.3695266008377075,
+ "num_input_tokens_seen": 424913632,
+ "step": 9710,
+ "train_runtime": 74985.8205,
+ "train_tokens_per_second": 5666.586
+ },
+ {
+ "epoch": 0.6541206571505521,
+ "grad_norm": 0.7263975698689714,
+ "learning_rate": 7.585296468720243e-06,
+ "loss": 0.3619475126266479,
+ "num_input_tokens_seen": 425139040,
+ "step": 9715,
+ "train_runtime": 75026.5575,
+ "train_tokens_per_second": 5666.514
+ },
+ {
+ "epoch": 0.6544573121465123,
+ "grad_norm": 0.712706963298683,
+ "learning_rate": 7.5830329063254705e-06,
+ "loss": 0.38277277946472166,
+ "num_input_tokens_seen": 425352824,
+ "step": 9720,
+ "train_runtime": 75065.2226,
+ "train_tokens_per_second": 5666.443
+ },
+ {
+ "epoch": 0.6547939671424724,
+ "grad_norm": 0.6764513937336157,
+ "learning_rate": 7.5807686215937835e-06,
+ "loss": 0.38765199184417726,
+ "num_input_tokens_seen": 425565784,
+ "step": 9725,
+ "train_runtime": 75105.3744,
+ "train_tokens_per_second": 5666.249
+ },
+ {
+ "epoch": 0.6551306221384325,
+ "grad_norm": 0.7196173883927505,
+ "learning_rate": 7.578503615158379e-06,
+ "loss": 0.3398409366607666,
+ "num_input_tokens_seen": 425778128,
+ "step": 9730,
+ "train_runtime": 75149.2065,
+ "train_tokens_per_second": 5665.77
+ },
+ {
+ "epoch": 0.6554672771343927,
+ "grad_norm": 0.646960134820036,
+ "learning_rate": 7.576237887652662e-06,
+ "loss": 0.38698830604553225,
+ "num_input_tokens_seen": 426007040,
+ "step": 9735,
+ "train_runtime": 75188.2698,
+ "train_tokens_per_second": 5665.871
+ },
+ {
+ "epoch": 0.6558039321303528,
+ "grad_norm": 0.8063318146524463,
+ "learning_rate": 7.573971439710235e-06,
+ "loss": 0.34668645858764646,
+ "num_input_tokens_seen": 426223488,
+ "step": 9740,
+ "train_runtime": 75225.1526,
+ "train_tokens_per_second": 5665.97
+ },
+ {
+ "epoch": 0.656140587126313,
+ "grad_norm": 0.8258091202754753,
+ "learning_rate": 7.571704271964902e-06,
+ "loss": 0.3675506591796875,
+ "num_input_tokens_seen": 426433528,
+ "step": 9745,
+ "train_runtime": 75260.6499,
+ "train_tokens_per_second": 5666.089
+ },
+ {
+ "epoch": 0.6564772421222731,
+ "grad_norm": 0.8027675726296949,
+ "learning_rate": 7.56943638505067e-06,
+ "loss": 0.3636334419250488,
+ "num_input_tokens_seen": 426658696,
+ "step": 9750,
+ "train_runtime": 75296.5611,
+ "train_tokens_per_second": 5666.377
+ },
+ {
+ "epoch": 0.6568138971182332,
+ "grad_norm": 0.7190477185688757,
+ "learning_rate": 7.567167779601746e-06,
+ "loss": 0.39123508930206297,
+ "num_input_tokens_seen": 426886792,
+ "step": 9755,
+ "train_runtime": 75338.0495,
+ "train_tokens_per_second": 5666.284
+ },
+ {
+ "epoch": 0.6571505521141934,
+ "grad_norm": 0.7906692823017727,
+ "learning_rate": 7.56489845625254e-06,
+ "loss": 0.413257360458374,
+ "num_input_tokens_seen": 427104192,
+ "step": 9760,
+ "train_runtime": 75386.2254,
+ "train_tokens_per_second": 5665.547
+ },
+ {
+ "epoch": 0.6574872071101535,
+ "grad_norm": 0.7301115269654299,
+ "learning_rate": 7.562628415637659e-06,
+ "loss": 0.369996976852417,
+ "num_input_tokens_seen": 427312504,
+ "step": 9765,
+ "train_runtime": 75426.5577,
+ "train_tokens_per_second": 5665.279
+ },
+ {
+ "epoch": 0.6578238621061137,
+ "grad_norm": 0.7224806256488935,
+ "learning_rate": 7.560357658391913e-06,
+ "loss": 0.4071043491363525,
+ "num_input_tokens_seen": 427533792,
+ "step": 9770,
+ "train_runtime": 75465.4869,
+ "train_tokens_per_second": 5665.289
+ },
+ {
+ "epoch": 0.6581605171020738,
+ "grad_norm": 0.7268098674561758,
+ "learning_rate": 7.558086185150311e-06,
+ "loss": 0.3804465770721436,
+ "num_input_tokens_seen": 427747608,
+ "step": 9775,
+ "train_runtime": 75507.0805,
+ "train_tokens_per_second": 5665.0
+ },
+ {
+ "epoch": 0.6584971720980339,
+ "grad_norm": 0.7850729912855515,
+ "learning_rate": 7.555813996548065e-06,
+ "loss": 0.40397090911865235,
+ "num_input_tokens_seen": 427972008,
+ "step": 9780,
+ "train_runtime": 75543.708,
+ "train_tokens_per_second": 5665.224
+ },
+ {
+ "epoch": 0.6588338270939941,
+ "grad_norm": 0.7662329566093665,
+ "learning_rate": 7.553541093220586e-06,
+ "loss": 0.3842537641525269,
+ "num_input_tokens_seen": 428186312,
+ "step": 9785,
+ "train_runtime": 75579.4178,
+ "train_tokens_per_second": 5665.383
+ },
+ {
+ "epoch": 0.6591704820899542,
+ "grad_norm": 0.70627363493871,
+ "learning_rate": 7.551267475803481e-06,
+ "loss": 0.37559702396392824,
+ "num_input_tokens_seen": 428413952,
+ "step": 9790,
+ "train_runtime": 75619.0124,
+ "train_tokens_per_second": 5665.426
+ },
+ {
+ "epoch": 0.6595071370859144,
+ "grad_norm": 0.7354101668176468,
+ "learning_rate": 7.5489931449325635e-06,
+ "loss": 0.3736457347869873,
+ "num_input_tokens_seen": 428630728,
+ "step": 9795,
+ "train_runtime": 75658.6215,
+ "train_tokens_per_second": 5665.326
+ },
+ {
+ "epoch": 0.6598437920818745,
+ "grad_norm": 0.7892531883838505,
+ "learning_rate": 7.546718101243842e-06,
+ "loss": 0.39138045310974123,
+ "num_input_tokens_seen": 428842760,
+ "step": 9800,
+ "train_runtime": 75693.8787,
+ "train_tokens_per_second": 5665.488
+ },
+ {
+ "epoch": 0.6601804470778346,
+ "grad_norm": 0.8775677410278766,
+ "learning_rate": 7.544442345373522e-06,
+ "loss": 0.3698113918304443,
+ "num_input_tokens_seen": 429051072,
+ "step": 9805,
+ "train_runtime": 75734.5916,
+ "train_tokens_per_second": 5665.193
+ },
+ {
+ "epoch": 0.6605171020737948,
+ "grad_norm": 0.6944135512445987,
+ "learning_rate": 7.542165877958014e-06,
+ "loss": 0.3717064380645752,
+ "num_input_tokens_seen": 429266232,
+ "step": 9810,
+ "train_runtime": 75771.7628,
+ "train_tokens_per_second": 5665.253
+ },
+ {
+ "epoch": 0.6608537570697549,
+ "grad_norm": 0.7361334690779209,
+ "learning_rate": 7.539888699633926e-06,
+ "loss": 0.38160529136657717,
+ "num_input_tokens_seen": 429495032,
+ "step": 9815,
+ "train_runtime": 75808.8176,
+ "train_tokens_per_second": 5665.502
+ },
+ {
+ "epoch": 0.661190412065715,
+ "grad_norm": 0.8126000202476551,
+ "learning_rate": 7.5376108110380606e-06,
+ "loss": 0.34899187088012695,
+ "num_input_tokens_seen": 429723384,
+ "step": 9820,
+ "train_runtime": 75847.4225,
+ "train_tokens_per_second": 5665.629
+ },
+ {
+ "epoch": 0.6615270670616752,
+ "grad_norm": 0.6957748548678448,
+ "learning_rate": 7.535332212807426e-06,
+ "loss": 0.3428369998931885,
+ "num_input_tokens_seen": 429924744,
+ "step": 9825,
+ "train_runtime": 75886.3897,
+ "train_tokens_per_second": 5665.374
+ },
+ {
+ "epoch": 0.6618637220576353,
+ "grad_norm": 0.7233083648231398,
+ "learning_rate": 7.53305290557922e-06,
+ "loss": 0.39725618362426757,
+ "num_input_tokens_seen": 430154928,
+ "step": 9830,
+ "train_runtime": 75925.3068,
+ "train_tokens_per_second": 5665.501
+ },
+ {
+ "epoch": 0.6622003770535955,
+ "grad_norm": 0.7049996354735961,
+ "learning_rate": 7.530772889990848e-06,
+ "loss": 0.3740149974822998,
+ "num_input_tokens_seen": 430386800,
+ "step": 9835,
+ "train_runtime": 75959.5503,
+ "train_tokens_per_second": 5666.0
+ },
+ {
+ "epoch": 0.6625370320495556,
+ "grad_norm": 0.6720383481048683,
+ "learning_rate": 7.528492166679907e-06,
+ "loss": 0.35007429122924805,
+ "num_input_tokens_seen": 430616592,
+ "step": 9840,
+ "train_runtime": 76001.567,
+ "train_tokens_per_second": 5665.891
+ },
+ {
+ "epoch": 0.6628736870455157,
+ "grad_norm": 0.6809307882652357,
+ "learning_rate": 7.526210736284194e-06,
+ "loss": 0.38547792434692385,
+ "num_input_tokens_seen": 430820368,
+ "step": 9845,
+ "train_runtime": 76036.0252,
+ "train_tokens_per_second": 5666.003
+ },
+ {
+ "epoch": 0.6632103420414759,
+ "grad_norm": 0.7241468747622457,
+ "learning_rate": 7.523928599441706e-06,
+ "loss": 0.3851293087005615,
+ "num_input_tokens_seen": 431045232,
+ "step": 9850,
+ "train_runtime": 76076.642,
+ "train_tokens_per_second": 5665.934
+ },
+ {
+ "epoch": 0.663546997037436,
+ "grad_norm": 0.8065697031009399,
+ "learning_rate": 7.5216457567906295e-06,
+ "loss": 0.3921393632888794,
+ "num_input_tokens_seen": 431272728,
+ "step": 9855,
+ "train_runtime": 76119.1303,
+ "train_tokens_per_second": 5665.76
+ },
+ {
+ "epoch": 0.6638836520333962,
+ "grad_norm": 0.7363581812525671,
+ "learning_rate": 7.519362208969362e-06,
+ "loss": 0.3589025974273682,
+ "num_input_tokens_seen": 431497576,
+ "step": 9860,
+ "train_runtime": 76164.9303,
+ "train_tokens_per_second": 5665.305
+ },
+ {
+ "epoch": 0.6642203070293563,
+ "grad_norm": 0.7028175355387544,
+ "learning_rate": 7.5170779566164855e-06,
+ "loss": 0.39649462699890137,
+ "num_input_tokens_seen": 431716768,
+ "step": 9865,
+ "train_runtime": 76207.7038,
+ "train_tokens_per_second": 5665.002
+ },
+ {
+ "epoch": 0.6645569620253164,
+ "grad_norm": 0.8083377822452007,
+ "learning_rate": 7.514793000370783e-06,
+ "loss": 0.39783761501312254,
+ "num_input_tokens_seen": 431927680,
+ "step": 9870,
+ "train_runtime": 76252.3393,
+ "train_tokens_per_second": 5664.452
+ },
+ {
+ "epoch": 0.6648936170212766,
+ "grad_norm": 0.7361462900334614,
+ "learning_rate": 7.512507340871239e-06,
+ "loss": 0.3664602756500244,
+ "num_input_tokens_seen": 432144784,
+ "step": 9875,
+ "train_runtime": 76293.2197,
+ "train_tokens_per_second": 5664.262
+ },
+ {
+ "epoch": 0.6652302720172367,
+ "grad_norm": 0.7771322596278541,
+ "learning_rate": 7.510220978757028e-06,
+ "loss": 0.39647390842437746,
+ "num_input_tokens_seen": 432361504,
+ "step": 9880,
+ "train_runtime": 76335.1474,
+ "train_tokens_per_second": 5663.99
+ },
+ {
+ "epoch": 0.6655669270131969,
+ "grad_norm": 0.9351257512839886,
+ "learning_rate": 7.507933914667524e-06,
+ "loss": 0.4201164722442627,
+ "num_input_tokens_seen": 432586992,
+ "step": 9885,
+ "train_runtime": 76375.8314,
+ "train_tokens_per_second": 5663.925
+ },
+ {
+ "epoch": 0.665903582009157,
+ "grad_norm": 0.7531976870702812,
+ "learning_rate": 7.505646149242299e-06,
+ "loss": 0.37460851669311523,
+ "num_input_tokens_seen": 432809080,
+ "step": 9890,
+ "train_runtime": 76415.9118,
+ "train_tokens_per_second": 5663.861
+ },
+ {
+ "epoch": 0.6662402370051171,
+ "grad_norm": 0.719734237163997,
+ "learning_rate": 7.503357683121119e-06,
+ "loss": 0.34871246814727785,
+ "num_input_tokens_seen": 433035120,
+ "step": 9895,
+ "train_runtime": 76462.7297,
+ "train_tokens_per_second": 5663.349
+ },
+ {
+ "epoch": 0.6665768920010773,
+ "grad_norm": 0.7130561120719486,
+ "learning_rate": 7.501068516943944e-06,
+ "loss": 0.3231820583343506,
+ "num_input_tokens_seen": 433248272,
+ "step": 9900,
+ "train_runtime": 76502.2615,
+ "train_tokens_per_second": 5663.209
+ },
+ {
+ "epoch": 0.6669135469970374,
+ "grad_norm": 0.6874890275442226,
+ "learning_rate": 7.498778651350934e-06,
+ "loss": 0.3865694284439087,
+ "num_input_tokens_seen": 433472920,
+ "step": 9905,
+ "train_runtime": 76539.0929,
+ "train_tokens_per_second": 5663.419
+ },
+ {
+ "epoch": 0.6672502019929976,
+ "grad_norm": 0.6837593508656135,
+ "learning_rate": 7.496488086982443e-06,
+ "loss": 0.3751338720321655,
+ "num_input_tokens_seen": 433694440,
+ "step": 9910,
+ "train_runtime": 76582.0454,
+ "train_tokens_per_second": 5663.135
+ },
+ {
+ "epoch": 0.6675868569889577,
+ "grad_norm": 0.834960347653905,
+ "learning_rate": 7.494196824479019e-06,
+ "loss": 0.37354316711425783,
+ "num_input_tokens_seen": 433909064,
+ "step": 9915,
+ "train_runtime": 76619.6521,
+ "train_tokens_per_second": 5663.156
+ },
+ {
+ "epoch": 0.6679235119849178,
+ "grad_norm": 0.6905943083082271,
+ "learning_rate": 7.491904864481406e-06,
+ "loss": 0.3818557024002075,
+ "num_input_tokens_seen": 434122768,
+ "step": 9920,
+ "train_runtime": 76661.7307,
+ "train_tokens_per_second": 5662.835
+ },
+ {
+ "epoch": 0.668260166980878,
+ "grad_norm": 0.6608540738794552,
+ "learning_rate": 7.489612207630543e-06,
+ "loss": 0.34175124168396,
+ "num_input_tokens_seen": 434354528,
+ "step": 9925,
+ "train_runtime": 76695.5171,
+ "train_tokens_per_second": 5663.363
+ },
+ {
+ "epoch": 0.6685968219768381,
+ "grad_norm": 0.747448764956391,
+ "learning_rate": 7.487318854567566e-06,
+ "loss": 0.37988905906677245,
+ "num_input_tokens_seen": 434564096,
+ "step": 9930,
+ "train_runtime": 76729.4177,
+ "train_tokens_per_second": 5663.592
+ },
+ {
+ "epoch": 0.6689334769727983,
+ "grad_norm": 0.7551042289550628,
+ "learning_rate": 7.485024805933803e-06,
+ "loss": 0.3599559307098389,
+ "num_input_tokens_seen": 434767584,
+ "step": 9935,
+ "train_runtime": 76765.8819,
+ "train_tokens_per_second": 5663.552
+ },
+ {
+ "epoch": 0.6692701319687584,
+ "grad_norm": 0.6631253047912142,
+ "learning_rate": 7.482730062370776e-06,
+ "loss": 0.3699958086013794,
+ "num_input_tokens_seen": 434981392,
+ "step": 9940,
+ "train_runtime": 76810.3453,
+ "train_tokens_per_second": 5663.057
+ },
+ {
+ "epoch": 0.6696067869647185,
+ "grad_norm": 0.7419282684235264,
+ "learning_rate": 7.480434624520205e-06,
+ "loss": 0.3593663215637207,
+ "num_input_tokens_seen": 435217648,
+ "step": 9945,
+ "train_runtime": 76843.0552,
+ "train_tokens_per_second": 5663.721
+ },
+ {
+ "epoch": 0.6699434419606787,
+ "grad_norm": 0.7057608866633803,
+ "learning_rate": 7.478138493024001e-06,
+ "loss": 0.3879536151885986,
+ "num_input_tokens_seen": 435444608,
+ "step": 9950,
+ "train_runtime": 76882.5546,
+ "train_tokens_per_second": 5663.764
+ },
+ {
+ "epoch": 0.6702800969566388,
+ "grad_norm": 0.740329239724491,
+ "learning_rate": 7.475841668524268e-06,
+ "loss": 0.34809188842773436,
+ "num_input_tokens_seen": 435656280,
+ "step": 9955,
+ "train_runtime": 76923.9627,
+ "train_tokens_per_second": 5663.466
+ },
+ {
+ "epoch": 0.670616751952599,
+ "grad_norm": 0.8121045332443452,
+ "learning_rate": 7.473544151663309e-06,
+ "loss": 0.4048673152923584,
+ "num_input_tokens_seen": 435875456,
+ "step": 9960,
+ "train_runtime": 76958.8069,
+ "train_tokens_per_second": 5663.75
+ },
+ {
+ "epoch": 0.6709534069485591,
+ "grad_norm": 0.8149107981141157,
+ "learning_rate": 7.471245943083615e-06,
+ "loss": 0.3677277326583862,
+ "num_input_tokens_seen": 436097568,
+ "step": 9965,
+ "train_runtime": 76998.2656,
+ "train_tokens_per_second": 5663.732
+ },
+ {
+ "epoch": 0.6712900619445192,
+ "grad_norm": 0.7104981842358721,
+ "learning_rate": 7.468947043427873e-06,
+ "loss": 0.3423744201660156,
+ "num_input_tokens_seen": 436314872,
+ "step": 9970,
+ "train_runtime": 77032.0267,
+ "train_tokens_per_second": 5664.071
+ },
+ {
+ "epoch": 0.6716267169404794,
+ "grad_norm": 0.8060621663622469,
+ "learning_rate": 7.466647453338965e-06,
+ "loss": 0.38165833950042727,
+ "num_input_tokens_seen": 436524088,
+ "step": 9975,
+ "train_runtime": 77070.1105,
+ "train_tokens_per_second": 5663.987
+ },
+ {
+ "epoch": 0.6719633719364395,
+ "grad_norm": 0.6817255880743095,
+ "learning_rate": 7.464347173459958e-06,
+ "loss": 0.3900636911392212,
+ "num_input_tokens_seen": 436763792,
+ "step": 9980,
+ "train_runtime": 77108.4939,
+ "train_tokens_per_second": 5664.276
+ },
+ {
+ "epoch": 0.6723000269323997,
+ "grad_norm": 0.5849966120883113,
+ "learning_rate": 7.462046204434126e-06,
+ "loss": 0.35174126625061036,
+ "num_input_tokens_seen": 437001592,
+ "step": 9985,
+ "train_runtime": 77143.495,
+ "train_tokens_per_second": 5664.789
+ },
+ {
+ "epoch": 0.6726366819283598,
+ "grad_norm": 0.819711833914903,
+ "learning_rate": 7.459744546904922e-06,
+ "loss": 0.37253947257995607,
+ "num_input_tokens_seen": 437217472,
+ "step": 9990,
+ "train_runtime": 77177.4484,
+ "train_tokens_per_second": 5665.094
+ },
+ {
+ "epoch": 0.6729733369243199,
+ "grad_norm": 0.7103670160362342,
+ "learning_rate": 7.457442201516001e-06,
+ "loss": 0.34635467529296876,
+ "num_input_tokens_seen": 437432256,
+ "step": 9995,
+ "train_runtime": 77222.7561,
+ "train_tokens_per_second": 5664.551
+ },
+ {
+ "epoch": 0.6733099919202801,
+ "grad_norm": 0.7772791959365979,
+ "learning_rate": 7.455139168911203e-06,
+ "loss": 0.3714813232421875,
+ "num_input_tokens_seen": 437632184,
+ "step": 10000,
+ "train_runtime": 77259.692,
+ "train_tokens_per_second": 5664.431
+ },
+ {
+ "epoch": 0.6736466469162402,
+ "grad_norm": 0.8663644820452563,
+ "learning_rate": 7.452835449734564e-06,
+ "loss": 0.37085671424865724,
+ "num_input_tokens_seen": 437852448,
+ "step": 10005,
+ "train_runtime": 77297.557,
+ "train_tokens_per_second": 5664.506
+ },
+ {
+ "epoch": 0.6739833019122003,
+ "grad_norm": 0.6628154115799231,
+ "learning_rate": 7.450531044630315e-06,
+ "loss": 0.35476658344268797,
+ "num_input_tokens_seen": 438072720,
+ "step": 10010,
+ "train_runtime": 77330.8077,
+ "train_tokens_per_second": 5664.918
+ },
+ {
+ "epoch": 0.6743199569081605,
+ "grad_norm": 0.7435766802454172,
+ "learning_rate": 7.448225954242874e-06,
+ "loss": 0.3909281253814697,
+ "num_input_tokens_seen": 438300752,
+ "step": 10015,
+ "train_runtime": 77372.9883,
+ "train_tokens_per_second": 5664.777
+ },
+ {
+ "epoch": 0.6746566119041206,
+ "grad_norm": 0.7076966807661021,
+ "learning_rate": 7.44592017921685e-06,
+ "loss": 0.37700977325439455,
+ "num_input_tokens_seen": 438521952,
+ "step": 10020,
+ "train_runtime": 77408.1985,
+ "train_tokens_per_second": 5665.058
+ },
+ {
+ "epoch": 0.6749932669000808,
+ "grad_norm": 0.7283120289124543,
+ "learning_rate": 7.443613720197048e-06,
+ "loss": 0.39572868347167967,
+ "num_input_tokens_seen": 438742144,
+ "step": 10025,
+ "train_runtime": 77451.6366,
+ "train_tokens_per_second": 5664.724
+ },
+ {
+ "epoch": 0.6753299218960409,
+ "grad_norm": 0.8184114176248545,
+ "learning_rate": 7.44130657782846e-06,
+ "loss": 0.35796802043914794,
+ "num_input_tokens_seen": 438946944,
+ "step": 10030,
+ "train_runtime": 77487.5196,
+ "train_tokens_per_second": 5664.744
+ },
+ {
+ "epoch": 0.675666576892001,
+ "grad_norm": 0.7168914852518904,
+ "learning_rate": 7.438998752756274e-06,
+ "loss": 0.3807486057281494,
+ "num_input_tokens_seen": 439164968,
+ "step": 10035,
+ "train_runtime": 77530.1548,
+ "train_tokens_per_second": 5664.441
+ },
+ {
+ "epoch": 0.6760032318879612,
+ "grad_norm": 0.7003642669524813,
+ "learning_rate": 7.436690245625864e-06,
+ "loss": 0.36953206062316896,
+ "num_input_tokens_seen": 439382032,
+ "step": 10040,
+ "train_runtime": 77568.57,
+ "train_tokens_per_second": 5664.434
+ },
+ {
+ "epoch": 0.6763398868839213,
+ "grad_norm": 0.8332693301417188,
+ "learning_rate": 7.434381057082797e-06,
+ "loss": 0.369028639793396,
+ "num_input_tokens_seen": 439590072,
+ "step": 10045,
+ "train_runtime": 77608.9784,
+ "train_tokens_per_second": 5664.165
+ },
+ {
+ "epoch": 0.6766765418798815,
+ "grad_norm": 0.7659413901414965,
+ "learning_rate": 7.4320711877728305e-06,
+ "loss": 0.3970766067504883,
+ "num_input_tokens_seen": 439814224,
+ "step": 10050,
+ "train_runtime": 77655.5893,
+ "train_tokens_per_second": 5663.652
+ },
+ {
+ "epoch": 0.6770131968758416,
+ "grad_norm": 0.7616766803709787,
+ "learning_rate": 7.429760638341909e-06,
+ "loss": 0.34726481437683104,
+ "num_input_tokens_seen": 440013360,
+ "step": 10055,
+ "train_runtime": 77694.3971,
+ "train_tokens_per_second": 5663.386
+ },
+ {
+ "epoch": 0.6773498518718017,
+ "grad_norm": 0.7214303745868783,
+ "learning_rate": 7.427449409436176e-06,
+ "loss": 0.42118000984191895,
+ "num_input_tokens_seen": 440232216,
+ "step": 10060,
+ "train_runtime": 77735.8693,
+ "train_tokens_per_second": 5663.18
+ },
+ {
+ "epoch": 0.6776865068677619,
+ "grad_norm": 0.6480309395843215,
+ "learning_rate": 7.425137501701955e-06,
+ "loss": 0.35333929061889646,
+ "num_input_tokens_seen": 440465120,
+ "step": 10065,
+ "train_runtime": 77778.1439,
+ "train_tokens_per_second": 5663.096
+ },
+ {
+ "epoch": 0.678023161863722,
+ "grad_norm": 0.8425975984479226,
+ "learning_rate": 7.422824915785765e-06,
+ "loss": 0.3732592582702637,
+ "num_input_tokens_seen": 440684520,
+ "step": 10070,
+ "train_runtime": 77813.6788,
+ "train_tokens_per_second": 5663.33
+ },
+ {
+ "epoch": 0.6783598168596822,
+ "grad_norm": 0.7751732244738783,
+ "learning_rate": 7.420511652334313e-06,
+ "loss": 0.3922509431838989,
+ "num_input_tokens_seen": 440888096,
+ "step": 10075,
+ "train_runtime": 77856.5574,
+ "train_tokens_per_second": 5662.825
+ },
+ {
+ "epoch": 0.6786964718556423,
+ "grad_norm": 0.746917656110605,
+ "learning_rate": 7.418197711994495e-06,
+ "loss": 0.3609792947769165,
+ "num_input_tokens_seen": 441099168,
+ "step": 10080,
+ "train_runtime": 77896.5389,
+ "train_tokens_per_second": 5662.629
+ },
+ {
+ "epoch": 0.6790331268516024,
+ "grad_norm": 0.7979148491415948,
+ "learning_rate": 7.4158830954133984e-06,
+ "loss": 0.37356295585632326,
+ "num_input_tokens_seen": 441329664,
+ "step": 10085,
+ "train_runtime": 77936.3536,
+ "train_tokens_per_second": 5662.693
+ },
+ {
+ "epoch": 0.6793697818475626,
+ "grad_norm": 0.7047169214626602,
+ "learning_rate": 7.413567803238298e-06,
+ "loss": 0.36446294784545896,
+ "num_input_tokens_seen": 441545160,
+ "step": 10090,
+ "train_runtime": 77973.8265,
+ "train_tokens_per_second": 5662.736
+ },
+ {
+ "epoch": 0.6797064368435227,
+ "grad_norm": 0.6385696916431561,
+ "learning_rate": 7.411251836116658e-06,
+ "loss": 0.3654582977294922,
+ "num_input_tokens_seen": 441770600,
+ "step": 10095,
+ "train_runtime": 78021.2974,
+ "train_tokens_per_second": 5662.18
+ },
+ {
+ "epoch": 0.6800430918394829,
+ "grad_norm": 0.675501708680357,
+ "learning_rate": 7.408935194696132e-06,
+ "loss": 0.38489809036254885,
+ "num_input_tokens_seen": 441989608,
+ "step": 10100,
+ "train_runtime": 78062.7748,
+ "train_tokens_per_second": 5661.977
+ },
+ {
+ "epoch": 0.680379746835443,
+ "grad_norm": 0.6976086448861777,
+ "learning_rate": 7.406617879624558e-06,
+ "loss": 0.39519965648651123,
+ "num_input_tokens_seen": 442219464,
+ "step": 10105,
+ "train_runtime": 78104.0438,
+ "train_tokens_per_second": 5661.928
+ },
+ {
+ "epoch": 0.6807164018314031,
+ "grad_norm": 0.6737767286617782,
+ "learning_rate": 7.40429989154997e-06,
+ "loss": 0.35456171035766604,
+ "num_input_tokens_seen": 442434504,
+ "step": 10110,
+ "train_runtime": 78146.9907,
+ "train_tokens_per_second": 5661.568
+ },
+ {
+ "epoch": 0.6810530568273633,
+ "grad_norm": 0.6824592254845565,
+ "learning_rate": 7.4019812311205826e-06,
+ "loss": 0.396740984916687,
+ "num_input_tokens_seen": 442663960,
+ "step": 10115,
+ "train_runtime": 78186.9172,
+ "train_tokens_per_second": 5661.612
+ },
+ {
+ "epoch": 0.6813897118233234,
+ "grad_norm": 0.6995329567504307,
+ "learning_rate": 7.399661898984805e-06,
+ "loss": 0.39820547103881837,
+ "num_input_tokens_seen": 442904544,
+ "step": 10120,
+ "train_runtime": 78228.9374,
+ "train_tokens_per_second": 5661.646
+ },
+ {
+ "epoch": 0.6817263668192836,
+ "grad_norm": 0.7146974067721376,
+ "learning_rate": 7.397341895791228e-06,
+ "loss": 0.3785873889923096,
+ "num_input_tokens_seen": 443119656,
+ "step": 10125,
+ "train_runtime": 78271.2656,
+ "train_tokens_per_second": 5661.332
+ },
+ {
+ "epoch": 0.6820630218152437,
+ "grad_norm": 0.7979315787216683,
+ "learning_rate": 7.395021222188634e-06,
+ "loss": 0.37588326930999755,
+ "num_input_tokens_seen": 443345048,
+ "step": 10130,
+ "train_runtime": 78307.7048,
+ "train_tokens_per_second": 5661.576
+ },
+ {
+ "epoch": 0.6823996768112038,
+ "grad_norm": 0.8569372783831029,
+ "learning_rate": 7.392699878825993e-06,
+ "loss": 0.4026289939880371,
+ "num_input_tokens_seen": 443550416,
+ "step": 10135,
+ "train_runtime": 78348.8182,
+ "train_tokens_per_second": 5661.227
+ },
+ {
+ "epoch": 0.682736331807164,
+ "grad_norm": 0.6799351048003278,
+ "learning_rate": 7.390377866352462e-06,
+ "loss": 0.3739679574966431,
+ "num_input_tokens_seen": 443760728,
+ "step": 10140,
+ "train_runtime": 78391.1203,
+ "train_tokens_per_second": 5660.855
+ },
+ {
+ "epoch": 0.6830729868031241,
+ "grad_norm": 0.7679162845866374,
+ "learning_rate": 7.388055185417382e-06,
+ "loss": 0.36473128795623777,
+ "num_input_tokens_seen": 443986040,
+ "step": 10145,
+ "train_runtime": 78429.781,
+ "train_tokens_per_second": 5660.937
+ },
+ {
+ "epoch": 0.6834096417990843,
+ "grad_norm": 0.7045812738235366,
+ "learning_rate": 7.3857318366702844e-06,
+ "loss": 0.35809712409973143,
+ "num_input_tokens_seen": 444206480,
+ "step": 10150,
+ "train_runtime": 78472.8507,
+ "train_tokens_per_second": 5660.639
+ },
+ {
+ "epoch": 0.6837462967950444,
+ "grad_norm": 0.7004993125342208,
+ "learning_rate": 7.383407820760885e-06,
+ "loss": 0.3713981151580811,
+ "num_input_tokens_seen": 444432096,
+ "step": 10155,
+ "train_runtime": 78515.6453,
+ "train_tokens_per_second": 5660.427
+ },
+ {
+ "epoch": 0.6840829517910045,
+ "grad_norm": 0.7774580479484972,
+ "learning_rate": 7.381083138339089e-06,
+ "loss": 0.38779325485229493,
+ "num_input_tokens_seen": 444643008,
+ "step": 10160,
+ "train_runtime": 78555.8979,
+ "train_tokens_per_second": 5660.212
+ },
+ {
+ "epoch": 0.6844196067869647,
+ "grad_norm": 0.7163976333149118,
+ "learning_rate": 7.378757790054988e-06,
+ "loss": 0.39658608436584475,
+ "num_input_tokens_seen": 444875528,
+ "step": 10165,
+ "train_runtime": 78592.4434,
+ "train_tokens_per_second": 5660.538
+ },
+ {
+ "epoch": 0.6847562617829248,
+ "grad_norm": 0.7470536381083663,
+ "learning_rate": 7.376431776558855e-06,
+ "loss": 0.3956130504608154,
+ "num_input_tokens_seen": 445097312,
+ "step": 10170,
+ "train_runtime": 78630.7483,
+ "train_tokens_per_second": 5660.601
+ },
+ {
+ "epoch": 0.685092916778885,
+ "grad_norm": 0.7408806601027974,
+ "learning_rate": 7.3741050985011546e-06,
+ "loss": 0.3585235118865967,
+ "num_input_tokens_seen": 445319152,
+ "step": 10175,
+ "train_runtime": 78670.4341,
+ "train_tokens_per_second": 5660.566
+ },
+ {
+ "epoch": 0.6854295717748451,
+ "grad_norm": 0.7086975222942481,
+ "learning_rate": 7.371777756532531e-06,
+ "loss": 0.34885153770446775,
+ "num_input_tokens_seen": 445544848,
+ "step": 10180,
+ "train_runtime": 78704.6359,
+ "train_tokens_per_second": 5660.973
+ },
+ {
+ "epoch": 0.6857662267708052,
+ "grad_norm": 0.7956956482990667,
+ "learning_rate": 7.369449751303823e-06,
+ "loss": 0.3933322191238403,
+ "num_input_tokens_seen": 445764768,
+ "step": 10185,
+ "train_runtime": 78741.4834,
+ "train_tokens_per_second": 5661.117
+ },
+ {
+ "epoch": 0.6861028817667654,
+ "grad_norm": 0.5998016401864463,
+ "learning_rate": 7.367121083466048e-06,
+ "loss": 0.3657109498977661,
+ "num_input_tokens_seen": 445988536,
+ "step": 10190,
+ "train_runtime": 78776.8049,
+ "train_tokens_per_second": 5661.419
+ },
+ {
+ "epoch": 0.6864395367627255,
+ "grad_norm": 0.8579022323586973,
+ "learning_rate": 7.364791753670409e-06,
+ "loss": 0.3847865104675293,
+ "num_input_tokens_seen": 446213600,
+ "step": 10195,
+ "train_runtime": 78810.0378,
+ "train_tokens_per_second": 5661.888
+ },
+ {
+ "epoch": 0.6867761917586857,
+ "grad_norm": 0.729478615131132,
+ "learning_rate": 7.362461762568298e-06,
+ "loss": 0.372448205947876,
+ "num_input_tokens_seen": 446432912,
+ "step": 10200,
+ "train_runtime": 78848.3501,
+ "train_tokens_per_second": 5661.918
+ },
+ {
+ "epoch": 0.6871128467546458,
+ "grad_norm": 0.67715047966365,
+ "learning_rate": 7.360131110811287e-06,
+ "loss": 0.3532183408737183,
+ "num_input_tokens_seen": 446656416,
+ "step": 10205,
+ "train_runtime": 78888.3323,
+ "train_tokens_per_second": 5661.882
+ },
+ {
+ "epoch": 0.6874495017506059,
+ "grad_norm": 0.7095724500319497,
+ "learning_rate": 7.357799799051138e-06,
+ "loss": 0.3632460355758667,
+ "num_input_tokens_seen": 446871712,
+ "step": 10210,
+ "train_runtime": 78927.4603,
+ "train_tokens_per_second": 5661.803
+ },
+ {
+ "epoch": 0.6877861567465661,
+ "grad_norm": 0.9476488222825441,
+ "learning_rate": 7.355467827939794e-06,
+ "loss": 0.38349783420562744,
+ "num_input_tokens_seen": 447082192,
+ "step": 10215,
+ "train_runtime": 78967.9316,
+ "train_tokens_per_second": 5661.566
+ },
+ {
+ "epoch": 0.6881228117425262,
+ "grad_norm": 0.7980347922355381,
+ "learning_rate": 7.353135198129382e-06,
+ "loss": 0.3912900686264038,
+ "num_input_tokens_seen": 447291168,
+ "step": 10220,
+ "train_runtime": 79002.7871,
+ "train_tokens_per_second": 5661.714
+ },
+ {
+ "epoch": 0.6884594667384863,
+ "grad_norm": 0.796413758238857,
+ "learning_rate": 7.350801910272216e-06,
+ "loss": 0.38165950775146484,
+ "num_input_tokens_seen": 447512392,
+ "step": 10225,
+ "train_runtime": 79046.6142,
+ "train_tokens_per_second": 5661.373
+ },
+ {
+ "epoch": 0.6887961217344465,
+ "grad_norm": 0.7746118196960639,
+ "learning_rate": 7.348467965020791e-06,
+ "loss": 0.34427103996276853,
+ "num_input_tokens_seen": 447731304,
+ "step": 10230,
+ "train_runtime": 79082.2762,
+ "train_tokens_per_second": 5661.588
+ },
+ {
+ "epoch": 0.6891327767304066,
+ "grad_norm": 0.8166556060856294,
+ "learning_rate": 7.3461333630277885e-06,
+ "loss": 0.3613192081451416,
+ "num_input_tokens_seen": 447922416,
+ "step": 10235,
+ "train_runtime": 79117.4476,
+ "train_tokens_per_second": 5661.487
+ },
+ {
+ "epoch": 0.6894694317263668,
+ "grad_norm": 0.6976640506234788,
+ "learning_rate": 7.343798104946074e-06,
+ "loss": 0.3512859344482422,
+ "num_input_tokens_seen": 448148520,
+ "step": 10240,
+ "train_runtime": 79156.1315,
+ "train_tokens_per_second": 5661.577
+ },
+ {
+ "epoch": 0.6898060867223269,
+ "grad_norm": 0.74195160383717,
+ "learning_rate": 7.341462191428692e-06,
+ "loss": 0.381636905670166,
+ "num_input_tokens_seen": 448364664,
+ "step": 10245,
+ "train_runtime": 79199.1829,
+ "train_tokens_per_second": 5661.228
+ },
+ {
+ "epoch": 0.690142741718287,
+ "grad_norm": 0.7527010867181296,
+ "learning_rate": 7.339125623128874e-06,
+ "loss": 0.3913971662521362,
+ "num_input_tokens_seen": 448587016,
+ "step": 10250,
+ "train_runtime": 79248.9634,
+ "train_tokens_per_second": 5660.478
+ },
+ {
+ "epoch": 0.6904793967142472,
+ "grad_norm": 0.7210165301469443,
+ "learning_rate": 7.336788400700035e-06,
+ "loss": 0.36610395908355714,
+ "num_input_tokens_seen": 448798728,
+ "step": 10255,
+ "train_runtime": 79292.6113,
+ "train_tokens_per_second": 5660.032
+ },
+ {
+ "epoch": 0.6908160517102074,
+ "grad_norm": 0.9096784607094398,
+ "learning_rate": 7.334450524795771e-06,
+ "loss": 0.408566951751709,
+ "num_input_tokens_seen": 449023728,
+ "step": 10260,
+ "train_runtime": 79333.8153,
+ "train_tokens_per_second": 5659.929
+ },
+ {
+ "epoch": 0.6911527067061676,
+ "grad_norm": 0.6784070074683265,
+ "learning_rate": 7.332111996069863e-06,
+ "loss": 0.3746421575546265,
+ "num_input_tokens_seen": 449245560,
+ "step": 10265,
+ "train_runtime": 79370.4334,
+ "train_tokens_per_second": 5660.112
+ },
+ {
+ "epoch": 0.6914893617021277,
+ "grad_norm": 0.8780985630278836,
+ "learning_rate": 7.32977281517627e-06,
+ "loss": 0.39997851848602295,
+ "num_input_tokens_seen": 449446632,
+ "step": 10270,
+ "train_runtime": 79406.7923,
+ "train_tokens_per_second": 5660.053
+ },
+ {
+ "epoch": 0.6918260166980879,
+ "grad_norm": 0.7238138438393178,
+ "learning_rate": 7.3274329827691406e-06,
+ "loss": 0.3859497308731079,
+ "num_input_tokens_seen": 449659944,
+ "step": 10275,
+ "train_runtime": 79449.7577,
+ "train_tokens_per_second": 5659.677
+ },
+ {
+ "epoch": 0.692162671694048,
+ "grad_norm": 0.6417831874350313,
+ "learning_rate": 7.325092499502796e-06,
+ "loss": 0.35598082542419435,
+ "num_input_tokens_seen": 449892656,
+ "step": 10280,
+ "train_runtime": 79489.514,
+ "train_tokens_per_second": 5659.774
+ },
+ {
+ "epoch": 0.6924993266900081,
+ "grad_norm": 0.6886975793380838,
+ "learning_rate": 7.322751366031751e-06,
+ "loss": 0.3739547967910767,
+ "num_input_tokens_seen": 450115160,
+ "step": 10285,
+ "train_runtime": 79528.7582,
+ "train_tokens_per_second": 5659.779
+ },
+ {
+ "epoch": 0.6928359816859683,
+ "grad_norm": 0.8067129855431404,
+ "learning_rate": 7.320409583010691e-06,
+ "loss": 0.3593981027603149,
+ "num_input_tokens_seen": 450339952,
+ "step": 10290,
+ "train_runtime": 79567.7249,
+ "train_tokens_per_second": 5659.832
+ },
+ {
+ "epoch": 0.6931726366819284,
+ "grad_norm": 0.720222949795357,
+ "learning_rate": 7.318067151094493e-06,
+ "loss": 0.4208518981933594,
+ "num_input_tokens_seen": 450562568,
+ "step": 10295,
+ "train_runtime": 79603.3438,
+ "train_tokens_per_second": 5660.096
+ },
+ {
+ "epoch": 0.6935092916778886,
+ "grad_norm": 0.6662609649687479,
+ "learning_rate": 7.3157240709382046e-06,
+ "loss": 0.3863611936569214,
+ "num_input_tokens_seen": 450794024,
+ "step": 10300,
+ "train_runtime": 79634.484,
+ "train_tokens_per_second": 5660.789
+ },
+ {
+ "epoch": 0.6938459466738487,
+ "grad_norm": 0.7343868155686462,
+ "learning_rate": 7.313380343197066e-06,
+ "loss": 0.3639280080795288,
+ "num_input_tokens_seen": 451027856,
+ "step": 10305,
+ "train_runtime": 79674.0903,
+ "train_tokens_per_second": 5660.91
+ },
+ {
+ "epoch": 0.6941826016698088,
+ "grad_norm": 0.7926474138155267,
+ "learning_rate": 7.311035968526492e-06,
+ "loss": 0.3794968605041504,
+ "num_input_tokens_seen": 451247432,
+ "step": 10310,
+ "train_runtime": 79715.3108,
+ "train_tokens_per_second": 5660.737
+ },
+ {
+ "epoch": 0.694519256665769,
+ "grad_norm": 0.8428213530948624,
+ "learning_rate": 7.308690947582078e-06,
+ "loss": 0.38078649044036866,
+ "num_input_tokens_seen": 451456696,
+ "step": 10315,
+ "train_runtime": 79750.434,
+ "train_tokens_per_second": 5660.868
+ },
+ {
+ "epoch": 0.6948559116617291,
+ "grad_norm": 0.7185545573840266,
+ "learning_rate": 7.306345281019602e-06,
+ "loss": 0.37413511276245115,
+ "num_input_tokens_seen": 451677704,
+ "step": 10320,
+ "train_runtime": 79790.1325,
+ "train_tokens_per_second": 5660.822
+ },
+ {
+ "epoch": 0.6951925666576892,
+ "grad_norm": 0.7678020136994413,
+ "learning_rate": 7.303998969495023e-06,
+ "loss": 0.40517454147338866,
+ "num_input_tokens_seen": 451892384,
+ "step": 10325,
+ "train_runtime": 79832.0377,
+ "train_tokens_per_second": 5660.539
+ },
+ {
+ "epoch": 0.6955292216536494,
+ "grad_norm": 0.7116247415151229,
+ "learning_rate": 7.301652013664479e-06,
+ "loss": 0.39863314628601076,
+ "num_input_tokens_seen": 452118752,
+ "step": 10330,
+ "train_runtime": 79868.7262,
+ "train_tokens_per_second": 5660.773
+ },
+ {
+ "epoch": 0.6958658766496095,
+ "grad_norm": 0.7656084452706404,
+ "learning_rate": 7.299304414184288e-06,
+ "loss": 0.40618147850036623,
+ "num_input_tokens_seen": 452338552,
+ "step": 10335,
+ "train_runtime": 79909.7465,
+ "train_tokens_per_second": 5660.618
+ },
+ {
+ "epoch": 0.6962025316455697,
+ "grad_norm": 0.6668364749984544,
+ "learning_rate": 7.296956171710951e-06,
+ "loss": 0.3879892587661743,
+ "num_input_tokens_seen": 452573592,
+ "step": 10340,
+ "train_runtime": 79946.2615,
+ "train_tokens_per_second": 5660.973
+ },
+ {
+ "epoch": 0.6965391866415298,
+ "grad_norm": 0.8038854323996706,
+ "learning_rate": 7.294607286901146e-06,
+ "loss": 0.3926121711730957,
+ "num_input_tokens_seen": 452795696,
+ "step": 10345,
+ "train_runtime": 79987.6707,
+ "train_tokens_per_second": 5660.819
+ },
+ {
+ "epoch": 0.69687584163749,
+ "grad_norm": 0.7769622893093939,
+ "learning_rate": 7.292257760411726e-06,
+ "loss": 0.3886493444442749,
+ "num_input_tokens_seen": 453009328,
+ "step": 10350,
+ "train_runtime": 80023.6895,
+ "train_tokens_per_second": 5660.94
+ },
+ {
+ "epoch": 0.6972124966334501,
+ "grad_norm": 0.7478012896643832,
+ "learning_rate": 7.289907592899737e-06,
+ "loss": 0.37845487594604493,
+ "num_input_tokens_seen": 453214464,
+ "step": 10355,
+ "train_runtime": 80062.9601,
+ "train_tokens_per_second": 5660.726
+ },
+ {
+ "epoch": 0.6975491516294102,
+ "grad_norm": 0.7047299342802962,
+ "learning_rate": 7.287556785022389e-06,
+ "loss": 0.3894767761230469,
+ "num_input_tokens_seen": 453437368,
+ "step": 10360,
+ "train_runtime": 80102.9256,
+ "train_tokens_per_second": 5660.684
+ },
+ {
+ "epoch": 0.6978858066253704,
+ "grad_norm": 0.7604979984461925,
+ "learning_rate": 7.285205337437082e-06,
+ "loss": 0.3892651557922363,
+ "num_input_tokens_seen": 453647680,
+ "step": 10365,
+ "train_runtime": 80141.1595,
+ "train_tokens_per_second": 5660.608
+ },
+ {
+ "epoch": 0.6982224616213305,
+ "grad_norm": 0.7970440490683166,
+ "learning_rate": 7.2828532508013895e-06,
+ "loss": 0.399712347984314,
+ "num_input_tokens_seen": 453867592,
+ "step": 10370,
+ "train_runtime": 80178.1608,
+ "train_tokens_per_second": 5660.738
+ },
+ {
+ "epoch": 0.6985591166172906,
+ "grad_norm": 0.7584286596844286,
+ "learning_rate": 7.2805005257730625e-06,
+ "loss": 0.3772284030914307,
+ "num_input_tokens_seen": 454087008,
+ "step": 10375,
+ "train_runtime": 80211.2544,
+ "train_tokens_per_second": 5661.138
+ },
+ {
+ "epoch": 0.6988957716132508,
+ "grad_norm": 0.6810825697097003,
+ "learning_rate": 7.278147163010037e-06,
+ "loss": 0.36341052055358886,
+ "num_input_tokens_seen": 454313224,
+ "step": 10380,
+ "train_runtime": 80250.7125,
+ "train_tokens_per_second": 5661.174
+ },
+ {
+ "epoch": 0.6992324266092109,
+ "grad_norm": 0.7849426670217893,
+ "learning_rate": 7.275793163170423e-06,
+ "loss": 0.36755235195159913,
+ "num_input_tokens_seen": 454539992,
+ "step": 10385,
+ "train_runtime": 80288.0816,
+ "train_tokens_per_second": 5661.363
+ },
+ {
+ "epoch": 0.6995690816051711,
+ "grad_norm": 0.7022327990344984,
+ "learning_rate": 7.273438526912506e-06,
+ "loss": 0.3956979990005493,
+ "num_input_tokens_seen": 454748824,
+ "step": 10390,
+ "train_runtime": 80325.0195,
+ "train_tokens_per_second": 5661.36
+ },
+ {
+ "epoch": 0.6999057366011312,
+ "grad_norm": 0.7244301627454164,
+ "learning_rate": 7.271083254894757e-06,
+ "loss": 0.36125593185424804,
+ "num_input_tokens_seen": 454962456,
+ "step": 10395,
+ "train_runtime": 80364.6433,
+ "train_tokens_per_second": 5661.227
+ },
+ {
+ "epoch": 0.7002423915970913,
+ "grad_norm": 0.7856991901608305,
+ "learning_rate": 7.268727347775815e-06,
+ "loss": 0.3766735076904297,
+ "num_input_tokens_seen": 455189544,
+ "step": 10400,
+ "train_runtime": 80398.3718,
+ "train_tokens_per_second": 5661.676
+ },
+ {
+ "epoch": 0.7005790465930515,
+ "grad_norm": 0.7076340782949039,
+ "learning_rate": 7.266370806214506e-06,
+ "loss": 0.32742795944213865,
+ "num_input_tokens_seen": 455409928,
+ "step": 10405,
+ "train_runtime": 80446.5155,
+ "train_tokens_per_second": 5661.027
+ },
+ {
+ "epoch": 0.7009157015890116,
+ "grad_norm": 0.6900523877236565,
+ "learning_rate": 7.264013630869827e-06,
+ "loss": 0.37685844898223875,
+ "num_input_tokens_seen": 455626104,
+ "step": 10410,
+ "train_runtime": 80491.2258,
+ "train_tokens_per_second": 5660.569
+ },
+ {
+ "epoch": 0.7012523565849718,
+ "grad_norm": 0.6710950318779684,
+ "learning_rate": 7.261655822400957e-06,
+ "loss": 0.39275636672973635,
+ "num_input_tokens_seen": 455857232,
+ "step": 10415,
+ "train_runtime": 80526.04,
+ "train_tokens_per_second": 5660.992
+ },
+ {
+ "epoch": 0.7015890115809319,
+ "grad_norm": 0.7198657361358307,
+ "learning_rate": 7.259297381467248e-06,
+ "loss": 0.38013458251953125,
+ "num_input_tokens_seen": 456077832,
+ "step": 10420,
+ "train_runtime": 80569.4365,
+ "train_tokens_per_second": 5660.68
+ },
+ {
+ "epoch": 0.701925666576892,
+ "grad_norm": 0.7005819646574629,
+ "learning_rate": 7.2569383087282285e-06,
+ "loss": 0.34854826927185056,
+ "num_input_tokens_seen": 456272960,
+ "step": 10425,
+ "train_runtime": 80611.1139,
+ "train_tokens_per_second": 5660.174
+ },
+ {
+ "epoch": 0.7022623215728522,
+ "grad_norm": 0.7072800232396435,
+ "learning_rate": 7.25457860484361e-06,
+ "loss": 0.39771556854248047,
+ "num_input_tokens_seen": 456493600,
+ "step": 10430,
+ "train_runtime": 80651.8402,
+ "train_tokens_per_second": 5660.052
+ },
+ {
+ "epoch": 0.7025989765688123,
+ "grad_norm": 0.7445696785976343,
+ "learning_rate": 7.252218270473274e-06,
+ "loss": 0.3760495662689209,
+ "num_input_tokens_seen": 456704912,
+ "step": 10435,
+ "train_runtime": 80692.0569,
+ "train_tokens_per_second": 5659.85
+ },
+ {
+ "epoch": 0.7029356315647725,
+ "grad_norm": 0.7457222056984448,
+ "learning_rate": 7.249857306277279e-06,
+ "loss": 0.369310188293457,
+ "num_input_tokens_seen": 456921864,
+ "step": 10440,
+ "train_runtime": 80726.6828,
+ "train_tokens_per_second": 5660.109
+ },
+ {
+ "epoch": 0.7032722865607326,
+ "grad_norm": 0.9221604408221112,
+ "learning_rate": 7.247495712915864e-06,
+ "loss": 0.35996944904327394,
+ "num_input_tokens_seen": 457143032,
+ "step": 10445,
+ "train_runtime": 80760.6102,
+ "train_tokens_per_second": 5660.47
+ },
+ {
+ "epoch": 0.7036089415566927,
+ "grad_norm": 0.6862985079879047,
+ "learning_rate": 7.245133491049439e-06,
+ "loss": 0.3736717939376831,
+ "num_input_tokens_seen": 457363400,
+ "step": 10450,
+ "train_runtime": 80799.8061,
+ "train_tokens_per_second": 5660.452
+ },
+ {
+ "epoch": 0.7039455965526529,
+ "grad_norm": 0.6935497313758425,
+ "learning_rate": 7.242770641338592e-06,
+ "loss": 0.4129630088806152,
+ "num_input_tokens_seen": 457584008,
+ "step": 10455,
+ "train_runtime": 80841.1453,
+ "train_tokens_per_second": 5660.286
+ },
+ {
+ "epoch": 0.704282251548613,
+ "grad_norm": 0.7202390162158359,
+ "learning_rate": 7.240407164444088e-06,
+ "loss": 0.36982312202453616,
+ "num_input_tokens_seen": 457798528,
+ "step": 10460,
+ "train_runtime": 80879.602,
+ "train_tokens_per_second": 5660.247
+ },
+ {
+ "epoch": 0.7046189065445732,
+ "grad_norm": 0.7863736874260931,
+ "learning_rate": 7.238043061026862e-06,
+ "loss": 0.34938392639160154,
+ "num_input_tokens_seen": 458006080,
+ "step": 10465,
+ "train_runtime": 80921.1996,
+ "train_tokens_per_second": 5659.902
+ },
+ {
+ "epoch": 0.7049555615405333,
+ "grad_norm": 0.7013473236004093,
+ "learning_rate": 7.235678331748034e-06,
+ "loss": 0.3546873092651367,
+ "num_input_tokens_seen": 458203208,
+ "step": 10470,
+ "train_runtime": 80956.6805,
+ "train_tokens_per_second": 5659.857
+ },
+ {
+ "epoch": 0.7052922165364934,
+ "grad_norm": 0.7707435582118967,
+ "learning_rate": 7.233312977268887e-06,
+ "loss": 0.40183773040771487,
+ "num_input_tokens_seen": 458427952,
+ "step": 10475,
+ "train_runtime": 80993.4895,
+ "train_tokens_per_second": 5660.059
+ },
+ {
+ "epoch": 0.7056288715324536,
+ "grad_norm": 0.7366337541289394,
+ "learning_rate": 7.230946998250889e-06,
+ "loss": 0.36978516578674314,
+ "num_input_tokens_seen": 458631504,
+ "step": 10480,
+ "train_runtime": 81028.6939,
+ "train_tokens_per_second": 5660.112
+ },
+ {
+ "epoch": 0.7059655265284137,
+ "grad_norm": 0.7121052684917011,
+ "learning_rate": 7.228580395355676e-06,
+ "loss": 0.34577772617340086,
+ "num_input_tokens_seen": 458845056,
+ "step": 10485,
+ "train_runtime": 81072.1146,
+ "train_tokens_per_second": 5659.715
+ },
+ {
+ "epoch": 0.7063021815243739,
+ "grad_norm": 0.7741334918584597,
+ "learning_rate": 7.226213169245062e-06,
+ "loss": 0.38037862777709963,
+ "num_input_tokens_seen": 459063872,
+ "step": 10490,
+ "train_runtime": 81113.0274,
+ "train_tokens_per_second": 5659.558
+ },
+ {
+ "epoch": 0.706638836520334,
+ "grad_norm": 0.7871663193272351,
+ "learning_rate": 7.223845320581035e-06,
+ "loss": 0.3649828672409058,
+ "num_input_tokens_seen": 459276768,
+ "step": 10495,
+ "train_runtime": 81158.0461,
+ "train_tokens_per_second": 5659.042
+ },
+ {
+ "epoch": 0.7069754915162941,
+ "grad_norm": 0.8550643751341503,
+ "learning_rate": 7.221476850025754e-06,
+ "loss": 0.3691579818725586,
+ "num_input_tokens_seen": 459504320,
+ "step": 10500,
+ "train_runtime": 81197.4298,
+ "train_tokens_per_second": 5659.099
+ },
+ {
+ "epoch": 0.7073121465122543,
+ "grad_norm": 0.6802216728098077,
+ "learning_rate": 7.219107758241557e-06,
+ "loss": 0.381423807144165,
+ "num_input_tokens_seen": 459726544,
+ "step": 10505,
+ "train_runtime": 81239.3131,
+ "train_tokens_per_second": 5658.917
+ },
+ {
+ "epoch": 0.7076488015082144,
+ "grad_norm": 0.7476015422415041,
+ "learning_rate": 7.2167380458909505e-06,
+ "loss": 0.3745572090148926,
+ "num_input_tokens_seen": 459949656,
+ "step": 10510,
+ "train_runtime": 81274.379,
+ "train_tokens_per_second": 5659.221
+ },
+ {
+ "epoch": 0.7079854565041745,
+ "grad_norm": 0.6742333237089346,
+ "learning_rate": 7.214367713636619e-06,
+ "loss": 0.375703763961792,
+ "num_input_tokens_seen": 460172128,
+ "step": 10515,
+ "train_runtime": 81312.8461,
+ "train_tokens_per_second": 5659.28
+ },
+ {
+ "epoch": 0.7083221115001347,
+ "grad_norm": 0.6954522791161296,
+ "learning_rate": 7.211996762141418e-06,
+ "loss": 0.37557039260864256,
+ "num_input_tokens_seen": 460390848,
+ "step": 10520,
+ "train_runtime": 81345.9866,
+ "train_tokens_per_second": 5659.663
+ },
+ {
+ "epoch": 0.7086587664960948,
+ "grad_norm": 0.7339967212433919,
+ "learning_rate": 7.209625192068374e-06,
+ "loss": 0.4028931617736816,
+ "num_input_tokens_seen": 460611832,
+ "step": 10525,
+ "train_runtime": 81379.8389,
+ "train_tokens_per_second": 5660.024
+ },
+ {
+ "epoch": 0.708995421492055,
+ "grad_norm": 0.701987120656059,
+ "learning_rate": 7.207253004080692e-06,
+ "loss": 0.37696285247802735,
+ "num_input_tokens_seen": 460844344,
+ "step": 10530,
+ "train_runtime": 81418.5196,
+ "train_tokens_per_second": 5660.191
+ },
+ {
+ "epoch": 0.7093320764880151,
+ "grad_norm": 0.7755594445584074,
+ "learning_rate": 7.204880198841746e-06,
+ "loss": 0.34694790840148926,
+ "num_input_tokens_seen": 461058504,
+ "step": 10535,
+ "train_runtime": 81453.7678,
+ "train_tokens_per_second": 5660.37
+ },
+ {
+ "epoch": 0.7096687314839752,
+ "grad_norm": 0.8426741622679441,
+ "learning_rate": 7.2025067770150835e-06,
+ "loss": 0.4015054225921631,
+ "num_input_tokens_seen": 461286472,
+ "step": 10540,
+ "train_runtime": 81493.9088,
+ "train_tokens_per_second": 5660.38
+ },
+ {
+ "epoch": 0.7100053864799354,
+ "grad_norm": 0.7557428415974053,
+ "learning_rate": 7.2001327392644235e-06,
+ "loss": 0.3497929573059082,
+ "num_input_tokens_seen": 461515496,
+ "step": 10545,
+ "train_runtime": 81528.2463,
+ "train_tokens_per_second": 5660.805
+ },
+ {
+ "epoch": 0.7103420414758955,
+ "grad_norm": 0.7845759323690452,
+ "learning_rate": 7.197758086253659e-06,
+ "loss": 0.39971365928649905,
+ "num_input_tokens_seen": 461738504,
+ "step": 10550,
+ "train_runtime": 81568.9369,
+ "train_tokens_per_second": 5660.715
+ },
+ {
+ "epoch": 0.7106786964718557,
+ "grad_norm": 0.7112358056324952,
+ "learning_rate": 7.195382818646854e-06,
+ "loss": 0.39182677268981936,
+ "num_input_tokens_seen": 461947016,
+ "step": 10555,
+ "train_runtime": 81608.3983,
+ "train_tokens_per_second": 5660.533
+ },
+ {
+ "epoch": 0.7110153514678158,
+ "grad_norm": 0.772603293296723,
+ "learning_rate": 7.1930069371082445e-06,
+ "loss": 0.36522927284240725,
+ "num_input_tokens_seen": 462152784,
+ "step": 10560,
+ "train_runtime": 81652.6367,
+ "train_tokens_per_second": 5659.986
+ },
+ {
+ "epoch": 0.7113520064637759,
+ "grad_norm": 0.7156871566744746,
+ "learning_rate": 7.190630442302238e-06,
+ "loss": 0.3833275079727173,
+ "num_input_tokens_seen": 462379128,
+ "step": 10565,
+ "train_runtime": 81694.2563,
+ "train_tokens_per_second": 5659.873
+ },
+ {
+ "epoch": 0.7116886614597361,
+ "grad_norm": 0.7033879162554953,
+ "learning_rate": 7.188253334893414e-06,
+ "loss": 0.3999313831329346,
+ "num_input_tokens_seen": 462598648,
+ "step": 10570,
+ "train_runtime": 81730.3714,
+ "train_tokens_per_second": 5660.058
+ },
+ {
+ "epoch": 0.7120253164556962,
+ "grad_norm": 0.8834420627247958,
+ "learning_rate": 7.18587561554652e-06,
+ "loss": 0.3910887956619263,
+ "num_input_tokens_seen": 462807208,
+ "step": 10575,
+ "train_runtime": 81771.4349,
+ "train_tokens_per_second": 5659.766
+ },
+ {
+ "epoch": 0.7123619714516564,
+ "grad_norm": 0.6918016219416666,
+ "learning_rate": 7.183497284926483e-06,
+ "loss": 0.36777448654174805,
+ "num_input_tokens_seen": 463042928,
+ "step": 10580,
+ "train_runtime": 81807.4141,
+ "train_tokens_per_second": 5660.158
+ },
+ {
+ "epoch": 0.7126986264476165,
+ "grad_norm": 0.7599277759479904,
+ "learning_rate": 7.181118343698392e-06,
+ "loss": 0.36260690689086916,
+ "num_input_tokens_seen": 463252424,
+ "step": 10585,
+ "train_runtime": 81851.2205,
+ "train_tokens_per_second": 5659.689
+ },
+ {
+ "epoch": 0.7130352814435766,
+ "grad_norm": 0.7224431307545425,
+ "learning_rate": 7.178738792527512e-06,
+ "loss": 0.39423255920410155,
+ "num_input_tokens_seen": 463469984,
+ "step": 10590,
+ "train_runtime": 81891.039,
+ "train_tokens_per_second": 5659.593
+ },
+ {
+ "epoch": 0.7133719364395368,
+ "grad_norm": 0.7536526351492208,
+ "learning_rate": 7.176358632079277e-06,
+ "loss": 0.37251434326171873,
+ "num_input_tokens_seen": 463682208,
+ "step": 10595,
+ "train_runtime": 81929.5391,
+ "train_tokens_per_second": 5659.524
+ },
+ {
+ "epoch": 0.7137085914354969,
+ "grad_norm": 0.6793493050362683,
+ "learning_rate": 7.173977863019288e-06,
+ "loss": 0.37181549072265624,
+ "num_input_tokens_seen": 463907464,
+ "step": 10600,
+ "train_runtime": 81970.1484,
+ "train_tokens_per_second": 5659.468
+ },
+ {
+ "epoch": 0.7140452464314571,
+ "grad_norm": 0.6822436534253042,
+ "learning_rate": 7.171596486013323e-06,
+ "loss": 0.34790878295898436,
+ "num_input_tokens_seen": 464138256,
+ "step": 10605,
+ "train_runtime": 82007.4481,
+ "train_tokens_per_second": 5659.709
+ },
+ {
+ "epoch": 0.7143819014274172,
+ "grad_norm": 0.8051213804159209,
+ "learning_rate": 7.169214501727326e-06,
+ "loss": 0.37795867919921877,
+ "num_input_tokens_seen": 464338936,
+ "step": 10610,
+ "train_runtime": 82043.2488,
+ "train_tokens_per_second": 5659.685
+ },
+ {
+ "epoch": 0.7147185564233773,
+ "grad_norm": 0.686749483511892,
+ "learning_rate": 7.16683191082741e-06,
+ "loss": 0.36943345069885253,
+ "num_input_tokens_seen": 464560144,
+ "step": 10615,
+ "train_runtime": 82079.0859,
+ "train_tokens_per_second": 5659.909
+ },
+ {
+ "epoch": 0.7150552114193375,
+ "grad_norm": 0.8746419009334871,
+ "learning_rate": 7.16444871397986e-06,
+ "loss": 0.368468713760376,
+ "num_input_tokens_seen": 464765496,
+ "step": 10620,
+ "train_runtime": 82117.3295,
+ "train_tokens_per_second": 5659.774
+ },
+ {
+ "epoch": 0.7153918664152976,
+ "grad_norm": 1.042013296164164,
+ "learning_rate": 7.162064911851129e-06,
+ "loss": 0.36743860244750975,
+ "num_input_tokens_seen": 464978408,
+ "step": 10625,
+ "train_runtime": 82160.6887,
+ "train_tokens_per_second": 5659.378
+ },
+ {
+ "epoch": 0.7157285214112578,
+ "grad_norm": 0.7321276232893942,
+ "learning_rate": 7.15968050510784e-06,
+ "loss": 0.40234060287475587,
+ "num_input_tokens_seen": 465192704,
+ "step": 10630,
+ "train_runtime": 82198.1606,
+ "train_tokens_per_second": 5659.405
+ },
+ {
+ "epoch": 0.7160651764072179,
+ "grad_norm": 0.8428920187826066,
+ "learning_rate": 7.1572954944167834e-06,
+ "loss": 0.3755232810974121,
+ "num_input_tokens_seen": 465408616,
+ "step": 10635,
+ "train_runtime": 82239.586,
+ "train_tokens_per_second": 5659.18
+ },
+ {
+ "epoch": 0.716401831403178,
+ "grad_norm": 0.679996578354662,
+ "learning_rate": 7.154909880444922e-06,
+ "loss": 0.3969583988189697,
+ "num_input_tokens_seen": 465633232,
+ "step": 10640,
+ "train_runtime": 82282.7769,
+ "train_tokens_per_second": 5658.939
+ },
+ {
+ "epoch": 0.7167384863991382,
+ "grad_norm": 1.5740350922344166,
+ "learning_rate": 7.152523663859384e-06,
+ "loss": 0.37885403633117676,
+ "num_input_tokens_seen": 465845256,
+ "step": 10645,
+ "train_runtime": 82324.9766,
+ "train_tokens_per_second": 5658.614
+ },
+ {
+ "epoch": 0.7170751413950983,
+ "grad_norm": 0.8328362306471117,
+ "learning_rate": 7.150136845327466e-06,
+ "loss": 0.3639181613922119,
+ "num_input_tokens_seen": 466056584,
+ "step": 10650,
+ "train_runtime": 82369.823,
+ "train_tokens_per_second": 5658.099
+ },
+ {
+ "epoch": 0.7174117963910585,
+ "grad_norm": 0.737238654335971,
+ "learning_rate": 7.147749425516637e-06,
+ "loss": 0.3720954418182373,
+ "num_input_tokens_seen": 466287992,
+ "step": 10655,
+ "train_runtime": 82407.102,
+ "train_tokens_per_second": 5658.347
+ },
+ {
+ "epoch": 0.7177484513870186,
+ "grad_norm": 0.7078523897720794,
+ "learning_rate": 7.145361405094529e-06,
+ "loss": 0.3578562021255493,
+ "num_input_tokens_seen": 466523176,
+ "step": 10660,
+ "train_runtime": 82439.2799,
+ "train_tokens_per_second": 5658.991
+ },
+ {
+ "epoch": 0.7180851063829787,
+ "grad_norm": 0.7362901383213949,
+ "learning_rate": 7.142972784728945e-06,
+ "loss": 0.3409886360168457,
+ "num_input_tokens_seen": 466734112,
+ "step": 10665,
+ "train_runtime": 82483.0478,
+ "train_tokens_per_second": 5658.546
+ },
+ {
+ "epoch": 0.7184217613789389,
+ "grad_norm": 0.8146924235270421,
+ "learning_rate": 7.140583565087856e-06,
+ "loss": 0.35494747161865237,
+ "num_input_tokens_seen": 466952312,
+ "step": 10670,
+ "train_runtime": 82519.1924,
+ "train_tokens_per_second": 5658.712
+ },
+ {
+ "epoch": 0.718758416374899,
+ "grad_norm": 0.7593849226158301,
+ "learning_rate": 7.138193746839399e-06,
+ "loss": 0.3933510065078735,
+ "num_input_tokens_seen": 467185248,
+ "step": 10675,
+ "train_runtime": 82565.1418,
+ "train_tokens_per_second": 5658.384
+ },
+ {
+ "epoch": 0.7190950713708592,
+ "grad_norm": 0.7169033125551184,
+ "learning_rate": 7.135803330651878e-06,
+ "loss": 0.37902536392211916,
+ "num_input_tokens_seen": 467421448,
+ "step": 10680,
+ "train_runtime": 82599.409,
+ "train_tokens_per_second": 5658.896
+ },
+ {
+ "epoch": 0.7194317263668193,
+ "grad_norm": 0.6927480355842544,
+ "learning_rate": 7.133412317193766e-06,
+ "loss": 0.3649748802185059,
+ "num_input_tokens_seen": 467648088,
+ "step": 10685,
+ "train_runtime": 82647.9029,
+ "train_tokens_per_second": 5658.318
+ },
+ {
+ "epoch": 0.7197683813627794,
+ "grad_norm": 0.8368737935217295,
+ "learning_rate": 7.131020707133703e-06,
+ "loss": 0.39012324810028076,
+ "num_input_tokens_seen": 467852784,
+ "step": 10690,
+ "train_runtime": 82685.2143,
+ "train_tokens_per_second": 5658.24
+ },
+ {
+ "epoch": 0.7201050363587396,
+ "grad_norm": 0.7246874489536738,
+ "learning_rate": 7.128628501140495e-06,
+ "loss": 0.3994007587432861,
+ "num_input_tokens_seen": 468062360,
+ "step": 10695,
+ "train_runtime": 82719.4298,
+ "train_tokens_per_second": 5658.433
+ },
+ {
+ "epoch": 0.7204416913546997,
+ "grad_norm": 0.7034507707765917,
+ "learning_rate": 7.126235699883115e-06,
+ "loss": 0.3543556213378906,
+ "num_input_tokens_seen": 468260752,
+ "step": 10700,
+ "train_runtime": 82762.0259,
+ "train_tokens_per_second": 5657.918
+ },
+ {
+ "epoch": 0.7207783463506598,
+ "grad_norm": 0.6713729985803661,
+ "learning_rate": 7.1238423040306995e-06,
+ "loss": 0.3659778594970703,
+ "num_input_tokens_seen": 468480296,
+ "step": 10705,
+ "train_runtime": 82807.3194,
+ "train_tokens_per_second": 5657.474
+ },
+ {
+ "epoch": 0.72111500134662,
+ "grad_norm": 0.6714547248755331,
+ "learning_rate": 7.121448314252557e-06,
+ "loss": 0.33471217155456545,
+ "num_input_tokens_seen": 468714640,
+ "step": 10710,
+ "train_runtime": 82849.0161,
+ "train_tokens_per_second": 5657.456
+ },
+ {
+ "epoch": 0.7214516563425801,
+ "grad_norm": 0.7542079623814812,
+ "learning_rate": 7.1190537312181586e-06,
+ "loss": 0.3590662479400635,
+ "num_input_tokens_seen": 468920200,
+ "step": 10715,
+ "train_runtime": 82883.6718,
+ "train_tokens_per_second": 5657.57
+ },
+ {
+ "epoch": 0.7217883113385403,
+ "grad_norm": 0.788689096156777,
+ "learning_rate": 7.116658555597141e-06,
+ "loss": 0.35546276569366453,
+ "num_input_tokens_seen": 469105088,
+ "step": 10720,
+ "train_runtime": 82920.6765,
+ "train_tokens_per_second": 5657.275
+ },
+ {
+ "epoch": 0.7221249663345004,
+ "grad_norm": 0.7720777482295464,
+ "learning_rate": 7.114262788059308e-06,
+ "loss": 0.36028232574462893,
+ "num_input_tokens_seen": 469316144,
+ "step": 10725,
+ "train_runtime": 82956.2558,
+ "train_tokens_per_second": 5657.393
+ },
+ {
+ "epoch": 0.7224616213304605,
+ "grad_norm": 0.7595426301739953,
+ "learning_rate": 7.111866429274627e-06,
+ "loss": 0.3734895706176758,
+ "num_input_tokens_seen": 469533344,
+ "step": 10730,
+ "train_runtime": 82995.9693,
+ "train_tokens_per_second": 5657.303
+ },
+ {
+ "epoch": 0.7227982763264207,
+ "grad_norm": 0.7895201856991956,
+ "learning_rate": 7.1094694799132356e-06,
+ "loss": 0.38394627571105955,
+ "num_input_tokens_seen": 469757776,
+ "step": 10735,
+ "train_runtime": 83039.5254,
+ "train_tokens_per_second": 5657.038
+ },
+ {
+ "epoch": 0.7231349313223808,
+ "grad_norm": 0.7170404350301777,
+ "learning_rate": 7.10707194064543e-06,
+ "loss": 0.355570125579834,
+ "num_input_tokens_seen": 469973048,
+ "step": 10740,
+ "train_runtime": 83082.4999,
+ "train_tokens_per_second": 5656.703
+ },
+ {
+ "epoch": 0.723471586318341,
+ "grad_norm": 0.751527765717788,
+ "learning_rate": 7.104673812141676e-06,
+ "loss": 0.4048286437988281,
+ "num_input_tokens_seen": 470200352,
+ "step": 10745,
+ "train_runtime": 83123.5825,
+ "train_tokens_per_second": 5656.642
+ },
+ {
+ "epoch": 0.7238082413143011,
+ "grad_norm": 0.7209513181054711,
+ "learning_rate": 7.102275095072601e-06,
+ "loss": 0.3921640157699585,
+ "num_input_tokens_seen": 470434416,
+ "step": 10750,
+ "train_runtime": 83170.1972,
+ "train_tokens_per_second": 5656.286
+ },
+ {
+ "epoch": 0.7241448963102612,
+ "grad_norm": 0.7141189337990387,
+ "learning_rate": 7.099875790108998e-06,
+ "loss": 0.36566712856292727,
+ "num_input_tokens_seen": 470662208,
+ "step": 10755,
+ "train_runtime": 83219.7805,
+ "train_tokens_per_second": 5655.653
+ },
+ {
+ "epoch": 0.7244815513062214,
+ "grad_norm": 0.6677452072067016,
+ "learning_rate": 7.097475897921827e-06,
+ "loss": 0.37137765884399415,
+ "num_input_tokens_seen": 470876192,
+ "step": 10760,
+ "train_runtime": 83259.577,
+ "train_tokens_per_second": 5655.52
+ },
+ {
+ "epoch": 0.7248182063021815,
+ "grad_norm": 0.8654836254283663,
+ "learning_rate": 7.09507541918221e-06,
+ "loss": 0.3979093790054321,
+ "num_input_tokens_seen": 471091920,
+ "step": 10765,
+ "train_runtime": 83298.0964,
+ "train_tokens_per_second": 5655.494
+ },
+ {
+ "epoch": 0.7251548612981417,
+ "grad_norm": 0.6072122160998616,
+ "learning_rate": 7.092674354561433e-06,
+ "loss": 0.3879514217376709,
+ "num_input_tokens_seen": 471316336,
+ "step": 10770,
+ "train_runtime": 83339.0663,
+ "train_tokens_per_second": 5655.407
+ },
+ {
+ "epoch": 0.7254915162941018,
+ "grad_norm": 0.753636374019264,
+ "learning_rate": 7.090272704730945e-06,
+ "loss": 0.3511503219604492,
+ "num_input_tokens_seen": 471518800,
+ "step": 10775,
+ "train_runtime": 83375.4431,
+ "train_tokens_per_second": 5655.368
+ },
+ {
+ "epoch": 0.7258281712900619,
+ "grad_norm": 0.7575044992088539,
+ "learning_rate": 7.08787047036236e-06,
+ "loss": 0.35472354888916013,
+ "num_input_tokens_seen": 471754616,
+ "step": 10780,
+ "train_runtime": 83418.0251,
+ "train_tokens_per_second": 5655.308
+ },
+ {
+ "epoch": 0.7261648262860221,
+ "grad_norm": 0.822516802699336,
+ "learning_rate": 7.085467652127457e-06,
+ "loss": 0.3795314788818359,
+ "num_input_tokens_seen": 471981872,
+ "step": 10785,
+ "train_runtime": 83450.133,
+ "train_tokens_per_second": 5655.855
+ },
+ {
+ "epoch": 0.7265014812819822,
+ "grad_norm": 0.6385132066085284,
+ "learning_rate": 7.083064250698175e-06,
+ "loss": 0.3529500961303711,
+ "num_input_tokens_seen": 472213144,
+ "step": 10790,
+ "train_runtime": 83491.9091,
+ "train_tokens_per_second": 5655.795
+ },
+ {
+ "epoch": 0.7268381362779424,
+ "grad_norm": 0.6595130665327932,
+ "learning_rate": 7.0806602667466175e-06,
+ "loss": 0.37914767265319826,
+ "num_input_tokens_seen": 472443584,
+ "step": 10795,
+ "train_runtime": 83534.7223,
+ "train_tokens_per_second": 5655.655
+ },
+ {
+ "epoch": 0.7271747912739025,
+ "grad_norm": 0.6988131699183094,
+ "learning_rate": 7.078255700945051e-06,
+ "loss": 0.38105590343475343,
+ "num_input_tokens_seen": 472655528,
+ "step": 10800,
+ "train_runtime": 83566.9737,
+ "train_tokens_per_second": 5656.009
+ },
+ {
+ "epoch": 0.7275114462698626,
+ "grad_norm": 0.7284700984007974,
+ "learning_rate": 7.075850553965904e-06,
+ "loss": 0.366654634475708,
+ "num_input_tokens_seen": 472863848,
+ "step": 10805,
+ "train_runtime": 83605.9871,
+ "train_tokens_per_second": 5655.861
+ },
+ {
+ "epoch": 0.7278481012658228,
+ "grad_norm": 0.7093309792405998,
+ "learning_rate": 7.07344482648177e-06,
+ "loss": 0.3818976879119873,
+ "num_input_tokens_seen": 473087584,
+ "step": 10810,
+ "train_runtime": 83637.9528,
+ "train_tokens_per_second": 5656.375
+ },
+ {
+ "epoch": 0.7281847562617829,
+ "grad_norm": 0.7655242160543194,
+ "learning_rate": 7.071038519165402e-06,
+ "loss": 0.3804640293121338,
+ "num_input_tokens_seen": 473302440,
+ "step": 10815,
+ "train_runtime": 83680.6746,
+ "train_tokens_per_second": 5656.054
+ },
+ {
+ "epoch": 0.7285214112577431,
+ "grad_norm": 0.7319764295438455,
+ "learning_rate": 7.068631632689716e-06,
+ "loss": 0.3819848299026489,
+ "num_input_tokens_seen": 473528888,
+ "step": 10820,
+ "train_runtime": 83719.4607,
+ "train_tokens_per_second": 5656.139
+ },
+ {
+ "epoch": 0.7288580662537032,
+ "grad_norm": 9.329714081971986,
+ "learning_rate": 7.066224167727791e-06,
+ "loss": 0.3868781566619873,
+ "num_input_tokens_seen": 473749232,
+ "step": 10825,
+ "train_runtime": 83753.7938,
+ "train_tokens_per_second": 5656.451
+ },
+ {
+ "epoch": 0.7291947212496633,
+ "grad_norm": 0.841160080727543,
+ "learning_rate": 7.063816124952865e-06,
+ "loss": 0.36863956451416013,
+ "num_input_tokens_seen": 473966688,
+ "step": 10830,
+ "train_runtime": 83791.8425,
+ "train_tokens_per_second": 5656.478
+ },
+ {
+ "epoch": 0.7295313762456235,
+ "grad_norm": 0.6549972032651611,
+ "learning_rate": 7.061407505038341e-06,
+ "loss": 0.3608518362045288,
+ "num_input_tokens_seen": 474183856,
+ "step": 10835,
+ "train_runtime": 83830.8355,
+ "train_tokens_per_second": 5656.437
+ },
+ {
+ "epoch": 0.7298680312415836,
+ "grad_norm": 0.6949899381534901,
+ "learning_rate": 7.058998308657782e-06,
+ "loss": 0.3866151809692383,
+ "num_input_tokens_seen": 474414672,
+ "step": 10840,
+ "train_runtime": 83863.1172,
+ "train_tokens_per_second": 5657.012
+ },
+ {
+ "epoch": 0.7302046862375438,
+ "grad_norm": 0.6794084004266899,
+ "learning_rate": 7.056588536484912e-06,
+ "loss": 0.35731284618377684,
+ "num_input_tokens_seen": 474622840,
+ "step": 10845,
+ "train_runtime": 83892.8777,
+ "train_tokens_per_second": 5657.487
+ },
+ {
+ "epoch": 0.7305413412335039,
+ "grad_norm": 1.3762092654034057,
+ "learning_rate": 7.0541781891936145e-06,
+ "loss": 0.35231194496154783,
+ "num_input_tokens_seen": 474842808,
+ "step": 10850,
+ "train_runtime": 83931.0911,
+ "train_tokens_per_second": 5657.532
+ },
+ {
+ "epoch": 0.730877996229464,
+ "grad_norm": 0.8182830044909798,
+ "learning_rate": 7.0517672674579364e-06,
+ "loss": 0.3739452838897705,
+ "num_input_tokens_seen": 475060040,
+ "step": 10855,
+ "train_runtime": 83971.3377,
+ "train_tokens_per_second": 5657.407
+ },
+ {
+ "epoch": 0.7312146512254242,
+ "grad_norm": 0.7405063263799381,
+ "learning_rate": 7.049355771952087e-06,
+ "loss": 0.367911171913147,
+ "num_input_tokens_seen": 475270960,
+ "step": 10860,
+ "train_runtime": 84012.5713,
+ "train_tokens_per_second": 5657.141
+ },
+ {
+ "epoch": 0.7315513062213843,
+ "grad_norm": 0.743519981806223,
+ "learning_rate": 7.046943703350429e-06,
+ "loss": 0.3706895112991333,
+ "num_input_tokens_seen": 475496232,
+ "step": 10865,
+ "train_runtime": 84056.0111,
+ "train_tokens_per_second": 5656.897
+ },
+ {
+ "epoch": 0.7318879612173445,
+ "grad_norm": 0.6596197156889615,
+ "learning_rate": 7.044531062327492e-06,
+ "loss": 0.356654691696167,
+ "num_input_tokens_seen": 475709832,
+ "step": 10870,
+ "train_runtime": 84089.5859,
+ "train_tokens_per_second": 5657.179
+ },
+ {
+ "epoch": 0.7322246162133046,
+ "grad_norm": 0.6526147436786898,
+ "learning_rate": 7.042117849557963e-06,
+ "loss": 0.3708888053894043,
+ "num_input_tokens_seen": 475932072,
+ "step": 10875,
+ "train_runtime": 84133.6504,
+ "train_tokens_per_second": 5656.858
+ },
+ {
+ "epoch": 0.7325612712092647,
+ "grad_norm": 0.744802883666199,
+ "learning_rate": 7.0397040657166885e-06,
+ "loss": 0.36887059211730955,
+ "num_input_tokens_seen": 476162776,
+ "step": 10880,
+ "train_runtime": 84169.2025,
+ "train_tokens_per_second": 5657.209
+ },
+ {
+ "epoch": 0.7328979262052249,
+ "grad_norm": 0.7068220775813847,
+ "learning_rate": 7.037289711478676e-06,
+ "loss": 0.34908170700073243,
+ "num_input_tokens_seen": 476376384,
+ "step": 10885,
+ "train_runtime": 84204.347,
+ "train_tokens_per_second": 5657.385
+ },
+ {
+ "epoch": 0.733234581201185,
+ "grad_norm": 0.7211408903358524,
+ "learning_rate": 7.034874787519095e-06,
+ "loss": 0.3759069204330444,
+ "num_input_tokens_seen": 476599400,
+ "step": 10890,
+ "train_runtime": 84242.2013,
+ "train_tokens_per_second": 5657.49
+ },
+ {
+ "epoch": 0.7335712361971451,
+ "grad_norm": 0.6965350386219624,
+ "learning_rate": 7.0324592945132655e-06,
+ "loss": 0.4069540023803711,
+ "num_input_tokens_seen": 476821304,
+ "step": 10895,
+ "train_runtime": 84276.594,
+ "train_tokens_per_second": 5657.814
+ },
+ {
+ "epoch": 0.7339078911931053,
+ "grad_norm": 0.7259339581846704,
+ "learning_rate": 7.0300432331366764e-06,
+ "loss": 0.3802814245223999,
+ "num_input_tokens_seen": 477036424,
+ "step": 10900,
+ "train_runtime": 84309.7162,
+ "train_tokens_per_second": 5658.143
+ },
+ {
+ "epoch": 0.7342445461890654,
+ "grad_norm": 0.726630192334144,
+ "learning_rate": 7.02762660406497e-06,
+ "loss": 0.3741856813430786,
+ "num_input_tokens_seen": 477257288,
+ "step": 10905,
+ "train_runtime": 84348.0236,
+ "train_tokens_per_second": 5658.192
+ },
+ {
+ "epoch": 0.7345812011850256,
+ "grad_norm": 0.6353064783947603,
+ "learning_rate": 7.02520940797395e-06,
+ "loss": 0.3721340179443359,
+ "num_input_tokens_seen": 477464672,
+ "step": 10910,
+ "train_runtime": 84385.7451,
+ "train_tokens_per_second": 5658.12
+ },
+ {
+ "epoch": 0.7349178561809857,
+ "grad_norm": 0.6280488461396918,
+ "learning_rate": 7.022791645539576e-06,
+ "loss": 0.37077898979187013,
+ "num_input_tokens_seen": 477670600,
+ "step": 10915,
+ "train_runtime": 84423.4019,
+ "train_tokens_per_second": 5658.035
+ },
+ {
+ "epoch": 0.7352545111769458,
+ "grad_norm": 0.7582582970874048,
+ "learning_rate": 7.02037331743797e-06,
+ "loss": 0.4068927764892578,
+ "num_input_tokens_seen": 477891112,
+ "step": 10920,
+ "train_runtime": 84461.3924,
+ "train_tokens_per_second": 5658.101
+ },
+ {
+ "epoch": 0.735591166172906,
+ "grad_norm": 0.6908837097564919,
+ "learning_rate": 7.0179544243454076e-06,
+ "loss": 0.3525757551193237,
+ "num_input_tokens_seen": 478120960,
+ "step": 10925,
+ "train_runtime": 84493.3059,
+ "train_tokens_per_second": 5658.684
+ },
+ {
+ "epoch": 0.7359278211688661,
+ "grad_norm": 0.8025906131360969,
+ "learning_rate": 7.0155349669383245e-06,
+ "loss": 0.36997184753417967,
+ "num_input_tokens_seen": 478344240,
+ "step": 10930,
+ "train_runtime": 84534.5442,
+ "train_tokens_per_second": 5658.565
+ },
+ {
+ "epoch": 0.7362644761648263,
+ "grad_norm": 0.691884724659927,
+ "learning_rate": 7.013114945893316e-06,
+ "loss": 0.36500933170318606,
+ "num_input_tokens_seen": 478567880,
+ "step": 10935,
+ "train_runtime": 84579.7591,
+ "train_tokens_per_second": 5658.184
+ },
+ {
+ "epoch": 0.7366011311607864,
+ "grad_norm": 0.7278702815389843,
+ "learning_rate": 7.010694361887131e-06,
+ "loss": 0.3656963109970093,
+ "num_input_tokens_seen": 478795792,
+ "step": 10940,
+ "train_runtime": 84624.1001,
+ "train_tokens_per_second": 5657.913
+ },
+ {
+ "epoch": 0.7369377861567465,
+ "grad_norm": 0.6970616747051929,
+ "learning_rate": 7.008273215596681e-06,
+ "loss": 0.3773782253265381,
+ "num_input_tokens_seen": 479014424,
+ "step": 10945,
+ "train_runtime": 84661.672,
+ "train_tokens_per_second": 5657.984
+ },
+ {
+ "epoch": 0.7372744411527067,
+ "grad_norm": 0.7390738693319422,
+ "learning_rate": 7.005851507699029e-06,
+ "loss": 0.3759654760360718,
+ "num_input_tokens_seen": 479243632,
+ "step": 10950,
+ "train_runtime": 84694.665,
+ "train_tokens_per_second": 5658.487
+ },
+ {
+ "epoch": 0.7376110961486668,
+ "grad_norm": 0.6808946119767743,
+ "learning_rate": 7.003429238871399e-06,
+ "loss": 0.3657572031021118,
+ "num_input_tokens_seen": 479473192,
+ "step": 10955,
+ "train_runtime": 84738.238,
+ "train_tokens_per_second": 5658.286
+ },
+ {
+ "epoch": 0.737947751144627,
+ "grad_norm": 0.7178504014412501,
+ "learning_rate": 7.001006409791171e-06,
+ "loss": 0.40654420852661133,
+ "num_input_tokens_seen": 479687920,
+ "step": 10960,
+ "train_runtime": 84776.0398,
+ "train_tokens_per_second": 5658.296
+ },
+ {
+ "epoch": 0.7382844061405871,
+ "grad_norm": 0.7628213758613887,
+ "learning_rate": 6.998583021135882e-06,
+ "loss": 0.3780509948730469,
+ "num_input_tokens_seen": 479906096,
+ "step": 10965,
+ "train_runtime": 84815.2532,
+ "train_tokens_per_second": 5658.252
+ },
+ {
+ "epoch": 0.7386210611365472,
+ "grad_norm": 0.8356715268382896,
+ "learning_rate": 6.996159073583224e-06,
+ "loss": 0.36737675666809083,
+ "num_input_tokens_seen": 480125696,
+ "step": 10970,
+ "train_runtime": 84852.9815,
+ "train_tokens_per_second": 5658.324
+ },
+ {
+ "epoch": 0.7389577161325074,
+ "grad_norm": 0.8055302742937802,
+ "learning_rate": 6.993734567811046e-06,
+ "loss": 0.34843399524688723,
+ "num_input_tokens_seen": 480343296,
+ "step": 10975,
+ "train_runtime": 84896.1427,
+ "train_tokens_per_second": 5658.011
+ },
+ {
+ "epoch": 0.7392943711284675,
+ "grad_norm": 0.6594571098967035,
+ "learning_rate": 6.991309504497352e-06,
+ "loss": 0.37664797306060793,
+ "num_input_tokens_seen": 480566544,
+ "step": 10980,
+ "train_runtime": 84934.0218,
+ "train_tokens_per_second": 5658.116
+ },
+ {
+ "epoch": 0.7396310261244277,
+ "grad_norm": 0.7694992058673898,
+ "learning_rate": 6.988883884320307e-06,
+ "loss": 0.3457742214202881,
+ "num_input_tokens_seen": 480774232,
+ "step": 10985,
+ "train_runtime": 84970.4518,
+ "train_tokens_per_second": 5658.134
+ },
+ {
+ "epoch": 0.7399676811203878,
+ "grad_norm": 0.7402878857378402,
+ "learning_rate": 6.9864577079582254e-06,
+ "loss": 0.39705471992492675,
+ "num_input_tokens_seen": 481000200,
+ "step": 10990,
+ "train_runtime": 85008.5973,
+ "train_tokens_per_second": 5658.254
+ },
+ {
+ "epoch": 0.7403043361163479,
+ "grad_norm": 0.8393816654846036,
+ "learning_rate": 6.984030976089581e-06,
+ "loss": 0.3594042301177979,
+ "num_input_tokens_seen": 481219544,
+ "step": 10995,
+ "train_runtime": 85043.9927,
+ "train_tokens_per_second": 5658.478
+ },
+ {
+ "epoch": 0.7406409911123081,
+ "grad_norm": 0.7035982175377999,
+ "learning_rate": 6.981603689392999e-06,
+ "loss": 0.3729727745056152,
+ "num_input_tokens_seen": 481433184,
+ "step": 11000,
+ "train_runtime": 85091.999,
+ "train_tokens_per_second": 5657.796
+ },
+ {
+ "epoch": 0.7409776461082682,
+ "grad_norm": 0.7063885505049592,
+ "learning_rate": 6.979175848547263e-06,
+ "loss": 0.3587386131286621,
+ "num_input_tokens_seen": 481648384,
+ "step": 11005,
+ "train_runtime": 85133.4023,
+ "train_tokens_per_second": 5657.572
+ },
+ {
+ "epoch": 0.7413143011042284,
+ "grad_norm": 0.818479059925005,
+ "learning_rate": 6.976747454231316e-06,
+ "loss": 0.36256041526794436,
+ "num_input_tokens_seen": 481853600,
+ "step": 11010,
+ "train_runtime": 85175.3796,
+ "train_tokens_per_second": 5657.193
+ },
+ {
+ "epoch": 0.7416509561001885,
+ "grad_norm": 0.7213537892974411,
+ "learning_rate": 6.974318507124245e-06,
+ "loss": 0.37796127796173096,
+ "num_input_tokens_seen": 482067848,
+ "step": 11015,
+ "train_runtime": 85211.3126,
+ "train_tokens_per_second": 5657.322
+ },
+ {
+ "epoch": 0.7419876110961486,
+ "grad_norm": 0.9553094338809663,
+ "learning_rate": 6.971889007905299e-06,
+ "loss": 0.38568458557128904,
+ "num_input_tokens_seen": 482272608,
+ "step": 11020,
+ "train_runtime": 85242.7243,
+ "train_tokens_per_second": 5657.64
+ },
+ {
+ "epoch": 0.7423242660921088,
+ "grad_norm": 0.6954588658200586,
+ "learning_rate": 6.9694589572538795e-06,
+ "loss": 0.4131626605987549,
+ "num_input_tokens_seen": 482500568,
+ "step": 11025,
+ "train_runtime": 85279.0869,
+ "train_tokens_per_second": 5657.9
+ },
+ {
+ "epoch": 0.7426609210880689,
+ "grad_norm": 0.6724190701417789,
+ "learning_rate": 6.9670283558495436e-06,
+ "loss": 0.3677205085754395,
+ "num_input_tokens_seen": 482727520,
+ "step": 11030,
+ "train_runtime": 85315.8408,
+ "train_tokens_per_second": 5658.123
+ },
+ {
+ "epoch": 0.742997576084029,
+ "grad_norm": 0.7736013776309341,
+ "learning_rate": 6.964597204371999e-06,
+ "loss": 0.3734050035476685,
+ "num_input_tokens_seen": 482943600,
+ "step": 11035,
+ "train_runtime": 85349.9838,
+ "train_tokens_per_second": 5658.391
+ },
+ {
+ "epoch": 0.7433342310799892,
+ "grad_norm": 0.7557593534274823,
+ "learning_rate": 6.962165503501111e-06,
+ "loss": 0.3419958591461182,
+ "num_input_tokens_seen": 483163040,
+ "step": 11040,
+ "train_runtime": 85392.7525,
+ "train_tokens_per_second": 5658.127
+ },
+ {
+ "epoch": 0.7436708860759493,
+ "grad_norm": 0.786614283633,
+ "learning_rate": 6.959733253916899e-06,
+ "loss": 0.39629015922546384,
+ "num_input_tokens_seen": 483387400,
+ "step": 11045,
+ "train_runtime": 85430.316,
+ "train_tokens_per_second": 5658.265
+ },
+ {
+ "epoch": 0.7440075410719095,
+ "grad_norm": 0.7118679551230915,
+ "learning_rate": 6.95730045629953e-06,
+ "loss": 0.3637987613677979,
+ "num_input_tokens_seen": 483586168,
+ "step": 11050,
+ "train_runtime": 85469.2598,
+ "train_tokens_per_second": 5658.013
+ },
+ {
+ "epoch": 0.7443441960678696,
+ "grad_norm": 0.775880823500199,
+ "learning_rate": 6.954867111329329e-06,
+ "loss": 0.3997496604919434,
+ "num_input_tokens_seen": 483812040,
+ "step": 11055,
+ "train_runtime": 85501.004,
+ "train_tokens_per_second": 5658.554
+ },
+ {
+ "epoch": 0.7446808510638298,
+ "grad_norm": 0.8104111912789507,
+ "learning_rate": 6.952433219686775e-06,
+ "loss": 0.36485040187835693,
+ "num_input_tokens_seen": 484033200,
+ "step": 11060,
+ "train_runtime": 85544.4427,
+ "train_tokens_per_second": 5658.266
+ },
+ {
+ "epoch": 0.7450175060597899,
+ "grad_norm": 0.745606714965393,
+ "learning_rate": 6.949998782052497e-06,
+ "loss": 0.38755149841308595,
+ "num_input_tokens_seen": 484251832,
+ "step": 11065,
+ "train_runtime": 85584.405,
+ "train_tokens_per_second": 5658.178
+ },
+ {
+ "epoch": 0.74535416105575,
+ "grad_norm": 0.7941311038883269,
+ "learning_rate": 6.947563799107277e-06,
+ "loss": 0.36263265609741213,
+ "num_input_tokens_seen": 484460952,
+ "step": 11070,
+ "train_runtime": 85617.994,
+ "train_tokens_per_second": 5658.401
+ },
+ {
+ "epoch": 0.7456908160517102,
+ "grad_norm": 0.795109174361431,
+ "learning_rate": 6.945128271532052e-06,
+ "loss": 0.37139124870300294,
+ "num_input_tokens_seen": 484696168,
+ "step": 11075,
+ "train_runtime": 85658.1904,
+ "train_tokens_per_second": 5658.492
+ },
+ {
+ "epoch": 0.7460274710476703,
+ "grad_norm": 0.7555629636652617,
+ "learning_rate": 6.942692200007908e-06,
+ "loss": 0.37421250343322754,
+ "num_input_tokens_seen": 484914304,
+ "step": 11080,
+ "train_runtime": 85693.7663,
+ "train_tokens_per_second": 5658.688
+ },
+ {
+ "epoch": 0.7463641260436304,
+ "grad_norm": 0.7710680161258006,
+ "learning_rate": 6.9402555852160845e-06,
+ "loss": 0.3868663787841797,
+ "num_input_tokens_seen": 485140952,
+ "step": 11085,
+ "train_runtime": 85733.9493,
+ "train_tokens_per_second": 5658.68
+ },
+ {
+ "epoch": 0.7467007810395906,
+ "grad_norm": 0.6618662744692935,
+ "learning_rate": 6.937818427837975e-06,
+ "loss": 0.35773367881774903,
+ "num_input_tokens_seen": 485364088,
+ "step": 11090,
+ "train_runtime": 85775.7043,
+ "train_tokens_per_second": 5658.526
+ },
+ {
+ "epoch": 0.7470374360355507,
+ "grad_norm": 0.8138674839150098,
+ "learning_rate": 6.93538072855512e-06,
+ "loss": 0.40668773651123047,
+ "num_input_tokens_seen": 485585456,
+ "step": 11095,
+ "train_runtime": 85811.2509,
+ "train_tokens_per_second": 5658.762
+ },
+ {
+ "epoch": 0.7473740910315109,
+ "grad_norm": 0.7237738810724516,
+ "learning_rate": 6.932942488049217e-06,
+ "loss": 0.3412586212158203,
+ "num_input_tokens_seen": 485799760,
+ "step": 11100,
+ "train_runtime": 85848.9089,
+ "train_tokens_per_second": 5658.776
+ },
+ {
+ "epoch": 0.747710746027471,
+ "grad_norm": 0.7192380582338419,
+ "learning_rate": 6.930503707002108e-06,
+ "loss": 0.39055523872375486,
+ "num_input_tokens_seen": 486017000,
+ "step": 11105,
+ "train_runtime": 85893.6142,
+ "train_tokens_per_second": 5658.36
+ },
+ {
+ "epoch": 0.7480474010234311,
+ "grad_norm": 0.7129551106844463,
+ "learning_rate": 6.928064386095796e-06,
+ "loss": 0.3696824789047241,
+ "num_input_tokens_seen": 486243416,
+ "step": 11110,
+ "train_runtime": 85934.5192,
+ "train_tokens_per_second": 5658.301
+ },
+ {
+ "epoch": 0.7483840560193913,
+ "grad_norm": 0.8296106328518675,
+ "learning_rate": 6.9256245260124264e-06,
+ "loss": 0.369446325302124,
+ "num_input_tokens_seen": 486469728,
+ "step": 11115,
+ "train_runtime": 85966.8967,
+ "train_tokens_per_second": 5658.803
+ },
+ {
+ "epoch": 0.7487207110153514,
+ "grad_norm": 0.8112267099332857,
+ "learning_rate": 6.923184127434298e-06,
+ "loss": 0.3864800453186035,
+ "num_input_tokens_seen": 486670960,
+ "step": 11120,
+ "train_runtime": 86001.8182,
+ "train_tokens_per_second": 5658.845
+ },
+ {
+ "epoch": 0.7490573660113116,
+ "grad_norm": 0.8933045922508607,
+ "learning_rate": 6.920743191043861e-06,
+ "loss": 0.35923840999603274,
+ "num_input_tokens_seen": 486881136,
+ "step": 11125,
+ "train_runtime": 86039.0507,
+ "train_tokens_per_second": 5658.839
+ },
+ {
+ "epoch": 0.7493940210072717,
+ "grad_norm": 0.6534193785965624,
+ "learning_rate": 6.918301717523714e-06,
+ "loss": 0.3482653617858887,
+ "num_input_tokens_seen": 487095072,
+ "step": 11130,
+ "train_runtime": 86077.5734,
+ "train_tokens_per_second": 5658.792
+ },
+ {
+ "epoch": 0.7497306760032318,
+ "grad_norm": 0.7160660990648459,
+ "learning_rate": 6.915859707556611e-06,
+ "loss": 0.3809635639190674,
+ "num_input_tokens_seen": 487312640,
+ "step": 11135,
+ "train_runtime": 86112.0956,
+ "train_tokens_per_second": 5659.05
+ },
+ {
+ "epoch": 0.750067330999192,
+ "grad_norm": 0.7880329367953478,
+ "learning_rate": 6.913417161825449e-06,
+ "loss": 0.40050134658813474,
+ "num_input_tokens_seen": 487525824,
+ "step": 11140,
+ "train_runtime": 86145.5605,
+ "train_tokens_per_second": 5659.326
+ },
+ {
+ "epoch": 0.7504039859951521,
+ "grad_norm": 0.7445156362898442,
+ "learning_rate": 6.91097408101328e-06,
+ "loss": 0.3789609432220459,
+ "num_input_tokens_seen": 487742328,
+ "step": 11145,
+ "train_runtime": 86186.4755,
+ "train_tokens_per_second": 5659.152
+ },
+ {
+ "epoch": 0.7507406409911123,
+ "grad_norm": 0.8940165351109152,
+ "learning_rate": 6.908530465803302e-06,
+ "loss": 0.370347785949707,
+ "num_input_tokens_seen": 487949632,
+ "step": 11150,
+ "train_runtime": 86217.4676,
+ "train_tokens_per_second": 5659.522
+ },
+ {
+ "epoch": 0.7510772959870724,
+ "grad_norm": 0.8568185419619945,
+ "learning_rate": 6.906086316878865e-06,
+ "loss": 0.3664633989334106,
+ "num_input_tokens_seen": 488152136,
+ "step": 11155,
+ "train_runtime": 86258.4672,
+ "train_tokens_per_second": 5659.179
+ },
+ {
+ "epoch": 0.7514139509830325,
+ "grad_norm": 0.6561053217546615,
+ "learning_rate": 6.903641634923468e-06,
+ "loss": 0.3477164268493652,
+ "num_input_tokens_seen": 488367048,
+ "step": 11160,
+ "train_runtime": 86297.7659,
+ "train_tokens_per_second": 5659.093
+ },
+ {
+ "epoch": 0.7517506059789927,
+ "grad_norm": 0.7394466907995108,
+ "learning_rate": 6.9011964206207585e-06,
+ "loss": 0.3994622230529785,
+ "num_input_tokens_seen": 488567560,
+ "step": 11165,
+ "train_runtime": 86330.4394,
+ "train_tokens_per_second": 5659.273
+ },
+ {
+ "epoch": 0.7520872609749528,
+ "grad_norm": 0.7890852129954726,
+ "learning_rate": 6.898750674654531e-06,
+ "loss": 0.36271369457244873,
+ "num_input_tokens_seen": 488788328,
+ "step": 11170,
+ "train_runtime": 86369.9487,
+ "train_tokens_per_second": 5659.241
+ },
+ {
+ "epoch": 0.752423915970913,
+ "grad_norm": 0.6327317994824692,
+ "learning_rate": 6.896304397708733e-06,
+ "loss": 0.35087008476257325,
+ "num_input_tokens_seen": 488995192,
+ "step": 11175,
+ "train_runtime": 86408.8053,
+ "train_tokens_per_second": 5659.09
+ },
+ {
+ "epoch": 0.7527605709668731,
+ "grad_norm": 0.7895910131617881,
+ "learning_rate": 6.893857590467456e-06,
+ "loss": 0.3524162292480469,
+ "num_input_tokens_seen": 489213016,
+ "step": 11180,
+ "train_runtime": 86445.1307,
+ "train_tokens_per_second": 5659.232
+ },
+ {
+ "epoch": 0.7530972259628332,
+ "grad_norm": 0.7380392223605537,
+ "learning_rate": 6.891410253614944e-06,
+ "loss": 0.34860227108001707,
+ "num_input_tokens_seen": 489418272,
+ "step": 11185,
+ "train_runtime": 86484.3079,
+ "train_tokens_per_second": 5659.041
+ },
+ {
+ "epoch": 0.7534338809587934,
+ "grad_norm": 0.702987221857376,
+ "learning_rate": 6.888962387835586e-06,
+ "loss": 0.384684419631958,
+ "num_input_tokens_seen": 489652568,
+ "step": 11190,
+ "train_runtime": 86517.2831,
+ "train_tokens_per_second": 5659.593
+ },
+ {
+ "epoch": 0.7537705359547535,
+ "grad_norm": 0.7485124246366612,
+ "learning_rate": 6.886513993813919e-06,
+ "loss": 0.38306422233581544,
+ "num_input_tokens_seen": 489864096,
+ "step": 11195,
+ "train_runtime": 86557.4583,
+ "train_tokens_per_second": 5659.409
+ },
+ {
+ "epoch": 0.7541071909507137,
+ "grad_norm": 0.7067279867145273,
+ "learning_rate": 6.88406507223463e-06,
+ "loss": 0.36569690704345703,
+ "num_input_tokens_seen": 490092880,
+ "step": 11200,
+ "train_runtime": 86594.5272,
+ "train_tokens_per_second": 5659.629
+ },
+ {
+ "epoch": 0.7544438459466738,
+ "grad_norm": 0.7595465228382173,
+ "learning_rate": 6.881615623782551e-06,
+ "loss": 0.35335478782653806,
+ "num_input_tokens_seen": 490301440,
+ "step": 11205,
+ "train_runtime": 86631.5997,
+ "train_tokens_per_second": 5659.614
+ },
+ {
+ "epoch": 0.7547805009426339,
+ "grad_norm": 0.7480898301661179,
+ "learning_rate": 6.879165649142663e-06,
+ "loss": 0.3659096956253052,
+ "num_input_tokens_seen": 490500792,
+ "step": 11210,
+ "train_runtime": 86669.1193,
+ "train_tokens_per_second": 5659.464
+ },
+ {
+ "epoch": 0.7551171559385941,
+ "grad_norm": 0.6467936248180215,
+ "learning_rate": 6.876715149000093e-06,
+ "loss": 0.36233794689178467,
+ "num_input_tokens_seen": 490705136,
+ "step": 11215,
+ "train_runtime": 86708.2047,
+ "train_tokens_per_second": 5659.27
+ },
+ {
+ "epoch": 0.7554538109345542,
+ "grad_norm": 0.7242780017786447,
+ "learning_rate": 6.874264124040117e-06,
+ "loss": 0.3791798114776611,
+ "num_input_tokens_seen": 490930904,
+ "step": 11220,
+ "train_runtime": 86744.2352,
+ "train_tokens_per_second": 5659.522
+ },
+ {
+ "epoch": 0.7557904659305144,
+ "grad_norm": 0.7534511874997005,
+ "learning_rate": 6.871812574948154e-06,
+ "loss": 0.394960880279541,
+ "num_input_tokens_seen": 491158544,
+ "step": 11225,
+ "train_runtime": 86777.1348,
+ "train_tokens_per_second": 5660.0
+ },
+ {
+ "epoch": 0.7561271209264745,
+ "grad_norm": 0.7151778561469025,
+ "learning_rate": 6.869360502409773e-06,
+ "loss": 0.3845766544342041,
+ "num_input_tokens_seen": 491385232,
+ "step": 11230,
+ "train_runtime": 86814.8579,
+ "train_tokens_per_second": 5660.151
+ },
+ {
+ "epoch": 0.7564637759224346,
+ "grad_norm": 0.7179238868789197,
+ "learning_rate": 6.866907907110689e-06,
+ "loss": 0.351468563079834,
+ "num_input_tokens_seen": 491593976,
+ "step": 11235,
+ "train_runtime": 86852.7406,
+ "train_tokens_per_second": 5660.086
+ },
+ {
+ "epoch": 0.7568004309183949,
+ "grad_norm": 0.6534240517661798,
+ "learning_rate": 6.864454789736763e-06,
+ "loss": 0.3472340822219849,
+ "num_input_tokens_seen": 491827504,
+ "step": 11240,
+ "train_runtime": 86890.0801,
+ "train_tokens_per_second": 5660.341
+ },
+ {
+ "epoch": 0.757137085914355,
+ "grad_norm": 0.6984388211704616,
+ "learning_rate": 6.8620011509740005e-06,
+ "loss": 0.3707188844680786,
+ "num_input_tokens_seen": 492042264,
+ "step": 11245,
+ "train_runtime": 86929.2774,
+ "train_tokens_per_second": 5660.259
+ },
+ {
+ "epoch": 0.7574737409103152,
+ "grad_norm": 0.8438964758084152,
+ "learning_rate": 6.859546991508552e-06,
+ "loss": 0.3541130542755127,
+ "num_input_tokens_seen": 492261592,
+ "step": 11250,
+ "train_runtime": 86971.6043,
+ "train_tokens_per_second": 5660.027
+ },
+ {
+ "epoch": 0.7578103959062753,
+ "grad_norm": 0.6972841594933251,
+ "learning_rate": 6.857092312026716e-06,
+ "loss": 0.3816833972930908,
+ "num_input_tokens_seen": 492495240,
+ "step": 11255,
+ "train_runtime": 87006.7607,
+ "train_tokens_per_second": 5660.425
+ },
+ {
+ "epoch": 0.7581470509022354,
+ "grad_norm": 0.7161736954331712,
+ "learning_rate": 6.8546371132149395e-06,
+ "loss": 0.39476380348205564,
+ "num_input_tokens_seen": 492712328,
+ "step": 11260,
+ "train_runtime": 87046.3192,
+ "train_tokens_per_second": 5660.346
+ },
+ {
+ "epoch": 0.7584837058981956,
+ "grad_norm": 0.7348844948766989,
+ "learning_rate": 6.852181395759807e-06,
+ "loss": 0.39330875873565674,
+ "num_input_tokens_seen": 492943248,
+ "step": 11265,
+ "train_runtime": 87087.1631,
+ "train_tokens_per_second": 5660.343
+ },
+ {
+ "epoch": 0.7588203608941557,
+ "grad_norm": 0.5683273529508269,
+ "learning_rate": 6.849725160348054e-06,
+ "loss": 0.3455686330795288,
+ "num_input_tokens_seen": 493152344,
+ "step": 11270,
+ "train_runtime": 87127.3625,
+ "train_tokens_per_second": 5660.132
+ },
+ {
+ "epoch": 0.7591570158901159,
+ "grad_norm": 0.7258002012386351,
+ "learning_rate": 6.847268407666557e-06,
+ "loss": 0.35138664245605467,
+ "num_input_tokens_seen": 493375584,
+ "step": 11275,
+ "train_runtime": 87171.3314,
+ "train_tokens_per_second": 5659.838
+ },
+ {
+ "epoch": 0.759493670886076,
+ "grad_norm": 0.7379175392341268,
+ "learning_rate": 6.844811138402341e-06,
+ "loss": 0.37623722553253175,
+ "num_input_tokens_seen": 493594016,
+ "step": 11280,
+ "train_runtime": 87212.4908,
+ "train_tokens_per_second": 5659.671
+ },
+ {
+ "epoch": 0.7598303258820361,
+ "grad_norm": 0.7131462906095328,
+ "learning_rate": 6.842353353242574e-06,
+ "loss": 0.36553401947021485,
+ "num_input_tokens_seen": 493803080,
+ "step": 11285,
+ "train_runtime": 87254.8113,
+ "train_tokens_per_second": 5659.322
+ },
+ {
+ "epoch": 0.7601669808779963,
+ "grad_norm": 0.7404989121841616,
+ "learning_rate": 6.839895052874565e-06,
+ "loss": 0.3759953737258911,
+ "num_input_tokens_seen": 494027672,
+ "step": 11290,
+ "train_runtime": 87290.4517,
+ "train_tokens_per_second": 5659.584
+ },
+ {
+ "epoch": 0.7605036358739564,
+ "grad_norm": 0.6838203757609232,
+ "learning_rate": 6.837436237985772e-06,
+ "loss": 0.3641194343566895,
+ "num_input_tokens_seen": 494238208,
+ "step": 11295,
+ "train_runtime": 87333.9419,
+ "train_tokens_per_second": 5659.177
+ },
+ {
+ "epoch": 0.7608402908699166,
+ "grad_norm": 0.850634796737577,
+ "learning_rate": 6.834976909263794e-06,
+ "loss": 0.37904021739959715,
+ "num_input_tokens_seen": 494450104,
+ "step": 11300,
+ "train_runtime": 87372.8846,
+ "train_tokens_per_second": 5659.08
+ },
+ {
+ "epoch": 0.7611769458658767,
+ "grad_norm": 0.7504214621662669,
+ "learning_rate": 6.832517067396374e-06,
+ "loss": 0.3586588859558105,
+ "num_input_tokens_seen": 494683264,
+ "step": 11305,
+ "train_runtime": 87416.6251,
+ "train_tokens_per_second": 5658.915
+ },
+ {
+ "epoch": 0.7615136008618368,
+ "grad_norm": 0.7181780717405477,
+ "learning_rate": 6.8300567130714e-06,
+ "loss": 0.39732561111450193,
+ "num_input_tokens_seen": 494912080,
+ "step": 11310,
+ "train_runtime": 87455.7279,
+ "train_tokens_per_second": 5659.001
+ },
+ {
+ "epoch": 0.761850255857797,
+ "grad_norm": 0.8326466499790485,
+ "learning_rate": 6.8275958469769035e-06,
+ "loss": 0.38528354167938234,
+ "num_input_tokens_seen": 495143768,
+ "step": 11315,
+ "train_runtime": 87496.2946,
+ "train_tokens_per_second": 5659.026
+ },
+ {
+ "epoch": 0.7621869108537571,
+ "grad_norm": 0.8043440720773757,
+ "learning_rate": 6.825134469801055e-06,
+ "loss": 0.3580173969268799,
+ "num_input_tokens_seen": 495373776,
+ "step": 11320,
+ "train_runtime": 87536.7056,
+ "train_tokens_per_second": 5659.041
+ },
+ {
+ "epoch": 0.7625235658497173,
+ "grad_norm": 0.7612366109464449,
+ "learning_rate": 6.8226725822321715e-06,
+ "loss": 0.4001301765441895,
+ "num_input_tokens_seen": 495581360,
+ "step": 11325,
+ "train_runtime": 87574.6303,
+ "train_tokens_per_second": 5658.96
+ },
+ {
+ "epoch": 0.7628602208456774,
+ "grad_norm": 0.7254035816528333,
+ "learning_rate": 6.820210184958712e-06,
+ "loss": 0.38148252964019774,
+ "num_input_tokens_seen": 495805664,
+ "step": 11330,
+ "train_runtime": 87611.2812,
+ "train_tokens_per_second": 5659.153
+ },
+ {
+ "epoch": 0.7631968758416375,
+ "grad_norm": 0.7816370531432046,
+ "learning_rate": 6.81774727866928e-06,
+ "loss": 0.38228158950805663,
+ "num_input_tokens_seen": 496015888,
+ "step": 11335,
+ "train_runtime": 87647.9698,
+ "train_tokens_per_second": 5659.183
+ },
+ {
+ "epoch": 0.7635335308375977,
+ "grad_norm": 0.9719681651052307,
+ "learning_rate": 6.815283864052617e-06,
+ "loss": 0.3840935468673706,
+ "num_input_tokens_seen": 496237176,
+ "step": 11340,
+ "train_runtime": 87687.7469,
+ "train_tokens_per_second": 5659.139
+ },
+ {
+ "epoch": 0.7638701858335578,
+ "grad_norm": 0.7821628809832356,
+ "learning_rate": 6.812819941797611e-06,
+ "loss": 0.38556504249572754,
+ "num_input_tokens_seen": 496442520,
+ "step": 11345,
+ "train_runtime": 87730.1254,
+ "train_tokens_per_second": 5658.746
+ },
+ {
+ "epoch": 0.764206840829518,
+ "grad_norm": 0.7793739338464833,
+ "learning_rate": 6.810355512593288e-06,
+ "loss": 0.3861824035644531,
+ "num_input_tokens_seen": 496669864,
+ "step": 11350,
+ "train_runtime": 87767.0891,
+ "train_tokens_per_second": 5658.953
+ },
+ {
+ "epoch": 0.7645434958254781,
+ "grad_norm": 0.7769049083652588,
+ "learning_rate": 6.807890577128817e-06,
+ "loss": 0.33897886276245115,
+ "num_input_tokens_seen": 496892112,
+ "step": 11355,
+ "train_runtime": 87813.3112,
+ "train_tokens_per_second": 5658.506
+ },
+ {
+ "epoch": 0.7648801508214382,
+ "grad_norm": 0.8235786764795517,
+ "learning_rate": 6.805425136093513e-06,
+ "loss": 0.3582730531692505,
+ "num_input_tokens_seen": 497111304,
+ "step": 11360,
+ "train_runtime": 87847.6951,
+ "train_tokens_per_second": 5658.786
+ },
+ {
+ "epoch": 0.7652168058173984,
+ "grad_norm": 0.825724303695909,
+ "learning_rate": 6.802959190176827e-06,
+ "loss": 0.38888885974884035,
+ "num_input_tokens_seen": 497318752,
+ "step": 11365,
+ "train_runtime": 87886.6126,
+ "train_tokens_per_second": 5658.641
+ },
+ {
+ "epoch": 0.7655534608133585,
+ "grad_norm": 0.7692237966789585,
+ "learning_rate": 6.800492740068352e-06,
+ "loss": 0.3655484676361084,
+ "num_input_tokens_seen": 497523256,
+ "step": 11370,
+ "train_runtime": 87923.25,
+ "train_tokens_per_second": 5658.609
+ },
+ {
+ "epoch": 0.7658901158093187,
+ "grad_norm": 0.8501602672533506,
+ "learning_rate": 6.798025786457825e-06,
+ "loss": 0.35955286026000977,
+ "num_input_tokens_seen": 497749392,
+ "step": 11375,
+ "train_runtime": 87963.9631,
+ "train_tokens_per_second": 5658.56
+ },
+ {
+ "epoch": 0.7662267708052788,
+ "grad_norm": 0.6654629734284263,
+ "learning_rate": 6.795558330035118e-06,
+ "loss": 0.3689566612243652,
+ "num_input_tokens_seen": 497978632,
+ "step": 11380,
+ "train_runtime": 88010.5357,
+ "train_tokens_per_second": 5658.171
+ },
+ {
+ "epoch": 0.7665634258012389,
+ "grad_norm": 0.571714653501898,
+ "learning_rate": 6.793090371490251e-06,
+ "loss": 0.37029447555541994,
+ "num_input_tokens_seen": 498211416,
+ "step": 11385,
+ "train_runtime": 88043.3102,
+ "train_tokens_per_second": 5658.708
+ },
+ {
+ "epoch": 0.7669000807971991,
+ "grad_norm": 0.6562894791866329,
+ "learning_rate": 6.790621911513378e-06,
+ "loss": 0.36082067489624026,
+ "num_input_tokens_seen": 498430376,
+ "step": 11390,
+ "train_runtime": 88076.9318,
+ "train_tokens_per_second": 5659.034
+ },
+ {
+ "epoch": 0.7672367357931592,
+ "grad_norm": 0.6830456569016362,
+ "learning_rate": 6.788152950794799e-06,
+ "loss": 0.3912408113479614,
+ "num_input_tokens_seen": 498646776,
+ "step": 11395,
+ "train_runtime": 88113.1004,
+ "train_tokens_per_second": 5659.167
+ },
+ {
+ "epoch": 0.7675733907891193,
+ "grad_norm": 0.8063375531181857,
+ "learning_rate": 6.7856834900249495e-06,
+ "loss": 0.3919992208480835,
+ "num_input_tokens_seen": 498865144,
+ "step": 11400,
+ "train_runtime": 88157.9923,
+ "train_tokens_per_second": 5658.763
+ },
+ {
+ "epoch": 0.7679100457850795,
+ "grad_norm": 0.7488462821998634,
+ "learning_rate": 6.783213529894405e-06,
+ "loss": 0.3473628044128418,
+ "num_input_tokens_seen": 499082328,
+ "step": 11405,
+ "train_runtime": 88202.6214,
+ "train_tokens_per_second": 5658.362
+ },
+ {
+ "epoch": 0.7682467007810396,
+ "grad_norm": 0.7845757383023176,
+ "learning_rate": 6.780743071093883e-06,
+ "loss": 0.35947589874267577,
+ "num_input_tokens_seen": 499306216,
+ "step": 11410,
+ "train_runtime": 88242.1656,
+ "train_tokens_per_second": 5658.363
+ },
+ {
+ "epoch": 0.7685833557769998,
+ "grad_norm": 1.4808480239457869,
+ "learning_rate": 6.77827211431424e-06,
+ "loss": 0.36936905384063723,
+ "num_input_tokens_seen": 499532072,
+ "step": 11415,
+ "train_runtime": 88280.0767,
+ "train_tokens_per_second": 5658.492
+ },
+ {
+ "epoch": 0.7689200107729599,
+ "grad_norm": 0.7193580497484903,
+ "learning_rate": 6.775800660246471e-06,
+ "loss": 0.36505095958709716,
+ "num_input_tokens_seen": 499732272,
+ "step": 11420,
+ "train_runtime": 88316.7765,
+ "train_tokens_per_second": 5658.407
+ },
+ {
+ "epoch": 0.76925666576892,
+ "grad_norm": 0.7225028793880158,
+ "learning_rate": 6.77332870958171e-06,
+ "loss": 0.3719770908355713,
+ "num_input_tokens_seen": 499963848,
+ "step": 11425,
+ "train_runtime": 88361.2969,
+ "train_tokens_per_second": 5658.177
+ },
+ {
+ "epoch": 0.7695933207648802,
+ "grad_norm": 0.7310175751597672,
+ "learning_rate": 6.770856263011229e-06,
+ "loss": 0.3673600196838379,
+ "num_input_tokens_seen": 500175216,
+ "step": 11430,
+ "train_runtime": 88394.2396,
+ "train_tokens_per_second": 5658.459
+ },
+ {
+ "epoch": 0.7699299757608403,
+ "grad_norm": 0.8011531340232532,
+ "learning_rate": 6.768383321226442e-06,
+ "loss": 0.38931114673614503,
+ "num_input_tokens_seen": 500380304,
+ "step": 11435,
+ "train_runtime": 88432.6121,
+ "train_tokens_per_second": 5658.323
+ },
+ {
+ "epoch": 0.7702666307568005,
+ "grad_norm": 1.540406850238615,
+ "learning_rate": 6.765909884918898e-06,
+ "loss": 0.3583209753036499,
+ "num_input_tokens_seen": 500590256,
+ "step": 11440,
+ "train_runtime": 88469.3611,
+ "train_tokens_per_second": 5658.346
+ },
+ {
+ "epoch": 0.7706032857527606,
+ "grad_norm": 0.6645503979893219,
+ "learning_rate": 6.763435954780286e-06,
+ "loss": 0.3920794725418091,
+ "num_input_tokens_seen": 500810896,
+ "step": 11445,
+ "train_runtime": 88504.9842,
+ "train_tokens_per_second": 5658.561
+ },
+ {
+ "epoch": 0.7709399407487207,
+ "grad_norm": 0.7669134143397409,
+ "learning_rate": 6.7609615315024325e-06,
+ "loss": 0.37000212669372556,
+ "num_input_tokens_seen": 501027440,
+ "step": 11450,
+ "train_runtime": 88543.2243,
+ "train_tokens_per_second": 5658.563
+ },
+ {
+ "epoch": 0.7712765957446809,
+ "grad_norm": 0.7415266590401455,
+ "learning_rate": 6.758486615777301e-06,
+ "loss": 0.37416849136352537,
+ "num_input_tokens_seen": 501241576,
+ "step": 11455,
+ "train_runtime": 88583.8287,
+ "train_tokens_per_second": 5658.387
+ },
+ {
+ "epoch": 0.771613250740641,
+ "grad_norm": 0.8407214839390016,
+ "learning_rate": 6.756011208296995e-06,
+ "loss": 0.3626854419708252,
+ "num_input_tokens_seen": 501454696,
+ "step": 11460,
+ "train_runtime": 88621.4278,
+ "train_tokens_per_second": 5658.391
+ },
+ {
+ "epoch": 0.7719499057366012,
+ "grad_norm": 0.7461388799032196,
+ "learning_rate": 6.753535309753755e-06,
+ "loss": 0.35544500350952146,
+ "num_input_tokens_seen": 501658528,
+ "step": 11465,
+ "train_runtime": 88662.546,
+ "train_tokens_per_second": 5658.066
+ },
+ {
+ "epoch": 0.7722865607325613,
+ "grad_norm": 0.6919457018289741,
+ "learning_rate": 6.7510589208399565e-06,
+ "loss": 0.3747244119644165,
+ "num_input_tokens_seen": 501877728,
+ "step": 11470,
+ "train_runtime": 88703.6872,
+ "train_tokens_per_second": 5657.913
+ },
+ {
+ "epoch": 0.7726232157285214,
+ "grad_norm": 0.7585317880232398,
+ "learning_rate": 6.748582042248115e-06,
+ "loss": 0.35962686538696287,
+ "num_input_tokens_seen": 502089032,
+ "step": 11475,
+ "train_runtime": 88740.6334,
+ "train_tokens_per_second": 5657.938
+ },
+ {
+ "epoch": 0.7729598707244816,
+ "grad_norm": 0.7198688949375671,
+ "learning_rate": 6.746104674670882e-06,
+ "loss": 0.38360817432403566,
+ "num_input_tokens_seen": 502313408,
+ "step": 11480,
+ "train_runtime": 88777.3916,
+ "train_tokens_per_second": 5658.123
+ },
+ {
+ "epoch": 0.7732965257204417,
+ "grad_norm": 0.6976205058667109,
+ "learning_rate": 6.743626818801042e-06,
+ "loss": 0.39516823291778563,
+ "num_input_tokens_seen": 502534360,
+ "step": 11485,
+ "train_runtime": 88818.4462,
+ "train_tokens_per_second": 5657.995
+ },
+ {
+ "epoch": 0.7736331807164019,
+ "grad_norm": 0.6984390114365983,
+ "learning_rate": 6.741148475331525e-06,
+ "loss": 0.37048795223236086,
+ "num_input_tokens_seen": 502757200,
+ "step": 11490,
+ "train_runtime": 88854.2343,
+ "train_tokens_per_second": 5658.224
+ },
+ {
+ "epoch": 0.773969835712362,
+ "grad_norm": 0.6953225919926989,
+ "learning_rate": 6.738669644955388e-06,
+ "loss": 0.3468164443969727,
+ "num_input_tokens_seen": 502970848,
+ "step": 11495,
+ "train_runtime": 88897.1339,
+ "train_tokens_per_second": 5657.897
+ },
+ {
+ "epoch": 0.7743064907083221,
+ "grad_norm": 0.7656686554212365,
+ "learning_rate": 6.736190328365828e-06,
+ "loss": 0.3672001838684082,
+ "num_input_tokens_seen": 503195432,
+ "step": 11500,
+ "train_runtime": 88938.4411,
+ "train_tokens_per_second": 5657.795
+ },
+ {
+ "epoch": 0.7746431457042823,
+ "grad_norm": 0.743345653178439,
+ "learning_rate": 6.73371052625618e-06,
+ "loss": 0.37738971710205077,
+ "num_input_tokens_seen": 503417816,
+ "step": 11505,
+ "train_runtime": 88979.7242,
+ "train_tokens_per_second": 5657.669
+ },
+ {
+ "epoch": 0.7749798007002424,
+ "grad_norm": 0.7491341805650183,
+ "learning_rate": 6.731230239319912e-06,
+ "loss": 0.37387096881866455,
+ "num_input_tokens_seen": 503642040,
+ "step": 11510,
+ "train_runtime": 89021.4956,
+ "train_tokens_per_second": 5657.533
+ },
+ {
+ "epoch": 0.7753164556962026,
+ "grad_norm": 0.6256746258505268,
+ "learning_rate": 6.728749468250627e-06,
+ "loss": 0.38176569938659666,
+ "num_input_tokens_seen": 503870440,
+ "step": 11515,
+ "train_runtime": 89058.3538,
+ "train_tokens_per_second": 5657.756
+ },
+ {
+ "epoch": 0.7756531106921627,
+ "grad_norm": 0.8527701221713213,
+ "learning_rate": 6.726268213742066e-06,
+ "loss": 0.375903582572937,
+ "num_input_tokens_seen": 504076280,
+ "step": 11520,
+ "train_runtime": 89100.0272,
+ "train_tokens_per_second": 5657.42
+ },
+ {
+ "epoch": 0.7759897656881228,
+ "grad_norm": 0.7746670396544606,
+ "learning_rate": 6.723786476488105e-06,
+ "loss": 0.38536639213562013,
+ "num_input_tokens_seen": 504303624,
+ "step": 11525,
+ "train_runtime": 89134.096,
+ "train_tokens_per_second": 5657.808
+ },
+ {
+ "epoch": 0.776326420684083,
+ "grad_norm": 0.6906699278173198,
+ "learning_rate": 6.72130425718275e-06,
+ "loss": 0.3504000186920166,
+ "num_input_tokens_seen": 504511120,
+ "step": 11530,
+ "train_runtime": 89170.4494,
+ "train_tokens_per_second": 5657.829
+ },
+ {
+ "epoch": 0.7766630756800431,
+ "grad_norm": 0.7640134079354167,
+ "learning_rate": 6.718821556520151e-06,
+ "loss": 0.3902331829071045,
+ "num_input_tokens_seen": 504720744,
+ "step": 11535,
+ "train_runtime": 89203.783,
+ "train_tokens_per_second": 5658.064
+ },
+ {
+ "epoch": 0.7769997306760033,
+ "grad_norm": 0.676852860064436,
+ "learning_rate": 6.716338375194583e-06,
+ "loss": 0.3558895826339722,
+ "num_input_tokens_seen": 504945672,
+ "step": 11540,
+ "train_runtime": 89239.0086,
+ "train_tokens_per_second": 5658.351
+ },
+ {
+ "epoch": 0.7773363856719634,
+ "grad_norm": 0.7081316231079949,
+ "learning_rate": 6.713854713900463e-06,
+ "loss": 0.39481263160705565,
+ "num_input_tokens_seen": 505168112,
+ "step": 11545,
+ "train_runtime": 89280.0653,
+ "train_tokens_per_second": 5658.241
+ },
+ {
+ "epoch": 0.7776730406679235,
+ "grad_norm": 0.7771142918437934,
+ "learning_rate": 6.7113705733323365e-06,
+ "loss": 0.3915176153182983,
+ "num_input_tokens_seen": 505373904,
+ "step": 11550,
+ "train_runtime": 89321.9305,
+ "train_tokens_per_second": 5657.893
+ },
+ {
+ "epoch": 0.7780096956638837,
+ "grad_norm": 0.7862100764061776,
+ "learning_rate": 6.708885954184885e-06,
+ "loss": 0.36997599601745607,
+ "num_input_tokens_seen": 505592960,
+ "step": 11555,
+ "train_runtime": 89362.1395,
+ "train_tokens_per_second": 5657.798
+ },
+ {
+ "epoch": 0.7783463506598438,
+ "grad_norm": 0.8044048640519208,
+ "learning_rate": 6.7064008571529275e-06,
+ "loss": 0.3574796199798584,
+ "num_input_tokens_seen": 505815992,
+ "step": 11560,
+ "train_runtime": 89407.5364,
+ "train_tokens_per_second": 5657.42
+ },
+ {
+ "epoch": 0.778683005655804,
+ "grad_norm": 0.63451019489799,
+ "learning_rate": 6.703915282931411e-06,
+ "loss": 0.37357444763183595,
+ "num_input_tokens_seen": 506041792,
+ "step": 11565,
+ "train_runtime": 89452.1107,
+ "train_tokens_per_second": 5657.125
+ },
+ {
+ "epoch": 0.7790196606517641,
+ "grad_norm": 0.7400296332193441,
+ "learning_rate": 6.701429232215418e-06,
+ "loss": 0.3822374105453491,
+ "num_input_tokens_seen": 506244224,
+ "step": 11570,
+ "train_runtime": 89491.3476,
+ "train_tokens_per_second": 5656.907
+ },
+ {
+ "epoch": 0.7793563156477242,
+ "grad_norm": 0.8414435559233742,
+ "learning_rate": 6.6989427057001665e-06,
+ "loss": 0.3804741144180298,
+ "num_input_tokens_seen": 506470096,
+ "step": 11575,
+ "train_runtime": 89532.7906,
+ "train_tokens_per_second": 5656.811
+ },
+ {
+ "epoch": 0.7796929706436844,
+ "grad_norm": 0.6787888680159854,
+ "learning_rate": 6.696455704081006e-06,
+ "loss": 0.3787699699401855,
+ "num_input_tokens_seen": 506688376,
+ "step": 11580,
+ "train_runtime": 89574.3736,
+ "train_tokens_per_second": 5656.622
+ },
+ {
+ "epoch": 0.7800296256396445,
+ "grad_norm": 0.711221275891421,
+ "learning_rate": 6.693968228053416e-06,
+ "loss": 0.38877334594726565,
+ "num_input_tokens_seen": 506908432,
+ "step": 11585,
+ "train_runtime": 89609.8238,
+ "train_tokens_per_second": 5656.84
+ },
+ {
+ "epoch": 0.7803662806356046,
+ "grad_norm": 0.7045284132631514,
+ "learning_rate": 6.691480278313013e-06,
+ "loss": 0.37210679054260254,
+ "num_input_tokens_seen": 507123272,
+ "step": 11590,
+ "train_runtime": 89650.9072,
+ "train_tokens_per_second": 5656.644
+ },
+ {
+ "epoch": 0.7807029356315648,
+ "grad_norm": 0.718080559096589,
+ "learning_rate": 6.688991855555545e-06,
+ "loss": 0.35350847244262695,
+ "num_input_tokens_seen": 507339456,
+ "step": 11595,
+ "train_runtime": 89694.1289,
+ "train_tokens_per_second": 5656.328
+ },
+ {
+ "epoch": 0.7810395906275249,
+ "grad_norm": 0.692068019249801,
+ "learning_rate": 6.686502960476888e-06,
+ "loss": 0.3558246612548828,
+ "num_input_tokens_seen": 507577136,
+ "step": 11600,
+ "train_runtime": 89731.2991,
+ "train_tokens_per_second": 5656.634
+ },
+ {
+ "epoch": 0.7813762456234851,
+ "grad_norm": 0.7053421542399363,
+ "learning_rate": 6.684013593773058e-06,
+ "loss": 0.3307861089706421,
+ "num_input_tokens_seen": 507811288,
+ "step": 11605,
+ "train_runtime": 89766.4366,
+ "train_tokens_per_second": 5657.028
+ },
+ {
+ "epoch": 0.7817129006194452,
+ "grad_norm": 0.7619528494233906,
+ "learning_rate": 6.6815237561401955e-06,
+ "loss": 0.33051750659942625,
+ "num_input_tokens_seen": 508006304,
+ "step": 11610,
+ "train_runtime": 89802.2646,
+ "train_tokens_per_second": 5656.943
+ },
+ {
+ "epoch": 0.7820495556154053,
+ "grad_norm": 0.8056101436691785,
+ "learning_rate": 6.679033448274577e-06,
+ "loss": 0.35503997802734377,
+ "num_input_tokens_seen": 508223832,
+ "step": 11615,
+ "train_runtime": 89850.1477,
+ "train_tokens_per_second": 5656.349
+ },
+ {
+ "epoch": 0.7823862106113655,
+ "grad_norm": 0.6611022339527542,
+ "learning_rate": 6.676542670872609e-06,
+ "loss": 0.3701477527618408,
+ "num_input_tokens_seen": 508440024,
+ "step": 11620,
+ "train_runtime": 89890.5249,
+ "train_tokens_per_second": 5656.214
+ },
+ {
+ "epoch": 0.7827228656073256,
+ "grad_norm": 0.6851834640607047,
+ "learning_rate": 6.674051424630829e-06,
+ "loss": 0.36185245513916015,
+ "num_input_tokens_seen": 508666440,
+ "step": 11625,
+ "train_runtime": 89924.4436,
+ "train_tokens_per_second": 5656.598
+ },
+ {
+ "epoch": 0.7830595206032858,
+ "grad_norm": 0.7081381172789725,
+ "learning_rate": 6.671559710245905e-06,
+ "loss": 0.36898825168609617,
+ "num_input_tokens_seen": 508892304,
+ "step": 11630,
+ "train_runtime": 89970.0326,
+ "train_tokens_per_second": 5656.242
+ },
+ {
+ "epoch": 0.7833961755992459,
+ "grad_norm": 0.7864088827874502,
+ "learning_rate": 6.669067528414639e-06,
+ "loss": 0.36539783477783205,
+ "num_input_tokens_seen": 509118040,
+ "step": 11635,
+ "train_runtime": 90012.286,
+ "train_tokens_per_second": 5656.095
+ },
+ {
+ "epoch": 0.783732830595206,
+ "grad_norm": 0.7072947789302337,
+ "learning_rate": 6.6665748798339606e-06,
+ "loss": 0.39262051582336427,
+ "num_input_tokens_seen": 509346248,
+ "step": 11640,
+ "train_runtime": 90050.1585,
+ "train_tokens_per_second": 5656.25
+ },
+ {
+ "epoch": 0.7840694855911662,
+ "grad_norm": 0.7696238239442139,
+ "learning_rate": 6.664081765200932e-06,
+ "loss": 0.38922972679138185,
+ "num_input_tokens_seen": 509547192,
+ "step": 11645,
+ "train_runtime": 90082.5361,
+ "train_tokens_per_second": 5656.448
+ },
+ {
+ "epoch": 0.7844061405871263,
+ "grad_norm": 0.9564255408577432,
+ "learning_rate": 6.661588185212745e-06,
+ "loss": 0.38661224842071534,
+ "num_input_tokens_seen": 509767456,
+ "step": 11650,
+ "train_runtime": 90124.0819,
+ "train_tokens_per_second": 5656.285
+ },
+ {
+ "epoch": 0.7847427955830865,
+ "grad_norm": 0.6920498506689123,
+ "learning_rate": 6.65909414056672e-06,
+ "loss": 0.37398757934570315,
+ "num_input_tokens_seen": 509978544,
+ "step": 11655,
+ "train_runtime": 90162.88,
+ "train_tokens_per_second": 5656.192
+ },
+ {
+ "epoch": 0.7850794505790466,
+ "grad_norm": 0.7983355414544516,
+ "learning_rate": 6.6565996319603075e-06,
+ "loss": 0.36649136543273925,
+ "num_input_tokens_seen": 510194488,
+ "step": 11660,
+ "train_runtime": 90205.1793,
+ "train_tokens_per_second": 5655.933
+ },
+ {
+ "epoch": 0.7854161055750067,
+ "grad_norm": 0.705582165882806,
+ "learning_rate": 6.6541046600910915e-06,
+ "loss": 0.35945792198181153,
+ "num_input_tokens_seen": 510408368,
+ "step": 11665,
+ "train_runtime": 90242.6397,
+ "train_tokens_per_second": 5655.956
+ },
+ {
+ "epoch": 0.7857527605709669,
+ "grad_norm": 0.7659229146292734,
+ "learning_rate": 6.6516092256567824e-06,
+ "loss": 0.36875619888305666,
+ "num_input_tokens_seen": 510635536,
+ "step": 11670,
+ "train_runtime": 90279.7373,
+ "train_tokens_per_second": 5656.148
+ },
+ {
+ "epoch": 0.786089415566927,
+ "grad_norm": 0.8578375495363305,
+ "learning_rate": 6.64911332935522e-06,
+ "loss": 0.39651241302490237,
+ "num_input_tokens_seen": 510860904,
+ "step": 11675,
+ "train_runtime": 90320.5501,
+ "train_tokens_per_second": 5656.087
+ },
+ {
+ "epoch": 0.7864260705628872,
+ "grad_norm": 0.7134409162325845,
+ "learning_rate": 6.646616971884373e-06,
+ "loss": 0.3647106885910034,
+ "num_input_tokens_seen": 511085760,
+ "step": 11680,
+ "train_runtime": 90359.7652,
+ "train_tokens_per_second": 5656.121
+ },
+ {
+ "epoch": 0.7867627255588473,
+ "grad_norm": 0.6706622363930574,
+ "learning_rate": 6.64412015394234e-06,
+ "loss": 0.35194456577301025,
+ "num_input_tokens_seen": 511285624,
+ "step": 11685,
+ "train_runtime": 90397.4925,
+ "train_tokens_per_second": 5655.971
+ },
+ {
+ "epoch": 0.7870993805548074,
+ "grad_norm": 0.8444783042140578,
+ "learning_rate": 6.641622876227349e-06,
+ "loss": 0.3812446594238281,
+ "num_input_tokens_seen": 511491128,
+ "step": 11690,
+ "train_runtime": 90439.6501,
+ "train_tokens_per_second": 5655.607
+ },
+ {
+ "epoch": 0.7874360355507676,
+ "grad_norm": 0.7494745665567035,
+ "learning_rate": 6.639125139437756e-06,
+ "loss": 0.3623605012893677,
+ "num_input_tokens_seen": 511707944,
+ "step": 11695,
+ "train_runtime": 90482.4407,
+ "train_tokens_per_second": 5655.329
+ },
+ {
+ "epoch": 0.7877726905467277,
+ "grad_norm": 0.7902023749862866,
+ "learning_rate": 6.6366269442720446e-06,
+ "loss": 0.36579647064208987,
+ "num_input_tokens_seen": 511918704,
+ "step": 11700,
+ "train_runtime": 90522.6671,
+ "train_tokens_per_second": 5655.144
+ },
+ {
+ "epoch": 0.7881093455426879,
+ "grad_norm": 0.7084908330150312,
+ "learning_rate": 6.634128291428827e-06,
+ "loss": 0.4204265594482422,
+ "num_input_tokens_seen": 512130360,
+ "step": 11705,
+ "train_runtime": 90561.121,
+ "train_tokens_per_second": 5655.08
+ },
+ {
+ "epoch": 0.788446000538648,
+ "grad_norm": 0.7463863432263244,
+ "learning_rate": 6.631629181606842e-06,
+ "loss": 0.37847514152526857,
+ "num_input_tokens_seen": 512353072,
+ "step": 11710,
+ "train_runtime": 90604.8421,
+ "train_tokens_per_second": 5654.809
+ },
+ {
+ "epoch": 0.7887826555346081,
+ "grad_norm": 0.7252447688219217,
+ "learning_rate": 6.629129615504961e-06,
+ "loss": 0.3778711795806885,
+ "num_input_tokens_seen": 512578040,
+ "step": 11715,
+ "train_runtime": 90644.5653,
+ "train_tokens_per_second": 5654.813
+ },
+ {
+ "epoch": 0.7891193105305683,
+ "grad_norm": 0.6877935716049668,
+ "learning_rate": 6.626629593822178e-06,
+ "loss": 0.349456787109375,
+ "num_input_tokens_seen": 512795568,
+ "step": 11720,
+ "train_runtime": 90683.8764,
+ "train_tokens_per_second": 5654.76
+ },
+ {
+ "epoch": 0.7894559655265284,
+ "grad_norm": 0.7159875503391986,
+ "learning_rate": 6.624129117257616e-06,
+ "loss": 0.3877539157867432,
+ "num_input_tokens_seen": 513020704,
+ "step": 11725,
+ "train_runtime": 90721.6257,
+ "train_tokens_per_second": 5654.889
+ },
+ {
+ "epoch": 0.7897926205224886,
+ "grad_norm": 0.8198647095115159,
+ "learning_rate": 6.621628186510524e-06,
+ "loss": 0.37906270027160643,
+ "num_input_tokens_seen": 513238728,
+ "step": 11730,
+ "train_runtime": 90763.6807,
+ "train_tokens_per_second": 5654.671
+ },
+ {
+ "epoch": 0.7901292755184487,
+ "grad_norm": 0.6710230144644209,
+ "learning_rate": 6.619126802280281e-06,
+ "loss": 0.35970292091369627,
+ "num_input_tokens_seen": 513454240,
+ "step": 11735,
+ "train_runtime": 90806.8266,
+ "train_tokens_per_second": 5654.357
+ },
+ {
+ "epoch": 0.7904659305144088,
+ "grad_norm": 0.6839883499284557,
+ "learning_rate": 6.616624965266392e-06,
+ "loss": 0.3722025632858276,
+ "num_input_tokens_seen": 513674120,
+ "step": 11740,
+ "train_runtime": 90845.0403,
+ "train_tokens_per_second": 5654.399
+ },
+ {
+ "epoch": 0.790802585510369,
+ "grad_norm": 0.7422766392719585,
+ "learning_rate": 6.614122676168486e-06,
+ "loss": 0.39596946239471437,
+ "num_input_tokens_seen": 513889520,
+ "step": 11745,
+ "train_runtime": 90889.19,
+ "train_tokens_per_second": 5654.022
+ },
+ {
+ "epoch": 0.7911392405063291,
+ "grad_norm": 0.8574462720949246,
+ "learning_rate": 6.611619935686321e-06,
+ "loss": 0.34925665855407717,
+ "num_input_tokens_seen": 514094568,
+ "step": 11750,
+ "train_runtime": 90923.7862,
+ "train_tokens_per_second": 5654.126
+ },
+ {
+ "epoch": 0.7914758955022893,
+ "grad_norm": 0.7396446955629874,
+ "learning_rate": 6.609116744519779e-06,
+ "loss": 0.3882757663726807,
+ "num_input_tokens_seen": 514319376,
+ "step": 11755,
+ "train_runtime": 90961.7449,
+ "train_tokens_per_second": 5654.238
+ },
+ {
+ "epoch": 0.7918125504982494,
+ "grad_norm": 0.8047711473168563,
+ "learning_rate": 6.60661310336887e-06,
+ "loss": 0.3384627103805542,
+ "num_input_tokens_seen": 514527584,
+ "step": 11760,
+ "train_runtime": 91008.8104,
+ "train_tokens_per_second": 5653.602
+ },
+ {
+ "epoch": 0.7921492054942095,
+ "grad_norm": 0.7981581690317775,
+ "learning_rate": 6.604109012933732e-06,
+ "loss": 0.37465848922729494,
+ "num_input_tokens_seen": 514729464,
+ "step": 11765,
+ "train_runtime": 91040.9107,
+ "train_tokens_per_second": 5653.826
+ },
+ {
+ "epoch": 0.7924858604901697,
+ "grad_norm": 0.7333032087882904,
+ "learning_rate": 6.601604473914622e-06,
+ "loss": 0.39129223823547366,
+ "num_input_tokens_seen": 514948712,
+ "step": 11770,
+ "train_runtime": 91080.7609,
+ "train_tokens_per_second": 5653.759
+ },
+ {
+ "epoch": 0.7928225154861298,
+ "grad_norm": 0.8203529810583842,
+ "learning_rate": 6.599099487011928e-06,
+ "loss": 0.3703568458557129,
+ "num_input_tokens_seen": 515174152,
+ "step": 11775,
+ "train_runtime": 91123.7113,
+ "train_tokens_per_second": 5653.569
+ },
+ {
+ "epoch": 0.79315917048209,
+ "grad_norm": 0.7692920757506985,
+ "learning_rate": 6.596594052926163e-06,
+ "loss": 0.37824716567993166,
+ "num_input_tokens_seen": 515398616,
+ "step": 11780,
+ "train_runtime": 91161.9049,
+ "train_tokens_per_second": 5653.662
+ },
+ {
+ "epoch": 0.7934958254780501,
+ "grad_norm": 0.6620821907912957,
+ "learning_rate": 6.5940881723579595e-06,
+ "loss": 0.36647372245788573,
+ "num_input_tokens_seen": 515618208,
+ "step": 11785,
+ "train_runtime": 91199.6136,
+ "train_tokens_per_second": 5653.732
+ },
+ {
+ "epoch": 0.7938324804740102,
+ "grad_norm": 0.7308558763801775,
+ "learning_rate": 6.591581846008085e-06,
+ "loss": 0.34111249446868896,
+ "num_input_tokens_seen": 515838000,
+ "step": 11790,
+ "train_runtime": 91242.8853,
+ "train_tokens_per_second": 5653.46
+ },
+ {
+ "epoch": 0.7941691354699704,
+ "grad_norm": 0.7510405382733509,
+ "learning_rate": 6.589075074577421e-06,
+ "loss": 0.3614863634109497,
+ "num_input_tokens_seen": 516058872,
+ "step": 11795,
+ "train_runtime": 91288.1701,
+ "train_tokens_per_second": 5653.075
+ },
+ {
+ "epoch": 0.7945057904659305,
+ "grad_norm": 0.682167345785408,
+ "learning_rate": 6.5865678587669805e-06,
+ "loss": 0.35341219902038573,
+ "num_input_tokens_seen": 516288480,
+ "step": 11800,
+ "train_runtime": 91324.6105,
+ "train_tokens_per_second": 5653.334
+ },
+ {
+ "epoch": 0.7948424454618906,
+ "grad_norm": 0.8727392399581251,
+ "learning_rate": 6.584060199277897e-06,
+ "loss": 0.3922135829925537,
+ "num_input_tokens_seen": 516500456,
+ "step": 11805,
+ "train_runtime": 91356.4687,
+ "train_tokens_per_second": 5653.682
+ },
+ {
+ "epoch": 0.7951791004578508,
+ "grad_norm": 0.7263526124399556,
+ "learning_rate": 6.581552096811428e-06,
+ "loss": 0.3736285209655762,
+ "num_input_tokens_seen": 516717184,
+ "step": 11810,
+ "train_runtime": 91391.2945,
+ "train_tokens_per_second": 5653.899
+ },
+ {
+ "epoch": 0.7955157554538109,
+ "grad_norm": 0.7483107004591556,
+ "learning_rate": 6.57904355206896e-06,
+ "loss": 0.3753064632415771,
+ "num_input_tokens_seen": 516937256,
+ "step": 11815,
+ "train_runtime": 91433.3951,
+ "train_tokens_per_second": 5653.703
+ },
+ {
+ "epoch": 0.7958524104497711,
+ "grad_norm": 0.741351883349472,
+ "learning_rate": 6.576534565751997e-06,
+ "loss": 0.35328757762908936,
+ "num_input_tokens_seen": 517155784,
+ "step": 11820,
+ "train_runtime": 91472.4934,
+ "train_tokens_per_second": 5653.675
+ },
+ {
+ "epoch": 0.7961890654457312,
+ "grad_norm": 0.7392430177231608,
+ "learning_rate": 6.574025138562172e-06,
+ "loss": 0.3843714475631714,
+ "num_input_tokens_seen": 517361632,
+ "step": 11825,
+ "train_runtime": 91507.0999,
+ "train_tokens_per_second": 5653.787
+ },
+ {
+ "epoch": 0.7965257204416913,
+ "grad_norm": 0.6186336035373811,
+ "learning_rate": 6.571515271201234e-06,
+ "loss": 0.3463775634765625,
+ "num_input_tokens_seen": 517585680,
+ "step": 11830,
+ "train_runtime": 91547.5139,
+ "train_tokens_per_second": 5653.738
+ },
+ {
+ "epoch": 0.7968623754376515,
+ "grad_norm": 0.6495706347541744,
+ "learning_rate": 6.569004964371062e-06,
+ "loss": 0.3861451864242554,
+ "num_input_tokens_seen": 517810808,
+ "step": 11835,
+ "train_runtime": 91580.6678,
+ "train_tokens_per_second": 5654.15
+ },
+ {
+ "epoch": 0.7971990304336116,
+ "grad_norm": 0.7143293886304013,
+ "learning_rate": 6.566494218773656e-06,
+ "loss": 0.3636291265487671,
+ "num_input_tokens_seen": 518036944,
+ "step": 11840,
+ "train_runtime": 91623.7035,
+ "train_tokens_per_second": 5653.962
+ },
+ {
+ "epoch": 0.7975356854295718,
+ "grad_norm": 0.7702610503844717,
+ "learning_rate": 6.563983035111136e-06,
+ "loss": 0.376003098487854,
+ "num_input_tokens_seen": 518266216,
+ "step": 11845,
+ "train_runtime": 91660.207,
+ "train_tokens_per_second": 5654.212
+ },
+ {
+ "epoch": 0.7978723404255319,
+ "grad_norm": 0.7600176188341217,
+ "learning_rate": 6.561471414085748e-06,
+ "loss": 0.36989893913269045,
+ "num_input_tokens_seen": 518477336,
+ "step": 11850,
+ "train_runtime": 91696.4011,
+ "train_tokens_per_second": 5654.282
+ },
+ {
+ "epoch": 0.798208995421492,
+ "grad_norm": 0.7119305326521156,
+ "learning_rate": 6.558959356399859e-06,
+ "loss": 0.3683455944061279,
+ "num_input_tokens_seen": 518691120,
+ "step": 11855,
+ "train_runtime": 91735.153,
+ "train_tokens_per_second": 5654.224
+ },
+ {
+ "epoch": 0.7985456504174522,
+ "grad_norm": 0.6724381741463443,
+ "learning_rate": 6.5564468627559544e-06,
+ "loss": 0.3579908609390259,
+ "num_input_tokens_seen": 518897696,
+ "step": 11860,
+ "train_runtime": 91775.8953,
+ "train_tokens_per_second": 5653.965
+ },
+ {
+ "epoch": 0.7988823054134123,
+ "grad_norm": 0.7387321689727897,
+ "learning_rate": 6.5539339338566495e-06,
+ "loss": 0.38613746166229246,
+ "num_input_tokens_seen": 519112712,
+ "step": 11865,
+ "train_runtime": 91816.066,
+ "train_tokens_per_second": 5653.833
+ },
+ {
+ "epoch": 0.7992189604093725,
+ "grad_norm": 0.8155036594396956,
+ "learning_rate": 6.5514205704046754e-06,
+ "loss": 0.34462246894836424,
+ "num_input_tokens_seen": 519326360,
+ "step": 11870,
+ "train_runtime": 91857.9272,
+ "train_tokens_per_second": 5653.582
+ },
+ {
+ "epoch": 0.7995556154053326,
+ "grad_norm": 0.7554873227873684,
+ "learning_rate": 6.548906773102886e-06,
+ "loss": 0.34732234477996826,
+ "num_input_tokens_seen": 519538040,
+ "step": 11875,
+ "train_runtime": 91892.7088,
+ "train_tokens_per_second": 5653.746
+ },
+ {
+ "epoch": 0.7998922704012927,
+ "grad_norm": 0.6865110717768457,
+ "learning_rate": 6.546392542654257e-06,
+ "loss": 0.3555797576904297,
+ "num_input_tokens_seen": 519754440,
+ "step": 11880,
+ "train_runtime": 91931.8443,
+ "train_tokens_per_second": 5653.693
+ },
+ {
+ "epoch": 0.8002289253972529,
+ "grad_norm": 0.7775168324740175,
+ "learning_rate": 6.543877879761882e-06,
+ "loss": 0.39305408000946046,
+ "num_input_tokens_seen": 519971256,
+ "step": 11885,
+ "train_runtime": 91972.3549,
+ "train_tokens_per_second": 5653.56
+ },
+ {
+ "epoch": 0.800565580393213,
+ "grad_norm": 0.6425158999557459,
+ "learning_rate": 6.5413627851289825e-06,
+ "loss": 0.3501169681549072,
+ "num_input_tokens_seen": 520197464,
+ "step": 11890,
+ "train_runtime": 92010.0574,
+ "train_tokens_per_second": 5653.702
+ },
+ {
+ "epoch": 0.8009022353891732,
+ "grad_norm": 0.7227536117889709,
+ "learning_rate": 6.538847259458896e-06,
+ "loss": 0.3551483631134033,
+ "num_input_tokens_seen": 520422616,
+ "step": 11895,
+ "train_runtime": 92042.9843,
+ "train_tokens_per_second": 5654.126
+ },
+ {
+ "epoch": 0.8012388903851333,
+ "grad_norm": 0.7042669271875746,
+ "learning_rate": 6.5363313034550805e-06,
+ "loss": 0.34134459495544434,
+ "num_input_tokens_seen": 520633392,
+ "step": 11900,
+ "train_runtime": 92087.6293,
+ "train_tokens_per_second": 5653.674
+ },
+ {
+ "epoch": 0.8015755453810934,
+ "grad_norm": 0.739311915732445,
+ "learning_rate": 6.533814917821114e-06,
+ "loss": 0.3754891395568848,
+ "num_input_tokens_seen": 520855784,
+ "step": 11905,
+ "train_runtime": 92127.8168,
+ "train_tokens_per_second": 5653.621
+ },
+ {
+ "epoch": 0.8019122003770536,
+ "grad_norm": 0.6302684231473084,
+ "learning_rate": 6.531298103260696e-06,
+ "loss": 0.3585964202880859,
+ "num_input_tokens_seen": 521066736,
+ "step": 11910,
+ "train_runtime": 92163.5443,
+ "train_tokens_per_second": 5653.719
+ },
+ {
+ "epoch": 0.8022488553730137,
+ "grad_norm": 0.7417671721607515,
+ "learning_rate": 6.528780860477648e-06,
+ "loss": 0.3797733783721924,
+ "num_input_tokens_seen": 521300416,
+ "step": 11915,
+ "train_runtime": 92200.3318,
+ "train_tokens_per_second": 5653.997
+ },
+ {
+ "epoch": 0.8025855103689739,
+ "grad_norm": 0.6890068438400968,
+ "learning_rate": 6.526263190175907e-06,
+ "loss": 0.3375946044921875,
+ "num_input_tokens_seen": 521532280,
+ "step": 11920,
+ "train_runtime": 92234.4047,
+ "train_tokens_per_second": 5654.422
+ },
+ {
+ "epoch": 0.802922165364934,
+ "grad_norm": 0.6823445253166819,
+ "learning_rate": 6.523745093059531e-06,
+ "loss": 0.3637096643447876,
+ "num_input_tokens_seen": 521751944,
+ "step": 11925,
+ "train_runtime": 92276.588,
+ "train_tokens_per_second": 5654.218
+ },
+ {
+ "epoch": 0.8032588203608941,
+ "grad_norm": 0.8198308621507191,
+ "learning_rate": 6.521226569832699e-06,
+ "loss": 0.3677360057830811,
+ "num_input_tokens_seen": 521974016,
+ "step": 11930,
+ "train_runtime": 92311.2607,
+ "train_tokens_per_second": 5654.5
+ },
+ {
+ "epoch": 0.8035954753568543,
+ "grad_norm": 0.7478623966170844,
+ "learning_rate": 6.518707621199707e-06,
+ "loss": 0.355226993560791,
+ "num_input_tokens_seen": 522190040,
+ "step": 11935,
+ "train_runtime": 92353.1434,
+ "train_tokens_per_second": 5654.275
+ },
+ {
+ "epoch": 0.8039321303528144,
+ "grad_norm": 0.7995961908090801,
+ "learning_rate": 6.516188247864972e-06,
+ "loss": 0.3752752304077148,
+ "num_input_tokens_seen": 522418072,
+ "step": 11940,
+ "train_runtime": 92392.9504,
+ "train_tokens_per_second": 5654.307
+ },
+ {
+ "epoch": 0.8042687853487746,
+ "grad_norm": 0.8084497458000778,
+ "learning_rate": 6.513668450533028e-06,
+ "loss": 0.3633857250213623,
+ "num_input_tokens_seen": 522645296,
+ "step": 11945,
+ "train_runtime": 92428.2209,
+ "train_tokens_per_second": 5654.607
+ },
+ {
+ "epoch": 0.8046054403447347,
+ "grad_norm": 0.7480700245413793,
+ "learning_rate": 6.511148229908531e-06,
+ "loss": 0.39172661304473877,
+ "num_input_tokens_seen": 522868752,
+ "step": 11950,
+ "train_runtime": 92461.7106,
+ "train_tokens_per_second": 5654.976
+ },
+ {
+ "epoch": 0.8049420953406948,
+ "grad_norm": 0.8708977640452219,
+ "learning_rate": 6.508627586696248e-06,
+ "loss": 0.3619298696517944,
+ "num_input_tokens_seen": 523082504,
+ "step": 11955,
+ "train_runtime": 92499.8318,
+ "train_tokens_per_second": 5654.956
+ },
+ {
+ "epoch": 0.805278750336655,
+ "grad_norm": 0.7261115677502934,
+ "learning_rate": 6.5061065216010695e-06,
+ "loss": 0.3667879581451416,
+ "num_input_tokens_seen": 523299872,
+ "step": 11960,
+ "train_runtime": 92542.5023,
+ "train_tokens_per_second": 5654.698
+ },
+ {
+ "epoch": 0.8056154053326151,
+ "grad_norm": 0.7422593678216647,
+ "learning_rate": 6.5035850353280075e-06,
+ "loss": 0.3360764503479004,
+ "num_input_tokens_seen": 523516992,
+ "step": 11965,
+ "train_runtime": 92581.276,
+ "train_tokens_per_second": 5654.675
+ },
+ {
+ "epoch": 0.8059520603285752,
+ "grad_norm": 0.7763633318074254,
+ "learning_rate": 6.501063128582183e-06,
+ "loss": 0.35983996391296386,
+ "num_input_tokens_seen": 523724536,
+ "step": 11970,
+ "train_runtime": 92615.6401,
+ "train_tokens_per_second": 5654.817
+ },
+ {
+ "epoch": 0.8062887153245354,
+ "grad_norm": 0.8599327424362851,
+ "learning_rate": 6.498540802068843e-06,
+ "loss": 0.36988580226898193,
+ "num_input_tokens_seen": 523941648,
+ "step": 11975,
+ "train_runtime": 92650.751,
+ "train_tokens_per_second": 5655.018
+ },
+ {
+ "epoch": 0.8066253703204955,
+ "grad_norm": 0.7313407988142705,
+ "learning_rate": 6.4960180564933454e-06,
+ "loss": 0.3595852851867676,
+ "num_input_tokens_seen": 524162560,
+ "step": 11980,
+ "train_runtime": 92691.1408,
+ "train_tokens_per_second": 5654.937
+ },
+ {
+ "epoch": 0.8069620253164557,
+ "grad_norm": 0.7521095206194225,
+ "learning_rate": 6.493494892561167e-06,
+ "loss": 0.3594093084335327,
+ "num_input_tokens_seen": 524374312,
+ "step": 11985,
+ "train_runtime": 92728.3173,
+ "train_tokens_per_second": 5654.953
+ },
+ {
+ "epoch": 0.8072986803124158,
+ "grad_norm": 0.8430159653893533,
+ "learning_rate": 6.490971310977907e-06,
+ "loss": 0.3831813812255859,
+ "num_input_tokens_seen": 524596088,
+ "step": 11990,
+ "train_runtime": 92765.8417,
+ "train_tokens_per_second": 5655.057
+ },
+ {
+ "epoch": 0.807635335308376,
+ "grad_norm": 0.7203196408716581,
+ "learning_rate": 6.488447312449273e-06,
+ "loss": 0.384428596496582,
+ "num_input_tokens_seen": 524821680,
+ "step": 11995,
+ "train_runtime": 92796.5099,
+ "train_tokens_per_second": 5655.619
+ },
+ {
+ "epoch": 0.8079719903043361,
+ "grad_norm": 0.7501646695656614,
+ "learning_rate": 6.485922897681095e-06,
+ "loss": 0.3725404739379883,
+ "num_input_tokens_seen": 525045696,
+ "step": 12000,
+ "train_runtime": 92832.3227,
+ "train_tokens_per_second": 5655.85
+ },
+ {
+ "epoch": 0.8083086453002962,
+ "grad_norm": 0.6490994037223768,
+ "learning_rate": 6.483398067379316e-06,
+ "loss": 0.4056462287902832,
+ "num_input_tokens_seen": 525279952,
+ "step": 12005,
+ "train_runtime": 92862.4479,
+ "train_tokens_per_second": 5656.538
+ },
+ {
+ "epoch": 0.8086453002962564,
+ "grad_norm": 0.719666275525647,
+ "learning_rate": 6.480872822249998e-06,
+ "loss": 0.3556082248687744,
+ "num_input_tokens_seen": 525506864,
+ "step": 12010,
+ "train_runtime": 92904.5673,
+ "train_tokens_per_second": 5656.416
+ },
+ {
+ "epoch": 0.8089819552922165,
+ "grad_norm": 0.764344635124093,
+ "learning_rate": 6.478347162999318e-06,
+ "loss": 0.3839968919754028,
+ "num_input_tokens_seen": 525719016,
+ "step": 12015,
+ "train_runtime": 92947.2954,
+ "train_tokens_per_second": 5656.098
+ },
+ {
+ "epoch": 0.8093186102881766,
+ "grad_norm": 0.6701860780093635,
+ "learning_rate": 6.475821090333568e-06,
+ "loss": 0.3664535999298096,
+ "num_input_tokens_seen": 525939808,
+ "step": 12020,
+ "train_runtime": 92994.8165,
+ "train_tokens_per_second": 5655.582
+ },
+ {
+ "epoch": 0.8096552652841368,
+ "grad_norm": 0.7604615691324128,
+ "learning_rate": 6.473294604959156e-06,
+ "loss": 0.38412063121795653,
+ "num_input_tokens_seen": 526150200,
+ "step": 12025,
+ "train_runtime": 93031.4012,
+ "train_tokens_per_second": 5655.619
+ },
+ {
+ "epoch": 0.8099919202800969,
+ "grad_norm": 0.7314264087625333,
+ "learning_rate": 6.470767707582605e-06,
+ "loss": 0.3502906322479248,
+ "num_input_tokens_seen": 526371776,
+ "step": 12030,
+ "train_runtime": 93073.7526,
+ "train_tokens_per_second": 5655.427
+ },
+ {
+ "epoch": 0.8103285752760571,
+ "grad_norm": 0.8031066646434316,
+ "learning_rate": 6.468240398910555e-06,
+ "loss": 0.36304702758789065,
+ "num_input_tokens_seen": 526566912,
+ "step": 12035,
+ "train_runtime": 93113.568,
+ "train_tokens_per_second": 5655.104
+ },
+ {
+ "epoch": 0.8106652302720172,
+ "grad_norm": 0.7056043206827206,
+ "learning_rate": 6.4657126796497595e-06,
+ "loss": 0.3545745849609375,
+ "num_input_tokens_seen": 526798856,
+ "step": 12040,
+ "train_runtime": 93156.2331,
+ "train_tokens_per_second": 5655.004
+ },
+ {
+ "epoch": 0.8110018852679773,
+ "grad_norm": 0.8107764954705456,
+ "learning_rate": 6.463184550507087e-06,
+ "loss": 0.37404961585998536,
+ "num_input_tokens_seen": 527012296,
+ "step": 12045,
+ "train_runtime": 93201.4479,
+ "train_tokens_per_second": 5654.551
+ },
+ {
+ "epoch": 0.8113385402639375,
+ "grad_norm": 0.6987203165103377,
+ "learning_rate": 6.46065601218952e-06,
+ "loss": 0.34106624126434326,
+ "num_input_tokens_seen": 527230672,
+ "step": 12050,
+ "train_runtime": 93238.6997,
+ "train_tokens_per_second": 5654.633
+ },
+ {
+ "epoch": 0.8116751952598976,
+ "grad_norm": 0.6797008656627719,
+ "learning_rate": 6.458127065404156e-06,
+ "loss": 0.3812570095062256,
+ "num_input_tokens_seen": 527453272,
+ "step": 12055,
+ "train_runtime": 93278.2031,
+ "train_tokens_per_second": 5654.625
+ },
+ {
+ "epoch": 0.8120118502558578,
+ "grad_norm": 0.6832021023137033,
+ "learning_rate": 6.45559771085821e-06,
+ "loss": 0.35934345722198485,
+ "num_input_tokens_seen": 527679752,
+ "step": 12060,
+ "train_runtime": 93316.0381,
+ "train_tokens_per_second": 5654.759
+ },
+ {
+ "epoch": 0.8123485052518179,
+ "grad_norm": 0.7280405860607188,
+ "learning_rate": 6.453067949259005e-06,
+ "loss": 0.35833373069763186,
+ "num_input_tokens_seen": 527905280,
+ "step": 12065,
+ "train_runtime": 93350.8689,
+ "train_tokens_per_second": 5655.066
+ },
+ {
+ "epoch": 0.812685160247778,
+ "grad_norm": 0.7042450928577643,
+ "learning_rate": 6.450537781313982e-06,
+ "loss": 0.33934841156005857,
+ "num_input_tokens_seen": 528118504,
+ "step": 12070,
+ "train_runtime": 93398.5502,
+ "train_tokens_per_second": 5654.461
+ },
+ {
+ "epoch": 0.8130218152437382,
+ "grad_norm": 0.8291985738023927,
+ "learning_rate": 6.448007207730693e-06,
+ "loss": 0.3756300926208496,
+ "num_input_tokens_seen": 528340504,
+ "step": 12075,
+ "train_runtime": 93436.5515,
+ "train_tokens_per_second": 5654.538
+ },
+ {
+ "epoch": 0.8133584702396983,
+ "grad_norm": 0.7204057314942438,
+ "learning_rate": 6.445476229216805e-06,
+ "loss": 0.35174586772918703,
+ "num_input_tokens_seen": 528575776,
+ "step": 12080,
+ "train_runtime": 93474.5152,
+ "train_tokens_per_second": 5654.758
+ },
+ {
+ "epoch": 0.8136951252356585,
+ "grad_norm": 0.7511444963498415,
+ "learning_rate": 6.442944846480099e-06,
+ "loss": 0.39254226684570315,
+ "num_input_tokens_seen": 528786904,
+ "step": 12085,
+ "train_runtime": 93515.4145,
+ "train_tokens_per_second": 5654.543
+ },
+ {
+ "epoch": 0.8140317802316186,
+ "grad_norm": 0.7816657648101096,
+ "learning_rate": 6.44041306022847e-06,
+ "loss": 0.39618897438049316,
+ "num_input_tokens_seen": 529015944,
+ "step": 12090,
+ "train_runtime": 93554.4561,
+ "train_tokens_per_second": 5654.631
+ },
+ {
+ "epoch": 0.8143684352275787,
+ "grad_norm": 0.8141588683065977,
+ "learning_rate": 6.43788087116992e-06,
+ "loss": 0.39859035015106203,
+ "num_input_tokens_seen": 529246224,
+ "step": 12095,
+ "train_runtime": 93595.1858,
+ "train_tokens_per_second": 5654.631
+ },
+ {
+ "epoch": 0.8147050902235389,
+ "grad_norm": 0.7932908932160116,
+ "learning_rate": 6.43534828001257e-06,
+ "loss": 0.3658564567565918,
+ "num_input_tokens_seen": 529476248,
+ "step": 12100,
+ "train_runtime": 93626.3021,
+ "train_tokens_per_second": 5655.208
+ },
+ {
+ "epoch": 0.815041745219499,
+ "grad_norm": 0.79019755327484,
+ "learning_rate": 6.43281528746465e-06,
+ "loss": 0.372475266456604,
+ "num_input_tokens_seen": 529683080,
+ "step": 12105,
+ "train_runtime": 93660.6895,
+ "train_tokens_per_second": 5655.34
+ },
+ {
+ "epoch": 0.8153784002154592,
+ "grad_norm": 0.6339454129210152,
+ "learning_rate": 6.430281894234504e-06,
+ "loss": 0.36925411224365234,
+ "num_input_tokens_seen": 529906632,
+ "step": 12110,
+ "train_runtime": 93709.4206,
+ "train_tokens_per_second": 5654.785
+ },
+ {
+ "epoch": 0.8157150552114193,
+ "grad_norm": 0.8085146313091093,
+ "learning_rate": 6.427748101030587e-06,
+ "loss": 0.3794065237045288,
+ "num_input_tokens_seen": 530122608,
+ "step": 12115,
+ "train_runtime": 93746.5407,
+ "train_tokens_per_second": 5654.85
+ },
+ {
+ "epoch": 0.8160517102073794,
+ "grad_norm": 0.7189030374735652,
+ "learning_rate": 6.425213908561468e-06,
+ "loss": 0.3644280433654785,
+ "num_input_tokens_seen": 530344760,
+ "step": 12120,
+ "train_runtime": 93782.6096,
+ "train_tokens_per_second": 5655.044
+ },
+ {
+ "epoch": 0.8163883652033396,
+ "grad_norm": 0.7379513095206941,
+ "learning_rate": 6.422679317535823e-06,
+ "loss": 0.3801795721054077,
+ "num_input_tokens_seen": 530554320,
+ "step": 12125,
+ "train_runtime": 93818.6237,
+ "train_tokens_per_second": 5655.107
+ },
+ {
+ "epoch": 0.8167250201992997,
+ "grad_norm": 0.7361384056056562,
+ "learning_rate": 6.420144328662441e-06,
+ "loss": 0.4110711097717285,
+ "num_input_tokens_seen": 530784216,
+ "step": 12130,
+ "train_runtime": 93858.9763,
+ "train_tokens_per_second": 5655.125
+ },
+ {
+ "epoch": 0.8170616751952599,
+ "grad_norm": 0.6984222353139848,
+ "learning_rate": 6.417608942650229e-06,
+ "loss": 0.36923315525054934,
+ "num_input_tokens_seen": 530985400,
+ "step": 12135,
+ "train_runtime": 93900.3001,
+ "train_tokens_per_second": 5654.779
+ },
+ {
+ "epoch": 0.81739833019122,
+ "grad_norm": 0.7696394382831785,
+ "learning_rate": 6.415073160208196e-06,
+ "loss": 0.368359637260437,
+ "num_input_tokens_seen": 531219992,
+ "step": 12140,
+ "train_runtime": 93938.5272,
+ "train_tokens_per_second": 5654.975
+ },
+ {
+ "epoch": 0.8177349851871801,
+ "grad_norm": 0.8778188370382649,
+ "learning_rate": 6.412536982045465e-06,
+ "loss": 0.38031327724456787,
+ "num_input_tokens_seen": 531416856,
+ "step": 12145,
+ "train_runtime": 93975.0195,
+ "train_tokens_per_second": 5654.874
+ },
+ {
+ "epoch": 0.8180716401831403,
+ "grad_norm": 0.7668958438378366,
+ "learning_rate": 6.410000408871273e-06,
+ "loss": 0.3821054458618164,
+ "num_input_tokens_seen": 531631960,
+ "step": 12150,
+ "train_runtime": 94010.159,
+ "train_tokens_per_second": 5655.048
+ },
+ {
+ "epoch": 0.8184082951791004,
+ "grad_norm": 0.7888864971284683,
+ "learning_rate": 6.407463441394961e-06,
+ "loss": 0.37601168155670167,
+ "num_input_tokens_seen": 531859288,
+ "step": 12155,
+ "train_runtime": 94046.1948,
+ "train_tokens_per_second": 5655.298
+ },
+ {
+ "epoch": 0.8187449501750605,
+ "grad_norm": 0.8191097049266588,
+ "learning_rate": 6.4049260803259874e-06,
+ "loss": 0.38929071426391604,
+ "num_input_tokens_seen": 532082152,
+ "step": 12160,
+ "train_runtime": 94088.2586,
+ "train_tokens_per_second": 5655.139
+ },
+ {
+ "epoch": 0.8190816051710207,
+ "grad_norm": 0.8221152177897472,
+ "learning_rate": 6.402388326373914e-06,
+ "loss": 0.3606018304824829,
+ "num_input_tokens_seen": 532304432,
+ "step": 12165,
+ "train_runtime": 94128.8133,
+ "train_tokens_per_second": 5655.064
+ },
+ {
+ "epoch": 0.8194182601669808,
+ "grad_norm": 0.6644138882185266,
+ "learning_rate": 6.399850180248418e-06,
+ "loss": 0.4054881572723389,
+ "num_input_tokens_seen": 532538000,
+ "step": 12170,
+ "train_runtime": 94167.2148,
+ "train_tokens_per_second": 5655.238
+ },
+ {
+ "epoch": 0.819754915162941,
+ "grad_norm": 0.7674282579500541,
+ "learning_rate": 6.397311642659283e-06,
+ "loss": 0.3806436538696289,
+ "num_input_tokens_seen": 532749856,
+ "step": 12175,
+ "train_runtime": 94203.2208,
+ "train_tokens_per_second": 5655.325
+ },
+ {
+ "epoch": 0.8200915701589011,
+ "grad_norm": 0.7532047074504019,
+ "learning_rate": 6.394772714316402e-06,
+ "loss": 0.3875831842422485,
+ "num_input_tokens_seen": 532981096,
+ "step": 12180,
+ "train_runtime": 94238.5397,
+ "train_tokens_per_second": 5655.66
+ },
+ {
+ "epoch": 0.8204282251548612,
+ "grad_norm": 0.691733853853237,
+ "learning_rate": 6.392233395929779e-06,
+ "loss": 0.34835638999938967,
+ "num_input_tokens_seen": 533193808,
+ "step": 12185,
+ "train_runtime": 94267.7706,
+ "train_tokens_per_second": 5656.162
+ },
+ {
+ "epoch": 0.8207648801508214,
+ "grad_norm": 0.6752837485492823,
+ "learning_rate": 6.389693688209526e-06,
+ "loss": 0.362264609336853,
+ "num_input_tokens_seen": 533413896,
+ "step": 12190,
+ "train_runtime": 94307.5026,
+ "train_tokens_per_second": 5656.113
+ },
+ {
+ "epoch": 0.8211015351467815,
+ "grad_norm": 0.6640861040569711,
+ "learning_rate": 6.387153591865867e-06,
+ "loss": 0.3703956604003906,
+ "num_input_tokens_seen": 533642648,
+ "step": 12195,
+ "train_runtime": 94351.3757,
+ "train_tokens_per_second": 5655.907
+ },
+ {
+ "epoch": 0.8214381901427417,
+ "grad_norm": 0.8639133957294087,
+ "learning_rate": 6.384613107609127e-06,
+ "loss": 0.38781538009643557,
+ "num_input_tokens_seen": 533855672,
+ "step": 12200,
+ "train_runtime": 94388.037,
+ "train_tokens_per_second": 5655.968
+ },
+ {
+ "epoch": 0.8217748451387018,
+ "grad_norm": 0.6461164084347707,
+ "learning_rate": 6.382072236149747e-06,
+ "loss": 0.3636336326599121,
+ "num_input_tokens_seen": 534086688,
+ "step": 12205,
+ "train_runtime": 94418.101,
+ "train_tokens_per_second": 5656.613
+ },
+ {
+ "epoch": 0.8221115001346619,
+ "grad_norm": 1.259849346653607,
+ "learning_rate": 6.379530978198273e-06,
+ "loss": 0.33962430953979494,
+ "num_input_tokens_seen": 534295064,
+ "step": 12210,
+ "train_runtime": 94452.6928,
+ "train_tokens_per_second": 5656.748
+ },
+ {
+ "epoch": 0.8224481551306222,
+ "grad_norm": 0.6840892751944132,
+ "learning_rate": 6.376989334465361e-06,
+ "loss": 0.34957318305969237,
+ "num_input_tokens_seen": 534518376,
+ "step": 12215,
+ "train_runtime": 94487.4762,
+ "train_tokens_per_second": 5657.029
+ },
+ {
+ "epoch": 0.8227848101265823,
+ "grad_norm": 0.7080671395957931,
+ "learning_rate": 6.374447305661772e-06,
+ "loss": 0.359330415725708,
+ "num_input_tokens_seen": 534753352,
+ "step": 12220,
+ "train_runtime": 94526.7452,
+ "train_tokens_per_second": 5657.165
+ },
+ {
+ "epoch": 0.8231214651225425,
+ "grad_norm": 0.7000813266501178,
+ "learning_rate": 6.3719048924983776e-06,
+ "loss": 0.3536032199859619,
+ "num_input_tokens_seen": 534959200,
+ "step": 12225,
+ "train_runtime": 94565.9434,
+ "train_tokens_per_second": 5656.996
+ },
+ {
+ "epoch": 0.8234581201185026,
+ "grad_norm": 0.7684419131244044,
+ "learning_rate": 6.369362095686152e-06,
+ "loss": 0.3803597688674927,
+ "num_input_tokens_seen": 535175808,
+ "step": 12230,
+ "train_runtime": 94601.305,
+ "train_tokens_per_second": 5657.172
+ },
+ {
+ "epoch": 0.8237947751144628,
+ "grad_norm": 0.6924107525351223,
+ "learning_rate": 6.366818915936185e-06,
+ "loss": 0.34799120426177976,
+ "num_input_tokens_seen": 535394008,
+ "step": 12235,
+ "train_runtime": 94637.8962,
+ "train_tokens_per_second": 5657.29
+ },
+ {
+ "epoch": 0.8241314301104229,
+ "grad_norm": 0.7585078799576096,
+ "learning_rate": 6.364275353959667e-06,
+ "loss": 0.3779747486114502,
+ "num_input_tokens_seen": 535603560,
+ "step": 12240,
+ "train_runtime": 94679.4987,
+ "train_tokens_per_second": 5657.017
+ },
+ {
+ "epoch": 0.824468085106383,
+ "grad_norm": 0.7685591544772424,
+ "learning_rate": 6.3617314104678966e-06,
+ "loss": 0.36225066184997556,
+ "num_input_tokens_seen": 535838080,
+ "step": 12245,
+ "train_runtime": 94720.1071,
+ "train_tokens_per_second": 5657.068
+ },
+ {
+ "epoch": 0.8248047401023432,
+ "grad_norm": 0.76161666147799,
+ "learning_rate": 6.359187086172278e-06,
+ "loss": 0.3977229118347168,
+ "num_input_tokens_seen": 536041408,
+ "step": 12250,
+ "train_runtime": 94755.4024,
+ "train_tokens_per_second": 5657.107
+ },
+ {
+ "epoch": 0.8251413950983033,
+ "grad_norm": 0.6732965305210872,
+ "learning_rate": 6.356642381784326e-06,
+ "loss": 0.3828271389007568,
+ "num_input_tokens_seen": 536255712,
+ "step": 12255,
+ "train_runtime": 94789.5374,
+ "train_tokens_per_second": 5657.33
+ },
+ {
+ "epoch": 0.8254780500942634,
+ "grad_norm": 0.9216909752320883,
+ "learning_rate": 6.354097298015658e-06,
+ "loss": 0.3526939392089844,
+ "num_input_tokens_seen": 536450456,
+ "step": 12260,
+ "train_runtime": 94828.1952,
+ "train_tokens_per_second": 5657.078
+ },
+ {
+ "epoch": 0.8258147050902236,
+ "grad_norm": 0.7392662084519794,
+ "learning_rate": 6.3515518355779985e-06,
+ "loss": 0.3932209014892578,
+ "num_input_tokens_seen": 536655032,
+ "step": 12265,
+ "train_runtime": 94865.9619,
+ "train_tokens_per_second": 5656.982
+ },
+ {
+ "epoch": 0.8261513600861837,
+ "grad_norm": 0.630920453324516,
+ "learning_rate": 6.349005995183177e-06,
+ "loss": 0.3485422134399414,
+ "num_input_tokens_seen": 536883576,
+ "step": 12270,
+ "train_runtime": 94899.668,
+ "train_tokens_per_second": 5657.381
+ },
+ {
+ "epoch": 0.8264880150821439,
+ "grad_norm": 0.6584126043910694,
+ "learning_rate": 6.346459777543131e-06,
+ "loss": 0.3577850341796875,
+ "num_input_tokens_seen": 537102112,
+ "step": 12275,
+ "train_runtime": 94933.4199,
+ "train_tokens_per_second": 5657.672
+ },
+ {
+ "epoch": 0.826824670078104,
+ "grad_norm": 0.69670427508635,
+ "learning_rate": 6.343913183369902e-06,
+ "loss": 0.3586315155029297,
+ "num_input_tokens_seen": 537306360,
+ "step": 12280,
+ "train_runtime": 94970.6969,
+ "train_tokens_per_second": 5657.602
+ },
+ {
+ "epoch": 0.8271613250740641,
+ "grad_norm": 0.6795435309811332,
+ "learning_rate": 6.341366213375637e-06,
+ "loss": 0.3537776231765747,
+ "num_input_tokens_seen": 537534592,
+ "step": 12285,
+ "train_runtime": 95003.3121,
+ "train_tokens_per_second": 5658.062
+ },
+ {
+ "epoch": 0.8274979800700243,
+ "grad_norm": 0.7334092343426839,
+ "learning_rate": 6.338818868272587e-06,
+ "loss": 0.3713988780975342,
+ "num_input_tokens_seen": 537747240,
+ "step": 12290,
+ "train_runtime": 95041.1213,
+ "train_tokens_per_second": 5658.048
+ },
+ {
+ "epoch": 0.8278346350659844,
+ "grad_norm": 0.7029457386112722,
+ "learning_rate": 6.336271148773111e-06,
+ "loss": 0.3642918109893799,
+ "num_input_tokens_seen": 537982080,
+ "step": 12295,
+ "train_runtime": 95085.5253,
+ "train_tokens_per_second": 5657.876
+ },
+ {
+ "epoch": 0.8281712900619446,
+ "grad_norm": 0.664085080042037,
+ "learning_rate": 6.333723055589669e-06,
+ "loss": 0.4124462127685547,
+ "num_input_tokens_seen": 538202248,
+ "step": 12300,
+ "train_runtime": 95123.8409,
+ "train_tokens_per_second": 5657.911
+ },
+ {
+ "epoch": 0.8285079450579047,
+ "grad_norm": 0.7689236487654127,
+ "learning_rate": 6.331174589434826e-06,
+ "loss": 0.3763313293457031,
+ "num_input_tokens_seen": 538420056,
+ "step": 12305,
+ "train_runtime": 95170.7251,
+ "train_tokens_per_second": 5657.413
+ },
+ {
+ "epoch": 0.8288446000538648,
+ "grad_norm": 0.7341729953622874,
+ "learning_rate": 6.328625751021254e-06,
+ "loss": 0.35441150665283205,
+ "num_input_tokens_seen": 538627360,
+ "step": 12310,
+ "train_runtime": 95214.1667,
+ "train_tokens_per_second": 5657.009
+ },
+ {
+ "epoch": 0.829181255049825,
+ "grad_norm": 0.7056989368636907,
+ "learning_rate": 6.326076541061729e-06,
+ "loss": 0.334460711479187,
+ "num_input_tokens_seen": 538844328,
+ "step": 12315,
+ "train_runtime": 95255.2861,
+ "train_tokens_per_second": 5656.844
+ },
+ {
+ "epoch": 0.8295179100457851,
+ "grad_norm": 0.7430926488416755,
+ "learning_rate": 6.323526960269127e-06,
+ "loss": 0.3949697971343994,
+ "num_input_tokens_seen": 539054992,
+ "step": 12320,
+ "train_runtime": 95291.3412,
+ "train_tokens_per_second": 5656.915
+ },
+ {
+ "epoch": 0.8298545650417453,
+ "grad_norm": 0.7296769222033309,
+ "learning_rate": 6.3209770093564315e-06,
+ "loss": 0.3561330556869507,
+ "num_input_tokens_seen": 539280216,
+ "step": 12325,
+ "train_runtime": 95332.3002,
+ "train_tokens_per_second": 5656.847
+ },
+ {
+ "epoch": 0.8301912200377054,
+ "grad_norm": 0.8506709358501154,
+ "learning_rate": 6.318426689036727e-06,
+ "loss": 0.3534719705581665,
+ "num_input_tokens_seen": 539494040,
+ "step": 12330,
+ "train_runtime": 95376.0518,
+ "train_tokens_per_second": 5656.494
+ },
+ {
+ "epoch": 0.8305278750336655,
+ "grad_norm": 0.7341499431044767,
+ "learning_rate": 6.315876000023202e-06,
+ "loss": 0.37946810722351076,
+ "num_input_tokens_seen": 539706008,
+ "step": 12335,
+ "train_runtime": 95416.5574,
+ "train_tokens_per_second": 5656.314
+ },
+ {
+ "epoch": 0.8308645300296257,
+ "grad_norm": 3.744199811122827,
+ "learning_rate": 6.313324943029151e-06,
+ "loss": 0.3778734922409058,
+ "num_input_tokens_seen": 539917272,
+ "step": 12340,
+ "train_runtime": 95460.2877,
+ "train_tokens_per_second": 5655.936
+ },
+ {
+ "epoch": 0.8312011850255858,
+ "grad_norm": 0.742412638594704,
+ "learning_rate": 6.310773518767967e-06,
+ "loss": 0.35590362548828125,
+ "num_input_tokens_seen": 540130136,
+ "step": 12345,
+ "train_runtime": 95493.567,
+ "train_tokens_per_second": 5656.194
+ },
+ {
+ "epoch": 0.831537840021546,
+ "grad_norm": 0.702835896250019,
+ "learning_rate": 6.30822172795315e-06,
+ "loss": 0.36078743934631347,
+ "num_input_tokens_seen": 540336200,
+ "step": 12350,
+ "train_runtime": 95528.2473,
+ "train_tokens_per_second": 5656.298
+ },
+ {
+ "epoch": 0.8318744950175061,
+ "grad_norm": 0.7843017900068701,
+ "learning_rate": 6.3056695712982965e-06,
+ "loss": 0.4154792785644531,
+ "num_input_tokens_seen": 540561200,
+ "step": 12355,
+ "train_runtime": 95559.8481,
+ "train_tokens_per_second": 5656.782
+ },
+ {
+ "epoch": 0.8322111500134662,
+ "grad_norm": 0.6757106331074613,
+ "learning_rate": 6.303117049517111e-06,
+ "loss": 0.3399226427078247,
+ "num_input_tokens_seen": 540776592,
+ "step": 12360,
+ "train_runtime": 95600.3642,
+ "train_tokens_per_second": 5656.637
+ },
+ {
+ "epoch": 0.8325478050094264,
+ "grad_norm": 0.7135184795154971,
+ "learning_rate": 6.300564163323398e-06,
+ "loss": 0.35541400909423826,
+ "num_input_tokens_seen": 540989568,
+ "step": 12365,
+ "train_runtime": 95640.6371,
+ "train_tokens_per_second": 5656.482
+ },
+ {
+ "epoch": 0.8328844600053865,
+ "grad_norm": 0.7511263678823955,
+ "learning_rate": 6.298010913431065e-06,
+ "loss": 0.36992754936218264,
+ "num_input_tokens_seen": 541205456,
+ "step": 12370,
+ "train_runtime": 95679.064,
+ "train_tokens_per_second": 5656.467
+ },
+ {
+ "epoch": 0.8332211150013467,
+ "grad_norm": 0.735359416996797,
+ "learning_rate": 6.295457300554119e-06,
+ "loss": 0.3586021184921265,
+ "num_input_tokens_seen": 541434640,
+ "step": 12375,
+ "train_runtime": 95717.4462,
+ "train_tokens_per_second": 5656.593
+ },
+ {
+ "epoch": 0.8335577699973068,
+ "grad_norm": 0.8567636961499998,
+ "learning_rate": 6.292903325406668e-06,
+ "loss": 0.38977057933807374,
+ "num_input_tokens_seen": 541635304,
+ "step": 12380,
+ "train_runtime": 95751.5012,
+ "train_tokens_per_second": 5656.677
+ },
+ {
+ "epoch": 0.8338944249932669,
+ "grad_norm": 0.7502505399469903,
+ "learning_rate": 6.2903489887029255e-06,
+ "loss": 0.34544246196746825,
+ "num_input_tokens_seen": 541858656,
+ "step": 12385,
+ "train_runtime": 95786.9128,
+ "train_tokens_per_second": 5656.917
+ },
+ {
+ "epoch": 0.8342310799892271,
+ "grad_norm": 0.6987916307726929,
+ "learning_rate": 6.287794291157204e-06,
+ "loss": 0.36299178600311277,
+ "num_input_tokens_seen": 542076592,
+ "step": 12390,
+ "train_runtime": 95820.7633,
+ "train_tokens_per_second": 5657.193
+ },
+ {
+ "epoch": 0.8345677349851872,
+ "grad_norm": 0.8150952905885787,
+ "learning_rate": 6.285239233483915e-06,
+ "loss": 0.36720116138458253,
+ "num_input_tokens_seen": 542309048,
+ "step": 12395,
+ "train_runtime": 95859.9652,
+ "train_tokens_per_second": 5657.305
+ },
+ {
+ "epoch": 0.8349043899811474,
+ "grad_norm": 0.683112485827057,
+ "learning_rate": 6.2826838163975724e-06,
+ "loss": 0.3783855438232422,
+ "num_input_tokens_seen": 542526280,
+ "step": 12400,
+ "train_runtime": 95898.4045,
+ "train_tokens_per_second": 5657.302
+ },
+ {
+ "epoch": 0.8352410449771075,
+ "grad_norm": 0.8138279320478745,
+ "learning_rate": 6.2801280406127905e-06,
+ "loss": 0.35349900722503663,
+ "num_input_tokens_seen": 542750064,
+ "step": 12405,
+ "train_runtime": 95934.8168,
+ "train_tokens_per_second": 5657.488
+ },
+ {
+ "epoch": 0.8355776999730676,
+ "grad_norm": 0.7615550252521133,
+ "learning_rate": 6.2775719068442845e-06,
+ "loss": 0.36704421043395996,
+ "num_input_tokens_seen": 542968240,
+ "step": 12410,
+ "train_runtime": 95972.339,
+ "train_tokens_per_second": 5657.549
+ },
+ {
+ "epoch": 0.8359143549690278,
+ "grad_norm": 0.7219105847794507,
+ "learning_rate": 6.275015415806869e-06,
+ "loss": 0.3532416343688965,
+ "num_input_tokens_seen": 543175408,
+ "step": 12415,
+ "train_runtime": 96016.4623,
+ "train_tokens_per_second": 5657.107
+ },
+ {
+ "epoch": 0.8362510099649879,
+ "grad_norm": 0.6790545597022978,
+ "learning_rate": 6.272458568215458e-06,
+ "loss": 0.3773714303970337,
+ "num_input_tokens_seen": 543389992,
+ "step": 12420,
+ "train_runtime": 96057.0753,
+ "train_tokens_per_second": 5656.949
+ },
+ {
+ "epoch": 0.836587664960948,
+ "grad_norm": 0.758978760003948,
+ "learning_rate": 6.269901364785066e-06,
+ "loss": 0.37525081634521484,
+ "num_input_tokens_seen": 543616944,
+ "step": 12425,
+ "train_runtime": 96095.6171,
+ "train_tokens_per_second": 5657.042
+ },
+ {
+ "epoch": 0.8369243199569082,
+ "grad_norm": 0.6823393482935673,
+ "learning_rate": 6.2673438062308055e-06,
+ "loss": 0.37595961093902586,
+ "num_input_tokens_seen": 543859984,
+ "step": 12430,
+ "train_runtime": 96131.3734,
+ "train_tokens_per_second": 5657.466
+ },
+ {
+ "epoch": 0.8372609749528683,
+ "grad_norm": 0.9689604562670696,
+ "learning_rate": 6.264785893267892e-06,
+ "loss": 0.3472005367279053,
+ "num_input_tokens_seen": 544063976,
+ "step": 12435,
+ "train_runtime": 96162.9294,
+ "train_tokens_per_second": 5657.731
+ },
+ {
+ "epoch": 0.8375976299488285,
+ "grad_norm": 0.7213840812199924,
+ "learning_rate": 6.2622276266116366e-06,
+ "loss": 0.36200590133666993,
+ "num_input_tokens_seen": 544288640,
+ "step": 12440,
+ "train_runtime": 96199.7795,
+ "train_tokens_per_second": 5657.899
+ },
+ {
+ "epoch": 0.8379342849447886,
+ "grad_norm": 0.6737267934047665,
+ "learning_rate": 6.25966900697745e-06,
+ "loss": 0.37799394130706787,
+ "num_input_tokens_seen": 544514152,
+ "step": 12445,
+ "train_runtime": 96237.3914,
+ "train_tokens_per_second": 5658.031
+ },
+ {
+ "epoch": 0.8382709399407487,
+ "grad_norm": 0.6929403643911132,
+ "learning_rate": 6.257110035080843e-06,
+ "loss": 0.3491935729980469,
+ "num_input_tokens_seen": 544741824,
+ "step": 12450,
+ "train_runtime": 96274.7948,
+ "train_tokens_per_second": 5658.198
+ },
+ {
+ "epoch": 0.8386075949367089,
+ "grad_norm": 0.6942505611698238,
+ "learning_rate": 6.254550711637422e-06,
+ "loss": 0.40814781188964844,
+ "num_input_tokens_seen": 544950448,
+ "step": 12455,
+ "train_runtime": 96314.5199,
+ "train_tokens_per_second": 5658.03
+ },
+ {
+ "epoch": 0.838944249932669,
+ "grad_norm": 0.7575909904692931,
+ "learning_rate": 6.251991037362897e-06,
+ "loss": 0.38295414447784426,
+ "num_input_tokens_seen": 545159072,
+ "step": 12460,
+ "train_runtime": 96349.3253,
+ "train_tokens_per_second": 5658.151
+ },
+ {
+ "epoch": 0.8392809049286292,
+ "grad_norm": 0.7581123803912536,
+ "learning_rate": 6.249431012973068e-06,
+ "loss": 0.37713871002197263,
+ "num_input_tokens_seen": 545367120,
+ "step": 12465,
+ "train_runtime": 96388.6672,
+ "train_tokens_per_second": 5658.0
+ },
+ {
+ "epoch": 0.8396175599245893,
+ "grad_norm": 0.7683081862438146,
+ "learning_rate": 6.246870639183843e-06,
+ "loss": 0.37398107051849366,
+ "num_input_tokens_seen": 545578664,
+ "step": 12470,
+ "train_runtime": 96426.5472,
+ "train_tokens_per_second": 5657.972
+ },
+ {
+ "epoch": 0.8399542149205494,
+ "grad_norm": 0.8223266832633437,
+ "learning_rate": 6.2443099167112185e-06,
+ "loss": 0.39701476097106936,
+ "num_input_tokens_seen": 545803912,
+ "step": 12475,
+ "train_runtime": 96462.5666,
+ "train_tokens_per_second": 5658.194
+ },
+ {
+ "epoch": 0.8402908699165096,
+ "grad_norm": 0.6969953488947742,
+ "learning_rate": 6.241748846271293e-06,
+ "loss": 0.3778507709503174,
+ "num_input_tokens_seen": 546030536,
+ "step": 12480,
+ "train_runtime": 96496.5077,
+ "train_tokens_per_second": 5658.552
+ },
+ {
+ "epoch": 0.8406275249124697,
+ "grad_norm": 0.7778677325039889,
+ "learning_rate": 6.239187428580263e-06,
+ "loss": 0.37636547088623046,
+ "num_input_tokens_seen": 546236768,
+ "step": 12485,
+ "train_runtime": 96538.6051,
+ "train_tokens_per_second": 5658.221
+ },
+ {
+ "epoch": 0.8409641799084299,
+ "grad_norm": 0.8059997970084556,
+ "learning_rate": 6.236625664354421e-06,
+ "loss": 0.3784951686859131,
+ "num_input_tokens_seen": 546462096,
+ "step": 12490,
+ "train_runtime": 96574.6617,
+ "train_tokens_per_second": 5658.442
+ },
+ {
+ "epoch": 0.84130083490439,
+ "grad_norm": 0.8520532583827878,
+ "learning_rate": 6.2340635543101545e-06,
+ "loss": 0.3676713466644287,
+ "num_input_tokens_seen": 546657168,
+ "step": 12495,
+ "train_runtime": 96614.4259,
+ "train_tokens_per_second": 5658.132
+ },
+ {
+ "epoch": 0.8416374899003501,
+ "grad_norm": 0.6539078891022843,
+ "learning_rate": 6.23150109916395e-06,
+ "loss": 0.33987085819244384,
+ "num_input_tokens_seen": 546881208,
+ "step": 12500,
+ "train_runtime": 96655.3189,
+ "train_tokens_per_second": 5658.056
+ },
+ {
+ "epoch": 0.8419741448963103,
+ "grad_norm": 0.6624040098906347,
+ "learning_rate": 6.22893829963239e-06,
+ "loss": 0.3859384059906006,
+ "num_input_tokens_seen": 547092816,
+ "step": 12505,
+ "train_runtime": 96691.3494,
+ "train_tokens_per_second": 5658.136
+ },
+ {
+ "epoch": 0.8423107998922704,
+ "grad_norm": 0.7298942090178581,
+ "learning_rate": 6.226375156432153e-06,
+ "loss": 0.357137393951416,
+ "num_input_tokens_seen": 547322656,
+ "step": 12510,
+ "train_runtime": 96731.546,
+ "train_tokens_per_second": 5658.161
+ },
+ {
+ "epoch": 0.8426474548882306,
+ "grad_norm": 0.7316486599091492,
+ "learning_rate": 6.2238116702800145e-06,
+ "loss": 0.3710299491882324,
+ "num_input_tokens_seen": 547537944,
+ "step": 12515,
+ "train_runtime": 96762.4833,
+ "train_tokens_per_second": 5658.577
+ },
+ {
+ "epoch": 0.8429841098841907,
+ "grad_norm": 0.7998109937320629,
+ "learning_rate": 6.221247841892846e-06,
+ "loss": 0.3879734992980957,
+ "num_input_tokens_seen": 547746624,
+ "step": 12520,
+ "train_runtime": 96805.6441,
+ "train_tokens_per_second": 5658.21
+ },
+ {
+ "epoch": 0.8433207648801508,
+ "grad_norm": 0.7307238083679223,
+ "learning_rate": 6.218683671987611e-06,
+ "loss": 0.3691586971282959,
+ "num_input_tokens_seen": 547983928,
+ "step": 12525,
+ "train_runtime": 96840.3547,
+ "train_tokens_per_second": 5658.632
+ },
+ {
+ "epoch": 0.843657419876111,
+ "grad_norm": 0.7271187490915773,
+ "learning_rate": 6.216119161281373e-06,
+ "loss": 0.37589943408966064,
+ "num_input_tokens_seen": 548209600,
+ "step": 12530,
+ "train_runtime": 96879.0798,
+ "train_tokens_per_second": 5658.699
+ },
+ {
+ "epoch": 0.8439940748720711,
+ "grad_norm": 0.7311632772691732,
+ "learning_rate": 6.213554310491291e-06,
+ "loss": 0.3672966003417969,
+ "num_input_tokens_seen": 548421688,
+ "step": 12535,
+ "train_runtime": 96928.6015,
+ "train_tokens_per_second": 5657.996
+ },
+ {
+ "epoch": 0.8443307298680313,
+ "grad_norm": 0.7450184498891292,
+ "learning_rate": 6.210989120334616e-06,
+ "loss": 0.362609338760376,
+ "num_input_tokens_seen": 548644072,
+ "step": 12540,
+ "train_runtime": 96971.5458,
+ "train_tokens_per_second": 5657.784
+ },
+ {
+ "epoch": 0.8446673848639914,
+ "grad_norm": 0.7758427572495085,
+ "learning_rate": 6.208423591528694e-06,
+ "loss": 0.38643317222595214,
+ "num_input_tokens_seen": 548854336,
+ "step": 12545,
+ "train_runtime": 97009.7887,
+ "train_tokens_per_second": 5657.721
+ },
+ {
+ "epoch": 0.8450040398599515,
+ "grad_norm": 0.688742133220875,
+ "learning_rate": 6.205857724790969e-06,
+ "loss": 0.3612468481063843,
+ "num_input_tokens_seen": 549072600,
+ "step": 12550,
+ "train_runtime": 97048.5889,
+ "train_tokens_per_second": 5657.708
+ },
+ {
+ "epoch": 0.8453406948559117,
+ "grad_norm": 0.7926232068697935,
+ "learning_rate": 6.203291520838974e-06,
+ "loss": 0.409790563583374,
+ "num_input_tokens_seen": 549303984,
+ "step": 12555,
+ "train_runtime": 97086.2605,
+ "train_tokens_per_second": 5657.896
+ },
+ {
+ "epoch": 0.8456773498518718,
+ "grad_norm": 0.7041401739211672,
+ "learning_rate": 6.200724980390344e-06,
+ "loss": 0.34305951595306394,
+ "num_input_tokens_seen": 549497696,
+ "step": 12560,
+ "train_runtime": 97125.3875,
+ "train_tokens_per_second": 5657.611
+ },
+ {
+ "epoch": 0.846014004847832,
+ "grad_norm": 0.6959581809215943,
+ "learning_rate": 6.198158104162804e-06,
+ "loss": 0.3178701400756836,
+ "num_input_tokens_seen": 549713600,
+ "step": 12565,
+ "train_runtime": 97158.9336,
+ "train_tokens_per_second": 5657.88
+ },
+ {
+ "epoch": 0.8463506598437921,
+ "grad_norm": 0.6766745725154476,
+ "learning_rate": 6.1955908928741685e-06,
+ "loss": 0.37602977752685546,
+ "num_input_tokens_seen": 549953184,
+ "step": 12570,
+ "train_runtime": 97195.8557,
+ "train_tokens_per_second": 5658.196
+ },
+ {
+ "epoch": 0.8466873148397522,
+ "grad_norm": 0.6881093744203187,
+ "learning_rate": 6.193023347242353e-06,
+ "loss": 0.38083457946777344,
+ "num_input_tokens_seen": 550184352,
+ "step": 12575,
+ "train_runtime": 97232.314,
+ "train_tokens_per_second": 5658.452
+ },
+ {
+ "epoch": 0.8470239698357124,
+ "grad_norm": 0.741931440728507,
+ "learning_rate": 6.190455467985361e-06,
+ "loss": 0.3633177042007446,
+ "num_input_tokens_seen": 550405552,
+ "step": 12580,
+ "train_runtime": 97272.8862,
+ "train_tokens_per_second": 5658.366
+ },
+ {
+ "epoch": 0.8473606248316725,
+ "grad_norm": 0.7792859259983645,
+ "learning_rate": 6.187887255821295e-06,
+ "loss": 0.3954836368560791,
+ "num_input_tokens_seen": 550625896,
+ "step": 12585,
+ "train_runtime": 97305.8601,
+ "train_tokens_per_second": 5658.713
+ },
+ {
+ "epoch": 0.8476972798276327,
+ "grad_norm": 0.7394004400669401,
+ "learning_rate": 6.185318711468345e-06,
+ "loss": 0.3501850128173828,
+ "num_input_tokens_seen": 550827088,
+ "step": 12590,
+ "train_runtime": 97346.1248,
+ "train_tokens_per_second": 5658.439
+ },
+ {
+ "epoch": 0.8480339348235928,
+ "grad_norm": 0.7906419360212633,
+ "learning_rate": 6.182749835644799e-06,
+ "loss": 0.3587897777557373,
+ "num_input_tokens_seen": 551036712,
+ "step": 12595,
+ "train_runtime": 97378.9075,
+ "train_tokens_per_second": 5658.687
+ },
+ {
+ "epoch": 0.8483705898195529,
+ "grad_norm": 0.6720676543784547,
+ "learning_rate": 6.180180629069031e-06,
+ "loss": 0.3534966468811035,
+ "num_input_tokens_seen": 551269008,
+ "step": 12600,
+ "train_runtime": 97416.9713,
+ "train_tokens_per_second": 5658.86
+ },
+ {
+ "epoch": 0.8487072448155131,
+ "grad_norm": 1.0844161686609408,
+ "learning_rate": 6.177611092459514e-06,
+ "loss": 0.3404214859008789,
+ "num_input_tokens_seen": 551486584,
+ "step": 12605,
+ "train_runtime": 97454.9229,
+ "train_tokens_per_second": 5658.889
+ },
+ {
+ "epoch": 0.8490438998114732,
+ "grad_norm": 0.8339846261840644,
+ "learning_rate": 6.175041226534809e-06,
+ "loss": 0.3584805965423584,
+ "num_input_tokens_seen": 551707952,
+ "step": 12610,
+ "train_runtime": 97487.377,
+ "train_tokens_per_second": 5659.276
+ },
+ {
+ "epoch": 0.8493805548074334,
+ "grad_norm": 0.688269893800755,
+ "learning_rate": 6.172471032013574e-06,
+ "loss": 0.327656888961792,
+ "num_input_tokens_seen": 551932560,
+ "step": 12615,
+ "train_runtime": 97526.9729,
+ "train_tokens_per_second": 5659.281
+ },
+ {
+ "epoch": 0.8497172098033935,
+ "grad_norm": 0.6595384550469138,
+ "learning_rate": 6.169900509614553e-06,
+ "loss": 0.3601629972457886,
+ "num_input_tokens_seen": 552163800,
+ "step": 12620,
+ "train_runtime": 97577.6237,
+ "train_tokens_per_second": 5658.713
+ },
+ {
+ "epoch": 0.8500538647993536,
+ "grad_norm": 0.8030731667400421,
+ "learning_rate": 6.167329660056585e-06,
+ "loss": 0.36720144748687744,
+ "num_input_tokens_seen": 552375472,
+ "step": 12625,
+ "train_runtime": 97617.7309,
+ "train_tokens_per_second": 5658.557
+ },
+ {
+ "epoch": 0.8503905197953138,
+ "grad_norm": 0.6580730799233167,
+ "learning_rate": 6.1647584840586e-06,
+ "loss": 0.36503446102142334,
+ "num_input_tokens_seen": 552599576,
+ "step": 12630,
+ "train_runtime": 97665.6991,
+ "train_tokens_per_second": 5658.072
+ },
+ {
+ "epoch": 0.8507271747912739,
+ "grad_norm": 0.7006154386468065,
+ "learning_rate": 6.1621869823396175e-06,
+ "loss": 0.37128496170043945,
+ "num_input_tokens_seen": 552832696,
+ "step": 12635,
+ "train_runtime": 97703.9353,
+ "train_tokens_per_second": 5658.244
+ },
+ {
+ "epoch": 0.851063829787234,
+ "grad_norm": 0.7225238597915238,
+ "learning_rate": 6.159615155618752e-06,
+ "loss": 0.37408173084259033,
+ "num_input_tokens_seen": 553059920,
+ "step": 12640,
+ "train_runtime": 97737.0325,
+ "train_tokens_per_second": 5658.653
+ },
+ {
+ "epoch": 0.8514004847831942,
+ "grad_norm": 0.8136467242343185,
+ "learning_rate": 6.157043004615207e-06,
+ "loss": 0.38596835136413576,
+ "num_input_tokens_seen": 553300640,
+ "step": 12645,
+ "train_runtime": 97780.3263,
+ "train_tokens_per_second": 5658.609
+ },
+ {
+ "epoch": 0.8517371397791543,
+ "grad_norm": 0.7939080541143891,
+ "learning_rate": 6.1544705300482755e-06,
+ "loss": 0.3616677761077881,
+ "num_input_tokens_seen": 553532032,
+ "step": 12650,
+ "train_runtime": 97820.9636,
+ "train_tokens_per_second": 5658.624
+ },
+ {
+ "epoch": 0.8520737947751145,
+ "grad_norm": 0.6610977560801387,
+ "learning_rate": 6.151897732637341e-06,
+ "loss": 0.3330955982208252,
+ "num_input_tokens_seen": 553739416,
+ "step": 12655,
+ "train_runtime": 97857.6515,
+ "train_tokens_per_second": 5658.622
+ },
+ {
+ "epoch": 0.8524104497710746,
+ "grad_norm": 0.8508497854348367,
+ "learning_rate": 6.149324613101877e-06,
+ "loss": 0.37527806758880616,
+ "num_input_tokens_seen": 553959976,
+ "step": 12660,
+ "train_runtime": 97897.0387,
+ "train_tokens_per_second": 5658.598
+ },
+ {
+ "epoch": 0.8527471047670347,
+ "grad_norm": 0.6193562313644957,
+ "learning_rate": 6.146751172161453e-06,
+ "loss": 0.33863201141357424,
+ "num_input_tokens_seen": 554189776,
+ "step": 12665,
+ "train_runtime": 97940.3461,
+ "train_tokens_per_second": 5658.442
+ },
+ {
+ "epoch": 0.8530837597629949,
+ "grad_norm": 0.7796518940843955,
+ "learning_rate": 6.144177410535719e-06,
+ "loss": 0.3652313470840454,
+ "num_input_tokens_seen": 554414232,
+ "step": 12670,
+ "train_runtime": 97979.2461,
+ "train_tokens_per_second": 5658.486
+ },
+ {
+ "epoch": 0.853420414758955,
+ "grad_norm": 0.7186651171193743,
+ "learning_rate": 6.14160332894442e-06,
+ "loss": 0.3651700973510742,
+ "num_input_tokens_seen": 554625192,
+ "step": 12675,
+ "train_runtime": 98015.4768,
+ "train_tokens_per_second": 5658.547
+ },
+ {
+ "epoch": 0.8537570697549152,
+ "grad_norm": 0.724709697937217,
+ "learning_rate": 6.139028928107391e-06,
+ "loss": 0.36736459732055665,
+ "num_input_tokens_seen": 554842240,
+ "step": 12680,
+ "train_runtime": 98056.2763,
+ "train_tokens_per_second": 5658.406
+ },
+ {
+ "epoch": 0.8540937247508753,
+ "grad_norm": 0.7018554477013618,
+ "learning_rate": 6.136454208744553e-06,
+ "loss": 0.3479247808456421,
+ "num_input_tokens_seen": 555060144,
+ "step": 12685,
+ "train_runtime": 98097.7581,
+ "train_tokens_per_second": 5658.235
+ },
+ {
+ "epoch": 0.8544303797468354,
+ "grad_norm": 0.7210589341441987,
+ "learning_rate": 6.133879171575921e-06,
+ "loss": 0.3790309906005859,
+ "num_input_tokens_seen": 555276648,
+ "step": 12690,
+ "train_runtime": 98135.4374,
+ "train_tokens_per_second": 5658.268
+ },
+ {
+ "epoch": 0.8547670347427956,
+ "grad_norm": 0.8012051731436295,
+ "learning_rate": 6.131303817321595e-06,
+ "loss": 0.3976943254470825,
+ "num_input_tokens_seen": 555506600,
+ "step": 12695,
+ "train_runtime": 98173.0099,
+ "train_tokens_per_second": 5658.445
+ },
+ {
+ "epoch": 0.8551036897387557,
+ "grad_norm": 0.6686702415410488,
+ "learning_rate": 6.128728146701763e-06,
+ "loss": 0.3907043933868408,
+ "num_input_tokens_seen": 555735640,
+ "step": 12700,
+ "train_runtime": 98208.5683,
+ "train_tokens_per_second": 5658.729
+ },
+ {
+ "epoch": 0.8554403447347159,
+ "grad_norm": 0.6745127609405563,
+ "learning_rate": 6.126152160436705e-06,
+ "loss": 0.3401200294494629,
+ "num_input_tokens_seen": 555954936,
+ "step": 12705,
+ "train_runtime": 98247.4191,
+ "train_tokens_per_second": 5658.723
+ },
+ {
+ "epoch": 0.855776999730676,
+ "grad_norm": 0.9952487891058109,
+ "learning_rate": 6.123575859246784e-06,
+ "loss": 0.3820429801940918,
+ "num_input_tokens_seen": 556162568,
+ "step": 12710,
+ "train_runtime": 98285.6421,
+ "train_tokens_per_second": 5658.635
+ },
+ {
+ "epoch": 0.8561136547266361,
+ "grad_norm": 0.7343846388254841,
+ "learning_rate": 6.120999243852459e-06,
+ "loss": 0.3488877773284912,
+ "num_input_tokens_seen": 556374088,
+ "step": 12715,
+ "train_runtime": 98325.3673,
+ "train_tokens_per_second": 5658.5
+ },
+ {
+ "epoch": 0.8564503097225963,
+ "grad_norm": 0.8403978577775663,
+ "learning_rate": 6.118422314974269e-06,
+ "loss": 0.3801596164703369,
+ "num_input_tokens_seen": 556596000,
+ "step": 12720,
+ "train_runtime": 98368.0103,
+ "train_tokens_per_second": 5658.303
+ },
+ {
+ "epoch": 0.8567869647185564,
+ "grad_norm": 0.7544062671681226,
+ "learning_rate": 6.115845073332845e-06,
+ "loss": 0.36686029434204104,
+ "num_input_tokens_seen": 556828456,
+ "step": 12725,
+ "train_runtime": 98413.3452,
+ "train_tokens_per_second": 5658.058
+ },
+ {
+ "epoch": 0.8571236197145166,
+ "grad_norm": 0.8566960890477198,
+ "learning_rate": 6.113267519648905e-06,
+ "loss": 0.39199421405792234,
+ "num_input_tokens_seen": 557067040,
+ "step": 12730,
+ "train_runtime": 98457.4483,
+ "train_tokens_per_second": 5657.947
+ },
+ {
+ "epoch": 0.8574602747104767,
+ "grad_norm": 0.6753240704632734,
+ "learning_rate": 6.110689654643253e-06,
+ "loss": 0.36382412910461426,
+ "num_input_tokens_seen": 557274264,
+ "step": 12735,
+ "train_runtime": 98497.8079,
+ "train_tokens_per_second": 5657.733
+ },
+ {
+ "epoch": 0.8577969297064368,
+ "grad_norm": 0.7623447924884228,
+ "learning_rate": 6.1081114790367805e-06,
+ "loss": 0.37306723594665525,
+ "num_input_tokens_seen": 557484064,
+ "step": 12740,
+ "train_runtime": 98531.9864,
+ "train_tokens_per_second": 5657.899
+ },
+ {
+ "epoch": 0.858133584702397,
+ "grad_norm": 0.7690727585148884,
+ "learning_rate": 6.105532993550467e-06,
+ "loss": 0.37079343795776365,
+ "num_input_tokens_seen": 557707168,
+ "step": 12745,
+ "train_runtime": 98570.6779,
+ "train_tokens_per_second": 5657.942
+ },
+ {
+ "epoch": 0.8584702396983571,
+ "grad_norm": 0.7227829734774792,
+ "learning_rate": 6.1029541989053765e-06,
+ "loss": 0.3675063610076904,
+ "num_input_tokens_seen": 557923328,
+ "step": 12750,
+ "train_runtime": 98611.8723,
+ "train_tokens_per_second": 5657.77
+ },
+ {
+ "epoch": 0.8588068946943173,
+ "grad_norm": 0.6216338336006675,
+ "learning_rate": 6.100375095822661e-06,
+ "loss": 0.3660135746002197,
+ "num_input_tokens_seen": 558141912,
+ "step": 12755,
+ "train_runtime": 98653.6004,
+ "train_tokens_per_second": 5657.593
+ },
+ {
+ "epoch": 0.8591435496902774,
+ "grad_norm": 0.7752180098632147,
+ "learning_rate": 6.097795685023558e-06,
+ "loss": 0.3654344081878662,
+ "num_input_tokens_seen": 558363664,
+ "step": 12760,
+ "train_runtime": 98689.3362,
+ "train_tokens_per_second": 5657.791
+ },
+ {
+ "epoch": 0.8594802046862375,
+ "grad_norm": 0.7139656897700937,
+ "learning_rate": 6.095215967229396e-06,
+ "loss": 0.3468952655792236,
+ "num_input_tokens_seen": 558569128,
+ "step": 12765,
+ "train_runtime": 98726.8939,
+ "train_tokens_per_second": 5657.72
+ },
+ {
+ "epoch": 0.8598168596821977,
+ "grad_norm": 0.6751611340302733,
+ "learning_rate": 6.092635943161578e-06,
+ "loss": 0.37009525299072266,
+ "num_input_tokens_seen": 558798144,
+ "step": 12770,
+ "train_runtime": 98770.45,
+ "train_tokens_per_second": 5657.544
+ },
+ {
+ "epoch": 0.8601535146781578,
+ "grad_norm": 0.7202350000001299,
+ "learning_rate": 6.090055613541603e-06,
+ "loss": 0.34846246242523193,
+ "num_input_tokens_seen": 559026000,
+ "step": 12775,
+ "train_runtime": 98813.4858,
+ "train_tokens_per_second": 5657.386
+ },
+ {
+ "epoch": 0.860490169674118,
+ "grad_norm": 0.7513541698981121,
+ "learning_rate": 6.087474979091052e-06,
+ "loss": 0.3586388111114502,
+ "num_input_tokens_seen": 559233040,
+ "step": 12780,
+ "train_runtime": 98860.409,
+ "train_tokens_per_second": 5656.795
+ },
+ {
+ "epoch": 0.8608268246700781,
+ "grad_norm": 0.7207910613458428,
+ "learning_rate": 6.084894040531591e-06,
+ "loss": 0.3598465919494629,
+ "num_input_tokens_seen": 559456240,
+ "step": 12785,
+ "train_runtime": 98903.5566,
+ "train_tokens_per_second": 5656.584
+ },
+ {
+ "epoch": 0.8611634796660382,
+ "grad_norm": 0.7450858357428686,
+ "learning_rate": 6.0823127985849705e-06,
+ "loss": 0.35327887535095215,
+ "num_input_tokens_seen": 559662696,
+ "step": 12790,
+ "train_runtime": 98942.3562,
+ "train_tokens_per_second": 5656.452
+ },
+ {
+ "epoch": 0.8615001346619984,
+ "grad_norm": 0.7413157970110851,
+ "learning_rate": 6.079731253973028e-06,
+ "loss": 0.3576143741607666,
+ "num_input_tokens_seen": 559886624,
+ "step": 12795,
+ "train_runtime": 98988.6398,
+ "train_tokens_per_second": 5656.069
+ },
+ {
+ "epoch": 0.8618367896579585,
+ "grad_norm": 0.7990871684605166,
+ "learning_rate": 6.077149407417683e-06,
+ "loss": 0.36043851375579833,
+ "num_input_tokens_seen": 560101432,
+ "step": 12800,
+ "train_runtime": 99024.1718,
+ "train_tokens_per_second": 5656.209
+ },
+ {
+ "epoch": 0.8621734446539187,
+ "grad_norm": 0.7364962437233615,
+ "learning_rate": 6.07456725964094e-06,
+ "loss": 0.3793959617614746,
+ "num_input_tokens_seen": 560309472,
+ "step": 12805,
+ "train_runtime": 99059.299,
+ "train_tokens_per_second": 5656.304
+ },
+ {
+ "epoch": 0.8625100996498788,
+ "grad_norm": 0.680598341183089,
+ "learning_rate": 6.07198481136489e-06,
+ "loss": 0.3436178684234619,
+ "num_input_tokens_seen": 560520408,
+ "step": 12810,
+ "train_runtime": 99092.681,
+ "train_tokens_per_second": 5656.527
+ },
+ {
+ "epoch": 0.8628467546458389,
+ "grad_norm": 0.765274818340009,
+ "learning_rate": 6.069402063311706e-06,
+ "loss": 0.3549325704574585,
+ "num_input_tokens_seen": 560731064,
+ "step": 12815,
+ "train_runtime": 99135.7209,
+ "train_tokens_per_second": 5656.196
+ },
+ {
+ "epoch": 0.8631834096417991,
+ "grad_norm": 0.8395744171280569,
+ "learning_rate": 6.066819016203646e-06,
+ "loss": 0.36306018829345704,
+ "num_input_tokens_seen": 560944360,
+ "step": 12820,
+ "train_runtime": 99173.0123,
+ "train_tokens_per_second": 5656.22
+ },
+ {
+ "epoch": 0.8635200646377592,
+ "grad_norm": 0.7436029507337527,
+ "learning_rate": 6.064235670763048e-06,
+ "loss": 0.36905884742736816,
+ "num_input_tokens_seen": 561164928,
+ "step": 12825,
+ "train_runtime": 99209.242,
+ "train_tokens_per_second": 5656.378
+ },
+ {
+ "epoch": 0.8638567196337193,
+ "grad_norm": 0.6763369772275482,
+ "learning_rate": 6.06165202771234e-06,
+ "loss": 0.3765472412109375,
+ "num_input_tokens_seen": 561394056,
+ "step": 12830,
+ "train_runtime": 99250.4089,
+ "train_tokens_per_second": 5656.34
+ },
+ {
+ "epoch": 0.8641933746296795,
+ "grad_norm": 0.6931570065850123,
+ "learning_rate": 6.059068087774026e-06,
+ "loss": 0.34474341869354247,
+ "num_input_tokens_seen": 561593424,
+ "step": 12835,
+ "train_runtime": 99291.1127,
+ "train_tokens_per_second": 5656.029
+ },
+ {
+ "epoch": 0.8645300296256396,
+ "grad_norm": 0.7306127243942253,
+ "learning_rate": 6.0564838516707005e-06,
+ "loss": 0.38959674835205077,
+ "num_input_tokens_seen": 561821784,
+ "step": 12840,
+ "train_runtime": 99328.0768,
+ "train_tokens_per_second": 5656.223
+ },
+ {
+ "epoch": 0.8648666846215998,
+ "grad_norm": 0.6692585369253214,
+ "learning_rate": 6.053899320125033e-06,
+ "loss": 0.34748311042785646,
+ "num_input_tokens_seen": 562042168,
+ "step": 12845,
+ "train_runtime": 99368.2816,
+ "train_tokens_per_second": 5656.153
+ },
+ {
+ "epoch": 0.8652033396175599,
+ "grad_norm": 0.731387810949299,
+ "learning_rate": 6.051314493859782e-06,
+ "loss": 0.37701342105865476,
+ "num_input_tokens_seen": 562268616,
+ "step": 12850,
+ "train_runtime": 99405.5774,
+ "train_tokens_per_second": 5656.309
+ },
+ {
+ "epoch": 0.86553999461352,
+ "grad_norm": 0.744983156104233,
+ "learning_rate": 6.048729373597786e-06,
+ "loss": 0.36400351524353025,
+ "num_input_tokens_seen": 562488728,
+ "step": 12855,
+ "train_runtime": 99441.1097,
+ "train_tokens_per_second": 5656.501
+ },
+ {
+ "epoch": 0.8658766496094802,
+ "grad_norm": 0.7052968569468162,
+ "learning_rate": 6.046143960061963e-06,
+ "loss": 0.38647964000701907,
+ "num_input_tokens_seen": 562715632,
+ "step": 12860,
+ "train_runtime": 99479.679,
+ "train_tokens_per_second": 5656.589
+ },
+ {
+ "epoch": 0.8662133046054403,
+ "grad_norm": 0.8490020320396742,
+ "learning_rate": 6.043558253975319e-06,
+ "loss": 0.3893282413482666,
+ "num_input_tokens_seen": 562949544,
+ "step": 12865,
+ "train_runtime": 99522.1085,
+ "train_tokens_per_second": 5656.528
+ },
+ {
+ "epoch": 0.8665499596014005,
+ "grad_norm": 0.7443300106386931,
+ "learning_rate": 6.040972256060936e-06,
+ "loss": 0.35498225688934326,
+ "num_input_tokens_seen": 563161168,
+ "step": 12870,
+ "train_runtime": 99567.3086,
+ "train_tokens_per_second": 5656.085
+ },
+ {
+ "epoch": 0.8668866145973606,
+ "grad_norm": 0.6849806622056446,
+ "learning_rate": 6.038385967041982e-06,
+ "loss": 0.3538069248199463,
+ "num_input_tokens_seen": 563369656,
+ "step": 12875,
+ "train_runtime": 99613.8289,
+ "train_tokens_per_second": 5655.537
+ },
+ {
+ "epoch": 0.8672232695933207,
+ "grad_norm": 0.6384169452977918,
+ "learning_rate": 6.035799387641703e-06,
+ "loss": 0.36878552436828616,
+ "num_input_tokens_seen": 563594552,
+ "step": 12880,
+ "train_runtime": 99655.4177,
+ "train_tokens_per_second": 5655.433
+ },
+ {
+ "epoch": 0.8675599245892809,
+ "grad_norm": 0.7636357717419208,
+ "learning_rate": 6.033212518583427e-06,
+ "loss": 0.3786178112030029,
+ "num_input_tokens_seen": 563803512,
+ "step": 12885,
+ "train_runtime": 99695.3116,
+ "train_tokens_per_second": 5655.266
+ },
+ {
+ "epoch": 0.867896579585241,
+ "grad_norm": 0.8323239828663773,
+ "learning_rate": 6.030625360590567e-06,
+ "loss": 0.35876643657684326,
+ "num_input_tokens_seen": 564030520,
+ "step": 12890,
+ "train_runtime": 99734.3991,
+ "train_tokens_per_second": 5655.326
+ },
+ {
+ "epoch": 0.8682332345812012,
+ "grad_norm": 0.7006071245579706,
+ "learning_rate": 6.028037914386609e-06,
+ "loss": 0.3620233297348022,
+ "num_input_tokens_seen": 564252688,
+ "step": 12895,
+ "train_runtime": 99773.9473,
+ "train_tokens_per_second": 5655.311
+ },
+ {
+ "epoch": 0.8685698895771613,
+ "grad_norm": 0.7617834642572114,
+ "learning_rate": 6.025450180695128e-06,
+ "loss": 0.38210210800170896,
+ "num_input_tokens_seen": 564471472,
+ "step": 12900,
+ "train_runtime": 99808.9271,
+ "train_tokens_per_second": 5655.521
+ },
+ {
+ "epoch": 0.8689065445731214,
+ "grad_norm": 0.8560111437943349,
+ "learning_rate": 6.022862160239774e-06,
+ "loss": 0.3630260467529297,
+ "num_input_tokens_seen": 564683208,
+ "step": 12905,
+ "train_runtime": 99850.9033,
+ "train_tokens_per_second": 5655.264
+ },
+ {
+ "epoch": 0.8692431995690816,
+ "grad_norm": 0.6306746195476955,
+ "learning_rate": 6.020273853744277e-06,
+ "loss": 0.3535295009613037,
+ "num_input_tokens_seen": 564903760,
+ "step": 12910,
+ "train_runtime": 99888.6442,
+ "train_tokens_per_second": 5655.335
+ },
+ {
+ "epoch": 0.8695798545650417,
+ "grad_norm": 0.6756821029513189,
+ "learning_rate": 6.017685261932452e-06,
+ "loss": 0.35518722534179686,
+ "num_input_tokens_seen": 565105760,
+ "step": 12915,
+ "train_runtime": 99923.6252,
+ "train_tokens_per_second": 5655.377
+ },
+ {
+ "epoch": 0.8699165095610019,
+ "grad_norm": 0.75004570307814,
+ "learning_rate": 6.015096385528189e-06,
+ "loss": 0.37440195083618166,
+ "num_input_tokens_seen": 565318768,
+ "step": 12920,
+ "train_runtime": 99960.4635,
+ "train_tokens_per_second": 5655.424
+ },
+ {
+ "epoch": 0.870253164556962,
+ "grad_norm": 0.725287349099034,
+ "learning_rate": 6.012507225255457e-06,
+ "loss": 0.3945824146270752,
+ "num_input_tokens_seen": 565515560,
+ "step": 12925,
+ "train_runtime": 99995.6077,
+ "train_tokens_per_second": 5655.404
+ },
+ {
+ "epoch": 0.8705898195529221,
+ "grad_norm": 0.6923174837526764,
+ "learning_rate": 6.009917781838309e-06,
+ "loss": 0.3706216812133789,
+ "num_input_tokens_seen": 565736648,
+ "step": 12930,
+ "train_runtime": 100027.8477,
+ "train_tokens_per_second": 5655.791
+ },
+ {
+ "epoch": 0.8709264745488823,
+ "grad_norm": 0.6934584410887805,
+ "learning_rate": 6.007328056000873e-06,
+ "loss": 0.37667386531829833,
+ "num_input_tokens_seen": 565960800,
+ "step": 12935,
+ "train_runtime": 100063.8699,
+ "train_tokens_per_second": 5655.996
+ },
+ {
+ "epoch": 0.8712631295448424,
+ "grad_norm": 0.7940274414990574,
+ "learning_rate": 6.004738048467359e-06,
+ "loss": 0.379984712600708,
+ "num_input_tokens_seen": 566180672,
+ "step": 12940,
+ "train_runtime": 100101.3217,
+ "train_tokens_per_second": 5656.076
+ },
+ {
+ "epoch": 0.8715997845408026,
+ "grad_norm": 0.7344390394871104,
+ "learning_rate": 6.002147759962054e-06,
+ "loss": 0.37138075828552247,
+ "num_input_tokens_seen": 566401496,
+ "step": 12945,
+ "train_runtime": 100139.2781,
+ "train_tokens_per_second": 5656.137
+ },
+ {
+ "epoch": 0.8719364395367627,
+ "grad_norm": 0.7421475803173249,
+ "learning_rate": 5.999557191209323e-06,
+ "loss": 0.39212737083435056,
+ "num_input_tokens_seen": 566630216,
+ "step": 12950,
+ "train_runtime": 100175.2546,
+ "train_tokens_per_second": 5656.389
+ },
+ {
+ "epoch": 0.8722730945327228,
+ "grad_norm": 0.8003491943698586,
+ "learning_rate": 5.996966342933611e-06,
+ "loss": 0.33372251987457274,
+ "num_input_tokens_seen": 566848176,
+ "step": 12955,
+ "train_runtime": 100212.8634,
+ "train_tokens_per_second": 5656.441
+ },
+ {
+ "epoch": 0.872609749528683,
+ "grad_norm": 0.7450010892993638,
+ "learning_rate": 5.99437521585944e-06,
+ "loss": 0.37457475662231443,
+ "num_input_tokens_seen": 567056408,
+ "step": 12960,
+ "train_runtime": 100250.1105,
+ "train_tokens_per_second": 5656.417
+ },
+ {
+ "epoch": 0.8729464045246431,
+ "grad_norm": 0.7217432409540151,
+ "learning_rate": 5.99178381071141e-06,
+ "loss": 0.3829956531524658,
+ "num_input_tokens_seen": 567275168,
+ "step": 12965,
+ "train_runtime": 100287.1326,
+ "train_tokens_per_second": 5656.51
+ },
+ {
+ "epoch": 0.8732830595206033,
+ "grad_norm": 0.8055483175054227,
+ "learning_rate": 5.989192128214201e-06,
+ "loss": 0.36388020515441893,
+ "num_input_tokens_seen": 567493544,
+ "step": 12970,
+ "train_runtime": 100327.3129,
+ "train_tokens_per_second": 5656.421
+ },
+ {
+ "epoch": 0.8736197145165634,
+ "grad_norm": 0.7033510804995021,
+ "learning_rate": 5.986600169092568e-06,
+ "loss": 0.350065803527832,
+ "num_input_tokens_seen": 567708176,
+ "step": 12975,
+ "train_runtime": 100364.1125,
+ "train_tokens_per_second": 5656.486
+ },
+ {
+ "epoch": 0.8739563695125235,
+ "grad_norm": 0.7167335785321889,
+ "learning_rate": 5.984007934071343e-06,
+ "loss": 0.3742049217224121,
+ "num_input_tokens_seen": 567941160,
+ "step": 12980,
+ "train_runtime": 100407.0573,
+ "train_tokens_per_second": 5656.387
+ },
+ {
+ "epoch": 0.8742930245084837,
+ "grad_norm": 0.7554361554619513,
+ "learning_rate": 5.981415423875436e-06,
+ "loss": 0.36548290252685545,
+ "num_input_tokens_seen": 568166040,
+ "step": 12985,
+ "train_runtime": 100442.582,
+ "train_tokens_per_second": 5656.625
+ },
+ {
+ "epoch": 0.8746296795044438,
+ "grad_norm": 0.9815756056409435,
+ "learning_rate": 5.978822639229836e-06,
+ "loss": 0.3484132528305054,
+ "num_input_tokens_seen": 568367984,
+ "step": 12990,
+ "train_runtime": 100479.4893,
+ "train_tokens_per_second": 5656.557
+ },
+ {
+ "epoch": 0.874966334500404,
+ "grad_norm": 0.6864898867694621,
+ "learning_rate": 5.976229580859607e-06,
+ "loss": 0.36139798164367676,
+ "num_input_tokens_seen": 568596000,
+ "step": 12995,
+ "train_runtime": 100516.4986,
+ "train_tokens_per_second": 5656.743
+ },
+ {
+ "epoch": 0.8753029894963641,
+ "grad_norm": 0.7186739597846133,
+ "learning_rate": 5.9736362494898895e-06,
+ "loss": 0.3539880275726318,
+ "num_input_tokens_seen": 568814440,
+ "step": 13000,
+ "train_runtime": 100555.0904,
+ "train_tokens_per_second": 5656.744
+ },
+ {
+ "epoch": 0.8756396444923242,
+ "grad_norm": 0.6506103454572264,
+ "learning_rate": 5.971042645845897e-06,
+ "loss": 0.3666597843170166,
+ "num_input_tokens_seen": 569029824,
+ "step": 13005,
+ "train_runtime": 100595.3102,
+ "train_tokens_per_second": 5656.624
+ },
+ {
+ "epoch": 0.8759762994882844,
+ "grad_norm": 0.7706124062857171,
+ "learning_rate": 5.968448770652926e-06,
+ "loss": 0.36510615348815917,
+ "num_input_tokens_seen": 569265096,
+ "step": 13010,
+ "train_runtime": 100627.6528,
+ "train_tokens_per_second": 5657.144
+ },
+ {
+ "epoch": 0.8763129544842445,
+ "grad_norm": 0.7419816582789281,
+ "learning_rate": 5.965854624636345e-06,
+ "loss": 0.3825402736663818,
+ "num_input_tokens_seen": 569479936,
+ "step": 13015,
+ "train_runtime": 100669.5038,
+ "train_tokens_per_second": 5656.926
+ },
+ {
+ "epoch": 0.8766496094802047,
+ "grad_norm": 0.7717112571968371,
+ "learning_rate": 5.963260208521598e-06,
+ "loss": 0.3698084354400635,
+ "num_input_tokens_seen": 569691888,
+ "step": 13020,
+ "train_runtime": 100708.4391,
+ "train_tokens_per_second": 5656.844
+ },
+ {
+ "epoch": 0.8769862644761648,
+ "grad_norm": 0.6761324100161831,
+ "learning_rate": 5.960665523034205e-06,
+ "loss": 0.38090927600860597,
+ "num_input_tokens_seen": 569901576,
+ "step": 13025,
+ "train_runtime": 100749.3523,
+ "train_tokens_per_second": 5656.628
+ },
+ {
+ "epoch": 0.8773229194721249,
+ "grad_norm": 0.7359973742678347,
+ "learning_rate": 5.958070568899762e-06,
+ "loss": 0.33312511444091797,
+ "num_input_tokens_seen": 570123424,
+ "step": 13030,
+ "train_runtime": 100790.5444,
+ "train_tokens_per_second": 5656.517
+ },
+ {
+ "epoch": 0.8776595744680851,
+ "grad_norm": 1.2136155199409513,
+ "learning_rate": 5.955475346843939e-06,
+ "loss": 0.3555187225341797,
+ "num_input_tokens_seen": 570337248,
+ "step": 13035,
+ "train_runtime": 100824.0572,
+ "train_tokens_per_second": 5656.758
+ },
+ {
+ "epoch": 0.8779962294640452,
+ "grad_norm": 0.6546063737811157,
+ "learning_rate": 5.952879857592482e-06,
+ "loss": 0.3440871000289917,
+ "num_input_tokens_seen": 570549824,
+ "step": 13040,
+ "train_runtime": 100867.9537,
+ "train_tokens_per_second": 5656.403
+ },
+ {
+ "epoch": 0.8783328844600053,
+ "grad_norm": 0.6844892975959102,
+ "learning_rate": 5.950284101871212e-06,
+ "loss": 0.38202319145202634,
+ "num_input_tokens_seen": 570757152,
+ "step": 13045,
+ "train_runtime": 100906.3146,
+ "train_tokens_per_second": 5656.308
+ },
+ {
+ "epoch": 0.8786695394559655,
+ "grad_norm": 0.747172974034045,
+ "learning_rate": 5.947688080406023e-06,
+ "loss": 0.3463796615600586,
+ "num_input_tokens_seen": 570960928,
+ "step": 13050,
+ "train_runtime": 100940.5074,
+ "train_tokens_per_second": 5656.41
+ },
+ {
+ "epoch": 0.8790061944519256,
+ "grad_norm": 0.7561590864609555,
+ "learning_rate": 5.945091793922885e-06,
+ "loss": 0.3679834842681885,
+ "num_input_tokens_seen": 571177360,
+ "step": 13055,
+ "train_runtime": 100974.1104,
+ "train_tokens_per_second": 5656.671
+ },
+ {
+ "epoch": 0.8793428494478858,
+ "grad_norm": 0.6782284385723677,
+ "learning_rate": 5.942495243147839e-06,
+ "loss": 0.3929899215698242,
+ "num_input_tokens_seen": 571399192,
+ "step": 13060,
+ "train_runtime": 101005.759,
+ "train_tokens_per_second": 5657.095
+ },
+ {
+ "epoch": 0.8796795044438459,
+ "grad_norm": 0.7143338491446609,
+ "learning_rate": 5.939898428807004e-06,
+ "loss": 0.3815101146697998,
+ "num_input_tokens_seen": 571612808,
+ "step": 13065,
+ "train_runtime": 101043.3454,
+ "train_tokens_per_second": 5657.105
+ },
+ {
+ "epoch": 0.880016159439806,
+ "grad_norm": 0.8449030830666151,
+ "learning_rate": 5.937301351626571e-06,
+ "loss": 0.379590368270874,
+ "num_input_tokens_seen": 571824200,
+ "step": 13070,
+ "train_runtime": 101081.8232,
+ "train_tokens_per_second": 5657.043
+ },
+ {
+ "epoch": 0.8803528144357662,
+ "grad_norm": 0.6830378970909649,
+ "learning_rate": 5.934704012332803e-06,
+ "loss": 0.35865001678466796,
+ "num_input_tokens_seen": 572052552,
+ "step": 13075,
+ "train_runtime": 101122.8991,
+ "train_tokens_per_second": 5657.003
+ },
+ {
+ "epoch": 0.8806894694317263,
+ "grad_norm": 0.8010624909706995,
+ "learning_rate": 5.93210641165204e-06,
+ "loss": 0.4059410095214844,
+ "num_input_tokens_seen": 572274280,
+ "step": 13080,
+ "train_runtime": 101165.6483,
+ "train_tokens_per_second": 5656.804
+ },
+ {
+ "epoch": 0.8810261244276865,
+ "grad_norm": 0.7989521029331256,
+ "learning_rate": 5.92950855031069e-06,
+ "loss": 0.36967930793762205,
+ "num_input_tokens_seen": 572493808,
+ "step": 13085,
+ "train_runtime": 101202.3764,
+ "train_tokens_per_second": 5656.921
+ },
+ {
+ "epoch": 0.8813627794236466,
+ "grad_norm": 0.8258054867788488,
+ "learning_rate": 5.926910429035239e-06,
+ "loss": 0.36150131225585935,
+ "num_input_tokens_seen": 572706072,
+ "step": 13090,
+ "train_runtime": 101239.3964,
+ "train_tokens_per_second": 5656.949
+ },
+ {
+ "epoch": 0.8816994344196067,
+ "grad_norm": 0.7636365729370889,
+ "learning_rate": 5.924312048552241e-06,
+ "loss": 0.3525059223175049,
+ "num_input_tokens_seen": 572932096,
+ "step": 13095,
+ "train_runtime": 101273.4837,
+ "train_tokens_per_second": 5657.276
+ },
+ {
+ "epoch": 0.8820360894155669,
+ "grad_norm": 0.7191821908315644,
+ "learning_rate": 5.9217134095883265e-06,
+ "loss": 0.3641670227050781,
+ "num_input_tokens_seen": 573150384,
+ "step": 13100,
+ "train_runtime": 101314.4917,
+ "train_tokens_per_second": 5657.141
+ },
+ {
+ "epoch": 0.882372744411527,
+ "grad_norm": 0.7363436743411855,
+ "learning_rate": 5.919114512870197e-06,
+ "loss": 0.3603955268859863,
+ "num_input_tokens_seen": 573381144,
+ "step": 13105,
+ "train_runtime": 101359.314,
+ "train_tokens_per_second": 5656.916
+ },
+ {
+ "epoch": 0.8827093994074872,
+ "grad_norm": 0.7568233135879606,
+ "learning_rate": 5.916515359124623e-06,
+ "loss": 0.3425842046737671,
+ "num_input_tokens_seen": 573581832,
+ "step": 13110,
+ "train_runtime": 101399.4218,
+ "train_tokens_per_second": 5656.658
+ },
+ {
+ "epoch": 0.8830460544034473,
+ "grad_norm": 0.7215795683330004,
+ "learning_rate": 5.913915949078453e-06,
+ "loss": 0.36979331970214846,
+ "num_input_tokens_seen": 573801512,
+ "step": 13115,
+ "train_runtime": 101438.6834,
+ "train_tokens_per_second": 5656.634
+ },
+ {
+ "epoch": 0.8833827093994074,
+ "grad_norm": 0.7764516396496984,
+ "learning_rate": 5.911316283458601e-06,
+ "loss": 0.3616600513458252,
+ "num_input_tokens_seen": 574007544,
+ "step": 13120,
+ "train_runtime": 101478.4371,
+ "train_tokens_per_second": 5656.448
+ },
+ {
+ "epoch": 0.8837193643953676,
+ "grad_norm": 0.7391044514688772,
+ "learning_rate": 5.908716362992057e-06,
+ "loss": 0.36803760528564455,
+ "num_input_tokens_seen": 574215752,
+ "step": 13125,
+ "train_runtime": 101514.9279,
+ "train_tokens_per_second": 5656.466
+ },
+ {
+ "epoch": 0.8840560193913277,
+ "grad_norm": 0.6863603049824145,
+ "learning_rate": 5.90611618840588e-06,
+ "loss": 0.379679536819458,
+ "num_input_tokens_seen": 574434272,
+ "step": 13130,
+ "train_runtime": 101554.971,
+ "train_tokens_per_second": 5656.388
+ },
+ {
+ "epoch": 0.8843926743872879,
+ "grad_norm": 0.9353489354148398,
+ "learning_rate": 5.903515760427198e-06,
+ "loss": 0.4000962257385254,
+ "num_input_tokens_seen": 574653864,
+ "step": 13135,
+ "train_runtime": 101588.2234,
+ "train_tokens_per_second": 5656.698
+ },
+ {
+ "epoch": 0.884729329383248,
+ "grad_norm": 0.732273202002582,
+ "learning_rate": 5.900915079783218e-06,
+ "loss": 0.3558906316757202,
+ "num_input_tokens_seen": 574887384,
+ "step": 13140,
+ "train_runtime": 101622.3485,
+ "train_tokens_per_second": 5657.096
+ },
+ {
+ "epoch": 0.8850659843792081,
+ "grad_norm": 0.7460917342917756,
+ "learning_rate": 5.898314147201207e-06,
+ "loss": 0.3416097640991211,
+ "num_input_tokens_seen": 575101544,
+ "step": 13145,
+ "train_runtime": 101664.6483,
+ "train_tokens_per_second": 5656.849
+ },
+ {
+ "epoch": 0.8854026393751683,
+ "grad_norm": 1.0430878064810838,
+ "learning_rate": 5.895712963408511e-06,
+ "loss": 0.3921904325485229,
+ "num_input_tokens_seen": 575309032,
+ "step": 13150,
+ "train_runtime": 101704.7114,
+ "train_tokens_per_second": 5656.661
+ },
+ {
+ "epoch": 0.8857392943711284,
+ "grad_norm": 0.7135808350871835,
+ "learning_rate": 5.893111529132539e-06,
+ "loss": 0.38542513847351073,
+ "num_input_tokens_seen": 575518184,
+ "step": 13155,
+ "train_runtime": 101742.2738,
+ "train_tokens_per_second": 5656.628
+ },
+ {
+ "epoch": 0.8860759493670886,
+ "grad_norm": 0.7661970012022706,
+ "learning_rate": 5.8905098451007755e-06,
+ "loss": 0.37914443016052246,
+ "num_input_tokens_seen": 575721848,
+ "step": 13160,
+ "train_runtime": 101781.9439,
+ "train_tokens_per_second": 5656.424
+ },
+ {
+ "epoch": 0.8864126043630487,
+ "grad_norm": 0.7203329040983677,
+ "learning_rate": 5.887907912040774e-06,
+ "loss": 0.3751526355743408,
+ "num_input_tokens_seen": 575946104,
+ "step": 13165,
+ "train_runtime": 101814.7009,
+ "train_tokens_per_second": 5656.807
+ },
+ {
+ "epoch": 0.8867492593590088,
+ "grad_norm": 0.7424424433197104,
+ "learning_rate": 5.885305730680158e-06,
+ "loss": 0.3788419723510742,
+ "num_input_tokens_seen": 576173784,
+ "step": 13170,
+ "train_runtime": 101847.3914,
+ "train_tokens_per_second": 5657.227
+ },
+ {
+ "epoch": 0.887085914354969,
+ "grad_norm": 0.7289665105864628,
+ "learning_rate": 5.882703301746616e-06,
+ "loss": 0.36290099620819094,
+ "num_input_tokens_seen": 576389120,
+ "step": 13175,
+ "train_runtime": 101889.8836,
+ "train_tokens_per_second": 5656.981
+ },
+ {
+ "epoch": 0.8874225693509291,
+ "grad_norm": 0.7262294984371385,
+ "learning_rate": 5.88010062596791e-06,
+ "loss": 0.35980806350708006,
+ "num_input_tokens_seen": 576598016,
+ "step": 13180,
+ "train_runtime": 101930.8807,
+ "train_tokens_per_second": 5656.755
+ },
+ {
+ "epoch": 0.8877592243468893,
+ "grad_norm": 0.601051210488872,
+ "learning_rate": 5.877497704071871e-06,
+ "loss": 0.3475220203399658,
+ "num_input_tokens_seen": 576831392,
+ "step": 13185,
+ "train_runtime": 101969.0556,
+ "train_tokens_per_second": 5656.926
+ },
+ {
+ "epoch": 0.8880958793428494,
+ "grad_norm": 0.7662456138801742,
+ "learning_rate": 5.874894536786396e-06,
+ "loss": 0.3340578079223633,
+ "num_input_tokens_seen": 577044568,
+ "step": 13190,
+ "train_runtime": 102009.2868,
+ "train_tokens_per_second": 5656.785
+ },
+ {
+ "epoch": 0.8884325343388096,
+ "grad_norm": 0.7991035565617537,
+ "learning_rate": 5.8722911248394555e-06,
+ "loss": 0.33777599334716796,
+ "num_input_tokens_seen": 577243936,
+ "step": 13195,
+ "train_runtime": 102050.722,
+ "train_tokens_per_second": 5656.441
+ },
+ {
+ "epoch": 0.8887691893347698,
+ "grad_norm": 0.7706395477682242,
+ "learning_rate": 5.8696874689590824e-06,
+ "loss": 0.3437623977661133,
+ "num_input_tokens_seen": 577453888,
+ "step": 13200,
+ "train_runtime": 102083.4149,
+ "train_tokens_per_second": 5656.687
+ },
+ {
+ "epoch": 0.8891058443307299,
+ "grad_norm": 0.8134775525528564,
+ "learning_rate": 5.867083569873383e-06,
+ "loss": 0.3634350776672363,
+ "num_input_tokens_seen": 577679688,
+ "step": 13205,
+ "train_runtime": 102117.5448,
+ "train_tokens_per_second": 5657.007
+ },
+ {
+ "epoch": 0.8894424993266901,
+ "grad_norm": 0.7673260905685215,
+ "learning_rate": 5.8644794283105255e-06,
+ "loss": 0.36774277687072754,
+ "num_input_tokens_seen": 577900960,
+ "step": 13210,
+ "train_runtime": 102155.9766,
+ "train_tokens_per_second": 5657.045
+ },
+ {
+ "epoch": 0.8897791543226502,
+ "grad_norm": 0.7882719202387701,
+ "learning_rate": 5.861875044998755e-06,
+ "loss": 0.40239391326904295,
+ "num_input_tokens_seen": 578125704,
+ "step": 13215,
+ "train_runtime": 102190.4444,
+ "train_tokens_per_second": 5657.336
+ },
+ {
+ "epoch": 0.8901158093186103,
+ "grad_norm": 0.7563037927913291,
+ "learning_rate": 5.859270420666377e-06,
+ "loss": 0.36860146522521975,
+ "num_input_tokens_seen": 578343816,
+ "step": 13220,
+ "train_runtime": 102228.533,
+ "train_tokens_per_second": 5657.362
+ },
+ {
+ "epoch": 0.8904524643145705,
+ "grad_norm": 0.7756762347173053,
+ "learning_rate": 5.856665556041764e-06,
+ "loss": 0.3481595993041992,
+ "num_input_tokens_seen": 578548144,
+ "step": 13225,
+ "train_runtime": 102265.8667,
+ "train_tokens_per_second": 5657.295
+ },
+ {
+ "epoch": 0.8907891193105306,
+ "grad_norm": 0.7183886665375777,
+ "learning_rate": 5.8540604518533605e-06,
+ "loss": 0.3859154224395752,
+ "num_input_tokens_seen": 578772776,
+ "step": 13230,
+ "train_runtime": 102299.6045,
+ "train_tokens_per_second": 5657.625
+ },
+ {
+ "epoch": 0.8911257743064908,
+ "grad_norm": 1.4271805656006054,
+ "learning_rate": 5.851455108829675e-06,
+ "loss": 0.34519567489624026,
+ "num_input_tokens_seen": 578981024,
+ "step": 13235,
+ "train_runtime": 102340.8563,
+ "train_tokens_per_second": 5657.379
+ },
+ {
+ "epoch": 0.8914624293024509,
+ "grad_norm": 0.6789149460339803,
+ "learning_rate": 5.848849527699283e-06,
+ "loss": 0.3698782205581665,
+ "num_input_tokens_seen": 579201768,
+ "step": 13240,
+ "train_runtime": 102379.4065,
+ "train_tokens_per_second": 5657.405
+ },
+ {
+ "epoch": 0.891799084298411,
+ "grad_norm": 0.7639170132777864,
+ "learning_rate": 5.846243709190827e-06,
+ "loss": 0.35757324695587156,
+ "num_input_tokens_seen": 579419880,
+ "step": 13245,
+ "train_runtime": 102417.7033,
+ "train_tokens_per_second": 5657.419
+ },
+ {
+ "epoch": 0.8921357392943712,
+ "grad_norm": 0.8837562529973417,
+ "learning_rate": 5.843637654033019e-06,
+ "loss": 0.37553791999816893,
+ "num_input_tokens_seen": 579620760,
+ "step": 13250,
+ "train_runtime": 102459.8571,
+ "train_tokens_per_second": 5657.052
+ },
+ {
+ "epoch": 0.8924723942903313,
+ "grad_norm": 0.676305723344451,
+ "learning_rate": 5.841031362954629e-06,
+ "loss": 0.39575533866882323,
+ "num_input_tokens_seen": 579849048,
+ "step": 13255,
+ "train_runtime": 102495.0631,
+ "train_tokens_per_second": 5657.336
+ },
+ {
+ "epoch": 0.8928090492862915,
+ "grad_norm": 0.6628658100615732,
+ "learning_rate": 5.8384248366845e-06,
+ "loss": 0.37047085762023924,
+ "num_input_tokens_seen": 580077352,
+ "step": 13260,
+ "train_runtime": 102540.1881,
+ "train_tokens_per_second": 5657.073
+ },
+ {
+ "epoch": 0.8931457042822516,
+ "grad_norm": 0.6825809402512039,
+ "learning_rate": 5.8358180759515396e-06,
+ "loss": 0.37846667766571046,
+ "num_input_tokens_seen": 580288200,
+ "step": 13265,
+ "train_runtime": 102577.1371,
+ "train_tokens_per_second": 5657.091
+ },
+ {
+ "epoch": 0.8934823592782117,
+ "grad_norm": 0.7471824938587698,
+ "learning_rate": 5.833211081484719e-06,
+ "loss": 0.37510037422180176,
+ "num_input_tokens_seen": 580506904,
+ "step": 13270,
+ "train_runtime": 102618.782,
+ "train_tokens_per_second": 5656.926
+ },
+ {
+ "epoch": 0.8938190142741719,
+ "grad_norm": 0.641717053888179,
+ "learning_rate": 5.830603854013077e-06,
+ "loss": 0.3673901081085205,
+ "num_input_tokens_seen": 580723408,
+ "step": 13275,
+ "train_runtime": 102660.3376,
+ "train_tokens_per_second": 5656.746
+ },
+ {
+ "epoch": 0.894155669270132,
+ "grad_norm": 0.8523312033525736,
+ "learning_rate": 5.827996394265716e-06,
+ "loss": 0.35067110061645507,
+ "num_input_tokens_seen": 580935112,
+ "step": 13280,
+ "train_runtime": 102701.6639,
+ "train_tokens_per_second": 5656.531
+ },
+ {
+ "epoch": 0.8944923242660922,
+ "grad_norm": 0.7794670905441262,
+ "learning_rate": 5.825388702971802e-06,
+ "loss": 0.35003342628479006,
+ "num_input_tokens_seen": 581151304,
+ "step": 13285,
+ "train_runtime": 102733.2999,
+ "train_tokens_per_second": 5656.893
+ },
+ {
+ "epoch": 0.8948289792620523,
+ "grad_norm": 0.7697508084044369,
+ "learning_rate": 5.822780780860571e-06,
+ "loss": 0.36042060852050783,
+ "num_input_tokens_seen": 581367952,
+ "step": 13290,
+ "train_runtime": 102770.8878,
+ "train_tokens_per_second": 5656.932
+ },
+ {
+ "epoch": 0.8951656342580124,
+ "grad_norm": 0.7094962387963322,
+ "learning_rate": 5.820172628661317e-06,
+ "loss": 0.3688685178756714,
+ "num_input_tokens_seen": 581603552,
+ "step": 13295,
+ "train_runtime": 102804.4905,
+ "train_tokens_per_second": 5657.375
+ },
+ {
+ "epoch": 0.8955022892539726,
+ "grad_norm": 0.7277970714037896,
+ "learning_rate": 5.8175642471034045e-06,
+ "loss": 0.34978251457214354,
+ "num_input_tokens_seen": 581823928,
+ "step": 13300,
+ "train_runtime": 102838.1084,
+ "train_tokens_per_second": 5657.669
+ },
+ {
+ "epoch": 0.8958389442499327,
+ "grad_norm": 0.7417598590201532,
+ "learning_rate": 5.8149556369162565e-06,
+ "loss": 0.3728696346282959,
+ "num_input_tokens_seen": 582031800,
+ "step": 13305,
+ "train_runtime": 102877.0099,
+ "train_tokens_per_second": 5657.55
+ },
+ {
+ "epoch": 0.8961755992458929,
+ "grad_norm": 0.7192634087406454,
+ "learning_rate": 5.812346798829361e-06,
+ "loss": 0.37227132320404055,
+ "num_input_tokens_seen": 582257072,
+ "step": 13310,
+ "train_runtime": 102918.7234,
+ "train_tokens_per_second": 5657.446
+ },
+ {
+ "epoch": 0.896512254241853,
+ "grad_norm": 0.7396821495476227,
+ "learning_rate": 5.809737733572276e-06,
+ "loss": 0.35695157051086424,
+ "num_input_tokens_seen": 582470280,
+ "step": 13315,
+ "train_runtime": 102957.1094,
+ "train_tokens_per_second": 5657.407
+ },
+ {
+ "epoch": 0.8968489092378131,
+ "grad_norm": 0.8286698107403867,
+ "learning_rate": 5.8071284418746166e-06,
+ "loss": 0.3608556747436523,
+ "num_input_tokens_seen": 582689968,
+ "step": 13320,
+ "train_runtime": 102997.8183,
+ "train_tokens_per_second": 5657.304
+ },
+ {
+ "epoch": 0.8971855642337733,
+ "grad_norm": 0.7429942677291922,
+ "learning_rate": 5.80451892446606e-06,
+ "loss": 0.38565864562988283,
+ "num_input_tokens_seen": 582917944,
+ "step": 13325,
+ "train_runtime": 103036.6727,
+ "train_tokens_per_second": 5657.383
+ },
+ {
+ "epoch": 0.8975222192297334,
+ "grad_norm": 0.6807484219365846,
+ "learning_rate": 5.801909182076354e-06,
+ "loss": 0.3657494068145752,
+ "num_input_tokens_seen": 583143528,
+ "step": 13330,
+ "train_runtime": 103071.4117,
+ "train_tokens_per_second": 5657.665
+ },
+ {
+ "epoch": 0.8978588742256935,
+ "grad_norm": 0.7911812244430602,
+ "learning_rate": 5.7992992154353e-06,
+ "loss": 0.3965157985687256,
+ "num_input_tokens_seen": 583360120,
+ "step": 13335,
+ "train_runtime": 103106.3833,
+ "train_tokens_per_second": 5657.847
+ },
+ {
+ "epoch": 0.8981955292216537,
+ "grad_norm": 0.7134344977635636,
+ "learning_rate": 5.79668902527277e-06,
+ "loss": 0.3728632926940918,
+ "num_input_tokens_seen": 583576392,
+ "step": 13340,
+ "train_runtime": 103143.3779,
+ "train_tokens_per_second": 5657.914
+ },
+ {
+ "epoch": 0.8985321842176138,
+ "grad_norm": 0.7068010881578768,
+ "learning_rate": 5.794078612318695e-06,
+ "loss": 0.36152513027191163,
+ "num_input_tokens_seen": 583793888,
+ "step": 13345,
+ "train_runtime": 103187.6757,
+ "train_tokens_per_second": 5657.593
+ },
+ {
+ "epoch": 0.898868839213574,
+ "grad_norm": 1.018985024468395,
+ "learning_rate": 5.791467977303068e-06,
+ "loss": 0.3539185285568237,
+ "num_input_tokens_seen": 584007880,
+ "step": 13350,
+ "train_runtime": 103221.6694,
+ "train_tokens_per_second": 5657.803
+ },
+ {
+ "epoch": 0.8992054942095341,
+ "grad_norm": 0.8787085802082749,
+ "learning_rate": 5.788857120955945e-06,
+ "loss": 0.3753284692764282,
+ "num_input_tokens_seen": 584222720,
+ "step": 13355,
+ "train_runtime": 103259.3517,
+ "train_tokens_per_second": 5657.819
+ },
+ {
+ "epoch": 0.8995421492054942,
+ "grad_norm": 0.7526792175486189,
+ "learning_rate": 5.786246044007443e-06,
+ "loss": 0.3408811092376709,
+ "num_input_tokens_seen": 584430032,
+ "step": 13360,
+ "train_runtime": 103303.2518,
+ "train_tokens_per_second": 5657.421
+ },
+ {
+ "epoch": 0.8998788042014544,
+ "grad_norm": 0.7042471293534047,
+ "learning_rate": 5.783634747187743e-06,
+ "loss": 0.38433132171630857,
+ "num_input_tokens_seen": 584657272,
+ "step": 13365,
+ "train_runtime": 103336.5354,
+ "train_tokens_per_second": 5657.798
+ },
+ {
+ "epoch": 0.9002154591974145,
+ "grad_norm": 0.7283278180613847,
+ "learning_rate": 5.781023231227084e-06,
+ "loss": 0.3663713693618774,
+ "num_input_tokens_seen": 584873064,
+ "step": 13370,
+ "train_runtime": 103383.5959,
+ "train_tokens_per_second": 5657.31
+ },
+ {
+ "epoch": 0.9005521141933747,
+ "grad_norm": 0.6987374788445736,
+ "learning_rate": 5.778411496855769e-06,
+ "loss": 0.3779282093048096,
+ "num_input_tokens_seen": 585095664,
+ "step": 13375,
+ "train_runtime": 103422.1497,
+ "train_tokens_per_second": 5657.354
+ },
+ {
+ "epoch": 0.9008887691893348,
+ "grad_norm": 0.7973255027303662,
+ "learning_rate": 5.775799544804161e-06,
+ "loss": 0.3911623477935791,
+ "num_input_tokens_seen": 585323224,
+ "step": 13380,
+ "train_runtime": 103458.2471,
+ "train_tokens_per_second": 5657.579
+ },
+ {
+ "epoch": 0.9012254241852949,
+ "grad_norm": 0.7376867678943736,
+ "learning_rate": 5.7731873758026845e-06,
+ "loss": 0.3567417860031128,
+ "num_input_tokens_seen": 585539800,
+ "step": 13385,
+ "train_runtime": 103499.6202,
+ "train_tokens_per_second": 5657.41
+ },
+ {
+ "epoch": 0.9015620791812551,
+ "grad_norm": 0.8164983344754712,
+ "learning_rate": 5.770574990581823e-06,
+ "loss": 0.3617391109466553,
+ "num_input_tokens_seen": 585756160,
+ "step": 13390,
+ "train_runtime": 103538.5541,
+ "train_tokens_per_second": 5657.372
+ },
+ {
+ "epoch": 0.9018987341772152,
+ "grad_norm": 0.7386008524112068,
+ "learning_rate": 5.767962389872124e-06,
+ "loss": 0.34462101459503175,
+ "num_input_tokens_seen": 585972312,
+ "step": 13395,
+ "train_runtime": 103578.7054,
+ "train_tokens_per_second": 5657.266
+ },
+ {
+ "epoch": 0.9022353891731754,
+ "grad_norm": 0.7749458765170661,
+ "learning_rate": 5.765349574404191e-06,
+ "loss": 0.3628375053405762,
+ "num_input_tokens_seen": 586175624,
+ "step": 13400,
+ "train_runtime": 103611.757,
+ "train_tokens_per_second": 5657.424
+ },
+ {
+ "epoch": 0.9025720441691355,
+ "grad_norm": 0.7292901746189114,
+ "learning_rate": 5.7627365449086905e-06,
+ "loss": 0.364549446105957,
+ "num_input_tokens_seen": 586370072,
+ "step": 13405,
+ "train_runtime": 103645.7307,
+ "train_tokens_per_second": 5657.445
+ },
+ {
+ "epoch": 0.9029086991650956,
+ "grad_norm": 0.7620757507465177,
+ "learning_rate": 5.760123302116345e-06,
+ "loss": 0.3901118040084839,
+ "num_input_tokens_seen": 586593560,
+ "step": 13410,
+ "train_runtime": 103677.957,
+ "train_tokens_per_second": 5657.843
+ },
+ {
+ "epoch": 0.9032453541610558,
+ "grad_norm": 0.7598225601483857,
+ "learning_rate": 5.757509846757945e-06,
+ "loss": 0.36999359130859377,
+ "num_input_tokens_seen": 586819408,
+ "step": 13415,
+ "train_runtime": 103716.3963,
+ "train_tokens_per_second": 5657.923
+ },
+ {
+ "epoch": 0.9035820091570159,
+ "grad_norm": 1.147309642657896,
+ "learning_rate": 5.75489617956433e-06,
+ "loss": 0.3971453428268433,
+ "num_input_tokens_seen": 587040712,
+ "step": 13420,
+ "train_runtime": 103762.395,
+ "train_tokens_per_second": 5657.548
+ },
+ {
+ "epoch": 0.9039186641529761,
+ "grad_norm": 0.6772540219988885,
+ "learning_rate": 5.752282301266406e-06,
+ "loss": 0.37274601459503176,
+ "num_input_tokens_seen": 587260024,
+ "step": 13425,
+ "train_runtime": 103799.3896,
+ "train_tokens_per_second": 5657.644
+ },
+ {
+ "epoch": 0.9042553191489362,
+ "grad_norm": 0.6865687048122345,
+ "learning_rate": 5.749668212595136e-06,
+ "loss": 0.36958420276641846,
+ "num_input_tokens_seen": 587475152,
+ "step": 13430,
+ "train_runtime": 103838.2234,
+ "train_tokens_per_second": 5657.6
+ },
+ {
+ "epoch": 0.9045919741448963,
+ "grad_norm": 0.7480660958831125,
+ "learning_rate": 5.747053914281539e-06,
+ "loss": 0.36184403896331785,
+ "num_input_tokens_seen": 587679656,
+ "step": 13435,
+ "train_runtime": 103881.14,
+ "train_tokens_per_second": 5657.231
+ },
+ {
+ "epoch": 0.9049286291408565,
+ "grad_norm": 0.7323661070573674,
+ "learning_rate": 5.744439407056699e-06,
+ "loss": 0.3511162757873535,
+ "num_input_tokens_seen": 587888648,
+ "step": 13440,
+ "train_runtime": 103918.9245,
+ "train_tokens_per_second": 5657.186
+ },
+ {
+ "epoch": 0.9052652841368166,
+ "grad_norm": 0.6204712137626076,
+ "learning_rate": 5.741824691651752e-06,
+ "loss": 0.3408759832382202,
+ "num_input_tokens_seen": 588114264,
+ "step": 13445,
+ "train_runtime": 103954.7773,
+ "train_tokens_per_second": 5657.405
+ },
+ {
+ "epoch": 0.9056019391327768,
+ "grad_norm": 0.8116090187953918,
+ "learning_rate": 5.739209768797896e-06,
+ "loss": 0.36322770118713377,
+ "num_input_tokens_seen": 588322904,
+ "step": 13450,
+ "train_runtime": 103991.6356,
+ "train_tokens_per_second": 5657.406
+ },
+ {
+ "epoch": 0.9059385941287369,
+ "grad_norm": 0.7146086921110422,
+ "learning_rate": 5.736594639226385e-06,
+ "loss": 0.32691378593444825,
+ "num_input_tokens_seen": 588542624,
+ "step": 13455,
+ "train_runtime": 104026.9412,
+ "train_tokens_per_second": 5657.598
+ },
+ {
+ "epoch": 0.906275249124697,
+ "grad_norm": 0.6758364668603787,
+ "learning_rate": 5.7339793036685306e-06,
+ "loss": 0.3772242546081543,
+ "num_input_tokens_seen": 588766448,
+ "step": 13460,
+ "train_runtime": 104068.0394,
+ "train_tokens_per_second": 5657.515
+ },
+ {
+ "epoch": 0.9066119041206572,
+ "grad_norm": 0.7332497573577879,
+ "learning_rate": 5.731363762855706e-06,
+ "loss": 0.3471216678619385,
+ "num_input_tokens_seen": 588981896,
+ "step": 13465,
+ "train_runtime": 104109.0854,
+ "train_tokens_per_second": 5657.353
+ },
+ {
+ "epoch": 0.9069485591166173,
+ "grad_norm": 0.8052750996260324,
+ "learning_rate": 5.728748017519337e-06,
+ "loss": 0.3998070001602173,
+ "num_input_tokens_seen": 589199016,
+ "step": 13470,
+ "train_runtime": 104146.4665,
+ "train_tokens_per_second": 5657.408
+ },
+ {
+ "epoch": 0.9072852141125775,
+ "grad_norm": 0.7072206601280142,
+ "learning_rate": 5.726132068390908e-06,
+ "loss": 0.3687425136566162,
+ "num_input_tokens_seen": 589402952,
+ "step": 13475,
+ "train_runtime": 104180.5112,
+ "train_tokens_per_second": 5657.516
+ },
+ {
+ "epoch": 0.9076218691085376,
+ "grad_norm": 0.7386177069645364,
+ "learning_rate": 5.723515916201962e-06,
+ "loss": 0.4100666046142578,
+ "num_input_tokens_seen": 589630440,
+ "step": 13480,
+ "train_runtime": 104218.7952,
+ "train_tokens_per_second": 5657.621
+ },
+ {
+ "epoch": 0.9079585241044977,
+ "grad_norm": 0.6255645042824444,
+ "learning_rate": 5.720899561684098e-06,
+ "loss": 0.3434640645980835,
+ "num_input_tokens_seen": 589849368,
+ "step": 13485,
+ "train_runtime": 104262.0828,
+ "train_tokens_per_second": 5657.372
+ },
+ {
+ "epoch": 0.9082951791004579,
+ "grad_norm": 0.7488072842696437,
+ "learning_rate": 5.71828300556897e-06,
+ "loss": 0.36190190315246584,
+ "num_input_tokens_seen": 590075552,
+ "step": 13490,
+ "train_runtime": 104304.1746,
+ "train_tokens_per_second": 5657.257
+ },
+ {
+ "epoch": 0.908631834096418,
+ "grad_norm": 0.595597879983287,
+ "learning_rate": 5.715666248588289e-06,
+ "loss": 0.3661900520324707,
+ "num_input_tokens_seen": 590300848,
+ "step": 13495,
+ "train_runtime": 104350.0273,
+ "train_tokens_per_second": 5656.93
+ },
+ {
+ "epoch": 0.9089684890923782,
+ "grad_norm": 0.7843852676552852,
+ "learning_rate": 5.713049291473825e-06,
+ "loss": 0.39053096771240237,
+ "num_input_tokens_seen": 590529208,
+ "step": 13500,
+ "train_runtime": 104390.3684,
+ "train_tokens_per_second": 5656.932
+ },
+ {
+ "epoch": 0.9093051440883383,
+ "grad_norm": 0.7754635800602863,
+ "learning_rate": 5.710432134957399e-06,
+ "loss": 0.3421627998352051,
+ "num_input_tokens_seen": 590729424,
+ "step": 13505,
+ "train_runtime": 104437.8595,
+ "train_tokens_per_second": 5656.277
+ },
+ {
+ "epoch": 0.9096417990842984,
+ "grad_norm": 0.7329858480957279,
+ "learning_rate": 5.707814779770892e-06,
+ "loss": 0.34261083602905273,
+ "num_input_tokens_seen": 590940928,
+ "step": 13510,
+ "train_runtime": 104484.5885,
+ "train_tokens_per_second": 5655.771
+ },
+ {
+ "epoch": 0.9099784540802586,
+ "grad_norm": 0.7345589839622834,
+ "learning_rate": 5.7051972266462395e-06,
+ "loss": 0.35787391662597656,
+ "num_input_tokens_seen": 591158192,
+ "step": 13515,
+ "train_runtime": 104521.8258,
+ "train_tokens_per_second": 5655.835
+ },
+ {
+ "epoch": 0.9103151090762187,
+ "grad_norm": 0.8008449716401149,
+ "learning_rate": 5.702579476315431e-06,
+ "loss": 0.3310979127883911,
+ "num_input_tokens_seen": 591376656,
+ "step": 13520,
+ "train_runtime": 104566.7223,
+ "train_tokens_per_second": 5655.496
+ },
+ {
+ "epoch": 0.9106517640721788,
+ "grad_norm": 0.7423340401003228,
+ "learning_rate": 5.699961529510513e-06,
+ "loss": 0.3578958988189697,
+ "num_input_tokens_seen": 591581336,
+ "step": 13525,
+ "train_runtime": 104607.1051,
+ "train_tokens_per_second": 5655.269
+ },
+ {
+ "epoch": 0.910988419068139,
+ "grad_norm": 0.7857402111034575,
+ "learning_rate": 5.697343386963586e-06,
+ "loss": 0.3766601800918579,
+ "num_input_tokens_seen": 591797696,
+ "step": 13530,
+ "train_runtime": 104651.9901,
+ "train_tokens_per_second": 5654.911
+ },
+ {
+ "epoch": 0.9113250740640991,
+ "grad_norm": 0.6657292352842593,
+ "learning_rate": 5.694725049406801e-06,
+ "loss": 0.3710474967956543,
+ "num_input_tokens_seen": 592014888,
+ "step": 13535,
+ "train_runtime": 104697.3063,
+ "train_tokens_per_second": 5654.538
+ },
+ {
+ "epoch": 0.9116617290600593,
+ "grad_norm": 0.6966081276914712,
+ "learning_rate": 5.692106517572373e-06,
+ "loss": 0.38451037406921384,
+ "num_input_tokens_seen": 592229672,
+ "step": 13540,
+ "train_runtime": 104731.9456,
+ "train_tokens_per_second": 5654.719
+ },
+ {
+ "epoch": 0.9119983840560194,
+ "grad_norm": 0.7724805146942423,
+ "learning_rate": 5.689487792192566e-06,
+ "loss": 0.3990262746810913,
+ "num_input_tokens_seen": 592451696,
+ "step": 13545,
+ "train_runtime": 104773.5682,
+ "train_tokens_per_second": 5654.591
+ },
+ {
+ "epoch": 0.9123350390519795,
+ "grad_norm": 0.7342026376072939,
+ "learning_rate": 5.686868873999696e-06,
+ "loss": 0.3797611236572266,
+ "num_input_tokens_seen": 592668960,
+ "step": 13550,
+ "train_runtime": 104816.4545,
+ "train_tokens_per_second": 5654.35
+ },
+ {
+ "epoch": 0.9126716940479397,
+ "grad_norm": 0.666648695750785,
+ "learning_rate": 5.684249763726135e-06,
+ "loss": 0.3764026641845703,
+ "num_input_tokens_seen": 592890144,
+ "step": 13555,
+ "train_runtime": 104853.017,
+ "train_tokens_per_second": 5654.488
+ },
+ {
+ "epoch": 0.9130083490438998,
+ "grad_norm": 0.796183302090131,
+ "learning_rate": 5.681630462104308e-06,
+ "loss": 0.3508880615234375,
+ "num_input_tokens_seen": 593092200,
+ "step": 13560,
+ "train_runtime": 104890.6314,
+ "train_tokens_per_second": 5654.387
+ },
+ {
+ "epoch": 0.91334500403986,
+ "grad_norm": 0.6941135923026022,
+ "learning_rate": 5.679010969866697e-06,
+ "loss": 0.3628666877746582,
+ "num_input_tokens_seen": 593304400,
+ "step": 13565,
+ "train_runtime": 104927.8794,
+ "train_tokens_per_second": 5654.402
+ },
+ {
+ "epoch": 0.9136816590358201,
+ "grad_norm": 0.7543532537354826,
+ "learning_rate": 5.676391287745835e-06,
+ "loss": 0.35071396827697754,
+ "num_input_tokens_seen": 593537816,
+ "step": 13570,
+ "train_runtime": 104959.059,
+ "train_tokens_per_second": 5654.946
+ },
+ {
+ "epoch": 0.9140183140317802,
+ "grad_norm": 0.6731321122510332,
+ "learning_rate": 5.673771416474305e-06,
+ "loss": 0.3439845085144043,
+ "num_input_tokens_seen": 593761520,
+ "step": 13575,
+ "train_runtime": 104997.0967,
+ "train_tokens_per_second": 5655.028
+ },
+ {
+ "epoch": 0.9143549690277404,
+ "grad_norm": 0.7587431284561748,
+ "learning_rate": 5.671151356784747e-06,
+ "loss": 0.3460673809051514,
+ "num_input_tokens_seen": 593970768,
+ "step": 13580,
+ "train_runtime": 105034.4592,
+ "train_tokens_per_second": 5655.009
+ },
+ {
+ "epoch": 0.9146916240237005,
+ "grad_norm": 0.6614026194133914,
+ "learning_rate": 5.668531109409851e-06,
+ "loss": 0.3715926170349121,
+ "num_input_tokens_seen": 594193064,
+ "step": 13585,
+ "train_runtime": 105071.7331,
+ "train_tokens_per_second": 5655.118
+ },
+ {
+ "epoch": 0.9150282790196607,
+ "grad_norm": 0.7852840297073502,
+ "learning_rate": 5.665910675082362e-06,
+ "loss": 0.36374545097351074,
+ "num_input_tokens_seen": 594409288,
+ "step": 13590,
+ "train_runtime": 105109.8348,
+ "train_tokens_per_second": 5655.125
+ },
+ {
+ "epoch": 0.9153649340156208,
+ "grad_norm": 0.7103415655589641,
+ "learning_rate": 5.663290054535078e-06,
+ "loss": 0.34845700263977053,
+ "num_input_tokens_seen": 594620336,
+ "step": 13595,
+ "train_runtime": 105150.7645,
+ "train_tokens_per_second": 5654.931
+ },
+ {
+ "epoch": 0.9157015890115809,
+ "grad_norm": 0.8295807713349123,
+ "learning_rate": 5.660669248500846e-06,
+ "loss": 0.39510331153869627,
+ "num_input_tokens_seen": 594838312,
+ "step": 13600,
+ "train_runtime": 105189.0392,
+ "train_tokens_per_second": 5654.946
+ },
+ {
+ "epoch": 0.9160382440075411,
+ "grad_norm": 0.7158029671235298,
+ "learning_rate": 5.658048257712563e-06,
+ "loss": 0.3466895818710327,
+ "num_input_tokens_seen": 595060816,
+ "step": 13605,
+ "train_runtime": 105234.3634,
+ "train_tokens_per_second": 5654.625
+ },
+ {
+ "epoch": 0.9163748990035012,
+ "grad_norm": 0.7282803359887597,
+ "learning_rate": 5.655427082903184e-06,
+ "loss": 0.3633090019226074,
+ "num_input_tokens_seen": 595279584,
+ "step": 13610,
+ "train_runtime": 105277.1838,
+ "train_tokens_per_second": 5654.403
+ },
+ {
+ "epoch": 0.9167115539994614,
+ "grad_norm": 0.6738750845313296,
+ "learning_rate": 5.652805724805711e-06,
+ "loss": 0.35727887153625487,
+ "num_input_tokens_seen": 595499192,
+ "step": 13615,
+ "train_runtime": 105314.5327,
+ "train_tokens_per_second": 5654.483
+ },
+ {
+ "epoch": 0.9170482089954215,
+ "grad_norm": 0.8264756955958018,
+ "learning_rate": 5.650184184153199e-06,
+ "loss": 0.3687946319580078,
+ "num_input_tokens_seen": 595723264,
+ "step": 13620,
+ "train_runtime": 105352.1128,
+ "train_tokens_per_second": 5654.592
+ },
+ {
+ "epoch": 0.9173848639913816,
+ "grad_norm": 0.7363635027786596,
+ "learning_rate": 5.647562461678754e-06,
+ "loss": 0.39588251113891604,
+ "num_input_tokens_seen": 595947688,
+ "step": 13625,
+ "train_runtime": 105388.882,
+ "train_tokens_per_second": 5654.749
+ },
+ {
+ "epoch": 0.9177215189873418,
+ "grad_norm": 0.7671112595712233,
+ "learning_rate": 5.64494055811553e-06,
+ "loss": 0.35439510345458985,
+ "num_input_tokens_seen": 596167280,
+ "step": 13630,
+ "train_runtime": 105432.2515,
+ "train_tokens_per_second": 5654.506
+ },
+ {
+ "epoch": 0.9180581739833019,
+ "grad_norm": 0.7391097804364258,
+ "learning_rate": 5.642318474196737e-06,
+ "loss": 0.3450098991394043,
+ "num_input_tokens_seen": 596389280,
+ "step": 13635,
+ "train_runtime": 105470.2709,
+ "train_tokens_per_second": 5654.572
+ },
+ {
+ "epoch": 0.9183948289792621,
+ "grad_norm": 0.7394314350713648,
+ "learning_rate": 5.63969621065563e-06,
+ "loss": 0.36347208023071287,
+ "num_input_tokens_seen": 596603992,
+ "step": 13640,
+ "train_runtime": 105510.1258,
+ "train_tokens_per_second": 5654.471
+ },
+ {
+ "epoch": 0.9187314839752222,
+ "grad_norm": 0.6523195737553367,
+ "learning_rate": 5.637073768225517e-06,
+ "loss": 0.3404127836227417,
+ "num_input_tokens_seen": 596814560,
+ "step": 13645,
+ "train_runtime": 105555.5698,
+ "train_tokens_per_second": 5654.032
+ },
+ {
+ "epoch": 0.9190681389711823,
+ "grad_norm": 0.7286950960994242,
+ "learning_rate": 5.634451147639758e-06,
+ "loss": 0.3368164300918579,
+ "num_input_tokens_seen": 597016352,
+ "step": 13650,
+ "train_runtime": 105596.0923,
+ "train_tokens_per_second": 5653.773
+ },
+ {
+ "epoch": 0.9194047939671425,
+ "grad_norm": 0.7381840389633337,
+ "learning_rate": 5.631828349631757e-06,
+ "loss": 0.34926304817199705,
+ "num_input_tokens_seen": 597241424,
+ "step": 13655,
+ "train_runtime": 105633.1844,
+ "train_tokens_per_second": 5653.919
+ },
+ {
+ "epoch": 0.9197414489631026,
+ "grad_norm": 0.7377168394438081,
+ "learning_rate": 5.629205374934974e-06,
+ "loss": 0.3418566703796387,
+ "num_input_tokens_seen": 597449976,
+ "step": 13660,
+ "train_runtime": 105670.7222,
+ "train_tokens_per_second": 5653.884
+ },
+ {
+ "epoch": 0.9200781039590628,
+ "grad_norm": 0.7062836493823704,
+ "learning_rate": 5.626582224282913e-06,
+ "loss": 0.36661295890808104,
+ "num_input_tokens_seen": 597675264,
+ "step": 13665,
+ "train_runtime": 105714.0467,
+ "train_tokens_per_second": 5653.698
+ },
+ {
+ "epoch": 0.9204147589550229,
+ "grad_norm": 0.7520494279886026,
+ "learning_rate": 5.623958898409132e-06,
+ "loss": 0.39702858924865725,
+ "num_input_tokens_seen": 597900752,
+ "step": 13670,
+ "train_runtime": 105743.9982,
+ "train_tokens_per_second": 5654.229
+ },
+ {
+ "epoch": 0.920751413950983,
+ "grad_norm": 0.6987426458318066,
+ "learning_rate": 5.621335398047234e-06,
+ "loss": 0.3848641157150269,
+ "num_input_tokens_seen": 598115680,
+ "step": 13675,
+ "train_runtime": 105787.0252,
+ "train_tokens_per_second": 5653.961
+ },
+ {
+ "epoch": 0.9210880689469432,
+ "grad_norm": 0.7307345712178913,
+ "learning_rate": 5.618711723930874e-06,
+ "loss": 0.36865854263305664,
+ "num_input_tokens_seen": 598333664,
+ "step": 13680,
+ "train_runtime": 105817.9321,
+ "train_tokens_per_second": 5654.369
+ },
+ {
+ "epoch": 0.9214247239429033,
+ "grad_norm": 0.7963356189127992,
+ "learning_rate": 5.616087876793753e-06,
+ "loss": 0.3818777561187744,
+ "num_input_tokens_seen": 598560984,
+ "step": 13685,
+ "train_runtime": 105853.2283,
+ "train_tokens_per_second": 5654.631
+ },
+ {
+ "epoch": 0.9217613789388635,
+ "grad_norm": 0.6658404353690085,
+ "learning_rate": 5.613463857369621e-06,
+ "loss": 0.3648639678955078,
+ "num_input_tokens_seen": 598784192,
+ "step": 13690,
+ "train_runtime": 105895.3239,
+ "train_tokens_per_second": 5654.491
+ },
+ {
+ "epoch": 0.9220980339348236,
+ "grad_norm": 0.7245035749434348,
+ "learning_rate": 5.610839666392278e-06,
+ "loss": 0.35527966022491453,
+ "num_input_tokens_seen": 599005456,
+ "step": 13695,
+ "train_runtime": 105941.0406,
+ "train_tokens_per_second": 5654.14
+ },
+ {
+ "epoch": 0.9224346889307837,
+ "grad_norm": 0.7564271755937382,
+ "learning_rate": 5.608215304595571e-06,
+ "loss": 0.3818879842758179,
+ "num_input_tokens_seen": 599222856,
+ "step": 13700,
+ "train_runtime": 105987.7905,
+ "train_tokens_per_second": 5653.697
+ },
+ {
+ "epoch": 0.9227713439267439,
+ "grad_norm": 0.71469916144582,
+ "learning_rate": 5.605590772713393e-06,
+ "loss": 0.35137481689453126,
+ "num_input_tokens_seen": 599421784,
+ "step": 13705,
+ "train_runtime": 106025.2333,
+ "train_tokens_per_second": 5653.577
+ },
+ {
+ "epoch": 0.923107998922704,
+ "grad_norm": 0.8680683000673615,
+ "learning_rate": 5.602966071479687e-06,
+ "loss": 0.3719889402389526,
+ "num_input_tokens_seen": 599626912,
+ "step": 13710,
+ "train_runtime": 106057.3995,
+ "train_tokens_per_second": 5653.796
+ },
+ {
+ "epoch": 0.9234446539186641,
+ "grad_norm": 0.7763025193380707,
+ "learning_rate": 5.600341201628439e-06,
+ "loss": 0.37358889579772947,
+ "num_input_tokens_seen": 599854192,
+ "step": 13715,
+ "train_runtime": 106090.67,
+ "train_tokens_per_second": 5654.165
+ },
+ {
+ "epoch": 0.9237813089146243,
+ "grad_norm": 0.6656080663010198,
+ "learning_rate": 5.597716163893692e-06,
+ "loss": 0.31069388389587405,
+ "num_input_tokens_seen": 600085080,
+ "step": 13720,
+ "train_runtime": 106134.8208,
+ "train_tokens_per_second": 5653.989
+ },
+ {
+ "epoch": 0.9241179639105844,
+ "grad_norm": 0.7327001774115326,
+ "learning_rate": 5.595090959009525e-06,
+ "loss": 0.34327847957611085,
+ "num_input_tokens_seen": 600310096,
+ "step": 13725,
+ "train_runtime": 106180.1722,
+ "train_tokens_per_second": 5653.693
+ },
+ {
+ "epoch": 0.9244546189065446,
+ "grad_norm": 0.6125120195863898,
+ "learning_rate": 5.592465587710068e-06,
+ "loss": 0.34142630100250243,
+ "num_input_tokens_seen": 600545032,
+ "step": 13730,
+ "train_runtime": 106218.5007,
+ "train_tokens_per_second": 5653.865
+ },
+ {
+ "epoch": 0.9247912739025047,
+ "grad_norm": 0.6167159170504363,
+ "learning_rate": 5.589840050729498e-06,
+ "loss": 0.3286783218383789,
+ "num_input_tokens_seen": 600768064,
+ "step": 13735,
+ "train_runtime": 106258.0895,
+ "train_tokens_per_second": 5653.857
+ },
+ {
+ "epoch": 0.9251279288984648,
+ "grad_norm": 0.7881839614694872,
+ "learning_rate": 5.587214348802039e-06,
+ "loss": 0.3909108638763428,
+ "num_input_tokens_seen": 600980632,
+ "step": 13740,
+ "train_runtime": 106298.4773,
+ "train_tokens_per_second": 5653.709
+ },
+ {
+ "epoch": 0.925464583894425,
+ "grad_norm": 0.7022898228386897,
+ "learning_rate": 5.5845884826619615e-06,
+ "loss": 0.36343495845794677,
+ "num_input_tokens_seen": 601207448,
+ "step": 13745,
+ "train_runtime": 106333.7277,
+ "train_tokens_per_second": 5653.968
+ },
+ {
+ "epoch": 0.9258012388903851,
+ "grad_norm": 0.6378832684361027,
+ "learning_rate": 5.5819624530435775e-06,
+ "loss": 0.3585641622543335,
+ "num_input_tokens_seen": 601419704,
+ "step": 13750,
+ "train_runtime": 106373.8251,
+ "train_tokens_per_second": 5653.832
+ },
+ {
+ "epoch": 0.9261378938863453,
+ "grad_norm": 0.7242428710010252,
+ "learning_rate": 5.57933626068125e-06,
+ "loss": 0.3933718681335449,
+ "num_input_tokens_seen": 601639824,
+ "step": 13755,
+ "train_runtime": 106414.6703,
+ "train_tokens_per_second": 5653.73
+ },
+ {
+ "epoch": 0.9264745488823054,
+ "grad_norm": 0.7473031695779977,
+ "learning_rate": 5.5767099063093835e-06,
+ "loss": 0.3709197998046875,
+ "num_input_tokens_seen": 601868536,
+ "step": 13760,
+ "train_runtime": 106452.2637,
+ "train_tokens_per_second": 5653.882
+ },
+ {
+ "epoch": 0.9268112038782655,
+ "grad_norm": 0.803164002917514,
+ "learning_rate": 5.574083390662431e-06,
+ "loss": 0.373383617401123,
+ "num_input_tokens_seen": 602098696,
+ "step": 13765,
+ "train_runtime": 106489.0364,
+ "train_tokens_per_second": 5654.091
+ },
+ {
+ "epoch": 0.9271478588742257,
+ "grad_norm": 0.7108278884261701,
+ "learning_rate": 5.57145671447489e-06,
+ "loss": 0.3655261993408203,
+ "num_input_tokens_seen": 602302480,
+ "step": 13770,
+ "train_runtime": 106525.8625,
+ "train_tokens_per_second": 5654.049
+ },
+ {
+ "epoch": 0.9274845138701858,
+ "grad_norm": 0.7613900230491544,
+ "learning_rate": 5.5688298784813e-06,
+ "loss": 0.3984712600708008,
+ "num_input_tokens_seen": 602522208,
+ "step": 13775,
+ "train_runtime": 106564.2964,
+ "train_tokens_per_second": 5654.072
+ },
+ {
+ "epoch": 0.927821168866146,
+ "grad_norm": 0.6708141182451115,
+ "learning_rate": 5.566202883416249e-06,
+ "loss": 0.3970597743988037,
+ "num_input_tokens_seen": 602750424,
+ "step": 13780,
+ "train_runtime": 106606.7053,
+ "train_tokens_per_second": 5653.964
+ },
+ {
+ "epoch": 0.9281578238621061,
+ "grad_norm": 0.7821647487746413,
+ "learning_rate": 5.563575730014366e-06,
+ "loss": 0.39171276092529295,
+ "num_input_tokens_seen": 602965320,
+ "step": 13785,
+ "train_runtime": 106643.4726,
+ "train_tokens_per_second": 5654.029
+ },
+ {
+ "epoch": 0.9284944788580662,
+ "grad_norm": 0.7438040498615117,
+ "learning_rate": 5.560948419010328e-06,
+ "loss": 0.3737471342086792,
+ "num_input_tokens_seen": 603186136,
+ "step": 13790,
+ "train_runtime": 106681.291,
+ "train_tokens_per_second": 5654.095
+ },
+ {
+ "epoch": 0.9288311338540264,
+ "grad_norm": 0.7446914306967766,
+ "learning_rate": 5.558320951138853e-06,
+ "loss": 0.34050416946411133,
+ "num_input_tokens_seen": 603397288,
+ "step": 13795,
+ "train_runtime": 106722.5651,
+ "train_tokens_per_second": 5653.887
+ },
+ {
+ "epoch": 0.9291677888499865,
+ "grad_norm": 0.7759821070900785,
+ "learning_rate": 5.5556933271347046e-06,
+ "loss": 0.3723226308822632,
+ "num_input_tokens_seen": 603606088,
+ "step": 13800,
+ "train_runtime": 106763.5139,
+ "train_tokens_per_second": 5653.674
+ },
+ {
+ "epoch": 0.9295044438459467,
+ "grad_norm": 0.7098216673132232,
+ "learning_rate": 5.55306554773269e-06,
+ "loss": 0.3821061611175537,
+ "num_input_tokens_seen": 603816144,
+ "step": 13805,
+ "train_runtime": 106805.761,
+ "train_tokens_per_second": 5653.404
+ },
+ {
+ "epoch": 0.9298410988419068,
+ "grad_norm": 0.9275669407149835,
+ "learning_rate": 5.550437613667658e-06,
+ "loss": 0.3593113660812378,
+ "num_input_tokens_seen": 604004552,
+ "step": 13810,
+ "train_runtime": 106836.865,
+ "train_tokens_per_second": 5653.522
+ },
+ {
+ "epoch": 0.9301777538378669,
+ "grad_norm": 0.6896177858346889,
+ "learning_rate": 5.547809525674501e-06,
+ "loss": 0.35454933643341063,
+ "num_input_tokens_seen": 604212440,
+ "step": 13815,
+ "train_runtime": 106875.1406,
+ "train_tokens_per_second": 5653.442
+ },
+ {
+ "epoch": 0.9305144088338271,
+ "grad_norm": 0.7927279590495904,
+ "learning_rate": 5.545181284488159e-06,
+ "loss": 0.38926100730895996,
+ "num_input_tokens_seen": 604439792,
+ "step": 13820,
+ "train_runtime": 106917.4476,
+ "train_tokens_per_second": 5653.332
+ },
+ {
+ "epoch": 0.9308510638297872,
+ "grad_norm": 0.8386569241979678,
+ "learning_rate": 5.5425528908436085e-06,
+ "loss": 0.35577256679534913,
+ "num_input_tokens_seen": 604643600,
+ "step": 13825,
+ "train_runtime": 106953.2903,
+ "train_tokens_per_second": 5653.343
+ },
+ {
+ "epoch": 0.9311877188257474,
+ "grad_norm": 0.7592336539171605,
+ "learning_rate": 5.539924345475873e-06,
+ "loss": 0.3907893180847168,
+ "num_input_tokens_seen": 604839736,
+ "step": 13830,
+ "train_runtime": 106988.9857,
+ "train_tokens_per_second": 5653.29
+ },
+ {
+ "epoch": 0.9315243738217075,
+ "grad_norm": 0.8126152719075438,
+ "learning_rate": 5.5372956491200145e-06,
+ "loss": 0.37813870906829833,
+ "num_input_tokens_seen": 605051728,
+ "step": 13835,
+ "train_runtime": 107025.6706,
+ "train_tokens_per_second": 5653.333
+ },
+ {
+ "epoch": 0.9318610288176676,
+ "grad_norm": 0.7879475237888696,
+ "learning_rate": 5.534666802511143e-06,
+ "loss": 0.36364994049072263,
+ "num_input_tokens_seen": 605251488,
+ "step": 13840,
+ "train_runtime": 107062.9642,
+ "train_tokens_per_second": 5653.229
+ },
+ {
+ "epoch": 0.9321976838136278,
+ "grad_norm": 0.6977984667014386,
+ "learning_rate": 5.532037806384404e-06,
+ "loss": 0.372878885269165,
+ "num_input_tokens_seen": 605478256,
+ "step": 13845,
+ "train_runtime": 107101.7476,
+ "train_tokens_per_second": 5653.299
+ },
+ {
+ "epoch": 0.9325343388095879,
+ "grad_norm": 0.7377782721135303,
+ "learning_rate": 5.529408661474989e-06,
+ "loss": 0.33782191276550294,
+ "num_input_tokens_seen": 605678152,
+ "step": 13850,
+ "train_runtime": 107140.4741,
+ "train_tokens_per_second": 5653.122
+ },
+ {
+ "epoch": 0.932870993805548,
+ "grad_norm": 0.7120793619791788,
+ "learning_rate": 5.526779368518129e-06,
+ "loss": 0.3900485038757324,
+ "num_input_tokens_seen": 605902976,
+ "step": 13855,
+ "train_runtime": 107174.2345,
+ "train_tokens_per_second": 5653.439
+ },
+ {
+ "epoch": 0.9332076488015082,
+ "grad_norm": 0.7800631196768846,
+ "learning_rate": 5.5241499282491e-06,
+ "loss": 0.35591723918914797,
+ "num_input_tokens_seen": 606103560,
+ "step": 13860,
+ "train_runtime": 107212.7655,
+ "train_tokens_per_second": 5653.278
+ },
+ {
+ "epoch": 0.9335443037974683,
+ "grad_norm": 0.6995200522098299,
+ "learning_rate": 5.5215203414032135e-06,
+ "loss": 0.3359647274017334,
+ "num_input_tokens_seen": 606332784,
+ "step": 13865,
+ "train_runtime": 107261.2791,
+ "train_tokens_per_second": 5652.858
+ },
+ {
+ "epoch": 0.9338809587934285,
+ "grad_norm": 0.7315710239059265,
+ "learning_rate": 5.518890608715828e-06,
+ "loss": 0.3636192798614502,
+ "num_input_tokens_seen": 606549080,
+ "step": 13870,
+ "train_runtime": 107306.9882,
+ "train_tokens_per_second": 5652.466
+ },
+ {
+ "epoch": 0.9342176137893886,
+ "grad_norm": 0.8443286259971129,
+ "learning_rate": 5.516260730922337e-06,
+ "loss": 0.3732717275619507,
+ "num_input_tokens_seen": 606754664,
+ "step": 13875,
+ "train_runtime": 107344.4184,
+ "train_tokens_per_second": 5652.41
+ },
+ {
+ "epoch": 0.9345542687853488,
+ "grad_norm": 0.7163150661251769,
+ "learning_rate": 5.51363070875818e-06,
+ "loss": 0.3665364503860474,
+ "num_input_tokens_seen": 606971336,
+ "step": 13880,
+ "train_runtime": 107380.8754,
+ "train_tokens_per_second": 5652.509
+ },
+ {
+ "epoch": 0.9348909237813089,
+ "grad_norm": 0.7826404929432437,
+ "learning_rate": 5.511000542958832e-06,
+ "loss": 0.3829598903656006,
+ "num_input_tokens_seen": 607182888,
+ "step": 13885,
+ "train_runtime": 107425.4489,
+ "train_tokens_per_second": 5652.133
+ },
+ {
+ "epoch": 0.935227578777269,
+ "grad_norm": 0.7974161880488172,
+ "learning_rate": 5.508370234259812e-06,
+ "loss": 0.3617416858673096,
+ "num_input_tokens_seen": 607395376,
+ "step": 13890,
+ "train_runtime": 107467.1083,
+ "train_tokens_per_second": 5651.919
+ },
+ {
+ "epoch": 0.9355642337732292,
+ "grad_norm": 0.7468198908899866,
+ "learning_rate": 5.505739783396678e-06,
+ "loss": 0.4197434425354004,
+ "num_input_tokens_seen": 607606560,
+ "step": 13895,
+ "train_runtime": 107501.3365,
+ "train_tokens_per_second": 5652.084
+ },
+ {
+ "epoch": 0.9359008887691893,
+ "grad_norm": 0.8018465922168921,
+ "learning_rate": 5.503109191105026e-06,
+ "loss": 0.368869423866272,
+ "num_input_tokens_seen": 607818184,
+ "step": 13900,
+ "train_runtime": 107542.256,
+ "train_tokens_per_second": 5651.901
+ },
+ {
+ "epoch": 0.9362375437651494,
+ "grad_norm": 0.6896918447153622,
+ "learning_rate": 5.500478458120493e-06,
+ "loss": 0.3878501892089844,
+ "num_input_tokens_seen": 608046144,
+ "step": 13905,
+ "train_runtime": 107578.9863,
+ "train_tokens_per_second": 5652.09
+ },
+ {
+ "epoch": 0.9365741987611096,
+ "grad_norm": 0.6502629376388114,
+ "learning_rate": 5.497847585178754e-06,
+ "loss": 0.37720532417297364,
+ "num_input_tokens_seen": 608274496,
+ "step": 13910,
+ "train_runtime": 107614.2489,
+ "train_tokens_per_second": 5652.36
+ },
+ {
+ "epoch": 0.9369108537570697,
+ "grad_norm": 0.671150617120138,
+ "learning_rate": 5.495216573015526e-06,
+ "loss": 0.34202401638031005,
+ "num_input_tokens_seen": 608487072,
+ "step": 13915,
+ "train_runtime": 107649.7272,
+ "train_tokens_per_second": 5652.472
+ },
+ {
+ "epoch": 0.9372475087530299,
+ "grad_norm": 0.7302905322159697,
+ "learning_rate": 5.492585422366562e-06,
+ "loss": 0.3395219802856445,
+ "num_input_tokens_seen": 608702136,
+ "step": 13920,
+ "train_runtime": 107690.4313,
+ "train_tokens_per_second": 5652.333
+ },
+ {
+ "epoch": 0.93758416374899,
+ "grad_norm": 0.7232012493162839,
+ "learning_rate": 5.489954133967657e-06,
+ "loss": 0.3353360652923584,
+ "num_input_tokens_seen": 608907648,
+ "step": 13925,
+ "train_runtime": 107730.6462,
+ "train_tokens_per_second": 5652.13
+ },
+ {
+ "epoch": 0.9379208187449501,
+ "grad_norm": 0.6700324546008681,
+ "learning_rate": 5.48732270855464e-06,
+ "loss": 0.3447211027145386,
+ "num_input_tokens_seen": 609137704,
+ "step": 13930,
+ "train_runtime": 107770.2254,
+ "train_tokens_per_second": 5652.189
+ },
+ {
+ "epoch": 0.9382574737409103,
+ "grad_norm": 0.744553222908747,
+ "learning_rate": 5.48469114686338e-06,
+ "loss": 0.36331336498260497,
+ "num_input_tokens_seen": 609342520,
+ "step": 13935,
+ "train_runtime": 107808.8795,
+ "train_tokens_per_second": 5652.062
+ },
+ {
+ "epoch": 0.9385941287368704,
+ "grad_norm": 0.7229277790841906,
+ "learning_rate": 5.482059449629788e-06,
+ "loss": 0.3539914131164551,
+ "num_input_tokens_seen": 609571440,
+ "step": 13940,
+ "train_runtime": 107847.7103,
+ "train_tokens_per_second": 5652.15
+ },
+ {
+ "epoch": 0.9389307837328306,
+ "grad_norm": 0.7003088104586751,
+ "learning_rate": 5.479427617589808e-06,
+ "loss": 0.3414954662322998,
+ "num_input_tokens_seen": 609787136,
+ "step": 13945,
+ "train_runtime": 107885.5574,
+ "train_tokens_per_second": 5652.167
+ },
+ {
+ "epoch": 0.9392674387287907,
+ "grad_norm": 0.6891150838884817,
+ "learning_rate": 5.4767956514794226e-06,
+ "loss": 0.3689284801483154,
+ "num_input_tokens_seen": 610015912,
+ "step": 13950,
+ "train_runtime": 107918.1469,
+ "train_tokens_per_second": 5652.58
+ },
+ {
+ "epoch": 0.9396040937247508,
+ "grad_norm": 0.8843526225142938,
+ "learning_rate": 5.474163552034655e-06,
+ "loss": 0.3568282127380371,
+ "num_input_tokens_seen": 610224872,
+ "step": 13955,
+ "train_runtime": 107958.8593,
+ "train_tokens_per_second": 5652.383
+ },
+ {
+ "epoch": 0.939940748720711,
+ "grad_norm": 0.7953669523701649,
+ "learning_rate": 5.4715313199915595e-06,
+ "loss": 0.33748788833618165,
+ "num_input_tokens_seen": 610433896,
+ "step": 13960,
+ "train_runtime": 107998.8796,
+ "train_tokens_per_second": 5652.224
+ },
+ {
+ "epoch": 0.9402774037166711,
+ "grad_norm": 0.8436524922865566,
+ "learning_rate": 5.468898956086236e-06,
+ "loss": 0.36934537887573243,
+ "num_input_tokens_seen": 610657928,
+ "step": 13965,
+ "train_runtime": 108037.0093,
+ "train_tokens_per_second": 5652.303
+ },
+ {
+ "epoch": 0.9406140587126313,
+ "grad_norm": 0.7062142551000551,
+ "learning_rate": 5.466266461054815e-06,
+ "loss": 0.3693251132965088,
+ "num_input_tokens_seen": 610879072,
+ "step": 13970,
+ "train_runtime": 108078.9259,
+ "train_tokens_per_second": 5652.157
+ },
+ {
+ "epoch": 0.9409507137085914,
+ "grad_norm": 0.7405368560790437,
+ "learning_rate": 5.463633835633464e-06,
+ "loss": 0.38323063850402833,
+ "num_input_tokens_seen": 611111488,
+ "step": 13975,
+ "train_runtime": 108116.6852,
+ "train_tokens_per_second": 5652.333
+ },
+ {
+ "epoch": 0.9412873687045515,
+ "grad_norm": 0.6906471137787829,
+ "learning_rate": 5.46100108055839e-06,
+ "loss": 0.37401399612426756,
+ "num_input_tokens_seen": 611338984,
+ "step": 13980,
+ "train_runtime": 108153.7009,
+ "train_tokens_per_second": 5652.502
+ },
+ {
+ "epoch": 0.9416240237005117,
+ "grad_norm": 0.699157303658586,
+ "learning_rate": 5.458368196565833e-06,
+ "loss": 0.34747042655944826,
+ "num_input_tokens_seen": 611557384,
+ "step": 13985,
+ "train_runtime": 108187.4002,
+ "train_tokens_per_second": 5652.76
+ },
+ {
+ "epoch": 0.9419606786964718,
+ "grad_norm": 0.6950434602213443,
+ "learning_rate": 5.455735184392073e-06,
+ "loss": 0.39068779945373533,
+ "num_input_tokens_seen": 611785288,
+ "step": 13990,
+ "train_runtime": 108222.1017,
+ "train_tokens_per_second": 5653.053
+ },
+ {
+ "epoch": 0.942297333692432,
+ "grad_norm": 0.8222775495837801,
+ "learning_rate": 5.4531020447734204e-06,
+ "loss": 0.3776577949523926,
+ "num_input_tokens_seen": 611999096,
+ "step": 13995,
+ "train_runtime": 108255.2451,
+ "train_tokens_per_second": 5653.297
+ },
+ {
+ "epoch": 0.9426339886883921,
+ "grad_norm": 0.6942977966354065,
+ "learning_rate": 5.450468778446226e-06,
+ "loss": 0.38014535903930663,
+ "num_input_tokens_seen": 612229616,
+ "step": 14000,
+ "train_runtime": 108302.2562,
+ "train_tokens_per_second": 5652.972
+ },
+ {
+ "epoch": 0.9429706436843522,
+ "grad_norm": 0.8181331571770365,
+ "learning_rate": 5.447835386146875e-06,
+ "loss": 0.3620244264602661,
+ "num_input_tokens_seen": 612457288,
+ "step": 14005,
+ "train_runtime": 108339.7787,
+ "train_tokens_per_second": 5653.116
+ },
+ {
+ "epoch": 0.9433072986803124,
+ "grad_norm": 0.7249133494502058,
+ "learning_rate": 5.445201868611784e-06,
+ "loss": 0.4034558296203613,
+ "num_input_tokens_seen": 612668352,
+ "step": 14010,
+ "train_runtime": 108376.897,
+ "train_tokens_per_second": 5653.127
+ },
+ {
+ "epoch": 0.9436439536762725,
+ "grad_norm": 0.7004943272407984,
+ "learning_rate": 5.442568226577413e-06,
+ "loss": 0.37315356731414795,
+ "num_input_tokens_seen": 612885272,
+ "step": 14015,
+ "train_runtime": 108415.5424,
+ "train_tokens_per_second": 5653.113
+ },
+ {
+ "epoch": 0.9439806086722327,
+ "grad_norm": 0.6490441295997635,
+ "learning_rate": 5.439934460780247e-06,
+ "loss": 0.3591346740722656,
+ "num_input_tokens_seen": 613107944,
+ "step": 14020,
+ "train_runtime": 108453.9122,
+ "train_tokens_per_second": 5653.166
+ },
+ {
+ "epoch": 0.9443172636681928,
+ "grad_norm": 0.6918392163792407,
+ "learning_rate": 5.437300571956813e-06,
+ "loss": 0.3745762586593628,
+ "num_input_tokens_seen": 613308168,
+ "step": 14025,
+ "train_runtime": 108489.5473,
+ "train_tokens_per_second": 5653.154
+ },
+ {
+ "epoch": 0.9446539186641529,
+ "grad_norm": 0.654742109915237,
+ "learning_rate": 5.434666560843667e-06,
+ "loss": 0.3551823616027832,
+ "num_input_tokens_seen": 613516376,
+ "step": 14030,
+ "train_runtime": 108525.2798,
+ "train_tokens_per_second": 5653.212
+ },
+ {
+ "epoch": 0.9449905736601131,
+ "grad_norm": 0.7025947454416406,
+ "learning_rate": 5.432032428177405e-06,
+ "loss": 0.3733973979949951,
+ "num_input_tokens_seen": 613743992,
+ "step": 14035,
+ "train_runtime": 108565.1399,
+ "train_tokens_per_second": 5653.233
+ },
+ {
+ "epoch": 0.9453272286560732,
+ "grad_norm": 0.7446809227599923,
+ "learning_rate": 5.429398174694651e-06,
+ "loss": 0.35695390701293944,
+ "num_input_tokens_seen": 613975032,
+ "step": 14040,
+ "train_runtime": 108600.6982,
+ "train_tokens_per_second": 5653.509
+ },
+ {
+ "epoch": 0.9456638836520334,
+ "grad_norm": 0.6898581042276205,
+ "learning_rate": 5.426763801132067e-06,
+ "loss": 0.3617269039154053,
+ "num_input_tokens_seen": 614162568,
+ "step": 14045,
+ "train_runtime": 108637.4509,
+ "train_tokens_per_second": 5653.323
+ },
+ {
+ "epoch": 0.9460005386479935,
+ "grad_norm": 0.7678862695446375,
+ "learning_rate": 5.424129308226347e-06,
+ "loss": 0.3843861103057861,
+ "num_input_tokens_seen": 614370656,
+ "step": 14050,
+ "train_runtime": 108675.5655,
+ "train_tokens_per_second": 5653.255
+ },
+ {
+ "epoch": 0.9463371936439536,
+ "grad_norm": 0.7654766109095031,
+ "learning_rate": 5.421494696714219e-06,
+ "loss": 0.3902265548706055,
+ "num_input_tokens_seen": 614608864,
+ "step": 14055,
+ "train_runtime": 108715.2383,
+ "train_tokens_per_second": 5653.383
+ },
+ {
+ "epoch": 0.9466738486399138,
+ "grad_norm": 0.7609882464523344,
+ "learning_rate": 5.418859967332442e-06,
+ "loss": 0.3707080841064453,
+ "num_input_tokens_seen": 614836608,
+ "step": 14060,
+ "train_runtime": 108755.1102,
+ "train_tokens_per_second": 5653.404
+ },
+ {
+ "epoch": 0.9470105036358739,
+ "grad_norm": 0.9394544427990754,
+ "learning_rate": 5.416225120817811e-06,
+ "loss": 0.3681238412857056,
+ "num_input_tokens_seen": 615027064,
+ "step": 14065,
+ "train_runtime": 108790.9955,
+ "train_tokens_per_second": 5653.29
+ },
+ {
+ "epoch": 0.947347158631834,
+ "grad_norm": 0.6539346876126566,
+ "learning_rate": 5.413590157907153e-06,
+ "loss": 0.3346075534820557,
+ "num_input_tokens_seen": 615270632,
+ "step": 14070,
+ "train_runtime": 108826.6066,
+ "train_tokens_per_second": 5653.678
+ },
+ {
+ "epoch": 0.9476838136277942,
+ "grad_norm": 0.7338889594818504,
+ "learning_rate": 5.4109550793373255e-06,
+ "loss": 0.355800724029541,
+ "num_input_tokens_seen": 615493608,
+ "step": 14075,
+ "train_runtime": 108868.2257,
+ "train_tokens_per_second": 5653.565
+ },
+ {
+ "epoch": 0.9480204686237543,
+ "grad_norm": 0.7464991259586956,
+ "learning_rate": 5.40831988584522e-06,
+ "loss": 0.38836262226104734,
+ "num_input_tokens_seen": 615719216,
+ "step": 14080,
+ "train_runtime": 108904.0303,
+ "train_tokens_per_second": 5653.778
+ },
+ {
+ "epoch": 0.9483571236197145,
+ "grad_norm": 0.7069248942260824,
+ "learning_rate": 5.40568457816776e-06,
+ "loss": 0.37587385177612304,
+ "num_input_tokens_seen": 615951848,
+ "step": 14085,
+ "train_runtime": 108947.057,
+ "train_tokens_per_second": 5653.68
+ },
+ {
+ "epoch": 0.9486937786156746,
+ "grad_norm": 0.6655009611716466,
+ "learning_rate": 5.403049157041903e-06,
+ "loss": 0.35560946464538573,
+ "num_input_tokens_seen": 616172928,
+ "step": 14090,
+ "train_runtime": 108984.6239,
+ "train_tokens_per_second": 5653.76
+ },
+ {
+ "epoch": 0.9490304336116347,
+ "grad_norm": 0.7730572202934075,
+ "learning_rate": 5.400413623204635e-06,
+ "loss": 0.35712380409240724,
+ "num_input_tokens_seen": 616390224,
+ "step": 14095,
+ "train_runtime": 109023.2559,
+ "train_tokens_per_second": 5653.75
+ },
+ {
+ "epoch": 0.9493670886075949,
+ "grad_norm": 0.7946160574063879,
+ "learning_rate": 5.397777977392973e-06,
+ "loss": 0.39092726707458497,
+ "num_input_tokens_seen": 616612064,
+ "step": 14100,
+ "train_runtime": 109071.1955,
+ "train_tokens_per_second": 5653.299
+ },
+ {
+ "epoch": 0.949703743603555,
+ "grad_norm": 0.7261331777714058,
+ "learning_rate": 5.395142220343967e-06,
+ "loss": 0.37581348419189453,
+ "num_input_tokens_seen": 616835560,
+ "step": 14105,
+ "train_runtime": 109114.6463,
+ "train_tokens_per_second": 5653.096
+ },
+ {
+ "epoch": 0.9500403985995152,
+ "grad_norm": 0.7685617426175168,
+ "learning_rate": 5.3925063527947e-06,
+ "loss": 0.34862990379333497,
+ "num_input_tokens_seen": 617051200,
+ "step": 14110,
+ "train_runtime": 109161.8492,
+ "train_tokens_per_second": 5652.627
+ },
+ {
+ "epoch": 0.9503770535954753,
+ "grad_norm": 0.771130071661086,
+ "learning_rate": 5.389870375482283e-06,
+ "loss": 0.37904350757598876,
+ "num_input_tokens_seen": 617248976,
+ "step": 14115,
+ "train_runtime": 109194.634,
+ "train_tokens_per_second": 5652.741
+ },
+ {
+ "epoch": 0.9507137085914354,
+ "grad_norm": 0.7205683359655156,
+ "learning_rate": 5.387234289143859e-06,
+ "loss": 0.331813383102417,
+ "num_input_tokens_seen": 617469104,
+ "step": 14120,
+ "train_runtime": 109226.9679,
+ "train_tokens_per_second": 5653.083
+ },
+ {
+ "epoch": 0.9510503635873956,
+ "grad_norm": 0.7693964177160706,
+ "learning_rate": 5.384598094516601e-06,
+ "loss": 0.3729023218154907,
+ "num_input_tokens_seen": 617692328,
+ "step": 14125,
+ "train_runtime": 109271.6235,
+ "train_tokens_per_second": 5652.816
+ },
+ {
+ "epoch": 0.9513870185833557,
+ "grad_norm": 0.6640672951882179,
+ "learning_rate": 5.381961792337712e-06,
+ "loss": 0.3759523153305054,
+ "num_input_tokens_seen": 617921872,
+ "step": 14130,
+ "train_runtime": 109307.5627,
+ "train_tokens_per_second": 5653.057
+ },
+ {
+ "epoch": 0.9517236735793159,
+ "grad_norm": 0.6678437973528463,
+ "learning_rate": 5.3793253833444245e-06,
+ "loss": 0.3401124715805054,
+ "num_input_tokens_seen": 618147624,
+ "step": 14135,
+ "train_runtime": 109342.9092,
+ "train_tokens_per_second": 5653.294
+ },
+ {
+ "epoch": 0.952060328575276,
+ "grad_norm": 0.6761734645912748,
+ "learning_rate": 5.3766888682740045e-06,
+ "loss": 0.3542716264724731,
+ "num_input_tokens_seen": 618375336,
+ "step": 14140,
+ "train_runtime": 109378.6235,
+ "train_tokens_per_second": 5653.53
+ },
+ {
+ "epoch": 0.9523969835712361,
+ "grad_norm": 0.7917338745636189,
+ "learning_rate": 5.374052247863745e-06,
+ "loss": 0.3719951152801514,
+ "num_input_tokens_seen": 618588568,
+ "step": 14145,
+ "train_runtime": 109412.8524,
+ "train_tokens_per_second": 5653.71
+ },
+ {
+ "epoch": 0.9527336385671963,
+ "grad_norm": 0.7643089958554735,
+ "learning_rate": 5.371415522850966e-06,
+ "loss": 0.34739990234375,
+ "num_input_tokens_seen": 618807136,
+ "step": 14150,
+ "train_runtime": 109455.6515,
+ "train_tokens_per_second": 5653.496
+ },
+ {
+ "epoch": 0.9530702935631564,
+ "grad_norm": 0.7334076755125779,
+ "learning_rate": 5.368778693973019e-06,
+ "loss": 0.35079226493835447,
+ "num_input_tokens_seen": 619020192,
+ "step": 14155,
+ "train_runtime": 109497.2351,
+ "train_tokens_per_second": 5653.295
+ },
+ {
+ "epoch": 0.9534069485591166,
+ "grad_norm": 0.6518007663729203,
+ "learning_rate": 5.366141761967287e-06,
+ "loss": 0.3731173753738403,
+ "num_input_tokens_seen": 619234312,
+ "step": 14160,
+ "train_runtime": 109535.8084,
+ "train_tokens_per_second": 5653.259
+ },
+ {
+ "epoch": 0.9537436035550767,
+ "grad_norm": 0.884684695372941,
+ "learning_rate": 5.363504727571179e-06,
+ "loss": 0.38480224609375,
+ "num_input_tokens_seen": 619455112,
+ "step": 14165,
+ "train_runtime": 109574.3358,
+ "train_tokens_per_second": 5653.286
+ },
+ {
+ "epoch": 0.954080258551037,
+ "grad_norm": 0.7444612393985186,
+ "learning_rate": 5.360867591522133e-06,
+ "loss": 0.4258725166320801,
+ "num_input_tokens_seen": 619686368,
+ "step": 14170,
+ "train_runtime": 109614.4844,
+ "train_tokens_per_second": 5653.326
+ },
+ {
+ "epoch": 0.9544169135469971,
+ "grad_norm": 0.766088989031395,
+ "learning_rate": 5.358230354557617e-06,
+ "loss": 0.34602572917938235,
+ "num_input_tokens_seen": 619912184,
+ "step": 14175,
+ "train_runtime": 109647.5189,
+ "train_tokens_per_second": 5653.682
+ },
+ {
+ "epoch": 0.9547535685429572,
+ "grad_norm": 0.8426997667284004,
+ "learning_rate": 5.355593017415123e-06,
+ "loss": 0.39844369888305664,
+ "num_input_tokens_seen": 620138400,
+ "step": 14180,
+ "train_runtime": 109681.0251,
+ "train_tokens_per_second": 5654.017
+ },
+ {
+ "epoch": 0.9550902235389174,
+ "grad_norm": 0.6855456694998082,
+ "learning_rate": 5.352955580832176e-06,
+ "loss": 0.36531069278717043,
+ "num_input_tokens_seen": 620363336,
+ "step": 14185,
+ "train_runtime": 109720.9773,
+ "train_tokens_per_second": 5654.008
+ },
+ {
+ "epoch": 0.9554268785348775,
+ "grad_norm": 0.6489568473620863,
+ "learning_rate": 5.350318045546326e-06,
+ "loss": 0.378409218788147,
+ "num_input_tokens_seen": 620586608,
+ "step": 14190,
+ "train_runtime": 109762.864,
+ "train_tokens_per_second": 5653.885
+ },
+ {
+ "epoch": 0.9557635335308377,
+ "grad_norm": 0.690203133783407,
+ "learning_rate": 5.347680412295152e-06,
+ "loss": 0.3777376413345337,
+ "num_input_tokens_seen": 620803016,
+ "step": 14195,
+ "train_runtime": 109803.0202,
+ "train_tokens_per_second": 5653.788
+ },
+ {
+ "epoch": 0.9561001885267978,
+ "grad_norm": 0.8763898868581147,
+ "learning_rate": 5.34504268181626e-06,
+ "loss": 0.37516131401062014,
+ "num_input_tokens_seen": 621009712,
+ "step": 14200,
+ "train_runtime": 109842.1757,
+ "train_tokens_per_second": 5653.655
+ },
+ {
+ "epoch": 0.9564368435227579,
+ "grad_norm": 0.8351648882525854,
+ "learning_rate": 5.342404854847282e-06,
+ "loss": 0.36751456260681153,
+ "num_input_tokens_seen": 621209912,
+ "step": 14205,
+ "train_runtime": 109878.6528,
+ "train_tokens_per_second": 5653.6
+ },
+ {
+ "epoch": 0.9567734985187181,
+ "grad_norm": 0.695513608466477,
+ "learning_rate": 5.3397669321258776e-06,
+ "loss": 0.3533796310424805,
+ "num_input_tokens_seen": 621439136,
+ "step": 14210,
+ "train_runtime": 109913.5419,
+ "train_tokens_per_second": 5653.891
+ },
+ {
+ "epoch": 0.9571101535146782,
+ "grad_norm": 0.7212650216714175,
+ "learning_rate": 5.337128914389734e-06,
+ "loss": 0.3940083980560303,
+ "num_input_tokens_seen": 621670256,
+ "step": 14215,
+ "train_runtime": 109948.6255,
+ "train_tokens_per_second": 5654.189
+ },
+ {
+ "epoch": 0.9574468085106383,
+ "grad_norm": 0.7045386935112185,
+ "learning_rate": 5.334490802376565e-06,
+ "loss": 0.3630239009857178,
+ "num_input_tokens_seen": 621896344,
+ "step": 14220,
+ "train_runtime": 109992.4071,
+ "train_tokens_per_second": 5653.993
+ },
+ {
+ "epoch": 0.9577834635065985,
+ "grad_norm": 1.0648426671565956,
+ "learning_rate": 5.33185259682411e-06,
+ "loss": 0.38904123306274413,
+ "num_input_tokens_seen": 622111488,
+ "step": 14225,
+ "train_runtime": 110028.3421,
+ "train_tokens_per_second": 5654.102
+ },
+ {
+ "epoch": 0.9581201185025586,
+ "grad_norm": 0.6854092676166333,
+ "learning_rate": 5.329214298470134e-06,
+ "loss": 0.3952740907669067,
+ "num_input_tokens_seen": 622335752,
+ "step": 14230,
+ "train_runtime": 110067.3201,
+ "train_tokens_per_second": 5654.137
+ },
+ {
+ "epoch": 0.9584567734985188,
+ "grad_norm": 0.6504683472772627,
+ "learning_rate": 5.326575908052428e-06,
+ "loss": 0.34938879013061525,
+ "num_input_tokens_seen": 622563512,
+ "step": 14235,
+ "train_runtime": 110103.0055,
+ "train_tokens_per_second": 5654.373
+ },
+ {
+ "epoch": 0.9587934284944789,
+ "grad_norm": 0.8743571550942602,
+ "learning_rate": 5.323937426308813e-06,
+ "loss": 0.4064329147338867,
+ "num_input_tokens_seen": 622789680,
+ "step": 14240,
+ "train_runtime": 110140.2833,
+ "train_tokens_per_second": 5654.513
+ },
+ {
+ "epoch": 0.959130083490439,
+ "grad_norm": 0.7364135099684984,
+ "learning_rate": 5.321298853977128e-06,
+ "loss": 0.3482068538665771,
+ "num_input_tokens_seen": 623030288,
+ "step": 14245,
+ "train_runtime": 110172.6755,
+ "train_tokens_per_second": 5655.035
+ },
+ {
+ "epoch": 0.9594667384863992,
+ "grad_norm": 1.04514521012973,
+ "learning_rate": 5.318660191795244e-06,
+ "loss": 0.35120339393615724,
+ "num_input_tokens_seen": 623250248,
+ "step": 14250,
+ "train_runtime": 110210.2613,
+ "train_tokens_per_second": 5655.102
+ },
+ {
+ "epoch": 0.9598033934823593,
+ "grad_norm": 0.6859815669971231,
+ "learning_rate": 5.316021440501053e-06,
+ "loss": 0.32753448486328124,
+ "num_input_tokens_seen": 623468032,
+ "step": 14255,
+ "train_runtime": 110254.6316,
+ "train_tokens_per_second": 5654.801
+ },
+ {
+ "epoch": 0.9601400484783195,
+ "grad_norm": 0.6696719492054304,
+ "learning_rate": 5.313382600832474e-06,
+ "loss": 0.32216849327087405,
+ "num_input_tokens_seen": 623687536,
+ "step": 14260,
+ "train_runtime": 110295.1281,
+ "train_tokens_per_second": 5654.715
+ },
+ {
+ "epoch": 0.9604767034742796,
+ "grad_norm": 0.6796334454419847,
+ "learning_rate": 5.310743673527451e-06,
+ "loss": 0.3820059776306152,
+ "num_input_tokens_seen": 623916888,
+ "step": 14265,
+ "train_runtime": 110337.4663,
+ "train_tokens_per_second": 5654.624
+ },
+ {
+ "epoch": 0.9608133584702397,
+ "grad_norm": 0.7257174247083755,
+ "learning_rate": 5.3081046593239496e-06,
+ "loss": 0.38223018646240237,
+ "num_input_tokens_seen": 624136416,
+ "step": 14270,
+ "train_runtime": 110378.6302,
+ "train_tokens_per_second": 5654.504
+ },
+ {
+ "epoch": 0.9611500134661999,
+ "grad_norm": 0.7285256715224561,
+ "learning_rate": 5.305465558959964e-06,
+ "loss": 0.366811728477478,
+ "num_input_tokens_seen": 624349152,
+ "step": 14275,
+ "train_runtime": 110421.354,
+ "train_tokens_per_second": 5654.243
+ },
+ {
+ "epoch": 0.96148666846216,
+ "grad_norm": 0.7064345566990189,
+ "learning_rate": 5.302826373173506e-06,
+ "loss": 0.3593759536743164,
+ "num_input_tokens_seen": 624577352,
+ "step": 14280,
+ "train_runtime": 110460.1396,
+ "train_tokens_per_second": 5654.323
+ },
+ {
+ "epoch": 0.9618233234581202,
+ "grad_norm": 0.7256414932456383,
+ "learning_rate": 5.3001871027026194e-06,
+ "loss": 0.3493657112121582,
+ "num_input_tokens_seen": 624788664,
+ "step": 14285,
+ "train_runtime": 110502.0095,
+ "train_tokens_per_second": 5654.093
+ },
+ {
+ "epoch": 0.9621599784540803,
+ "grad_norm": 0.7017567146140217,
+ "learning_rate": 5.2975477482853685e-06,
+ "loss": 0.3475620746612549,
+ "num_input_tokens_seen": 625013976,
+ "step": 14290,
+ "train_runtime": 110543.2223,
+ "train_tokens_per_second": 5654.023
+ },
+ {
+ "epoch": 0.9624966334500404,
+ "grad_norm": 0.8025787225363398,
+ "learning_rate": 5.294908310659838e-06,
+ "loss": 0.3636843204498291,
+ "num_input_tokens_seen": 625243416,
+ "step": 14295,
+ "train_runtime": 110579.9936,
+ "train_tokens_per_second": 5654.218
+ },
+ {
+ "epoch": 0.9628332884460006,
+ "grad_norm": 0.796752899535943,
+ "learning_rate": 5.292268790564138e-06,
+ "loss": 0.3458188772201538,
+ "num_input_tokens_seen": 625456472,
+ "step": 14300,
+ "train_runtime": 110621.9221,
+ "train_tokens_per_second": 5654.001
+ },
+ {
+ "epoch": 0.9631699434419607,
+ "grad_norm": 0.7622734681695542,
+ "learning_rate": 5.289629188736403e-06,
+ "loss": 0.39026923179626466,
+ "num_input_tokens_seen": 625676128,
+ "step": 14305,
+ "train_runtime": 110667.8056,
+ "train_tokens_per_second": 5653.642
+ },
+ {
+ "epoch": 0.9635065984379209,
+ "grad_norm": 0.973852297067902,
+ "learning_rate": 5.286989505914788e-06,
+ "loss": 0.38088412284851075,
+ "num_input_tokens_seen": 625904384,
+ "step": 14310,
+ "train_runtime": 110702.6608,
+ "train_tokens_per_second": 5653.924
+ },
+ {
+ "epoch": 0.963843253433881,
+ "grad_norm": 0.8147025206812348,
+ "learning_rate": 5.284349742837475e-06,
+ "loss": 0.3758919954299927,
+ "num_input_tokens_seen": 626134784,
+ "step": 14315,
+ "train_runtime": 110744.3819,
+ "train_tokens_per_second": 5653.874
+ },
+ {
+ "epoch": 0.9641799084298411,
+ "grad_norm": 0.8865610211887707,
+ "learning_rate": 5.281709900242662e-06,
+ "loss": 0.35522916316986086,
+ "num_input_tokens_seen": 626357296,
+ "step": 14320,
+ "train_runtime": 110789.1337,
+ "train_tokens_per_second": 5653.599
+ },
+ {
+ "epoch": 0.9645165634258013,
+ "grad_norm": 0.7685303095133553,
+ "learning_rate": 5.279069978868574e-06,
+ "loss": 0.3454322099685669,
+ "num_input_tokens_seen": 626576680,
+ "step": 14325,
+ "train_runtime": 110823.2787,
+ "train_tokens_per_second": 5653.836
+ },
+ {
+ "epoch": 0.9648532184217614,
+ "grad_norm": 0.7582279578264484,
+ "learning_rate": 5.276429979453455e-06,
+ "loss": 0.36722531318664553,
+ "num_input_tokens_seen": 626795800,
+ "step": 14330,
+ "train_runtime": 110855.2605,
+ "train_tokens_per_second": 5654.182
+ },
+ {
+ "epoch": 0.9651898734177216,
+ "grad_norm": 0.6609643186180265,
+ "learning_rate": 5.273789902735575e-06,
+ "loss": 0.36404809951782224,
+ "num_input_tokens_seen": 627011600,
+ "step": 14335,
+ "train_runtime": 110890.1684,
+ "train_tokens_per_second": 5654.348
+ },
+ {
+ "epoch": 0.9655265284136817,
+ "grad_norm": 0.7160660322102089,
+ "learning_rate": 5.27114974945322e-06,
+ "loss": 0.33545889854431155,
+ "num_input_tokens_seen": 627226888,
+ "step": 14340,
+ "train_runtime": 110937.4224,
+ "train_tokens_per_second": 5653.88
+ },
+ {
+ "epoch": 0.9658631834096418,
+ "grad_norm": 0.7371545286431122,
+ "learning_rate": 5.2685095203447046e-06,
+ "loss": 0.3596933841705322,
+ "num_input_tokens_seen": 627453768,
+ "step": 14345,
+ "train_runtime": 110976.5578,
+ "train_tokens_per_second": 5653.931
+ },
+ {
+ "epoch": 0.966199838405602,
+ "grad_norm": 0.6944067783154944,
+ "learning_rate": 5.265869216148357e-06,
+ "loss": 0.352811861038208,
+ "num_input_tokens_seen": 627673216,
+ "step": 14350,
+ "train_runtime": 111014.9066,
+ "train_tokens_per_second": 5653.954
+ },
+ {
+ "epoch": 0.9665364934015621,
+ "grad_norm": 0.7335669224993876,
+ "learning_rate": 5.263228837602532e-06,
+ "loss": 0.42299542427062986,
+ "num_input_tokens_seen": 627887552,
+ "step": 14355,
+ "train_runtime": 111050.8572,
+ "train_tokens_per_second": 5654.054
+ },
+ {
+ "epoch": 0.9668731483975223,
+ "grad_norm": 0.83794134450925,
+ "learning_rate": 5.2605883854456e-06,
+ "loss": 0.3737421989440918,
+ "num_input_tokens_seen": 628108328,
+ "step": 14360,
+ "train_runtime": 111089.5086,
+ "train_tokens_per_second": 5654.074
+ },
+ {
+ "epoch": 0.9672098033934824,
+ "grad_norm": 0.6342873275404923,
+ "learning_rate": 5.257947860415957e-06,
+ "loss": 0.3357997894287109,
+ "num_input_tokens_seen": 628335808,
+ "step": 14365,
+ "train_runtime": 111122.6788,
+ "train_tokens_per_second": 5654.434
+ },
+ {
+ "epoch": 0.9675464583894425,
+ "grad_norm": 0.8822334071658794,
+ "learning_rate": 5.255307263252019e-06,
+ "loss": 0.3541123867034912,
+ "num_input_tokens_seen": 628566792,
+ "step": 14370,
+ "train_runtime": 111156.9119,
+ "train_tokens_per_second": 5654.77
+ },
+ {
+ "epoch": 0.9678831133854027,
+ "grad_norm": 0.674709973682241,
+ "learning_rate": 5.2526665946922185e-06,
+ "loss": 0.3823538303375244,
+ "num_input_tokens_seen": 628784336,
+ "step": 14375,
+ "train_runtime": 111196.5312,
+ "train_tokens_per_second": 5654.712
+ },
+ {
+ "epoch": 0.9682197683813628,
+ "grad_norm": 0.658823173011451,
+ "learning_rate": 5.25002585547501e-06,
+ "loss": 0.35442066192626953,
+ "num_input_tokens_seen": 629008320,
+ "step": 14380,
+ "train_runtime": 111234.2162,
+ "train_tokens_per_second": 5654.81
+ },
+ {
+ "epoch": 0.968556423377323,
+ "grad_norm": 0.723111666266317,
+ "learning_rate": 5.2473850463388675e-06,
+ "loss": 0.3943564176559448,
+ "num_input_tokens_seen": 629232256,
+ "step": 14385,
+ "train_runtime": 111274.7039,
+ "train_tokens_per_second": 5654.765
+ },
+ {
+ "epoch": 0.9688930783732831,
+ "grad_norm": 0.6809256630686321,
+ "learning_rate": 5.244744168022288e-06,
+ "loss": 0.3620326995849609,
+ "num_input_tokens_seen": 629455456,
+ "step": 14390,
+ "train_runtime": 111314.9521,
+ "train_tokens_per_second": 5654.725
+ },
+ {
+ "epoch": 0.9692297333692432,
+ "grad_norm": 0.6956920321641313,
+ "learning_rate": 5.24210322126378e-06,
+ "loss": 0.36865873336791993,
+ "num_input_tokens_seen": 629691848,
+ "step": 14395,
+ "train_runtime": 111353.3824,
+ "train_tokens_per_second": 5654.896
+ },
+ {
+ "epoch": 0.9695663883652034,
+ "grad_norm": 0.7152647788363251,
+ "learning_rate": 5.23946220680188e-06,
+ "loss": 0.3680101871490479,
+ "num_input_tokens_seen": 629928368,
+ "step": 14400,
+ "train_runtime": 111391.3207,
+ "train_tokens_per_second": 5655.094
+ },
+ {
+ "epoch": 0.9699030433611635,
+ "grad_norm": 0.7371345942960084,
+ "learning_rate": 5.236821125375136e-06,
+ "loss": 0.3584733486175537,
+ "num_input_tokens_seen": 630150304,
+ "step": 14405,
+ "train_runtime": 111429.6066,
+ "train_tokens_per_second": 5655.142
+ },
+ {
+ "epoch": 0.9702396983571236,
+ "grad_norm": 0.6525643955944114,
+ "learning_rate": 5.234179977722119e-06,
+ "loss": 0.31973896026611326,
+ "num_input_tokens_seen": 630389640,
+ "step": 14410,
+ "train_runtime": 111465.3624,
+ "train_tokens_per_second": 5655.476
+ },
+ {
+ "epoch": 0.9705763533530838,
+ "grad_norm": 0.6728032683654267,
+ "learning_rate": 5.23153876458142e-06,
+ "loss": 0.36993358135223386,
+ "num_input_tokens_seen": 630599464,
+ "step": 14415,
+ "train_runtime": 111509.2822,
+ "train_tokens_per_second": 5655.13
+ },
+ {
+ "epoch": 0.9709130083490439,
+ "grad_norm": 0.7435706382156235,
+ "learning_rate": 5.2288974866916435e-06,
+ "loss": 0.36161553859710693,
+ "num_input_tokens_seen": 630794912,
+ "step": 14420,
+ "train_runtime": 111547.8849,
+ "train_tokens_per_second": 5654.925
+ },
+ {
+ "epoch": 0.9712496633450041,
+ "grad_norm": 0.702712902975055,
+ "learning_rate": 5.226256144791414e-06,
+ "loss": 0.36026706695556643,
+ "num_input_tokens_seen": 631000992,
+ "step": 14425,
+ "train_runtime": 111582.5203,
+ "train_tokens_per_second": 5655.016
+ },
+ {
+ "epoch": 0.9715863183409642,
+ "grad_norm": 0.7340375348827854,
+ "learning_rate": 5.223614739619375e-06,
+ "loss": 0.37513973712921145,
+ "num_input_tokens_seen": 631202192,
+ "step": 14430,
+ "train_runtime": 111620.6231,
+ "train_tokens_per_second": 5654.889
+ },
+ {
+ "epoch": 0.9719229733369243,
+ "grad_norm": 0.6505020134915022,
+ "learning_rate": 5.2209732719141855e-06,
+ "loss": 0.33454012870788574,
+ "num_input_tokens_seen": 631411712,
+ "step": 14435,
+ "train_runtime": 111660.3902,
+ "train_tokens_per_second": 5654.751
+ },
+ {
+ "epoch": 0.9722596283328845,
+ "grad_norm": 0.6834948719980648,
+ "learning_rate": 5.218331742414528e-06,
+ "loss": 0.3455739736557007,
+ "num_input_tokens_seen": 631641904,
+ "step": 14440,
+ "train_runtime": 111694.5696,
+ "train_tokens_per_second": 5655.082
+ },
+ {
+ "epoch": 0.9725962833288446,
+ "grad_norm": 0.7031614396342304,
+ "learning_rate": 5.215690151859094e-06,
+ "loss": 0.3380829572677612,
+ "num_input_tokens_seen": 631848056,
+ "step": 14445,
+ "train_runtime": 111733.8193,
+ "train_tokens_per_second": 5654.94
+ },
+ {
+ "epoch": 0.9729329383248048,
+ "grad_norm": 0.737272924329504,
+ "learning_rate": 5.213048500986597e-06,
+ "loss": 0.3667490243911743,
+ "num_input_tokens_seen": 632068664,
+ "step": 14450,
+ "train_runtime": 111782.1877,
+ "train_tokens_per_second": 5654.467
+ },
+ {
+ "epoch": 0.9732695933207649,
+ "grad_norm": 0.8384823634165054,
+ "learning_rate": 5.210406790535766e-06,
+ "loss": 0.3663896083831787,
+ "num_input_tokens_seen": 632279544,
+ "step": 14455,
+ "train_runtime": 111817.1549,
+ "train_tokens_per_second": 5654.584
+ },
+ {
+ "epoch": 0.973606248316725,
+ "grad_norm": 0.7339723372461647,
+ "learning_rate": 5.207765021245346e-06,
+ "loss": 0.3660223722457886,
+ "num_input_tokens_seen": 632492984,
+ "step": 14460,
+ "train_runtime": 111858.5212,
+ "train_tokens_per_second": 5654.401
+ },
+ {
+ "epoch": 0.9739429033126852,
+ "grad_norm": 0.7329956341586107,
+ "learning_rate": 5.2051231938541015e-06,
+ "loss": 0.3708993911743164,
+ "num_input_tokens_seen": 632694568,
+ "step": 14465,
+ "train_runtime": 111894.0881,
+ "train_tokens_per_second": 5654.406
+ },
+ {
+ "epoch": 0.9742795583086453,
+ "grad_norm": 0.7751460530884079,
+ "learning_rate": 5.202481309100808e-06,
+ "loss": 0.3687734127044678,
+ "num_input_tokens_seen": 632904360,
+ "step": 14470,
+ "train_runtime": 111935.2636,
+ "train_tokens_per_second": 5654.2
+ },
+ {
+ "epoch": 0.9746162133046055,
+ "grad_norm": 0.6898579692907033,
+ "learning_rate": 5.199839367724266e-06,
+ "loss": 0.3596921920776367,
+ "num_input_tokens_seen": 633132800,
+ "step": 14475,
+ "train_runtime": 111976.0175,
+ "train_tokens_per_second": 5654.182
+ },
+ {
+ "epoch": 0.9749528683005656,
+ "grad_norm": 0.7400137611963173,
+ "learning_rate": 5.197197370463278e-06,
+ "loss": 0.343227481842041,
+ "num_input_tokens_seen": 633370648,
+ "step": 14480,
+ "train_runtime": 112017.4444,
+ "train_tokens_per_second": 5654.214
+ },
+ {
+ "epoch": 0.9752895232965257,
+ "grad_norm": 0.6958590027819616,
+ "learning_rate": 5.194555318056675e-06,
+ "loss": 0.39154577255249023,
+ "num_input_tokens_seen": 633605816,
+ "step": 14485,
+ "train_runtime": 112050.1996,
+ "train_tokens_per_second": 5654.66
+ },
+ {
+ "epoch": 0.9756261782924859,
+ "grad_norm": 0.9461003642017359,
+ "learning_rate": 5.191913211243299e-06,
+ "loss": 0.37716875076293943,
+ "num_input_tokens_seen": 633832456,
+ "step": 14490,
+ "train_runtime": 112095.5877,
+ "train_tokens_per_second": 5654.393
+ },
+ {
+ "epoch": 0.975962833288446,
+ "grad_norm": 0.7296195050735405,
+ "learning_rate": 5.1892710507620055e-06,
+ "loss": 0.36735081672668457,
+ "num_input_tokens_seen": 634052208,
+ "step": 14495,
+ "train_runtime": 112132.123,
+ "train_tokens_per_second": 5654.51
+ },
+ {
+ "epoch": 0.9762994882844062,
+ "grad_norm": 0.713237266588106,
+ "learning_rate": 5.186628837351666e-06,
+ "loss": 0.3544626235961914,
+ "num_input_tokens_seen": 634282120,
+ "step": 14500,
+ "train_runtime": 112168.6707,
+ "train_tokens_per_second": 5654.717
+ },
+ {
+ "epoch": 0.9766361432803663,
+ "grad_norm": 0.7152784729674379,
+ "learning_rate": 5.183986571751167e-06,
+ "loss": 0.3681643962860107,
+ "num_input_tokens_seen": 634503040,
+ "step": 14505,
+ "train_runtime": 112208.6433,
+ "train_tokens_per_second": 5654.672
+ },
+ {
+ "epoch": 0.9769727982763264,
+ "grad_norm": 0.7143902821508025,
+ "learning_rate": 5.181344254699408e-06,
+ "loss": 0.384121036529541,
+ "num_input_tokens_seen": 634718224,
+ "step": 14510,
+ "train_runtime": 112246.9754,
+ "train_tokens_per_second": 5654.658
+ },
+ {
+ "epoch": 0.9773094532722866,
+ "grad_norm": 0.7563338962960594,
+ "learning_rate": 5.178701886935309e-06,
+ "loss": 0.3466785907745361,
+ "num_input_tokens_seen": 634926976,
+ "step": 14515,
+ "train_runtime": 112283.3188,
+ "train_tokens_per_second": 5654.687
+ },
+ {
+ "epoch": 0.9776461082682467,
+ "grad_norm": 0.6709897911941165,
+ "learning_rate": 5.176059469197796e-06,
+ "loss": 0.3563254356384277,
+ "num_input_tokens_seen": 635147104,
+ "step": 14520,
+ "train_runtime": 112318.4084,
+ "train_tokens_per_second": 5654.88
+ },
+ {
+ "epoch": 0.9779827632642069,
+ "grad_norm": 0.84998527536875,
+ "learning_rate": 5.173417002225815e-06,
+ "loss": 0.3569037914276123,
+ "num_input_tokens_seen": 635366792,
+ "step": 14525,
+ "train_runtime": 112357.718,
+ "train_tokens_per_second": 5654.857
+ },
+ {
+ "epoch": 0.978319418260167,
+ "grad_norm": 0.7892134632795219,
+ "learning_rate": 5.170774486758322e-06,
+ "loss": 0.36590094566345216,
+ "num_input_tokens_seen": 635560568,
+ "step": 14530,
+ "train_runtime": 112393.8137,
+ "train_tokens_per_second": 5654.765
+ },
+ {
+ "epoch": 0.9786560732561271,
+ "grad_norm": 0.6842756660337252,
+ "learning_rate": 5.168131923534287e-06,
+ "loss": 0.38362579345703124,
+ "num_input_tokens_seen": 635787144,
+ "step": 14535,
+ "train_runtime": 112428.7709,
+ "train_tokens_per_second": 5655.022
+ },
+ {
+ "epoch": 0.9789927282520873,
+ "grad_norm": 0.6691518942357926,
+ "learning_rate": 5.165489313292696e-06,
+ "loss": 0.35508084297180176,
+ "num_input_tokens_seen": 636000120,
+ "step": 14540,
+ "train_runtime": 112469.9454,
+ "train_tokens_per_second": 5654.845
+ },
+ {
+ "epoch": 0.9793293832480474,
+ "grad_norm": 0.7564357234909826,
+ "learning_rate": 5.162846656772548e-06,
+ "loss": 0.35072224140167235,
+ "num_input_tokens_seen": 636214280,
+ "step": 14545,
+ "train_runtime": 112512.921,
+ "train_tokens_per_second": 5654.589
+ },
+ {
+ "epoch": 0.9796660382440076,
+ "grad_norm": 0.7450101360653049,
+ "learning_rate": 5.16020395471285e-06,
+ "loss": 0.3438509464263916,
+ "num_input_tokens_seen": 636435816,
+ "step": 14550,
+ "train_runtime": 112554.2979,
+ "train_tokens_per_second": 5654.478
+ },
+ {
+ "epoch": 0.9800026932399677,
+ "grad_norm": 0.8137120525286944,
+ "learning_rate": 5.157561207852628e-06,
+ "loss": 0.3831623077392578,
+ "num_input_tokens_seen": 636640224,
+ "step": 14555,
+ "train_runtime": 112591.2086,
+ "train_tokens_per_second": 5654.44
+ },
+ {
+ "epoch": 0.9803393482359278,
+ "grad_norm": 0.7297592438951429,
+ "learning_rate": 5.154918416930915e-06,
+ "loss": 0.3411484479904175,
+ "num_input_tokens_seen": 636858520,
+ "step": 14560,
+ "train_runtime": 112629.598,
+ "train_tokens_per_second": 5654.451
+ },
+ {
+ "epoch": 0.980676003231888,
+ "grad_norm": 0.760019959304734,
+ "learning_rate": 5.152275582686761e-06,
+ "loss": 0.3839298486709595,
+ "num_input_tokens_seen": 637073856,
+ "step": 14565,
+ "train_runtime": 112666.3482,
+ "train_tokens_per_second": 5654.518
+ },
+ {
+ "epoch": 0.9810126582278481,
+ "grad_norm": 0.6959300895402221,
+ "learning_rate": 5.149632705859227e-06,
+ "loss": 0.36935174465179443,
+ "num_input_tokens_seen": 637299440,
+ "step": 14570,
+ "train_runtime": 112707.9899,
+ "train_tokens_per_second": 5654.43
+ },
+ {
+ "epoch": 0.9813493132238083,
+ "grad_norm": 0.7586977280516466,
+ "learning_rate": 5.146989787187383e-06,
+ "loss": 0.37570605278015134,
+ "num_input_tokens_seen": 637529440,
+ "step": 14575,
+ "train_runtime": 112743.4679,
+ "train_tokens_per_second": 5654.691
+ },
+ {
+ "epoch": 0.9816859682197684,
+ "grad_norm": 0.6825264977223254,
+ "learning_rate": 5.144346827410314e-06,
+ "loss": 0.3703890562057495,
+ "num_input_tokens_seen": 637754408,
+ "step": 14580,
+ "train_runtime": 112773.1649,
+ "train_tokens_per_second": 5655.197
+ },
+ {
+ "epoch": 0.9820226232157285,
+ "grad_norm": 0.826290290377258,
+ "learning_rate": 5.141703827267117e-06,
+ "loss": 0.3702056646347046,
+ "num_input_tokens_seen": 637982336,
+ "step": 14585,
+ "train_runtime": 112804.5446,
+ "train_tokens_per_second": 5655.644
+ },
+ {
+ "epoch": 0.9823592782116887,
+ "grad_norm": 0.7212214874544945,
+ "learning_rate": 5.139060787496893e-06,
+ "loss": 0.3754692077636719,
+ "num_input_tokens_seen": 638203872,
+ "step": 14590,
+ "train_runtime": 112841.7495,
+ "train_tokens_per_second": 5655.742
+ },
+ {
+ "epoch": 0.9826959332076488,
+ "grad_norm": 0.6942239873292527,
+ "learning_rate": 5.136417708838766e-06,
+ "loss": 0.3775701284408569,
+ "num_input_tokens_seen": 638407968,
+ "step": 14595,
+ "train_runtime": 112885.1355,
+ "train_tokens_per_second": 5655.377
+ },
+ {
+ "epoch": 0.983032588203609,
+ "grad_norm": 0.7575191844718635,
+ "learning_rate": 5.13377459203186e-06,
+ "loss": 0.3628222465515137,
+ "num_input_tokens_seen": 638616040,
+ "step": 14600,
+ "train_runtime": 112923.0341,
+ "train_tokens_per_second": 5655.321
+ },
+ {
+ "epoch": 0.9833692431995691,
+ "grad_norm": 0.7100885514863359,
+ "learning_rate": 5.131131437815318e-06,
+ "loss": 0.3371574401855469,
+ "num_input_tokens_seen": 638837176,
+ "step": 14605,
+ "train_runtime": 112963.1632,
+ "train_tokens_per_second": 5655.27
+ },
+ {
+ "epoch": 0.9837058981955292,
+ "grad_norm": 0.7208738472250812,
+ "learning_rate": 5.128488246928287e-06,
+ "loss": 0.36636881828308104,
+ "num_input_tokens_seen": 639065552,
+ "step": 14610,
+ "train_runtime": 113000.1458,
+ "train_tokens_per_second": 5655.44
+ },
+ {
+ "epoch": 0.9840425531914894,
+ "grad_norm": 0.7630145727500303,
+ "learning_rate": 5.125845020109925e-06,
+ "loss": 0.37039480209350584,
+ "num_input_tokens_seen": 639285272,
+ "step": 14615,
+ "train_runtime": 113047.3523,
+ "train_tokens_per_second": 5655.022
+ },
+ {
+ "epoch": 0.9843792081874495,
+ "grad_norm": 0.7720868800544706,
+ "learning_rate": 5.123201758099406e-06,
+ "loss": 0.36545164585113527,
+ "num_input_tokens_seen": 639499208,
+ "step": 14620,
+ "train_runtime": 113085.6686,
+ "train_tokens_per_second": 5654.998
+ },
+ {
+ "epoch": 0.9847158631834096,
+ "grad_norm": 0.7035125490360992,
+ "learning_rate": 5.12055846163591e-06,
+ "loss": 0.37010979652404785,
+ "num_input_tokens_seen": 639714312,
+ "step": 14625,
+ "train_runtime": 113123.7227,
+ "train_tokens_per_second": 5654.997
+ },
+ {
+ "epoch": 0.9850525181793698,
+ "grad_norm": 0.6519589735545973,
+ "learning_rate": 5.117915131458623e-06,
+ "loss": 0.3291551828384399,
+ "num_input_tokens_seen": 639938488,
+ "step": 14630,
+ "train_runtime": 113162.111,
+ "train_tokens_per_second": 5655.06
+ },
+ {
+ "epoch": 0.9853891731753299,
+ "grad_norm": 0.8145405370454951,
+ "learning_rate": 5.115271768306745e-06,
+ "loss": 0.36716470718383787,
+ "num_input_tokens_seen": 640152840,
+ "step": 14635,
+ "train_runtime": 113199.5845,
+ "train_tokens_per_second": 5655.081
+ },
+ {
+ "epoch": 0.9857258281712901,
+ "grad_norm": 0.790933823693933,
+ "learning_rate": 5.112628372919485e-06,
+ "loss": 0.3815732717514038,
+ "num_input_tokens_seen": 640369952,
+ "step": 14640,
+ "train_runtime": 113234.1823,
+ "train_tokens_per_second": 5655.271
+ },
+ {
+ "epoch": 0.9860624831672502,
+ "grad_norm": 0.9699782510514777,
+ "learning_rate": 5.1099849460360595e-06,
+ "loss": 0.34613332748413084,
+ "num_input_tokens_seen": 640567240,
+ "step": 14645,
+ "train_runtime": 113271.7361,
+ "train_tokens_per_second": 5655.137
+ },
+ {
+ "epoch": 0.9863991381632103,
+ "grad_norm": 0.6723948378020855,
+ "learning_rate": 5.107341488395695e-06,
+ "loss": 0.3379997730255127,
+ "num_input_tokens_seen": 640788384,
+ "step": 14650,
+ "train_runtime": 113310.4257,
+ "train_tokens_per_second": 5655.158
+ },
+ {
+ "epoch": 0.9867357931591705,
+ "grad_norm": 0.7000157996530737,
+ "learning_rate": 5.104698000737626e-06,
+ "loss": 0.37780165672302246,
+ "num_input_tokens_seen": 641011792,
+ "step": 14655,
+ "train_runtime": 113351.4983,
+ "train_tokens_per_second": 5655.08
+ },
+ {
+ "epoch": 0.9870724481551306,
+ "grad_norm": 0.7543705648632995,
+ "learning_rate": 5.102054483801094e-06,
+ "loss": 0.38533506393432615,
+ "num_input_tokens_seen": 641239984,
+ "step": 14660,
+ "train_runtime": 113390.3486,
+ "train_tokens_per_second": 5655.155
+ },
+ {
+ "epoch": 0.9874091031510908,
+ "grad_norm": 0.7193117605335543,
+ "learning_rate": 5.099410938325351e-06,
+ "loss": 0.3746023178100586,
+ "num_input_tokens_seen": 641468200,
+ "step": 14665,
+ "train_runtime": 113424.7218,
+ "train_tokens_per_second": 5655.453
+ },
+ {
+ "epoch": 0.9877457581470509,
+ "grad_norm": 0.8146622438199124,
+ "learning_rate": 5.096767365049657e-06,
+ "loss": 0.3502329111099243,
+ "num_input_tokens_seen": 641672792,
+ "step": 14670,
+ "train_runtime": 113464.964,
+ "train_tokens_per_second": 5655.25
+ },
+ {
+ "epoch": 0.988082413143011,
+ "grad_norm": 0.6875126263724606,
+ "learning_rate": 5.094123764713277e-06,
+ "loss": 0.3421436309814453,
+ "num_input_tokens_seen": 641904736,
+ "step": 14675,
+ "train_runtime": 113500.904,
+ "train_tokens_per_second": 5655.503
+ },
+ {
+ "epoch": 0.9884190681389712,
+ "grad_norm": 0.7258139194348168,
+ "learning_rate": 5.091480138055487e-06,
+ "loss": 0.36416258811950686,
+ "num_input_tokens_seen": 642136592,
+ "step": 14680,
+ "train_runtime": 113531.8425,
+ "train_tokens_per_second": 5656.004
+ },
+ {
+ "epoch": 0.9887557231349313,
+ "grad_norm": 0.7250476516557031,
+ "learning_rate": 5.088836485815568e-06,
+ "loss": 0.35428576469421386,
+ "num_input_tokens_seen": 642334648,
+ "step": 14685,
+ "train_runtime": 113566.8784,
+ "train_tokens_per_second": 5656.003
+ },
+ {
+ "epoch": 0.9890923781308915,
+ "grad_norm": 0.7385563837095339,
+ "learning_rate": 5.086192808732809e-06,
+ "loss": 0.35656602382659913,
+ "num_input_tokens_seen": 642538480,
+ "step": 14690,
+ "train_runtime": 113608.5759,
+ "train_tokens_per_second": 5655.722
+ },
+ {
+ "epoch": 0.9894290331268516,
+ "grad_norm": 0.7946980704838973,
+ "learning_rate": 5.083549107546505e-06,
+ "loss": 0.36766040325164795,
+ "num_input_tokens_seen": 642744792,
+ "step": 14695,
+ "train_runtime": 113647.6019,
+ "train_tokens_per_second": 5655.595
+ },
+ {
+ "epoch": 0.9897656881228117,
+ "grad_norm": 0.771038431226611,
+ "learning_rate": 5.080905382995961e-06,
+ "loss": 0.38895292282104493,
+ "num_input_tokens_seen": 642974608,
+ "step": 14700,
+ "train_runtime": 113683.9469,
+ "train_tokens_per_second": 5655.808
+ },
+ {
+ "epoch": 0.9901023431187719,
+ "grad_norm": 0.7343676499490034,
+ "learning_rate": 5.0782616358204835e-06,
+ "loss": 0.37378993034362795,
+ "num_input_tokens_seen": 643191928,
+ "step": 14705,
+ "train_runtime": 113723.5317,
+ "train_tokens_per_second": 5655.751
+ },
+ {
+ "epoch": 0.990438998114732,
+ "grad_norm": 0.8524692289540274,
+ "learning_rate": 5.075617866759389e-06,
+ "loss": 0.33492867946624755,
+ "num_input_tokens_seen": 643403376,
+ "step": 14710,
+ "train_runtime": 113766.092,
+ "train_tokens_per_second": 5655.493
+ },
+ {
+ "epoch": 0.9907756531106922,
+ "grad_norm": 0.7901150595595263,
+ "learning_rate": 5.0729740765519995e-06,
+ "loss": 0.4053798675537109,
+ "num_input_tokens_seen": 643617760,
+ "step": 14715,
+ "train_runtime": 113802.6369,
+ "train_tokens_per_second": 5655.561
+ },
+ {
+ "epoch": 0.9911123081066523,
+ "grad_norm": 0.7850703195564729,
+ "learning_rate": 5.070330265937643e-06,
+ "loss": 0.3686873435974121,
+ "num_input_tokens_seen": 643827376,
+ "step": 14720,
+ "train_runtime": 113838.4716,
+ "train_tokens_per_second": 5655.622
+ },
+ {
+ "epoch": 0.9914489631026124,
+ "grad_norm": 0.7339072075707702,
+ "learning_rate": 5.067686435655651e-06,
+ "loss": 0.3935819625854492,
+ "num_input_tokens_seen": 644051992,
+ "step": 14725,
+ "train_runtime": 113875.0085,
+ "train_tokens_per_second": 5655.78
+ },
+ {
+ "epoch": 0.9917856180985726,
+ "grad_norm": 0.7717913369926017,
+ "learning_rate": 5.065042586445363e-06,
+ "loss": 0.39838461875915526,
+ "num_input_tokens_seen": 644255512,
+ "step": 14730,
+ "train_runtime": 113907.4094,
+ "train_tokens_per_second": 5655.958
+ },
+ {
+ "epoch": 0.9921222730945327,
+ "grad_norm": 0.7852173126276834,
+ "learning_rate": 5.0623987190461245e-06,
+ "loss": 0.30464193820953367,
+ "num_input_tokens_seen": 644460824,
+ "step": 14735,
+ "train_runtime": 113944.9392,
+ "train_tokens_per_second": 5655.897
+ },
+ {
+ "epoch": 0.9924589280904929,
+ "grad_norm": 0.7913343599249734,
+ "learning_rate": 5.059754834197281e-06,
+ "loss": 0.3790989875793457,
+ "num_input_tokens_seen": 644675152,
+ "step": 14740,
+ "train_runtime": 113988.6481,
+ "train_tokens_per_second": 5655.608
+ },
+ {
+ "epoch": 0.992795583086453,
+ "grad_norm": 0.6313018834314664,
+ "learning_rate": 5.05711093263819e-06,
+ "loss": 0.32619643211364746,
+ "num_input_tokens_seen": 644890856,
+ "step": 14745,
+ "train_runtime": 114026.2898,
+ "train_tokens_per_second": 5655.633
+ },
+ {
+ "epoch": 0.9931322380824131,
+ "grad_norm": 0.7700737076224284,
+ "learning_rate": 5.054467015108208e-06,
+ "loss": 0.37743263244628905,
+ "num_input_tokens_seen": 645124328,
+ "step": 14750,
+ "train_runtime": 114061.9307,
+ "train_tokens_per_second": 5655.913
+ },
+ {
+ "epoch": 0.9934688930783733,
+ "grad_norm": 0.8548358470170475,
+ "learning_rate": 5.051823082346701e-06,
+ "loss": 0.37684605121612547,
+ "num_input_tokens_seen": 645342896,
+ "step": 14755,
+ "train_runtime": 114095.225,
+ "train_tokens_per_second": 5656.178
+ },
+ {
+ "epoch": 0.9938055480743334,
+ "grad_norm": 0.7487000463516748,
+ "learning_rate": 5.049179135093033e-06,
+ "loss": 0.3667085409164429,
+ "num_input_tokens_seen": 645572616,
+ "step": 14760,
+ "train_runtime": 114128.9739,
+ "train_tokens_per_second": 5656.518
+ },
+ {
+ "epoch": 0.9941422030702936,
+ "grad_norm": 0.7554586359729867,
+ "learning_rate": 5.046535174086575e-06,
+ "loss": 0.37410502433776854,
+ "num_input_tokens_seen": 645784112,
+ "step": 14765,
+ "train_runtime": 114171.4105,
+ "train_tokens_per_second": 5656.268
+ },
+ {
+ "epoch": 0.9944788580662537,
+ "grad_norm": 0.806201449115856,
+ "learning_rate": 5.043891200066706e-06,
+ "loss": 0.35855536460876464,
+ "num_input_tokens_seen": 646000480,
+ "step": 14770,
+ "train_runtime": 114211.5644,
+ "train_tokens_per_second": 5656.174
+ },
+ {
+ "epoch": 0.9948155130622138,
+ "grad_norm": 0.7970070550101245,
+ "learning_rate": 5.041247213772802e-06,
+ "loss": 0.3647798538208008,
+ "num_input_tokens_seen": 646228928,
+ "step": 14775,
+ "train_runtime": 114248.8453,
+ "train_tokens_per_second": 5656.328
+ },
+ {
+ "epoch": 0.995152168058174,
+ "grad_norm": 0.9274665922453337,
+ "learning_rate": 5.038603215944247e-06,
+ "loss": 0.3697342872619629,
+ "num_input_tokens_seen": 646427096,
+ "step": 14780,
+ "train_runtime": 114283.9377,
+ "train_tokens_per_second": 5656.325
+ },
+ {
+ "epoch": 0.9954888230541341,
+ "grad_norm": 0.7200893083392556,
+ "learning_rate": 5.0359592073204254e-06,
+ "loss": 0.34227933883666994,
+ "num_input_tokens_seen": 646643272,
+ "step": 14785,
+ "train_runtime": 114327.3802,
+ "train_tokens_per_second": 5656.067
+ },
+ {
+ "epoch": 0.9958254780500942,
+ "grad_norm": 1.0261926014877596,
+ "learning_rate": 5.033315188640725e-06,
+ "loss": 0.3627659797668457,
+ "num_input_tokens_seen": 646872200,
+ "step": 14790,
+ "train_runtime": 114366.7864,
+ "train_tokens_per_second": 5656.119
+ },
+ {
+ "epoch": 0.9961621330460544,
+ "grad_norm": 0.6681278213263397,
+ "learning_rate": 5.03067116064454e-06,
+ "loss": 0.3450484752655029,
+ "num_input_tokens_seen": 647094832,
+ "step": 14795,
+ "train_runtime": 114408.9697,
+ "train_tokens_per_second": 5655.98
+ },
+ {
+ "epoch": 0.9964987880420145,
+ "grad_norm": 0.6613774299958016,
+ "learning_rate": 5.028027124071262e-06,
+ "loss": 0.36185712814331056,
+ "num_input_tokens_seen": 647301568,
+ "step": 14800,
+ "train_runtime": 114449.8831,
+ "train_tokens_per_second": 5655.764
+ },
+ {
+ "epoch": 0.9968354430379747,
+ "grad_norm": 0.7005897120089515,
+ "learning_rate": 5.025383079660289e-06,
+ "loss": 0.3468491554260254,
+ "num_input_tokens_seen": 647531264,
+ "step": 14805,
+ "train_runtime": 114488.5353,
+ "train_tokens_per_second": 5655.861
+ },
+ {
+ "epoch": 0.9971720980339348,
+ "grad_norm": 0.7061938748348358,
+ "learning_rate": 5.0227390281510166e-06,
+ "loss": 0.3663314342498779,
+ "num_input_tokens_seen": 647753976,
+ "step": 14810,
+ "train_runtime": 114526.0357,
+ "train_tokens_per_second": 5655.954
+ },
+ {
+ "epoch": 0.997508753029895,
+ "grad_norm": 0.7212945140660744,
+ "learning_rate": 5.020094970282849e-06,
+ "loss": 0.3663675308227539,
+ "num_input_tokens_seen": 647955768,
+ "step": 14815,
+ "train_runtime": 114566.6744,
+ "train_tokens_per_second": 5655.709
+ },
+ {
+ "epoch": 0.9978454080258551,
+ "grad_norm": 0.7579147457391882,
+ "learning_rate": 5.017450906795187e-06,
+ "loss": 0.36784412860870364,
+ "num_input_tokens_seen": 648181384,
+ "step": 14820,
+ "train_runtime": 114601.4001,
+ "train_tokens_per_second": 5655.964
+ },
+ {
+ "epoch": 0.9981820630218152,
+ "grad_norm": 0.7496488759479608,
+ "learning_rate": 5.014806838427434e-06,
+ "loss": 0.36796035766601565,
+ "num_input_tokens_seen": 648382672,
+ "step": 14825,
+ "train_runtime": 114646.9011,
+ "train_tokens_per_second": 5655.475
+ },
+ {
+ "epoch": 0.9985187180177754,
+ "grad_norm": 0.6851905965880742,
+ "learning_rate": 5.0121627659189964e-06,
+ "loss": 0.3591431140899658,
+ "num_input_tokens_seen": 648612168,
+ "step": 14830,
+ "train_runtime": 114682.2822,
+ "train_tokens_per_second": 5655.731
+ },
+ {
+ "epoch": 0.9988553730137355,
+ "grad_norm": 0.6696970669236173,
+ "learning_rate": 5.009518690009281e-06,
+ "loss": 0.3566739082336426,
+ "num_input_tokens_seen": 648821208,
+ "step": 14835,
+ "train_runtime": 114724.3707,
+ "train_tokens_per_second": 5655.478
+ },
+ {
+ "epoch": 0.9991920280096956,
+ "grad_norm": 0.7614065048649256,
+ "learning_rate": 5.006874611437692e-06,
+ "loss": 0.3776280879974365,
+ "num_input_tokens_seen": 649046288,
+ "step": 14840,
+ "train_runtime": 114764.8122,
+ "train_tokens_per_second": 5655.447
+ },
+ {
+ "epoch": 0.9995286830056558,
+ "grad_norm": 0.6492499944472357,
+ "learning_rate": 5.004230530943641e-06,
+ "loss": 0.35755281448364257,
+ "num_input_tokens_seen": 649262232,
+ "step": 14845,
+ "train_runtime": 114804.5614,
+ "train_tokens_per_second": 5655.37
+ },
+ {
+ "epoch": 0.9998653380016159,
+ "grad_norm": 0.7184203567667401,
+ "learning_rate": 5.001586449266536e-06,
+ "loss": 0.3260646343231201,
+ "num_input_tokens_seen": 649487024,
+ "step": 14850,
+ "train_runtime": 114845.5757,
+ "train_tokens_per_second": 5655.307
+ },
+ {
+ "epoch": 1.000201992997576,
+ "grad_norm": 0.6358891287683051,
+ "learning_rate": 4.998942367145787e-06,
+ "loss": 0.35132179260253904,
+ "num_input_tokens_seen": 649706888,
+ "step": 14855,
+ "train_runtime": 114929.2136,
+ "train_tokens_per_second": 5653.105
+ },
+ {
+ "epoch": 1.0005386479935363,
+ "grad_norm": 0.7034269548859629,
+ "learning_rate": 4.996298285320798e-06,
+ "loss": 0.30704555511474607,
+ "num_input_tokens_seen": 649911248,
+ "step": 14860,
+ "train_runtime": 114965.2665,
+ "train_tokens_per_second": 5653.11
+ },
+ {
+ "epoch": 1.0008753029894963,
+ "grad_norm": 0.7771207057358878,
+ "learning_rate": 4.993654204530983e-06,
+ "loss": 0.2908186435699463,
+ "num_input_tokens_seen": 650136336,
+ "step": 14865,
+ "train_runtime": 115009.0891,
+ "train_tokens_per_second": 5652.913
+ },
+ {
+ "epoch": 1.0012119579854566,
+ "grad_norm": 0.7800505661917783,
+ "learning_rate": 4.99101012551575e-06,
+ "loss": 0.3061246633529663,
+ "num_input_tokens_seen": 650324944,
+ "step": 14870,
+ "train_runtime": 115047.7835,
+ "train_tokens_per_second": 5652.651
+ },
+ {
+ "epoch": 1.0015486129814166,
+ "grad_norm": 0.7451144290465576,
+ "learning_rate": 4.9883660490145075e-06,
+ "loss": 0.31706483364105226,
+ "num_input_tokens_seen": 650556192,
+ "step": 14875,
+ "train_runtime": 115083.1511,
+ "train_tokens_per_second": 5652.923
+ },
+ {
+ "epoch": 1.0018852679773769,
+ "grad_norm": 0.6924866106934799,
+ "learning_rate": 4.985721975766661e-06,
+ "loss": 0.32414817810058594,
+ "num_input_tokens_seen": 650775960,
+ "step": 14880,
+ "train_runtime": 115118.3264,
+ "train_tokens_per_second": 5653.105
+ },
+ {
+ "epoch": 1.002221922973337,
+ "grad_norm": 1.0051470668362659,
+ "learning_rate": 4.983077906511619e-06,
+ "loss": 0.32347331047058103,
+ "num_input_tokens_seen": 650983240,
+ "step": 14885,
+ "train_runtime": 115161.3444,
+ "train_tokens_per_second": 5652.793
+ },
+ {
+ "epoch": 1.0025585779692971,
+ "grad_norm": 0.7329517809959321,
+ "learning_rate": 4.980433841988786e-06,
+ "loss": 0.30762290954589844,
+ "num_input_tokens_seen": 651204568,
+ "step": 14890,
+ "train_runtime": 115197.7799,
+ "train_tokens_per_second": 5652.926
+ },
+ {
+ "epoch": 1.0028952329652572,
+ "grad_norm": 0.7126218012313592,
+ "learning_rate": 4.977789782937568e-06,
+ "loss": 0.3130316257476807,
+ "num_input_tokens_seen": 651431064,
+ "step": 14895,
+ "train_runtime": 115231.9768,
+ "train_tokens_per_second": 5653.214
+ },
+ {
+ "epoch": 1.0032318879612174,
+ "grad_norm": 0.7709912103285202,
+ "learning_rate": 4.975145730097364e-06,
+ "loss": 0.3236842155456543,
+ "num_input_tokens_seen": 651636984,
+ "step": 14900,
+ "train_runtime": 115269.0999,
+ "train_tokens_per_second": 5653.18
+ },
+ {
+ "epoch": 1.0035685429571775,
+ "grad_norm": 0.6705219562784165,
+ "learning_rate": 4.97250168420758e-06,
+ "loss": 0.3103510618209839,
+ "num_input_tokens_seen": 651849816,
+ "step": 14905,
+ "train_runtime": 115304.0536,
+ "train_tokens_per_second": 5653.312
+ },
+ {
+ "epoch": 1.0039051979531377,
+ "grad_norm": 0.737858641027806,
+ "learning_rate": 4.96985764600761e-06,
+ "loss": 0.3144733428955078,
+ "num_input_tokens_seen": 652078544,
+ "step": 14910,
+ "train_runtime": 115343.8311,
+ "train_tokens_per_second": 5653.346
+ },
+ {
+ "epoch": 1.0042418529490977,
+ "grad_norm": 0.7152864263979617,
+ "learning_rate": 4.967213616236855e-06,
+ "loss": 0.32990431785583496,
+ "num_input_tokens_seen": 652288816,
+ "step": 14915,
+ "train_runtime": 115379.8129,
+ "train_tokens_per_second": 5653.405
+ },
+ {
+ "epoch": 1.004578507945058,
+ "grad_norm": 0.750830922325084,
+ "learning_rate": 4.964569595634706e-06,
+ "loss": 0.3154884338378906,
+ "num_input_tokens_seen": 652509832,
+ "step": 14920,
+ "train_runtime": 115422.4071,
+ "train_tokens_per_second": 5653.234
+ },
+ {
+ "epoch": 1.004915162941018,
+ "grad_norm": 0.7045131043182794,
+ "learning_rate": 4.961925584940559e-06,
+ "loss": 0.3344886779785156,
+ "num_input_tokens_seen": 652737176,
+ "step": 14925,
+ "train_runtime": 115469.7957,
+ "train_tokens_per_second": 5652.882
+ },
+ {
+ "epoch": 1.0052518179369783,
+ "grad_norm": 0.7769156856429779,
+ "learning_rate": 4.959281584893799e-06,
+ "loss": 0.28301913738250734,
+ "num_input_tokens_seen": 652949264,
+ "step": 14930,
+ "train_runtime": 115507.9351,
+ "train_tokens_per_second": 5652.852
+ },
+ {
+ "epoch": 1.0055884729329383,
+ "grad_norm": 0.7698614993301071,
+ "learning_rate": 4.956637596233814e-06,
+ "loss": 0.3222982883453369,
+ "num_input_tokens_seen": 653149432,
+ "step": 14935,
+ "train_runtime": 115542.574,
+ "train_tokens_per_second": 5652.89
+ },
+ {
+ "epoch": 1.0059251279288985,
+ "grad_norm": 0.7927923790873723,
+ "learning_rate": 4.953993619699986e-06,
+ "loss": 0.29294121265411377,
+ "num_input_tokens_seen": 653371256,
+ "step": 14940,
+ "train_runtime": 115587.1554,
+ "train_tokens_per_second": 5652.629
+ },
+ {
+ "epoch": 1.0062617829248586,
+ "grad_norm": 0.7524347822729336,
+ "learning_rate": 4.951349656031698e-06,
+ "loss": 0.30971436500549315,
+ "num_input_tokens_seen": 653582728,
+ "step": 14945,
+ "train_runtime": 115628.4503,
+ "train_tokens_per_second": 5652.439
+ },
+ {
+ "epoch": 1.0065984379208188,
+ "grad_norm": 0.7572597064872421,
+ "learning_rate": 4.948705705968322e-06,
+ "loss": 0.3157053470611572,
+ "num_input_tokens_seen": 653808728,
+ "step": 14950,
+ "train_runtime": 115667.4033,
+ "train_tokens_per_second": 5652.489
+ },
+ {
+ "epoch": 1.0069350929167789,
+ "grad_norm": 0.7025308350419313,
+ "learning_rate": 4.946061770249233e-06,
+ "loss": 0.31903715133666993,
+ "num_input_tokens_seen": 654028536,
+ "step": 14955,
+ "train_runtime": 115709.3504,
+ "train_tokens_per_second": 5652.34
+ },
+ {
+ "epoch": 1.007271747912739,
+ "grad_norm": 0.746389149349819,
+ "learning_rate": 4.9434178496137955e-06,
+ "loss": 0.3309792518615723,
+ "num_input_tokens_seen": 654245296,
+ "step": 14960,
+ "train_runtime": 115746.6069,
+ "train_tokens_per_second": 5652.393
+ },
+ {
+ "epoch": 1.0076084029086991,
+ "grad_norm": 0.6143946299920816,
+ "learning_rate": 4.940773944801378e-06,
+ "loss": 0.31335494518280027,
+ "num_input_tokens_seen": 654464232,
+ "step": 14965,
+ "train_runtime": 115779.8261,
+ "train_tokens_per_second": 5652.662
+ },
+ {
+ "epoch": 1.0079450579046594,
+ "grad_norm": 0.8846423078456821,
+ "learning_rate": 4.938130056551337e-06,
+ "loss": 0.3319870948791504,
+ "num_input_tokens_seen": 654675824,
+ "step": 14970,
+ "train_runtime": 115819.5601,
+ "train_tokens_per_second": 5652.55
+ },
+ {
+ "epoch": 1.0082817129006194,
+ "grad_norm": 0.7105076447650682,
+ "learning_rate": 4.9354861856030305e-06,
+ "loss": 0.3344868183135986,
+ "num_input_tokens_seen": 654897216,
+ "step": 14975,
+ "train_runtime": 115859.9373,
+ "train_tokens_per_second": 5652.491
+ },
+ {
+ "epoch": 1.0086183678965797,
+ "grad_norm": 0.7528024963191167,
+ "learning_rate": 4.932842332695803e-06,
+ "loss": 0.28916707038879397,
+ "num_input_tokens_seen": 655099488,
+ "step": 14980,
+ "train_runtime": 115894.1166,
+ "train_tokens_per_second": 5652.569
+ },
+ {
+ "epoch": 1.0089550228925397,
+ "grad_norm": 0.6760065554104225,
+ "learning_rate": 4.9301984985690055e-06,
+ "loss": 0.31049747467041017,
+ "num_input_tokens_seen": 655334416,
+ "step": 14985,
+ "train_runtime": 115935.8108,
+ "train_tokens_per_second": 5652.562
+ },
+ {
+ "epoch": 1.0092916778885,
+ "grad_norm": 0.8434100907552518,
+ "learning_rate": 4.927554683961974e-06,
+ "loss": 0.3354443073272705,
+ "num_input_tokens_seen": 655543112,
+ "step": 14990,
+ "train_runtime": 115970.966,
+ "train_tokens_per_second": 5652.649
+ },
+ {
+ "epoch": 1.00962833288446,
+ "grad_norm": 0.7283088356363585,
+ "learning_rate": 4.924910889614046e-06,
+ "loss": 0.3087354898452759,
+ "num_input_tokens_seen": 655763808,
+ "step": 14995,
+ "train_runtime": 116009.5582,
+ "train_tokens_per_second": 5652.671
+ },
+ {
+ "epoch": 1.0099649878804202,
+ "grad_norm": 0.7618425432790579,
+ "learning_rate": 4.922267116264546e-06,
+ "loss": 0.33311076164245607,
+ "num_input_tokens_seen": 655973824,
+ "step": 15000,
+ "train_runtime": 116046.9467,
+ "train_tokens_per_second": 5652.659
+ },
+ {
+ "epoch": 1.0103016428763802,
+ "grad_norm": 0.8482961991460483,
+ "learning_rate": 4.919623364652801e-06,
+ "loss": 0.3059119462966919,
+ "num_input_tokens_seen": 656189544,
+ "step": 15005,
+ "train_runtime": 116087.885,
+ "train_tokens_per_second": 5652.524
+ },
+ {
+ "epoch": 1.0106382978723405,
+ "grad_norm": 0.7996774528917011,
+ "learning_rate": 4.916979635518124e-06,
+ "loss": 0.31961765289306643,
+ "num_input_tokens_seen": 656393600,
+ "step": 15010,
+ "train_runtime": 116127.8068,
+ "train_tokens_per_second": 5652.338
+ },
+ {
+ "epoch": 1.0109749528683005,
+ "grad_norm": 0.7485929128993108,
+ "learning_rate": 4.914335929599828e-06,
+ "loss": 0.3200997829437256,
+ "num_input_tokens_seen": 656618032,
+ "step": 15015,
+ "train_runtime": 116165.2315,
+ "train_tokens_per_second": 5652.449
+ },
+ {
+ "epoch": 1.0113116078642608,
+ "grad_norm": 0.8695129879078439,
+ "learning_rate": 4.9116922476372155e-06,
+ "loss": 0.32773504257202146,
+ "num_input_tokens_seen": 656838496,
+ "step": 15020,
+ "train_runtime": 116207.5873,
+ "train_tokens_per_second": 5652.286
+ },
+ {
+ "epoch": 1.0116482628602208,
+ "grad_norm": 0.8589619543857242,
+ "learning_rate": 4.909048590369586e-06,
+ "loss": 0.31574647426605223,
+ "num_input_tokens_seen": 657057672,
+ "step": 15025,
+ "train_runtime": 116240.7138,
+ "train_tokens_per_second": 5652.561
+ },
+ {
+ "epoch": 1.011984917856181,
+ "grad_norm": 0.7207841320471391,
+ "learning_rate": 4.906404958536227e-06,
+ "loss": 0.3148768901824951,
+ "num_input_tokens_seen": 657282040,
+ "step": 15030,
+ "train_runtime": 116286.0079,
+ "train_tokens_per_second": 5652.288
+ },
+ {
+ "epoch": 1.012321572852141,
+ "grad_norm": 0.7575770351222048,
+ "learning_rate": 4.903761352876422e-06,
+ "loss": 0.31647629737854005,
+ "num_input_tokens_seen": 657477448,
+ "step": 15035,
+ "train_runtime": 116329.5257,
+ "train_tokens_per_second": 5651.854
+ },
+ {
+ "epoch": 1.0126582278481013,
+ "grad_norm": 0.8492818928373452,
+ "learning_rate": 4.9011177741294485e-06,
+ "loss": 0.3354557275772095,
+ "num_input_tokens_seen": 657686648,
+ "step": 15040,
+ "train_runtime": 116368.5371,
+ "train_tokens_per_second": 5651.757
+ },
+ {
+ "epoch": 1.0129948828440614,
+ "grad_norm": 0.6899675763130493,
+ "learning_rate": 4.898474223034576e-06,
+ "loss": 0.3180086612701416,
+ "num_input_tokens_seen": 657917808,
+ "step": 15045,
+ "train_runtime": 116410.5268,
+ "train_tokens_per_second": 5651.704
+ },
+ {
+ "epoch": 1.0133315378400216,
+ "grad_norm": 0.7682260059994229,
+ "learning_rate": 4.895830700331061e-06,
+ "loss": 0.31407153606414795,
+ "num_input_tokens_seen": 658124800,
+ "step": 15050,
+ "train_runtime": 116451.6904,
+ "train_tokens_per_second": 5651.483
+ },
+ {
+ "epoch": 1.0136681928359816,
+ "grad_norm": 0.7119767891756066,
+ "learning_rate": 4.893187206758158e-06,
+ "loss": 0.3087860822677612,
+ "num_input_tokens_seen": 658352392,
+ "step": 15055,
+ "train_runtime": 116484.7713,
+ "train_tokens_per_second": 5651.832
+ },
+ {
+ "epoch": 1.014004847831942,
+ "grad_norm": 0.7727900894930202,
+ "learning_rate": 4.890543743055115e-06,
+ "loss": 0.3047621250152588,
+ "num_input_tokens_seen": 658570240,
+ "step": 15060,
+ "train_runtime": 116523.7373,
+ "train_tokens_per_second": 5651.812
+ },
+ {
+ "epoch": 1.014341502827902,
+ "grad_norm": 0.7153606277605876,
+ "learning_rate": 4.887900309961163e-06,
+ "loss": 0.310034441947937,
+ "num_input_tokens_seen": 658791472,
+ "step": 15065,
+ "train_runtime": 116562.1854,
+ "train_tokens_per_second": 5651.846
+ },
+ {
+ "epoch": 1.0146781578238622,
+ "grad_norm": 0.7071012991080629,
+ "learning_rate": 4.885256908215532e-06,
+ "loss": 0.32690412998199464,
+ "num_input_tokens_seen": 659014128,
+ "step": 15070,
+ "train_runtime": 116590.8598,
+ "train_tokens_per_second": 5652.365
+ },
+ {
+ "epoch": 1.0150148128198222,
+ "grad_norm": 0.7712968829018512,
+ "learning_rate": 4.88261353855744e-06,
+ "loss": 0.3006293296813965,
+ "num_input_tokens_seen": 659239712,
+ "step": 15075,
+ "train_runtime": 116632.4209,
+ "train_tokens_per_second": 5652.285
+ },
+ {
+ "epoch": 1.0153514678157824,
+ "grad_norm": 0.6949970194408363,
+ "learning_rate": 4.8799702017261e-06,
+ "loss": 0.28703746795654295,
+ "num_input_tokens_seen": 659455440,
+ "step": 15080,
+ "train_runtime": 116668.3446,
+ "train_tokens_per_second": 5652.394
+ },
+ {
+ "epoch": 1.0156881228117425,
+ "grad_norm": 0.7902536702292094,
+ "learning_rate": 4.877326898460709e-06,
+ "loss": 0.31946167945861814,
+ "num_input_tokens_seen": 659666248,
+ "step": 15085,
+ "train_runtime": 116710.1355,
+ "train_tokens_per_second": 5652.176
+ },
+ {
+ "epoch": 1.0160247778077027,
+ "grad_norm": 0.705619839622007,
+ "learning_rate": 4.87468362950046e-06,
+ "loss": 0.28933143615722656,
+ "num_input_tokens_seen": 659891120,
+ "step": 15090,
+ "train_runtime": 116756.1991,
+ "train_tokens_per_second": 5651.872
+ },
+ {
+ "epoch": 1.0163614328036628,
+ "grad_norm": 0.7513443644619009,
+ "learning_rate": 4.872040395584533e-06,
+ "loss": 0.3063974380493164,
+ "num_input_tokens_seen": 660099960,
+ "step": 15095,
+ "train_runtime": 116792.0331,
+ "train_tokens_per_second": 5651.926
+ },
+ {
+ "epoch": 1.016698087799623,
+ "grad_norm": 0.7643641805946824,
+ "learning_rate": 4.869397197452104e-06,
+ "loss": 0.2840407848358154,
+ "num_input_tokens_seen": 660319352,
+ "step": 15100,
+ "train_runtime": 116827.7471,
+ "train_tokens_per_second": 5652.076
+ },
+ {
+ "epoch": 1.017034742795583,
+ "grad_norm": 0.8009081378183989,
+ "learning_rate": 4.866754035842329e-06,
+ "loss": 0.3019397258758545,
+ "num_input_tokens_seen": 660540440,
+ "step": 15105,
+ "train_runtime": 116864.8835,
+ "train_tokens_per_second": 5652.172
+ },
+ {
+ "epoch": 1.0173713977915433,
+ "grad_norm": 0.7774094479500263,
+ "learning_rate": 4.8641109114943666e-06,
+ "loss": 0.30378923416137693,
+ "num_input_tokens_seen": 660755584,
+ "step": 15110,
+ "train_runtime": 116900.7001,
+ "train_tokens_per_second": 5652.281
+ },
+ {
+ "epoch": 1.0177080527875033,
+ "grad_norm": 0.6851052748232287,
+ "learning_rate": 4.861467825147352e-06,
+ "loss": 0.3153241157531738,
+ "num_input_tokens_seen": 660976736,
+ "step": 15115,
+ "train_runtime": 116933.8046,
+ "train_tokens_per_second": 5652.572
+ },
+ {
+ "epoch": 1.0180447077834636,
+ "grad_norm": 0.7308126890818304,
+ "learning_rate": 4.858824777540419e-06,
+ "loss": 0.3002683401107788,
+ "num_input_tokens_seen": 661194120,
+ "step": 15120,
+ "train_runtime": 116976.4465,
+ "train_tokens_per_second": 5652.37
+ },
+ {
+ "epoch": 1.0183813627794236,
+ "grad_norm": 0.7320711876294524,
+ "learning_rate": 4.856181769412686e-06,
+ "loss": 0.3039540767669678,
+ "num_input_tokens_seen": 661406256,
+ "step": 15125,
+ "train_runtime": 117021.2899,
+ "train_tokens_per_second": 5652.016
+ },
+ {
+ "epoch": 1.0187180177753838,
+ "grad_norm": 0.8338813052362888,
+ "learning_rate": 4.8535388015032644e-06,
+ "loss": 0.30543341636657717,
+ "num_input_tokens_seen": 661615064,
+ "step": 15130,
+ "train_runtime": 117059.1584,
+ "train_tokens_per_second": 5651.972
+ },
+ {
+ "epoch": 1.0190546727713439,
+ "grad_norm": 0.6951071164752909,
+ "learning_rate": 4.850895874551248e-06,
+ "loss": 0.32107832431793215,
+ "num_input_tokens_seen": 661843768,
+ "step": 15135,
+ "train_runtime": 117098.7017,
+ "train_tokens_per_second": 5652.016
+ },
+ {
+ "epoch": 1.0193913277673041,
+ "grad_norm": 0.6999606429191154,
+ "learning_rate": 4.848252989295726e-06,
+ "loss": 0.3267232894897461,
+ "num_input_tokens_seen": 662043496,
+ "step": 15140,
+ "train_runtime": 117135.0098,
+ "train_tokens_per_second": 5651.969
+ },
+ {
+ "epoch": 1.0197279827632642,
+ "grad_norm": 0.7298299606738377,
+ "learning_rate": 4.845610146475771e-06,
+ "loss": 0.2995804786682129,
+ "num_input_tokens_seen": 662264176,
+ "step": 15145,
+ "train_runtime": 117169.4311,
+ "train_tokens_per_second": 5652.192
+ },
+ {
+ "epoch": 1.0200646377592244,
+ "grad_norm": 0.7687173415714617,
+ "learning_rate": 4.842967346830447e-06,
+ "loss": 0.2979107141494751,
+ "num_input_tokens_seen": 662484464,
+ "step": 15150,
+ "train_runtime": 117205.462,
+ "train_tokens_per_second": 5652.334
+ },
+ {
+ "epoch": 1.0204012927551844,
+ "grad_norm": 0.8483763855881413,
+ "learning_rate": 4.840324591098803e-06,
+ "loss": 0.3103074073791504,
+ "num_input_tokens_seen": 662691688,
+ "step": 15155,
+ "train_runtime": 117245.4475,
+ "train_tokens_per_second": 5652.174
+ },
+ {
+ "epoch": 1.0207379477511447,
+ "grad_norm": 0.8319667788025759,
+ "learning_rate": 4.83768188001988e-06,
+ "loss": 0.3106090068817139,
+ "num_input_tokens_seen": 662895200,
+ "step": 15160,
+ "train_runtime": 117290.7287,
+ "train_tokens_per_second": 5651.727
+ },
+ {
+ "epoch": 1.0210746027471047,
+ "grad_norm": 0.7991516150174239,
+ "learning_rate": 4.835039214332699e-06,
+ "loss": 0.30320937633514405,
+ "num_input_tokens_seen": 663097200,
+ "step": 15165,
+ "train_runtime": 117331.5062,
+ "train_tokens_per_second": 5651.485
+ },
+ {
+ "epoch": 1.021411257743065,
+ "grad_norm": 0.7494866477414706,
+ "learning_rate": 4.832396594776277e-06,
+ "loss": 0.3088703155517578,
+ "num_input_tokens_seen": 663319712,
+ "step": 15170,
+ "train_runtime": 117375.1688,
+ "train_tokens_per_second": 5651.278
+ },
+ {
+ "epoch": 1.021747912739025,
+ "grad_norm": 0.7704595139249664,
+ "learning_rate": 4.82975402208961e-06,
+ "loss": 0.2921560764312744,
+ "num_input_tokens_seen": 663537384,
+ "step": 15175,
+ "train_runtime": 117413.8519,
+ "train_tokens_per_second": 5651.27
+ },
+ {
+ "epoch": 1.0220845677349852,
+ "grad_norm": 0.7282417186604674,
+ "learning_rate": 4.827111497011692e-06,
+ "loss": 0.284089994430542,
+ "num_input_tokens_seen": 663777040,
+ "step": 15180,
+ "train_runtime": 117452.9484,
+ "train_tokens_per_second": 5651.429
+ },
+ {
+ "epoch": 1.0224212227309453,
+ "grad_norm": 0.9566982948749372,
+ "learning_rate": 4.82446902028149e-06,
+ "loss": 0.3103042125701904,
+ "num_input_tokens_seen": 663998952,
+ "step": 15185,
+ "train_runtime": 117490.5025,
+ "train_tokens_per_second": 5651.512
+ },
+ {
+ "epoch": 1.0227578777269055,
+ "grad_norm": 0.8002891674482239,
+ "learning_rate": 4.821826592637968e-06,
+ "loss": 0.3256609201431274,
+ "num_input_tokens_seen": 664216976,
+ "step": 15190,
+ "train_runtime": 117525.3723,
+ "train_tokens_per_second": 5651.69
+ },
+ {
+ "epoch": 1.0230945327228655,
+ "grad_norm": 0.6734170041621031,
+ "learning_rate": 4.8191842148200694e-06,
+ "loss": 0.30333666801452636,
+ "num_input_tokens_seen": 664447704,
+ "step": 15195,
+ "train_runtime": 117563.4341,
+ "train_tokens_per_second": 5651.823
+ },
+ {
+ "epoch": 1.0234311877188258,
+ "grad_norm": 0.7333137152011099,
+ "learning_rate": 4.8165418875667315e-06,
+ "loss": 0.3009442090988159,
+ "num_input_tokens_seen": 664678616,
+ "step": 15200,
+ "train_runtime": 117609.1613,
+ "train_tokens_per_second": 5651.589
+ },
+ {
+ "epoch": 1.0237678427147858,
+ "grad_norm": 0.7015384084375479,
+ "learning_rate": 4.813899611616867e-06,
+ "loss": 0.3210268020629883,
+ "num_input_tokens_seen": 664897488,
+ "step": 15205,
+ "train_runtime": 117645.0557,
+ "train_tokens_per_second": 5651.725
+ },
+ {
+ "epoch": 1.024104497710746,
+ "grad_norm": 1.4380458993791456,
+ "learning_rate": 4.8112573877093855e-06,
+ "loss": 0.3092465400695801,
+ "num_input_tokens_seen": 665131288,
+ "step": 15210,
+ "train_runtime": 117686.9545,
+ "train_tokens_per_second": 5651.699
+ },
+ {
+ "epoch": 1.024441152706706,
+ "grad_norm": 0.7003471237764793,
+ "learning_rate": 4.808615216583171e-06,
+ "loss": 0.3192662000656128,
+ "num_input_tokens_seen": 665344872,
+ "step": 15215,
+ "train_runtime": 117720.8851,
+ "train_tokens_per_second": 5651.885
+ },
+ {
+ "epoch": 1.0247778077026664,
+ "grad_norm": 0.8436628825159369,
+ "learning_rate": 4.805973098977102e-06,
+ "loss": 0.32068495750427245,
+ "num_input_tokens_seen": 665556568,
+ "step": 15220,
+ "train_runtime": 117755.2974,
+ "train_tokens_per_second": 5652.031
+ },
+ {
+ "epoch": 1.0251144626986264,
+ "grad_norm": 0.7109267791884789,
+ "learning_rate": 4.803331035630037e-06,
+ "loss": 0.32355175018310545,
+ "num_input_tokens_seen": 665786432,
+ "step": 15225,
+ "train_runtime": 117788.2016,
+ "train_tokens_per_second": 5652.403
+ },
+ {
+ "epoch": 1.0254511176945866,
+ "grad_norm": 0.6877005988993926,
+ "learning_rate": 4.800689027280822e-06,
+ "loss": 0.29448685646057127,
+ "num_input_tokens_seen": 666001856,
+ "step": 15230,
+ "train_runtime": 117828.4813,
+ "train_tokens_per_second": 5652.299
+ },
+ {
+ "epoch": 1.0257877726905467,
+ "grad_norm": 0.7724290603372872,
+ "learning_rate": 4.798047074668283e-06,
+ "loss": 0.2994450330734253,
+ "num_input_tokens_seen": 666220128,
+ "step": 15235,
+ "train_runtime": 117866.0261,
+ "train_tokens_per_second": 5652.351
+ },
+ {
+ "epoch": 1.026124427686507,
+ "grad_norm": 0.6832618434512723,
+ "learning_rate": 4.795405178531236e-06,
+ "loss": 0.29120640754699706,
+ "num_input_tokens_seen": 666430824,
+ "step": 15240,
+ "train_runtime": 117899.98,
+ "train_tokens_per_second": 5652.51
+ },
+ {
+ "epoch": 1.026461082682467,
+ "grad_norm": 0.7900874797047402,
+ "learning_rate": 4.792763339608478e-06,
+ "loss": 0.31427597999572754,
+ "num_input_tokens_seen": 666637408,
+ "step": 15245,
+ "train_runtime": 117941.2447,
+ "train_tokens_per_second": 5652.284
+ },
+ {
+ "epoch": 1.0267977376784272,
+ "grad_norm": 0.830479588475876,
+ "learning_rate": 4.790121558638793e-06,
+ "loss": 0.2697123050689697,
+ "num_input_tokens_seen": 666856712,
+ "step": 15250,
+ "train_runtime": 117983.2012,
+ "train_tokens_per_second": 5652.133
+ },
+ {
+ "epoch": 1.0271343926743872,
+ "grad_norm": 0.702781552620645,
+ "learning_rate": 4.787479836360944e-06,
+ "loss": 0.33616783618927004,
+ "num_input_tokens_seen": 667072880,
+ "step": 15255,
+ "train_runtime": 118015.753,
+ "train_tokens_per_second": 5652.405
+ },
+ {
+ "epoch": 1.0274710476703475,
+ "grad_norm": 0.752937484002602,
+ "learning_rate": 4.784838173513682e-06,
+ "loss": 0.308072566986084,
+ "num_input_tokens_seen": 667290960,
+ "step": 15260,
+ "train_runtime": 118055.7427,
+ "train_tokens_per_second": 5652.338
+ },
+ {
+ "epoch": 1.0278077026663075,
+ "grad_norm": 0.8169678756972052,
+ "learning_rate": 4.782196570835737e-06,
+ "loss": 0.3036174297332764,
+ "num_input_tokens_seen": 667517912,
+ "step": 15265,
+ "train_runtime": 118092.0571,
+ "train_tokens_per_second": 5652.522
+ },
+ {
+ "epoch": 1.0281443576622677,
+ "grad_norm": 0.8398308050050739,
+ "learning_rate": 4.779555029065827e-06,
+ "loss": 0.29191629886627196,
+ "num_input_tokens_seen": 667717400,
+ "step": 15270,
+ "train_runtime": 118131.9033,
+ "train_tokens_per_second": 5652.304
+ },
+ {
+ "epoch": 1.0284810126582278,
+ "grad_norm": 0.7571501727231587,
+ "learning_rate": 4.776913548942649e-06,
+ "loss": 0.31882758140563966,
+ "num_input_tokens_seen": 667943704,
+ "step": 15275,
+ "train_runtime": 118171.6607,
+ "train_tokens_per_second": 5652.317
+ },
+ {
+ "epoch": 1.028817667654188,
+ "grad_norm": 0.7323168837010766,
+ "learning_rate": 4.774272131204889e-06,
+ "loss": 0.31909172534942626,
+ "num_input_tokens_seen": 668171240,
+ "step": 15280,
+ "train_runtime": 118214.9165,
+ "train_tokens_per_second": 5652.174
+ },
+ {
+ "epoch": 1.029154322650148,
+ "grad_norm": 0.8168602912557673,
+ "learning_rate": 4.7716307765912045e-06,
+ "loss": 0.3175906419754028,
+ "num_input_tokens_seen": 668389056,
+ "step": 15285,
+ "train_runtime": 118259.4761,
+ "train_tokens_per_second": 5651.886
+ },
+ {
+ "epoch": 1.0294909776461083,
+ "grad_norm": 0.686585468117989,
+ "learning_rate": 4.768989485840246e-06,
+ "loss": 0.3248094081878662,
+ "num_input_tokens_seen": 668619360,
+ "step": 15290,
+ "train_runtime": 118295.1221,
+ "train_tokens_per_second": 5652.13
+ },
+ {
+ "epoch": 1.0298276326420683,
+ "grad_norm": 0.7842217672393594,
+ "learning_rate": 4.766348259690641e-06,
+ "loss": 0.2917296886444092,
+ "num_input_tokens_seen": 668845224,
+ "step": 15295,
+ "train_runtime": 118327.5434,
+ "train_tokens_per_second": 5652.49
+ },
+ {
+ "epoch": 1.0301642876380286,
+ "grad_norm": 0.8697416761883611,
+ "learning_rate": 4.763707098881002e-06,
+ "loss": 0.3045883893966675,
+ "num_input_tokens_seen": 669059584,
+ "step": 15300,
+ "train_runtime": 118366.6141,
+ "train_tokens_per_second": 5652.435
+ },
+ {
+ "epoch": 1.0305009426339886,
+ "grad_norm": 0.8048541483957821,
+ "learning_rate": 4.761066004149917e-06,
+ "loss": 0.3286751270294189,
+ "num_input_tokens_seen": 669264992,
+ "step": 15305,
+ "train_runtime": 118405.4204,
+ "train_tokens_per_second": 5652.317
+ },
+ {
+ "epoch": 1.0308375976299489,
+ "grad_norm": 0.7590289473967594,
+ "learning_rate": 4.758424976235964e-06,
+ "loss": 0.3184011936187744,
+ "num_input_tokens_seen": 669485880,
+ "step": 15310,
+ "train_runtime": 118445.1028,
+ "train_tokens_per_second": 5652.288
+ },
+ {
+ "epoch": 1.031174252625909,
+ "grad_norm": 0.8012934218342038,
+ "learning_rate": 4.755784015877694e-06,
+ "loss": 0.33234100341796874,
+ "num_input_tokens_seen": 669718792,
+ "step": 15315,
+ "train_runtime": 118479.6014,
+ "train_tokens_per_second": 5652.608
+ },
+ {
+ "epoch": 1.0315109076218691,
+ "grad_norm": 0.7664598006252185,
+ "learning_rate": 4.753143123813646e-06,
+ "loss": 0.3180215358734131,
+ "num_input_tokens_seen": 669934720,
+ "step": 15320,
+ "train_runtime": 118516.7599,
+ "train_tokens_per_second": 5652.658
+ },
+ {
+ "epoch": 1.0318475626178292,
+ "grad_norm": 0.7487344131067605,
+ "learning_rate": 4.7505023007823365e-06,
+ "loss": 0.3165398597717285,
+ "num_input_tokens_seen": 670156656,
+ "step": 15325,
+ "train_runtime": 118558.1181,
+ "train_tokens_per_second": 5652.558
+ },
+ {
+ "epoch": 1.0321842176137894,
+ "grad_norm": 0.6711552412916334,
+ "learning_rate": 4.747861547522264e-06,
+ "loss": 0.28763623237609864,
+ "num_input_tokens_seen": 670377432,
+ "step": 15330,
+ "train_runtime": 118592.5418,
+ "train_tokens_per_second": 5652.779
+ },
+ {
+ "epoch": 1.0325208726097495,
+ "grad_norm": 0.7093961443291146,
+ "learning_rate": 4.745220864771904e-06,
+ "loss": 0.29185004234313966,
+ "num_input_tokens_seen": 670582320,
+ "step": 15335,
+ "train_runtime": 118635.1864,
+ "train_tokens_per_second": 5652.474
+ },
+ {
+ "epoch": 1.0328575276057097,
+ "grad_norm": 0.685756160543914,
+ "learning_rate": 4.742580253269718e-06,
+ "loss": 0.31283178329467776,
+ "num_input_tokens_seen": 670787368,
+ "step": 15340,
+ "train_runtime": 118672.7657,
+ "train_tokens_per_second": 5652.412
+ },
+ {
+ "epoch": 1.0331941826016697,
+ "grad_norm": 0.8426009554601754,
+ "learning_rate": 4.739939713754141e-06,
+ "loss": 0.28745934963226316,
+ "num_input_tokens_seen": 670998288,
+ "step": 15345,
+ "train_runtime": 118718.3964,
+ "train_tokens_per_second": 5652.016
+ },
+ {
+ "epoch": 1.03353083759763,
+ "grad_norm": 0.7932262204171826,
+ "learning_rate": 4.737299246963597e-06,
+ "loss": 0.3006887912750244,
+ "num_input_tokens_seen": 671207256,
+ "step": 15350,
+ "train_runtime": 118756.3685,
+ "train_tokens_per_second": 5651.969
+ },
+ {
+ "epoch": 1.03386749259359,
+ "grad_norm": 0.6568880971934117,
+ "learning_rate": 4.734658853636479e-06,
+ "loss": 0.3070893526077271,
+ "num_input_tokens_seen": 671429312,
+ "step": 15355,
+ "train_runtime": 118793.5087,
+ "train_tokens_per_second": 5652.071
+ },
+ {
+ "epoch": 1.0342041475895503,
+ "grad_norm": 0.8138534659581989,
+ "learning_rate": 4.732018534511167e-06,
+ "loss": 0.3325775146484375,
+ "num_input_tokens_seen": 671648736,
+ "step": 15360,
+ "train_runtime": 118828.7625,
+ "train_tokens_per_second": 5652.24
+ },
+ {
+ "epoch": 1.0345408025855103,
+ "grad_norm": 1.1941290710370875,
+ "learning_rate": 4.729378290326017e-06,
+ "loss": 0.2876759052276611,
+ "num_input_tokens_seen": 671854064,
+ "step": 15365,
+ "train_runtime": 118861.8808,
+ "train_tokens_per_second": 5652.393
+ },
+ {
+ "epoch": 1.0348774575814705,
+ "grad_norm": 0.6797186656709692,
+ "learning_rate": 4.726738121819365e-06,
+ "loss": 0.3200536727905273,
+ "num_input_tokens_seen": 672072384,
+ "step": 15370,
+ "train_runtime": 118895.929,
+ "train_tokens_per_second": 5652.611
+ },
+ {
+ "epoch": 1.0352141125774306,
+ "grad_norm": 0.8461920537238418,
+ "learning_rate": 4.7240980297295255e-06,
+ "loss": 0.2775277137756348,
+ "num_input_tokens_seen": 672283880,
+ "step": 15375,
+ "train_runtime": 118932.7251,
+ "train_tokens_per_second": 5652.64
+ },
+ {
+ "epoch": 1.0355507675733908,
+ "grad_norm": 0.6169331576620465,
+ "learning_rate": 4.721458014794795e-06,
+ "loss": 0.30312108993530273,
+ "num_input_tokens_seen": 672508584,
+ "step": 15380,
+ "train_runtime": 118965.3642,
+ "train_tokens_per_second": 5652.978
+ },
+ {
+ "epoch": 1.0358874225693508,
+ "grad_norm": 0.830035914264151,
+ "learning_rate": 4.718818077753439e-06,
+ "loss": 0.32550725936889646,
+ "num_input_tokens_seen": 672737840,
+ "step": 15385,
+ "train_runtime": 119001.6804,
+ "train_tokens_per_second": 5653.179
+ },
+ {
+ "epoch": 1.036224077565311,
+ "grad_norm": 0.7363851453554499,
+ "learning_rate": 4.716178219343712e-06,
+ "loss": 0.310680890083313,
+ "num_input_tokens_seen": 672951456,
+ "step": 15390,
+ "train_runtime": 119037.6094,
+ "train_tokens_per_second": 5653.268
+ },
+ {
+ "epoch": 1.0365607325612711,
+ "grad_norm": 0.9510291073057959,
+ "learning_rate": 4.713538440303841e-06,
+ "loss": 0.3095111846923828,
+ "num_input_tokens_seen": 673143664,
+ "step": 15395,
+ "train_runtime": 119069.9192,
+ "train_tokens_per_second": 5653.348
+ },
+ {
+ "epoch": 1.0368973875572314,
+ "grad_norm": 0.7829230030914317,
+ "learning_rate": 4.7108987413720335e-06,
+ "loss": 0.3282665491104126,
+ "num_input_tokens_seen": 673355776,
+ "step": 15400,
+ "train_runtime": 119108.1422,
+ "train_tokens_per_second": 5653.314
+ },
+ {
+ "epoch": 1.0372340425531914,
+ "grad_norm": 0.9090900544664394,
+ "learning_rate": 4.708259123286469e-06,
+ "loss": 0.31505331993103025,
+ "num_input_tokens_seen": 673568088,
+ "step": 15405,
+ "train_runtime": 119152.501,
+ "train_tokens_per_second": 5652.992
+ },
+ {
+ "epoch": 1.0375706975491517,
+ "grad_norm": 0.7723625134036335,
+ "learning_rate": 4.7056195867853105e-06,
+ "loss": 0.30636959075927733,
+ "num_input_tokens_seen": 673795064,
+ "step": 15410,
+ "train_runtime": 119195.1556,
+ "train_tokens_per_second": 5652.873
+ },
+ {
+ "epoch": 1.0379073525451117,
+ "grad_norm": 0.7552569039187913,
+ "learning_rate": 4.702980132606696e-06,
+ "loss": 0.2860846996307373,
+ "num_input_tokens_seen": 674021712,
+ "step": 15415,
+ "train_runtime": 119229.7446,
+ "train_tokens_per_second": 5653.134
+ },
+ {
+ "epoch": 1.038244007541072,
+ "grad_norm": 0.7757156949433505,
+ "learning_rate": 4.700340761488741e-06,
+ "loss": 0.297334623336792,
+ "num_input_tokens_seen": 674242240,
+ "step": 15420,
+ "train_runtime": 119259.447,
+ "train_tokens_per_second": 5653.575
+ },
+ {
+ "epoch": 1.038580662537032,
+ "grad_norm": 0.845930651529408,
+ "learning_rate": 4.6977014741695345e-06,
+ "loss": 0.3014695167541504,
+ "num_input_tokens_seen": 674464624,
+ "step": 15425,
+ "train_runtime": 119301.6217,
+ "train_tokens_per_second": 5653.441
+ },
+ {
+ "epoch": 1.0389173175329922,
+ "grad_norm": 0.6863209085436818,
+ "learning_rate": 4.69506227138715e-06,
+ "loss": 0.31081855297088623,
+ "num_input_tokens_seen": 674697456,
+ "step": 15430,
+ "train_runtime": 119344.807,
+ "train_tokens_per_second": 5653.346
+ },
+ {
+ "epoch": 1.0392539725289522,
+ "grad_norm": 0.8027138711324319,
+ "learning_rate": 4.6924231538796265e-06,
+ "loss": 0.3378600120544434,
+ "num_input_tokens_seen": 674921088,
+ "step": 15435,
+ "train_runtime": 119379.2034,
+ "train_tokens_per_second": 5653.59
+ },
+ {
+ "epoch": 1.0395906275249125,
+ "grad_norm": 0.7714672936432438,
+ "learning_rate": 4.6897841223849885e-06,
+ "loss": 0.3055275917053223,
+ "num_input_tokens_seen": 675136400,
+ "step": 15440,
+ "train_runtime": 119423.2981,
+ "train_tokens_per_second": 5653.306
+ },
+ {
+ "epoch": 1.0399272825208725,
+ "grad_norm": 0.7820475842478094,
+ "learning_rate": 4.68714517764123e-06,
+ "loss": 0.2911995887756348,
+ "num_input_tokens_seen": 675354608,
+ "step": 15445,
+ "train_runtime": 119463.4783,
+ "train_tokens_per_second": 5653.231
+ },
+ {
+ "epoch": 1.0402639375168328,
+ "grad_norm": 0.6960997029967657,
+ "learning_rate": 4.684506320386327e-06,
+ "loss": 0.2882180690765381,
+ "num_input_tokens_seen": 675560352,
+ "step": 15450,
+ "train_runtime": 119502.7651,
+ "train_tokens_per_second": 5653.094
+ },
+ {
+ "epoch": 1.0406005925127928,
+ "grad_norm": 0.7240646804471922,
+ "learning_rate": 4.6818675513582235e-06,
+ "loss": 0.2950881481170654,
+ "num_input_tokens_seen": 675780176,
+ "step": 15455,
+ "train_runtime": 119549.6104,
+ "train_tokens_per_second": 5652.718
+ },
+ {
+ "epoch": 1.040937247508753,
+ "grad_norm": 0.7871482559840214,
+ "learning_rate": 4.679228871294846e-06,
+ "loss": 0.2878568649291992,
+ "num_input_tokens_seen": 676001504,
+ "step": 15460,
+ "train_runtime": 119590.0384,
+ "train_tokens_per_second": 5652.657
+ },
+ {
+ "epoch": 1.041273902504713,
+ "grad_norm": 0.7023040971791071,
+ "learning_rate": 4.67659028093409e-06,
+ "loss": 0.31134207248687745,
+ "num_input_tokens_seen": 676224760,
+ "step": 15465,
+ "train_runtime": 119626.8329,
+ "train_tokens_per_second": 5652.785
+ },
+ {
+ "epoch": 1.0416105575006733,
+ "grad_norm": 0.7287009704666462,
+ "learning_rate": 4.673951781013834e-06,
+ "loss": 0.3257226228713989,
+ "num_input_tokens_seen": 676437448,
+ "step": 15470,
+ "train_runtime": 119663.8355,
+ "train_tokens_per_second": 5652.814
+ },
+ {
+ "epoch": 1.0419472124966334,
+ "grad_norm": 0.694252013991942,
+ "learning_rate": 4.67131337227192e-06,
+ "loss": 0.32440500259399413,
+ "num_input_tokens_seen": 676661152,
+ "step": 15475,
+ "train_runtime": 119696.4836,
+ "train_tokens_per_second": 5653.141
+ },
+ {
+ "epoch": 1.0422838674925936,
+ "grad_norm": 0.742357464257659,
+ "learning_rate": 4.668675055446175e-06,
+ "loss": 0.3229750394821167,
+ "num_input_tokens_seen": 676878032,
+ "step": 15480,
+ "train_runtime": 119731.4831,
+ "train_tokens_per_second": 5653.3
+ },
+ {
+ "epoch": 1.0426205224885536,
+ "grad_norm": 0.8706983479448045,
+ "learning_rate": 4.666036831274392e-06,
+ "loss": 0.27671065330505373,
+ "num_input_tokens_seen": 677092008,
+ "step": 15485,
+ "train_runtime": 119767.1182,
+ "train_tokens_per_second": 5653.405
+ },
+ {
+ "epoch": 1.0429571774845139,
+ "grad_norm": 0.7127829768191473,
+ "learning_rate": 4.663398700494346e-06,
+ "loss": 0.30718288421630857,
+ "num_input_tokens_seen": 677302040,
+ "step": 15490,
+ "train_runtime": 119808.142,
+ "train_tokens_per_second": 5653.222
+ },
+ {
+ "epoch": 1.043293832480474,
+ "grad_norm": 0.7433197799578122,
+ "learning_rate": 4.66076066384378e-06,
+ "loss": 0.3258528232574463,
+ "num_input_tokens_seen": 677525784,
+ "step": 15495,
+ "train_runtime": 119841.1792,
+ "train_tokens_per_second": 5653.531
+ },
+ {
+ "epoch": 1.0436304874764342,
+ "grad_norm": 0.6844094807815899,
+ "learning_rate": 4.658122722060413e-06,
+ "loss": 0.3136504888534546,
+ "num_input_tokens_seen": 677745832,
+ "step": 15500,
+ "train_runtime": 119873.5793,
+ "train_tokens_per_second": 5653.838
+ },
+ {
+ "epoch": 1.0439671424723942,
+ "grad_norm": 0.7516388636804747,
+ "learning_rate": 4.655484875881935e-06,
+ "loss": 0.3184205055236816,
+ "num_input_tokens_seen": 677956888,
+ "step": 15505,
+ "train_runtime": 119919.1449,
+ "train_tokens_per_second": 5653.45
+ },
+ {
+ "epoch": 1.0443037974683544,
+ "grad_norm": 0.7798917548462908,
+ "learning_rate": 4.652847126046014e-06,
+ "loss": 0.335153865814209,
+ "num_input_tokens_seen": 678150664,
+ "step": 15510,
+ "train_runtime": 119955.6333,
+ "train_tokens_per_second": 5653.346
+ },
+ {
+ "epoch": 1.0446404524643145,
+ "grad_norm": 0.7432288033280057,
+ "learning_rate": 4.650209473290286e-06,
+ "loss": 0.33000261783599855,
+ "num_input_tokens_seen": 678377032,
+ "step": 15515,
+ "train_runtime": 119990.1304,
+ "train_tokens_per_second": 5653.607
+ },
+ {
+ "epoch": 1.0449771074602747,
+ "grad_norm": 0.7736787887096265,
+ "learning_rate": 4.647571918352366e-06,
+ "loss": 0.31568212509155275,
+ "num_input_tokens_seen": 678595080,
+ "step": 15520,
+ "train_runtime": 120030.7386,
+ "train_tokens_per_second": 5653.511
+ },
+ {
+ "epoch": 1.0453137624562348,
+ "grad_norm": 0.8603539995040905,
+ "learning_rate": 4.644934461969833e-06,
+ "loss": 0.305867600440979,
+ "num_input_tokens_seen": 678818928,
+ "step": 15525,
+ "train_runtime": 120078.8018,
+ "train_tokens_per_second": 5653.112
+ },
+ {
+ "epoch": 1.045650417452195,
+ "grad_norm": 0.802437774953463,
+ "learning_rate": 4.642297104880247e-06,
+ "loss": 0.2917011260986328,
+ "num_input_tokens_seen": 679044840,
+ "step": 15530,
+ "train_runtime": 120117.9302,
+ "train_tokens_per_second": 5653.151
+ },
+ {
+ "epoch": 1.045987072448155,
+ "grad_norm": 0.8097716033620479,
+ "learning_rate": 4.639659847821133e-06,
+ "loss": 0.29921350479125974,
+ "num_input_tokens_seen": 679260504,
+ "step": 15535,
+ "train_runtime": 120155.5327,
+ "train_tokens_per_second": 5653.177
+ },
+ {
+ "epoch": 1.0463237274441153,
+ "grad_norm": 0.7081626022799157,
+ "learning_rate": 4.6370226915299945e-06,
+ "loss": 0.2979047060012817,
+ "num_input_tokens_seen": 679478208,
+ "step": 15540,
+ "train_runtime": 120188.8745,
+ "train_tokens_per_second": 5653.42
+ },
+ {
+ "epoch": 1.0466603824400753,
+ "grad_norm": 0.6768088643103284,
+ "learning_rate": 4.634385636744302e-06,
+ "loss": 0.3120341539382935,
+ "num_input_tokens_seen": 679708344,
+ "step": 15545,
+ "train_runtime": 120226.4885,
+ "train_tokens_per_second": 5653.566
+ },
+ {
+ "epoch": 1.0469970374360356,
+ "grad_norm": 1.008615167669033,
+ "learning_rate": 4.631748684201503e-06,
+ "loss": 0.3280055522918701,
+ "num_input_tokens_seen": 679937752,
+ "step": 15550,
+ "train_runtime": 120267.2677,
+ "train_tokens_per_second": 5653.556
+ },
+ {
+ "epoch": 1.0473336924319956,
+ "grad_norm": 0.7693743879402827,
+ "learning_rate": 4.629111834639008e-06,
+ "loss": 0.3046785593032837,
+ "num_input_tokens_seen": 680165064,
+ "step": 15555,
+ "train_runtime": 120305.2426,
+ "train_tokens_per_second": 5653.661
+ },
+ {
+ "epoch": 1.0476703474279558,
+ "grad_norm": 0.7036312918875381,
+ "learning_rate": 4.626475088794206e-06,
+ "loss": 0.3018321990966797,
+ "num_input_tokens_seen": 680389272,
+ "step": 15560,
+ "train_runtime": 120342.4244,
+ "train_tokens_per_second": 5653.777
+ },
+ {
+ "epoch": 1.0480070024239159,
+ "grad_norm": 0.7427333305833319,
+ "learning_rate": 4.623838447404454e-06,
+ "loss": 0.3129916191101074,
+ "num_input_tokens_seen": 680620144,
+ "step": 15565,
+ "train_runtime": 120385.4131,
+ "train_tokens_per_second": 5653.676
+ },
+ {
+ "epoch": 1.0483436574198761,
+ "grad_norm": 0.6803553069904376,
+ "learning_rate": 4.621201911207082e-06,
+ "loss": 0.29700002670288084,
+ "num_input_tokens_seen": 680832464,
+ "step": 15570,
+ "train_runtime": 120420.9855,
+ "train_tokens_per_second": 5653.769
+ },
+ {
+ "epoch": 1.0486803124158361,
+ "grad_norm": 0.6877375090372397,
+ "learning_rate": 4.618565480939387e-06,
+ "loss": 0.3175528526306152,
+ "num_input_tokens_seen": 681059856,
+ "step": 15575,
+ "train_runtime": 120454.7328,
+ "train_tokens_per_second": 5654.073
+ },
+ {
+ "epoch": 1.0490169674117964,
+ "grad_norm": 0.8334120115163696,
+ "learning_rate": 4.6159291573386395e-06,
+ "loss": 0.3175390958786011,
+ "num_input_tokens_seen": 681279128,
+ "step": 15580,
+ "train_runtime": 120491.4228,
+ "train_tokens_per_second": 5654.171
+ },
+ {
+ "epoch": 1.0493536224077564,
+ "grad_norm": 0.7481436859437062,
+ "learning_rate": 4.613292941142077e-06,
+ "loss": 0.30297222137451174,
+ "num_input_tokens_seen": 681498080,
+ "step": 15585,
+ "train_runtime": 120532.1304,
+ "train_tokens_per_second": 5654.078
+ },
+ {
+ "epoch": 1.0496902774037167,
+ "grad_norm": 0.7558163128951925,
+ "learning_rate": 4.610656833086912e-06,
+ "loss": 0.31702475547790526,
+ "num_input_tokens_seen": 681707736,
+ "step": 15590,
+ "train_runtime": 120571.1677,
+ "train_tokens_per_second": 5653.986
+ },
+ {
+ "epoch": 1.0500269323996767,
+ "grad_norm": 0.8628106650373581,
+ "learning_rate": 4.60802083391032e-06,
+ "loss": 0.3515254259109497,
+ "num_input_tokens_seen": 681929096,
+ "step": 15595,
+ "train_runtime": 120607.6322,
+ "train_tokens_per_second": 5654.112
+ },
+ {
+ "epoch": 1.050363587395637,
+ "grad_norm": 0.8187294344342202,
+ "learning_rate": 4.605384944349454e-06,
+ "loss": 0.29188697338104247,
+ "num_input_tokens_seen": 682138848,
+ "step": 15600,
+ "train_runtime": 120643.9645,
+ "train_tokens_per_second": 5654.148
+ },
+ {
+ "epoch": 1.050700242391597,
+ "grad_norm": 0.762024230166568,
+ "learning_rate": 4.602749165141429e-06,
+ "loss": 0.2967026948928833,
+ "num_input_tokens_seen": 682341504,
+ "step": 15605,
+ "train_runtime": 120680.7614,
+ "train_tokens_per_second": 5654.103
+ },
+ {
+ "epoch": 1.0510368973875572,
+ "grad_norm": 0.770669064131956,
+ "learning_rate": 4.600113497023333e-06,
+ "loss": 0.3200972080230713,
+ "num_input_tokens_seen": 682558744,
+ "step": 15610,
+ "train_runtime": 120718.1512,
+ "train_tokens_per_second": 5654.152
+ },
+ {
+ "epoch": 1.0513735523835175,
+ "grad_norm": 0.8428983487418248,
+ "learning_rate": 4.597477940732222e-06,
+ "loss": 0.2950571537017822,
+ "num_input_tokens_seen": 682762744,
+ "step": 15615,
+ "train_runtime": 120758.4342,
+ "train_tokens_per_second": 5653.955
+ },
+ {
+ "epoch": 1.0517102073794775,
+ "grad_norm": 0.8298281512031902,
+ "learning_rate": 4.594842497005124e-06,
+ "loss": 0.2987983226776123,
+ "num_input_tokens_seen": 682994336,
+ "step": 15620,
+ "train_runtime": 120791.1567,
+ "train_tokens_per_second": 5654.341
+ },
+ {
+ "epoch": 1.0520468623754375,
+ "grad_norm": 0.6720517085495755,
+ "learning_rate": 4.5922071665790264e-06,
+ "loss": 0.3123783111572266,
+ "num_input_tokens_seen": 683207456,
+ "step": 15625,
+ "train_runtime": 120830.8588,
+ "train_tokens_per_second": 5654.246
+ },
+ {
+ "epoch": 1.0523835173713978,
+ "grad_norm": 0.719767612563415,
+ "learning_rate": 4.589571950190897e-06,
+ "loss": 0.3103508949279785,
+ "num_input_tokens_seen": 683437384,
+ "step": 15630,
+ "train_runtime": 120873.1823,
+ "train_tokens_per_second": 5654.169
+ },
+ {
+ "epoch": 1.052720172367358,
+ "grad_norm": 0.7097307825735949,
+ "learning_rate": 4.586936848577661e-06,
+ "loss": 0.3118293285369873,
+ "num_input_tokens_seen": 683657552,
+ "step": 15635,
+ "train_runtime": 120907.5119,
+ "train_tokens_per_second": 5654.384
+ },
+ {
+ "epoch": 1.053056827363318,
+ "grad_norm": 0.7044277499214124,
+ "learning_rate": 4.584301862476219e-06,
+ "loss": 0.30995869636535645,
+ "num_input_tokens_seen": 683876008,
+ "step": 15640,
+ "train_runtime": 120944.8259,
+ "train_tokens_per_second": 5654.446
+ },
+ {
+ "epoch": 1.053393482359278,
+ "grad_norm": 0.7010017324067445,
+ "learning_rate": 4.581666992623435e-06,
+ "loss": 0.2860646963119507,
+ "num_input_tokens_seen": 684104824,
+ "step": 15645,
+ "train_runtime": 120984.5659,
+ "train_tokens_per_second": 5654.48
+ },
+ {
+ "epoch": 1.0537301373552383,
+ "grad_norm": 0.7902236702010951,
+ "learning_rate": 4.579032239756144e-06,
+ "loss": 0.3144806146621704,
+ "num_input_tokens_seen": 684326216,
+ "step": 15650,
+ "train_runtime": 121026.0735,
+ "train_tokens_per_second": 5654.37
+ },
+ {
+ "epoch": 1.0540667923511986,
+ "grad_norm": 0.7099398655983817,
+ "learning_rate": 4.576397604611143e-06,
+ "loss": 0.29763078689575195,
+ "num_input_tokens_seen": 684548424,
+ "step": 15655,
+ "train_runtime": 121060.9714,
+ "train_tokens_per_second": 5654.576
+ },
+ {
+ "epoch": 1.0544034473471586,
+ "grad_norm": 0.7151110990015658,
+ "learning_rate": 4.573763087925202e-06,
+ "loss": 0.30714902877807615,
+ "num_input_tokens_seen": 684772760,
+ "step": 15660,
+ "train_runtime": 121105.9773,
+ "train_tokens_per_second": 5654.327
+ },
+ {
+ "epoch": 1.0547401023431189,
+ "grad_norm": 0.7472326006657539,
+ "learning_rate": 4.571128690435053e-06,
+ "loss": 0.3167744159698486,
+ "num_input_tokens_seen": 684995472,
+ "step": 15665,
+ "train_runtime": 121149.0474,
+ "train_tokens_per_second": 5654.155
+ },
+ {
+ "epoch": 1.055076757339079,
+ "grad_norm": 0.8728278378658701,
+ "learning_rate": 4.568494412877401e-06,
+ "loss": 0.31170904636383057,
+ "num_input_tokens_seen": 685198200,
+ "step": 15670,
+ "train_runtime": 121178.4485,
+ "train_tokens_per_second": 5654.456
+ },
+ {
+ "epoch": 1.0554134123350392,
+ "grad_norm": 0.7744504434630247,
+ "learning_rate": 4.565860255988908e-06,
+ "loss": 0.31058855056762696,
+ "num_input_tokens_seen": 685420416,
+ "step": 15675,
+ "train_runtime": 121210.5142,
+ "train_tokens_per_second": 5654.793
+ },
+ {
+ "epoch": 1.0557500673309992,
+ "grad_norm": 0.7252974178645615,
+ "learning_rate": 4.563226220506212e-06,
+ "loss": 0.31565635204315184,
+ "num_input_tokens_seen": 685659776,
+ "step": 15680,
+ "train_runtime": 121247.6601,
+ "train_tokens_per_second": 5655.035
+ },
+ {
+ "epoch": 1.0560867223269594,
+ "grad_norm": 0.6919225790448248,
+ "learning_rate": 4.5605923071659085e-06,
+ "loss": 0.3112238645553589,
+ "num_input_tokens_seen": 685871672,
+ "step": 15685,
+ "train_runtime": 121283.9917,
+ "train_tokens_per_second": 5655.088
+ },
+ {
+ "epoch": 1.0564233773229195,
+ "grad_norm": 0.7233596710854652,
+ "learning_rate": 4.557958516704566e-06,
+ "loss": 0.2988229990005493,
+ "num_input_tokens_seen": 686092808,
+ "step": 15690,
+ "train_runtime": 121330.4161,
+ "train_tokens_per_second": 5654.747
+ },
+ {
+ "epoch": 1.0567600323188797,
+ "grad_norm": 0.8148391336751524,
+ "learning_rate": 4.5553248498587144e-06,
+ "loss": 0.3125143051147461,
+ "num_input_tokens_seen": 686309592,
+ "step": 15695,
+ "train_runtime": 121360.8019,
+ "train_tokens_per_second": 5655.117
+ },
+ {
+ "epoch": 1.0570966873148397,
+ "grad_norm": 0.8471401120479854,
+ "learning_rate": 4.552691307364853e-06,
+ "loss": 0.31246707439422605,
+ "num_input_tokens_seen": 686520592,
+ "step": 15700,
+ "train_runtime": 121401.4752,
+ "train_tokens_per_second": 5654.961
+ },
+ {
+ "epoch": 1.0574333423108,
+ "grad_norm": 0.8240701136904154,
+ "learning_rate": 4.5500578899594375e-06,
+ "loss": 0.33534066677093505,
+ "num_input_tokens_seen": 686743480,
+ "step": 15705,
+ "train_runtime": 121443.4876,
+ "train_tokens_per_second": 5654.84
+ },
+ {
+ "epoch": 1.05776999730676,
+ "grad_norm": 0.7338513258395319,
+ "learning_rate": 4.5474245983789e-06,
+ "loss": 0.3143591403961182,
+ "num_input_tokens_seen": 686964184,
+ "step": 15710,
+ "train_runtime": 121482.8751,
+ "train_tokens_per_second": 5654.823
+ },
+ {
+ "epoch": 1.0581066523027203,
+ "grad_norm": 0.7196154099105955,
+ "learning_rate": 4.54479143335963e-06,
+ "loss": 0.29776525497436523,
+ "num_input_tokens_seen": 687183192,
+ "step": 15715,
+ "train_runtime": 121528.6289,
+ "train_tokens_per_second": 5654.496
+ },
+ {
+ "epoch": 1.0584433072986803,
+ "grad_norm": 0.8284203342820305,
+ "learning_rate": 4.5421583956379855e-06,
+ "loss": 0.32465856075286864,
+ "num_input_tokens_seen": 687394088,
+ "step": 15720,
+ "train_runtime": 121563.9668,
+ "train_tokens_per_second": 5654.588
+ },
+ {
+ "epoch": 1.0587799622946406,
+ "grad_norm": 0.7128364600635472,
+ "learning_rate": 4.539525485950284e-06,
+ "loss": 0.29742043018341063,
+ "num_input_tokens_seen": 687621296,
+ "step": 15725,
+ "train_runtime": 121603.4696,
+ "train_tokens_per_second": 5654.619
+ },
+ {
+ "epoch": 1.0591166172906006,
+ "grad_norm": 0.8468437589127955,
+ "learning_rate": 4.536892705032815e-06,
+ "loss": 0.3188284397125244,
+ "num_input_tokens_seen": 687846312,
+ "step": 15730,
+ "train_runtime": 121637.3598,
+ "train_tokens_per_second": 5654.893
+ },
+ {
+ "epoch": 1.0594532722865608,
+ "grad_norm": 0.7689731083753425,
+ "learning_rate": 4.534260053621821e-06,
+ "loss": 0.29900522232055665,
+ "num_input_tokens_seen": 688062864,
+ "step": 15735,
+ "train_runtime": 121671.2264,
+ "train_tokens_per_second": 5655.099
+ },
+ {
+ "epoch": 1.0597899272825209,
+ "grad_norm": 0.7661666075190292,
+ "learning_rate": 4.531627532453519e-06,
+ "loss": 0.2740945816040039,
+ "num_input_tokens_seen": 688281760,
+ "step": 15740,
+ "train_runtime": 121719.5434,
+ "train_tokens_per_second": 5654.653
+ },
+ {
+ "epoch": 1.0601265822784811,
+ "grad_norm": 0.7197588574824276,
+ "learning_rate": 4.5289951422640826e-06,
+ "loss": 0.3090967655181885,
+ "num_input_tokens_seen": 688482288,
+ "step": 15745,
+ "train_runtime": 121757.3536,
+ "train_tokens_per_second": 5654.544
+ },
+ {
+ "epoch": 1.0604632372744411,
+ "grad_norm": 0.7581368758886011,
+ "learning_rate": 4.526362883789655e-06,
+ "loss": 0.29933021068572996,
+ "num_input_tokens_seen": 688706904,
+ "step": 15750,
+ "train_runtime": 121799.5328,
+ "train_tokens_per_second": 5654.43
+ },
+ {
+ "epoch": 1.0607998922704014,
+ "grad_norm": 0.7999491894487345,
+ "learning_rate": 4.5237307577663345e-06,
+ "loss": 0.32481019496917723,
+ "num_input_tokens_seen": 688934432,
+ "step": 15755,
+ "train_runtime": 121838.9872,
+ "train_tokens_per_second": 5654.466
+ },
+ {
+ "epoch": 1.0611365472663614,
+ "grad_norm": 0.668582256136596,
+ "learning_rate": 4.52109876493019e-06,
+ "loss": 0.3126739501953125,
+ "num_input_tokens_seen": 689154176,
+ "step": 15760,
+ "train_runtime": 121881.6703,
+ "train_tokens_per_second": 5654.289
+ },
+ {
+ "epoch": 1.0614732022623217,
+ "grad_norm": 0.7253320849615243,
+ "learning_rate": 4.518466906017249e-06,
+ "loss": 0.3287156343460083,
+ "num_input_tokens_seen": 689365536,
+ "step": 15765,
+ "train_runtime": 121920.7449,
+ "train_tokens_per_second": 5654.21
+ },
+ {
+ "epoch": 1.0618098572582817,
+ "grad_norm": 0.9312066051011625,
+ "learning_rate": 4.515835181763503e-06,
+ "loss": 0.29747960567474363,
+ "num_input_tokens_seen": 689584480,
+ "step": 15770,
+ "train_runtime": 121960.8859,
+ "train_tokens_per_second": 5654.145
+ },
+ {
+ "epoch": 1.062146512254242,
+ "grad_norm": 0.779107627410124,
+ "learning_rate": 4.513203592904903e-06,
+ "loss": 0.3129979848861694,
+ "num_input_tokens_seen": 689802248,
+ "step": 15775,
+ "train_runtime": 121996.8975,
+ "train_tokens_per_second": 5654.261
+ },
+ {
+ "epoch": 1.062483167250202,
+ "grad_norm": 0.7066715035398882,
+ "learning_rate": 4.510572140177368e-06,
+ "loss": 0.3110002040863037,
+ "num_input_tokens_seen": 690027296,
+ "step": 15780,
+ "train_runtime": 122041.4381,
+ "train_tokens_per_second": 5654.041
+ },
+ {
+ "epoch": 1.0628198222461622,
+ "grad_norm": 0.7657323857261208,
+ "learning_rate": 4.507940824316771e-06,
+ "loss": 0.31113996505737307,
+ "num_input_tokens_seen": 690244872,
+ "step": 15785,
+ "train_runtime": 122082.9515,
+ "train_tokens_per_second": 5653.901
+ },
+ {
+ "epoch": 1.0631564772421223,
+ "grad_norm": 0.778188387346321,
+ "learning_rate": 4.505309646058954e-06,
+ "loss": 0.306329345703125,
+ "num_input_tokens_seen": 690457904,
+ "step": 15790,
+ "train_runtime": 122123.6242,
+ "train_tokens_per_second": 5653.762
+ },
+ {
+ "epoch": 1.0634931322380825,
+ "grad_norm": 0.8721432879238993,
+ "learning_rate": 4.502678606139716e-06,
+ "loss": 0.290619421005249,
+ "num_input_tokens_seen": 690669592,
+ "step": 15795,
+ "train_runtime": 122161.8006,
+ "train_tokens_per_second": 5653.728
+ },
+ {
+ "epoch": 1.0638297872340425,
+ "grad_norm": 0.6919291628903627,
+ "learning_rate": 4.500047705294823e-06,
+ "loss": 0.28872241973876955,
+ "num_input_tokens_seen": 690892416,
+ "step": 15800,
+ "train_runtime": 122194.5033,
+ "train_tokens_per_second": 5654.038
+ },
+ {
+ "epoch": 1.0641664422300028,
+ "grad_norm": 0.7336900413158081,
+ "learning_rate": 4.497416944259993e-06,
+ "loss": 0.3109187364578247,
+ "num_input_tokens_seen": 691111312,
+ "step": 15805,
+ "train_runtime": 122227.6554,
+ "train_tokens_per_second": 5654.296
+ },
+ {
+ "epoch": 1.0645030972259628,
+ "grad_norm": 0.766581872834498,
+ "learning_rate": 4.494786323770912e-06,
+ "loss": 0.3253108263015747,
+ "num_input_tokens_seen": 691322056,
+ "step": 15810,
+ "train_runtime": 122271.3236,
+ "train_tokens_per_second": 5654.0
+ },
+ {
+ "epoch": 1.064839752221923,
+ "grad_norm": 0.6778315955670481,
+ "learning_rate": 4.492155844563224e-06,
+ "loss": 0.30454020500183104,
+ "num_input_tokens_seen": 691548176,
+ "step": 15815,
+ "train_runtime": 122307.168,
+ "train_tokens_per_second": 5654.192
+ },
+ {
+ "epoch": 1.065176407217883,
+ "grad_norm": 0.7604711413614016,
+ "learning_rate": 4.489525507372537e-06,
+ "loss": 0.3255009651184082,
+ "num_input_tokens_seen": 691767320,
+ "step": 15820,
+ "train_runtime": 122349.9913,
+ "train_tokens_per_second": 5654.004
+ },
+ {
+ "epoch": 1.0655130622138433,
+ "grad_norm": 0.7292388149896591,
+ "learning_rate": 4.486895312934411e-06,
+ "loss": 0.3073612689971924,
+ "num_input_tokens_seen": 691981536,
+ "step": 15825,
+ "train_runtime": 122391.6825,
+ "train_tokens_per_second": 5653.828
+ },
+ {
+ "epoch": 1.0658497172098034,
+ "grad_norm": 0.799787823954383,
+ "learning_rate": 4.484265261984378e-06,
+ "loss": 0.3512305974960327,
+ "num_input_tokens_seen": 692199872,
+ "step": 15830,
+ "train_runtime": 122434.3321,
+ "train_tokens_per_second": 5653.642
+ },
+ {
+ "epoch": 1.0661863722057636,
+ "grad_norm": 0.7027476090234593,
+ "learning_rate": 4.4816353552579165e-06,
+ "loss": 0.292151141166687,
+ "num_input_tokens_seen": 692412728,
+ "step": 15835,
+ "train_runtime": 122476.3937,
+ "train_tokens_per_second": 5653.438
+ },
+ {
+ "epoch": 1.0665230272017237,
+ "grad_norm": 0.8159747493895196,
+ "learning_rate": 4.479005593490477e-06,
+ "loss": 0.3161210536956787,
+ "num_input_tokens_seen": 692630960,
+ "step": 15840,
+ "train_runtime": 122513.6448,
+ "train_tokens_per_second": 5653.501
+ },
+ {
+ "epoch": 1.066859682197684,
+ "grad_norm": 0.7411839574837649,
+ "learning_rate": 4.476375977417461e-06,
+ "loss": 0.28519272804260254,
+ "num_input_tokens_seen": 692854048,
+ "step": 15845,
+ "train_runtime": 122555.3496,
+ "train_tokens_per_second": 5653.397
+ },
+ {
+ "epoch": 1.067196337193644,
+ "grad_norm": 0.7780714296454672,
+ "learning_rate": 4.473746507774235e-06,
+ "loss": 0.3064600944519043,
+ "num_input_tokens_seen": 693081408,
+ "step": 15850,
+ "train_runtime": 122591.5693,
+ "train_tokens_per_second": 5653.581
+ },
+ {
+ "epoch": 1.0675329921896042,
+ "grad_norm": 0.6762102135441501,
+ "learning_rate": 4.471117185296117e-06,
+ "loss": 0.28726582527160643,
+ "num_input_tokens_seen": 693299008,
+ "step": 15855,
+ "train_runtime": 122633.3438,
+ "train_tokens_per_second": 5653.43
+ },
+ {
+ "epoch": 1.0678696471855642,
+ "grad_norm": 0.6740157289152823,
+ "learning_rate": 4.468488010718394e-06,
+ "loss": 0.2859110593795776,
+ "num_input_tokens_seen": 693503552,
+ "step": 15860,
+ "train_runtime": 122669.1306,
+ "train_tokens_per_second": 5653.448
+ },
+ {
+ "epoch": 1.0682063021815245,
+ "grad_norm": 0.8206292602575027,
+ "learning_rate": 4.465858984776302e-06,
+ "loss": 0.31541666984558103,
+ "num_input_tokens_seen": 693719200,
+ "step": 15865,
+ "train_runtime": 122705.5825,
+ "train_tokens_per_second": 5653.526
+ },
+ {
+ "epoch": 1.0685429571774845,
+ "grad_norm": 0.8178983469469215,
+ "learning_rate": 4.463230108205044e-06,
+ "loss": 0.30719547271728515,
+ "num_input_tokens_seen": 693910832,
+ "step": 15870,
+ "train_runtime": 122745.905,
+ "train_tokens_per_second": 5653.23
+ },
+ {
+ "epoch": 1.0688796121734447,
+ "grad_norm": 0.7105445134284586,
+ "learning_rate": 4.460601381739772e-06,
+ "loss": 0.3158994674682617,
+ "num_input_tokens_seen": 694130912,
+ "step": 15875,
+ "train_runtime": 122784.3767,
+ "train_tokens_per_second": 5653.251
+ },
+ {
+ "epoch": 1.0692162671694048,
+ "grad_norm": 0.8182490674352466,
+ "learning_rate": 4.457972806115607e-06,
+ "loss": 0.33564248085021975,
+ "num_input_tokens_seen": 694345664,
+ "step": 15880,
+ "train_runtime": 122826.7003,
+ "train_tokens_per_second": 5653.052
+ },
+ {
+ "epoch": 1.069552922165365,
+ "grad_norm": 0.6520703883013584,
+ "learning_rate": 4.455344382067616e-06,
+ "loss": 0.31251869201660154,
+ "num_input_tokens_seen": 694563184,
+ "step": 15885,
+ "train_runtime": 122869.8883,
+ "train_tokens_per_second": 5652.835
+ },
+ {
+ "epoch": 1.069889577161325,
+ "grad_norm": 0.7437291287192324,
+ "learning_rate": 4.452716110330832e-06,
+ "loss": 0.3079527378082275,
+ "num_input_tokens_seen": 694784912,
+ "step": 15890,
+ "train_runtime": 122904.1033,
+ "train_tokens_per_second": 5653.065
+ },
+ {
+ "epoch": 1.0702262321572853,
+ "grad_norm": 0.8863362961200022,
+ "learning_rate": 4.450087991640242e-06,
+ "loss": 0.31774344444274905,
+ "num_input_tokens_seen": 694988912,
+ "step": 15895,
+ "train_runtime": 122939.5922,
+ "train_tokens_per_second": 5653.093
+ },
+ {
+ "epoch": 1.0705628871532453,
+ "grad_norm": 0.7816781374609185,
+ "learning_rate": 4.447460026730794e-06,
+ "loss": 0.3184868812561035,
+ "num_input_tokens_seen": 695201352,
+ "step": 15900,
+ "train_runtime": 122980.7158,
+ "train_tokens_per_second": 5652.93
+ },
+ {
+ "epoch": 1.0708995421492056,
+ "grad_norm": 0.7651374199427906,
+ "learning_rate": 4.444832216337384e-06,
+ "loss": 0.30800135135650636,
+ "num_input_tokens_seen": 695423576,
+ "step": 15905,
+ "train_runtime": 123019.1946,
+ "train_tokens_per_second": 5652.968
+ },
+ {
+ "epoch": 1.0712361971451656,
+ "grad_norm": 0.9298545556387856,
+ "learning_rate": 4.442204561194877e-06,
+ "loss": 0.29862637519836427,
+ "num_input_tokens_seen": 695642552,
+ "step": 15910,
+ "train_runtime": 123054.9881,
+ "train_tokens_per_second": 5653.103
+ },
+ {
+ "epoch": 1.0715728521411259,
+ "grad_norm": 0.679838258169619,
+ "learning_rate": 4.439577062038085e-06,
+ "loss": 0.29393310546875,
+ "num_input_tokens_seen": 695878256,
+ "step": 15915,
+ "train_runtime": 123095.8131,
+ "train_tokens_per_second": 5653.143
+ },
+ {
+ "epoch": 1.0719095071370859,
+ "grad_norm": 0.7450262188866336,
+ "learning_rate": 4.436949719601782e-06,
+ "loss": 0.3315741062164307,
+ "num_input_tokens_seen": 696096832,
+ "step": 15920,
+ "train_runtime": 123128.696,
+ "train_tokens_per_second": 5653.409
+ },
+ {
+ "epoch": 1.0722461621330461,
+ "grad_norm": 0.9095527446118012,
+ "learning_rate": 4.434322534620692e-06,
+ "loss": 0.3238795757293701,
+ "num_input_tokens_seen": 696300144,
+ "step": 15925,
+ "train_runtime": 123165.6915,
+ "train_tokens_per_second": 5653.361
+ },
+ {
+ "epoch": 1.0725828171290062,
+ "grad_norm": 0.6843805205148538,
+ "learning_rate": 4.431695507829504e-06,
+ "loss": 0.2765275716781616,
+ "num_input_tokens_seen": 696521264,
+ "step": 15930,
+ "train_runtime": 123205.1311,
+ "train_tokens_per_second": 5653.346
+ },
+ {
+ "epoch": 1.0729194721249664,
+ "grad_norm": 0.7473289414033233,
+ "learning_rate": 4.429068639962853e-06,
+ "loss": 0.3107874870300293,
+ "num_input_tokens_seen": 696737432,
+ "step": 15935,
+ "train_runtime": 123245.4597,
+ "train_tokens_per_second": 5653.25
+ },
+ {
+ "epoch": 1.0732561271209264,
+ "grad_norm": 0.7774027394469076,
+ "learning_rate": 4.426441931755337e-06,
+ "loss": 0.3024940490722656,
+ "num_input_tokens_seen": 696961616,
+ "step": 15940,
+ "train_runtime": 123286.7421,
+ "train_tokens_per_second": 5653.176
+ },
+ {
+ "epoch": 1.0735927821168867,
+ "grad_norm": 0.706587053052343,
+ "learning_rate": 4.4238153839415045e-06,
+ "loss": 0.32619800567626955,
+ "num_input_tokens_seen": 697193784,
+ "step": 15945,
+ "train_runtime": 123322.6979,
+ "train_tokens_per_second": 5653.41
+ },
+ {
+ "epoch": 1.0739294371128467,
+ "grad_norm": 0.8216197763150708,
+ "learning_rate": 4.421188997255865e-06,
+ "loss": 0.3372673749923706,
+ "num_input_tokens_seen": 697403080,
+ "step": 15950,
+ "train_runtime": 123358.9718,
+ "train_tokens_per_second": 5653.444
+ },
+ {
+ "epoch": 1.074266092108807,
+ "grad_norm": 0.7673913799601232,
+ "learning_rate": 4.418562772432874e-06,
+ "loss": 0.310252046585083,
+ "num_input_tokens_seen": 697621144,
+ "step": 15955,
+ "train_runtime": 123396.4433,
+ "train_tokens_per_second": 5653.495
+ },
+ {
+ "epoch": 1.074602747104767,
+ "grad_norm": 0.7842485455879602,
+ "learning_rate": 4.415936710206952e-06,
+ "loss": 0.31459197998046873,
+ "num_input_tokens_seen": 697855736,
+ "step": 15960,
+ "train_runtime": 123434.5672,
+ "train_tokens_per_second": 5653.649
+ },
+ {
+ "epoch": 1.0749394021007272,
+ "grad_norm": 0.8604702018619234,
+ "learning_rate": 4.413310811312465e-06,
+ "loss": 0.31325018405914307,
+ "num_input_tokens_seen": 698071128,
+ "step": 15965,
+ "train_runtime": 123468.5762,
+ "train_tokens_per_second": 5653.836
+ },
+ {
+ "epoch": 1.0752760570966873,
+ "grad_norm": 0.746270251899186,
+ "learning_rate": 4.41068507648374e-06,
+ "loss": 0.2819394826889038,
+ "num_input_tokens_seen": 698297384,
+ "step": 15970,
+ "train_runtime": 123503.2252,
+ "train_tokens_per_second": 5654.082
+ },
+ {
+ "epoch": 1.0756127120926475,
+ "grad_norm": 0.7078829630924763,
+ "learning_rate": 4.408059506455053e-06,
+ "loss": 0.29783015251159667,
+ "num_input_tokens_seen": 698536768,
+ "step": 15975,
+ "train_runtime": 123541.9583,
+ "train_tokens_per_second": 5654.247
+ },
+ {
+ "epoch": 1.0759493670886076,
+ "grad_norm": 0.6756334175310369,
+ "learning_rate": 4.40543410196064e-06,
+ "loss": 0.3107433795928955,
+ "num_input_tokens_seen": 698759848,
+ "step": 15980,
+ "train_runtime": 123579.0597,
+ "train_tokens_per_second": 5654.355
+ },
+ {
+ "epoch": 1.0762860220845678,
+ "grad_norm": 0.8743925394845493,
+ "learning_rate": 4.402808863734681e-06,
+ "loss": 0.32532343864440916,
+ "num_input_tokens_seen": 698978376,
+ "step": 15985,
+ "train_runtime": 123625.0383,
+ "train_tokens_per_second": 5654.019
+ },
+ {
+ "epoch": 1.0766226770805278,
+ "grad_norm": 0.7689380089547567,
+ "learning_rate": 4.40018379251132e-06,
+ "loss": 0.3170433044433594,
+ "num_input_tokens_seen": 699198368,
+ "step": 15990,
+ "train_runtime": 123668.6043,
+ "train_tokens_per_second": 5653.807
+ },
+ {
+ "epoch": 1.076959332076488,
+ "grad_norm": 0.8662735325666648,
+ "learning_rate": 4.397558889024649e-06,
+ "loss": 0.350922966003418,
+ "num_input_tokens_seen": 699398432,
+ "step": 15995,
+ "train_runtime": 123704.5755,
+ "train_tokens_per_second": 5653.78
+ },
+ {
+ "epoch": 1.0772959870724481,
+ "grad_norm": 0.6339471211052848,
+ "learning_rate": 4.394934154008715e-06,
+ "loss": 0.3178732395172119,
+ "num_input_tokens_seen": 699622136,
+ "step": 16000,
+ "train_runtime": 123751.9275,
+ "train_tokens_per_second": 5653.424
+ },
+ {
+ "epoch": 1.0776326420684084,
+ "grad_norm": 0.7488843367203586,
+ "learning_rate": 4.392309588197514e-06,
+ "loss": 0.3029392957687378,
+ "num_input_tokens_seen": 699844832,
+ "step": 16005,
+ "train_runtime": 123789.2847,
+ "train_tokens_per_second": 5653.517
+ },
+ {
+ "epoch": 1.0779692970643684,
+ "grad_norm": 0.7019722238512428,
+ "learning_rate": 4.389685192324999e-06,
+ "loss": 0.30281388759613037,
+ "num_input_tokens_seen": 700062776,
+ "step": 16010,
+ "train_runtime": 123831.9289,
+ "train_tokens_per_second": 5653.33
+ },
+ {
+ "epoch": 1.0783059520603286,
+ "grad_norm": 0.6922589184425343,
+ "learning_rate": 4.387060967125073e-06,
+ "loss": 0.3248000144958496,
+ "num_input_tokens_seen": 700270472,
+ "step": 16015,
+ "train_runtime": 123866.5766,
+ "train_tokens_per_second": 5653.426
+ },
+ {
+ "epoch": 1.0786426070562887,
+ "grad_norm": 0.8480144418378566,
+ "learning_rate": 4.384436913331596e-06,
+ "loss": 0.3387366533279419,
+ "num_input_tokens_seen": 700490600,
+ "step": 16020,
+ "train_runtime": 123908.611,
+ "train_tokens_per_second": 5653.284
+ },
+ {
+ "epoch": 1.078979262052249,
+ "grad_norm": 0.7822749047395506,
+ "learning_rate": 4.38181303167837e-06,
+ "loss": 0.3053700923919678,
+ "num_input_tokens_seen": 700708944,
+ "step": 16025,
+ "train_runtime": 123949.1805,
+ "train_tokens_per_second": 5653.195
+ },
+ {
+ "epoch": 1.079315917048209,
+ "grad_norm": 0.7151897163772738,
+ "learning_rate": 4.379189322899159e-06,
+ "loss": 0.29577248096466063,
+ "num_input_tokens_seen": 700917296,
+ "step": 16030,
+ "train_runtime": 123988.781,
+ "train_tokens_per_second": 5653.07
+ },
+ {
+ "epoch": 1.0796525720441692,
+ "grad_norm": 0.6994370560748215,
+ "learning_rate": 4.376565787727676e-06,
+ "loss": 0.30431132316589354,
+ "num_input_tokens_seen": 701140832,
+ "step": 16035,
+ "train_runtime": 124020.3139,
+ "train_tokens_per_second": 5653.435
+ },
+ {
+ "epoch": 1.0799892270401292,
+ "grad_norm": 0.7403775219598513,
+ "learning_rate": 4.373942426897581e-06,
+ "loss": 0.30581619739532473,
+ "num_input_tokens_seen": 701344984,
+ "step": 16040,
+ "train_runtime": 124056.6795,
+ "train_tokens_per_second": 5653.424
+ },
+ {
+ "epoch": 1.0803258820360895,
+ "grad_norm": 0.8197194371899179,
+ "learning_rate": 4.37131924114249e-06,
+ "loss": 0.32230820655822756,
+ "num_input_tokens_seen": 701550320,
+ "step": 16045,
+ "train_runtime": 124096.2964,
+ "train_tokens_per_second": 5653.274
+ },
+ {
+ "epoch": 1.0806625370320495,
+ "grad_norm": 0.7798066192591843,
+ "learning_rate": 4.368696231195968e-06,
+ "loss": 0.3164224624633789,
+ "num_input_tokens_seen": 701773408,
+ "step": 16050,
+ "train_runtime": 124133.9174,
+ "train_tokens_per_second": 5653.357
+ },
+ {
+ "epoch": 1.0809991920280098,
+ "grad_norm": 0.8232305871133864,
+ "learning_rate": 4.366073397791534e-06,
+ "loss": 0.29906353950500486,
+ "num_input_tokens_seen": 701989056,
+ "step": 16055,
+ "train_runtime": 124171.9449,
+ "train_tokens_per_second": 5653.363
+ },
+ {
+ "epoch": 1.0813358470239698,
+ "grad_norm": 0.6949559219075683,
+ "learning_rate": 4.363450741662651e-06,
+ "loss": 0.28748531341552735,
+ "num_input_tokens_seen": 702206288,
+ "step": 16060,
+ "train_runtime": 124208.7109,
+ "train_tokens_per_second": 5653.438
+ },
+ {
+ "epoch": 1.08167250201993,
+ "grad_norm": 0.7364548266367604,
+ "learning_rate": 4.360828263542739e-06,
+ "loss": 0.2912368059158325,
+ "num_input_tokens_seen": 702427720,
+ "step": 16065,
+ "train_runtime": 124248.5519,
+ "train_tokens_per_second": 5653.408
+ },
+ {
+ "epoch": 1.08200915701589,
+ "grad_norm": 0.7936024878904875,
+ "learning_rate": 4.358205964165164e-06,
+ "loss": 0.3195831298828125,
+ "num_input_tokens_seen": 702647704,
+ "step": 16070,
+ "train_runtime": 124288.6557,
+ "train_tokens_per_second": 5653.353
+ },
+ {
+ "epoch": 1.0823458120118503,
+ "grad_norm": 0.7229626629457014,
+ "learning_rate": 4.355583844263247e-06,
+ "loss": 0.32871217727661134,
+ "num_input_tokens_seen": 702872568,
+ "step": 16075,
+ "train_runtime": 124334.326,
+ "train_tokens_per_second": 5653.085
+ },
+ {
+ "epoch": 1.0826824670078103,
+ "grad_norm": 0.8615005299982201,
+ "learning_rate": 4.352961904570251e-06,
+ "loss": 0.3125025272369385,
+ "num_input_tokens_seen": 703089672,
+ "step": 16080,
+ "train_runtime": 124374.4941,
+ "train_tokens_per_second": 5653.005
+ },
+ {
+ "epoch": 1.0830191220037706,
+ "grad_norm": 0.7771590178180461,
+ "learning_rate": 4.350340145819397e-06,
+ "loss": 0.31962432861328127,
+ "num_input_tokens_seen": 703303976,
+ "step": 16085,
+ "train_runtime": 124407.3656,
+ "train_tokens_per_second": 5653.234
+ },
+ {
+ "epoch": 1.0833557769997306,
+ "grad_norm": 0.7052567726864358,
+ "learning_rate": 4.347718568743847e-06,
+ "loss": 0.31935825347900393,
+ "num_input_tokens_seen": 703525192,
+ "step": 16090,
+ "train_runtime": 124447.9997,
+ "train_tokens_per_second": 5653.166
+ },
+ {
+ "epoch": 1.0836924319956909,
+ "grad_norm": 0.6884996142175045,
+ "learning_rate": 4.34509717407672e-06,
+ "loss": 0.32219245433807375,
+ "num_input_tokens_seen": 703758184,
+ "step": 16095,
+ "train_runtime": 124485.0197,
+ "train_tokens_per_second": 5653.356
+ },
+ {
+ "epoch": 1.084029086991651,
+ "grad_norm": 0.7945771259917277,
+ "learning_rate": 4.342475962551081e-06,
+ "loss": 0.3197646141052246,
+ "num_input_tokens_seen": 703976600,
+ "step": 16100,
+ "train_runtime": 124520.4953,
+ "train_tokens_per_second": 5653.5
+ },
+ {
+ "epoch": 1.0843657419876112,
+ "grad_norm": 0.7193933737753864,
+ "learning_rate": 4.339854934899943e-06,
+ "loss": 0.2914067268371582,
+ "num_input_tokens_seen": 704189712,
+ "step": 16105,
+ "train_runtime": 124558.1197,
+ "train_tokens_per_second": 5653.503
+ },
+ {
+ "epoch": 1.0847023969835712,
+ "grad_norm": 0.6245767712435051,
+ "learning_rate": 4.337234091856265e-06,
+ "loss": 0.2758800983428955,
+ "num_input_tokens_seen": 704420104,
+ "step": 16110,
+ "train_runtime": 124598.8261,
+ "train_tokens_per_second": 5653.505
+ },
+ {
+ "epoch": 1.0850390519795314,
+ "grad_norm": 0.7731352613680776,
+ "learning_rate": 4.334613434152961e-06,
+ "loss": 0.3056293249130249,
+ "num_input_tokens_seen": 704628384,
+ "step": 16115,
+ "train_runtime": 124633.8287,
+ "train_tokens_per_second": 5653.589
+ },
+ {
+ "epoch": 1.0853757069754915,
+ "grad_norm": 0.8100290972348857,
+ "learning_rate": 4.331992962522888e-06,
+ "loss": 0.3311766147613525,
+ "num_input_tokens_seen": 704848816,
+ "step": 16120,
+ "train_runtime": 124667.6455,
+ "train_tokens_per_second": 5653.823
+ },
+ {
+ "epoch": 1.0857123619714517,
+ "grad_norm": 0.7302554603263116,
+ "learning_rate": 4.329372677698854e-06,
+ "loss": 0.3087399244308472,
+ "num_input_tokens_seen": 705076168,
+ "step": 16125,
+ "train_runtime": 124705.2351,
+ "train_tokens_per_second": 5653.942
+ },
+ {
+ "epoch": 1.0860490169674117,
+ "grad_norm": 0.7471321256250719,
+ "learning_rate": 4.3267525804136114e-06,
+ "loss": 0.3147927761077881,
+ "num_input_tokens_seen": 705296784,
+ "step": 16130,
+ "train_runtime": 124743.6802,
+ "train_tokens_per_second": 5653.968
+ },
+ {
+ "epoch": 1.086385671963372,
+ "grad_norm": 0.795048965234782,
+ "learning_rate": 4.324132671399864e-06,
+ "loss": 0.3137171268463135,
+ "num_input_tokens_seen": 705500048,
+ "step": 16135,
+ "train_runtime": 124785.7851,
+ "train_tokens_per_second": 5653.689
+ },
+ {
+ "epoch": 1.086722326959332,
+ "grad_norm": 0.7470988295180053,
+ "learning_rate": 4.321512951390258e-06,
+ "loss": 0.3282973289489746,
+ "num_input_tokens_seen": 705720120,
+ "step": 16140,
+ "train_runtime": 124823.8383,
+ "train_tokens_per_second": 5653.729
+ },
+ {
+ "epoch": 1.0870589819552923,
+ "grad_norm": 0.7567687111417263,
+ "learning_rate": 4.318893421117393e-06,
+ "loss": 0.3061366081237793,
+ "num_input_tokens_seen": 705941352,
+ "step": 16145,
+ "train_runtime": 124862.2769,
+ "train_tokens_per_second": 5653.76
+ },
+ {
+ "epoch": 1.0873956369512523,
+ "grad_norm": 0.7636628451001315,
+ "learning_rate": 4.31627408131381e-06,
+ "loss": 0.2968632698059082,
+ "num_input_tokens_seen": 706150776,
+ "step": 16150,
+ "train_runtime": 124908.0524,
+ "train_tokens_per_second": 5653.365
+ },
+ {
+ "epoch": 1.0877322919472125,
+ "grad_norm": 0.7225164738453893,
+ "learning_rate": 4.313654932712001e-06,
+ "loss": 0.2813676357269287,
+ "num_input_tokens_seen": 706367696,
+ "step": 16155,
+ "train_runtime": 124945.3002,
+ "train_tokens_per_second": 5653.415
+ },
+ {
+ "epoch": 1.0880689469431726,
+ "grad_norm": 0.7624100309618296,
+ "learning_rate": 4.3110359760444e-06,
+ "loss": 0.3084303617477417,
+ "num_input_tokens_seen": 706598168,
+ "step": 16160,
+ "train_runtime": 124977.2044,
+ "train_tokens_per_second": 5653.816
+ },
+ {
+ "epoch": 1.0884056019391328,
+ "grad_norm": 0.8194505401549996,
+ "learning_rate": 4.308417212043392e-06,
+ "loss": 0.2846923589706421,
+ "num_input_tokens_seen": 706810256,
+ "step": 16165,
+ "train_runtime": 125015.2796,
+ "train_tokens_per_second": 5653.791
+ },
+ {
+ "epoch": 1.0887422569350929,
+ "grad_norm": 1.0425941904759917,
+ "learning_rate": 4.305798641441304e-06,
+ "loss": 0.2945555210113525,
+ "num_input_tokens_seen": 707037952,
+ "step": 16170,
+ "train_runtime": 125057.1296,
+ "train_tokens_per_second": 5653.72
+ },
+ {
+ "epoch": 1.089078911931053,
+ "grad_norm": 0.7001009592549052,
+ "learning_rate": 4.303180264970416e-06,
+ "loss": 0.3075662851333618,
+ "num_input_tokens_seen": 707251272,
+ "step": 16175,
+ "train_runtime": 125103.0693,
+ "train_tokens_per_second": 5653.349
+ },
+ {
+ "epoch": 1.0894155669270131,
+ "grad_norm": 0.6728629250761784,
+ "learning_rate": 4.3005620833629415e-06,
+ "loss": 0.2843637466430664,
+ "num_input_tokens_seen": 707466896,
+ "step": 16180,
+ "train_runtime": 125144.1227,
+ "train_tokens_per_second": 5653.217
+ },
+ {
+ "epoch": 1.0897522219229734,
+ "grad_norm": 0.8207014522019808,
+ "learning_rate": 4.297944097351053e-06,
+ "loss": 0.3128535270690918,
+ "num_input_tokens_seen": 707669312,
+ "step": 16185,
+ "train_runtime": 125187.0812,
+ "train_tokens_per_second": 5652.894
+ },
+ {
+ "epoch": 1.0900888769189334,
+ "grad_norm": 0.7403515636473879,
+ "learning_rate": 4.295326307666856e-06,
+ "loss": 0.32403197288513186,
+ "num_input_tokens_seen": 707901856,
+ "step": 16190,
+ "train_runtime": 125223.3339,
+ "train_tokens_per_second": 5653.115
+ },
+ {
+ "epoch": 1.0904255319148937,
+ "grad_norm": 0.7331845978184556,
+ "learning_rate": 4.292708715042411e-06,
+ "loss": 0.3054851531982422,
+ "num_input_tokens_seen": 708121472,
+ "step": 16195,
+ "train_runtime": 125262.4718,
+ "train_tokens_per_second": 5653.102
+ },
+ {
+ "epoch": 1.0907621869108537,
+ "grad_norm": 0.8734471364431988,
+ "learning_rate": 4.290091320209718e-06,
+ "loss": 0.32215137481689454,
+ "num_input_tokens_seen": 708352952,
+ "step": 16200,
+ "train_runtime": 125297.9813,
+ "train_tokens_per_second": 5653.347
+ },
+ {
+ "epoch": 1.091098841906814,
+ "grad_norm": 0.814424549291456,
+ "learning_rate": 4.287474123900726e-06,
+ "loss": 0.3153803825378418,
+ "num_input_tokens_seen": 708559152,
+ "step": 16205,
+ "train_runtime": 125333.9985,
+ "train_tokens_per_second": 5653.367
+ },
+ {
+ "epoch": 1.091435496902774,
+ "grad_norm": 0.7017503991057531,
+ "learning_rate": 4.284857126847321e-06,
+ "loss": 0.3124639272689819,
+ "num_input_tokens_seen": 708773424,
+ "step": 16210,
+ "train_runtime": 125375.3591,
+ "train_tokens_per_second": 5653.212
+ },
+ {
+ "epoch": 1.0917721518987342,
+ "grad_norm": 0.7054843976620787,
+ "learning_rate": 4.28224032978134e-06,
+ "loss": 0.32616453170776366,
+ "num_input_tokens_seen": 709001032,
+ "step": 16215,
+ "train_runtime": 125415.5584,
+ "train_tokens_per_second": 5653.214
+ },
+ {
+ "epoch": 1.0921088068946943,
+ "grad_norm": 0.6285131408838311,
+ "learning_rate": 4.279623733434561e-06,
+ "loss": 0.29432287216186526,
+ "num_input_tokens_seen": 709220648,
+ "step": 16220,
+ "train_runtime": 125453.1068,
+ "train_tokens_per_second": 5653.273
+ },
+ {
+ "epoch": 1.0924454618906545,
+ "grad_norm": 0.7070132737008057,
+ "learning_rate": 4.277007338538711e-06,
+ "loss": 0.28021812438964844,
+ "num_input_tokens_seen": 709458880,
+ "step": 16225,
+ "train_runtime": 125489.4284,
+ "train_tokens_per_second": 5653.535
+ },
+ {
+ "epoch": 1.0927821168866145,
+ "grad_norm": 0.6568772287438737,
+ "learning_rate": 4.27439114582545e-06,
+ "loss": 0.278457498550415,
+ "num_input_tokens_seen": 709678888,
+ "step": 16230,
+ "train_runtime": 125528.1241,
+ "train_tokens_per_second": 5653.545
+ },
+ {
+ "epoch": 1.0931187718825748,
+ "grad_norm": 0.7614047551407584,
+ "learning_rate": 4.271775156026395e-06,
+ "loss": 0.3240532636642456,
+ "num_input_tokens_seen": 709881120,
+ "step": 16235,
+ "train_runtime": 125570.4013,
+ "train_tokens_per_second": 5653.252
+ },
+ {
+ "epoch": 1.0934554268785348,
+ "grad_norm": 0.6948637574677056,
+ "learning_rate": 4.269159369873093e-06,
+ "loss": 0.30025343894958495,
+ "num_input_tokens_seen": 710087584,
+ "step": 16240,
+ "train_runtime": 125608.5048,
+ "train_tokens_per_second": 5653.181
+ },
+ {
+ "epoch": 1.093792081874495,
+ "grad_norm": 0.6766287089415108,
+ "learning_rate": 4.266543788097043e-06,
+ "loss": 0.30265159606933595,
+ "num_input_tokens_seen": 710323016,
+ "step": 16245,
+ "train_runtime": 125642.5699,
+ "train_tokens_per_second": 5653.522
+ },
+ {
+ "epoch": 1.094128736870455,
+ "grad_norm": 0.775588372137403,
+ "learning_rate": 4.263928411429683e-06,
+ "loss": 0.34309751987457277,
+ "num_input_tokens_seen": 710535008,
+ "step": 16250,
+ "train_runtime": 125685.9241,
+ "train_tokens_per_second": 5653.258
+ },
+ {
+ "epoch": 1.0944653918664153,
+ "grad_norm": 0.7548771221924332,
+ "learning_rate": 4.261313240602398e-06,
+ "loss": 0.32427158355712893,
+ "num_input_tokens_seen": 710748160,
+ "step": 16255,
+ "train_runtime": 125721.4997,
+ "train_tokens_per_second": 5653.354
+ },
+ {
+ "epoch": 1.0948020468623754,
+ "grad_norm": 0.7250716525242203,
+ "learning_rate": 4.2586982763465064e-06,
+ "loss": 0.3176891326904297,
+ "num_input_tokens_seen": 710976808,
+ "step": 16260,
+ "train_runtime": 125755.3891,
+ "train_tokens_per_second": 5653.649
+ },
+ {
+ "epoch": 1.0951387018583356,
+ "grad_norm": 0.7160308232765491,
+ "learning_rate": 4.256083519393281e-06,
+ "loss": 0.30221972465515134,
+ "num_input_tokens_seen": 711195160,
+ "step": 16265,
+ "train_runtime": 125791.4899,
+ "train_tokens_per_second": 5653.762
+ },
+ {
+ "epoch": 1.0954753568542956,
+ "grad_norm": 0.7399122315680735,
+ "learning_rate": 4.253468970473925e-06,
+ "loss": 0.32169513702392577,
+ "num_input_tokens_seen": 711418152,
+ "step": 16270,
+ "train_runtime": 125823.4265,
+ "train_tokens_per_second": 5654.099
+ },
+ {
+ "epoch": 1.095812011850256,
+ "grad_norm": 0.7365264997369239,
+ "learning_rate": 4.250854630319593e-06,
+ "loss": 0.345171594619751,
+ "num_input_tokens_seen": 711640424,
+ "step": 16275,
+ "train_runtime": 125864.4895,
+ "train_tokens_per_second": 5654.021
+ },
+ {
+ "epoch": 1.096148666846216,
+ "grad_norm": 0.7088190228061766,
+ "learning_rate": 4.248240499661373e-06,
+ "loss": 0.3070409059524536,
+ "num_input_tokens_seen": 711848088,
+ "step": 16280,
+ "train_runtime": 125903.3761,
+ "train_tokens_per_second": 5653.924
+ },
+ {
+ "epoch": 1.0964853218421762,
+ "grad_norm": 0.7753388506945056,
+ "learning_rate": 4.245626579230302e-06,
+ "loss": 0.2884035587310791,
+ "num_input_tokens_seen": 712069664,
+ "step": 16285,
+ "train_runtime": 125935.0217,
+ "train_tokens_per_second": 5654.262
+ },
+ {
+ "epoch": 1.0968219768381362,
+ "grad_norm": 0.898320442866567,
+ "learning_rate": 4.243012869757351e-06,
+ "loss": 0.296952223777771,
+ "num_input_tokens_seen": 712306208,
+ "step": 16290,
+ "train_runtime": 125975.6337,
+ "train_tokens_per_second": 5654.317
+ },
+ {
+ "epoch": 1.0971586318340965,
+ "grad_norm": 0.6948335232020322,
+ "learning_rate": 4.240399371973438e-06,
+ "loss": 0.31108913421630857,
+ "num_input_tokens_seen": 712536912,
+ "step": 16295,
+ "train_runtime": 126015.4524,
+ "train_tokens_per_second": 5654.361
+ },
+ {
+ "epoch": 1.0974952868300565,
+ "grad_norm": 0.7536053030674238,
+ "learning_rate": 4.2377860866094165e-06,
+ "loss": 0.3221109390258789,
+ "num_input_tokens_seen": 712758800,
+ "step": 16300,
+ "train_runtime": 126050.8307,
+ "train_tokens_per_second": 5654.535
+ },
+ {
+ "epoch": 1.0978319418260167,
+ "grad_norm": 0.757140397293046,
+ "learning_rate": 4.235173014396088e-06,
+ "loss": 0.3079149007797241,
+ "num_input_tokens_seen": 712977904,
+ "step": 16305,
+ "train_runtime": 126086.8838,
+ "train_tokens_per_second": 5654.656
+ },
+ {
+ "epoch": 1.0981685968219768,
+ "grad_norm": 0.7077551119296236,
+ "learning_rate": 4.2325601560641854e-06,
+ "loss": 0.3280055046081543,
+ "num_input_tokens_seen": 713204864,
+ "step": 16310,
+ "train_runtime": 126127.6082,
+ "train_tokens_per_second": 5654.629
+ },
+ {
+ "epoch": 1.098505251817937,
+ "grad_norm": 0.742995716486435,
+ "learning_rate": 4.229947512344389e-06,
+ "loss": 0.3023355484008789,
+ "num_input_tokens_seen": 713422520,
+ "step": 16315,
+ "train_runtime": 126168.2023,
+ "train_tokens_per_second": 5654.535
+ },
+ {
+ "epoch": 1.098841906813897,
+ "grad_norm": 0.7574565860803252,
+ "learning_rate": 4.227335083967314e-06,
+ "loss": 0.3024130344390869,
+ "num_input_tokens_seen": 713640344,
+ "step": 16320,
+ "train_runtime": 126208.7161,
+ "train_tokens_per_second": 5654.446
+ },
+ {
+ "epoch": 1.0991785618098573,
+ "grad_norm": 0.6713642233636719,
+ "learning_rate": 4.224722871663522e-06,
+ "loss": 0.3010653018951416,
+ "num_input_tokens_seen": 713862120,
+ "step": 16325,
+ "train_runtime": 126241.4058,
+ "train_tokens_per_second": 5654.738
+ },
+ {
+ "epoch": 1.0995152168058173,
+ "grad_norm": 0.7224131004458003,
+ "learning_rate": 4.222110876163503e-06,
+ "loss": 0.3363258600234985,
+ "num_input_tokens_seen": 714091328,
+ "step": 16330,
+ "train_runtime": 126279.2124,
+ "train_tokens_per_second": 5654.86
+ },
+ {
+ "epoch": 1.0998518718017776,
+ "grad_norm": 0.788806961337755,
+ "learning_rate": 4.2194990981977e-06,
+ "loss": 0.2941534757614136,
+ "num_input_tokens_seen": 714318104,
+ "step": 16335,
+ "train_runtime": 126315.8939,
+ "train_tokens_per_second": 5655.014
+ },
+ {
+ "epoch": 1.1001885267977376,
+ "grad_norm": 0.7340574826834825,
+ "learning_rate": 4.216887538496482e-06,
+ "loss": 0.2995609760284424,
+ "num_input_tokens_seen": 714533544,
+ "step": 16340,
+ "train_runtime": 126356.39,
+ "train_tokens_per_second": 5654.906
+ },
+ {
+ "epoch": 1.1005251817936978,
+ "grad_norm": 0.7563051863525939,
+ "learning_rate": 4.2142761977901695e-06,
+ "loss": 0.3597081184387207,
+ "num_input_tokens_seen": 714751432,
+ "step": 16345,
+ "train_runtime": 126392.7269,
+ "train_tokens_per_second": 5655.004
+ },
+ {
+ "epoch": 1.1008618367896579,
+ "grad_norm": 0.7088542234512013,
+ "learning_rate": 4.21166507680901e-06,
+ "loss": 0.3166736364364624,
+ "num_input_tokens_seen": 714962360,
+ "step": 16350,
+ "train_runtime": 126422.3158,
+ "train_tokens_per_second": 5655.349
+ },
+ {
+ "epoch": 1.1011984917856181,
+ "grad_norm": 0.7396624505443076,
+ "learning_rate": 4.209054176283202e-06,
+ "loss": 0.31155028343200686,
+ "num_input_tokens_seen": 715172784,
+ "step": 16355,
+ "train_runtime": 126456.1758,
+ "train_tokens_per_second": 5655.499
+ },
+ {
+ "epoch": 1.1015351467815782,
+ "grad_norm": 0.762754435287207,
+ "learning_rate": 4.206443496942868e-06,
+ "loss": 0.3161408185958862,
+ "num_input_tokens_seen": 715390344,
+ "step": 16360,
+ "train_runtime": 126495.2215,
+ "train_tokens_per_second": 5655.473
+ },
+ {
+ "epoch": 1.1018718017775384,
+ "grad_norm": 0.976771706771798,
+ "learning_rate": 4.203833039518081e-06,
+ "loss": 0.2889822006225586,
+ "num_input_tokens_seen": 715601072,
+ "step": 16365,
+ "train_runtime": 126533.131,
+ "train_tokens_per_second": 5655.444
+ },
+ {
+ "epoch": 1.1022084567734984,
+ "grad_norm": 0.7390990491778792,
+ "learning_rate": 4.201222804738845e-06,
+ "loss": 0.29911801815032957,
+ "num_input_tokens_seen": 715832864,
+ "step": 16370,
+ "train_runtime": 126567.5723,
+ "train_tokens_per_second": 5655.737
+ },
+ {
+ "epoch": 1.1025451117694587,
+ "grad_norm": 0.7688021513330408,
+ "learning_rate": 4.198612793335107e-06,
+ "loss": 0.28797073364257814,
+ "num_input_tokens_seen": 716055080,
+ "step": 16375,
+ "train_runtime": 126611.3125,
+ "train_tokens_per_second": 5655.538
+ },
+ {
+ "epoch": 1.1028817667654187,
+ "grad_norm": 0.6787164905475817,
+ "learning_rate": 4.196003006036743e-06,
+ "loss": 0.3057976722717285,
+ "num_input_tokens_seen": 716266872,
+ "step": 16380,
+ "train_runtime": 126648.5655,
+ "train_tokens_per_second": 5655.547
+ },
+ {
+ "epoch": 1.103218421761379,
+ "grad_norm": 0.7945554316956748,
+ "learning_rate": 4.193393443573576e-06,
+ "loss": 0.3109589576721191,
+ "num_input_tokens_seen": 716490328,
+ "step": 16385,
+ "train_runtime": 126684.476,
+ "train_tokens_per_second": 5655.707
+ },
+ {
+ "epoch": 1.103555076757339,
+ "grad_norm": 0.7930645433989376,
+ "learning_rate": 4.19078410667536e-06,
+ "loss": 0.3338963985443115,
+ "num_input_tokens_seen": 716702856,
+ "step": 16390,
+ "train_runtime": 126724.8797,
+ "train_tokens_per_second": 5655.581
+ },
+ {
+ "epoch": 1.1038917317532992,
+ "grad_norm": 0.8073944896858126,
+ "learning_rate": 4.188174996071788e-06,
+ "loss": 0.32840120792388916,
+ "num_input_tokens_seen": 716893944,
+ "step": 16395,
+ "train_runtime": 126758.4514,
+ "train_tokens_per_second": 5655.591
+ },
+ {
+ "epoch": 1.1042283867492593,
+ "grad_norm": 0.7471631872859732,
+ "learning_rate": 4.18556611249249e-06,
+ "loss": 0.30612549781799314,
+ "num_input_tokens_seen": 717111816,
+ "step": 16400,
+ "train_runtime": 126792.6813,
+ "train_tokens_per_second": 5655.782
+ },
+ {
+ "epoch": 1.1045650417452195,
+ "grad_norm": 0.8308844608789487,
+ "learning_rate": 4.182957456667032e-06,
+ "loss": 0.3163015127182007,
+ "num_input_tokens_seen": 717335736,
+ "step": 16405,
+ "train_runtime": 126836.6466,
+ "train_tokens_per_second": 5655.587
+ },
+ {
+ "epoch": 1.1049016967411796,
+ "grad_norm": 0.7689707726304607,
+ "learning_rate": 4.1803490293249145e-06,
+ "loss": 0.3357135057449341,
+ "num_input_tokens_seen": 717557144,
+ "step": 16410,
+ "train_runtime": 126873.1786,
+ "train_tokens_per_second": 5655.704
+ },
+ {
+ "epoch": 1.1052383517371398,
+ "grad_norm": 0.866900091357806,
+ "learning_rate": 4.1777408311955785e-06,
+ "loss": 0.3057410717010498,
+ "num_input_tokens_seen": 717761472,
+ "step": 16415,
+ "train_runtime": 126909.3722,
+ "train_tokens_per_second": 5655.701
+ },
+ {
+ "epoch": 1.1055750067330998,
+ "grad_norm": 0.7300512821916881,
+ "learning_rate": 4.175132863008395e-06,
+ "loss": 0.33652937412261963,
+ "num_input_tokens_seen": 717986048,
+ "step": 16420,
+ "train_runtime": 126949.4267,
+ "train_tokens_per_second": 5655.686
+ },
+ {
+ "epoch": 1.10591166172906,
+ "grad_norm": 0.9194268863297489,
+ "learning_rate": 4.17252512549268e-06,
+ "loss": 0.3076220273971558,
+ "num_input_tokens_seen": 718179376,
+ "step": 16425,
+ "train_runtime": 126987.7182,
+ "train_tokens_per_second": 5655.503
+ },
+ {
+ "epoch": 1.10624831672502,
+ "grad_norm": 0.7503337161594836,
+ "learning_rate": 4.169917619377671e-06,
+ "loss": 0.3038554906845093,
+ "num_input_tokens_seen": 718392416,
+ "step": 16430,
+ "train_runtime": 127030.797,
+ "train_tokens_per_second": 5655.262
+ },
+ {
+ "epoch": 1.1065849717209804,
+ "grad_norm": 0.8669099466716282,
+ "learning_rate": 4.167310345392557e-06,
+ "loss": 0.34373929500579836,
+ "num_input_tokens_seen": 718606440,
+ "step": 16435,
+ "train_runtime": 127066.9236,
+ "train_tokens_per_second": 5655.338
+ },
+ {
+ "epoch": 1.1069216267169404,
+ "grad_norm": 0.7525899229778,
+ "learning_rate": 4.1647033042664455e-06,
+ "loss": 0.3425747871398926,
+ "num_input_tokens_seen": 718818792,
+ "step": 16440,
+ "train_runtime": 127105.1604,
+ "train_tokens_per_second": 5655.308
+ },
+ {
+ "epoch": 1.1072582817129006,
+ "grad_norm": 0.7395303188295309,
+ "learning_rate": 4.162096496728394e-06,
+ "loss": 0.3010833263397217,
+ "num_input_tokens_seen": 719039416,
+ "step": 16445,
+ "train_runtime": 127146.3098,
+ "train_tokens_per_second": 5655.213
+ },
+ {
+ "epoch": 1.1075949367088607,
+ "grad_norm": 0.7264360847991507,
+ "learning_rate": 4.1594899235073835e-06,
+ "loss": 0.2992356300354004,
+ "num_input_tokens_seen": 719268528,
+ "step": 16450,
+ "train_runtime": 127190.5278,
+ "train_tokens_per_second": 5655.048
+ },
+ {
+ "epoch": 1.107931591704821,
+ "grad_norm": 0.7474728546722258,
+ "learning_rate": 4.156883585332338e-06,
+ "loss": 0.33159723281860354,
+ "num_input_tokens_seen": 719497592,
+ "step": 16455,
+ "train_runtime": 127229.1668,
+ "train_tokens_per_second": 5655.131
+ },
+ {
+ "epoch": 1.108268246700781,
+ "grad_norm": 0.6867746878864566,
+ "learning_rate": 4.154277482932107e-06,
+ "loss": 0.31698899269104003,
+ "num_input_tokens_seen": 719703328,
+ "step": 16460,
+ "train_runtime": 127263.452,
+ "train_tokens_per_second": 5655.224
+ },
+ {
+ "epoch": 1.1086049016967412,
+ "grad_norm": 0.7188428455380581,
+ "learning_rate": 4.151671617035482e-06,
+ "loss": 0.30119948387145995,
+ "num_input_tokens_seen": 719915352,
+ "step": 16465,
+ "train_runtime": 127303.2856,
+ "train_tokens_per_second": 5655.12
+ },
+ {
+ "epoch": 1.1089415566927012,
+ "grad_norm": 0.7701204509330969,
+ "learning_rate": 4.149065988371184e-06,
+ "loss": 0.2912993669509888,
+ "num_input_tokens_seen": 720134040,
+ "step": 16470,
+ "train_runtime": 127336.2618,
+ "train_tokens_per_second": 5655.373
+ },
+ {
+ "epoch": 1.1092782116886615,
+ "grad_norm": 0.7513429280225311,
+ "learning_rate": 4.14646059766787e-06,
+ "loss": 0.30213978290557864,
+ "num_input_tokens_seen": 720337824,
+ "step": 16475,
+ "train_runtime": 127369.5743,
+ "train_tokens_per_second": 5655.494
+ },
+ {
+ "epoch": 1.1096148666846215,
+ "grad_norm": 0.6711604996544339,
+ "learning_rate": 4.143855445654127e-06,
+ "loss": 0.29933414459228513,
+ "num_input_tokens_seen": 720567936,
+ "step": 16480,
+ "train_runtime": 127409.6507,
+ "train_tokens_per_second": 5655.521
+ },
+ {
+ "epoch": 1.1099515216805818,
+ "grad_norm": 0.8475234440498322,
+ "learning_rate": 4.14125053305848e-06,
+ "loss": 0.2931232213973999,
+ "num_input_tokens_seen": 720770472,
+ "step": 16485,
+ "train_runtime": 127446.8624,
+ "train_tokens_per_second": 5655.459
+ },
+ {
+ "epoch": 1.1102881766765418,
+ "grad_norm": 0.7010224781863986,
+ "learning_rate": 4.1386458606093805e-06,
+ "loss": 0.30602593421936036,
+ "num_input_tokens_seen": 720990928,
+ "step": 16490,
+ "train_runtime": 127487.302,
+ "train_tokens_per_second": 5655.394
+ },
+ {
+ "epoch": 1.110624831672502,
+ "grad_norm": 0.7499446880256623,
+ "learning_rate": 4.1360414290352214e-06,
+ "loss": 0.3390794038772583,
+ "num_input_tokens_seen": 721222304,
+ "step": 16495,
+ "train_runtime": 127531.9109,
+ "train_tokens_per_second": 5655.23
+ },
+ {
+ "epoch": 1.110961486668462,
+ "grad_norm": 0.7785896374993136,
+ "learning_rate": 4.1334372390643205e-06,
+ "loss": 0.30769715309143064,
+ "num_input_tokens_seen": 721451296,
+ "step": 16500,
+ "train_runtime": 127572.4988,
+ "train_tokens_per_second": 5655.226
+ },
+ {
+ "epoch": 1.1112981416644223,
+ "grad_norm": 0.6703802322842832,
+ "learning_rate": 4.130833291424934e-06,
+ "loss": 0.32526416778564454,
+ "num_input_tokens_seen": 721672688,
+ "step": 16505,
+ "train_runtime": 127609.7092,
+ "train_tokens_per_second": 5655.312
+ },
+ {
+ "epoch": 1.1116347966603823,
+ "grad_norm": 0.8080822522673832,
+ "learning_rate": 4.128229586845244e-06,
+ "loss": 0.30047011375427246,
+ "num_input_tokens_seen": 721893296,
+ "step": 16510,
+ "train_runtime": 127652.4127,
+ "train_tokens_per_second": 5655.148
+ },
+ {
+ "epoch": 1.1119714516563426,
+ "grad_norm": 0.6398810192890917,
+ "learning_rate": 4.125626126053372e-06,
+ "loss": 0.2854645013809204,
+ "num_input_tokens_seen": 722117224,
+ "step": 16515,
+ "train_runtime": 127693.4791,
+ "train_tokens_per_second": 5655.083
+ },
+ {
+ "epoch": 1.1123081066523026,
+ "grad_norm": 0.7428717458853391,
+ "learning_rate": 4.123022909777364e-06,
+ "loss": 0.31157827377319336,
+ "num_input_tokens_seen": 722338424,
+ "step": 16520,
+ "train_runtime": 127724.0973,
+ "train_tokens_per_second": 5655.459
+ },
+ {
+ "epoch": 1.1126447616482629,
+ "grad_norm": 0.8547681854868342,
+ "learning_rate": 4.120419938745205e-06,
+ "loss": 0.31059443950653076,
+ "num_input_tokens_seen": 722572192,
+ "step": 16525,
+ "train_runtime": 127768.6413,
+ "train_tokens_per_second": 5655.317
+ },
+ {
+ "epoch": 1.112981416644223,
+ "grad_norm": 0.8535266815266339,
+ "learning_rate": 4.1178172136848035e-06,
+ "loss": 0.30442132949829104,
+ "num_input_tokens_seen": 722777960,
+ "step": 16530,
+ "train_runtime": 127805.8834,
+ "train_tokens_per_second": 5655.279
+ },
+ {
+ "epoch": 1.1133180716401831,
+ "grad_norm": 0.7715289136002668,
+ "learning_rate": 4.115214735324007e-06,
+ "loss": 0.29233555793762206,
+ "num_input_tokens_seen": 723011784,
+ "step": 16535,
+ "train_runtime": 127845.3251,
+ "train_tokens_per_second": 5655.363
+ },
+ {
+ "epoch": 1.1136547266361432,
+ "grad_norm": 0.6308478276358274,
+ "learning_rate": 4.112612504390586e-06,
+ "loss": 0.2731541872024536,
+ "num_input_tokens_seen": 723233976,
+ "step": 16540,
+ "train_runtime": 127887.594,
+ "train_tokens_per_second": 5655.232
+ },
+ {
+ "epoch": 1.1139913816321034,
+ "grad_norm": 0.7595556917317189,
+ "learning_rate": 4.11001052161225e-06,
+ "loss": 0.3273204803466797,
+ "num_input_tokens_seen": 723450040,
+ "step": 16545,
+ "train_runtime": 127924.3668,
+ "train_tokens_per_second": 5655.295
+ },
+ {
+ "epoch": 1.1143280366280637,
+ "grad_norm": 0.7539458359188388,
+ "learning_rate": 4.107408787716632e-06,
+ "loss": 0.3083867788314819,
+ "num_input_tokens_seen": 723674424,
+ "step": 16550,
+ "train_runtime": 127966.9008,
+ "train_tokens_per_second": 5655.169
+ },
+ {
+ "epoch": 1.1146646916240237,
+ "grad_norm": 0.6933389096869937,
+ "learning_rate": 4.104807303431303e-06,
+ "loss": 0.30321598052978516,
+ "num_input_tokens_seen": 723898176,
+ "step": 16555,
+ "train_runtime": 128013.6102,
+ "train_tokens_per_second": 5654.853
+ },
+ {
+ "epoch": 1.1150013466199837,
+ "grad_norm": 0.8314161994474075,
+ "learning_rate": 4.102206069483754e-06,
+ "loss": 0.32450571060180666,
+ "num_input_tokens_seen": 724110632,
+ "step": 16560,
+ "train_runtime": 128050.9045,
+ "train_tokens_per_second": 5654.865
+ },
+ {
+ "epoch": 1.115338001615944,
+ "grad_norm": 0.7582409248110241,
+ "learning_rate": 4.099605086601417e-06,
+ "loss": 0.28465938568115234,
+ "num_input_tokens_seen": 724312128,
+ "step": 16565,
+ "train_runtime": 128098.5725,
+ "train_tokens_per_second": 5654.334
+ },
+ {
+ "epoch": 1.1156746566119042,
+ "grad_norm": 0.7723226944103582,
+ "learning_rate": 4.097004355511645e-06,
+ "loss": 0.2872473239898682,
+ "num_input_tokens_seen": 724532224,
+ "step": 16570,
+ "train_runtime": 128135.0872,
+ "train_tokens_per_second": 5654.44
+ },
+ {
+ "epoch": 1.1160113116078643,
+ "grad_norm": 0.80551126446832,
+ "learning_rate": 4.094403876941728e-06,
+ "loss": 0.2786348819732666,
+ "num_input_tokens_seen": 724749712,
+ "step": 16575,
+ "train_runtime": 128172.2871,
+ "train_tokens_per_second": 5654.496
+ },
+ {
+ "epoch": 1.1163479666038243,
+ "grad_norm": 0.7081089427677039,
+ "learning_rate": 4.0918036516188755e-06,
+ "loss": 0.3147793531417847,
+ "num_input_tokens_seen": 724953832,
+ "step": 16580,
+ "train_runtime": 128215.8528,
+ "train_tokens_per_second": 5654.167
+ },
+ {
+ "epoch": 1.1166846215997845,
+ "grad_norm": 0.7234281848413817,
+ "learning_rate": 4.0892036802702395e-06,
+ "loss": 0.3022450923919678,
+ "num_input_tokens_seen": 725178424,
+ "step": 16585,
+ "train_runtime": 128258.0871,
+ "train_tokens_per_second": 5654.056
+ },
+ {
+ "epoch": 1.1170212765957448,
+ "grad_norm": 0.6801507964833571,
+ "learning_rate": 4.086603963622887e-06,
+ "loss": 0.2855158090591431,
+ "num_input_tokens_seen": 725402000,
+ "step": 16590,
+ "train_runtime": 128299.5805,
+ "train_tokens_per_second": 5653.97
+ },
+ {
+ "epoch": 1.1173579315917048,
+ "grad_norm": 0.712093664671178,
+ "learning_rate": 4.084004502403825e-06,
+ "loss": 0.32064228057861327,
+ "num_input_tokens_seen": 725622480,
+ "step": 16595,
+ "train_runtime": 128341.8001,
+ "train_tokens_per_second": 5653.828
+ },
+ {
+ "epoch": 1.1176945865876649,
+ "grad_norm": 0.7068967114722002,
+ "learning_rate": 4.081405297339982e-06,
+ "loss": 0.30630931854248045,
+ "num_input_tokens_seen": 725856392,
+ "step": 16600,
+ "train_runtime": 128375.2241,
+ "train_tokens_per_second": 5654.178
+ },
+ {
+ "epoch": 1.118031241583625,
+ "grad_norm": 0.7257036824000926,
+ "learning_rate": 4.0788063491582205e-06,
+ "loss": 0.3028933048248291,
+ "num_input_tokens_seen": 726063792,
+ "step": 16605,
+ "train_runtime": 128411.5161,
+ "train_tokens_per_second": 5654.195
+ },
+ {
+ "epoch": 1.1183678965795854,
+ "grad_norm": 0.8569267027392826,
+ "learning_rate": 4.076207658585323e-06,
+ "loss": 0.3076032638549805,
+ "num_input_tokens_seen": 726293720,
+ "step": 16610,
+ "train_runtime": 128453.0536,
+ "train_tokens_per_second": 5654.157
+ },
+ {
+ "epoch": 1.1187045515755454,
+ "grad_norm": 0.6859953010109835,
+ "learning_rate": 4.07360922634801e-06,
+ "loss": 0.31174705028533933,
+ "num_input_tokens_seen": 726522048,
+ "step": 16615,
+ "train_runtime": 128490.1199,
+ "train_tokens_per_second": 5654.303
+ },
+ {
+ "epoch": 1.1190412065715054,
+ "grad_norm": 0.8064627601454211,
+ "learning_rate": 4.071011053172922e-06,
+ "loss": 0.3052211284637451,
+ "num_input_tokens_seen": 726722760,
+ "step": 16620,
+ "train_runtime": 128528.4186,
+ "train_tokens_per_second": 5654.18
+ },
+ {
+ "epoch": 1.1193778615674657,
+ "grad_norm": 0.7719182024204673,
+ "learning_rate": 4.068413139786631e-06,
+ "loss": 0.29850754737854,
+ "num_input_tokens_seen": 726939496,
+ "step": 16625,
+ "train_runtime": 128567.5214,
+ "train_tokens_per_second": 5654.146
+ },
+ {
+ "epoch": 1.119714516563426,
+ "grad_norm": 0.7949931481273091,
+ "learning_rate": 4.065815486915635e-06,
+ "loss": 0.30380845069885254,
+ "num_input_tokens_seen": 727169648,
+ "step": 16630,
+ "train_runtime": 128605.7452,
+ "train_tokens_per_second": 5654.255
+ },
+ {
+ "epoch": 1.120051171559386,
+ "grad_norm": 0.7231190372167858,
+ "learning_rate": 4.063218095286359e-06,
+ "loss": 0.31917145252227785,
+ "num_input_tokens_seen": 727399600,
+ "step": 16635,
+ "train_runtime": 128642.9376,
+ "train_tokens_per_second": 5654.408
+ },
+ {
+ "epoch": 1.120387826555346,
+ "grad_norm": 0.7667748207910186,
+ "learning_rate": 4.0606209656251555e-06,
+ "loss": 0.30777583122253416,
+ "num_input_tokens_seen": 727623584,
+ "step": 16640,
+ "train_runtime": 128679.2099,
+ "train_tokens_per_second": 5654.554
+ },
+ {
+ "epoch": 1.1207244815513062,
+ "grad_norm": 0.7358231463616071,
+ "learning_rate": 4.058024098658304e-06,
+ "loss": 0.32628979682922366,
+ "num_input_tokens_seen": 727849768,
+ "step": 16645,
+ "train_runtime": 128721.1978,
+ "train_tokens_per_second": 5654.467
+ },
+ {
+ "epoch": 1.1210611365472665,
+ "grad_norm": 0.7289574965983563,
+ "learning_rate": 4.055427495112008e-06,
+ "loss": 0.33713221549987793,
+ "num_input_tokens_seen": 728072752,
+ "step": 16650,
+ "train_runtime": 128764.41,
+ "train_tokens_per_second": 5654.301
+ },
+ {
+ "epoch": 1.1213977915432265,
+ "grad_norm": 0.7302453800120214,
+ "learning_rate": 4.052831155712403e-06,
+ "loss": 0.28829092979431153,
+ "num_input_tokens_seen": 728297032,
+ "step": 16655,
+ "train_runtime": 128798.8316,
+ "train_tokens_per_second": 5654.531
+ },
+ {
+ "epoch": 1.1217344465391865,
+ "grad_norm": 0.6736314639659831,
+ "learning_rate": 4.050235081185544e-06,
+ "loss": 0.27930502891540526,
+ "num_input_tokens_seen": 728518648,
+ "step": 16660,
+ "train_runtime": 128837.9436,
+ "train_tokens_per_second": 5654.535
+ },
+ {
+ "epoch": 1.1220711015351468,
+ "grad_norm": 0.778226843935622,
+ "learning_rate": 4.047639272257415e-06,
+ "loss": 0.31568381786346433,
+ "num_input_tokens_seen": 728747776,
+ "step": 16665,
+ "train_runtime": 128881.478,
+ "train_tokens_per_second": 5654.403
+ },
+ {
+ "epoch": 1.122407756531107,
+ "grad_norm": 0.616922697066709,
+ "learning_rate": 4.045043729653927e-06,
+ "loss": 0.2794900894165039,
+ "num_input_tokens_seen": 728975240,
+ "step": 16670,
+ "train_runtime": 128920.0081,
+ "train_tokens_per_second": 5654.477
+ },
+ {
+ "epoch": 1.122744411527067,
+ "grad_norm": 0.7407200139322001,
+ "learning_rate": 4.0424484541009175e-06,
+ "loss": 0.3211031436920166,
+ "num_input_tokens_seen": 729207240,
+ "step": 16675,
+ "train_runtime": 128961.9626,
+ "train_tokens_per_second": 5654.437
+ },
+ {
+ "epoch": 1.1230810665230273,
+ "grad_norm": 0.8208779512348747,
+ "learning_rate": 4.039853446324141e-06,
+ "loss": 0.2974057197570801,
+ "num_input_tokens_seen": 729416088,
+ "step": 16680,
+ "train_runtime": 129004.1657,
+ "train_tokens_per_second": 5654.206
+ },
+ {
+ "epoch": 1.1234177215189873,
+ "grad_norm": 0.7898967563608909,
+ "learning_rate": 4.037258707049289e-06,
+ "loss": 0.2974997043609619,
+ "num_input_tokens_seen": 729633568,
+ "step": 16685,
+ "train_runtime": 129048.8248,
+ "train_tokens_per_second": 5653.934
+ },
+ {
+ "epoch": 1.1237543765149476,
+ "grad_norm": 0.6741454248192755,
+ "learning_rate": 4.034664237001969e-06,
+ "loss": 0.29579739570617675,
+ "num_input_tokens_seen": 729855424,
+ "step": 16690,
+ "train_runtime": 129087.6374,
+ "train_tokens_per_second": 5653.953
+ },
+ {
+ "epoch": 1.1240910315109076,
+ "grad_norm": 0.8311896058049114,
+ "learning_rate": 4.032070036907716e-06,
+ "loss": 0.3193542003631592,
+ "num_input_tokens_seen": 730062656,
+ "step": 16695,
+ "train_runtime": 129126.9171,
+ "train_tokens_per_second": 5653.838
+ },
+ {
+ "epoch": 1.1244276865068679,
+ "grad_norm": 0.7991665442055015,
+ "learning_rate": 4.029476107491991e-06,
+ "loss": 0.3203211784362793,
+ "num_input_tokens_seen": 730282736,
+ "step": 16700,
+ "train_runtime": 129171.6017,
+ "train_tokens_per_second": 5653.586
+ },
+ {
+ "epoch": 1.124764341502828,
+ "grad_norm": 0.7190813419301547,
+ "learning_rate": 4.0268824494801784e-06,
+ "loss": 0.30711615085601807,
+ "num_input_tokens_seen": 730503648,
+ "step": 16705,
+ "train_runtime": 129209.3295,
+ "train_tokens_per_second": 5653.645
+ },
+ {
+ "epoch": 1.1251009964987881,
+ "grad_norm": 0.7285352578803672,
+ "learning_rate": 4.024289063597585e-06,
+ "loss": 0.31215999126434324,
+ "num_input_tokens_seen": 730706704,
+ "step": 16710,
+ "train_runtime": 129251.1282,
+ "train_tokens_per_second": 5653.387
+ },
+ {
+ "epoch": 1.1254376514947482,
+ "grad_norm": 0.7346612645471816,
+ "learning_rate": 4.021695950569443e-06,
+ "loss": 0.29782519340515134,
+ "num_input_tokens_seen": 730929536,
+ "step": 16715,
+ "train_runtime": 129286.421,
+ "train_tokens_per_second": 5653.568
+ },
+ {
+ "epoch": 1.1257743064907082,
+ "grad_norm": 0.7837756415475194,
+ "learning_rate": 4.01910311112091e-06,
+ "loss": 0.3227576017379761,
+ "num_input_tokens_seen": 731146120,
+ "step": 16720,
+ "train_runtime": 129325.3701,
+ "train_tokens_per_second": 5653.54
+ },
+ {
+ "epoch": 1.1261109614866684,
+ "grad_norm": 0.742625616948461,
+ "learning_rate": 4.016510545977064e-06,
+ "loss": 0.30303683280944826,
+ "num_input_tokens_seen": 731374192,
+ "step": 16725,
+ "train_runtime": 129369.3334,
+ "train_tokens_per_second": 5653.381
+ },
+ {
+ "epoch": 1.1264476164826287,
+ "grad_norm": 0.8331891535993596,
+ "learning_rate": 4.013918255862908e-06,
+ "loss": 0.3152627944946289,
+ "num_input_tokens_seen": 731580072,
+ "step": 16730,
+ "train_runtime": 129413.4377,
+ "train_tokens_per_second": 5653.046
+ },
+ {
+ "epoch": 1.1267842714785887,
+ "grad_norm": 0.8423878577043675,
+ "learning_rate": 4.011326241503369e-06,
+ "loss": 0.2968592643737793,
+ "num_input_tokens_seen": 731789896,
+ "step": 16735,
+ "train_runtime": 129458.1727,
+ "train_tokens_per_second": 5652.713
+ },
+ {
+ "epoch": 1.127120926474549,
+ "grad_norm": 0.8161928635173408,
+ "learning_rate": 4.00873450362329e-06,
+ "loss": 0.32601251602172854,
+ "num_input_tokens_seen": 732012656,
+ "step": 16740,
+ "train_runtime": 129494.0456,
+ "train_tokens_per_second": 5652.867
+ },
+ {
+ "epoch": 1.127457581470509,
+ "grad_norm": 0.7114507295708142,
+ "learning_rate": 4.0061430429474476e-06,
+ "loss": 0.31506369113922117,
+ "num_input_tokens_seen": 732216280,
+ "step": 16745,
+ "train_runtime": 129530.2218,
+ "train_tokens_per_second": 5652.861
+ },
+ {
+ "epoch": 1.1277942364664693,
+ "grad_norm": 0.6959294825982753,
+ "learning_rate": 4.003551860200534e-06,
+ "loss": 0.3079033851623535,
+ "num_input_tokens_seen": 732421912,
+ "step": 16750,
+ "train_runtime": 129574.2624,
+ "train_tokens_per_second": 5652.526
+ },
+ {
+ "epoch": 1.1281308914624293,
+ "grad_norm": 0.7729992928736847,
+ "learning_rate": 4.000960956107167e-06,
+ "loss": 0.3259746551513672,
+ "num_input_tokens_seen": 732628200,
+ "step": 16755,
+ "train_runtime": 129616.0713,
+ "train_tokens_per_second": 5652.294
+ },
+ {
+ "epoch": 1.1284675464583895,
+ "grad_norm": 0.7253296171356767,
+ "learning_rate": 3.998370331391881e-06,
+ "loss": 0.3024310111999512,
+ "num_input_tokens_seen": 732828608,
+ "step": 16760,
+ "train_runtime": 129653.8465,
+ "train_tokens_per_second": 5652.193
+ },
+ {
+ "epoch": 1.1288042014543496,
+ "grad_norm": 0.8768820968683303,
+ "learning_rate": 3.995779986779139e-06,
+ "loss": 0.3032145738601685,
+ "num_input_tokens_seen": 733030240,
+ "step": 16765,
+ "train_runtime": 129696.2491,
+ "train_tokens_per_second": 5651.9
+ },
+ {
+ "epoch": 1.1291408564503098,
+ "grad_norm": 0.7571543524283587,
+ "learning_rate": 3.9931899229933204e-06,
+ "loss": 0.32830450534820554,
+ "num_input_tokens_seen": 733249296,
+ "step": 16770,
+ "train_runtime": 129729.1458,
+ "train_tokens_per_second": 5652.155
+ },
+ {
+ "epoch": 1.1294775114462698,
+ "grad_norm": 0.7464218260015815,
+ "learning_rate": 3.990600140758731e-06,
+ "loss": 0.3006105422973633,
+ "num_input_tokens_seen": 733474720,
+ "step": 16775,
+ "train_runtime": 129763.0451,
+ "train_tokens_per_second": 5652.416
+ },
+ {
+ "epoch": 1.12981416644223,
+ "grad_norm": 0.6221590957693296,
+ "learning_rate": 3.988010640799591e-06,
+ "loss": 0.2877049922943115,
+ "num_input_tokens_seen": 733694216,
+ "step": 16780,
+ "train_runtime": 129806.6164,
+ "train_tokens_per_second": 5652.21
+ },
+ {
+ "epoch": 1.1301508214381901,
+ "grad_norm": 0.8483974664089013,
+ "learning_rate": 3.985421423840051e-06,
+ "loss": 0.33561320304870607,
+ "num_input_tokens_seen": 733906352,
+ "step": 16785,
+ "train_runtime": 129844.4587,
+ "train_tokens_per_second": 5652.196
+ },
+ {
+ "epoch": 1.1304874764341504,
+ "grad_norm": 0.7448892261292801,
+ "learning_rate": 3.982832490604173e-06,
+ "loss": 0.28998355865478515,
+ "num_input_tokens_seen": 734133280,
+ "step": 16790,
+ "train_runtime": 129888.8092,
+ "train_tokens_per_second": 5652.013
+ },
+ {
+ "epoch": 1.1308241314301104,
+ "grad_norm": 0.8527357235171165,
+ "learning_rate": 3.9802438418159465e-06,
+ "loss": 0.3185850143432617,
+ "num_input_tokens_seen": 734371888,
+ "step": 16795,
+ "train_runtime": 129928.7672,
+ "train_tokens_per_second": 5652.112
+ },
+ {
+ "epoch": 1.1311607864260707,
+ "grad_norm": 0.7819044120460309,
+ "learning_rate": 3.977655478199275e-06,
+ "loss": 0.32431769371032715,
+ "num_input_tokens_seen": 734599608,
+ "step": 16800,
+ "train_runtime": 129969.4055,
+ "train_tokens_per_second": 5652.096
+ },
+ {
+ "epoch": 1.1314974414220307,
+ "grad_norm": 0.7998665202892938,
+ "learning_rate": 3.975067400477993e-06,
+ "loss": 0.3372927188873291,
+ "num_input_tokens_seen": 734795752,
+ "step": 16805,
+ "train_runtime": 130003.7743,
+ "train_tokens_per_second": 5652.111
+ },
+ {
+ "epoch": 1.131834096417991,
+ "grad_norm": 0.7814987669774647,
+ "learning_rate": 3.972479609375842e-06,
+ "loss": 0.31949615478515625,
+ "num_input_tokens_seen": 735021664,
+ "step": 16810,
+ "train_runtime": 130042.9188,
+ "train_tokens_per_second": 5652.147
+ },
+ {
+ "epoch": 1.132170751413951,
+ "grad_norm": 0.7598227623059015,
+ "learning_rate": 3.969892105616492e-06,
+ "loss": 0.31243975162506105,
+ "num_input_tokens_seen": 735246192,
+ "step": 16815,
+ "train_runtime": 130079.0722,
+ "train_tokens_per_second": 5652.302
+ },
+ {
+ "epoch": 1.1325074064099112,
+ "grad_norm": 1.1084494733823549,
+ "learning_rate": 3.967304889923529e-06,
+ "loss": 0.30039591789245607,
+ "num_input_tokens_seen": 735462560,
+ "step": 16820,
+ "train_runtime": 130118.6173,
+ "train_tokens_per_second": 5652.247
+ },
+ {
+ "epoch": 1.1328440614058712,
+ "grad_norm": 0.8395955074548532,
+ "learning_rate": 3.964717963020463e-06,
+ "loss": 0.3070765256881714,
+ "num_input_tokens_seen": 735678824,
+ "step": 16825,
+ "train_runtime": 130155.0309,
+ "train_tokens_per_second": 5652.327
+ },
+ {
+ "epoch": 1.1331807164018315,
+ "grad_norm": 0.7687120040297668,
+ "learning_rate": 3.962131325630715e-06,
+ "loss": 0.29083225727081297,
+ "num_input_tokens_seen": 735911416,
+ "step": 16830,
+ "train_runtime": 130203.9338,
+ "train_tokens_per_second": 5651.991
+ },
+ {
+ "epoch": 1.1335173713977915,
+ "grad_norm": 0.8425980087790549,
+ "learning_rate": 3.959544978477634e-06,
+ "loss": 0.33116788864135743,
+ "num_input_tokens_seen": 736118352,
+ "step": 16835,
+ "train_runtime": 130242.9795,
+ "train_tokens_per_second": 5651.885
+ },
+ {
+ "epoch": 1.1338540263937518,
+ "grad_norm": 0.6639463454020782,
+ "learning_rate": 3.9569589222844805e-06,
+ "loss": 0.31282644271850585,
+ "num_input_tokens_seen": 736333424,
+ "step": 16840,
+ "train_runtime": 130287.4137,
+ "train_tokens_per_second": 5651.608
+ },
+ {
+ "epoch": 1.1341906813897118,
+ "grad_norm": 0.7757792448332916,
+ "learning_rate": 3.954373157774439e-06,
+ "loss": 0.3130516052246094,
+ "num_input_tokens_seen": 736558888,
+ "step": 16845,
+ "train_runtime": 130324.2351,
+ "train_tokens_per_second": 5651.742
+ },
+ {
+ "epoch": 1.134527336385672,
+ "grad_norm": 0.8187765783249485,
+ "learning_rate": 3.951787685670609e-06,
+ "loss": 0.3079373359680176,
+ "num_input_tokens_seen": 736782672,
+ "step": 16850,
+ "train_runtime": 130363.4678,
+ "train_tokens_per_second": 5651.757
+ },
+ {
+ "epoch": 1.134863991381632,
+ "grad_norm": 0.7794193428094371,
+ "learning_rate": 3.949202506696012e-06,
+ "loss": 0.315033483505249,
+ "num_input_tokens_seen": 736994032,
+ "step": 16855,
+ "train_runtime": 130400.0871,
+ "train_tokens_per_second": 5651.791
+ },
+ {
+ "epoch": 1.1352006463775923,
+ "grad_norm": 0.7253716724703653,
+ "learning_rate": 3.946617621573581e-06,
+ "loss": 0.3076453685760498,
+ "num_input_tokens_seen": 737227128,
+ "step": 16860,
+ "train_runtime": 130441.2278,
+ "train_tokens_per_second": 5651.795
+ },
+ {
+ "epoch": 1.1355373013735524,
+ "grad_norm": 0.8240959096061262,
+ "learning_rate": 3.944033031026175e-06,
+ "loss": 0.2819002628326416,
+ "num_input_tokens_seen": 737425512,
+ "step": 16865,
+ "train_runtime": 130480.7005,
+ "train_tokens_per_second": 5651.606
+ },
+ {
+ "epoch": 1.1358739563695126,
+ "grad_norm": 0.7359564971034284,
+ "learning_rate": 3.941448735776563e-06,
+ "loss": 0.3037100791931152,
+ "num_input_tokens_seen": 737645856,
+ "step": 16870,
+ "train_runtime": 130515.6474,
+ "train_tokens_per_second": 5651.781
+ },
+ {
+ "epoch": 1.1362106113654726,
+ "grad_norm": 0.7004505644549724,
+ "learning_rate": 3.938864736547439e-06,
+ "loss": 0.31839513778686523,
+ "num_input_tokens_seen": 737874128,
+ "step": 16875,
+ "train_runtime": 130552.8691,
+ "train_tokens_per_second": 5651.918
+ },
+ {
+ "epoch": 1.1365472663614329,
+ "grad_norm": 1.9350046761541768,
+ "learning_rate": 3.936281034061405e-06,
+ "loss": 0.2992375135421753,
+ "num_input_tokens_seen": 738099488,
+ "step": 16880,
+ "train_runtime": 130593.8162,
+ "train_tokens_per_second": 5651.872
+ },
+ {
+ "epoch": 1.136883921357393,
+ "grad_norm": 0.7114720599403541,
+ "learning_rate": 3.93369762904099e-06,
+ "loss": 0.3111462354660034,
+ "num_input_tokens_seen": 738305920,
+ "step": 16885,
+ "train_runtime": 130639.1413,
+ "train_tokens_per_second": 5651.491
+ },
+ {
+ "epoch": 1.1372205763533532,
+ "grad_norm": 0.7657629856292947,
+ "learning_rate": 3.931114522208631e-06,
+ "loss": 0.3166337251663208,
+ "num_input_tokens_seen": 738508624,
+ "step": 16890,
+ "train_runtime": 130682.7388,
+ "train_tokens_per_second": 5651.157
+ },
+ {
+ "epoch": 1.1375572313493132,
+ "grad_norm": 0.7692834919287869,
+ "learning_rate": 3.928531714286689e-06,
+ "loss": 0.30945444107055664,
+ "num_input_tokens_seen": 738732992,
+ "step": 16895,
+ "train_runtime": 130719.4165,
+ "train_tokens_per_second": 5651.287
+ },
+ {
+ "epoch": 1.1378938863452734,
+ "grad_norm": 0.8278288727680423,
+ "learning_rate": 3.925949205997434e-06,
+ "loss": 0.32941608428955077,
+ "num_input_tokens_seen": 738942840,
+ "step": 16900,
+ "train_runtime": 130761.9137,
+ "train_tokens_per_second": 5651.056
+ },
+ {
+ "epoch": 1.1382305413412335,
+ "grad_norm": 0.7724135944921471,
+ "learning_rate": 3.923366998063062e-06,
+ "loss": 0.2874737739562988,
+ "num_input_tokens_seen": 739155240,
+ "step": 16905,
+ "train_runtime": 130801.6692,
+ "train_tokens_per_second": 5650.962
+ },
+ {
+ "epoch": 1.1385671963371937,
+ "grad_norm": 0.6482256296875065,
+ "learning_rate": 3.920785091205674e-06,
+ "loss": 0.27861971855163575,
+ "num_input_tokens_seen": 739385256,
+ "step": 16910,
+ "train_runtime": 130835.4441,
+ "train_tokens_per_second": 5651.261
+ },
+ {
+ "epoch": 1.1389038513331537,
+ "grad_norm": 0.8189363341827849,
+ "learning_rate": 3.918203486147294e-06,
+ "loss": 0.34842791557312014,
+ "num_input_tokens_seen": 739594936,
+ "step": 16915,
+ "train_runtime": 130876.0263,
+ "train_tokens_per_second": 5651.111
+ },
+ {
+ "epoch": 1.139240506329114,
+ "grad_norm": 0.8291380348485267,
+ "learning_rate": 3.9156221836098594e-06,
+ "loss": 0.3238081455230713,
+ "num_input_tokens_seen": 739819304,
+ "step": 16920,
+ "train_runtime": 130912.2289,
+ "train_tokens_per_second": 5651.262
+ },
+ {
+ "epoch": 1.139577161325074,
+ "grad_norm": 0.7047778409887597,
+ "learning_rate": 3.913041184315224e-06,
+ "loss": 0.3074013710021973,
+ "num_input_tokens_seen": 740033032,
+ "step": 16925,
+ "train_runtime": 130954.6922,
+ "train_tokens_per_second": 5651.062
+ },
+ {
+ "epoch": 1.1399138163210343,
+ "grad_norm": 0.8050080201528949,
+ "learning_rate": 3.910460488985154e-06,
+ "loss": 0.3118597030639648,
+ "num_input_tokens_seen": 740249656,
+ "step": 16930,
+ "train_runtime": 131003.4754,
+ "train_tokens_per_second": 5650.611
+ },
+ {
+ "epoch": 1.1402504713169943,
+ "grad_norm": 0.727734150500772,
+ "learning_rate": 3.9078800983413355e-06,
+ "loss": 0.3097679138183594,
+ "num_input_tokens_seen": 740474224,
+ "step": 16935,
+ "train_runtime": 131039.7311,
+ "train_tokens_per_second": 5650.761
+ },
+ {
+ "epoch": 1.1405871263129546,
+ "grad_norm": 0.7396507242655731,
+ "learning_rate": 3.905300013105365e-06,
+ "loss": 0.2923974752426147,
+ "num_input_tokens_seen": 740697112,
+ "step": 16940,
+ "train_runtime": 131079.5559,
+ "train_tokens_per_second": 5650.745
+ },
+ {
+ "epoch": 1.1409237813089146,
+ "grad_norm": 0.8383241709134363,
+ "learning_rate": 3.902720233998754e-06,
+ "loss": 0.30002388954162595,
+ "num_input_tokens_seen": 740910376,
+ "step": 16945,
+ "train_runtime": 131116.3686,
+ "train_tokens_per_second": 5650.785
+ },
+ {
+ "epoch": 1.1412604363048748,
+ "grad_norm": 0.7028544654776611,
+ "learning_rate": 3.90014076174293e-06,
+ "loss": 0.31175391674041747,
+ "num_input_tokens_seen": 741133216,
+ "step": 16950,
+ "train_runtime": 131156.2994,
+ "train_tokens_per_second": 5650.763
+ },
+ {
+ "epoch": 1.1415970913008349,
+ "grad_norm": 0.6785576189006745,
+ "learning_rate": 3.897561597059237e-06,
+ "loss": 0.2938354253768921,
+ "num_input_tokens_seen": 741357712,
+ "step": 16955,
+ "train_runtime": 131193.385,
+ "train_tokens_per_second": 5650.877
+ },
+ {
+ "epoch": 1.1419337462967951,
+ "grad_norm": 0.720906429135837,
+ "learning_rate": 3.894982740668927e-06,
+ "loss": 0.29590883255004885,
+ "num_input_tokens_seen": 741562560,
+ "step": 16960,
+ "train_runtime": 131227.1172,
+ "train_tokens_per_second": 5650.986
+ },
+ {
+ "epoch": 1.1422704012927551,
+ "grad_norm": 0.7501425891472119,
+ "learning_rate": 3.89240419329317e-06,
+ "loss": 0.3243727207183838,
+ "num_input_tokens_seen": 741804536,
+ "step": 16965,
+ "train_runtime": 131263.585,
+ "train_tokens_per_second": 5651.259
+ },
+ {
+ "epoch": 1.1426070562887154,
+ "grad_norm": 0.7110651173816877,
+ "learning_rate": 3.889825955653046e-06,
+ "loss": 0.31747913360595703,
+ "num_input_tokens_seen": 742023480,
+ "step": 16970,
+ "train_runtime": 131299.3451,
+ "train_tokens_per_second": 5651.388
+ },
+ {
+ "epoch": 1.1429437112846754,
+ "grad_norm": 0.7195903483502709,
+ "learning_rate": 3.887248028469558e-06,
+ "loss": 0.2832490921020508,
+ "num_input_tokens_seen": 742258408,
+ "step": 16975,
+ "train_runtime": 131333.0196,
+ "train_tokens_per_second": 5651.727
+ },
+ {
+ "epoch": 1.1432803662806357,
+ "grad_norm": 1.0763044494541907,
+ "learning_rate": 3.884670412463607e-06,
+ "loss": 0.2930821657180786,
+ "num_input_tokens_seen": 742474896,
+ "step": 16980,
+ "train_runtime": 131371.6188,
+ "train_tokens_per_second": 5651.715
+ },
+ {
+ "epoch": 1.1436170212765957,
+ "grad_norm": 0.7486899797692392,
+ "learning_rate": 3.8820931083560194e-06,
+ "loss": 0.31524713039398194,
+ "num_input_tokens_seen": 742680248,
+ "step": 16985,
+ "train_runtime": 131410.9057,
+ "train_tokens_per_second": 5651.588
+ },
+ {
+ "epoch": 1.143953676272556,
+ "grad_norm": 0.7165743405717724,
+ "learning_rate": 3.879516116867531e-06,
+ "loss": 0.29150216579437255,
+ "num_input_tokens_seen": 742894640,
+ "step": 16990,
+ "train_runtime": 131445.9029,
+ "train_tokens_per_second": 5651.714
+ },
+ {
+ "epoch": 1.144290331268516,
+ "grad_norm": 0.8121082083066077,
+ "learning_rate": 3.876939438718786e-06,
+ "loss": 0.30789361000061033,
+ "num_input_tokens_seen": 743112224,
+ "step": 16995,
+ "train_runtime": 131487.3253,
+ "train_tokens_per_second": 5651.588
+ },
+ {
+ "epoch": 1.1446269862644762,
+ "grad_norm": 0.7621739149221717,
+ "learning_rate": 3.874363074630345e-06,
+ "loss": 0.28911523818969725,
+ "num_input_tokens_seen": 743319176,
+ "step": 17000,
+ "train_runtime": 131524.9073,
+ "train_tokens_per_second": 5651.547
+ },
+ {
+ "epoch": 1.1449636412604363,
+ "grad_norm": 0.7737775948010384,
+ "learning_rate": 3.871787025322681e-06,
+ "loss": 0.3129148006439209,
+ "num_input_tokens_seen": 743545736,
+ "step": 17005,
+ "train_runtime": 131561.9387,
+ "train_tokens_per_second": 5651.678
+ },
+ {
+ "epoch": 1.1453002962563965,
+ "grad_norm": 0.7555308668544749,
+ "learning_rate": 3.86921129151618e-06,
+ "loss": 0.3087345361709595,
+ "num_input_tokens_seen": 743756824,
+ "step": 17010,
+ "train_runtime": 131596.877,
+ "train_tokens_per_second": 5651.782
+ },
+ {
+ "epoch": 1.1456369512523565,
+ "grad_norm": 0.7822771296052244,
+ "learning_rate": 3.866635873931133e-06,
+ "loss": 0.3323315143585205,
+ "num_input_tokens_seen": 743968848,
+ "step": 17015,
+ "train_runtime": 131635.333,
+ "train_tokens_per_second": 5651.741
+ },
+ {
+ "epoch": 1.1459736062483168,
+ "grad_norm": 0.750133051806931,
+ "learning_rate": 3.864060773287751e-06,
+ "loss": 0.3157732725143433,
+ "num_input_tokens_seen": 744191936,
+ "step": 17020,
+ "train_runtime": 131677.9874,
+ "train_tokens_per_second": 5651.605
+ },
+ {
+ "epoch": 1.1463102612442768,
+ "grad_norm": 0.750658874065227,
+ "learning_rate": 3.861485990306151e-06,
+ "loss": 0.3161588191986084,
+ "num_input_tokens_seen": 744416472,
+ "step": 17025,
+ "train_runtime": 131722.5155,
+ "train_tokens_per_second": 5651.399
+ },
+ {
+ "epoch": 1.146646916240237,
+ "grad_norm": 0.7657095158045504,
+ "learning_rate": 3.858911525706365e-06,
+ "loss": 0.3021253108978271,
+ "num_input_tokens_seen": 744645048,
+ "step": 17030,
+ "train_runtime": 131759.7192,
+ "train_tokens_per_second": 5651.538
+ },
+ {
+ "epoch": 1.146983571236197,
+ "grad_norm": 0.8025594627650351,
+ "learning_rate": 3.85633738020833e-06,
+ "loss": 0.34988272190093994,
+ "num_input_tokens_seen": 744878048,
+ "step": 17035,
+ "train_runtime": 131794.0086,
+ "train_tokens_per_second": 5651.835
+ },
+ {
+ "epoch": 1.1473202262321573,
+ "grad_norm": 0.7709002150890326,
+ "learning_rate": 3.853763554531902e-06,
+ "loss": 0.3174929618835449,
+ "num_input_tokens_seen": 745082920,
+ "step": 17040,
+ "train_runtime": 131829.24,
+ "train_tokens_per_second": 5651.879
+ },
+ {
+ "epoch": 1.1476568812281174,
+ "grad_norm": 0.715665166170693,
+ "learning_rate": 3.851190049396839e-06,
+ "loss": 0.2886343479156494,
+ "num_input_tokens_seen": 745302104,
+ "step": 17045,
+ "train_runtime": 131865.7716,
+ "train_tokens_per_second": 5651.975
+ },
+ {
+ "epoch": 1.1479935362240776,
+ "grad_norm": 0.7364852648051009,
+ "learning_rate": 3.848616865522815e-06,
+ "loss": 0.31344852447509763,
+ "num_input_tokens_seen": 745522704,
+ "step": 17050,
+ "train_runtime": 131907.8212,
+ "train_tokens_per_second": 5651.846
+ },
+ {
+ "epoch": 1.1483301912200377,
+ "grad_norm": 0.7409016478894407,
+ "learning_rate": 3.846044003629414e-06,
+ "loss": 0.30931525230407714,
+ "num_input_tokens_seen": 745742992,
+ "step": 17055,
+ "train_runtime": 131945.4567,
+ "train_tokens_per_second": 5651.904
+ },
+ {
+ "epoch": 1.148666846215998,
+ "grad_norm": 0.7178354231276903,
+ "learning_rate": 3.843471464436128e-06,
+ "loss": 0.2803271055221558,
+ "num_input_tokens_seen": 745942912,
+ "step": 17060,
+ "train_runtime": 131989.4344,
+ "train_tokens_per_second": 5651.535
+ },
+ {
+ "epoch": 1.149003501211958,
+ "grad_norm": 0.787297382407357,
+ "learning_rate": 3.840899248662358e-06,
+ "loss": 0.33543803691864016,
+ "num_input_tokens_seen": 746154376,
+ "step": 17065,
+ "train_runtime": 132025.9108,
+ "train_tokens_per_second": 5651.575
+ },
+ {
+ "epoch": 1.1493401562079182,
+ "grad_norm": 0.7696262614580517,
+ "learning_rate": 3.838327357027419e-06,
+ "loss": 0.32558441162109375,
+ "num_input_tokens_seen": 746367240,
+ "step": 17070,
+ "train_runtime": 132064.2315,
+ "train_tokens_per_second": 5651.547
+ },
+ {
+ "epoch": 1.1496768112038782,
+ "grad_norm": 0.726552691405123,
+ "learning_rate": 3.835755790250528e-06,
+ "loss": 0.3117891550064087,
+ "num_input_tokens_seen": 746592480,
+ "step": 17075,
+ "train_runtime": 132101.1976,
+ "train_tokens_per_second": 5651.671
+ },
+ {
+ "epoch": 1.1500134661998385,
+ "grad_norm": 0.650793107544695,
+ "learning_rate": 3.83318454905082e-06,
+ "loss": 0.2934911251068115,
+ "num_input_tokens_seen": 746818640,
+ "step": 17080,
+ "train_runtime": 132143.7046,
+ "train_tokens_per_second": 5651.564
+ },
+ {
+ "epoch": 1.1503501211957985,
+ "grad_norm": 0.7704916837868131,
+ "learning_rate": 3.830613634147331e-06,
+ "loss": 0.32155680656433105,
+ "num_input_tokens_seen": 747039496,
+ "step": 17085,
+ "train_runtime": 132184.8265,
+ "train_tokens_per_second": 5651.477
+ },
+ {
+ "epoch": 1.1506867761917587,
+ "grad_norm": 0.720317163934302,
+ "learning_rate": 3.82804304625901e-06,
+ "loss": 0.3073138236999512,
+ "num_input_tokens_seen": 747272792,
+ "step": 17090,
+ "train_runtime": 132220.525,
+ "train_tokens_per_second": 5651.716
+ },
+ {
+ "epoch": 1.1510234311877188,
+ "grad_norm": 0.7743031429870126,
+ "learning_rate": 3.825472786104715e-06,
+ "loss": 0.2988899707794189,
+ "num_input_tokens_seen": 747507048,
+ "step": 17095,
+ "train_runtime": 132256.7682,
+ "train_tokens_per_second": 5651.938
+ },
+ {
+ "epoch": 1.151360086183679,
+ "grad_norm": 0.8974847495064373,
+ "learning_rate": 3.8229028544032095e-06,
+ "loss": 0.3167643308639526,
+ "num_input_tokens_seen": 747707608,
+ "step": 17100,
+ "train_runtime": 132298.3666,
+ "train_tokens_per_second": 5651.677
+ },
+ {
+ "epoch": 1.151696741179639,
+ "grad_norm": 0.8824423710352778,
+ "learning_rate": 3.8203332518731666e-06,
+ "loss": 0.3144956111907959,
+ "num_input_tokens_seen": 747928560,
+ "step": 17105,
+ "train_runtime": 132333.3063,
+ "train_tokens_per_second": 5651.854
+ },
+ {
+ "epoch": 1.1520333961755993,
+ "grad_norm": 0.6533415062560854,
+ "learning_rate": 3.8177639792331695e-06,
+ "loss": 0.3034801483154297,
+ "num_input_tokens_seen": 748157856,
+ "step": 17110,
+ "train_runtime": 132372.8474,
+ "train_tokens_per_second": 5651.898
+ },
+ {
+ "epoch": 1.1523700511715593,
+ "grad_norm": 0.9418120724183658,
+ "learning_rate": 3.815195037201704e-06,
+ "loss": 0.34812791347503663,
+ "num_input_tokens_seen": 748364016,
+ "step": 17115,
+ "train_runtime": 132412.7976,
+ "train_tokens_per_second": 5651.75
+ },
+ {
+ "epoch": 1.1527067061675196,
+ "grad_norm": 0.7569217027983804,
+ "learning_rate": 3.8126264264971694e-06,
+ "loss": 0.321405029296875,
+ "num_input_tokens_seen": 748586232,
+ "step": 17120,
+ "train_runtime": 132454.6523,
+ "train_tokens_per_second": 5651.642
+ },
+ {
+ "epoch": 1.1530433611634796,
+ "grad_norm": 0.7524862294811008,
+ "learning_rate": 3.8100581478378673e-06,
+ "loss": 0.3137218952178955,
+ "num_input_tokens_seen": 748805128,
+ "step": 17125,
+ "train_runtime": 132492.9601,
+ "train_tokens_per_second": 5651.66
+ },
+ {
+ "epoch": 1.1533800161594399,
+ "grad_norm": 0.7796331102294818,
+ "learning_rate": 3.807490201942012e-06,
+ "loss": 0.3160563468933105,
+ "num_input_tokens_seen": 749027264,
+ "step": 17130,
+ "train_runtime": 132527.0569,
+ "train_tokens_per_second": 5651.882
+ },
+ {
+ "epoch": 1.1537166711553999,
+ "grad_norm": 0.774378899106255,
+ "learning_rate": 3.804922589527717e-06,
+ "loss": 0.29862165451049805,
+ "num_input_tokens_seen": 749243080,
+ "step": 17135,
+ "train_runtime": 132567.1494,
+ "train_tokens_per_second": 5651.8
+ },
+ {
+ "epoch": 1.1540533261513601,
+ "grad_norm": 0.7096741310187692,
+ "learning_rate": 3.8023553113130096e-06,
+ "loss": 0.3242902040481567,
+ "num_input_tokens_seen": 749478816,
+ "step": 17140,
+ "train_runtime": 132602.8293,
+ "train_tokens_per_second": 5652.058
+ },
+ {
+ "epoch": 1.1543899811473202,
+ "grad_norm": 0.7242850071624176,
+ "learning_rate": 3.799788368015821e-06,
+ "loss": 0.28427910804748535,
+ "num_input_tokens_seen": 749699336,
+ "step": 17145,
+ "train_runtime": 132642.0928,
+ "train_tokens_per_second": 5652.047
+ },
+ {
+ "epoch": 1.1547266361432804,
+ "grad_norm": 0.8411805719100419,
+ "learning_rate": 3.7972217603539873e-06,
+ "loss": 0.30901274681091306,
+ "num_input_tokens_seen": 749905168,
+ "step": 17150,
+ "train_runtime": 132681.4222,
+ "train_tokens_per_second": 5651.923
+ },
+ {
+ "epoch": 1.1550632911392404,
+ "grad_norm": 0.7527979126590726,
+ "learning_rate": 3.7946554890452524e-06,
+ "loss": 0.3213947772979736,
+ "num_input_tokens_seen": 750135464,
+ "step": 17155,
+ "train_runtime": 132719.6511,
+ "train_tokens_per_second": 5652.03
+ },
+ {
+ "epoch": 1.1553999461352007,
+ "grad_norm": 0.7550688889586102,
+ "learning_rate": 3.792089554807269e-06,
+ "loss": 0.3054837226867676,
+ "num_input_tokens_seen": 750362056,
+ "step": 17160,
+ "train_runtime": 132761.8356,
+ "train_tokens_per_second": 5651.941
+ },
+ {
+ "epoch": 1.1557366011311607,
+ "grad_norm": 0.8245885062079921,
+ "learning_rate": 3.789523958357587e-06,
+ "loss": 0.3161867618560791,
+ "num_input_tokens_seen": 750572160,
+ "step": 17165,
+ "train_runtime": 132802.089,
+ "train_tokens_per_second": 5651.81
+ },
+ {
+ "epoch": 1.156073256127121,
+ "grad_norm": 0.700436201047616,
+ "learning_rate": 3.7869587004136714e-06,
+ "loss": 0.2760681390762329,
+ "num_input_tokens_seen": 750802552,
+ "step": 17170,
+ "train_runtime": 132839.9318,
+ "train_tokens_per_second": 5651.934
+ },
+ {
+ "epoch": 1.156409911123081,
+ "grad_norm": 0.7117777773350229,
+ "learning_rate": 3.7843937816928864e-06,
+ "loss": 0.3197911262512207,
+ "num_input_tokens_seen": 751031672,
+ "step": 17175,
+ "train_runtime": 132884.5398,
+ "train_tokens_per_second": 5651.761
+ },
+ {
+ "epoch": 1.1567465661190413,
+ "grad_norm": 0.7963520839657313,
+ "learning_rate": 3.781829202912506e-06,
+ "loss": 0.3260814666748047,
+ "num_input_tokens_seen": 751243800,
+ "step": 17180,
+ "train_runtime": 132921.2705,
+ "train_tokens_per_second": 5651.795
+ },
+ {
+ "epoch": 1.1570832211150013,
+ "grad_norm": 0.735123307982365,
+ "learning_rate": 3.7792649647897023e-06,
+ "loss": 0.3070995807647705,
+ "num_input_tokens_seen": 751468240,
+ "step": 17185,
+ "train_runtime": 132963.3105,
+ "train_tokens_per_second": 5651.696
+ },
+ {
+ "epoch": 1.1574198761109615,
+ "grad_norm": 0.7383532177231918,
+ "learning_rate": 3.7767010680415595e-06,
+ "loss": 0.30820441246032715,
+ "num_input_tokens_seen": 751680248,
+ "step": 17190,
+ "train_runtime": 133002.5463,
+ "train_tokens_per_second": 5651.623
+ },
+ {
+ "epoch": 1.1577565311069216,
+ "grad_norm": 0.8213108428390675,
+ "learning_rate": 3.774137513385061e-06,
+ "loss": 0.3007169008255005,
+ "num_input_tokens_seen": 751897104,
+ "step": 17195,
+ "train_runtime": 133050.5633,
+ "train_tokens_per_second": 5651.213
+ },
+ {
+ "epoch": 1.1580931861028818,
+ "grad_norm": 0.6991961330303099,
+ "learning_rate": 3.7715743015371005e-06,
+ "loss": 0.316827392578125,
+ "num_input_tokens_seen": 752126600,
+ "step": 17200,
+ "train_runtime": 133089.0112,
+ "train_tokens_per_second": 5651.305
+ },
+ {
+ "epoch": 1.1584298410988418,
+ "grad_norm": 0.6752546760035666,
+ "learning_rate": 3.769011433214467e-06,
+ "loss": 0.28647899627685547,
+ "num_input_tokens_seen": 752350528,
+ "step": 17205,
+ "train_runtime": 133130.5351,
+ "train_tokens_per_second": 5651.224
+ },
+ {
+ "epoch": 1.158766496094802,
+ "grad_norm": 0.7193811538704217,
+ "learning_rate": 3.766448909133863e-06,
+ "loss": 0.32767200469970703,
+ "num_input_tokens_seen": 752583080,
+ "step": 17210,
+ "train_runtime": 133161.9958,
+ "train_tokens_per_second": 5651.636
+ },
+ {
+ "epoch": 1.1591031510907621,
+ "grad_norm": 0.774174630991004,
+ "learning_rate": 3.7638867300118864e-06,
+ "loss": 0.29516313076019285,
+ "num_input_tokens_seen": 752794792,
+ "step": 17215,
+ "train_runtime": 133198.9108,
+ "train_tokens_per_second": 5651.659
+ },
+ {
+ "epoch": 1.1594398060867224,
+ "grad_norm": 0.810354987252184,
+ "learning_rate": 3.7613248965650447e-06,
+ "loss": 0.2927243709564209,
+ "num_input_tokens_seen": 753018152,
+ "step": 17220,
+ "train_runtime": 133236.819,
+ "train_tokens_per_second": 5651.727
+ },
+ {
+ "epoch": 1.1597764610826824,
+ "grad_norm": 0.7109008764685462,
+ "learning_rate": 3.758763409509745e-06,
+ "loss": 0.3014249324798584,
+ "num_input_tokens_seen": 753245648,
+ "step": 17225,
+ "train_runtime": 133276.7031,
+ "train_tokens_per_second": 5651.743
+ },
+ {
+ "epoch": 1.1601131160786426,
+ "grad_norm": 0.8029128503628568,
+ "learning_rate": 3.7562022695623013e-06,
+ "loss": 0.3143294334411621,
+ "num_input_tokens_seen": 753463144,
+ "step": 17230,
+ "train_runtime": 133315.2382,
+ "train_tokens_per_second": 5651.741
+ },
+ {
+ "epoch": 1.1604497710746027,
+ "grad_norm": 0.6493145264817488,
+ "learning_rate": 3.7536414774389252e-06,
+ "loss": 0.3114176273345947,
+ "num_input_tokens_seen": 753687504,
+ "step": 17235,
+ "train_runtime": 133354.768,
+ "train_tokens_per_second": 5651.748
+ },
+ {
+ "epoch": 1.160786426070563,
+ "grad_norm": 0.7972007226513281,
+ "learning_rate": 3.7510810338557363e-06,
+ "loss": 0.2993874788284302,
+ "num_input_tokens_seen": 753893032,
+ "step": 17240,
+ "train_runtime": 133399.0853,
+ "train_tokens_per_second": 5651.411
+ },
+ {
+ "epoch": 1.161123081066523,
+ "grad_norm": 0.7446879758207031,
+ "learning_rate": 3.748520939528753e-06,
+ "loss": 0.30895147323608396,
+ "num_input_tokens_seen": 754112432,
+ "step": 17245,
+ "train_runtime": 133440.2549,
+ "train_tokens_per_second": 5651.311
+ },
+ {
+ "epoch": 1.1614597360624832,
+ "grad_norm": 0.7281872851418484,
+ "learning_rate": 3.7459611951739007e-06,
+ "loss": 0.3002096891403198,
+ "num_input_tokens_seen": 754338736,
+ "step": 17250,
+ "train_runtime": 133481.8765,
+ "train_tokens_per_second": 5651.245
+ },
+ {
+ "epoch": 1.1617963910584432,
+ "grad_norm": 0.7431634131632837,
+ "learning_rate": 3.7434018015069984e-06,
+ "loss": 0.3243889093399048,
+ "num_input_tokens_seen": 754559056,
+ "step": 17255,
+ "train_runtime": 133525.754,
+ "train_tokens_per_second": 5651.038
+ },
+ {
+ "epoch": 1.1621330460544035,
+ "grad_norm": 0.8324688332558176,
+ "learning_rate": 3.740842759243778e-06,
+ "loss": 0.3150045871734619,
+ "num_input_tokens_seen": 754778800,
+ "step": 17260,
+ "train_runtime": 133565.7133,
+ "train_tokens_per_second": 5650.992
+ },
+ {
+ "epoch": 1.1624697010503635,
+ "grad_norm": 0.6921071488758873,
+ "learning_rate": 3.738284069099862e-06,
+ "loss": 0.28803367614746095,
+ "num_input_tokens_seen": 755011760,
+ "step": 17265,
+ "train_runtime": 133605.0409,
+ "train_tokens_per_second": 5651.072
+ },
+ {
+ "epoch": 1.1628063560463238,
+ "grad_norm": 0.7280978950943039,
+ "learning_rate": 3.735725731790785e-06,
+ "loss": 0.31182174682617186,
+ "num_input_tokens_seen": 755236808,
+ "step": 17270,
+ "train_runtime": 133638.2275,
+ "train_tokens_per_second": 5651.353
+ },
+ {
+ "epoch": 1.1631430110422838,
+ "grad_norm": 0.7658534531818101,
+ "learning_rate": 3.733167748031974e-06,
+ "loss": 0.3179009914398193,
+ "num_input_tokens_seen": 755464248,
+ "step": 17275,
+ "train_runtime": 133676.324,
+ "train_tokens_per_second": 5651.444
+ },
+ {
+ "epoch": 1.163479666038244,
+ "grad_norm": 0.7211234703667541,
+ "learning_rate": 3.7306101185387646e-06,
+ "loss": 0.3087700366973877,
+ "num_input_tokens_seen": 755687464,
+ "step": 17280,
+ "train_runtime": 133717.1549,
+ "train_tokens_per_second": 5651.388
+ },
+ {
+ "epoch": 1.163816321034204,
+ "grad_norm": 0.7824239357354816,
+ "learning_rate": 3.728052844026386e-06,
+ "loss": 0.3141305923461914,
+ "num_input_tokens_seen": 755907672,
+ "step": 17285,
+ "train_runtime": 133760.0292,
+ "train_tokens_per_second": 5651.222
+ },
+ {
+ "epoch": 1.1641529760301643,
+ "grad_norm": 0.7951750585384382,
+ "learning_rate": 3.7254959252099745e-06,
+ "loss": 0.3007349014282227,
+ "num_input_tokens_seen": 756114320,
+ "step": 17290,
+ "train_runtime": 133800.4505,
+ "train_tokens_per_second": 5651.06
+ },
+ {
+ "epoch": 1.1644896310261243,
+ "grad_norm": 0.7638584288959342,
+ "learning_rate": 3.7229393628045617e-06,
+ "loss": 0.2961156368255615,
+ "num_input_tokens_seen": 756338136,
+ "step": 17295,
+ "train_runtime": 133836.5594,
+ "train_tokens_per_second": 5651.207
+ },
+ {
+ "epoch": 1.1648262860220846,
+ "grad_norm": 0.7194065626433747,
+ "learning_rate": 3.720383157525087e-06,
+ "loss": 0.2917820453643799,
+ "num_input_tokens_seen": 756556624,
+ "step": 17300,
+ "train_runtime": 133876.3071,
+ "train_tokens_per_second": 5651.161
+ },
+ {
+ "epoch": 1.1651629410180446,
+ "grad_norm": 0.8023605053763685,
+ "learning_rate": 3.717827310086379e-06,
+ "loss": 0.3004648685455322,
+ "num_input_tokens_seen": 756770512,
+ "step": 17305,
+ "train_runtime": 133914.1223,
+ "train_tokens_per_second": 5651.163
+ },
+ {
+ "epoch": 1.1654995960140049,
+ "grad_norm": 0.778357180423363,
+ "learning_rate": 3.7152718212031778e-06,
+ "loss": 0.31314473152160643,
+ "num_input_tokens_seen": 756987488,
+ "step": 17310,
+ "train_runtime": 133957.8798,
+ "train_tokens_per_second": 5650.937
+ },
+ {
+ "epoch": 1.165836251009965,
+ "grad_norm": 0.8275289683679823,
+ "learning_rate": 3.7127166915901123e-06,
+ "loss": 0.33815641403198243,
+ "num_input_tokens_seen": 757190888,
+ "step": 17315,
+ "train_runtime": 133995.0668,
+ "train_tokens_per_second": 5650.886
+ },
+ {
+ "epoch": 1.1661729060059252,
+ "grad_norm": 0.8112349556897875,
+ "learning_rate": 3.710161921961721e-06,
+ "loss": 0.3213828086853027,
+ "num_input_tokens_seen": 757406736,
+ "step": 17320,
+ "train_runtime": 134033.0013,
+ "train_tokens_per_second": 5650.897
+ },
+ {
+ "epoch": 1.1665095610018852,
+ "grad_norm": 0.7412368352707359,
+ "learning_rate": 3.707607513032434e-06,
+ "loss": 0.29859418869018556,
+ "num_input_tokens_seen": 757621448,
+ "step": 17325,
+ "train_runtime": 134070.8112,
+ "train_tokens_per_second": 5650.905
+ },
+ {
+ "epoch": 1.1668462159978454,
+ "grad_norm": 0.8530421631271973,
+ "learning_rate": 3.705053465516587e-06,
+ "loss": 0.2971299171447754,
+ "num_input_tokens_seen": 757842136,
+ "step": 17330,
+ "train_runtime": 134106.2945,
+ "train_tokens_per_second": 5651.056
+ },
+ {
+ "epoch": 1.1671828709938055,
+ "grad_norm": 0.7217367075989789,
+ "learning_rate": 3.702499780128407e-06,
+ "loss": 0.2950276851654053,
+ "num_input_tokens_seen": 758048560,
+ "step": 17335,
+ "train_runtime": 134146.644,
+ "train_tokens_per_second": 5650.895
+ },
+ {
+ "epoch": 1.1675195259897657,
+ "grad_norm": 1.8887784795675984,
+ "learning_rate": 3.6999464575820266e-06,
+ "loss": 0.3086616277694702,
+ "num_input_tokens_seen": 758266576,
+ "step": 17340,
+ "train_runtime": 134185.9605,
+ "train_tokens_per_second": 5650.864
+ },
+ {
+ "epoch": 1.1678561809857257,
+ "grad_norm": 0.6843592201114646,
+ "learning_rate": 3.697393498591473e-06,
+ "loss": 0.3014980792999268,
+ "num_input_tokens_seen": 758487384,
+ "step": 17345,
+ "train_runtime": 134230.8121,
+ "train_tokens_per_second": 5650.621
+ },
+ {
+ "epoch": 1.168192835981686,
+ "grad_norm": 0.8394323507587605,
+ "learning_rate": 3.694840903870676e-06,
+ "loss": 0.32812952995300293,
+ "num_input_tokens_seen": 758704816,
+ "step": 17350,
+ "train_runtime": 134266.4562,
+ "train_tokens_per_second": 5650.74
+ },
+ {
+ "epoch": 1.168529490977646,
+ "grad_norm": 0.7479174116708699,
+ "learning_rate": 3.6922886741334564e-06,
+ "loss": 0.3267998695373535,
+ "num_input_tokens_seen": 758924344,
+ "step": 17355,
+ "train_runtime": 134305.3103,
+ "train_tokens_per_second": 5650.74
+ },
+ {
+ "epoch": 1.1688661459736063,
+ "grad_norm": 0.7136771552457577,
+ "learning_rate": 3.689736810093541e-06,
+ "loss": 0.2991864442825317,
+ "num_input_tokens_seen": 759152032,
+ "step": 17360,
+ "train_runtime": 134342.9756,
+ "train_tokens_per_second": 5650.85
+ },
+ {
+ "epoch": 1.1692028009695663,
+ "grad_norm": 0.850190780847832,
+ "learning_rate": 3.6871853124645464e-06,
+ "loss": 0.31711387634277344,
+ "num_input_tokens_seen": 759378848,
+ "step": 17365,
+ "train_runtime": 134382.3965,
+ "train_tokens_per_second": 5650.88
+ },
+ {
+ "epoch": 1.1695394559655266,
+ "grad_norm": 0.6756862427689906,
+ "learning_rate": 3.684634181959994e-06,
+ "loss": 0.30228276252746583,
+ "num_input_tokens_seen": 759614152,
+ "step": 17370,
+ "train_runtime": 134422.1735,
+ "train_tokens_per_second": 5650.959
+ },
+ {
+ "epoch": 1.1698761109614866,
+ "grad_norm": 0.9293848953693671,
+ "learning_rate": 3.6820834192932974e-06,
+ "loss": 0.3172401666641235,
+ "num_input_tokens_seen": 759830040,
+ "step": 17375,
+ "train_runtime": 134460.803,
+ "train_tokens_per_second": 5650.941
+ },
+ {
+ "epoch": 1.1702127659574468,
+ "grad_norm": 0.7492817955305227,
+ "learning_rate": 3.679533025177772e-06,
+ "loss": 0.3354950904846191,
+ "num_input_tokens_seen": 760073280,
+ "step": 17380,
+ "train_runtime": 134496.7214,
+ "train_tokens_per_second": 5651.24
+ },
+ {
+ "epoch": 1.1705494209534069,
+ "grad_norm": 0.7424725386974647,
+ "learning_rate": 3.6769830003266234e-06,
+ "loss": 0.31007990837097166,
+ "num_input_tokens_seen": 760294432,
+ "step": 17385,
+ "train_runtime": 134528.7303,
+ "train_tokens_per_second": 5651.539
+ },
+ {
+ "epoch": 1.1708860759493671,
+ "grad_norm": 0.7790858025441731,
+ "learning_rate": 3.6744333454529607e-06,
+ "loss": 0.3161858081817627,
+ "num_input_tokens_seen": 760522824,
+ "step": 17390,
+ "train_runtime": 134571.7613,
+ "train_tokens_per_second": 5651.43
+ },
+ {
+ "epoch": 1.1712227309453271,
+ "grad_norm": 0.6903404727708745,
+ "learning_rate": 3.6718840612697847e-06,
+ "loss": 0.3075212001800537,
+ "num_input_tokens_seen": 760757728,
+ "step": 17395,
+ "train_runtime": 134611.0442,
+ "train_tokens_per_second": 5651.525
+ },
+ {
+ "epoch": 1.1715593859412874,
+ "grad_norm": 0.7365357926688636,
+ "learning_rate": 3.6693351484899986e-06,
+ "loss": 0.3011133670806885,
+ "num_input_tokens_seen": 760969040,
+ "step": 17400,
+ "train_runtime": 134643.3087,
+ "train_tokens_per_second": 5651.74
+ },
+ {
+ "epoch": 1.1718960409372474,
+ "grad_norm": 1.0260127216999277,
+ "learning_rate": 3.6667866078263937e-06,
+ "loss": 0.30281815528869627,
+ "num_input_tokens_seen": 761178944,
+ "step": 17405,
+ "train_runtime": 134680.0119,
+ "train_tokens_per_second": 5651.759
+ },
+ {
+ "epoch": 1.1722326959332077,
+ "grad_norm": 0.8315108294093327,
+ "learning_rate": 3.664238439991663e-06,
+ "loss": 0.34009361267089844,
+ "num_input_tokens_seen": 761396128,
+ "step": 17410,
+ "train_runtime": 134717.5677,
+ "train_tokens_per_second": 5651.795
+ },
+ {
+ "epoch": 1.1725693509291677,
+ "grad_norm": 0.7676177184346961,
+ "learning_rate": 3.6616906456983925e-06,
+ "loss": 0.3160113334655762,
+ "num_input_tokens_seen": 761617984,
+ "step": 17415,
+ "train_runtime": 134752.1162,
+ "train_tokens_per_second": 5651.993
+ },
+ {
+ "epoch": 1.172906005925128,
+ "grad_norm": 0.7720738410000789,
+ "learning_rate": 3.659143225659065e-06,
+ "loss": 0.32563726902008056,
+ "num_input_tokens_seen": 761835752,
+ "step": 17420,
+ "train_runtime": 134789.1882,
+ "train_tokens_per_second": 5652.054
+ },
+ {
+ "epoch": 1.173242660921088,
+ "grad_norm": 0.6926475609016922,
+ "learning_rate": 3.6565961805860595e-06,
+ "loss": 0.30316720008850095,
+ "num_input_tokens_seen": 762067720,
+ "step": 17425,
+ "train_runtime": 134833.4769,
+ "train_tokens_per_second": 5651.918
+ },
+ {
+ "epoch": 1.1735793159170482,
+ "grad_norm": 0.6768651518731568,
+ "learning_rate": 3.654049511191649e-06,
+ "loss": 0.3012946605682373,
+ "num_input_tokens_seen": 762275792,
+ "step": 17430,
+ "train_runtime": 134874.1565,
+ "train_tokens_per_second": 5651.756
+ },
+ {
+ "epoch": 1.1739159709130083,
+ "grad_norm": 0.7655463622291874,
+ "learning_rate": 3.6515032181879995e-06,
+ "loss": 0.3098176956176758,
+ "num_input_tokens_seen": 762491920,
+ "step": 17435,
+ "train_runtime": 134914.91,
+ "train_tokens_per_second": 5651.651
+ },
+ {
+ "epoch": 1.1742526259089685,
+ "grad_norm": 0.7242442794388552,
+ "learning_rate": 3.6489573022871755e-06,
+ "loss": 0.28149569034576416,
+ "num_input_tokens_seen": 762712840,
+ "step": 17440,
+ "train_runtime": 134952.0365,
+ "train_tokens_per_second": 5651.733
+ },
+ {
+ "epoch": 1.1745892809049285,
+ "grad_norm": 0.6756916904062972,
+ "learning_rate": 3.6464117642011333e-06,
+ "loss": 0.31446964740753175,
+ "num_input_tokens_seen": 762919536,
+ "step": 17445,
+ "train_runtime": 134991.2971,
+ "train_tokens_per_second": 5651.62
+ },
+ {
+ "epoch": 1.1749259359008888,
+ "grad_norm": 0.7799367018501008,
+ "learning_rate": 3.643866604641726e-06,
+ "loss": 0.31410171985626223,
+ "num_input_tokens_seen": 763152304,
+ "step": 17450,
+ "train_runtime": 135025.582,
+ "train_tokens_per_second": 5651.909
+ },
+ {
+ "epoch": 1.1752625908968488,
+ "grad_norm": 0.7436018486594765,
+ "learning_rate": 3.641321824320697e-06,
+ "loss": 0.29862222671508787,
+ "num_input_tokens_seen": 763371656,
+ "step": 17455,
+ "train_runtime": 135071.3304,
+ "train_tokens_per_second": 5651.619
+ },
+ {
+ "epoch": 1.175599245892809,
+ "grad_norm": 0.6881929025831984,
+ "learning_rate": 3.6387774239496893e-06,
+ "loss": 0.2896681785583496,
+ "num_input_tokens_seen": 763607368,
+ "step": 17460,
+ "train_runtime": 135110.4495,
+ "train_tokens_per_second": 5651.727
+ },
+ {
+ "epoch": 1.175935900888769,
+ "grad_norm": 0.7211281131146224,
+ "learning_rate": 3.6362334042402326e-06,
+ "loss": 0.3099272012710571,
+ "num_input_tokens_seen": 763825752,
+ "step": 17465,
+ "train_runtime": 135156.7264,
+ "train_tokens_per_second": 5651.408
+ },
+ {
+ "epoch": 1.1762725558847293,
+ "grad_norm": 0.7082063726811787,
+ "learning_rate": 3.6336897659037563e-06,
+ "loss": 0.2802994966506958,
+ "num_input_tokens_seen": 764035624,
+ "step": 17470,
+ "train_runtime": 135188.9582,
+ "train_tokens_per_second": 5651.613
+ },
+ {
+ "epoch": 1.1766092108806894,
+ "grad_norm": 0.721477429769724,
+ "learning_rate": 3.6311465096515787e-06,
+ "loss": 0.29577035903930665,
+ "num_input_tokens_seen": 764261488,
+ "step": 17475,
+ "train_runtime": 135224.5379,
+ "train_tokens_per_second": 5651.796
+ },
+ {
+ "epoch": 1.1769458658766496,
+ "grad_norm": 0.7174604232091135,
+ "learning_rate": 3.6286036361949175e-06,
+ "loss": 0.2771427154541016,
+ "num_input_tokens_seen": 764486984,
+ "step": 17480,
+ "train_runtime": 135262.7158,
+ "train_tokens_per_second": 5651.868
+ },
+ {
+ "epoch": 1.1772825208726099,
+ "grad_norm": 0.6898907133835245,
+ "learning_rate": 3.6260611462448736e-06,
+ "loss": 0.2587808847427368,
+ "num_input_tokens_seen": 764711328,
+ "step": 17485,
+ "train_runtime": 135302.057,
+ "train_tokens_per_second": 5651.882
+ },
+ {
+ "epoch": 1.17761917586857,
+ "grad_norm": 0.9411699136352466,
+ "learning_rate": 3.623519040512451e-06,
+ "loss": 0.3395826816558838,
+ "num_input_tokens_seen": 764916768,
+ "step": 17490,
+ "train_runtime": 135338.873,
+ "train_tokens_per_second": 5651.863
+ },
+ {
+ "epoch": 1.17795583086453,
+ "grad_norm": 0.6299891222606505,
+ "learning_rate": 3.6209773197085373e-06,
+ "loss": 0.2713226556777954,
+ "num_input_tokens_seen": 765139336,
+ "step": 17495,
+ "train_runtime": 135380.6388,
+ "train_tokens_per_second": 5651.763
+ },
+ {
+ "epoch": 1.1782924858604902,
+ "grad_norm": 0.7488235191732048,
+ "learning_rate": 3.6184359845439214e-06,
+ "loss": 0.31320526599884035,
+ "num_input_tokens_seen": 765373584,
+ "step": 17500,
+ "train_runtime": 135413.0388,
+ "train_tokens_per_second": 5652.141
+ },
+ {
+ "epoch": 1.1786291408564504,
+ "grad_norm": 0.7715560770139029,
+ "learning_rate": 3.615895035729275e-06,
+ "loss": 0.32944774627685547,
+ "num_input_tokens_seen": 765601768,
+ "step": 17505,
+ "train_runtime": 135459.3007,
+ "train_tokens_per_second": 5651.895
+ },
+ {
+ "epoch": 1.1789657958524105,
+ "grad_norm": 0.7026976712835434,
+ "learning_rate": 3.6133544739751703e-06,
+ "loss": 0.2919255256652832,
+ "num_input_tokens_seen": 765813632,
+ "step": 17510,
+ "train_runtime": 135499.1227,
+ "train_tokens_per_second": 5651.798
+ },
+ {
+ "epoch": 1.1793024508483705,
+ "grad_norm": 0.7763204505182639,
+ "learning_rate": 3.6108142999920633e-06,
+ "loss": 0.3174884796142578,
+ "num_input_tokens_seen": 766030160,
+ "step": 17515,
+ "train_runtime": 135540.5033,
+ "train_tokens_per_second": 5651.67
+ },
+ {
+ "epoch": 1.1796391058443307,
+ "grad_norm": 0.7315888831681889,
+ "learning_rate": 3.608274514490308e-06,
+ "loss": 0.30716781616210936,
+ "num_input_tokens_seen": 766254576,
+ "step": 17520,
+ "train_runtime": 135580.8511,
+ "train_tokens_per_second": 5651.643
+ },
+ {
+ "epoch": 1.179975760840291,
+ "grad_norm": 0.8351647015979019,
+ "learning_rate": 3.605735118180146e-06,
+ "loss": 0.2879929542541504,
+ "num_input_tokens_seen": 766474424,
+ "step": 17525,
+ "train_runtime": 135626.1303,
+ "train_tokens_per_second": 5651.377
+ },
+ {
+ "epoch": 1.180312415836251,
+ "grad_norm": 0.7061934846953144,
+ "learning_rate": 3.6031961117717144e-06,
+ "loss": 0.2810915470123291,
+ "num_input_tokens_seen": 766686368,
+ "step": 17530,
+ "train_runtime": 135665.2165,
+ "train_tokens_per_second": 5651.311
+ },
+ {
+ "epoch": 1.180649070832211,
+ "grad_norm": 0.6723872835220442,
+ "learning_rate": 3.600657495975034e-06,
+ "loss": 0.29715049266815186,
+ "num_input_tokens_seen": 766912752,
+ "step": 17535,
+ "train_runtime": 135706.1892,
+ "train_tokens_per_second": 5651.273
+ },
+ {
+ "epoch": 1.1809857258281713,
+ "grad_norm": 0.7032026671628777,
+ "learning_rate": 3.598119271500024e-06,
+ "loss": 0.30086421966552734,
+ "num_input_tokens_seen": 767131128,
+ "step": 17540,
+ "train_runtime": 135738.6404,
+ "train_tokens_per_second": 5651.531
+ },
+ {
+ "epoch": 1.1813223808241315,
+ "grad_norm": 0.7440024692945637,
+ "learning_rate": 3.5955814390564882e-06,
+ "loss": 0.30273914337158203,
+ "num_input_tokens_seen": 767352664,
+ "step": 17545,
+ "train_runtime": 135778.2276,
+ "train_tokens_per_second": 5651.515
+ },
+ {
+ "epoch": 1.1816590358200916,
+ "grad_norm": 0.7568009197320091,
+ "learning_rate": 3.5930439993541265e-06,
+ "loss": 0.3133152961730957,
+ "num_input_tokens_seen": 767564224,
+ "step": 17550,
+ "train_runtime": 135812.8257,
+ "train_tokens_per_second": 5651.633
+ },
+ {
+ "epoch": 1.1819956908160516,
+ "grad_norm": 0.8131849513084429,
+ "learning_rate": 3.590506953102522e-06,
+ "loss": 0.3256534576416016,
+ "num_input_tokens_seen": 767781400,
+ "step": 17555,
+ "train_runtime": 135853.5865,
+ "train_tokens_per_second": 5651.536
+ },
+ {
+ "epoch": 1.1823323458120119,
+ "grad_norm": 0.7021756461446376,
+ "learning_rate": 3.5879703010111554e-06,
+ "loss": 0.3131730318069458,
+ "num_input_tokens_seen": 767996072,
+ "step": 17560,
+ "train_runtime": 135897.3639,
+ "train_tokens_per_second": 5651.295
+ },
+ {
+ "epoch": 1.182669000807972,
+ "grad_norm": 0.6947691337677638,
+ "learning_rate": 3.5854340437893898e-06,
+ "loss": 0.29679055213928224,
+ "num_input_tokens_seen": 768218592,
+ "step": 17565,
+ "train_runtime": 135933.4998,
+ "train_tokens_per_second": 5651.43
+ },
+ {
+ "epoch": 1.1830056558039321,
+ "grad_norm": 0.7588414880559752,
+ "learning_rate": 3.5828981821464838e-06,
+ "loss": 0.31975440979003905,
+ "num_input_tokens_seen": 768427128,
+ "step": 17570,
+ "train_runtime": 135972.4916,
+ "train_tokens_per_second": 5651.343
+ },
+ {
+ "epoch": 1.1833423107998922,
+ "grad_norm": 0.708470213652483,
+ "learning_rate": 3.580362716791582e-06,
+ "loss": 0.31680521965026853,
+ "num_input_tokens_seen": 768660704,
+ "step": 17575,
+ "train_runtime": 136004.1836,
+ "train_tokens_per_second": 5651.743
+ },
+ {
+ "epoch": 1.1836789657958524,
+ "grad_norm": 0.8012972832226694,
+ "learning_rate": 3.5778276484337205e-06,
+ "loss": 0.3270276069641113,
+ "num_input_tokens_seen": 768871496,
+ "step": 17580,
+ "train_runtime": 136039.0476,
+ "train_tokens_per_second": 5651.844
+ },
+ {
+ "epoch": 1.1840156207918127,
+ "grad_norm": 0.6791587926903686,
+ "learning_rate": 3.575292977781821e-06,
+ "loss": 0.2917544603347778,
+ "num_input_tokens_seen": 769081176,
+ "step": 17585,
+ "train_runtime": 136078.1123,
+ "train_tokens_per_second": 5651.763
+ },
+ {
+ "epoch": 1.1843522757877727,
+ "grad_norm": 0.774600266373974,
+ "learning_rate": 3.5727587055446976e-06,
+ "loss": 0.30277276039123535,
+ "num_input_tokens_seen": 769310368,
+ "step": 17590,
+ "train_runtime": 136115.3881,
+ "train_tokens_per_second": 5651.899
+ },
+ {
+ "epoch": 1.1846889307837327,
+ "grad_norm": 0.707413300444675,
+ "learning_rate": 3.570224832431051e-06,
+ "loss": 0.3234860420227051,
+ "num_input_tokens_seen": 769534008,
+ "step": 17595,
+ "train_runtime": 136150.4904,
+ "train_tokens_per_second": 5652.084
+ },
+ {
+ "epoch": 1.185025585779693,
+ "grad_norm": 0.8625892185888415,
+ "learning_rate": 3.5676913591494715e-06,
+ "loss": 0.3072806358337402,
+ "num_input_tokens_seen": 769757384,
+ "step": 17600,
+ "train_runtime": 136188.6995,
+ "train_tokens_per_second": 5652.138
+ },
+ {
+ "epoch": 1.1853622407756532,
+ "grad_norm": 0.6997799283337246,
+ "learning_rate": 3.5651582864084345e-06,
+ "loss": 0.32632970809936523,
+ "num_input_tokens_seen": 769975064,
+ "step": 17605,
+ "train_runtime": 136230.7878,
+ "train_tokens_per_second": 5651.99
+ },
+ {
+ "epoch": 1.1856988957716132,
+ "grad_norm": 0.7412339983103909,
+ "learning_rate": 3.5626256149163096e-06,
+ "loss": 0.3052994966506958,
+ "num_input_tokens_seen": 770193168,
+ "step": 17610,
+ "train_runtime": 136266.0259,
+ "train_tokens_per_second": 5652.129
+ },
+ {
+ "epoch": 1.1860355507675733,
+ "grad_norm": 0.7442659857634908,
+ "learning_rate": 3.5600933453813464e-06,
+ "loss": 0.3075841188430786,
+ "num_input_tokens_seen": 770408384,
+ "step": 17615,
+ "train_runtime": 136300.5956,
+ "train_tokens_per_second": 5652.275
+ },
+ {
+ "epoch": 1.1863722057635335,
+ "grad_norm": 0.7199031556232233,
+ "learning_rate": 3.5575614785116886e-06,
+ "loss": 0.28684291839599607,
+ "num_input_tokens_seen": 770620728,
+ "step": 17620,
+ "train_runtime": 136345.9364,
+ "train_tokens_per_second": 5651.952
+ },
+ {
+ "epoch": 1.1867088607594938,
+ "grad_norm": 0.8819942138485812,
+ "learning_rate": 3.5550300150153618e-06,
+ "loss": 0.3115447998046875,
+ "num_input_tokens_seen": 770842752,
+ "step": 17625,
+ "train_runtime": 136383.5236,
+ "train_tokens_per_second": 5652.023
+ },
+ {
+ "epoch": 1.1870455157554538,
+ "grad_norm": 0.7246608362400021,
+ "learning_rate": 3.5524989556002863e-06,
+ "loss": 0.29783077239990235,
+ "num_input_tokens_seen": 771060848,
+ "step": 17630,
+ "train_runtime": 136428.5453,
+ "train_tokens_per_second": 5651.756
+ },
+ {
+ "epoch": 1.1873821707514138,
+ "grad_norm": 0.7193175427066145,
+ "learning_rate": 3.5499683009742604e-06,
+ "loss": 0.30118565559387206,
+ "num_input_tokens_seen": 771263296,
+ "step": 17635,
+ "train_runtime": 136462.119,
+ "train_tokens_per_second": 5651.849
+ },
+ {
+ "epoch": 1.187718825747374,
+ "grad_norm": 0.7869452467124062,
+ "learning_rate": 3.5474380518449757e-06,
+ "loss": 0.2976966381072998,
+ "num_input_tokens_seen": 771484736,
+ "step": 17640,
+ "train_runtime": 136496.9783,
+ "train_tokens_per_second": 5652.028
+ },
+ {
+ "epoch": 1.1880554807433343,
+ "grad_norm": 0.7932366417755232,
+ "learning_rate": 3.5449082089200072e-06,
+ "loss": 0.301554536819458,
+ "num_input_tokens_seen": 771704632,
+ "step": 17645,
+ "train_runtime": 136536.9073,
+ "train_tokens_per_second": 5651.986
+ },
+ {
+ "epoch": 1.1883921357392944,
+ "grad_norm": 0.6938403571190614,
+ "learning_rate": 3.5423787729068206e-06,
+ "loss": 0.32130672931671145,
+ "num_input_tokens_seen": 771927880,
+ "step": 17650,
+ "train_runtime": 136573.1716,
+ "train_tokens_per_second": 5652.119
+ },
+ {
+ "epoch": 1.1887287907352544,
+ "grad_norm": 0.82729296746894,
+ "learning_rate": 3.539849744512759e-06,
+ "loss": 0.29790964126586916,
+ "num_input_tokens_seen": 772137016,
+ "step": 17655,
+ "train_runtime": 136621.1016,
+ "train_tokens_per_second": 5651.667
+ },
+ {
+ "epoch": 1.1890654457312146,
+ "grad_norm": 0.7549562336474883,
+ "learning_rate": 3.5373211244450633e-06,
+ "loss": 0.29845280647277833,
+ "num_input_tokens_seen": 772361904,
+ "step": 17660,
+ "train_runtime": 136662.555,
+ "train_tokens_per_second": 5651.599
+ },
+ {
+ "epoch": 1.189402100727175,
+ "grad_norm": 0.7357102335902723,
+ "learning_rate": 3.5347929134108483e-06,
+ "loss": 0.31483676433563235,
+ "num_input_tokens_seen": 772575080,
+ "step": 17665,
+ "train_runtime": 136699.3985,
+ "train_tokens_per_second": 5651.635
+ },
+ {
+ "epoch": 1.189738755723135,
+ "grad_norm": 0.8439288436808993,
+ "learning_rate": 3.532265112117125e-06,
+ "loss": 0.33406968116760255,
+ "num_input_tokens_seen": 772803264,
+ "step": 17670,
+ "train_runtime": 136740.7168,
+ "train_tokens_per_second": 5651.596
+ },
+ {
+ "epoch": 1.190075410719095,
+ "grad_norm": 0.7786740028253488,
+ "learning_rate": 3.5297377212707805e-06,
+ "loss": 0.2623422861099243,
+ "num_input_tokens_seen": 773015632,
+ "step": 17675,
+ "train_runtime": 136775.9051,
+ "train_tokens_per_second": 5651.695
+ },
+ {
+ "epoch": 1.1904120657150552,
+ "grad_norm": 0.7870032609519265,
+ "learning_rate": 3.527210741578596e-06,
+ "loss": 0.3026540279388428,
+ "num_input_tokens_seen": 773239976,
+ "step": 17680,
+ "train_runtime": 136808.3986,
+ "train_tokens_per_second": 5651.992
+ },
+ {
+ "epoch": 1.1907487207110155,
+ "grad_norm": 0.7671210970920903,
+ "learning_rate": 3.5246841737472304e-06,
+ "loss": 0.3219871520996094,
+ "num_input_tokens_seen": 773448912,
+ "step": 17685,
+ "train_runtime": 136855.0232,
+ "train_tokens_per_second": 5651.593
+ },
+ {
+ "epoch": 1.1910853757069755,
+ "grad_norm": 0.7580157770145218,
+ "learning_rate": 3.522158018483231e-06,
+ "loss": 0.2920974254608154,
+ "num_input_tokens_seen": 773675944,
+ "step": 17690,
+ "train_runtime": 136891.1287,
+ "train_tokens_per_second": 5651.761
+ },
+ {
+ "epoch": 1.1914220307029355,
+ "grad_norm": 0.7290657363390212,
+ "learning_rate": 3.519632276493029e-06,
+ "loss": 0.31039018630981446,
+ "num_input_tokens_seen": 773894272,
+ "step": 17695,
+ "train_runtime": 136937.1303,
+ "train_tokens_per_second": 5651.457
+ },
+ {
+ "epoch": 1.1917586856988958,
+ "grad_norm": 0.7233052465646747,
+ "learning_rate": 3.5171069484829417e-06,
+ "loss": 0.32564287185668944,
+ "num_input_tokens_seen": 774110520,
+ "step": 17700,
+ "train_runtime": 136981.8929,
+ "train_tokens_per_second": 5651.189
+ },
+ {
+ "epoch": 1.192095340694856,
+ "grad_norm": 0.8073757654663947,
+ "learning_rate": 3.5145820351591663e-06,
+ "loss": 0.28891453742980955,
+ "num_input_tokens_seen": 774319632,
+ "step": 17705,
+ "train_runtime": 137022.015,
+ "train_tokens_per_second": 5651.06
+ },
+ {
+ "epoch": 1.192431995690816,
+ "grad_norm": 0.7652622798191697,
+ "learning_rate": 3.51205753722779e-06,
+ "loss": 0.3185939073562622,
+ "num_input_tokens_seen": 774550784,
+ "step": 17710,
+ "train_runtime": 137056.4812,
+ "train_tokens_per_second": 5651.325
+ },
+ {
+ "epoch": 1.1927686506867763,
+ "grad_norm": 0.8416381367688279,
+ "learning_rate": 3.5095334553947757e-06,
+ "loss": 0.33892183303833007,
+ "num_input_tokens_seen": 774781848,
+ "step": 17715,
+ "train_runtime": 137094.7795,
+ "train_tokens_per_second": 5651.432
+ },
+ {
+ "epoch": 1.1931053056827363,
+ "grad_norm": 0.7454120054728669,
+ "learning_rate": 3.50700979036598e-06,
+ "loss": 0.2988421440124512,
+ "num_input_tokens_seen": 774997592,
+ "step": 17720,
+ "train_runtime": 137134.1311,
+ "train_tokens_per_second": 5651.384
+ },
+ {
+ "epoch": 1.1934419606786966,
+ "grad_norm": 0.7748610171469188,
+ "learning_rate": 3.5044865428471343e-06,
+ "loss": 0.28873891830444337,
+ "num_input_tokens_seen": 775203648,
+ "step": 17725,
+ "train_runtime": 137170.5574,
+ "train_tokens_per_second": 5651.385
+ },
+ {
+ "epoch": 1.1937786156746566,
+ "grad_norm": 0.7302857347443209,
+ "learning_rate": 3.50196371354386e-06,
+ "loss": 0.30784637928009034,
+ "num_input_tokens_seen": 775418528,
+ "step": 17730,
+ "train_runtime": 137207.7458,
+ "train_tokens_per_second": 5651.419
+ },
+ {
+ "epoch": 1.1941152706706168,
+ "grad_norm": 0.8321187550816316,
+ "learning_rate": 3.499441303161655e-06,
+ "loss": 0.3373323202133179,
+ "num_input_tokens_seen": 775644000,
+ "step": 17735,
+ "train_runtime": 137243.8108,
+ "train_tokens_per_second": 5651.577
+ },
+ {
+ "epoch": 1.1944519256665769,
+ "grad_norm": 0.6939534771748743,
+ "learning_rate": 3.4969193124059042e-06,
+ "loss": 0.29674830436706545,
+ "num_input_tokens_seen": 775872464,
+ "step": 17740,
+ "train_runtime": 137280.866,
+ "train_tokens_per_second": 5651.716
+ },
+ {
+ "epoch": 1.1947885806625371,
+ "grad_norm": 0.6781828643380845,
+ "learning_rate": 3.4943977419818748e-06,
+ "loss": 0.2963244438171387,
+ "num_input_tokens_seen": 776107304,
+ "step": 17745,
+ "train_runtime": 137317.7425,
+ "train_tokens_per_second": 5651.908
+ },
+ {
+ "epoch": 1.1951252356584972,
+ "grad_norm": 0.7205121541036261,
+ "learning_rate": 3.4918765925947173e-06,
+ "loss": 0.2947703838348389,
+ "num_input_tokens_seen": 776333360,
+ "step": 17750,
+ "train_runtime": 137361.1596,
+ "train_tokens_per_second": 5651.768
+ },
+ {
+ "epoch": 1.1954618906544574,
+ "grad_norm": 0.7958380388514483,
+ "learning_rate": 3.4893558649494595e-06,
+ "loss": 0.32445640563964845,
+ "num_input_tokens_seen": 776565352,
+ "step": 17755,
+ "train_runtime": 137407.4556,
+ "train_tokens_per_second": 5651.552
+ },
+ {
+ "epoch": 1.1957985456504174,
+ "grad_norm": 0.7848842866143078,
+ "learning_rate": 3.4868355597510194e-06,
+ "loss": 0.29216394424438474,
+ "num_input_tokens_seen": 776782064,
+ "step": 17760,
+ "train_runtime": 137451.8285,
+ "train_tokens_per_second": 5651.304
+ },
+ {
+ "epoch": 1.1961352006463777,
+ "grad_norm": 0.7869373477637197,
+ "learning_rate": 3.4843156777041887e-06,
+ "loss": 0.27714879512786866,
+ "num_input_tokens_seen": 777006896,
+ "step": 17765,
+ "train_runtime": 137487.3294,
+ "train_tokens_per_second": 5651.48
+ },
+ {
+ "epoch": 1.1964718556423377,
+ "grad_norm": 0.7742008308908815,
+ "learning_rate": 3.4817962195136457e-06,
+ "loss": 0.3087809085845947,
+ "num_input_tokens_seen": 777222560,
+ "step": 17770,
+ "train_runtime": 137530.2067,
+ "train_tokens_per_second": 5651.286
+ },
+ {
+ "epoch": 1.196808510638298,
+ "grad_norm": 0.8109631152985817,
+ "learning_rate": 3.479277185883947e-06,
+ "loss": 0.3070957183837891,
+ "num_input_tokens_seen": 777436752,
+ "step": 17775,
+ "train_runtime": 137571.1937,
+ "train_tokens_per_second": 5651.159
+ },
+ {
+ "epoch": 1.197145165634258,
+ "grad_norm": 0.956312966639188,
+ "learning_rate": 3.476758577519537e-06,
+ "loss": 0.29088459014892576,
+ "num_input_tokens_seen": 777654880,
+ "step": 17780,
+ "train_runtime": 137605.082,
+ "train_tokens_per_second": 5651.353
+ },
+ {
+ "epoch": 1.1974818206302182,
+ "grad_norm": 0.7882126606326822,
+ "learning_rate": 3.4742403951247318e-06,
+ "loss": 0.3081394672393799,
+ "num_input_tokens_seen": 777876192,
+ "step": 17785,
+ "train_runtime": 137646.6579,
+ "train_tokens_per_second": 5651.254
+ },
+ {
+ "epoch": 1.1978184756261783,
+ "grad_norm": 0.7275982569821285,
+ "learning_rate": 3.471722639403735e-06,
+ "loss": 0.29077887535095215,
+ "num_input_tokens_seen": 778090376,
+ "step": 17790,
+ "train_runtime": 137688.0934,
+ "train_tokens_per_second": 5651.109
+ },
+ {
+ "epoch": 1.1981551306221385,
+ "grad_norm": 0.7256020434598892,
+ "learning_rate": 3.4692053110606277e-06,
+ "loss": 0.2993917942047119,
+ "num_input_tokens_seen": 778289376,
+ "step": 17795,
+ "train_runtime": 137728.124,
+ "train_tokens_per_second": 5650.911
+ },
+ {
+ "epoch": 1.1984917856180985,
+ "grad_norm": 0.6738742049588294,
+ "learning_rate": 3.4666884107993766e-06,
+ "loss": 0.2942534923553467,
+ "num_input_tokens_seen": 778502056,
+ "step": 17800,
+ "train_runtime": 137764.3256,
+ "train_tokens_per_second": 5650.97
+ },
+ {
+ "epoch": 1.1988284406140588,
+ "grad_norm": 0.6924946642523322,
+ "learning_rate": 3.464171939323819e-06,
+ "loss": 0.30054686069488523,
+ "num_input_tokens_seen": 778716688,
+ "step": 17805,
+ "train_runtime": 137804.6948,
+ "train_tokens_per_second": 5650.872
+ },
+ {
+ "epoch": 1.1991650956100188,
+ "grad_norm": 0.7464221349536175,
+ "learning_rate": 3.4616558973376848e-06,
+ "loss": 0.3028707981109619,
+ "num_input_tokens_seen": 778927512,
+ "step": 17810,
+ "train_runtime": 137840.357,
+ "train_tokens_per_second": 5650.939
+ },
+ {
+ "epoch": 1.199501750605979,
+ "grad_norm": 0.8392280191250697,
+ "learning_rate": 3.4591402855445697e-06,
+ "loss": 0.30633254051208497,
+ "num_input_tokens_seen": 779150112,
+ "step": 17815,
+ "train_runtime": 137873.0033,
+ "train_tokens_per_second": 5651.216
+ },
+ {
+ "epoch": 1.199838405601939,
+ "grad_norm": 0.7385160748310561,
+ "learning_rate": 3.4566251046479625e-06,
+ "loss": 0.3079712152481079,
+ "num_input_tokens_seen": 779369912,
+ "step": 17820,
+ "train_runtime": 137915.2772,
+ "train_tokens_per_second": 5651.077
+ },
+ {
+ "epoch": 1.2001750605978994,
+ "grad_norm": 0.7785026542998826,
+ "learning_rate": 3.454110355351221e-06,
+ "loss": 0.3142574310302734,
+ "num_input_tokens_seen": 779590624,
+ "step": 17825,
+ "train_runtime": 137954.41,
+ "train_tokens_per_second": 5651.074
+ },
+ {
+ "epoch": 1.2005117155938594,
+ "grad_norm": 0.7070886799527935,
+ "learning_rate": 3.451596038357593e-06,
+ "loss": 0.3034769773483276,
+ "num_input_tokens_seen": 779821712,
+ "step": 17830,
+ "train_runtime": 137993.4334,
+ "train_tokens_per_second": 5651.151
+ },
+ {
+ "epoch": 1.2008483705898196,
+ "grad_norm": 0.8286609047748951,
+ "learning_rate": 3.4490821543701924e-06,
+ "loss": 0.32932286262512206,
+ "num_input_tokens_seen": 780045176,
+ "step": 17835,
+ "train_runtime": 138029.8077,
+ "train_tokens_per_second": 5651.281
+ },
+ {
+ "epoch": 1.2011850255857797,
+ "grad_norm": 0.7276434685048679,
+ "learning_rate": 3.4465687040920227e-06,
+ "loss": 0.30574448108673097,
+ "num_input_tokens_seen": 780268064,
+ "step": 17840,
+ "train_runtime": 138072.0423,
+ "train_tokens_per_second": 5651.166
+ },
+ {
+ "epoch": 1.20152168058174,
+ "grad_norm": 0.6824095105587578,
+ "learning_rate": 3.44405568822596e-06,
+ "loss": 0.278488302230835,
+ "num_input_tokens_seen": 780486632,
+ "step": 17845,
+ "train_runtime": 138113.2062,
+ "train_tokens_per_second": 5651.064
+ },
+ {
+ "epoch": 1.2018583355777,
+ "grad_norm": 0.833968026923619,
+ "learning_rate": 3.441543107474765e-06,
+ "loss": 0.3007037162780762,
+ "num_input_tokens_seen": 780694864,
+ "step": 17850,
+ "train_runtime": 138148.5762,
+ "train_tokens_per_second": 5651.125
+ },
+ {
+ "epoch": 1.2021949905736602,
+ "grad_norm": 0.7390135533381705,
+ "learning_rate": 3.439030962541069e-06,
+ "loss": 0.293853759765625,
+ "num_input_tokens_seen": 780910776,
+ "step": 17855,
+ "train_runtime": 138187.3343,
+ "train_tokens_per_second": 5651.102
+ },
+ {
+ "epoch": 1.2025316455696202,
+ "grad_norm": 0.7449577348762936,
+ "learning_rate": 3.4365192541273885e-06,
+ "loss": 0.30453169345855713,
+ "num_input_tokens_seen": 781111576,
+ "step": 17860,
+ "train_runtime": 138223.1035,
+ "train_tokens_per_second": 5651.093
+ },
+ {
+ "epoch": 1.2028683005655805,
+ "grad_norm": 0.7209045215476029,
+ "learning_rate": 3.4340079829361108e-06,
+ "loss": 0.3177401304244995,
+ "num_input_tokens_seen": 781317608,
+ "step": 17865,
+ "train_runtime": 138256.2897,
+ "train_tokens_per_second": 5651.226
+ },
+ {
+ "epoch": 1.2032049555615405,
+ "grad_norm": 0.7033530226860483,
+ "learning_rate": 3.4314971496695078e-06,
+ "loss": 0.31711881160736083,
+ "num_input_tokens_seen": 781545368,
+ "step": 17870,
+ "train_runtime": 138290.4596,
+ "train_tokens_per_second": 5651.477
+ },
+ {
+ "epoch": 1.2035416105575008,
+ "grad_norm": 0.7085516748330253,
+ "learning_rate": 3.428986755029724e-06,
+ "loss": 0.2896330118179321,
+ "num_input_tokens_seen": 781774192,
+ "step": 17875,
+ "train_runtime": 138336.6049,
+ "train_tokens_per_second": 5651.246
+ },
+ {
+ "epoch": 1.2038782655534608,
+ "grad_norm": 0.7125099135090971,
+ "learning_rate": 3.4264767997187874e-06,
+ "loss": 0.2978813171386719,
+ "num_input_tokens_seen": 782004920,
+ "step": 17880,
+ "train_runtime": 138372.9121,
+ "train_tokens_per_second": 5651.431
+ },
+ {
+ "epoch": 1.204214920549421,
+ "grad_norm": 0.7572580628329799,
+ "learning_rate": 3.423967284438594e-06,
+ "loss": 0.3156701326370239,
+ "num_input_tokens_seen": 782217384,
+ "step": 17885,
+ "train_runtime": 138408.8011,
+ "train_tokens_per_second": 5651.5
+ },
+ {
+ "epoch": 1.204551575545381,
+ "grad_norm": 0.6615339386741577,
+ "learning_rate": 3.421458209890925e-06,
+ "loss": 0.2962948799133301,
+ "num_input_tokens_seen": 782450248,
+ "step": 17890,
+ "train_runtime": 138443.1781,
+ "train_tokens_per_second": 5651.779
+ },
+ {
+ "epoch": 1.2048882305413413,
+ "grad_norm": 0.700079916053248,
+ "learning_rate": 3.418949576777433e-06,
+ "loss": 0.28663947582244875,
+ "num_input_tokens_seen": 782681456,
+ "step": 17895,
+ "train_runtime": 138480.6285,
+ "train_tokens_per_second": 5651.92
+ },
+ {
+ "epoch": 1.2052248855373013,
+ "grad_norm": 0.775430671158068,
+ "learning_rate": 3.4164413857996513e-06,
+ "loss": 0.30020842552185056,
+ "num_input_tokens_seen": 782891192,
+ "step": 17900,
+ "train_runtime": 138522.9513,
+ "train_tokens_per_second": 5651.707
+ },
+ {
+ "epoch": 1.2055615405332616,
+ "grad_norm": 0.7147718137595105,
+ "learning_rate": 3.4139336376589853e-06,
+ "loss": 0.291434121131897,
+ "num_input_tokens_seen": 783119112,
+ "step": 17905,
+ "train_runtime": 138554.6985,
+ "train_tokens_per_second": 5652.057
+ },
+ {
+ "epoch": 1.2058981955292216,
+ "grad_norm": 0.8762127003247577,
+ "learning_rate": 3.411426333056722e-06,
+ "loss": 0.30569398403167725,
+ "num_input_tokens_seen": 783321392,
+ "step": 17910,
+ "train_runtime": 138597.0294,
+ "train_tokens_per_second": 5651.791
+ },
+ {
+ "epoch": 1.2062348505251819,
+ "grad_norm": 0.7685183853693496,
+ "learning_rate": 3.408919472694017e-06,
+ "loss": 0.321109676361084,
+ "num_input_tokens_seen": 783543848,
+ "step": 17915,
+ "train_runtime": 138637.4131,
+ "train_tokens_per_second": 5651.749
+ },
+ {
+ "epoch": 1.206571505521142,
+ "grad_norm": 0.7046016721897478,
+ "learning_rate": 3.406413057271909e-06,
+ "loss": 0.30394299030303956,
+ "num_input_tokens_seen": 783758592,
+ "step": 17920,
+ "train_runtime": 138676.5892,
+ "train_tokens_per_second": 5651.701
+ },
+ {
+ "epoch": 1.2069081605171021,
+ "grad_norm": 0.7508302781476457,
+ "learning_rate": 3.403907087491306e-06,
+ "loss": 0.303491997718811,
+ "num_input_tokens_seen": 783976568,
+ "step": 17925,
+ "train_runtime": 138712.5584,
+ "train_tokens_per_second": 5651.807
+ },
+ {
+ "epoch": 1.2072448155130622,
+ "grad_norm": 0.7413375117669119,
+ "learning_rate": 3.4014015640529996e-06,
+ "loss": 0.30534958839416504,
+ "num_input_tokens_seen": 784193960,
+ "step": 17930,
+ "train_runtime": 138748.3843,
+ "train_tokens_per_second": 5651.914
+ },
+ {
+ "epoch": 1.2075814705090224,
+ "grad_norm": 0.7576411855768873,
+ "learning_rate": 3.398896487657646e-06,
+ "loss": 0.31456284523010253,
+ "num_input_tokens_seen": 784414136,
+ "step": 17935,
+ "train_runtime": 138791.0719,
+ "train_tokens_per_second": 5651.762
+ },
+ {
+ "epoch": 1.2079181255049825,
+ "grad_norm": 0.7960278015173246,
+ "learning_rate": 3.396391859005785e-06,
+ "loss": 0.2981652975082397,
+ "num_input_tokens_seen": 784638336,
+ "step": 17940,
+ "train_runtime": 138825.2652,
+ "train_tokens_per_second": 5651.985
+ },
+ {
+ "epoch": 1.2082547805009427,
+ "grad_norm": 0.6970907266821021,
+ "learning_rate": 3.393887678797826e-06,
+ "loss": 0.3018413305282593,
+ "num_input_tokens_seen": 784844856,
+ "step": 17945,
+ "train_runtime": 138863.6497,
+ "train_tokens_per_second": 5651.91
+ },
+ {
+ "epoch": 1.2085914354969027,
+ "grad_norm": 0.7592630519307038,
+ "learning_rate": 3.3913839477340575e-06,
+ "loss": 0.30618746280670167,
+ "num_input_tokens_seen": 785058936,
+ "step": 17950,
+ "train_runtime": 138902.6182,
+ "train_tokens_per_second": 5651.866
+ },
+ {
+ "epoch": 1.208928090492863,
+ "grad_norm": 0.7766479409436013,
+ "learning_rate": 3.388880666514637e-06,
+ "loss": 0.28571882247924807,
+ "num_input_tokens_seen": 785286832,
+ "step": 17955,
+ "train_runtime": 138937.3108,
+ "train_tokens_per_second": 5652.095
+ },
+ {
+ "epoch": 1.209264745488823,
+ "grad_norm": 0.7267074307937196,
+ "learning_rate": 3.3863778358396033e-06,
+ "loss": 0.3155225276947021,
+ "num_input_tokens_seen": 785504144,
+ "step": 17960,
+ "train_runtime": 138972.8709,
+ "train_tokens_per_second": 5652.212
+ },
+ {
+ "epoch": 1.2096014004847833,
+ "grad_norm": 0.7097174073868221,
+ "learning_rate": 3.3838754564088592e-06,
+ "loss": 0.3005061149597168,
+ "num_input_tokens_seen": 785729392,
+ "step": 17965,
+ "train_runtime": 139011.4611,
+ "train_tokens_per_second": 5652.263
+ },
+ {
+ "epoch": 1.2099380554807433,
+ "grad_norm": 0.6597216885356778,
+ "learning_rate": 3.3813735289221923e-06,
+ "loss": 0.29288544654846194,
+ "num_input_tokens_seen": 785948848,
+ "step": 17970,
+ "train_runtime": 139049.8184,
+ "train_tokens_per_second": 5652.282
+ },
+ {
+ "epoch": 1.2102747104767035,
+ "grad_norm": 0.6269995153972749,
+ "learning_rate": 3.378872054079255e-06,
+ "loss": 0.29325079917907715,
+ "num_input_tokens_seen": 786168672,
+ "step": 17975,
+ "train_runtime": 139089.6764,
+ "train_tokens_per_second": 5652.243
+ },
+ {
+ "epoch": 1.2106113654726636,
+ "grad_norm": 0.7838620260591436,
+ "learning_rate": 3.3763710325795784e-06,
+ "loss": 0.3322885513305664,
+ "num_input_tokens_seen": 786371976,
+ "step": 17980,
+ "train_runtime": 139126.705,
+ "train_tokens_per_second": 5652.2
+ },
+ {
+ "epoch": 1.2109480204686238,
+ "grad_norm": 0.6771354234860218,
+ "learning_rate": 3.3738704651225673e-06,
+ "loss": 0.3040419340133667,
+ "num_input_tokens_seen": 786600776,
+ "step": 17985,
+ "train_runtime": 139162.6054,
+ "train_tokens_per_second": 5652.386
+ },
+ {
+ "epoch": 1.2112846754645838,
+ "grad_norm": 0.7878835736125291,
+ "learning_rate": 3.3713703524074936e-06,
+ "loss": 0.2779910326004028,
+ "num_input_tokens_seen": 786804904,
+ "step": 17990,
+ "train_runtime": 139199.5468,
+ "train_tokens_per_second": 5652.352
+ },
+ {
+ "epoch": 1.211621330460544,
+ "grad_norm": 0.8227585744222804,
+ "learning_rate": 3.3688706951335082e-06,
+ "loss": 0.31528596878051757,
+ "num_input_tokens_seen": 787008024,
+ "step": 17995,
+ "train_runtime": 139235.8527,
+ "train_tokens_per_second": 5652.337
+ },
+ {
+ "epoch": 1.2119579854565041,
+ "grad_norm": 0.7422262732589775,
+ "learning_rate": 3.3663714939996305e-06,
+ "loss": 0.30718555450439455,
+ "num_input_tokens_seen": 787245096,
+ "step": 18000,
+ "train_runtime": 139272.1742,
+ "train_tokens_per_second": 5652.566
+ },
+ {
+ "epoch": 1.2122946404524644,
+ "grad_norm": 0.7541583112520067,
+ "learning_rate": 3.3638727497047575e-06,
+ "loss": 0.2900678634643555,
+ "num_input_tokens_seen": 787449376,
+ "step": 18005,
+ "train_runtime": 139315.1398,
+ "train_tokens_per_second": 5652.289
+ },
+ {
+ "epoch": 1.2126312954484244,
+ "grad_norm": 0.7220759362847328,
+ "learning_rate": 3.3613744629476507e-06,
+ "loss": 0.29748926162719724,
+ "num_input_tokens_seen": 787661624,
+ "step": 18010,
+ "train_runtime": 139356.0904,
+ "train_tokens_per_second": 5652.151
+ },
+ {
+ "epoch": 1.2129679504443847,
+ "grad_norm": 0.7372103337616008,
+ "learning_rate": 3.358876634426952e-06,
+ "loss": 0.2816985845565796,
+ "num_input_tokens_seen": 787888360,
+ "step": 18015,
+ "train_runtime": 139393.2354,
+ "train_tokens_per_second": 5652.271
+ },
+ {
+ "epoch": 1.2133046054403447,
+ "grad_norm": 0.6737993858813106,
+ "learning_rate": 3.3563792648411676e-06,
+ "loss": 0.32491137981414797,
+ "num_input_tokens_seen": 788091136,
+ "step": 18020,
+ "train_runtime": 139429.8296,
+ "train_tokens_per_second": 5652.242
+ },
+ {
+ "epoch": 1.213641260436305,
+ "grad_norm": 0.7236596806287144,
+ "learning_rate": 3.353882354888681e-06,
+ "loss": 0.288041353225708,
+ "num_input_tokens_seen": 788312048,
+ "step": 18025,
+ "train_runtime": 139470.5579,
+ "train_tokens_per_second": 5652.175
+ },
+ {
+ "epoch": 1.213977915432265,
+ "grad_norm": 0.7377693831933381,
+ "learning_rate": 3.3513859052677444e-06,
+ "loss": 0.2752676486968994,
+ "num_input_tokens_seen": 788512224,
+ "step": 18030,
+ "train_runtime": 139508.9565,
+ "train_tokens_per_second": 5652.054
+ },
+ {
+ "epoch": 1.2143145704282252,
+ "grad_norm": 0.7799869819088461,
+ "learning_rate": 3.3488899166764842e-06,
+ "loss": 0.3178478479385376,
+ "num_input_tokens_seen": 788737240,
+ "step": 18035,
+ "train_runtime": 139542.6128,
+ "train_tokens_per_second": 5652.304
+ },
+ {
+ "epoch": 1.2146512254241852,
+ "grad_norm": 0.8463476604239449,
+ "learning_rate": 3.346394389812892e-06,
+ "loss": 0.3054813385009766,
+ "num_input_tokens_seen": 788959120,
+ "step": 18040,
+ "train_runtime": 139577.1377,
+ "train_tokens_per_second": 5652.495
+ },
+ {
+ "epoch": 1.2149878804201455,
+ "grad_norm": 0.7048797167467936,
+ "learning_rate": 3.3438993253748365e-06,
+ "loss": 0.2954627752304077,
+ "num_input_tokens_seen": 789168384,
+ "step": 18045,
+ "train_runtime": 139620.335,
+ "train_tokens_per_second": 5652.245
+ },
+ {
+ "epoch": 1.2153245354161055,
+ "grad_norm": 0.7007505922657411,
+ "learning_rate": 3.3414047240600526e-06,
+ "loss": 0.3028736591339111,
+ "num_input_tokens_seen": 789390496,
+ "step": 18050,
+ "train_runtime": 139660.3558,
+ "train_tokens_per_second": 5652.216
+ },
+ {
+ "epoch": 1.2156611904120658,
+ "grad_norm": 0.8100526851783998,
+ "learning_rate": 3.338910586566151e-06,
+ "loss": 0.3098297119140625,
+ "num_input_tokens_seen": 789600216,
+ "step": 18055,
+ "train_runtime": 139695.7464,
+ "train_tokens_per_second": 5652.285
+ },
+ {
+ "epoch": 1.2159978454080258,
+ "grad_norm": 0.7531507425187864,
+ "learning_rate": 3.336416913590605e-06,
+ "loss": 0.3021399974822998,
+ "num_input_tokens_seen": 789807928,
+ "step": 18060,
+ "train_runtime": 139731.0332,
+ "train_tokens_per_second": 5652.344
+ },
+ {
+ "epoch": 1.216334500403986,
+ "grad_norm": 0.8809448489320266,
+ "learning_rate": 3.333923705830766e-06,
+ "loss": 0.3216712474822998,
+ "num_input_tokens_seen": 790032400,
+ "step": 18065,
+ "train_runtime": 139770.9889,
+ "train_tokens_per_second": 5652.335
+ },
+ {
+ "epoch": 1.216671155399946,
+ "grad_norm": 0.7668922416234147,
+ "learning_rate": 3.3314309639838484e-06,
+ "loss": 0.3249736547470093,
+ "num_input_tokens_seen": 790253040,
+ "step": 18070,
+ "train_runtime": 139801.8412,
+ "train_tokens_per_second": 5652.665
+ },
+ {
+ "epoch": 1.2170078103959063,
+ "grad_norm": 0.7501530270251235,
+ "learning_rate": 3.328938688746942e-06,
+ "loss": 0.30673856735229493,
+ "num_input_tokens_seen": 790481600,
+ "step": 18075,
+ "train_runtime": 139848.0026,
+ "train_tokens_per_second": 5652.434
+ },
+ {
+ "epoch": 1.2173444653918664,
+ "grad_norm": 0.7405483975043581,
+ "learning_rate": 3.3264468808170012e-06,
+ "loss": 0.3439304828643799,
+ "num_input_tokens_seen": 790707832,
+ "step": 18080,
+ "train_runtime": 139884.7103,
+ "train_tokens_per_second": 5652.568
+ },
+ {
+ "epoch": 1.2176811203878266,
+ "grad_norm": 0.7461154895905937,
+ "learning_rate": 3.3239555408908562e-06,
+ "loss": 0.3145643711090088,
+ "num_input_tokens_seen": 790904072,
+ "step": 18085,
+ "train_runtime": 139927.9767,
+ "train_tokens_per_second": 5652.223
+ },
+ {
+ "epoch": 1.2180177753837866,
+ "grad_norm": 0.7450961063246211,
+ "learning_rate": 3.3214646696651974e-06,
+ "loss": 0.2987529277801514,
+ "num_input_tokens_seen": 791119712,
+ "step": 18090,
+ "train_runtime": 139967.4008,
+ "train_tokens_per_second": 5652.171
+ },
+ {
+ "epoch": 1.2183544303797469,
+ "grad_norm": 0.7553620001734184,
+ "learning_rate": 3.3189742678365923e-06,
+ "loss": 0.31772985458374026,
+ "num_input_tokens_seen": 791340456,
+ "step": 18095,
+ "train_runtime": 140002.9288,
+ "train_tokens_per_second": 5652.314
+ },
+ {
+ "epoch": 1.218691085375707,
+ "grad_norm": 0.6740962905213788,
+ "learning_rate": 3.316484336101472e-06,
+ "loss": 0.28296403884887694,
+ "num_input_tokens_seen": 791546784,
+ "step": 18100,
+ "train_runtime": 140038.0985,
+ "train_tokens_per_second": 5652.367
+ },
+ {
+ "epoch": 1.2190277403716672,
+ "grad_norm": 0.7648105865576331,
+ "learning_rate": 3.313994875156141e-06,
+ "loss": 0.31280903816223143,
+ "num_input_tokens_seen": 791764024,
+ "step": 18105,
+ "train_runtime": 140081.5209,
+ "train_tokens_per_second": 5652.166
+ },
+ {
+ "epoch": 1.2193643953676272,
+ "grad_norm": 0.7674425963272242,
+ "learning_rate": 3.3115058856967643e-06,
+ "loss": 0.31598572731018065,
+ "num_input_tokens_seen": 791993616,
+ "step": 18110,
+ "train_runtime": 140122.4895,
+ "train_tokens_per_second": 5652.152
+ },
+ {
+ "epoch": 1.2197010503635874,
+ "grad_norm": 0.7813158906447104,
+ "learning_rate": 3.3090173684193845e-06,
+ "loss": 0.27434990406036375,
+ "num_input_tokens_seen": 792207856,
+ "step": 18115,
+ "train_runtime": 140159.8456,
+ "train_tokens_per_second": 5652.174
+ },
+ {
+ "epoch": 1.2200377053595475,
+ "grad_norm": 0.7015793012796032,
+ "learning_rate": 3.3065293240199037e-06,
+ "loss": 0.31843407154083253,
+ "num_input_tokens_seen": 792431928,
+ "step": 18120,
+ "train_runtime": 140202.8473,
+ "train_tokens_per_second": 5652.039
+ },
+ {
+ "epoch": 1.2203743603555077,
+ "grad_norm": 0.8362321813673071,
+ "learning_rate": 3.3040417531940982e-06,
+ "loss": 0.3226888179779053,
+ "num_input_tokens_seen": 792655832,
+ "step": 18125,
+ "train_runtime": 140238.3502,
+ "train_tokens_per_second": 5652.204
+ },
+ {
+ "epoch": 1.2207110153514678,
+ "grad_norm": 0.691317135330766,
+ "learning_rate": 3.301554656637607e-06,
+ "loss": 0.2980408906936646,
+ "num_input_tokens_seen": 792868664,
+ "step": 18130,
+ "train_runtime": 140271.1305,
+ "train_tokens_per_second": 5652.401
+ },
+ {
+ "epoch": 1.221047670347428,
+ "grad_norm": 0.7033787564345391,
+ "learning_rate": 3.2990680350459427e-06,
+ "loss": 0.29863677024841306,
+ "num_input_tokens_seen": 793080312,
+ "step": 18135,
+ "train_runtime": 140311.0004,
+ "train_tokens_per_second": 5652.303
+ },
+ {
+ "epoch": 1.221384325343388,
+ "grad_norm": 0.7340650029847289,
+ "learning_rate": 3.2965818891144763e-06,
+ "loss": 0.3160408020019531,
+ "num_input_tokens_seen": 793303920,
+ "step": 18140,
+ "train_runtime": 140340.856,
+ "train_tokens_per_second": 5652.694
+ },
+ {
+ "epoch": 1.2217209803393483,
+ "grad_norm": 0.7408714024481081,
+ "learning_rate": 3.294096219538454e-06,
+ "loss": 0.322130012512207,
+ "num_input_tokens_seen": 793536920,
+ "step": 18145,
+ "train_runtime": 140387.2152,
+ "train_tokens_per_second": 5652.487
+ },
+ {
+ "epoch": 1.2220576353353083,
+ "grad_norm": 0.725033878974291,
+ "learning_rate": 3.291611027012983e-06,
+ "loss": 0.3116767406463623,
+ "num_input_tokens_seen": 793761088,
+ "step": 18150,
+ "train_runtime": 140429.0399,
+ "train_tokens_per_second": 5652.4
+ },
+ {
+ "epoch": 1.2223942903312686,
+ "grad_norm": 0.8302224573076524,
+ "learning_rate": 3.289126312233043e-06,
+ "loss": 0.29644060134887695,
+ "num_input_tokens_seen": 793993376,
+ "step": 18155,
+ "train_runtime": 140469.6701,
+ "train_tokens_per_second": 5652.419
+ },
+ {
+ "epoch": 1.2227309453272286,
+ "grad_norm": 0.6981694762570436,
+ "learning_rate": 3.2866420758934714e-06,
+ "loss": 0.3092994213104248,
+ "num_input_tokens_seen": 794216288,
+ "step": 18160,
+ "train_runtime": 140503.4652,
+ "train_tokens_per_second": 5652.646
+ },
+ {
+ "epoch": 1.2230676003231888,
+ "grad_norm": 0.6856222569708069,
+ "learning_rate": 3.2841583186889826e-06,
+ "loss": 0.30644276142120364,
+ "num_input_tokens_seen": 794442800,
+ "step": 18165,
+ "train_runtime": 140540.8201,
+ "train_tokens_per_second": 5652.755
+ },
+ {
+ "epoch": 1.2234042553191489,
+ "grad_norm": 0.76204653073967,
+ "learning_rate": 3.281675041314146e-06,
+ "loss": 0.2917419195175171,
+ "num_input_tokens_seen": 794656160,
+ "step": 18170,
+ "train_runtime": 140583.221,
+ "train_tokens_per_second": 5652.568
+ },
+ {
+ "epoch": 1.2237409103151091,
+ "grad_norm": 0.7110260250075573,
+ "learning_rate": 3.279192244463406e-06,
+ "loss": 0.29529433250427245,
+ "num_input_tokens_seen": 794869448,
+ "step": 18175,
+ "train_runtime": 140624.4232,
+ "train_tokens_per_second": 5652.428
+ },
+ {
+ "epoch": 1.2240775653110691,
+ "grad_norm": 0.7343704316676943,
+ "learning_rate": 3.2767099288310667e-06,
+ "loss": 0.29960105419158933,
+ "num_input_tokens_seen": 795093624,
+ "step": 18180,
+ "train_runtime": 140664.5318,
+ "train_tokens_per_second": 5652.41
+ },
+ {
+ "epoch": 1.2244142203070294,
+ "grad_norm": 0.639609983163744,
+ "learning_rate": 3.2742280951113025e-06,
+ "loss": 0.2962838888168335,
+ "num_input_tokens_seen": 795322792,
+ "step": 18185,
+ "train_runtime": 140702.7441,
+ "train_tokens_per_second": 5652.504
+ },
+ {
+ "epoch": 1.2247508753029894,
+ "grad_norm": 0.743438894841407,
+ "learning_rate": 3.2717467439981464e-06,
+ "loss": 0.30564584732055666,
+ "num_input_tokens_seen": 795532552,
+ "step": 18190,
+ "train_runtime": 140745.8939,
+ "train_tokens_per_second": 5652.261
+ },
+ {
+ "epoch": 1.2250875302989497,
+ "grad_norm": 0.7489547177708321,
+ "learning_rate": 3.2692658761855033e-06,
+ "loss": 0.27375977039337157,
+ "num_input_tokens_seen": 795735568,
+ "step": 18195,
+ "train_runtime": 140781.1568,
+ "train_tokens_per_second": 5652.287
+ },
+ {
+ "epoch": 1.2254241852949097,
+ "grad_norm": 0.6829987732324682,
+ "learning_rate": 3.266785492367138e-06,
+ "loss": 0.2808785915374756,
+ "num_input_tokens_seen": 795955960,
+ "step": 18200,
+ "train_runtime": 140815.8619,
+ "train_tokens_per_second": 5652.46
+ },
+ {
+ "epoch": 1.22576084029087,
+ "grad_norm": 0.7035746454375664,
+ "learning_rate": 3.2643055932366852e-06,
+ "loss": 0.28676323890686034,
+ "num_input_tokens_seen": 796165624,
+ "step": 18205,
+ "train_runtime": 140854.0679,
+ "train_tokens_per_second": 5652.415
+ },
+ {
+ "epoch": 1.22609749528683,
+ "grad_norm": 0.8214707476930566,
+ "learning_rate": 3.2618261794876365e-06,
+ "loss": 0.3010567665100098,
+ "num_input_tokens_seen": 796395392,
+ "step": 18210,
+ "train_runtime": 140889.0235,
+ "train_tokens_per_second": 5652.643
+ },
+ {
+ "epoch": 1.2264341502827902,
+ "grad_norm": 0.6985742460828861,
+ "learning_rate": 3.259347251813356e-06,
+ "loss": 0.2966732978820801,
+ "num_input_tokens_seen": 796617856,
+ "step": 18215,
+ "train_runtime": 140927.9467,
+ "train_tokens_per_second": 5652.661
+ },
+ {
+ "epoch": 1.2267708052787503,
+ "grad_norm": 0.727228038188911,
+ "learning_rate": 3.256868810907064e-06,
+ "loss": 0.3215049982070923,
+ "num_input_tokens_seen": 796852648,
+ "step": 18220,
+ "train_runtime": 140967.9622,
+ "train_tokens_per_second": 5652.722
+ },
+ {
+ "epoch": 1.2271074602747105,
+ "grad_norm": 0.7512872700129827,
+ "learning_rate": 3.2543908574618514e-06,
+ "loss": 0.3135005712509155,
+ "num_input_tokens_seen": 797070736,
+ "step": 18225,
+ "train_runtime": 141012.3331,
+ "train_tokens_per_second": 5652.49
+ },
+ {
+ "epoch": 1.2274441152706705,
+ "grad_norm": 0.6464160299656531,
+ "learning_rate": 3.251913392170668e-06,
+ "loss": 0.2944124460220337,
+ "num_input_tokens_seen": 797290688,
+ "step": 18230,
+ "train_runtime": 141057.431,
+ "train_tokens_per_second": 5652.242
+ },
+ {
+ "epoch": 1.2277807702666308,
+ "grad_norm": 0.7214923202638721,
+ "learning_rate": 3.249436415726333e-06,
+ "loss": 0.29227581024169924,
+ "num_input_tokens_seen": 797516256,
+ "step": 18235,
+ "train_runtime": 141095.1719,
+ "train_tokens_per_second": 5652.328
+ },
+ {
+ "epoch": 1.2281174252625908,
+ "grad_norm": 0.8506838334171156,
+ "learning_rate": 3.24695992882152e-06,
+ "loss": 0.3139004707336426,
+ "num_input_tokens_seen": 797725232,
+ "step": 18240,
+ "train_runtime": 141134.7796,
+ "train_tokens_per_second": 5652.223
+ },
+ {
+ "epoch": 1.228454080258551,
+ "grad_norm": 0.6887059927032259,
+ "learning_rate": 3.244483932148773e-06,
+ "loss": 0.3086080551147461,
+ "num_input_tokens_seen": 797943968,
+ "step": 18245,
+ "train_runtime": 141166.1,
+ "train_tokens_per_second": 5652.518
+ },
+ {
+ "epoch": 1.228790735254511,
+ "grad_norm": 0.7231712696156054,
+ "learning_rate": 3.2420084264004966e-06,
+ "loss": 0.30695390701293945,
+ "num_input_tokens_seen": 798161800,
+ "step": 18250,
+ "train_runtime": 141201.6515,
+ "train_tokens_per_second": 5652.638
+ },
+ {
+ "epoch": 1.2291273902504714,
+ "grad_norm": 0.7340695565349038,
+ "learning_rate": 3.2395334122689594e-06,
+ "loss": 0.30524253845214844,
+ "num_input_tokens_seen": 798379968,
+ "step": 18255,
+ "train_runtime": 141240.9452,
+ "train_tokens_per_second": 5652.61
+ },
+ {
+ "epoch": 1.2294640452464314,
+ "grad_norm": 0.8802102442010598,
+ "learning_rate": 3.2370588904462873e-06,
+ "loss": 0.3120718955993652,
+ "num_input_tokens_seen": 798609312,
+ "step": 18260,
+ "train_runtime": 141276.169,
+ "train_tokens_per_second": 5652.824
+ },
+ {
+ "epoch": 1.2298007002423916,
+ "grad_norm": 0.7640356901782597,
+ "learning_rate": 3.2345848616244775e-06,
+ "loss": 0.3038747549057007,
+ "num_input_tokens_seen": 798804840,
+ "step": 18265,
+ "train_runtime": 141315.8853,
+ "train_tokens_per_second": 5652.619
+ },
+ {
+ "epoch": 1.2301373552383517,
+ "grad_norm": 0.8040217158342741,
+ "learning_rate": 3.2321113264953797e-06,
+ "loss": 0.307041072845459,
+ "num_input_tokens_seen": 799027080,
+ "step": 18270,
+ "train_runtime": 141347.3832,
+ "train_tokens_per_second": 5652.932
+ },
+ {
+ "epoch": 1.230474010234312,
+ "grad_norm": 0.8528931267954475,
+ "learning_rate": 3.2296382857507124e-06,
+ "loss": 0.2948000431060791,
+ "num_input_tokens_seen": 799238584,
+ "step": 18275,
+ "train_runtime": 141386.4656,
+ "train_tokens_per_second": 5652.865
+ },
+ {
+ "epoch": 1.230810665230272,
+ "grad_norm": 0.7187404855989243,
+ "learning_rate": 3.227165740082053e-06,
+ "loss": 0.31221818923950195,
+ "num_input_tokens_seen": 799452544,
+ "step": 18280,
+ "train_runtime": 141428.6252,
+ "train_tokens_per_second": 5652.693
+ },
+ {
+ "epoch": 1.2311473202262322,
+ "grad_norm": 0.762382560173622,
+ "learning_rate": 3.2246936901808424e-06,
+ "loss": 0.2914170980453491,
+ "num_input_tokens_seen": 799675552,
+ "step": 18285,
+ "train_runtime": 141467.594,
+ "train_tokens_per_second": 5652.712
+ },
+ {
+ "epoch": 1.2314839752221922,
+ "grad_norm": 0.7606748879884372,
+ "learning_rate": 3.2222221367383786e-06,
+ "loss": 0.32509727478027345,
+ "num_input_tokens_seen": 799898272,
+ "step": 18290,
+ "train_runtime": 141515.3162,
+ "train_tokens_per_second": 5652.379
+ },
+ {
+ "epoch": 1.2318206302181525,
+ "grad_norm": 0.7608834756975011,
+ "learning_rate": 3.2197510804458265e-06,
+ "loss": 0.30308525562286376,
+ "num_input_tokens_seen": 800105152,
+ "step": 18295,
+ "train_runtime": 141554.2953,
+ "train_tokens_per_second": 5652.285
+ },
+ {
+ "epoch": 1.2321572852141125,
+ "grad_norm": 0.6370726326925514,
+ "learning_rate": 3.217280521994207e-06,
+ "loss": 0.2986292362213135,
+ "num_input_tokens_seen": 800331312,
+ "step": 18300,
+ "train_runtime": 141590.2519,
+ "train_tokens_per_second": 5652.446
+ },
+ {
+ "epoch": 1.2324939402100727,
+ "grad_norm": 0.7624529476398764,
+ "learning_rate": 3.214810462074407e-06,
+ "loss": 0.3016390562057495,
+ "num_input_tokens_seen": 800551688,
+ "step": 18305,
+ "train_runtime": 141623.377,
+ "train_tokens_per_second": 5652.68
+ },
+ {
+ "epoch": 1.2328305952060328,
+ "grad_norm": 0.7693052320987479,
+ "learning_rate": 3.2123409013771666e-06,
+ "loss": 0.30267770290374757,
+ "num_input_tokens_seen": 800777344,
+ "step": 18310,
+ "train_runtime": 141660.7953,
+ "train_tokens_per_second": 5652.78
+ },
+ {
+ "epoch": 1.233167250201993,
+ "grad_norm": 0.6772837658060004,
+ "learning_rate": 3.209871840593095e-06,
+ "loss": 0.31267609596252444,
+ "num_input_tokens_seen": 801009256,
+ "step": 18315,
+ "train_runtime": 141706.1002,
+ "train_tokens_per_second": 5652.61
+ },
+ {
+ "epoch": 1.233503905197953,
+ "grad_norm": 0.7125458851611337,
+ "learning_rate": 3.207403280412652e-06,
+ "loss": 0.30893242359161377,
+ "num_input_tokens_seen": 801228256,
+ "step": 18320,
+ "train_runtime": 141743.076,
+ "train_tokens_per_second": 5652.68
+ },
+ {
+ "epoch": 1.2338405601939133,
+ "grad_norm": 0.6945330021115061,
+ "learning_rate": 3.2049352215261677e-06,
+ "loss": 0.28535709381103513,
+ "num_input_tokens_seen": 801458800,
+ "step": 18325,
+ "train_runtime": 141780.6123,
+ "train_tokens_per_second": 5652.81
+ },
+ {
+ "epoch": 1.2341772151898733,
+ "grad_norm": 0.6960194289052382,
+ "learning_rate": 3.2024676646238222e-06,
+ "loss": 0.29358530044555664,
+ "num_input_tokens_seen": 801679448,
+ "step": 18330,
+ "train_runtime": 141816.9621,
+ "train_tokens_per_second": 5652.917
+ },
+ {
+ "epoch": 1.2345138701858336,
+ "grad_norm": 0.8171703953778717,
+ "learning_rate": 3.2000006103956653e-06,
+ "loss": 0.3098346948623657,
+ "num_input_tokens_seen": 801900976,
+ "step": 18335,
+ "train_runtime": 141857.1454,
+ "train_tokens_per_second": 5652.877
+ },
+ {
+ "epoch": 1.2348505251817936,
+ "grad_norm": 0.7299590696983723,
+ "learning_rate": 3.1975340595315956e-06,
+ "loss": 0.28231058120727537,
+ "num_input_tokens_seen": 802111128,
+ "step": 18340,
+ "train_runtime": 141893.2475,
+ "train_tokens_per_second": 5652.92
+ },
+ {
+ "epoch": 1.2351871801777539,
+ "grad_norm": 0.7644787058330598,
+ "learning_rate": 3.1950680127213796e-06,
+ "loss": 0.2976810455322266,
+ "num_input_tokens_seen": 802330912,
+ "step": 18345,
+ "train_runtime": 141932.7239,
+ "train_tokens_per_second": 5652.896
+ },
+ {
+ "epoch": 1.235523835173714,
+ "grad_norm": 0.7526412783611733,
+ "learning_rate": 3.192602470654638e-06,
+ "loss": 0.32010388374328613,
+ "num_input_tokens_seen": 802544448,
+ "step": 18350,
+ "train_runtime": 141974.2746,
+ "train_tokens_per_second": 5652.746
+ },
+ {
+ "epoch": 1.2358604901696741,
+ "grad_norm": 0.7865649130269652,
+ "learning_rate": 3.190137434020853e-06,
+ "loss": 0.2869425296783447,
+ "num_input_tokens_seen": 802766184,
+ "step": 18355,
+ "train_runtime": 142021.1129,
+ "train_tokens_per_second": 5652.443
+ },
+ {
+ "epoch": 1.2361971451656342,
+ "grad_norm": 0.8600627915587952,
+ "learning_rate": 3.1876729035093616e-06,
+ "loss": 0.29856894016265867,
+ "num_input_tokens_seen": 802986112,
+ "step": 18360,
+ "train_runtime": 142056.2119,
+ "train_tokens_per_second": 5652.594
+ },
+ {
+ "epoch": 1.2365338001615944,
+ "grad_norm": 0.7568409452496826,
+ "learning_rate": 3.185208879809366e-06,
+ "loss": 0.29802355766296384,
+ "num_input_tokens_seen": 803214056,
+ "step": 18365,
+ "train_runtime": 142100.6909,
+ "train_tokens_per_second": 5652.429
+ },
+ {
+ "epoch": 1.2368704551575544,
+ "grad_norm": 0.7529248541722069,
+ "learning_rate": 3.1827453636099183e-06,
+ "loss": 0.3028829574584961,
+ "num_input_tokens_seen": 803449464,
+ "step": 18370,
+ "train_runtime": 142136.647,
+ "train_tokens_per_second": 5652.655
+ },
+ {
+ "epoch": 1.2372071101535147,
+ "grad_norm": 0.9260284783600946,
+ "learning_rate": 3.1802823555999356e-06,
+ "loss": 0.2902355194091797,
+ "num_input_tokens_seen": 803675088,
+ "step": 18375,
+ "train_runtime": 142173.9871,
+ "train_tokens_per_second": 5652.758
+ },
+ {
+ "epoch": 1.2375437651494747,
+ "grad_norm": 0.7733188398886425,
+ "learning_rate": 3.1778198564681885e-06,
+ "loss": 0.32343668937683107,
+ "num_input_tokens_seen": 803903288,
+ "step": 18380,
+ "train_runtime": 142205.7209,
+ "train_tokens_per_second": 5653.101
+ },
+ {
+ "epoch": 1.237880420145435,
+ "grad_norm": 0.7192786766579897,
+ "learning_rate": 3.17535786690331e-06,
+ "loss": 0.30882773399353025,
+ "num_input_tokens_seen": 804121616,
+ "step": 18385,
+ "train_runtime": 142246.3225,
+ "train_tokens_per_second": 5653.022
+ },
+ {
+ "epoch": 1.238217075141395,
+ "grad_norm": 0.7729011461174,
+ "learning_rate": 3.172896387593784e-06,
+ "loss": 0.304299521446228,
+ "num_input_tokens_seen": 804348264,
+ "step": 18390,
+ "train_runtime": 142288.5576,
+ "train_tokens_per_second": 5652.937
+ },
+ {
+ "epoch": 1.2385537301373553,
+ "grad_norm": 0.8194549580251287,
+ "learning_rate": 3.1704354192279573e-06,
+ "loss": 0.31928482055664065,
+ "num_input_tokens_seen": 804564656,
+ "step": 18395,
+ "train_runtime": 142323.8728,
+ "train_tokens_per_second": 5653.055
+ },
+ {
+ "epoch": 1.2388903851333153,
+ "grad_norm": 0.7962633908276077,
+ "learning_rate": 3.1679749624940316e-06,
+ "loss": 0.318212103843689,
+ "num_input_tokens_seen": 804786400,
+ "step": 18400,
+ "train_runtime": 142360.3102,
+ "train_tokens_per_second": 5653.166
+ },
+ {
+ "epoch": 1.2392270401292755,
+ "grad_norm": 0.7855805493097535,
+ "learning_rate": 3.165515018080067e-06,
+ "loss": 0.3026642084121704,
+ "num_input_tokens_seen": 805005768,
+ "step": 18405,
+ "train_runtime": 142394.1464,
+ "train_tokens_per_second": 5653.363
+ },
+ {
+ "epoch": 1.2395636951252356,
+ "grad_norm": 0.938130030820677,
+ "learning_rate": 3.163055586673976e-06,
+ "loss": 0.29436650276184084,
+ "num_input_tokens_seen": 805226016,
+ "step": 18410,
+ "train_runtime": 142427.7333,
+ "train_tokens_per_second": 5653.576
+ },
+ {
+ "epoch": 1.2399003501211958,
+ "grad_norm": 0.7559775139780297,
+ "learning_rate": 3.1605966689635348e-06,
+ "loss": 0.32002964019775393,
+ "num_input_tokens_seen": 805453752,
+ "step": 18415,
+ "train_runtime": 142464.7236,
+ "train_tokens_per_second": 5653.707
+ },
+ {
+ "epoch": 1.2402370051171558,
+ "grad_norm": 0.8613984382442207,
+ "learning_rate": 3.1581382656363666e-06,
+ "loss": 0.30423684120178224,
+ "num_input_tokens_seen": 805658856,
+ "step": 18420,
+ "train_runtime": 142494.9465,
+ "train_tokens_per_second": 5653.947
+ },
+ {
+ "epoch": 1.240573660113116,
+ "grad_norm": 0.8697803980165589,
+ "learning_rate": 3.1556803773799616e-06,
+ "loss": 0.3185387372970581,
+ "num_input_tokens_seen": 805880136,
+ "step": 18425,
+ "train_runtime": 142536.9031,
+ "train_tokens_per_second": 5653.835
+ },
+ {
+ "epoch": 1.2409103151090761,
+ "grad_norm": 0.7422635875140698,
+ "learning_rate": 3.1532230048816564e-06,
+ "loss": 0.3022879123687744,
+ "num_input_tokens_seen": 806097056,
+ "step": 18430,
+ "train_runtime": 142575.9432,
+ "train_tokens_per_second": 5653.808
+ },
+ {
+ "epoch": 1.2412469701050364,
+ "grad_norm": 0.6776328969826404,
+ "learning_rate": 3.150766148828651e-06,
+ "loss": 0.3036804676055908,
+ "num_input_tokens_seen": 806315480,
+ "step": 18435,
+ "train_runtime": 142613.0736,
+ "train_tokens_per_second": 5653.868
+ },
+ {
+ "epoch": 1.2415836251009964,
+ "grad_norm": 0.7561483870082674,
+ "learning_rate": 3.1483098099079934e-06,
+ "loss": 0.2874735116958618,
+ "num_input_tokens_seen": 806538456,
+ "step": 18440,
+ "train_runtime": 142661.6297,
+ "train_tokens_per_second": 5653.507
+ },
+ {
+ "epoch": 1.2419202800969567,
+ "grad_norm": 0.7257415829423263,
+ "learning_rate": 3.1458539888065937e-06,
+ "loss": 0.3029060363769531,
+ "num_input_tokens_seen": 806750424,
+ "step": 18445,
+ "train_runtime": 142702.778,
+ "train_tokens_per_second": 5653.362
+ },
+ {
+ "epoch": 1.2422569350929167,
+ "grad_norm": 0.84249717889109,
+ "learning_rate": 3.143398686211212e-06,
+ "loss": 0.29645981788635256,
+ "num_input_tokens_seen": 806958552,
+ "step": 18450,
+ "train_runtime": 142740.2229,
+ "train_tokens_per_second": 5653.337
+ },
+ {
+ "epoch": 1.242593590088877,
+ "grad_norm": 0.7359383150490025,
+ "learning_rate": 3.1409439028084706e-06,
+ "loss": 0.3129459857940674,
+ "num_input_tokens_seen": 807178464,
+ "step": 18455,
+ "train_runtime": 142779.9615,
+ "train_tokens_per_second": 5653.304
+ },
+ {
+ "epoch": 1.2429302450848372,
+ "grad_norm": 0.7356032513359525,
+ "learning_rate": 3.1384896392848363e-06,
+ "loss": 0.28752105236053466,
+ "num_input_tokens_seen": 807397576,
+ "step": 18460,
+ "train_runtime": 142821.147,
+ "train_tokens_per_second": 5653.207
+ },
+ {
+ "epoch": 1.2432669000807972,
+ "grad_norm": 0.7330504631691293,
+ "learning_rate": 3.1360358963266404e-06,
+ "loss": 0.32807331085205077,
+ "num_input_tokens_seen": 807624936,
+ "step": 18465,
+ "train_runtime": 142852.5383,
+ "train_tokens_per_second": 5653.557
+ },
+ {
+ "epoch": 1.2436035550767572,
+ "grad_norm": 0.7686799242262236,
+ "learning_rate": 3.1335826746200594e-06,
+ "loss": 0.3099084377288818,
+ "num_input_tokens_seen": 807820280,
+ "step": 18470,
+ "train_runtime": 142890.1797,
+ "train_tokens_per_second": 5653.435
+ },
+ {
+ "epoch": 1.2439402100727175,
+ "grad_norm": 0.7644480228425453,
+ "learning_rate": 3.1311299748511335e-06,
+ "loss": 0.32409205436706545,
+ "num_input_tokens_seen": 808042832,
+ "step": 18475,
+ "train_runtime": 142931.8612,
+ "train_tokens_per_second": 5653.343
+ },
+ {
+ "epoch": 1.2442768650686777,
+ "grad_norm": 0.7485842376379487,
+ "learning_rate": 3.1286777977057504e-06,
+ "loss": 0.29155697822570803,
+ "num_input_tokens_seen": 808252592,
+ "step": 18480,
+ "train_runtime": 142970.1026,
+ "train_tokens_per_second": 5653.298
+ },
+ {
+ "epoch": 1.2446135200646378,
+ "grad_norm": 0.8186958870732818,
+ "learning_rate": 3.126226143869655e-06,
+ "loss": 0.3249779224395752,
+ "num_input_tokens_seen": 808465720,
+ "step": 18485,
+ "train_runtime": 143009.3714,
+ "train_tokens_per_second": 5653.236
+ },
+ {
+ "epoch": 1.2449501750605978,
+ "grad_norm": 0.6799895472703079,
+ "learning_rate": 3.1237750140284424e-06,
+ "loss": 0.27743706703186033,
+ "num_input_tokens_seen": 808678680,
+ "step": 18490,
+ "train_runtime": 143043.1574,
+ "train_tokens_per_second": 5653.389
+ },
+ {
+ "epoch": 1.245286830056558,
+ "grad_norm": 0.8374856400249484,
+ "learning_rate": 3.1213244088675647e-06,
+ "loss": 0.30376696586608887,
+ "num_input_tokens_seen": 808892656,
+ "step": 18495,
+ "train_runtime": 143081.3078,
+ "train_tokens_per_second": 5653.378
+ },
+ {
+ "epoch": 1.2456234850525183,
+ "grad_norm": 0.7662091689060238,
+ "learning_rate": 3.1188743290723246e-06,
+ "loss": 0.317877197265625,
+ "num_input_tokens_seen": 809111464,
+ "step": 18500,
+ "train_runtime": 143125.016,
+ "train_tokens_per_second": 5653.18
+ },
+ {
+ "epoch": 1.2459601400484783,
+ "grad_norm": 0.8647091169854892,
+ "learning_rate": 3.116424775327882e-06,
+ "loss": 0.3008237361907959,
+ "num_input_tokens_seen": 809328360,
+ "step": 18505,
+ "train_runtime": 143173.0534,
+ "train_tokens_per_second": 5652.798
+ },
+ {
+ "epoch": 1.2462967950444384,
+ "grad_norm": 0.6945551074504395,
+ "learning_rate": 3.1139757483192423e-06,
+ "loss": 0.3034308433532715,
+ "num_input_tokens_seen": 809557648,
+ "step": 18510,
+ "train_runtime": 143207.7193,
+ "train_tokens_per_second": 5653.031
+ },
+ {
+ "epoch": 1.2466334500403986,
+ "grad_norm": 0.8318013387227233,
+ "learning_rate": 3.1115272487312727e-06,
+ "loss": 0.29793879985809324,
+ "num_input_tokens_seen": 809766040,
+ "step": 18515,
+ "train_runtime": 143239.1385,
+ "train_tokens_per_second": 5653.246
+ },
+ {
+ "epoch": 1.2469701050363589,
+ "grad_norm": 0.8696915186885799,
+ "learning_rate": 3.109079277248684e-06,
+ "loss": 0.29328839778900145,
+ "num_input_tokens_seen": 809972912,
+ "step": 18520,
+ "train_runtime": 143278.4616,
+ "train_tokens_per_second": 5653.138
+ },
+ {
+ "epoch": 1.2473067600323189,
+ "grad_norm": 0.7294833612317175,
+ "learning_rate": 3.1066318345560463e-06,
+ "loss": 0.32238657474517823,
+ "num_input_tokens_seen": 810187728,
+ "step": 18525,
+ "train_runtime": 143312.263,
+ "train_tokens_per_second": 5653.304
+ },
+ {
+ "epoch": 1.247643415028279,
+ "grad_norm": 0.7391236990985328,
+ "learning_rate": 3.1041849213377782e-06,
+ "loss": 0.3177293300628662,
+ "num_input_tokens_seen": 810419880,
+ "step": 18530,
+ "train_runtime": 143347.6909,
+ "train_tokens_per_second": 5653.526
+ },
+ {
+ "epoch": 1.2479800700242392,
+ "grad_norm": 0.7271664362688767,
+ "learning_rate": 3.101738538278154e-06,
+ "loss": 0.29890563488006594,
+ "num_input_tokens_seen": 810642576,
+ "step": 18535,
+ "train_runtime": 143382.8102,
+ "train_tokens_per_second": 5653.694
+ },
+ {
+ "epoch": 1.2483167250201994,
+ "grad_norm": 0.7689863894703435,
+ "learning_rate": 3.0992926860612914e-06,
+ "loss": 0.30640869140625,
+ "num_input_tokens_seen": 810855296,
+ "step": 18540,
+ "train_runtime": 143422.92,
+ "train_tokens_per_second": 5653.596
+ },
+ {
+ "epoch": 1.2486533800161594,
+ "grad_norm": 0.7776443662476371,
+ "learning_rate": 3.0968473653711693e-06,
+ "loss": 0.2990746259689331,
+ "num_input_tokens_seen": 811094136,
+ "step": 18545,
+ "train_runtime": 143462.1432,
+ "train_tokens_per_second": 5653.715
+ },
+ {
+ "epoch": 1.2489900350121195,
+ "grad_norm": 0.7266832056756357,
+ "learning_rate": 3.094402576891613e-06,
+ "loss": 0.3187594413757324,
+ "num_input_tokens_seen": 811315008,
+ "step": 18550,
+ "train_runtime": 143497.5164,
+ "train_tokens_per_second": 5653.861
+ },
+ {
+ "epoch": 1.2493266900080797,
+ "grad_norm": 0.6721184594034306,
+ "learning_rate": 3.0919583213063004e-06,
+ "loss": 0.3195340156555176,
+ "num_input_tokens_seen": 811546128,
+ "step": 18555,
+ "train_runtime": 143531.9724,
+ "train_tokens_per_second": 5654.114
+ },
+ {
+ "epoch": 1.24966334500404,
+ "grad_norm": 0.75584993986847,
+ "learning_rate": 3.0895145992987574e-06,
+ "loss": 0.30277261734008787,
+ "num_input_tokens_seen": 811776888,
+ "step": 18560,
+ "train_runtime": 143569.4998,
+ "train_tokens_per_second": 5654.243
+ },
+ {
+ "epoch": 1.25,
+ "grad_norm": 0.8137422225809046,
+ "learning_rate": 3.087071411552366e-06,
+ "loss": 0.2947868824005127,
+ "num_input_tokens_seen": 811992848,
+ "step": 18565,
+ "train_runtime": 143605.6285,
+ "train_tokens_per_second": 5654.325
+ },
+ {
+ "epoch": 1.25033665499596,
+ "grad_norm": 0.7305460030824085,
+ "learning_rate": 3.084628758750353e-06,
+ "loss": 0.31966290473937986,
+ "num_input_tokens_seen": 812211480,
+ "step": 18570,
+ "train_runtime": 143644.3856,
+ "train_tokens_per_second": 5654.321
+ },
+ {
+ "epoch": 1.2506733099919203,
+ "grad_norm": 0.9296238525143283,
+ "learning_rate": 3.082186641575801e-06,
+ "loss": 0.3240236759185791,
+ "num_input_tokens_seen": 812427616,
+ "step": 18575,
+ "train_runtime": 143686.3373,
+ "train_tokens_per_second": 5654.174
+ },
+ {
+ "epoch": 1.2510099649878805,
+ "grad_norm": 0.7503570194002699,
+ "learning_rate": 3.0797450607116384e-06,
+ "loss": 0.30055992603302,
+ "num_input_tokens_seen": 812646144,
+ "step": 18580,
+ "train_runtime": 143727.8546,
+ "train_tokens_per_second": 5654.062
+ },
+ {
+ "epoch": 1.2513466199838406,
+ "grad_norm": 0.7124035622045924,
+ "learning_rate": 3.0773040168406475e-06,
+ "loss": 0.31271209716796877,
+ "num_input_tokens_seen": 812878472,
+ "step": 18585,
+ "train_runtime": 143761.8158,
+ "train_tokens_per_second": 5654.342
+ },
+ {
+ "epoch": 1.2516832749798006,
+ "grad_norm": 0.7943563134878205,
+ "learning_rate": 3.074863510645456e-06,
+ "loss": 0.31382598876953127,
+ "num_input_tokens_seen": 813099832,
+ "step": 18590,
+ "train_runtime": 143801.9795,
+ "train_tokens_per_second": 5654.302
+ },
+ {
+ "epoch": 1.2520199299757608,
+ "grad_norm": 0.6636865747124504,
+ "learning_rate": 3.072423542808546e-06,
+ "loss": 0.30039525032043457,
+ "num_input_tokens_seen": 813316560,
+ "step": 18595,
+ "train_runtime": 143835.7935,
+ "train_tokens_per_second": 5654.48
+ },
+ {
+ "epoch": 1.252356584971721,
+ "grad_norm": 0.7917515381442752,
+ "learning_rate": 3.0699841140122443e-06,
+ "loss": 0.3078362464904785,
+ "num_input_tokens_seen": 813538248,
+ "step": 18600,
+ "train_runtime": 143877.45,
+ "train_tokens_per_second": 5654.383
+ },
+ {
+ "epoch": 1.2526932399676811,
+ "grad_norm": 0.7542366254079307,
+ "learning_rate": 3.0675452249387327e-06,
+ "loss": 0.2891314744949341,
+ "num_input_tokens_seen": 813765928,
+ "step": 18605,
+ "train_runtime": 143916.9674,
+ "train_tokens_per_second": 5654.413
+ },
+ {
+ "epoch": 1.2530298949636411,
+ "grad_norm": 0.769680983787988,
+ "learning_rate": 3.0651068762700354e-06,
+ "loss": 0.3373257637023926,
+ "num_input_tokens_seen": 813974344,
+ "step": 18610,
+ "train_runtime": 143956.4928,
+ "train_tokens_per_second": 5654.308
+ },
+ {
+ "epoch": 1.2533665499596014,
+ "grad_norm": 0.7698186645906399,
+ "learning_rate": 3.0626690686880313e-06,
+ "loss": 0.32206950187683103,
+ "num_input_tokens_seen": 814211744,
+ "step": 18615,
+ "train_runtime": 143994.9815,
+ "train_tokens_per_second": 5654.445
+ },
+ {
+ "epoch": 1.2537032049555616,
+ "grad_norm": 0.7376028562088249,
+ "learning_rate": 3.060231802874443e-06,
+ "loss": 0.3022738218307495,
+ "num_input_tokens_seen": 814438872,
+ "step": 18620,
+ "train_runtime": 144031.2128,
+ "train_tokens_per_second": 5654.6
+ },
+ {
+ "epoch": 1.2540398599515217,
+ "grad_norm": 0.7155062346403165,
+ "learning_rate": 3.0577950795108462e-06,
+ "loss": 0.30109267234802245,
+ "num_input_tokens_seen": 814661808,
+ "step": 18625,
+ "train_runtime": 144071.0237,
+ "train_tokens_per_second": 5654.585
+ },
+ {
+ "epoch": 1.2543765149474817,
+ "grad_norm": 0.7986947159381109,
+ "learning_rate": 3.055358899278662e-06,
+ "loss": 0.2987940788269043,
+ "num_input_tokens_seen": 814871016,
+ "step": 18630,
+ "train_runtime": 144109.5084,
+ "train_tokens_per_second": 5654.526
+ },
+ {
+ "epoch": 1.254713169943442,
+ "grad_norm": 0.8209503548573042,
+ "learning_rate": 3.052923262859162e-06,
+ "loss": 0.3024166107177734,
+ "num_input_tokens_seen": 815098200,
+ "step": 18635,
+ "train_runtime": 144149.6967,
+ "train_tokens_per_second": 5654.526
+ },
+ {
+ "epoch": 1.2550498249394022,
+ "grad_norm": 0.7146717304289526,
+ "learning_rate": 3.0504881709334603e-06,
+ "loss": 0.29818005561828614,
+ "num_input_tokens_seen": 815319256,
+ "step": 18640,
+ "train_runtime": 144189.1823,
+ "train_tokens_per_second": 5654.511
+ },
+ {
+ "epoch": 1.2553864799353622,
+ "grad_norm": 0.7132265699815106,
+ "learning_rate": 3.0480536241825263e-06,
+ "loss": 0.28137338161468506,
+ "num_input_tokens_seen": 815537448,
+ "step": 18645,
+ "train_runtime": 144230.2973,
+ "train_tokens_per_second": 5654.411
+ },
+ {
+ "epoch": 1.2557231349313223,
+ "grad_norm": 0.7429234432398791,
+ "learning_rate": 3.0456196232871715e-06,
+ "loss": 0.31012353897094724,
+ "num_input_tokens_seen": 815749224,
+ "step": 18650,
+ "train_runtime": 144268.7463,
+ "train_tokens_per_second": 5654.372
+ },
+ {
+ "epoch": 1.2560597899272825,
+ "grad_norm": 0.8026609003022944,
+ "learning_rate": 3.043186168928059e-06,
+ "loss": 0.30425825119018557,
+ "num_input_tokens_seen": 815954440,
+ "step": 18655,
+ "train_runtime": 144316.2977,
+ "train_tokens_per_second": 5653.931
+ },
+ {
+ "epoch": 1.2563964449232428,
+ "grad_norm": 0.9158794619186253,
+ "learning_rate": 3.040753261785692e-06,
+ "loss": 0.2835689067840576,
+ "num_input_tokens_seen": 816164248,
+ "step": 18660,
+ "train_runtime": 144353.3412,
+ "train_tokens_per_second": 5653.934
+ },
+ {
+ "epoch": 1.2567330999192028,
+ "grad_norm": 0.7533631787720405,
+ "learning_rate": 3.038320902540429e-06,
+ "loss": 0.28765387535095216,
+ "num_input_tokens_seen": 816379936,
+ "step": 18665,
+ "train_runtime": 144394.4225,
+ "train_tokens_per_second": 5653.819
+ },
+ {
+ "epoch": 1.2570697549151628,
+ "grad_norm": 0.7215132949101682,
+ "learning_rate": 3.0358890918724706e-06,
+ "loss": 0.303347110748291,
+ "num_input_tokens_seen": 816611832,
+ "step": 18670,
+ "train_runtime": 144441.5414,
+ "train_tokens_per_second": 5653.58
+ },
+ {
+ "epoch": 1.257406409911123,
+ "grad_norm": 0.8254057548339484,
+ "learning_rate": 3.0334578304618646e-06,
+ "loss": 0.30634298324584963,
+ "num_input_tokens_seen": 816814520,
+ "step": 18675,
+ "train_runtime": 144483.9095,
+ "train_tokens_per_second": 5653.325
+ },
+ {
+ "epoch": 1.2577430649070833,
+ "grad_norm": 0.7191371330865693,
+ "learning_rate": 3.0310271189885037e-06,
+ "loss": 0.3063030481338501,
+ "num_input_tokens_seen": 817024696,
+ "step": 18680,
+ "train_runtime": 144517.3756,
+ "train_tokens_per_second": 5653.47
+ },
+ {
+ "epoch": 1.2580797199030433,
+ "grad_norm": 0.7073217630646664,
+ "learning_rate": 3.0285969581321328e-06,
+ "loss": 0.296706485748291,
+ "num_input_tokens_seen": 817244872,
+ "step": 18685,
+ "train_runtime": 144554.5892,
+ "train_tokens_per_second": 5653.538
+ },
+ {
+ "epoch": 1.2584163748990034,
+ "grad_norm": 0.7607230653856618,
+ "learning_rate": 3.0261673485723343e-06,
+ "loss": 0.3513062000274658,
+ "num_input_tokens_seen": 817456408,
+ "step": 18690,
+ "train_runtime": 144592.758,
+ "train_tokens_per_second": 5653.509
+ },
+ {
+ "epoch": 1.2587530298949636,
+ "grad_norm": 0.7009162874744629,
+ "learning_rate": 3.023738290988544e-06,
+ "loss": 0.28501205444335936,
+ "num_input_tokens_seen": 817681112,
+ "step": 18695,
+ "train_runtime": 144634.3333,
+ "train_tokens_per_second": 5653.437
+ },
+ {
+ "epoch": 1.2590896848909239,
+ "grad_norm": 0.8160907857921498,
+ "learning_rate": 3.021309786060037e-06,
+ "loss": 0.30846569538116453,
+ "num_input_tokens_seen": 817885488,
+ "step": 18700,
+ "train_runtime": 144674.0076,
+ "train_tokens_per_second": 5653.299
+ },
+ {
+ "epoch": 1.259426339886884,
+ "grad_norm": 0.7466687057656971,
+ "learning_rate": 3.018881834465941e-06,
+ "loss": 0.290647029876709,
+ "num_input_tokens_seen": 818098808,
+ "step": 18705,
+ "train_runtime": 144712.1439,
+ "train_tokens_per_second": 5653.284
+ },
+ {
+ "epoch": 1.259762994882844,
+ "grad_norm": 0.7496967934224382,
+ "learning_rate": 3.016454436885221e-06,
+ "loss": 0.3127418041229248,
+ "num_input_tokens_seen": 818317480,
+ "step": 18710,
+ "train_runtime": 144748.3739,
+ "train_tokens_per_second": 5653.379
+ },
+ {
+ "epoch": 1.2600996498788042,
+ "grad_norm": 0.7415188653692539,
+ "learning_rate": 3.0140275939966934e-06,
+ "loss": 0.3062695026397705,
+ "num_input_tokens_seen": 818539128,
+ "step": 18715,
+ "train_runtime": 144789.8872,
+ "train_tokens_per_second": 5653.289
+ },
+ {
+ "epoch": 1.2604363048747644,
+ "grad_norm": 0.7926838386125326,
+ "learning_rate": 3.0116013064790166e-06,
+ "loss": 0.3510124921798706,
+ "num_input_tokens_seen": 818770136,
+ "step": 18720,
+ "train_runtime": 144822.5073,
+ "train_tokens_per_second": 5653.611
+ },
+ {
+ "epoch": 1.2607729598707245,
+ "grad_norm": 0.838754094026344,
+ "learning_rate": 3.0091755750106937e-06,
+ "loss": 0.31281888484954834,
+ "num_input_tokens_seen": 818987192,
+ "step": 18725,
+ "train_runtime": 144857.9964,
+ "train_tokens_per_second": 5653.724
+ },
+ {
+ "epoch": 1.2611096148666845,
+ "grad_norm": 0.7065966064188791,
+ "learning_rate": 3.006750400270072e-06,
+ "loss": 0.2828061580657959,
+ "num_input_tokens_seen": 819201568,
+ "step": 18730,
+ "train_runtime": 144893.0397,
+ "train_tokens_per_second": 5653.837
+ },
+ {
+ "epoch": 1.2614462698626447,
+ "grad_norm": 0.6761766202190553,
+ "learning_rate": 3.0043257829353466e-06,
+ "loss": 0.2915013790130615,
+ "num_input_tokens_seen": 819423736,
+ "step": 18735,
+ "train_runtime": 144932.2044,
+ "train_tokens_per_second": 5653.842
+ },
+ {
+ "epoch": 1.261782924858605,
+ "grad_norm": 0.7399210906210721,
+ "learning_rate": 3.0019017236845504e-06,
+ "loss": 0.3179666042327881,
+ "num_input_tokens_seen": 819640880,
+ "step": 18740,
+ "train_runtime": 144969.2626,
+ "train_tokens_per_second": 5653.894
+ },
+ {
+ "epoch": 1.262119579854565,
+ "grad_norm": 1.0185583570035832,
+ "learning_rate": 2.9994782231955673e-06,
+ "loss": 0.3084989547729492,
+ "num_input_tokens_seen": 819858960,
+ "step": 18745,
+ "train_runtime": 145006.0911,
+ "train_tokens_per_second": 5653.962
+ },
+ {
+ "epoch": 1.262456234850525,
+ "grad_norm": 0.6872092978916245,
+ "learning_rate": 2.9970552821461185e-06,
+ "loss": 0.28854634761810305,
+ "num_input_tokens_seen": 820070184,
+ "step": 18750,
+ "train_runtime": 145045.6889,
+ "train_tokens_per_second": 5653.875
+ },
+ {
+ "epoch": 1.2627928898464853,
+ "grad_norm": 0.8684996226363714,
+ "learning_rate": 2.994632901213775e-06,
+ "loss": 0.2994524955749512,
+ "num_input_tokens_seen": 820280600,
+ "step": 18755,
+ "train_runtime": 145083.5094,
+ "train_tokens_per_second": 5653.851
+ },
+ {
+ "epoch": 1.2631295448424456,
+ "grad_norm": 0.8319193699909638,
+ "learning_rate": 2.9922110810759443e-06,
+ "loss": 0.3167592763900757,
+ "num_input_tokens_seen": 820492832,
+ "step": 18760,
+ "train_runtime": 145127.9985,
+ "train_tokens_per_second": 5653.581
+ },
+ {
+ "epoch": 1.2634661998384056,
+ "grad_norm": 0.7753044461490659,
+ "learning_rate": 2.989789822409883e-06,
+ "loss": 0.3189209461212158,
+ "num_input_tokens_seen": 820717400,
+ "step": 18765,
+ "train_runtime": 145162.0094,
+ "train_tokens_per_second": 5653.803
+ },
+ {
+ "epoch": 1.2638028548343656,
+ "grad_norm": 0.7322817419625537,
+ "learning_rate": 2.9873691258926864e-06,
+ "loss": 0.3269504547119141,
+ "num_input_tokens_seen": 820956960,
+ "step": 18770,
+ "train_runtime": 145198.2506,
+ "train_tokens_per_second": 5654.042
+ },
+ {
+ "epoch": 1.2641395098303259,
+ "grad_norm": 0.696541167423223,
+ "learning_rate": 2.9849489922012985e-06,
+ "loss": 0.3042433261871338,
+ "num_input_tokens_seen": 821165096,
+ "step": 18775,
+ "train_runtime": 145239.0804,
+ "train_tokens_per_second": 5653.885
+ },
+ {
+ "epoch": 1.264476164826286,
+ "grad_norm": 0.7452468533447384,
+ "learning_rate": 2.982529422012498e-06,
+ "loss": 0.338847017288208,
+ "num_input_tokens_seen": 821398992,
+ "step": 18780,
+ "train_runtime": 145277.8958,
+ "train_tokens_per_second": 5653.985
+ },
+ {
+ "epoch": 1.2648128198222461,
+ "grad_norm": 0.7815954898874985,
+ "learning_rate": 2.9801104160029127e-06,
+ "loss": 0.2665702819824219,
+ "num_input_tokens_seen": 821606024,
+ "step": 18785,
+ "train_runtime": 145319.4203,
+ "train_tokens_per_second": 5653.794
+ },
+ {
+ "epoch": 1.2651494748182062,
+ "grad_norm": 0.7521512397733022,
+ "learning_rate": 2.9776919748490063e-06,
+ "loss": 0.31261477470397947,
+ "num_input_tokens_seen": 821822040,
+ "step": 18790,
+ "train_runtime": 145359.9037,
+ "train_tokens_per_second": 5653.705
+ },
+ {
+ "epoch": 1.2654861298141664,
+ "grad_norm": 0.8383923854339487,
+ "learning_rate": 2.975274099227092e-06,
+ "loss": 0.32703323364257814,
+ "num_input_tokens_seen": 822027760,
+ "step": 18795,
+ "train_runtime": 145403.5301,
+ "train_tokens_per_second": 5653.424
+ },
+ {
+ "epoch": 1.2658227848101267,
+ "grad_norm": 0.8931253914400595,
+ "learning_rate": 2.9728567898133183e-06,
+ "loss": 0.31384882926940916,
+ "num_input_tokens_seen": 822249072,
+ "step": 18800,
+ "train_runtime": 145447.7238,
+ "train_tokens_per_second": 5653.227
+ },
+ {
+ "epoch": 1.2661594398060867,
+ "grad_norm": 0.7239471432554654,
+ "learning_rate": 2.9704400472836816e-06,
+ "loss": 0.3009427309036255,
+ "num_input_tokens_seen": 822468088,
+ "step": 18805,
+ "train_runtime": 145491.919,
+ "train_tokens_per_second": 5653.016
+ },
+ {
+ "epoch": 1.266496094802047,
+ "grad_norm": 0.8248452170258889,
+ "learning_rate": 2.9680238723140108e-06,
+ "loss": 0.33100879192352295,
+ "num_input_tokens_seen": 822682520,
+ "step": 18810,
+ "train_runtime": 145531.6555,
+ "train_tokens_per_second": 5652.946
+ },
+ {
+ "epoch": 1.266832749798007,
+ "grad_norm": 0.7312225706048872,
+ "learning_rate": 2.9656082655799866e-06,
+ "loss": 0.3179032325744629,
+ "num_input_tokens_seen": 822894336,
+ "step": 18815,
+ "train_runtime": 145569.5437,
+ "train_tokens_per_second": 5652.929
+ },
+ {
+ "epoch": 1.2671694047939672,
+ "grad_norm": 0.8914282863081362,
+ "learning_rate": 2.9631932277571225e-06,
+ "loss": 0.31571259498596194,
+ "num_input_tokens_seen": 823116688,
+ "step": 18820,
+ "train_runtime": 145602.2926,
+ "train_tokens_per_second": 5653.185
+ },
+ {
+ "epoch": 1.2675060597899273,
+ "grad_norm": 0.7436639756998058,
+ "learning_rate": 2.9607787595207794e-06,
+ "loss": 0.2941970109939575,
+ "num_input_tokens_seen": 823334680,
+ "step": 18825,
+ "train_runtime": 145642.6854,
+ "train_tokens_per_second": 5653.114
+ },
+ {
+ "epoch": 1.2678427147858875,
+ "grad_norm": 0.6064296816907161,
+ "learning_rate": 2.9583648615461514e-06,
+ "loss": 0.2951462507247925,
+ "num_input_tokens_seen": 823548976,
+ "step": 18830,
+ "train_runtime": 145684.7379,
+ "train_tokens_per_second": 5652.953
+ },
+ {
+ "epoch": 1.2681793697818475,
+ "grad_norm": 0.8015900434605868,
+ "learning_rate": 2.955951534508281e-06,
+ "loss": 0.32911660671234133,
+ "num_input_tokens_seen": 823759784,
+ "step": 18835,
+ "train_runtime": 145720.3213,
+ "train_tokens_per_second": 5653.019
+ },
+ {
+ "epoch": 1.2685160247778078,
+ "grad_norm": 0.6806231841224424,
+ "learning_rate": 2.9535387790820452e-06,
+ "loss": 0.3123172283172607,
+ "num_input_tokens_seen": 823989856,
+ "step": 18840,
+ "train_runtime": 145763.7657,
+ "train_tokens_per_second": 5652.913
+ },
+ {
+ "epoch": 1.2688526797737678,
+ "grad_norm": 0.827759809808618,
+ "learning_rate": 2.951126595942165e-06,
+ "loss": 0.29312863349914553,
+ "num_input_tokens_seen": 824217960,
+ "step": 18845,
+ "train_runtime": 145800.553,
+ "train_tokens_per_second": 5653.051
+ },
+ {
+ "epoch": 1.269189334769728,
+ "grad_norm": 0.7164470604004327,
+ "learning_rate": 2.948714985763197e-06,
+ "loss": 0.2757392883300781,
+ "num_input_tokens_seen": 824421568,
+ "step": 18850,
+ "train_runtime": 145841.8634,
+ "train_tokens_per_second": 5652.846
+ },
+ {
+ "epoch": 1.269525989765688,
+ "grad_norm": 0.8496090603821623,
+ "learning_rate": 2.9463039492195445e-06,
+ "loss": 0.3131323575973511,
+ "num_input_tokens_seen": 824655760,
+ "step": 18855,
+ "train_runtime": 145880.5542,
+ "train_tokens_per_second": 5652.952
+ },
+ {
+ "epoch": 1.2698626447616483,
+ "grad_norm": 0.7352726813474353,
+ "learning_rate": 2.9438934869854417e-06,
+ "loss": 0.2902811527252197,
+ "num_input_tokens_seen": 824879032,
+ "step": 18860,
+ "train_runtime": 145923.738,
+ "train_tokens_per_second": 5652.809
+ },
+ {
+ "epoch": 1.2701992997576084,
+ "grad_norm": 0.7811715951881948,
+ "learning_rate": 2.9414835997349705e-06,
+ "loss": 0.2894813060760498,
+ "num_input_tokens_seen": 825095288,
+ "step": 18865,
+ "train_runtime": 145960.032,
+ "train_tokens_per_second": 5652.885
+ },
+ {
+ "epoch": 1.2705359547535686,
+ "grad_norm": 0.6562307917959052,
+ "learning_rate": 2.939074288142045e-06,
+ "loss": 0.32269885540008547,
+ "num_input_tokens_seen": 825326376,
+ "step": 18870,
+ "train_runtime": 145998.5987,
+ "train_tokens_per_second": 5652.975
+ },
+ {
+ "epoch": 1.2708726097495286,
+ "grad_norm": 0.7677795002971816,
+ "learning_rate": 2.9366655528804257e-06,
+ "loss": 0.29350924491882324,
+ "num_input_tokens_seen": 825544816,
+ "step": 18875,
+ "train_runtime": 146042.0539,
+ "train_tokens_per_second": 5652.788
+ },
+ {
+ "epoch": 1.271209264745489,
+ "grad_norm": 0.6905930617244008,
+ "learning_rate": 2.934257394623702e-06,
+ "loss": 0.2879796504974365,
+ "num_input_tokens_seen": 825768752,
+ "step": 18880,
+ "train_runtime": 146079.3121,
+ "train_tokens_per_second": 5652.88
+ },
+ {
+ "epoch": 1.271545919741449,
+ "grad_norm": 0.7151332172339365,
+ "learning_rate": 2.9318498140453134e-06,
+ "loss": 0.32280969619750977,
+ "num_input_tokens_seen": 825987712,
+ "step": 18885,
+ "train_runtime": 146122.0243,
+ "train_tokens_per_second": 5652.726
+ },
+ {
+ "epoch": 1.2718825747374092,
+ "grad_norm": 0.8504766443077242,
+ "learning_rate": 2.929442811818527e-06,
+ "loss": 0.3143021106719971,
+ "num_input_tokens_seen": 826196048,
+ "step": 18890,
+ "train_runtime": 146161.5263,
+ "train_tokens_per_second": 5652.623
+ },
+ {
+ "epoch": 1.2722192297333692,
+ "grad_norm": 0.797394168026311,
+ "learning_rate": 2.927036388616457e-06,
+ "loss": 0.29908573627471924,
+ "num_input_tokens_seen": 826399384,
+ "step": 18895,
+ "train_runtime": 146195.9133,
+ "train_tokens_per_second": 5652.685
+ },
+ {
+ "epoch": 1.2725558847293295,
+ "grad_norm": 0.8320805975958718,
+ "learning_rate": 2.9246305451120494e-06,
+ "loss": 0.3228297233581543,
+ "num_input_tokens_seen": 826603896,
+ "step": 18900,
+ "train_runtime": 146234.1471,
+ "train_tokens_per_second": 5652.605
+ },
+ {
+ "epoch": 1.2728925397252895,
+ "grad_norm": 0.7307797880302476,
+ "learning_rate": 2.922225281978095e-06,
+ "loss": 0.3007101058959961,
+ "num_input_tokens_seen": 826821328,
+ "step": 18905,
+ "train_runtime": 146279.6181,
+ "train_tokens_per_second": 5652.334
+ },
+ {
+ "epoch": 1.2732291947212497,
+ "grad_norm": 0.7406348258598292,
+ "learning_rate": 2.919820599887212e-06,
+ "loss": 0.3024411916732788,
+ "num_input_tokens_seen": 827039792,
+ "step": 18910,
+ "train_runtime": 146321.5475,
+ "train_tokens_per_second": 5652.208
+ },
+ {
+ "epoch": 1.2735658497172098,
+ "grad_norm": 0.8447453780821402,
+ "learning_rate": 2.917416499511866e-06,
+ "loss": 0.30447540283203123,
+ "num_input_tokens_seen": 827242576,
+ "step": 18915,
+ "train_runtime": 146363.2378,
+ "train_tokens_per_second": 5651.983
+ },
+ {
+ "epoch": 1.27390250471317,
+ "grad_norm": 0.7936589269610094,
+ "learning_rate": 2.9150129815243553e-06,
+ "loss": 0.288972806930542,
+ "num_input_tokens_seen": 827457896,
+ "step": 18920,
+ "train_runtime": 146404.1905,
+ "train_tokens_per_second": 5651.873
+ },
+ {
+ "epoch": 1.27423915970913,
+ "grad_norm": 0.850907349149432,
+ "learning_rate": 2.912610046596816e-06,
+ "loss": 0.2999333620071411,
+ "num_input_tokens_seen": 827666848,
+ "step": 18925,
+ "train_runtime": 146446.094,
+ "train_tokens_per_second": 5651.683
+ },
+ {
+ "epoch": 1.2745758147050903,
+ "grad_norm": 0.6871438604531044,
+ "learning_rate": 2.91020769540122e-06,
+ "loss": 0.3152765274047852,
+ "num_input_tokens_seen": 827866000,
+ "step": 18930,
+ "train_runtime": 146485.6336,
+ "train_tokens_per_second": 5651.517
+ },
+ {
+ "epoch": 1.2749124697010503,
+ "grad_norm": 0.679666459947843,
+ "learning_rate": 2.9078059286093795e-06,
+ "loss": 0.2963433027267456,
+ "num_input_tokens_seen": 828072488,
+ "step": 18935,
+ "train_runtime": 146524.0873,
+ "train_tokens_per_second": 5651.443
+ },
+ {
+ "epoch": 1.2752491246970106,
+ "grad_norm": 0.6861610947075387,
+ "learning_rate": 2.905404746892937e-06,
+ "loss": 0.28579273223876955,
+ "num_input_tokens_seen": 828287296,
+ "step": 18940,
+ "train_runtime": 146560.9928,
+ "train_tokens_per_second": 5651.485
+ },
+ {
+ "epoch": 1.2755857796929706,
+ "grad_norm": 0.7154785961890833,
+ "learning_rate": 2.903004150923382e-06,
+ "loss": 0.3039486169815063,
+ "num_input_tokens_seen": 828510696,
+ "step": 18945,
+ "train_runtime": 146598.3451,
+ "train_tokens_per_second": 5651.569
+ },
+ {
+ "epoch": 1.2759224346889309,
+ "grad_norm": 0.7669932436785718,
+ "learning_rate": 2.9006041413720252e-06,
+ "loss": 0.3282867431640625,
+ "num_input_tokens_seen": 828734304,
+ "step": 18950,
+ "train_runtime": 146645.3601,
+ "train_tokens_per_second": 5651.282
+ },
+ {
+ "epoch": 1.2762590896848909,
+ "grad_norm": 0.6719242223720826,
+ "learning_rate": 2.8982047189100276e-06,
+ "loss": 0.30024127960205077,
+ "num_input_tokens_seen": 828967864,
+ "step": 18955,
+ "train_runtime": 146680.0954,
+ "train_tokens_per_second": 5651.536
+ },
+ {
+ "epoch": 1.2765957446808511,
+ "grad_norm": 0.7852643316945973,
+ "learning_rate": 2.895805884208378e-06,
+ "loss": 0.3082295894622803,
+ "num_input_tokens_seen": 829167224,
+ "step": 18960,
+ "train_runtime": 146717.6947,
+ "train_tokens_per_second": 5651.447
+ },
+ {
+ "epoch": 1.2769323996768112,
+ "grad_norm": 0.8516866669455334,
+ "learning_rate": 2.8934076379379016e-06,
+ "loss": 0.31248860359191893,
+ "num_input_tokens_seen": 829396808,
+ "step": 18965,
+ "train_runtime": 146755.3528,
+ "train_tokens_per_second": 5651.561
+ },
+ {
+ "epoch": 1.2772690546727714,
+ "grad_norm": 0.7752902430798366,
+ "learning_rate": 2.891009980769266e-06,
+ "loss": 0.3060382127761841,
+ "num_input_tokens_seen": 829619304,
+ "step": 18970,
+ "train_runtime": 146798.1501,
+ "train_tokens_per_second": 5651.429
+ },
+ {
+ "epoch": 1.2776057096687314,
+ "grad_norm": 0.8744465210673298,
+ "learning_rate": 2.88861291337296e-06,
+ "loss": 0.29961209297180175,
+ "num_input_tokens_seen": 829837784,
+ "step": 18975,
+ "train_runtime": 146836.9106,
+ "train_tokens_per_second": 5651.425
+ },
+ {
+ "epoch": 1.2779423646646917,
+ "grad_norm": 0.7351797146412703,
+ "learning_rate": 2.8862164364193236e-06,
+ "loss": 0.2904377460479736,
+ "num_input_tokens_seen": 830058208,
+ "step": 18980,
+ "train_runtime": 146873.1011,
+ "train_tokens_per_second": 5651.533
+ },
+ {
+ "epoch": 1.2782790196606517,
+ "grad_norm": 0.7312431717294958,
+ "learning_rate": 2.8838205505785168e-06,
+ "loss": 0.3134331703186035,
+ "num_input_tokens_seen": 830283008,
+ "step": 18985,
+ "train_runtime": 146914.2968,
+ "train_tokens_per_second": 5651.479
+ },
+ {
+ "epoch": 1.278615674656612,
+ "grad_norm": 0.663622797522999,
+ "learning_rate": 2.881425256520547e-06,
+ "loss": 0.2921187400817871,
+ "num_input_tokens_seen": 830500192,
+ "step": 18990,
+ "train_runtime": 146951.0522,
+ "train_tokens_per_second": 5651.543
+ },
+ {
+ "epoch": 1.278952329652572,
+ "grad_norm": 0.7237194461676807,
+ "learning_rate": 2.8790305549152487e-06,
+ "loss": 0.26155910491943357,
+ "num_input_tokens_seen": 830725544,
+ "step": 18995,
+ "train_runtime": 146990.9253,
+ "train_tokens_per_second": 5651.543
+ },
+ {
+ "epoch": 1.2792889846485322,
+ "grad_norm": 0.7631364393415948,
+ "learning_rate": 2.876636446432292e-06,
+ "loss": 0.3240873098373413,
+ "num_input_tokens_seen": 830948824,
+ "step": 19000,
+ "train_runtime": 147025.5963,
+ "train_tokens_per_second": 5651.729
+ },
+ {
+ "epoch": 1.2796256396444923,
+ "grad_norm": 0.7251110163689186,
+ "learning_rate": 2.8742429317411826e-06,
+ "loss": 0.30750136375427245,
+ "num_input_tokens_seen": 831173768,
+ "step": 19005,
+ "train_runtime": 147062.3862,
+ "train_tokens_per_second": 5651.845
+ },
+ {
+ "epoch": 1.2799622946404525,
+ "grad_norm": 0.9679509381325586,
+ "learning_rate": 2.8718500115112585e-06,
+ "loss": 0.30974340438842773,
+ "num_input_tokens_seen": 831380944,
+ "step": 19010,
+ "train_runtime": 147096.7442,
+ "train_tokens_per_second": 5651.933
+ },
+ {
+ "epoch": 1.2802989496364126,
+ "grad_norm": 0.7640168557000633,
+ "learning_rate": 2.869457686411691e-06,
+ "loss": 0.33095221519470214,
+ "num_input_tokens_seen": 831589040,
+ "step": 19015,
+ "train_runtime": 147137.7505,
+ "train_tokens_per_second": 5651.772
+ },
+ {
+ "epoch": 1.2806356046323728,
+ "grad_norm": 0.7459470398332059,
+ "learning_rate": 2.8670659571114912e-06,
+ "loss": 0.30541186332702636,
+ "num_input_tokens_seen": 831816448,
+ "step": 19020,
+ "train_runtime": 147176.0251,
+ "train_tokens_per_second": 5651.847
+ },
+ {
+ "epoch": 1.2809722596283328,
+ "grad_norm": 0.6423476554873053,
+ "learning_rate": 2.864674824279491e-06,
+ "loss": 0.2939921855926514,
+ "num_input_tokens_seen": 832035736,
+ "step": 19025,
+ "train_runtime": 147213.464,
+ "train_tokens_per_second": 5651.9
+ },
+ {
+ "epoch": 1.281308914624293,
+ "grad_norm": 0.7355078108437945,
+ "learning_rate": 2.862284288584371e-06,
+ "loss": 0.2885020971298218,
+ "num_input_tokens_seen": 832264232,
+ "step": 19030,
+ "train_runtime": 147251.1103,
+ "train_tokens_per_second": 5652.006
+ },
+ {
+ "epoch": 1.2816455696202531,
+ "grad_norm": 0.8373034806587798,
+ "learning_rate": 2.8598943506946297e-06,
+ "loss": 0.300329327583313,
+ "num_input_tokens_seen": 832482744,
+ "step": 19035,
+ "train_runtime": 147288.7273,
+ "train_tokens_per_second": 5652.047
+ },
+ {
+ "epoch": 1.2819822246162134,
+ "grad_norm": 0.7138141742816384,
+ "learning_rate": 2.85750501127861e-06,
+ "loss": 0.277694034576416,
+ "num_input_tokens_seen": 832697464,
+ "step": 19040,
+ "train_runtime": 147329.6861,
+ "train_tokens_per_second": 5651.933
+ },
+ {
+ "epoch": 1.2823188796121734,
+ "grad_norm": 0.6608599560904919,
+ "learning_rate": 2.8551162710044827e-06,
+ "loss": 0.30247793197631834,
+ "num_input_tokens_seen": 832912640,
+ "step": 19045,
+ "train_runtime": 147365.9859,
+ "train_tokens_per_second": 5652.001
+ },
+ {
+ "epoch": 1.2826555346081336,
+ "grad_norm": 0.7589903744520924,
+ "learning_rate": 2.8527281305402503e-06,
+ "loss": 0.32385740280151365,
+ "num_input_tokens_seen": 833133224,
+ "step": 19050,
+ "train_runtime": 147403.1332,
+ "train_tokens_per_second": 5652.073
+ },
+ {
+ "epoch": 1.2829921896040937,
+ "grad_norm": 0.8057473648914902,
+ "learning_rate": 2.8503405905537496e-06,
+ "loss": 0.29269657135009763,
+ "num_input_tokens_seen": 833346936,
+ "step": 19055,
+ "train_runtime": 147442.1312,
+ "train_tokens_per_second": 5652.027
+ },
+ {
+ "epoch": 1.283328844600054,
+ "grad_norm": 0.7212257083936001,
+ "learning_rate": 2.847953651712647e-06,
+ "loss": 0.3061253309249878,
+ "num_input_tokens_seen": 833581464,
+ "step": 19060,
+ "train_runtime": 147477.4905,
+ "train_tokens_per_second": 5652.262
+ },
+ {
+ "epoch": 1.283665499596014,
+ "grad_norm": 0.8465650818228247,
+ "learning_rate": 2.8455673146844426e-06,
+ "loss": 0.30112695693969727,
+ "num_input_tokens_seen": 833784496,
+ "step": 19065,
+ "train_runtime": 147521.4424,
+ "train_tokens_per_second": 5651.955
+ },
+ {
+ "epoch": 1.2840021545919742,
+ "grad_norm": 0.7588100951853031,
+ "learning_rate": 2.8431815801364717e-06,
+ "loss": 0.3190593719482422,
+ "num_input_tokens_seen": 834001896,
+ "step": 19070,
+ "train_runtime": 147563.6694,
+ "train_tokens_per_second": 5651.811
+ },
+ {
+ "epoch": 1.2843388095879342,
+ "grad_norm": 0.8096844572768415,
+ "learning_rate": 2.8407964487358906e-06,
+ "loss": 0.3083981513977051,
+ "num_input_tokens_seen": 834226528,
+ "step": 19075,
+ "train_runtime": 147606.1012,
+ "train_tokens_per_second": 5651.708
+ },
+ {
+ "epoch": 1.2846754645838945,
+ "grad_norm": 0.8121583620232876,
+ "learning_rate": 2.838411921149702e-06,
+ "loss": 0.30108425617218015,
+ "num_input_tokens_seen": 834428112,
+ "step": 19080,
+ "train_runtime": 147643.2603,
+ "train_tokens_per_second": 5651.651
+ },
+ {
+ "epoch": 1.2850121195798545,
+ "grad_norm": 0.7654588268965005,
+ "learning_rate": 2.836027998044722e-06,
+ "loss": 0.3017698287963867,
+ "num_input_tokens_seen": 834638200,
+ "step": 19085,
+ "train_runtime": 147687.2931,
+ "train_tokens_per_second": 5651.388
+ },
+ {
+ "epoch": 1.2853487745758148,
+ "grad_norm": 0.7769786248168525,
+ "learning_rate": 2.8336446800876143e-06,
+ "loss": 0.3067366600036621,
+ "num_input_tokens_seen": 834833144,
+ "step": 19090,
+ "train_runtime": 147722.5693,
+ "train_tokens_per_second": 5651.358
+ },
+ {
+ "epoch": 1.2856854295717748,
+ "grad_norm": 0.8520512402767025,
+ "learning_rate": 2.8312619679448637e-06,
+ "loss": 0.2865460872650146,
+ "num_input_tokens_seen": 835055128,
+ "step": 19095,
+ "train_runtime": 147762.1093,
+ "train_tokens_per_second": 5651.348
+ },
+ {
+ "epoch": 1.286022084567735,
+ "grad_norm": 0.8168300340689107,
+ "learning_rate": 2.8288798622827884e-06,
+ "loss": 0.31798462867736815,
+ "num_input_tokens_seen": 835252248,
+ "step": 19100,
+ "train_runtime": 147802.5202,
+ "train_tokens_per_second": 5651.137
+ },
+ {
+ "epoch": 1.286358739563695,
+ "grad_norm": 0.9604203872867969,
+ "learning_rate": 2.826498363767536e-06,
+ "loss": 0.316135311126709,
+ "num_input_tokens_seen": 835451680,
+ "step": 19105,
+ "train_runtime": 147840.3751,
+ "train_tokens_per_second": 5651.039
+ },
+ {
+ "epoch": 1.2866953945596553,
+ "grad_norm": 0.8232657359064803,
+ "learning_rate": 2.8241174730650855e-06,
+ "loss": 0.3065540313720703,
+ "num_input_tokens_seen": 835675080,
+ "step": 19110,
+ "train_runtime": 147880.4439,
+ "train_tokens_per_second": 5651.018
+ },
+ {
+ "epoch": 1.2870320495556153,
+ "grad_norm": 0.7796058041196822,
+ "learning_rate": 2.821737190841244e-06,
+ "loss": 0.29743661880493166,
+ "num_input_tokens_seen": 835891040,
+ "step": 19115,
+ "train_runtime": 147915.5543,
+ "train_tokens_per_second": 5651.137
+ },
+ {
+ "epoch": 1.2873687045515756,
+ "grad_norm": 0.7312714159541509,
+ "learning_rate": 2.8193575177616537e-06,
+ "loss": 0.2965965986251831,
+ "num_input_tokens_seen": 836111592,
+ "step": 19120,
+ "train_runtime": 147959.8125,
+ "train_tokens_per_second": 5650.937
+ },
+ {
+ "epoch": 1.2877053595475356,
+ "grad_norm": 0.7000449200994382,
+ "learning_rate": 2.816978454491777e-06,
+ "loss": 0.2831472396850586,
+ "num_input_tokens_seen": 836342224,
+ "step": 19125,
+ "train_runtime": 147994.2807,
+ "train_tokens_per_second": 5651.179
+ },
+ {
+ "epoch": 1.2880420145434959,
+ "grad_norm": 0.8484575372461595,
+ "learning_rate": 2.814600001696919e-06,
+ "loss": 0.31551055908203124,
+ "num_input_tokens_seen": 836555312,
+ "step": 19130,
+ "train_runtime": 148030.887,
+ "train_tokens_per_second": 5651.221
+ },
+ {
+ "epoch": 1.288378669539456,
+ "grad_norm": 0.7939086544066017,
+ "learning_rate": 2.812222160042196e-06,
+ "loss": 0.3081847906112671,
+ "num_input_tokens_seen": 836781384,
+ "step": 19135,
+ "train_runtime": 148071.8667,
+ "train_tokens_per_second": 5651.184
+ },
+ {
+ "epoch": 1.2887153245354162,
+ "grad_norm": 0.7293883099186741,
+ "learning_rate": 2.809844930192573e-06,
+ "loss": 0.31632413864135744,
+ "num_input_tokens_seen": 836996448,
+ "step": 19140,
+ "train_runtime": 148110.3971,
+ "train_tokens_per_second": 5651.166
+ },
+ {
+ "epoch": 1.2890519795313762,
+ "grad_norm": 0.7845647411236363,
+ "learning_rate": 2.807468312812831e-06,
+ "loss": 0.3241159677505493,
+ "num_input_tokens_seen": 837211800,
+ "step": 19145,
+ "train_runtime": 148145.8757,
+ "train_tokens_per_second": 5651.266
+ },
+ {
+ "epoch": 1.2893886345273364,
+ "grad_norm": 0.7305402054857416,
+ "learning_rate": 2.805092308567584e-06,
+ "loss": 0.2902689933776855,
+ "num_input_tokens_seen": 837439120,
+ "step": 19150,
+ "train_runtime": 148177.9274,
+ "train_tokens_per_second": 5651.578
+ },
+ {
+ "epoch": 1.2897252895232965,
+ "grad_norm": 0.706253740316737,
+ "learning_rate": 2.802716918121273e-06,
+ "loss": 0.2991164207458496,
+ "num_input_tokens_seen": 837665040,
+ "step": 19155,
+ "train_runtime": 148222.7975,
+ "train_tokens_per_second": 5651.391
+ },
+ {
+ "epoch": 1.2900619445192567,
+ "grad_norm": 0.7258823029517253,
+ "learning_rate": 2.8003421421381704e-06,
+ "loss": 0.29270741939544676,
+ "num_input_tokens_seen": 837893216,
+ "step": 19160,
+ "train_runtime": 148255.3261,
+ "train_tokens_per_second": 5651.69
+ },
+ {
+ "epoch": 1.2903985995152167,
+ "grad_norm": 0.7437197036098059,
+ "learning_rate": 2.797967981282371e-06,
+ "loss": 0.3194019317626953,
+ "num_input_tokens_seen": 838115864,
+ "step": 19165,
+ "train_runtime": 148297.7873,
+ "train_tokens_per_second": 5651.574
+ },
+ {
+ "epoch": 1.290735254511177,
+ "grad_norm": 0.7205908995167096,
+ "learning_rate": 2.795594436217808e-06,
+ "loss": 0.29573659896850585,
+ "num_input_tokens_seen": 838332696,
+ "step": 19170,
+ "train_runtime": 148338.6228,
+ "train_tokens_per_second": 5651.48
+ },
+ {
+ "epoch": 1.291071909507137,
+ "grad_norm": 0.6798097348635077,
+ "learning_rate": 2.7932215076082277e-06,
+ "loss": 0.29705004692077636,
+ "num_input_tokens_seen": 838541408,
+ "step": 19175,
+ "train_runtime": 148382.7784,
+ "train_tokens_per_second": 5651.204
+ },
+ {
+ "epoch": 1.2914085645030973,
+ "grad_norm": 0.7142148877160069,
+ "learning_rate": 2.790849196117219e-06,
+ "loss": 0.30868139266967776,
+ "num_input_tokens_seen": 838777616,
+ "step": 19180,
+ "train_runtime": 148425.245,
+ "train_tokens_per_second": 5651.179
+ },
+ {
+ "epoch": 1.2917452194990573,
+ "grad_norm": 0.7147263101151003,
+ "learning_rate": 2.7884775024081856e-06,
+ "loss": 0.30340185165405276,
+ "num_input_tokens_seen": 838994528,
+ "step": 19185,
+ "train_runtime": 148461.9522,
+ "train_tokens_per_second": 5651.243
+ },
+ {
+ "epoch": 1.2920818744950175,
+ "grad_norm": 0.7411183792886813,
+ "learning_rate": 2.786106427144367e-06,
+ "loss": 0.292337703704834,
+ "num_input_tokens_seen": 839211816,
+ "step": 19190,
+ "train_runtime": 148497.1246,
+ "train_tokens_per_second": 5651.367
+ },
+ {
+ "epoch": 1.2924185294909776,
+ "grad_norm": 0.784444211412864,
+ "learning_rate": 2.783735970988828e-06,
+ "loss": 0.27770419120788575,
+ "num_input_tokens_seen": 839425992,
+ "step": 19195,
+ "train_runtime": 148536.5284,
+ "train_tokens_per_second": 5651.31
+ },
+ {
+ "epoch": 1.2927551844869378,
+ "grad_norm": 0.7627571829537649,
+ "learning_rate": 2.7813661346044566e-06,
+ "loss": 0.31344926357269287,
+ "num_input_tokens_seen": 839639736,
+ "step": 19200,
+ "train_runtime": 148570.4019,
+ "train_tokens_per_second": 5651.46
+ },
+ {
+ "epoch": 1.2930918394828979,
+ "grad_norm": 0.8342081884001865,
+ "learning_rate": 2.7789969186539713e-06,
+ "loss": 0.30696806907653806,
+ "num_input_tokens_seen": 839843624,
+ "step": 19205,
+ "train_runtime": 148611.1575,
+ "train_tokens_per_second": 5651.282
+ },
+ {
+ "epoch": 1.293428494478858,
+ "grad_norm": 0.7673215752721638,
+ "learning_rate": 2.776628323799915e-06,
+ "loss": 0.32203989028930663,
+ "num_input_tokens_seen": 840061080,
+ "step": 19210,
+ "train_runtime": 148644.4524,
+ "train_tokens_per_second": 5651.48
+ },
+ {
+ "epoch": 1.2937651494748181,
+ "grad_norm": 0.8083666833563818,
+ "learning_rate": 2.774260350704657e-06,
+ "loss": 0.3181305408477783,
+ "num_input_tokens_seen": 840276920,
+ "step": 19215,
+ "train_runtime": 148679.4894,
+ "train_tokens_per_second": 5651.599
+ },
+ {
+ "epoch": 1.2941018044707784,
+ "grad_norm": 0.7678745273221902,
+ "learning_rate": 2.771893000030398e-06,
+ "loss": 0.3186459541320801,
+ "num_input_tokens_seen": 840489952,
+ "step": 19220,
+ "train_runtime": 148716.8291,
+ "train_tokens_per_second": 5651.613
+ },
+ {
+ "epoch": 1.2944384594667384,
+ "grad_norm": 0.7272017085399577,
+ "learning_rate": 2.7695262724391526e-06,
+ "loss": 0.29137458801269533,
+ "num_input_tokens_seen": 840702144,
+ "step": 19225,
+ "train_runtime": 148759.5159,
+ "train_tokens_per_second": 5651.418
+ },
+ {
+ "epoch": 1.2947751144626987,
+ "grad_norm": 0.7630834988917379,
+ "learning_rate": 2.767160168592777e-06,
+ "loss": 0.33965349197387695,
+ "num_input_tokens_seen": 840931688,
+ "step": 19230,
+ "train_runtime": 148806.4026,
+ "train_tokens_per_second": 5651.179
+ },
+ {
+ "epoch": 1.2951117694586587,
+ "grad_norm": 0.7684748875099028,
+ "learning_rate": 2.7647946891529355e-06,
+ "loss": 0.3225222110748291,
+ "num_input_tokens_seen": 841129968,
+ "step": 19235,
+ "train_runtime": 148842.7883,
+ "train_tokens_per_second": 5651.13
+ },
+ {
+ "epoch": 1.295448424454619,
+ "grad_norm": 0.728042153912699,
+ "learning_rate": 2.7624298347811342e-06,
+ "loss": 0.3056331157684326,
+ "num_input_tokens_seen": 841331904,
+ "step": 19240,
+ "train_runtime": 148884.4042,
+ "train_tokens_per_second": 5650.907
+ },
+ {
+ "epoch": 1.295785079450579,
+ "grad_norm": 0.7854514457019474,
+ "learning_rate": 2.7600656061386945e-06,
+ "loss": 0.31122641563415526,
+ "num_input_tokens_seen": 841562064,
+ "step": 19245,
+ "train_runtime": 148919.1545,
+ "train_tokens_per_second": 5651.134
+ },
+ {
+ "epoch": 1.2961217344465392,
+ "grad_norm": 0.7775133786016957,
+ "learning_rate": 2.757702003886765e-06,
+ "loss": 0.2915956974029541,
+ "num_input_tokens_seen": 841780328,
+ "step": 19250,
+ "train_runtime": 148953.299,
+ "train_tokens_per_second": 5651.304
+ },
+ {
+ "epoch": 1.2964583894424992,
+ "grad_norm": 1.058051599935181,
+ "learning_rate": 2.7553390286863205e-06,
+ "loss": 0.29277448654174804,
+ "num_input_tokens_seen": 841982320,
+ "step": 19255,
+ "train_runtime": 148990.52,
+ "train_tokens_per_second": 5651.248
+ },
+ {
+ "epoch": 1.2967950444384595,
+ "grad_norm": 0.7544624265455544,
+ "learning_rate": 2.752976681198159e-06,
+ "loss": 0.3195603609085083,
+ "num_input_tokens_seen": 842210920,
+ "step": 19260,
+ "train_runtime": 149028.3612,
+ "train_tokens_per_second": 5651.347
+ },
+ {
+ "epoch": 1.2971316994344195,
+ "grad_norm": 0.8113254179825854,
+ "learning_rate": 2.750614962082901e-06,
+ "loss": 0.28080661296844484,
+ "num_input_tokens_seen": 842433856,
+ "step": 19265,
+ "train_runtime": 149066.4868,
+ "train_tokens_per_second": 5651.397
+ },
+ {
+ "epoch": 1.2974683544303798,
+ "grad_norm": 0.7219059530126368,
+ "learning_rate": 2.748253872001e-06,
+ "loss": 0.3025611162185669,
+ "num_input_tokens_seen": 842651984,
+ "step": 19270,
+ "train_runtime": 149103.7773,
+ "train_tokens_per_second": 5651.446
+ },
+ {
+ "epoch": 1.2978050094263398,
+ "grad_norm": 0.7933392458748796,
+ "learning_rate": 2.74589341161272e-06,
+ "loss": 0.30447611808776853,
+ "num_input_tokens_seen": 842849968,
+ "step": 19275,
+ "train_runtime": 149143.5367,
+ "train_tokens_per_second": 5651.267
+ },
+ {
+ "epoch": 1.2981416644223,
+ "grad_norm": 0.8775197487075084,
+ "learning_rate": 2.743533581578163e-06,
+ "loss": 0.31205015182495116,
+ "num_input_tokens_seen": 843059320,
+ "step": 19280,
+ "train_runtime": 149180.2644,
+ "train_tokens_per_second": 5651.279
+ },
+ {
+ "epoch": 1.29847831941826,
+ "grad_norm": 0.7923357381559762,
+ "learning_rate": 2.7411743825572407e-06,
+ "loss": 0.3455479145050049,
+ "num_input_tokens_seen": 843291856,
+ "step": 19285,
+ "train_runtime": 149215.3246,
+ "train_tokens_per_second": 5651.51
+ },
+ {
+ "epoch": 1.2988149744142203,
+ "grad_norm": 0.7093696654952529,
+ "learning_rate": 2.7388158152097015e-06,
+ "loss": 0.30530009269714353,
+ "num_input_tokens_seen": 843521128,
+ "step": 19290,
+ "train_runtime": 149245.9801,
+ "train_tokens_per_second": 5651.885
+ },
+ {
+ "epoch": 1.2991516294101804,
+ "grad_norm": 0.7343455792050803,
+ "learning_rate": 2.736457880195109e-06,
+ "loss": 0.3188295841217041,
+ "num_input_tokens_seen": 843733872,
+ "step": 19295,
+ "train_runtime": 149283.513,
+ "train_tokens_per_second": 5651.889
+ },
+ {
+ "epoch": 1.2994882844061406,
+ "grad_norm": 0.7396876861043844,
+ "learning_rate": 2.7341005781728523e-06,
+ "loss": 0.3281911849975586,
+ "num_input_tokens_seen": 843964944,
+ "step": 19300,
+ "train_runtime": 149329.5165,
+ "train_tokens_per_second": 5651.695
+ },
+ {
+ "epoch": 1.2998249394021006,
+ "grad_norm": 0.752304677269571,
+ "learning_rate": 2.7317439098021444e-06,
+ "loss": 0.3072381496429443,
+ "num_input_tokens_seen": 844183152,
+ "step": 19305,
+ "train_runtime": 149373.377,
+ "train_tokens_per_second": 5651.497
+ },
+ {
+ "epoch": 1.300161594398061,
+ "grad_norm": 0.7866331506852371,
+ "learning_rate": 2.7293878757420185e-06,
+ "loss": 0.32221331596374514,
+ "num_input_tokens_seen": 844400152,
+ "step": 19310,
+ "train_runtime": 149415.5476,
+ "train_tokens_per_second": 5651.354
+ },
+ {
+ "epoch": 1.3004982493940211,
+ "grad_norm": 0.7241103171317756,
+ "learning_rate": 2.7270324766513306e-06,
+ "loss": 0.3104707717895508,
+ "num_input_tokens_seen": 844631680,
+ "step": 19315,
+ "train_runtime": 149454.6137,
+ "train_tokens_per_second": 5651.426
+ },
+ {
+ "epoch": 1.3008349043899812,
+ "grad_norm": 0.7716528370818122,
+ "learning_rate": 2.724677713188767e-06,
+ "loss": 0.3056204795837402,
+ "num_input_tokens_seen": 844836448,
+ "step": 19320,
+ "train_runtime": 149496.2843,
+ "train_tokens_per_second": 5651.22
+ },
+ {
+ "epoch": 1.3011715593859412,
+ "grad_norm": 0.7945005896939524,
+ "learning_rate": 2.722323586012821e-06,
+ "loss": 0.313628101348877,
+ "num_input_tokens_seen": 845053136,
+ "step": 19325,
+ "train_runtime": 149533.6826,
+ "train_tokens_per_second": 5651.256
+ },
+ {
+ "epoch": 1.3015082143819015,
+ "grad_norm": 0.7917610556705151,
+ "learning_rate": 2.7199700957818252e-06,
+ "loss": 0.3013652801513672,
+ "num_input_tokens_seen": 845276504,
+ "step": 19330,
+ "train_runtime": 149575.9646,
+ "train_tokens_per_second": 5651.152
+ },
+ {
+ "epoch": 1.3018448693778617,
+ "grad_norm": 0.8720139344385164,
+ "learning_rate": 2.717617243153918e-06,
+ "loss": 0.3019449710845947,
+ "num_input_tokens_seen": 845488208,
+ "step": 19335,
+ "train_runtime": 149622.3515,
+ "train_tokens_per_second": 5650.815
+ },
+ {
+ "epoch": 1.3021815243738217,
+ "grad_norm": 0.7397233088744621,
+ "learning_rate": 2.715265028787073e-06,
+ "loss": 0.31660406589508056,
+ "num_input_tokens_seen": 845695240,
+ "step": 19340,
+ "train_runtime": 149660.9196,
+ "train_tokens_per_second": 5650.742
+ },
+ {
+ "epoch": 1.3025181793697818,
+ "grad_norm": 0.8121569406562228,
+ "learning_rate": 2.7129134533390765e-06,
+ "loss": 0.33323040008544924,
+ "num_input_tokens_seen": 845925768,
+ "step": 19345,
+ "train_runtime": 149697.7788,
+ "train_tokens_per_second": 5650.891
+ },
+ {
+ "epoch": 1.302854834365742,
+ "grad_norm": 0.7696563717601685,
+ "learning_rate": 2.7105625174675404e-06,
+ "loss": 0.2952426433563232,
+ "num_input_tokens_seen": 846155368,
+ "step": 19350,
+ "train_runtime": 149736.5358,
+ "train_tokens_per_second": 5650.961
+ },
+ {
+ "epoch": 1.3031914893617023,
+ "grad_norm": 0.7453975160800163,
+ "learning_rate": 2.7082122218298957e-06,
+ "loss": 0.2969479560852051,
+ "num_input_tokens_seen": 846367080,
+ "step": 19355,
+ "train_runtime": 149773.8436,
+ "train_tokens_per_second": 5650.967
+ },
+ {
+ "epoch": 1.3035281443576623,
+ "grad_norm": 0.7913525098809976,
+ "learning_rate": 2.7058625670833947e-06,
+ "loss": 0.3046135902404785,
+ "num_input_tokens_seen": 846584096,
+ "step": 19360,
+ "train_runtime": 149810.105,
+ "train_tokens_per_second": 5651.048
+ },
+ {
+ "epoch": 1.3038647993536223,
+ "grad_norm": 0.7456322958807796,
+ "learning_rate": 2.70351355388511e-06,
+ "loss": 0.306339168548584,
+ "num_input_tokens_seen": 846800416,
+ "step": 19365,
+ "train_runtime": 149844.731,
+ "train_tokens_per_second": 5651.186
+ },
+ {
+ "epoch": 1.3042014543495826,
+ "grad_norm": 0.7705924730860086,
+ "learning_rate": 2.7011651828919407e-06,
+ "loss": 0.30881786346435547,
+ "num_input_tokens_seen": 847024880,
+ "step": 19370,
+ "train_runtime": 149886.5908,
+ "train_tokens_per_second": 5651.105
+ },
+ {
+ "epoch": 1.3045381093455428,
+ "grad_norm": 0.7634554372376511,
+ "learning_rate": 2.698817454760593e-06,
+ "loss": 0.305006217956543,
+ "num_input_tokens_seen": 847244184,
+ "step": 19375,
+ "train_runtime": 149919.4984,
+ "train_tokens_per_second": 5651.327
+ },
+ {
+ "epoch": 1.3048747643415028,
+ "grad_norm": 0.7420517594807788,
+ "learning_rate": 2.696470370147611e-06,
+ "loss": 0.3149790048599243,
+ "num_input_tokens_seen": 847461096,
+ "step": 19380,
+ "train_runtime": 149955.0762,
+ "train_tokens_per_second": 5651.433
+ },
+ {
+ "epoch": 1.3052114193374629,
+ "grad_norm": 0.7739344518736725,
+ "learning_rate": 2.69412392970934e-06,
+ "loss": 0.3325794219970703,
+ "num_input_tokens_seen": 847680648,
+ "step": 19385,
+ "train_runtime": 149993.1514,
+ "train_tokens_per_second": 5651.462
+ },
+ {
+ "epoch": 1.3055480743334231,
+ "grad_norm": 0.6530513267019178,
+ "learning_rate": 2.6917781341019615e-06,
+ "loss": 0.2991641044616699,
+ "num_input_tokens_seen": 847901040,
+ "step": 19390,
+ "train_runtime": 150031.3101,
+ "train_tokens_per_second": 5651.494
+ },
+ {
+ "epoch": 1.3058847293293834,
+ "grad_norm": 0.8035047874324559,
+ "learning_rate": 2.689432983981467e-06,
+ "loss": 0.2986927032470703,
+ "num_input_tokens_seen": 848128768,
+ "step": 19395,
+ "train_runtime": 150080.1123,
+ "train_tokens_per_second": 5651.174
+ },
+ {
+ "epoch": 1.3062213843253434,
+ "grad_norm": 0.7325503260453592,
+ "learning_rate": 2.6870884800036713e-06,
+ "loss": 0.3080791234970093,
+ "num_input_tokens_seen": 848340072,
+ "step": 19400,
+ "train_runtime": 150113.6518,
+ "train_tokens_per_second": 5651.319
+ },
+ {
+ "epoch": 1.3065580393213034,
+ "grad_norm": 0.948401336461252,
+ "learning_rate": 2.684744622824207e-06,
+ "loss": 0.32254228591918943,
+ "num_input_tokens_seen": 848548544,
+ "step": 19405,
+ "train_runtime": 150152.8491,
+ "train_tokens_per_second": 5651.232
+ },
+ {
+ "epoch": 1.3068946943172637,
+ "grad_norm": 0.7795604426950599,
+ "learning_rate": 2.6824014130985255e-06,
+ "loss": 0.31671667098999023,
+ "num_input_tokens_seen": 848765608,
+ "step": 19410,
+ "train_runtime": 150190.0674,
+ "train_tokens_per_second": 5651.277
+ },
+ {
+ "epoch": 1.307231349313224,
+ "grad_norm": 0.7275328657604889,
+ "learning_rate": 2.680058851481896e-06,
+ "loss": 0.2992823600769043,
+ "num_input_tokens_seen": 848984336,
+ "step": 19415,
+ "train_runtime": 150225.1677,
+ "train_tokens_per_second": 5651.412
+ },
+ {
+ "epoch": 1.307568004309184,
+ "grad_norm": 0.8231036715155717,
+ "learning_rate": 2.6777169386294154e-06,
+ "loss": 0.3124098777770996,
+ "num_input_tokens_seen": 849202320,
+ "step": 19420,
+ "train_runtime": 150263.4801,
+ "train_tokens_per_second": 5651.422
+ },
+ {
+ "epoch": 1.307904659305144,
+ "grad_norm": 0.784487047120117,
+ "learning_rate": 2.6753756751959846e-06,
+ "loss": 0.30565834045410156,
+ "num_input_tokens_seen": 849432016,
+ "step": 19425,
+ "train_runtime": 150304.88,
+ "train_tokens_per_second": 5651.393
+ },
+ {
+ "epoch": 1.3082413143011042,
+ "grad_norm": 0.7160689409622006,
+ "learning_rate": 2.6730350618363367e-06,
+ "loss": 0.3007196426391602,
+ "num_input_tokens_seen": 849652096,
+ "step": 19430,
+ "train_runtime": 150339.418,
+ "train_tokens_per_second": 5651.559
+ },
+ {
+ "epoch": 1.3085779692970645,
+ "grad_norm": 1.0693121673231818,
+ "learning_rate": 2.6706950992050097e-06,
+ "loss": 0.2976806640625,
+ "num_input_tokens_seen": 849873624,
+ "step": 19435,
+ "train_runtime": 150377.7065,
+ "train_tokens_per_second": 5651.593
+ },
+ {
+ "epoch": 1.3089146242930245,
+ "grad_norm": 0.7521055380512028,
+ "learning_rate": 2.6683557879563726e-06,
+ "loss": 0.3227806091308594,
+ "num_input_tokens_seen": 850094824,
+ "step": 19440,
+ "train_runtime": 150411.5882,
+ "train_tokens_per_second": 5651.791
+ },
+ {
+ "epoch": 1.3092512792889845,
+ "grad_norm": 0.7577610588393768,
+ "learning_rate": 2.6660171287446045e-06,
+ "loss": 0.33275818824768066,
+ "num_input_tokens_seen": 850318184,
+ "step": 19445,
+ "train_runtime": 150448.3342,
+ "train_tokens_per_second": 5651.895
+ },
+ {
+ "epoch": 1.3095879342849448,
+ "grad_norm": 0.7044234887744943,
+ "learning_rate": 2.6636791222237034e-06,
+ "loss": 0.3194304943084717,
+ "num_input_tokens_seen": 850537512,
+ "step": 19450,
+ "train_runtime": 150485.0966,
+ "train_tokens_per_second": 5651.972
+ },
+ {
+ "epoch": 1.309924589280905,
+ "grad_norm": 0.7351419738193509,
+ "learning_rate": 2.6613417690474864e-06,
+ "loss": 0.3084381580352783,
+ "num_input_tokens_seen": 850762520,
+ "step": 19455,
+ "train_runtime": 150516.0117,
+ "train_tokens_per_second": 5652.306
+ },
+ {
+ "epoch": 1.310261244276865,
+ "grad_norm": 0.6957251747847025,
+ "learning_rate": 2.6590050698695856e-06,
+ "loss": 0.30487637519836425,
+ "num_input_tokens_seen": 850980888,
+ "step": 19460,
+ "train_runtime": 150552.8661,
+ "train_tokens_per_second": 5652.373
+ },
+ {
+ "epoch": 1.310597899272825,
+ "grad_norm": 0.7124189387924086,
+ "learning_rate": 2.6566690253434508e-06,
+ "loss": 0.31617035865783694,
+ "num_input_tokens_seen": 851212384,
+ "step": 19465,
+ "train_runtime": 150591.0113,
+ "train_tokens_per_second": 5652.478
+ },
+ {
+ "epoch": 1.3109345542687854,
+ "grad_norm": 0.7885036457021949,
+ "learning_rate": 2.654333636122354e-06,
+ "loss": 0.29657797813415526,
+ "num_input_tokens_seen": 851438520,
+ "step": 19470,
+ "train_runtime": 150631.5054,
+ "train_tokens_per_second": 5652.46
+ },
+ {
+ "epoch": 1.3112712092647456,
+ "grad_norm": 0.7024735867575697,
+ "learning_rate": 2.6519989028593728e-06,
+ "loss": 0.3158660173416138,
+ "num_input_tokens_seen": 851666592,
+ "step": 19475,
+ "train_runtime": 150671.4222,
+ "train_tokens_per_second": 5652.476
+ },
+ {
+ "epoch": 1.3116078642607056,
+ "grad_norm": 0.8346584854469274,
+ "learning_rate": 2.649664826207414e-06,
+ "loss": 0.31950523853302004,
+ "num_input_tokens_seen": 851881728,
+ "step": 19480,
+ "train_runtime": 150710.0418,
+ "train_tokens_per_second": 5652.455
+ },
+ {
+ "epoch": 1.3119445192566657,
+ "grad_norm": 0.8221325321759024,
+ "learning_rate": 2.647331406819189e-06,
+ "loss": 0.3350534915924072,
+ "num_input_tokens_seen": 852092288,
+ "step": 19485,
+ "train_runtime": 150747.5697,
+ "train_tokens_per_second": 5652.445
+ },
+ {
+ "epoch": 1.312281174252626,
+ "grad_norm": 0.7024744149478919,
+ "learning_rate": 2.644998645347236e-06,
+ "loss": 0.29921770095825195,
+ "num_input_tokens_seen": 852306504,
+ "step": 19490,
+ "train_runtime": 150791.8993,
+ "train_tokens_per_second": 5652.204
+ },
+ {
+ "epoch": 1.3126178292485862,
+ "grad_norm": 0.7284467267787711,
+ "learning_rate": 2.6426665424439025e-06,
+ "loss": 0.316993260383606,
+ "num_input_tokens_seen": 852532944,
+ "step": 19495,
+ "train_runtime": 150836.5301,
+ "train_tokens_per_second": 5652.032
+ },
+ {
+ "epoch": 1.3129544842445462,
+ "grad_norm": 0.8271670284793109,
+ "learning_rate": 2.6403350987613526e-06,
+ "loss": 0.31356039047241213,
+ "num_input_tokens_seen": 852756160,
+ "step": 19500,
+ "train_runtime": 150879.3547,
+ "train_tokens_per_second": 5651.908
+ },
+ {
+ "epoch": 1.3132911392405062,
+ "grad_norm": 0.7670722817590104,
+ "learning_rate": 2.6380043149515676e-06,
+ "loss": 0.3076605796813965,
+ "num_input_tokens_seen": 852971624,
+ "step": 19505,
+ "train_runtime": 150913.3012,
+ "train_tokens_per_second": 5652.064
+ },
+ {
+ "epoch": 1.3136277942364665,
+ "grad_norm": 0.7538754898846585,
+ "learning_rate": 2.6356741916663453e-06,
+ "loss": 0.3078343868255615,
+ "num_input_tokens_seen": 853194624,
+ "step": 19510,
+ "train_runtime": 150952.8524,
+ "train_tokens_per_second": 5652.06
+ },
+ {
+ "epoch": 1.3139644492324267,
+ "grad_norm": 0.8299463216741857,
+ "learning_rate": 2.6333447295572933e-06,
+ "loss": 0.310660195350647,
+ "num_input_tokens_seen": 853414920,
+ "step": 19515,
+ "train_runtime": 150985.9986,
+ "train_tokens_per_second": 5652.279
+ },
+ {
+ "epoch": 1.3143011042283868,
+ "grad_norm": 0.7460970673432722,
+ "learning_rate": 2.631015929275845e-06,
+ "loss": 0.3059165239334106,
+ "num_input_tokens_seen": 853622712,
+ "step": 19520,
+ "train_runtime": 151026.6956,
+ "train_tokens_per_second": 5652.131
+ },
+ {
+ "epoch": 1.3146377592243468,
+ "grad_norm": 0.770739755897705,
+ "learning_rate": 2.628687791473234e-06,
+ "loss": 0.29071602821350095,
+ "num_input_tokens_seen": 853853296,
+ "step": 19525,
+ "train_runtime": 151057.0111,
+ "train_tokens_per_second": 5652.523
+ },
+ {
+ "epoch": 1.314974414220307,
+ "grad_norm": 0.8302523376916937,
+ "learning_rate": 2.6263603168005224e-06,
+ "loss": 0.2854055643081665,
+ "num_input_tokens_seen": 854070296,
+ "step": 19530,
+ "train_runtime": 151102.3064,
+ "train_tokens_per_second": 5652.265
+ },
+ {
+ "epoch": 1.3153110692162673,
+ "grad_norm": 0.6658939434607352,
+ "learning_rate": 2.6240335059085797e-06,
+ "loss": 0.30446481704711914,
+ "num_input_tokens_seen": 854298216,
+ "step": 19535,
+ "train_runtime": 151143.3911,
+ "train_tokens_per_second": 5652.237
+ },
+ {
+ "epoch": 1.3156477242122273,
+ "grad_norm": 0.7452286529155747,
+ "learning_rate": 2.62170735944809e-06,
+ "loss": 0.3120380163192749,
+ "num_input_tokens_seen": 854517360,
+ "step": 19540,
+ "train_runtime": 151180.803,
+ "train_tokens_per_second": 5652.287
+ },
+ {
+ "epoch": 1.3159843792081873,
+ "grad_norm": 0.7794163551487556,
+ "learning_rate": 2.6193818780695537e-06,
+ "loss": 0.298751163482666,
+ "num_input_tokens_seen": 854745560,
+ "step": 19545,
+ "train_runtime": 151214.1843,
+ "train_tokens_per_second": 5652.549
+ },
+ {
+ "epoch": 1.3163210342041476,
+ "grad_norm": 0.7606004553161916,
+ "learning_rate": 2.6170570624232837e-06,
+ "loss": 0.3143803119659424,
+ "num_input_tokens_seen": 854941592,
+ "step": 19550,
+ "train_runtime": 151257.5538,
+ "train_tokens_per_second": 5652.224
+ },
+ {
+ "epoch": 1.3166576892001078,
+ "grad_norm": 0.7540362996447706,
+ "learning_rate": 2.6147329131594078e-06,
+ "loss": 0.2849498510360718,
+ "num_input_tokens_seen": 855170576,
+ "step": 19555,
+ "train_runtime": 151298.4772,
+ "train_tokens_per_second": 5652.209
+ },
+ {
+ "epoch": 1.3169943441960679,
+ "grad_norm": 0.6379893471729663,
+ "learning_rate": 2.612409430927867e-06,
+ "loss": 0.2712059020996094,
+ "num_input_tokens_seen": 855390328,
+ "step": 19560,
+ "train_runtime": 151335.988,
+ "train_tokens_per_second": 5652.26
+ },
+ {
+ "epoch": 1.317330999192028,
+ "grad_norm": 1.705002769226515,
+ "learning_rate": 2.610086616378413e-06,
+ "loss": 0.29099502563476565,
+ "num_input_tokens_seen": 855599680,
+ "step": 19565,
+ "train_runtime": 151369.3355,
+ "train_tokens_per_second": 5652.398
+ },
+ {
+ "epoch": 1.3176676541879881,
+ "grad_norm": 0.6619066534695688,
+ "learning_rate": 2.60776447016062e-06,
+ "loss": 0.3147557497024536,
+ "num_input_tokens_seen": 855821216,
+ "step": 19570,
+ "train_runtime": 151406.8413,
+ "train_tokens_per_second": 5652.461
+ },
+ {
+ "epoch": 1.3180043091839484,
+ "grad_norm": 0.7847891473320998,
+ "learning_rate": 2.6054429929238607e-06,
+ "loss": 0.31384851932525637,
+ "num_input_tokens_seen": 856037064,
+ "step": 19575,
+ "train_runtime": 151440.0977,
+ "train_tokens_per_second": 5652.645
+ },
+ {
+ "epoch": 1.3183409641799084,
+ "grad_norm": 0.6462398801805597,
+ "learning_rate": 2.603122185317335e-06,
+ "loss": 0.29209444522857664,
+ "num_input_tokens_seen": 856240856,
+ "step": 19580,
+ "train_runtime": 151475.745,
+ "train_tokens_per_second": 5652.66
+ },
+ {
+ "epoch": 1.3186776191758685,
+ "grad_norm": 0.7747243284477985,
+ "learning_rate": 2.6008020479900465e-06,
+ "loss": 0.2708132743835449,
+ "num_input_tokens_seen": 856453280,
+ "step": 19585,
+ "train_runtime": 151515.0863,
+ "train_tokens_per_second": 5652.594
+ },
+ {
+ "epoch": 1.3190142741718287,
+ "grad_norm": 0.6709060817752821,
+ "learning_rate": 2.598482581590815e-06,
+ "loss": 0.30856051445007326,
+ "num_input_tokens_seen": 856685160,
+ "step": 19590,
+ "train_runtime": 151550.6826,
+ "train_tokens_per_second": 5652.796
+ },
+ {
+ "epoch": 1.319350929167789,
+ "grad_norm": 0.698540481486877,
+ "learning_rate": 2.5961637867682715e-06,
+ "loss": 0.280717658996582,
+ "num_input_tokens_seen": 856916440,
+ "step": 19595,
+ "train_runtime": 151591.3298,
+ "train_tokens_per_second": 5652.806
+ },
+ {
+ "epoch": 1.319687584163749,
+ "grad_norm": 0.6957573651733252,
+ "learning_rate": 2.5938456641708598e-06,
+ "loss": 0.28789021968841555,
+ "num_input_tokens_seen": 857134672,
+ "step": 19600,
+ "train_runtime": 151620.8055,
+ "train_tokens_per_second": 5653.147
+ },
+ {
+ "epoch": 1.320024239159709,
+ "grad_norm": 0.772599706020318,
+ "learning_rate": 2.591528214446836e-06,
+ "loss": 0.3085463523864746,
+ "num_input_tokens_seen": 857366576,
+ "step": 19605,
+ "train_runtime": 151656.731,
+ "train_tokens_per_second": 5653.337
+ },
+ {
+ "epoch": 1.3203608941556693,
+ "grad_norm": 0.7024918619799805,
+ "learning_rate": 2.5892114382442664e-06,
+ "loss": 0.3224550724029541,
+ "num_input_tokens_seen": 857583784,
+ "step": 19610,
+ "train_runtime": 151693.8363,
+ "train_tokens_per_second": 5653.386
+ },
+ {
+ "epoch": 1.3206975491516295,
+ "grad_norm": 0.632625020427823,
+ "learning_rate": 2.5868953362110287e-06,
+ "loss": 0.29941411018371583,
+ "num_input_tokens_seen": 857806240,
+ "step": 19615,
+ "train_runtime": 151729.7798,
+ "train_tokens_per_second": 5653.513
+ },
+ {
+ "epoch": 1.3210342041475895,
+ "grad_norm": 0.6986436139568517,
+ "learning_rate": 2.5845799089948196e-06,
+ "loss": 0.29595515727996824,
+ "num_input_tokens_seen": 858015600,
+ "step": 19620,
+ "train_runtime": 151767.663,
+ "train_tokens_per_second": 5653.481
+ },
+ {
+ "epoch": 1.3213708591435496,
+ "grad_norm": 0.7085148449641908,
+ "learning_rate": 2.5822651572431323e-06,
+ "loss": 0.30276174545288087,
+ "num_input_tokens_seen": 858241536,
+ "step": 19625,
+ "train_runtime": 151807.9703,
+ "train_tokens_per_second": 5653.468
+ },
+ {
+ "epoch": 1.3217075141395098,
+ "grad_norm": 0.7897947876767634,
+ "learning_rate": 2.5799510816032867e-06,
+ "loss": 0.3044028043746948,
+ "num_input_tokens_seen": 858431176,
+ "step": 19630,
+ "train_runtime": 151849.2654,
+ "train_tokens_per_second": 5653.18
+ },
+ {
+ "epoch": 1.32204416913547,
+ "grad_norm": 0.7421651364229699,
+ "learning_rate": 2.577637682722404e-06,
+ "loss": 0.31405105590820315,
+ "num_input_tokens_seen": 858649160,
+ "step": 19635,
+ "train_runtime": 151893.2046,
+ "train_tokens_per_second": 5652.979
+ },
+ {
+ "epoch": 1.32238082413143,
+ "grad_norm": 0.9243365365036531,
+ "learning_rate": 2.5753249612474195e-06,
+ "loss": 0.3185321807861328,
+ "num_input_tokens_seen": 858864688,
+ "step": 19640,
+ "train_runtime": 151935.4081,
+ "train_tokens_per_second": 5652.828
+ },
+ {
+ "epoch": 1.3227174791273901,
+ "grad_norm": 0.714378409765336,
+ "learning_rate": 2.5730129178250774e-06,
+ "loss": 0.3150054693222046,
+ "num_input_tokens_seen": 859090984,
+ "step": 19645,
+ "train_runtime": 151973.3637,
+ "train_tokens_per_second": 5652.905
+ },
+ {
+ "epoch": 1.3230541341233504,
+ "grad_norm": 0.7478713199114753,
+ "learning_rate": 2.570701553101933e-06,
+ "loss": 0.3087807893753052,
+ "num_input_tokens_seen": 859303816,
+ "step": 19650,
+ "train_runtime": 152016.815,
+ "train_tokens_per_second": 5652.689
+ },
+ {
+ "epoch": 1.3233907891193106,
+ "grad_norm": 0.7180167108657562,
+ "learning_rate": 2.568390867724354e-06,
+ "loss": 0.3278617858886719,
+ "num_input_tokens_seen": 859526840,
+ "step": 19655,
+ "train_runtime": 152059.4571,
+ "train_tokens_per_second": 5652.571
+ },
+ {
+ "epoch": 1.3237274441152707,
+ "grad_norm": 0.8472549748110341,
+ "learning_rate": 2.5660808623385143e-06,
+ "loss": 0.3053846836090088,
+ "num_input_tokens_seen": 859736800,
+ "step": 19660,
+ "train_runtime": 152097.7055,
+ "train_tokens_per_second": 5652.53
+ },
+ {
+ "epoch": 1.3240640991112307,
+ "grad_norm": 0.6854888856572094,
+ "learning_rate": 2.5637715375903984e-06,
+ "loss": 0.2824453830718994,
+ "num_input_tokens_seen": 859955360,
+ "step": 19665,
+ "train_runtime": 152139.8395,
+ "train_tokens_per_second": 5652.401
+ },
+ {
+ "epoch": 1.324400754107191,
+ "grad_norm": 0.8425691445759216,
+ "learning_rate": 2.5614628941258075e-06,
+ "loss": 0.3014561176300049,
+ "num_input_tokens_seen": 860163408,
+ "step": 19670,
+ "train_runtime": 152179.5969,
+ "train_tokens_per_second": 5652.291
+ },
+ {
+ "epoch": 1.3247374091031512,
+ "grad_norm": 0.785890365158059,
+ "learning_rate": 2.5591549325903388e-06,
+ "loss": 0.3199615478515625,
+ "num_input_tokens_seen": 860403016,
+ "step": 19675,
+ "train_runtime": 152220.8565,
+ "train_tokens_per_second": 5652.333
+ },
+ {
+ "epoch": 1.3250740640991112,
+ "grad_norm": 0.7049194900739367,
+ "learning_rate": 2.5568476536294105e-06,
+ "loss": 0.3173452615737915,
+ "num_input_tokens_seen": 860626712,
+ "step": 19680,
+ "train_runtime": 152253.5963,
+ "train_tokens_per_second": 5652.587
+ },
+ {
+ "epoch": 1.3254107190950712,
+ "grad_norm": 0.7458641790263951,
+ "learning_rate": 2.5545410578882466e-06,
+ "loss": 0.2950143814086914,
+ "num_input_tokens_seen": 860857936,
+ "step": 19685,
+ "train_runtime": 152289.1953,
+ "train_tokens_per_second": 5652.784
+ },
+ {
+ "epoch": 1.3257473740910315,
+ "grad_norm": 0.8429589187979019,
+ "learning_rate": 2.5522351460118776e-06,
+ "loss": 0.2903186321258545,
+ "num_input_tokens_seen": 861063800,
+ "step": 19690,
+ "train_runtime": 152319.4076,
+ "train_tokens_per_second": 5653.014
+ },
+ {
+ "epoch": 1.3260840290869917,
+ "grad_norm": 0.6773218912537953,
+ "learning_rate": 2.549929918645144e-06,
+ "loss": 0.31392974853515626,
+ "num_input_tokens_seen": 861297432,
+ "step": 19695,
+ "train_runtime": 152355.3898,
+ "train_tokens_per_second": 5653.213
+ },
+ {
+ "epoch": 1.3264206840829518,
+ "grad_norm": 0.7356765211271618,
+ "learning_rate": 2.5476253764326963e-06,
+ "loss": 0.2875689506530762,
+ "num_input_tokens_seen": 861503656,
+ "step": 19700,
+ "train_runtime": 152391.6663,
+ "train_tokens_per_second": 5653.22
+ },
+ {
+ "epoch": 1.3267573390789118,
+ "grad_norm": 0.7016223117638092,
+ "learning_rate": 2.5453215200189927e-06,
+ "loss": 0.2943393707275391,
+ "num_input_tokens_seen": 861717016,
+ "step": 19705,
+ "train_runtime": 152429.9059,
+ "train_tokens_per_second": 5653.202
+ },
+ {
+ "epoch": 1.327093994074872,
+ "grad_norm": 0.7845624304921898,
+ "learning_rate": 2.543018350048298e-06,
+ "loss": 0.29305644035339357,
+ "num_input_tokens_seen": 861937640,
+ "step": 19710,
+ "train_runtime": 152468.0419,
+ "train_tokens_per_second": 5653.235
+ },
+ {
+ "epoch": 1.3274306490708323,
+ "grad_norm": 0.7711375220547478,
+ "learning_rate": 2.540715867164685e-06,
+ "loss": 0.3149756908416748,
+ "num_input_tokens_seen": 862156848,
+ "step": 19715,
+ "train_runtime": 152506.2922,
+ "train_tokens_per_second": 5653.254
+ },
+ {
+ "epoch": 1.3277673040667923,
+ "grad_norm": 0.8404428785990262,
+ "learning_rate": 2.538414072012042e-06,
+ "loss": 0.30890326499938964,
+ "num_input_tokens_seen": 862362672,
+ "step": 19720,
+ "train_runtime": 152547.9973,
+ "train_tokens_per_second": 5653.058
+ },
+ {
+ "epoch": 1.3281039590627524,
+ "grad_norm": 0.8169832466518516,
+ "learning_rate": 2.53611296523405e-06,
+ "loss": 0.3164846420288086,
+ "num_input_tokens_seen": 862573904,
+ "step": 19725,
+ "train_runtime": 152580.4337,
+ "train_tokens_per_second": 5653.241
+ },
+ {
+ "epoch": 1.3284406140587126,
+ "grad_norm": 0.793059236401201,
+ "learning_rate": 2.5338125474742127e-06,
+ "loss": 0.2981910228729248,
+ "num_input_tokens_seen": 862787064,
+ "step": 19730,
+ "train_runtime": 152618.0709,
+ "train_tokens_per_second": 5653.243
+ },
+ {
+ "epoch": 1.3287772690546729,
+ "grad_norm": 0.6549096029319272,
+ "learning_rate": 2.531512819375832e-06,
+ "loss": 0.2908968448638916,
+ "num_input_tokens_seen": 863007040,
+ "step": 19735,
+ "train_runtime": 152662.0691,
+ "train_tokens_per_second": 5653.055
+ },
+ {
+ "epoch": 1.3291139240506329,
+ "grad_norm": 0.7643581112810446,
+ "learning_rate": 2.52921378158202e-06,
+ "loss": 0.2935811996459961,
+ "num_input_tokens_seen": 863219112,
+ "step": 19740,
+ "train_runtime": 152697.5407,
+ "train_tokens_per_second": 5653.13
+ },
+ {
+ "epoch": 1.329450579046593,
+ "grad_norm": 0.8140047984203814,
+ "learning_rate": 2.5269154347356957e-06,
+ "loss": 0.2813416957855225,
+ "num_input_tokens_seen": 863441536,
+ "step": 19745,
+ "train_runtime": 152736.5725,
+ "train_tokens_per_second": 5653.142
+ },
+ {
+ "epoch": 1.3297872340425532,
+ "grad_norm": 0.8712638461066203,
+ "learning_rate": 2.5246177794795834e-06,
+ "loss": 0.30994884967803954,
+ "num_input_tokens_seen": 863641736,
+ "step": 19750,
+ "train_runtime": 152783.7657,
+ "train_tokens_per_second": 5652.706
+ },
+ {
+ "epoch": 1.3301238890385134,
+ "grad_norm": 0.6601157103213744,
+ "learning_rate": 2.5223208164562163e-06,
+ "loss": 0.30829753875732424,
+ "num_input_tokens_seen": 863857152,
+ "step": 19755,
+ "train_runtime": 152823.124,
+ "train_tokens_per_second": 5652.66
+ },
+ {
+ "epoch": 1.3304605440344734,
+ "grad_norm": 0.7196822514488103,
+ "learning_rate": 2.5200245463079315e-06,
+ "loss": 0.31928205490112305,
+ "num_input_tokens_seen": 864070800,
+ "step": 19760,
+ "train_runtime": 152859.6114,
+ "train_tokens_per_second": 5652.708
+ },
+ {
+ "epoch": 1.3307971990304335,
+ "grad_norm": 0.7776014244495211,
+ "learning_rate": 2.5177289696768725e-06,
+ "loss": 0.32152862548828126,
+ "num_input_tokens_seen": 864291928,
+ "step": 19765,
+ "train_runtime": 152893.0449,
+ "train_tokens_per_second": 5652.919
+ },
+ {
+ "epoch": 1.3311338540263937,
+ "grad_norm": 0.7457812155053926,
+ "learning_rate": 2.515434087204996e-06,
+ "loss": 0.29552688598632815,
+ "num_input_tokens_seen": 864526144,
+ "step": 19770,
+ "train_runtime": 152933.1454,
+ "train_tokens_per_second": 5652.968
+ },
+ {
+ "epoch": 1.331470509022354,
+ "grad_norm": 0.628218974067636,
+ "learning_rate": 2.51313989953405e-06,
+ "loss": 0.3004995107650757,
+ "num_input_tokens_seen": 864756000,
+ "step": 19775,
+ "train_runtime": 152967.0366,
+ "train_tokens_per_second": 5653.218
+ },
+ {
+ "epoch": 1.331807164018314,
+ "grad_norm": 0.7135273557700192,
+ "learning_rate": 2.510846407305603e-06,
+ "loss": 0.298743748664856,
+ "num_input_tokens_seen": 864976528,
+ "step": 19780,
+ "train_runtime": 153002.9046,
+ "train_tokens_per_second": 5653.334
+ },
+ {
+ "epoch": 1.332143819014274,
+ "grad_norm": 0.678096834485325,
+ "learning_rate": 2.508553611161021e-06,
+ "loss": 0.3048709392547607,
+ "num_input_tokens_seen": 865197968,
+ "step": 19785,
+ "train_runtime": 153045.6277,
+ "train_tokens_per_second": 5653.203
+ },
+ {
+ "epoch": 1.3324804740102343,
+ "grad_norm": 0.7272471213502572,
+ "learning_rate": 2.5062615117414773e-06,
+ "loss": 0.3029000759124756,
+ "num_input_tokens_seen": 865431584,
+ "step": 19790,
+ "train_runtime": 153089.6551,
+ "train_tokens_per_second": 5653.103
+ },
+ {
+ "epoch": 1.3328171290061945,
+ "grad_norm": 0.7477356879857912,
+ "learning_rate": 2.50397010968795e-06,
+ "loss": 0.2859670400619507,
+ "num_input_tokens_seen": 865646552,
+ "step": 19795,
+ "train_runtime": 153133.4962,
+ "train_tokens_per_second": 5652.888
+ },
+ {
+ "epoch": 1.3331537840021546,
+ "grad_norm": 0.7549750199167686,
+ "learning_rate": 2.501679405641222e-06,
+ "loss": 0.32290475368499755,
+ "num_input_tokens_seen": 865868968,
+ "step": 19800,
+ "train_runtime": 153168.7301,
+ "train_tokens_per_second": 5653.04
+ },
+ {
+ "epoch": 1.3334904389981148,
+ "grad_norm": 0.7654205877160946,
+ "learning_rate": 2.499389400241881e-06,
+ "loss": 0.30891990661621094,
+ "num_input_tokens_seen": 866072568,
+ "step": 19805,
+ "train_runtime": 153205.9089,
+ "train_tokens_per_second": 5652.997
+ },
+ {
+ "epoch": 1.3338270939940748,
+ "grad_norm": 0.6968224775968341,
+ "learning_rate": 2.497100094130321e-06,
+ "loss": 0.27096853256225584,
+ "num_input_tokens_seen": 866291048,
+ "step": 19810,
+ "train_runtime": 153242.9077,
+ "train_tokens_per_second": 5653.058
+ },
+ {
+ "epoch": 1.334163748990035,
+ "grad_norm": 0.6748296546699505,
+ "learning_rate": 2.494811487946737e-06,
+ "loss": 0.2911532402038574,
+ "num_input_tokens_seen": 866515272,
+ "step": 19815,
+ "train_runtime": 153277.1916,
+ "train_tokens_per_second": 5653.256
+ },
+ {
+ "epoch": 1.3345004039859951,
+ "grad_norm": 0.7534075118050922,
+ "learning_rate": 2.492523582331135e-06,
+ "loss": 0.3089088201522827,
+ "num_input_tokens_seen": 866738320,
+ "step": 19820,
+ "train_runtime": 153318.6584,
+ "train_tokens_per_second": 5653.182
+ },
+ {
+ "epoch": 1.3348370589819554,
+ "grad_norm": 0.7981812627641985,
+ "learning_rate": 2.4902363779233134e-06,
+ "loss": 0.29610767364501955,
+ "num_input_tokens_seen": 866959032,
+ "step": 19825,
+ "train_runtime": 153358.9779,
+ "train_tokens_per_second": 5653.135
+ },
+ {
+ "epoch": 1.3351737139779154,
+ "grad_norm": 0.7457013614340272,
+ "learning_rate": 2.4879498753628885e-06,
+ "loss": 0.31332998275756835,
+ "num_input_tokens_seen": 867155816,
+ "step": 19830,
+ "train_runtime": 153400.6658,
+ "train_tokens_per_second": 5652.882
+ },
+ {
+ "epoch": 1.3355103689738757,
+ "grad_norm": 0.6543399547857148,
+ "learning_rate": 2.4856640752892702e-06,
+ "loss": 0.27421958446502687,
+ "num_input_tokens_seen": 867382880,
+ "step": 19835,
+ "train_runtime": 153438.0486,
+ "train_tokens_per_second": 5652.984
+ },
+ {
+ "epoch": 1.3358470239698357,
+ "grad_norm": 0.7540236121688988,
+ "learning_rate": 2.4833789783416757e-06,
+ "loss": 0.28227849006652833,
+ "num_input_tokens_seen": 867604504,
+ "step": 19840,
+ "train_runtime": 153478.0421,
+ "train_tokens_per_second": 5652.955
+ },
+ {
+ "epoch": 1.336183678965796,
+ "grad_norm": 0.7797596388766936,
+ "learning_rate": 2.4810945851591245e-06,
+ "loss": 0.3176878452301025,
+ "num_input_tokens_seen": 867810880,
+ "step": 19845,
+ "train_runtime": 153514.5274,
+ "train_tokens_per_second": 5652.956
+ },
+ {
+ "epoch": 1.336520333961756,
+ "grad_norm": 0.7762325041420014,
+ "learning_rate": 2.4788108963804406e-06,
+ "loss": 0.2944268941879272,
+ "num_input_tokens_seen": 868031240,
+ "step": 19850,
+ "train_runtime": 153555.0069,
+ "train_tokens_per_second": 5652.901
+ },
+ {
+ "epoch": 1.3368569889577162,
+ "grad_norm": 0.7515387143609674,
+ "learning_rate": 2.4765279126442502e-06,
+ "loss": 0.30596926212310793,
+ "num_input_tokens_seen": 868253728,
+ "step": 19855,
+ "train_runtime": 153594.4952,
+ "train_tokens_per_second": 5652.896
+ },
+ {
+ "epoch": 1.3371936439536762,
+ "grad_norm": 0.7609490688259355,
+ "learning_rate": 2.4742456345889828e-06,
+ "loss": 0.3069468975067139,
+ "num_input_tokens_seen": 868466312,
+ "step": 19860,
+ "train_runtime": 153633.742,
+ "train_tokens_per_second": 5652.836
+ },
+ {
+ "epoch": 1.3375302989496365,
+ "grad_norm": 0.7746964460091734,
+ "learning_rate": 2.471964062852867e-06,
+ "loss": 0.30750603675842286,
+ "num_input_tokens_seen": 868689096,
+ "step": 19865,
+ "train_runtime": 153665.3062,
+ "train_tokens_per_second": 5653.124
+ },
+ {
+ "epoch": 1.3378669539455965,
+ "grad_norm": 0.6717681437842503,
+ "learning_rate": 2.469683198073943e-06,
+ "loss": 0.29093983173370364,
+ "num_input_tokens_seen": 868917088,
+ "step": 19870,
+ "train_runtime": 153696.1235,
+ "train_tokens_per_second": 5653.474
+ },
+ {
+ "epoch": 1.3382036089415568,
+ "grad_norm": 0.9244971353347443,
+ "learning_rate": 2.4674030408900414e-06,
+ "loss": 0.2891664505004883,
+ "num_input_tokens_seen": 869105376,
+ "step": 19875,
+ "train_runtime": 153735.3162,
+ "train_tokens_per_second": 5653.258
+ },
+ {
+ "epoch": 1.3385402639375168,
+ "grad_norm": 0.8136223037135817,
+ "learning_rate": 2.465123591938805e-06,
+ "loss": 0.3123000144958496,
+ "num_input_tokens_seen": 869319584,
+ "step": 19880,
+ "train_runtime": 153771.8119,
+ "train_tokens_per_second": 5653.309
+ },
+ {
+ "epoch": 1.338876918933477,
+ "grad_norm": 0.73409884799906,
+ "learning_rate": 2.462844851857673e-06,
+ "loss": 0.31596615314483645,
+ "num_input_tokens_seen": 869526576,
+ "step": 19885,
+ "train_runtime": 153815.0765,
+ "train_tokens_per_second": 5653.065
+ },
+ {
+ "epoch": 1.339213573929437,
+ "grad_norm": 0.7895572970153802,
+ "learning_rate": 2.4605668212838878e-06,
+ "loss": 0.31440134048461915,
+ "num_input_tokens_seen": 869742096,
+ "step": 19890,
+ "train_runtime": 153855.5507,
+ "train_tokens_per_second": 5652.978
+ },
+ {
+ "epoch": 1.3395502289253973,
+ "grad_norm": 0.786706111069309,
+ "learning_rate": 2.458289500854493e-06,
+ "loss": 0.34726858139038086,
+ "num_input_tokens_seen": 869961536,
+ "step": 19895,
+ "train_runtime": 153890.7632,
+ "train_tokens_per_second": 5653.111
+ },
+ {
+ "epoch": 1.3398868839213574,
+ "grad_norm": 0.7370468195124786,
+ "learning_rate": 2.4560128912063353e-06,
+ "loss": 0.3165955305099487,
+ "num_input_tokens_seen": 870165224,
+ "step": 19900,
+ "train_runtime": 153932.5225,
+ "train_tokens_per_second": 5652.9
+ },
+ {
+ "epoch": 1.3402235389173176,
+ "grad_norm": 0.7650689001742789,
+ "learning_rate": 2.4537369929760596e-06,
+ "loss": 0.34606313705444336,
+ "num_input_tokens_seen": 870396672,
+ "step": 19905,
+ "train_runtime": 153969.9255,
+ "train_tokens_per_second": 5653.03
+ },
+ {
+ "epoch": 1.3405601939132776,
+ "grad_norm": 0.7725397454862113,
+ "learning_rate": 2.4514618068001144e-06,
+ "loss": 0.305745005607605,
+ "num_input_tokens_seen": 870610560,
+ "step": 19910,
+ "train_runtime": 154008.56,
+ "train_tokens_per_second": 5653.001
+ },
+ {
+ "epoch": 1.3408968489092379,
+ "grad_norm": 0.7591073107632155,
+ "learning_rate": 2.4491873333147463e-06,
+ "loss": 0.29233665466308595,
+ "num_input_tokens_seen": 870825304,
+ "step": 19915,
+ "train_runtime": 154050.5577,
+ "train_tokens_per_second": 5652.854
+ },
+ {
+ "epoch": 1.341233503905198,
+ "grad_norm": 0.8155190700964781,
+ "learning_rate": 2.4469135731560106e-06,
+ "loss": 0.2776491641998291,
+ "num_input_tokens_seen": 871033624,
+ "step": 19920,
+ "train_runtime": 154088.9114,
+ "train_tokens_per_second": 5652.799
+ },
+ {
+ "epoch": 1.3415701589011582,
+ "grad_norm": 0.7835805613524144,
+ "learning_rate": 2.4446405269597494e-06,
+ "loss": 0.31336369514465334,
+ "num_input_tokens_seen": 871261480,
+ "step": 19925,
+ "train_runtime": 154138.0941,
+ "train_tokens_per_second": 5652.473
+ },
+ {
+ "epoch": 1.3419068138971182,
+ "grad_norm": 0.7131677555675145,
+ "learning_rate": 2.442368195361618e-06,
+ "loss": 0.27323174476623535,
+ "num_input_tokens_seen": 871484784,
+ "step": 19930,
+ "train_runtime": 154175.5255,
+ "train_tokens_per_second": 5652.549
+ },
+ {
+ "epoch": 1.3422434688930784,
+ "grad_norm": 0.7614586629617044,
+ "learning_rate": 2.4400965789970647e-06,
+ "loss": 0.2858338594436646,
+ "num_input_tokens_seen": 871694432,
+ "step": 19935,
+ "train_runtime": 154214.3119,
+ "train_tokens_per_second": 5652.487
+ },
+ {
+ "epoch": 1.3425801238890385,
+ "grad_norm": 0.7849206092021981,
+ "learning_rate": 2.437825678501339e-06,
+ "loss": 0.3141868352890015,
+ "num_input_tokens_seen": 871916776,
+ "step": 19940,
+ "train_runtime": 154256.1387,
+ "train_tokens_per_second": 5652.396
+ },
+ {
+ "epoch": 1.3429167788849987,
+ "grad_norm": 0.6972967949553635,
+ "learning_rate": 2.4355554945094956e-06,
+ "loss": 0.3040180683135986,
+ "num_input_tokens_seen": 872136592,
+ "step": 19945,
+ "train_runtime": 154298.8906,
+ "train_tokens_per_second": 5652.254
+ },
+ {
+ "epoch": 1.3432534338809587,
+ "grad_norm": 0.7860171584750419,
+ "learning_rate": 2.4332860276563763e-06,
+ "loss": 0.2974998950958252,
+ "num_input_tokens_seen": 872357128,
+ "step": 19950,
+ "train_runtime": 154335.8822,
+ "train_tokens_per_second": 5652.329
+ },
+ {
+ "epoch": 1.343590088876919,
+ "grad_norm": 0.678809783107802,
+ "learning_rate": 2.4310172785766396e-06,
+ "loss": 0.30943219661712645,
+ "num_input_tokens_seen": 872573624,
+ "step": 19955,
+ "train_runtime": 154378.2017,
+ "train_tokens_per_second": 5652.182
+ },
+ {
+ "epoch": 1.343926743872879,
+ "grad_norm": 0.8435675441786462,
+ "learning_rate": 2.428749247904724e-06,
+ "loss": 0.31881489753723147,
+ "num_input_tokens_seen": 872789824,
+ "step": 19960,
+ "train_runtime": 154415.7323,
+ "train_tokens_per_second": 5652.208
+ },
+ {
+ "epoch": 1.3442633988688393,
+ "grad_norm": 0.7510529942793552,
+ "learning_rate": 2.4264819362748856e-06,
+ "loss": 0.3125936031341553,
+ "num_input_tokens_seen": 873009864,
+ "step": 19965,
+ "train_runtime": 154452.1402,
+ "train_tokens_per_second": 5652.3
+ },
+ {
+ "epoch": 1.3446000538647993,
+ "grad_norm": 0.8178579575509619,
+ "learning_rate": 2.4242153443211663e-06,
+ "loss": 0.3280599594116211,
+ "num_input_tokens_seen": 873230264,
+ "step": 19970,
+ "train_runtime": 154483.7052,
+ "train_tokens_per_second": 5652.572
+ },
+ {
+ "epoch": 1.3449367088607596,
+ "grad_norm": 0.7189596121215553,
+ "learning_rate": 2.4219494726774124e-06,
+ "loss": 0.2943082809448242,
+ "num_input_tokens_seen": 873453312,
+ "step": 19975,
+ "train_runtime": 154519.1177,
+ "train_tokens_per_second": 5652.72
+ },
+ {
+ "epoch": 1.3452733638567196,
+ "grad_norm": 0.7355041064223901,
+ "learning_rate": 2.419684321977268e-06,
+ "loss": 0.2926133632659912,
+ "num_input_tokens_seen": 873660984,
+ "step": 19980,
+ "train_runtime": 154553.7989,
+ "train_tokens_per_second": 5652.795
+ },
+ {
+ "epoch": 1.3456100188526798,
+ "grad_norm": 0.6865570643674106,
+ "learning_rate": 2.417419892854175e-06,
+ "loss": 0.29560866355896,
+ "num_input_tokens_seen": 873892712,
+ "step": 19985,
+ "train_runtime": 154598.2899,
+ "train_tokens_per_second": 5652.667
+ },
+ {
+ "epoch": 1.3459466738486399,
+ "grad_norm": 0.885274027016748,
+ "learning_rate": 2.415156185941371e-06,
+ "loss": 0.29713964462280273,
+ "num_input_tokens_seen": 874102984,
+ "step": 19990,
+ "train_runtime": 154633.4222,
+ "train_tokens_per_second": 5652.743
+ },
+ {
+ "epoch": 1.3462833288446001,
+ "grad_norm": 0.7645675693693532,
+ "learning_rate": 2.412893201871902e-06,
+ "loss": 0.30781164169311526,
+ "num_input_tokens_seen": 874317048,
+ "step": 19995,
+ "train_runtime": 154674.4314,
+ "train_tokens_per_second": 5652.628
+ },
+ {
+ "epoch": 1.3466199838405601,
+ "grad_norm": 0.7929208093171044,
+ "learning_rate": 2.410630941278595e-06,
+ "loss": 0.31746225357055663,
+ "num_input_tokens_seen": 874537536,
+ "step": 20000,
+ "train_runtime": 154717.3345,
+ "train_tokens_per_second": 5652.486
+ },
+ {
+ "epoch": 1.3469566388365204,
+ "grad_norm": 0.8665886750624643,
+ "learning_rate": 2.408369404794092e-06,
+ "loss": 0.3095715045928955,
+ "num_input_tokens_seen": 874752264,
+ "step": 20005,
+ "train_runtime": 154752.6238,
+ "train_tokens_per_second": 5652.584
+ },
+ {
+ "epoch": 1.3472932938324804,
+ "grad_norm": 0.7516875018971654,
+ "learning_rate": 2.4061085930508176e-06,
+ "loss": 0.30551714897155763,
+ "num_input_tokens_seen": 874964968,
+ "step": 20010,
+ "train_runtime": 154789.9589,
+ "train_tokens_per_second": 5652.595
+ },
+ {
+ "epoch": 1.3476299488284407,
+ "grad_norm": 0.8041861814830855,
+ "learning_rate": 2.4038485066810058e-06,
+ "loss": 0.3143286228179932,
+ "num_input_tokens_seen": 875176368,
+ "step": 20015,
+ "train_runtime": 154836.4632,
+ "train_tokens_per_second": 5652.263
+ },
+ {
+ "epoch": 1.3479666038244007,
+ "grad_norm": 0.7821727058895627,
+ "learning_rate": 2.40158914631668e-06,
+ "loss": 0.276894736289978,
+ "num_input_tokens_seen": 875385288,
+ "step": 20020,
+ "train_runtime": 154870.1836,
+ "train_tokens_per_second": 5652.381
+ },
+ {
+ "epoch": 1.348303258820361,
+ "grad_norm": 0.723635333292079,
+ "learning_rate": 2.399330512589664e-06,
+ "loss": 0.29750776290893555,
+ "num_input_tokens_seen": 875609880,
+ "step": 20025,
+ "train_runtime": 154911.8697,
+ "train_tokens_per_second": 5652.31
+ },
+ {
+ "epoch": 1.348639913816321,
+ "grad_norm": 0.7243913395939594,
+ "learning_rate": 2.3970726061315775e-06,
+ "loss": 0.30971436500549315,
+ "num_input_tokens_seen": 875831888,
+ "step": 20030,
+ "train_runtime": 154948.8908,
+ "train_tokens_per_second": 5652.392
+ },
+ {
+ "epoch": 1.3489765688122812,
+ "grad_norm": 0.7399215488091135,
+ "learning_rate": 2.394815427573836e-06,
+ "loss": 0.30228476524353026,
+ "num_input_tokens_seen": 876061376,
+ "step": 20035,
+ "train_runtime": 154992.8431,
+ "train_tokens_per_second": 5652.27
+ },
+ {
+ "epoch": 1.3493132238082413,
+ "grad_norm": 0.7012240220598052,
+ "learning_rate": 2.3925589775476516e-06,
+ "loss": 0.2862797021865845,
+ "num_input_tokens_seen": 876272016,
+ "step": 20040,
+ "train_runtime": 155032.043,
+ "train_tokens_per_second": 5652.199
+ },
+ {
+ "epoch": 1.3496498788042015,
+ "grad_norm": 0.7194801409568423,
+ "learning_rate": 2.3903032566840377e-06,
+ "loss": 0.31173529624938967,
+ "num_input_tokens_seen": 876494336,
+ "step": 20045,
+ "train_runtime": 155073.9927,
+ "train_tokens_per_second": 5652.104
+ },
+ {
+ "epoch": 1.3499865338001615,
+ "grad_norm": 0.7007455859084586,
+ "learning_rate": 2.3880482656137926e-06,
+ "loss": 0.31239540576934816,
+ "num_input_tokens_seen": 876727136,
+ "step": 20050,
+ "train_runtime": 155114.2632,
+ "train_tokens_per_second": 5652.137
+ },
+ {
+ "epoch": 1.3503231887961218,
+ "grad_norm": 0.7248173637285731,
+ "learning_rate": 2.3857940049675254e-06,
+ "loss": 0.3017667055130005,
+ "num_input_tokens_seen": 876951616,
+ "step": 20055,
+ "train_runtime": 155154.1439,
+ "train_tokens_per_second": 5652.131
+ },
+ {
+ "epoch": 1.3506598437920818,
+ "grad_norm": 0.7421571706340755,
+ "learning_rate": 2.383540475375625e-06,
+ "loss": 0.2906331062316895,
+ "num_input_tokens_seen": 877172960,
+ "step": 20060,
+ "train_runtime": 155195.6002,
+ "train_tokens_per_second": 5652.048
+ },
+ {
+ "epoch": 1.350996498788042,
+ "grad_norm": 0.7848990495009128,
+ "learning_rate": 2.3812876774682886e-06,
+ "loss": 0.30210440158843993,
+ "num_input_tokens_seen": 877405296,
+ "step": 20065,
+ "train_runtime": 155229.0295,
+ "train_tokens_per_second": 5652.327
+ },
+ {
+ "epoch": 1.351333153784002,
+ "grad_norm": 0.7518085342917976,
+ "learning_rate": 2.379035611875503e-06,
+ "loss": 0.3308712959289551,
+ "num_input_tokens_seen": 877636392,
+ "step": 20070,
+ "train_runtime": 155261.3122,
+ "train_tokens_per_second": 5652.641
+ },
+ {
+ "epoch": 1.3516698087799623,
+ "grad_norm": 0.7278323335362945,
+ "learning_rate": 2.376784279227051e-06,
+ "loss": 0.2974379062652588,
+ "num_input_tokens_seen": 877871256,
+ "step": 20075,
+ "train_runtime": 155301.2047,
+ "train_tokens_per_second": 5652.701
+ },
+ {
+ "epoch": 1.3520064637759224,
+ "grad_norm": 0.7168837215922106,
+ "learning_rate": 2.3745336801525105e-06,
+ "loss": 0.3024725437164307,
+ "num_input_tokens_seen": 878104688,
+ "step": 20080,
+ "train_runtime": 155345.4421,
+ "train_tokens_per_second": 5652.594
+ },
+ {
+ "epoch": 1.3523431187718826,
+ "grad_norm": 0.7794337072066637,
+ "learning_rate": 2.3722838152812543e-06,
+ "loss": 0.3230156898498535,
+ "num_input_tokens_seen": 878325104,
+ "step": 20085,
+ "train_runtime": 155391.9834,
+ "train_tokens_per_second": 5652.319
+ },
+ {
+ "epoch": 1.3526797737678427,
+ "grad_norm": 0.607568547988963,
+ "learning_rate": 2.3700346852424485e-06,
+ "loss": 0.2935077428817749,
+ "num_input_tokens_seen": 878541320,
+ "step": 20090,
+ "train_runtime": 155423.8708,
+ "train_tokens_per_second": 5652.551
+ },
+ {
+ "epoch": 1.353016428763803,
+ "grad_norm": 0.7909886746041502,
+ "learning_rate": 2.3677862906650594e-06,
+ "loss": 0.3263228416442871,
+ "num_input_tokens_seen": 878755824,
+ "step": 20095,
+ "train_runtime": 155457.5114,
+ "train_tokens_per_second": 5652.707
+ },
+ {
+ "epoch": 1.353353083759763,
+ "grad_norm": 0.8833083944310808,
+ "learning_rate": 2.365538632177837e-06,
+ "loss": 0.3379992961883545,
+ "num_input_tokens_seen": 878961448,
+ "step": 20100,
+ "train_runtime": 155499.7479,
+ "train_tokens_per_second": 5652.494
+ },
+ {
+ "epoch": 1.3536897387557232,
+ "grad_norm": 0.7244128821181046,
+ "learning_rate": 2.36329171040934e-06,
+ "loss": 0.3040963649749756,
+ "num_input_tokens_seen": 879189696,
+ "step": 20105,
+ "train_runtime": 155531.2684,
+ "train_tokens_per_second": 5652.816
+ },
+ {
+ "epoch": 1.3540263937516832,
+ "grad_norm": 0.7531385915612084,
+ "learning_rate": 2.3610455259879045e-06,
+ "loss": 0.2924042224884033,
+ "num_input_tokens_seen": 879399600,
+ "step": 20110,
+ "train_runtime": 155567.1802,
+ "train_tokens_per_second": 5652.861
+ },
+ {
+ "epoch": 1.3543630487476435,
+ "grad_norm": 0.8365089742949873,
+ "learning_rate": 2.358800079541674e-06,
+ "loss": 0.3243632078170776,
+ "num_input_tokens_seen": 879611920,
+ "step": 20115,
+ "train_runtime": 155607.878,
+ "train_tokens_per_second": 5652.747
+ },
+ {
+ "epoch": 1.3546997037436035,
+ "grad_norm": 0.8166182835087867,
+ "learning_rate": 2.3565553716985794e-06,
+ "loss": 0.3272987365722656,
+ "num_input_tokens_seen": 879834920,
+ "step": 20120,
+ "train_runtime": 155641.8573,
+ "train_tokens_per_second": 5652.945
+ },
+ {
+ "epoch": 1.3550363587395637,
+ "grad_norm": 0.8219130483551738,
+ "learning_rate": 2.3543114030863457e-06,
+ "loss": 0.2879811763763428,
+ "num_input_tokens_seen": 880035976,
+ "step": 20125,
+ "train_runtime": 155679.3392,
+ "train_tokens_per_second": 5652.876
+ },
+ {
+ "epoch": 1.3553730137355238,
+ "grad_norm": 0.778830195040845,
+ "learning_rate": 2.352068174332491e-06,
+ "loss": 0.2902420997619629,
+ "num_input_tokens_seen": 880240280,
+ "step": 20130,
+ "train_runtime": 155721.9307,
+ "train_tokens_per_second": 5652.642
+ },
+ {
+ "epoch": 1.355709668731484,
+ "grad_norm": 0.7246075057224634,
+ "learning_rate": 2.349825686064329e-06,
+ "loss": 0.2633873462677002,
+ "num_input_tokens_seen": 880457816,
+ "step": 20135,
+ "train_runtime": 155760.9984,
+ "train_tokens_per_second": 5652.621
+ },
+ {
+ "epoch": 1.356046323727444,
+ "grad_norm": 0.7082973857821522,
+ "learning_rate": 2.3475839389089605e-06,
+ "loss": 0.31766154766082766,
+ "num_input_tokens_seen": 880688632,
+ "step": 20140,
+ "train_runtime": 155802.16,
+ "train_tokens_per_second": 5652.609
+ },
+ {
+ "epoch": 1.3563829787234043,
+ "grad_norm": 0.7717628495827882,
+ "learning_rate": 2.345342933493289e-06,
+ "loss": 0.3041571617126465,
+ "num_input_tokens_seen": 880900704,
+ "step": 20145,
+ "train_runtime": 155842.8065,
+ "train_tokens_per_second": 5652.495
+ },
+ {
+ "epoch": 1.3567196337193643,
+ "grad_norm": 0.6914688947775744,
+ "learning_rate": 2.343102670443998e-06,
+ "loss": 0.2988900661468506,
+ "num_input_tokens_seen": 881116800,
+ "step": 20150,
+ "train_runtime": 155885.0508,
+ "train_tokens_per_second": 5652.35
+ },
+ {
+ "epoch": 1.3570562887153246,
+ "grad_norm": 0.742550661860349,
+ "learning_rate": 2.3408631503875763e-06,
+ "loss": 0.32421116828918456,
+ "num_input_tokens_seen": 881349184,
+ "step": 20155,
+ "train_runtime": 155925.098,
+ "train_tokens_per_second": 5652.388
+ },
+ {
+ "epoch": 1.3573929437112846,
+ "grad_norm": 0.878343673750393,
+ "learning_rate": 2.338624373950291e-06,
+ "loss": 0.2701118946075439,
+ "num_input_tokens_seen": 881566872,
+ "step": 20160,
+ "train_runtime": 155957.1429,
+ "train_tokens_per_second": 5652.623
+ },
+ {
+ "epoch": 1.3577295987072449,
+ "grad_norm": 0.755841373165112,
+ "learning_rate": 2.336386341758216e-06,
+ "loss": 0.30777089595794677,
+ "num_input_tokens_seen": 881802424,
+ "step": 20165,
+ "train_runtime": 155994.1426,
+ "train_tokens_per_second": 5652.792
+ },
+ {
+ "epoch": 1.3580662537032049,
+ "grad_norm": 0.8144086492913879,
+ "learning_rate": 2.3341490544372063e-06,
+ "loss": 0.3082592964172363,
+ "num_input_tokens_seen": 882020264,
+ "step": 20170,
+ "train_runtime": 156033.3282,
+ "train_tokens_per_second": 5652.768
+ },
+ {
+ "epoch": 1.3584029086991651,
+ "grad_norm": 0.7512527703728554,
+ "learning_rate": 2.3319125126129128e-06,
+ "loss": 0.3048335790634155,
+ "num_input_tokens_seen": 882240304,
+ "step": 20175,
+ "train_runtime": 156070.3454,
+ "train_tokens_per_second": 5652.838
+ },
+ {
+ "epoch": 1.3587395636951252,
+ "grad_norm": 0.7584329932938183,
+ "learning_rate": 2.329676716910777e-06,
+ "loss": 0.32310731410980226,
+ "num_input_tokens_seen": 882456760,
+ "step": 20180,
+ "train_runtime": 156107.9369,
+ "train_tokens_per_second": 5652.863
+ },
+ {
+ "epoch": 1.3590762186910854,
+ "grad_norm": 0.791075197267678,
+ "learning_rate": 2.3274416679560325e-06,
+ "loss": 0.3185749053955078,
+ "num_input_tokens_seen": 882677128,
+ "step": 20185,
+ "train_runtime": 156139.8506,
+ "train_tokens_per_second": 5653.119
+ },
+ {
+ "epoch": 1.3594128736870454,
+ "grad_norm": 0.789149126017219,
+ "learning_rate": 2.3252073663737016e-06,
+ "loss": 0.31891803741455077,
+ "num_input_tokens_seen": 882909016,
+ "step": 20190,
+ "train_runtime": 156176.5363,
+ "train_tokens_per_second": 5653.276
+ },
+ {
+ "epoch": 1.3597495286830057,
+ "grad_norm": 0.8825428696038273,
+ "learning_rate": 2.3229738127886054e-06,
+ "loss": 0.312581205368042,
+ "num_input_tokens_seen": 883114832,
+ "step": 20195,
+ "train_runtime": 156221.1973,
+ "train_tokens_per_second": 5652.977
+ },
+ {
+ "epoch": 1.3600861836789657,
+ "grad_norm": 0.666411052579731,
+ "learning_rate": 2.3207410078253416e-06,
+ "loss": 0.30325491428375245,
+ "num_input_tokens_seen": 883339080,
+ "step": 20200,
+ "train_runtime": 156268.3408,
+ "train_tokens_per_second": 5652.707
+ },
+ {
+ "epoch": 1.360422838674926,
+ "grad_norm": 0.7726612121097549,
+ "learning_rate": 2.3185089521083154e-06,
+ "loss": 0.3062901496887207,
+ "num_input_tokens_seen": 883551600,
+ "step": 20205,
+ "train_runtime": 156305.042,
+ "train_tokens_per_second": 5652.739
+ },
+ {
+ "epoch": 1.360759493670886,
+ "grad_norm": 0.7956523367599065,
+ "learning_rate": 2.3162776462617067e-06,
+ "loss": 0.3058837175369263,
+ "num_input_tokens_seen": 883769552,
+ "step": 20210,
+ "train_runtime": 156351.0682,
+ "train_tokens_per_second": 5652.469
+ },
+ {
+ "epoch": 1.3610961486668463,
+ "grad_norm": 0.8171843690979657,
+ "learning_rate": 2.314047090909498e-06,
+ "loss": 0.28980307579040526,
+ "num_input_tokens_seen": 883987024,
+ "step": 20215,
+ "train_runtime": 156388.3039,
+ "train_tokens_per_second": 5652.514
+ },
+ {
+ "epoch": 1.3614328036628063,
+ "grad_norm": 0.7366060863576702,
+ "learning_rate": 2.3118172866754557e-06,
+ "loss": 0.289400577545166,
+ "num_input_tokens_seen": 884204584,
+ "step": 20220,
+ "train_runtime": 156436.651,
+ "train_tokens_per_second": 5652.157
+ },
+ {
+ "epoch": 1.3617694586587665,
+ "grad_norm": 0.8418632778595053,
+ "learning_rate": 2.309588234183137e-06,
+ "loss": 0.29900097846984863,
+ "num_input_tokens_seen": 884430104,
+ "step": 20225,
+ "train_runtime": 156479.2021,
+ "train_tokens_per_second": 5652.062
+ },
+ {
+ "epoch": 1.3621061136547266,
+ "grad_norm": 0.7596247998423821,
+ "learning_rate": 2.3073599340558896e-06,
+ "loss": 0.3193103313446045,
+ "num_input_tokens_seen": 884638136,
+ "step": 20230,
+ "train_runtime": 156513.4901,
+ "train_tokens_per_second": 5652.153
+ },
+ {
+ "epoch": 1.3624427686506868,
+ "grad_norm": 0.794889195242685,
+ "learning_rate": 2.3051323869168506e-06,
+ "loss": 0.31695847511291503,
+ "num_input_tokens_seen": 884867312,
+ "step": 20235,
+ "train_runtime": 156552.9308,
+ "train_tokens_per_second": 5652.193
+ },
+ {
+ "epoch": 1.3627794236466468,
+ "grad_norm": 0.6894100091265923,
+ "learning_rate": 2.3029055933889443e-06,
+ "loss": 0.299372124671936,
+ "num_input_tokens_seen": 885078648,
+ "step": 20240,
+ "train_runtime": 156591.403,
+ "train_tokens_per_second": 5652.154
+ },
+ {
+ "epoch": 1.363116078642607,
+ "grad_norm": 0.679829103468669,
+ "learning_rate": 2.300679554094892e-06,
+ "loss": 0.27194240093231203,
+ "num_input_tokens_seen": 885288824,
+ "step": 20245,
+ "train_runtime": 156624.8066,
+ "train_tokens_per_second": 5652.29
+ },
+ {
+ "epoch": 1.3634527336385671,
+ "grad_norm": 0.7428557022690266,
+ "learning_rate": 2.298454269657191e-06,
+ "loss": 0.2977342367172241,
+ "num_input_tokens_seen": 885509192,
+ "step": 20250,
+ "train_runtime": 156665.816,
+ "train_tokens_per_second": 5652.217
+ },
+ {
+ "epoch": 1.3637893886345274,
+ "grad_norm": 0.7495727549443245,
+ "learning_rate": 2.2962297406981432e-06,
+ "loss": 0.30438592433929446,
+ "num_input_tokens_seen": 885728368,
+ "step": 20255,
+ "train_runtime": 156704.8388,
+ "train_tokens_per_second": 5652.208
+ },
+ {
+ "epoch": 1.3641260436304874,
+ "grad_norm": 0.7491077517728351,
+ "learning_rate": 2.2940059678398223e-06,
+ "loss": 0.2986233949661255,
+ "num_input_tokens_seen": 885946024,
+ "step": 20260,
+ "train_runtime": 156744.7021,
+ "train_tokens_per_second": 5652.159
+ },
+ {
+ "epoch": 1.3644626986264476,
+ "grad_norm": 0.7341728818382711,
+ "learning_rate": 2.291782951704106e-06,
+ "loss": 0.30042304992675783,
+ "num_input_tokens_seen": 886159456,
+ "step": 20265,
+ "train_runtime": 156781.0468,
+ "train_tokens_per_second": 5652.21
+ },
+ {
+ "epoch": 1.3647993536224077,
+ "grad_norm": 0.7856616622328413,
+ "learning_rate": 2.289560692912651e-06,
+ "loss": 0.30829734802246095,
+ "num_input_tokens_seen": 886378184,
+ "step": 20270,
+ "train_runtime": 156822.443,
+ "train_tokens_per_second": 5652.113
+ },
+ {
+ "epoch": 1.365136008618368,
+ "grad_norm": 0.722302764482047,
+ "learning_rate": 2.287339192086906e-06,
+ "loss": 0.30705411434173585,
+ "num_input_tokens_seen": 886596912,
+ "step": 20275,
+ "train_runtime": 156862.8796,
+ "train_tokens_per_second": 5652.05
+ },
+ {
+ "epoch": 1.365472663614328,
+ "grad_norm": 0.7818607969471453,
+ "learning_rate": 2.2851184498481056e-06,
+ "loss": 0.30739502906799315,
+ "num_input_tokens_seen": 886810320,
+ "step": 20280,
+ "train_runtime": 156900.5368,
+ "train_tokens_per_second": 5652.054
+ },
+ {
+ "epoch": 1.3658093186102882,
+ "grad_norm": 0.8017159372977292,
+ "learning_rate": 2.2828984668172732e-06,
+ "loss": 0.29115149974822996,
+ "num_input_tokens_seen": 887018752,
+ "step": 20285,
+ "train_runtime": 156938.5878,
+ "train_tokens_per_second": 5652.012
+ },
+ {
+ "epoch": 1.3661459736062485,
+ "grad_norm": 0.8305588542630102,
+ "learning_rate": 2.2806792436152185e-06,
+ "loss": 0.3051044464111328,
+ "num_input_tokens_seen": 887236112,
+ "step": 20290,
+ "train_runtime": 156975.4855,
+ "train_tokens_per_second": 5652.068
+ },
+ {
+ "epoch": 1.3664826286022085,
+ "grad_norm": 0.80831246019114,
+ "learning_rate": 2.2784607808625463e-06,
+ "loss": 0.30942864418029786,
+ "num_input_tokens_seen": 887448744,
+ "step": 20295,
+ "train_runtime": 157014.5677,
+ "train_tokens_per_second": 5652.015
+ },
+ {
+ "epoch": 1.3668192835981685,
+ "grad_norm": 0.7675600985733588,
+ "learning_rate": 2.276243079179634e-06,
+ "loss": 0.2999575138092041,
+ "num_input_tokens_seen": 887665496,
+ "step": 20300,
+ "train_runtime": 157053.8678,
+ "train_tokens_per_second": 5651.981
+ },
+ {
+ "epoch": 1.3671559385941288,
+ "grad_norm": 0.811512893486603,
+ "learning_rate": 2.2740261391866634e-06,
+ "loss": 0.31462917327880857,
+ "num_input_tokens_seen": 887885544,
+ "step": 20305,
+ "train_runtime": 157085.2659,
+ "train_tokens_per_second": 5652.252
+ },
+ {
+ "epoch": 1.367492593590089,
+ "grad_norm": 0.7496759995290464,
+ "learning_rate": 2.2718099615035865e-06,
+ "loss": 0.2609941720962524,
+ "num_input_tokens_seen": 888107128,
+ "step": 20310,
+ "train_runtime": 157126.9618,
+ "train_tokens_per_second": 5652.163
+ },
+ {
+ "epoch": 1.367829248586049,
+ "grad_norm": 0.6960078882906409,
+ "learning_rate": 2.2695945467501567e-06,
+ "loss": 0.3140725612640381,
+ "num_input_tokens_seen": 888330336,
+ "step": 20315,
+ "train_runtime": 157165.5698,
+ "train_tokens_per_second": 5652.194
+ },
+ {
+ "epoch": 1.368165903582009,
+ "grad_norm": 0.736910787731632,
+ "learning_rate": 2.267379895545905e-06,
+ "loss": 0.2964684009552002,
+ "num_input_tokens_seen": 888549184,
+ "step": 20320,
+ "train_runtime": 157205.25,
+ "train_tokens_per_second": 5652.16
+ },
+ {
+ "epoch": 1.3685025585779693,
+ "grad_norm": 0.6504305837119558,
+ "learning_rate": 2.2651660085101513e-06,
+ "loss": 0.3142958641052246,
+ "num_input_tokens_seen": 888773632,
+ "step": 20325,
+ "train_runtime": 157243.6633,
+ "train_tokens_per_second": 5652.206
+ },
+ {
+ "epoch": 1.3688392135739296,
+ "grad_norm": 0.7317039701175643,
+ "learning_rate": 2.262952886262003e-06,
+ "loss": 0.30502166748046877,
+ "num_input_tokens_seen": 889003680,
+ "step": 20330,
+ "train_runtime": 157285.4107,
+ "train_tokens_per_second": 5652.169
+ },
+ {
+ "epoch": 1.3691758685698896,
+ "grad_norm": 0.6676476248897819,
+ "learning_rate": 2.260740529420352e-06,
+ "loss": 0.29352798461914065,
+ "num_input_tokens_seen": 889224368,
+ "step": 20335,
+ "train_runtime": 157329.8859,
+ "train_tokens_per_second": 5651.974
+ },
+ {
+ "epoch": 1.3695125235658496,
+ "grad_norm": 0.8139919345906689,
+ "learning_rate": 2.2585289386038754e-06,
+ "loss": 0.29394705295562745,
+ "num_input_tokens_seen": 889456760,
+ "step": 20340,
+ "train_runtime": 157369.7584,
+ "train_tokens_per_second": 5652.018
+ },
+ {
+ "epoch": 1.3698491785618099,
+ "grad_norm": 0.6922241241359759,
+ "learning_rate": 2.256318114431043e-06,
+ "loss": 0.3354970932006836,
+ "num_input_tokens_seen": 889693056,
+ "step": 20345,
+ "train_runtime": 157404.9626,
+ "train_tokens_per_second": 5652.255
+ },
+ {
+ "epoch": 1.3701858335577701,
+ "grad_norm": 0.8121467144758657,
+ "learning_rate": 2.2541080575200973e-06,
+ "loss": 0.2972100734710693,
+ "num_input_tokens_seen": 889902256,
+ "step": 20350,
+ "train_runtime": 157447.764,
+ "train_tokens_per_second": 5652.048
+ },
+ {
+ "epoch": 1.3705224885537302,
+ "grad_norm": 0.7645000522207971,
+ "learning_rate": 2.251898768489081e-06,
+ "loss": 0.3170339822769165,
+ "num_input_tokens_seen": 890133696,
+ "step": 20355,
+ "train_runtime": 157481.0705,
+ "train_tokens_per_second": 5652.322
+ },
+ {
+ "epoch": 1.3708591435496902,
+ "grad_norm": 0.7817916261633198,
+ "learning_rate": 2.249690247955807e-06,
+ "loss": 0.31004133224487307,
+ "num_input_tokens_seen": 890349704,
+ "step": 20360,
+ "train_runtime": 157525.1026,
+ "train_tokens_per_second": 5652.113
+ },
+ {
+ "epoch": 1.3711957985456504,
+ "grad_norm": 1.0121499916406171,
+ "learning_rate": 2.247482496537887e-06,
+ "loss": 0.32703423500061035,
+ "num_input_tokens_seen": 890570904,
+ "step": 20365,
+ "train_runtime": 157561.604,
+ "train_tokens_per_second": 5652.208
+ },
+ {
+ "epoch": 1.3715324535416107,
+ "grad_norm": 0.7908431166811735,
+ "learning_rate": 2.24527551485271e-06,
+ "loss": 0.3082599639892578,
+ "num_input_tokens_seen": 890789232,
+ "step": 20370,
+ "train_runtime": 157598.6794,
+ "train_tokens_per_second": 5652.263
+ },
+ {
+ "epoch": 1.3718691085375707,
+ "grad_norm": 0.739831014749458,
+ "learning_rate": 2.243069303517451e-06,
+ "loss": 0.3278525352478027,
+ "num_input_tokens_seen": 891003792,
+ "step": 20375,
+ "train_runtime": 157631.5929,
+ "train_tokens_per_second": 5652.444
+ },
+ {
+ "epoch": 1.3722057635335307,
+ "grad_norm": 0.6731804830965303,
+ "learning_rate": 2.2408638631490696e-06,
+ "loss": 0.2803332328796387,
+ "num_input_tokens_seen": 891222136,
+ "step": 20380,
+ "train_runtime": 157671.5132,
+ "train_tokens_per_second": 5652.398
+ },
+ {
+ "epoch": 1.372542418529491,
+ "grad_norm": 0.7577807518738551,
+ "learning_rate": 2.2386591943643117e-06,
+ "loss": 0.2959303617477417,
+ "num_input_tokens_seen": 891447208,
+ "step": 20385,
+ "train_runtime": 157709.6607,
+ "train_tokens_per_second": 5652.458
+ },
+ {
+ "epoch": 1.3728790735254512,
+ "grad_norm": 0.740650904611458,
+ "learning_rate": 2.2364552977797023e-06,
+ "loss": 0.3103938579559326,
+ "num_input_tokens_seen": 891668296,
+ "step": 20390,
+ "train_runtime": 157751.8454,
+ "train_tokens_per_second": 5652.348
+ },
+ {
+ "epoch": 1.3732157285214113,
+ "grad_norm": 0.8801339361162619,
+ "learning_rate": 2.23425217401156e-06,
+ "loss": 0.3227031707763672,
+ "num_input_tokens_seen": 891888096,
+ "step": 20395,
+ "train_runtime": 157787.2294,
+ "train_tokens_per_second": 5652.473
+ },
+ {
+ "epoch": 1.3735523835173713,
+ "grad_norm": 0.766299417789016,
+ "learning_rate": 2.2320498236759746e-06,
+ "loss": 0.2954218864440918,
+ "num_input_tokens_seen": 892110048,
+ "step": 20400,
+ "train_runtime": 157817.9321,
+ "train_tokens_per_second": 5652.78
+ },
+ {
+ "epoch": 1.3738890385133316,
+ "grad_norm": 0.6837353560545685,
+ "learning_rate": 2.2298482473888335e-06,
+ "loss": 0.2996779441833496,
+ "num_input_tokens_seen": 892325640,
+ "step": 20405,
+ "train_runtime": 157859.5289,
+ "train_tokens_per_second": 5652.656
+ },
+ {
+ "epoch": 1.3742256935092918,
+ "grad_norm": 0.7214654235542152,
+ "learning_rate": 2.227647445765792e-06,
+ "loss": 0.3093862533569336,
+ "num_input_tokens_seen": 892551144,
+ "step": 20410,
+ "train_runtime": 157899.5496,
+ "train_tokens_per_second": 5652.652
+ },
+ {
+ "epoch": 1.3745623485052518,
+ "grad_norm": 0.9006378011875812,
+ "learning_rate": 2.225447419422305e-06,
+ "loss": 0.2831476926803589,
+ "num_input_tokens_seen": 892763912,
+ "step": 20415,
+ "train_runtime": 157937.9848,
+ "train_tokens_per_second": 5652.623
+ },
+ {
+ "epoch": 1.3748990035012119,
+ "grad_norm": 0.751847783359078,
+ "learning_rate": 2.223248168973599e-06,
+ "loss": 0.29417757987976073,
+ "num_input_tokens_seen": 892984176,
+ "step": 20420,
+ "train_runtime": 157976.464,
+ "train_tokens_per_second": 5652.641
+ },
+ {
+ "epoch": 1.375235658497172,
+ "grad_norm": 0.762461921564537,
+ "learning_rate": 2.221049695034688e-06,
+ "loss": 0.2913064479827881,
+ "num_input_tokens_seen": 893210312,
+ "step": 20425,
+ "train_runtime": 158009.5129,
+ "train_tokens_per_second": 5652.889
+ },
+ {
+ "epoch": 1.3755723134931324,
+ "grad_norm": 0.7837668271345243,
+ "learning_rate": 2.218851998220369e-06,
+ "loss": 0.3103857755661011,
+ "num_input_tokens_seen": 893433344,
+ "step": 20430,
+ "train_runtime": 158043.6487,
+ "train_tokens_per_second": 5653.08
+ },
+ {
+ "epoch": 1.3759089684890924,
+ "grad_norm": 0.7225824849007935,
+ "learning_rate": 2.21665507914522e-06,
+ "loss": 0.2799984931945801,
+ "num_input_tokens_seen": 893630528,
+ "step": 20435,
+ "train_runtime": 158082.8071,
+ "train_tokens_per_second": 5652.927
+ },
+ {
+ "epoch": 1.3762456234850524,
+ "grad_norm": 0.7975885820333312,
+ "learning_rate": 2.214458938423601e-06,
+ "loss": 0.311299204826355,
+ "num_input_tokens_seen": 893854672,
+ "step": 20440,
+ "train_runtime": 158125.2748,
+ "train_tokens_per_second": 5652.826
+ },
+ {
+ "epoch": 1.3765822784810127,
+ "grad_norm": 0.6830394522468479,
+ "learning_rate": 2.212263576669661e-06,
+ "loss": 0.2979740142822266,
+ "num_input_tokens_seen": 894084224,
+ "step": 20445,
+ "train_runtime": 158169.3177,
+ "train_tokens_per_second": 5652.703
+ },
+ {
+ "epoch": 1.376918933476973,
+ "grad_norm": 0.8072743303598326,
+ "learning_rate": 2.2100689944973182e-06,
+ "loss": 0.3165121555328369,
+ "num_input_tokens_seen": 894311056,
+ "step": 20450,
+ "train_runtime": 158215.4891,
+ "train_tokens_per_second": 5652.487
+ },
+ {
+ "epoch": 1.377255588472933,
+ "grad_norm": 0.7485842516852395,
+ "learning_rate": 2.2078751925202886e-06,
+ "loss": 0.34031267166137696,
+ "num_input_tokens_seen": 894543864,
+ "step": 20455,
+ "train_runtime": 158263.0938,
+ "train_tokens_per_second": 5652.258
+ },
+ {
+ "epoch": 1.377592243468893,
+ "grad_norm": 0.878465392483748,
+ "learning_rate": 2.205682171352054e-06,
+ "loss": 0.30671300888061526,
+ "num_input_tokens_seen": 894767288,
+ "step": 20460,
+ "train_runtime": 158302.6816,
+ "train_tokens_per_second": 5652.256
+ },
+ {
+ "epoch": 1.3779288984648532,
+ "grad_norm": 0.7614994645886192,
+ "learning_rate": 2.2034899316058926e-06,
+ "loss": 0.287479043006897,
+ "num_input_tokens_seen": 894979648,
+ "step": 20465,
+ "train_runtime": 158343.8582,
+ "train_tokens_per_second": 5652.127
+ },
+ {
+ "epoch": 1.3782655534608135,
+ "grad_norm": 0.7877267867305047,
+ "learning_rate": 2.201298473894854e-06,
+ "loss": 0.29344854354858396,
+ "num_input_tokens_seen": 895182680,
+ "step": 20470,
+ "train_runtime": 158380.2408,
+ "train_tokens_per_second": 5652.111
+ },
+ {
+ "epoch": 1.3786022084567735,
+ "grad_norm": 0.7361412279331169,
+ "learning_rate": 2.1991077988317723e-06,
+ "loss": 0.30328798294067383,
+ "num_input_tokens_seen": 895414024,
+ "step": 20475,
+ "train_runtime": 158422.7502,
+ "train_tokens_per_second": 5652.055
+ },
+ {
+ "epoch": 1.3789388634527335,
+ "grad_norm": 0.751463034805612,
+ "learning_rate": 2.1969179070292637e-06,
+ "loss": 0.2803157329559326,
+ "num_input_tokens_seen": 895624392,
+ "step": 20480,
+ "train_runtime": 158463.0376,
+ "train_tokens_per_second": 5651.945
+ },
+ {
+ "epoch": 1.3792755184486938,
+ "grad_norm": 0.7411499573407221,
+ "learning_rate": 2.1947287990997236e-06,
+ "loss": 0.3004035711288452,
+ "num_input_tokens_seen": 895840912,
+ "step": 20485,
+ "train_runtime": 158494.4467,
+ "train_tokens_per_second": 5652.191
+ },
+ {
+ "epoch": 1.379612173444654,
+ "grad_norm": 0.7715375986241239,
+ "learning_rate": 2.192540475655328e-06,
+ "loss": 0.299238920211792,
+ "num_input_tokens_seen": 896053216,
+ "step": 20490,
+ "train_runtime": 158532.5086,
+ "train_tokens_per_second": 5652.173
+ },
+ {
+ "epoch": 1.379948828440614,
+ "grad_norm": 0.8053384795009297,
+ "learning_rate": 2.1903529373080397e-06,
+ "loss": 0.3175162076950073,
+ "num_input_tokens_seen": 896270408,
+ "step": 20495,
+ "train_runtime": 158571.7359,
+ "train_tokens_per_second": 5652.145
+ },
+ {
+ "epoch": 1.380285483436574,
+ "grad_norm": 0.8121078413056436,
+ "learning_rate": 2.1881661846695895e-06,
+ "loss": 0.2798121452331543,
+ "num_input_tokens_seen": 896464640,
+ "step": 20500,
+ "train_runtime": 158604.3337,
+ "train_tokens_per_second": 5652.208
+ },
+ {
+ "epoch": 1.3806221384325343,
+ "grad_norm": 0.8616911766005805,
+ "learning_rate": 2.185980218351503e-06,
+ "loss": 0.31658909320831297,
+ "num_input_tokens_seen": 896678032,
+ "step": 20505,
+ "train_runtime": 158643.7892,
+ "train_tokens_per_second": 5652.147
+ },
+ {
+ "epoch": 1.3809587934284946,
+ "grad_norm": 0.7365179371963805,
+ "learning_rate": 2.1837950389650714e-06,
+ "loss": 0.3018453598022461,
+ "num_input_tokens_seen": 896910472,
+ "step": 20510,
+ "train_runtime": 158684.9601,
+ "train_tokens_per_second": 5652.145
+ },
+ {
+ "epoch": 1.3812954484244546,
+ "grad_norm": 0.6736956456373172,
+ "learning_rate": 2.1816106471213787e-06,
+ "loss": 0.30421650409698486,
+ "num_input_tokens_seen": 897142376,
+ "step": 20515,
+ "train_runtime": 158727.7032,
+ "train_tokens_per_second": 5652.084
+ },
+ {
+ "epoch": 1.3816321034204146,
+ "grad_norm": 0.747870924932963,
+ "learning_rate": 2.179427043431282e-06,
+ "loss": 0.32655086517333987,
+ "num_input_tokens_seen": 897368312,
+ "step": 20520,
+ "train_runtime": 158760.2315,
+ "train_tokens_per_second": 5652.349
+ },
+ {
+ "epoch": 1.381968758416375,
+ "grad_norm": 0.697584938785786,
+ "learning_rate": 2.177244228505418e-06,
+ "loss": 0.32046022415161135,
+ "num_input_tokens_seen": 897570856,
+ "step": 20525,
+ "train_runtime": 158800.7072,
+ "train_tokens_per_second": 5652.184
+ },
+ {
+ "epoch": 1.3823054134123351,
+ "grad_norm": 0.8926152387651795,
+ "learning_rate": 2.1750622029542047e-06,
+ "loss": 0.2898330926895142,
+ "num_input_tokens_seen": 897801520,
+ "step": 20530,
+ "train_runtime": 158836.9929,
+ "train_tokens_per_second": 5652.345
+ },
+ {
+ "epoch": 1.3826420684082952,
+ "grad_norm": 0.8132938665366518,
+ "learning_rate": 2.1728809673878374e-06,
+ "loss": 0.31139211654663085,
+ "num_input_tokens_seen": 898013200,
+ "step": 20535,
+ "train_runtime": 158877.4038,
+ "train_tokens_per_second": 5652.24
+ },
+ {
+ "epoch": 1.3829787234042552,
+ "grad_norm": 0.7797244695242583,
+ "learning_rate": 2.1707005224162915e-06,
+ "loss": 0.2919374227523804,
+ "num_input_tokens_seen": 898234880,
+ "step": 20540,
+ "train_runtime": 158916.8206,
+ "train_tokens_per_second": 5652.233
+ },
+ {
+ "epoch": 1.3833153784002155,
+ "grad_norm": 0.7364923903516897,
+ "learning_rate": 2.1685208686493257e-06,
+ "loss": 0.30305407047271726,
+ "num_input_tokens_seen": 898436912,
+ "step": 20545,
+ "train_runtime": 158958.896,
+ "train_tokens_per_second": 5652.008
+ },
+ {
+ "epoch": 1.3836520333961757,
+ "grad_norm": 0.735813175372172,
+ "learning_rate": 2.166342006696467e-06,
+ "loss": 0.30848112106323244,
+ "num_input_tokens_seen": 898651112,
+ "step": 20550,
+ "train_runtime": 158991.467,
+ "train_tokens_per_second": 5652.197
+ },
+ {
+ "epoch": 1.3839886883921357,
+ "grad_norm": 0.7637368092940925,
+ "learning_rate": 2.1641639371670335e-06,
+ "loss": 0.3011596202850342,
+ "num_input_tokens_seen": 898866064,
+ "step": 20555,
+ "train_runtime": 159024.145,
+ "train_tokens_per_second": 5652.387
+ },
+ {
+ "epoch": 1.3843253433880958,
+ "grad_norm": 0.9110668442903609,
+ "learning_rate": 2.161986660670109e-06,
+ "loss": 0.3262584686279297,
+ "num_input_tokens_seen": 899090152,
+ "step": 20560,
+ "train_runtime": 159055.0547,
+ "train_tokens_per_second": 5652.698
+ },
+ {
+ "epoch": 1.384661998384056,
+ "grad_norm": 0.6343085979931745,
+ "learning_rate": 2.159810177814566e-06,
+ "loss": 0.2925774574279785,
+ "num_input_tokens_seen": 899299560,
+ "step": 20565,
+ "train_runtime": 159088.6315,
+ "train_tokens_per_second": 5652.821
+ },
+ {
+ "epoch": 1.3849986533800163,
+ "grad_norm": 0.7193981263314468,
+ "learning_rate": 2.1576344892090507e-06,
+ "loss": 0.3194908618927002,
+ "num_input_tokens_seen": 899514728,
+ "step": 20570,
+ "train_runtime": 159127.1256,
+ "train_tokens_per_second": 5652.806
+ },
+ {
+ "epoch": 1.3853353083759763,
+ "grad_norm": 0.7312989213302724,
+ "learning_rate": 2.1554595954619866e-06,
+ "loss": 0.30469086170196535,
+ "num_input_tokens_seen": 899730168,
+ "step": 20575,
+ "train_runtime": 159168.3547,
+ "train_tokens_per_second": 5652.695
+ },
+ {
+ "epoch": 1.3856719633719363,
+ "grad_norm": 0.7885283712306526,
+ "learning_rate": 2.1532854971815763e-06,
+ "loss": 0.30314340591430666,
+ "num_input_tokens_seen": 899940496,
+ "step": 20580,
+ "train_runtime": 159210.4856,
+ "train_tokens_per_second": 5652.52
+ },
+ {
+ "epoch": 1.3860086183678966,
+ "grad_norm": 0.7120887950890258,
+ "learning_rate": 2.151112194975799e-06,
+ "loss": 0.2792366504669189,
+ "num_input_tokens_seen": 900163032,
+ "step": 20585,
+ "train_runtime": 159256.144,
+ "train_tokens_per_second": 5652.297
+ },
+ {
+ "epoch": 1.3863452733638568,
+ "grad_norm": 0.6730346420964233,
+ "learning_rate": 2.1489396894524106e-06,
+ "loss": 0.2858273506164551,
+ "num_input_tokens_seen": 900392200,
+ "step": 20590,
+ "train_runtime": 159303.4591,
+ "train_tokens_per_second": 5652.057
+ },
+ {
+ "epoch": 1.3866819283598169,
+ "grad_norm": 0.6233675139772793,
+ "learning_rate": 2.1467679812189497e-06,
+ "loss": 0.3015708923339844,
+ "num_input_tokens_seen": 900610928,
+ "step": 20595,
+ "train_runtime": 159340.5828,
+ "train_tokens_per_second": 5652.113
+ },
+ {
+ "epoch": 1.3870185833557769,
+ "grad_norm": 1.0595463617660366,
+ "learning_rate": 2.1445970708827213e-06,
+ "loss": 0.30586059093475343,
+ "num_input_tokens_seen": 900814704,
+ "step": 20600,
+ "train_runtime": 159380.6541,
+ "train_tokens_per_second": 5651.97
+ },
+ {
+ "epoch": 1.3873552383517371,
+ "grad_norm": 0.6602500948938497,
+ "learning_rate": 2.14242695905082e-06,
+ "loss": 0.289659309387207,
+ "num_input_tokens_seen": 901036328,
+ "step": 20605,
+ "train_runtime": 159412.9764,
+ "train_tokens_per_second": 5652.214
+ },
+ {
+ "epoch": 1.3876918933476974,
+ "grad_norm": 0.7593497019073732,
+ "learning_rate": 2.1402576463301044e-06,
+ "loss": 0.3204467535018921,
+ "num_input_tokens_seen": 901271056,
+ "step": 20610,
+ "train_runtime": 159456.2832,
+ "train_tokens_per_second": 5652.151
+ },
+ {
+ "epoch": 1.3880285483436574,
+ "grad_norm": 0.8198802661040533,
+ "learning_rate": 2.13808913332722e-06,
+ "loss": 0.28775959014892577,
+ "num_input_tokens_seen": 901485976,
+ "step": 20615,
+ "train_runtime": 159494.9253,
+ "train_tokens_per_second": 5652.13
+ },
+ {
+ "epoch": 1.3883652033396174,
+ "grad_norm": 0.7717803821263688,
+ "learning_rate": 2.1359214206485845e-06,
+ "loss": 0.32880635261535646,
+ "num_input_tokens_seen": 901703104,
+ "step": 20620,
+ "train_runtime": 159533.389,
+ "train_tokens_per_second": 5652.128
+ },
+ {
+ "epoch": 1.3887018583355777,
+ "grad_norm": 0.7172069769325149,
+ "learning_rate": 2.1337545089003905e-06,
+ "loss": 0.2626586675643921,
+ "num_input_tokens_seen": 901902224,
+ "step": 20625,
+ "train_runtime": 159566.5463,
+ "train_tokens_per_second": 5652.201
+ },
+ {
+ "epoch": 1.389038513331538,
+ "grad_norm": 0.764053491586925,
+ "learning_rate": 2.131588398688608e-06,
+ "loss": 0.31684534549713134,
+ "num_input_tokens_seen": 902125584,
+ "step": 20630,
+ "train_runtime": 159611.5339,
+ "train_tokens_per_second": 5652.007
+ },
+ {
+ "epoch": 1.389375168327498,
+ "grad_norm": 0.7374395690303461,
+ "learning_rate": 2.1294230906189833e-06,
+ "loss": 0.28316431045532225,
+ "num_input_tokens_seen": 902333832,
+ "step": 20635,
+ "train_runtime": 159653.8037,
+ "train_tokens_per_second": 5651.815
+ },
+ {
+ "epoch": 1.389711823323458,
+ "grad_norm": 0.7355709013834859,
+ "learning_rate": 2.1272585852970358e-06,
+ "loss": 0.29244768619537354,
+ "num_input_tokens_seen": 902541064,
+ "step": 20640,
+ "train_runtime": 159692.4958,
+ "train_tokens_per_second": 5651.744
+ },
+ {
+ "epoch": 1.3900484783194182,
+ "grad_norm": 0.7160331969470104,
+ "learning_rate": 2.125094883328068e-06,
+ "loss": 0.2933189392089844,
+ "num_input_tokens_seen": 902745264,
+ "step": 20645,
+ "train_runtime": 159733.1324,
+ "train_tokens_per_second": 5651.584
+ },
+ {
+ "epoch": 1.3903851333153785,
+ "grad_norm": 0.8057277953126102,
+ "learning_rate": 2.1229319853171447e-06,
+ "loss": 0.2766158103942871,
+ "num_input_tokens_seen": 902957784,
+ "step": 20650,
+ "train_runtime": 159767.4906,
+ "train_tokens_per_second": 5651.699
+ },
+ {
+ "epoch": 1.3907217883113385,
+ "grad_norm": 0.8447973197859374,
+ "learning_rate": 2.1207698918691217e-06,
+ "loss": 0.3175971508026123,
+ "num_input_tokens_seen": 903176368,
+ "step": 20655,
+ "train_runtime": 159809.1069,
+ "train_tokens_per_second": 5651.595
+ },
+ {
+ "epoch": 1.3910584433072986,
+ "grad_norm": 0.7509364554421408,
+ "learning_rate": 2.1186086035886133e-06,
+ "loss": 0.28777422904968264,
+ "num_input_tokens_seen": 903403424,
+ "step": 20660,
+ "train_runtime": 159847.5357,
+ "train_tokens_per_second": 5651.657
+ },
+ {
+ "epoch": 1.3913950983032588,
+ "grad_norm": 0.7719727435620117,
+ "learning_rate": 2.116448121080023e-06,
+ "loss": 0.309741735458374,
+ "num_input_tokens_seen": 903643016,
+ "step": 20665,
+ "train_runtime": 159888.1841,
+ "train_tokens_per_second": 5651.719
+ },
+ {
+ "epoch": 1.391731753299219,
+ "grad_norm": 0.7650108977382473,
+ "learning_rate": 2.1142884449475205e-06,
+ "loss": 0.30854334831237795,
+ "num_input_tokens_seen": 903861496,
+ "step": 20670,
+ "train_runtime": 159925.9335,
+ "train_tokens_per_second": 5651.751
+ },
+ {
+ "epoch": 1.392068408295179,
+ "grad_norm": 0.8857585605046503,
+ "learning_rate": 2.112129575795052e-06,
+ "loss": 0.2980179309844971,
+ "num_input_tokens_seen": 904082696,
+ "step": 20675,
+ "train_runtime": 159970.8097,
+ "train_tokens_per_second": 5651.548
+ },
+ {
+ "epoch": 1.3924050632911391,
+ "grad_norm": 0.6579126875016472,
+ "learning_rate": 2.1099715142263394e-06,
+ "loss": 0.2740116357803345,
+ "num_input_tokens_seen": 904295832,
+ "step": 20680,
+ "train_runtime": 160012.6441,
+ "train_tokens_per_second": 5651.402
+ },
+ {
+ "epoch": 1.3927417182870994,
+ "grad_norm": 0.7692183867198543,
+ "learning_rate": 2.1078142608448773e-06,
+ "loss": 0.28600759506225587,
+ "num_input_tokens_seen": 904516064,
+ "step": 20685,
+ "train_runtime": 160046.5372,
+ "train_tokens_per_second": 5651.582
+ },
+ {
+ "epoch": 1.3930783732830596,
+ "grad_norm": 0.8142763298016777,
+ "learning_rate": 2.1056578162539327e-06,
+ "loss": 0.32352604866027834,
+ "num_input_tokens_seen": 904737152,
+ "step": 20690,
+ "train_runtime": 160088.4432,
+ "train_tokens_per_second": 5651.483
+ },
+ {
+ "epoch": 1.3934150282790196,
+ "grad_norm": 0.871875644642827,
+ "learning_rate": 2.103502181056554e-06,
+ "loss": 0.30927181243896484,
+ "num_input_tokens_seen": 904951984,
+ "step": 20695,
+ "train_runtime": 160130.5591,
+ "train_tokens_per_second": 5651.338
+ },
+ {
+ "epoch": 1.3937516832749797,
+ "grad_norm": 0.704322984008134,
+ "learning_rate": 2.1013473558555503e-06,
+ "loss": 0.2932461738586426,
+ "num_input_tokens_seen": 905188432,
+ "step": 20700,
+ "train_runtime": 160177.8633,
+ "train_tokens_per_second": 5651.146
+ },
+ {
+ "epoch": 1.39408833827094,
+ "grad_norm": 0.7189726909358929,
+ "learning_rate": 2.099193341253519e-06,
+ "loss": 0.29349417686462403,
+ "num_input_tokens_seen": 905395280,
+ "step": 20705,
+ "train_runtime": 160214.1418,
+ "train_tokens_per_second": 5651.157
+ },
+ {
+ "epoch": 1.3944249932669002,
+ "grad_norm": 0.6639597772905609,
+ "learning_rate": 2.0970401378528167e-06,
+ "loss": 0.286403751373291,
+ "num_input_tokens_seen": 905601488,
+ "step": 20710,
+ "train_runtime": 160261.6101,
+ "train_tokens_per_second": 5650.77
+ },
+ {
+ "epoch": 1.3947616482628602,
+ "grad_norm": 0.702286400471809,
+ "learning_rate": 2.0948877462555844e-06,
+ "loss": 0.2749554872512817,
+ "num_input_tokens_seen": 905827264,
+ "step": 20715,
+ "train_runtime": 160304.5265,
+ "train_tokens_per_second": 5650.666
+ },
+ {
+ "epoch": 1.3950983032588202,
+ "grad_norm": 0.7033676607004119,
+ "learning_rate": 2.09273616706373e-06,
+ "loss": 0.32062878608703616,
+ "num_input_tokens_seen": 906057568,
+ "step": 20720,
+ "train_runtime": 160337.276,
+ "train_tokens_per_second": 5650.948
+ },
+ {
+ "epoch": 1.3954349582547805,
+ "grad_norm": 0.8756497939875741,
+ "learning_rate": 2.090585400878936e-06,
+ "loss": 0.3149780035018921,
+ "num_input_tokens_seen": 906277360,
+ "step": 20725,
+ "train_runtime": 160380.6002,
+ "train_tokens_per_second": 5650.792
+ },
+ {
+ "epoch": 1.3957716132507407,
+ "grad_norm": 0.7980623610419325,
+ "learning_rate": 2.0884354483026576e-06,
+ "loss": 0.3101172924041748,
+ "num_input_tokens_seen": 906499640,
+ "step": 20730,
+ "train_runtime": 160413.5552,
+ "train_tokens_per_second": 5651.016
+ },
+ {
+ "epoch": 1.3961082682467008,
+ "grad_norm": 0.6980186229936438,
+ "learning_rate": 2.0862863099361214e-06,
+ "loss": 0.30629117488861085,
+ "num_input_tokens_seen": 906734072,
+ "step": 20735,
+ "train_runtime": 160445.4472,
+ "train_tokens_per_second": 5651.354
+ },
+ {
+ "epoch": 1.3964449232426608,
+ "grad_norm": 0.7214735662789337,
+ "learning_rate": 2.084137986380326e-06,
+ "loss": 0.3239948272705078,
+ "num_input_tokens_seen": 906945840,
+ "step": 20740,
+ "train_runtime": 160486.1913,
+ "train_tokens_per_second": 5651.239
+ },
+ {
+ "epoch": 1.396781578238621,
+ "grad_norm": 0.7503921044839111,
+ "learning_rate": 2.081990478236048e-06,
+ "loss": 0.28044352531433103,
+ "num_input_tokens_seen": 907146456,
+ "step": 20745,
+ "train_runtime": 160522.0232,
+ "train_tokens_per_second": 5651.227
+ },
+ {
+ "epoch": 1.3971182332345813,
+ "grad_norm": 0.7170605848167944,
+ "learning_rate": 2.079843786103824e-06,
+ "loss": 0.2951197624206543,
+ "num_input_tokens_seen": 907356536,
+ "step": 20750,
+ "train_runtime": 160565.1412,
+ "train_tokens_per_second": 5651.018
+ },
+ {
+ "epoch": 1.3974548882305413,
+ "grad_norm": 0.7858210180931035,
+ "learning_rate": 2.0776979105839774e-06,
+ "loss": 0.31067094802856443,
+ "num_input_tokens_seen": 907585320,
+ "step": 20755,
+ "train_runtime": 160605.9994,
+ "train_tokens_per_second": 5651.005
+ },
+ {
+ "epoch": 1.3977915432265013,
+ "grad_norm": 0.811518268399912,
+ "learning_rate": 2.075552852276588e-06,
+ "loss": 0.33663692474365237,
+ "num_input_tokens_seen": 907812376,
+ "step": 20760,
+ "train_runtime": 160647.5068,
+ "train_tokens_per_second": 5650.958
+ },
+ {
+ "epoch": 1.3981281982224616,
+ "grad_norm": 0.843500732222522,
+ "learning_rate": 2.07340861178152e-06,
+ "loss": 0.3083709716796875,
+ "num_input_tokens_seen": 908030480,
+ "step": 20765,
+ "train_runtime": 160684.2393,
+ "train_tokens_per_second": 5651.024
+ },
+ {
+ "epoch": 1.3984648532184218,
+ "grad_norm": 0.7409888586409921,
+ "learning_rate": 2.0712651896984013e-06,
+ "loss": 0.30435686111450194,
+ "num_input_tokens_seen": 908257064,
+ "step": 20770,
+ "train_runtime": 160718.9627,
+ "train_tokens_per_second": 5651.213
+ },
+ {
+ "epoch": 1.3988015082143819,
+ "grad_norm": 0.7528366267193225,
+ "learning_rate": 2.0691225866266335e-06,
+ "loss": 0.3061100244522095,
+ "num_input_tokens_seen": 908466856,
+ "step": 20775,
+ "train_runtime": 160758.0637,
+ "train_tokens_per_second": 5651.143
+ },
+ {
+ "epoch": 1.3991381632103421,
+ "grad_norm": 0.8274013015791201,
+ "learning_rate": 2.066980803165388e-06,
+ "loss": 0.3086406707763672,
+ "num_input_tokens_seen": 908690664,
+ "step": 20780,
+ "train_runtime": 160789.4191,
+ "train_tokens_per_second": 5651.433
+ },
+ {
+ "epoch": 1.3994748182063022,
+ "grad_norm": 0.7972254073123841,
+ "learning_rate": 2.0648398399136088e-06,
+ "loss": 0.30843343734741213,
+ "num_input_tokens_seen": 908916400,
+ "step": 20785,
+ "train_runtime": 160833.2267,
+ "train_tokens_per_second": 5651.297
+ },
+ {
+ "epoch": 1.3998114732022624,
+ "grad_norm": 0.7390702899298262,
+ "learning_rate": 2.062699697470007e-06,
+ "loss": 0.295151948928833,
+ "num_input_tokens_seen": 909128960,
+ "step": 20790,
+ "train_runtime": 160877.6756,
+ "train_tokens_per_second": 5651.057
+ },
+ {
+ "epoch": 1.4001481281982224,
+ "grad_norm": 0.6966665870930083,
+ "learning_rate": 2.060560376433072e-06,
+ "loss": 0.33457579612731936,
+ "num_input_tokens_seen": 909357768,
+ "step": 20795,
+ "train_runtime": 160920.3143,
+ "train_tokens_per_second": 5650.982
+ },
+ {
+ "epoch": 1.4004847831941827,
+ "grad_norm": 0.6567105821545446,
+ "learning_rate": 2.05842187740105e-06,
+ "loss": 0.3098001956939697,
+ "num_input_tokens_seen": 909572480,
+ "step": 20800,
+ "train_runtime": 160963.4825,
+ "train_tokens_per_second": 5650.8
+ },
+ {
+ "epoch": 1.4008214381901427,
+ "grad_norm": 0.8379053674392297,
+ "learning_rate": 2.0562842009719745e-06,
+ "loss": 0.2820146560668945,
+ "num_input_tokens_seen": 909757296,
+ "step": 20805,
+ "train_runtime": 161008.9129,
+ "train_tokens_per_second": 5650.354
+ },
+ {
+ "epoch": 1.401158093186103,
+ "grad_norm": 0.8065231699303244,
+ "learning_rate": 2.0541473477436317e-06,
+ "loss": 0.2844945669174194,
+ "num_input_tokens_seen": 909975984,
+ "step": 20810,
+ "train_runtime": 161043.343,
+ "train_tokens_per_second": 5650.504
+ },
+ {
+ "epoch": 1.401494748182063,
+ "grad_norm": 0.7694176813911128,
+ "learning_rate": 2.052011318313591e-06,
+ "loss": 0.28799991607666015,
+ "num_input_tokens_seen": 910199280,
+ "step": 20815,
+ "train_runtime": 161080.8517,
+ "train_tokens_per_second": 5650.574
+ },
+ {
+ "epoch": 1.4018314031780232,
+ "grad_norm": 0.7254285849515797,
+ "learning_rate": 2.049876113279185e-06,
+ "loss": 0.32419145107269287,
+ "num_input_tokens_seen": 910419968,
+ "step": 20820,
+ "train_runtime": 161115.1641,
+ "train_tokens_per_second": 5650.74
+ },
+ {
+ "epoch": 1.4021680581739833,
+ "grad_norm": 0.663299004599166,
+ "learning_rate": 2.0477417332375165e-06,
+ "loss": 0.28672337532043457,
+ "num_input_tokens_seen": 910648216,
+ "step": 20825,
+ "train_runtime": 161146.0692,
+ "train_tokens_per_second": 5651.073
+ },
+ {
+ "epoch": 1.4025047131699435,
+ "grad_norm": 0.7218766432038832,
+ "learning_rate": 2.045608178785458e-06,
+ "loss": 0.29271349906921384,
+ "num_input_tokens_seen": 910871288,
+ "step": 20830,
+ "train_runtime": 161185.4894,
+ "train_tokens_per_second": 5651.075
+ },
+ {
+ "epoch": 1.4028413681659035,
+ "grad_norm": 0.670423396219908,
+ "learning_rate": 2.043475450519652e-06,
+ "loss": 0.29650259017944336,
+ "num_input_tokens_seen": 911087016,
+ "step": 20835,
+ "train_runtime": 161217.0941,
+ "train_tokens_per_second": 5651.305
+ },
+ {
+ "epoch": 1.4031780231618638,
+ "grad_norm": 0.7755384152878778,
+ "learning_rate": 2.0413435490365064e-06,
+ "loss": 0.31277635097503664,
+ "num_input_tokens_seen": 911295624,
+ "step": 20840,
+ "train_runtime": 161255.7945,
+ "train_tokens_per_second": 5651.243
+ },
+ {
+ "epoch": 1.4035146781578238,
+ "grad_norm": 0.7431255146543578,
+ "learning_rate": 2.0392124749322064e-06,
+ "loss": 0.31907684803009034,
+ "num_input_tokens_seen": 911522672,
+ "step": 20845,
+ "train_runtime": 161298.269,
+ "train_tokens_per_second": 5651.162
+ },
+ {
+ "epoch": 1.403851333153784,
+ "grad_norm": 0.8355426134565388,
+ "learning_rate": 2.037082228802693e-06,
+ "loss": 0.28636276721954346,
+ "num_input_tokens_seen": 911742024,
+ "step": 20850,
+ "train_runtime": 161332.2795,
+ "train_tokens_per_second": 5651.33
+ },
+ {
+ "epoch": 1.404187988149744,
+ "grad_norm": 0.7290848719366411,
+ "learning_rate": 2.03495281124369e-06,
+ "loss": 0.279879093170166,
+ "num_input_tokens_seen": 911958464,
+ "step": 20855,
+ "train_runtime": 161371.514,
+ "train_tokens_per_second": 5651.298
+ },
+ {
+ "epoch": 1.4045246431457044,
+ "grad_norm": 0.7216593366596518,
+ "learning_rate": 2.0328242228506746e-06,
+ "loss": 0.28923535346984863,
+ "num_input_tokens_seen": 912172576,
+ "step": 20860,
+ "train_runtime": 161410.452,
+ "train_tokens_per_second": 5651.261
+ },
+ {
+ "epoch": 1.4048612981416644,
+ "grad_norm": 0.7898442019168748,
+ "learning_rate": 2.0306964642189053e-06,
+ "loss": 0.3036557674407959,
+ "num_input_tokens_seen": 912389440,
+ "step": 20865,
+ "train_runtime": 161449.3775,
+ "train_tokens_per_second": 5651.242
+ },
+ {
+ "epoch": 1.4051979531376246,
+ "grad_norm": 0.7311964977804852,
+ "learning_rate": 2.028569535943402e-06,
+ "loss": 0.2902818202972412,
+ "num_input_tokens_seen": 912599400,
+ "step": 20870,
+ "train_runtime": 161487.5229,
+ "train_tokens_per_second": 5651.207
+ },
+ {
+ "epoch": 1.4055346081335847,
+ "grad_norm": 0.7771281147529382,
+ "learning_rate": 2.0264434386189524e-06,
+ "loss": 0.30182933807373047,
+ "num_input_tokens_seen": 912816512,
+ "step": 20875,
+ "train_runtime": 161522.8112,
+ "train_tokens_per_second": 5651.316
+ },
+ {
+ "epoch": 1.405871263129545,
+ "grad_norm": 0.6711489743793063,
+ "learning_rate": 2.0243181728401136e-06,
+ "loss": 0.29143590927124025,
+ "num_input_tokens_seen": 913038160,
+ "step": 20880,
+ "train_runtime": 161565.9228,
+ "train_tokens_per_second": 5651.18
+ },
+ {
+ "epoch": 1.406207918125505,
+ "grad_norm": 0.7946566060485273,
+ "learning_rate": 2.022193739201208e-06,
+ "loss": 0.31396236419677737,
+ "num_input_tokens_seen": 913238344,
+ "step": 20885,
+ "train_runtime": 161599.9418,
+ "train_tokens_per_second": 5651.229
+ },
+ {
+ "epoch": 1.4065445731214652,
+ "grad_norm": 0.7053560827598276,
+ "learning_rate": 2.0200701382963267e-06,
+ "loss": 0.2820024013519287,
+ "num_input_tokens_seen": 913456008,
+ "step": 20890,
+ "train_runtime": 161640.9986,
+ "train_tokens_per_second": 5651.141
+ },
+ {
+ "epoch": 1.4068812281174252,
+ "grad_norm": 0.8293767478590064,
+ "learning_rate": 2.0179473707193324e-06,
+ "loss": 0.2967421531677246,
+ "num_input_tokens_seen": 913682016,
+ "step": 20895,
+ "train_runtime": 161681.1903,
+ "train_tokens_per_second": 5651.134
+ },
+ {
+ "epoch": 1.4072178831133855,
+ "grad_norm": 0.681242664001568,
+ "learning_rate": 2.0158254370638435e-06,
+ "loss": 0.30292630195617676,
+ "num_input_tokens_seen": 913903464,
+ "step": 20900,
+ "train_runtime": 161723.6924,
+ "train_tokens_per_second": 5651.018
+ },
+ {
+ "epoch": 1.4075545381093455,
+ "grad_norm": 0.6662676288835188,
+ "learning_rate": 2.0137043379232585e-06,
+ "loss": 0.3011344909667969,
+ "num_input_tokens_seen": 914128144,
+ "step": 20905,
+ "train_runtime": 161765.7508,
+ "train_tokens_per_second": 5650.937
+ },
+ {
+ "epoch": 1.4078911931053057,
+ "grad_norm": 0.7275678076333139,
+ "learning_rate": 2.0115840738907293e-06,
+ "loss": 0.2961131572723389,
+ "num_input_tokens_seen": 914332560,
+ "step": 20910,
+ "train_runtime": 161803.517,
+ "train_tokens_per_second": 5650.882
+ },
+ {
+ "epoch": 1.4082278481012658,
+ "grad_norm": 0.8438338939411848,
+ "learning_rate": 2.0094646455591865e-06,
+ "loss": 0.31174521446228026,
+ "num_input_tokens_seen": 914554936,
+ "step": 20915,
+ "train_runtime": 161846.6205,
+ "train_tokens_per_second": 5650.751
+ },
+ {
+ "epoch": 1.408564503097226,
+ "grad_norm": 0.7029690243486251,
+ "learning_rate": 2.007346053521319e-06,
+ "loss": 0.3245561599731445,
+ "num_input_tokens_seen": 914776376,
+ "step": 20920,
+ "train_runtime": 161885.6534,
+ "train_tokens_per_second": 5650.756
+ },
+ {
+ "epoch": 1.408901158093186,
+ "grad_norm": 0.7673612026423596,
+ "learning_rate": 2.005228298369583e-06,
+ "loss": 0.3163960695266724,
+ "num_input_tokens_seen": 914979152,
+ "step": 20925,
+ "train_runtime": 161924.9473,
+ "train_tokens_per_second": 5650.637
+ },
+ {
+ "epoch": 1.4092378130891463,
+ "grad_norm": 0.8203680546301054,
+ "learning_rate": 2.003111380696208e-06,
+ "loss": 0.282145619392395,
+ "num_input_tokens_seen": 915204416,
+ "step": 20930,
+ "train_runtime": 161972.9634,
+ "train_tokens_per_second": 5650.353
+ },
+ {
+ "epoch": 1.4095744680851063,
+ "grad_norm": 0.826854532975397,
+ "learning_rate": 2.0009953010931747e-06,
+ "loss": 0.2879758834838867,
+ "num_input_tokens_seen": 915424616,
+ "step": 20935,
+ "train_runtime": 162018.6585,
+ "train_tokens_per_second": 5650.118
+ },
+ {
+ "epoch": 1.4099111230810666,
+ "grad_norm": 0.8165582356041448,
+ "learning_rate": 1.998880060152244e-06,
+ "loss": 0.32476277351379396,
+ "num_input_tokens_seen": 915651288,
+ "step": 20940,
+ "train_runtime": 162057.5342,
+ "train_tokens_per_second": 5650.162
+ },
+ {
+ "epoch": 1.4102477780770266,
+ "grad_norm": 0.7599467895290581,
+ "learning_rate": 1.9967656584649354e-06,
+ "loss": 0.3187078952789307,
+ "num_input_tokens_seen": 915870040,
+ "step": 20945,
+ "train_runtime": 162094.7907,
+ "train_tokens_per_second": 5650.213
+ },
+ {
+ "epoch": 1.4105844330729869,
+ "grad_norm": 0.7168782752872359,
+ "learning_rate": 1.994652096622533e-06,
+ "loss": 0.3237470149993896,
+ "num_input_tokens_seen": 916090120,
+ "step": 20950,
+ "train_runtime": 162135.1509,
+ "train_tokens_per_second": 5650.164
+ },
+ {
+ "epoch": 1.410921088068947,
+ "grad_norm": 0.8403414065661872,
+ "learning_rate": 1.992539375216089e-06,
+ "loss": 0.3227944612503052,
+ "num_input_tokens_seen": 916315064,
+ "step": 20955,
+ "train_runtime": 162172.94,
+ "train_tokens_per_second": 5650.234
+ },
+ {
+ "epoch": 1.4112577430649071,
+ "grad_norm": 0.7048786926884193,
+ "learning_rate": 1.9904274948364184e-06,
+ "loss": 0.30767481327056884,
+ "num_input_tokens_seen": 916539152,
+ "step": 20960,
+ "train_runtime": 162213.0654,
+ "train_tokens_per_second": 5650.218
+ },
+ {
+ "epoch": 1.4115943980608672,
+ "grad_norm": 0.7655887084377769,
+ "learning_rate": 1.988316456074102e-06,
+ "loss": 0.30903284549713134,
+ "num_input_tokens_seen": 916766256,
+ "step": 20965,
+ "train_runtime": 162250.1786,
+ "train_tokens_per_second": 5650.325
+ },
+ {
+ "epoch": 1.4119310530568274,
+ "grad_norm": 0.7205405993664067,
+ "learning_rate": 1.9862062595194854e-06,
+ "loss": 0.31888184547424314,
+ "num_input_tokens_seen": 916978512,
+ "step": 20970,
+ "train_runtime": 162288.8233,
+ "train_tokens_per_second": 5650.288
+ },
+ {
+ "epoch": 1.4122677080527875,
+ "grad_norm": 0.8681175003128851,
+ "learning_rate": 1.984096905762676e-06,
+ "loss": 0.2966637372970581,
+ "num_input_tokens_seen": 917190424,
+ "step": 20975,
+ "train_runtime": 162331.7616,
+ "train_tokens_per_second": 5650.098
+ },
+ {
+ "epoch": 1.4126043630487477,
+ "grad_norm": 0.7477822351336473,
+ "learning_rate": 1.9819883953935533e-06,
+ "loss": 0.3042705535888672,
+ "num_input_tokens_seen": 917420304,
+ "step": 20980,
+ "train_runtime": 162368.0639,
+ "train_tokens_per_second": 5650.251
+ },
+ {
+ "epoch": 1.4129410180447077,
+ "grad_norm": 0.8100441236586595,
+ "learning_rate": 1.9798807290017485e-06,
+ "loss": 0.2959493637084961,
+ "num_input_tokens_seen": 917628664,
+ "step": 20985,
+ "train_runtime": 162403.0968,
+ "train_tokens_per_second": 5650.315
+ },
+ {
+ "epoch": 1.413277673040668,
+ "grad_norm": 0.7188686347683017,
+ "learning_rate": 1.9777739071766693e-06,
+ "loss": 0.30003719329833983,
+ "num_input_tokens_seen": 917841064,
+ "step": 20990,
+ "train_runtime": 162445.3947,
+ "train_tokens_per_second": 5650.151
+ },
+ {
+ "epoch": 1.413614328036628,
+ "grad_norm": 0.7153354531724283,
+ "learning_rate": 1.97566793050748e-06,
+ "loss": 0.2987996578216553,
+ "num_input_tokens_seen": 918043080,
+ "step": 20995,
+ "train_runtime": 162480.8482,
+ "train_tokens_per_second": 5650.162
+ },
+ {
+ "epoch": 1.4139509830325883,
+ "grad_norm": 0.8260680916935863,
+ "learning_rate": 1.9735627995831096e-06,
+ "loss": 0.2807154178619385,
+ "num_input_tokens_seen": 918268936,
+ "step": 21000,
+ "train_runtime": 162511.5804,
+ "train_tokens_per_second": 5650.483
+ },
+ {
+ "epoch": 1.4142876380285483,
+ "grad_norm": 0.8413087670817464,
+ "learning_rate": 1.9714585149922523e-06,
+ "loss": 0.322419261932373,
+ "num_input_tokens_seen": 918495704,
+ "step": 21005,
+ "train_runtime": 162548.3519,
+ "train_tokens_per_second": 5650.6
+ },
+ {
+ "epoch": 1.4146242930245085,
+ "grad_norm": 0.7701984865141125,
+ "learning_rate": 1.9693550773233634e-06,
+ "loss": 0.2878579139709473,
+ "num_input_tokens_seen": 918721536,
+ "step": 21010,
+ "train_runtime": 162584.2427,
+ "train_tokens_per_second": 5650.742
+ },
+ {
+ "epoch": 1.4149609480204686,
+ "grad_norm": 0.7629047291809269,
+ "learning_rate": 1.967252487164663e-06,
+ "loss": 0.27617802619934084,
+ "num_input_tokens_seen": 918949576,
+ "step": 21015,
+ "train_runtime": 162625.7531,
+ "train_tokens_per_second": 5650.701
+ },
+ {
+ "epoch": 1.4152976030164288,
+ "grad_norm": 0.8987812118267688,
+ "learning_rate": 1.9651507451041344e-06,
+ "loss": 0.31108310222625735,
+ "num_input_tokens_seen": 919161544,
+ "step": 21020,
+ "train_runtime": 162664.4459,
+ "train_tokens_per_second": 5650.66
+ },
+ {
+ "epoch": 1.4156342580123888,
+ "grad_norm": 0.7347239736627051,
+ "learning_rate": 1.9630498517295204e-06,
+ "loss": 0.3056536436080933,
+ "num_input_tokens_seen": 919390784,
+ "step": 21025,
+ "train_runtime": 162697.0609,
+ "train_tokens_per_second": 5650.937
+ },
+ {
+ "epoch": 1.415970913008349,
+ "grad_norm": 0.7770976786280579,
+ "learning_rate": 1.960949807628335e-06,
+ "loss": 0.31153440475463867,
+ "num_input_tokens_seen": 919610680,
+ "step": 21030,
+ "train_runtime": 162731.7215,
+ "train_tokens_per_second": 5651.084
+ },
+ {
+ "epoch": 1.4163075680043091,
+ "grad_norm": 0.797661242102853,
+ "learning_rate": 1.9588506133878415e-06,
+ "loss": 0.33806021213531495,
+ "num_input_tokens_seen": 919830696,
+ "step": 21035,
+ "train_runtime": 162768.2447,
+ "train_tokens_per_second": 5651.168
+ },
+ {
+ "epoch": 1.4166442230002694,
+ "grad_norm": 0.7171305458599561,
+ "learning_rate": 1.956752269595078e-06,
+ "loss": 0.31885838508605957,
+ "num_input_tokens_seen": 920055416,
+ "step": 21040,
+ "train_runtime": 162811.2347,
+ "train_tokens_per_second": 5651.056
+ },
+ {
+ "epoch": 1.4169808779962294,
+ "grad_norm": 0.795769696712416,
+ "learning_rate": 1.9546547768368384e-06,
+ "loss": 0.3016183853149414,
+ "num_input_tokens_seen": 920286312,
+ "step": 21045,
+ "train_runtime": 162846.1897,
+ "train_tokens_per_second": 5651.261
+ },
+ {
+ "epoch": 1.4173175329921897,
+ "grad_norm": 0.9186674926118773,
+ "learning_rate": 1.9525581356996793e-06,
+ "loss": 0.28722732067108153,
+ "num_input_tokens_seen": 920500576,
+ "step": 21050,
+ "train_runtime": 162885.5207,
+ "train_tokens_per_second": 5651.212
+ },
+ {
+ "epoch": 1.4176541879881497,
+ "grad_norm": 0.7889483479656868,
+ "learning_rate": 1.95046234676992e-06,
+ "loss": 0.30141215324401854,
+ "num_input_tokens_seen": 920738648,
+ "step": 21055,
+ "train_runtime": 162931.4756,
+ "train_tokens_per_second": 5651.079
+ },
+ {
+ "epoch": 1.41799084298411,
+ "grad_norm": 0.8574903736733213,
+ "learning_rate": 1.9483674106336416e-06,
+ "loss": 0.2801323413848877,
+ "num_input_tokens_seen": 920949800,
+ "step": 21060,
+ "train_runtime": 162973.0326,
+ "train_tokens_per_second": 5650.934
+ },
+ {
+ "epoch": 1.41832749798007,
+ "grad_norm": 0.7178113150312732,
+ "learning_rate": 1.9462733278766853e-06,
+ "loss": 0.2972491025924683,
+ "num_input_tokens_seen": 921159152,
+ "step": 21065,
+ "train_runtime": 163013.3989,
+ "train_tokens_per_second": 5650.819
+ },
+ {
+ "epoch": 1.4186641529760302,
+ "grad_norm": 0.704039276053053,
+ "learning_rate": 1.9441800990846553e-06,
+ "loss": 0.29517397880554197,
+ "num_input_tokens_seen": 921390072,
+ "step": 21070,
+ "train_runtime": 163044.8203,
+ "train_tokens_per_second": 5651.146
+ },
+ {
+ "epoch": 1.4190008079719902,
+ "grad_norm": 0.9714518989901114,
+ "learning_rate": 1.942087724842915e-06,
+ "loss": 0.3248338460922241,
+ "num_input_tokens_seen": 921598928,
+ "step": 21075,
+ "train_runtime": 163079.8467,
+ "train_tokens_per_second": 5651.213
+ },
+ {
+ "epoch": 1.4193374629679505,
+ "grad_norm": 0.8145914643998322,
+ "learning_rate": 1.9399962057365944e-06,
+ "loss": 0.30252177715301515,
+ "num_input_tokens_seen": 921803544,
+ "step": 21080,
+ "train_runtime": 163114.8336,
+ "train_tokens_per_second": 5651.255
+ },
+ {
+ "epoch": 1.4196741179639105,
+ "grad_norm": 0.7419159789443908,
+ "learning_rate": 1.937905542350574e-06,
+ "loss": 0.32372136116027833,
+ "num_input_tokens_seen": 922027736,
+ "step": 21085,
+ "train_runtime": 163156.4557,
+ "train_tokens_per_second": 5651.188
+ },
+ {
+ "epoch": 1.4200107729598708,
+ "grad_norm": 0.7910617461936362,
+ "learning_rate": 1.935815735269506e-06,
+ "loss": 0.30223875045776366,
+ "num_input_tokens_seen": 922247592,
+ "step": 21090,
+ "train_runtime": 163190.2127,
+ "train_tokens_per_second": 5651.366
+ },
+ {
+ "epoch": 1.4203474279558308,
+ "grad_norm": 0.8193345009678259,
+ "learning_rate": 1.933726785077797e-06,
+ "loss": 0.3228907108306885,
+ "num_input_tokens_seen": 922462256,
+ "step": 21095,
+ "train_runtime": 163230.7674,
+ "train_tokens_per_second": 5651.277
+ },
+ {
+ "epoch": 1.420684082951791,
+ "grad_norm": 0.8060301943601257,
+ "learning_rate": 1.9316386923596146e-06,
+ "loss": 0.3193122386932373,
+ "num_input_tokens_seen": 922683024,
+ "step": 21100,
+ "train_runtime": 163264.9379,
+ "train_tokens_per_second": 5651.446
+ },
+ {
+ "epoch": 1.421020737947751,
+ "grad_norm": 0.9170237047964257,
+ "learning_rate": 1.929551457698888e-06,
+ "loss": 0.31147422790527346,
+ "num_input_tokens_seen": 922907808,
+ "step": 21105,
+ "train_runtime": 163294.9124,
+ "train_tokens_per_second": 5651.785
+ },
+ {
+ "epoch": 1.4213573929437113,
+ "grad_norm": 0.7054636815045637,
+ "learning_rate": 1.9274650816793044e-06,
+ "loss": 0.3035153388977051,
+ "num_input_tokens_seen": 923140672,
+ "step": 21110,
+ "train_runtime": 163327.9489,
+ "train_tokens_per_second": 5652.068
+ },
+ {
+ "epoch": 1.4216940479396714,
+ "grad_norm": 0.7667361817429569,
+ "learning_rate": 1.9253795648843136e-06,
+ "loss": 0.3144373893737793,
+ "num_input_tokens_seen": 923369128,
+ "step": 21115,
+ "train_runtime": 163362.795,
+ "train_tokens_per_second": 5652.261
+ },
+ {
+ "epoch": 1.4220307029356316,
+ "grad_norm": 0.8143437243469857,
+ "learning_rate": 1.9232949078971236e-06,
+ "loss": 0.3086548805236816,
+ "num_input_tokens_seen": 923585136,
+ "step": 21120,
+ "train_runtime": 163401.3138,
+ "train_tokens_per_second": 5652.25
+ },
+ {
+ "epoch": 1.4223673579315916,
+ "grad_norm": 0.6659717947233721,
+ "learning_rate": 1.9212111113006998e-06,
+ "loss": 0.28868756294250486,
+ "num_input_tokens_seen": 923819592,
+ "step": 21125,
+ "train_runtime": 163444.6265,
+ "train_tokens_per_second": 5652.187
+ },
+ {
+ "epoch": 1.4227040129275519,
+ "grad_norm": 0.8511283738043982,
+ "learning_rate": 1.919128175677775e-06,
+ "loss": 0.2978044986724854,
+ "num_input_tokens_seen": 924024304,
+ "step": 21130,
+ "train_runtime": 163482.4066,
+ "train_tokens_per_second": 5652.133
+ },
+ {
+ "epoch": 1.423040667923512,
+ "grad_norm": 0.7092905997999482,
+ "learning_rate": 1.917046101610827e-06,
+ "loss": 0.30063571929931643,
+ "num_input_tokens_seen": 924252928,
+ "step": 21135,
+ "train_runtime": 163517.7525,
+ "train_tokens_per_second": 5652.309
+ },
+ {
+ "epoch": 1.4233773229194722,
+ "grad_norm": 0.6893485352461994,
+ "learning_rate": 1.9149648896821087e-06,
+ "loss": 0.3125055551528931,
+ "num_input_tokens_seen": 924467024,
+ "step": 21140,
+ "train_runtime": 163562.0209,
+ "train_tokens_per_second": 5652.089
+ },
+ {
+ "epoch": 1.4237139779154322,
+ "grad_norm": 0.8321835207068441,
+ "learning_rate": 1.912884540473621e-06,
+ "loss": 0.2941263675689697,
+ "num_input_tokens_seen": 924669616,
+ "step": 21145,
+ "train_runtime": 163600.3041,
+ "train_tokens_per_second": 5652.004
+ },
+ {
+ "epoch": 1.4240506329113924,
+ "grad_norm": 0.8081464869797703,
+ "learning_rate": 1.9108050545671283e-06,
+ "loss": 0.27542452812194823,
+ "num_input_tokens_seen": 924875728,
+ "step": 21150,
+ "train_runtime": 163635.3214,
+ "train_tokens_per_second": 5652.054
+ },
+ {
+ "epoch": 1.4243872879073525,
+ "grad_norm": 0.787738441929922,
+ "learning_rate": 1.9087264325441507e-06,
+ "loss": 0.27084195613861084,
+ "num_input_tokens_seen": 925083640,
+ "step": 21155,
+ "train_runtime": 163676.6739,
+ "train_tokens_per_second": 5651.897
+ },
+ {
+ "epoch": 1.4247239429033127,
+ "grad_norm": 0.7323327553262341,
+ "learning_rate": 1.9066486749859686e-06,
+ "loss": 0.3030634880065918,
+ "num_input_tokens_seen": 925295656,
+ "step": 21160,
+ "train_runtime": 163710.5311,
+ "train_tokens_per_second": 5652.023
+ },
+ {
+ "epoch": 1.4250605978992728,
+ "grad_norm": 0.6814532189083007,
+ "learning_rate": 1.9045717824736209e-06,
+ "loss": 0.3159012794494629,
+ "num_input_tokens_seen": 925523192,
+ "step": 21165,
+ "train_runtime": 163748.7467,
+ "train_tokens_per_second": 5652.093
+ },
+ {
+ "epoch": 1.425397252895233,
+ "grad_norm": 0.7944933793290381,
+ "learning_rate": 1.9024957555879037e-06,
+ "loss": 0.30693175792694094,
+ "num_input_tokens_seen": 925733280,
+ "step": 21170,
+ "train_runtime": 163785.8609,
+ "train_tokens_per_second": 5652.095
+ },
+ {
+ "epoch": 1.425733907891193,
+ "grad_norm": 0.7669788666621243,
+ "learning_rate": 1.9004205949093695e-06,
+ "loss": 0.31174449920654296,
+ "num_input_tokens_seen": 925952408,
+ "step": 21175,
+ "train_runtime": 163819.8869,
+ "train_tokens_per_second": 5652.259
+ },
+ {
+ "epoch": 1.4260705628871533,
+ "grad_norm": 0.686815248193178,
+ "learning_rate": 1.8983463010183355e-06,
+ "loss": 0.3182375907897949,
+ "num_input_tokens_seen": 926166584,
+ "step": 21180,
+ "train_runtime": 163852.9598,
+ "train_tokens_per_second": 5652.425
+ },
+ {
+ "epoch": 1.4264072178831133,
+ "grad_norm": 0.7134163125741153,
+ "learning_rate": 1.8962728744948644e-06,
+ "loss": 0.30806000232696534,
+ "num_input_tokens_seen": 926394392,
+ "step": 21185,
+ "train_runtime": 163892.3417,
+ "train_tokens_per_second": 5652.457
+ },
+ {
+ "epoch": 1.4267438728790736,
+ "grad_norm": 0.7517845683613481,
+ "learning_rate": 1.894200315918789e-06,
+ "loss": 0.3054067850112915,
+ "num_input_tokens_seen": 926630992,
+ "step": 21190,
+ "train_runtime": 163926.2105,
+ "train_tokens_per_second": 5652.732
+ },
+ {
+ "epoch": 1.4270805278750336,
+ "grad_norm": 0.8347312289874191,
+ "learning_rate": 1.8921286258696913e-06,
+ "loss": 0.2958783388137817,
+ "num_input_tokens_seen": 926834656,
+ "step": 21195,
+ "train_runtime": 163965.1465,
+ "train_tokens_per_second": 5652.632
+ },
+ {
+ "epoch": 1.4274171828709938,
+ "grad_norm": 0.695621933573107,
+ "learning_rate": 1.890057804926913e-06,
+ "loss": 0.2869250297546387,
+ "num_input_tokens_seen": 927061800,
+ "step": 21200,
+ "train_runtime": 164003.1629,
+ "train_tokens_per_second": 5652.707
+ },
+ {
+ "epoch": 1.4277538378669539,
+ "grad_norm": 0.6966348686251759,
+ "learning_rate": 1.887987853669553e-06,
+ "loss": 0.2620058536529541,
+ "num_input_tokens_seen": 927286432,
+ "step": 21205,
+ "train_runtime": 164046.6271,
+ "train_tokens_per_second": 5652.578
+ },
+ {
+ "epoch": 1.4280904928629141,
+ "grad_norm": 0.830983696417087,
+ "learning_rate": 1.8859187726764654e-06,
+ "loss": 0.3177152633666992,
+ "num_input_tokens_seen": 927512824,
+ "step": 21210,
+ "train_runtime": 164084.865,
+ "train_tokens_per_second": 5652.641
+ },
+ {
+ "epoch": 1.4284271478588741,
+ "grad_norm": 0.7634247104794691,
+ "learning_rate": 1.8838505625262638e-06,
+ "loss": 0.29890553951263427,
+ "num_input_tokens_seen": 927741616,
+ "step": 21215,
+ "train_runtime": 164125.1634,
+ "train_tokens_per_second": 5652.647
+ },
+ {
+ "epoch": 1.4287638028548344,
+ "grad_norm": 0.8416951428048186,
+ "learning_rate": 1.8817832237973154e-06,
+ "loss": 0.2987877368927002,
+ "num_input_tokens_seen": 927959432,
+ "step": 21220,
+ "train_runtime": 164164.0626,
+ "train_tokens_per_second": 5652.634
+ },
+ {
+ "epoch": 1.4291004578507944,
+ "grad_norm": 0.7473862936478309,
+ "learning_rate": 1.8797167570677439e-06,
+ "loss": 0.28604328632354736,
+ "num_input_tokens_seen": 928181056,
+ "step": 21225,
+ "train_runtime": 164213.2671,
+ "train_tokens_per_second": 5652.29
+ },
+ {
+ "epoch": 1.4294371128467547,
+ "grad_norm": 0.7676020809790378,
+ "learning_rate": 1.8776511629154342e-06,
+ "loss": 0.3128670215606689,
+ "num_input_tokens_seen": 928412792,
+ "step": 21230,
+ "train_runtime": 164257.191,
+ "train_tokens_per_second": 5652.19
+ },
+ {
+ "epoch": 1.4297737678427147,
+ "grad_norm": 0.7060907183473086,
+ "learning_rate": 1.8755864419180176e-06,
+ "loss": 0.30298452377319335,
+ "num_input_tokens_seen": 928637368,
+ "step": 21235,
+ "train_runtime": 164304.219,
+ "train_tokens_per_second": 5651.939
+ },
+ {
+ "epoch": 1.430110422838675,
+ "grad_norm": 0.7454649225429147,
+ "learning_rate": 1.8735225946528906e-06,
+ "loss": 0.3099356174468994,
+ "num_input_tokens_seen": 928865136,
+ "step": 21240,
+ "train_runtime": 164340.2939,
+ "train_tokens_per_second": 5652.084
+ },
+ {
+ "epoch": 1.430447077834635,
+ "grad_norm": 0.8028586780145847,
+ "learning_rate": 1.8714596216972008e-06,
+ "loss": 0.3137706518173218,
+ "num_input_tokens_seen": 929082920,
+ "step": 21245,
+ "train_runtime": 164371.8405,
+ "train_tokens_per_second": 5652.324
+ },
+ {
+ "epoch": 1.4307837328305952,
+ "grad_norm": 0.7853913811139581,
+ "learning_rate": 1.8693975236278512e-06,
+ "loss": 0.29973363876342773,
+ "num_input_tokens_seen": 929295744,
+ "step": 21250,
+ "train_runtime": 164413.6617,
+ "train_tokens_per_second": 5652.181
+ },
+ {
+ "epoch": 1.4311203878265553,
+ "grad_norm": 0.8450932910736807,
+ "learning_rate": 1.8673363010215017e-06,
+ "loss": 0.32357001304626465,
+ "num_input_tokens_seen": 929509608,
+ "step": 21255,
+ "train_runtime": 164454.4421,
+ "train_tokens_per_second": 5652.08
+ },
+ {
+ "epoch": 1.4314570428225155,
+ "grad_norm": 0.9839403436936655,
+ "learning_rate": 1.8652759544545661e-06,
+ "loss": 0.3077038288116455,
+ "num_input_tokens_seen": 929730176,
+ "step": 21260,
+ "train_runtime": 164491.0098,
+ "train_tokens_per_second": 5652.164
+ },
+ {
+ "epoch": 1.4317936978184758,
+ "grad_norm": 0.7604757449953816,
+ "learning_rate": 1.863216484503214e-06,
+ "loss": 0.31038436889648435,
+ "num_input_tokens_seen": 929946784,
+ "step": 21265,
+ "train_runtime": 164534.2411,
+ "train_tokens_per_second": 5651.995
+ },
+ {
+ "epoch": 1.4321303528144358,
+ "grad_norm": 0.7883194464998438,
+ "learning_rate": 1.8611578917433698e-06,
+ "loss": 0.3216505527496338,
+ "num_input_tokens_seen": 930172032,
+ "step": 21270,
+ "train_runtime": 164567.2868,
+ "train_tokens_per_second": 5652.229
+ },
+ {
+ "epoch": 1.4324670078103958,
+ "grad_norm": 0.7899893399884377,
+ "learning_rate": 1.85910017675071e-06,
+ "loss": 0.32173566818237304,
+ "num_input_tokens_seen": 930388552,
+ "step": 21275,
+ "train_runtime": 164609.009,
+ "train_tokens_per_second": 5652.112
+ },
+ {
+ "epoch": 1.432803662806356,
+ "grad_norm": 0.7217767584914241,
+ "learning_rate": 1.8570433401006737e-06,
+ "loss": 0.2714304208755493,
+ "num_input_tokens_seen": 930614664,
+ "step": 21280,
+ "train_runtime": 164643.9676,
+ "train_tokens_per_second": 5652.285
+ },
+ {
+ "epoch": 1.4331403178023163,
+ "grad_norm": 0.6750386059644308,
+ "learning_rate": 1.8549873823684418e-06,
+ "loss": 0.2765933990478516,
+ "num_input_tokens_seen": 930827792,
+ "step": 21285,
+ "train_runtime": 164684.8033,
+ "train_tokens_per_second": 5652.178
+ },
+ {
+ "epoch": 1.4334769727982763,
+ "grad_norm": 0.6817317088910789,
+ "learning_rate": 1.8529323041289616e-06,
+ "loss": 0.32932586669921876,
+ "num_input_tokens_seen": 931052808,
+ "step": 21290,
+ "train_runtime": 164719.7957,
+ "train_tokens_per_second": 5652.343
+ },
+ {
+ "epoch": 1.4338136277942364,
+ "grad_norm": 0.7091930055386446,
+ "learning_rate": 1.850878105956927e-06,
+ "loss": 0.30021870136260986,
+ "num_input_tokens_seen": 931280240,
+ "step": 21295,
+ "train_runtime": 164748.8947,
+ "train_tokens_per_second": 5652.725
+ },
+ {
+ "epoch": 1.4341502827901966,
+ "grad_norm": 0.7271899404201645,
+ "learning_rate": 1.8488247884267885e-06,
+ "loss": 0.29010815620422364,
+ "num_input_tokens_seen": 931510352,
+ "step": 21300,
+ "train_runtime": 164786.0486,
+ "train_tokens_per_second": 5652.847
+ },
+ {
+ "epoch": 1.4344869377861569,
+ "grad_norm": 0.7528962182100777,
+ "learning_rate": 1.8467723521127495e-06,
+ "loss": 0.31166844367980956,
+ "num_input_tokens_seen": 931718464,
+ "step": 21305,
+ "train_runtime": 164828.9311,
+ "train_tokens_per_second": 5652.639
+ },
+ {
+ "epoch": 1.434823592782117,
+ "grad_norm": 0.7144128400418023,
+ "learning_rate": 1.8447207975887671e-06,
+ "loss": 0.28693480491638185,
+ "num_input_tokens_seen": 931935064,
+ "step": 21310,
+ "train_runtime": 164871.6945,
+ "train_tokens_per_second": 5652.487
+ },
+ {
+ "epoch": 1.435160247778077,
+ "grad_norm": 0.7690115630964821,
+ "learning_rate": 1.8426701254285522e-06,
+ "loss": 0.3131574630737305,
+ "num_input_tokens_seen": 932161648,
+ "step": 21315,
+ "train_runtime": 164919.0598,
+ "train_tokens_per_second": 5652.237
+ },
+ {
+ "epoch": 1.4354969027740372,
+ "grad_norm": 0.7673435561570116,
+ "learning_rate": 1.840620336205569e-06,
+ "loss": 0.3200539588928223,
+ "num_input_tokens_seen": 932386520,
+ "step": 21320,
+ "train_runtime": 164962.3028,
+ "train_tokens_per_second": 5652.119
+ },
+ {
+ "epoch": 1.4358335577699974,
+ "grad_norm": 0.7307859186284956,
+ "learning_rate": 1.8385714304930324e-06,
+ "loss": 0.2745990753173828,
+ "num_input_tokens_seen": 932614272,
+ "step": 21325,
+ "train_runtime": 164999.3023,
+ "train_tokens_per_second": 5652.232
+ },
+ {
+ "epoch": 1.4361702127659575,
+ "grad_norm": 0.6957468061230739,
+ "learning_rate": 1.8365234088639178e-06,
+ "loss": 0.2942768096923828,
+ "num_input_tokens_seen": 932835368,
+ "step": 21330,
+ "train_runtime": 165036.968,
+ "train_tokens_per_second": 5652.281
+ },
+ {
+ "epoch": 1.4365068677619175,
+ "grad_norm": 0.7649094984689576,
+ "learning_rate": 1.834476271890941e-06,
+ "loss": 0.2779007196426392,
+ "num_input_tokens_seen": 933035248,
+ "step": 21335,
+ "train_runtime": 165075.7307,
+ "train_tokens_per_second": 5652.165
+ },
+ {
+ "epoch": 1.4368435227578777,
+ "grad_norm": 0.8467030298955615,
+ "learning_rate": 1.8324300201465823e-06,
+ "loss": 0.28595268726348877,
+ "num_input_tokens_seen": 933258792,
+ "step": 21340,
+ "train_runtime": 165110.7752,
+ "train_tokens_per_second": 5652.319
+ },
+ {
+ "epoch": 1.437180177753838,
+ "grad_norm": 0.9120057814047616,
+ "learning_rate": 1.8303846542030674e-06,
+ "loss": 0.3224499225616455,
+ "num_input_tokens_seen": 933469416,
+ "step": 21345,
+ "train_runtime": 165149.1219,
+ "train_tokens_per_second": 5652.282
+ },
+ {
+ "epoch": 1.437516832749798,
+ "grad_norm": 0.6674633214236769,
+ "learning_rate": 1.828340174632377e-06,
+ "loss": 0.28315880298614504,
+ "num_input_tokens_seen": 933691024,
+ "step": 21350,
+ "train_runtime": 165182.0127,
+ "train_tokens_per_second": 5652.498
+ },
+ {
+ "epoch": 1.437853487745758,
+ "grad_norm": 0.8225297648308803,
+ "learning_rate": 1.8262965820062434e-06,
+ "loss": 0.29636235237121583,
+ "num_input_tokens_seen": 933911248,
+ "step": 21355,
+ "train_runtime": 165216.2714,
+ "train_tokens_per_second": 5652.659
+ },
+ {
+ "epoch": 1.4381901427417183,
+ "grad_norm": 0.6767595245186808,
+ "learning_rate": 1.8242538768961499e-06,
+ "loss": 0.2795313835144043,
+ "num_input_tokens_seen": 934130456,
+ "step": 21360,
+ "train_runtime": 165249.3613,
+ "train_tokens_per_second": 5652.854
+ },
+ {
+ "epoch": 1.4385267977376786,
+ "grad_norm": 0.7672480972220348,
+ "learning_rate": 1.8222120598733317e-06,
+ "loss": 0.28819353580474855,
+ "num_input_tokens_seen": 934359280,
+ "step": 21365,
+ "train_runtime": 165289.8341,
+ "train_tokens_per_second": 5652.854
+ },
+ {
+ "epoch": 1.4388634527336386,
+ "grad_norm": 0.6461035328546648,
+ "learning_rate": 1.820171131508781e-06,
+ "loss": 0.32363121509552,
+ "num_input_tokens_seen": 934570728,
+ "step": 21370,
+ "train_runtime": 165324.9001,
+ "train_tokens_per_second": 5652.934
+ },
+ {
+ "epoch": 1.4392001077295986,
+ "grad_norm": 0.7231126098153248,
+ "learning_rate": 1.8181310923732304e-06,
+ "loss": 0.2775664091110229,
+ "num_input_tokens_seen": 934787104,
+ "step": 21375,
+ "train_runtime": 165367.2208,
+ "train_tokens_per_second": 5652.796
+ },
+ {
+ "epoch": 1.4395367627255589,
+ "grad_norm": 0.7398980756791921,
+ "learning_rate": 1.8160919430371764e-06,
+ "loss": 0.3209282875061035,
+ "num_input_tokens_seen": 934994512,
+ "step": 21380,
+ "train_runtime": 165407.8952,
+ "train_tokens_per_second": 5652.66
+ },
+ {
+ "epoch": 1.439873417721519,
+ "grad_norm": 0.7357422737413674,
+ "learning_rate": 1.814053684070855e-06,
+ "loss": 0.29017982482910154,
+ "num_input_tokens_seen": 935221480,
+ "step": 21385,
+ "train_runtime": 165444.6956,
+ "train_tokens_per_second": 5652.774
+ },
+ {
+ "epoch": 1.4402100727174791,
+ "grad_norm": 0.6759617094745846,
+ "learning_rate": 1.8120163160442633e-06,
+ "loss": 0.28193049430847167,
+ "num_input_tokens_seen": 935448880,
+ "step": 21390,
+ "train_runtime": 165481.2265,
+ "train_tokens_per_second": 5652.9
+ },
+ {
+ "epoch": 1.4405467277134392,
+ "grad_norm": 0.7834824232339231,
+ "learning_rate": 1.8099798395271428e-06,
+ "loss": 0.31296658515930176,
+ "num_input_tokens_seen": 935684528,
+ "step": 21395,
+ "train_runtime": 165519.1626,
+ "train_tokens_per_second": 5653.028
+ },
+ {
+ "epoch": 1.4408833827093994,
+ "grad_norm": 0.8498195490700376,
+ "learning_rate": 1.807944255088988e-06,
+ "loss": 0.28637051582336426,
+ "num_input_tokens_seen": 935896728,
+ "step": 21400,
+ "train_runtime": 165555.3887,
+ "train_tokens_per_second": 5653.073
+ },
+ {
+ "epoch": 1.4412200377053597,
+ "grad_norm": 0.748373198382384,
+ "learning_rate": 1.8059095632990437e-06,
+ "loss": 0.29571921825408937,
+ "num_input_tokens_seen": 936114728,
+ "step": 21405,
+ "train_runtime": 165587.8193,
+ "train_tokens_per_second": 5653.283
+ },
+ {
+ "epoch": 1.4415566927013197,
+ "grad_norm": 0.7727984893995797,
+ "learning_rate": 1.8038757647263045e-06,
+ "loss": 0.312378978729248,
+ "num_input_tokens_seen": 936331896,
+ "step": 21410,
+ "train_runtime": 165625.2428,
+ "train_tokens_per_second": 5653.316
+ },
+ {
+ "epoch": 1.4418933476972797,
+ "grad_norm": 0.7518266240149807,
+ "learning_rate": 1.8018428599395143e-06,
+ "loss": 0.28726487159729003,
+ "num_input_tokens_seen": 936563512,
+ "step": 21415,
+ "train_runtime": 165667.7133,
+ "train_tokens_per_second": 5653.265
+ },
+ {
+ "epoch": 1.44223000269324,
+ "grad_norm": 0.7820091979263111,
+ "learning_rate": 1.7998108495071743e-06,
+ "loss": 0.2904759645462036,
+ "num_input_tokens_seen": 936786872,
+ "step": 21420,
+ "train_runtime": 165709.7298,
+ "train_tokens_per_second": 5653.18
+ },
+ {
+ "epoch": 1.4425666576892002,
+ "grad_norm": 0.6751633930702853,
+ "learning_rate": 1.797779733997521e-06,
+ "loss": 0.2962934970855713,
+ "num_input_tokens_seen": 937007176,
+ "step": 21425,
+ "train_runtime": 165749.8126,
+ "train_tokens_per_second": 5653.142
+ },
+ {
+ "epoch": 1.4429033126851603,
+ "grad_norm": 0.8092929943597413,
+ "learning_rate": 1.7957495139785585e-06,
+ "loss": 0.3217646837234497,
+ "num_input_tokens_seen": 937228952,
+ "step": 21430,
+ "train_runtime": 165788.6583,
+ "train_tokens_per_second": 5653.155
+ },
+ {
+ "epoch": 1.4432399676811203,
+ "grad_norm": 0.7953086319181529,
+ "learning_rate": 1.7937201900180228e-06,
+ "loss": 0.3023351192474365,
+ "num_input_tokens_seen": 937432152,
+ "step": 21435,
+ "train_runtime": 165829.5232,
+ "train_tokens_per_second": 5652.987
+ },
+ {
+ "epoch": 1.4435766226770805,
+ "grad_norm": 0.779971454215353,
+ "learning_rate": 1.7916917626834134e-06,
+ "loss": 0.339884877204895,
+ "num_input_tokens_seen": 937653160,
+ "step": 21440,
+ "train_runtime": 165865.7943,
+ "train_tokens_per_second": 5653.083
+ },
+ {
+ "epoch": 1.4439132776730408,
+ "grad_norm": 0.7683843965957411,
+ "learning_rate": 1.7896642325419722e-06,
+ "loss": 0.31574015617370604,
+ "num_input_tokens_seen": 937871808,
+ "step": 21445,
+ "train_runtime": 165902.9614,
+ "train_tokens_per_second": 5653.135
+ },
+ {
+ "epoch": 1.4442499326690008,
+ "grad_norm": 0.8231498052304078,
+ "learning_rate": 1.7876376001606915e-06,
+ "loss": 0.30422613620758054,
+ "num_input_tokens_seen": 938077432,
+ "step": 21450,
+ "train_runtime": 165937.6243,
+ "train_tokens_per_second": 5653.193
+ },
+ {
+ "epoch": 1.4445865876649608,
+ "grad_norm": 0.7326842394414514,
+ "learning_rate": 1.7856118661063128e-06,
+ "loss": 0.2780045747756958,
+ "num_input_tokens_seen": 938306192,
+ "step": 21455,
+ "train_runtime": 165979.0063,
+ "train_tokens_per_second": 5653.162
+ },
+ {
+ "epoch": 1.444923242660921,
+ "grad_norm": 0.7610326872210833,
+ "learning_rate": 1.7835870309453251e-06,
+ "loss": 0.3040392637252808,
+ "num_input_tokens_seen": 938518824,
+ "step": 21460,
+ "train_runtime": 166016.1945,
+ "train_tokens_per_second": 5653.176
+ },
+ {
+ "epoch": 1.4452598976568813,
+ "grad_norm": 0.8223404953379267,
+ "learning_rate": 1.7815630952439667e-06,
+ "loss": 0.29809937477111814,
+ "num_input_tokens_seen": 938731144,
+ "step": 21465,
+ "train_runtime": 166055.7098,
+ "train_tokens_per_second": 5653.11
+ },
+ {
+ "epoch": 1.4455965526528414,
+ "grad_norm": 0.6879100802906614,
+ "learning_rate": 1.7795400595682289e-06,
+ "loss": 0.2990201473236084,
+ "num_input_tokens_seen": 938939520,
+ "step": 21470,
+ "train_runtime": 166093.1609,
+ "train_tokens_per_second": 5653.09
+ },
+ {
+ "epoch": 1.4459332076488014,
+ "grad_norm": 0.813015218983521,
+ "learning_rate": 1.7775179244838408e-06,
+ "loss": 0.342014741897583,
+ "num_input_tokens_seen": 939148712,
+ "step": 21475,
+ "train_runtime": 166128.378,
+ "train_tokens_per_second": 5653.15
+ },
+ {
+ "epoch": 1.4462698626447617,
+ "grad_norm": 0.9260324228291221,
+ "learning_rate": 1.7754966905562925e-06,
+ "loss": 0.29918684959411623,
+ "num_input_tokens_seen": 939352016,
+ "step": 21480,
+ "train_runtime": 166162.1908,
+ "train_tokens_per_second": 5653.224
+ },
+ {
+ "epoch": 1.446606517640722,
+ "grad_norm": 0.8748856406745196,
+ "learning_rate": 1.7734763583508098e-06,
+ "loss": 0.29963269233703616,
+ "num_input_tokens_seen": 939569112,
+ "step": 21485,
+ "train_runtime": 166205.207,
+ "train_tokens_per_second": 5653.067
+ },
+ {
+ "epoch": 1.446943172636682,
+ "grad_norm": 0.7627386667181976,
+ "learning_rate": 1.7714569284323757e-06,
+ "loss": 0.31328535079956055,
+ "num_input_tokens_seen": 939800608,
+ "step": 21490,
+ "train_runtime": 166243.8986,
+ "train_tokens_per_second": 5653.143
+ },
+ {
+ "epoch": 1.447279827632642,
+ "grad_norm": 0.8015396995375884,
+ "learning_rate": 1.7694384013657173e-06,
+ "loss": 0.29224376678466796,
+ "num_input_tokens_seen": 940016152,
+ "step": 21495,
+ "train_runtime": 166286.6326,
+ "train_tokens_per_second": 5652.987
+ },
+ {
+ "epoch": 1.4476164826286022,
+ "grad_norm": 0.8776810745850877,
+ "learning_rate": 1.767420777715308e-06,
+ "loss": 0.31572024822235106,
+ "num_input_tokens_seen": 940248920,
+ "step": 21500,
+ "train_runtime": 166324.4801,
+ "train_tokens_per_second": 5653.1
+ },
+ {
+ "epoch": 1.4479531376245625,
+ "grad_norm": 0.7580449081333255,
+ "learning_rate": 1.7654040580453702e-06,
+ "loss": 0.2916224241256714,
+ "num_input_tokens_seen": 940476808,
+ "step": 21505,
+ "train_runtime": 166359.4179,
+ "train_tokens_per_second": 5653.283
+ },
+ {
+ "epoch": 1.4482897926205225,
+ "grad_norm": 0.639413137355145,
+ "learning_rate": 1.7633882429198734e-06,
+ "loss": 0.3200731039047241,
+ "num_input_tokens_seen": 940704272,
+ "step": 21510,
+ "train_runtime": 166400.2408,
+ "train_tokens_per_second": 5653.263
+ },
+ {
+ "epoch": 1.4486264476164825,
+ "grad_norm": 0.7602925689681508,
+ "learning_rate": 1.7613733329025318e-06,
+ "loss": 0.29833736419677737,
+ "num_input_tokens_seen": 940937904,
+ "step": 21515,
+ "train_runtime": 166440.5197,
+ "train_tokens_per_second": 5653.298
+ },
+ {
+ "epoch": 1.4489631026124428,
+ "grad_norm": 0.7097451959902662,
+ "learning_rate": 1.759359328556814e-06,
+ "loss": 0.3069751739501953,
+ "num_input_tokens_seen": 941164048,
+ "step": 21520,
+ "train_runtime": 166482.7242,
+ "train_tokens_per_second": 5653.223
+ },
+ {
+ "epoch": 1.449299757608403,
+ "grad_norm": 0.710759458078187,
+ "learning_rate": 1.7573462304459227e-06,
+ "loss": 0.3017618179321289,
+ "num_input_tokens_seen": 941384664,
+ "step": 21525,
+ "train_runtime": 166515.4855,
+ "train_tokens_per_second": 5653.436
+ },
+ {
+ "epoch": 1.449636412604363,
+ "grad_norm": 0.7685693129079793,
+ "learning_rate": 1.7553340391328216e-06,
+ "loss": 0.28963470458984375,
+ "num_input_tokens_seen": 941603416,
+ "step": 21530,
+ "train_runtime": 166553.5695,
+ "train_tokens_per_second": 5653.457
+ },
+ {
+ "epoch": 1.449973067600323,
+ "grad_norm": 0.738257634670664,
+ "learning_rate": 1.7533227551802068e-06,
+ "loss": 0.3050259590148926,
+ "num_input_tokens_seen": 941831232,
+ "step": 21535,
+ "train_runtime": 166590.2874,
+ "train_tokens_per_second": 5653.578
+ },
+ {
+ "epoch": 1.4503097225962833,
+ "grad_norm": 0.754114134507954,
+ "learning_rate": 1.7513123791505321e-06,
+ "loss": 0.32262425422668456,
+ "num_input_tokens_seen": 942039264,
+ "step": 21540,
+ "train_runtime": 166635.0494,
+ "train_tokens_per_second": 5653.308
+ },
+ {
+ "epoch": 1.4506463775922436,
+ "grad_norm": 0.7204342650845901,
+ "learning_rate": 1.7493029116059913e-06,
+ "loss": 0.3137768268585205,
+ "num_input_tokens_seen": 942254168,
+ "step": 21545,
+ "train_runtime": 166675.6352,
+ "train_tokens_per_second": 5653.221
+ },
+ {
+ "epoch": 1.4509830325882036,
+ "grad_norm": 0.697936428942608,
+ "learning_rate": 1.7472943531085252e-06,
+ "loss": 0.29879701137542725,
+ "num_input_tokens_seen": 942479744,
+ "step": 21550,
+ "train_runtime": 166714.327,
+ "train_tokens_per_second": 5653.262
+ },
+ {
+ "epoch": 1.4513196875841636,
+ "grad_norm": 0.7740871394950247,
+ "learning_rate": 1.7452867042198212e-06,
+ "loss": 0.28659663200378416,
+ "num_input_tokens_seen": 942691224,
+ "step": 21555,
+ "train_runtime": 166751.6931,
+ "train_tokens_per_second": 5653.263
+ },
+ {
+ "epoch": 1.4516563425801239,
+ "grad_norm": 0.6502354382689284,
+ "learning_rate": 1.7432799655013122e-06,
+ "loss": 0.2767528533935547,
+ "num_input_tokens_seen": 942915144,
+ "step": 21560,
+ "train_runtime": 166790.4888,
+ "train_tokens_per_second": 5653.291
+ },
+ {
+ "epoch": 1.4519929975760841,
+ "grad_norm": 0.779285918218484,
+ "learning_rate": 1.741274137514174e-06,
+ "loss": 0.31976196765899656,
+ "num_input_tokens_seen": 943124816,
+ "step": 21565,
+ "train_runtime": 166827.7671,
+ "train_tokens_per_second": 5653.284
+ },
+ {
+ "epoch": 1.4523296525720442,
+ "grad_norm": 0.730492770357248,
+ "learning_rate": 1.7392692208193351e-06,
+ "loss": 0.28882317543029784,
+ "num_input_tokens_seen": 943341696,
+ "step": 21570,
+ "train_runtime": 166874.6737,
+ "train_tokens_per_second": 5652.995
+ },
+ {
+ "epoch": 1.4526663075680042,
+ "grad_norm": 0.7541365898746943,
+ "learning_rate": 1.7372652159774578e-06,
+ "loss": 0.27709264755249025,
+ "num_input_tokens_seen": 943566704,
+ "step": 21575,
+ "train_runtime": 166921.1117,
+ "train_tokens_per_second": 5652.77
+ },
+ {
+ "epoch": 1.4530029625639644,
+ "grad_norm": 0.7928699130255181,
+ "learning_rate": 1.7352621235489625e-06,
+ "loss": 0.2953044414520264,
+ "num_input_tokens_seen": 943783040,
+ "step": 21580,
+ "train_runtime": 166963.3208,
+ "train_tokens_per_second": 5652.637
+ },
+ {
+ "epoch": 1.4533396175599247,
+ "grad_norm": 0.9056536373531542,
+ "learning_rate": 1.733259944094e-06,
+ "loss": 0.3123426914215088,
+ "num_input_tokens_seen": 944002144,
+ "step": 21585,
+ "train_runtime": 167004.2159,
+ "train_tokens_per_second": 5652.565
+ },
+ {
+ "epoch": 1.4536762725558847,
+ "grad_norm": 0.8189634122392393,
+ "learning_rate": 1.7312586781724789e-06,
+ "loss": 0.2972878456115723,
+ "num_input_tokens_seen": 944220536,
+ "step": 21590,
+ "train_runtime": 167035.0784,
+ "train_tokens_per_second": 5652.828
+ },
+ {
+ "epoch": 1.4540129275518447,
+ "grad_norm": 0.732986430525691,
+ "learning_rate": 1.7292583263440454e-06,
+ "loss": 0.31235198974609374,
+ "num_input_tokens_seen": 944434048,
+ "step": 21595,
+ "train_runtime": 167074.9914,
+ "train_tokens_per_second": 5652.755
+ },
+ {
+ "epoch": 1.454349582547805,
+ "grad_norm": 0.7844415288989635,
+ "learning_rate": 1.7272588891680914e-06,
+ "loss": 0.28976993560791015,
+ "num_input_tokens_seen": 944637656,
+ "step": 21600,
+ "train_runtime": 167114.3019,
+ "train_tokens_per_second": 5652.644
+ },
+ {
+ "epoch": 1.4546862375437652,
+ "grad_norm": 0.7510452926746763,
+ "learning_rate": 1.7252603672037527e-06,
+ "loss": 0.29179136753082274,
+ "num_input_tokens_seen": 944849080,
+ "step": 21605,
+ "train_runtime": 167151.8596,
+ "train_tokens_per_second": 5652.639
+ },
+ {
+ "epoch": 1.4550228925397253,
+ "grad_norm": 0.7069433117953492,
+ "learning_rate": 1.72326276100991e-06,
+ "loss": 0.2792097568511963,
+ "num_input_tokens_seen": 945079208,
+ "step": 21610,
+ "train_runtime": 167188.5022,
+ "train_tokens_per_second": 5652.776
+ },
+ {
+ "epoch": 1.4553595475356853,
+ "grad_norm": 0.7362903218179828,
+ "learning_rate": 1.721266071145186e-06,
+ "loss": 0.3040107250213623,
+ "num_input_tokens_seen": 945305216,
+ "step": 21615,
+ "train_runtime": 167227.9102,
+ "train_tokens_per_second": 5652.796
+ },
+ {
+ "epoch": 1.4556962025316456,
+ "grad_norm": 0.6951834691585765,
+ "learning_rate": 1.7192702981679531e-06,
+ "loss": 0.30785455703735354,
+ "num_input_tokens_seen": 945532656,
+ "step": 21620,
+ "train_runtime": 167262.6432,
+ "train_tokens_per_second": 5652.982
+ },
+ {
+ "epoch": 1.4560328575276058,
+ "grad_norm": 0.821045679169631,
+ "learning_rate": 1.717275442636317e-06,
+ "loss": 0.27479963302612304,
+ "num_input_tokens_seen": 945753800,
+ "step": 21625,
+ "train_runtime": 167298.9162,
+ "train_tokens_per_second": 5653.078
+ },
+ {
+ "epoch": 1.4563695125235658,
+ "grad_norm": 0.8193382447695793,
+ "learning_rate": 1.7152815051081384e-06,
+ "loss": 0.30873837471008303,
+ "num_input_tokens_seen": 945972456,
+ "step": 21630,
+ "train_runtime": 167336.3125,
+ "train_tokens_per_second": 5653.121
+ },
+ {
+ "epoch": 1.4567061675195259,
+ "grad_norm": 0.638886315430458,
+ "learning_rate": 1.7132884861410093e-06,
+ "loss": 0.3021131753921509,
+ "num_input_tokens_seen": 946181664,
+ "step": 21635,
+ "train_runtime": 167372.1705,
+ "train_tokens_per_second": 5653.16
+ },
+ {
+ "epoch": 1.4570428225154861,
+ "grad_norm": 0.718262205464768,
+ "learning_rate": 1.7112963862922766e-06,
+ "loss": 0.28493921756744384,
+ "num_input_tokens_seen": 946392928,
+ "step": 21640,
+ "train_runtime": 167406.6012,
+ "train_tokens_per_second": 5653.259
+ },
+ {
+ "epoch": 1.4573794775114464,
+ "grad_norm": 0.7426374869134673,
+ "learning_rate": 1.7093052061190224e-06,
+ "loss": 0.30887959003448484,
+ "num_input_tokens_seen": 946623584,
+ "step": 21645,
+ "train_runtime": 167442.3926,
+ "train_tokens_per_second": 5653.428
+ },
+ {
+ "epoch": 1.4577161325074064,
+ "grad_norm": 0.772847955691719,
+ "learning_rate": 1.7073149461780742e-06,
+ "loss": 0.3057863712310791,
+ "num_input_tokens_seen": 946857512,
+ "step": 21650,
+ "train_runtime": 167480.3097,
+ "train_tokens_per_second": 5653.545
+ },
+ {
+ "epoch": 1.4580527875033664,
+ "grad_norm": 0.8197339025409829,
+ "learning_rate": 1.705325607026001e-06,
+ "loss": 0.31071517467498777,
+ "num_input_tokens_seen": 947062648,
+ "step": 21655,
+ "train_runtime": 167515.2805,
+ "train_tokens_per_second": 5653.59
+ },
+ {
+ "epoch": 1.4583894424993267,
+ "grad_norm": 0.8087443650582465,
+ "learning_rate": 1.7033371892191164e-06,
+ "loss": 0.30583949089050294,
+ "num_input_tokens_seen": 947284464,
+ "step": 21660,
+ "train_runtime": 167551.6527,
+ "train_tokens_per_second": 5653.686
+ },
+ {
+ "epoch": 1.458726097495287,
+ "grad_norm": 0.8458934309863756,
+ "learning_rate": 1.7013496933134726e-06,
+ "loss": 0.3118964433670044,
+ "num_input_tokens_seen": 947494920,
+ "step": 21665,
+ "train_runtime": 167591.8547,
+ "train_tokens_per_second": 5653.586
+ },
+ {
+ "epoch": 1.459062752491247,
+ "grad_norm": 0.7530131041192967,
+ "learning_rate": 1.6993631198648724e-06,
+ "loss": 0.34239089488983154,
+ "num_input_tokens_seen": 947716632,
+ "step": 21670,
+ "train_runtime": 167625.9367,
+ "train_tokens_per_second": 5653.759
+ },
+ {
+ "epoch": 1.459399407487207,
+ "grad_norm": 0.7606267649483974,
+ "learning_rate": 1.697377469428847e-06,
+ "loss": 0.29693849086761476,
+ "num_input_tokens_seen": 947941544,
+ "step": 21675,
+ "train_runtime": 167662.1721,
+ "train_tokens_per_second": 5653.878
+ },
+ {
+ "epoch": 1.4597360624831672,
+ "grad_norm": 0.759611478078839,
+ "learning_rate": 1.6953927425606842e-06,
+ "loss": 0.32032740116119385,
+ "num_input_tokens_seen": 948159704,
+ "step": 21680,
+ "train_runtime": 167706.2866,
+ "train_tokens_per_second": 5653.692
+ },
+ {
+ "epoch": 1.4600727174791275,
+ "grad_norm": 0.6996571917020317,
+ "learning_rate": 1.6934089398154002e-06,
+ "loss": 0.2828591585159302,
+ "num_input_tokens_seen": 948375840,
+ "step": 21685,
+ "train_runtime": 167745.1778,
+ "train_tokens_per_second": 5653.67
+ },
+ {
+ "epoch": 1.4604093724750875,
+ "grad_norm": 0.6597372627621453,
+ "learning_rate": 1.6914260617477634e-06,
+ "loss": 0.284234356880188,
+ "num_input_tokens_seen": 948601288,
+ "step": 21690,
+ "train_runtime": 167778.0303,
+ "train_tokens_per_second": 5653.906
+ },
+ {
+ "epoch": 1.4607460274710475,
+ "grad_norm": 0.630497034262492,
+ "learning_rate": 1.6894441089122788e-06,
+ "loss": 0.2686470985412598,
+ "num_input_tokens_seen": 948816664,
+ "step": 21695,
+ "train_runtime": 167813.8414,
+ "train_tokens_per_second": 5653.983
+ },
+ {
+ "epoch": 1.4610826824670078,
+ "grad_norm": 0.8999505558321718,
+ "learning_rate": 1.6874630818631926e-06,
+ "loss": 0.31239607334136965,
+ "num_input_tokens_seen": 949036440,
+ "step": 21700,
+ "train_runtime": 167856.1735,
+ "train_tokens_per_second": 5653.867
+ },
+ {
+ "epoch": 1.461419337462968,
+ "grad_norm": 0.7748359789057865,
+ "learning_rate": 1.685482981154492e-06,
+ "loss": 0.31434266567230223,
+ "num_input_tokens_seen": 949257872,
+ "step": 21705,
+ "train_runtime": 167886.9016,
+ "train_tokens_per_second": 5654.151
+ },
+ {
+ "epoch": 1.461755992458928,
+ "grad_norm": 0.6957529322275174,
+ "learning_rate": 1.683503807339907e-06,
+ "loss": 0.29442682266235354,
+ "num_input_tokens_seen": 949486984,
+ "step": 21710,
+ "train_runtime": 167930.9226,
+ "train_tokens_per_second": 5654.033
+ },
+ {
+ "epoch": 1.462092647454888,
+ "grad_norm": 0.7012919327076385,
+ "learning_rate": 1.6815255609729047e-06,
+ "loss": 0.3005738496780396,
+ "num_input_tokens_seen": 949713336,
+ "step": 21715,
+ "train_runtime": 167966.4841,
+ "train_tokens_per_second": 5654.184
+ },
+ {
+ "epoch": 1.4624293024508483,
+ "grad_norm": 0.8858327435153073,
+ "learning_rate": 1.6795482426067012e-06,
+ "loss": 0.30016207695007324,
+ "num_input_tokens_seen": 949920736,
+ "step": 21720,
+ "train_runtime": 168001.7371,
+ "train_tokens_per_second": 5654.232
+ },
+ {
+ "epoch": 1.4627659574468086,
+ "grad_norm": 0.8206508131279838,
+ "learning_rate": 1.6775718527942403e-06,
+ "loss": 0.2928857564926147,
+ "num_input_tokens_seen": 950115240,
+ "step": 21725,
+ "train_runtime": 168039.1435,
+ "train_tokens_per_second": 5654.13
+ },
+ {
+ "epoch": 1.4631026124427686,
+ "grad_norm": 0.8111515519361249,
+ "learning_rate": 1.675596392088219e-06,
+ "loss": 0.3051966667175293,
+ "num_input_tokens_seen": 950343040,
+ "step": 21730,
+ "train_runtime": 168073.7856,
+ "train_tokens_per_second": 5654.32
+ },
+ {
+ "epoch": 1.4634392674387287,
+ "grad_norm": 0.68703851940888,
+ "learning_rate": 1.6736218610410633e-06,
+ "loss": 0.28461332321166993,
+ "num_input_tokens_seen": 950576064,
+ "step": 21735,
+ "train_runtime": 168112.6372,
+ "train_tokens_per_second": 5654.4
+ },
+ {
+ "epoch": 1.463775922434689,
+ "grad_norm": 0.7370612171362331,
+ "learning_rate": 1.6716482602049483e-06,
+ "loss": 0.30020885467529296,
+ "num_input_tokens_seen": 950795848,
+ "step": 21740,
+ "train_runtime": 168149.267,
+ "train_tokens_per_second": 5654.475
+ },
+ {
+ "epoch": 1.4641125774306492,
+ "grad_norm": 0.8143604362673187,
+ "learning_rate": 1.669675590131784e-06,
+ "loss": 0.3058682203292847,
+ "num_input_tokens_seen": 950999432,
+ "step": 21745,
+ "train_runtime": 168183.0542,
+ "train_tokens_per_second": 5654.55
+ },
+ {
+ "epoch": 1.4644492324266092,
+ "grad_norm": 0.8042094006923218,
+ "learning_rate": 1.6677038513732214e-06,
+ "loss": 0.29998371601104734,
+ "num_input_tokens_seen": 951224688,
+ "step": 21750,
+ "train_runtime": 168221.299,
+ "train_tokens_per_second": 5654.603
+ },
+ {
+ "epoch": 1.4647858874225694,
+ "grad_norm": 0.7585438265622613,
+ "learning_rate": 1.6657330444806508e-06,
+ "loss": 0.3094784259796143,
+ "num_input_tokens_seen": 951438456,
+ "step": 21755,
+ "train_runtime": 168256.7081,
+ "train_tokens_per_second": 5654.684
+ },
+ {
+ "epoch": 1.4651225424185295,
+ "grad_norm": 0.7062718544616226,
+ "learning_rate": 1.6637631700052021e-06,
+ "loss": 0.29812400341033934,
+ "num_input_tokens_seen": 951670352,
+ "step": 21760,
+ "train_runtime": 168294.5166,
+ "train_tokens_per_second": 5654.791
+ },
+ {
+ "epoch": 1.4654591974144897,
+ "grad_norm": 0.7939195773072991,
+ "learning_rate": 1.6617942284977428e-06,
+ "loss": 0.2911461591720581,
+ "num_input_tokens_seen": 951904024,
+ "step": 21765,
+ "train_runtime": 168329.8691,
+ "train_tokens_per_second": 5654.992
+ },
+ {
+ "epoch": 1.4657958524104497,
+ "grad_norm": 0.8580033189722117,
+ "learning_rate": 1.6598262205088862e-06,
+ "loss": 0.3206942081451416,
+ "num_input_tokens_seen": 952128864,
+ "step": 21770,
+ "train_runtime": 168367.7998,
+ "train_tokens_per_second": 5655.053
+ },
+ {
+ "epoch": 1.46613250740641,
+ "grad_norm": 0.8326314241061893,
+ "learning_rate": 1.6578591465889722e-06,
+ "loss": 0.306864070892334,
+ "num_input_tokens_seen": 952357128,
+ "step": 21775,
+ "train_runtime": 168407.1966,
+ "train_tokens_per_second": 5655.086
+ },
+ {
+ "epoch": 1.46646916240237,
+ "grad_norm": 0.788622072145791,
+ "learning_rate": 1.6558930072880942e-06,
+ "loss": 0.3149855136871338,
+ "num_input_tokens_seen": 952581712,
+ "step": 21780,
+ "train_runtime": 168443.31,
+ "train_tokens_per_second": 5655.207
+ },
+ {
+ "epoch": 1.4668058173983303,
+ "grad_norm": 0.7690789166422503,
+ "learning_rate": 1.6539278031560695e-06,
+ "loss": 0.3216851234436035,
+ "num_input_tokens_seen": 952795432,
+ "step": 21785,
+ "train_runtime": 168482.8899,
+ "train_tokens_per_second": 5655.147
+ },
+ {
+ "epoch": 1.4671424723942903,
+ "grad_norm": 0.8893746538486333,
+ "learning_rate": 1.6519635347424672e-06,
+ "loss": 0.29906625747680665,
+ "num_input_tokens_seen": 953010928,
+ "step": 21790,
+ "train_runtime": 168519.6074,
+ "train_tokens_per_second": 5655.193
+ },
+ {
+ "epoch": 1.4674791273902505,
+ "grad_norm": 0.8072532370811785,
+ "learning_rate": 1.650000202596586e-06,
+ "loss": 0.3279500246047974,
+ "num_input_tokens_seen": 953243200,
+ "step": 21795,
+ "train_runtime": 168562.3622,
+ "train_tokens_per_second": 5655.137
+ },
+ {
+ "epoch": 1.4678157823862106,
+ "grad_norm": 0.7974849910062847,
+ "learning_rate": 1.6480378072674659e-06,
+ "loss": 0.3130605220794678,
+ "num_input_tokens_seen": 953466792,
+ "step": 21800,
+ "train_runtime": 168598.3186,
+ "train_tokens_per_second": 5655.257
+ },
+ {
+ "epoch": 1.4681524373821708,
+ "grad_norm": 0.7397056696771876,
+ "learning_rate": 1.646076349303884e-06,
+ "loss": 0.29395854473114014,
+ "num_input_tokens_seen": 953688232,
+ "step": 21805,
+ "train_runtime": 168638.0597,
+ "train_tokens_per_second": 5655.237
+ },
+ {
+ "epoch": 1.4684890923781309,
+ "grad_norm": 0.8246357078420278,
+ "learning_rate": 1.6441158292543557e-06,
+ "loss": 0.29840645790100095,
+ "num_input_tokens_seen": 953899232,
+ "step": 21810,
+ "train_runtime": 168676.8656,
+ "train_tokens_per_second": 5655.187
+ },
+ {
+ "epoch": 1.468825747374091,
+ "grad_norm": 0.6910590596780838,
+ "learning_rate": 1.6421562476671339e-06,
+ "loss": 0.325870156288147,
+ "num_input_tokens_seen": 954116752,
+ "step": 21815,
+ "train_runtime": 168714.8794,
+ "train_tokens_per_second": 5655.202
+ },
+ {
+ "epoch": 1.4691624023700511,
+ "grad_norm": 0.6572349294585963,
+ "learning_rate": 1.640197605090213e-06,
+ "loss": 0.27675824165344237,
+ "num_input_tokens_seen": 954338024,
+ "step": 21820,
+ "train_runtime": 168751.1585,
+ "train_tokens_per_second": 5655.298
+ },
+ {
+ "epoch": 1.4694990573660114,
+ "grad_norm": 0.7330863432474112,
+ "learning_rate": 1.6382399020713146e-06,
+ "loss": 0.29128522872924806,
+ "num_input_tokens_seen": 954538472,
+ "step": 21825,
+ "train_runtime": 168789.4804,
+ "train_tokens_per_second": 5655.201
+ },
+ {
+ "epoch": 1.4698357123619714,
+ "grad_norm": 0.8821884108200732,
+ "learning_rate": 1.6362831391579115e-06,
+ "loss": 0.2947737455368042,
+ "num_input_tokens_seen": 954753480,
+ "step": 21830,
+ "train_runtime": 168828.8811,
+ "train_tokens_per_second": 5655.155
+ },
+ {
+ "epoch": 1.4701723673579317,
+ "grad_norm": 0.7790169887599169,
+ "learning_rate": 1.634327316897199e-06,
+ "loss": 0.3025893926620483,
+ "num_input_tokens_seen": 954963128,
+ "step": 21835,
+ "train_runtime": 168866.0318,
+ "train_tokens_per_second": 5655.152
+ },
+ {
+ "epoch": 1.4705090223538917,
+ "grad_norm": 0.7612078391890548,
+ "learning_rate": 1.6323724358361216e-06,
+ "loss": 0.2903714656829834,
+ "num_input_tokens_seen": 955171016,
+ "step": 21840,
+ "train_runtime": 168904.9902,
+ "train_tokens_per_second": 5655.079
+ },
+ {
+ "epoch": 1.470845677349852,
+ "grad_norm": 0.740158711633362,
+ "learning_rate": 1.6304184965213537e-06,
+ "loss": 0.29531302452087405,
+ "num_input_tokens_seen": 955381568,
+ "step": 21845,
+ "train_runtime": 168943.7731,
+ "train_tokens_per_second": 5655.027
+ },
+ {
+ "epoch": 1.471182332345812,
+ "grad_norm": 0.7896656186315674,
+ "learning_rate": 1.6284654994993088e-06,
+ "loss": 0.30909888744354247,
+ "num_input_tokens_seen": 955612384,
+ "step": 21850,
+ "train_runtime": 168982.4971,
+ "train_tokens_per_second": 5655.097
+ },
+ {
+ "epoch": 1.4715189873417722,
+ "grad_norm": 0.6615277768870618,
+ "learning_rate": 1.6265134453161358e-06,
+ "loss": 0.29757180213928225,
+ "num_input_tokens_seen": 955824680,
+ "step": 21855,
+ "train_runtime": 169021.4477,
+ "train_tokens_per_second": 5655.05
+ },
+ {
+ "epoch": 1.4718556423377323,
+ "grad_norm": 0.8019048450650291,
+ "learning_rate": 1.6245623345177203e-06,
+ "loss": 0.3056344985961914,
+ "num_input_tokens_seen": 956030528,
+ "step": 21860,
+ "train_runtime": 169060.2008,
+ "train_tokens_per_second": 5654.971
+ },
+ {
+ "epoch": 1.4721922973336925,
+ "grad_norm": 0.6742658933683465,
+ "learning_rate": 1.622612167649683e-06,
+ "loss": 0.28513193130493164,
+ "num_input_tokens_seen": 956238824,
+ "step": 21865,
+ "train_runtime": 169094.7819,
+ "train_tokens_per_second": 5655.046
+ },
+ {
+ "epoch": 1.4725289523296525,
+ "grad_norm": 0.6663845283359301,
+ "learning_rate": 1.6206629452573868e-06,
+ "loss": 0.2704298496246338,
+ "num_input_tokens_seen": 956461272,
+ "step": 21870,
+ "train_runtime": 169128.5661,
+ "train_tokens_per_second": 5655.232
+ },
+ {
+ "epoch": 1.4728656073256128,
+ "grad_norm": 0.719412564793699,
+ "learning_rate": 1.618714667885919e-06,
+ "loss": 0.2946579933166504,
+ "num_input_tokens_seen": 956679360,
+ "step": 21875,
+ "train_runtime": 169169.36,
+ "train_tokens_per_second": 5655.157
+ },
+ {
+ "epoch": 1.4732022623215728,
+ "grad_norm": 0.7215311614958766,
+ "learning_rate": 1.6167673360801151e-06,
+ "loss": 0.29600162506103517,
+ "num_input_tokens_seen": 956909072,
+ "step": 21880,
+ "train_runtime": 169211.1899,
+ "train_tokens_per_second": 5655.117
+ },
+ {
+ "epoch": 1.473538917317533,
+ "grad_norm": 0.8145724755385428,
+ "learning_rate": 1.6148209503845336e-06,
+ "loss": 0.33778862953186034,
+ "num_input_tokens_seen": 957127864,
+ "step": 21885,
+ "train_runtime": 169247.483,
+ "train_tokens_per_second": 5655.197
+ },
+ {
+ "epoch": 1.473875572313493,
+ "grad_norm": 0.831154654556167,
+ "learning_rate": 1.6128755113434808e-06,
+ "loss": 0.2987597703933716,
+ "num_input_tokens_seen": 957342536,
+ "step": 21890,
+ "train_runtime": 169286.3383,
+ "train_tokens_per_second": 5655.167
+ },
+ {
+ "epoch": 1.4742122273094533,
+ "grad_norm": 0.6930146071437139,
+ "learning_rate": 1.6109310195009908e-06,
+ "loss": 0.303321099281311,
+ "num_input_tokens_seen": 957573936,
+ "step": 21895,
+ "train_runtime": 169325.1989,
+ "train_tokens_per_second": 5655.236
+ },
+ {
+ "epoch": 1.4745488823054134,
+ "grad_norm": 0.8269128549821896,
+ "learning_rate": 1.6089874754008323e-06,
+ "loss": 0.29223430156707764,
+ "num_input_tokens_seen": 957779504,
+ "step": 21900,
+ "train_runtime": 169360.8254,
+ "train_tokens_per_second": 5655.26
+ },
+ {
+ "epoch": 1.4748855373013736,
+ "grad_norm": 0.9691951191701637,
+ "learning_rate": 1.6070448795865162e-06,
+ "loss": 0.3116729974746704,
+ "num_input_tokens_seen": 957995792,
+ "step": 21905,
+ "train_runtime": 169396.4164,
+ "train_tokens_per_second": 5655.349
+ },
+ {
+ "epoch": 1.4752221922973336,
+ "grad_norm": 0.7309284982091603,
+ "learning_rate": 1.605103232601277e-06,
+ "loss": 0.29828710556030275,
+ "num_input_tokens_seen": 958213080,
+ "step": 21910,
+ "train_runtime": 169439.4719,
+ "train_tokens_per_second": 5655.194
+ },
+ {
+ "epoch": 1.475558847293294,
+ "grad_norm": 0.716113695585292,
+ "learning_rate": 1.603162534988096e-06,
+ "loss": 0.2677397966384888,
+ "num_input_tokens_seen": 958431216,
+ "step": 21915,
+ "train_runtime": 169477.1448,
+ "train_tokens_per_second": 5655.224
+ },
+ {
+ "epoch": 1.475895502289254,
+ "grad_norm": 0.6939181157623007,
+ "learning_rate": 1.6012227872896791e-06,
+ "loss": 0.28536086082458495,
+ "num_input_tokens_seen": 958642264,
+ "step": 21920,
+ "train_runtime": 169510.665,
+ "train_tokens_per_second": 5655.351
+ },
+ {
+ "epoch": 1.4762321572852142,
+ "grad_norm": 0.7628495263736654,
+ "learning_rate": 1.599283990048472e-06,
+ "loss": 0.2980980396270752,
+ "num_input_tokens_seen": 958850528,
+ "step": 21925,
+ "train_runtime": 169545.9545,
+ "train_tokens_per_second": 5655.402
+ },
+ {
+ "epoch": 1.4765688122811742,
+ "grad_norm": 0.7345760703144733,
+ "learning_rate": 1.5973461438066533e-06,
+ "loss": 0.2987915754318237,
+ "num_input_tokens_seen": 959072368,
+ "step": 21930,
+ "train_runtime": 169591.6892,
+ "train_tokens_per_second": 5655.185
+ },
+ {
+ "epoch": 1.4769054672771345,
+ "grad_norm": 0.7604661278462718,
+ "learning_rate": 1.5954092491061351e-06,
+ "loss": 0.28679566383361815,
+ "num_input_tokens_seen": 959287640,
+ "step": 21935,
+ "train_runtime": 169633.6673,
+ "train_tokens_per_second": 5655.055
+ },
+ {
+ "epoch": 1.4772421222730945,
+ "grad_norm": 0.6981553171592313,
+ "learning_rate": 1.5934733064885637e-06,
+ "loss": 0.30909249782562254,
+ "num_input_tokens_seen": 959515464,
+ "step": 21940,
+ "train_runtime": 169674.7339,
+ "train_tokens_per_second": 5655.029
+ },
+ {
+ "epoch": 1.4775787772690547,
+ "grad_norm": 0.7357160098109768,
+ "learning_rate": 1.5915383164953203e-06,
+ "loss": 0.3021413803100586,
+ "num_input_tokens_seen": 959722168,
+ "step": 21945,
+ "train_runtime": 169712.7213,
+ "train_tokens_per_second": 5654.981
+ },
+ {
+ "epoch": 1.4779154322650148,
+ "grad_norm": 0.8743355744735111,
+ "learning_rate": 1.5896042796675155e-06,
+ "loss": 0.2926707983016968,
+ "num_input_tokens_seen": 959926384,
+ "step": 21950,
+ "train_runtime": 169749.0885,
+ "train_tokens_per_second": 5654.972
+ },
+ {
+ "epoch": 1.478252087260975,
+ "grad_norm": 0.7663596131904904,
+ "learning_rate": 1.5876711965460023e-06,
+ "loss": 0.3000811576843262,
+ "num_input_tokens_seen": 960147848,
+ "step": 21955,
+ "train_runtime": 169793.9752,
+ "train_tokens_per_second": 5654.782
+ },
+ {
+ "epoch": 1.478588742256935,
+ "grad_norm": 0.7506605855075004,
+ "learning_rate": 1.5857390676713546e-06,
+ "loss": 0.3030038118362427,
+ "num_input_tokens_seen": 960354248,
+ "step": 21960,
+ "train_runtime": 169826.7418,
+ "train_tokens_per_second": 5654.906
+ },
+ {
+ "epoch": 1.4789253972528953,
+ "grad_norm": 0.6893712500395903,
+ "learning_rate": 1.5838078935838913e-06,
+ "loss": 0.2610928058624268,
+ "num_input_tokens_seen": 960559904,
+ "step": 21965,
+ "train_runtime": 169865.4378,
+ "train_tokens_per_second": 5654.828
+ },
+ {
+ "epoch": 1.4792620522488553,
+ "grad_norm": 0.7970970921801804,
+ "learning_rate": 1.581877674823657e-06,
+ "loss": 0.286395788192749,
+ "num_input_tokens_seen": 960763880,
+ "step": 21970,
+ "train_runtime": 169904.7926,
+ "train_tokens_per_second": 5654.719
+ },
+ {
+ "epoch": 1.4795987072448156,
+ "grad_norm": 0.7469532730634993,
+ "learning_rate": 1.5799484119304308e-06,
+ "loss": 0.3111164093017578,
+ "num_input_tokens_seen": 960983256,
+ "step": 21975,
+ "train_runtime": 169948.1325,
+ "train_tokens_per_second": 5654.568
+ },
+ {
+ "epoch": 1.4799353622407756,
+ "grad_norm": 0.7131812605057424,
+ "learning_rate": 1.578020105443726e-06,
+ "loss": 0.2867883205413818,
+ "num_input_tokens_seen": 961204296,
+ "step": 21980,
+ "train_runtime": 169989.5356,
+ "train_tokens_per_second": 5654.491
+ },
+ {
+ "epoch": 1.4802720172367358,
+ "grad_norm": 0.8560168052021411,
+ "learning_rate": 1.576092755902787e-06,
+ "loss": 0.3115290641784668,
+ "num_input_tokens_seen": 961419104,
+ "step": 21985,
+ "train_runtime": 170025.6173,
+ "train_tokens_per_second": 5654.554
+ },
+ {
+ "epoch": 1.4806086722326959,
+ "grad_norm": 0.7910019975942408,
+ "learning_rate": 1.5741663638465903e-06,
+ "loss": 0.32115159034729,
+ "num_input_tokens_seen": 961624552,
+ "step": 21990,
+ "train_runtime": 170059.2178,
+ "train_tokens_per_second": 5654.645
+ },
+ {
+ "epoch": 1.4809453272286561,
+ "grad_norm": 0.7740362774361383,
+ "learning_rate": 1.5722409298138463e-06,
+ "loss": 0.2831075668334961,
+ "num_input_tokens_seen": 961828112,
+ "step": 21995,
+ "train_runtime": 170097.4303,
+ "train_tokens_per_second": 5654.572
+ },
+ {
+ "epoch": 1.4812819822246162,
+ "grad_norm": 0.887435473761687,
+ "learning_rate": 1.570316454342994e-06,
+ "loss": 0.3216114044189453,
+ "num_input_tokens_seen": 962043280,
+ "step": 22000,
+ "train_runtime": 170135.4044,
+ "train_tokens_per_second": 5654.574
+ },
+ {
+ "epoch": 1.4816186372205764,
+ "grad_norm": 0.7331141791279817,
+ "learning_rate": 1.568392937972212e-06,
+ "loss": 0.30264816284179685,
+ "num_input_tokens_seen": 962277624,
+ "step": 22005,
+ "train_runtime": 170177.2637,
+ "train_tokens_per_second": 5654.56
+ },
+ {
+ "epoch": 1.4819552922165364,
+ "grad_norm": 0.7730593120342526,
+ "learning_rate": 1.5664703812393988e-06,
+ "loss": 0.30075869560241697,
+ "num_input_tokens_seen": 962510352,
+ "step": 22010,
+ "train_runtime": 170215.2224,
+ "train_tokens_per_second": 5654.667
+ },
+ {
+ "epoch": 1.4822919472124967,
+ "grad_norm": 0.7777724901976949,
+ "learning_rate": 1.5645487846821967e-06,
+ "loss": 0.2980914831161499,
+ "num_input_tokens_seen": 962736264,
+ "step": 22015,
+ "train_runtime": 170256.3094,
+ "train_tokens_per_second": 5654.629
+ },
+ {
+ "epoch": 1.4826286022084567,
+ "grad_norm": 0.7843674786251581,
+ "learning_rate": 1.5626281488379724e-06,
+ "loss": 0.30644855499267576,
+ "num_input_tokens_seen": 962954856,
+ "step": 22020,
+ "train_runtime": 170290.947,
+ "train_tokens_per_second": 5654.762
+ },
+ {
+ "epoch": 1.482965257204417,
+ "grad_norm": 0.7409514511968761,
+ "learning_rate": 1.560708474243825e-06,
+ "loss": 0.28392848968505857,
+ "num_input_tokens_seen": 963172840,
+ "step": 22025,
+ "train_runtime": 170330.7319,
+ "train_tokens_per_second": 5654.721
+ },
+ {
+ "epoch": 1.483301912200377,
+ "grad_norm": 0.8527324584193815,
+ "learning_rate": 1.5587897614365854e-06,
+ "loss": 0.3137376070022583,
+ "num_input_tokens_seen": 963402704,
+ "step": 22030,
+ "train_runtime": 170363.8783,
+ "train_tokens_per_second": 5654.97
+ },
+ {
+ "epoch": 1.4836385671963372,
+ "grad_norm": 0.817954831707004,
+ "learning_rate": 1.5568720109528163e-06,
+ "loss": 0.29897422790527345,
+ "num_input_tokens_seen": 963622656,
+ "step": 22035,
+ "train_runtime": 170407.9186,
+ "train_tokens_per_second": 5654.8
+ },
+ {
+ "epoch": 1.4839752221922973,
+ "grad_norm": 0.9090048918137543,
+ "learning_rate": 1.5549552233288095e-06,
+ "loss": 0.32768850326538085,
+ "num_input_tokens_seen": 963839112,
+ "step": 22040,
+ "train_runtime": 170444.0634,
+ "train_tokens_per_second": 5654.871
+ },
+ {
+ "epoch": 1.4843118771882575,
+ "grad_norm": 0.7108214887496512,
+ "learning_rate": 1.5530393991005893e-06,
+ "loss": 0.2966951370239258,
+ "num_input_tokens_seen": 964056312,
+ "step": 22045,
+ "train_runtime": 170495.1004,
+ "train_tokens_per_second": 5654.452
+ },
+ {
+ "epoch": 1.4846485321842176,
+ "grad_norm": 0.7628438783439455,
+ "learning_rate": 1.551124538803908e-06,
+ "loss": 0.3010308265686035,
+ "num_input_tokens_seen": 964282488,
+ "step": 22050,
+ "train_runtime": 170534.3626,
+ "train_tokens_per_second": 5654.476
+ },
+ {
+ "epoch": 1.4849851871801778,
+ "grad_norm": 0.766886301451967,
+ "learning_rate": 1.5492106429742547e-06,
+ "loss": 0.30379014015197753,
+ "num_input_tokens_seen": 964505592,
+ "step": 22055,
+ "train_runtime": 170574.175,
+ "train_tokens_per_second": 5654.464
+ },
+ {
+ "epoch": 1.4853218421761378,
+ "grad_norm": 0.8077546385043058,
+ "learning_rate": 1.5472977121468386e-06,
+ "loss": 0.3085505962371826,
+ "num_input_tokens_seen": 964734320,
+ "step": 22060,
+ "train_runtime": 170605.4495,
+ "train_tokens_per_second": 5654.768
+ },
+ {
+ "epoch": 1.485658497172098,
+ "grad_norm": 0.8224102700502363,
+ "learning_rate": 1.545385746856609e-06,
+ "loss": 0.2914187669754028,
+ "num_input_tokens_seen": 964946888,
+ "step": 22065,
+ "train_runtime": 170641.0717,
+ "train_tokens_per_second": 5654.834
+ },
+ {
+ "epoch": 1.485995152168058,
+ "grad_norm": 0.73340844909282,
+ "learning_rate": 1.5434747476382388e-06,
+ "loss": 0.2726666212081909,
+ "num_input_tokens_seen": 965178856,
+ "step": 22070,
+ "train_runtime": 170676.1913,
+ "train_tokens_per_second": 5655.029
+ },
+ {
+ "epoch": 1.4863318071640184,
+ "grad_norm": 0.8499683640218707,
+ "learning_rate": 1.5415647150261337e-06,
+ "loss": 0.2993777751922607,
+ "num_input_tokens_seen": 965382264,
+ "step": 22075,
+ "train_runtime": 170714.8439,
+ "train_tokens_per_second": 5654.94
+ },
+ {
+ "epoch": 1.4866684621599784,
+ "grad_norm": 0.8088856371624081,
+ "learning_rate": 1.539655649554428e-06,
+ "loss": 0.2940176010131836,
+ "num_input_tokens_seen": 965597136,
+ "step": 22080,
+ "train_runtime": 170755.1927,
+ "train_tokens_per_second": 5654.862
+ },
+ {
+ "epoch": 1.4870051171559386,
+ "grad_norm": 0.7548642358843536,
+ "learning_rate": 1.5377475517569856e-06,
+ "loss": 0.29267857074737547,
+ "num_input_tokens_seen": 965803184,
+ "step": 22085,
+ "train_runtime": 170791.9589,
+ "train_tokens_per_second": 5654.852
+ },
+ {
+ "epoch": 1.4873417721518987,
+ "grad_norm": 0.765784318943008,
+ "learning_rate": 1.5358404221674e-06,
+ "loss": 0.30087296962738036,
+ "num_input_tokens_seen": 966009728,
+ "step": 22090,
+ "train_runtime": 170831.4433,
+ "train_tokens_per_second": 5654.754
+ },
+ {
+ "epoch": 1.487678427147859,
+ "grad_norm": 0.7556445487670026,
+ "learning_rate": 1.533934261318994e-06,
+ "loss": 0.30022130012512205,
+ "num_input_tokens_seen": 966233096,
+ "step": 22095,
+ "train_runtime": 170870.3846,
+ "train_tokens_per_second": 5654.772
+ },
+ {
+ "epoch": 1.488015082143819,
+ "grad_norm": 0.7251976781164082,
+ "learning_rate": 1.5320290697448177e-06,
+ "loss": 0.2852438449859619,
+ "num_input_tokens_seen": 966445288,
+ "step": 22100,
+ "train_runtime": 170911.1982,
+ "train_tokens_per_second": 5654.663
+ },
+ {
+ "epoch": 1.4883517371397792,
+ "grad_norm": 0.859114292106197,
+ "learning_rate": 1.530124847977657e-06,
+ "loss": 0.320376992225647,
+ "num_input_tokens_seen": 966659272,
+ "step": 22105,
+ "train_runtime": 170948.93,
+ "train_tokens_per_second": 5654.667
+ },
+ {
+ "epoch": 1.4886883921357392,
+ "grad_norm": 0.7697511007972789,
+ "learning_rate": 1.528221596550014e-06,
+ "loss": 0.2888146638870239,
+ "num_input_tokens_seen": 966874112,
+ "step": 22110,
+ "train_runtime": 170988.9473,
+ "train_tokens_per_second": 5654.6
+ },
+ {
+ "epoch": 1.4890250471316995,
+ "grad_norm": 0.7792399302642892,
+ "learning_rate": 1.5263193159941331e-06,
+ "loss": 0.2921769618988037,
+ "num_input_tokens_seen": 967092032,
+ "step": 22115,
+ "train_runtime": 171028.9685,
+ "train_tokens_per_second": 5654.551
+ },
+ {
+ "epoch": 1.4893617021276595,
+ "grad_norm": 0.7804399534999346,
+ "learning_rate": 1.5244180068419783e-06,
+ "loss": 0.28116235733032224,
+ "num_input_tokens_seen": 967308280,
+ "step": 22120,
+ "train_runtime": 171077.0397,
+ "train_tokens_per_second": 5654.226
+ },
+ {
+ "epoch": 1.4896983571236198,
+ "grad_norm": 0.8657428518226007,
+ "learning_rate": 1.5225176696252453e-06,
+ "loss": 0.32404913902282717,
+ "num_input_tokens_seen": 967517656,
+ "step": 22125,
+ "train_runtime": 171124.2522,
+ "train_tokens_per_second": 5653.89
+ },
+ {
+ "epoch": 1.4900350121195798,
+ "grad_norm": 0.7792695604281658,
+ "learning_rate": 1.5206183048753575e-06,
+ "loss": 0.30272302627563474,
+ "num_input_tokens_seen": 967731128,
+ "step": 22130,
+ "train_runtime": 171161.8669,
+ "train_tokens_per_second": 5653.894
+ },
+ {
+ "epoch": 1.49037166711554,
+ "grad_norm": 0.8874127014480637,
+ "learning_rate": 1.5187199131234653e-06,
+ "loss": 0.34148373603820803,
+ "num_input_tokens_seen": 967953912,
+ "step": 22135,
+ "train_runtime": 171197.9915,
+ "train_tokens_per_second": 5654.003
+ },
+ {
+ "epoch": 1.4907083221115,
+ "grad_norm": 0.8158011732743956,
+ "learning_rate": 1.5168224949004484e-06,
+ "loss": 0.2946127653121948,
+ "num_input_tokens_seen": 968163840,
+ "step": 22140,
+ "train_runtime": 171239.0138,
+ "train_tokens_per_second": 5653.874
+ },
+ {
+ "epoch": 1.4910449771074603,
+ "grad_norm": 0.7094644713372431,
+ "learning_rate": 1.514926050736914e-06,
+ "loss": 0.2730919361114502,
+ "num_input_tokens_seen": 968387800,
+ "step": 22145,
+ "train_runtime": 171288.0971,
+ "train_tokens_per_second": 5653.562
+ },
+ {
+ "epoch": 1.4913816321034203,
+ "grad_norm": 0.7887316321901531,
+ "learning_rate": 1.5130305811631945e-06,
+ "loss": 0.3045715570449829,
+ "num_input_tokens_seen": 968613216,
+ "step": 22150,
+ "train_runtime": 171326.4957,
+ "train_tokens_per_second": 5653.61
+ },
+ {
+ "epoch": 1.4917182870993806,
+ "grad_norm": 0.8209012844760429,
+ "learning_rate": 1.5111360867093571e-06,
+ "loss": 0.3100686550140381,
+ "num_input_tokens_seen": 968837880,
+ "step": 22155,
+ "train_runtime": 171360.4409,
+ "train_tokens_per_second": 5653.801
+ },
+ {
+ "epoch": 1.4920549420953406,
+ "grad_norm": 0.7096360424168036,
+ "learning_rate": 1.5092425679051848e-06,
+ "loss": 0.2817687511444092,
+ "num_input_tokens_seen": 969060560,
+ "step": 22160,
+ "train_runtime": 171394.8343,
+ "train_tokens_per_second": 5653.966
+ },
+ {
+ "epoch": 1.4923915970913009,
+ "grad_norm": 0.74898568213781,
+ "learning_rate": 1.5073500252801988e-06,
+ "loss": 0.31572632789611815,
+ "num_input_tokens_seen": 969287272,
+ "step": 22165,
+ "train_runtime": 171423.1343,
+ "train_tokens_per_second": 5654.355
+ },
+ {
+ "epoch": 1.492728252087261,
+ "grad_norm": 0.7194688541867551,
+ "learning_rate": 1.5054584593636412e-06,
+ "loss": 0.2992145776748657,
+ "num_input_tokens_seen": 969517872,
+ "step": 22170,
+ "train_runtime": 171459.1961,
+ "train_tokens_per_second": 5654.511
+ },
+ {
+ "epoch": 1.4930649070832211,
+ "grad_norm": 0.8380897247829227,
+ "learning_rate": 1.5035678706844814e-06,
+ "loss": 0.3096607208251953,
+ "num_input_tokens_seen": 969734072,
+ "step": 22175,
+ "train_runtime": 171496.6903,
+ "train_tokens_per_second": 5654.535
+ },
+ {
+ "epoch": 1.4934015620791812,
+ "grad_norm": 0.719052290185219,
+ "learning_rate": 1.501678259771418e-06,
+ "loss": 0.311397123336792,
+ "num_input_tokens_seen": 969956280,
+ "step": 22180,
+ "train_runtime": 171541.9334,
+ "train_tokens_per_second": 5654.339
+ },
+ {
+ "epoch": 1.4937382170751414,
+ "grad_norm": 0.8005847376324827,
+ "learning_rate": 1.499789627152874e-06,
+ "loss": 0.3107875108718872,
+ "num_input_tokens_seen": 970180640,
+ "step": 22185,
+ "train_runtime": 171584.1364,
+ "train_tokens_per_second": 5654.256
+ },
+ {
+ "epoch": 1.4940748720711015,
+ "grad_norm": 0.7488870093377558,
+ "learning_rate": 1.497901973356999e-06,
+ "loss": 0.26401643753051757,
+ "num_input_tokens_seen": 970405336,
+ "step": 22190,
+ "train_runtime": 171621.5048,
+ "train_tokens_per_second": 5654.334
+ },
+ {
+ "epoch": 1.4944115270670617,
+ "grad_norm": 0.6586409758641915,
+ "learning_rate": 1.4960152989116694e-06,
+ "loss": 0.29029550552368166,
+ "num_input_tokens_seen": 970628528,
+ "step": 22195,
+ "train_runtime": 171659.7512,
+ "train_tokens_per_second": 5654.375
+ },
+ {
+ "epoch": 1.4947481820630217,
+ "grad_norm": 0.7846315012803541,
+ "learning_rate": 1.4941296043444869e-06,
+ "loss": 0.31101787090301514,
+ "num_input_tokens_seen": 970833256,
+ "step": 22200,
+ "train_runtime": 171700.279,
+ "train_tokens_per_second": 5654.232
+ },
+ {
+ "epoch": 1.495084837058982,
+ "grad_norm": 0.6807750672867574,
+ "learning_rate": 1.4922448901827847e-06,
+ "loss": 0.2868825912475586,
+ "num_input_tokens_seen": 971056192,
+ "step": 22205,
+ "train_runtime": 171746.7294,
+ "train_tokens_per_second": 5654.001
+ },
+ {
+ "epoch": 1.495421492054942,
+ "grad_norm": 0.9017046482285462,
+ "learning_rate": 1.4903611569536103e-06,
+ "loss": 0.27285065650939944,
+ "num_input_tokens_seen": 971251664,
+ "step": 22210,
+ "train_runtime": 171784.3198,
+ "train_tokens_per_second": 5653.902
+ },
+ {
+ "epoch": 1.4957581470509023,
+ "grad_norm": 0.7546903370960244,
+ "learning_rate": 1.488478405183748e-06,
+ "loss": 0.3076265096664429,
+ "num_input_tokens_seen": 971461824,
+ "step": 22215,
+ "train_runtime": 171819.829,
+ "train_tokens_per_second": 5653.956
+ },
+ {
+ "epoch": 1.4960948020468623,
+ "grad_norm": 0.7286152762441251,
+ "learning_rate": 1.4865966353997024e-06,
+ "loss": 0.2991771221160889,
+ "num_input_tokens_seen": 971691984,
+ "step": 22220,
+ "train_runtime": 171853.9282,
+ "train_tokens_per_second": 5654.174
+ },
+ {
+ "epoch": 1.4964314570428225,
+ "grad_norm": 0.6811343761957527,
+ "learning_rate": 1.4847158481277047e-06,
+ "loss": 0.27810006141662597,
+ "num_input_tokens_seen": 971910752,
+ "step": 22225,
+ "train_runtime": 171892.8709,
+ "train_tokens_per_second": 5654.166
+ },
+ {
+ "epoch": 1.4967681120387826,
+ "grad_norm": 0.8778824220267035,
+ "learning_rate": 1.48283604389371e-06,
+ "loss": 0.274126935005188,
+ "num_input_tokens_seen": 972116168,
+ "step": 22230,
+ "train_runtime": 171929.1838,
+ "train_tokens_per_second": 5654.166
+ },
+ {
+ "epoch": 1.4971047670347428,
+ "grad_norm": 0.6703777252052292,
+ "learning_rate": 1.480957223223401e-06,
+ "loss": 0.28500995635986326,
+ "num_input_tokens_seen": 972336272,
+ "step": 22235,
+ "train_runtime": 171966.7609,
+ "train_tokens_per_second": 5654.211
+ },
+ {
+ "epoch": 1.497441422030703,
+ "grad_norm": 0.8037319519905057,
+ "learning_rate": 1.4790793866421826e-06,
+ "loss": 0.32386600971221924,
+ "num_input_tokens_seen": 972548536,
+ "step": 22240,
+ "train_runtime": 172003.2466,
+ "train_tokens_per_second": 5654.245
+ },
+ {
+ "epoch": 1.497778077026663,
+ "grad_norm": 0.8609976203519863,
+ "learning_rate": 1.4772025346751862e-06,
+ "loss": 0.30510640144348145,
+ "num_input_tokens_seen": 972767336,
+ "step": 22245,
+ "train_runtime": 172035.4661,
+ "train_tokens_per_second": 5654.458
+ },
+ {
+ "epoch": 1.4981147320226231,
+ "grad_norm": 0.6833659788420334,
+ "learning_rate": 1.475326667847266e-06,
+ "loss": 0.3076331615447998,
+ "num_input_tokens_seen": 972987240,
+ "step": 22250,
+ "train_runtime": 172073.245,
+ "train_tokens_per_second": 5654.495
+ },
+ {
+ "epoch": 1.4984513870185834,
+ "grad_norm": 0.7212227153724603,
+ "learning_rate": 1.4734517866830073e-06,
+ "loss": 0.30481748580932616,
+ "num_input_tokens_seen": 973215336,
+ "step": 22255,
+ "train_runtime": 172107.067,
+ "train_tokens_per_second": 5654.709
+ },
+ {
+ "epoch": 1.4987880420145436,
+ "grad_norm": 0.7269815306579177,
+ "learning_rate": 1.4715778917067081e-06,
+ "loss": 0.303895378112793,
+ "num_input_tokens_seen": 973425600,
+ "step": 22260,
+ "train_runtime": 172149.3633,
+ "train_tokens_per_second": 5654.541
+ },
+ {
+ "epoch": 1.4991246970105037,
+ "grad_norm": 0.6350255641888927,
+ "learning_rate": 1.4697049834424016e-06,
+ "loss": 0.29153730869293215,
+ "num_input_tokens_seen": 973643560,
+ "step": 22265,
+ "train_runtime": 172188.7254,
+ "train_tokens_per_second": 5654.514
+ },
+ {
+ "epoch": 1.4994613520064637,
+ "grad_norm": 0.9326241952838099,
+ "learning_rate": 1.4678330624138387e-06,
+ "loss": 0.285983943939209,
+ "num_input_tokens_seen": 973854824,
+ "step": 22270,
+ "train_runtime": 172221.8921,
+ "train_tokens_per_second": 5654.652
+ },
+ {
+ "epoch": 1.499798007002424,
+ "grad_norm": 0.6824859115358656,
+ "learning_rate": 1.465962129144497e-06,
+ "loss": 0.2954267978668213,
+ "num_input_tokens_seen": 974073608,
+ "step": 22275,
+ "train_runtime": 172256.0105,
+ "train_tokens_per_second": 5654.802
+ },
+ {
+ "epoch": 1.5001346619983842,
+ "grad_norm": 0.8186629815319795,
+ "learning_rate": 1.4640921841575766e-06,
+ "loss": 0.32407686710357664,
+ "num_input_tokens_seen": 974307744,
+ "step": 22280,
+ "train_runtime": 172292.4565,
+ "train_tokens_per_second": 5654.965
+ },
+ {
+ "epoch": 1.5004713169943442,
+ "grad_norm": 0.6972447701320164,
+ "learning_rate": 1.4622232279760018e-06,
+ "loss": 0.29003033638000486,
+ "num_input_tokens_seen": 974527448,
+ "step": 22285,
+ "train_runtime": 172326.5609,
+ "train_tokens_per_second": 5655.12
+ },
+ {
+ "epoch": 1.5008079719903042,
+ "grad_norm": 0.8003233393761351,
+ "learning_rate": 1.4603552611224202e-06,
+ "loss": 0.3260036468505859,
+ "num_input_tokens_seen": 974732192,
+ "step": 22290,
+ "train_runtime": 172364.9722,
+ "train_tokens_per_second": 5655.048
+ },
+ {
+ "epoch": 1.5011446269862645,
+ "grad_norm": 0.7801634179627978,
+ "learning_rate": 1.4584882841192027e-06,
+ "loss": 0.3202426195144653,
+ "num_input_tokens_seen": 974950184,
+ "step": 22295,
+ "train_runtime": 172402.4948,
+ "train_tokens_per_second": 5655.082
+ },
+ {
+ "epoch": 1.5014812819822247,
+ "grad_norm": 0.7560222588514484,
+ "learning_rate": 1.456622297488442e-06,
+ "loss": 0.2875128030776978,
+ "num_input_tokens_seen": 975160280,
+ "step": 22300,
+ "train_runtime": 172436.8178,
+ "train_tokens_per_second": 5655.174
+ },
+ {
+ "epoch": 1.5018179369781848,
+ "grad_norm": 0.8392814865686347,
+ "learning_rate": 1.4547573017519595e-06,
+ "loss": 0.29420223236083987,
+ "num_input_tokens_seen": 975368840,
+ "step": 22305,
+ "train_runtime": 172476.8287,
+ "train_tokens_per_second": 5655.072
+ },
+ {
+ "epoch": 1.5021545919741448,
+ "grad_norm": 0.7889313036819893,
+ "learning_rate": 1.4528932974312897e-06,
+ "loss": 0.28504467010498047,
+ "num_input_tokens_seen": 975600768,
+ "step": 22310,
+ "train_runtime": 172519.0021,
+ "train_tokens_per_second": 5655.034
+ },
+ {
+ "epoch": 1.502491246970105,
+ "grad_norm": 0.8405102432636552,
+ "learning_rate": 1.4510302850477005e-06,
+ "loss": 0.29707143306732176,
+ "num_input_tokens_seen": 975818888,
+ "step": 22315,
+ "train_runtime": 172557.3322,
+ "train_tokens_per_second": 5655.042
+ },
+ {
+ "epoch": 1.5028279019660653,
+ "grad_norm": 0.7839583808405233,
+ "learning_rate": 1.4491682651221746e-06,
+ "loss": 0.289170503616333,
+ "num_input_tokens_seen": 976024992,
+ "step": 22320,
+ "train_runtime": 172593.7247,
+ "train_tokens_per_second": 5655.043
+ },
+ {
+ "epoch": 1.5031645569620253,
+ "grad_norm": 0.7899317516379861,
+ "learning_rate": 1.447307238175421e-06,
+ "loss": 0.298907732963562,
+ "num_input_tokens_seen": 976231792,
+ "step": 22325,
+ "train_runtime": 172634.9012,
+ "train_tokens_per_second": 5654.892
+ },
+ {
+ "epoch": 1.5035012119579854,
+ "grad_norm": 0.7367720389125396,
+ "learning_rate": 1.4454472047278688e-06,
+ "loss": 0.29713735580444334,
+ "num_input_tokens_seen": 976442768,
+ "step": 22330,
+ "train_runtime": 172676.2447,
+ "train_tokens_per_second": 5654.76
+ },
+ {
+ "epoch": 1.5038378669539456,
+ "grad_norm": 0.7850128485368816,
+ "learning_rate": 1.443588165299672e-06,
+ "loss": 0.2721333265304565,
+ "num_input_tokens_seen": 976670000,
+ "step": 22335,
+ "train_runtime": 172721.7292,
+ "train_tokens_per_second": 5654.587
+ },
+ {
+ "epoch": 1.5041745219499059,
+ "grad_norm": 0.7958028090619667,
+ "learning_rate": 1.4417301204107037e-06,
+ "loss": 0.31305713653564454,
+ "num_input_tokens_seen": 976892192,
+ "step": 22340,
+ "train_runtime": 172757.5302,
+ "train_tokens_per_second": 5654.701
+ },
+ {
+ "epoch": 1.504511176945866,
+ "grad_norm": 0.7865067258558222,
+ "learning_rate": 1.4398730705805607e-06,
+ "loss": 0.3005058288574219,
+ "num_input_tokens_seen": 977116432,
+ "step": 22345,
+ "train_runtime": 172795.4681,
+ "train_tokens_per_second": 5654.757
+ },
+ {
+ "epoch": 1.504847831941826,
+ "grad_norm": 0.6971438552549611,
+ "learning_rate": 1.43801701632856e-06,
+ "loss": 0.3000321388244629,
+ "num_input_tokens_seen": 977343248,
+ "step": 22350,
+ "train_runtime": 172831.4992,
+ "train_tokens_per_second": 5654.891
+ },
+ {
+ "epoch": 1.5051844869377862,
+ "grad_norm": 0.7852554320922348,
+ "learning_rate": 1.436161958173745e-06,
+ "loss": 0.32420189380645753,
+ "num_input_tokens_seen": 977567152,
+ "step": 22355,
+ "train_runtime": 172868.2683,
+ "train_tokens_per_second": 5654.983
+ },
+ {
+ "epoch": 1.5055211419337464,
+ "grad_norm": 0.7387859896989003,
+ "learning_rate": 1.4343078966348718e-06,
+ "loss": 0.3067188262939453,
+ "num_input_tokens_seen": 977777560,
+ "step": 22360,
+ "train_runtime": 172910.0558,
+ "train_tokens_per_second": 5654.833
+ },
+ {
+ "epoch": 1.5058577969297064,
+ "grad_norm": 0.7529457682896904,
+ "learning_rate": 1.432454832230426e-06,
+ "loss": 0.28587291240692136,
+ "num_input_tokens_seen": 978009320,
+ "step": 22365,
+ "train_runtime": 172953.4465,
+ "train_tokens_per_second": 5654.755
+ },
+ {
+ "epoch": 1.5061944519256665,
+ "grad_norm": 0.6792382751676833,
+ "learning_rate": 1.4306027654786103e-06,
+ "loss": 0.28476083278656006,
+ "num_input_tokens_seen": 978236352,
+ "step": 22370,
+ "train_runtime": 172991.2837,
+ "train_tokens_per_second": 5654.83
+ },
+ {
+ "epoch": 1.5065311069216267,
+ "grad_norm": 0.7207520305966407,
+ "learning_rate": 1.4287516968973491e-06,
+ "loss": 0.3085874080657959,
+ "num_input_tokens_seen": 978453896,
+ "step": 22375,
+ "train_runtime": 173028.6858,
+ "train_tokens_per_second": 5654.865
+ },
+ {
+ "epoch": 1.506867761917587,
+ "grad_norm": 0.816106376696343,
+ "learning_rate": 1.4269016270042884e-06,
+ "loss": 0.3164313077926636,
+ "num_input_tokens_seen": 978673624,
+ "step": 22380,
+ "train_runtime": 173069.187,
+ "train_tokens_per_second": 5654.811
+ },
+ {
+ "epoch": 1.507204416913547,
+ "grad_norm": 0.7895982117278676,
+ "learning_rate": 1.425052556316794e-06,
+ "loss": 0.2872413873672485,
+ "num_input_tokens_seen": 978891952,
+ "step": 22385,
+ "train_runtime": 173110.0103,
+ "train_tokens_per_second": 5654.739
+ },
+ {
+ "epoch": 1.507541071909507,
+ "grad_norm": 0.9301659133996552,
+ "learning_rate": 1.4232044853519522e-06,
+ "loss": 0.3223658323287964,
+ "num_input_tokens_seen": 979097208,
+ "step": 22390,
+ "train_runtime": 173149.0058,
+ "train_tokens_per_second": 5654.651
+ },
+ {
+ "epoch": 1.5078777269054673,
+ "grad_norm": 0.6813929579263441,
+ "learning_rate": 1.4213574146265708e-06,
+ "loss": 0.28975491523742675,
+ "num_input_tokens_seen": 979304416,
+ "step": 22395,
+ "train_runtime": 173184.2372,
+ "train_tokens_per_second": 5654.697
+ },
+ {
+ "epoch": 1.5082143819014275,
+ "grad_norm": 0.7646303429188873,
+ "learning_rate": 1.4195113446571752e-06,
+ "loss": 0.32438082695007325,
+ "num_input_tokens_seen": 979527856,
+ "step": 22400,
+ "train_runtime": 173219.7965,
+ "train_tokens_per_second": 5654.826
+ },
+ {
+ "epoch": 1.5085510368973876,
+ "grad_norm": 0.7391053441959036,
+ "learning_rate": 1.417666275960018e-06,
+ "loss": 0.2848972797393799,
+ "num_input_tokens_seen": 979761880,
+ "step": 22405,
+ "train_runtime": 173250.7995,
+ "train_tokens_per_second": 5655.165
+ },
+ {
+ "epoch": 1.5088876918933476,
+ "grad_norm": 0.6903765320662747,
+ "learning_rate": 1.4158222090510605e-06,
+ "loss": 0.27116708755493163,
+ "num_input_tokens_seen": 979990616,
+ "step": 22410,
+ "train_runtime": 173289.5475,
+ "train_tokens_per_second": 5655.221
+ },
+ {
+ "epoch": 1.5092243468893078,
+ "grad_norm": 0.7236456392467262,
+ "learning_rate": 1.4139791444459945e-06,
+ "loss": 0.3101805686950684,
+ "num_input_tokens_seen": 980209928,
+ "step": 22415,
+ "train_runtime": 173323.1137,
+ "train_tokens_per_second": 5655.391
+ },
+ {
+ "epoch": 1.509561001885268,
+ "grad_norm": 0.8526682680286438,
+ "learning_rate": 1.4121370826602254e-06,
+ "loss": 0.3069303512573242,
+ "num_input_tokens_seen": 980417304,
+ "step": 22420,
+ "train_runtime": 173363.0188,
+ "train_tokens_per_second": 5655.285
+ },
+ {
+ "epoch": 1.5098976568812281,
+ "grad_norm": 0.7042288015488988,
+ "learning_rate": 1.4102960242088803e-06,
+ "loss": 0.3054825782775879,
+ "num_input_tokens_seen": 980633968,
+ "step": 22425,
+ "train_runtime": 173407.4082,
+ "train_tokens_per_second": 5655.087
+ },
+ {
+ "epoch": 1.5102343118771882,
+ "grad_norm": 0.7504743822892517,
+ "learning_rate": 1.4084559696068057e-06,
+ "loss": 0.28913288116455077,
+ "num_input_tokens_seen": 980850136,
+ "step": 22430,
+ "train_runtime": 173445.4607,
+ "train_tokens_per_second": 5655.093
+ },
+ {
+ "epoch": 1.5105709668731484,
+ "grad_norm": 0.7252926106988145,
+ "learning_rate": 1.406616919368567e-06,
+ "loss": 0.2985984325408936,
+ "num_input_tokens_seen": 981066808,
+ "step": 22435,
+ "train_runtime": 173488.0196,
+ "train_tokens_per_second": 5654.954
+ },
+ {
+ "epoch": 1.5109076218691087,
+ "grad_norm": 0.801327813808047,
+ "learning_rate": 1.4047788740084474e-06,
+ "loss": 0.29711313247680665,
+ "num_input_tokens_seen": 981290632,
+ "step": 22440,
+ "train_runtime": 173530.0228,
+ "train_tokens_per_second": 5654.875
+ },
+ {
+ "epoch": 1.5112442768650687,
+ "grad_norm": 0.6984253644671601,
+ "learning_rate": 1.402941834040452e-06,
+ "loss": 0.2857701539993286,
+ "num_input_tokens_seen": 981524520,
+ "step": 22445,
+ "train_runtime": 173571.2058,
+ "train_tokens_per_second": 5654.881
+ },
+ {
+ "epoch": 1.5115809318610287,
+ "grad_norm": 0.7193805868113159,
+ "learning_rate": 1.4011057999783006e-06,
+ "loss": 0.28029732704162597,
+ "num_input_tokens_seen": 981733352,
+ "step": 22450,
+ "train_runtime": 173613.2142,
+ "train_tokens_per_second": 5654.716
+ },
+ {
+ "epoch": 1.511917586856989,
+ "grad_norm": 0.7608894196010775,
+ "learning_rate": 1.3992707723354398e-06,
+ "loss": 0.3050365924835205,
+ "num_input_tokens_seen": 981955216,
+ "step": 22455,
+ "train_runtime": 173651.5688,
+ "train_tokens_per_second": 5654.744
+ },
+ {
+ "epoch": 1.5122542418529492,
+ "grad_norm": 0.7183358803573994,
+ "learning_rate": 1.397436751625022e-06,
+ "loss": 0.2846760988235474,
+ "num_input_tokens_seen": 982166184,
+ "step": 22460,
+ "train_runtime": 173692.0688,
+ "train_tokens_per_second": 5654.64
+ },
+ {
+ "epoch": 1.5125908968489092,
+ "grad_norm": 0.8684762367144154,
+ "learning_rate": 1.3956037383599308e-06,
+ "loss": 0.2913294792175293,
+ "num_input_tokens_seen": 982385952,
+ "step": 22465,
+ "train_runtime": 173731.1563,
+ "train_tokens_per_second": 5654.633
+ },
+ {
+ "epoch": 1.5129275518448693,
+ "grad_norm": 0.7292786314768155,
+ "learning_rate": 1.39377173305276e-06,
+ "loss": 0.27708570957183837,
+ "num_input_tokens_seen": 982592624,
+ "step": 22470,
+ "train_runtime": 173769.1128,
+ "train_tokens_per_second": 5654.587
+ },
+ {
+ "epoch": 1.5132642068408295,
+ "grad_norm": 0.7362638583219614,
+ "learning_rate": 1.3919407362158254e-06,
+ "loss": 0.3209893465042114,
+ "num_input_tokens_seen": 982825112,
+ "step": 22475,
+ "train_runtime": 173806.2328,
+ "train_tokens_per_second": 5654.717
+ },
+ {
+ "epoch": 1.5136008618367898,
+ "grad_norm": 0.8685451291862974,
+ "learning_rate": 1.3901107483611582e-06,
+ "loss": 0.2871732711791992,
+ "num_input_tokens_seen": 983031648,
+ "step": 22480,
+ "train_runtime": 173844.6509,
+ "train_tokens_per_second": 5654.656
+ },
+ {
+ "epoch": 1.5139375168327498,
+ "grad_norm": 0.7432181034545295,
+ "learning_rate": 1.3882817700005091e-06,
+ "loss": 0.2984759330749512,
+ "num_input_tokens_seen": 983247128,
+ "step": 22485,
+ "train_runtime": 173892.4519,
+ "train_tokens_per_second": 5654.34
+ },
+ {
+ "epoch": 1.5142741718287098,
+ "grad_norm": 0.7851212386445382,
+ "learning_rate": 1.3864538016453466e-06,
+ "loss": 0.3003609418869019,
+ "num_input_tokens_seen": 983445264,
+ "step": 22490,
+ "train_runtime": 173929.1801,
+ "train_tokens_per_second": 5654.286
+ },
+ {
+ "epoch": 1.51461082682467,
+ "grad_norm": 0.929685801907135,
+ "learning_rate": 1.384626843806855e-06,
+ "loss": 0.304855489730835,
+ "num_input_tokens_seen": 983646472,
+ "step": 22495,
+ "train_runtime": 173969.1302,
+ "train_tokens_per_second": 5654.144
+ },
+ {
+ "epoch": 1.5149474818206303,
+ "grad_norm": 0.7557647415601482,
+ "learning_rate": 1.3828008969959367e-06,
+ "loss": 0.31099758148193357,
+ "num_input_tokens_seen": 983875960,
+ "step": 22500,
+ "train_runtime": 174009.2793,
+ "train_tokens_per_second": 5654.158
+ },
+ {
+ "epoch": 1.5152841368165904,
+ "grad_norm": 0.850008061053931,
+ "learning_rate": 1.3809759617232166e-06,
+ "loss": 0.2885129928588867,
+ "num_input_tokens_seen": 984089952,
+ "step": 22505,
+ "train_runtime": 174050.4492,
+ "train_tokens_per_second": 5654.05
+ },
+ {
+ "epoch": 1.5156207918125504,
+ "grad_norm": 0.7124328658676645,
+ "learning_rate": 1.3791520384990247e-06,
+ "loss": 0.29050498008728026,
+ "num_input_tokens_seen": 984303984,
+ "step": 22510,
+ "train_runtime": 174085.4118,
+ "train_tokens_per_second": 5654.144
+ },
+ {
+ "epoch": 1.5159574468085106,
+ "grad_norm": 0.6783534298885123,
+ "learning_rate": 1.3773291278334216e-06,
+ "loss": 0.2980940818786621,
+ "num_input_tokens_seen": 984518832,
+ "step": 22515,
+ "train_runtime": 174122.6841,
+ "train_tokens_per_second": 5654.168
+ },
+ {
+ "epoch": 1.5162941018044709,
+ "grad_norm": 0.772415132712806,
+ "learning_rate": 1.3755072302361754e-06,
+ "loss": 0.3086568355560303,
+ "num_input_tokens_seen": 984751312,
+ "step": 22520,
+ "train_runtime": 174163.3837,
+ "train_tokens_per_second": 5654.181
+ },
+ {
+ "epoch": 1.516630756800431,
+ "grad_norm": 0.7777502241934342,
+ "learning_rate": 1.3736863462167749e-06,
+ "loss": 0.28640379905700686,
+ "num_input_tokens_seen": 984953136,
+ "step": 22525,
+ "train_runtime": 174203.7264,
+ "train_tokens_per_second": 5654.03
+ },
+ {
+ "epoch": 1.516967411796391,
+ "grad_norm": 0.7811964811897554,
+ "learning_rate": 1.3718664762844242e-06,
+ "loss": 0.28820300102233887,
+ "num_input_tokens_seen": 985162064,
+ "step": 22530,
+ "train_runtime": 174238.0664,
+ "train_tokens_per_second": 5654.115
+ },
+ {
+ "epoch": 1.5173040667923512,
+ "grad_norm": 0.7667814818121892,
+ "learning_rate": 1.370047620948044e-06,
+ "loss": 0.2942743062973022,
+ "num_input_tokens_seen": 985382480,
+ "step": 22535,
+ "train_runtime": 174279.1667,
+ "train_tokens_per_second": 5654.046
+ },
+ {
+ "epoch": 1.5176407217883114,
+ "grad_norm": 0.7745949128302583,
+ "learning_rate": 1.368229780716272e-06,
+ "loss": 0.3100618839263916,
+ "num_input_tokens_seen": 985602688,
+ "step": 22540,
+ "train_runtime": 174310.4991,
+ "train_tokens_per_second": 5654.293
+ },
+ {
+ "epoch": 1.5179773767842715,
+ "grad_norm": 0.8932217604026287,
+ "learning_rate": 1.3664129560974604e-06,
+ "loss": 0.2899731397628784,
+ "num_input_tokens_seen": 985820032,
+ "step": 22545,
+ "train_runtime": 174346.7623,
+ "train_tokens_per_second": 5654.364
+ },
+ {
+ "epoch": 1.5183140317802315,
+ "grad_norm": 0.7820206063355695,
+ "learning_rate": 1.3645971475996777e-06,
+ "loss": 0.29310288429260256,
+ "num_input_tokens_seen": 986025552,
+ "step": 22550,
+ "train_runtime": 174382.5794,
+ "train_tokens_per_second": 5654.381
+ },
+ {
+ "epoch": 1.5186506867761917,
+ "grad_norm": 0.8104637399237582,
+ "learning_rate": 1.3627823557307135e-06,
+ "loss": 0.29608545303344724,
+ "num_input_tokens_seen": 986257600,
+ "step": 22555,
+ "train_runtime": 174424.2399,
+ "train_tokens_per_second": 5654.361
+ },
+ {
+ "epoch": 1.518987341772152,
+ "grad_norm": 0.7046023346650935,
+ "learning_rate": 1.3609685809980616e-06,
+ "loss": 0.3058366298675537,
+ "num_input_tokens_seen": 986486440,
+ "step": 22560,
+ "train_runtime": 174466.7289,
+ "train_tokens_per_second": 5654.295
+ },
+ {
+ "epoch": 1.519323996768112,
+ "grad_norm": 0.7418847854253624,
+ "learning_rate": 1.3591558239089443e-06,
+ "loss": 0.2807319164276123,
+ "num_input_tokens_seen": 986707576,
+ "step": 22565,
+ "train_runtime": 174512.4598,
+ "train_tokens_per_second": 5654.081
+ },
+ {
+ "epoch": 1.519660651764072,
+ "grad_norm": 0.8450682344923751,
+ "learning_rate": 1.3573440849702902e-06,
+ "loss": 0.317004919052124,
+ "num_input_tokens_seen": 986937912,
+ "step": 22570,
+ "train_runtime": 174554.2288,
+ "train_tokens_per_second": 5654.048
+ },
+ {
+ "epoch": 1.5199973067600323,
+ "grad_norm": 0.7386948536050674,
+ "learning_rate": 1.3555333646887476e-06,
+ "loss": 0.3014618635177612,
+ "num_input_tokens_seen": 987163824,
+ "step": 22575,
+ "train_runtime": 174589.7315,
+ "train_tokens_per_second": 5654.192
+ },
+ {
+ "epoch": 1.5203339617559926,
+ "grad_norm": 0.7377369303219051,
+ "learning_rate": 1.3537236635706779e-06,
+ "loss": 0.29477739334106445,
+ "num_input_tokens_seen": 987396616,
+ "step": 22580,
+ "train_runtime": 174623.2763,
+ "train_tokens_per_second": 5654.439
+ },
+ {
+ "epoch": 1.5206706167519526,
+ "grad_norm": 0.8578989985964252,
+ "learning_rate": 1.351914982122159e-06,
+ "loss": 0.27982220649719236,
+ "num_input_tokens_seen": 987623048,
+ "step": 22585,
+ "train_runtime": 174659.6001,
+ "train_tokens_per_second": 5654.559
+ },
+ {
+ "epoch": 1.5210072717479126,
+ "grad_norm": 0.7750049777273713,
+ "learning_rate": 1.3501073208489823e-06,
+ "loss": 0.3012231349945068,
+ "num_input_tokens_seen": 987842064,
+ "step": 22590,
+ "train_runtime": 174696.8944,
+ "train_tokens_per_second": 5654.606
+ },
+ {
+ "epoch": 1.5213439267438729,
+ "grad_norm": 0.7792774723783644,
+ "learning_rate": 1.3483006802566546e-06,
+ "loss": 0.29522716999053955,
+ "num_input_tokens_seen": 988065680,
+ "step": 22595,
+ "train_runtime": 174733.3309,
+ "train_tokens_per_second": 5654.706
+ },
+ {
+ "epoch": 1.5216805817398331,
+ "grad_norm": 0.8105588967824191,
+ "learning_rate": 1.3464950608503957e-06,
+ "loss": 0.3093279838562012,
+ "num_input_tokens_seen": 988279680,
+ "step": 22600,
+ "train_runtime": 174772.1153,
+ "train_tokens_per_second": 5654.676
+ },
+ {
+ "epoch": 1.5220172367357931,
+ "grad_norm": 0.8485409938047664,
+ "learning_rate": 1.344690463135146e-06,
+ "loss": 0.30634384155273436,
+ "num_input_tokens_seen": 988495592,
+ "step": 22605,
+ "train_runtime": 174817.3086,
+ "train_tokens_per_second": 5654.449
+ },
+ {
+ "epoch": 1.5223538917317532,
+ "grad_norm": 0.7715064487874997,
+ "learning_rate": 1.3428868876155493e-06,
+ "loss": 0.3291673183441162,
+ "num_input_tokens_seen": 988698000,
+ "step": 22610,
+ "train_runtime": 174853.7041,
+ "train_tokens_per_second": 5654.43
+ },
+ {
+ "epoch": 1.5226905467277134,
+ "grad_norm": 0.6665959654227848,
+ "learning_rate": 1.3410843347959745e-06,
+ "loss": 0.2833380699157715,
+ "num_input_tokens_seen": 988914088,
+ "step": 22615,
+ "train_runtime": 174891.5132,
+ "train_tokens_per_second": 5654.443
+ },
+ {
+ "epoch": 1.5230272017236737,
+ "grad_norm": 0.7178804958210366,
+ "learning_rate": 1.3392828051804979e-06,
+ "loss": 0.3043062686920166,
+ "num_input_tokens_seen": 989141848,
+ "step": 22620,
+ "train_runtime": 174934.5815,
+ "train_tokens_per_second": 5654.353
+ },
+ {
+ "epoch": 1.5233638567196337,
+ "grad_norm": 0.7316276122750418,
+ "learning_rate": 1.3374822992729114e-06,
+ "loss": 0.3176832437515259,
+ "num_input_tokens_seen": 989359560,
+ "step": 22625,
+ "train_runtime": 174977.0909,
+ "train_tokens_per_second": 5654.223
+ },
+ {
+ "epoch": 1.5237005117155937,
+ "grad_norm": 0.8151737151768202,
+ "learning_rate": 1.3356828175767212e-06,
+ "loss": 0.30327634811401366,
+ "num_input_tokens_seen": 989566784,
+ "step": 22630,
+ "train_runtime": 175016.0102,
+ "train_tokens_per_second": 5654.15
+ },
+ {
+ "epoch": 1.524037166711554,
+ "grad_norm": 0.7458938184161533,
+ "learning_rate": 1.3338843605951462e-06,
+ "loss": 0.27550592422485354,
+ "num_input_tokens_seen": 989793688,
+ "step": 22635,
+ "train_runtime": 175058.1324,
+ "train_tokens_per_second": 5654.086
+ },
+ {
+ "epoch": 1.5243738217075142,
+ "grad_norm": 0.7254475228515918,
+ "learning_rate": 1.33208692883112e-06,
+ "loss": 0.2927609205245972,
+ "num_input_tokens_seen": 990030416,
+ "step": 22640,
+ "train_runtime": 175099.372,
+ "train_tokens_per_second": 5654.106
+ },
+ {
+ "epoch": 1.5247104767034743,
+ "grad_norm": 0.6451473463172038,
+ "learning_rate": 1.3302905227872876e-06,
+ "loss": 0.2735952377319336,
+ "num_input_tokens_seen": 990266880,
+ "step": 22645,
+ "train_runtime": 175138.0414,
+ "train_tokens_per_second": 5654.208
+ },
+ {
+ "epoch": 1.5250471316994343,
+ "grad_norm": 0.7523183374053246,
+ "learning_rate": 1.328495142966007e-06,
+ "loss": 0.3053655862808228,
+ "num_input_tokens_seen": 990479584,
+ "step": 22650,
+ "train_runtime": 175176.6743,
+ "train_tokens_per_second": 5654.175
+ },
+ {
+ "epoch": 1.5253837866953945,
+ "grad_norm": 0.7119868993623765,
+ "learning_rate": 1.3267007898693552e-06,
+ "loss": 0.29980125427246096,
+ "num_input_tokens_seen": 990712176,
+ "step": 22655,
+ "train_runtime": 175223.5228,
+ "train_tokens_per_second": 5653.991
+ },
+ {
+ "epoch": 1.5257204416913548,
+ "grad_norm": 0.7695418964162357,
+ "learning_rate": 1.324907463999111e-06,
+ "loss": 0.3055927276611328,
+ "num_input_tokens_seen": 990938448,
+ "step": 22660,
+ "train_runtime": 175261.0683,
+ "train_tokens_per_second": 5654.071
+ },
+ {
+ "epoch": 1.5260570966873148,
+ "grad_norm": 0.750866798684191,
+ "learning_rate": 1.3231151658567776e-06,
+ "loss": 0.3090615510940552,
+ "num_input_tokens_seen": 991152664,
+ "step": 22665,
+ "train_runtime": 175294.7905,
+ "train_tokens_per_second": 5654.205
+ },
+ {
+ "epoch": 1.5263937516832748,
+ "grad_norm": 0.7700423401788759,
+ "learning_rate": 1.321323895943563e-06,
+ "loss": 0.3019981861114502,
+ "num_input_tokens_seen": 991370424,
+ "step": 22670,
+ "train_runtime": 175335.3762,
+ "train_tokens_per_second": 5654.138
+ },
+ {
+ "epoch": 1.526730406679235,
+ "grad_norm": 0.6944631282349509,
+ "learning_rate": 1.3195336547603905e-06,
+ "loss": 0.31381664276123045,
+ "num_input_tokens_seen": 991588192,
+ "step": 22675,
+ "train_runtime": 175378.1626,
+ "train_tokens_per_second": 5654.0
+ },
+ {
+ "epoch": 1.5270670616751953,
+ "grad_norm": 0.711842537575573,
+ "learning_rate": 1.3177444428078951e-06,
+ "loss": 0.28606882095336916,
+ "num_input_tokens_seen": 991798856,
+ "step": 22680,
+ "train_runtime": 175413.7546,
+ "train_tokens_per_second": 5654.054
+ },
+ {
+ "epoch": 1.5274037166711554,
+ "grad_norm": 0.815411302830813,
+ "learning_rate": 1.3159562605864245e-06,
+ "loss": 0.30185737609863283,
+ "num_input_tokens_seen": 992026624,
+ "step": 22685,
+ "train_runtime": 175457.606,
+ "train_tokens_per_second": 5653.939
+ },
+ {
+ "epoch": 1.5277403716671154,
+ "grad_norm": 0.7981498875678553,
+ "learning_rate": 1.3141691085960385e-06,
+ "loss": 0.29338903427124025,
+ "num_input_tokens_seen": 992265864,
+ "step": 22690,
+ "train_runtime": 175496.9155,
+ "train_tokens_per_second": 5654.036
+ },
+ {
+ "epoch": 1.5280770266630757,
+ "grad_norm": 0.7568915202824599,
+ "learning_rate": 1.3123829873365073e-06,
+ "loss": 0.2825155258178711,
+ "num_input_tokens_seen": 992478536,
+ "step": 22695,
+ "train_runtime": 175535.4671,
+ "train_tokens_per_second": 5654.006
+ },
+ {
+ "epoch": 1.528413681659036,
+ "grad_norm": 0.7962079639131174,
+ "learning_rate": 1.3105978973073136e-06,
+ "loss": 0.2889784097671509,
+ "num_input_tokens_seen": 992692016,
+ "step": 22700,
+ "train_runtime": 175577.4777,
+ "train_tokens_per_second": 5653.869
+ },
+ {
+ "epoch": 1.528750336654996,
+ "grad_norm": 0.7986814122905842,
+ "learning_rate": 1.3088138390076565e-06,
+ "loss": 0.3144156694412231,
+ "num_input_tokens_seen": 992915000,
+ "step": 22705,
+ "train_runtime": 175610.667,
+ "train_tokens_per_second": 5654.07
+ },
+ {
+ "epoch": 1.529086991650956,
+ "grad_norm": 0.784101134617306,
+ "learning_rate": 1.3070308129364357e-06,
+ "loss": 0.28853483200073243,
+ "num_input_tokens_seen": 993131448,
+ "step": 22710,
+ "train_runtime": 175653.2502,
+ "train_tokens_per_second": 5653.932
+ },
+ {
+ "epoch": 1.5294236466469162,
+ "grad_norm": 0.8613062066807416,
+ "learning_rate": 1.3052488195922736e-06,
+ "loss": 0.28805036544799806,
+ "num_input_tokens_seen": 993337608,
+ "step": 22715,
+ "train_runtime": 175693.4521,
+ "train_tokens_per_second": 5653.811
+ },
+ {
+ "epoch": 1.5297603016428765,
+ "grad_norm": 0.7554420455835269,
+ "learning_rate": 1.3034678594734972e-06,
+ "loss": 0.2852388620376587,
+ "num_input_tokens_seen": 993560896,
+ "step": 22720,
+ "train_runtime": 175731.737,
+ "train_tokens_per_second": 5653.85
+ },
+ {
+ "epoch": 1.5300969566388365,
+ "grad_norm": 0.8715494762310279,
+ "learning_rate": 1.3016879330781468e-06,
+ "loss": 0.30547394752502444,
+ "num_input_tokens_seen": 993766904,
+ "step": 22725,
+ "train_runtime": 175768.7081,
+ "train_tokens_per_second": 5653.833
+ },
+ {
+ "epoch": 1.5304336116347965,
+ "grad_norm": 0.7476831667764691,
+ "learning_rate": 1.2999090409039728e-06,
+ "loss": 0.29537832736968994,
+ "num_input_tokens_seen": 993988440,
+ "step": 22730,
+ "train_runtime": 175806.2319,
+ "train_tokens_per_second": 5653.886
+ },
+ {
+ "epoch": 1.5307702666307568,
+ "grad_norm": 0.7728242495534967,
+ "learning_rate": 1.2981311834484367e-06,
+ "loss": 0.30858845710754396,
+ "num_input_tokens_seen": 994207440,
+ "step": 22735,
+ "train_runtime": 175851.9938,
+ "train_tokens_per_second": 5653.66
+ },
+ {
+ "epoch": 1.531106921626717,
+ "grad_norm": 0.7739386600524095,
+ "learning_rate": 1.296354361208711e-06,
+ "loss": 0.28605356216430666,
+ "num_input_tokens_seen": 994421664,
+ "step": 22740,
+ "train_runtime": 175891.4348,
+ "train_tokens_per_second": 5653.611
+ },
+ {
+ "epoch": 1.531443576622677,
+ "grad_norm": 0.8012741513918233,
+ "learning_rate": 1.294578574681678e-06,
+ "loss": 0.3216970920562744,
+ "num_input_tokens_seen": 994629248,
+ "step": 22745,
+ "train_runtime": 175928.8973,
+ "train_tokens_per_second": 5653.587
+ },
+ {
+ "epoch": 1.531780231618637,
+ "grad_norm": 0.7817355354702294,
+ "learning_rate": 1.292803824363929e-06,
+ "loss": 0.28469209671020507,
+ "num_input_tokens_seen": 994850920,
+ "step": 22750,
+ "train_runtime": 175967.4254,
+ "train_tokens_per_second": 5653.608
+ },
+ {
+ "epoch": 1.5321168866145973,
+ "grad_norm": 0.767294892049272,
+ "learning_rate": 1.2910301107517726e-06,
+ "loss": 0.3232132434844971,
+ "num_input_tokens_seen": 995074552,
+ "step": 22755,
+ "train_runtime": 176008.5181,
+ "train_tokens_per_second": 5653.559
+ },
+ {
+ "epoch": 1.5324535416105576,
+ "grad_norm": 0.8124734532106183,
+ "learning_rate": 1.2892574343412156e-06,
+ "loss": 0.319903826713562,
+ "num_input_tokens_seen": 995294736,
+ "step": 22760,
+ "train_runtime": 176043.0639,
+ "train_tokens_per_second": 5653.7
+ },
+ {
+ "epoch": 1.5327901966065176,
+ "grad_norm": 0.7450430227623678,
+ "learning_rate": 1.2874857956279858e-06,
+ "loss": 0.3038845777511597,
+ "num_input_tokens_seen": 995505008,
+ "step": 22765,
+ "train_runtime": 176081.1467,
+ "train_tokens_per_second": 5653.672
+ },
+ {
+ "epoch": 1.5331268516024776,
+ "grad_norm": 0.7872978056072006,
+ "learning_rate": 1.2857151951075147e-06,
+ "loss": 0.3022770404815674,
+ "num_input_tokens_seen": 995729520,
+ "step": 22770,
+ "train_runtime": 176120.5406,
+ "train_tokens_per_second": 5653.682
+ },
+ {
+ "epoch": 1.5334635065984379,
+ "grad_norm": 0.7298470919630958,
+ "learning_rate": 1.2839456332749456e-06,
+ "loss": 0.2998102903366089,
+ "num_input_tokens_seen": 995943808,
+ "step": 22775,
+ "train_runtime": 176158.1432,
+ "train_tokens_per_second": 5653.692
+ },
+ {
+ "epoch": 1.5338001615943981,
+ "grad_norm": 0.8031049157762769,
+ "learning_rate": 1.2821771106251308e-06,
+ "loss": 0.2987299919128418,
+ "num_input_tokens_seen": 996160472,
+ "step": 22780,
+ "train_runtime": 176194.5973,
+ "train_tokens_per_second": 5653.752
+ },
+ {
+ "epoch": 1.5341368165903582,
+ "grad_norm": 0.7596275846497683,
+ "learning_rate": 1.2804096276526312e-06,
+ "loss": 0.27775073051452637,
+ "num_input_tokens_seen": 996378168,
+ "step": 22785,
+ "train_runtime": 176231.0577,
+ "train_tokens_per_second": 5653.817
+ },
+ {
+ "epoch": 1.5344734715863182,
+ "grad_norm": 0.8239381536070142,
+ "learning_rate": 1.2786431848517184e-06,
+ "loss": 0.2855050086975098,
+ "num_input_tokens_seen": 996584112,
+ "step": 22790,
+ "train_runtime": 176267.1349,
+ "train_tokens_per_second": 5653.828
+ },
+ {
+ "epoch": 1.5348101265822784,
+ "grad_norm": 0.7589447815362402,
+ "learning_rate": 1.2768777827163725e-06,
+ "loss": 0.29274682998657225,
+ "num_input_tokens_seen": 996797912,
+ "step": 22795,
+ "train_runtime": 176305.2925,
+ "train_tokens_per_second": 5653.817
+ },
+ {
+ "epoch": 1.5351467815782387,
+ "grad_norm": 0.8140404323951491,
+ "learning_rate": 1.2751134217402811e-06,
+ "loss": 0.29671099185943606,
+ "num_input_tokens_seen": 997003488,
+ "step": 22800,
+ "train_runtime": 176347.2741,
+ "train_tokens_per_second": 5653.637
+ },
+ {
+ "epoch": 1.5354834365741987,
+ "grad_norm": 0.7258016591513569,
+ "learning_rate": 1.2733501024168465e-06,
+ "loss": 0.2848902940750122,
+ "num_input_tokens_seen": 997225312,
+ "step": 22805,
+ "train_runtime": 176381.4588,
+ "train_tokens_per_second": 5653.799
+ },
+ {
+ "epoch": 1.5358200915701588,
+ "grad_norm": 0.8298549659043899,
+ "learning_rate": 1.2715878252391695e-06,
+ "loss": 0.30264275074005126,
+ "num_input_tokens_seen": 997430736,
+ "step": 22810,
+ "train_runtime": 176421.3628,
+ "train_tokens_per_second": 5653.685
+ },
+ {
+ "epoch": 1.536156746566119,
+ "grad_norm": 0.7905575392804943,
+ "learning_rate": 1.269826590700069e-06,
+ "loss": 0.3026357412338257,
+ "num_input_tokens_seen": 997650400,
+ "step": 22815,
+ "train_runtime": 176454.177,
+ "train_tokens_per_second": 5653.878
+ },
+ {
+ "epoch": 1.5364934015620793,
+ "grad_norm": 0.7167147078330245,
+ "learning_rate": 1.2680663992920684e-06,
+ "loss": 0.2799959659576416,
+ "num_input_tokens_seen": 997872240,
+ "step": 22820,
+ "train_runtime": 176495.7097,
+ "train_tokens_per_second": 5653.805
+ },
+ {
+ "epoch": 1.5368300565580393,
+ "grad_norm": 0.8394660285508826,
+ "learning_rate": 1.2663072515073988e-06,
+ "loss": 0.30673794746398925,
+ "num_input_tokens_seen": 998084992,
+ "step": 22825,
+ "train_runtime": 176535.1094,
+ "train_tokens_per_second": 5653.748
+ },
+ {
+ "epoch": 1.5371667115539993,
+ "grad_norm": 0.819012798672818,
+ "learning_rate": 1.2645491478380002e-06,
+ "loss": 0.29927687644958495,
+ "num_input_tokens_seen": 998297904,
+ "step": 22830,
+ "train_runtime": 176577.7119,
+ "train_tokens_per_second": 5653.59
+ },
+ {
+ "epoch": 1.5375033665499596,
+ "grad_norm": 0.8389462503007052,
+ "learning_rate": 1.2627920887755212e-06,
+ "loss": 0.3054951190948486,
+ "num_input_tokens_seen": 998508592,
+ "step": 22835,
+ "train_runtime": 176613.3774,
+ "train_tokens_per_second": 5653.641
+ },
+ {
+ "epoch": 1.5378400215459198,
+ "grad_norm": 0.7157688346659142,
+ "learning_rate": 1.261036074811317e-06,
+ "loss": 0.3088068962097168,
+ "num_input_tokens_seen": 998733664,
+ "step": 22840,
+ "train_runtime": 176645.3708,
+ "train_tokens_per_second": 5653.891
+ },
+ {
+ "epoch": 1.5381766765418798,
+ "grad_norm": 0.7683443200543241,
+ "learning_rate": 1.2592811064364524e-06,
+ "loss": 0.2938983917236328,
+ "num_input_tokens_seen": 998954872,
+ "step": 22845,
+ "train_runtime": 176689.3099,
+ "train_tokens_per_second": 5653.737
+ },
+ {
+ "epoch": 1.5385133315378399,
+ "grad_norm": 1.0039919983392351,
+ "learning_rate": 1.2575271841416963e-06,
+ "loss": 0.3196579456329346,
+ "num_input_tokens_seen": 999145416,
+ "step": 22850,
+ "train_runtime": 176727.2813,
+ "train_tokens_per_second": 5653.6
+ },
+ {
+ "epoch": 1.5388499865338001,
+ "grad_norm": 0.7489841192691534,
+ "learning_rate": 1.2557743084175327e-06,
+ "loss": 0.27573988437652586,
+ "num_input_tokens_seen": 999360320,
+ "step": 22855,
+ "train_runtime": 176762.6012,
+ "train_tokens_per_second": 5653.686
+ },
+ {
+ "epoch": 1.5391866415297604,
+ "grad_norm": 0.7473129975200795,
+ "learning_rate": 1.254022479754141e-06,
+ "loss": 0.2842655420303345,
+ "num_input_tokens_seen": 999588440,
+ "step": 22860,
+ "train_runtime": 176797.2452,
+ "train_tokens_per_second": 5653.869
+ },
+ {
+ "epoch": 1.5395232965257204,
+ "grad_norm": 0.7459689277622236,
+ "learning_rate": 1.252271698641419e-06,
+ "loss": 0.28894410133361814,
+ "num_input_tokens_seen": 999804216,
+ "step": 22865,
+ "train_runtime": 176835.9913,
+ "train_tokens_per_second": 5653.85
+ },
+ {
+ "epoch": 1.5398599515216804,
+ "grad_norm": 0.7252321531655411,
+ "learning_rate": 1.2505219655689659e-06,
+ "loss": 0.310650110244751,
+ "num_input_tokens_seen": 1000033680,
+ "step": 22870,
+ "train_runtime": 176868.215,
+ "train_tokens_per_second": 5654.118
+ },
+ {
+ "epoch": 1.5401966065176407,
+ "grad_norm": 0.7797088059008812,
+ "learning_rate": 1.2487732810260877e-06,
+ "loss": 0.3046865701675415,
+ "num_input_tokens_seen": 1000253168,
+ "step": 22875,
+ "train_runtime": 176907.967,
+ "train_tokens_per_second": 5654.088
+ },
+ {
+ "epoch": 1.540533261513601,
+ "grad_norm": 0.6927104452813211,
+ "learning_rate": 1.2470256455018027e-06,
+ "loss": 0.2814844846725464,
+ "num_input_tokens_seen": 1000469584,
+ "step": 22880,
+ "train_runtime": 176948.9419,
+ "train_tokens_per_second": 5654.002
+ },
+ {
+ "epoch": 1.540869916509561,
+ "grad_norm": 0.7567591660729376,
+ "learning_rate": 1.2452790594848257e-06,
+ "loss": 0.28022444248199463,
+ "num_input_tokens_seen": 1000680152,
+ "step": 22885,
+ "train_runtime": 176994.3264,
+ "train_tokens_per_second": 5653.741
+ },
+ {
+ "epoch": 1.541206571505521,
+ "grad_norm": 0.8836138624724967,
+ "learning_rate": 1.2435335234635892e-06,
+ "loss": 0.3237956762313843,
+ "num_input_tokens_seen": 1000898768,
+ "step": 22890,
+ "train_runtime": 177031.9969,
+ "train_tokens_per_second": 5653.773
+ },
+ {
+ "epoch": 1.5415432265014812,
+ "grad_norm": 0.8098774986288102,
+ "learning_rate": 1.2417890379262221e-06,
+ "loss": 0.29699585437774656,
+ "num_input_tokens_seen": 1001118440,
+ "step": 22895,
+ "train_runtime": 177077.1356,
+ "train_tokens_per_second": 5653.573
+ },
+ {
+ "epoch": 1.5418798814974415,
+ "grad_norm": 0.7965414488198174,
+ "learning_rate": 1.2400456033605672e-06,
+ "loss": 0.26516332626342776,
+ "num_input_tokens_seen": 1001331680,
+ "step": 22900,
+ "train_runtime": 177118.7589,
+ "train_tokens_per_second": 5653.448
+ },
+ {
+ "epoch": 1.5422165364934015,
+ "grad_norm": 0.6817693862207875,
+ "learning_rate": 1.2383032202541706e-06,
+ "loss": 0.3010143756866455,
+ "num_input_tokens_seen": 1001546592,
+ "step": 22905,
+ "train_runtime": 177156.0589,
+ "train_tokens_per_second": 5653.471
+ },
+ {
+ "epoch": 1.5425531914893615,
+ "grad_norm": 0.7874018271811691,
+ "learning_rate": 1.2365618890942827e-06,
+ "loss": 0.32308430671691896,
+ "num_input_tokens_seen": 1001766568,
+ "step": 22910,
+ "train_runtime": 177191.6126,
+ "train_tokens_per_second": 5653.578
+ },
+ {
+ "epoch": 1.5428898464853218,
+ "grad_norm": 0.7366938406257898,
+ "learning_rate": 1.2348216103678618e-06,
+ "loss": 0.3221110820770264,
+ "num_input_tokens_seen": 1001986120,
+ "step": 22915,
+ "train_runtime": 177222.995,
+ "train_tokens_per_second": 5653.816
+ },
+ {
+ "epoch": 1.543226501481282,
+ "grad_norm": 0.8883683471081493,
+ "learning_rate": 1.2330823845615713e-06,
+ "loss": 0.2979999780654907,
+ "num_input_tokens_seen": 1002184920,
+ "step": 22920,
+ "train_runtime": 177260.8843,
+ "train_tokens_per_second": 5653.729
+ },
+ {
+ "epoch": 1.543563156477242,
+ "grad_norm": 0.7325915491270013,
+ "learning_rate": 1.2313442121617785e-06,
+ "loss": 0.29711506366729734,
+ "num_input_tokens_seen": 1002416768,
+ "step": 22925,
+ "train_runtime": 177301.116,
+ "train_tokens_per_second": 5653.753
+ },
+ {
+ "epoch": 1.543899811473202,
+ "grad_norm": 0.7697827606669048,
+ "learning_rate": 1.229607093654563e-06,
+ "loss": 0.2836009502410889,
+ "num_input_tokens_seen": 1002644952,
+ "step": 22930,
+ "train_runtime": 177333.8735,
+ "train_tokens_per_second": 5653.996
+ },
+ {
+ "epoch": 1.5442364664691623,
+ "grad_norm": 0.6685394671070748,
+ "learning_rate": 1.2278710295256974e-06,
+ "loss": 0.30063090324401853,
+ "num_input_tokens_seen": 1002875328,
+ "step": 22935,
+ "train_runtime": 177368.0874,
+ "train_tokens_per_second": 5654.204
+ },
+ {
+ "epoch": 1.5445731214651226,
+ "grad_norm": 0.707781339272189,
+ "learning_rate": 1.2261360202606725e-06,
+ "loss": 0.2941617965698242,
+ "num_input_tokens_seen": 1003108632,
+ "step": 22940,
+ "train_runtime": 177406.2713,
+ "train_tokens_per_second": 5654.302
+ },
+ {
+ "epoch": 1.5449097764610826,
+ "grad_norm": 0.7708196807094642,
+ "learning_rate": 1.2244020663446731e-06,
+ "loss": 0.3176748752593994,
+ "num_input_tokens_seen": 1003320896,
+ "step": 22945,
+ "train_runtime": 177449.0984,
+ "train_tokens_per_second": 5654.134
+ },
+ {
+ "epoch": 1.5452464314570427,
+ "grad_norm": 0.9256755732496396,
+ "learning_rate": 1.222669168262598e-06,
+ "loss": 0.29861812591552733,
+ "num_input_tokens_seen": 1003542056,
+ "step": 22950,
+ "train_runtime": 177486.5774,
+ "train_tokens_per_second": 5654.186
+ },
+ {
+ "epoch": 1.545583086453003,
+ "grad_norm": 0.7130093827585298,
+ "learning_rate": 1.2209373264990448e-06,
+ "loss": 0.29040777683258057,
+ "num_input_tokens_seen": 1003762224,
+ "step": 22955,
+ "train_runtime": 177519.7915,
+ "train_tokens_per_second": 5654.368
+ },
+ {
+ "epoch": 1.5459197414489632,
+ "grad_norm": 0.807014861280199,
+ "learning_rate": 1.2192065415383176e-06,
+ "loss": 0.29322617053985595,
+ "num_input_tokens_seen": 1003976720,
+ "step": 22960,
+ "train_runtime": 177559.4075,
+ "train_tokens_per_second": 5654.314
+ },
+ {
+ "epoch": 1.5462563964449232,
+ "grad_norm": 0.7816291079417402,
+ "learning_rate": 1.217476813864425e-06,
+ "loss": 0.3234560966491699,
+ "num_input_tokens_seen": 1004209864,
+ "step": 22965,
+ "train_runtime": 177604.7394,
+ "train_tokens_per_second": 5654.184
+ },
+ {
+ "epoch": 1.5465930514408834,
+ "grad_norm": 0.7404020286277221,
+ "learning_rate": 1.2157481439610801e-06,
+ "loss": 0.27009384632110595,
+ "num_input_tokens_seen": 1004428792,
+ "step": 22970,
+ "train_runtime": 177647.5783,
+ "train_tokens_per_second": 5654.053
+ },
+ {
+ "epoch": 1.5469297064368437,
+ "grad_norm": 0.8293778191938409,
+ "learning_rate": 1.2140205323116976e-06,
+ "loss": 0.2998237609863281,
+ "num_input_tokens_seen": 1004626616,
+ "step": 22975,
+ "train_runtime": 177682.0158,
+ "train_tokens_per_second": 5654.07
+ },
+ {
+ "epoch": 1.5472663614328037,
+ "grad_norm": 0.8562313154245355,
+ "learning_rate": 1.2122939793994043e-06,
+ "loss": 0.2867993116378784,
+ "num_input_tokens_seen": 1004841312,
+ "step": 22980,
+ "train_runtime": 177716.4916,
+ "train_tokens_per_second": 5654.182
+ },
+ {
+ "epoch": 1.5476030164287637,
+ "grad_norm": 0.7604165950877463,
+ "learning_rate": 1.2105684857070182e-06,
+ "loss": 0.292344331741333,
+ "num_input_tokens_seen": 1005062880,
+ "step": 22985,
+ "train_runtime": 177754.2016,
+ "train_tokens_per_second": 5654.229
+ },
+ {
+ "epoch": 1.547939671424724,
+ "grad_norm": 0.7094120408682225,
+ "learning_rate": 1.2088440517170729e-06,
+ "loss": 0.30412988662719725,
+ "num_input_tokens_seen": 1005287360,
+ "step": 22990,
+ "train_runtime": 177790.0242,
+ "train_tokens_per_second": 5654.352
+ },
+ {
+ "epoch": 1.5482763264206842,
+ "grad_norm": 0.8682158123872002,
+ "learning_rate": 1.2071206779117994e-06,
+ "loss": 0.3056650161743164,
+ "num_input_tokens_seen": 1005509112,
+ "step": 22995,
+ "train_runtime": 177827.549,
+ "train_tokens_per_second": 5654.406
+ },
+ {
+ "epoch": 1.5486129814166443,
+ "grad_norm": 0.7363938814599378,
+ "learning_rate": 1.2053983647731337e-06,
+ "loss": 0.29947643280029296,
+ "num_input_tokens_seen": 1005742432,
+ "step": 23000,
+ "train_runtime": 177862.3567,
+ "train_tokens_per_second": 5654.611
+ },
+ {
+ "epoch": 1.5489496364126043,
+ "grad_norm": 0.818386721631332,
+ "learning_rate": 1.2036771127827152e-06,
+ "loss": 0.3036951541900635,
+ "num_input_tokens_seen": 1005948408,
+ "step": 23005,
+ "train_runtime": 177896.9071,
+ "train_tokens_per_second": 5654.671
+ },
+ {
+ "epoch": 1.5492862914085646,
+ "grad_norm": 0.7210690822091025,
+ "learning_rate": 1.2019569224218869e-06,
+ "loss": 0.3014035224914551,
+ "num_input_tokens_seen": 1006184672,
+ "step": 23010,
+ "train_runtime": 177940.4278,
+ "train_tokens_per_second": 5654.615
+ },
+ {
+ "epoch": 1.5496229464045248,
+ "grad_norm": 0.7336526052679622,
+ "learning_rate": 1.2002377941716936e-06,
+ "loss": 0.27741825580596924,
+ "num_input_tokens_seen": 1006397816,
+ "step": 23015,
+ "train_runtime": 177978.0712,
+ "train_tokens_per_second": 5654.617
+ },
+ {
+ "epoch": 1.5499596014004848,
+ "grad_norm": 0.8612322343025608,
+ "learning_rate": 1.1985197285128853e-06,
+ "loss": 0.30047073364257815,
+ "num_input_tokens_seen": 1006606872,
+ "step": 23020,
+ "train_runtime": 178024.3698,
+ "train_tokens_per_second": 5654.321
+ },
+ {
+ "epoch": 1.5502962563964449,
+ "grad_norm": 0.7706187528021623,
+ "learning_rate": 1.1968027259259107e-06,
+ "loss": 0.29756712913513184,
+ "num_input_tokens_seen": 1006810960,
+ "step": 23025,
+ "train_runtime": 178061.8932,
+ "train_tokens_per_second": 5654.275
+ },
+ {
+ "epoch": 1.5506329113924051,
+ "grad_norm": 0.7669195811239113,
+ "learning_rate": 1.1950867868909293e-06,
+ "loss": 0.27481400966644287,
+ "num_input_tokens_seen": 1007021520,
+ "step": 23030,
+ "train_runtime": 178104.3,
+ "train_tokens_per_second": 5654.111
+ },
+ {
+ "epoch": 1.5509695663883654,
+ "grad_norm": 0.8825553327883765,
+ "learning_rate": 1.1933719118877923e-06,
+ "loss": 0.30696120262146,
+ "num_input_tokens_seen": 1007265144,
+ "step": 23035,
+ "train_runtime": 178134.9312,
+ "train_tokens_per_second": 5654.507
+ },
+ {
+ "epoch": 1.5513062213843254,
+ "grad_norm": 0.6530163696183422,
+ "learning_rate": 1.1916581013960632e-06,
+ "loss": 0.2742566347122192,
+ "num_input_tokens_seen": 1007502816,
+ "step": 23040,
+ "train_runtime": 178176.3835,
+ "train_tokens_per_second": 5654.525
+ },
+ {
+ "epoch": 1.5516428763802854,
+ "grad_norm": 0.8012588977520706,
+ "learning_rate": 1.189945355895002e-06,
+ "loss": 0.31151793003082273,
+ "num_input_tokens_seen": 1007718880,
+ "step": 23045,
+ "train_runtime": 178217.168,
+ "train_tokens_per_second": 5654.443
+ },
+ {
+ "epoch": 1.5519795313762457,
+ "grad_norm": 0.7331978765364463,
+ "learning_rate": 1.1882336758635727e-06,
+ "loss": 0.27026982307434083,
+ "num_input_tokens_seen": 1007951160,
+ "step": 23050,
+ "train_runtime": 178252.7434,
+ "train_tokens_per_second": 5654.618
+ },
+ {
+ "epoch": 1.552316186372206,
+ "grad_norm": 0.8000089411786985,
+ "learning_rate": 1.1865230617804412e-06,
+ "loss": 0.32370221614837646,
+ "num_input_tokens_seen": 1008186432,
+ "step": 23055,
+ "train_runtime": 178288.4347,
+ "train_tokens_per_second": 5654.806
+ },
+ {
+ "epoch": 1.552652841368166,
+ "grad_norm": 0.7285802544354917,
+ "learning_rate": 1.1848135141239753e-06,
+ "loss": 0.2949348211288452,
+ "num_input_tokens_seen": 1008403400,
+ "step": 23060,
+ "train_runtime": 178329.6018,
+ "train_tokens_per_second": 5654.717
+ },
+ {
+ "epoch": 1.552989496364126,
+ "grad_norm": 0.8133624441552181,
+ "learning_rate": 1.1831050333722438e-06,
+ "loss": 0.3110039710998535,
+ "num_input_tokens_seen": 1008619128,
+ "step": 23065,
+ "train_runtime": 178364.4191,
+ "train_tokens_per_second": 5654.822
+ },
+ {
+ "epoch": 1.5533261513600862,
+ "grad_norm": 0.7041630401525898,
+ "learning_rate": 1.1813976200030191e-06,
+ "loss": 0.29149832725524905,
+ "num_input_tokens_seen": 1008834768,
+ "step": 23070,
+ "train_runtime": 178400.3205,
+ "train_tokens_per_second": 5654.893
+ },
+ {
+ "epoch": 1.5536628063560465,
+ "grad_norm": 0.7088406276262311,
+ "learning_rate": 1.1796912744937717e-06,
+ "loss": 0.31109683513641356,
+ "num_input_tokens_seen": 1009058816,
+ "step": 23075,
+ "train_runtime": 178437.2135,
+ "train_tokens_per_second": 5654.98
+ },
+ {
+ "epoch": 1.5539994613520065,
+ "grad_norm": 0.8006442613984758,
+ "learning_rate": 1.1779859973216802e-06,
+ "loss": 0.2948103904724121,
+ "num_input_tokens_seen": 1009270824,
+ "step": 23080,
+ "train_runtime": 178479.8877,
+ "train_tokens_per_second": 5654.815
+ },
+ {
+ "epoch": 1.5543361163479665,
+ "grad_norm": 0.8721550220509111,
+ "learning_rate": 1.1762817889636142e-06,
+ "loss": 0.29932336807250975,
+ "num_input_tokens_seen": 1009470192,
+ "step": 23085,
+ "train_runtime": 178518.382,
+ "train_tokens_per_second": 5654.713
+ },
+ {
+ "epoch": 1.5546727713439268,
+ "grad_norm": 0.7952095441240394,
+ "learning_rate": 1.1745786498961542e-06,
+ "loss": 0.31789186000823977,
+ "num_input_tokens_seen": 1009708792,
+ "step": 23090,
+ "train_runtime": 178557.7116,
+ "train_tokens_per_second": 5654.804
+ },
+ {
+ "epoch": 1.555009426339887,
+ "grad_norm": 0.7311486152172996,
+ "learning_rate": 1.1728765805955766e-06,
+ "loss": 0.2769877195358276,
+ "num_input_tokens_seen": 1009922120,
+ "step": 23095,
+ "train_runtime": 178592.2137,
+ "train_tokens_per_second": 5654.906
+ },
+ {
+ "epoch": 1.555346081335847,
+ "grad_norm": 0.8415310169008027,
+ "learning_rate": 1.1711755815378595e-06,
+ "loss": 0.32232234477996824,
+ "num_input_tokens_seen": 1010130920,
+ "step": 23100,
+ "train_runtime": 178626.0283,
+ "train_tokens_per_second": 5655.004
+ },
+ {
+ "epoch": 1.555682736331807,
+ "grad_norm": 0.6963455165526707,
+ "learning_rate": 1.1694756531986818e-06,
+ "loss": 0.293909478187561,
+ "num_input_tokens_seen": 1010358016,
+ "step": 23105,
+ "train_runtime": 178670.1633,
+ "train_tokens_per_second": 5654.878
+ },
+ {
+ "epoch": 1.5560193913277673,
+ "grad_norm": 0.7864173646107584,
+ "learning_rate": 1.1677767960534232e-06,
+ "loss": 0.3079185962677002,
+ "num_input_tokens_seen": 1010578448,
+ "step": 23110,
+ "train_runtime": 178714.6961,
+ "train_tokens_per_second": 5654.703
+ },
+ {
+ "epoch": 1.5563560463237276,
+ "grad_norm": 0.7334560465799055,
+ "learning_rate": 1.166079010577163e-06,
+ "loss": 0.31560115814208983,
+ "num_input_tokens_seen": 1010787032,
+ "step": 23115,
+ "train_runtime": 178752.3631,
+ "train_tokens_per_second": 5654.678
+ },
+ {
+ "epoch": 1.5566927013196876,
+ "grad_norm": 0.8841111887561086,
+ "learning_rate": 1.1643822972446823e-06,
+ "loss": 0.3215439796447754,
+ "num_input_tokens_seen": 1011011240,
+ "step": 23120,
+ "train_runtime": 178786.2102,
+ "train_tokens_per_second": 5654.861
+ },
+ {
+ "epoch": 1.5570293563156477,
+ "grad_norm": 0.7142397763016451,
+ "learning_rate": 1.1626866565304594e-06,
+ "loss": 0.2977125883102417,
+ "num_input_tokens_seen": 1011244280,
+ "step": 23125,
+ "train_runtime": 178828.8659,
+ "train_tokens_per_second": 5654.816
+ },
+ {
+ "epoch": 1.557366011311608,
+ "grad_norm": 0.7277792379867785,
+ "learning_rate": 1.16099208890868e-06,
+ "loss": 0.2924909830093384,
+ "num_input_tokens_seen": 1011459816,
+ "step": 23130,
+ "train_runtime": 178867.1737,
+ "train_tokens_per_second": 5654.81
+ },
+ {
+ "epoch": 1.5577026663075682,
+ "grad_norm": 0.7072278904651554,
+ "learning_rate": 1.1592985948532187e-06,
+ "loss": 0.29182374477386475,
+ "num_input_tokens_seen": 1011668760,
+ "step": 23135,
+ "train_runtime": 178901.1993,
+ "train_tokens_per_second": 5654.902
+ },
+ {
+ "epoch": 1.5580393213035282,
+ "grad_norm": 0.7122762244099401,
+ "learning_rate": 1.1576061748376594e-06,
+ "loss": 0.28834197521209715,
+ "num_input_tokens_seen": 1011885552,
+ "step": 23140,
+ "train_runtime": 178941.129,
+ "train_tokens_per_second": 5654.852
+ },
+ {
+ "epoch": 1.5583759762994882,
+ "grad_norm": 0.7581459438101124,
+ "learning_rate": 1.1559148293352805e-06,
+ "loss": 0.293502950668335,
+ "num_input_tokens_seen": 1012108496,
+ "step": 23145,
+ "train_runtime": 178981.2231,
+ "train_tokens_per_second": 5654.831
+ },
+ {
+ "epoch": 1.5587126312954485,
+ "grad_norm": 0.6988252432908294,
+ "learning_rate": 1.154224558819062e-06,
+ "loss": 0.2947830677032471,
+ "num_input_tokens_seen": 1012331792,
+ "step": 23150,
+ "train_runtime": 179021.8936,
+ "train_tokens_per_second": 5654.793
+ },
+ {
+ "epoch": 1.5590492862914087,
+ "grad_norm": 0.7771507807990884,
+ "learning_rate": 1.1525353637616821e-06,
+ "loss": 0.29758148193359374,
+ "num_input_tokens_seen": 1012552976,
+ "step": 23155,
+ "train_runtime": 179058.4207,
+ "train_tokens_per_second": 5654.875
+ },
+ {
+ "epoch": 1.5593859412873687,
+ "grad_norm": 0.7491488781107557,
+ "learning_rate": 1.1508472446355196e-06,
+ "loss": 0.3053610324859619,
+ "num_input_tokens_seen": 1012785584,
+ "step": 23160,
+ "train_runtime": 179099.292,
+ "train_tokens_per_second": 5654.883
+ },
+ {
+ "epoch": 1.5597225962833288,
+ "grad_norm": 0.697759562324447,
+ "learning_rate": 1.1491602019126501e-06,
+ "loss": 0.2998149871826172,
+ "num_input_tokens_seen": 1013018136,
+ "step": 23165,
+ "train_runtime": 179129.3593,
+ "train_tokens_per_second": 5655.232
+ },
+ {
+ "epoch": 1.560059251279289,
+ "grad_norm": 0.7568078853356832,
+ "learning_rate": 1.1474742360648515e-06,
+ "loss": 0.31585729122161865,
+ "num_input_tokens_seen": 1013244520,
+ "step": 23170,
+ "train_runtime": 179174.5036,
+ "train_tokens_per_second": 5655.071
+ },
+ {
+ "epoch": 1.5603959062752493,
+ "grad_norm": 0.8520933038767174,
+ "learning_rate": 1.1457893475635968e-06,
+ "loss": 0.3203889846801758,
+ "num_input_tokens_seen": 1013457472,
+ "step": 23175,
+ "train_runtime": 179219.5043,
+ "train_tokens_per_second": 5654.839
+ },
+ {
+ "epoch": 1.5607325612712093,
+ "grad_norm": 0.6916094191053404,
+ "learning_rate": 1.144105536880063e-06,
+ "loss": 0.2820396900177002,
+ "num_input_tokens_seen": 1013672336,
+ "step": 23180,
+ "train_runtime": 179252.287,
+ "train_tokens_per_second": 5655.004
+ },
+ {
+ "epoch": 1.5610692162671693,
+ "grad_norm": 0.7609947532948533,
+ "learning_rate": 1.1424228044851176e-06,
+ "loss": 0.29037587642669677,
+ "num_input_tokens_seen": 1013859016,
+ "step": 23185,
+ "train_runtime": 179288.0236,
+ "train_tokens_per_second": 5654.918
+ },
+ {
+ "epoch": 1.5614058712631296,
+ "grad_norm": 0.6991374139098863,
+ "learning_rate": 1.1407411508493355e-06,
+ "loss": 0.3020007133483887,
+ "num_input_tokens_seen": 1014067352,
+ "step": 23190,
+ "train_runtime": 179325.7985,
+ "train_tokens_per_second": 5654.888
+ },
+ {
+ "epoch": 1.5617425262590898,
+ "grad_norm": 0.7525302404327049,
+ "learning_rate": 1.1390605764429846e-06,
+ "loss": 0.2727111101150513,
+ "num_input_tokens_seen": 1014277952,
+ "step": 23195,
+ "train_runtime": 179365.198,
+ "train_tokens_per_second": 5654.82
+ },
+ {
+ "epoch": 1.5620791812550499,
+ "grad_norm": 0.755912903107464,
+ "learning_rate": 1.1373810817360314e-06,
+ "loss": 0.29233598709106445,
+ "num_input_tokens_seen": 1014509144,
+ "step": 23200,
+ "train_runtime": 179403.2626,
+ "train_tokens_per_second": 5654.909
+ },
+ {
+ "epoch": 1.5624158362510099,
+ "grad_norm": 0.8490364171656128,
+ "learning_rate": 1.135702667198142e-06,
+ "loss": 0.3186174392700195,
+ "num_input_tokens_seen": 1014733760,
+ "step": 23205,
+ "train_runtime": 179436.0347,
+ "train_tokens_per_second": 5655.128
+ },
+ {
+ "epoch": 1.5627524912469701,
+ "grad_norm": 0.7833432896868294,
+ "learning_rate": 1.134025333298679e-06,
+ "loss": 0.2658507823944092,
+ "num_input_tokens_seen": 1014933448,
+ "step": 23210,
+ "train_runtime": 179473.8421,
+ "train_tokens_per_second": 5655.049
+ },
+ {
+ "epoch": 1.5630891462429304,
+ "grad_norm": 0.6767171989222308,
+ "learning_rate": 1.1323490805067045e-06,
+ "loss": 0.27293944358825684,
+ "num_input_tokens_seen": 1015151416,
+ "step": 23215,
+ "train_runtime": 179514.1502,
+ "train_tokens_per_second": 5654.994
+ },
+ {
+ "epoch": 1.5634258012388904,
+ "grad_norm": 0.7564520547243825,
+ "learning_rate": 1.1306739092909757e-06,
+ "loss": 0.27212421894073485,
+ "num_input_tokens_seen": 1015357104,
+ "step": 23220,
+ "train_runtime": 179558.059,
+ "train_tokens_per_second": 5654.757
+ },
+ {
+ "epoch": 1.5637624562348504,
+ "grad_norm": 0.7556836048466959,
+ "learning_rate": 1.1289998201199493e-06,
+ "loss": 0.28475046157836914,
+ "num_input_tokens_seen": 1015578216,
+ "step": 23225,
+ "train_runtime": 179598.1851,
+ "train_tokens_per_second": 5654.724
+ },
+ {
+ "epoch": 1.5640991112308107,
+ "grad_norm": 0.7771413672780467,
+ "learning_rate": 1.1273268134617816e-06,
+ "loss": 0.3049982786178589,
+ "num_input_tokens_seen": 1015803864,
+ "step": 23230,
+ "train_runtime": 179637.2745,
+ "train_tokens_per_second": 5654.75
+ },
+ {
+ "epoch": 1.564435766226771,
+ "grad_norm": 0.6387132976491532,
+ "learning_rate": 1.1256548897843189e-06,
+ "loss": 0.2998204708099365,
+ "num_input_tokens_seen": 1016035144,
+ "step": 23235,
+ "train_runtime": 179676.6807,
+ "train_tokens_per_second": 5654.797
+ },
+ {
+ "epoch": 1.564772421222731,
+ "grad_norm": 0.677517945115812,
+ "learning_rate": 1.1239840495551136e-06,
+ "loss": 0.2948010921478271,
+ "num_input_tokens_seen": 1016256464,
+ "step": 23240,
+ "train_runtime": 179716.2122,
+ "train_tokens_per_second": 5654.785
+ },
+ {
+ "epoch": 1.565109076218691,
+ "grad_norm": 0.8189763586455887,
+ "learning_rate": 1.1223142932414087e-06,
+ "loss": 0.3241511583328247,
+ "num_input_tokens_seen": 1016483368,
+ "step": 23245,
+ "train_runtime": 179758.4552,
+ "train_tokens_per_second": 5654.718
+ },
+ {
+ "epoch": 1.5654457312146512,
+ "grad_norm": 0.6822186571021763,
+ "learning_rate": 1.120645621310147e-06,
+ "loss": 0.2887975454330444,
+ "num_input_tokens_seen": 1016700976,
+ "step": 23250,
+ "train_runtime": 179800.864,
+ "train_tokens_per_second": 5654.594
+ },
+ {
+ "epoch": 1.5657823862106115,
+ "grad_norm": 0.8030985846370237,
+ "learning_rate": 1.1189780342279666e-06,
+ "loss": 0.2927762508392334,
+ "num_input_tokens_seen": 1016907680,
+ "step": 23255,
+ "train_runtime": 179835.7597,
+ "train_tokens_per_second": 5654.647
+ },
+ {
+ "epoch": 1.5661190412065715,
+ "grad_norm": 0.7041599888250881,
+ "learning_rate": 1.1173115324612033e-06,
+ "loss": 0.30506865978240966,
+ "num_input_tokens_seen": 1017110400,
+ "step": 23260,
+ "train_runtime": 179876.7454,
+ "train_tokens_per_second": 5654.485
+ },
+ {
+ "epoch": 1.5664556962025316,
+ "grad_norm": 0.8207082370707505,
+ "learning_rate": 1.1156461164758891e-06,
+ "loss": 0.3330164909362793,
+ "num_input_tokens_seen": 1017339120,
+ "step": 23265,
+ "train_runtime": 179913.8123,
+ "train_tokens_per_second": 5654.592
+ },
+ {
+ "epoch": 1.5667923511984918,
+ "grad_norm": 0.6985440775195174,
+ "learning_rate": 1.1139817867377522e-06,
+ "loss": 0.2630832433700562,
+ "num_input_tokens_seen": 1017568520,
+ "step": 23270,
+ "train_runtime": 179951.8501,
+ "train_tokens_per_second": 5654.671
+ },
+ {
+ "epoch": 1.567129006194452,
+ "grad_norm": 0.7538202300333461,
+ "learning_rate": 1.1123185437122153e-06,
+ "loss": 0.29378499984741213,
+ "num_input_tokens_seen": 1017797256,
+ "step": 23275,
+ "train_runtime": 179995.1502,
+ "train_tokens_per_second": 5654.582
+ },
+ {
+ "epoch": 1.567465661190412,
+ "grad_norm": 0.8033728394499458,
+ "learning_rate": 1.1106563878644038e-06,
+ "loss": 0.30549144744873047,
+ "num_input_tokens_seen": 1018024880,
+ "step": 23280,
+ "train_runtime": 180035.9497,
+ "train_tokens_per_second": 5654.564
+ },
+ {
+ "epoch": 1.5678023161863721,
+ "grad_norm": 0.7853141000994098,
+ "learning_rate": 1.1089953196591286e-06,
+ "loss": 0.30190200805664064,
+ "num_input_tokens_seen": 1018241712,
+ "step": 23285,
+ "train_runtime": 180077.9944,
+ "train_tokens_per_second": 5654.448
+ },
+ {
+ "epoch": 1.5681389711823324,
+ "grad_norm": 0.7859724149410223,
+ "learning_rate": 1.107335339560906e-06,
+ "loss": 0.3117091417312622,
+ "num_input_tokens_seen": 1018448032,
+ "step": 23290,
+ "train_runtime": 180115.2413,
+ "train_tokens_per_second": 5654.424
+ },
+ {
+ "epoch": 1.5684756261782926,
+ "grad_norm": 0.7863228372315338,
+ "learning_rate": 1.1056764480339427e-06,
+ "loss": 0.3014981269836426,
+ "num_input_tokens_seen": 1018674080,
+ "step": 23295,
+ "train_runtime": 180158.4273,
+ "train_tokens_per_second": 5654.324
+ },
+ {
+ "epoch": 1.5688122811742526,
+ "grad_norm": 0.7789425478674455,
+ "learning_rate": 1.1040186455421425e-06,
+ "loss": 0.33310580253601074,
+ "num_input_tokens_seen": 1018908720,
+ "step": 23300,
+ "train_runtime": 180194.8091,
+ "train_tokens_per_second": 5654.484
+ },
+ {
+ "epoch": 1.5691489361702127,
+ "grad_norm": 0.8291198693575462,
+ "learning_rate": 1.1023619325491052e-06,
+ "loss": 0.310105037689209,
+ "num_input_tokens_seen": 1019122152,
+ "step": 23305,
+ "train_runtime": 180234.0997,
+ "train_tokens_per_second": 5654.436
+ },
+ {
+ "epoch": 1.569485591166173,
+ "grad_norm": 0.6253815970834292,
+ "learning_rate": 1.1007063095181248e-06,
+ "loss": 0.27553586959838866,
+ "num_input_tokens_seen": 1019342832,
+ "step": 23310,
+ "train_runtime": 180278.8765,
+ "train_tokens_per_second": 5654.256
+ },
+ {
+ "epoch": 1.5698222461621332,
+ "grad_norm": 0.8786406413208016,
+ "learning_rate": 1.0990517769121905e-06,
+ "loss": 0.29708888530731203,
+ "num_input_tokens_seen": 1019546240,
+ "step": 23315,
+ "train_runtime": 180314.3153,
+ "train_tokens_per_second": 5654.272
+ },
+ {
+ "epoch": 1.5701589011580932,
+ "grad_norm": 0.8222072276717016,
+ "learning_rate": 1.0973983351939876e-06,
+ "loss": 0.3073767900466919,
+ "num_input_tokens_seen": 1019767384,
+ "step": 23320,
+ "train_runtime": 180352.2362,
+ "train_tokens_per_second": 5654.31
+ },
+ {
+ "epoch": 1.5704955561540532,
+ "grad_norm": 0.7054404712911847,
+ "learning_rate": 1.095745984825895e-06,
+ "loss": 0.2895050048828125,
+ "num_input_tokens_seen": 1019996880,
+ "step": 23325,
+ "train_runtime": 180386.2267,
+ "train_tokens_per_second": 5654.516
+ },
+ {
+ "epoch": 1.5708322111500135,
+ "grad_norm": 0.8092139157721354,
+ "learning_rate": 1.0940947262699902e-06,
+ "loss": 0.30101957321166994,
+ "num_input_tokens_seen": 1020200320,
+ "step": 23330,
+ "train_runtime": 180428.7048,
+ "train_tokens_per_second": 5654.313
+ },
+ {
+ "epoch": 1.5711688661459737,
+ "grad_norm": 0.7111960834790324,
+ "learning_rate": 1.092444559988038e-06,
+ "loss": 0.28864340782165526,
+ "num_input_tokens_seen": 1020417768,
+ "step": 23335,
+ "train_runtime": 180466.2384,
+ "train_tokens_per_second": 5654.342
+ },
+ {
+ "epoch": 1.5715055211419338,
+ "grad_norm": 0.660638688067613,
+ "learning_rate": 1.0907954864415048e-06,
+ "loss": 0.29140138626098633,
+ "num_input_tokens_seen": 1020645240,
+ "step": 23340,
+ "train_runtime": 180507.8532,
+ "train_tokens_per_second": 5654.298
+ },
+ {
+ "epoch": 1.5718421761378938,
+ "grad_norm": 0.8672767958383719,
+ "learning_rate": 1.089147506091549e-06,
+ "loss": 0.28663978576660154,
+ "num_input_tokens_seen": 1020864992,
+ "step": 23345,
+ "train_runtime": 180542.9652,
+ "train_tokens_per_second": 5654.416
+ },
+ {
+ "epoch": 1.572178831133854,
+ "grad_norm": 0.8719743776750728,
+ "learning_rate": 1.0875006193990217e-06,
+ "loss": 0.3154577732086182,
+ "num_input_tokens_seen": 1021076760,
+ "step": 23350,
+ "train_runtime": 180584.1055,
+ "train_tokens_per_second": 5654.3
+ },
+ {
+ "epoch": 1.5725154861298143,
+ "grad_norm": 0.7880930108374784,
+ "learning_rate": 1.0858548268244706e-06,
+ "loss": 0.2915508270263672,
+ "num_input_tokens_seen": 1021295960,
+ "step": 23355,
+ "train_runtime": 180623.241,
+ "train_tokens_per_second": 5654.289
+ },
+ {
+ "epoch": 1.5728521411257743,
+ "grad_norm": 0.8007349258213046,
+ "learning_rate": 1.0842101288281359e-06,
+ "loss": 0.293584418296814,
+ "num_input_tokens_seen": 1021516488,
+ "step": 23360,
+ "train_runtime": 180662.7604,
+ "train_tokens_per_second": 5654.273
+ },
+ {
+ "epoch": 1.5731887961217343,
+ "grad_norm": 0.7656889523789541,
+ "learning_rate": 1.082566525869952e-06,
+ "loss": 0.28592419624328613,
+ "num_input_tokens_seen": 1021745864,
+ "step": 23365,
+ "train_runtime": 180702.808,
+ "train_tokens_per_second": 5654.289
+ },
+ {
+ "epoch": 1.5735254511176946,
+ "grad_norm": 0.8722272978728876,
+ "learning_rate": 1.0809240184095476e-06,
+ "loss": 0.3062607288360596,
+ "num_input_tokens_seen": 1021959216,
+ "step": 23370,
+ "train_runtime": 180737.0204,
+ "train_tokens_per_second": 5654.399
+ },
+ {
+ "epoch": 1.5738621061136548,
+ "grad_norm": 0.6745653891846737,
+ "learning_rate": 1.0792826069062429e-06,
+ "loss": 0.26804351806640625,
+ "num_input_tokens_seen": 1022178816,
+ "step": 23375,
+ "train_runtime": 180775.1712,
+ "train_tokens_per_second": 5654.42
+ },
+ {
+ "epoch": 1.5741987611096149,
+ "grad_norm": 0.6985205733037011,
+ "learning_rate": 1.0776422918190576e-06,
+ "loss": 0.31443793773651124,
+ "num_input_tokens_seen": 1022402224,
+ "step": 23380,
+ "train_runtime": 180806.1938,
+ "train_tokens_per_second": 5654.686
+ },
+ {
+ "epoch": 1.574535416105575,
+ "grad_norm": 0.8000413255647302,
+ "learning_rate": 1.0760030736066952e-06,
+ "loss": 0.28395633697509765,
+ "num_input_tokens_seen": 1022607104,
+ "step": 23385,
+ "train_runtime": 180846.4339,
+ "train_tokens_per_second": 5654.561
+ },
+ {
+ "epoch": 1.5748720711015352,
+ "grad_norm": 0.7301802635358972,
+ "learning_rate": 1.0743649527275619e-06,
+ "loss": 0.3159285068511963,
+ "num_input_tokens_seen": 1022834416,
+ "step": 23390,
+ "train_runtime": 180888.2807,
+ "train_tokens_per_second": 5654.509
+ },
+ {
+ "epoch": 1.5752087260974954,
+ "grad_norm": 0.8194485612881878,
+ "learning_rate": 1.0727279296397514e-06,
+ "loss": 0.3103215456008911,
+ "num_input_tokens_seen": 1023064384,
+ "step": 23395,
+ "train_runtime": 180932.3993,
+ "train_tokens_per_second": 5654.401
+ },
+ {
+ "epoch": 1.5755453810934554,
+ "grad_norm": 0.8975237265847585,
+ "learning_rate": 1.0710920048010526e-06,
+ "loss": 0.3091814279556274,
+ "num_input_tokens_seen": 1023286312,
+ "step": 23400,
+ "train_runtime": 180973.834,
+ "train_tokens_per_second": 5654.333
+ },
+ {
+ "epoch": 1.5758820360894155,
+ "grad_norm": 0.7551422109604019,
+ "learning_rate": 1.0694571786689462e-06,
+ "loss": 0.29776315689086913,
+ "num_input_tokens_seen": 1023528528,
+ "step": 23405,
+ "train_runtime": 181012.8344,
+ "train_tokens_per_second": 5654.453
+ },
+ {
+ "epoch": 1.5762186910853757,
+ "grad_norm": 0.7923202761413153,
+ "learning_rate": 1.067823451700606e-06,
+ "loss": 0.29261269569396975,
+ "num_input_tokens_seen": 1023748656,
+ "step": 23410,
+ "train_runtime": 181047.5433,
+ "train_tokens_per_second": 5654.585
+ },
+ {
+ "epoch": 1.576555346081336,
+ "grad_norm": 0.7616310783411705,
+ "learning_rate": 1.066190824352899e-06,
+ "loss": 0.29994802474975585,
+ "num_input_tokens_seen": 1023974480,
+ "step": 23415,
+ "train_runtime": 181087.6631,
+ "train_tokens_per_second": 5654.579
+ },
+ {
+ "epoch": 1.576892001077296,
+ "grad_norm": 0.6523336414139479,
+ "learning_rate": 1.0645592970823837e-06,
+ "loss": 0.27131869792938235,
+ "num_input_tokens_seen": 1024177832,
+ "step": 23420,
+ "train_runtime": 181123.7413,
+ "train_tokens_per_second": 5654.575
+ },
+ {
+ "epoch": 1.577228656073256,
+ "grad_norm": 0.6772527419522956,
+ "learning_rate": 1.0629288703453105e-06,
+ "loss": 0.3091407299041748,
+ "num_input_tokens_seen": 1024403608,
+ "step": 23425,
+ "train_runtime": 181163.8389,
+ "train_tokens_per_second": 5654.57
+ },
+ {
+ "epoch": 1.5775653110692163,
+ "grad_norm": 0.8870052312783957,
+ "learning_rate": 1.061299544597627e-06,
+ "loss": 0.29852406978607177,
+ "num_input_tokens_seen": 1024608360,
+ "step": 23430,
+ "train_runtime": 181201.9913,
+ "train_tokens_per_second": 5654.509
+ },
+ {
+ "epoch": 1.5779019660651765,
+ "grad_norm": 0.7385143164519493,
+ "learning_rate": 1.0596713202949637e-06,
+ "loss": 0.3082388162612915,
+ "num_input_tokens_seen": 1024836472,
+ "step": 23435,
+ "train_runtime": 181244.3758,
+ "train_tokens_per_second": 5654.446
+ },
+ {
+ "epoch": 1.5782386210611365,
+ "grad_norm": 0.7117628849410049,
+ "learning_rate": 1.0580441978926527e-06,
+ "loss": 0.3000252485275269,
+ "num_input_tokens_seen": 1025047408,
+ "step": 23440,
+ "train_runtime": 181284.1102,
+ "train_tokens_per_second": 5654.37
+ },
+ {
+ "epoch": 1.5785752760570966,
+ "grad_norm": 0.8021413628123183,
+ "learning_rate": 1.056418177845711e-06,
+ "loss": 0.2908348083496094,
+ "num_input_tokens_seen": 1025252528,
+ "step": 23445,
+ "train_runtime": 181326.465,
+ "train_tokens_per_second": 5654.18
+ },
+ {
+ "epoch": 1.5789119310530568,
+ "grad_norm": 0.7083241483136948,
+ "learning_rate": 1.0547932606088506e-06,
+ "loss": 0.3012411594390869,
+ "num_input_tokens_seen": 1025462120,
+ "step": 23450,
+ "train_runtime": 181365.3534,
+ "train_tokens_per_second": 5654.124
+ },
+ {
+ "epoch": 1.579248586049017,
+ "grad_norm": 0.749765463777595,
+ "learning_rate": 1.053169446636475e-06,
+ "loss": 0.29930336475372316,
+ "num_input_tokens_seen": 1025676472,
+ "step": 23455,
+ "train_runtime": 181402.7027,
+ "train_tokens_per_second": 5654.141
+ },
+ {
+ "epoch": 1.579585241044977,
+ "grad_norm": 0.870290848746088,
+ "learning_rate": 1.0515467363826782e-06,
+ "loss": 0.3104248046875,
+ "num_input_tokens_seen": 1025904840,
+ "step": 23460,
+ "train_runtime": 181436.5696,
+ "train_tokens_per_second": 5654.344
+ },
+ {
+ "epoch": 1.5799218960409371,
+ "grad_norm": 0.7377471679217029,
+ "learning_rate": 1.0499251303012454e-06,
+ "loss": 0.2840702533721924,
+ "num_input_tokens_seen": 1026129016,
+ "step": 23465,
+ "train_runtime": 181472.8949,
+ "train_tokens_per_second": 5654.448
+ },
+ {
+ "epoch": 1.5802585510368974,
+ "grad_norm": 0.7923429690046462,
+ "learning_rate": 1.0483046288456544e-06,
+ "loss": 0.30331013202667234,
+ "num_input_tokens_seen": 1026336600,
+ "step": 23470,
+ "train_runtime": 181508.7471,
+ "train_tokens_per_second": 5654.475
+ },
+ {
+ "epoch": 1.5805952060328576,
+ "grad_norm": 0.7779123855214451,
+ "learning_rate": 1.0466852324690719e-06,
+ "loss": 0.30037317276000974,
+ "num_input_tokens_seen": 1026549792,
+ "step": 23475,
+ "train_runtime": 181547.6766,
+ "train_tokens_per_second": 5654.436
+ },
+ {
+ "epoch": 1.5809318610288177,
+ "grad_norm": 0.812157038252082,
+ "learning_rate": 1.0450669416243596e-06,
+ "loss": 0.2942227363586426,
+ "num_input_tokens_seen": 1026774584,
+ "step": 23480,
+ "train_runtime": 181585.7621,
+ "train_tokens_per_second": 5654.488
+ },
+ {
+ "epoch": 1.5812685160247777,
+ "grad_norm": 0.7370282264500914,
+ "learning_rate": 1.0434497567640639e-06,
+ "loss": 0.3083514213562012,
+ "num_input_tokens_seen": 1026996640,
+ "step": 23485,
+ "train_runtime": 181622.5429,
+ "train_tokens_per_second": 5654.566
+ },
+ {
+ "epoch": 1.581605171020738,
+ "grad_norm": 0.8340284951059088,
+ "learning_rate": 1.0418336783404282e-06,
+ "loss": 0.3067812204360962,
+ "num_input_tokens_seen": 1027217736,
+ "step": 23490,
+ "train_runtime": 181660.7149,
+ "train_tokens_per_second": 5654.595
+ },
+ {
+ "epoch": 1.5819418260166982,
+ "grad_norm": 0.694715006207709,
+ "learning_rate": 1.0402187068053825e-06,
+ "loss": 0.3050783395767212,
+ "num_input_tokens_seen": 1027433032,
+ "step": 23495,
+ "train_runtime": 181694.8222,
+ "train_tokens_per_second": 5654.718
+ },
+ {
+ "epoch": 1.5822784810126582,
+ "grad_norm": 0.6925827695690472,
+ "learning_rate": 1.0386048426105495e-06,
+ "loss": 0.28964962959289553,
+ "num_input_tokens_seen": 1027649992,
+ "step": 23500,
+ "train_runtime": 181732.3475,
+ "train_tokens_per_second": 5654.744
+ },
+ {
+ "epoch": 1.5826151360086183,
+ "grad_norm": 0.7154905995519729,
+ "learning_rate": 1.0369920862072396e-06,
+ "loss": 0.28251526355743406,
+ "num_input_tokens_seen": 1027858824,
+ "step": 23505,
+ "train_runtime": 181776.3615,
+ "train_tokens_per_second": 5654.524
+ },
+ {
+ "epoch": 1.5829517910045785,
+ "grad_norm": 0.7732969474103675,
+ "learning_rate": 1.0353804380464556e-06,
+ "loss": 0.28255496025085447,
+ "num_input_tokens_seen": 1028079384,
+ "step": 23510,
+ "train_runtime": 181818.5723,
+ "train_tokens_per_second": 5654.424
+ },
+ {
+ "epoch": 1.5832884460005388,
+ "grad_norm": 0.7444357762125697,
+ "learning_rate": 1.0337698985788903e-06,
+ "loss": 0.2897408246994019,
+ "num_input_tokens_seen": 1028306808,
+ "step": 23515,
+ "train_runtime": 181859.8638,
+ "train_tokens_per_second": 5654.391
+ },
+ {
+ "epoch": 1.5836251009964988,
+ "grad_norm": 0.7396832464568777,
+ "learning_rate": 1.0321604682549247e-06,
+ "loss": 0.27561049461364745,
+ "num_input_tokens_seen": 1028540048,
+ "step": 23520,
+ "train_runtime": 181903.878,
+ "train_tokens_per_second": 5654.305
+ },
+ {
+ "epoch": 1.5839617559924588,
+ "grad_norm": 0.7843713186397172,
+ "learning_rate": 1.0305521475246311e-06,
+ "loss": 0.3138261318206787,
+ "num_input_tokens_seen": 1028749872,
+ "step": 23525,
+ "train_runtime": 181940.9808,
+ "train_tokens_per_second": 5654.305
+ },
+ {
+ "epoch": 1.584298410988419,
+ "grad_norm": 0.811370185654451,
+ "learning_rate": 1.028944936837774e-06,
+ "loss": 0.31621651649475097,
+ "num_input_tokens_seen": 1028951280,
+ "step": 23530,
+ "train_runtime": 181974.7241,
+ "train_tokens_per_second": 5654.364
+ },
+ {
+ "epoch": 1.5846350659843793,
+ "grad_norm": 0.7315538832578915,
+ "learning_rate": 1.0273388366438003e-06,
+ "loss": 0.30298333168029784,
+ "num_input_tokens_seen": 1029175832,
+ "step": 23535,
+ "train_runtime": 182014.8854,
+ "train_tokens_per_second": 5654.35
+ },
+ {
+ "epoch": 1.5849717209803393,
+ "grad_norm": 0.6822526332510949,
+ "learning_rate": 1.0257338473918538e-06,
+ "loss": 0.2973572015762329,
+ "num_input_tokens_seen": 1029399728,
+ "step": 23540,
+ "train_runtime": 182052.6379,
+ "train_tokens_per_second": 5654.407
+ },
+ {
+ "epoch": 1.5853083759762994,
+ "grad_norm": 0.7092783516071793,
+ "learning_rate": 1.0241299695307644e-06,
+ "loss": 0.3015413761138916,
+ "num_input_tokens_seen": 1029616800,
+ "step": 23545,
+ "train_runtime": 182093.0118,
+ "train_tokens_per_second": 5654.345
+ },
+ {
+ "epoch": 1.5856450309722596,
+ "grad_norm": 0.6440724656411346,
+ "learning_rate": 1.0225272035090506e-06,
+ "loss": 0.282063889503479,
+ "num_input_tokens_seen": 1029848952,
+ "step": 23550,
+ "train_runtime": 182128.0658,
+ "train_tokens_per_second": 5654.532
+ },
+ {
+ "epoch": 1.5859816859682199,
+ "grad_norm": 0.7978640593248636,
+ "learning_rate": 1.0209255497749209e-06,
+ "loss": 0.30583910942077636,
+ "num_input_tokens_seen": 1030073088,
+ "step": 23555,
+ "train_runtime": 182170.9808,
+ "train_tokens_per_second": 5654.43
+ },
+ {
+ "epoch": 1.58631834096418,
+ "grad_norm": 0.7044002037356724,
+ "learning_rate": 1.0193250087762735e-06,
+ "loss": 0.28136019706726073,
+ "num_input_tokens_seen": 1030294120,
+ "step": 23560,
+ "train_runtime": 182212.5582,
+ "train_tokens_per_second": 5654.353
+ },
+ {
+ "epoch": 1.58665499596014,
+ "grad_norm": 0.626174961728994,
+ "learning_rate": 1.0177255809606934e-06,
+ "loss": 0.29512906074523926,
+ "num_input_tokens_seen": 1030518512,
+ "step": 23565,
+ "train_runtime": 182245.4993,
+ "train_tokens_per_second": 5654.562
+ },
+ {
+ "epoch": 1.5869916509561002,
+ "grad_norm": 0.7958252805077025,
+ "learning_rate": 1.0161272667754562e-06,
+ "loss": 0.2986447334289551,
+ "num_input_tokens_seen": 1030753488,
+ "step": 23570,
+ "train_runtime": 182291.1583,
+ "train_tokens_per_second": 5654.435
+ },
+ {
+ "epoch": 1.5873283059520604,
+ "grad_norm": 0.6503902978361567,
+ "learning_rate": 1.014530066667524e-06,
+ "loss": 0.2920704364776611,
+ "num_input_tokens_seen": 1030980864,
+ "step": 23575,
+ "train_runtime": 182338.3663,
+ "train_tokens_per_second": 5654.218
+ },
+ {
+ "epoch": 1.5876649609480205,
+ "grad_norm": 0.7822423572349132,
+ "learning_rate": 1.0129339810835526e-06,
+ "loss": 0.26558356285095214,
+ "num_input_tokens_seen": 1031184200,
+ "step": 23580,
+ "train_runtime": 182380.0639,
+ "train_tokens_per_second": 5654.04
+ },
+ {
+ "epoch": 1.5880016159439805,
+ "grad_norm": 0.7740865888482027,
+ "learning_rate": 1.0113390104698767e-06,
+ "loss": 0.2880537986755371,
+ "num_input_tokens_seen": 1031415416,
+ "step": 23585,
+ "train_runtime": 182415.6709,
+ "train_tokens_per_second": 5654.204
+ },
+ {
+ "epoch": 1.5883382709399407,
+ "grad_norm": 0.6444406633812003,
+ "learning_rate": 1.009745155272529e-06,
+ "loss": 0.32445645332336426,
+ "num_input_tokens_seen": 1031642064,
+ "step": 23590,
+ "train_runtime": 182452.8245,
+ "train_tokens_per_second": 5654.295
+ },
+ {
+ "epoch": 1.588674925935901,
+ "grad_norm": 0.716940757316124,
+ "learning_rate": 1.0081524159372246e-06,
+ "loss": 0.30813934803009035,
+ "num_input_tokens_seen": 1031852072,
+ "step": 23595,
+ "train_runtime": 182489.063,
+ "train_tokens_per_second": 5654.323
+ },
+ {
+ "epoch": 1.589011580931861,
+ "grad_norm": 0.7929537083603673,
+ "learning_rate": 1.0065607929093685e-06,
+ "loss": 0.2949239253997803,
+ "num_input_tokens_seen": 1032066320,
+ "step": 23600,
+ "train_runtime": 182530.2797,
+ "train_tokens_per_second": 5654.22
+ },
+ {
+ "epoch": 1.589348235927821,
+ "grad_norm": 0.7144421636045533,
+ "learning_rate": 1.0049702866340521e-06,
+ "loss": 0.2600874423980713,
+ "num_input_tokens_seen": 1032291736,
+ "step": 23605,
+ "train_runtime": 182571.9611,
+ "train_tokens_per_second": 5654.164
+ },
+ {
+ "epoch": 1.5896848909237813,
+ "grad_norm": 0.7689607902349916,
+ "learning_rate": 1.0033808975560556e-06,
+ "loss": 0.28646035194396974,
+ "num_input_tokens_seen": 1032506240,
+ "step": 23610,
+ "train_runtime": 182609.5452,
+ "train_tokens_per_second": 5654.175
+ },
+ {
+ "epoch": 1.5900215459197415,
+ "grad_norm": 0.7558437464479234,
+ "learning_rate": 1.0017926261198473e-06,
+ "loss": 0.29119462966918946,
+ "num_input_tokens_seen": 1032720936,
+ "step": 23615,
+ "train_runtime": 182649.2323,
+ "train_tokens_per_second": 5654.121
+ },
+ {
+ "epoch": 1.5903582009157016,
+ "grad_norm": 0.8842680705984803,
+ "learning_rate": 1.0002054727695814e-06,
+ "loss": 0.30643343925476074,
+ "num_input_tokens_seen": 1032929224,
+ "step": 23620,
+ "train_runtime": 182690.3573,
+ "train_tokens_per_second": 5653.989
+ },
+ {
+ "epoch": 1.5906948559116616,
+ "grad_norm": 0.6572152521341675,
+ "learning_rate": 9.986194379490993e-07,
+ "loss": 0.29436190128326417,
+ "num_input_tokens_seen": 1033140248,
+ "step": 23625,
+ "train_runtime": 182726.5478,
+ "train_tokens_per_second": 5654.024
+ },
+ {
+ "epoch": 1.5910315109076218,
+ "grad_norm": 0.7264405013312064,
+ "learning_rate": 9.970345221019345e-07,
+ "loss": 0.30518579483032227,
+ "num_input_tokens_seen": 1033354080,
+ "step": 23630,
+ "train_runtime": 182768.4987,
+ "train_tokens_per_second": 5653.896
+ },
+ {
+ "epoch": 1.591368165903582,
+ "grad_norm": 0.9173611218409357,
+ "learning_rate": 9.954507256712987e-07,
+ "loss": 0.30224905014038084,
+ "num_input_tokens_seen": 1033566992,
+ "step": 23635,
+ "train_runtime": 182805.0001,
+ "train_tokens_per_second": 5653.932
+ },
+ {
+ "epoch": 1.5917048208995421,
+ "grad_norm": 0.9443978197431155,
+ "learning_rate": 9.938680491000991e-07,
+ "loss": 0.3042686700820923,
+ "num_input_tokens_seen": 1033777624,
+ "step": 23640,
+ "train_runtime": 182847.8991,
+ "train_tokens_per_second": 5653.757
+ },
+ {
+ "epoch": 1.5920414758955022,
+ "grad_norm": 0.7432434366050173,
+ "learning_rate": 9.92286492830925e-07,
+ "loss": 0.2958357810974121,
+ "num_input_tokens_seen": 1034004184,
+ "step": 23645,
+ "train_runtime": 182887.2125,
+ "train_tokens_per_second": 5653.781
+ },
+ {
+ "epoch": 1.5923781308914624,
+ "grad_norm": 0.7076131103520957,
+ "learning_rate": 9.907060573060535e-07,
+ "loss": 0.31322293281555175,
+ "num_input_tokens_seen": 1034231736,
+ "step": 23650,
+ "train_runtime": 182921.8257,
+ "train_tokens_per_second": 5653.955
+ },
+ {
+ "epoch": 1.5927147858874227,
+ "grad_norm": 0.7877155616601498,
+ "learning_rate": 9.891267429674484e-07,
+ "loss": 0.3153421401977539,
+ "num_input_tokens_seen": 1034440472,
+ "step": 23655,
+ "train_runtime": 182960.1809,
+ "train_tokens_per_second": 5653.91
+ },
+ {
+ "epoch": 1.5930514408833827,
+ "grad_norm": 0.7335260153611561,
+ "learning_rate": 9.8754855025676e-07,
+ "loss": 0.28733248710632325,
+ "num_input_tokens_seen": 1034659216,
+ "step": 23660,
+ "train_runtime": 182996.9749,
+ "train_tokens_per_second": 5653.969
+ },
+ {
+ "epoch": 1.5933880958793427,
+ "grad_norm": 0.7659546851429526,
+ "learning_rate": 9.85971479615324e-07,
+ "loss": 0.29557249546051023,
+ "num_input_tokens_seen": 1034888536,
+ "step": 23665,
+ "train_runtime": 183037.5139,
+ "train_tokens_per_second": 5653.97
+ },
+ {
+ "epoch": 1.593724750875303,
+ "grad_norm": 0.7118609901015792,
+ "learning_rate": 9.843955314841647e-07,
+ "loss": 0.28273298740386965,
+ "num_input_tokens_seen": 1035112904,
+ "step": 23670,
+ "train_runtime": 183070.8286,
+ "train_tokens_per_second": 5654.166
+ },
+ {
+ "epoch": 1.5940614058712632,
+ "grad_norm": 0.7513624785773189,
+ "learning_rate": 9.82820706303989e-07,
+ "loss": 0.29895691871643065,
+ "num_input_tokens_seen": 1035338280,
+ "step": 23675,
+ "train_runtime": 183114.0185,
+ "train_tokens_per_second": 5654.063
+ },
+ {
+ "epoch": 1.5943980608672232,
+ "grad_norm": 0.8383529935018983,
+ "learning_rate": 9.812470045151952e-07,
+ "loss": 0.3054283618927002,
+ "num_input_tokens_seen": 1035542816,
+ "step": 23680,
+ "train_runtime": 183151.8635,
+ "train_tokens_per_second": 5654.012
+ },
+ {
+ "epoch": 1.5947347158631833,
+ "grad_norm": 0.7538875645456946,
+ "learning_rate": 9.79674426557859e-07,
+ "loss": 0.29663498401641847,
+ "num_input_tokens_seen": 1035752536,
+ "step": 23685,
+ "train_runtime": 183193.8197,
+ "train_tokens_per_second": 5653.862
+ },
+ {
+ "epoch": 1.5950713708591435,
+ "grad_norm": 0.6945464407471583,
+ "learning_rate": 9.781029728717513e-07,
+ "loss": 0.2824697732925415,
+ "num_input_tokens_seen": 1035967072,
+ "step": 23690,
+ "train_runtime": 183231.5083,
+ "train_tokens_per_second": 5653.87
+ },
+ {
+ "epoch": 1.5954080258551038,
+ "grad_norm": 0.7148723682038612,
+ "learning_rate": 9.765326438963218e-07,
+ "loss": 0.31726827621459963,
+ "num_input_tokens_seen": 1036183056,
+ "step": 23695,
+ "train_runtime": 183268.6587,
+ "train_tokens_per_second": 5653.902
+ },
+ {
+ "epoch": 1.5957446808510638,
+ "grad_norm": 0.7520572590185046,
+ "learning_rate": 9.749634400707087e-07,
+ "loss": 0.29088969230651857,
+ "num_input_tokens_seen": 1036386224,
+ "step": 23700,
+ "train_runtime": 183304.0574,
+ "train_tokens_per_second": 5653.919
+ },
+ {
+ "epoch": 1.5960813358470238,
+ "grad_norm": 0.691989732417355,
+ "learning_rate": 9.733953618337344e-07,
+ "loss": 0.3045504570007324,
+ "num_input_tokens_seen": 1036590880,
+ "step": 23705,
+ "train_runtime": 183345.8827,
+ "train_tokens_per_second": 5653.745
+ },
+ {
+ "epoch": 1.596417990842984,
+ "grad_norm": 0.7360691080511433,
+ "learning_rate": 9.718284096239077e-07,
+ "loss": 0.315799617767334,
+ "num_input_tokens_seen": 1036831896,
+ "step": 23710,
+ "train_runtime": 183384.7735,
+ "train_tokens_per_second": 5653.86
+ },
+ {
+ "epoch": 1.5967546458389443,
+ "grad_norm": 0.780467411583108,
+ "learning_rate": 9.702625838794215e-07,
+ "loss": 0.28366994857788086,
+ "num_input_tokens_seen": 1037051528,
+ "step": 23715,
+ "train_runtime": 183424.345,
+ "train_tokens_per_second": 5653.838
+ },
+ {
+ "epoch": 1.5970913008349044,
+ "grad_norm": 0.7083055249265116,
+ "learning_rate": 9.686978850381535e-07,
+ "loss": 0.28309922218322753,
+ "num_input_tokens_seen": 1037274432,
+ "step": 23720,
+ "train_runtime": 183463.0678,
+ "train_tokens_per_second": 5653.86
+ },
+ {
+ "epoch": 1.5974279558308644,
+ "grad_norm": 0.7040502210402872,
+ "learning_rate": 9.67134313537666e-07,
+ "loss": 0.28775806427001954,
+ "num_input_tokens_seen": 1037487448,
+ "step": 23725,
+ "train_runtime": 183498.8739,
+ "train_tokens_per_second": 5653.917
+ },
+ {
+ "epoch": 1.5977646108268246,
+ "grad_norm": 0.7648445160032851,
+ "learning_rate": 9.6557186981521e-07,
+ "loss": 0.2938704013824463,
+ "num_input_tokens_seen": 1037711184,
+ "step": 23730,
+ "train_runtime": 183532.5183,
+ "train_tokens_per_second": 5654.1
+ },
+ {
+ "epoch": 1.5981012658227849,
+ "grad_norm": 0.7590116723636553,
+ "learning_rate": 9.640105543077133e-07,
+ "loss": 0.29409012794494627,
+ "num_input_tokens_seen": 1037930696,
+ "step": 23735,
+ "train_runtime": 183569.9815,
+ "train_tokens_per_second": 5654.142
+ },
+ {
+ "epoch": 1.598437920818745,
+ "grad_norm": 0.6999829490635439,
+ "learning_rate": 9.624503674517972e-07,
+ "loss": 0.31141953468322753,
+ "num_input_tokens_seen": 1038151728,
+ "step": 23740,
+ "train_runtime": 183616.0849,
+ "train_tokens_per_second": 5653.926
+ },
+ {
+ "epoch": 1.598774575814705,
+ "grad_norm": 0.6931872830922609,
+ "learning_rate": 9.608913096837603e-07,
+ "loss": 0.2785954475402832,
+ "num_input_tokens_seen": 1038375320,
+ "step": 23745,
+ "train_runtime": 183658.3574,
+ "train_tokens_per_second": 5653.842
+ },
+ {
+ "epoch": 1.5991112308106652,
+ "grad_norm": 0.7826269007767029,
+ "learning_rate": 9.593333814395889e-07,
+ "loss": 0.30640616416931155,
+ "num_input_tokens_seen": 1038579600,
+ "step": 23750,
+ "train_runtime": 183690.2536,
+ "train_tokens_per_second": 5653.972
+ },
+ {
+ "epoch": 1.5994478858066254,
+ "grad_norm": 0.659801248076501,
+ "learning_rate": 9.577765831549529e-07,
+ "loss": 0.29427356719970704,
+ "num_input_tokens_seen": 1038803840,
+ "step": 23755,
+ "train_runtime": 183723.8696,
+ "train_tokens_per_second": 5654.158
+ },
+ {
+ "epoch": 1.5997845408025855,
+ "grad_norm": 0.7271704349083015,
+ "learning_rate": 9.562209152652058e-07,
+ "loss": 0.30373120307922363,
+ "num_input_tokens_seen": 1039013760,
+ "step": 23760,
+ "train_runtime": 183766.5421,
+ "train_tokens_per_second": 5653.988
+ },
+ {
+ "epoch": 1.6001211957985455,
+ "grad_norm": 0.8875347712272086,
+ "learning_rate": 9.54666378205385e-07,
+ "loss": 0.3227171182632446,
+ "num_input_tokens_seen": 1039242488,
+ "step": 23765,
+ "train_runtime": 183813.6297,
+ "train_tokens_per_second": 5653.784
+ },
+ {
+ "epoch": 1.6004578507945058,
+ "grad_norm": 0.8111840363520786,
+ "learning_rate": 9.531129724102117e-07,
+ "loss": 0.2936609983444214,
+ "num_input_tokens_seen": 1039454848,
+ "step": 23770,
+ "train_runtime": 183852.3211,
+ "train_tokens_per_second": 5653.749
+ },
+ {
+ "epoch": 1.600794505790466,
+ "grad_norm": 0.7937834563085631,
+ "learning_rate": 9.515606983140896e-07,
+ "loss": 0.2982936859130859,
+ "num_input_tokens_seen": 1039677264,
+ "step": 23775,
+ "train_runtime": 183886.8713,
+ "train_tokens_per_second": 5653.896
+ },
+ {
+ "epoch": 1.601131160786426,
+ "grad_norm": 0.7845551974583694,
+ "learning_rate": 9.500095563511119e-07,
+ "loss": 0.3003803253173828,
+ "num_input_tokens_seen": 1039906360,
+ "step": 23780,
+ "train_runtime": 183931.0004,
+ "train_tokens_per_second": 5653.785
+ },
+ {
+ "epoch": 1.601467815782386,
+ "grad_norm": 0.6656969784743623,
+ "learning_rate": 9.484595469550445e-07,
+ "loss": 0.28896503448486327,
+ "num_input_tokens_seen": 1040131192,
+ "step": 23785,
+ "train_runtime": 183970.5264,
+ "train_tokens_per_second": 5653.793
+ },
+ {
+ "epoch": 1.6018044707783463,
+ "grad_norm": 0.6946762642734486,
+ "learning_rate": 9.469106705593462e-07,
+ "loss": 0.27666046619415285,
+ "num_input_tokens_seen": 1040333032,
+ "step": 23790,
+ "train_runtime": 184009.1267,
+ "train_tokens_per_second": 5653.703
+ },
+ {
+ "epoch": 1.6021411257743066,
+ "grad_norm": 0.7736179336799578,
+ "learning_rate": 9.453629275971549e-07,
+ "loss": 0.29598097801208495,
+ "num_input_tokens_seen": 1040552408,
+ "step": 23795,
+ "train_runtime": 184047.3995,
+ "train_tokens_per_second": 5653.72
+ },
+ {
+ "epoch": 1.6024777807702666,
+ "grad_norm": 0.6664557365960093,
+ "learning_rate": 9.438163185012916e-07,
+ "loss": 0.28930160999298093,
+ "num_input_tokens_seen": 1040764544,
+ "step": 23800,
+ "train_runtime": 184091.6726,
+ "train_tokens_per_second": 5653.512
+ },
+ {
+ "epoch": 1.6028144357662266,
+ "grad_norm": 0.7303797761962147,
+ "learning_rate": 9.422708437042604e-07,
+ "loss": 0.2819626808166504,
+ "num_input_tokens_seen": 1040990752,
+ "step": 23805,
+ "train_runtime": 184135.0608,
+ "train_tokens_per_second": 5653.409
+ },
+ {
+ "epoch": 1.6031510907621869,
+ "grad_norm": 0.7712489855582633,
+ "learning_rate": 9.40726503638249e-07,
+ "loss": 0.31176731586456297,
+ "num_input_tokens_seen": 1041206528,
+ "step": 23810,
+ "train_runtime": 184167.4154,
+ "train_tokens_per_second": 5653.587
+ },
+ {
+ "epoch": 1.6034877457581471,
+ "grad_norm": 0.8231902257626785,
+ "learning_rate": 9.391832987351268e-07,
+ "loss": 0.29704809188842773,
+ "num_input_tokens_seen": 1041421728,
+ "step": 23815,
+ "train_runtime": 184202.3075,
+ "train_tokens_per_second": 5653.684
+ },
+ {
+ "epoch": 1.6038244007541071,
+ "grad_norm": 0.8241864917839168,
+ "learning_rate": 9.376412294264458e-07,
+ "loss": 0.29626688957214353,
+ "num_input_tokens_seen": 1041643056,
+ "step": 23820,
+ "train_runtime": 184245.0385,
+ "train_tokens_per_second": 5653.575
+ },
+ {
+ "epoch": 1.6041610557500672,
+ "grad_norm": 0.6892308812819989,
+ "learning_rate": 9.361002961434401e-07,
+ "loss": 0.29646663665771483,
+ "num_input_tokens_seen": 1041859688,
+ "step": 23825,
+ "train_runtime": 184286.6378,
+ "train_tokens_per_second": 5653.474
+ },
+ {
+ "epoch": 1.6044977107460274,
+ "grad_norm": 0.8147077105803181,
+ "learning_rate": 9.345604993170299e-07,
+ "loss": 0.31340832710266114,
+ "num_input_tokens_seen": 1042088840,
+ "step": 23830,
+ "train_runtime": 184324.2778,
+ "train_tokens_per_second": 5653.563
+ },
+ {
+ "epoch": 1.6048343657419877,
+ "grad_norm": 0.7853709628304697,
+ "learning_rate": 9.330218393778101e-07,
+ "loss": 0.29099485874176023,
+ "num_input_tokens_seen": 1042295560,
+ "step": 23835,
+ "train_runtime": 184359.1879,
+ "train_tokens_per_second": 5653.613
+ },
+ {
+ "epoch": 1.6051710207379477,
+ "grad_norm": 0.8167824477623031,
+ "learning_rate": 9.314843167560656e-07,
+ "loss": 0.2913378715515137,
+ "num_input_tokens_seen": 1042510744,
+ "step": 23840,
+ "train_runtime": 184393.2673,
+ "train_tokens_per_second": 5653.735
+ },
+ {
+ "epoch": 1.6055076757339077,
+ "grad_norm": 0.7205357729428846,
+ "learning_rate": 9.299479318817578e-07,
+ "loss": 0.29028098583221434,
+ "num_input_tokens_seen": 1042725176,
+ "step": 23845,
+ "train_runtime": 184433.6182,
+ "train_tokens_per_second": 5653.661
+ },
+ {
+ "epoch": 1.605844330729868,
+ "grad_norm": 0.766245897972215,
+ "learning_rate": 9.284126851845316e-07,
+ "loss": 0.2781973361968994,
+ "num_input_tokens_seen": 1042960480,
+ "step": 23850,
+ "train_runtime": 184473.0014,
+ "train_tokens_per_second": 5653.73
+ },
+ {
+ "epoch": 1.6061809857258282,
+ "grad_norm": 0.7749016859550374,
+ "learning_rate": 9.268785770937172e-07,
+ "loss": 0.3008106231689453,
+ "num_input_tokens_seen": 1043174632,
+ "step": 23855,
+ "train_runtime": 184513.6412,
+ "train_tokens_per_second": 5653.645
+ },
+ {
+ "epoch": 1.6065176407217883,
+ "grad_norm": 0.8470162466279486,
+ "learning_rate": 9.253456080383178e-07,
+ "loss": 0.3095564842224121,
+ "num_input_tokens_seen": 1043395416,
+ "step": 23860,
+ "train_runtime": 184550.3827,
+ "train_tokens_per_second": 5653.716
+ },
+ {
+ "epoch": 1.6068542957177483,
+ "grad_norm": 0.806209780755035,
+ "learning_rate": 9.238137784470286e-07,
+ "loss": 0.32625675201416016,
+ "num_input_tokens_seen": 1043602960,
+ "step": 23865,
+ "train_runtime": 184582.4698,
+ "train_tokens_per_second": 5653.857
+ },
+ {
+ "epoch": 1.6071909507137085,
+ "grad_norm": 0.7896528179568558,
+ "learning_rate": 9.222830887482153e-07,
+ "loss": 0.2661672353744507,
+ "num_input_tokens_seen": 1043826208,
+ "step": 23870,
+ "train_runtime": 184619.7546,
+ "train_tokens_per_second": 5653.925
+ },
+ {
+ "epoch": 1.6075276057096688,
+ "grad_norm": 0.8946960868592196,
+ "learning_rate": 9.207535393699351e-07,
+ "loss": 0.3128221035003662,
+ "num_input_tokens_seen": 1044034312,
+ "step": 23875,
+ "train_runtime": 184658.2466,
+ "train_tokens_per_second": 5653.873
+ },
+ {
+ "epoch": 1.6078642607056288,
+ "grad_norm": 0.8295488010306658,
+ "learning_rate": 9.192251307399197e-07,
+ "loss": 0.29769034385681153,
+ "num_input_tokens_seen": 1044252784,
+ "step": 23880,
+ "train_runtime": 184694.2905,
+ "train_tokens_per_second": 5653.953
+ },
+ {
+ "epoch": 1.6082009157015889,
+ "grad_norm": 0.7652692381284338,
+ "learning_rate": 9.176978632855842e-07,
+ "loss": 0.2840586185455322,
+ "num_input_tokens_seen": 1044471704,
+ "step": 23885,
+ "train_runtime": 184732.0494,
+ "train_tokens_per_second": 5653.982
+ },
+ {
+ "epoch": 1.608537570697549,
+ "grad_norm": 0.7068737921677537,
+ "learning_rate": 9.161717374340235e-07,
+ "loss": 0.2887522459030151,
+ "num_input_tokens_seen": 1044709256,
+ "step": 23890,
+ "train_runtime": 184770.2163,
+ "train_tokens_per_second": 5654.1
+ },
+ {
+ "epoch": 1.6088742256935094,
+ "grad_norm": 0.865885020527099,
+ "learning_rate": 9.14646753612014e-07,
+ "loss": 0.3057317495346069,
+ "num_input_tokens_seen": 1044932856,
+ "step": 23895,
+ "train_runtime": 184802.9042,
+ "train_tokens_per_second": 5654.31
+ },
+ {
+ "epoch": 1.6092108806894694,
+ "grad_norm": 0.7260701569352984,
+ "learning_rate": 9.131229122460112e-07,
+ "loss": 0.29195594787597656,
+ "num_input_tokens_seen": 1045169176,
+ "step": 23900,
+ "train_runtime": 184847.6389,
+ "train_tokens_per_second": 5654.22
+ },
+ {
+ "epoch": 1.6095475356854294,
+ "grad_norm": 0.7232801129725709,
+ "learning_rate": 9.116002137621566e-07,
+ "loss": 0.30215911865234374,
+ "num_input_tokens_seen": 1045388344,
+ "step": 23905,
+ "train_runtime": 184888.5248,
+ "train_tokens_per_second": 5654.155
+ },
+ {
+ "epoch": 1.6098841906813897,
+ "grad_norm": 0.8623009349911321,
+ "learning_rate": 9.100786585862626e-07,
+ "loss": 0.33268914222717283,
+ "num_input_tokens_seen": 1045590472,
+ "step": 23910,
+ "train_runtime": 184932.0284,
+ "train_tokens_per_second": 5653.918
+ },
+ {
+ "epoch": 1.61022084567735,
+ "grad_norm": 0.8126716994808482,
+ "learning_rate": 9.085582471438326e-07,
+ "loss": 0.3007976531982422,
+ "num_input_tokens_seen": 1045807448,
+ "step": 23915,
+ "train_runtime": 184968.2847,
+ "train_tokens_per_second": 5653.983
+ },
+ {
+ "epoch": 1.61055750067331,
+ "grad_norm": 0.9161097317548726,
+ "learning_rate": 9.070389798600394e-07,
+ "loss": 0.3141302585601807,
+ "num_input_tokens_seen": 1046022008,
+ "step": 23920,
+ "train_runtime": 185008.228,
+ "train_tokens_per_second": 5653.922
+ },
+ {
+ "epoch": 1.61089415566927,
+ "grad_norm": 0.7937859830875522,
+ "learning_rate": 9.055208571597451e-07,
+ "loss": 0.3039546489715576,
+ "num_input_tokens_seen": 1046240136,
+ "step": 23925,
+ "train_runtime": 185046.349,
+ "train_tokens_per_second": 5653.936
+ },
+ {
+ "epoch": 1.6112308106652302,
+ "grad_norm": 0.7922862286264548,
+ "learning_rate": 9.040038794674872e-07,
+ "loss": 0.30896403789520266,
+ "num_input_tokens_seen": 1046460096,
+ "step": 23930,
+ "train_runtime": 185087.8953,
+ "train_tokens_per_second": 5653.855
+ },
+ {
+ "epoch": 1.6115674656611905,
+ "grad_norm": 0.7338711700679107,
+ "learning_rate": 9.024880472074831e-07,
+ "loss": 0.28701066970825195,
+ "num_input_tokens_seen": 1046662408,
+ "step": 23935,
+ "train_runtime": 185132.2278,
+ "train_tokens_per_second": 5653.594
+ },
+ {
+ "epoch": 1.6119041206571505,
+ "grad_norm": 0.7141524995931083,
+ "learning_rate": 9.009733608036308e-07,
+ "loss": 0.30707645416259766,
+ "num_input_tokens_seen": 1046889064,
+ "step": 23940,
+ "train_runtime": 185169.7651,
+ "train_tokens_per_second": 5653.672
+ },
+ {
+ "epoch": 1.6122407756531107,
+ "grad_norm": 0.7277986509792717,
+ "learning_rate": 8.994598206795068e-07,
+ "loss": 0.29622645378112794,
+ "num_input_tokens_seen": 1047104032,
+ "step": 23945,
+ "train_runtime": 185205.0679,
+ "train_tokens_per_second": 5653.755
+ },
+ {
+ "epoch": 1.612577430649071,
+ "grad_norm": 0.735906736712434,
+ "learning_rate": 8.979474272583672e-07,
+ "loss": 0.28798418045043944,
+ "num_input_tokens_seen": 1047312440,
+ "step": 23950,
+ "train_runtime": 185247.3251,
+ "train_tokens_per_second": 5653.59
+ },
+ {
+ "epoch": 1.612914085645031,
+ "grad_norm": 0.7572638632134422,
+ "learning_rate": 8.964361809631517e-07,
+ "loss": 0.3181246280670166,
+ "num_input_tokens_seen": 1047539480,
+ "step": 23955,
+ "train_runtime": 185289.8341,
+ "train_tokens_per_second": 5653.518
+ },
+ {
+ "epoch": 1.613250740640991,
+ "grad_norm": 0.8041739758256639,
+ "learning_rate": 8.949260822164707e-07,
+ "loss": 0.31255478858947755,
+ "num_input_tokens_seen": 1047761560,
+ "step": 23960,
+ "train_runtime": 185326.4519,
+ "train_tokens_per_second": 5653.6
+ },
+ {
+ "epoch": 1.6135873956369513,
+ "grad_norm": 0.7237860013192642,
+ "learning_rate": 8.934171314406226e-07,
+ "loss": 0.2988784074783325,
+ "num_input_tokens_seen": 1047987312,
+ "step": 23965,
+ "train_runtime": 185361.8266,
+ "train_tokens_per_second": 5653.739
+ },
+ {
+ "epoch": 1.6139240506329116,
+ "grad_norm": 0.7942034276838248,
+ "learning_rate": 8.919093290575765e-07,
+ "loss": 0.30140008926391604,
+ "num_input_tokens_seen": 1048194320,
+ "step": 23970,
+ "train_runtime": 185405.0815,
+ "train_tokens_per_second": 5653.536
+ },
+ {
+ "epoch": 1.6142607056288716,
+ "grad_norm": 0.8222295997811078,
+ "learning_rate": 8.904026754889877e-07,
+ "loss": 0.30414538383483886,
+ "num_input_tokens_seen": 1048404056,
+ "step": 23975,
+ "train_runtime": 185444.5814,
+ "train_tokens_per_second": 5653.463
+ },
+ {
+ "epoch": 1.6145973606248316,
+ "grad_norm": 0.8609222631961138,
+ "learning_rate": 8.888971711561867e-07,
+ "loss": 0.32020394802093505,
+ "num_input_tokens_seen": 1048613480,
+ "step": 23980,
+ "train_runtime": 185486.5715,
+ "train_tokens_per_second": 5653.312
+ },
+ {
+ "epoch": 1.6149340156207919,
+ "grad_norm": 0.8930736881841469,
+ "learning_rate": 8.873928164801821e-07,
+ "loss": 0.32930703163146974,
+ "num_input_tokens_seen": 1048836768,
+ "step": 23985,
+ "train_runtime": 185525.6042,
+ "train_tokens_per_second": 5653.326
+ },
+ {
+ "epoch": 1.6152706706167521,
+ "grad_norm": 0.7714631943477365,
+ "learning_rate": 8.858896118816624e-07,
+ "loss": 0.2910914421081543,
+ "num_input_tokens_seen": 1049026440,
+ "step": 23990,
+ "train_runtime": 185568.2914,
+ "train_tokens_per_second": 5653.048
+ },
+ {
+ "epoch": 1.6156073256127121,
+ "grad_norm": 0.759228215290319,
+ "learning_rate": 8.84387557780994e-07,
+ "loss": 0.2792020797729492,
+ "num_input_tokens_seen": 1049235720,
+ "step": 23995,
+ "train_runtime": 185611.4217,
+ "train_tokens_per_second": 5652.862
+ },
+ {
+ "epoch": 1.6159439806086722,
+ "grad_norm": 0.7757710745934641,
+ "learning_rate": 8.828866545982196e-07,
+ "loss": 0.2844292163848877,
+ "num_input_tokens_seen": 1049455984,
+ "step": 24000,
+ "train_runtime": 185652.6171,
+ "train_tokens_per_second": 5652.794
+ },
+ {
+ "epoch": 1.6162806356046324,
+ "grad_norm": 0.7617714598274377,
+ "learning_rate": 8.813869027530664e-07,
+ "loss": 0.28838918209075926,
+ "num_input_tokens_seen": 1049680856,
+ "step": 24005,
+ "train_runtime": 185698.9704,
+ "train_tokens_per_second": 5652.594
+ },
+ {
+ "epoch": 1.6166172906005927,
+ "grad_norm": 0.7142148786302278,
+ "learning_rate": 8.798883026649302e-07,
+ "loss": 0.31560914516448973,
+ "num_input_tokens_seen": 1049914184,
+ "step": 24010,
+ "train_runtime": 185735.3739,
+ "train_tokens_per_second": 5652.742
+ },
+ {
+ "epoch": 1.6169539455965527,
+ "grad_norm": 0.6136150732667002,
+ "learning_rate": 8.783908547528947e-07,
+ "loss": 0.2625638246536255,
+ "num_input_tokens_seen": 1050126648,
+ "step": 24015,
+ "train_runtime": 185774.3978,
+ "train_tokens_per_second": 5652.698
+ },
+ {
+ "epoch": 1.6172906005925127,
+ "grad_norm": 0.7756733989123253,
+ "learning_rate": 8.768945594357109e-07,
+ "loss": 0.29282054901123045,
+ "num_input_tokens_seen": 1050339456,
+ "step": 24020,
+ "train_runtime": 185813.7393,
+ "train_tokens_per_second": 5652.647
+ },
+ {
+ "epoch": 1.617627255588473,
+ "grad_norm": 0.900598117098456,
+ "learning_rate": 8.753994171318175e-07,
+ "loss": 0.3160220146179199,
+ "num_input_tokens_seen": 1050547328,
+ "step": 24025,
+ "train_runtime": 185851.1561,
+ "train_tokens_per_second": 5652.627
+ },
+ {
+ "epoch": 1.6179639105844332,
+ "grad_norm": 0.7405411228891803,
+ "learning_rate": 8.739054282593245e-07,
+ "loss": 0.3164478063583374,
+ "num_input_tokens_seen": 1050771952,
+ "step": 24030,
+ "train_runtime": 185890.3938,
+ "train_tokens_per_second": 5652.643
+ },
+ {
+ "epoch": 1.6183005655803933,
+ "grad_norm": 0.835344576020692,
+ "learning_rate": 8.724125932360217e-07,
+ "loss": 0.3083709716796875,
+ "num_input_tokens_seen": 1050984192,
+ "step": 24035,
+ "train_runtime": 185929.4023,
+ "train_tokens_per_second": 5652.598
+ },
+ {
+ "epoch": 1.6186372205763533,
+ "grad_norm": 0.7799187017517342,
+ "learning_rate": 8.70920912479375e-07,
+ "loss": 0.3174473762512207,
+ "num_input_tokens_seen": 1051196368,
+ "step": 24040,
+ "train_runtime": 185964.2309,
+ "train_tokens_per_second": 5652.68
+ },
+ {
+ "epoch": 1.6189738755723135,
+ "grad_norm": 0.7850184374405595,
+ "learning_rate": 8.694303864065285e-07,
+ "loss": 0.29400835037231443,
+ "num_input_tokens_seen": 1051425808,
+ "step": 24045,
+ "train_runtime": 186005.0047,
+ "train_tokens_per_second": 5652.675
+ },
+ {
+ "epoch": 1.6193105305682738,
+ "grad_norm": 0.7182128587421389,
+ "learning_rate": 8.679410154343015e-07,
+ "loss": 0.29439725875854494,
+ "num_input_tokens_seen": 1051646216,
+ "step": 24050,
+ "train_runtime": 186042.8773,
+ "train_tokens_per_second": 5652.709
+ },
+ {
+ "epoch": 1.6196471855642338,
+ "grad_norm": 0.7969951778334732,
+ "learning_rate": 8.664527999791961e-07,
+ "loss": 0.3235488414764404,
+ "num_input_tokens_seen": 1051864880,
+ "step": 24055,
+ "train_runtime": 186082.635,
+ "train_tokens_per_second": 5652.676
+ },
+ {
+ "epoch": 1.6199838405601938,
+ "grad_norm": 0.7800713274850012,
+ "learning_rate": 8.649657404573819e-07,
+ "loss": 0.29723272323608396,
+ "num_input_tokens_seen": 1052093008,
+ "step": 24060,
+ "train_runtime": 186126.3919,
+ "train_tokens_per_second": 5652.573
+ },
+ {
+ "epoch": 1.620320495556154,
+ "grad_norm": 0.8274231375759122,
+ "learning_rate": 8.634798372847148e-07,
+ "loss": 0.317378568649292,
+ "num_input_tokens_seen": 1052316944,
+ "step": 24065,
+ "train_runtime": 186162.2307,
+ "train_tokens_per_second": 5652.688
+ },
+ {
+ "epoch": 1.6206571505521143,
+ "grad_norm": 0.8056126386548218,
+ "learning_rate": 8.619950908767189e-07,
+ "loss": 0.2666431665420532,
+ "num_input_tokens_seen": 1052516136,
+ "step": 24070,
+ "train_runtime": 186205.2929,
+ "train_tokens_per_second": 5652.45
+ },
+ {
+ "epoch": 1.6209938055480744,
+ "grad_norm": 0.6724696382638301,
+ "learning_rate": 8.605115016486016e-07,
+ "loss": 0.2880539894104004,
+ "num_input_tokens_seen": 1052748400,
+ "step": 24075,
+ "train_runtime": 186243.1874,
+ "train_tokens_per_second": 5652.547
+ },
+ {
+ "epoch": 1.6213304605440344,
+ "grad_norm": 0.9066394567633849,
+ "learning_rate": 8.590290700152426e-07,
+ "loss": 0.2833811044692993,
+ "num_input_tokens_seen": 1052963936,
+ "step": 24080,
+ "train_runtime": 186289.0516,
+ "train_tokens_per_second": 5652.313
+ },
+ {
+ "epoch": 1.6216671155399947,
+ "grad_norm": 0.799201189010084,
+ "learning_rate": 8.575477963912005e-07,
+ "loss": 0.29882135391235354,
+ "num_input_tokens_seen": 1053188096,
+ "step": 24085,
+ "train_runtime": 186329.9225,
+ "train_tokens_per_second": 5652.276
+ },
+ {
+ "epoch": 1.622003770535955,
+ "grad_norm": 0.8755519590986088,
+ "learning_rate": 8.560676811907071e-07,
+ "loss": 0.3093993663787842,
+ "num_input_tokens_seen": 1053409680,
+ "step": 24090,
+ "train_runtime": 186370.6512,
+ "train_tokens_per_second": 5652.229
+ },
+ {
+ "epoch": 1.622340425531915,
+ "grad_norm": 1.5513430020002343,
+ "learning_rate": 8.54588724827673e-07,
+ "loss": 0.29143998622894285,
+ "num_input_tokens_seen": 1053604304,
+ "step": 24095,
+ "train_runtime": 186412.1977,
+ "train_tokens_per_second": 5652.014
+ },
+ {
+ "epoch": 1.622677080527875,
+ "grad_norm": 0.7491596315112452,
+ "learning_rate": 8.531109277156818e-07,
+ "loss": 0.2952460050582886,
+ "num_input_tokens_seen": 1053836472,
+ "step": 24100,
+ "train_runtime": 186445.3192,
+ "train_tokens_per_second": 5652.255
+ },
+ {
+ "epoch": 1.6230137355238352,
+ "grad_norm": 0.6130646121890772,
+ "learning_rate": 8.516342902679986e-07,
+ "loss": 0.26687233448028563,
+ "num_input_tokens_seen": 1054042928,
+ "step": 24105,
+ "train_runtime": 186487.115,
+ "train_tokens_per_second": 5652.095
+ },
+ {
+ "epoch": 1.6233503905197955,
+ "grad_norm": 0.7118183132847752,
+ "learning_rate": 8.501588128975557e-07,
+ "loss": 0.274417781829834,
+ "num_input_tokens_seen": 1054266256,
+ "step": 24110,
+ "train_runtime": 186525.4577,
+ "train_tokens_per_second": 5652.131
+ },
+ {
+ "epoch": 1.6236870455157555,
+ "grad_norm": 0.7790848429598987,
+ "learning_rate": 8.486844960169704e-07,
+ "loss": 0.28249123096466067,
+ "num_input_tokens_seen": 1054484008,
+ "step": 24115,
+ "train_runtime": 186565.9108,
+ "train_tokens_per_second": 5652.072
+ },
+ {
+ "epoch": 1.6240237005117155,
+ "grad_norm": 0.7894998002519212,
+ "learning_rate": 8.472113400385257e-07,
+ "loss": 0.2758941650390625,
+ "num_input_tokens_seen": 1054711848,
+ "step": 24120,
+ "train_runtime": 186605.8621,
+ "train_tokens_per_second": 5652.083
+ },
+ {
+ "epoch": 1.6243603555076758,
+ "grad_norm": 0.7543857092013571,
+ "learning_rate": 8.457393453741886e-07,
+ "loss": 0.3019557476043701,
+ "num_input_tokens_seen": 1054923912,
+ "step": 24125,
+ "train_runtime": 186648.2364,
+ "train_tokens_per_second": 5651.936
+ },
+ {
+ "epoch": 1.624697010503636,
+ "grad_norm": 0.7442690058985666,
+ "learning_rate": 8.442685124355965e-07,
+ "loss": 0.28956966400146483,
+ "num_input_tokens_seen": 1055138184,
+ "step": 24130,
+ "train_runtime": 186689.7766,
+ "train_tokens_per_second": 5651.826
+ },
+ {
+ "epoch": 1.625033665499596,
+ "grad_norm": 0.7707121718508636,
+ "learning_rate": 8.427988416340633e-07,
+ "loss": 0.31919891834259034,
+ "num_input_tokens_seen": 1055341616,
+ "step": 24135,
+ "train_runtime": 186736.5031,
+ "train_tokens_per_second": 5651.501
+ },
+ {
+ "epoch": 1.625370320495556,
+ "grad_norm": 0.8345320866926412,
+ "learning_rate": 8.413303333805777e-07,
+ "loss": 0.2868987560272217,
+ "num_input_tokens_seen": 1055555328,
+ "step": 24140,
+ "train_runtime": 186778.9844,
+ "train_tokens_per_second": 5651.36
+ },
+ {
+ "epoch": 1.6257069754915163,
+ "grad_norm": 0.832357010463937,
+ "learning_rate": 8.398629880858034e-07,
+ "loss": 0.3065673589706421,
+ "num_input_tokens_seen": 1055788776,
+ "step": 24145,
+ "train_runtime": 186814.9017,
+ "train_tokens_per_second": 5651.523
+ },
+ {
+ "epoch": 1.6260436304874766,
+ "grad_norm": 0.7264477849943358,
+ "learning_rate": 8.383968061600778e-07,
+ "loss": 0.30824875831604004,
+ "num_input_tokens_seen": 1055998904,
+ "step": 24150,
+ "train_runtime": 186851.2697,
+ "train_tokens_per_second": 5651.548
+ },
+ {
+ "epoch": 1.6263802854834366,
+ "grad_norm": 0.6983483917587298,
+ "learning_rate": 8.36931788013417e-07,
+ "loss": 0.2875575304031372,
+ "num_input_tokens_seen": 1056224128,
+ "step": 24155,
+ "train_runtime": 186885.9871,
+ "train_tokens_per_second": 5651.703
+ },
+ {
+ "epoch": 1.6267169404793966,
+ "grad_norm": 0.8238457852516533,
+ "learning_rate": 8.354679340555044e-07,
+ "loss": 0.30007798671722413,
+ "num_input_tokens_seen": 1056436296,
+ "step": 24160,
+ "train_runtime": 186918.9055,
+ "train_tokens_per_second": 5651.843
+ },
+ {
+ "epoch": 1.6270535954753569,
+ "grad_norm": 0.7716224609530045,
+ "learning_rate": 8.34005244695707e-07,
+ "loss": 0.28886971473693845,
+ "num_input_tokens_seen": 1056648824,
+ "step": 24165,
+ "train_runtime": 186953.1199,
+ "train_tokens_per_second": 5651.945
+ },
+ {
+ "epoch": 1.6273902504713171,
+ "grad_norm": 0.6680421386267813,
+ "learning_rate": 8.325437203430558e-07,
+ "loss": 0.29457979202270507,
+ "num_input_tokens_seen": 1056867768,
+ "step": 24170,
+ "train_runtime": 186995.0671,
+ "train_tokens_per_second": 5651.848
+ },
+ {
+ "epoch": 1.6277269054672772,
+ "grad_norm": 0.7051397648416529,
+ "learning_rate": 8.310833614062652e-07,
+ "loss": 0.295581316947937,
+ "num_input_tokens_seen": 1057083272,
+ "step": 24175,
+ "train_runtime": 187032.373,
+ "train_tokens_per_second": 5651.873
+ },
+ {
+ "epoch": 1.6280635604632372,
+ "grad_norm": 0.8804185228424846,
+ "learning_rate": 8.296241682937189e-07,
+ "loss": 0.2983792543411255,
+ "num_input_tokens_seen": 1057309512,
+ "step": 24180,
+ "train_runtime": 187074.9311,
+ "train_tokens_per_second": 5651.797
+ },
+ {
+ "epoch": 1.6284002154591974,
+ "grad_norm": 0.7277060223716454,
+ "learning_rate": 8.281661414134761e-07,
+ "loss": 0.2972233295440674,
+ "num_input_tokens_seen": 1057522792,
+ "step": 24185,
+ "train_runtime": 187108.976,
+ "train_tokens_per_second": 5651.908
+ },
+ {
+ "epoch": 1.6287368704551577,
+ "grad_norm": 0.7892086033151113,
+ "learning_rate": 8.267092811732686e-07,
+ "loss": 0.31084840297698973,
+ "num_input_tokens_seen": 1057745072,
+ "step": 24190,
+ "train_runtime": 187148.6775,
+ "train_tokens_per_second": 5651.897
+ },
+ {
+ "epoch": 1.6290735254511177,
+ "grad_norm": 0.730388450752643,
+ "learning_rate": 8.252535879805029e-07,
+ "loss": 0.2759894847869873,
+ "num_input_tokens_seen": 1057972008,
+ "step": 24195,
+ "train_runtime": 187187.3763,
+ "train_tokens_per_second": 5651.941
+ },
+ {
+ "epoch": 1.6294101804470777,
+ "grad_norm": 0.863594940750271,
+ "learning_rate": 8.237990622422576e-07,
+ "loss": 0.289616322517395,
+ "num_input_tokens_seen": 1058191104,
+ "step": 24200,
+ "train_runtime": 187223.8174,
+ "train_tokens_per_second": 5652.011
+ },
+ {
+ "epoch": 1.629746835443038,
+ "grad_norm": 0.7266468720434132,
+ "learning_rate": 8.223457043652905e-07,
+ "loss": 0.3101085662841797,
+ "num_input_tokens_seen": 1058412784,
+ "step": 24205,
+ "train_runtime": 187256.143,
+ "train_tokens_per_second": 5652.219
+ },
+ {
+ "epoch": 1.6300834904389983,
+ "grad_norm": 0.9420820301682334,
+ "learning_rate": 8.208935147560227e-07,
+ "loss": 0.2965271472930908,
+ "num_input_tokens_seen": 1058609720,
+ "step": 24210,
+ "train_runtime": 187288.1131,
+ "train_tokens_per_second": 5652.306
+ },
+ {
+ "epoch": 1.6304201454349583,
+ "grad_norm": 0.7345571411845794,
+ "learning_rate": 8.194424938205597e-07,
+ "loss": 0.3259326696395874,
+ "num_input_tokens_seen": 1058845344,
+ "step": 24215,
+ "train_runtime": 187328.0703,
+ "train_tokens_per_second": 5652.358
+ },
+ {
+ "epoch": 1.6307568004309183,
+ "grad_norm": 0.7486771661318714,
+ "learning_rate": 8.179926419646705e-07,
+ "loss": 0.28400454521179197,
+ "num_input_tokens_seen": 1059062096,
+ "step": 24220,
+ "train_runtime": 187371.6715,
+ "train_tokens_per_second": 5652.2
+ },
+ {
+ "epoch": 1.6310934554268786,
+ "grad_norm": 0.8281741934458996,
+ "learning_rate": 8.165439595938047e-07,
+ "loss": 0.32569429874420164,
+ "num_input_tokens_seen": 1059262648,
+ "step": 24225,
+ "train_runtime": 187414.9864,
+ "train_tokens_per_second": 5651.963
+ },
+ {
+ "epoch": 1.6314301104228388,
+ "grad_norm": 0.8133179081823259,
+ "learning_rate": 8.150964471130801e-07,
+ "loss": 0.2735768795013428,
+ "num_input_tokens_seen": 1059491256,
+ "step": 24230,
+ "train_runtime": 187448.4434,
+ "train_tokens_per_second": 5652.174
+ },
+ {
+ "epoch": 1.6317667654187988,
+ "grad_norm": 0.7466167890530678,
+ "learning_rate": 8.136501049272899e-07,
+ "loss": 0.30480170249938965,
+ "num_input_tokens_seen": 1059706760,
+ "step": 24235,
+ "train_runtime": 187485.3647,
+ "train_tokens_per_second": 5652.211
+ },
+ {
+ "epoch": 1.6321034204147589,
+ "grad_norm": 0.7655169967068027,
+ "learning_rate": 8.122049334408983e-07,
+ "loss": 0.29867599010467527,
+ "num_input_tokens_seen": 1059920888,
+ "step": 24240,
+ "train_runtime": 187522.586,
+ "train_tokens_per_second": 5652.231
+ },
+ {
+ "epoch": 1.6324400754107191,
+ "grad_norm": 0.7048483211838898,
+ "learning_rate": 8.107609330580429e-07,
+ "loss": 0.29615049362182616,
+ "num_input_tokens_seen": 1060137784,
+ "step": 24245,
+ "train_runtime": 187563.0743,
+ "train_tokens_per_second": 5652.167
+ },
+ {
+ "epoch": 1.6327767304066794,
+ "grad_norm": 0.7788326112043563,
+ "learning_rate": 8.093181041825332e-07,
+ "loss": 0.3028998613357544,
+ "num_input_tokens_seen": 1060360840,
+ "step": 24250,
+ "train_runtime": 187603.6663,
+ "train_tokens_per_second": 5652.133
+ },
+ {
+ "epoch": 1.6331133854026394,
+ "grad_norm": 0.8246354271264723,
+ "learning_rate": 8.078764472178552e-07,
+ "loss": 0.29851455688476564,
+ "num_input_tokens_seen": 1060573408,
+ "step": 24255,
+ "train_runtime": 187643.6101,
+ "train_tokens_per_second": 5652.062
+ },
+ {
+ "epoch": 1.6334500403985994,
+ "grad_norm": 0.7136275615926145,
+ "learning_rate": 8.06435962567158e-07,
+ "loss": 0.32530434131622316,
+ "num_input_tokens_seen": 1060766968,
+ "step": 24260,
+ "train_runtime": 187685.5425,
+ "train_tokens_per_second": 5651.831
+ },
+ {
+ "epoch": 1.6337866953945597,
+ "grad_norm": 0.7353400312549586,
+ "learning_rate": 8.049966506332746e-07,
+ "loss": 0.26018400192260743,
+ "num_input_tokens_seen": 1060985712,
+ "step": 24265,
+ "train_runtime": 187718.4957,
+ "train_tokens_per_second": 5652.004
+ },
+ {
+ "epoch": 1.63412335039052,
+ "grad_norm": 0.736074674468452,
+ "learning_rate": 8.035585118186984e-07,
+ "loss": 0.2798137664794922,
+ "num_input_tokens_seen": 1061195896,
+ "step": 24270,
+ "train_runtime": 187751.2703,
+ "train_tokens_per_second": 5652.137
+ },
+ {
+ "epoch": 1.63446000538648,
+ "grad_norm": 0.7545879942996399,
+ "learning_rate": 8.021215465256038e-07,
+ "loss": 0.2788818836212158,
+ "num_input_tokens_seen": 1061420544,
+ "step": 24275,
+ "train_runtime": 187795.6161,
+ "train_tokens_per_second": 5651.999
+ },
+ {
+ "epoch": 1.63479666038244,
+ "grad_norm": 0.7515197813453762,
+ "learning_rate": 8.006857551558328e-07,
+ "loss": 0.30008883476257325,
+ "num_input_tokens_seen": 1061636496,
+ "step": 24280,
+ "train_runtime": 187842.4419,
+ "train_tokens_per_second": 5651.739
+ },
+ {
+ "epoch": 1.6351333153784002,
+ "grad_norm": 0.8049426928966876,
+ "learning_rate": 7.992511381108997e-07,
+ "loss": 0.3101008415222168,
+ "num_input_tokens_seen": 1061878752,
+ "step": 24285,
+ "train_runtime": 187875.4149,
+ "train_tokens_per_second": 5652.037
+ },
+ {
+ "epoch": 1.6354699703743605,
+ "grad_norm": 0.7928126374708939,
+ "learning_rate": 7.978176957919909e-07,
+ "loss": 0.3050835132598877,
+ "num_input_tokens_seen": 1062088984,
+ "step": 24290,
+ "train_runtime": 187921.6004,
+ "train_tokens_per_second": 5651.766
+ },
+ {
+ "epoch": 1.6358066253703205,
+ "grad_norm": 0.660553687086565,
+ "learning_rate": 7.963854285999634e-07,
+ "loss": 0.27643623352050783,
+ "num_input_tokens_seen": 1062308496,
+ "step": 24295,
+ "train_runtime": 187966.5083,
+ "train_tokens_per_second": 5651.584
+ },
+ {
+ "epoch": 1.6361432803662805,
+ "grad_norm": 0.8664436436523667,
+ "learning_rate": 7.949543369353452e-07,
+ "loss": 0.30392279624938967,
+ "num_input_tokens_seen": 1062508472,
+ "step": 24300,
+ "train_runtime": 188008.6162,
+ "train_tokens_per_second": 5651.382
+ },
+ {
+ "epoch": 1.6364799353622408,
+ "grad_norm": 0.7681401573545563,
+ "learning_rate": 7.9352442119834e-07,
+ "loss": 0.3222590446472168,
+ "num_input_tokens_seen": 1062721224,
+ "step": 24305,
+ "train_runtime": 188049.5257,
+ "train_tokens_per_second": 5651.284
+ },
+ {
+ "epoch": 1.636816590358201,
+ "grad_norm": 0.7578044675754301,
+ "learning_rate": 7.920956817888148e-07,
+ "loss": 0.3079147577285767,
+ "num_input_tokens_seen": 1062938784,
+ "step": 24310,
+ "train_runtime": 188085.9323,
+ "train_tokens_per_second": 5651.347
+ },
+ {
+ "epoch": 1.637153245354161,
+ "grad_norm": 0.8771063708461605,
+ "learning_rate": 7.906681191063165e-07,
+ "loss": 0.2923039436340332,
+ "num_input_tokens_seen": 1063160648,
+ "step": 24315,
+ "train_runtime": 188122.7957,
+ "train_tokens_per_second": 5651.418
+ },
+ {
+ "epoch": 1.637489900350121,
+ "grad_norm": 0.7603147520393745,
+ "learning_rate": 7.892417335500541e-07,
+ "loss": 0.3094671964645386,
+ "num_input_tokens_seen": 1063371504,
+ "step": 24320,
+ "train_runtime": 188158.9471,
+ "train_tokens_per_second": 5651.453
+ },
+ {
+ "epoch": 1.6378265553460813,
+ "grad_norm": 0.828515228743802,
+ "learning_rate": 7.878165255189146e-07,
+ "loss": 0.32140753269195554,
+ "num_input_tokens_seen": 1063574440,
+ "step": 24325,
+ "train_runtime": 188196.1254,
+ "train_tokens_per_second": 5651.415
+ },
+ {
+ "epoch": 1.6381632103420416,
+ "grad_norm": 0.8506425251973367,
+ "learning_rate": 7.863924954114522e-07,
+ "loss": 0.28619048595428465,
+ "num_input_tokens_seen": 1063782584,
+ "step": 24330,
+ "train_runtime": 188232.3371,
+ "train_tokens_per_second": 5651.434
+ },
+ {
+ "epoch": 1.6384998653380016,
+ "grad_norm": 0.6750210682265397,
+ "learning_rate": 7.849696436258919e-07,
+ "loss": 0.2759582042694092,
+ "num_input_tokens_seen": 1063995392,
+ "step": 24335,
+ "train_runtime": 188272.2346,
+ "train_tokens_per_second": 5651.366
+ },
+ {
+ "epoch": 1.6388365203339617,
+ "grad_norm": 0.71118679601102,
+ "learning_rate": 7.835479705601307e-07,
+ "loss": 0.30858869552612306,
+ "num_input_tokens_seen": 1064201992,
+ "step": 24340,
+ "train_runtime": 188310.5847,
+ "train_tokens_per_second": 5651.313
+ },
+ {
+ "epoch": 1.639173175329922,
+ "grad_norm": 0.7067326301637101,
+ "learning_rate": 7.821274766117337e-07,
+ "loss": 0.3018312692642212,
+ "num_input_tokens_seen": 1064418504,
+ "step": 24345,
+ "train_runtime": 188354.2732,
+ "train_tokens_per_second": 5651.151
+ },
+ {
+ "epoch": 1.6395098303258822,
+ "grad_norm": 0.7701230824468515,
+ "learning_rate": 7.807081621779367e-07,
+ "loss": 0.3180276393890381,
+ "num_input_tokens_seen": 1064640776,
+ "step": 24350,
+ "train_runtime": 188396.5704,
+ "train_tokens_per_second": 5651.062
+ },
+ {
+ "epoch": 1.6398464853218422,
+ "grad_norm": 0.7453739839923318,
+ "learning_rate": 7.792900276556509e-07,
+ "loss": 0.30577759742736815,
+ "num_input_tokens_seen": 1064866144,
+ "step": 24355,
+ "train_runtime": 188437.6782,
+ "train_tokens_per_second": 5651.026
+ },
+ {
+ "epoch": 1.6401831403178022,
+ "grad_norm": 0.7840785820426818,
+ "learning_rate": 7.778730734414469e-07,
+ "loss": 0.3030556201934814,
+ "num_input_tokens_seen": 1065082080,
+ "step": 24360,
+ "train_runtime": 188478.6851,
+ "train_tokens_per_second": 5650.942
+ },
+ {
+ "epoch": 1.6405197953137625,
+ "grad_norm": 0.7537497153937553,
+ "learning_rate": 7.764572999315772e-07,
+ "loss": 0.2959816217422485,
+ "num_input_tokens_seen": 1065308184,
+ "step": 24365,
+ "train_runtime": 188514.0839,
+ "train_tokens_per_second": 5651.08
+ },
+ {
+ "epoch": 1.6408564503097227,
+ "grad_norm": 0.7520641802122896,
+ "learning_rate": 7.750427075219536e-07,
+ "loss": 0.3010838031768799,
+ "num_input_tokens_seen": 1065519384,
+ "step": 24370,
+ "train_runtime": 188554.9706,
+ "train_tokens_per_second": 5650.975
+ },
+ {
+ "epoch": 1.6411931053056827,
+ "grad_norm": 0.8548907262270895,
+ "learning_rate": 7.736292966081655e-07,
+ "loss": 0.28882603645324706,
+ "num_input_tokens_seen": 1065712464,
+ "step": 24375,
+ "train_runtime": 188592.4974,
+ "train_tokens_per_second": 5650.874
+ },
+ {
+ "epoch": 1.6415297603016428,
+ "grad_norm": 0.6533945003956135,
+ "learning_rate": 7.722170675854673e-07,
+ "loss": 0.26800405979156494,
+ "num_input_tokens_seen": 1065939064,
+ "step": 24380,
+ "train_runtime": 188629.38,
+ "train_tokens_per_second": 5650.971
+ },
+ {
+ "epoch": 1.641866415297603,
+ "grad_norm": 0.7624504132460348,
+ "learning_rate": 7.70806020848785e-07,
+ "loss": 0.2844860076904297,
+ "num_input_tokens_seen": 1066159000,
+ "step": 24385,
+ "train_runtime": 188660.9475,
+ "train_tokens_per_second": 5651.191
+ },
+ {
+ "epoch": 1.6422030702935633,
+ "grad_norm": 0.7515316499838955,
+ "learning_rate": 7.693961567927133e-07,
+ "loss": 0.2823920726776123,
+ "num_input_tokens_seen": 1066381848,
+ "step": 24390,
+ "train_runtime": 188700.2202,
+ "train_tokens_per_second": 5651.196
+ },
+ {
+ "epoch": 1.6425397252895233,
+ "grad_norm": 0.7063503273233152,
+ "learning_rate": 7.679874758115153e-07,
+ "loss": 0.2843328475952148,
+ "num_input_tokens_seen": 1066612256,
+ "step": 24395,
+ "train_runtime": 188738.2146,
+ "train_tokens_per_second": 5651.279
+ },
+ {
+ "epoch": 1.6428763802854833,
+ "grad_norm": 0.697327098375531,
+ "learning_rate": 7.66579978299124e-07,
+ "loss": 0.30366389751434325,
+ "num_input_tokens_seen": 1066840032,
+ "step": 24400,
+ "train_runtime": 188775.3375,
+ "train_tokens_per_second": 5651.374
+ },
+ {
+ "epoch": 1.6432130352814436,
+ "grad_norm": 1.0022326064476843,
+ "learning_rate": 7.651736646491447e-07,
+ "loss": 0.32686214447021483,
+ "num_input_tokens_seen": 1067042480,
+ "step": 24405,
+ "train_runtime": 188812.9262,
+ "train_tokens_per_second": 5651.321
+ },
+ {
+ "epoch": 1.6435496902774038,
+ "grad_norm": 0.623590025551479,
+ "learning_rate": 7.63768535254844e-07,
+ "loss": 0.27104051113128663,
+ "num_input_tokens_seen": 1067263472,
+ "step": 24410,
+ "train_runtime": 188849.6348,
+ "train_tokens_per_second": 5651.393
+ },
+ {
+ "epoch": 1.6438863452733639,
+ "grad_norm": 0.830785224698193,
+ "learning_rate": 7.623645905091664e-07,
+ "loss": 0.3036444902420044,
+ "num_input_tokens_seen": 1067483856,
+ "step": 24415,
+ "train_runtime": 188892.9648,
+ "train_tokens_per_second": 5651.263
+ },
+ {
+ "epoch": 1.6442230002693239,
+ "grad_norm": 0.8051008708219101,
+ "learning_rate": 7.609618308047156e-07,
+ "loss": 0.3053473472595215,
+ "num_input_tokens_seen": 1067700112,
+ "step": 24420,
+ "train_runtime": 188925.2763,
+ "train_tokens_per_second": 5651.441
+ },
+ {
+ "epoch": 1.6445596552652841,
+ "grad_norm": 0.7189604815408457,
+ "learning_rate": 7.595602565337729e-07,
+ "loss": 0.29787912368774416,
+ "num_input_tokens_seen": 1067928768,
+ "step": 24425,
+ "train_runtime": 188973.0073,
+ "train_tokens_per_second": 5651.224
+ },
+ {
+ "epoch": 1.6448963102612444,
+ "grad_norm": 0.8282739062629259,
+ "learning_rate": 7.581598680882829e-07,
+ "loss": 0.31963634490966797,
+ "num_input_tokens_seen": 1068150624,
+ "step": 24430,
+ "train_runtime": 189012.1223,
+ "train_tokens_per_second": 5651.228
+ },
+ {
+ "epoch": 1.6452329652572044,
+ "grad_norm": 0.8570504742841146,
+ "learning_rate": 7.567606658598592e-07,
+ "loss": 0.29708251953125,
+ "num_input_tokens_seen": 1068367600,
+ "step": 24435,
+ "train_runtime": 189050.0223,
+ "train_tokens_per_second": 5651.243
+ },
+ {
+ "epoch": 1.6455696202531644,
+ "grad_norm": 0.7274203559032685,
+ "learning_rate": 7.553626502397848e-07,
+ "loss": 0.3157409906387329,
+ "num_input_tokens_seen": 1068589480,
+ "step": 24440,
+ "train_runtime": 189094.3938,
+ "train_tokens_per_second": 5651.09
+ },
+ {
+ "epoch": 1.6459062752491247,
+ "grad_norm": 0.7307127250711936,
+ "learning_rate": 7.539658216190099e-07,
+ "loss": 0.2788280248641968,
+ "num_input_tokens_seen": 1068810256,
+ "step": 24445,
+ "train_runtime": 189133.9599,
+ "train_tokens_per_second": 5651.075
+ },
+ {
+ "epoch": 1.646242930245085,
+ "grad_norm": 0.8440106593735498,
+ "learning_rate": 7.525701803881525e-07,
+ "loss": 0.305745792388916,
+ "num_input_tokens_seen": 1069020600,
+ "step": 24450,
+ "train_runtime": 189174.0585,
+ "train_tokens_per_second": 5650.989
+ },
+ {
+ "epoch": 1.646579585241045,
+ "grad_norm": 0.9140651733358113,
+ "learning_rate": 7.511757269375019e-07,
+ "loss": 0.2875369548797607,
+ "num_input_tokens_seen": 1069228640,
+ "step": 24455,
+ "train_runtime": 189220.4481,
+ "train_tokens_per_second": 5650.703
+ },
+ {
+ "epoch": 1.646916240237005,
+ "grad_norm": 0.767638448585773,
+ "learning_rate": 7.497824616570087e-07,
+ "loss": 0.31629951000213624,
+ "num_input_tokens_seen": 1069428344,
+ "step": 24460,
+ "train_runtime": 189254.1834,
+ "train_tokens_per_second": 5650.751
+ },
+ {
+ "epoch": 1.6472528952329653,
+ "grad_norm": 0.7331009881607108,
+ "learning_rate": 7.483903849362995e-07,
+ "loss": 0.3073263168334961,
+ "num_input_tokens_seen": 1069649936,
+ "step": 24465,
+ "train_runtime": 189293.0368,
+ "train_tokens_per_second": 5650.762
+ },
+ {
+ "epoch": 1.6475895502289255,
+ "grad_norm": 0.7388708195492701,
+ "learning_rate": 7.469994971646594e-07,
+ "loss": 0.3005094289779663,
+ "num_input_tokens_seen": 1069880992,
+ "step": 24470,
+ "train_runtime": 189337.0814,
+ "train_tokens_per_second": 5650.668
+ },
+ {
+ "epoch": 1.6479262052248855,
+ "grad_norm": 0.7085554942280863,
+ "learning_rate": 7.456097987310495e-07,
+ "loss": 0.3060446262359619,
+ "num_input_tokens_seen": 1070091864,
+ "step": 24475,
+ "train_runtime": 189375.3021,
+ "train_tokens_per_second": 5650.641
+ },
+ {
+ "epoch": 1.6482628602208456,
+ "grad_norm": 0.9191804958073158,
+ "learning_rate": 7.442212900240931e-07,
+ "loss": 0.3100743770599365,
+ "num_input_tokens_seen": 1070303776,
+ "step": 24480,
+ "train_runtime": 189415.09,
+ "train_tokens_per_second": 5650.573
+ },
+ {
+ "epoch": 1.6485995152168058,
+ "grad_norm": 0.7298994648086058,
+ "learning_rate": 7.42833971432082e-07,
+ "loss": 0.2700855255126953,
+ "num_input_tokens_seen": 1070514192,
+ "step": 24485,
+ "train_runtime": 189447.1005,
+ "train_tokens_per_second": 5650.729
+ },
+ {
+ "epoch": 1.648936170212766,
+ "grad_norm": 0.7999546159714445,
+ "learning_rate": 7.414478433429756e-07,
+ "loss": 0.3064833641052246,
+ "num_input_tokens_seen": 1070735968,
+ "step": 24490,
+ "train_runtime": 189484.8803,
+ "train_tokens_per_second": 5650.773
+ },
+ {
+ "epoch": 1.649272825208726,
+ "grad_norm": 0.7881065041204545,
+ "learning_rate": 7.400629061444003e-07,
+ "loss": 0.29147264957427976,
+ "num_input_tokens_seen": 1070953008,
+ "step": 24495,
+ "train_runtime": 189528.9666,
+ "train_tokens_per_second": 5650.603
+ },
+ {
+ "epoch": 1.6496094802046861,
+ "grad_norm": 0.6964365331001868,
+ "learning_rate": 7.386791602236482e-07,
+ "loss": 0.32079496383666994,
+ "num_input_tokens_seen": 1071182864,
+ "step": 24500,
+ "train_runtime": 189563.7574,
+ "train_tokens_per_second": 5650.779
+ },
+ {
+ "epoch": 1.6499461352006464,
+ "grad_norm": 0.7757976985222431,
+ "learning_rate": 7.372966059676834e-07,
+ "loss": 0.34608256816864014,
+ "num_input_tokens_seen": 1071404416,
+ "step": 24505,
+ "train_runtime": 189604.7721,
+ "train_tokens_per_second": 5650.725
+ },
+ {
+ "epoch": 1.6502827901966066,
+ "grad_norm": 0.7009325679276259,
+ "learning_rate": 7.359152437631273e-07,
+ "loss": 0.31566901206970216,
+ "num_input_tokens_seen": 1071628384,
+ "step": 24510,
+ "train_runtime": 189642.0172,
+ "train_tokens_per_second": 5650.796
+ },
+ {
+ "epoch": 1.6506194451925666,
+ "grad_norm": 0.7512626090141364,
+ "learning_rate": 7.345350739962781e-07,
+ "loss": 0.29537224769592285,
+ "num_input_tokens_seen": 1071849024,
+ "step": 24515,
+ "train_runtime": 189682.4667,
+ "train_tokens_per_second": 5650.754
+ },
+ {
+ "epoch": 1.6509561001885267,
+ "grad_norm": 0.7652445412623243,
+ "learning_rate": 7.331560970530921e-07,
+ "loss": 0.29905951023101807,
+ "num_input_tokens_seen": 1072071032,
+ "step": 24520,
+ "train_runtime": 189724.0471,
+ "train_tokens_per_second": 5650.686
+ },
+ {
+ "epoch": 1.651292755184487,
+ "grad_norm": 0.6948758340555818,
+ "learning_rate": 7.317783133191986e-07,
+ "loss": 0.30941028594970704,
+ "num_input_tokens_seen": 1072293992,
+ "step": 24525,
+ "train_runtime": 189754.0392,
+ "train_tokens_per_second": 5650.968
+ },
+ {
+ "epoch": 1.6516294101804472,
+ "grad_norm": 0.7212176612484024,
+ "learning_rate": 7.304017231798893e-07,
+ "loss": 0.2879075050354004,
+ "num_input_tokens_seen": 1072519464,
+ "step": 24530,
+ "train_runtime": 189793.2385,
+ "train_tokens_per_second": 5650.989
+ },
+ {
+ "epoch": 1.6519660651764072,
+ "grad_norm": 0.8807710558741058,
+ "learning_rate": 7.290263270201231e-07,
+ "loss": 0.29528236389160156,
+ "num_input_tokens_seen": 1072738280,
+ "step": 24535,
+ "train_runtime": 189830.8758,
+ "train_tokens_per_second": 5651.021
+ },
+ {
+ "epoch": 1.6523027201723672,
+ "grad_norm": 0.7720732199868177,
+ "learning_rate": 7.276521252245261e-07,
+ "loss": 0.28215551376342773,
+ "num_input_tokens_seen": 1072948632,
+ "step": 24540,
+ "train_runtime": 189866.7872,
+ "train_tokens_per_second": 5651.06
+ },
+ {
+ "epoch": 1.6526393751683275,
+ "grad_norm": 0.8981368488565471,
+ "learning_rate": 7.262791181773882e-07,
+ "loss": 0.29338951110839845,
+ "num_input_tokens_seen": 1073174384,
+ "step": 24545,
+ "train_runtime": 189911.0819,
+ "train_tokens_per_second": 5650.931
+ },
+ {
+ "epoch": 1.6529760301642877,
+ "grad_norm": 0.8332745819366663,
+ "learning_rate": 7.24907306262666e-07,
+ "loss": 0.30844817161560056,
+ "num_input_tokens_seen": 1073386448,
+ "step": 24550,
+ "train_runtime": 189955.2277,
+ "train_tokens_per_second": 5650.734
+ },
+ {
+ "epoch": 1.6533126851602478,
+ "grad_norm": 0.7405815929469131,
+ "learning_rate": 7.235366898639856e-07,
+ "loss": 0.3130614995956421,
+ "num_input_tokens_seen": 1073610192,
+ "step": 24555,
+ "train_runtime": 189996.8678,
+ "train_tokens_per_second": 5650.673
+ },
+ {
+ "epoch": 1.6536493401562078,
+ "grad_norm": 0.7968518861817051,
+ "learning_rate": 7.221672693646309e-07,
+ "loss": 0.3168208122253418,
+ "num_input_tokens_seen": 1073834968,
+ "step": 24560,
+ "train_runtime": 190036.3942,
+ "train_tokens_per_second": 5650.681
+ },
+ {
+ "epoch": 1.653985995152168,
+ "grad_norm": 0.7509915696488056,
+ "learning_rate": 7.207990451475594e-07,
+ "loss": 0.30944645404815674,
+ "num_input_tokens_seen": 1074068032,
+ "step": 24565,
+ "train_runtime": 190069.3229,
+ "train_tokens_per_second": 5650.928
+ },
+ {
+ "epoch": 1.6543226501481283,
+ "grad_norm": 0.8218450305806543,
+ "learning_rate": 7.194320175953901e-07,
+ "loss": 0.3064124584197998,
+ "num_input_tokens_seen": 1074280488,
+ "step": 24570,
+ "train_runtime": 190114.2338,
+ "train_tokens_per_second": 5650.71
+ },
+ {
+ "epoch": 1.6546593051440883,
+ "grad_norm": 0.7421711436175688,
+ "learning_rate": 7.180661870904066e-07,
+ "loss": 0.285313081741333,
+ "num_input_tokens_seen": 1074502024,
+ "step": 24575,
+ "train_runtime": 190155.2205,
+ "train_tokens_per_second": 5650.658
+ },
+ {
+ "epoch": 1.6549959601400483,
+ "grad_norm": 0.7026987080451035,
+ "learning_rate": 7.167015540145599e-07,
+ "loss": 0.2826228618621826,
+ "num_input_tokens_seen": 1074710072,
+ "step": 24580,
+ "train_runtime": 190194.1052,
+ "train_tokens_per_second": 5650.596
+ },
+ {
+ "epoch": 1.6553326151360086,
+ "grad_norm": 0.7787998439821507,
+ "learning_rate": 7.153381187494657e-07,
+ "loss": 0.3077234745025635,
+ "num_input_tokens_seen": 1074926392,
+ "step": 24585,
+ "train_runtime": 190228.7504,
+ "train_tokens_per_second": 5650.704
+ },
+ {
+ "epoch": 1.6556692701319689,
+ "grad_norm": 0.7756508425427905,
+ "learning_rate": 7.139758816764036e-07,
+ "loss": 0.28848824501037595,
+ "num_input_tokens_seen": 1075131776,
+ "step": 24590,
+ "train_runtime": 190266.2138,
+ "train_tokens_per_second": 5650.671
+ },
+ {
+ "epoch": 1.6560059251279289,
+ "grad_norm": 0.6980517503857505,
+ "learning_rate": 7.126148431763191e-07,
+ "loss": 0.28489990234375,
+ "num_input_tokens_seen": 1075362760,
+ "step": 24595,
+ "train_runtime": 190301.6247,
+ "train_tokens_per_second": 5650.833
+ },
+ {
+ "epoch": 1.656342580123889,
+ "grad_norm": 0.7359183935227057,
+ "learning_rate": 7.11255003629821e-07,
+ "loss": 0.30712342262268066,
+ "num_input_tokens_seen": 1075591552,
+ "step": 24600,
+ "train_runtime": 190339.4831,
+ "train_tokens_per_second": 5650.911
+ },
+ {
+ "epoch": 1.6566792351198492,
+ "grad_norm": 0.7209263296219092,
+ "learning_rate": 7.098963634171874e-07,
+ "loss": 0.3017916679382324,
+ "num_input_tokens_seen": 1075828544,
+ "step": 24605,
+ "train_runtime": 190376.2279,
+ "train_tokens_per_second": 5651.066
+ },
+ {
+ "epoch": 1.6570158901158094,
+ "grad_norm": 0.8863531557142312,
+ "learning_rate": 7.085389229183537e-07,
+ "loss": 0.3059419631958008,
+ "num_input_tokens_seen": 1076024056,
+ "step": 24610,
+ "train_runtime": 190418.8793,
+ "train_tokens_per_second": 5650.827
+ },
+ {
+ "epoch": 1.6573525451117694,
+ "grad_norm": 0.812027087160403,
+ "learning_rate": 7.071826825129264e-07,
+ "loss": 0.28927106857299806,
+ "num_input_tokens_seen": 1076235696,
+ "step": 24615,
+ "train_runtime": 190461.2181,
+ "train_tokens_per_second": 5650.682
+ },
+ {
+ "epoch": 1.6576892001077295,
+ "grad_norm": 0.7711258281303948,
+ "learning_rate": 7.058276425801735e-07,
+ "loss": 0.30096821784973143,
+ "num_input_tokens_seen": 1076456872,
+ "step": 24620,
+ "train_runtime": 190502.0918,
+ "train_tokens_per_second": 5650.63
+ },
+ {
+ "epoch": 1.6580258551036897,
+ "grad_norm": 0.7870890841283721,
+ "learning_rate": 7.044738034990273e-07,
+ "loss": 0.29817709922790525,
+ "num_input_tokens_seen": 1076662016,
+ "step": 24625,
+ "train_runtime": 190533.526,
+ "train_tokens_per_second": 5650.775
+ },
+ {
+ "epoch": 1.65836251009965,
+ "grad_norm": 0.7826364707728534,
+ "learning_rate": 7.031211656480841e-07,
+ "loss": 0.3162134885787964,
+ "num_input_tokens_seen": 1076875184,
+ "step": 24630,
+ "train_runtime": 190573.3113,
+ "train_tokens_per_second": 5650.714
+ },
+ {
+ "epoch": 1.65869916509561,
+ "grad_norm": 0.7671703072883405,
+ "learning_rate": 7.017697294056058e-07,
+ "loss": 0.27853074073791506,
+ "num_input_tokens_seen": 1077089776,
+ "step": 24635,
+ "train_runtime": 190615.0578,
+ "train_tokens_per_second": 5650.602
+ },
+ {
+ "epoch": 1.65903582009157,
+ "grad_norm": 0.8067540848795106,
+ "learning_rate": 7.004194951495163e-07,
+ "loss": 0.3217106819152832,
+ "num_input_tokens_seen": 1077318536,
+ "step": 24640,
+ "train_runtime": 190649.8372,
+ "train_tokens_per_second": 5650.771
+ },
+ {
+ "epoch": 1.6593724750875303,
+ "grad_norm": 0.7719725402304747,
+ "learning_rate": 6.990704632574046e-07,
+ "loss": 0.29686970710754396,
+ "num_input_tokens_seen": 1077547440,
+ "step": 24645,
+ "train_runtime": 190685.7981,
+ "train_tokens_per_second": 5650.906
+ },
+ {
+ "epoch": 1.6597091300834905,
+ "grad_norm": 0.8088510895933726,
+ "learning_rate": 6.977226341065219e-07,
+ "loss": 0.299147891998291,
+ "num_input_tokens_seen": 1077764672,
+ "step": 24650,
+ "train_runtime": 190723.2912,
+ "train_tokens_per_second": 5650.934
+ },
+ {
+ "epoch": 1.6600457850794506,
+ "grad_norm": 0.7515417585035541,
+ "learning_rate": 6.96376008073788e-07,
+ "loss": 0.3067599296569824,
+ "num_input_tokens_seen": 1077969096,
+ "step": 24655,
+ "train_runtime": 190758.1707,
+ "train_tokens_per_second": 5650.972
+ },
+ {
+ "epoch": 1.6603824400754106,
+ "grad_norm": 0.7301895968368473,
+ "learning_rate": 6.950305855357781e-07,
+ "loss": 0.27429049015045165,
+ "num_input_tokens_seen": 1078181392,
+ "step": 24660,
+ "train_runtime": 190792.7332,
+ "train_tokens_per_second": 5651.061
+ },
+ {
+ "epoch": 1.6607190950713708,
+ "grad_norm": 0.7577040604392636,
+ "learning_rate": 6.936863668687383e-07,
+ "loss": 0.30942230224609374,
+ "num_input_tokens_seen": 1078397824,
+ "step": 24665,
+ "train_runtime": 190830.5609,
+ "train_tokens_per_second": 5651.075
+ },
+ {
+ "epoch": 1.661055750067331,
+ "grad_norm": 0.7766234601269146,
+ "learning_rate": 6.923433524485745e-07,
+ "loss": 0.323834753036499,
+ "num_input_tokens_seen": 1078622016,
+ "step": 24670,
+ "train_runtime": 190872.8273,
+ "train_tokens_per_second": 5650.998
+ },
+ {
+ "epoch": 1.6613924050632911,
+ "grad_norm": 0.7623394697980529,
+ "learning_rate": 6.910015426508554e-07,
+ "loss": 0.296146297454834,
+ "num_input_tokens_seen": 1078847744,
+ "step": 24675,
+ "train_runtime": 190912.4664,
+ "train_tokens_per_second": 5651.007
+ },
+ {
+ "epoch": 1.6617290600592511,
+ "grad_norm": 0.7664054805943663,
+ "learning_rate": 6.896609378508152e-07,
+ "loss": 0.309325647354126,
+ "num_input_tokens_seen": 1079050512,
+ "step": 24680,
+ "train_runtime": 190950.4474,
+ "train_tokens_per_second": 5650.945
+ },
+ {
+ "epoch": 1.6620657150552114,
+ "grad_norm": 0.7252601816921024,
+ "learning_rate": 6.883215384233488e-07,
+ "loss": 0.2978062629699707,
+ "num_input_tokens_seen": 1079282736,
+ "step": 24685,
+ "train_runtime": 190991.7978,
+ "train_tokens_per_second": 5650.938
+ },
+ {
+ "epoch": 1.6624023700511716,
+ "grad_norm": 0.8104244979601277,
+ "learning_rate": 6.869833447430152e-07,
+ "loss": 0.28819808959960935,
+ "num_input_tokens_seen": 1079488008,
+ "step": 24690,
+ "train_runtime": 191037.7882,
+ "train_tokens_per_second": 5650.652
+ },
+ {
+ "epoch": 1.6627390250471317,
+ "grad_norm": 0.7648578275299457,
+ "learning_rate": 6.856463571840361e-07,
+ "loss": 0.2954545021057129,
+ "num_input_tokens_seen": 1079693528,
+ "step": 24695,
+ "train_runtime": 191074.9276,
+ "train_tokens_per_second": 5650.629
+ },
+ {
+ "epoch": 1.6630756800430917,
+ "grad_norm": 0.7776160702348354,
+ "learning_rate": 6.843105761202945e-07,
+ "loss": 0.3160905599594116,
+ "num_input_tokens_seen": 1079922112,
+ "step": 24700,
+ "train_runtime": 191116.2384,
+ "train_tokens_per_second": 5650.604
+ },
+ {
+ "epoch": 1.663412335039052,
+ "grad_norm": 0.795996876323385,
+ "learning_rate": 6.829760019253412e-07,
+ "loss": 0.3114809513092041,
+ "num_input_tokens_seen": 1080138072,
+ "step": 24705,
+ "train_runtime": 191153.6294,
+ "train_tokens_per_second": 5650.628
+ },
+ {
+ "epoch": 1.6637489900350122,
+ "grad_norm": 0.7322911069723348,
+ "learning_rate": 6.816426349723804e-07,
+ "loss": 0.30275850296020507,
+ "num_input_tokens_seen": 1080349800,
+ "step": 24710,
+ "train_runtime": 191191.3873,
+ "train_tokens_per_second": 5650.62
+ },
+ {
+ "epoch": 1.6640856450309722,
+ "grad_norm": 0.6721252860586514,
+ "learning_rate": 6.803104756342877e-07,
+ "loss": 0.2700210094451904,
+ "num_input_tokens_seen": 1080576104,
+ "step": 24715,
+ "train_runtime": 191227.8925,
+ "train_tokens_per_second": 5650.724
+ },
+ {
+ "epoch": 1.6644223000269323,
+ "grad_norm": 0.7184313716768772,
+ "learning_rate": 6.789795242835962e-07,
+ "loss": 0.26948862075805663,
+ "num_input_tokens_seen": 1080793368,
+ "step": 24720,
+ "train_runtime": 191269.297,
+ "train_tokens_per_second": 5650.637
+ },
+ {
+ "epoch": 1.6647589550228925,
+ "grad_norm": 0.7596106189238437,
+ "learning_rate": 6.776497812925015e-07,
+ "loss": 0.3047426700592041,
+ "num_input_tokens_seen": 1081001160,
+ "step": 24725,
+ "train_runtime": 191307.1705,
+ "train_tokens_per_second": 5650.605
+ },
+ {
+ "epoch": 1.6650956100188528,
+ "grad_norm": 0.8078999876649536,
+ "learning_rate": 6.763212470328628e-07,
+ "loss": 0.3334406614303589,
+ "num_input_tokens_seen": 1081221944,
+ "step": 24730,
+ "train_runtime": 191348.8888,
+ "train_tokens_per_second": 5650.526
+ },
+ {
+ "epoch": 1.6654322650148128,
+ "grad_norm": 0.7785010647558146,
+ "learning_rate": 6.749939218762003e-07,
+ "loss": 0.2895099878311157,
+ "num_input_tokens_seen": 1081439048,
+ "step": 24735,
+ "train_runtime": 191390.3574,
+ "train_tokens_per_second": 5650.436
+ },
+ {
+ "epoch": 1.6657689200107728,
+ "grad_norm": 0.7473396001618127,
+ "learning_rate": 6.736678061936958e-07,
+ "loss": 0.3205715656280518,
+ "num_input_tokens_seen": 1081670648,
+ "step": 24740,
+ "train_runtime": 191432.0544,
+ "train_tokens_per_second": 5650.416
+ },
+ {
+ "epoch": 1.666105575006733,
+ "grad_norm": 0.8395952302334243,
+ "learning_rate": 6.723429003561937e-07,
+ "loss": 0.325761604309082,
+ "num_input_tokens_seen": 1081907840,
+ "step": 24745,
+ "train_runtime": 191466.0452,
+ "train_tokens_per_second": 5650.651
+ },
+ {
+ "epoch": 1.6664422300026933,
+ "grad_norm": 0.7553522775373878,
+ "learning_rate": 6.710192047341984e-07,
+ "loss": 0.2971907615661621,
+ "num_input_tokens_seen": 1082129744,
+ "step": 24750,
+ "train_runtime": 191505.5101,
+ "train_tokens_per_second": 5650.645
+ },
+ {
+ "epoch": 1.6667788849986533,
+ "grad_norm": 0.7670068129458077,
+ "learning_rate": 6.69696719697881e-07,
+ "loss": 0.29376702308654784,
+ "num_input_tokens_seen": 1082352688,
+ "step": 24755,
+ "train_runtime": 191544.0035,
+ "train_tokens_per_second": 5650.674
+ },
+ {
+ "epoch": 1.6671155399946134,
+ "grad_norm": 0.7937246542094001,
+ "learning_rate": 6.683754456170655e-07,
+ "loss": 0.27995355129241944,
+ "num_input_tokens_seen": 1082568000,
+ "step": 24760,
+ "train_runtime": 191576.7889,
+ "train_tokens_per_second": 5650.831
+ },
+ {
+ "epoch": 1.6674521949905736,
+ "grad_norm": 0.7863716640239266,
+ "learning_rate": 6.670553828612448e-07,
+ "loss": 0.28558523654937745,
+ "num_input_tokens_seen": 1082762664,
+ "step": 24765,
+ "train_runtime": 191615.9134,
+ "train_tokens_per_second": 5650.693
+ },
+ {
+ "epoch": 1.6677888499865339,
+ "grad_norm": 0.6755324668287188,
+ "learning_rate": 6.657365317995696e-07,
+ "loss": 0.2899418592453003,
+ "num_input_tokens_seen": 1082980808,
+ "step": 24770,
+ "train_runtime": 191656.4559,
+ "train_tokens_per_second": 5650.636
+ },
+ {
+ "epoch": 1.668125504982494,
+ "grad_norm": 0.7089937331683779,
+ "learning_rate": 6.644188928008521e-07,
+ "loss": 0.31790828704833984,
+ "num_input_tokens_seen": 1083191992,
+ "step": 24775,
+ "train_runtime": 191696.5978,
+ "train_tokens_per_second": 5650.554
+ },
+ {
+ "epoch": 1.668462159978454,
+ "grad_norm": 0.8389878605630636,
+ "learning_rate": 6.631024662335667e-07,
+ "loss": 0.31216111183166506,
+ "num_input_tokens_seen": 1083409256,
+ "step": 24780,
+ "train_runtime": 191745.9321,
+ "train_tokens_per_second": 5650.233
+ },
+ {
+ "epoch": 1.6687988149744142,
+ "grad_norm": 0.7518345608186,
+ "learning_rate": 6.617872524658469e-07,
+ "loss": 0.29103267192840576,
+ "num_input_tokens_seen": 1083638560,
+ "step": 24785,
+ "train_runtime": 191781.1772,
+ "train_tokens_per_second": 5650.391
+ },
+ {
+ "epoch": 1.6691354699703744,
+ "grad_norm": 0.8193494154451724,
+ "learning_rate": 6.604732518654882e-07,
+ "loss": 0.3304905414581299,
+ "num_input_tokens_seen": 1083853344,
+ "step": 24790,
+ "train_runtime": 191823.5191,
+ "train_tokens_per_second": 5650.263
+ },
+ {
+ "epoch": 1.6694721249663345,
+ "grad_norm": 0.8968284580615762,
+ "learning_rate": 6.591604647999473e-07,
+ "loss": 0.32892765998840334,
+ "num_input_tokens_seen": 1084083632,
+ "step": 24795,
+ "train_runtime": 191860.6602,
+ "train_tokens_per_second": 5650.37
+ },
+ {
+ "epoch": 1.6698087799622945,
+ "grad_norm": 0.6910607916103046,
+ "learning_rate": 6.57848891636339e-07,
+ "loss": 0.2843611478805542,
+ "num_input_tokens_seen": 1084309072,
+ "step": 24800,
+ "train_runtime": 191903.5578,
+ "train_tokens_per_second": 5650.281
+ },
+ {
+ "epoch": 1.6701454349582547,
+ "grad_norm": 0.7582706445392589,
+ "learning_rate": 6.565385327414441e-07,
+ "loss": 0.30639631748199464,
+ "num_input_tokens_seen": 1084535760,
+ "step": 24805,
+ "train_runtime": 191937.737,
+ "train_tokens_per_second": 5650.456
+ },
+ {
+ "epoch": 1.670482089954215,
+ "grad_norm": 0.8230500682984503,
+ "learning_rate": 6.552293884816968e-07,
+ "loss": 0.2900113582611084,
+ "num_input_tokens_seen": 1084761472,
+ "step": 24810,
+ "train_runtime": 191977.452,
+ "train_tokens_per_second": 5650.463
+ },
+ {
+ "epoch": 1.670818744950175,
+ "grad_norm": 0.6579326078890422,
+ "learning_rate": 6.539214592231968e-07,
+ "loss": 0.29627652168273927,
+ "num_input_tokens_seen": 1084987552,
+ "step": 24815,
+ "train_runtime": 192018.4764,
+ "train_tokens_per_second": 5650.433
+ },
+ {
+ "epoch": 1.671155399946135,
+ "grad_norm": 0.7194038569880106,
+ "learning_rate": 6.526147453317028e-07,
+ "loss": 0.26366333961486815,
+ "num_input_tokens_seen": 1085201944,
+ "step": 24820,
+ "train_runtime": 192058.3704,
+ "train_tokens_per_second": 5650.376
+ },
+ {
+ "epoch": 1.6714920549420953,
+ "grad_norm": 0.7597781908310334,
+ "learning_rate": 6.513092471726307e-07,
+ "loss": 0.299773645401001,
+ "num_input_tokens_seen": 1085434936,
+ "step": 24825,
+ "train_runtime": 192099.0636,
+ "train_tokens_per_second": 5650.392
+ },
+ {
+ "epoch": 1.6718287099380555,
+ "grad_norm": 0.7278296909083918,
+ "learning_rate": 6.500049651110635e-07,
+ "loss": 0.2928581714630127,
+ "num_input_tokens_seen": 1085656760,
+ "step": 24830,
+ "train_runtime": 192139.5855,
+ "train_tokens_per_second": 5650.354
+ },
+ {
+ "epoch": 1.6721653649340156,
+ "grad_norm": 0.7587076551124298,
+ "learning_rate": 6.487018995117339e-07,
+ "loss": 0.2924539089202881,
+ "num_input_tokens_seen": 1085894200,
+ "step": 24835,
+ "train_runtime": 192174.6364,
+ "train_tokens_per_second": 5650.559
+ },
+ {
+ "epoch": 1.6725020199299756,
+ "grad_norm": 0.7823318054145896,
+ "learning_rate": 6.474000507390455e-07,
+ "loss": 0.2911822319030762,
+ "num_input_tokens_seen": 1086112280,
+ "step": 24840,
+ "train_runtime": 192208.7768,
+ "train_tokens_per_second": 5650.69
+ },
+ {
+ "epoch": 1.6728386749259359,
+ "grad_norm": 0.7008110558611228,
+ "learning_rate": 6.460994191570513e-07,
+ "loss": 0.3034616470336914,
+ "num_input_tokens_seen": 1086334184,
+ "step": 24845,
+ "train_runtime": 192246.1123,
+ "train_tokens_per_second": 5650.747
+ },
+ {
+ "epoch": 1.673175329921896,
+ "grad_norm": 0.7292792811187331,
+ "learning_rate": 6.448000051294717e-07,
+ "loss": 0.2885445594787598,
+ "num_input_tokens_seen": 1086553800,
+ "step": 24850,
+ "train_runtime": 192284.9452,
+ "train_tokens_per_second": 5650.748
+ },
+ {
+ "epoch": 1.6735119849178561,
+ "grad_norm": 0.6923122611597387,
+ "learning_rate": 6.435018090196832e-07,
+ "loss": 0.30068202018737794,
+ "num_input_tokens_seen": 1086783696,
+ "step": 24855,
+ "train_runtime": 192321.9072,
+ "train_tokens_per_second": 5650.858
+ },
+ {
+ "epoch": 1.6738486399138162,
+ "grad_norm": 1.2221971616080631,
+ "learning_rate": 6.42204831190722e-07,
+ "loss": 0.2921877861022949,
+ "num_input_tokens_seen": 1087006792,
+ "step": 24860,
+ "train_runtime": 192356.0704,
+ "train_tokens_per_second": 5651.014
+ },
+ {
+ "epoch": 1.6741852949097764,
+ "grad_norm": 0.7759288823452887,
+ "learning_rate": 6.409090720052835e-07,
+ "loss": 0.30269834995269773,
+ "num_input_tokens_seen": 1087221944,
+ "step": 24865,
+ "train_runtime": 192396.3099,
+ "train_tokens_per_second": 5650.95
+ },
+ {
+ "epoch": 1.6745219499057367,
+ "grad_norm": 0.7489484593312903,
+ "learning_rate": 6.396145318257229e-07,
+ "loss": 0.27388882637023926,
+ "num_input_tokens_seen": 1087443824,
+ "step": 24870,
+ "train_runtime": 192433.614,
+ "train_tokens_per_second": 5651.008
+ },
+ {
+ "epoch": 1.6748586049016967,
+ "grad_norm": 0.7471747560717638,
+ "learning_rate": 6.383212110140535e-07,
+ "loss": 0.31777362823486327,
+ "num_input_tokens_seen": 1087659568,
+ "step": 24875,
+ "train_runtime": 192469.464,
+ "train_tokens_per_second": 5651.076
+ },
+ {
+ "epoch": 1.6751952598976567,
+ "grad_norm": 0.9939208015357779,
+ "learning_rate": 6.370291099319509e-07,
+ "loss": 0.3115004301071167,
+ "num_input_tokens_seen": 1087852096,
+ "step": 24880,
+ "train_runtime": 192505.3832,
+ "train_tokens_per_second": 5651.022
+ },
+ {
+ "epoch": 1.675531914893617,
+ "grad_norm": 0.7589414568505637,
+ "learning_rate": 6.357382289407433e-07,
+ "loss": 0.29903695583343504,
+ "num_input_tokens_seen": 1088082304,
+ "step": 24885,
+ "train_runtime": 192540.2873,
+ "train_tokens_per_second": 5651.193
+ },
+ {
+ "epoch": 1.6758685698895772,
+ "grad_norm": 0.7522829988493125,
+ "learning_rate": 6.344485684014251e-07,
+ "loss": 0.3122910499572754,
+ "num_input_tokens_seen": 1088296296,
+ "step": 24890,
+ "train_runtime": 192578.9655,
+ "train_tokens_per_second": 5651.169
+ },
+ {
+ "epoch": 1.6762052248855372,
+ "grad_norm": 0.7718700735790412,
+ "learning_rate": 6.33160128674642e-07,
+ "loss": 0.31540684700012206,
+ "num_input_tokens_seen": 1088501864,
+ "step": 24895,
+ "train_runtime": 192615.6028,
+ "train_tokens_per_second": 5651.161
+ },
+ {
+ "epoch": 1.6765418798814973,
+ "grad_norm": 0.7172203437810806,
+ "learning_rate": 6.318729101207055e-07,
+ "loss": 0.2925147533416748,
+ "num_input_tokens_seen": 1088727288,
+ "step": 24900,
+ "train_runtime": 192657.5288,
+ "train_tokens_per_second": 5651.102
+ },
+ {
+ "epoch": 1.6768785348774575,
+ "grad_norm": 0.6793461071435637,
+ "learning_rate": 6.305869130995807e-07,
+ "loss": 0.30594768524169924,
+ "num_input_tokens_seen": 1088947848,
+ "step": 24905,
+ "train_runtime": 192692.2858,
+ "train_tokens_per_second": 5651.227
+ },
+ {
+ "epoch": 1.6772151898734178,
+ "grad_norm": 0.8065807442493926,
+ "learning_rate": 6.293021379708924e-07,
+ "loss": 0.30313916206359864,
+ "num_input_tokens_seen": 1089171240,
+ "step": 24910,
+ "train_runtime": 192732.2584,
+ "train_tokens_per_second": 5651.214
+ },
+ {
+ "epoch": 1.6775518448693778,
+ "grad_norm": 0.812051527961556,
+ "learning_rate": 6.280185850939241e-07,
+ "loss": 0.31810295581817627,
+ "num_input_tokens_seen": 1089387216,
+ "step": 24915,
+ "train_runtime": 192774.0539,
+ "train_tokens_per_second": 5651.109
+ },
+ {
+ "epoch": 1.677888499865338,
+ "grad_norm": 0.8171798449848718,
+ "learning_rate": 6.267362548276173e-07,
+ "loss": 0.30693349838256834,
+ "num_input_tokens_seen": 1089597552,
+ "step": 24920,
+ "train_runtime": 192811.4332,
+ "train_tokens_per_second": 5651.104
+ },
+ {
+ "epoch": 1.6782251548612983,
+ "grad_norm": 0.9137416038709986,
+ "learning_rate": 6.254551475305703e-07,
+ "loss": 0.30268676280975343,
+ "num_input_tokens_seen": 1089790616,
+ "step": 24925,
+ "train_runtime": 192849.8209,
+ "train_tokens_per_second": 5650.981
+ },
+ {
+ "epoch": 1.6785618098572583,
+ "grad_norm": 0.80355688108705,
+ "learning_rate": 6.241752635610443e-07,
+ "loss": 0.31978979110717776,
+ "num_input_tokens_seen": 1090007856,
+ "step": 24930,
+ "train_runtime": 192886.1002,
+ "train_tokens_per_second": 5651.044
+ },
+ {
+ "epoch": 1.6788984648532184,
+ "grad_norm": 0.9368900433932702,
+ "learning_rate": 6.228966032769496e-07,
+ "loss": 0.31137659549713137,
+ "num_input_tokens_seen": 1090222152,
+ "step": 24935,
+ "train_runtime": 192922.3211,
+ "train_tokens_per_second": 5651.094
+ },
+ {
+ "epoch": 1.6792351198491786,
+ "grad_norm": 0.8015786582216125,
+ "learning_rate": 6.216191670358651e-07,
+ "loss": 0.29123806953430176,
+ "num_input_tokens_seen": 1090437952,
+ "step": 24940,
+ "train_runtime": 192965.473,
+ "train_tokens_per_second": 5650.949
+ },
+ {
+ "epoch": 1.6795717748451389,
+ "grad_norm": 0.7556404800034242,
+ "learning_rate": 6.203429551950163e-07,
+ "loss": 0.32757062911987306,
+ "num_input_tokens_seen": 1090662752,
+ "step": 24945,
+ "train_runtime": 193011.4944,
+ "train_tokens_per_second": 5650.766
+ },
+ {
+ "epoch": 1.679908429841099,
+ "grad_norm": 0.8290622091556459,
+ "learning_rate": 6.19067968111296e-07,
+ "loss": 0.28361237049102783,
+ "num_input_tokens_seen": 1090869792,
+ "step": 24950,
+ "train_runtime": 193048.7244,
+ "train_tokens_per_second": 5650.749
+ },
+ {
+ "epoch": 1.680245084837059,
+ "grad_norm": 0.7116384964210178,
+ "learning_rate": 6.177942061412478e-07,
+ "loss": 0.2768688678741455,
+ "num_input_tokens_seen": 1091083944,
+ "step": 24955,
+ "train_runtime": 193088.2417,
+ "train_tokens_per_second": 5650.701
+ },
+ {
+ "epoch": 1.6805817398330192,
+ "grad_norm": 0.7009278311455667,
+ "learning_rate": 6.16521669641077e-07,
+ "loss": 0.2864533424377441,
+ "num_input_tokens_seen": 1091313248,
+ "step": 24960,
+ "train_runtime": 193120.829,
+ "train_tokens_per_second": 5650.935
+ },
+ {
+ "epoch": 1.6809183948289794,
+ "grad_norm": 0.7261926667581532,
+ "learning_rate": 6.152503589666426e-07,
+ "loss": 0.30624988079071047,
+ "num_input_tokens_seen": 1091544272,
+ "step": 24965,
+ "train_runtime": 193157.6847,
+ "train_tokens_per_second": 5651.053
+ },
+ {
+ "epoch": 1.6812550498249395,
+ "grad_norm": 0.7142592236914898,
+ "learning_rate": 6.139802744734636e-07,
+ "loss": 0.26999435424804685,
+ "num_input_tokens_seen": 1091761648,
+ "step": 24970,
+ "train_runtime": 193193.4372,
+ "train_tokens_per_second": 5651.132
+ },
+ {
+ "epoch": 1.6815917048208995,
+ "grad_norm": 0.8017434808157823,
+ "learning_rate": 6.127114165167142e-07,
+ "loss": 0.26949467658996584,
+ "num_input_tokens_seen": 1091969864,
+ "step": 24975,
+ "train_runtime": 193228.9139,
+ "train_tokens_per_second": 5651.172
+ },
+ {
+ "epoch": 1.6819283598168597,
+ "grad_norm": 0.8597502504635929,
+ "learning_rate": 6.11443785451229e-07,
+ "loss": 0.30141243934631345,
+ "num_input_tokens_seen": 1092190952,
+ "step": 24980,
+ "train_runtime": 193269.634,
+ "train_tokens_per_second": 5651.125
+ },
+ {
+ "epoch": 1.68226501481282,
+ "grad_norm": 0.7558509297387581,
+ "learning_rate": 6.101773816314926e-07,
+ "loss": 0.29298968315124513,
+ "num_input_tokens_seen": 1092417104,
+ "step": 24985,
+ "train_runtime": 193306.8721,
+ "train_tokens_per_second": 5651.207
+ },
+ {
+ "epoch": 1.68260166980878,
+ "grad_norm": 0.7852414890479306,
+ "learning_rate": 6.089122054116558e-07,
+ "loss": 0.3086401462554932,
+ "num_input_tokens_seen": 1092645416,
+ "step": 24990,
+ "train_runtime": 193342.0721,
+ "train_tokens_per_second": 5651.359
+ },
+ {
+ "epoch": 1.68293832480474,
+ "grad_norm": 0.8486422869311524,
+ "learning_rate": 6.076482571455161e-07,
+ "loss": 0.2916858196258545,
+ "num_input_tokens_seen": 1092868304,
+ "step": 24995,
+ "train_runtime": 193383.0382,
+ "train_tokens_per_second": 5651.314
+ },
+ {
+ "epoch": 1.6832749798007003,
+ "grad_norm": 0.8736609060726518,
+ "learning_rate": 6.063855371865369e-07,
+ "loss": 0.30973148345947266,
+ "num_input_tokens_seen": 1093073112,
+ "step": 25000,
+ "train_runtime": 193420.2455,
+ "train_tokens_per_second": 5651.286
+ },
+ {
+ "epoch": 1.6836116347966605,
+ "grad_norm": 0.8270587506153307,
+ "learning_rate": 6.051240458878316e-07,
+ "loss": 0.3318582057952881,
+ "num_input_tokens_seen": 1093292280,
+ "step": 25005,
+ "train_runtime": 193460.4002,
+ "train_tokens_per_second": 5651.246
+ },
+ {
+ "epoch": 1.6839482897926206,
+ "grad_norm": 0.7861426638271823,
+ "learning_rate": 6.038637836021721e-07,
+ "loss": 0.30891571044921873,
+ "num_input_tokens_seen": 1093512488,
+ "step": 25010,
+ "train_runtime": 193500.033,
+ "train_tokens_per_second": 5651.226
+ },
+ {
+ "epoch": 1.6842849447885806,
+ "grad_norm": 0.7570592424235427,
+ "learning_rate": 6.026047506819882e-07,
+ "loss": 0.29696998596191404,
+ "num_input_tokens_seen": 1093724280,
+ "step": 25015,
+ "train_runtime": 193535.9844,
+ "train_tokens_per_second": 5651.271
+ },
+ {
+ "epoch": 1.6846215997845408,
+ "grad_norm": 0.8023108005761309,
+ "learning_rate": 6.013469474793632e-07,
+ "loss": 0.2949728012084961,
+ "num_input_tokens_seen": 1093945480,
+ "step": 25020,
+ "train_runtime": 193575.5488,
+ "train_tokens_per_second": 5651.259
+ },
+ {
+ "epoch": 1.684958254780501,
+ "grad_norm": 0.792969947232272,
+ "learning_rate": 6.000903743460368e-07,
+ "loss": 0.2946459770202637,
+ "num_input_tokens_seen": 1094164496,
+ "step": 25025,
+ "train_runtime": 193619.1001,
+ "train_tokens_per_second": 5651.119
+ },
+ {
+ "epoch": 1.6852949097764611,
+ "grad_norm": 0.7249873667459386,
+ "learning_rate": 5.98835031633409e-07,
+ "loss": 0.296980881690979,
+ "num_input_tokens_seen": 1094387696,
+ "step": 25030,
+ "train_runtime": 193658.0253,
+ "train_tokens_per_second": 5651.135
+ },
+ {
+ "epoch": 1.6856315647724212,
+ "grad_norm": 0.7187877526122118,
+ "learning_rate": 5.97580919692528e-07,
+ "loss": 0.30903480052947996,
+ "num_input_tokens_seen": 1094621216,
+ "step": 25035,
+ "train_runtime": 193690.3611,
+ "train_tokens_per_second": 5651.397
+ },
+ {
+ "epoch": 1.6859682197683814,
+ "grad_norm": 0.7738587622533499,
+ "learning_rate": 5.963280388741072e-07,
+ "loss": 0.2778623104095459,
+ "num_input_tokens_seen": 1094843312,
+ "step": 25040,
+ "train_runtime": 193726.0711,
+ "train_tokens_per_second": 5651.502
+ },
+ {
+ "epoch": 1.6863048747643417,
+ "grad_norm": 0.8921293625641102,
+ "learning_rate": 5.950763895285055e-07,
+ "loss": 0.3040936946868896,
+ "num_input_tokens_seen": 1095048760,
+ "step": 25045,
+ "train_runtime": 193766.9375,
+ "train_tokens_per_second": 5651.371
+ },
+ {
+ "epoch": 1.6866415297603017,
+ "grad_norm": 0.6626481867816244,
+ "learning_rate": 5.938259720057465e-07,
+ "loss": 0.29845240116119387,
+ "num_input_tokens_seen": 1095275144,
+ "step": 25050,
+ "train_runtime": 193803.6462,
+ "train_tokens_per_second": 5651.468
+ },
+ {
+ "epoch": 1.6869781847562617,
+ "grad_norm": 0.7608946538453196,
+ "learning_rate": 5.925767866555043e-07,
+ "loss": 0.28672802448272705,
+ "num_input_tokens_seen": 1095488208,
+ "step": 25055,
+ "train_runtime": 193846.2174,
+ "train_tokens_per_second": 5651.326
+ },
+ {
+ "epoch": 1.687314839752222,
+ "grad_norm": 0.8656369475130261,
+ "learning_rate": 5.913288338271095e-07,
+ "loss": 0.31831324100494385,
+ "num_input_tokens_seen": 1095711504,
+ "step": 25060,
+ "train_runtime": 193884.7583,
+ "train_tokens_per_second": 5651.355
+ },
+ {
+ "epoch": 1.6876514947481822,
+ "grad_norm": 0.7023811573666321,
+ "learning_rate": 5.900821138695478e-07,
+ "loss": 0.29593048095703123,
+ "num_input_tokens_seen": 1095931728,
+ "step": 25065,
+ "train_runtime": 193924.6678,
+ "train_tokens_per_second": 5651.327
+ },
+ {
+ "epoch": 1.6879881497441422,
+ "grad_norm": 0.760116486923957,
+ "learning_rate": 5.888366271314605e-07,
+ "loss": 0.2997978448867798,
+ "num_input_tokens_seen": 1096158920,
+ "step": 25070,
+ "train_runtime": 193970.5562,
+ "train_tokens_per_second": 5651.161
+ },
+ {
+ "epoch": 1.6883248047401023,
+ "grad_norm": 0.6921190819902673,
+ "learning_rate": 5.875923739611439e-07,
+ "loss": 0.3102496385574341,
+ "num_input_tokens_seen": 1096386400,
+ "step": 25075,
+ "train_runtime": 194005.6003,
+ "train_tokens_per_second": 5651.313
+ },
+ {
+ "epoch": 1.6886614597360625,
+ "grad_norm": 0.708393747684828,
+ "learning_rate": 5.863493547065513e-07,
+ "loss": 0.288178825378418,
+ "num_input_tokens_seen": 1096598208,
+ "step": 25080,
+ "train_runtime": 194046.4638,
+ "train_tokens_per_second": 5651.215
+ },
+ {
+ "epoch": 1.6889981147320228,
+ "grad_norm": 0.7325173487118091,
+ "learning_rate": 5.851075697152858e-07,
+ "loss": 0.2890732765197754,
+ "num_input_tokens_seen": 1096830840,
+ "step": 25085,
+ "train_runtime": 194089.2226,
+ "train_tokens_per_second": 5651.168
+ },
+ {
+ "epoch": 1.6893347697279828,
+ "grad_norm": 1.084295866928116,
+ "learning_rate": 5.838670193346118e-07,
+ "loss": 0.3075998067855835,
+ "num_input_tokens_seen": 1097042768,
+ "step": 25090,
+ "train_runtime": 194131.2034,
+ "train_tokens_per_second": 5651.038
+ },
+ {
+ "epoch": 1.6896714247239428,
+ "grad_norm": 0.708754132607845,
+ "learning_rate": 5.826277039114426e-07,
+ "loss": 0.2823533773422241,
+ "num_input_tokens_seen": 1097258936,
+ "step": 25095,
+ "train_runtime": 194173.4531,
+ "train_tokens_per_second": 5650.921
+ },
+ {
+ "epoch": 1.690008079719903,
+ "grad_norm": 0.6971867426568854,
+ "learning_rate": 5.813896237923505e-07,
+ "loss": 0.33135075569152833,
+ "num_input_tokens_seen": 1097479600,
+ "step": 25100,
+ "train_runtime": 194218.3152,
+ "train_tokens_per_second": 5650.752
+ },
+ {
+ "epoch": 1.6903447347158633,
+ "grad_norm": 0.7270149386031489,
+ "learning_rate": 5.801527793235606e-07,
+ "loss": 0.2957238912582397,
+ "num_input_tokens_seen": 1097708984,
+ "step": 25105,
+ "train_runtime": 194258.1829,
+ "train_tokens_per_second": 5650.773
+ },
+ {
+ "epoch": 1.6906813897118234,
+ "grad_norm": 0.882744897146511,
+ "learning_rate": 5.789171708509516e-07,
+ "loss": 0.28771932125091554,
+ "num_input_tokens_seen": 1097906256,
+ "step": 25110,
+ "train_runtime": 194299.2332,
+ "train_tokens_per_second": 5650.595
+ },
+ {
+ "epoch": 1.6910180447077834,
+ "grad_norm": 0.7410230789420439,
+ "learning_rate": 5.776827987200584e-07,
+ "loss": 0.30194315910339353,
+ "num_input_tokens_seen": 1098128136,
+ "step": 25115,
+ "train_runtime": 194336.6611,
+ "train_tokens_per_second": 5650.648
+ },
+ {
+ "epoch": 1.6913546997037436,
+ "grad_norm": 0.7896654861282312,
+ "learning_rate": 5.764496632760691e-07,
+ "loss": 0.2923448085784912,
+ "num_input_tokens_seen": 1098334664,
+ "step": 25120,
+ "train_runtime": 194373.4203,
+ "train_tokens_per_second": 5650.642
+ },
+ {
+ "epoch": 1.6916913546997039,
+ "grad_norm": 0.7314826062008781,
+ "learning_rate": 5.752177648638241e-07,
+ "loss": 0.27889845371246336,
+ "num_input_tokens_seen": 1098559728,
+ "step": 25125,
+ "train_runtime": 194408.4815,
+ "train_tokens_per_second": 5650.781
+ },
+ {
+ "epoch": 1.692028009695664,
+ "grad_norm": 0.8906343966024878,
+ "learning_rate": 5.739871038278239e-07,
+ "loss": 0.3101338863372803,
+ "num_input_tokens_seen": 1098782712,
+ "step": 25130,
+ "train_runtime": 194440.3466,
+ "train_tokens_per_second": 5651.002
+ },
+ {
+ "epoch": 1.692364664691624,
+ "grad_norm": 0.7281493386147246,
+ "learning_rate": 5.727576805122142e-07,
+ "loss": 0.28267052173614504,
+ "num_input_tokens_seen": 1098991872,
+ "step": 25135,
+ "train_runtime": 194480.4862,
+ "train_tokens_per_second": 5650.911
+ },
+ {
+ "epoch": 1.6927013196875842,
+ "grad_norm": 0.6893488772809538,
+ "learning_rate": 5.715294952608041e-07,
+ "loss": 0.30869436264038086,
+ "num_input_tokens_seen": 1099224528,
+ "step": 25140,
+ "train_runtime": 194512.1799,
+ "train_tokens_per_second": 5651.186
+ },
+ {
+ "epoch": 1.6930379746835444,
+ "grad_norm": 0.8413021899719699,
+ "learning_rate": 5.703025484170471e-07,
+ "loss": 0.3042963981628418,
+ "num_input_tokens_seen": 1099455640,
+ "step": 25145,
+ "train_runtime": 194548.5571,
+ "train_tokens_per_second": 5651.317
+ },
+ {
+ "epoch": 1.6933746296795045,
+ "grad_norm": 0.760681189560044,
+ "learning_rate": 5.690768403240587e-07,
+ "loss": 0.3153714656829834,
+ "num_input_tokens_seen": 1099676720,
+ "step": 25150,
+ "train_runtime": 194585.4115,
+ "train_tokens_per_second": 5651.383
+ },
+ {
+ "epoch": 1.6937112846754645,
+ "grad_norm": 0.7173964449350725,
+ "learning_rate": 5.67852371324602e-07,
+ "loss": 0.30834920406341554,
+ "num_input_tokens_seen": 1099907704,
+ "step": 25155,
+ "train_runtime": 194620.3712,
+ "train_tokens_per_second": 5651.555
+ },
+ {
+ "epoch": 1.6940479396714248,
+ "grad_norm": 0.7329549717993644,
+ "learning_rate": 5.66629141761097e-07,
+ "loss": 0.2930551528930664,
+ "num_input_tokens_seen": 1100142256,
+ "step": 25160,
+ "train_runtime": 194657.3311,
+ "train_tokens_per_second": 5651.687
+ },
+ {
+ "epoch": 1.694384594667385,
+ "grad_norm": 0.7342138353588805,
+ "learning_rate": 5.654071519756154e-07,
+ "loss": 0.30519957542419435,
+ "num_input_tokens_seen": 1100366192,
+ "step": 25165,
+ "train_runtime": 194704.0482,
+ "train_tokens_per_second": 5651.481
+ },
+ {
+ "epoch": 1.694721249663345,
+ "grad_norm": 0.7791350168211661,
+ "learning_rate": 5.641864023098831e-07,
+ "loss": 0.33337717056274413,
+ "num_input_tokens_seen": 1100584800,
+ "step": 25170,
+ "train_runtime": 194745.9485,
+ "train_tokens_per_second": 5651.387
+ },
+ {
+ "epoch": 1.695057904659305,
+ "grad_norm": 0.7179815820915966,
+ "learning_rate": 5.629668931052773e-07,
+ "loss": 0.29319195747375487,
+ "num_input_tokens_seen": 1100817816,
+ "step": 25175,
+ "train_runtime": 194785.1895,
+ "train_tokens_per_second": 5651.445
+ },
+ {
+ "epoch": 1.6953945596552653,
+ "grad_norm": 0.73809182276721,
+ "learning_rate": 5.61748624702833e-07,
+ "loss": 0.31841015815734863,
+ "num_input_tokens_seen": 1101043616,
+ "step": 25180,
+ "train_runtime": 194820.4275,
+ "train_tokens_per_second": 5651.582
+ },
+ {
+ "epoch": 1.6957312146512256,
+ "grad_norm": 0.7597610918234663,
+ "learning_rate": 5.605315974432313e-07,
+ "loss": 0.29940404891967776,
+ "num_input_tokens_seen": 1101262632,
+ "step": 25185,
+ "train_runtime": 194863.2161,
+ "train_tokens_per_second": 5651.465
+ },
+ {
+ "epoch": 1.6960678696471856,
+ "grad_norm": 0.7151694523792368,
+ "learning_rate": 5.593158116668146e-07,
+ "loss": 0.2952594757080078,
+ "num_input_tokens_seen": 1101502480,
+ "step": 25190,
+ "train_runtime": 194904.3942,
+ "train_tokens_per_second": 5651.502
+ },
+ {
+ "epoch": 1.6964045246431456,
+ "grad_norm": 0.8157887300040912,
+ "learning_rate": 5.581012677135683e-07,
+ "loss": 0.2999080181121826,
+ "num_input_tokens_seen": 1101706720,
+ "step": 25195,
+ "train_runtime": 194939.3381,
+ "train_tokens_per_second": 5651.536
+ },
+ {
+ "epoch": 1.6967411796391059,
+ "grad_norm": 0.7188880993897309,
+ "learning_rate": 5.56887965923139e-07,
+ "loss": 0.30147387981414797,
+ "num_input_tokens_seen": 1101936320,
+ "step": 25200,
+ "train_runtime": 194980.9761,
+ "train_tokens_per_second": 5651.507
+ },
+ {
+ "epoch": 1.6970778346350661,
+ "grad_norm": 0.8551105269669304,
+ "learning_rate": 5.556759066348227e-07,
+ "loss": 0.29673826694488525,
+ "num_input_tokens_seen": 1102155368,
+ "step": 25205,
+ "train_runtime": 195018.2983,
+ "train_tokens_per_second": 5651.548
+ },
+ {
+ "epoch": 1.6974144896310261,
+ "grad_norm": 0.7898708049199309,
+ "learning_rate": 5.54465090187567e-07,
+ "loss": 0.3003216743469238,
+ "num_input_tokens_seen": 1102367352,
+ "step": 25210,
+ "train_runtime": 195053.3847,
+ "train_tokens_per_second": 5651.619
+ },
+ {
+ "epoch": 1.6977511446269862,
+ "grad_norm": 0.7892657487617118,
+ "learning_rate": 5.532555169199727e-07,
+ "loss": 0.29103655815124513,
+ "num_input_tokens_seen": 1102583040,
+ "step": 25215,
+ "train_runtime": 195087.2838,
+ "train_tokens_per_second": 5651.742
+ },
+ {
+ "epoch": 1.6980877996229464,
+ "grad_norm": 0.729451521498232,
+ "learning_rate": 5.520471871702937e-07,
+ "loss": 0.28641371726989745,
+ "num_input_tokens_seen": 1102810576,
+ "step": 25220,
+ "train_runtime": 195127.6822,
+ "train_tokens_per_second": 5651.738
+ },
+ {
+ "epoch": 1.6984244546189067,
+ "grad_norm": 0.8574198765928966,
+ "learning_rate": 5.508401012764341e-07,
+ "loss": 0.2999880313873291,
+ "num_input_tokens_seen": 1103028192,
+ "step": 25225,
+ "train_runtime": 195157.8665,
+ "train_tokens_per_second": 5651.979
+ },
+ {
+ "epoch": 1.6987611096148667,
+ "grad_norm": 0.7201969737459759,
+ "learning_rate": 5.496342595759552e-07,
+ "loss": 0.28559165000915526,
+ "num_input_tokens_seen": 1103256944,
+ "step": 25230,
+ "train_runtime": 195204.2497,
+ "train_tokens_per_second": 5651.808
+ },
+ {
+ "epoch": 1.6990977646108267,
+ "grad_norm": 0.719407135820255,
+ "learning_rate": 5.484296624060626e-07,
+ "loss": 0.2883117198944092,
+ "num_input_tokens_seen": 1103482616,
+ "step": 25235,
+ "train_runtime": 195249.5833,
+ "train_tokens_per_second": 5651.652
+ },
+ {
+ "epoch": 1.699434419606787,
+ "grad_norm": 0.8370197571942842,
+ "learning_rate": 5.472263101036212e-07,
+ "loss": 0.29885265827178953,
+ "num_input_tokens_seen": 1103706952,
+ "step": 25240,
+ "train_runtime": 195285.64,
+ "train_tokens_per_second": 5651.757
+ },
+ {
+ "epoch": 1.6997710746027472,
+ "grad_norm": 0.839267857665248,
+ "learning_rate": 5.460242030051421e-07,
+ "loss": 0.3131263732910156,
+ "num_input_tokens_seen": 1103918168,
+ "step": 25245,
+ "train_runtime": 195322.8829,
+ "train_tokens_per_second": 5651.761
+ },
+ {
+ "epoch": 1.7001077295987073,
+ "grad_norm": 0.7353115375828058,
+ "learning_rate": 5.448233414467924e-07,
+ "loss": 0.3075102806091309,
+ "num_input_tokens_seen": 1104153064,
+ "step": 25250,
+ "train_runtime": 195358.9751,
+ "train_tokens_per_second": 5651.919
+ },
+ {
+ "epoch": 1.7004443845946673,
+ "grad_norm": 0.73339929516101,
+ "learning_rate": 5.43623725764389e-07,
+ "loss": 0.3154341697692871,
+ "num_input_tokens_seen": 1104367200,
+ "step": 25255,
+ "train_runtime": 195396.1799,
+ "train_tokens_per_second": 5651.939
+ },
+ {
+ "epoch": 1.7007810395906275,
+ "grad_norm": 0.7555499440673459,
+ "learning_rate": 5.424253562934001e-07,
+ "loss": 0.29883978366851804,
+ "num_input_tokens_seen": 1104597088,
+ "step": 25260,
+ "train_runtime": 195441.1857,
+ "train_tokens_per_second": 5651.813
+ },
+ {
+ "epoch": 1.7011176945865878,
+ "grad_norm": 0.7445236468484644,
+ "learning_rate": 5.412282333689461e-07,
+ "loss": 0.29061620235443114,
+ "num_input_tokens_seen": 1104824056,
+ "step": 25265,
+ "train_runtime": 195480.4688,
+ "train_tokens_per_second": 5651.839
+ },
+ {
+ "epoch": 1.7014543495825478,
+ "grad_norm": 0.7030179795843015,
+ "learning_rate": 5.400323573257987e-07,
+ "loss": 0.29516527652740476,
+ "num_input_tokens_seen": 1105049880,
+ "step": 25270,
+ "train_runtime": 195517.0208,
+ "train_tokens_per_second": 5651.937
+ },
+ {
+ "epoch": 1.7017910045785078,
+ "grad_norm": 0.7158916904045084,
+ "learning_rate": 5.388377284983803e-07,
+ "loss": 0.287148642539978,
+ "num_input_tokens_seen": 1105277608,
+ "step": 25275,
+ "train_runtime": 195553.5443,
+ "train_tokens_per_second": 5652.046
+ },
+ {
+ "epoch": 1.702127659574468,
+ "grad_norm": 0.7855552714949071,
+ "learning_rate": 5.37644347220767e-07,
+ "loss": 0.2909886360168457,
+ "num_input_tokens_seen": 1105492920,
+ "step": 25280,
+ "train_runtime": 195593.7926,
+ "train_tokens_per_second": 5651.984
+ },
+ {
+ "epoch": 1.7024643145704283,
+ "grad_norm": 0.7992003312183278,
+ "learning_rate": 5.364522138266809e-07,
+ "loss": 0.30830817222595214,
+ "num_input_tokens_seen": 1105725216,
+ "step": 25285,
+ "train_runtime": 195629.7512,
+ "train_tokens_per_second": 5652.132
+ },
+ {
+ "epoch": 1.7028009695663884,
+ "grad_norm": 0.7548214170551567,
+ "learning_rate": 5.352613286495018e-07,
+ "loss": 0.2986649990081787,
+ "num_input_tokens_seen": 1105936272,
+ "step": 25290,
+ "train_runtime": 195672.9754,
+ "train_tokens_per_second": 5651.962
+ },
+ {
+ "epoch": 1.7031376245623484,
+ "grad_norm": 0.6650218012220956,
+ "learning_rate": 5.340716920222527e-07,
+ "loss": 0.28066387176513674,
+ "num_input_tokens_seen": 1106155736,
+ "step": 25295,
+ "train_runtime": 195711.2334,
+ "train_tokens_per_second": 5651.979
+ },
+ {
+ "epoch": 1.7034742795583087,
+ "grad_norm": 0.8192022199015255,
+ "learning_rate": 5.32883304277616e-07,
+ "loss": 0.3220468521118164,
+ "num_input_tokens_seen": 1106367880,
+ "step": 25300,
+ "train_runtime": 195754.2351,
+ "train_tokens_per_second": 5651.821
+ },
+ {
+ "epoch": 1.703810934554269,
+ "grad_norm": 0.8220156836674571,
+ "learning_rate": 5.31696165747918e-07,
+ "loss": 0.2808229446411133,
+ "num_input_tokens_seen": 1106591160,
+ "step": 25305,
+ "train_runtime": 195793.8848,
+ "train_tokens_per_second": 5651.817
+ },
+ {
+ "epoch": 1.704147589550229,
+ "grad_norm": 0.7943490070647735,
+ "learning_rate": 5.305102767651393e-07,
+ "loss": 0.28664450645446776,
+ "num_input_tokens_seen": 1106820392,
+ "step": 25310,
+ "train_runtime": 195840.7495,
+ "train_tokens_per_second": 5651.635
+ },
+ {
+ "epoch": 1.704484244546189,
+ "grad_norm": 0.7913194850774943,
+ "learning_rate": 5.293256376609091e-07,
+ "loss": 0.30523157119750977,
+ "num_input_tokens_seen": 1107043152,
+ "step": 25315,
+ "train_runtime": 195882.9879,
+ "train_tokens_per_second": 5651.553
+ },
+ {
+ "epoch": 1.7048208995421492,
+ "grad_norm": 0.7081852025829084,
+ "learning_rate": 5.281422487665089e-07,
+ "loss": 0.27099974155426027,
+ "num_input_tokens_seen": 1107273600,
+ "step": 25320,
+ "train_runtime": 195922.0686,
+ "train_tokens_per_second": 5651.602
+ },
+ {
+ "epoch": 1.7051575545381095,
+ "grad_norm": 0.7673688868553783,
+ "learning_rate": 5.269601104128674e-07,
+ "loss": 0.2887356042861938,
+ "num_input_tokens_seen": 1107500568,
+ "step": 25325,
+ "train_runtime": 195968.4871,
+ "train_tokens_per_second": 5651.422
+ },
+ {
+ "epoch": 1.7054942095340695,
+ "grad_norm": 0.7390671103279993,
+ "learning_rate": 5.257792229305702e-07,
+ "loss": 0.3096314430236816,
+ "num_input_tokens_seen": 1107721408,
+ "step": 25330,
+ "train_runtime": 196009.11,
+ "train_tokens_per_second": 5651.377
+ },
+ {
+ "epoch": 1.7058308645300295,
+ "grad_norm": 0.9154523312754923,
+ "learning_rate": 5.245995866498438e-07,
+ "loss": 0.27949187755584715,
+ "num_input_tokens_seen": 1107927904,
+ "step": 25335,
+ "train_runtime": 196050.2257,
+ "train_tokens_per_second": 5651.245
+ },
+ {
+ "epoch": 1.7061675195259898,
+ "grad_norm": 0.779691321745416,
+ "learning_rate": 5.234212019005741e-07,
+ "loss": 0.3069715738296509,
+ "num_input_tokens_seen": 1108137040,
+ "step": 25340,
+ "train_runtime": 196092.6734,
+ "train_tokens_per_second": 5651.088
+ },
+ {
+ "epoch": 1.70650417452195,
+ "grad_norm": 0.7743911753132502,
+ "learning_rate": 5.222440690122882e-07,
+ "loss": 0.3008026838302612,
+ "num_input_tokens_seen": 1108352000,
+ "step": 25345,
+ "train_runtime": 196133.018,
+ "train_tokens_per_second": 5651.022
+ },
+ {
+ "epoch": 1.70684082951791,
+ "grad_norm": 0.771105877517414,
+ "learning_rate": 5.210681883141716e-07,
+ "loss": 0.2749685764312744,
+ "num_input_tokens_seen": 1108571744,
+ "step": 25350,
+ "train_runtime": 196169.374,
+ "train_tokens_per_second": 5651.095
+ },
+ {
+ "epoch": 1.70717748451387,
+ "grad_norm": 0.7792881082810094,
+ "learning_rate": 5.19893560135053e-07,
+ "loss": 0.3137876749038696,
+ "num_input_tokens_seen": 1108803560,
+ "step": 25355,
+ "train_runtime": 196203.8165,
+ "train_tokens_per_second": 5651.284
+ },
+ {
+ "epoch": 1.7075141395098303,
+ "grad_norm": 0.7995549624581778,
+ "learning_rate": 5.187201848034146e-07,
+ "loss": 0.3042750835418701,
+ "num_input_tokens_seen": 1109023488,
+ "step": 25360,
+ "train_runtime": 196245.4211,
+ "train_tokens_per_second": 5651.207
+ },
+ {
+ "epoch": 1.7078507945057906,
+ "grad_norm": 0.7706690517207955,
+ "learning_rate": 5.175480626473873e-07,
+ "loss": 0.290797758102417,
+ "num_input_tokens_seen": 1109234112,
+ "step": 25365,
+ "train_runtime": 196284.3763,
+ "train_tokens_per_second": 5651.158
+ },
+ {
+ "epoch": 1.7081874495017506,
+ "grad_norm": 0.888485346761317,
+ "learning_rate": 5.163771939947504e-07,
+ "loss": 0.30904016494750974,
+ "num_input_tokens_seen": 1109454048,
+ "step": 25370,
+ "train_runtime": 196324.1237,
+ "train_tokens_per_second": 5651.135
+ },
+ {
+ "epoch": 1.7085241044977106,
+ "grad_norm": 0.6868307931042189,
+ "learning_rate": 5.152075791729327e-07,
+ "loss": 0.3244988203048706,
+ "num_input_tokens_seen": 1109687648,
+ "step": 25375,
+ "train_runtime": 196364.0378,
+ "train_tokens_per_second": 5651.176
+ },
+ {
+ "epoch": 1.7088607594936709,
+ "grad_norm": 0.7101255126564905,
+ "learning_rate": 5.140392185090165e-07,
+ "loss": 0.2983409881591797,
+ "num_input_tokens_seen": 1109893064,
+ "step": 25380,
+ "train_runtime": 196405.8629,
+ "train_tokens_per_second": 5651.018
+ },
+ {
+ "epoch": 1.7091974144896311,
+ "grad_norm": 0.8655143566241545,
+ "learning_rate": 5.128721123297265e-07,
+ "loss": 0.3025853157043457,
+ "num_input_tokens_seen": 1110127392,
+ "step": 25385,
+ "train_runtime": 196448.4648,
+ "train_tokens_per_second": 5650.985
+ },
+ {
+ "epoch": 1.7095340694855912,
+ "grad_norm": 0.7765483897971529,
+ "learning_rate": 5.117062609614438e-07,
+ "loss": 0.2883389949798584,
+ "num_input_tokens_seen": 1110349016,
+ "step": 25390,
+ "train_runtime": 196484.3558,
+ "train_tokens_per_second": 5651.081
+ },
+ {
+ "epoch": 1.7098707244815512,
+ "grad_norm": 0.8068466453475462,
+ "learning_rate": 5.105416647301909e-07,
+ "loss": 0.2935046195983887,
+ "num_input_tokens_seen": 1110572696,
+ "step": 25395,
+ "train_runtime": 196527.8258,
+ "train_tokens_per_second": 5650.969
+ },
+ {
+ "epoch": 1.7102073794775114,
+ "grad_norm": 0.8272038940843907,
+ "learning_rate": 5.093783239616468e-07,
+ "loss": 0.26650362014770507,
+ "num_input_tokens_seen": 1110779632,
+ "step": 25400,
+ "train_runtime": 196565.2985,
+ "train_tokens_per_second": 5650.945
+ },
+ {
+ "epoch": 1.7105440344734717,
+ "grad_norm": 0.8912009620998278,
+ "learning_rate": 5.082162389811352e-07,
+ "loss": 0.28749451637268064,
+ "num_input_tokens_seen": 1110997128,
+ "step": 25405,
+ "train_runtime": 196600.4482,
+ "train_tokens_per_second": 5651.041
+ },
+ {
+ "epoch": 1.7108806894694317,
+ "grad_norm": 0.8227134244564941,
+ "learning_rate": 5.070554101136288e-07,
+ "loss": 0.2878044605255127,
+ "num_input_tokens_seen": 1111200552,
+ "step": 25410,
+ "train_runtime": 196637.1184,
+ "train_tokens_per_second": 5651.021
+ },
+ {
+ "epoch": 1.7112173444653918,
+ "grad_norm": 0.6957933161324612,
+ "learning_rate": 5.058958376837497e-07,
+ "loss": 0.2986098289489746,
+ "num_input_tokens_seen": 1111427560,
+ "step": 25415,
+ "train_runtime": 196678.1193,
+ "train_tokens_per_second": 5650.997
+ },
+ {
+ "epoch": 1.711553999461352,
+ "grad_norm": 0.7375797617626252,
+ "learning_rate": 5.047375220157691e-07,
+ "loss": 0.3050894021987915,
+ "num_input_tokens_seen": 1111652600,
+ "step": 25420,
+ "train_runtime": 196718.6261,
+ "train_tokens_per_second": 5650.978
+ },
+ {
+ "epoch": 1.7118906544573123,
+ "grad_norm": 0.7728802385790842,
+ "learning_rate": 5.035804634336056e-07,
+ "loss": 0.3212940216064453,
+ "num_input_tokens_seen": 1111871248,
+ "step": 25425,
+ "train_runtime": 196760.4219,
+ "train_tokens_per_second": 5650.889
+ },
+ {
+ "epoch": 1.7122273094532723,
+ "grad_norm": 0.7147851806060915,
+ "learning_rate": 5.024246622608292e-07,
+ "loss": 0.293403172492981,
+ "num_input_tokens_seen": 1112108600,
+ "step": 25430,
+ "train_runtime": 196799.535,
+ "train_tokens_per_second": 5650.972
+ },
+ {
+ "epoch": 1.7125639644492323,
+ "grad_norm": 0.8753202500856931,
+ "learning_rate": 5.01270118820652e-07,
+ "loss": 0.3186427354812622,
+ "num_input_tokens_seen": 1112335328,
+ "step": 25435,
+ "train_runtime": 196834.1465,
+ "train_tokens_per_second": 5651.13
+ },
+ {
+ "epoch": 1.7129006194451926,
+ "grad_norm": 0.8169092701177705,
+ "learning_rate": 5.001168334359429e-07,
+ "loss": 0.3163825511932373,
+ "num_input_tokens_seen": 1112566808,
+ "step": 25440,
+ "train_runtime": 196867.059,
+ "train_tokens_per_second": 5651.361
+ },
+ {
+ "epoch": 1.7132372744411528,
+ "grad_norm": 0.7130198658873356,
+ "learning_rate": 4.989648064292102e-07,
+ "loss": 0.3128488540649414,
+ "num_input_tokens_seen": 1112795792,
+ "step": 25445,
+ "train_runtime": 196905.6927,
+ "train_tokens_per_second": 5651.415
+ },
+ {
+ "epoch": 1.7135739294371128,
+ "grad_norm": 0.8272498786011496,
+ "learning_rate": 4.978140381226177e-07,
+ "loss": 0.31338427066802976,
+ "num_input_tokens_seen": 1113015816,
+ "step": 25450,
+ "train_runtime": 196950.2891,
+ "train_tokens_per_second": 5651.253
+ },
+ {
+ "epoch": 1.7139105844330729,
+ "grad_norm": 0.6817930969478092,
+ "learning_rate": 4.966645288379729e-07,
+ "loss": 0.2886411428451538,
+ "num_input_tokens_seen": 1113242208,
+ "step": 25455,
+ "train_runtime": 196986.8049,
+ "train_tokens_per_second": 5651.354
+ },
+ {
+ "epoch": 1.7142472394290331,
+ "grad_norm": 0.8193216088135635,
+ "learning_rate": 4.955162788967327e-07,
+ "loss": 0.3108631134033203,
+ "num_input_tokens_seen": 1113433336,
+ "step": 25460,
+ "train_runtime": 197028.9568,
+ "train_tokens_per_second": 5651.115
+ },
+ {
+ "epoch": 1.7145838944249934,
+ "grad_norm": 0.7410886402030697,
+ "learning_rate": 4.943692886200013e-07,
+ "loss": 0.30652599334716796,
+ "num_input_tokens_seen": 1113646408,
+ "step": 25465,
+ "train_runtime": 197065.6766,
+ "train_tokens_per_second": 5651.143
+ },
+ {
+ "epoch": 1.7149205494209534,
+ "grad_norm": 0.7490199419500515,
+ "learning_rate": 4.932235583285305e-07,
+ "loss": 0.31150734424591064,
+ "num_input_tokens_seen": 1113881008,
+ "step": 25470,
+ "train_runtime": 197104.2655,
+ "train_tokens_per_second": 5651.227
+ },
+ {
+ "epoch": 1.7152572044169134,
+ "grad_norm": 0.8168471127164367,
+ "learning_rate": 4.920790883427201e-07,
+ "loss": 0.2943288326263428,
+ "num_input_tokens_seen": 1114079888,
+ "step": 25475,
+ "train_runtime": 197148.0427,
+ "train_tokens_per_second": 5650.981
+ },
+ {
+ "epoch": 1.7155938594128737,
+ "grad_norm": 0.8217572900775035,
+ "learning_rate": 4.909358789826196e-07,
+ "loss": 0.3120744228363037,
+ "num_input_tokens_seen": 1114268728,
+ "step": 25480,
+ "train_runtime": 197186.0762,
+ "train_tokens_per_second": 5650.849
+ },
+ {
+ "epoch": 1.715930514408834,
+ "grad_norm": 0.842144785636648,
+ "learning_rate": 4.897939305679206e-07,
+ "loss": 0.30831499099731446,
+ "num_input_tokens_seen": 1114498760,
+ "step": 25485,
+ "train_runtime": 197217.8311,
+ "train_tokens_per_second": 5651.105
+ },
+ {
+ "epoch": 1.716267169404794,
+ "grad_norm": 0.7113758542252159,
+ "learning_rate": 4.886532434179686e-07,
+ "loss": 0.2864993572235107,
+ "num_input_tokens_seen": 1114714920,
+ "step": 25490,
+ "train_runtime": 197261.5876,
+ "train_tokens_per_second": 5650.948
+ },
+ {
+ "epoch": 1.716603824400754,
+ "grad_norm": 0.7070024392396191,
+ "learning_rate": 4.875138178517497e-07,
+ "loss": 0.2663728713989258,
+ "num_input_tokens_seen": 1114931832,
+ "step": 25495,
+ "train_runtime": 197300.3097,
+ "train_tokens_per_second": 5650.938
+ },
+ {
+ "epoch": 1.7169404793967142,
+ "grad_norm": 0.8884464092351633,
+ "learning_rate": 4.863756541879028e-07,
+ "loss": 0.31972756385803225,
+ "num_input_tokens_seen": 1115156536,
+ "step": 25500,
+ "train_runtime": 197338.9855,
+ "train_tokens_per_second": 5650.969
+ },
+ {
+ "epoch": 1.7172771343926745,
+ "grad_norm": 0.693402281770043,
+ "learning_rate": 4.852387527447117e-07,
+ "loss": 0.2761407375335693,
+ "num_input_tokens_seen": 1115391624,
+ "step": 25505,
+ "train_runtime": 197375.3791,
+ "train_tokens_per_second": 5651.118
+ },
+ {
+ "epoch": 1.7176137893886345,
+ "grad_norm": 0.7175083758215528,
+ "learning_rate": 4.84103113840107e-07,
+ "loss": 0.2971747398376465,
+ "num_input_tokens_seen": 1115621104,
+ "step": 25510,
+ "train_runtime": 197410.3395,
+ "train_tokens_per_second": 5651.28
+ },
+ {
+ "epoch": 1.7179504443845945,
+ "grad_norm": 0.7429321287522253,
+ "learning_rate": 4.829687377916664e-07,
+ "loss": 0.2947702407836914,
+ "num_input_tokens_seen": 1115833104,
+ "step": 25515,
+ "train_runtime": 197441.5837,
+ "train_tokens_per_second": 5651.459
+ },
+ {
+ "epoch": 1.7182870993805548,
+ "grad_norm": 0.8107818792796975,
+ "learning_rate": 4.81835624916614e-07,
+ "loss": 0.2821854591369629,
+ "num_input_tokens_seen": 1116048240,
+ "step": 25520,
+ "train_runtime": 197479.0855,
+ "train_tokens_per_second": 5651.476
+ },
+ {
+ "epoch": 1.718623754376515,
+ "grad_norm": 0.802522255622259,
+ "learning_rate": 4.807037755318211e-07,
+ "loss": 0.28911077976226807,
+ "num_input_tokens_seen": 1116280624,
+ "step": 25525,
+ "train_runtime": 197512.9072,
+ "train_tokens_per_second": 5651.684
+ },
+ {
+ "epoch": 1.718960409372475,
+ "grad_norm": 0.6592568584202634,
+ "learning_rate": 4.795731899538086e-07,
+ "loss": 0.3094839096069336,
+ "num_input_tokens_seen": 1116492320,
+ "step": 25530,
+ "train_runtime": 197560.5492,
+ "train_tokens_per_second": 5651.393
+ },
+ {
+ "epoch": 1.719297064368435,
+ "grad_norm": 0.7237886999876593,
+ "learning_rate": 4.78443868498737e-07,
+ "loss": 0.2828222274780273,
+ "num_input_tokens_seen": 1116713608,
+ "step": 25535,
+ "train_runtime": 197596.8732,
+ "train_tokens_per_second": 5651.474
+ },
+ {
+ "epoch": 1.7196337193643954,
+ "grad_norm": 0.7185713199478717,
+ "learning_rate": 4.773158114824206e-07,
+ "loss": 0.28137338161468506,
+ "num_input_tokens_seen": 1116938536,
+ "step": 25540,
+ "train_runtime": 197632.4385,
+ "train_tokens_per_second": 5651.595
+ },
+ {
+ "epoch": 1.7199703743603556,
+ "grad_norm": 1.1879567747502189,
+ "learning_rate": 4.7618901922031445e-07,
+ "loss": 0.2812469959259033,
+ "num_input_tokens_seen": 1117161488,
+ "step": 25545,
+ "train_runtime": 197670.8796,
+ "train_tokens_per_second": 5651.624
+ },
+ {
+ "epoch": 1.7203070293563156,
+ "grad_norm": 0.8094448913417437,
+ "learning_rate": 4.750634920275243e-07,
+ "loss": 0.3022423267364502,
+ "num_input_tokens_seen": 1117374592,
+ "step": 25550,
+ "train_runtime": 197712.1996,
+ "train_tokens_per_second": 5651.521
+ },
+ {
+ "epoch": 1.7206436843522757,
+ "grad_norm": 0.752864591432035,
+ "learning_rate": 4.739392302188001e-07,
+ "loss": 0.319273042678833,
+ "num_input_tokens_seen": 1117603384,
+ "step": 25555,
+ "train_runtime": 197750.9243,
+ "train_tokens_per_second": 5651.571
+ },
+ {
+ "epoch": 1.720980339348236,
+ "grad_norm": 0.8294658057367533,
+ "learning_rate": 4.728162341085368e-07,
+ "loss": 0.31023268699645995,
+ "num_input_tokens_seen": 1117820784,
+ "step": 25560,
+ "train_runtime": 197793.1161,
+ "train_tokens_per_second": 5651.465
+ },
+ {
+ "epoch": 1.7213169943441962,
+ "grad_norm": 0.7053663821231337,
+ "learning_rate": 4.7169450401077843e-07,
+ "loss": 0.312652063369751,
+ "num_input_tokens_seen": 1118048160,
+ "step": 25565,
+ "train_runtime": 197836.8692,
+ "train_tokens_per_second": 5651.364
+ },
+ {
+ "epoch": 1.7216536493401562,
+ "grad_norm": 0.7700362624042942,
+ "learning_rate": 4.7057404023921174e-07,
+ "loss": 0.3073885440826416,
+ "num_input_tokens_seen": 1118274672,
+ "step": 25570,
+ "train_runtime": 197872.3987,
+ "train_tokens_per_second": 5651.494
+ },
+ {
+ "epoch": 1.7219903043361162,
+ "grad_norm": 0.8048237014417249,
+ "learning_rate": 4.694548431071705e-07,
+ "loss": 0.31366183757781985,
+ "num_input_tokens_seen": 1118492208,
+ "step": 25575,
+ "train_runtime": 197919.5839,
+ "train_tokens_per_second": 5651.246
+ },
+ {
+ "epoch": 1.7223269593320765,
+ "grad_norm": 0.7167405150431256,
+ "learning_rate": 4.683369129276377e-07,
+ "loss": 0.3049338340759277,
+ "num_input_tokens_seen": 1118716936,
+ "step": 25580,
+ "train_runtime": 197953.2698,
+ "train_tokens_per_second": 5651.419
+ },
+ {
+ "epoch": 1.7226636143280367,
+ "grad_norm": 0.663480470579991,
+ "learning_rate": 4.6722025001323444e-07,
+ "loss": 0.2772920846939087,
+ "num_input_tokens_seen": 1118917560,
+ "step": 25585,
+ "train_runtime": 197998.4722,
+ "train_tokens_per_second": 5651.142
+ },
+ {
+ "epoch": 1.7230002693239967,
+ "grad_norm": 0.8472714336560921,
+ "learning_rate": 4.6610485467623643e-07,
+ "loss": 0.3002648830413818,
+ "num_input_tokens_seen": 1119122208,
+ "step": 25590,
+ "train_runtime": 198037.4639,
+ "train_tokens_per_second": 5651.063
+ },
+ {
+ "epoch": 1.7233369243199568,
+ "grad_norm": 0.73662714701174,
+ "learning_rate": 4.649907272285564e-07,
+ "loss": 0.29230647087097167,
+ "num_input_tokens_seen": 1119341064,
+ "step": 25595,
+ "train_runtime": 198077.5592,
+ "train_tokens_per_second": 5651.024
+ },
+ {
+ "epoch": 1.723673579315917,
+ "grad_norm": 0.7420030033003574,
+ "learning_rate": 4.638778679817596e-07,
+ "loss": 0.28714590072631835,
+ "num_input_tokens_seen": 1119570192,
+ "step": 25600,
+ "train_runtime": 198120.1877,
+ "train_tokens_per_second": 5650.965
+ },
+ {
+ "epoch": 1.7240102343118773,
+ "grad_norm": 0.824030176081278,
+ "learning_rate": 4.6276627724705204e-07,
+ "loss": 0.3071947813034058,
+ "num_input_tokens_seen": 1119786040,
+ "step": 25605,
+ "train_runtime": 198161.3927,
+ "train_tokens_per_second": 5650.879
+ },
+ {
+ "epoch": 1.7243468893078373,
+ "grad_norm": 0.7154654354680215,
+ "learning_rate": 4.616559553352873e-07,
+ "loss": 0.2772822380065918,
+ "num_input_tokens_seen": 1120003688,
+ "step": 25610,
+ "train_runtime": 198197.9173,
+ "train_tokens_per_second": 5650.936
+ },
+ {
+ "epoch": 1.7246835443037973,
+ "grad_norm": 0.7332839778498565,
+ "learning_rate": 4.6054690255696256e-07,
+ "loss": 0.3121823310852051,
+ "num_input_tokens_seen": 1120225000,
+ "step": 25615,
+ "train_runtime": 198232.3778,
+ "train_tokens_per_second": 5651.07
+ },
+ {
+ "epoch": 1.7250201992997576,
+ "grad_norm": 0.8304618732220158,
+ "learning_rate": 4.594391192222214e-07,
+ "loss": 0.299558162689209,
+ "num_input_tokens_seen": 1120436736,
+ "step": 25620,
+ "train_runtime": 198275.4196,
+ "train_tokens_per_second": 5650.911
+ },
+ {
+ "epoch": 1.7253568542957178,
+ "grad_norm": 0.7210943628722162,
+ "learning_rate": 4.583326056408505e-07,
+ "loss": 0.2828629732131958,
+ "num_input_tokens_seen": 1120656400,
+ "step": 25625,
+ "train_runtime": 198318.2954,
+ "train_tokens_per_second": 5650.797
+ },
+ {
+ "epoch": 1.7256935092916779,
+ "grad_norm": 0.8701484329791821,
+ "learning_rate": 4.5722736212228613e-07,
+ "loss": 0.3062956094741821,
+ "num_input_tokens_seen": 1120853112,
+ "step": 25630,
+ "train_runtime": 198360.6419,
+ "train_tokens_per_second": 5650.582
+ },
+ {
+ "epoch": 1.726030164287638,
+ "grad_norm": 0.7789016257709495,
+ "learning_rate": 4.5612338897560227e-07,
+ "loss": 0.3126507759094238,
+ "num_input_tokens_seen": 1121085288,
+ "step": 25635,
+ "train_runtime": 198403.9135,
+ "train_tokens_per_second": 5650.52
+ },
+ {
+ "epoch": 1.7263668192835981,
+ "grad_norm": 0.8101928313460407,
+ "learning_rate": 4.5502068650952527e-07,
+ "loss": 0.31250529289245604,
+ "num_input_tokens_seen": 1121299992,
+ "step": 25640,
+ "train_runtime": 198443.451,
+ "train_tokens_per_second": 5650.476
+ },
+ {
+ "epoch": 1.7267034742795584,
+ "grad_norm": 0.8289277705819492,
+ "learning_rate": 4.5391925503241894e-07,
+ "loss": 0.2995558023452759,
+ "num_input_tokens_seen": 1121526960,
+ "step": 25645,
+ "train_runtime": 198482.6522,
+ "train_tokens_per_second": 5650.504
+ },
+ {
+ "epoch": 1.7270401292755184,
+ "grad_norm": 0.7957658163232814,
+ "learning_rate": 4.5281909485229704e-07,
+ "loss": 0.29502243995666505,
+ "num_input_tokens_seen": 1121740648,
+ "step": 25650,
+ "train_runtime": 198523.6274,
+ "train_tokens_per_second": 5650.414
+ },
+ {
+ "epoch": 1.7273767842714784,
+ "grad_norm": 0.8283561928306208,
+ "learning_rate": 4.51720206276815e-07,
+ "loss": 0.27261307239532473,
+ "num_input_tokens_seen": 1121954360,
+ "step": 25655,
+ "train_runtime": 198559.855,
+ "train_tokens_per_second": 5650.459
+ },
+ {
+ "epoch": 1.7277134392674387,
+ "grad_norm": 0.7838264977883859,
+ "learning_rate": 4.506225896132743e-07,
+ "loss": 0.2796020984649658,
+ "num_input_tokens_seen": 1122168368,
+ "step": 25660,
+ "train_runtime": 198597.1547,
+ "train_tokens_per_second": 5650.476
+ },
+ {
+ "epoch": 1.728050094263399,
+ "grad_norm": 0.7911207040134044,
+ "learning_rate": 4.4952624516861944e-07,
+ "loss": 0.2953200101852417,
+ "num_input_tokens_seen": 1122372424,
+ "step": 25665,
+ "train_runtime": 198632.1048,
+ "train_tokens_per_second": 5650.509
+ },
+ {
+ "epoch": 1.728386749259359,
+ "grad_norm": 0.7188871555974835,
+ "learning_rate": 4.4843117324944005e-07,
+ "loss": 0.2733217716217041,
+ "num_input_tokens_seen": 1122608584,
+ "step": 25670,
+ "train_runtime": 198671.3821,
+ "train_tokens_per_second": 5650.58
+ },
+ {
+ "epoch": 1.728723404255319,
+ "grad_norm": 0.8236529240166243,
+ "learning_rate": 4.4733737416196745e-07,
+ "loss": 0.2921212911605835,
+ "num_input_tokens_seen": 1122786752,
+ "step": 25675,
+ "train_runtime": 198705.9189,
+ "train_tokens_per_second": 5650.495
+ },
+ {
+ "epoch": 1.7290600592512793,
+ "grad_norm": 0.6891675836585306,
+ "learning_rate": 4.4624484821208346e-07,
+ "loss": 0.29300737380981445,
+ "num_input_tokens_seen": 1123000352,
+ "step": 25680,
+ "train_runtime": 198743.7451,
+ "train_tokens_per_second": 5650.494
+ },
+ {
+ "epoch": 1.7293967142472395,
+ "grad_norm": 0.7600535970631421,
+ "learning_rate": 4.451535957053044e-07,
+ "loss": 0.26680479049682615,
+ "num_input_tokens_seen": 1123218520,
+ "step": 25685,
+ "train_runtime": 198786.9925,
+ "train_tokens_per_second": 5650.362
+ },
+ {
+ "epoch": 1.7297333692431995,
+ "grad_norm": 0.7452499443166702,
+ "learning_rate": 4.440636169467999e-07,
+ "loss": 0.28789448738098145,
+ "num_input_tokens_seen": 1123429840,
+ "step": 25690,
+ "train_runtime": 198828.3828,
+ "train_tokens_per_second": 5650.249
+ },
+ {
+ "epoch": 1.7300700242391596,
+ "grad_norm": 0.801697711689962,
+ "learning_rate": 4.429749122413757e-07,
+ "loss": 0.2902831077575684,
+ "num_input_tokens_seen": 1123644168,
+ "step": 25695,
+ "train_runtime": 198871.0647,
+ "train_tokens_per_second": 5650.114
+ },
+ {
+ "epoch": 1.7304066792351198,
+ "grad_norm": 0.7857956805272411,
+ "learning_rate": 4.418874818934865e-07,
+ "loss": 0.29162254333496096,
+ "num_input_tokens_seen": 1123867240,
+ "step": 25700,
+ "train_runtime": 198907.5425,
+ "train_tokens_per_second": 5650.199
+ },
+ {
+ "epoch": 1.73074333423108,
+ "grad_norm": 0.829177173381663,
+ "learning_rate": 4.4080132620722806e-07,
+ "loss": 0.30682687759399413,
+ "num_input_tokens_seen": 1124064016,
+ "step": 25705,
+ "train_runtime": 198944.8498,
+ "train_tokens_per_second": 5650.129
+ },
+ {
+ "epoch": 1.73107998922704,
+ "grad_norm": 0.7300356398237123,
+ "learning_rate": 4.3971644548634116e-07,
+ "loss": 0.30285561084747314,
+ "num_input_tokens_seen": 1124276384,
+ "step": 25710,
+ "train_runtime": 198983.4747,
+ "train_tokens_per_second": 5650.099
+ },
+ {
+ "epoch": 1.7314166442230001,
+ "grad_norm": 1.020129891409066,
+ "learning_rate": 4.3863284003420826e-07,
+ "loss": 0.3177463531494141,
+ "num_input_tokens_seen": 1124494424,
+ "step": 25715,
+ "train_runtime": 199023.7261,
+ "train_tokens_per_second": 5650.052
+ },
+ {
+ "epoch": 1.7317532992189604,
+ "grad_norm": 0.7523401828963102,
+ "learning_rate": 4.375505101538563e-07,
+ "loss": 0.2902905702590942,
+ "num_input_tokens_seen": 1124712920,
+ "step": 25720,
+ "train_runtime": 199061.3982,
+ "train_tokens_per_second": 5650.08
+ },
+ {
+ "epoch": 1.7320899542149206,
+ "grad_norm": 0.8005644638652999,
+ "learning_rate": 4.3646945614795487e-07,
+ "loss": 0.3163978099822998,
+ "num_input_tokens_seen": 1124936520,
+ "step": 25725,
+ "train_runtime": 199095.1882,
+ "train_tokens_per_second": 5650.245
+ },
+ {
+ "epoch": 1.7324266092108807,
+ "grad_norm": 0.7714782744078659,
+ "learning_rate": 4.3538967831881995e-07,
+ "loss": 0.31358091831207274,
+ "num_input_tokens_seen": 1125167472,
+ "step": 25730,
+ "train_runtime": 199132.5807,
+ "train_tokens_per_second": 5650.343
+ },
+ {
+ "epoch": 1.7327632642068407,
+ "grad_norm": 0.7624866372005362,
+ "learning_rate": 4.343111769684039e-07,
+ "loss": 0.29058811664581297,
+ "num_input_tokens_seen": 1125373784,
+ "step": 25735,
+ "train_runtime": 199172.4308,
+ "train_tokens_per_second": 5650.249
+ },
+ {
+ "epoch": 1.733099919202801,
+ "grad_norm": 0.6799023332475078,
+ "learning_rate": 4.332339523983103e-07,
+ "loss": 0.27836108207702637,
+ "num_input_tokens_seen": 1125601280,
+ "step": 25740,
+ "train_runtime": 199212.9839,
+ "train_tokens_per_second": 5650.241
+ },
+ {
+ "epoch": 1.7334365741987612,
+ "grad_norm": 0.7561206882286643,
+ "learning_rate": 4.321580049097773e-07,
+ "loss": 0.28482909202575685,
+ "num_input_tokens_seen": 1125825176,
+ "step": 25745,
+ "train_runtime": 199246.8335,
+ "train_tokens_per_second": 5650.404
+ },
+ {
+ "epoch": 1.7337732291947212,
+ "grad_norm": 0.8118236999029766,
+ "learning_rate": 4.310833348036941e-07,
+ "loss": 0.27461097240447996,
+ "num_input_tokens_seen": 1126041128,
+ "step": 25750,
+ "train_runtime": 199290.7549,
+ "train_tokens_per_second": 5650.243
+ },
+ {
+ "epoch": 1.7341098841906812,
+ "grad_norm": 0.7118638487915145,
+ "learning_rate": 4.300099423805865e-07,
+ "loss": 0.2958423614501953,
+ "num_input_tokens_seen": 1126244560,
+ "step": 25755,
+ "train_runtime": 199333.5162,
+ "train_tokens_per_second": 5650.051
+ },
+ {
+ "epoch": 1.7344465391866415,
+ "grad_norm": 0.7506783418387937,
+ "learning_rate": 4.289378279406264e-07,
+ "loss": 0.2995717525482178,
+ "num_input_tokens_seen": 1126468312,
+ "step": 25760,
+ "train_runtime": 199366.8538,
+ "train_tokens_per_second": 5650.229
+ },
+ {
+ "epoch": 1.7347831941826017,
+ "grad_norm": 0.8246621666903383,
+ "learning_rate": 4.2786699178362647e-07,
+ "loss": 0.30245399475097656,
+ "num_input_tokens_seen": 1126689568,
+ "step": 25765,
+ "train_runtime": 199407.3048,
+ "train_tokens_per_second": 5650.192
+ },
+ {
+ "epoch": 1.7351198491785618,
+ "grad_norm": 0.8135304755844003,
+ "learning_rate": 4.2679743420904254e-07,
+ "loss": 0.3159316539764404,
+ "num_input_tokens_seen": 1126917656,
+ "step": 25770,
+ "train_runtime": 199441.5657,
+ "train_tokens_per_second": 5650.365
+ },
+ {
+ "epoch": 1.7354565041745218,
+ "grad_norm": 0.8164341987242794,
+ "learning_rate": 4.257291555159726e-07,
+ "loss": 0.26204819679260255,
+ "num_input_tokens_seen": 1127132592,
+ "step": 25775,
+ "train_runtime": 199474.9349,
+ "train_tokens_per_second": 5650.497
+ },
+ {
+ "epoch": 1.735793159170482,
+ "grad_norm": 0.7634016130973207,
+ "learning_rate": 4.246621560031594e-07,
+ "loss": 0.2924259424209595,
+ "num_input_tokens_seen": 1127339712,
+ "step": 25780,
+ "train_runtime": 199509.5428,
+ "train_tokens_per_second": 5650.555
+ },
+ {
+ "epoch": 1.7361298141664423,
+ "grad_norm": 0.7910333110781415,
+ "learning_rate": 4.235964359689837e-07,
+ "loss": 0.27674455642700196,
+ "num_input_tokens_seen": 1127566704,
+ "step": 25785,
+ "train_runtime": 199548.866,
+ "train_tokens_per_second": 5650.579
+ },
+ {
+ "epoch": 1.7364664691624023,
+ "grad_norm": 0.8840197427923319,
+ "learning_rate": 4.225319957114726e-07,
+ "loss": 0.3013903617858887,
+ "num_input_tokens_seen": 1127772128,
+ "step": 25790,
+ "train_runtime": 199588.6729,
+ "train_tokens_per_second": 5650.482
+ },
+ {
+ "epoch": 1.7368031241583624,
+ "grad_norm": 0.7325596847908312,
+ "learning_rate": 4.214688355282909e-07,
+ "loss": 0.28219919204711913,
+ "num_input_tokens_seen": 1127988664,
+ "step": 25795,
+ "train_runtime": 199631.9302,
+ "train_tokens_per_second": 5650.342
+ },
+ {
+ "epoch": 1.7371397791543226,
+ "grad_norm": 0.8173650460918741,
+ "learning_rate": 4.204069557167506e-07,
+ "loss": 0.2839889764785767,
+ "num_input_tokens_seen": 1128202648,
+ "step": 25800,
+ "train_runtime": 199667.9001,
+ "train_tokens_per_second": 5650.396
+ },
+ {
+ "epoch": 1.7374764341502829,
+ "grad_norm": 0.876625098531604,
+ "learning_rate": 4.1934635657380153e-07,
+ "loss": 0.304730224609375,
+ "num_input_tokens_seen": 1128423984,
+ "step": 25805,
+ "train_runtime": 199702.6037,
+ "train_tokens_per_second": 5650.522
+ },
+ {
+ "epoch": 1.7378130891462429,
+ "grad_norm": 0.7480060231681754,
+ "learning_rate": 4.1828703839603637e-07,
+ "loss": 0.2796186447143555,
+ "num_input_tokens_seen": 1128651136,
+ "step": 25810,
+ "train_runtime": 199740.7074,
+ "train_tokens_per_second": 5650.581
+ },
+ {
+ "epoch": 1.738149744142203,
+ "grad_norm": 0.8165869116083727,
+ "learning_rate": 4.172290014796926e-07,
+ "loss": 0.32165372371673584,
+ "num_input_tokens_seen": 1128867360,
+ "step": 25815,
+ "train_runtime": 199775.1172,
+ "train_tokens_per_second": 5650.691
+ },
+ {
+ "epoch": 1.7384863991381632,
+ "grad_norm": 0.8122569762079913,
+ "learning_rate": 4.1617224612064353e-07,
+ "loss": 0.29702818393707275,
+ "num_input_tokens_seen": 1129077456,
+ "step": 25820,
+ "train_runtime": 199815.7373,
+ "train_tokens_per_second": 5650.593
+ },
+ {
+ "epoch": 1.7388230541341234,
+ "grad_norm": 0.7151797283997641,
+ "learning_rate": 4.151167726144101e-07,
+ "loss": 0.2909691333770752,
+ "num_input_tokens_seen": 1129312768,
+ "step": 25825,
+ "train_runtime": 199850.2397,
+ "train_tokens_per_second": 5650.795
+ },
+ {
+ "epoch": 1.7391597091300834,
+ "grad_norm": 0.7870049200470505,
+ "learning_rate": 4.1406258125615106e-07,
+ "loss": 0.3070362567901611,
+ "num_input_tokens_seen": 1129537904,
+ "step": 25830,
+ "train_runtime": 199880.6637,
+ "train_tokens_per_second": 5651.061
+ },
+ {
+ "epoch": 1.7394963641260435,
+ "grad_norm": 0.6937152886038869,
+ "learning_rate": 4.130096723406674e-07,
+ "loss": 0.29037325382232665,
+ "num_input_tokens_seen": 1129757208,
+ "step": 25835,
+ "train_runtime": 199915.4723,
+ "train_tokens_per_second": 5651.174
+ },
+ {
+ "epoch": 1.7398330191220037,
+ "grad_norm": 0.8109854430003198,
+ "learning_rate": 4.119580461624023e-07,
+ "loss": 0.3026682615280151,
+ "num_input_tokens_seen": 1129966384,
+ "step": 25840,
+ "train_runtime": 199951.9393,
+ "train_tokens_per_second": 5651.19
+ },
+ {
+ "epoch": 1.740169674117964,
+ "grad_norm": 0.77895028602625,
+ "learning_rate": 4.109077030154396e-07,
+ "loss": 0.30320086479187014,
+ "num_input_tokens_seen": 1130186152,
+ "step": 25845,
+ "train_runtime": 199986.3251,
+ "train_tokens_per_second": 5651.317
+ },
+ {
+ "epoch": 1.740506329113924,
+ "grad_norm": 0.7121386019107564,
+ "learning_rate": 4.0985864319350366e-07,
+ "loss": 0.3050692081451416,
+ "num_input_tokens_seen": 1130406368,
+ "step": 25850,
+ "train_runtime": 200030.0945,
+ "train_tokens_per_second": 5651.181
+ },
+ {
+ "epoch": 1.740842984109884,
+ "grad_norm": 0.8089630935847927,
+ "learning_rate": 4.0881086698996154e-07,
+ "loss": 0.2959964513778687,
+ "num_input_tokens_seen": 1130629712,
+ "step": 25855,
+ "train_runtime": 200072.2089,
+ "train_tokens_per_second": 5651.108
+ },
+ {
+ "epoch": 1.7411796391058443,
+ "grad_norm": 0.6596855615199905,
+ "learning_rate": 4.077643746978194e-07,
+ "loss": 0.2970879316329956,
+ "num_input_tokens_seen": 1130854672,
+ "step": 25860,
+ "train_runtime": 200107.7191,
+ "train_tokens_per_second": 5651.23
+ },
+ {
+ "epoch": 1.7415162941018045,
+ "grad_norm": 0.7016863080677025,
+ "learning_rate": 4.067191666097281e-07,
+ "loss": 0.29431335926055907,
+ "num_input_tokens_seen": 1131071712,
+ "step": 25865,
+ "train_runtime": 200143.6703,
+ "train_tokens_per_second": 5651.299
+ },
+ {
+ "epoch": 1.7418529490977646,
+ "grad_norm": 0.721345869315641,
+ "learning_rate": 4.056752430179728e-07,
+ "loss": 0.28345699310302735,
+ "num_input_tokens_seen": 1131294176,
+ "step": 25870,
+ "train_runtime": 200187.8959,
+ "train_tokens_per_second": 5651.162
+ },
+ {
+ "epoch": 1.7421896040937246,
+ "grad_norm": 0.7756187950368285,
+ "learning_rate": 4.046326042144866e-07,
+ "loss": 0.2875986576080322,
+ "num_input_tokens_seen": 1131517072,
+ "step": 25875,
+ "train_runtime": 200228.1281,
+ "train_tokens_per_second": 5651.139
+ },
+ {
+ "epoch": 1.7425262590896848,
+ "grad_norm": 0.6831782860497421,
+ "learning_rate": 4.0359125049083956e-07,
+ "loss": 0.2836933135986328,
+ "num_input_tokens_seen": 1131734008,
+ "step": 25880,
+ "train_runtime": 200260.4735,
+ "train_tokens_per_second": 5651.31
+ },
+ {
+ "epoch": 1.742862914085645,
+ "grad_norm": 0.7787460128342357,
+ "learning_rate": 4.025511821382422e-07,
+ "loss": 0.30028643608093264,
+ "num_input_tokens_seen": 1131948944,
+ "step": 25885,
+ "train_runtime": 200306.9879,
+ "train_tokens_per_second": 5651.071
+ },
+ {
+ "epoch": 1.7431995690816051,
+ "grad_norm": 0.7541542349539324,
+ "learning_rate": 4.015123994475462e-07,
+ "loss": 0.28558173179626467,
+ "num_input_tokens_seen": 1132155112,
+ "step": 25890,
+ "train_runtime": 200346.4536,
+ "train_tokens_per_second": 5650.987
+ },
+ {
+ "epoch": 1.7435362240775654,
+ "grad_norm": 0.7540947221379003,
+ "learning_rate": 4.0047490270924417e-07,
+ "loss": 0.29801325798034667,
+ "num_input_tokens_seen": 1132391904,
+ "step": 25895,
+ "train_runtime": 200377.6526,
+ "train_tokens_per_second": 5651.288
+ },
+ {
+ "epoch": 1.7438728790735256,
+ "grad_norm": 0.7097292901302847,
+ "learning_rate": 3.994386922134691e-07,
+ "loss": 0.2968616247177124,
+ "num_input_tokens_seen": 1132613664,
+ "step": 25900,
+ "train_runtime": 200410.4186,
+ "train_tokens_per_second": 5651.471
+ },
+ {
+ "epoch": 1.7442095340694856,
+ "grad_norm": 0.6871287862369735,
+ "learning_rate": 3.984037682499936e-07,
+ "loss": 0.2899972438812256,
+ "num_input_tokens_seen": 1132836560,
+ "step": 25905,
+ "train_runtime": 200448.2196,
+ "train_tokens_per_second": 5651.517
+ },
+ {
+ "epoch": 1.7445461890654457,
+ "grad_norm": 0.7897783179799106,
+ "learning_rate": 3.973701311082301e-07,
+ "loss": 0.31725754737854006,
+ "num_input_tokens_seen": 1133050712,
+ "step": 25910,
+ "train_runtime": 200489.2159,
+ "train_tokens_per_second": 5651.43
+ },
+ {
+ "epoch": 1.744882844061406,
+ "grad_norm": 0.8896464404276014,
+ "learning_rate": 3.9633778107723454e-07,
+ "loss": 0.2900426387786865,
+ "num_input_tokens_seen": 1133254224,
+ "step": 25915,
+ "train_runtime": 200526.2526,
+ "train_tokens_per_second": 5651.401
+ },
+ {
+ "epoch": 1.7452194990573662,
+ "grad_norm": 0.7236782781112409,
+ "learning_rate": 3.953067184456966e-07,
+ "loss": 0.3090181350708008,
+ "num_input_tokens_seen": 1133470824,
+ "step": 25920,
+ "train_runtime": 200564.9033,
+ "train_tokens_per_second": 5651.392
+ },
+ {
+ "epoch": 1.7455561540533262,
+ "grad_norm": 0.8647141709585514,
+ "learning_rate": 3.9427694350195224e-07,
+ "loss": 0.31003541946411134,
+ "num_input_tokens_seen": 1133700592,
+ "step": 25925,
+ "train_runtime": 200601.8067,
+ "train_tokens_per_second": 5651.497
+ },
+ {
+ "epoch": 1.7458928090492862,
+ "grad_norm": 0.7588158649781692,
+ "learning_rate": 3.9324845653397446e-07,
+ "loss": 0.3075569152832031,
+ "num_input_tokens_seen": 1133925704,
+ "step": 25930,
+ "train_runtime": 200639.6339,
+ "train_tokens_per_second": 5651.554
+ },
+ {
+ "epoch": 1.7462294640452465,
+ "grad_norm": 0.7340500241842518,
+ "learning_rate": 3.922212578293755e-07,
+ "loss": 0.2702242374420166,
+ "num_input_tokens_seen": 1134140200,
+ "step": 25935,
+ "train_runtime": 200682.8182,
+ "train_tokens_per_second": 5651.407
+ },
+ {
+ "epoch": 1.7465661190412067,
+ "grad_norm": 0.82037212018234,
+ "learning_rate": 3.911953476754088e-07,
+ "loss": 0.2941112518310547,
+ "num_input_tokens_seen": 1134363752,
+ "step": 25940,
+ "train_runtime": 200715.6542,
+ "train_tokens_per_second": 5651.596
+ },
+ {
+ "epoch": 1.7469027740371668,
+ "grad_norm": 0.8812494234102816,
+ "learning_rate": 3.9017072635896716e-07,
+ "loss": 0.31668860912323,
+ "num_input_tokens_seen": 1134590888,
+ "step": 25945,
+ "train_runtime": 200756.942,
+ "train_tokens_per_second": 5651.565
+ },
+ {
+ "epoch": 1.7472394290331268,
+ "grad_norm": 0.76480797109403,
+ "learning_rate": 3.891473941665819e-07,
+ "loss": 0.3079123020172119,
+ "num_input_tokens_seen": 1134821944,
+ "step": 25950,
+ "train_runtime": 200795.9315,
+ "train_tokens_per_second": 5651.618
+ },
+ {
+ "epoch": 1.747576084029087,
+ "grad_norm": 0.8623020733146947,
+ "learning_rate": 3.881253513844246e-07,
+ "loss": 0.30869007110595703,
+ "num_input_tokens_seen": 1135026112,
+ "step": 25955,
+ "train_runtime": 200836.4857,
+ "train_tokens_per_second": 5651.494
+ },
+ {
+ "epoch": 1.7479127390250473,
+ "grad_norm": 0.7551876844288087,
+ "learning_rate": 3.871045982983063e-07,
+ "loss": 0.29468765258789065,
+ "num_input_tokens_seen": 1135250784,
+ "step": 25960,
+ "train_runtime": 200877.5576,
+ "train_tokens_per_second": 5651.457
+ },
+ {
+ "epoch": 1.7482493940210073,
+ "grad_norm": 0.7546664443489978,
+ "learning_rate": 3.8608513519367897e-07,
+ "loss": 0.2700496673583984,
+ "num_input_tokens_seen": 1135481416,
+ "step": 25965,
+ "train_runtime": 200922.6162,
+ "train_tokens_per_second": 5651.337
+ },
+ {
+ "epoch": 1.7485860490169673,
+ "grad_norm": 0.9863347976919302,
+ "learning_rate": 3.8506696235562915e-07,
+ "loss": 0.2968393087387085,
+ "num_input_tokens_seen": 1135693800,
+ "step": 25970,
+ "train_runtime": 200960.332,
+ "train_tokens_per_second": 5651.333
+ },
+ {
+ "epoch": 1.7489227040129276,
+ "grad_norm": 0.7481085355968223,
+ "learning_rate": 3.840500800688884e-07,
+ "loss": 0.2980025768280029,
+ "num_input_tokens_seen": 1135925744,
+ "step": 25975,
+ "train_runtime": 201006.8964,
+ "train_tokens_per_second": 5651.178
+ },
+ {
+ "epoch": 1.7492593590088878,
+ "grad_norm": 0.8214495132803978,
+ "learning_rate": 3.8303448861782333e-07,
+ "loss": 0.2952871322631836,
+ "num_input_tokens_seen": 1136130424,
+ "step": 25980,
+ "train_runtime": 201040.8482,
+ "train_tokens_per_second": 5651.242
+ },
+ {
+ "epoch": 1.7495960140048479,
+ "grad_norm": 0.8019257337743148,
+ "learning_rate": 3.8202018828644105e-07,
+ "loss": 0.27895636558532716,
+ "num_input_tokens_seen": 1136347560,
+ "step": 25985,
+ "train_runtime": 201079.8679,
+ "train_tokens_per_second": 5651.225
+ },
+ {
+ "epoch": 1.749932669000808,
+ "grad_norm": 0.9034770504035424,
+ "learning_rate": 3.8100717935838804e-07,
+ "loss": 0.32080807685852053,
+ "num_input_tokens_seen": 1136562648,
+ "step": 25990,
+ "train_runtime": 201121.1998,
+ "train_tokens_per_second": 5651.133
+ },
+ {
+ "epoch": 1.7502693239967682,
+ "grad_norm": 0.7869916616338762,
+ "learning_rate": 3.799954621169477e-07,
+ "loss": 0.27611637115478516,
+ "num_input_tokens_seen": 1136781352,
+ "step": 25995,
+ "train_runtime": 201160.3173,
+ "train_tokens_per_second": 5651.121
+ },
+ {
+ "epoch": 1.7506059789927284,
+ "grad_norm": 0.7949557024166627,
+ "learning_rate": 3.7898503684504483e-07,
+ "loss": 0.29705018997192384,
+ "num_input_tokens_seen": 1136994720,
+ "step": 26000,
+ "train_runtime": 201200.2712,
+ "train_tokens_per_second": 5651.06
+ },
+ {
+ "epoch": 1.7509426339886884,
+ "grad_norm": 0.7027696556054698,
+ "learning_rate": 3.7797590382524066e-07,
+ "loss": 0.28871927261352537,
+ "num_input_tokens_seen": 1137197696,
+ "step": 26005,
+ "train_runtime": 201242.4798,
+ "train_tokens_per_second": 5650.883
+ },
+ {
+ "epoch": 1.7512792889846485,
+ "grad_norm": 0.7864204996333114,
+ "learning_rate": 3.76968063339736e-07,
+ "loss": 0.3024580955505371,
+ "num_input_tokens_seen": 1137430488,
+ "step": 26010,
+ "train_runtime": 201274.5947,
+ "train_tokens_per_second": 5651.138
+ },
+ {
+ "epoch": 1.7516159439806087,
+ "grad_norm": 0.8310573898266609,
+ "learning_rate": 3.759615156703722e-07,
+ "loss": 0.28331038951873777,
+ "num_input_tokens_seen": 1137644120,
+ "step": 26015,
+ "train_runtime": 201312.2778,
+ "train_tokens_per_second": 5651.141
+ },
+ {
+ "epoch": 1.751952598976569,
+ "grad_norm": 0.7194041093377634,
+ "learning_rate": 3.749562610986235e-07,
+ "loss": 0.3102069854736328,
+ "num_input_tokens_seen": 1137861320,
+ "step": 26020,
+ "train_runtime": 201348.9701,
+ "train_tokens_per_second": 5651.19
+ },
+ {
+ "epoch": 1.752289253972529,
+ "grad_norm": 0.7426038956707869,
+ "learning_rate": 3.739522999056094e-07,
+ "loss": 0.28222415447235105,
+ "num_input_tokens_seen": 1138066152,
+ "step": 26025,
+ "train_runtime": 201385.6107,
+ "train_tokens_per_second": 5651.179
+ },
+ {
+ "epoch": 1.752625908968489,
+ "grad_norm": 0.7686985240205324,
+ "learning_rate": 3.7294963237208326e-07,
+ "loss": 0.3015716552734375,
+ "num_input_tokens_seen": 1138288896,
+ "step": 26030,
+ "train_runtime": 201430.5418,
+ "train_tokens_per_second": 5651.024
+ },
+ {
+ "epoch": 1.7529625639644493,
+ "grad_norm": 0.7939046805197044,
+ "learning_rate": 3.7194825877843786e-07,
+ "loss": 0.2880394458770752,
+ "num_input_tokens_seen": 1138513200,
+ "step": 26035,
+ "train_runtime": 201479.5121,
+ "train_tokens_per_second": 5650.764
+ },
+ {
+ "epoch": 1.7532992189604095,
+ "grad_norm": 0.7428618998986647,
+ "learning_rate": 3.7094817940470375e-07,
+ "loss": 0.30432782173156736,
+ "num_input_tokens_seen": 1138721632,
+ "step": 26040,
+ "train_runtime": 201520.1099,
+ "train_tokens_per_second": 5650.66
+ },
+ {
+ "epoch": 1.7536358739563696,
+ "grad_norm": 0.738527693119671,
+ "learning_rate": 3.6994939453055044e-07,
+ "loss": 0.2994675159454346,
+ "num_input_tokens_seen": 1138951136,
+ "step": 26045,
+ "train_runtime": 201561.9244,
+ "train_tokens_per_second": 5650.626
+ },
+ {
+ "epoch": 1.7539725289523296,
+ "grad_norm": 0.9322929614127222,
+ "learning_rate": 3.6895190443528515e-07,
+ "loss": 0.2943981170654297,
+ "num_input_tokens_seen": 1139175408,
+ "step": 26050,
+ "train_runtime": 201601.723,
+ "train_tokens_per_second": 5650.623
+ },
+ {
+ "epoch": 1.7543091839482898,
+ "grad_norm": 0.7070168137348763,
+ "learning_rate": 3.679557093978525e-07,
+ "loss": 0.312309193611145,
+ "num_input_tokens_seen": 1139396600,
+ "step": 26055,
+ "train_runtime": 201638.5926,
+ "train_tokens_per_second": 5650.687
+ },
+ {
+ "epoch": 1.75464583894425,
+ "grad_norm": 0.7986110214626218,
+ "learning_rate": 3.669608096968341e-07,
+ "loss": 0.29059772491455077,
+ "num_input_tokens_seen": 1139603752,
+ "step": 26060,
+ "train_runtime": 201681.4232,
+ "train_tokens_per_second": 5650.514
+ },
+ {
+ "epoch": 1.75498249394021,
+ "grad_norm": 0.8095452658769865,
+ "learning_rate": 3.6596720561045295e-07,
+ "loss": 0.2955317974090576,
+ "num_input_tokens_seen": 1139829360,
+ "step": 26065,
+ "train_runtime": 201718.99,
+ "train_tokens_per_second": 5650.58
+ },
+ {
+ "epoch": 1.7553191489361701,
+ "grad_norm": 0.791001274129923,
+ "learning_rate": 3.6497489741656455e-07,
+ "loss": 0.2989501476287842,
+ "num_input_tokens_seen": 1140053168,
+ "step": 26070,
+ "train_runtime": 201754.5113,
+ "train_tokens_per_second": 5650.695
+ },
+ {
+ "epoch": 1.7556558039321304,
+ "grad_norm": 0.6815029838821709,
+ "learning_rate": 3.639838853926669e-07,
+ "loss": 0.29015383720397947,
+ "num_input_tokens_seen": 1140284016,
+ "step": 26075,
+ "train_runtime": 201793.6507,
+ "train_tokens_per_second": 5650.743
+ },
+ {
+ "epoch": 1.7559924589280906,
+ "grad_norm": 0.7697158876317151,
+ "learning_rate": 3.629941698158923e-07,
+ "loss": 0.3135251998901367,
+ "num_input_tokens_seen": 1140503592,
+ "step": 26080,
+ "train_runtime": 201837.3895,
+ "train_tokens_per_second": 5650.606
+ },
+ {
+ "epoch": 1.7563291139240507,
+ "grad_norm": 0.6797973936849687,
+ "learning_rate": 3.6200575096301207e-07,
+ "loss": 0.2774399757385254,
+ "num_input_tokens_seen": 1140721216,
+ "step": 26085,
+ "train_runtime": 201873.4172,
+ "train_tokens_per_second": 5650.676
+ },
+ {
+ "epoch": 1.7566657689200107,
+ "grad_norm": 0.8393124553061458,
+ "learning_rate": 3.61018629110434e-07,
+ "loss": 0.32251529693603515,
+ "num_input_tokens_seen": 1140946680,
+ "step": 26090,
+ "train_runtime": 201907.6529,
+ "train_tokens_per_second": 5650.834
+ },
+ {
+ "epoch": 1.757002423915971,
+ "grad_norm": 0.7056858714134263,
+ "learning_rate": 3.6003280453420386e-07,
+ "loss": 0.2859219551086426,
+ "num_input_tokens_seen": 1141168176,
+ "step": 26095,
+ "train_runtime": 201940.7733,
+ "train_tokens_per_second": 5651.004
+ },
+ {
+ "epoch": 1.7573390789119312,
+ "grad_norm": 0.7740995437255472,
+ "learning_rate": 3.59048277510004e-07,
+ "loss": 0.28550260066986083,
+ "num_input_tokens_seen": 1141389904,
+ "step": 26100,
+ "train_runtime": 201981.2605,
+ "train_tokens_per_second": 5650.969
+ },
+ {
+ "epoch": 1.7576757339078912,
+ "grad_norm": 0.7658978202078977,
+ "learning_rate": 3.5806504831315423e-07,
+ "loss": 0.28779993057250974,
+ "num_input_tokens_seen": 1141608056,
+ "step": 26105,
+ "train_runtime": 202020.9745,
+ "train_tokens_per_second": 5650.938
+ },
+ {
+ "epoch": 1.7580123889038513,
+ "grad_norm": 0.7409514848212948,
+ "learning_rate": 3.570831172186112e-07,
+ "loss": 0.3048838138580322,
+ "num_input_tokens_seen": 1141832104,
+ "step": 26110,
+ "train_runtime": 202059.9812,
+ "train_tokens_per_second": 5650.956
+ },
+ {
+ "epoch": 1.7583490438998115,
+ "grad_norm": 0.7982423185812808,
+ "learning_rate": 3.5610248450097084e-07,
+ "loss": 0.30502030849456785,
+ "num_input_tokens_seen": 1142049928,
+ "step": 26115,
+ "train_runtime": 202100.022,
+ "train_tokens_per_second": 5650.914
+ },
+ {
+ "epoch": 1.7586856988957718,
+ "grad_norm": 0.7105646801738283,
+ "learning_rate": 3.551231504344604e-07,
+ "loss": 0.2915775775909424,
+ "num_input_tokens_seen": 1142270008,
+ "step": 26120,
+ "train_runtime": 202138.5884,
+ "train_tokens_per_second": 5650.925
+ },
+ {
+ "epoch": 1.7590223538917318,
+ "grad_norm": 0.6784969862052538,
+ "learning_rate": 3.5414511529295036e-07,
+ "loss": 0.30109832286834715,
+ "num_input_tokens_seen": 1142484296,
+ "step": 26125,
+ "train_runtime": 202171.0989,
+ "train_tokens_per_second": 5651.076
+ },
+ {
+ "epoch": 1.7593590088876918,
+ "grad_norm": 0.8313599079016897,
+ "learning_rate": 3.5316837934994465e-07,
+ "loss": 0.3126358509063721,
+ "num_input_tokens_seen": 1142703184,
+ "step": 26130,
+ "train_runtime": 202211.0287,
+ "train_tokens_per_second": 5651.043
+ },
+ {
+ "epoch": 1.759695663883652,
+ "grad_norm": 0.7462437378743789,
+ "learning_rate": 3.521929428785836e-07,
+ "loss": 0.3012665271759033,
+ "num_input_tokens_seen": 1142932000,
+ "step": 26135,
+ "train_runtime": 202249.327,
+ "train_tokens_per_second": 5651.104
+ },
+ {
+ "epoch": 1.7600323188796123,
+ "grad_norm": 0.7707749450724266,
+ "learning_rate": 3.5121880615164584e-07,
+ "loss": 0.3177218198776245,
+ "num_input_tokens_seen": 1143147936,
+ "step": 26140,
+ "train_runtime": 202288.9562,
+ "train_tokens_per_second": 5651.064
+ },
+ {
+ "epoch": 1.7603689738755723,
+ "grad_norm": 0.7201463000961772,
+ "learning_rate": 3.5024596944154445e-07,
+ "loss": 0.30140066146850586,
+ "num_input_tokens_seen": 1143365592,
+ "step": 26145,
+ "train_runtime": 202327.958,
+ "train_tokens_per_second": 5651.051
+ },
+ {
+ "epoch": 1.7607056288715324,
+ "grad_norm": 0.7940935130035773,
+ "learning_rate": 3.4927443302033127e-07,
+ "loss": 0.2727382659912109,
+ "num_input_tokens_seen": 1143591488,
+ "step": 26150,
+ "train_runtime": 202375.4812,
+ "train_tokens_per_second": 5650.84
+ },
+ {
+ "epoch": 1.7610422838674926,
+ "grad_norm": 0.7992454946312759,
+ "learning_rate": 3.4830419715969233e-07,
+ "loss": 0.2696976900100708,
+ "num_input_tokens_seen": 1143804104,
+ "step": 26155,
+ "train_runtime": 202410.9519,
+ "train_tokens_per_second": 5650.9
+ },
+ {
+ "epoch": 1.7613789388634529,
+ "grad_norm": 0.8375508909384706,
+ "learning_rate": 3.4733526213095114e-07,
+ "loss": 0.30320181846618655,
+ "num_input_tokens_seen": 1144024536,
+ "step": 26160,
+ "train_runtime": 202453.0452,
+ "train_tokens_per_second": 5650.814
+ },
+ {
+ "epoch": 1.761715593859413,
+ "grad_norm": 0.8560064029859269,
+ "learning_rate": 3.4636762820506876e-07,
+ "loss": 0.3035079002380371,
+ "num_input_tokens_seen": 1144244512,
+ "step": 26165,
+ "train_runtime": 202492.5015,
+ "train_tokens_per_second": 5650.799
+ },
+ {
+ "epoch": 1.762052248855373,
+ "grad_norm": 0.8794486904268961,
+ "learning_rate": 3.4540129565263867e-07,
+ "loss": 0.30885040760040283,
+ "num_input_tokens_seen": 1144457424,
+ "step": 26170,
+ "train_runtime": 202528.6337,
+ "train_tokens_per_second": 5650.843
+ },
+ {
+ "epoch": 1.7623889038513332,
+ "grad_norm": 0.7747667521193755,
+ "learning_rate": 3.444362647438942e-07,
+ "loss": 0.30580987930297854,
+ "num_input_tokens_seen": 1144685136,
+ "step": 26175,
+ "train_runtime": 202566.2912,
+ "train_tokens_per_second": 5650.916
+ },
+ {
+ "epoch": 1.7627255588472934,
+ "grad_norm": 0.749987526461497,
+ "learning_rate": 3.434725357487029e-07,
+ "loss": 0.29648699760437014,
+ "num_input_tokens_seen": 1144912488,
+ "step": 26180,
+ "train_runtime": 202608.8203,
+ "train_tokens_per_second": 5650.852
+ },
+ {
+ "epoch": 1.7630622138432535,
+ "grad_norm": 0.7973029748008016,
+ "learning_rate": 3.4251010893656844e-07,
+ "loss": 0.3241448402404785,
+ "num_input_tokens_seen": 1145132240,
+ "step": 26185,
+ "train_runtime": 202646.9809,
+ "train_tokens_per_second": 5650.872
+ },
+ {
+ "epoch": 1.7633988688392135,
+ "grad_norm": 0.8079881889044294,
+ "learning_rate": 3.4154898457663065e-07,
+ "loss": 0.2964663505554199,
+ "num_input_tokens_seen": 1145339072,
+ "step": 26190,
+ "train_runtime": 202684.9934,
+ "train_tokens_per_second": 5650.833
+ },
+ {
+ "epoch": 1.7637355238351737,
+ "grad_norm": 0.74358995342413,
+ "learning_rate": 3.40589162937664e-07,
+ "loss": 0.2934283256530762,
+ "num_input_tokens_seen": 1145537280,
+ "step": 26195,
+ "train_runtime": 202723.2034,
+ "train_tokens_per_second": 5650.746
+ },
+ {
+ "epoch": 1.764072178831134,
+ "grad_norm": 0.7784930364611832,
+ "learning_rate": 3.3963064428808034e-07,
+ "loss": 0.2671190738677979,
+ "num_input_tokens_seen": 1145744432,
+ "step": 26200,
+ "train_runtime": 202765.8103,
+ "train_tokens_per_second": 5650.58
+ },
+ {
+ "epoch": 1.764408833827094,
+ "grad_norm": 0.6964223743604306,
+ "learning_rate": 3.386734288959265e-07,
+ "loss": 0.29965527057647706,
+ "num_input_tokens_seen": 1145941000,
+ "step": 26205,
+ "train_runtime": 202807.8719,
+ "train_tokens_per_second": 5650.377
+ },
+ {
+ "epoch": 1.764745488823054,
+ "grad_norm": 0.7356574014692101,
+ "learning_rate": 3.377175170288838e-07,
+ "loss": 0.3009481430053711,
+ "num_input_tokens_seen": 1146164576,
+ "step": 26210,
+ "train_runtime": 202841.3847,
+ "train_tokens_per_second": 5650.546
+ },
+ {
+ "epoch": 1.7650821438190143,
+ "grad_norm": 0.7941269061669325,
+ "learning_rate": 3.367629089542712e-07,
+ "loss": 0.2912009239196777,
+ "num_input_tokens_seen": 1146397144,
+ "step": 26215,
+ "train_runtime": 202889.1986,
+ "train_tokens_per_second": 5650.361
+ },
+ {
+ "epoch": 1.7654187988149745,
+ "grad_norm": 0.7636350648934562,
+ "learning_rate": 3.3580960493904024e-07,
+ "loss": 0.3059426546096802,
+ "num_input_tokens_seen": 1146616808,
+ "step": 26220,
+ "train_runtime": 202930.3093,
+ "train_tokens_per_second": 5650.298
+ },
+ {
+ "epoch": 1.7657554538109346,
+ "grad_norm": 0.7999400483263438,
+ "learning_rate": 3.3485760524978093e-07,
+ "loss": 0.31472842693328856,
+ "num_input_tokens_seen": 1146822992,
+ "step": 26225,
+ "train_runtime": 202962.5617,
+ "train_tokens_per_second": 5650.416
+ },
+ {
+ "epoch": 1.7660921088068946,
+ "grad_norm": 0.826503171384712,
+ "learning_rate": 3.3390691015271646e-07,
+ "loss": 0.29501938819885254,
+ "num_input_tokens_seen": 1147041408,
+ "step": 26230,
+ "train_runtime": 202998.4752,
+ "train_tokens_per_second": 5650.493
+ },
+ {
+ "epoch": 1.7664287638028549,
+ "grad_norm": 0.7752263216759548,
+ "learning_rate": 3.329575199137053e-07,
+ "loss": 0.2826894521713257,
+ "num_input_tokens_seen": 1147263888,
+ "step": 26235,
+ "train_runtime": 203037.2378,
+ "train_tokens_per_second": 5650.51
+ },
+ {
+ "epoch": 1.766765418798815,
+ "grad_norm": 0.8585013195673774,
+ "learning_rate": 3.3200943479824113e-07,
+ "loss": 0.327055025100708,
+ "num_input_tokens_seen": 1147480536,
+ "step": 26240,
+ "train_runtime": 203079.6569,
+ "train_tokens_per_second": 5650.396
+ },
+ {
+ "epoch": 1.7671020737947751,
+ "grad_norm": 0.6517474645162997,
+ "learning_rate": 3.3106265507145406e-07,
+ "loss": 0.2585421562194824,
+ "num_input_tokens_seen": 1147695160,
+ "step": 26245,
+ "train_runtime": 203115.9008,
+ "train_tokens_per_second": 5650.445
+ },
+ {
+ "epoch": 1.7674387287907352,
+ "grad_norm": 0.7376728837137707,
+ "learning_rate": 3.3011718099810687e-07,
+ "loss": 0.28537108898162844,
+ "num_input_tokens_seen": 1147904984,
+ "step": 26250,
+ "train_runtime": 203155.1468,
+ "train_tokens_per_second": 5650.386
+ },
+ {
+ "epoch": 1.7677753837866954,
+ "grad_norm": 0.7607253054674292,
+ "learning_rate": 3.2917301284259904e-07,
+ "loss": 0.3015407085418701,
+ "num_input_tokens_seen": 1148133136,
+ "step": 26255,
+ "train_runtime": 203195.8171,
+ "train_tokens_per_second": 5650.378
+ },
+ {
+ "epoch": 1.7681120387826557,
+ "grad_norm": 0.7368433932513038,
+ "learning_rate": 3.2823015086896335e-07,
+ "loss": 0.31117079257965086,
+ "num_input_tokens_seen": 1148351808,
+ "step": 26260,
+ "train_runtime": 203231.7124,
+ "train_tokens_per_second": 5650.456
+ },
+ {
+ "epoch": 1.7684486937786157,
+ "grad_norm": 0.7455974069716451,
+ "learning_rate": 3.2728859534087e-07,
+ "loss": 0.28980340957641604,
+ "num_input_tokens_seen": 1148578048,
+ "step": 26265,
+ "train_runtime": 203270.0794,
+ "train_tokens_per_second": 5650.502
+ },
+ {
+ "epoch": 1.7687853487745757,
+ "grad_norm": 0.726415832314535,
+ "learning_rate": 3.263483465216194e-07,
+ "loss": 0.30661208629608155,
+ "num_input_tokens_seen": 1148808400,
+ "step": 26270,
+ "train_runtime": 203304.8486,
+ "train_tokens_per_second": 5650.669
+ },
+ {
+ "epoch": 1.769122003770536,
+ "grad_norm": 0.7198274225186322,
+ "learning_rate": 3.2540940467415126e-07,
+ "loss": 0.29364948272705077,
+ "num_input_tokens_seen": 1149026320,
+ "step": 26275,
+ "train_runtime": 203344.2573,
+ "train_tokens_per_second": 5650.646
+ },
+ {
+ "epoch": 1.7694586587664962,
+ "grad_norm": 0.675379666160097,
+ "learning_rate": 3.2447177006103726e-07,
+ "loss": 0.29972095489501954,
+ "num_input_tokens_seen": 1149241504,
+ "step": 26280,
+ "train_runtime": 203385.3582,
+ "train_tokens_per_second": 5650.562
+ },
+ {
+ "epoch": 1.7697953137624562,
+ "grad_norm": 0.7180951826553738,
+ "learning_rate": 3.235354429444831e-07,
+ "loss": 0.29863905906677246,
+ "num_input_tokens_seen": 1149442960,
+ "step": 26285,
+ "train_runtime": 203420.2428,
+ "train_tokens_per_second": 5650.583
+ },
+ {
+ "epoch": 1.7701319687584163,
+ "grad_norm": 0.7550877294810541,
+ "learning_rate": 3.226004235863306e-07,
+ "loss": 0.3004556655883789,
+ "num_input_tokens_seen": 1149664768,
+ "step": 26290,
+ "train_runtime": 203464.2499,
+ "train_tokens_per_second": 5650.451
+ },
+ {
+ "epoch": 1.7704686237543765,
+ "grad_norm": 0.7735355193352721,
+ "learning_rate": 3.2166671224805437e-07,
+ "loss": 0.2912909984588623,
+ "num_input_tokens_seen": 1149894072,
+ "step": 26295,
+ "train_runtime": 203504.8957,
+ "train_tokens_per_second": 5650.449
+ },
+ {
+ "epoch": 1.7708052787503368,
+ "grad_norm": 0.7860723514589903,
+ "learning_rate": 3.207343091907633e-07,
+ "loss": 0.28990769386291504,
+ "num_input_tokens_seen": 1150100408,
+ "step": 26300,
+ "train_runtime": 203547.7921,
+ "train_tokens_per_second": 5650.272
+ },
+ {
+ "epoch": 1.7711419337462968,
+ "grad_norm": 0.8144219098196478,
+ "learning_rate": 3.1980321467520226e-07,
+ "loss": 0.30245256423950195,
+ "num_input_tokens_seen": 1150300712,
+ "step": 26305,
+ "train_runtime": 203587.3111,
+ "train_tokens_per_second": 5650.159
+ },
+ {
+ "epoch": 1.7714785887422568,
+ "grad_norm": 0.8081994567096525,
+ "learning_rate": 3.1887342896174735e-07,
+ "loss": 0.31411175727844237,
+ "num_input_tokens_seen": 1150501592,
+ "step": 26310,
+ "train_runtime": 203630.5459,
+ "train_tokens_per_second": 5649.946
+ },
+ {
+ "epoch": 1.771815243738217,
+ "grad_norm": 0.6657629683425634,
+ "learning_rate": 3.1794495231041276e-07,
+ "loss": 0.28406760692596433,
+ "num_input_tokens_seen": 1150727232,
+ "step": 26315,
+ "train_runtime": 203671.5283,
+ "train_tokens_per_second": 5649.917
+ },
+ {
+ "epoch": 1.7721518987341773,
+ "grad_norm": 0.7568239836133204,
+ "learning_rate": 3.1701778498084035e-07,
+ "loss": 0.3084291458129883,
+ "num_input_tokens_seen": 1150932320,
+ "step": 26320,
+ "train_runtime": 203712.8008,
+ "train_tokens_per_second": 5649.779
+ },
+ {
+ "epoch": 1.7724885537301374,
+ "grad_norm": 0.7723629219802286,
+ "learning_rate": 3.160919272323121e-07,
+ "loss": 0.3075510263442993,
+ "num_input_tokens_seen": 1151162984,
+ "step": 26325,
+ "train_runtime": 203750.2121,
+ "train_tokens_per_second": 5649.874
+ },
+ {
+ "epoch": 1.7728252087260974,
+ "grad_norm": 0.7180286793509041,
+ "learning_rate": 3.1516737932374087e-07,
+ "loss": 0.30257492065429686,
+ "num_input_tokens_seen": 1151371272,
+ "step": 26330,
+ "train_runtime": 203784.7687,
+ "train_tokens_per_second": 5649.938
+ },
+ {
+ "epoch": 1.7731618637220576,
+ "grad_norm": 0.7988698082761704,
+ "learning_rate": 3.142441415136727e-07,
+ "loss": 0.29022831916809083,
+ "num_input_tokens_seen": 1151601408,
+ "step": 26335,
+ "train_runtime": 203817.7911,
+ "train_tokens_per_second": 5650.152
+ },
+ {
+ "epoch": 1.773498518718018,
+ "grad_norm": 0.7403905711500696,
+ "learning_rate": 3.133222140602893e-07,
+ "loss": 0.2794740438461304,
+ "num_input_tokens_seen": 1151807024,
+ "step": 26340,
+ "train_runtime": 203851.848,
+ "train_tokens_per_second": 5650.216
+ },
+ {
+ "epoch": 1.773835173713978,
+ "grad_norm": 0.7358244297636568,
+ "learning_rate": 3.124015972214034e-07,
+ "loss": 0.30714993476867675,
+ "num_input_tokens_seen": 1152030904,
+ "step": 26345,
+ "train_runtime": 203880.8052,
+ "train_tokens_per_second": 5650.512
+ },
+ {
+ "epoch": 1.774171828709938,
+ "grad_norm": 0.7987243174003847,
+ "learning_rate": 3.114822912544635e-07,
+ "loss": 0.31725311279296875,
+ "num_input_tokens_seen": 1152240824,
+ "step": 26350,
+ "train_runtime": 203921.8118,
+ "train_tokens_per_second": 5650.405
+ },
+ {
+ "epoch": 1.7745084837058982,
+ "grad_norm": 0.7884531597674449,
+ "learning_rate": 3.1056429641655007e-07,
+ "loss": 0.2949615716934204,
+ "num_input_tokens_seen": 1152447328,
+ "step": 26355,
+ "train_runtime": 203963.8804,
+ "train_tokens_per_second": 5650.252
+ },
+ {
+ "epoch": 1.7748451387018584,
+ "grad_norm": 0.8631624259760383,
+ "learning_rate": 3.0964761296437665e-07,
+ "loss": 0.2921284198760986,
+ "num_input_tokens_seen": 1152671024,
+ "step": 26360,
+ "train_runtime": 204005.1311,
+ "train_tokens_per_second": 5650.206
+ },
+ {
+ "epoch": 1.7751817936978185,
+ "grad_norm": 0.7806230963793127,
+ "learning_rate": 3.087322411542937e-07,
+ "loss": 0.3069296836853027,
+ "num_input_tokens_seen": 1152898216,
+ "step": 26365,
+ "train_runtime": 204043.7868,
+ "train_tokens_per_second": 5650.249
+ },
+ {
+ "epoch": 1.7755184486937785,
+ "grad_norm": 0.8045496935454175,
+ "learning_rate": 3.0781818124227867e-07,
+ "loss": 0.27917664051055907,
+ "num_input_tokens_seen": 1153109064,
+ "step": 26370,
+ "train_runtime": 204083.2918,
+ "train_tokens_per_second": 5650.188
+ },
+ {
+ "epoch": 1.7758551036897388,
+ "grad_norm": 0.7628962650971699,
+ "learning_rate": 3.0690543348394776e-07,
+ "loss": 0.2973165988922119,
+ "num_input_tokens_seen": 1153332104,
+ "step": 26375,
+ "train_runtime": 204124.6151,
+ "train_tokens_per_second": 5650.137
+ },
+ {
+ "epoch": 1.776191758685699,
+ "grad_norm": 0.735769438564152,
+ "learning_rate": 3.059939981345467e-07,
+ "loss": 0.2835052490234375,
+ "num_input_tokens_seen": 1153547680,
+ "step": 26380,
+ "train_runtime": 204153.8046,
+ "train_tokens_per_second": 5650.385
+ },
+ {
+ "epoch": 1.776528413681659,
+ "grad_norm": 0.7910223938595813,
+ "learning_rate": 3.050838754489566e-07,
+ "loss": 0.28968124389648436,
+ "num_input_tokens_seen": 1153764144,
+ "step": 26385,
+ "train_runtime": 204197.8486,
+ "train_tokens_per_second": 5650.227
+ },
+ {
+ "epoch": 1.776865068677619,
+ "grad_norm": 0.7189107694256088,
+ "learning_rate": 3.041750656816894e-07,
+ "loss": 0.29620354175567626,
+ "num_input_tokens_seen": 1153982600,
+ "step": 26390,
+ "train_runtime": 204235.7042,
+ "train_tokens_per_second": 5650.249
+ },
+ {
+ "epoch": 1.7772017236735793,
+ "grad_norm": 0.8335040707060406,
+ "learning_rate": 3.0326756908689135e-07,
+ "loss": 0.3176565647125244,
+ "num_input_tokens_seen": 1154204320,
+ "step": 26395,
+ "train_runtime": 204274.2047,
+ "train_tokens_per_second": 5650.27
+ },
+ {
+ "epoch": 1.7775383786695396,
+ "grad_norm": 0.687809411831832,
+ "learning_rate": 3.0236138591833995e-07,
+ "loss": 0.2815841197967529,
+ "num_input_tokens_seen": 1154426800,
+ "step": 26400,
+ "train_runtime": 204312.8632,
+ "train_tokens_per_second": 5650.289
+ },
+ {
+ "epoch": 1.7778750336654996,
+ "grad_norm": 0.7809149749160714,
+ "learning_rate": 3.0145651642944863e-07,
+ "loss": 0.31723904609680176,
+ "num_input_tokens_seen": 1154656600,
+ "step": 26405,
+ "train_runtime": 204357.2069,
+ "train_tokens_per_second": 5650.188
+ },
+ {
+ "epoch": 1.7782116886614596,
+ "grad_norm": 0.739025964030064,
+ "learning_rate": 3.005529608732588e-07,
+ "loss": 0.26880898475646975,
+ "num_input_tokens_seen": 1154869552,
+ "step": 26410,
+ "train_runtime": 204394.2963,
+ "train_tokens_per_second": 5650.204
+ },
+ {
+ "epoch": 1.7785483436574199,
+ "grad_norm": 0.7646452369868884,
+ "learning_rate": 2.996507195024495e-07,
+ "loss": 0.322553014755249,
+ "num_input_tokens_seen": 1155101520,
+ "step": 26415,
+ "train_runtime": 204435.1594,
+ "train_tokens_per_second": 5650.21
+ },
+ {
+ "epoch": 1.7788849986533801,
+ "grad_norm": 0.8474473108460225,
+ "learning_rate": 2.9874979256932614e-07,
+ "loss": 0.2975711107254028,
+ "num_input_tokens_seen": 1155322552,
+ "step": 26420,
+ "train_runtime": 204473.0585,
+ "train_tokens_per_second": 5650.243
+ },
+ {
+ "epoch": 1.7792216536493402,
+ "grad_norm": 0.8062221329371094,
+ "learning_rate": 2.9785018032583325e-07,
+ "loss": 0.3099265813827515,
+ "num_input_tokens_seen": 1155544168,
+ "step": 26425,
+ "train_runtime": 204516.1566,
+ "train_tokens_per_second": 5650.136
+ },
+ {
+ "epoch": 1.7795583086453002,
+ "grad_norm": 0.8324573418937758,
+ "learning_rate": 2.969518830235435e-07,
+ "loss": 0.298630952835083,
+ "num_input_tokens_seen": 1155762856,
+ "step": 26430,
+ "train_runtime": 204549.2337,
+ "train_tokens_per_second": 5650.292
+ },
+ {
+ "epoch": 1.7798949636412604,
+ "grad_norm": 0.7975157170801572,
+ "learning_rate": 2.960549009136626e-07,
+ "loss": 0.30402772426605223,
+ "num_input_tokens_seen": 1155983360,
+ "step": 26435,
+ "train_runtime": 204590.4666,
+ "train_tokens_per_second": 5650.231
+ },
+ {
+ "epoch": 1.7802316186372207,
+ "grad_norm": 0.8344419171870143,
+ "learning_rate": 2.951592342470289e-07,
+ "loss": 0.33509268760681155,
+ "num_input_tokens_seen": 1156211056,
+ "step": 26440,
+ "train_runtime": 204625.1719,
+ "train_tokens_per_second": 5650.385
+ },
+ {
+ "epoch": 1.7805682736331807,
+ "grad_norm": 0.7399401765355089,
+ "learning_rate": 2.942648832741124e-07,
+ "loss": 0.28286046981811525,
+ "num_input_tokens_seen": 1156416184,
+ "step": 26445,
+ "train_runtime": 204663.2394,
+ "train_tokens_per_second": 5650.337
+ },
+ {
+ "epoch": 1.7809049286291407,
+ "grad_norm": 0.6734864094384185,
+ "learning_rate": 2.933718482450154e-07,
+ "loss": 0.2894281387329102,
+ "num_input_tokens_seen": 1156635048,
+ "step": 26450,
+ "train_runtime": 204698.0899,
+ "train_tokens_per_second": 5650.444
+ },
+ {
+ "epoch": 1.781241583625101,
+ "grad_norm": 0.8706400424083963,
+ "learning_rate": 2.9248012940947423e-07,
+ "loss": 0.3163520097732544,
+ "num_input_tokens_seen": 1156864984,
+ "step": 26455,
+ "train_runtime": 204729.4569,
+ "train_tokens_per_second": 5650.701
+ },
+ {
+ "epoch": 1.7815782386210612,
+ "grad_norm": 0.7638164398502649,
+ "learning_rate": 2.915897270168522e-07,
+ "loss": 0.2976935625076294,
+ "num_input_tokens_seen": 1157071688,
+ "step": 26460,
+ "train_runtime": 204767.7005,
+ "train_tokens_per_second": 5650.655
+ },
+ {
+ "epoch": 1.7819148936170213,
+ "grad_norm": 0.8606236851862529,
+ "learning_rate": 2.9070064131615004e-07,
+ "loss": 0.30093798637390134,
+ "num_input_tokens_seen": 1157281584,
+ "step": 26465,
+ "train_runtime": 204809.9681,
+ "train_tokens_per_second": 5650.514
+ },
+ {
+ "epoch": 1.7822515486129813,
+ "grad_norm": 0.8072069252667485,
+ "learning_rate": 2.8981287255599567e-07,
+ "loss": 0.291579532623291,
+ "num_input_tokens_seen": 1157474720,
+ "step": 26470,
+ "train_runtime": 204843.9736,
+ "train_tokens_per_second": 5650.519
+ },
+ {
+ "epoch": 1.7825882036089415,
+ "grad_norm": 0.7787592014629802,
+ "learning_rate": 2.8892642098465205e-07,
+ "loss": 0.2964937210083008,
+ "num_input_tokens_seen": 1157694912,
+ "step": 26475,
+ "train_runtime": 204887.0752,
+ "train_tokens_per_second": 5650.405
+ },
+ {
+ "epoch": 1.7829248586049018,
+ "grad_norm": 0.741114575623905,
+ "learning_rate": 2.8804128685001263e-07,
+ "loss": 0.29356560707092283,
+ "num_input_tokens_seen": 1157923424,
+ "step": 26480,
+ "train_runtime": 204930.4827,
+ "train_tokens_per_second": 5650.323
+ },
+ {
+ "epoch": 1.7832615136008618,
+ "grad_norm": 0.8320550507276643,
+ "learning_rate": 2.8715747039960216e-07,
+ "loss": 0.2721066474914551,
+ "num_input_tokens_seen": 1158134336,
+ "step": 26485,
+ "train_runtime": 204962.95,
+ "train_tokens_per_second": 5650.457
+ },
+ {
+ "epoch": 1.7835981685968219,
+ "grad_norm": 0.8218511157843951,
+ "learning_rate": 2.8627497188057684e-07,
+ "loss": 0.30223267078399657,
+ "num_input_tokens_seen": 1158358632,
+ "step": 26490,
+ "train_runtime": 205002.9913,
+ "train_tokens_per_second": 5650.447
+ },
+ {
+ "epoch": 1.783934823592782,
+ "grad_norm": 0.7711493227331807,
+ "learning_rate": 2.853937915397248e-07,
+ "loss": 0.32183716297149656,
+ "num_input_tokens_seen": 1158575032,
+ "step": 26495,
+ "train_runtime": 205041.1723,
+ "train_tokens_per_second": 5650.451
+ },
+ {
+ "epoch": 1.7842714785887424,
+ "grad_norm": 0.7845009644246281,
+ "learning_rate": 2.845139296234639e-07,
+ "loss": 0.28379087448120116,
+ "num_input_tokens_seen": 1158803496,
+ "step": 26500,
+ "train_runtime": 205077.698,
+ "train_tokens_per_second": 5650.558
+ },
+ {
+ "epoch": 1.7846081335847024,
+ "grad_norm": 0.6932426732560467,
+ "learning_rate": 2.836353863778479e-07,
+ "loss": 0.29067995548248293,
+ "num_input_tokens_seen": 1159023072,
+ "step": 26505,
+ "train_runtime": 205114.9773,
+ "train_tokens_per_second": 5650.602
+ },
+ {
+ "epoch": 1.7849447885806624,
+ "grad_norm": 0.66733187111737,
+ "learning_rate": 2.8275816204855534e-07,
+ "loss": 0.3040937900543213,
+ "num_input_tokens_seen": 1159239184,
+ "step": 26510,
+ "train_runtime": 205163.8168,
+ "train_tokens_per_second": 5650.31
+ },
+ {
+ "epoch": 1.7852814435766227,
+ "grad_norm": 0.8094709155559127,
+ "learning_rate": 2.818822568809015e-07,
+ "loss": 0.31920318603515624,
+ "num_input_tokens_seen": 1159445280,
+ "step": 26515,
+ "train_runtime": 205199.8163,
+ "train_tokens_per_second": 5650.323
+ },
+ {
+ "epoch": 1.785618098572583,
+ "grad_norm": 0.7397186499607165,
+ "learning_rate": 2.810076711198284e-07,
+ "loss": 0.2857333183288574,
+ "num_input_tokens_seen": 1159658816,
+ "step": 26520,
+ "train_runtime": 205238.5515,
+ "train_tokens_per_second": 5650.297
+ },
+ {
+ "epoch": 1.785954753568543,
+ "grad_norm": 0.7573212272405271,
+ "learning_rate": 2.801344050099131e-07,
+ "loss": 0.3130965709686279,
+ "num_input_tokens_seen": 1159896520,
+ "step": 26525,
+ "train_runtime": 205275.2217,
+ "train_tokens_per_second": 5650.446
+ },
+ {
+ "epoch": 1.786291408564503,
+ "grad_norm": 0.7444021852294227,
+ "learning_rate": 2.7926245879536076e-07,
+ "loss": 0.3018080711364746,
+ "num_input_tokens_seen": 1160116400,
+ "step": 26530,
+ "train_runtime": 205318.3954,
+ "train_tokens_per_second": 5650.329
+ },
+ {
+ "epoch": 1.7866280635604632,
+ "grad_norm": 0.7159851006520612,
+ "learning_rate": 2.7839183272000857e-07,
+ "loss": 0.28679165840148924,
+ "num_input_tokens_seen": 1160350120,
+ "step": 26535,
+ "train_runtime": 205353.8407,
+ "train_tokens_per_second": 5650.491
+ },
+ {
+ "epoch": 1.7869647185564235,
+ "grad_norm": 0.9695258135131855,
+ "learning_rate": 2.775225270273241e-07,
+ "loss": 0.30443391799926756,
+ "num_input_tokens_seen": 1160577328,
+ "step": 26540,
+ "train_runtime": 205388.8223,
+ "train_tokens_per_second": 5650.635
+ },
+ {
+ "epoch": 1.7873013735523835,
+ "grad_norm": 0.6888517858579661,
+ "learning_rate": 2.7665454196040665e-07,
+ "loss": 0.2736891508102417,
+ "num_input_tokens_seen": 1160803320,
+ "step": 26545,
+ "train_runtime": 205426.5857,
+ "train_tokens_per_second": 5650.697
+ },
+ {
+ "epoch": 1.7876380285483435,
+ "grad_norm": 0.8050232282526216,
+ "learning_rate": 2.7578787776198325e-07,
+ "loss": 0.31282496452331543,
+ "num_input_tokens_seen": 1161013768,
+ "step": 26550,
+ "train_runtime": 205464.4875,
+ "train_tokens_per_second": 5650.679
+ },
+ {
+ "epoch": 1.7879746835443038,
+ "grad_norm": 0.7051417852744916,
+ "learning_rate": 2.7492253467441775e-07,
+ "loss": 0.2743107318878174,
+ "num_input_tokens_seen": 1161234120,
+ "step": 26555,
+ "train_runtime": 205500.8557,
+ "train_tokens_per_second": 5650.751
+ },
+ {
+ "epoch": 1.788311338540264,
+ "grad_norm": 0.7282394673289367,
+ "learning_rate": 2.740585129396961e-07,
+ "loss": 0.30170412063598634,
+ "num_input_tokens_seen": 1161458560,
+ "step": 26560,
+ "train_runtime": 205545.0743,
+ "train_tokens_per_second": 5650.627
+ },
+ {
+ "epoch": 1.788647993536224,
+ "grad_norm": 0.7602917600626389,
+ "learning_rate": 2.731958127994433e-07,
+ "loss": 0.2769881248474121,
+ "num_input_tokens_seen": 1161674768,
+ "step": 26565,
+ "train_runtime": 205586.7314,
+ "train_tokens_per_second": 5650.534
+ },
+ {
+ "epoch": 1.788984648532184,
+ "grad_norm": 0.8213091294594603,
+ "learning_rate": 2.7233443449490695e-07,
+ "loss": 0.307672905921936,
+ "num_input_tokens_seen": 1161890360,
+ "step": 26570,
+ "train_runtime": 205630.1097,
+ "train_tokens_per_second": 5650.39
+ },
+ {
+ "epoch": 1.7893213035281443,
+ "grad_norm": 0.7931628273831316,
+ "learning_rate": 2.7147437826697153e-07,
+ "loss": 0.29656333923339845,
+ "num_input_tokens_seen": 1162107656,
+ "step": 26575,
+ "train_runtime": 205671.7811,
+ "train_tokens_per_second": 5650.302
+ },
+ {
+ "epoch": 1.7896579585241046,
+ "grad_norm": 1.1103372341248579,
+ "learning_rate": 2.706156443561481e-07,
+ "loss": 0.3016365528106689,
+ "num_input_tokens_seen": 1162338136,
+ "step": 26580,
+ "train_runtime": 205709.1075,
+ "train_tokens_per_second": 5650.397
+ },
+ {
+ "epoch": 1.7899946135200646,
+ "grad_norm": 0.7165987576366394,
+ "learning_rate": 2.6975823300257884e-07,
+ "loss": 0.29532361030578613,
+ "num_input_tokens_seen": 1162551688,
+ "step": 26585,
+ "train_runtime": 205742.7144,
+ "train_tokens_per_second": 5650.512
+ },
+ {
+ "epoch": 1.7903312685160246,
+ "grad_norm": 0.6849565677358419,
+ "learning_rate": 2.6890214444603604e-07,
+ "loss": 0.29342079162597656,
+ "num_input_tokens_seen": 1162773064,
+ "step": 26590,
+ "train_runtime": 205775.5364,
+ "train_tokens_per_second": 5650.687
+ },
+ {
+ "epoch": 1.790667923511985,
+ "grad_norm": 0.8242687417078883,
+ "learning_rate": 2.6804737892592193e-07,
+ "loss": 0.30418787002563474,
+ "num_input_tokens_seen": 1162987848,
+ "step": 26595,
+ "train_runtime": 205824.7685,
+ "train_tokens_per_second": 5650.379
+ },
+ {
+ "epoch": 1.7910045785079451,
+ "grad_norm": 0.8268380789860988,
+ "learning_rate": 2.671939366812687e-07,
+ "loss": 0.30772976875305175,
+ "num_input_tokens_seen": 1163202312,
+ "step": 26600,
+ "train_runtime": 205862.7926,
+ "train_tokens_per_second": 5650.377
+ },
+ {
+ "epoch": 1.7913412335039052,
+ "grad_norm": 0.767877157549208,
+ "learning_rate": 2.66341817950741e-07,
+ "loss": 0.29974985122680664,
+ "num_input_tokens_seen": 1163438080,
+ "step": 26605,
+ "train_runtime": 205898.0882,
+ "train_tokens_per_second": 5650.553
+ },
+ {
+ "epoch": 1.7916778884998652,
+ "grad_norm": 0.7009232602836326,
+ "learning_rate": 2.6549102297262764e-07,
+ "loss": 0.29571521282196045,
+ "num_input_tokens_seen": 1163660200,
+ "step": 26610,
+ "train_runtime": 205939.8237,
+ "train_tokens_per_second": 5650.487
+ },
+ {
+ "epoch": 1.7920145434958255,
+ "grad_norm": 0.871322486063614,
+ "learning_rate": 2.6464155198485387e-07,
+ "loss": 0.30173745155334475,
+ "num_input_tokens_seen": 1163887248,
+ "step": 26615,
+ "train_runtime": 205979.0591,
+ "train_tokens_per_second": 5650.512
+ },
+ {
+ "epoch": 1.7923511984917857,
+ "grad_norm": 0.803169204691401,
+ "learning_rate": 2.6379340522496853e-07,
+ "loss": 0.29706802368164065,
+ "num_input_tokens_seen": 1164110704,
+ "step": 26620,
+ "train_runtime": 206013.7386,
+ "train_tokens_per_second": 5650.646
+ },
+ {
+ "epoch": 1.7926878534877457,
+ "grad_norm": 0.7782352100729737,
+ "learning_rate": 2.6294658293015584e-07,
+ "loss": 0.3205864429473877,
+ "num_input_tokens_seen": 1164337200,
+ "step": 26625,
+ "train_runtime": 206050.8157,
+ "train_tokens_per_second": 5650.728
+ },
+ {
+ "epoch": 1.7930245084837058,
+ "grad_norm": 0.8588354428150048,
+ "learning_rate": 2.621010853372258e-07,
+ "loss": 0.31758651733398435,
+ "num_input_tokens_seen": 1164544968,
+ "step": 26630,
+ "train_runtime": 206091.8464,
+ "train_tokens_per_second": 5650.612
+ },
+ {
+ "epoch": 1.793361163479666,
+ "grad_norm": 0.8321453808487478,
+ "learning_rate": 2.612569126826187e-07,
+ "loss": 0.3288443088531494,
+ "num_input_tokens_seen": 1164763392,
+ "step": 26635,
+ "train_runtime": 206129.4651,
+ "train_tokens_per_second": 5650.64
+ },
+ {
+ "epoch": 1.7936978184756263,
+ "grad_norm": 0.7646490512058014,
+ "learning_rate": 2.6041406520240563e-07,
+ "loss": 0.3208475589752197,
+ "num_input_tokens_seen": 1164996760,
+ "step": 26640,
+ "train_runtime": 206162.845,
+ "train_tokens_per_second": 5650.857
+ },
+ {
+ "epoch": 1.7940344734715863,
+ "grad_norm": 0.9618962883016744,
+ "learning_rate": 2.5957254313228595e-07,
+ "loss": 0.3083490371704102,
+ "num_input_tokens_seen": 1165209184,
+ "step": 26645,
+ "train_runtime": 206203.1869,
+ "train_tokens_per_second": 5650.782
+ },
+ {
+ "epoch": 1.7943711284675463,
+ "grad_norm": 0.6865797037478947,
+ "learning_rate": 2.5873234670758753e-07,
+ "loss": 0.2793558597564697,
+ "num_input_tokens_seen": 1165437296,
+ "step": 26650,
+ "train_runtime": 206245.195,
+ "train_tokens_per_second": 5650.737
+ },
+ {
+ "epoch": 1.7947077834635066,
+ "grad_norm": 0.743249619367278,
+ "learning_rate": 2.578934761632712e-07,
+ "loss": 0.29536702632904055,
+ "num_input_tokens_seen": 1165668480,
+ "step": 26655,
+ "train_runtime": 206276.3477,
+ "train_tokens_per_second": 5651.004
+ },
+ {
+ "epoch": 1.7950444384594668,
+ "grad_norm": 0.775356540520281,
+ "learning_rate": 2.570559317339211e-07,
+ "loss": 0.30143041610717775,
+ "num_input_tokens_seen": 1165893536,
+ "step": 26660,
+ "train_runtime": 206318.3881,
+ "train_tokens_per_second": 5650.943
+ },
+ {
+ "epoch": 1.7953810934554268,
+ "grad_norm": 0.6994994001366991,
+ "learning_rate": 2.56219713653757e-07,
+ "loss": 0.2995075941085815,
+ "num_input_tokens_seen": 1166122240,
+ "step": 26665,
+ "train_runtime": 206352.7461,
+ "train_tokens_per_second": 5651.111
+ },
+ {
+ "epoch": 1.7957177484513869,
+ "grad_norm": 0.8569243475539816,
+ "learning_rate": 2.5538482215662187e-07,
+ "loss": 0.3236057758331299,
+ "num_input_tokens_seen": 1166338248,
+ "step": 26670,
+ "train_runtime": 206386.7169,
+ "train_tokens_per_second": 5651.227
+ },
+ {
+ "epoch": 1.7960544034473471,
+ "grad_norm": 0.766892273093451,
+ "learning_rate": 2.545512574759923e-07,
+ "loss": 0.2701665163040161,
+ "num_input_tokens_seen": 1166555592,
+ "step": 26675,
+ "train_runtime": 206433.7742,
+ "train_tokens_per_second": 5650.992
+ },
+ {
+ "epoch": 1.7963910584433074,
+ "grad_norm": 0.7756837374057447,
+ "learning_rate": 2.537190198449713e-07,
+ "loss": 0.31007416248321534,
+ "num_input_tokens_seen": 1166772216,
+ "step": 26680,
+ "train_runtime": 206467.3628,
+ "train_tokens_per_second": 5651.122
+ },
+ {
+ "epoch": 1.7967277134392674,
+ "grad_norm": 0.7601571953732823,
+ "learning_rate": 2.5288810949629207e-07,
+ "loss": 0.3006875991821289,
+ "num_input_tokens_seen": 1166996856,
+ "step": 26685,
+ "train_runtime": 206509.338,
+ "train_tokens_per_second": 5651.061
+ },
+ {
+ "epoch": 1.7970643684352274,
+ "grad_norm": 0.7713983757495535,
+ "learning_rate": 2.5205852666231544e-07,
+ "loss": 0.2778863668441772,
+ "num_input_tokens_seen": 1167208832,
+ "step": 26690,
+ "train_runtime": 206552.5087,
+ "train_tokens_per_second": 5650.906
+ },
+ {
+ "epoch": 1.7974010234311877,
+ "grad_norm": 0.8187900042183005,
+ "learning_rate": 2.5123027157503144e-07,
+ "loss": 0.28261337280273435,
+ "num_input_tokens_seen": 1167408464,
+ "step": 26695,
+ "train_runtime": 206590.339,
+ "train_tokens_per_second": 5650.838
+ },
+ {
+ "epoch": 1.797737678427148,
+ "grad_norm": 0.7565722539390317,
+ "learning_rate": 2.504033444660592e-07,
+ "loss": 0.3025941848754883,
+ "num_input_tokens_seen": 1167643608,
+ "step": 26700,
+ "train_runtime": 206627.1922,
+ "train_tokens_per_second": 5650.968
+ },
+ {
+ "epoch": 1.798074333423108,
+ "grad_norm": 0.6833026230882182,
+ "learning_rate": 2.4957774556664706e-07,
+ "loss": 0.28271756172180174,
+ "num_input_tokens_seen": 1167868528,
+ "step": 26705,
+ "train_runtime": 206670.4456,
+ "train_tokens_per_second": 5650.873
+ },
+ {
+ "epoch": 1.798410988419068,
+ "grad_norm": 0.8308124821063052,
+ "learning_rate": 2.487534751076692e-07,
+ "loss": 0.29103024005889894,
+ "num_input_tokens_seen": 1168090128,
+ "step": 26710,
+ "train_runtime": 206715.3981,
+ "train_tokens_per_second": 5650.717
+ },
+ {
+ "epoch": 1.7987476434150282,
+ "grad_norm": 0.7543688224419066,
+ "learning_rate": 2.47930533319633e-07,
+ "loss": 0.2806957721710205,
+ "num_input_tokens_seen": 1168302824,
+ "step": 26715,
+ "train_runtime": 206757.8975,
+ "train_tokens_per_second": 5650.584
+ },
+ {
+ "epoch": 1.7990842984109885,
+ "grad_norm": 0.8079691544469572,
+ "learning_rate": 2.4710892043266823e-07,
+ "loss": 0.3256166219711304,
+ "num_input_tokens_seen": 1168526496,
+ "step": 26720,
+ "train_runtime": 206793.1384,
+ "train_tokens_per_second": 5650.702
+ },
+ {
+ "epoch": 1.7994209534069485,
+ "grad_norm": 0.7709686680304173,
+ "learning_rate": 2.462886366765388e-07,
+ "loss": 0.2893906354904175,
+ "num_input_tokens_seen": 1168747720,
+ "step": 26725,
+ "train_runtime": 206834.9728,
+ "train_tokens_per_second": 5650.629
+ },
+ {
+ "epoch": 1.7997576084029085,
+ "grad_norm": 0.861388639740031,
+ "learning_rate": 2.454696822806341e-07,
+ "loss": 0.3119854211807251,
+ "num_input_tokens_seen": 1168955928,
+ "step": 26730,
+ "train_runtime": 206876.7703,
+ "train_tokens_per_second": 5650.494
+ },
+ {
+ "epoch": 1.8000942633988688,
+ "grad_norm": 0.7333451652936307,
+ "learning_rate": 2.446520574739708e-07,
+ "loss": 0.2833410739898682,
+ "num_input_tokens_seen": 1169181640,
+ "step": 26735,
+ "train_runtime": 206910.3727,
+ "train_tokens_per_second": 5650.667
+ },
+ {
+ "epoch": 1.800430918394829,
+ "grad_norm": 0.8244240692612769,
+ "learning_rate": 2.438357624851967e-07,
+ "loss": 0.265745735168457,
+ "num_input_tokens_seen": 1169399504,
+ "step": 26740,
+ "train_runtime": 206944.8765,
+ "train_tokens_per_second": 5650.778
+ },
+ {
+ "epoch": 1.800767573390789,
+ "grad_norm": 0.7646663892198701,
+ "learning_rate": 2.430207975425847e-07,
+ "loss": 0.30634517669677735,
+ "num_input_tokens_seen": 1169627072,
+ "step": 26745,
+ "train_runtime": 206986.9735,
+ "train_tokens_per_second": 5650.728
+ },
+ {
+ "epoch": 1.801104228386749,
+ "grad_norm": 0.7977131343420332,
+ "learning_rate": 2.4220716287403744e-07,
+ "loss": 0.306353759765625,
+ "num_input_tokens_seen": 1169859232,
+ "step": 26750,
+ "train_runtime": 207024.3013,
+ "train_tokens_per_second": 5650.83
+ },
+ {
+ "epoch": 1.8014408833827094,
+ "grad_norm": 0.7225010612739647,
+ "learning_rate": 2.413948587070869e-07,
+ "loss": 0.2988481044769287,
+ "num_input_tokens_seen": 1170041408,
+ "step": 26755,
+ "train_runtime": 207063.8036,
+ "train_tokens_per_second": 5650.632
+ },
+ {
+ "epoch": 1.8017775383786696,
+ "grad_norm": 0.775593581305895,
+ "learning_rate": 2.405838852688891e-07,
+ "loss": 0.3097994327545166,
+ "num_input_tokens_seen": 1170267480,
+ "step": 26760,
+ "train_runtime": 207095.0651,
+ "train_tokens_per_second": 5650.871
+ },
+ {
+ "epoch": 1.8021141933746296,
+ "grad_norm": 0.8152686402481596,
+ "learning_rate": 2.397742427862315e-07,
+ "loss": 0.2666957378387451,
+ "num_input_tokens_seen": 1170495840,
+ "step": 26765,
+ "train_runtime": 207129.3856,
+ "train_tokens_per_second": 5651.037
+ },
+ {
+ "epoch": 1.8024508483705897,
+ "grad_norm": 0.7668475732368676,
+ "learning_rate": 2.3896593148552695e-07,
+ "loss": 0.3030317544937134,
+ "num_input_tokens_seen": 1170707968,
+ "step": 26770,
+ "train_runtime": 207164.983,
+ "train_tokens_per_second": 5651.09
+ },
+ {
+ "epoch": 1.80278750336655,
+ "grad_norm": 0.7115287772275755,
+ "learning_rate": 2.3815895159281844e-07,
+ "loss": 0.28127307891845704,
+ "num_input_tokens_seen": 1170932136,
+ "step": 26775,
+ "train_runtime": 207201.4624,
+ "train_tokens_per_second": 5651.177
+ },
+ {
+ "epoch": 1.8031241583625102,
+ "grad_norm": 0.7814253275211129,
+ "learning_rate": 2.3735330333377493e-07,
+ "loss": 0.2792118310928345,
+ "num_input_tokens_seen": 1171162408,
+ "step": 26780,
+ "train_runtime": 207238.2441,
+ "train_tokens_per_second": 5651.285
+ },
+ {
+ "epoch": 1.8034608133584702,
+ "grad_norm": 0.8200002096055589,
+ "learning_rate": 2.3654898693369232e-07,
+ "loss": 0.29738240242004393,
+ "num_input_tokens_seen": 1171377408,
+ "step": 26785,
+ "train_runtime": 207281.7346,
+ "train_tokens_per_second": 5651.137
+ },
+ {
+ "epoch": 1.8037974683544302,
+ "grad_norm": 0.7884729475059873,
+ "learning_rate": 2.3574600261749735e-07,
+ "loss": 0.30119688510894777,
+ "num_input_tokens_seen": 1171597216,
+ "step": 26790,
+ "train_runtime": 207321.2969,
+ "train_tokens_per_second": 5651.118
+ },
+ {
+ "epoch": 1.8041341233503905,
+ "grad_norm": 0.8577997537906676,
+ "learning_rate": 2.3494435060973875e-07,
+ "loss": 0.3168250322341919,
+ "num_input_tokens_seen": 1171822008,
+ "step": 26795,
+ "train_runtime": 207355.9199,
+ "train_tokens_per_second": 5651.259
+ },
+ {
+ "epoch": 1.8044707783463507,
+ "grad_norm": 0.9036192566903659,
+ "learning_rate": 2.3414403113459937e-07,
+ "loss": 0.30097527503967286,
+ "num_input_tokens_seen": 1172040272,
+ "step": 26800,
+ "train_runtime": 207392.3163,
+ "train_tokens_per_second": 5651.32
+ },
+ {
+ "epoch": 1.8048074333423108,
+ "grad_norm": 0.7489283622158472,
+ "learning_rate": 2.3334504441588413e-07,
+ "loss": 0.2781896352767944,
+ "num_input_tokens_seen": 1172260704,
+ "step": 26805,
+ "train_runtime": 207432.2609,
+ "train_tokens_per_second": 5651.294
+ },
+ {
+ "epoch": 1.8051440883382708,
+ "grad_norm": 0.7236248733853186,
+ "learning_rate": 2.3254739067702704e-07,
+ "loss": 0.2797673225402832,
+ "num_input_tokens_seen": 1172485680,
+ "step": 26810,
+ "train_runtime": 207469.9941,
+ "train_tokens_per_second": 5651.351
+ },
+ {
+ "epoch": 1.805480743334231,
+ "grad_norm": 0.8892645318062625,
+ "learning_rate": 2.3175107014109077e-07,
+ "loss": 0.2838028907775879,
+ "num_input_tokens_seen": 1172707720,
+ "step": 26815,
+ "train_runtime": 207507.3876,
+ "train_tokens_per_second": 5651.402
+ },
+ {
+ "epoch": 1.8058173983301913,
+ "grad_norm": 0.7651063250427357,
+ "learning_rate": 2.3095608303076223e-07,
+ "loss": 0.311909818649292,
+ "num_input_tokens_seen": 1172928712,
+ "step": 26820,
+ "train_runtime": 207551.5584,
+ "train_tokens_per_second": 5651.264
+ },
+ {
+ "epoch": 1.8061540533261513,
+ "grad_norm": 0.734890627438004,
+ "learning_rate": 2.3016242956835798e-07,
+ "loss": 0.31490211486816405,
+ "num_input_tokens_seen": 1173154336,
+ "step": 26825,
+ "train_runtime": 207594.3613,
+ "train_tokens_per_second": 5651.186
+ },
+ {
+ "epoch": 1.8064907083221113,
+ "grad_norm": 0.7733770590852225,
+ "learning_rate": 2.2937010997581999e-07,
+ "loss": 0.284251070022583,
+ "num_input_tokens_seen": 1173384008,
+ "step": 26830,
+ "train_runtime": 207634.0828,
+ "train_tokens_per_second": 5651.211
+ },
+ {
+ "epoch": 1.8068273633180716,
+ "grad_norm": 0.9236592739480588,
+ "learning_rate": 2.2857912447471818e-07,
+ "loss": 0.2854511260986328,
+ "num_input_tokens_seen": 1173591424,
+ "step": 26835,
+ "train_runtime": 207670.999,
+ "train_tokens_per_second": 5651.205
+ },
+ {
+ "epoch": 1.8071640183140318,
+ "grad_norm": 0.709148295682169,
+ "learning_rate": 2.2778947328625012e-07,
+ "loss": 0.2829423427581787,
+ "num_input_tokens_seen": 1173795056,
+ "step": 26840,
+ "train_runtime": 207707.7029,
+ "train_tokens_per_second": 5651.187
+ },
+ {
+ "epoch": 1.8075006733099919,
+ "grad_norm": 0.9439136755079099,
+ "learning_rate": 2.2700115663123745e-07,
+ "loss": 0.3229234218597412,
+ "num_input_tokens_seen": 1174010232,
+ "step": 26845,
+ "train_runtime": 207744.5589,
+ "train_tokens_per_second": 5651.22
+ },
+ {
+ "epoch": 1.807837328305952,
+ "grad_norm": 0.9241155338367218,
+ "learning_rate": 2.2621417473013164e-07,
+ "loss": 0.28131952285766604,
+ "num_input_tokens_seen": 1174230336,
+ "step": 26850,
+ "train_runtime": 207785.5145,
+ "train_tokens_per_second": 5651.166
+ },
+ {
+ "epoch": 1.8081739833019121,
+ "grad_norm": 0.8397274812785074,
+ "learning_rate": 2.254285278030094e-07,
+ "loss": 0.30530600547790526,
+ "num_input_tokens_seen": 1174459216,
+ "step": 26855,
+ "train_runtime": 207819.3658,
+ "train_tokens_per_second": 5651.346
+ },
+ {
+ "epoch": 1.8085106382978724,
+ "grad_norm": 0.8065249949734656,
+ "learning_rate": 2.24644216069575e-07,
+ "loss": 0.30002517700195314,
+ "num_input_tokens_seen": 1174667136,
+ "step": 26860,
+ "train_runtime": 207866.2773,
+ "train_tokens_per_second": 5651.071
+ },
+ {
+ "epoch": 1.8088472932938324,
+ "grad_norm": 0.6707793753515161,
+ "learning_rate": 2.238612397491574e-07,
+ "loss": 0.2729137420654297,
+ "num_input_tokens_seen": 1174893104,
+ "step": 26865,
+ "train_runtime": 207905.4724,
+ "train_tokens_per_second": 5651.093
+ },
+ {
+ "epoch": 1.8091839482897925,
+ "grad_norm": 0.7616275632457961,
+ "learning_rate": 2.230795990607143e-07,
+ "loss": 0.29819502830505373,
+ "num_input_tokens_seen": 1175120424,
+ "step": 26870,
+ "train_runtime": 207946.3976,
+ "train_tokens_per_second": 5651.074
+ },
+ {
+ "epoch": 1.809520603285753,
+ "grad_norm": 0.7786936080708119,
+ "learning_rate": 2.2229929422282915e-07,
+ "loss": 0.27527265548706054,
+ "num_input_tokens_seen": 1175334208,
+ "step": 26875,
+ "train_runtime": 207982.2651,
+ "train_tokens_per_second": 5651.127
+ },
+ {
+ "epoch": 1.809857258281713,
+ "grad_norm": 0.7904035393585753,
+ "learning_rate": 2.215203254537107e-07,
+ "loss": 0.3183107614517212,
+ "num_input_tokens_seen": 1175555296,
+ "step": 26880,
+ "train_runtime": 208022.7623,
+ "train_tokens_per_second": 5651.09
+ },
+ {
+ "epoch": 1.810193913277673,
+ "grad_norm": 0.7604104589550896,
+ "learning_rate": 2.2074269297119588e-07,
+ "loss": 0.3029907464981079,
+ "num_input_tokens_seen": 1175780944,
+ "step": 26885,
+ "train_runtime": 208056.1021,
+ "train_tokens_per_second": 5651.269
+ },
+ {
+ "epoch": 1.8105305682736332,
+ "grad_norm": 0.8292516099336918,
+ "learning_rate": 2.199663969927479e-07,
+ "loss": 0.29438438415527346,
+ "num_input_tokens_seen": 1175996056,
+ "step": 26890,
+ "train_runtime": 208096.3568,
+ "train_tokens_per_second": 5651.209
+ },
+ {
+ "epoch": 1.8108672232695935,
+ "grad_norm": 0.8271865279223709,
+ "learning_rate": 2.1919143773545316e-07,
+ "loss": 0.29373531341552733,
+ "num_input_tokens_seen": 1176219504,
+ "step": 26895,
+ "train_runtime": 208134.3303,
+ "train_tokens_per_second": 5651.252
+ },
+ {
+ "epoch": 1.8112038782655535,
+ "grad_norm": 0.9028884805515789,
+ "learning_rate": 2.184178154160288e-07,
+ "loss": 0.29395351409912107,
+ "num_input_tokens_seen": 1176436808,
+ "step": 26900,
+ "train_runtime": 208170.4974,
+ "train_tokens_per_second": 5651.314
+ },
+ {
+ "epoch": 1.8115405332615135,
+ "grad_norm": 0.7863867287916279,
+ "learning_rate": 2.1764553025081514e-07,
+ "loss": 0.2931424379348755,
+ "num_input_tokens_seen": 1176656768,
+ "step": 26905,
+ "train_runtime": 208207.3522,
+ "train_tokens_per_second": 5651.37
+ },
+ {
+ "epoch": 1.8118771882574738,
+ "grad_norm": 0.8798746009059921,
+ "learning_rate": 2.168745824557794e-07,
+ "loss": 0.28325557708740234,
+ "num_input_tokens_seen": 1176871952,
+ "step": 26910,
+ "train_runtime": 208250.7917,
+ "train_tokens_per_second": 5651.224
+ },
+ {
+ "epoch": 1.812213843253434,
+ "grad_norm": 0.8072441204009388,
+ "learning_rate": 2.161049722465136e-07,
+ "loss": 0.28316502571105956,
+ "num_input_tokens_seen": 1177088104,
+ "step": 26915,
+ "train_runtime": 208293.9071,
+ "train_tokens_per_second": 5651.092
+ },
+ {
+ "epoch": 1.812550498249394,
+ "grad_norm": 0.7553112409241435,
+ "learning_rate": 2.1533669983823835e-07,
+ "loss": 0.29533112049102783,
+ "num_input_tokens_seen": 1177320360,
+ "step": 26920,
+ "train_runtime": 208336.7997,
+ "train_tokens_per_second": 5651.044
+ },
+ {
+ "epoch": 1.812887153245354,
+ "grad_norm": 0.6386036144723565,
+ "learning_rate": 2.1456976544579732e-07,
+ "loss": 0.2751606464385986,
+ "num_input_tokens_seen": 1177560024,
+ "step": 26925,
+ "train_runtime": 208382.272,
+ "train_tokens_per_second": 5650.961
+ },
+ {
+ "epoch": 1.8132238082413143,
+ "grad_norm": 0.7537084466037821,
+ "learning_rate": 2.138041692836623e-07,
+ "loss": 0.29282474517822266,
+ "num_input_tokens_seen": 1177776744,
+ "step": 26930,
+ "train_runtime": 208414.4067,
+ "train_tokens_per_second": 5651.129
+ },
+ {
+ "epoch": 1.8135604632372746,
+ "grad_norm": 0.766246409563199,
+ "learning_rate": 2.1303991156592818e-07,
+ "loss": 0.29790616035461426,
+ "num_input_tokens_seen": 1177986848,
+ "step": 26935,
+ "train_runtime": 208452.677,
+ "train_tokens_per_second": 5651.1
+ },
+ {
+ "epoch": 1.8138971182332346,
+ "grad_norm": 0.7584006159785801,
+ "learning_rate": 2.122769925063195e-07,
+ "loss": 0.2797182559967041,
+ "num_input_tokens_seen": 1178200832,
+ "step": 26940,
+ "train_runtime": 208493.3034,
+ "train_tokens_per_second": 5651.025
+ },
+ {
+ "epoch": 1.8142337732291947,
+ "grad_norm": 0.8751221362570611,
+ "learning_rate": 2.1151541231818174e-07,
+ "loss": 0.2855216503143311,
+ "num_input_tokens_seen": 1178415944,
+ "step": 26945,
+ "train_runtime": 208536.0031,
+ "train_tokens_per_second": 5650.899
+ },
+ {
+ "epoch": 1.814570428225155,
+ "grad_norm": 0.8086208091486383,
+ "learning_rate": 2.1075517121449063e-07,
+ "loss": 0.31327362060546876,
+ "num_input_tokens_seen": 1178637904,
+ "step": 26950,
+ "train_runtime": 208579.4059,
+ "train_tokens_per_second": 5650.788
+ },
+ {
+ "epoch": 1.8149070832211152,
+ "grad_norm": 0.8301036812650167,
+ "learning_rate": 2.0999626940784334e-07,
+ "loss": 0.27339811325073243,
+ "num_input_tokens_seen": 1178843008,
+ "step": 26955,
+ "train_runtime": 208615.5879,
+ "train_tokens_per_second": 5650.791
+ },
+ {
+ "epoch": 1.8152437382170752,
+ "grad_norm": 0.8710877367090695,
+ "learning_rate": 2.0923870711046568e-07,
+ "loss": 0.2946287155151367,
+ "num_input_tokens_seen": 1179067688,
+ "step": 26960,
+ "train_runtime": 208653.4478,
+ "train_tokens_per_second": 5650.842
+ },
+ {
+ "epoch": 1.8155803932130352,
+ "grad_norm": 0.6554427095056514,
+ "learning_rate": 2.0848248453420704e-07,
+ "loss": 0.2847181797027588,
+ "num_input_tokens_seen": 1179293168,
+ "step": 26965,
+ "train_runtime": 208690.2441,
+ "train_tokens_per_second": 5650.926
+ },
+ {
+ "epoch": 1.8159170482089955,
+ "grad_norm": 0.8746736454560675,
+ "learning_rate": 2.077276018905422e-07,
+ "loss": 0.2915675163269043,
+ "num_input_tokens_seen": 1179511568,
+ "step": 26970,
+ "train_runtime": 208729.2372,
+ "train_tokens_per_second": 5650.917
+ },
+ {
+ "epoch": 1.8162537032049557,
+ "grad_norm": 0.7805453963758747,
+ "learning_rate": 2.0697405939057225e-07,
+ "loss": 0.3022061824798584,
+ "num_input_tokens_seen": 1179708384,
+ "step": 26975,
+ "train_runtime": 208763.205,
+ "train_tokens_per_second": 5650.94
+ },
+ {
+ "epoch": 1.8165903582009157,
+ "grad_norm": 0.6389930066356448,
+ "learning_rate": 2.0622185724502253e-07,
+ "loss": 0.27937405109405516,
+ "num_input_tokens_seen": 1179925616,
+ "step": 26980,
+ "train_runtime": 208797.998,
+ "train_tokens_per_second": 5651.039
+ },
+ {
+ "epoch": 1.8169270131968758,
+ "grad_norm": 0.8421462718448113,
+ "learning_rate": 2.0547099566424367e-07,
+ "loss": 0.2782083511352539,
+ "num_input_tokens_seen": 1180142704,
+ "step": 26985,
+ "train_runtime": 208842.2398,
+ "train_tokens_per_second": 5650.881
+ },
+ {
+ "epoch": 1.817263668192836,
+ "grad_norm": 0.8420718246560043,
+ "learning_rate": 2.0472147485821325e-07,
+ "loss": 0.27927365303039553,
+ "num_input_tokens_seen": 1180348248,
+ "step": 26990,
+ "train_runtime": 208879.8268,
+ "train_tokens_per_second": 5650.848
+ },
+ {
+ "epoch": 1.8176003231887963,
+ "grad_norm": 0.7460943379350691,
+ "learning_rate": 2.0397329503653084e-07,
+ "loss": 0.2773773670196533,
+ "num_input_tokens_seen": 1180563544,
+ "step": 26995,
+ "train_runtime": 208916.8282,
+ "train_tokens_per_second": 5650.878
+ },
+ {
+ "epoch": 1.8179369781847563,
+ "grad_norm": 0.8475284436859102,
+ "learning_rate": 2.03226456408423e-07,
+ "loss": 0.3106630802154541,
+ "num_input_tokens_seen": 1180762848,
+ "step": 27000,
+ "train_runtime": 208960.3449,
+ "train_tokens_per_second": 5650.655
+ },
+ {
+ "epoch": 1.8182736331807163,
+ "grad_norm": 0.7183644291988805,
+ "learning_rate": 2.024809591827409e-07,
+ "loss": 0.30523037910461426,
+ "num_input_tokens_seen": 1180988840,
+ "step": 27005,
+ "train_runtime": 209002.5667,
+ "train_tokens_per_second": 5650.595
+ },
+ {
+ "epoch": 1.8186102881766766,
+ "grad_norm": 0.801765388255765,
+ "learning_rate": 2.0173680356796066e-07,
+ "loss": 0.26985788345336914,
+ "num_input_tokens_seen": 1181189432,
+ "step": 27010,
+ "train_runtime": 209041.262,
+ "train_tokens_per_second": 5650.509
+ },
+ {
+ "epoch": 1.8189469431726368,
+ "grad_norm": 0.8150765259723309,
+ "learning_rate": 2.00993989772183e-07,
+ "loss": 0.2943392515182495,
+ "num_input_tokens_seen": 1181407992,
+ "step": 27015,
+ "train_runtime": 209073.5583,
+ "train_tokens_per_second": 5650.681
+ },
+ {
+ "epoch": 1.8192835981685969,
+ "grad_norm": 0.7914385589041727,
+ "learning_rate": 2.0025251800313284e-07,
+ "loss": 0.3112603187561035,
+ "num_input_tokens_seen": 1181622192,
+ "step": 27020,
+ "train_runtime": 209120.2047,
+ "train_tokens_per_second": 5650.445
+ },
+ {
+ "epoch": 1.8196202531645569,
+ "grad_norm": 0.7561624538382505,
+ "learning_rate": 1.9951238846816045e-07,
+ "loss": 0.3185489892959595,
+ "num_input_tokens_seen": 1181854088,
+ "step": 27025,
+ "train_runtime": 209152.1413,
+ "train_tokens_per_second": 5650.691
+ },
+ {
+ "epoch": 1.8199569081605171,
+ "grad_norm": 0.7921686074062922,
+ "learning_rate": 1.987736013742414e-07,
+ "loss": 0.3113109111785889,
+ "num_input_tokens_seen": 1182065984,
+ "step": 27030,
+ "train_runtime": 209191.0211,
+ "train_tokens_per_second": 5650.654
+ },
+ {
+ "epoch": 1.8202935631564774,
+ "grad_norm": 0.7653273331531749,
+ "learning_rate": 1.9803615692797375e-07,
+ "loss": 0.27271738052368166,
+ "num_input_tokens_seen": 1182287536,
+ "step": 27035,
+ "train_runtime": 209228.5228,
+ "train_tokens_per_second": 5650.7
+ },
+ {
+ "epoch": 1.8206302181524374,
+ "grad_norm": 0.7068384362133963,
+ "learning_rate": 1.9730005533558417e-07,
+ "loss": 0.29187893867492676,
+ "num_input_tokens_seen": 1182520232,
+ "step": 27040,
+ "train_runtime": 209272.9747,
+ "train_tokens_per_second": 5650.611
+ },
+ {
+ "epoch": 1.8209668731483974,
+ "grad_norm": 0.7399705032741323,
+ "learning_rate": 1.9656529680291802e-07,
+ "loss": 0.27612121105194093,
+ "num_input_tokens_seen": 1182742904,
+ "step": 27045,
+ "train_runtime": 209312.4766,
+ "train_tokens_per_second": 5650.609
+ },
+ {
+ "epoch": 1.8213035281443577,
+ "grad_norm": 0.7369131913718883,
+ "learning_rate": 1.9583188153544986e-07,
+ "loss": 0.28553037643432616,
+ "num_input_tokens_seen": 1182945672,
+ "step": 27050,
+ "train_runtime": 209347.8795,
+ "train_tokens_per_second": 5650.622
+ },
+ {
+ "epoch": 1.821640183140318,
+ "grad_norm": 1.1539568423391955,
+ "learning_rate": 1.9509980973827724e-07,
+ "loss": 0.2979405641555786,
+ "num_input_tokens_seen": 1183165608,
+ "step": 27055,
+ "train_runtime": 209386.946,
+ "train_tokens_per_second": 5650.618
+ },
+ {
+ "epoch": 1.821976838136278,
+ "grad_norm": 0.7640565156115381,
+ "learning_rate": 1.943690816161209e-07,
+ "loss": 0.3077999114990234,
+ "num_input_tokens_seen": 1183395432,
+ "step": 27060,
+ "train_runtime": 209424.1043,
+ "train_tokens_per_second": 5650.713
+ },
+ {
+ "epoch": 1.822313493132238,
+ "grad_norm": 0.7149421204366606,
+ "learning_rate": 1.9363969737332677e-07,
+ "loss": 0.2753284454345703,
+ "num_input_tokens_seen": 1183617608,
+ "step": 27065,
+ "train_runtime": 209460.0325,
+ "train_tokens_per_second": 5650.804
+ },
+ {
+ "epoch": 1.8226501481281983,
+ "grad_norm": 0.652062441800595,
+ "learning_rate": 1.9291165721386507e-07,
+ "loss": 0.27479655742645265,
+ "num_input_tokens_seen": 1183826768,
+ "step": 27070,
+ "train_runtime": 209495.4576,
+ "train_tokens_per_second": 5650.847
+ },
+ {
+ "epoch": 1.8229868031241585,
+ "grad_norm": 0.9271372389885489,
+ "learning_rate": 1.9218496134132957e-07,
+ "loss": 0.3049477100372314,
+ "num_input_tokens_seen": 1184052800,
+ "step": 27075,
+ "train_runtime": 209531.6276,
+ "train_tokens_per_second": 5650.95
+ },
+ {
+ "epoch": 1.8233234581201185,
+ "grad_norm": 0.6911714020764224,
+ "learning_rate": 1.9145960995893831e-07,
+ "loss": 0.282161283493042,
+ "num_input_tokens_seen": 1184270392,
+ "step": 27080,
+ "train_runtime": 209567.2517,
+ "train_tokens_per_second": 5651.028
+ },
+ {
+ "epoch": 1.8236601131160786,
+ "grad_norm": 0.7018056304999722,
+ "learning_rate": 1.9073560326953344e-07,
+ "loss": 0.3012186050415039,
+ "num_input_tokens_seen": 1184494848,
+ "step": 27085,
+ "train_runtime": 209601.3891,
+ "train_tokens_per_second": 5651.178
+ },
+ {
+ "epoch": 1.8239967681120388,
+ "grad_norm": 0.7756023060864273,
+ "learning_rate": 1.9001294147558303e-07,
+ "loss": 0.3109905242919922,
+ "num_input_tokens_seen": 1184709808,
+ "step": 27090,
+ "train_runtime": 209641.9441,
+ "train_tokens_per_second": 5651.111
+ },
+ {
+ "epoch": 1.824333423107999,
+ "grad_norm": 0.8354480678136353,
+ "learning_rate": 1.892916247791743e-07,
+ "loss": 0.3057056188583374,
+ "num_input_tokens_seen": 1184920824,
+ "step": 27095,
+ "train_runtime": 209678.6955,
+ "train_tokens_per_second": 5651.126
+ },
+ {
+ "epoch": 1.824670078103959,
+ "grad_norm": 0.7450481139042144,
+ "learning_rate": 1.885716533820231e-07,
+ "loss": 0.2988601207733154,
+ "num_input_tokens_seen": 1185141448,
+ "step": 27100,
+ "train_runtime": 209721.9756,
+ "train_tokens_per_second": 5651.012
+ },
+ {
+ "epoch": 1.8250067330999191,
+ "grad_norm": 0.7070713398300131,
+ "learning_rate": 1.878530274854662e-07,
+ "loss": 0.2867612838745117,
+ "num_input_tokens_seen": 1185358376,
+ "step": 27105,
+ "train_runtime": 209757.3918,
+ "train_tokens_per_second": 5651.092
+ },
+ {
+ "epoch": 1.8253433880958794,
+ "grad_norm": 0.7229583190371239,
+ "learning_rate": 1.8713574729046557e-07,
+ "loss": 0.3091785669326782,
+ "num_input_tokens_seen": 1185587344,
+ "step": 27110,
+ "train_runtime": 209793.505,
+ "train_tokens_per_second": 5651.211
+ },
+ {
+ "epoch": 1.8256800430918396,
+ "grad_norm": 0.7520860741774413,
+ "learning_rate": 1.8641981299760582e-07,
+ "loss": 0.29340548515319825,
+ "num_input_tokens_seen": 1185795904,
+ "step": 27115,
+ "train_runtime": 209826.9024,
+ "train_tokens_per_second": 5651.305
+ },
+ {
+ "epoch": 1.8260166980877997,
+ "grad_norm": 0.7995613128971399,
+ "learning_rate": 1.857052248070962e-07,
+ "loss": 0.30235800743103025,
+ "num_input_tokens_seen": 1186003992,
+ "step": 27120,
+ "train_runtime": 209868.1712,
+ "train_tokens_per_second": 5651.186
+ },
+ {
+ "epoch": 1.8263533530837597,
+ "grad_norm": 0.8063565780315961,
+ "learning_rate": 1.8499198291876907e-07,
+ "loss": 0.2838723659515381,
+ "num_input_tokens_seen": 1186206928,
+ "step": 27125,
+ "train_runtime": 209908.8621,
+ "train_tokens_per_second": 5651.057
+ },
+ {
+ "epoch": 1.82669000807972,
+ "grad_norm": 0.659822757313705,
+ "learning_rate": 1.8428008753207937e-07,
+ "loss": 0.3044004440307617,
+ "num_input_tokens_seen": 1186430512,
+ "step": 27130,
+ "train_runtime": 209952.433,
+ "train_tokens_per_second": 5650.949
+ },
+ {
+ "epoch": 1.8270266630756802,
+ "grad_norm": 0.7534355877635904,
+ "learning_rate": 1.8356953884610673e-07,
+ "loss": 0.3086956262588501,
+ "num_input_tokens_seen": 1186652392,
+ "step": 27135,
+ "train_runtime": 209996.3146,
+ "train_tokens_per_second": 5650.825
+ },
+ {
+ "epoch": 1.8273633180716402,
+ "grad_norm": 0.7536622469259906,
+ "learning_rate": 1.8286033705955554e-07,
+ "loss": 0.2920703411102295,
+ "num_input_tokens_seen": 1186878432,
+ "step": 27140,
+ "train_runtime": 210028.5754,
+ "train_tokens_per_second": 5651.033
+ },
+ {
+ "epoch": 1.8276999730676002,
+ "grad_norm": 0.8189703081756119,
+ "learning_rate": 1.8215248237074879e-07,
+ "loss": 0.2958489656448364,
+ "num_input_tokens_seen": 1187100664,
+ "step": 27145,
+ "train_runtime": 210067.8385,
+ "train_tokens_per_second": 5651.035
+ },
+ {
+ "epoch": 1.8280366280635605,
+ "grad_norm": 0.6908418891259124,
+ "learning_rate": 1.8144597497763816e-07,
+ "loss": 0.2915064334869385,
+ "num_input_tokens_seen": 1187314688,
+ "step": 27150,
+ "train_runtime": 210105.6707,
+ "train_tokens_per_second": 5651.036
+ },
+ {
+ "epoch": 1.8283732830595207,
+ "grad_norm": 0.7771306934508,
+ "learning_rate": 1.8074081507779506e-07,
+ "loss": 0.2950667142868042,
+ "num_input_tokens_seen": 1187543272,
+ "step": 27155,
+ "train_runtime": 210145.1035,
+ "train_tokens_per_second": 5651.063
+ },
+ {
+ "epoch": 1.8287099380554808,
+ "grad_norm": 0.7951975641517514,
+ "learning_rate": 1.8003700286841563e-07,
+ "loss": 0.3046703815460205,
+ "num_input_tokens_seen": 1187764808,
+ "step": 27160,
+ "train_runtime": 210185.6937,
+ "train_tokens_per_second": 5651.026
+ },
+ {
+ "epoch": 1.8290465930514408,
+ "grad_norm": 0.8247212903362624,
+ "learning_rate": 1.793345385463191e-07,
+ "loss": 0.2969427824020386,
+ "num_input_tokens_seen": 1187985984,
+ "step": 27165,
+ "train_runtime": 210225.4532,
+ "train_tokens_per_second": 5651.009
+ },
+ {
+ "epoch": 1.829383248047401,
+ "grad_norm": 0.8389229794347001,
+ "learning_rate": 1.7863342230794667e-07,
+ "loss": 0.306946849822998,
+ "num_input_tokens_seen": 1188205016,
+ "step": 27170,
+ "train_runtime": 210256.0102,
+ "train_tokens_per_second": 5651.23
+ },
+ {
+ "epoch": 1.8297199030433613,
+ "grad_norm": 0.6995002304359613,
+ "learning_rate": 1.7793365434936316e-07,
+ "loss": 0.2987263917922974,
+ "num_input_tokens_seen": 1188436064,
+ "step": 27175,
+ "train_runtime": 210296.1751,
+ "train_tokens_per_second": 5651.249
+ },
+ {
+ "epoch": 1.8300565580393213,
+ "grad_norm": 0.8195984422353864,
+ "learning_rate": 1.7723523486625704e-07,
+ "loss": 0.29759798049926756,
+ "num_input_tokens_seen": 1188648440,
+ "step": 27180,
+ "train_runtime": 210337.9552,
+ "train_tokens_per_second": 5651.136
+ },
+ {
+ "epoch": 1.8303932130352814,
+ "grad_norm": 0.7388330413443034,
+ "learning_rate": 1.7653816405393766e-07,
+ "loss": 0.3033345937728882,
+ "num_input_tokens_seen": 1188864712,
+ "step": 27185,
+ "train_runtime": 210377.2381,
+ "train_tokens_per_second": 5651.109
+ },
+ {
+ "epoch": 1.8307298680312416,
+ "grad_norm": 0.8567770406651678,
+ "learning_rate": 1.758424421073418e-07,
+ "loss": 0.308400821685791,
+ "num_input_tokens_seen": 1189093376,
+ "step": 27190,
+ "train_runtime": 210414.6866,
+ "train_tokens_per_second": 5651.19
+ },
+ {
+ "epoch": 1.8310665230272019,
+ "grad_norm": 0.7475393949595193,
+ "learning_rate": 1.7514806922102222e-07,
+ "loss": 0.2787260293960571,
+ "num_input_tokens_seen": 1189290992,
+ "step": 27195,
+ "train_runtime": 210452.2542,
+ "train_tokens_per_second": 5651.12
+ },
+ {
+ "epoch": 1.8314031780231619,
+ "grad_norm": 0.7779392879208484,
+ "learning_rate": 1.7445504558916027e-07,
+ "loss": 0.2868196964263916,
+ "num_input_tokens_seen": 1189504064,
+ "step": 27200,
+ "train_runtime": 210482.3919,
+ "train_tokens_per_second": 5651.323
+ },
+ {
+ "epoch": 1.831739833019122,
+ "grad_norm": 0.7793252529447385,
+ "learning_rate": 1.7376337140555754e-07,
+ "loss": 0.283552360534668,
+ "num_input_tokens_seen": 1189712800,
+ "step": 27205,
+ "train_runtime": 210518.9894,
+ "train_tokens_per_second": 5651.332
+ },
+ {
+ "epoch": 1.8320764880150822,
+ "grad_norm": 0.7136774949097324,
+ "learning_rate": 1.7307304686363825e-07,
+ "loss": 0.30210127830505373,
+ "num_input_tokens_seen": 1189932032,
+ "step": 27210,
+ "train_runtime": 210557.2729,
+ "train_tokens_per_second": 5651.346
+ },
+ {
+ "epoch": 1.8324131430110424,
+ "grad_norm": 0.6733587332513272,
+ "learning_rate": 1.723840721564496e-07,
+ "loss": 0.2896436214447021,
+ "num_input_tokens_seen": 1190162824,
+ "step": 27215,
+ "train_runtime": 210596.5269,
+ "train_tokens_per_second": 5651.389
+ },
+ {
+ "epoch": 1.8327497980070024,
+ "grad_norm": 0.7975602822015874,
+ "learning_rate": 1.7169644747666081e-07,
+ "loss": 0.29418466091156004,
+ "num_input_tokens_seen": 1190392976,
+ "step": 27220,
+ "train_runtime": 210635.9714,
+ "train_tokens_per_second": 5651.423
+ },
+ {
+ "epoch": 1.8330864530029625,
+ "grad_norm": 0.6803205298907224,
+ "learning_rate": 1.710101730165642e-07,
+ "loss": 0.30214219093322753,
+ "num_input_tokens_seen": 1190618688,
+ "step": 27225,
+ "train_runtime": 210673.7247,
+ "train_tokens_per_second": 5651.482
+ },
+ {
+ "epoch": 1.8334231079989227,
+ "grad_norm": 0.7468300724636248,
+ "learning_rate": 1.7032524896807402e-07,
+ "loss": 0.27968459129333495,
+ "num_input_tokens_seen": 1190825912,
+ "step": 27230,
+ "train_runtime": 210712.9044,
+ "train_tokens_per_second": 5651.414
+ },
+ {
+ "epoch": 1.833759762994883,
+ "grad_norm": 0.75132763867745,
+ "learning_rate": 1.6964167552272702e-07,
+ "loss": 0.3041102647781372,
+ "num_input_tokens_seen": 1191031648,
+ "step": 27235,
+ "train_runtime": 210750.2864,
+ "train_tokens_per_second": 5651.388
+ },
+ {
+ "epoch": 1.834096417990843,
+ "grad_norm": 0.8236438444986357,
+ "learning_rate": 1.6895945287168314e-07,
+ "loss": 0.30144453048706055,
+ "num_input_tokens_seen": 1191243504,
+ "step": 27240,
+ "train_runtime": 210787.0749,
+ "train_tokens_per_second": 5651.407
+ },
+ {
+ "epoch": 1.834433072986803,
+ "grad_norm": 0.8116073242293598,
+ "learning_rate": 1.682785812057225e-07,
+ "loss": 0.31502351760864256,
+ "num_input_tokens_seen": 1191444928,
+ "step": 27245,
+ "train_runtime": 210818.024,
+ "train_tokens_per_second": 5651.533
+ },
+ {
+ "epoch": 1.8347697279827633,
+ "grad_norm": 0.7705506997654222,
+ "learning_rate": 1.6759906071525e-07,
+ "loss": 0.2920320987701416,
+ "num_input_tokens_seen": 1191662640,
+ "step": 27250,
+ "train_runtime": 210862.0161,
+ "train_tokens_per_second": 5651.386
+ },
+ {
+ "epoch": 1.8351063829787235,
+ "grad_norm": 0.7712074370148209,
+ "learning_rate": 1.669208915902909e-07,
+ "loss": 0.33261871337890625,
+ "num_input_tokens_seen": 1191900592,
+ "step": 27255,
+ "train_runtime": 210897.8718,
+ "train_tokens_per_second": 5651.553
+ },
+ {
+ "epoch": 1.8354430379746836,
+ "grad_norm": 0.7737113976334139,
+ "learning_rate": 1.6624407402049237e-07,
+ "loss": 0.2747644901275635,
+ "num_input_tokens_seen": 1192112584,
+ "step": 27260,
+ "train_runtime": 210931.9902,
+ "train_tokens_per_second": 5651.644
+ },
+ {
+ "epoch": 1.8357796929706436,
+ "grad_norm": 0.7765514097289985,
+ "learning_rate": 1.655686081951252e-07,
+ "loss": 0.28942217826843264,
+ "num_input_tokens_seen": 1192328048,
+ "step": 27265,
+ "train_runtime": 210962.6642,
+ "train_tokens_per_second": 5651.844
+ },
+ {
+ "epoch": 1.8361163479666038,
+ "grad_norm": 0.7986187558729846,
+ "learning_rate": 1.6489449430308058e-07,
+ "loss": 0.32457714080810546,
+ "num_input_tokens_seen": 1192555872,
+ "step": 27270,
+ "train_runtime": 211006.3008,
+ "train_tokens_per_second": 5651.755
+ },
+ {
+ "epoch": 1.836453002962564,
+ "grad_norm": 0.7814336439979509,
+ "learning_rate": 1.6422173253287323e-07,
+ "loss": 0.31779980659484863,
+ "num_input_tokens_seen": 1192784792,
+ "step": 27275,
+ "train_runtime": 211044.4069,
+ "train_tokens_per_second": 5651.819
+ },
+ {
+ "epoch": 1.8367896579585241,
+ "grad_norm": 0.8231817333366194,
+ "learning_rate": 1.6355032307263764e-07,
+ "loss": 0.2909559726715088,
+ "num_input_tokens_seen": 1193008024,
+ "step": 27280,
+ "train_runtime": 211087.1712,
+ "train_tokens_per_second": 5651.732
+ },
+ {
+ "epoch": 1.8371263129544841,
+ "grad_norm": 0.7297603518728784,
+ "learning_rate": 1.628802661101314e-07,
+ "loss": 0.27639720439910886,
+ "num_input_tokens_seen": 1193222040,
+ "step": 27285,
+ "train_runtime": 211126.1642,
+ "train_tokens_per_second": 5651.701
+ },
+ {
+ "epoch": 1.8374629679504444,
+ "grad_norm": 0.7393851625213684,
+ "learning_rate": 1.6221156183273524e-07,
+ "loss": 0.2929367542266846,
+ "num_input_tokens_seen": 1193445752,
+ "step": 27290,
+ "train_runtime": 211166.2819,
+ "train_tokens_per_second": 5651.687
+ },
+ {
+ "epoch": 1.8377996229464046,
+ "grad_norm": 0.8626095075985437,
+ "learning_rate": 1.615442104274484e-07,
+ "loss": 0.2990311622619629,
+ "num_input_tokens_seen": 1193676024,
+ "step": 27295,
+ "train_runtime": 211203.0891,
+ "train_tokens_per_second": 5651.792
+ },
+ {
+ "epoch": 1.8381362779423647,
+ "grad_norm": 0.7989960018689986,
+ "learning_rate": 1.6087821208089492e-07,
+ "loss": 0.2992110252380371,
+ "num_input_tokens_seen": 1193898896,
+ "step": 27300,
+ "train_runtime": 211239.9537,
+ "train_tokens_per_second": 5651.861
+ },
+ {
+ "epoch": 1.8384729329383247,
+ "grad_norm": 0.675694068212015,
+ "learning_rate": 1.6021356697931866e-07,
+ "loss": 0.2737618684768677,
+ "num_input_tokens_seen": 1194124640,
+ "step": 27305,
+ "train_runtime": 211274.8199,
+ "train_tokens_per_second": 5651.997
+ },
+ {
+ "epoch": 1.838809587934285,
+ "grad_norm": 0.7414825082624487,
+ "learning_rate": 1.5955027530858537e-07,
+ "loss": 0.27455065250396726,
+ "num_input_tokens_seen": 1194345288,
+ "step": 27310,
+ "train_runtime": 211311.0878,
+ "train_tokens_per_second": 5652.071
+ },
+ {
+ "epoch": 1.8391462429302452,
+ "grad_norm": 0.6826873569949812,
+ "learning_rate": 1.5888833725418275e-07,
+ "loss": 0.2740659236907959,
+ "num_input_tokens_seen": 1194562672,
+ "step": 27315,
+ "train_runtime": 211347.4146,
+ "train_tokens_per_second": 5652.128
+ },
+ {
+ "epoch": 1.8394828979262052,
+ "grad_norm": 0.7227868616604537,
+ "learning_rate": 1.582277530012194e-07,
+ "loss": 0.26207771301269533,
+ "num_input_tokens_seen": 1194797416,
+ "step": 27320,
+ "train_runtime": 211380.4831,
+ "train_tokens_per_second": 5652.354
+ },
+ {
+ "epoch": 1.8398195529221653,
+ "grad_norm": 0.8448912125013232,
+ "learning_rate": 1.5756852273442536e-07,
+ "loss": 0.29030356407165525,
+ "num_input_tokens_seen": 1195005200,
+ "step": 27325,
+ "train_runtime": 211422.3023,
+ "train_tokens_per_second": 5652.219
+ },
+ {
+ "epoch": 1.8401562079181255,
+ "grad_norm": 0.697440308697788,
+ "learning_rate": 1.569106466381526e-07,
+ "loss": 0.3182931900024414,
+ "num_input_tokens_seen": 1195227200,
+ "step": 27330,
+ "train_runtime": 211462.6591,
+ "train_tokens_per_second": 5652.19
+ },
+ {
+ "epoch": 1.8404928629140858,
+ "grad_norm": 0.7422044369619762,
+ "learning_rate": 1.5625412489637337e-07,
+ "loss": 0.27002391815185545,
+ "num_input_tokens_seen": 1195454536,
+ "step": 27335,
+ "train_runtime": 211507.1847,
+ "train_tokens_per_second": 5652.075
+ },
+ {
+ "epoch": 1.8408295179100458,
+ "grad_norm": 0.7443593742932311,
+ "learning_rate": 1.555989576926842e-07,
+ "loss": 0.2888377904891968,
+ "num_input_tokens_seen": 1195681944,
+ "step": 27340,
+ "train_runtime": 211549.0676,
+ "train_tokens_per_second": 5652.031
+ },
+ {
+ "epoch": 1.8411661729060058,
+ "grad_norm": 0.801773457702004,
+ "learning_rate": 1.5494514521029736e-07,
+ "loss": 0.29750452041625974,
+ "num_input_tokens_seen": 1195905376,
+ "step": 27345,
+ "train_runtime": 211588.7755,
+ "train_tokens_per_second": 5652.027
+ },
+ {
+ "epoch": 1.841502827901966,
+ "grad_norm": 0.7523395284692173,
+ "learning_rate": 1.5429268763205108e-07,
+ "loss": 0.2998412370681763,
+ "num_input_tokens_seen": 1196127744,
+ "step": 27350,
+ "train_runtime": 211635.9655,
+ "train_tokens_per_second": 5651.817
+ },
+ {
+ "epoch": 1.8418394828979263,
+ "grad_norm": 0.8119149425227684,
+ "learning_rate": 1.5364158514040328e-07,
+ "loss": 0.3214261531829834,
+ "num_input_tokens_seen": 1196344424,
+ "step": 27355,
+ "train_runtime": 211676.2418,
+ "train_tokens_per_second": 5651.765
+ },
+ {
+ "epoch": 1.8421761378938863,
+ "grad_norm": 0.9267135845555328,
+ "learning_rate": 1.5299183791743223e-07,
+ "loss": 0.30311391353607176,
+ "num_input_tokens_seen": 1196551984,
+ "step": 27360,
+ "train_runtime": 211712.3977,
+ "train_tokens_per_second": 5651.78
+ },
+ {
+ "epoch": 1.8425127928898464,
+ "grad_norm": 0.79776476761183,
+ "learning_rate": 1.523434461448381e-07,
+ "loss": 0.31781580448150637,
+ "num_input_tokens_seen": 1196784808,
+ "step": 27365,
+ "train_runtime": 211750.7126,
+ "train_tokens_per_second": 5651.857
+ },
+ {
+ "epoch": 1.8428494478858066,
+ "grad_norm": 0.7201171985633329,
+ "learning_rate": 1.5169641000394086e-07,
+ "loss": 0.29927232265472414,
+ "num_input_tokens_seen": 1197018080,
+ "step": 27370,
+ "train_runtime": 211789.2549,
+ "train_tokens_per_second": 5651.93
+ },
+ {
+ "epoch": 1.8431861028817669,
+ "grad_norm": 0.730174194835056,
+ "learning_rate": 1.5105072967568246e-07,
+ "loss": 0.2902745246887207,
+ "num_input_tokens_seen": 1197229392,
+ "step": 27375,
+ "train_runtime": 211827.3609,
+ "train_tokens_per_second": 5651.911
+ },
+ {
+ "epoch": 1.843522757877727,
+ "grad_norm": 0.6910545329116042,
+ "learning_rate": 1.504064053406251e-07,
+ "loss": 0.29667439460754397,
+ "num_input_tokens_seen": 1197443992,
+ "step": 27380,
+ "train_runtime": 211869.9495,
+ "train_tokens_per_second": 5651.788
+ },
+ {
+ "epoch": 1.843859412873687,
+ "grad_norm": 0.7540793648170933,
+ "learning_rate": 1.4976343717895192e-07,
+ "loss": 0.3001372814178467,
+ "num_input_tokens_seen": 1197673640,
+ "step": 27385,
+ "train_runtime": 211902.7264,
+ "train_tokens_per_second": 5651.997
+ },
+ {
+ "epoch": 1.8441960678696472,
+ "grad_norm": 0.7814817716047526,
+ "learning_rate": 1.491218253704685e-07,
+ "loss": 0.30607056617736816,
+ "num_input_tokens_seen": 1197898760,
+ "step": 27390,
+ "train_runtime": 211943.3379,
+ "train_tokens_per_second": 5651.976
+ },
+ {
+ "epoch": 1.8445327228656074,
+ "grad_norm": 0.7515118407147103,
+ "learning_rate": 1.484815700945974e-07,
+ "loss": 0.30979294776916505,
+ "num_input_tokens_seen": 1198115856,
+ "step": 27395,
+ "train_runtime": 211982.4794,
+ "train_tokens_per_second": 5651.957
+ },
+ {
+ "epoch": 1.8448693778615675,
+ "grad_norm": 0.7192859496811161,
+ "learning_rate": 1.478426715303849e-07,
+ "loss": 0.2884451150894165,
+ "num_input_tokens_seen": 1198320608,
+ "step": 27400,
+ "train_runtime": 212021.1992,
+ "train_tokens_per_second": 5651.891
+ },
+ {
+ "epoch": 1.8452060328575275,
+ "grad_norm": 0.7754550279674132,
+ "learning_rate": 1.4720512985649694e-07,
+ "loss": 0.29123849868774415,
+ "num_input_tokens_seen": 1198532416,
+ "step": 27405,
+ "train_runtime": 212063.5064,
+ "train_tokens_per_second": 5651.762
+ },
+ {
+ "epoch": 1.8455426878534877,
+ "grad_norm": 0.9501576369289315,
+ "learning_rate": 1.4656894525122034e-07,
+ "loss": 0.30306599140167234,
+ "num_input_tokens_seen": 1198734552,
+ "step": 27410,
+ "train_runtime": 212105.602,
+ "train_tokens_per_second": 5651.593
+ },
+ {
+ "epoch": 1.845879342849448,
+ "grad_norm": 0.7974261229774522,
+ "learning_rate": 1.459341178924617e-07,
+ "loss": 0.30881667137145996,
+ "num_input_tokens_seen": 1198955512,
+ "step": 27415,
+ "train_runtime": 212144.96,
+ "train_tokens_per_second": 5651.586
+ },
+ {
+ "epoch": 1.846215997845408,
+ "grad_norm": 0.7988079730512095,
+ "learning_rate": 1.453006479577479e-07,
+ "loss": 0.28842964172363283,
+ "num_input_tokens_seen": 1199167896,
+ "step": 27420,
+ "train_runtime": 212180.9912,
+ "train_tokens_per_second": 5651.627
+ },
+ {
+ "epoch": 1.846552652841368,
+ "grad_norm": 0.7202060611505903,
+ "learning_rate": 1.4466853562422778e-07,
+ "loss": 0.3094759464263916,
+ "num_input_tokens_seen": 1199366136,
+ "step": 27425,
+ "train_runtime": 212214.506,
+ "train_tokens_per_second": 5651.669
+ },
+ {
+ "epoch": 1.8468893078373283,
+ "grad_norm": 0.8214802158441638,
+ "learning_rate": 1.440377810686694e-07,
+ "loss": 0.294615650177002,
+ "num_input_tokens_seen": 1199589824,
+ "step": 27430,
+ "train_runtime": 212247.8835,
+ "train_tokens_per_second": 5651.834
+ },
+ {
+ "epoch": 1.8472259628332885,
+ "grad_norm": 0.7935797505217701,
+ "learning_rate": 1.4340838446746052e-07,
+ "loss": 0.2963638067245483,
+ "num_input_tokens_seen": 1199816200,
+ "step": 27435,
+ "train_runtime": 212284.2378,
+ "train_tokens_per_second": 5651.933
+ },
+ {
+ "epoch": 1.8475626178292486,
+ "grad_norm": 0.7255235631748863,
+ "learning_rate": 1.4278034599661094e-07,
+ "loss": 0.2769758701324463,
+ "num_input_tokens_seen": 1200031328,
+ "step": 27440,
+ "train_runtime": 212314.5115,
+ "train_tokens_per_second": 5652.14
+ },
+ {
+ "epoch": 1.8478992728252086,
+ "grad_norm": 0.78374561254918,
+ "learning_rate": 1.421536658317485e-07,
+ "loss": 0.2689784288406372,
+ "num_input_tokens_seen": 1200227096,
+ "step": 27445,
+ "train_runtime": 212349.6526,
+ "train_tokens_per_second": 5652.126
+ },
+ {
+ "epoch": 1.8482359278211689,
+ "grad_norm": 0.7784687447417672,
+ "learning_rate": 1.41528344148123e-07,
+ "loss": 0.28493242263793944,
+ "num_input_tokens_seen": 1200462264,
+ "step": 27450,
+ "train_runtime": 212386.6742,
+ "train_tokens_per_second": 5652.249
+ },
+ {
+ "epoch": 1.848572582817129,
+ "grad_norm": 0.8647273624224125,
+ "learning_rate": 1.40904381120604e-07,
+ "loss": 0.2554482936859131,
+ "num_input_tokens_seen": 1200679968,
+ "step": 27455,
+ "train_runtime": 212417.8431,
+ "train_tokens_per_second": 5652.444
+ },
+ {
+ "epoch": 1.8489092378130891,
+ "grad_norm": 0.7867893330661792,
+ "learning_rate": 1.4028177692367973e-07,
+ "loss": 0.2979857921600342,
+ "num_input_tokens_seen": 1200888896,
+ "step": 27460,
+ "train_runtime": 212455.3142,
+ "train_tokens_per_second": 5652.43
+ },
+ {
+ "epoch": 1.8492458928090492,
+ "grad_norm": 0.9388552543962256,
+ "learning_rate": 1.3966053173146032e-07,
+ "loss": 0.29952003955841067,
+ "num_input_tokens_seen": 1201115544,
+ "step": 27465,
+ "train_runtime": 212489.4425,
+ "train_tokens_per_second": 5652.589
+ },
+ {
+ "epoch": 1.8495825478050094,
+ "grad_norm": 0.6831577223182406,
+ "learning_rate": 1.3904064571767462e-07,
+ "loss": 0.29733192920684814,
+ "num_input_tokens_seen": 1201343992,
+ "step": 27470,
+ "train_runtime": 212529.6948,
+ "train_tokens_per_second": 5652.594
+ },
+ {
+ "epoch": 1.8499192028009697,
+ "grad_norm": 0.7057151389200952,
+ "learning_rate": 1.3842211905567227e-07,
+ "loss": 0.2762616157531738,
+ "num_input_tokens_seen": 1201569224,
+ "step": 27475,
+ "train_runtime": 212561.6614,
+ "train_tokens_per_second": 5652.803
+ },
+ {
+ "epoch": 1.8502558577969297,
+ "grad_norm": 0.8452116887810223,
+ "learning_rate": 1.378049519184216e-07,
+ "loss": 0.2911623239517212,
+ "num_input_tokens_seen": 1201779840,
+ "step": 27480,
+ "train_runtime": 212599.5548,
+ "train_tokens_per_second": 5652.786
+ },
+ {
+ "epoch": 1.8505925127928897,
+ "grad_norm": 0.7069439995002406,
+ "learning_rate": 1.3718914447851116e-07,
+ "loss": 0.2653970718383789,
+ "num_input_tokens_seen": 1202007472,
+ "step": 27485,
+ "train_runtime": 212638.9418,
+ "train_tokens_per_second": 5652.81
+ },
+ {
+ "epoch": 1.85092916778885,
+ "grad_norm": 0.767853195090442,
+ "learning_rate": 1.3657469690815106e-07,
+ "loss": 0.28401851654052734,
+ "num_input_tokens_seen": 1202246288,
+ "step": 27490,
+ "train_runtime": 212671.529,
+ "train_tokens_per_second": 5653.066
+ },
+ {
+ "epoch": 1.8512658227848102,
+ "grad_norm": 0.7219513957308681,
+ "learning_rate": 1.359616093791677e-07,
+ "loss": 0.3170768737792969,
+ "num_input_tokens_seen": 1202478016,
+ "step": 27495,
+ "train_runtime": 212715.4107,
+ "train_tokens_per_second": 5652.99
+ },
+ {
+ "epoch": 1.8516024777807703,
+ "grad_norm": 0.7333849781257291,
+ "learning_rate": 1.353498820630106e-07,
+ "loss": 0.29583718776702883,
+ "num_input_tokens_seen": 1202698624,
+ "step": 27500,
+ "train_runtime": 212754.0646,
+ "train_tokens_per_second": 5653.0
+ },
+ {
+ "epoch": 1.8519391327767303,
+ "grad_norm": 0.7743102457437921,
+ "learning_rate": 1.3473951513074625e-07,
+ "loss": 0.31031112670898436,
+ "num_input_tokens_seen": 1202918504,
+ "step": 27505,
+ "train_runtime": 212795.3126,
+ "train_tokens_per_second": 5652.937
+ },
+ {
+ "epoch": 1.8522757877726905,
+ "grad_norm": 0.9893994757413681,
+ "learning_rate": 1.341305087530631e-07,
+ "loss": 0.29344775676727297,
+ "num_input_tokens_seen": 1203135384,
+ "step": 27510,
+ "train_runtime": 212826.4122,
+ "train_tokens_per_second": 5653.13
+ },
+ {
+ "epoch": 1.8526124427686508,
+ "grad_norm": 0.8824830489862262,
+ "learning_rate": 1.3352286310026663e-07,
+ "loss": 0.3105172634124756,
+ "num_input_tokens_seen": 1203348136,
+ "step": 27515,
+ "train_runtime": 212864.5316,
+ "train_tokens_per_second": 5653.117
+ },
+ {
+ "epoch": 1.8529490977646108,
+ "grad_norm": 0.8177611033215039,
+ "learning_rate": 1.329165783422831e-07,
+ "loss": 0.31962890625,
+ "num_input_tokens_seen": 1203560776,
+ "step": 27520,
+ "train_runtime": 212898.9135,
+ "train_tokens_per_second": 5653.203
+ },
+ {
+ "epoch": 1.8532857527605708,
+ "grad_norm": 0.8203874085361618,
+ "learning_rate": 1.3231165464865914e-07,
+ "loss": 0.29946393966674806,
+ "num_input_tokens_seen": 1203778736,
+ "step": 27525,
+ "train_runtime": 212929.6442,
+ "train_tokens_per_second": 5653.411
+ },
+ {
+ "epoch": 1.853622407756531,
+ "grad_norm": 0.6941728137863973,
+ "learning_rate": 1.3170809218855884e-07,
+ "loss": 0.29173548221588136,
+ "num_input_tokens_seen": 1204009952,
+ "step": 27530,
+ "train_runtime": 212966.9944,
+ "train_tokens_per_second": 5653.505
+ },
+ {
+ "epoch": 1.8539590627524913,
+ "grad_norm": 0.7342203887402693,
+ "learning_rate": 1.3110589113076554e-07,
+ "loss": 0.2878772258758545,
+ "num_input_tokens_seen": 1204230200,
+ "step": 27535,
+ "train_runtime": 213008.2192,
+ "train_tokens_per_second": 5653.445
+ },
+ {
+ "epoch": 1.8542957177484514,
+ "grad_norm": 0.7315091280926438,
+ "learning_rate": 1.3050505164368566e-07,
+ "loss": 0.29323248863220214,
+ "num_input_tokens_seen": 1204451192,
+ "step": 27540,
+ "train_runtime": 213039.3969,
+ "train_tokens_per_second": 5653.655
+ },
+ {
+ "epoch": 1.8546323727444114,
+ "grad_norm": 0.8014357486751242,
+ "learning_rate": 1.299055738953392e-07,
+ "loss": 0.2807599067687988,
+ "num_input_tokens_seen": 1204650768,
+ "step": 27545,
+ "train_runtime": 213080.939,
+ "train_tokens_per_second": 5653.489
+ },
+ {
+ "epoch": 1.8549690277403716,
+ "grad_norm": 0.7816529327087509,
+ "learning_rate": 1.293074580533704e-07,
+ "loss": 0.28776376247406005,
+ "num_input_tokens_seen": 1204864152,
+ "step": 27550,
+ "train_runtime": 213127.338,
+ "train_tokens_per_second": 5653.26
+ },
+ {
+ "epoch": 1.855305682736332,
+ "grad_norm": 0.7431632363177251,
+ "learning_rate": 1.2871070428503885e-07,
+ "loss": 0.3120842456817627,
+ "num_input_tokens_seen": 1205072184,
+ "step": 27555,
+ "train_runtime": 213166.7662,
+ "train_tokens_per_second": 5653.19
+ },
+ {
+ "epoch": 1.855642337732292,
+ "grad_norm": 0.6740193343425465,
+ "learning_rate": 1.2811531275722543e-07,
+ "loss": 0.3159989595413208,
+ "num_input_tokens_seen": 1205302840,
+ "step": 27560,
+ "train_runtime": 213202.2706,
+ "train_tokens_per_second": 5653.33
+ },
+ {
+ "epoch": 1.855978992728252,
+ "grad_norm": 0.850999270296055,
+ "learning_rate": 1.275212836364298e-07,
+ "loss": 0.30814902782440184,
+ "num_input_tokens_seen": 1205521608,
+ "step": 27565,
+ "train_runtime": 213246.1453,
+ "train_tokens_per_second": 5653.193
+ },
+ {
+ "epoch": 1.8563156477242122,
+ "grad_norm": 0.7919555393902858,
+ "learning_rate": 1.2692861708877014e-07,
+ "loss": 0.28333234786987305,
+ "num_input_tokens_seen": 1205730248,
+ "step": 27570,
+ "train_runtime": 213289.0034,
+ "train_tokens_per_second": 5653.035
+ },
+ {
+ "epoch": 1.8566523027201725,
+ "grad_norm": 0.7321483909457048,
+ "learning_rate": 1.2633731327998332e-07,
+ "loss": 0.2803944110870361,
+ "num_input_tokens_seen": 1205950784,
+ "step": 27575,
+ "train_runtime": 213322.9311,
+ "train_tokens_per_second": 5653.17
+ },
+ {
+ "epoch": 1.8569889577161325,
+ "grad_norm": 0.9024845578541033,
+ "learning_rate": 1.25747372375426e-07,
+ "loss": 0.2885521411895752,
+ "num_input_tokens_seen": 1206159224,
+ "step": 27580,
+ "train_runtime": 213368.2449,
+ "train_tokens_per_second": 5652.946
+ },
+ {
+ "epoch": 1.8573256127120925,
+ "grad_norm": 0.8389667628017906,
+ "learning_rate": 1.251587945400723e-07,
+ "loss": 0.3000046253204346,
+ "num_input_tokens_seen": 1206379280,
+ "step": 27585,
+ "train_runtime": 213408.9939,
+ "train_tokens_per_second": 5652.898
+ },
+ {
+ "epoch": 1.8576622677080528,
+ "grad_norm": 0.7189413123955991,
+ "learning_rate": 1.2457157993851832e-07,
+ "loss": 0.2905135154724121,
+ "num_input_tokens_seen": 1206604064,
+ "step": 27590,
+ "train_runtime": 213446.507,
+ "train_tokens_per_second": 5652.958
+ },
+ {
+ "epoch": 1.857998922704013,
+ "grad_norm": 0.756958619394132,
+ "learning_rate": 1.2398572873497438e-07,
+ "loss": 0.2975008487701416,
+ "num_input_tokens_seen": 1206821040,
+ "step": 27595,
+ "train_runtime": 213487.2557,
+ "train_tokens_per_second": 5652.895
+ },
+ {
+ "epoch": 1.858335577699973,
+ "grad_norm": 0.7936162080166628,
+ "learning_rate": 1.2340124109327334e-07,
+ "loss": 0.2905877590179443,
+ "num_input_tokens_seen": 1207029960,
+ "step": 27600,
+ "train_runtime": 213526.3845,
+ "train_tokens_per_second": 5652.838
+ },
+ {
+ "epoch": 1.858672232695933,
+ "grad_norm": 0.7617351149236978,
+ "learning_rate": 1.22818117176865e-07,
+ "loss": 0.3129683494567871,
+ "num_input_tokens_seen": 1207251408,
+ "step": 27605,
+ "train_runtime": 213571.4503,
+ "train_tokens_per_second": 5652.682
+ },
+ {
+ "epoch": 1.8590088876918933,
+ "grad_norm": 0.6863825753058197,
+ "learning_rate": 1.2223635714881777e-07,
+ "loss": 0.30173707008361816,
+ "num_input_tokens_seen": 1207460752,
+ "step": 27610,
+ "train_runtime": 213606.2956,
+ "train_tokens_per_second": 5652.74
+ },
+ {
+ "epoch": 1.8593455426878536,
+ "grad_norm": 0.6602683521069512,
+ "learning_rate": 1.2165596117181931e-07,
+ "loss": 0.28788735866546633,
+ "num_input_tokens_seen": 1207688200,
+ "step": 27615,
+ "train_runtime": 213640.7332,
+ "train_tokens_per_second": 5652.893
+ },
+ {
+ "epoch": 1.8596821976838136,
+ "grad_norm": 0.7244916772239959,
+ "learning_rate": 1.2107692940817594e-07,
+ "loss": 0.30665135383605957,
+ "num_input_tokens_seen": 1207904320,
+ "step": 27620,
+ "train_runtime": 213682.5989,
+ "train_tokens_per_second": 5652.797
+ },
+ {
+ "epoch": 1.8600188526797736,
+ "grad_norm": 0.8628325975987942,
+ "learning_rate": 1.2049926201981088e-07,
+ "loss": 0.2934282302856445,
+ "num_input_tokens_seen": 1208124344,
+ "step": 27625,
+ "train_runtime": 213723.4491,
+ "train_tokens_per_second": 5652.746
+ },
+ {
+ "epoch": 1.8603555076757339,
+ "grad_norm": 0.719201813819039,
+ "learning_rate": 1.199229591682677e-07,
+ "loss": 0.3092188835144043,
+ "num_input_tokens_seen": 1208348632,
+ "step": 27630,
+ "train_runtime": 213757.6689,
+ "train_tokens_per_second": 5652.89
+ },
+ {
+ "epoch": 1.8606921626716941,
+ "grad_norm": 0.8213286351201897,
+ "learning_rate": 1.193480210147069e-07,
+ "loss": 0.3101773262023926,
+ "num_input_tokens_seen": 1208576560,
+ "step": 27635,
+ "train_runtime": 213793.1342,
+ "train_tokens_per_second": 5653.019
+ },
+ {
+ "epoch": 1.8610288176676542,
+ "grad_norm": 0.7546904535243726,
+ "learning_rate": 1.1877444771991042e-07,
+ "loss": 0.2830918788909912,
+ "num_input_tokens_seen": 1208805448,
+ "step": 27640,
+ "train_runtime": 213830.9044,
+ "train_tokens_per_second": 5653.09
+ },
+ {
+ "epoch": 1.8613654726636142,
+ "grad_norm": 0.7654042884672665,
+ "learning_rate": 1.1820223944427279e-07,
+ "loss": 0.28035907745361327,
+ "num_input_tokens_seen": 1209015368,
+ "step": 27645,
+ "train_runtime": 213873.433,
+ "train_tokens_per_second": 5652.948
+ },
+ {
+ "epoch": 1.8617021276595744,
+ "grad_norm": 0.7531392542074884,
+ "learning_rate": 1.1763139634781262e-07,
+ "loss": 0.29651951789855957,
+ "num_input_tokens_seen": 1209240952,
+ "step": 27650,
+ "train_runtime": 213911.0698,
+ "train_tokens_per_second": 5653.008
+ },
+ {
+ "epoch": 1.8620387826555347,
+ "grad_norm": 0.7502735387914459,
+ "learning_rate": 1.1706191859016392e-07,
+ "loss": 0.3263824224472046,
+ "num_input_tokens_seen": 1209461064,
+ "step": 27655,
+ "train_runtime": 213953.9938,
+ "train_tokens_per_second": 5652.902
+ },
+ {
+ "epoch": 1.8623754376514947,
+ "grad_norm": 0.8331137280204224,
+ "learning_rate": 1.1649380633057927e-07,
+ "loss": 0.3037287950515747,
+ "num_input_tokens_seen": 1209677248,
+ "step": 27660,
+ "train_runtime": 213997.5656,
+ "train_tokens_per_second": 5652.762
+ },
+ {
+ "epoch": 1.8627120926474547,
+ "grad_norm": 0.7575852346311494,
+ "learning_rate": 1.1592705972792939e-07,
+ "loss": 0.27420315742492674,
+ "num_input_tokens_seen": 1209892576,
+ "step": 27665,
+ "train_runtime": 214039.3874,
+ "train_tokens_per_second": 5652.663
+ },
+ {
+ "epoch": 1.863048747643415,
+ "grad_norm": 0.7320879410064924,
+ "learning_rate": 1.1536167894070249e-07,
+ "loss": 0.3063500165939331,
+ "num_input_tokens_seen": 1210118104,
+ "step": 27670,
+ "train_runtime": 214082.5531,
+ "train_tokens_per_second": 5652.577
+ },
+ {
+ "epoch": 1.8633854026393752,
+ "grad_norm": 0.7458641244137307,
+ "learning_rate": 1.1479766412700655e-07,
+ "loss": 0.33033363819122313,
+ "num_input_tokens_seen": 1210345472,
+ "step": 27675,
+ "train_runtime": 214121.2922,
+ "train_tokens_per_second": 5652.616
+ },
+ {
+ "epoch": 1.8637220576353353,
+ "grad_norm": 0.7595248981178473,
+ "learning_rate": 1.1423501544456595e-07,
+ "loss": 0.31739757061004636,
+ "num_input_tokens_seen": 1210558512,
+ "step": 27680,
+ "train_runtime": 214156.2124,
+ "train_tokens_per_second": 5652.689
+ },
+ {
+ "epoch": 1.8640587126312953,
+ "grad_norm": 0.7625771638849843,
+ "learning_rate": 1.1367373305072315e-07,
+ "loss": 0.2853879451751709,
+ "num_input_tokens_seen": 1210779944,
+ "step": 27685,
+ "train_runtime": 214192.52,
+ "train_tokens_per_second": 5652.765
+ },
+ {
+ "epoch": 1.8643953676272556,
+ "grad_norm": 0.7246559600061148,
+ "learning_rate": 1.1311381710244096e-07,
+ "loss": 0.2670130729675293,
+ "num_input_tokens_seen": 1211003328,
+ "step": 27690,
+ "train_runtime": 214233.2878,
+ "train_tokens_per_second": 5652.732
+ },
+ {
+ "epoch": 1.8647320226232158,
+ "grad_norm": 0.8080414702851416,
+ "learning_rate": 1.1255526775629521e-07,
+ "loss": 0.27195014953613283,
+ "num_input_tokens_seen": 1211215040,
+ "step": 27695,
+ "train_runtime": 214278.5375,
+ "train_tokens_per_second": 5652.526
+ },
+ {
+ "epoch": 1.8650686776191758,
+ "grad_norm": 0.7987137339242992,
+ "learning_rate": 1.1199808516848487e-07,
+ "loss": 0.30287680625915525,
+ "num_input_tokens_seen": 1211429152,
+ "step": 27700,
+ "train_runtime": 214324.0073,
+ "train_tokens_per_second": 5652.326
+ },
+ {
+ "epoch": 1.8654053326151359,
+ "grad_norm": 0.7407863475801888,
+ "learning_rate": 1.1144226949482306e-07,
+ "loss": 0.28324642181396487,
+ "num_input_tokens_seen": 1211638376,
+ "step": 27705,
+ "train_runtime": 214360.0889,
+ "train_tokens_per_second": 5652.351
+ },
+ {
+ "epoch": 1.865741987611096,
+ "grad_norm": 0.6793586684952472,
+ "learning_rate": 1.1088782089074157e-07,
+ "loss": 0.27295451164245604,
+ "num_input_tokens_seen": 1211854440,
+ "step": 27710,
+ "train_runtime": 214400.1883,
+ "train_tokens_per_second": 5652.301
+ },
+ {
+ "epoch": 1.8660786426070564,
+ "grad_norm": 0.8291474455329964,
+ "learning_rate": 1.1033473951129137e-07,
+ "loss": 0.3024445056915283,
+ "num_input_tokens_seen": 1212077896,
+ "step": 27715,
+ "train_runtime": 214434.9545,
+ "train_tokens_per_second": 5652.427
+ },
+ {
+ "epoch": 1.8664152976030164,
+ "grad_norm": 0.7458532162025366,
+ "learning_rate": 1.0978302551113929e-07,
+ "loss": 0.3134200096130371,
+ "num_input_tokens_seen": 1212305392,
+ "step": 27720,
+ "train_runtime": 214476.0312,
+ "train_tokens_per_second": 5652.405
+ },
+ {
+ "epoch": 1.8667519525989764,
+ "grad_norm": 0.7681313743079988,
+ "learning_rate": 1.0923267904457025e-07,
+ "loss": 0.30133209228515623,
+ "num_input_tokens_seen": 1212533032,
+ "step": 27725,
+ "train_runtime": 214516.8657,
+ "train_tokens_per_second": 5652.39
+ },
+ {
+ "epoch": 1.8670886075949367,
+ "grad_norm": 0.7916528062724383,
+ "learning_rate": 1.086837002654867e-07,
+ "loss": 0.30214452743530273,
+ "num_input_tokens_seen": 1212743440,
+ "step": 27730,
+ "train_runtime": 214558.688,
+ "train_tokens_per_second": 5652.269
+ },
+ {
+ "epoch": 1.867425262590897,
+ "grad_norm": 0.7468188583517545,
+ "learning_rate": 1.0813608932740915e-07,
+ "loss": 0.3034868001937866,
+ "num_input_tokens_seen": 1212977168,
+ "step": 27735,
+ "train_runtime": 214594.7214,
+ "train_tokens_per_second": 5652.409
+ },
+ {
+ "epoch": 1.867761917586857,
+ "grad_norm": 0.7831956841711861,
+ "learning_rate": 1.0758984638347569e-07,
+ "loss": 0.2956989765167236,
+ "num_input_tokens_seen": 1213207904,
+ "step": 27740,
+ "train_runtime": 214629.3326,
+ "train_tokens_per_second": 5652.573
+ },
+ {
+ "epoch": 1.868098572582817,
+ "grad_norm": 0.7955988160557622,
+ "learning_rate": 1.0704497158644022e-07,
+ "loss": 0.32259302139282225,
+ "num_input_tokens_seen": 1213425128,
+ "step": 27745,
+ "train_runtime": 214673.9293,
+ "train_tokens_per_second": 5652.41
+ },
+ {
+ "epoch": 1.8684352275787772,
+ "grad_norm": 0.6842277034050441,
+ "learning_rate": 1.0650146508867642e-07,
+ "loss": 0.2915182590484619,
+ "num_input_tokens_seen": 1213645624,
+ "step": 27750,
+ "train_runtime": 214712.349,
+ "train_tokens_per_second": 5652.426
+ },
+ {
+ "epoch": 1.8687718825747375,
+ "grad_norm": 0.7772804158541299,
+ "learning_rate": 1.0595932704217327e-07,
+ "loss": 0.2860418796539307,
+ "num_input_tokens_seen": 1213868720,
+ "step": 27755,
+ "train_runtime": 214747.403,
+ "train_tokens_per_second": 5652.542
+ },
+ {
+ "epoch": 1.8691085375706975,
+ "grad_norm": 0.7387951063630381,
+ "learning_rate": 1.0541855759853837e-07,
+ "loss": 0.2798731327056885,
+ "num_input_tokens_seen": 1214099608,
+ "step": 27760,
+ "train_runtime": 214776.3703,
+ "train_tokens_per_second": 5652.855
+ },
+ {
+ "epoch": 1.8694451925666575,
+ "grad_norm": 0.6909728784946527,
+ "learning_rate": 1.0487915690899685e-07,
+ "loss": 0.29963858127593995,
+ "num_input_tokens_seen": 1214319216,
+ "step": 27765,
+ "train_runtime": 214814.2285,
+ "train_tokens_per_second": 5652.881
+ },
+ {
+ "epoch": 1.8697818475626178,
+ "grad_norm": 0.7184706301388402,
+ "learning_rate": 1.0434112512438921e-07,
+ "loss": 0.29176318645477295,
+ "num_input_tokens_seen": 1214533368,
+ "step": 27770,
+ "train_runtime": 214852.4866,
+ "train_tokens_per_second": 5652.871
+ },
+ {
+ "epoch": 1.870118502558578,
+ "grad_norm": 0.8307447666812611,
+ "learning_rate": 1.0380446239517506e-07,
+ "loss": 0.2727836608886719,
+ "num_input_tokens_seen": 1214730928,
+ "step": 27775,
+ "train_runtime": 214893.9185,
+ "train_tokens_per_second": 5652.7
+ },
+ {
+ "epoch": 1.870455157554538,
+ "grad_norm": 0.8916967588803947,
+ "learning_rate": 1.032691688714299e-07,
+ "loss": 0.2895357608795166,
+ "num_input_tokens_seen": 1214918680,
+ "step": 27780,
+ "train_runtime": 214939.4108,
+ "train_tokens_per_second": 5652.377
+ },
+ {
+ "epoch": 1.870791812550498,
+ "grad_norm": 0.7913603727738847,
+ "learning_rate": 1.0273524470284734e-07,
+ "loss": 0.3172266483306885,
+ "num_input_tokens_seen": 1215146136,
+ "step": 27785,
+ "train_runtime": 214976.005,
+ "train_tokens_per_second": 5652.473
+ },
+ {
+ "epoch": 1.8711284675464583,
+ "grad_norm": 0.9142096797553524,
+ "learning_rate": 1.0220269003873739e-07,
+ "loss": 0.29100470542907714,
+ "num_input_tokens_seen": 1215365024,
+ "step": 27790,
+ "train_runtime": 215011.3451,
+ "train_tokens_per_second": 5652.562
+ },
+ {
+ "epoch": 1.8714651225424186,
+ "grad_norm": 0.8041418730532507,
+ "learning_rate": 1.0167150502802703e-07,
+ "loss": 0.2920061111450195,
+ "num_input_tokens_seen": 1215587488,
+ "step": 27795,
+ "train_runtime": 215049.5211,
+ "train_tokens_per_second": 5652.593
+ },
+ {
+ "epoch": 1.8718017775383786,
+ "grad_norm": 0.7464357375845391,
+ "learning_rate": 1.0114168981926075e-07,
+ "loss": 0.31276350021362304,
+ "num_input_tokens_seen": 1215810000,
+ "step": 27800,
+ "train_runtime": 215091.2811,
+ "train_tokens_per_second": 5652.53
+ },
+ {
+ "epoch": 1.8721384325343386,
+ "grad_norm": 0.7898446735642499,
+ "learning_rate": 1.006132445605995e-07,
+ "loss": 0.3080112457275391,
+ "num_input_tokens_seen": 1216014256,
+ "step": 27805,
+ "train_runtime": 215124.1859,
+ "train_tokens_per_second": 5652.615
+ },
+ {
+ "epoch": 1.872475087530299,
+ "grad_norm": 0.7329730105281138,
+ "learning_rate": 1.0008616939982118e-07,
+ "loss": 0.28853521347045896,
+ "num_input_tokens_seen": 1216231488,
+ "step": 27810,
+ "train_runtime": 215162.2144,
+ "train_tokens_per_second": 5652.626
+ },
+ {
+ "epoch": 1.8728117425262591,
+ "grad_norm": 0.7289363854898081,
+ "learning_rate": 9.956046448432177e-08,
+ "loss": 0.283016300201416,
+ "num_input_tokens_seen": 1216451408,
+ "step": 27815,
+ "train_runtime": 215198.9756,
+ "train_tokens_per_second": 5652.682
+ },
+ {
+ "epoch": 1.8731483975222192,
+ "grad_norm": 0.7391182483442761,
+ "learning_rate": 9.90361299611109e-08,
+ "loss": 0.30707416534423826,
+ "num_input_tokens_seen": 1216680872,
+ "step": 27820,
+ "train_runtime": 215231.7273,
+ "train_tokens_per_second": 5652.888
+ },
+ {
+ "epoch": 1.8734850525181792,
+ "grad_norm": 0.8057353688260167,
+ "learning_rate": 9.851316597681959e-08,
+ "loss": 0.3057239055633545,
+ "num_input_tokens_seen": 1216892024,
+ "step": 27825,
+ "train_runtime": 215268.9874,
+ "train_tokens_per_second": 5652.891
+ },
+ {
+ "epoch": 1.8738217075141395,
+ "grad_norm": 0.8107611348319113,
+ "learning_rate": 9.799157267769088e-08,
+ "loss": 0.3025531768798828,
+ "num_input_tokens_seen": 1217114336,
+ "step": 27830,
+ "train_runtime": 215300.5052,
+ "train_tokens_per_second": 5653.096
+ },
+ {
+ "epoch": 1.8741583625100997,
+ "grad_norm": 0.6867341450134844,
+ "learning_rate": 9.747135020958753e-08,
+ "loss": 0.2927044153213501,
+ "num_input_tokens_seen": 1217335976,
+ "step": 27835,
+ "train_runtime": 215340.9059,
+ "train_tokens_per_second": 5653.064
+ },
+ {
+ "epoch": 1.8744950175060597,
+ "grad_norm": 0.7604489307015202,
+ "learning_rate": 9.695249871798818e-08,
+ "loss": 0.2803203582763672,
+ "num_input_tokens_seen": 1217552464,
+ "step": 27840,
+ "train_runtime": 215388.9005,
+ "train_tokens_per_second": 5652.81
+ },
+ {
+ "epoch": 1.8748316725020198,
+ "grad_norm": 0.7584373162773101,
+ "learning_rate": 9.643501834798786e-08,
+ "loss": 0.30305957794189453,
+ "num_input_tokens_seen": 1217765392,
+ "step": 27845,
+ "train_runtime": 215429.0423,
+ "train_tokens_per_second": 5652.745
+ },
+ {
+ "epoch": 1.8751683274979802,
+ "grad_norm": 0.7728951165449691,
+ "learning_rate": 9.591890924429858e-08,
+ "loss": 0.2971980094909668,
+ "num_input_tokens_seen": 1218000984,
+ "step": 27850,
+ "train_runtime": 215470.184,
+ "train_tokens_per_second": 5652.759
+ },
+ {
+ "epoch": 1.8755049824939403,
+ "grad_norm": 0.7960632667781874,
+ "learning_rate": 9.540417155124826e-08,
+ "loss": 0.278977370262146,
+ "num_input_tokens_seen": 1218198848,
+ "step": 27855,
+ "train_runtime": 215509.0395,
+ "train_tokens_per_second": 5652.658
+ },
+ {
+ "epoch": 1.8758416374899003,
+ "grad_norm": 0.73841359058749,
+ "learning_rate": 9.489080541278117e-08,
+ "loss": 0.2986736536026001,
+ "num_input_tokens_seen": 1218407032,
+ "step": 27860,
+ "train_runtime": 215554.3706,
+ "train_tokens_per_second": 5652.435
+ },
+ {
+ "epoch": 1.8761782924858605,
+ "grad_norm": 0.7691921305890641,
+ "learning_rate": 9.437881097246026e-08,
+ "loss": 0.3069639205932617,
+ "num_input_tokens_seen": 1218632248,
+ "step": 27865,
+ "train_runtime": 215593.7195,
+ "train_tokens_per_second": 5652.448
+ },
+ {
+ "epoch": 1.8765149474818208,
+ "grad_norm": 0.7084528384556897,
+ "learning_rate": 9.3868188373461e-08,
+ "loss": 0.2972809076309204,
+ "num_input_tokens_seen": 1218846656,
+ "step": 27870,
+ "train_runtime": 215628.931,
+ "train_tokens_per_second": 5652.519
+ },
+ {
+ "epoch": 1.8768516024777808,
+ "grad_norm": 0.8347410289645919,
+ "learning_rate": 9.335893775857918e-08,
+ "loss": 0.3106103420257568,
+ "num_input_tokens_seen": 1219056688,
+ "step": 27875,
+ "train_runtime": 215663.948,
+ "train_tokens_per_second": 5652.575
+ },
+ {
+ "epoch": 1.8771882574737409,
+ "grad_norm": 0.7646297115981436,
+ "learning_rate": 9.285105927022309e-08,
+ "loss": 0.28227670192718507,
+ "num_input_tokens_seen": 1219278728,
+ "step": 27880,
+ "train_runtime": 215702.0527,
+ "train_tokens_per_second": 5652.606
+ },
+ {
+ "epoch": 1.877524912469701,
+ "grad_norm": 0.8949871913293714,
+ "learning_rate": 9.234455305042133e-08,
+ "loss": 0.2822525978088379,
+ "num_input_tokens_seen": 1219485440,
+ "step": 27885,
+ "train_runtime": 215750.0835,
+ "train_tokens_per_second": 5652.306
+ },
+ {
+ "epoch": 1.8778615674656614,
+ "grad_norm": 0.8356384310040774,
+ "learning_rate": 9.183941924081563e-08,
+ "loss": 0.2904726266860962,
+ "num_input_tokens_seen": 1219702416,
+ "step": 27890,
+ "train_runtime": 215787.36,
+ "train_tokens_per_second": 5652.335
+ },
+ {
+ "epoch": 1.8781982224616214,
+ "grad_norm": 0.7934445345932645,
+ "learning_rate": 9.133565798266574e-08,
+ "loss": 0.2883297443389893,
+ "num_input_tokens_seen": 1219921880,
+ "step": 27895,
+ "train_runtime": 215820.3228,
+ "train_tokens_per_second": 5652.488
+ },
+ {
+ "epoch": 1.8785348774575814,
+ "grad_norm": 0.783421786892747,
+ "learning_rate": 9.083326941684622e-08,
+ "loss": 0.3204617500305176,
+ "num_input_tokens_seen": 1220148280,
+ "step": 27900,
+ "train_runtime": 215855.8328,
+ "train_tokens_per_second": 5652.607
+ },
+ {
+ "epoch": 1.8788715324535417,
+ "grad_norm": 0.7556141302042848,
+ "learning_rate": 9.033225368384912e-08,
+ "loss": 0.3104442358016968,
+ "num_input_tokens_seen": 1220358200,
+ "step": 27905,
+ "train_runtime": 215897.0384,
+ "train_tokens_per_second": 5652.501
+ },
+ {
+ "epoch": 1.879208187449502,
+ "grad_norm": 0.6646972540097822,
+ "learning_rate": 8.98326109237807e-08,
+ "loss": 0.27222867012023927,
+ "num_input_tokens_seen": 1220576488,
+ "step": 27910,
+ "train_runtime": 215939.5159,
+ "train_tokens_per_second": 5652.4
+ },
+ {
+ "epoch": 1.879544842445462,
+ "grad_norm": 0.73208942270682,
+ "learning_rate": 8.933434127636586e-08,
+ "loss": 0.31929314136505127,
+ "num_input_tokens_seen": 1220801344,
+ "step": 27915,
+ "train_runtime": 215983.142,
+ "train_tokens_per_second": 5652.299
+ },
+ {
+ "epoch": 1.879881497441422,
+ "grad_norm": 0.7379086364893702,
+ "learning_rate": 8.883744488094314e-08,
+ "loss": 0.3021604061126709,
+ "num_input_tokens_seen": 1221012952,
+ "step": 27920,
+ "train_runtime": 216020.1906,
+ "train_tokens_per_second": 5652.309
+ },
+ {
+ "epoch": 1.8802181524373822,
+ "grad_norm": 0.7315736019754095,
+ "learning_rate": 8.834192187646917e-08,
+ "loss": 0.30646262168884275,
+ "num_input_tokens_seen": 1221243752,
+ "step": 27925,
+ "train_runtime": 216054.4565,
+ "train_tokens_per_second": 5652.481
+ },
+ {
+ "epoch": 1.8805548074333425,
+ "grad_norm": 0.8178242002248947,
+ "learning_rate": 8.784777240151365e-08,
+ "loss": 0.30849490165710447,
+ "num_input_tokens_seen": 1221469512,
+ "step": 27930,
+ "train_runtime": 216085.3674,
+ "train_tokens_per_second": 5652.717
+ },
+ {
+ "epoch": 1.8808914624293025,
+ "grad_norm": 0.745298782226071,
+ "learning_rate": 8.73549965942655e-08,
+ "loss": 0.3049221754074097,
+ "num_input_tokens_seen": 1221695328,
+ "step": 27935,
+ "train_runtime": 216123.8064,
+ "train_tokens_per_second": 5652.757
+ },
+ {
+ "epoch": 1.8812281174252625,
+ "grad_norm": 0.7296969049019549,
+ "learning_rate": 8.686359459252725e-08,
+ "loss": 0.29515538215637205,
+ "num_input_tokens_seen": 1221906744,
+ "step": 27940,
+ "train_runtime": 216162.3363,
+ "train_tokens_per_second": 5652.727
+ },
+ {
+ "epoch": 1.8815647724212228,
+ "grad_norm": 0.6723805170647249,
+ "learning_rate": 8.637356653371787e-08,
+ "loss": 0.28774356842041016,
+ "num_input_tokens_seen": 1222134512,
+ "step": 27945,
+ "train_runtime": 216197.844,
+ "train_tokens_per_second": 5652.852
+ },
+ {
+ "epoch": 1.881901427417183,
+ "grad_norm": 0.8473708457816271,
+ "learning_rate": 8.588491255487163e-08,
+ "loss": 0.2720442056655884,
+ "num_input_tokens_seen": 1222344728,
+ "step": 27950,
+ "train_runtime": 216239.4622,
+ "train_tokens_per_second": 5652.737
+ },
+ {
+ "epoch": 1.882238082413143,
+ "grad_norm": 0.7624946480602264,
+ "learning_rate": 8.539763279264035e-08,
+ "loss": 0.2836963176727295,
+ "num_input_tokens_seen": 1222560856,
+ "step": 27955,
+ "train_runtime": 216284.8082,
+ "train_tokens_per_second": 5652.551
+ },
+ {
+ "epoch": 1.882574737409103,
+ "grad_norm": 0.7053326289175388,
+ "learning_rate": 8.49117273832889e-08,
+ "loss": 0.3163050651550293,
+ "num_input_tokens_seen": 1222789864,
+ "step": 27960,
+ "train_runtime": 216318.1103,
+ "train_tokens_per_second": 5652.739
+ },
+ {
+ "epoch": 1.8829113924050633,
+ "grad_norm": 0.6499749519322318,
+ "learning_rate": 8.442719646270082e-08,
+ "loss": 0.28265206813812255,
+ "num_input_tokens_seen": 1223005096,
+ "step": 27965,
+ "train_runtime": 216357.5685,
+ "train_tokens_per_second": 5652.703
+ },
+ {
+ "epoch": 1.8832480474010236,
+ "grad_norm": 0.7141334939696597,
+ "learning_rate": 8.394404016637159e-08,
+ "loss": 0.28957691192626955,
+ "num_input_tokens_seen": 1223223864,
+ "step": 27970,
+ "train_runtime": 216399.1348,
+ "train_tokens_per_second": 5652.628
+ },
+ {
+ "epoch": 1.8835847023969836,
+ "grad_norm": 0.7972265729262933,
+ "learning_rate": 8.346225862941592e-08,
+ "loss": 0.28935163021087645,
+ "num_input_tokens_seen": 1223430136,
+ "step": 27975,
+ "train_runtime": 216434.0905,
+ "train_tokens_per_second": 5652.668
+ },
+ {
+ "epoch": 1.8839213573929436,
+ "grad_norm": 0.8292085959022567,
+ "learning_rate": 8.298185198656161e-08,
+ "loss": 0.31040134429931643,
+ "num_input_tokens_seen": 1223661888,
+ "step": 27980,
+ "train_runtime": 216475.8962,
+ "train_tokens_per_second": 5652.647
+ },
+ {
+ "epoch": 1.884258012388904,
+ "grad_norm": 0.7497952405897995,
+ "learning_rate": 8.25028203721534e-08,
+ "loss": 0.2839097023010254,
+ "num_input_tokens_seen": 1223872392,
+ "step": 27985,
+ "train_runtime": 216516.2523,
+ "train_tokens_per_second": 5652.566
+ },
+ {
+ "epoch": 1.8845946673848641,
+ "grad_norm": 0.8204275222628281,
+ "learning_rate": 8.202516392015081e-08,
+ "loss": 0.30760276317596436,
+ "num_input_tokens_seen": 1224095496,
+ "step": 27990,
+ "train_runtime": 216553.2529,
+ "train_tokens_per_second": 5652.63
+ },
+ {
+ "epoch": 1.8849313223808242,
+ "grad_norm": 0.9261868943922493,
+ "learning_rate": 8.154888276412865e-08,
+ "loss": 0.31410627365112304,
+ "num_input_tokens_seen": 1224301616,
+ "step": 27995,
+ "train_runtime": 216587.9714,
+ "train_tokens_per_second": 5652.676
+ },
+ {
+ "epoch": 1.8852679773767842,
+ "grad_norm": 0.7105028791923415,
+ "learning_rate": 8.10739770372776e-08,
+ "loss": 0.2736332416534424,
+ "num_input_tokens_seen": 1224525064,
+ "step": 28000,
+ "train_runtime": 216624.6668,
+ "train_tokens_per_second": 5652.75
+ },
+ {
+ "epoch": 1.8856046323727444,
+ "grad_norm": 0.6544014559458133,
+ "learning_rate": 8.060044687240365e-08,
+ "loss": 0.2702982425689697,
+ "num_input_tokens_seen": 1224764712,
+ "step": 28005,
+ "train_runtime": 216655.2066,
+ "train_tokens_per_second": 5653.059
+ },
+ {
+ "epoch": 1.8859412873687047,
+ "grad_norm": 0.7293503705566087,
+ "learning_rate": 8.012829240192754e-08,
+ "loss": 0.31306607723236085,
+ "num_input_tokens_seen": 1224979848,
+ "step": 28010,
+ "train_runtime": 216695.9204,
+ "train_tokens_per_second": 5652.99
+ },
+ {
+ "epoch": 1.8862779423646647,
+ "grad_norm": 0.7181373637032492,
+ "learning_rate": 7.9657513757887e-08,
+ "loss": 0.2802762985229492,
+ "num_input_tokens_seen": 1225200744,
+ "step": 28015,
+ "train_runtime": 216735.8221,
+ "train_tokens_per_second": 5652.968
+ },
+ {
+ "epoch": 1.8866145973606248,
+ "grad_norm": 0.753661015701672,
+ "learning_rate": 7.918811107193225e-08,
+ "loss": 0.28357934951782227,
+ "num_input_tokens_seen": 1225430328,
+ "step": 28020,
+ "train_runtime": 216771.4604,
+ "train_tokens_per_second": 5653.098
+ },
+ {
+ "epoch": 1.886951252356585,
+ "grad_norm": 0.8998847487777263,
+ "learning_rate": 7.872008447533109e-08,
+ "loss": 0.3160097122192383,
+ "num_input_tokens_seen": 1225648792,
+ "step": 28025,
+ "train_runtime": 216808.3453,
+ "train_tokens_per_second": 5653.144
+ },
+ {
+ "epoch": 1.8872879073525453,
+ "grad_norm": 0.7934211783590108,
+ "learning_rate": 7.825343409896547e-08,
+ "loss": 0.3094397306442261,
+ "num_input_tokens_seen": 1225864272,
+ "step": 28030,
+ "train_runtime": 216844.6438,
+ "train_tokens_per_second": 5653.191
+ },
+ {
+ "epoch": 1.8876245623485053,
+ "grad_norm": 0.7748967589073342,
+ "learning_rate": 7.778816007333268e-08,
+ "loss": 0.3081014156341553,
+ "num_input_tokens_seen": 1226087184,
+ "step": 28035,
+ "train_runtime": 216887.2541,
+ "train_tokens_per_second": 5653.109
+ },
+ {
+ "epoch": 1.8879612173444653,
+ "grad_norm": 0.8144894017280363,
+ "learning_rate": 7.73242625285453e-08,
+ "loss": 0.29062952995300295,
+ "num_input_tokens_seen": 1226306464,
+ "step": 28040,
+ "train_runtime": 216925.8357,
+ "train_tokens_per_second": 5653.114
+ },
+ {
+ "epoch": 1.8882978723404256,
+ "grad_norm": 0.7574566262911345,
+ "learning_rate": 7.686174159433069e-08,
+ "loss": 0.3064660787582397,
+ "num_input_tokens_seen": 1226520296,
+ "step": 28045,
+ "train_runtime": 216966.9693,
+ "train_tokens_per_second": 5653.028
+ },
+ {
+ "epoch": 1.8886345273363858,
+ "grad_norm": 0.627537289336894,
+ "learning_rate": 7.640059740003091e-08,
+ "loss": 0.2722774982452393,
+ "num_input_tokens_seen": 1226746648,
+ "step": 28050,
+ "train_runtime": 217006.2516,
+ "train_tokens_per_second": 5653.047
+ },
+ {
+ "epoch": 1.8889711823323458,
+ "grad_norm": 0.7705422901820207,
+ "learning_rate": 7.594083007460396e-08,
+ "loss": 0.29419803619384766,
+ "num_input_tokens_seen": 1226953312,
+ "step": 28055,
+ "train_runtime": 217048.1468,
+ "train_tokens_per_second": 5652.908
+ },
+ {
+ "epoch": 1.8893078373283059,
+ "grad_norm": 0.7044024327241764,
+ "learning_rate": 7.54824397466225e-08,
+ "loss": 0.2896998167037964,
+ "num_input_tokens_seen": 1227175648,
+ "step": 28060,
+ "train_runtime": 217084.5331,
+ "train_tokens_per_second": 5652.985
+ },
+ {
+ "epoch": 1.8896444923242661,
+ "grad_norm": 0.715442564314742,
+ "learning_rate": 7.502542654427347e-08,
+ "loss": 0.28265087604522704,
+ "num_input_tokens_seen": 1227383304,
+ "step": 28065,
+ "train_runtime": 217116.2593,
+ "train_tokens_per_second": 5653.116
+ },
+ {
+ "epoch": 1.8899811473202264,
+ "grad_norm": 0.7174922222370509,
+ "learning_rate": 7.456979059535907e-08,
+ "loss": 0.29991984367370605,
+ "num_input_tokens_seen": 1227603976,
+ "step": 28070,
+ "train_runtime": 217161.5657,
+ "train_tokens_per_second": 5652.952
+ },
+ {
+ "epoch": 1.8903178023161864,
+ "grad_norm": 0.8821755978176467,
+ "learning_rate": 7.411553202729682e-08,
+ "loss": 0.2972327470779419,
+ "num_input_tokens_seen": 1227795992,
+ "step": 28075,
+ "train_runtime": 217202.7149,
+ "train_tokens_per_second": 5652.765
+ },
+ {
+ "epoch": 1.8906544573121464,
+ "grad_norm": 0.7547240032942905,
+ "learning_rate": 7.3662650967119e-08,
+ "loss": 0.2765605926513672,
+ "num_input_tokens_seen": 1228009920,
+ "step": 28080,
+ "train_runtime": 217242.1532,
+ "train_tokens_per_second": 5652.724
+ },
+ {
+ "epoch": 1.8909911123081067,
+ "grad_norm": 0.7381759970020896,
+ "learning_rate": 7.321114754147151e-08,
+ "loss": 0.28570716381072997,
+ "num_input_tokens_seen": 1228232816,
+ "step": 28085,
+ "train_runtime": 217280.844,
+ "train_tokens_per_second": 5652.743
+ },
+ {
+ "epoch": 1.891327767304067,
+ "grad_norm": 0.9586290374485588,
+ "learning_rate": 7.276102187661615e-08,
+ "loss": 0.31825072765350343,
+ "num_input_tokens_seen": 1228448704,
+ "step": 28090,
+ "train_runtime": 217319.8127,
+ "train_tokens_per_second": 5652.723
+ },
+ {
+ "epoch": 1.891664422300027,
+ "grad_norm": 0.7106822110807806,
+ "learning_rate": 7.231227409842944e-08,
+ "loss": 0.30093338489532473,
+ "num_input_tokens_seen": 1228675704,
+ "step": 28095,
+ "train_runtime": 217355.8762,
+ "train_tokens_per_second": 5652.829
+ },
+ {
+ "epoch": 1.892001077295987,
+ "grad_norm": 0.7479457461633977,
+ "learning_rate": 7.186490433240156e-08,
+ "loss": 0.30986011028289795,
+ "num_input_tokens_seen": 1228877808,
+ "step": 28100,
+ "train_runtime": 217390.778,
+ "train_tokens_per_second": 5652.852
+ },
+ {
+ "epoch": 1.8923377322919472,
+ "grad_norm": 0.7862198031676131,
+ "learning_rate": 7.14189127036391e-08,
+ "loss": 0.2832916259765625,
+ "num_input_tokens_seen": 1229095280,
+ "step": 28105,
+ "train_runtime": 217425.4702,
+ "train_tokens_per_second": 5652.95
+ },
+ {
+ "epoch": 1.8926743872879075,
+ "grad_norm": 0.7610693134796771,
+ "learning_rate": 7.097429933686118e-08,
+ "loss": 0.3074744701385498,
+ "num_input_tokens_seen": 1229309216,
+ "step": 28110,
+ "train_runtime": 217466.3238,
+ "train_tokens_per_second": 5652.872
+ },
+ {
+ "epoch": 1.8930110422838675,
+ "grad_norm": 0.9062932954935005,
+ "learning_rate": 7.053106435640333e-08,
+ "loss": 0.31830539703369143,
+ "num_input_tokens_seen": 1229541912,
+ "step": 28115,
+ "train_runtime": 217504.1355,
+ "train_tokens_per_second": 5652.959
+ },
+ {
+ "epoch": 1.8933476972798275,
+ "grad_norm": 0.7426318413849856,
+ "learning_rate": 7.008920788621366e-08,
+ "loss": 0.299678635597229,
+ "num_input_tokens_seen": 1229758840,
+ "step": 28120,
+ "train_runtime": 217545.4661,
+ "train_tokens_per_second": 5652.882
+ },
+ {
+ "epoch": 1.8936843522757878,
+ "grad_norm": 0.7413896799320939,
+ "learning_rate": 6.964873004985717e-08,
+ "loss": 0.3041519641876221,
+ "num_input_tokens_seen": 1229975280,
+ "step": 28125,
+ "train_runtime": 217582.095,
+ "train_tokens_per_second": 5652.925
+ },
+ {
+ "epoch": 1.894021007271748,
+ "grad_norm": 0.8140877400473608,
+ "learning_rate": 6.920963097051148e-08,
+ "loss": 0.3081005096435547,
+ "num_input_tokens_seen": 1230195248,
+ "step": 28130,
+ "train_runtime": 217622.0274,
+ "train_tokens_per_second": 5652.899
+ },
+ {
+ "epoch": 1.894357662267708,
+ "grad_norm": 0.8171774682734737,
+ "learning_rate": 6.877191077096945e-08,
+ "loss": 0.28580417633056643,
+ "num_input_tokens_seen": 1230403720,
+ "step": 28135,
+ "train_runtime": 217663.6137,
+ "train_tokens_per_second": 5652.776
+ },
+ {
+ "epoch": 1.894694317263668,
+ "grad_norm": 0.7796944173521188,
+ "learning_rate": 6.833556957363818e-08,
+ "loss": 0.31174564361572266,
+ "num_input_tokens_seen": 1230619400,
+ "step": 28140,
+ "train_runtime": 217701.7663,
+ "train_tokens_per_second": 5652.776
+ },
+ {
+ "epoch": 1.8950309722596284,
+ "grad_norm": 0.7394530047484759,
+ "learning_rate": 6.790060750053895e-08,
+ "loss": 0.2922555446624756,
+ "num_input_tokens_seen": 1230833368,
+ "step": 28145,
+ "train_runtime": 217738.7294,
+ "train_tokens_per_second": 5652.799
+ },
+ {
+ "epoch": 1.8953676272555886,
+ "grad_norm": 0.6213839509137753,
+ "learning_rate": 6.746702467330723e-08,
+ "loss": 0.2662560701370239,
+ "num_input_tokens_seen": 1231064032,
+ "step": 28150,
+ "train_runtime": 217778.1192,
+ "train_tokens_per_second": 5652.836
+ },
+ {
+ "epoch": 1.8957042822515486,
+ "grad_norm": 0.7209769414897946,
+ "learning_rate": 6.703482121319383e-08,
+ "loss": 0.2952676296234131,
+ "num_input_tokens_seen": 1231282944,
+ "step": 28155,
+ "train_runtime": 217815.4664,
+ "train_tokens_per_second": 5652.872
+ },
+ {
+ "epoch": 1.8960409372475087,
+ "grad_norm": 0.7719176273534697,
+ "learning_rate": 6.660399724106204e-08,
+ "loss": 0.31438963413238524,
+ "num_input_tokens_seen": 1231513400,
+ "step": 28160,
+ "train_runtime": 217855.297,
+ "train_tokens_per_second": 5652.896
+ },
+ {
+ "epoch": 1.896377592243469,
+ "grad_norm": 0.7554115178407177,
+ "learning_rate": 6.617455287739161e-08,
+ "loss": 0.28926453590393064,
+ "num_input_tokens_seen": 1231738296,
+ "step": 28165,
+ "train_runtime": 217889.6272,
+ "train_tokens_per_second": 5653.038
+ },
+ {
+ "epoch": 1.8967142472394292,
+ "grad_norm": 0.7869091031125452,
+ "learning_rate": 6.574648824227369e-08,
+ "loss": 0.29925999641418455,
+ "num_input_tokens_seen": 1231961800,
+ "step": 28170,
+ "train_runtime": 217924.705,
+ "train_tokens_per_second": 5653.153
+ },
+ {
+ "epoch": 1.8970509022353892,
+ "grad_norm": 0.7908040273802075,
+ "learning_rate": 6.5319803455417e-08,
+ "loss": 0.3044144153594971,
+ "num_input_tokens_seen": 1232184856,
+ "step": 28175,
+ "train_runtime": 217965.4264,
+ "train_tokens_per_second": 5653.121
+ },
+ {
+ "epoch": 1.8973875572313492,
+ "grad_norm": 0.7737887474560637,
+ "learning_rate": 6.489449863614106e-08,
+ "loss": 0.3226012706756592,
+ "num_input_tokens_seen": 1232404696,
+ "step": 28180,
+ "train_runtime": 218003.5275,
+ "train_tokens_per_second": 5653.141
+ },
+ {
+ "epoch": 1.8977242122273095,
+ "grad_norm": 0.7245308841502345,
+ "learning_rate": 6.44705739033824e-08,
+ "loss": 0.29819202423095703,
+ "num_input_tokens_seen": 1232620728,
+ "step": 28185,
+ "train_runtime": 218043.0047,
+ "train_tokens_per_second": 5653.108
+ },
+ {
+ "epoch": 1.8980608672232697,
+ "grad_norm": 0.883569648340652,
+ "learning_rate": 6.404802937568899e-08,
+ "loss": 0.283476448059082,
+ "num_input_tokens_seen": 1232822576,
+ "step": 28190,
+ "train_runtime": 218076.1787,
+ "train_tokens_per_second": 5653.174
+ },
+ {
+ "epoch": 1.8983975222192297,
+ "grad_norm": 0.6097934961985845,
+ "learning_rate": 6.362686517122463e-08,
+ "loss": 0.276761531829834,
+ "num_input_tokens_seen": 1233038648,
+ "step": 28195,
+ "train_runtime": 218115.3719,
+ "train_tokens_per_second": 5653.149
+ },
+ {
+ "epoch": 1.8987341772151898,
+ "grad_norm": 0.7436026418824351,
+ "learning_rate": 6.320708140776676e-08,
+ "loss": 0.28359341621398926,
+ "num_input_tokens_seen": 1233258352,
+ "step": 28200,
+ "train_runtime": 218158.6419,
+ "train_tokens_per_second": 5653.035
+ },
+ {
+ "epoch": 1.89907083221115,
+ "grad_norm": 0.7681913781823284,
+ "learning_rate": 6.278867820270596e-08,
+ "loss": 0.299153208732605,
+ "num_input_tokens_seen": 1233480288,
+ "step": 28205,
+ "train_runtime": 218195.6989,
+ "train_tokens_per_second": 5653.092
+ },
+ {
+ "epoch": 1.8994074872071103,
+ "grad_norm": 0.7830138695146227,
+ "learning_rate": 6.237165567304749e-08,
+ "loss": 0.29818384647369384,
+ "num_input_tokens_seen": 1233691368,
+ "step": 28210,
+ "train_runtime": 218237.5859,
+ "train_tokens_per_second": 5652.974
+ },
+ {
+ "epoch": 1.8997441422030703,
+ "grad_norm": 0.7003448488149383,
+ "learning_rate": 6.19560139354114e-08,
+ "loss": 0.30460848808288576,
+ "num_input_tokens_seen": 1233930016,
+ "step": 28215,
+ "train_runtime": 218270.3105,
+ "train_tokens_per_second": 5653.22
+ },
+ {
+ "epoch": 1.9000807971990303,
+ "grad_norm": 0.7370403299032923,
+ "learning_rate": 6.154175310602916e-08,
+ "loss": 0.30403151512146,
+ "num_input_tokens_seen": 1234154432,
+ "step": 28220,
+ "train_runtime": 218302.9228,
+ "train_tokens_per_second": 5653.403
+ },
+ {
+ "epoch": 1.9004174521949906,
+ "grad_norm": 0.7562586977964543,
+ "learning_rate": 6.112887330074812e-08,
+ "loss": 0.2886775255203247,
+ "num_input_tokens_seen": 1234363888,
+ "step": 28225,
+ "train_runtime": 218340.6417,
+ "train_tokens_per_second": 5653.386
+ },
+ {
+ "epoch": 1.9007541071909508,
+ "grad_norm": 0.8023435948804223,
+ "learning_rate": 6.07173746350287e-08,
+ "loss": 0.3039447069168091,
+ "num_input_tokens_seen": 1234586704,
+ "step": 28230,
+ "train_runtime": 218377.1498,
+ "train_tokens_per_second": 5653.461
+ },
+ {
+ "epoch": 1.9010907621869109,
+ "grad_norm": 0.7102448231458076,
+ "learning_rate": 6.030725722394548e-08,
+ "loss": 0.30956282615661623,
+ "num_input_tokens_seen": 1234811984,
+ "step": 28235,
+ "train_runtime": 218419.5266,
+ "train_tokens_per_second": 5653.396
+ },
+ {
+ "epoch": 1.901427417182871,
+ "grad_norm": 0.8917953406376619,
+ "learning_rate": 5.98985211821862e-08,
+ "loss": 0.31682801246643066,
+ "num_input_tokens_seen": 1235026864,
+ "step": 28240,
+ "train_runtime": 218463.2255,
+ "train_tokens_per_second": 5653.248
+ },
+ {
+ "epoch": 1.9017640721788311,
+ "grad_norm": 3.1213239281278913,
+ "learning_rate": 5.9491166624052185e-08,
+ "loss": 0.29059295654296874,
+ "num_input_tokens_seen": 1235257424,
+ "step": 28245,
+ "train_runtime": 218502.1071,
+ "train_tokens_per_second": 5653.298
+ },
+ {
+ "epoch": 1.9021007271747914,
+ "grad_norm": 0.800370397152095,
+ "learning_rate": 5.908519366345955e-08,
+ "loss": 0.2929649829864502,
+ "num_input_tokens_seen": 1235465968,
+ "step": 28250,
+ "train_runtime": 218531.1775,
+ "train_tokens_per_second": 5653.5
+ },
+ {
+ "epoch": 1.9024373821707514,
+ "grad_norm": 0.7899542839676547,
+ "learning_rate": 5.868060241393747e-08,
+ "loss": 0.32121825218200684,
+ "num_input_tokens_seen": 1235701752,
+ "step": 28255,
+ "train_runtime": 218574.1157,
+ "train_tokens_per_second": 5653.468
+ },
+ {
+ "epoch": 1.9027740371667115,
+ "grad_norm": 0.8945208045237926,
+ "learning_rate": 5.8277392988627665e-08,
+ "loss": 0.30273237228393557,
+ "num_input_tokens_seen": 1235918048,
+ "step": 28260,
+ "train_runtime": 218613.0158,
+ "train_tokens_per_second": 5653.451
+ },
+ {
+ "epoch": 1.9031106921626717,
+ "grad_norm": 1.0956824749607716,
+ "learning_rate": 5.787556550028772e-08,
+ "loss": 0.2830909967422485,
+ "num_input_tokens_seen": 1236133248,
+ "step": 28265,
+ "train_runtime": 218655.8574,
+ "train_tokens_per_second": 5653.328
+ },
+ {
+ "epoch": 1.903447347158632,
+ "grad_norm": 0.7232891426044542,
+ "learning_rate": 5.747512006128553e-08,
+ "loss": 0.3094156265258789,
+ "num_input_tokens_seen": 1236365056,
+ "step": 28270,
+ "train_runtime": 218690.6793,
+ "train_tokens_per_second": 5653.488
+ },
+ {
+ "epoch": 1.903784002154592,
+ "grad_norm": 0.825207510049438,
+ "learning_rate": 5.707605678360595e-08,
+ "loss": 0.30304322242736814,
+ "num_input_tokens_seen": 1236586568,
+ "step": 28275,
+ "train_runtime": 218729.618,
+ "train_tokens_per_second": 5653.494
+ },
+ {
+ "epoch": 1.904120657150552,
+ "grad_norm": 0.6746156299713596,
+ "learning_rate": 5.6678375778845276e-08,
+ "loss": 0.2991356611251831,
+ "num_input_tokens_seen": 1236816112,
+ "step": 28280,
+ "train_runtime": 218761.9014,
+ "train_tokens_per_second": 5653.709
+ },
+ {
+ "epoch": 1.9044573121465123,
+ "grad_norm": 0.7102199653388521,
+ "learning_rate": 5.6282077158213435e-08,
+ "loss": 0.29000029563903806,
+ "num_input_tokens_seen": 1237040088,
+ "step": 28285,
+ "train_runtime": 218799.9174,
+ "train_tokens_per_second": 5653.75
+ },
+ {
+ "epoch": 1.9047939671424725,
+ "grad_norm": 0.8451189106225591,
+ "learning_rate": 5.588716103253511e-08,
+ "loss": 0.29954440593719484,
+ "num_input_tokens_seen": 1237269408,
+ "step": 28290,
+ "train_runtime": 218845.6994,
+ "train_tokens_per_second": 5653.615
+ },
+ {
+ "epoch": 1.9051306221384325,
+ "grad_norm": 0.8190739439948348,
+ "learning_rate": 5.549362751224585e-08,
+ "loss": 0.31754138469696047,
+ "num_input_tokens_seen": 1237476144,
+ "step": 28295,
+ "train_runtime": 218883.6501,
+ "train_tokens_per_second": 5653.58
+ },
+ {
+ "epoch": 1.9054672771343926,
+ "grad_norm": 0.7766937595152473,
+ "learning_rate": 5.5101476707397075e-08,
+ "loss": 0.2879055976867676,
+ "num_input_tokens_seen": 1237699056,
+ "step": 28300,
+ "train_runtime": 218922.8288,
+ "train_tokens_per_second": 5653.586
+ },
+ {
+ "epoch": 1.9058039321303528,
+ "grad_norm": 0.6581310767139287,
+ "learning_rate": 5.471070872765216e-08,
+ "loss": 0.31143577098846437,
+ "num_input_tokens_seen": 1237922344,
+ "step": 28305,
+ "train_runtime": 218962.3366,
+ "train_tokens_per_second": 5653.586
+ },
+ {
+ "epoch": 1.906140587126313,
+ "grad_norm": 0.8595902577669746,
+ "learning_rate": 5.432132368228815e-08,
+ "loss": 0.3075572490692139,
+ "num_input_tokens_seen": 1238144360,
+ "step": 28310,
+ "train_runtime": 218996.6769,
+ "train_tokens_per_second": 5653.713
+ },
+ {
+ "epoch": 1.906477242122273,
+ "grad_norm": 0.7260821752374543,
+ "learning_rate": 5.393332168019572e-08,
+ "loss": 0.29381508827209474,
+ "num_input_tokens_seen": 1238376024,
+ "step": 28315,
+ "train_runtime": 219038.2854,
+ "train_tokens_per_second": 5653.697
+ },
+ {
+ "epoch": 1.9068138971182331,
+ "grad_norm": 0.7695214541702009,
+ "learning_rate": 5.354670282987695e-08,
+ "loss": 0.31864123344421386,
+ "num_input_tokens_seen": 1238615352,
+ "step": 28320,
+ "train_runtime": 219071.5756,
+ "train_tokens_per_second": 5653.93
+ },
+ {
+ "epoch": 1.9071505521141934,
+ "grad_norm": 0.7626491001109459,
+ "learning_rate": 5.316146723945037e-08,
+ "loss": 0.3147014379501343,
+ "num_input_tokens_seen": 1238839600,
+ "step": 28325,
+ "train_runtime": 219109.8816,
+ "train_tokens_per_second": 5653.965
+ },
+ {
+ "epoch": 1.9074872071101536,
+ "grad_norm": 0.7542385697702055,
+ "learning_rate": 5.277761501664535e-08,
+ "loss": 0.28926112651824953,
+ "num_input_tokens_seen": 1239055736,
+ "step": 28330,
+ "train_runtime": 219151.3861,
+ "train_tokens_per_second": 5653.88
+ },
+ {
+ "epoch": 1.9078238621061137,
+ "grad_norm": 0.8539701413858525,
+ "learning_rate": 5.2395146268804934e-08,
+ "loss": 0.3053768157958984,
+ "num_input_tokens_seen": 1239282360,
+ "step": 28335,
+ "train_runtime": 219183.9605,
+ "train_tokens_per_second": 5654.074
+ },
+ {
+ "epoch": 1.9081605171020737,
+ "grad_norm": 0.6987538717491121,
+ "learning_rate": 5.201406110288465e-08,
+ "loss": 0.2954925537109375,
+ "num_input_tokens_seen": 1239505344,
+ "step": 28340,
+ "train_runtime": 219222.4201,
+ "train_tokens_per_second": 5654.099
+ },
+ {
+ "epoch": 1.908497172098034,
+ "grad_norm": 0.7258797958708753,
+ "learning_rate": 5.1634359625454266e-08,
+ "loss": 0.26157379150390625,
+ "num_input_tokens_seen": 1239718552,
+ "step": 28345,
+ "train_runtime": 219261.4852,
+ "train_tokens_per_second": 5654.064
+ },
+ {
+ "epoch": 1.9088338270939942,
+ "grad_norm": 0.833392494175178,
+ "learning_rate": 5.125604194269606e-08,
+ "loss": 0.32343695163726804,
+ "num_input_tokens_seen": 1239933464,
+ "step": 28350,
+ "train_runtime": 219303.7788,
+ "train_tokens_per_second": 5653.954
+ },
+ {
+ "epoch": 1.9091704820899542,
+ "grad_norm": 0.8588011467833099,
+ "learning_rate": 5.087910816040542e-08,
+ "loss": 0.32165529727935793,
+ "num_input_tokens_seen": 1240153424,
+ "step": 28355,
+ "train_runtime": 219348.9617,
+ "train_tokens_per_second": 5653.792
+ },
+ {
+ "epoch": 1.9095071370859142,
+ "grad_norm": 0.7170515438107391,
+ "learning_rate": 5.0503558383990235e-08,
+ "loss": 0.29596068859100344,
+ "num_input_tokens_seen": 1240382984,
+ "step": 28360,
+ "train_runtime": 219391.3692,
+ "train_tokens_per_second": 5653.746
+ },
+ {
+ "epoch": 1.9098437920818745,
+ "grad_norm": 0.9056425664871629,
+ "learning_rate": 5.0129392718472614e-08,
+ "loss": 0.29495375156402587,
+ "num_input_tokens_seen": 1240602456,
+ "step": 28365,
+ "train_runtime": 219428.0549,
+ "train_tokens_per_second": 5653.801
+ },
+ {
+ "epoch": 1.9101804470778347,
+ "grad_norm": 0.7683712573518254,
+ "learning_rate": 4.9756611268486077e-08,
+ "loss": 0.3231898307800293,
+ "num_input_tokens_seen": 1240834056,
+ "step": 28370,
+ "train_runtime": 219470.1885,
+ "train_tokens_per_second": 5653.77
+ },
+ {
+ "epoch": 1.9105171020737948,
+ "grad_norm": 0.8473839083939165,
+ "learning_rate": 4.9385214138277795e-08,
+ "loss": 0.3085433006286621,
+ "num_input_tokens_seen": 1241053560,
+ "step": 28375,
+ "train_runtime": 219515.4349,
+ "train_tokens_per_second": 5653.605
+ },
+ {
+ "epoch": 1.9108537570697548,
+ "grad_norm": 0.7530840258849333,
+ "learning_rate": 4.901520143170857e-08,
+ "loss": 0.31458406448364257,
+ "num_input_tokens_seen": 1241278752,
+ "step": 28380,
+ "train_runtime": 219551.5142,
+ "train_tokens_per_second": 5653.702
+ },
+ {
+ "epoch": 1.911190412065715,
+ "grad_norm": 0.8440231962915516,
+ "learning_rate": 4.864657325225064e-08,
+ "loss": 0.2959612846374512,
+ "num_input_tokens_seen": 1241501272,
+ "step": 28385,
+ "train_runtime": 219592.0284,
+ "train_tokens_per_second": 5653.672
+ },
+ {
+ "epoch": 1.9115270670616753,
+ "grad_norm": 0.7788969789388622,
+ "learning_rate": 4.827932970298932e-08,
+ "loss": 0.28577983379364014,
+ "num_input_tokens_seen": 1241706992,
+ "step": 28390,
+ "train_runtime": 219631.7102,
+ "train_tokens_per_second": 5653.587
+ },
+ {
+ "epoch": 1.9118637220576353,
+ "grad_norm": 0.7570563184078223,
+ "learning_rate": 4.7913470886624125e-08,
+ "loss": 0.30133938789367676,
+ "num_input_tokens_seen": 1241929008,
+ "step": 28395,
+ "train_runtime": 219667.8047,
+ "train_tokens_per_second": 5653.669
+ },
+ {
+ "epoch": 1.9122003770535954,
+ "grad_norm": 0.7674867971456765,
+ "learning_rate": 4.754899690546544e-08,
+ "loss": 0.2801696538925171,
+ "num_input_tokens_seen": 1242147848,
+ "step": 28400,
+ "train_runtime": 219701.4788,
+ "train_tokens_per_second": 5653.798
+ },
+ {
+ "epoch": 1.9125370320495556,
+ "grad_norm": 0.7214457077784336,
+ "learning_rate": 4.718590786143729e-08,
+ "loss": 0.3026761531829834,
+ "num_input_tokens_seen": 1242368696,
+ "step": 28405,
+ "train_runtime": 219734.8895,
+ "train_tokens_per_second": 5653.944
+ },
+ {
+ "epoch": 1.9128736870455159,
+ "grad_norm": 0.7424922888900161,
+ "learning_rate": 4.6824203856076776e-08,
+ "loss": 0.3060077905654907,
+ "num_input_tokens_seen": 1242605672,
+ "step": 28410,
+ "train_runtime": 219770.358,
+ "train_tokens_per_second": 5654.11
+ },
+ {
+ "epoch": 1.9132103420414759,
+ "grad_norm": 0.7488946195290975,
+ "learning_rate": 4.646388499053356e-08,
+ "loss": 0.2775508642196655,
+ "num_input_tokens_seen": 1242816568,
+ "step": 28415,
+ "train_runtime": 219806.033,
+ "train_tokens_per_second": 5654.151
+ },
+ {
+ "epoch": 1.913546997037436,
+ "grad_norm": 0.7402940285995784,
+ "learning_rate": 4.6104951365568696e-08,
+ "loss": 0.2732607126235962,
+ "num_input_tokens_seen": 1243037664,
+ "step": 28420,
+ "train_runtime": 219842.7305,
+ "train_tokens_per_second": 5654.213
+ },
+ {
+ "epoch": 1.9138836520333962,
+ "grad_norm": 0.8182194298111709,
+ "learning_rate": 4.574740308155801e-08,
+ "loss": 0.2992802858352661,
+ "num_input_tokens_seen": 1243248672,
+ "step": 28425,
+ "train_runtime": 219884.9,
+ "train_tokens_per_second": 5654.088
+ },
+ {
+ "epoch": 1.9142203070293564,
+ "grad_norm": 0.7646860170343411,
+ "learning_rate": 4.539124023848762e-08,
+ "loss": 0.2957939624786377,
+ "num_input_tokens_seen": 1243481672,
+ "step": 28430,
+ "train_runtime": 219924.6589,
+ "train_tokens_per_second": 5654.126
+ },
+ {
+ "epoch": 1.9145569620253164,
+ "grad_norm": 0.7725331026025192,
+ "learning_rate": 4.503646293595787e-08,
+ "loss": 0.32006833553314207,
+ "num_input_tokens_seen": 1243698024,
+ "step": 28435,
+ "train_runtime": 219958.6796,
+ "train_tokens_per_second": 5654.235
+ },
+ {
+ "epoch": 1.9148936170212765,
+ "grad_norm": 0.7967463881139475,
+ "learning_rate": 4.468307127318105e-08,
+ "loss": 0.2883774280548096,
+ "num_input_tokens_seen": 1243923240,
+ "step": 28440,
+ "train_runtime": 219994.0164,
+ "train_tokens_per_second": 5654.35
+ },
+ {
+ "epoch": 1.9152302720172367,
+ "grad_norm": 0.8721239693546312,
+ "learning_rate": 4.4331065348982014e-08,
+ "loss": 0.2830281496047974,
+ "num_input_tokens_seen": 1244140864,
+ "step": 28445,
+ "train_runtime": 220025.1373,
+ "train_tokens_per_second": 5654.54
+ },
+ {
+ "epoch": 1.915566927013197,
+ "grad_norm": 0.7395255680690924,
+ "learning_rate": 4.3980445261798125e-08,
+ "loss": 0.3066460371017456,
+ "num_input_tokens_seen": 1244363064,
+ "step": 28450,
+ "train_runtime": 220071.6297,
+ "train_tokens_per_second": 5654.355
+ },
+ {
+ "epoch": 1.915903582009157,
+ "grad_norm": 0.8552277238547642,
+ "learning_rate": 4.363121110967927e-08,
+ "loss": 0.2869099140167236,
+ "num_input_tokens_seen": 1244567184,
+ "step": 28455,
+ "train_runtime": 220111.9824,
+ "train_tokens_per_second": 5654.245
+ },
+ {
+ "epoch": 1.916240237005117,
+ "grad_norm": 0.7565359558538893,
+ "learning_rate": 4.328336299028735e-08,
+ "loss": 0.27414395809173586,
+ "num_input_tokens_seen": 1244775824,
+ "step": 28460,
+ "train_runtime": 220156.5249,
+ "train_tokens_per_second": 5654.049
+ },
+ {
+ "epoch": 1.9165768920010773,
+ "grad_norm": 0.8537030065422263,
+ "learning_rate": 4.293690100089731e-08,
+ "loss": 0.30220975875854494,
+ "num_input_tokens_seen": 1244988464,
+ "step": 28465,
+ "train_runtime": 220195.8129,
+ "train_tokens_per_second": 5654.006
+ },
+ {
+ "epoch": 1.9169135469970375,
+ "grad_norm": 0.8148234793335632,
+ "learning_rate": 4.259182523839556e-08,
+ "loss": 0.29154458045959475,
+ "num_input_tokens_seen": 1245214008,
+ "step": 28470,
+ "train_runtime": 220234.437,
+ "train_tokens_per_second": 5654.039
+ },
+ {
+ "epoch": 1.9172502019929976,
+ "grad_norm": 0.7939908369257882,
+ "learning_rate": 4.224813579928211e-08,
+ "loss": 0.3210744857788086,
+ "num_input_tokens_seen": 1245434864,
+ "step": 28475,
+ "train_runtime": 220278.6573,
+ "train_tokens_per_second": 5653.906
+ },
+ {
+ "epoch": 1.9175868569889576,
+ "grad_norm": 0.7166123345087314,
+ "learning_rate": 4.190583277966842e-08,
+ "loss": 0.29584832191467286,
+ "num_input_tokens_seen": 1245653840,
+ "step": 28480,
+ "train_runtime": 220319.9507,
+ "train_tokens_per_second": 5653.84
+ },
+ {
+ "epoch": 1.9179235119849178,
+ "grad_norm": 0.7696523687780278,
+ "learning_rate": 4.1564916275278477e-08,
+ "loss": 0.2821512699127197,
+ "num_input_tokens_seen": 1245862296,
+ "step": 28485,
+ "train_runtime": 220358.5767,
+ "train_tokens_per_second": 5653.795
+ },
+ {
+ "epoch": 1.918260166980878,
+ "grad_norm": 0.7811203217746909,
+ "learning_rate": 4.122538638144824e-08,
+ "loss": 0.2951045989990234,
+ "num_input_tokens_seen": 1246082824,
+ "step": 28490,
+ "train_runtime": 220395.6576,
+ "train_tokens_per_second": 5653.845
+ },
+ {
+ "epoch": 1.9185968219768381,
+ "grad_norm": 0.7044342809522225,
+ "learning_rate": 4.08872431931262e-08,
+ "loss": 0.2757328271865845,
+ "num_input_tokens_seen": 1246295392,
+ "step": 28495,
+ "train_runtime": 220435.9021,
+ "train_tokens_per_second": 5653.777
+ },
+ {
+ "epoch": 1.9189334769727981,
+ "grad_norm": 0.8062793001735447,
+ "learning_rate": 4.055048680487339e-08,
+ "loss": 0.29179792404174804,
+ "num_input_tokens_seen": 1246522464,
+ "step": 28500,
+ "train_runtime": 220473.702,
+ "train_tokens_per_second": 5653.837
+ },
+ {
+ "epoch": 1.9192701319687584,
+ "grad_norm": 0.7448550831538739,
+ "learning_rate": 4.0215117310862806e-08,
+ "loss": 0.3171137809753418,
+ "num_input_tokens_seen": 1246721184,
+ "step": 28505,
+ "train_runtime": 220514.8027,
+ "train_tokens_per_second": 5653.685
+ },
+ {
+ "epoch": 1.9196067869647186,
+ "grad_norm": 0.8454561648217672,
+ "learning_rate": 3.9881134804878316e-08,
+ "loss": 0.3017882347106934,
+ "num_input_tokens_seen": 1246923736,
+ "step": 28510,
+ "train_runtime": 220555.7088,
+ "train_tokens_per_second": 5653.555
+ },
+ {
+ "epoch": 1.9199434419606787,
+ "grad_norm": 0.818261031372424,
+ "learning_rate": 3.9548539380318e-08,
+ "loss": 0.3179955005645752,
+ "num_input_tokens_seen": 1247152480,
+ "step": 28515,
+ "train_runtime": 220588.4295,
+ "train_tokens_per_second": 5653.753
+ },
+ {
+ "epoch": 1.9202800969566387,
+ "grad_norm": 0.7797530065110128,
+ "learning_rate": 3.921733113019077e-08,
+ "loss": 0.3016833305358887,
+ "num_input_tokens_seen": 1247354448,
+ "step": 28520,
+ "train_runtime": 220628.9625,
+ "train_tokens_per_second": 5653.63
+ },
+ {
+ "epoch": 1.920616751952599,
+ "grad_norm": 0.7210084261051174,
+ "learning_rate": 3.888751014711867e-08,
+ "loss": 0.3128508567810059,
+ "num_input_tokens_seen": 1247557312,
+ "step": 28525,
+ "train_runtime": 220662.2317,
+ "train_tokens_per_second": 5653.697
+ },
+ {
+ "epoch": 1.9209534069485592,
+ "grad_norm": 0.6727498501097389,
+ "learning_rate": 3.855907652333402e-08,
+ "loss": 0.28232321739196775,
+ "num_input_tokens_seen": 1247776920,
+ "step": 28530,
+ "train_runtime": 220702.1687,
+ "train_tokens_per_second": 5653.669
+ },
+ {
+ "epoch": 1.9212900619445192,
+ "grad_norm": 0.7326711637458728,
+ "learning_rate": 3.8232030350683344e-08,
+ "loss": 0.2884651184082031,
+ "num_input_tokens_seen": 1248006808,
+ "step": 28535,
+ "train_runtime": 220743.7071,
+ "train_tokens_per_second": 5653.646
+ },
+ {
+ "epoch": 1.9216267169404793,
+ "grad_norm": 0.749333938315018,
+ "learning_rate": 3.7906371720622946e-08,
+ "loss": 0.2990164041519165,
+ "num_input_tokens_seen": 1248234992,
+ "step": 28540,
+ "train_runtime": 220782.3848,
+ "train_tokens_per_second": 5653.689
+ },
+ {
+ "epoch": 1.9219633719364395,
+ "grad_norm": 0.7532923064142389,
+ "learning_rate": 3.758210072422275e-08,
+ "loss": 0.28374738693237306,
+ "num_input_tokens_seen": 1248453088,
+ "step": 28545,
+ "train_runtime": 220818.7593,
+ "train_tokens_per_second": 5653.746
+ },
+ {
+ "epoch": 1.9223000269323998,
+ "grad_norm": 0.8243046568432454,
+ "learning_rate": 3.72592174521641e-08,
+ "loss": 0.30987684726715087,
+ "num_input_tokens_seen": 1248674912,
+ "step": 28550,
+ "train_runtime": 220861.5131,
+ "train_tokens_per_second": 5653.656
+ },
+ {
+ "epoch": 1.9226366819283598,
+ "grad_norm": 0.7456873109655421,
+ "learning_rate": 3.693772199474033e-08,
+ "loss": 0.30152506828308107,
+ "num_input_tokens_seen": 1248896320,
+ "step": 28555,
+ "train_runtime": 220899.3543,
+ "train_tokens_per_second": 5653.689
+ },
+ {
+ "epoch": 1.9229733369243198,
+ "grad_norm": 0.8268089260806681,
+ "learning_rate": 3.661761444185674e-08,
+ "loss": 0.30233216285705566,
+ "num_input_tokens_seen": 1249110256,
+ "step": 28560,
+ "train_runtime": 220941.1918,
+ "train_tokens_per_second": 5653.587
+ },
+ {
+ "epoch": 1.92330999192028,
+ "grad_norm": 0.7596988638251234,
+ "learning_rate": 3.6298894883030066e-08,
+ "loss": 0.3038165807723999,
+ "num_input_tokens_seen": 1249337408,
+ "step": 28565,
+ "train_runtime": 220975.0974,
+ "train_tokens_per_second": 5653.748
+ },
+ {
+ "epoch": 1.9236466469162403,
+ "grad_norm": 0.8230343694103864,
+ "learning_rate": 3.5981563407389007e-08,
+ "loss": 0.3213151454925537,
+ "num_input_tokens_seen": 1249559176,
+ "step": 28570,
+ "train_runtime": 221018.9022,
+ "train_tokens_per_second": 5653.63
+ },
+ {
+ "epoch": 1.9239833019122003,
+ "grad_norm": 0.8032414208109102,
+ "learning_rate": 3.566562010367536e-08,
+ "loss": 0.2826304197311401,
+ "num_input_tokens_seen": 1249766312,
+ "step": 28575,
+ "train_runtime": 221051.4492,
+ "train_tokens_per_second": 5653.735
+ },
+ {
+ "epoch": 1.9243199569081604,
+ "grad_norm": 0.6984009548103529,
+ "learning_rate": 3.535106506024011e-08,
+ "loss": 0.2978407621383667,
+ "num_input_tokens_seen": 1249983696,
+ "step": 28580,
+ "train_runtime": 221088.9385,
+ "train_tokens_per_second": 5653.76
+ },
+ {
+ "epoch": 1.9246566119041206,
+ "grad_norm": 0.7453143461192999,
+ "learning_rate": 3.5037898365049005e-08,
+ "loss": 0.2644341230392456,
+ "num_input_tokens_seen": 1250204168,
+ "step": 28585,
+ "train_runtime": 221131.5643,
+ "train_tokens_per_second": 5653.667
+ },
+ {
+ "epoch": 1.9249932669000809,
+ "grad_norm": 0.8872235907895941,
+ "learning_rate": 3.472612010567755e-08,
+ "loss": 0.29390702247619627,
+ "num_input_tokens_seen": 1250412232,
+ "step": 28590,
+ "train_runtime": 221167.7612,
+ "train_tokens_per_second": 5653.682
+ },
+ {
+ "epoch": 1.925329921896041,
+ "grad_norm": 0.7629679742085492,
+ "learning_rate": 3.441573036931323e-08,
+ "loss": 0.3093965291976929,
+ "num_input_tokens_seen": 1250638880,
+ "step": 28595,
+ "train_runtime": 221205.5961,
+ "train_tokens_per_second": 5653.74
+ },
+ {
+ "epoch": 1.925666576892001,
+ "grad_norm": 0.9764963685352374,
+ "learning_rate": 3.410672924275604e-08,
+ "loss": 0.273404598236084,
+ "num_input_tokens_seen": 1250870096,
+ "step": 28600,
+ "train_runtime": 221249.308,
+ "train_tokens_per_second": 5653.668
+ },
+ {
+ "epoch": 1.9260032318879612,
+ "grad_norm": 0.8128740784863975,
+ "learning_rate": 3.3799116812416876e-08,
+ "loss": 0.3114511489868164,
+ "num_input_tokens_seen": 1251082376,
+ "step": 28605,
+ "train_runtime": 221285.4439,
+ "train_tokens_per_second": 5653.704
+ },
+ {
+ "epoch": 1.9263398868839214,
+ "grad_norm": 0.8683827269674369,
+ "learning_rate": 3.349289316431803e-08,
+ "loss": 0.31067667007446287,
+ "num_input_tokens_seen": 1251307384,
+ "step": 28610,
+ "train_runtime": 221322.0142,
+ "train_tokens_per_second": 5653.786
+ },
+ {
+ "epoch": 1.9266765418798815,
+ "grad_norm": 0.7478140934985664,
+ "learning_rate": 3.3188058384095446e-08,
+ "loss": 0.29280407428741456,
+ "num_input_tokens_seen": 1251525400,
+ "step": 28615,
+ "train_runtime": 221360.0032,
+ "train_tokens_per_second": 5653.801
+ },
+ {
+ "epoch": 1.9270131968758415,
+ "grad_norm": 0.8025594638672747,
+ "learning_rate": 3.2884612556993714e-08,
+ "loss": 0.32468104362487793,
+ "num_input_tokens_seen": 1251744936,
+ "step": 28620,
+ "train_runtime": 221403.9081,
+ "train_tokens_per_second": 5653.671
+ },
+ {
+ "epoch": 1.9273498518718017,
+ "grad_norm": 0.8491439802692531,
+ "learning_rate": 3.258255576787161e-08,
+ "loss": 0.3124075174331665,
+ "num_input_tokens_seen": 1251958328,
+ "step": 28625,
+ "train_runtime": 221443.8187,
+ "train_tokens_per_second": 5653.616
+ },
+ {
+ "epoch": 1.927686506867762,
+ "grad_norm": 0.8546654392090276,
+ "learning_rate": 3.22818881011977e-08,
+ "loss": 0.31823225021362306,
+ "num_input_tokens_seen": 1252192752,
+ "step": 28630,
+ "train_runtime": 221481.8825,
+ "train_tokens_per_second": 5653.703
+ },
+ {
+ "epoch": 1.928023161863722,
+ "grad_norm": 0.7973579543913553,
+ "learning_rate": 3.198260964105249e-08,
+ "loss": 0.27967541217803954,
+ "num_input_tokens_seen": 1252407792,
+ "step": 28635,
+ "train_runtime": 221519.1612,
+ "train_tokens_per_second": 5653.722
+ },
+ {
+ "epoch": 1.928359816859682,
+ "grad_norm": 0.7954284380902052,
+ "learning_rate": 3.168472047112903e-08,
+ "loss": 0.2996687412261963,
+ "num_input_tokens_seen": 1252623280,
+ "step": 28640,
+ "train_runtime": 221566.1063,
+ "train_tokens_per_second": 5653.497
+ },
+ {
+ "epoch": 1.9286964718556423,
+ "grad_norm": 0.7984207095399551,
+ "learning_rate": 3.138822067473068e-08,
+ "loss": 0.28585450649261473,
+ "num_input_tokens_seen": 1252835408,
+ "step": 28645,
+ "train_runtime": 221604.288,
+ "train_tokens_per_second": 5653.48
+ },
+ {
+ "epoch": 1.9290331268516026,
+ "grad_norm": 0.7748173909126452,
+ "learning_rate": 3.10931103347728e-08,
+ "loss": 0.30099103450775144,
+ "num_input_tokens_seen": 1253057512,
+ "step": 28650,
+ "train_runtime": 221646.7214,
+ "train_tokens_per_second": 5653.4
+ },
+ {
+ "epoch": 1.9293697818475626,
+ "grad_norm": 0.7586349259346754,
+ "learning_rate": 3.0799389533781587e-08,
+ "loss": 0.3233338356018066,
+ "num_input_tokens_seen": 1253285920,
+ "step": 28655,
+ "train_runtime": 221683.9501,
+ "train_tokens_per_second": 5653.481
+ },
+ {
+ "epoch": 1.9297064368435226,
+ "grad_norm": 0.7596388594233524,
+ "learning_rate": 3.050705835389578e-08,
+ "loss": 0.28039982318878176,
+ "num_input_tokens_seen": 1253502504,
+ "step": 28660,
+ "train_runtime": 221722.1476,
+ "train_tokens_per_second": 5653.484
+ },
+ {
+ "epoch": 1.9300430918394829,
+ "grad_norm": 0.8761598764006204,
+ "learning_rate": 3.021611687686443e-08,
+ "loss": 0.2949859380722046,
+ "num_input_tokens_seen": 1253714448,
+ "step": 28665,
+ "train_runtime": 221759.6137,
+ "train_tokens_per_second": 5653.484
+ },
+ {
+ "epoch": 1.9303797468354431,
+ "grad_norm": 0.8219444240328218,
+ "learning_rate": 2.992656518404857e-08,
+ "loss": 0.3135673999786377,
+ "num_input_tokens_seen": 1253921528,
+ "step": 28670,
+ "train_runtime": 221798.9051,
+ "train_tokens_per_second": 5653.416
+ },
+ {
+ "epoch": 1.9307164018314031,
+ "grad_norm": 0.661270135482443,
+ "learning_rate": 2.96384033564201e-08,
+ "loss": 0.28868966102600097,
+ "num_input_tokens_seen": 1254137480,
+ "step": 28675,
+ "train_runtime": 221835.7558,
+ "train_tokens_per_second": 5653.451
+ },
+ {
+ "epoch": 1.9310530568273632,
+ "grad_norm": 0.767415898484409,
+ "learning_rate": 2.935163147456288e-08,
+ "loss": 0.27773294448852537,
+ "num_input_tokens_seen": 1254360912,
+ "step": 28680,
+ "train_runtime": 221876.4785,
+ "train_tokens_per_second": 5653.42
+ },
+ {
+ "epoch": 1.9313897118233234,
+ "grad_norm": 0.8423049211038874,
+ "learning_rate": 2.906624961867166e-08,
+ "loss": 0.3049351215362549,
+ "num_input_tokens_seen": 1254581976,
+ "step": 28685,
+ "train_runtime": 221913.3786,
+ "train_tokens_per_second": 5653.476
+ },
+ {
+ "epoch": 1.9317263668192837,
+ "grad_norm": 0.8039083333559773,
+ "learning_rate": 2.8782257868553154e-08,
+ "loss": 0.2789355754852295,
+ "num_input_tokens_seen": 1254798040,
+ "step": 28690,
+ "train_runtime": 221950.8121,
+ "train_tokens_per_second": 5653.496
+ },
+ {
+ "epoch": 1.9320630218152437,
+ "grad_norm": 0.7939203667614896,
+ "learning_rate": 2.849965630362328e-08,
+ "loss": 0.2943142890930176,
+ "num_input_tokens_seen": 1255008552,
+ "step": 28695,
+ "train_runtime": 221991.6931,
+ "train_tokens_per_second": 5653.403
+ },
+ {
+ "epoch": 1.9323996768112037,
+ "grad_norm": 1.040825902674123,
+ "learning_rate": 2.8218445002912153e-08,
+ "loss": 0.2914332389831543,
+ "num_input_tokens_seen": 1255234552,
+ "step": 28700,
+ "train_runtime": 222032.4922,
+ "train_tokens_per_second": 5653.382
+ },
+ {
+ "epoch": 1.932736331807164,
+ "grad_norm": 0.7879730035255785,
+ "learning_rate": 2.7938624045059094e-08,
+ "loss": 0.2954190731048584,
+ "num_input_tokens_seen": 1255459816,
+ "step": 28705,
+ "train_runtime": 222067.1659,
+ "train_tokens_per_second": 5653.514
+ },
+ {
+ "epoch": 1.9330729868031242,
+ "grad_norm": 0.8446326753294154,
+ "learning_rate": 2.766019350831428e-08,
+ "loss": 0.3093540668487549,
+ "num_input_tokens_seen": 1255699664,
+ "step": 28710,
+ "train_runtime": 222106.7303,
+ "train_tokens_per_second": 5653.587
+ },
+ {
+ "epoch": 1.9334096417990843,
+ "grad_norm": 0.7880511342051334,
+ "learning_rate": 2.7383153470541546e-08,
+ "loss": 0.3093677282333374,
+ "num_input_tokens_seen": 1255930096,
+ "step": 28715,
+ "train_runtime": 222139.5127,
+ "train_tokens_per_second": 5653.79
+ },
+ {
+ "epoch": 1.9337462967950443,
+ "grad_norm": 0.7912632326756309,
+ "learning_rate": 2.710750400921336e-08,
+ "loss": 0.2810804843902588,
+ "num_input_tokens_seen": 1256129960,
+ "step": 28720,
+ "train_runtime": 222179.3382,
+ "train_tokens_per_second": 5653.676
+ },
+ {
+ "epoch": 1.9340829517910045,
+ "grad_norm": 0.7097484635650013,
+ "learning_rate": 2.683324520141417e-08,
+ "loss": 0.29244356155395507,
+ "num_input_tokens_seen": 1256355072,
+ "step": 28725,
+ "train_runtime": 222220.9674,
+ "train_tokens_per_second": 5653.63
+ },
+ {
+ "epoch": 1.9344196067869648,
+ "grad_norm": 0.8119430664301859,
+ "learning_rate": 2.65603771238393e-08,
+ "loss": 0.3020967483520508,
+ "num_input_tokens_seen": 1256563176,
+ "step": 28730,
+ "train_runtime": 222266.5769,
+ "train_tokens_per_second": 5653.406
+ },
+ {
+ "epoch": 1.9347562617829248,
+ "grad_norm": 0.7459006681748622,
+ "learning_rate": 2.6288899852796036e-08,
+ "loss": 0.2978071689605713,
+ "num_input_tokens_seen": 1256797928,
+ "step": 28735,
+ "train_runtime": 222306.2836,
+ "train_tokens_per_second": 5653.452
+ },
+ {
+ "epoch": 1.9350929167788848,
+ "grad_norm": 0.8087885741576003,
+ "learning_rate": 2.6018813464202543e-08,
+ "loss": 0.302975058555603,
+ "num_input_tokens_seen": 1257009296,
+ "step": 28740,
+ "train_runtime": 222341.5433,
+ "train_tokens_per_second": 5653.506
+ },
+ {
+ "epoch": 1.935429571774845,
+ "grad_norm": 0.727456579862242,
+ "learning_rate": 2.575011803358618e-08,
+ "loss": 0.29305691719055177,
+ "num_input_tokens_seen": 1257233712,
+ "step": 28745,
+ "train_runtime": 222380.0158,
+ "train_tokens_per_second": 5653.537
+ },
+ {
+ "epoch": 1.9357662267708053,
+ "grad_norm": 0.6572939384489086,
+ "learning_rate": 2.5482813636087955e-08,
+ "loss": 0.27554895877838137,
+ "num_input_tokens_seen": 1257442168,
+ "step": 28750,
+ "train_runtime": 222423.441,
+ "train_tokens_per_second": 5653.371
+ },
+ {
+ "epoch": 1.9361028817667654,
+ "grad_norm": 0.7993380569308158,
+ "learning_rate": 2.5216900346458073e-08,
+ "loss": 0.2871532678604126,
+ "num_input_tokens_seen": 1257660464,
+ "step": 28755,
+ "train_runtime": 222466.1845,
+ "train_tokens_per_second": 5653.266
+ },
+ {
+ "epoch": 1.9364395367627254,
+ "grad_norm": 0.766816269617841,
+ "learning_rate": 2.4952378239058716e-08,
+ "loss": 0.29356064796447756,
+ "num_input_tokens_seen": 1257889496,
+ "step": 28760,
+ "train_runtime": 222499.466,
+ "train_tokens_per_second": 5653.45
+ },
+ {
+ "epoch": 1.9367761917586857,
+ "grad_norm": 0.7954103166328962,
+ "learning_rate": 2.4689247387862934e-08,
+ "loss": 0.2963290214538574,
+ "num_input_tokens_seen": 1258103416,
+ "step": 28765,
+ "train_runtime": 222532.7425,
+ "train_tokens_per_second": 5653.565
+ },
+ {
+ "epoch": 1.937112846754646,
+ "grad_norm": 0.7997832806288906,
+ "learning_rate": 2.4427507866453537e-08,
+ "loss": 0.28399271965026857,
+ "num_input_tokens_seen": 1258294944,
+ "step": 28770,
+ "train_runtime": 222573.7553,
+ "train_tokens_per_second": 5653.384
+ },
+ {
+ "epoch": 1.937449501750606,
+ "grad_norm": 0.885443863513371,
+ "learning_rate": 2.416715974802586e-08,
+ "loss": 0.2864212989807129,
+ "num_input_tokens_seen": 1258505472,
+ "step": 28775,
+ "train_runtime": 222608.5067,
+ "train_tokens_per_second": 5653.447
+ },
+ {
+ "epoch": 1.937786156746566,
+ "grad_norm": 0.7164017486860768,
+ "learning_rate": 2.3908203105384997e-08,
+ "loss": 0.30834760665893557,
+ "num_input_tokens_seen": 1258727136,
+ "step": 28780,
+ "train_runtime": 222642.3336,
+ "train_tokens_per_second": 5653.584
+ },
+ {
+ "epoch": 1.9381228117425262,
+ "grad_norm": 0.7587198432537924,
+ "learning_rate": 2.365063801094747e-08,
+ "loss": 0.2865851402282715,
+ "num_input_tokens_seen": 1258942528,
+ "step": 28785,
+ "train_runtime": 222676.6074,
+ "train_tokens_per_second": 5653.681
+ },
+ {
+ "epoch": 1.9384594667384865,
+ "grad_norm": 0.706256825592826,
+ "learning_rate": 2.3394464536740658e-08,
+ "loss": 0.30954954624176023,
+ "num_input_tokens_seen": 1259175632,
+ "step": 28790,
+ "train_runtime": 222718.1525,
+ "train_tokens_per_second": 5653.673
+ },
+ {
+ "epoch": 1.9387961217344465,
+ "grad_norm": 0.8885250264575596,
+ "learning_rate": 2.3139682754402816e-08,
+ "loss": 0.3149877548217773,
+ "num_input_tokens_seen": 1259405216,
+ "step": 28795,
+ "train_runtime": 222755.681,
+ "train_tokens_per_second": 5653.751
+ },
+ {
+ "epoch": 1.9391327767304065,
+ "grad_norm": 0.7626354653879915,
+ "learning_rate": 2.2886292735182502e-08,
+ "loss": 0.2907993316650391,
+ "num_input_tokens_seen": 1259637856,
+ "step": 28800,
+ "train_runtime": 222792.0279,
+ "train_tokens_per_second": 5653.873
+ },
+ {
+ "epoch": 1.9394694317263668,
+ "grad_norm": 0.7549952496575574,
+ "learning_rate": 2.2634294549939705e-08,
+ "loss": 0.2934614896774292,
+ "num_input_tokens_seen": 1259866088,
+ "step": 28805,
+ "train_runtime": 222833.3252,
+ "train_tokens_per_second": 5653.85
+ },
+ {
+ "epoch": 1.939806086722327,
+ "grad_norm": 0.7716438013359953,
+ "learning_rate": 2.2383688269144723e-08,
+ "loss": 0.3583078145980835,
+ "num_input_tokens_seen": 1260088088,
+ "step": 28810,
+ "train_runtime": 222873.4346,
+ "train_tokens_per_second": 5653.828
+ },
+ {
+ "epoch": 1.940142741718287,
+ "grad_norm": 0.7760006769888061,
+ "learning_rate": 2.2134473962878712e-08,
+ "loss": 0.2929667949676514,
+ "num_input_tokens_seen": 1260302240,
+ "step": 28815,
+ "train_runtime": 222909.5666,
+ "train_tokens_per_second": 5653.872
+ },
+ {
+ "epoch": 1.940479396714247,
+ "grad_norm": 0.9788787208283655,
+ "learning_rate": 2.1886651700833705e-08,
+ "loss": 0.28723697662353515,
+ "num_input_tokens_seen": 1260489696,
+ "step": 28820,
+ "train_runtime": 222944.9358,
+ "train_tokens_per_second": 5653.816
+ },
+ {
+ "epoch": 1.9408160517102075,
+ "grad_norm": 0.6725253717418324,
+ "learning_rate": 2.1640221552312603e-08,
+ "loss": 0.3041691780090332,
+ "num_input_tokens_seen": 1260707328,
+ "step": 28825,
+ "train_runtime": 222983.6272,
+ "train_tokens_per_second": 5653.811
+ },
+ {
+ "epoch": 1.9411527067061676,
+ "grad_norm": 1.0073697163500734,
+ "learning_rate": 2.1395183586229163e-08,
+ "loss": 0.30638680458068845,
+ "num_input_tokens_seen": 1260942368,
+ "step": 28830,
+ "train_runtime": 223022.3981,
+ "train_tokens_per_second": 5653.882
+ },
+ {
+ "epoch": 1.9414893617021276,
+ "grad_norm": 0.7599893500344853,
+ "learning_rate": 2.1151537871106353e-08,
+ "loss": 0.28965296745300295,
+ "num_input_tokens_seen": 1261152728,
+ "step": 28835,
+ "train_runtime": 223058.9354,
+ "train_tokens_per_second": 5653.899
+ },
+ {
+ "epoch": 1.9418260166980879,
+ "grad_norm": 0.7939541327760055,
+ "learning_rate": 2.0909284475080226e-08,
+ "loss": 0.285279655456543,
+ "num_input_tokens_seen": 1261373808,
+ "step": 28840,
+ "train_runtime": 223107.3758,
+ "train_tokens_per_second": 5653.663
+ },
+ {
+ "epoch": 1.942162671694048,
+ "grad_norm": 0.8217446318942616,
+ "learning_rate": 2.0668423465894928e-08,
+ "loss": 0.2754975318908691,
+ "num_input_tokens_seen": 1261592168,
+ "step": 28845,
+ "train_runtime": 223148.9614,
+ "train_tokens_per_second": 5653.587
+ },
+ {
+ "epoch": 1.9424993266900081,
+ "grad_norm": 0.7874416991261182,
+ "learning_rate": 2.0428954910907684e-08,
+ "loss": 0.28652522563934324,
+ "num_input_tokens_seen": 1261787104,
+ "step": 28850,
+ "train_runtime": 223186.7342,
+ "train_tokens_per_second": 5653.504
+ },
+ {
+ "epoch": 1.9428359816859682,
+ "grad_norm": 0.8264865579012179,
+ "learning_rate": 2.019087887708382e-08,
+ "loss": 0.3006443977355957,
+ "num_input_tokens_seen": 1262007728,
+ "step": 28855,
+ "train_runtime": 223227.7335,
+ "train_tokens_per_second": 5653.454
+ },
+ {
+ "epoch": 1.9431726366819284,
+ "grad_norm": 0.7815316131606884,
+ "learning_rate": 1.9954195431001744e-08,
+ "loss": 0.29778175354003905,
+ "num_input_tokens_seen": 1262233872,
+ "step": 28860,
+ "train_runtime": 223262.3109,
+ "train_tokens_per_second": 5653.591
+ },
+ {
+ "epoch": 1.9435092916778887,
+ "grad_norm": 0.7625140991861158,
+ "learning_rate": 1.9718904638848513e-08,
+ "loss": 0.2813864231109619,
+ "num_input_tokens_seen": 1262445256,
+ "step": 28865,
+ "train_runtime": 223305.0134,
+ "train_tokens_per_second": 5653.457
+ },
+ {
+ "epoch": 1.9438459466738487,
+ "grad_norm": 0.819180819579518,
+ "learning_rate": 1.9485006566422602e-08,
+ "loss": 0.2955453395843506,
+ "num_input_tokens_seen": 1262680216,
+ "step": 28870,
+ "train_runtime": 223347.1343,
+ "train_tokens_per_second": 5653.443
+ },
+ {
+ "epoch": 1.9441826016698087,
+ "grad_norm": 0.7141341923837556,
+ "learning_rate": 1.9252501279132806e-08,
+ "loss": 0.3031654357910156,
+ "num_input_tokens_seen": 1262895920,
+ "step": 28875,
+ "train_runtime": 223391.4022,
+ "train_tokens_per_second": 5653.288
+ },
+ {
+ "epoch": 1.944519256665769,
+ "grad_norm": 0.7599819750131713,
+ "learning_rate": 1.902138884199878e-08,
+ "loss": 0.2994793176651001,
+ "num_input_tokens_seen": 1263122360,
+ "step": 28880,
+ "train_runtime": 223427.1868,
+ "train_tokens_per_second": 5653.396
+ },
+ {
+ "epoch": 1.9448559116617292,
+ "grad_norm": 0.7754812262214777,
+ "learning_rate": 1.879166931964993e-08,
+ "loss": 0.30690946578979494,
+ "num_input_tokens_seen": 1263337688,
+ "step": 28885,
+ "train_runtime": 223467.6117,
+ "train_tokens_per_second": 5653.337
+ },
+ {
+ "epoch": 1.9451925666576892,
+ "grad_norm": 0.7766593010208253,
+ "learning_rate": 1.856334277632765e-08,
+ "loss": 0.29476094245910645,
+ "num_input_tokens_seen": 1263574904,
+ "step": 28890,
+ "train_runtime": 223506.8009,
+ "train_tokens_per_second": 5653.407
+ },
+ {
+ "epoch": 1.9455292216536493,
+ "grad_norm": 0.6831350388874682,
+ "learning_rate": 1.8336409275880872e-08,
+ "loss": 0.3002762794494629,
+ "num_input_tokens_seen": 1263807760,
+ "step": 28895,
+ "train_runtime": 223552.6379,
+ "train_tokens_per_second": 5653.289
+ },
+ {
+ "epoch": 1.9458658766496095,
+ "grad_norm": 0.7320234419132584,
+ "learning_rate": 1.8110868881772713e-08,
+ "loss": 0.2916655302047729,
+ "num_input_tokens_seen": 1264016568,
+ "step": 28900,
+ "train_runtime": 223592.718,
+ "train_tokens_per_second": 5653.21
+ },
+ {
+ "epoch": 1.9462025316455698,
+ "grad_norm": 0.7177337064879638,
+ "learning_rate": 1.788672165707328e-08,
+ "loss": 0.2689183712005615,
+ "num_input_tokens_seen": 1264236144,
+ "step": 28905,
+ "train_runtime": 223637.4105,
+ "train_tokens_per_second": 5653.062
+ },
+ {
+ "epoch": 1.9465391866415298,
+ "grad_norm": 0.7335996504510445,
+ "learning_rate": 1.7663967664465763e-08,
+ "loss": 0.2718139886856079,
+ "num_input_tokens_seen": 1264443160,
+ "step": 28910,
+ "train_runtime": 223677.4438,
+ "train_tokens_per_second": 5652.976
+ },
+ {
+ "epoch": 1.9468758416374898,
+ "grad_norm": 0.7751823359014647,
+ "learning_rate": 1.7442606966242005e-08,
+ "loss": 0.3149592399597168,
+ "num_input_tokens_seen": 1264681216,
+ "step": 28915,
+ "train_runtime": 223720.453,
+ "train_tokens_per_second": 5652.953
+ },
+ {
+ "epoch": 1.94721249663345,
+ "grad_norm": 0.7519570365655589,
+ "learning_rate": 1.722263962430526e-08,
+ "loss": 0.3035134315490723,
+ "num_input_tokens_seen": 1264904128,
+ "step": 28920,
+ "train_runtime": 223755.8381,
+ "train_tokens_per_second": 5653.055
+ },
+ {
+ "epoch": 1.9475491516294103,
+ "grad_norm": 0.828398864844539,
+ "learning_rate": 1.7004065700168547e-08,
+ "loss": 0.28637185096740725,
+ "num_input_tokens_seen": 1265126560,
+ "step": 28925,
+ "train_runtime": 223788.0098,
+ "train_tokens_per_second": 5653.237
+ },
+ {
+ "epoch": 1.9478858066253704,
+ "grad_norm": 0.700009166031884,
+ "learning_rate": 1.6786885254954644e-08,
+ "loss": 0.3050400733947754,
+ "num_input_tokens_seen": 1265361040,
+ "step": 28930,
+ "train_runtime": 223824.3291,
+ "train_tokens_per_second": 5653.367
+ },
+ {
+ "epoch": 1.9482224616213304,
+ "grad_norm": 0.8369964157697908,
+ "learning_rate": 1.6571098349398296e-08,
+ "loss": 0.2956183910369873,
+ "num_input_tokens_seen": 1265565976,
+ "step": 28935,
+ "train_runtime": 223863.4274,
+ "train_tokens_per_second": 5653.295
+ },
+ {
+ "epoch": 1.9485591166172906,
+ "grad_norm": 0.8754073900671024,
+ "learning_rate": 1.635670504384346e-08,
+ "loss": 0.31615753173828126,
+ "num_input_tokens_seen": 1265765816,
+ "step": 28940,
+ "train_runtime": 223908.9438,
+ "train_tokens_per_second": 5653.038
+ },
+ {
+ "epoch": 1.948895771613251,
+ "grad_norm": 0.941511353921259,
+ "learning_rate": 1.6143705398243837e-08,
+ "loss": 0.29006495475769045,
+ "num_input_tokens_seen": 1265969312,
+ "step": 28945,
+ "train_runtime": 223952.0727,
+ "train_tokens_per_second": 5652.858
+ },
+ {
+ "epoch": 1.949232426609211,
+ "grad_norm": 0.7913567769706255,
+ "learning_rate": 1.5932099472165674e-08,
+ "loss": 0.29108805656433107,
+ "num_input_tokens_seen": 1266179488,
+ "step": 28950,
+ "train_runtime": 223990.7097,
+ "train_tokens_per_second": 5652.821
+ },
+ {
+ "epoch": 1.949569081605171,
+ "grad_norm": 0.7310227670184213,
+ "learning_rate": 1.572188732478164e-08,
+ "loss": 0.2916900157928467,
+ "num_input_tokens_seen": 1266409776,
+ "step": 28955,
+ "train_runtime": 224030.8188,
+ "train_tokens_per_second": 5652.837
+ },
+ {
+ "epoch": 1.9499057366011312,
+ "grad_norm": 0.7938589935789288,
+ "learning_rate": 1.551306901487859e-08,
+ "loss": 0.3035112857818604,
+ "num_input_tokens_seen": 1266624136,
+ "step": 28960,
+ "train_runtime": 224074.3669,
+ "train_tokens_per_second": 5652.695
+ },
+ {
+ "epoch": 1.9502423915970915,
+ "grad_norm": 0.6920718267717534,
+ "learning_rate": 1.5305644600852043e-08,
+ "loss": 0.2758214473724365,
+ "num_input_tokens_seen": 1266855056,
+ "step": 28965,
+ "train_runtime": 224114.5981,
+ "train_tokens_per_second": 5652.711
+ },
+ {
+ "epoch": 1.9505790465930515,
+ "grad_norm": 0.7554921758498927,
+ "learning_rate": 1.5099614140706154e-08,
+ "loss": 0.2854607105255127,
+ "num_input_tokens_seen": 1267072344,
+ "step": 28970,
+ "train_runtime": 224158.7838,
+ "train_tokens_per_second": 5652.566
+ },
+ {
+ "epoch": 1.9509157015890115,
+ "grad_norm": 0.8000972769551457,
+ "learning_rate": 1.489497769205761e-08,
+ "loss": 0.31021819114685056,
+ "num_input_tokens_seen": 1267277496,
+ "step": 28975,
+ "train_runtime": 224199.1913,
+ "train_tokens_per_second": 5652.462
+ },
+ {
+ "epoch": 1.9512523565849718,
+ "grad_norm": 0.7228114152004868,
+ "learning_rate": 1.4691735312132304e-08,
+ "loss": 0.2921774864196777,
+ "num_input_tokens_seen": 1267506744,
+ "step": 28980,
+ "train_runtime": 224237.6676,
+ "train_tokens_per_second": 5652.515
+ },
+ {
+ "epoch": 1.951589011580932,
+ "grad_norm": 0.7982672860216988,
+ "learning_rate": 1.4489887057766439e-08,
+ "loss": 0.30820393562316895,
+ "num_input_tokens_seen": 1267724224,
+ "step": 28985,
+ "train_runtime": 224272.862,
+ "train_tokens_per_second": 5652.598
+ },
+ {
+ "epoch": 1.951925666576892,
+ "grad_norm": 0.7019172612562962,
+ "learning_rate": 1.428943298540597e-08,
+ "loss": 0.29095611572265623,
+ "num_input_tokens_seen": 1267950120,
+ "step": 28990,
+ "train_runtime": 224309.6136,
+ "train_tokens_per_second": 5652.678
+ },
+ {
+ "epoch": 1.952262321572852,
+ "grad_norm": 0.6605820071712591,
+ "learning_rate": 1.4090373151107173e-08,
+ "loss": 0.3069890022277832,
+ "num_input_tokens_seen": 1268176376,
+ "step": 28995,
+ "train_runtime": 224342.3845,
+ "train_tokens_per_second": 5652.861
+ },
+ {
+ "epoch": 1.9525989765688123,
+ "grad_norm": 0.7607077702464043,
+ "learning_rate": 1.3892707610536627e-08,
+ "loss": 0.2565622329711914,
+ "num_input_tokens_seen": 1268384472,
+ "step": 29000,
+ "train_runtime": 224386.5442,
+ "train_tokens_per_second": 5652.676
+ },
+ {
+ "epoch": 1.9529356315647726,
+ "grad_norm": 0.806764475508148,
+ "learning_rate": 1.3696436418970672e-08,
+ "loss": 0.2830059051513672,
+ "num_input_tokens_seen": 1268601536,
+ "step": 29005,
+ "train_runtime": 224428.0534,
+ "train_tokens_per_second": 5652.598
+ },
+ {
+ "epoch": 1.9532722865607326,
+ "grad_norm": 0.8503139141015114,
+ "learning_rate": 1.3501559631296512e-08,
+ "loss": 0.31444830894470216,
+ "num_input_tokens_seen": 1268805592,
+ "step": 29010,
+ "train_runtime": 224471.139,
+ "train_tokens_per_second": 5652.422
+ },
+ {
+ "epoch": 1.9536089415566926,
+ "grad_norm": 0.7710744016084916,
+ "learning_rate": 1.3308077302010557e-08,
+ "loss": 0.31547060012817385,
+ "num_input_tokens_seen": 1269039096,
+ "step": 29015,
+ "train_runtime": 224509.2303,
+ "train_tokens_per_second": 5652.503
+ },
+ {
+ "epoch": 1.9539455965526529,
+ "grad_norm": 0.8026179441993274,
+ "learning_rate": 1.3115989485218972e-08,
+ "loss": 0.2989495754241943,
+ "num_input_tokens_seen": 1269246368,
+ "step": 29020,
+ "train_runtime": 224543.5644,
+ "train_tokens_per_second": 5652.562
+ },
+ {
+ "epoch": 1.9542822515486131,
+ "grad_norm": 0.805001102582023,
+ "learning_rate": 1.2925296234638784e-08,
+ "loss": 0.2807016849517822,
+ "num_input_tokens_seen": 1269464104,
+ "step": 29025,
+ "train_runtime": 224584.5623,
+ "train_tokens_per_second": 5652.499
+ },
+ {
+ "epoch": 1.9546189065445732,
+ "grad_norm": 0.6954871573935902,
+ "learning_rate": 1.2735997603597339e-08,
+ "loss": 0.29611780643463137,
+ "num_input_tokens_seen": 1269699728,
+ "step": 29030,
+ "train_runtime": 224616.5365,
+ "train_tokens_per_second": 5652.744
+ },
+ {
+ "epoch": 1.9549555615405332,
+ "grad_norm": 0.857067719947332,
+ "learning_rate": 1.2548093645030623e-08,
+ "loss": 0.2928828001022339,
+ "num_input_tokens_seen": 1269928480,
+ "step": 29035,
+ "train_runtime": 224656.3805,
+ "train_tokens_per_second": 5652.759
+ },
+ {
+ "epoch": 1.9552922165364934,
+ "grad_norm": 0.7481342627141493,
+ "learning_rate": 1.236158441148605e-08,
+ "loss": 0.28606395721435546,
+ "num_input_tokens_seen": 1270155816,
+ "step": 29040,
+ "train_runtime": 224692.812,
+ "train_tokens_per_second": 5652.855
+ },
+ {
+ "epoch": 1.9556288715324537,
+ "grad_norm": 0.8112388374034099,
+ "learning_rate": 1.2176469955119119e-08,
+ "loss": 0.32508511543273927,
+ "num_input_tokens_seen": 1270385632,
+ "step": 29045,
+ "train_runtime": 224733.026,
+ "train_tokens_per_second": 5652.866
+ },
+ {
+ "epoch": 1.9559655265284137,
+ "grad_norm": 0.8038650607300725,
+ "learning_rate": 1.199275032769842e-08,
+ "loss": 0.3148582696914673,
+ "num_input_tokens_seen": 1270610312,
+ "step": 29050,
+ "train_runtime": 224776.4013,
+ "train_tokens_per_second": 5652.775
+ },
+ {
+ "epoch": 1.9563021815243737,
+ "grad_norm": 0.8038054887828693,
+ "learning_rate": 1.1810425580598971e-08,
+ "loss": 0.32245888710021975,
+ "num_input_tokens_seen": 1270844152,
+ "step": 29055,
+ "train_runtime": 224817.2922,
+ "train_tokens_per_second": 5652.786
+ },
+ {
+ "epoch": 1.956638836520334,
+ "grad_norm": 0.8285772488267888,
+ "learning_rate": 1.1629495764807763e-08,
+ "loss": 0.29970650672912597,
+ "num_input_tokens_seen": 1271079472,
+ "step": 29060,
+ "train_runtime": 224861.1966,
+ "train_tokens_per_second": 5652.729
+ },
+ {
+ "epoch": 1.9569754915162942,
+ "grad_norm": 0.7500252998485927,
+ "learning_rate": 1.1449960930921544e-08,
+ "loss": 0.29777934551239016,
+ "num_input_tokens_seen": 1271292624,
+ "step": 29065,
+ "train_runtime": 224898.4968,
+ "train_tokens_per_second": 5652.74
+ },
+ {
+ "epoch": 1.9573121465122543,
+ "grad_norm": 0.7492263033388418,
+ "learning_rate": 1.1271821129146266e-08,
+ "loss": 0.28283419609069826,
+ "num_input_tokens_seen": 1271514960,
+ "step": 29070,
+ "train_runtime": 224932.5444,
+ "train_tokens_per_second": 5652.872
+ },
+ {
+ "epoch": 1.9576488015082143,
+ "grad_norm": 0.794721710026827,
+ "learning_rate": 1.1095076409298189e-08,
+ "loss": 0.28227386474609373,
+ "num_input_tokens_seen": 1271727712,
+ "step": 29075,
+ "train_runtime": 224970.8077,
+ "train_tokens_per_second": 5652.857
+ },
+ {
+ "epoch": 1.9579854565041745,
+ "grad_norm": 0.8361082033343284,
+ "learning_rate": 1.0919726820803889e-08,
+ "loss": 0.2864542484283447,
+ "num_input_tokens_seen": 1271937480,
+ "step": 29080,
+ "train_runtime": 225007.7296,
+ "train_tokens_per_second": 5652.861
+ },
+ {
+ "epoch": 1.9583221115001348,
+ "grad_norm": 0.7529640945800179,
+ "learning_rate": 1.0745772412698585e-08,
+ "loss": 0.2920374870300293,
+ "num_input_tokens_seen": 1272150888,
+ "step": 29085,
+ "train_runtime": 225047.2154,
+ "train_tokens_per_second": 5652.818
+ },
+ {
+ "epoch": 1.9586587664960948,
+ "grad_norm": 0.7725658212667705,
+ "learning_rate": 1.057321323362892e-08,
+ "loss": 0.28984389305114744,
+ "num_input_tokens_seen": 1272385000,
+ "step": 29090,
+ "train_runtime": 225088.6801,
+ "train_tokens_per_second": 5652.816
+ },
+ {
+ "epoch": 1.9589954214920549,
+ "grad_norm": 0.7321569817209439,
+ "learning_rate": 1.0402049331849629e-08,
+ "loss": 0.2945473432540894,
+ "num_input_tokens_seen": 1272608520,
+ "step": 29095,
+ "train_runtime": 225120.0817,
+ "train_tokens_per_second": 5653.021
+ },
+ {
+ "epoch": 1.959332076488015,
+ "grad_norm": 0.7684463547894764,
+ "learning_rate": 1.0232280755226865e-08,
+ "loss": 0.30609171390533446,
+ "num_input_tokens_seen": 1272824152,
+ "step": 29100,
+ "train_runtime": 225164.1788,
+ "train_tokens_per_second": 5652.871
+ },
+ {
+ "epoch": 1.9596687314839754,
+ "grad_norm": 0.6970891800486236,
+ "learning_rate": 1.006390755123543e-08,
+ "loss": 0.2889437675476074,
+ "num_input_tokens_seen": 1273039232,
+ "step": 29105,
+ "train_runtime": 225204.746,
+ "train_tokens_per_second": 5652.808
+ },
+ {
+ "epoch": 1.9600053864799354,
+ "grad_norm": 0.8951268014858879,
+ "learning_rate": 9.89692976696044e-09,
+ "loss": 0.3118439674377441,
+ "num_input_tokens_seen": 1273253968,
+ "step": 29110,
+ "train_runtime": 225250.2013,
+ "train_tokens_per_second": 5652.621
+ },
+ {
+ "epoch": 1.9603420414758954,
+ "grad_norm": 0.7116868073327594,
+ "learning_rate": 9.731347449097316e-09,
+ "loss": 0.31283159255981446,
+ "num_input_tokens_seen": 1273474224,
+ "step": 29115,
+ "train_runtime": 225292.9511,
+ "train_tokens_per_second": 5652.526
+ },
+ {
+ "epoch": 1.9606786964718557,
+ "grad_norm": 0.7624467705283283,
+ "learning_rate": 9.56716064394958e-09,
+ "loss": 0.30971055030822753,
+ "num_input_tokens_seen": 1273701776,
+ "step": 29120,
+ "train_runtime": 225324.6547,
+ "train_tokens_per_second": 5652.74
+ },
+ {
+ "epoch": 1.961015351467816,
+ "grad_norm": 0.8807522971861178,
+ "learning_rate": 9.404369397432166e-09,
+ "loss": 0.287065315246582,
+ "num_input_tokens_seen": 1273920400,
+ "step": 29125,
+ "train_runtime": 225360.8369,
+ "train_tokens_per_second": 5652.803
+ },
+ {
+ "epoch": 1.961352006463776,
+ "grad_norm": 0.766183012473875,
+ "learning_rate": 9.242973755068663e-09,
+ "loss": 0.2945993900299072,
+ "num_input_tokens_seen": 1274143544,
+ "step": 29130,
+ "train_runtime": 225396.198,
+ "train_tokens_per_second": 5652.906
+ },
+ {
+ "epoch": 1.961688661459736,
+ "grad_norm": 0.8619210725642108,
+ "learning_rate": 9.082973761993518e-09,
+ "loss": 0.2969123363494873,
+ "num_input_tokens_seen": 1274356376,
+ "step": 29135,
+ "train_runtime": 225437.914,
+ "train_tokens_per_second": 5652.804
+ },
+ {
+ "epoch": 1.9620253164556962,
+ "grad_norm": 0.878826623744484,
+ "learning_rate": 8.924369462949834e-09,
+ "loss": 0.3293659448623657,
+ "num_input_tokens_seen": 1274559280,
+ "step": 29140,
+ "train_runtime": 225478.4148,
+ "train_tokens_per_second": 5652.689
+ },
+ {
+ "epoch": 1.9623619714516565,
+ "grad_norm": 0.795997521175958,
+ "learning_rate": 8.767160902291016e-09,
+ "loss": 0.2915841817855835,
+ "num_input_tokens_seen": 1274776792,
+ "step": 29145,
+ "train_runtime": 225514.841,
+ "train_tokens_per_second": 5652.74
+ },
+ {
+ "epoch": 1.9626986264476165,
+ "grad_norm": 0.7676235515949815,
+ "learning_rate": 8.61134812397968e-09,
+ "loss": 0.30088181495666505,
+ "num_input_tokens_seen": 1274996960,
+ "step": 29150,
+ "train_runtime": 225551.2935,
+ "train_tokens_per_second": 5652.803
+ },
+ {
+ "epoch": 1.9630352814435765,
+ "grad_norm": 0.7939420108805821,
+ "learning_rate": 8.45693117158819e-09,
+ "loss": 0.2771022796630859,
+ "num_input_tokens_seen": 1275203528,
+ "step": 29155,
+ "train_runtime": 225587.6966,
+ "train_tokens_per_second": 5652.806
+ },
+ {
+ "epoch": 1.9633719364395368,
+ "grad_norm": 0.8267377385515322,
+ "learning_rate": 8.303910088299228e-09,
+ "loss": 0.2890202045440674,
+ "num_input_tokens_seen": 1275420104,
+ "step": 29160,
+ "train_runtime": 225627.9039,
+ "train_tokens_per_second": 5652.759
+ },
+ {
+ "epoch": 1.963708591435497,
+ "grad_norm": 0.7872969059306067,
+ "learning_rate": 8.152284916904674e-09,
+ "loss": 0.297850513458252,
+ "num_input_tokens_seen": 1275648864,
+ "step": 29165,
+ "train_runtime": 225665.3758,
+ "train_tokens_per_second": 5652.834
+ },
+ {
+ "epoch": 1.964045246431457,
+ "grad_norm": 0.7526544096353636,
+ "learning_rate": 8.002055699805612e-09,
+ "loss": 0.2845508098602295,
+ "num_input_tokens_seen": 1275859328,
+ "step": 29170,
+ "train_runtime": 225711.406,
+ "train_tokens_per_second": 5652.613
+ },
+ {
+ "epoch": 1.964381901427417,
+ "grad_norm": 0.7307972307788215,
+ "learning_rate": 7.853222479013434e-09,
+ "loss": 0.28490164279937746,
+ "num_input_tokens_seen": 1276098304,
+ "step": 29175,
+ "train_runtime": 225747.3603,
+ "train_tokens_per_second": 5652.772
+ },
+ {
+ "epoch": 1.9647185564233773,
+ "grad_norm": 0.7126777979052846,
+ "learning_rate": 7.705785296148737e-09,
+ "loss": 0.29223477840423584,
+ "num_input_tokens_seen": 1276313944,
+ "step": 29180,
+ "train_runtime": 225782.9902,
+ "train_tokens_per_second": 5652.835
+ },
+ {
+ "epoch": 1.9650552114193376,
+ "grad_norm": 0.7780760910871221,
+ "learning_rate": 7.55974419244132e-09,
+ "loss": 0.3048643350601196,
+ "num_input_tokens_seen": 1276534976,
+ "step": 29185,
+ "train_runtime": 225815.2561,
+ "train_tokens_per_second": 5653.006
+ },
+ {
+ "epoch": 1.9653918664152976,
+ "grad_norm": 0.7124335281108779,
+ "learning_rate": 7.415099208732402e-09,
+ "loss": 0.2795270919799805,
+ "num_input_tokens_seen": 1276756464,
+ "step": 29190,
+ "train_runtime": 225850.916,
+ "train_tokens_per_second": 5653.094
+ },
+ {
+ "epoch": 1.9657285214112576,
+ "grad_norm": 0.7090900998979741,
+ "learning_rate": 7.2718503854707355e-09,
+ "loss": 0.3065217971801758,
+ "num_input_tokens_seen": 1276983008,
+ "step": 29195,
+ "train_runtime": 225891.8245,
+ "train_tokens_per_second": 5653.073
+ },
+ {
+ "epoch": 1.966065176407218,
+ "grad_norm": 0.8593525853790884,
+ "learning_rate": 7.129997762715391e-09,
+ "loss": 0.26297545433044434,
+ "num_input_tokens_seen": 1277205144,
+ "step": 29200,
+ "train_runtime": 225929.616,
+ "train_tokens_per_second": 5653.111
+ },
+ {
+ "epoch": 1.9664018314031781,
+ "grad_norm": 0.8188907230159181,
+ "learning_rate": 6.98954138013519e-09,
+ "loss": 0.2864915132522583,
+ "num_input_tokens_seen": 1277414784,
+ "step": 29205,
+ "train_runtime": 225967.4368,
+ "train_tokens_per_second": 5653.092
+ },
+ {
+ "epoch": 1.9667384863991382,
+ "grad_norm": 0.7346378968927514,
+ "learning_rate": 6.850481277008159e-09,
+ "loss": 0.3049713373184204,
+ "num_input_tokens_seen": 1277643008,
+ "step": 29210,
+ "train_runtime": 226014.3298,
+ "train_tokens_per_second": 5652.929
+ },
+ {
+ "epoch": 1.9670751413950982,
+ "grad_norm": 0.7674532896602834,
+ "learning_rate": 6.712817492222079e-09,
+ "loss": 0.32462234497070314,
+ "num_input_tokens_seen": 1277862488,
+ "step": 29215,
+ "train_runtime": 226054.1024,
+ "train_tokens_per_second": 5652.906
+ },
+ {
+ "epoch": 1.9674117963910585,
+ "grad_norm": 0.6823823423034494,
+ "learning_rate": 6.576550064273934e-09,
+ "loss": 0.29960381984710693,
+ "num_input_tokens_seen": 1278083144,
+ "step": 29220,
+ "train_runtime": 226087.3475,
+ "train_tokens_per_second": 5653.05
+ },
+ {
+ "epoch": 1.9677484513870187,
+ "grad_norm": 0.6768114788982357,
+ "learning_rate": 6.44167903127102e-09,
+ "loss": 0.30608859062194826,
+ "num_input_tokens_seen": 1278306920,
+ "step": 29225,
+ "train_runtime": 226128.6262,
+ "train_tokens_per_second": 5653.008
+ },
+ {
+ "epoch": 1.9680851063829787,
+ "grad_norm": 0.7867457093292909,
+ "learning_rate": 6.3082044309287215e-09,
+ "loss": 0.2798271656036377,
+ "num_input_tokens_seen": 1278530936,
+ "step": 29230,
+ "train_runtime": 226163.2382,
+ "train_tokens_per_second": 5653.133
+ },
+ {
+ "epoch": 1.9684217613789388,
+ "grad_norm": 0.7363218232718188,
+ "learning_rate": 6.176126300573848e-09,
+ "loss": 0.29613399505615234,
+ "num_input_tokens_seen": 1278747232,
+ "step": 29235,
+ "train_runtime": 226204.0451,
+ "train_tokens_per_second": 5653.07
+ },
+ {
+ "epoch": 1.968758416374899,
+ "grad_norm": 0.772729635299263,
+ "learning_rate": 6.045444677140744e-09,
+ "loss": 0.28275036811828613,
+ "num_input_tokens_seen": 1278973944,
+ "step": 29240,
+ "train_runtime": 226241.1384,
+ "train_tokens_per_second": 5653.145
+ },
+ {
+ "epoch": 1.9690950713708593,
+ "grad_norm": 0.7892151646300906,
+ "learning_rate": 5.91615959717462e-09,
+ "loss": 0.2926478385925293,
+ "num_input_tokens_seen": 1279195664,
+ "step": 29245,
+ "train_runtime": 226280.1718,
+ "train_tokens_per_second": 5653.15
+ },
+ {
+ "epoch": 1.9694317263668193,
+ "grad_norm": 0.6905193077369461,
+ "learning_rate": 5.78827109682989e-09,
+ "loss": 0.27014522552490233,
+ "num_input_tokens_seen": 1279416864,
+ "step": 29250,
+ "train_runtime": 226314.3102,
+ "train_tokens_per_second": 5653.274
+ },
+ {
+ "epoch": 1.9697683813627793,
+ "grad_norm": 0.7987578510711066,
+ "learning_rate": 5.661779211869056e-09,
+ "loss": 0.2848729848861694,
+ "num_input_tokens_seen": 1279631848,
+ "step": 29255,
+ "train_runtime": 226352.8714,
+ "train_tokens_per_second": 5653.261
+ },
+ {
+ "epoch": 1.9701050363587396,
+ "grad_norm": 0.7803335673063395,
+ "learning_rate": 5.536683977666601e-09,
+ "loss": 0.3105973243713379,
+ "num_input_tokens_seen": 1279865120,
+ "step": 29260,
+ "train_runtime": 226385.8852,
+ "train_tokens_per_second": 5653.467
+ },
+ {
+ "epoch": 1.9704416913546998,
+ "grad_norm": 0.6890074934726101,
+ "learning_rate": 5.412985429203988e-09,
+ "loss": 0.28957743644714357,
+ "num_input_tokens_seen": 1280078376,
+ "step": 29265,
+ "train_runtime": 226423.2957,
+ "train_tokens_per_second": 5653.475
+ },
+ {
+ "epoch": 1.9707783463506598,
+ "grad_norm": 0.7625841866074922,
+ "learning_rate": 5.290683601073543e-09,
+ "loss": 0.2784770965576172,
+ "num_input_tokens_seen": 1280304192,
+ "step": 29270,
+ "train_runtime": 226459.6491,
+ "train_tokens_per_second": 5653.564
+ },
+ {
+ "epoch": 1.9711150013466199,
+ "grad_norm": 0.7840668749162247,
+ "learning_rate": 5.169778527476243e-09,
+ "loss": 0.299863862991333,
+ "num_input_tokens_seen": 1280532272,
+ "step": 29275,
+ "train_runtime": 226498.8885,
+ "train_tokens_per_second": 5653.592
+ },
+ {
+ "epoch": 1.9714516563425801,
+ "grad_norm": 0.8729160388891067,
+ "learning_rate": 5.050270242222821e-09,
+ "loss": 0.31504406929016116,
+ "num_input_tokens_seen": 1280731456,
+ "step": 29280,
+ "train_runtime": 226536.9074,
+ "train_tokens_per_second": 5653.522
+ },
+ {
+ "epoch": 1.9717883113385404,
+ "grad_norm": 0.8134946456239271,
+ "learning_rate": 4.932158778733765e-09,
+ "loss": 0.2948625087738037,
+ "num_input_tokens_seen": 1280954184,
+ "step": 29285,
+ "train_runtime": 226580.571,
+ "train_tokens_per_second": 5653.416
+ },
+ {
+ "epoch": 1.9721249663345004,
+ "grad_norm": 0.9015410062046074,
+ "learning_rate": 4.8154441700387636e-09,
+ "loss": 0.3150008201599121,
+ "num_input_tokens_seen": 1281153920,
+ "step": 29290,
+ "train_runtime": 226613.4169,
+ "train_tokens_per_second": 5653.478
+ },
+ {
+ "epoch": 1.9724616213304604,
+ "grad_norm": 0.7835120290516451,
+ "learning_rate": 4.7001264487761545e-09,
+ "loss": 0.31816887855529785,
+ "num_input_tokens_seen": 1281377960,
+ "step": 29295,
+ "train_runtime": 226653.002,
+ "train_tokens_per_second": 5653.479
+ },
+ {
+ "epoch": 1.9727982763264207,
+ "grad_norm": 0.7177281483739487,
+ "learning_rate": 4.586205647194031e-09,
+ "loss": 0.3077888250350952,
+ "num_input_tokens_seen": 1281598760,
+ "step": 29300,
+ "train_runtime": 226691.7054,
+ "train_tokens_per_second": 5653.488
+ },
+ {
+ "epoch": 1.973134931322381,
+ "grad_norm": 0.8554091191816099,
+ "learning_rate": 4.4736817971507974e-09,
+ "loss": 0.3019340991973877,
+ "num_input_tokens_seen": 1281813592,
+ "step": 29305,
+ "train_runtime": 226737.7122,
+ "train_tokens_per_second": 5653.288
+ },
+ {
+ "epoch": 1.973471586318341,
+ "grad_norm": 0.7693611386783255,
+ "learning_rate": 4.362554930112395e-09,
+ "loss": 0.30186340808868406,
+ "num_input_tokens_seen": 1282029480,
+ "step": 29310,
+ "train_runtime": 226775.6925,
+ "train_tokens_per_second": 5653.293
+ },
+ {
+ "epoch": 1.973808241314301,
+ "grad_norm": 0.7405776360693889,
+ "learning_rate": 4.252825077155631e-09,
+ "loss": 0.2891343593597412,
+ "num_input_tokens_seen": 1282250600,
+ "step": 29315,
+ "train_runtime": 226814.7389,
+ "train_tokens_per_second": 5653.295
+ },
+ {
+ "epoch": 1.9741448963102612,
+ "grad_norm": 0.7831694807451932,
+ "learning_rate": 4.1444922689665156e-09,
+ "loss": 0.29642043113708494,
+ "num_input_tokens_seen": 1282463216,
+ "step": 29320,
+ "train_runtime": 226855.4258,
+ "train_tokens_per_second": 5653.218
+ },
+ {
+ "epoch": 1.9744815513062215,
+ "grad_norm": 0.8733709826420818,
+ "learning_rate": 4.037556535839149e-09,
+ "loss": 0.2692773103713989,
+ "num_input_tokens_seen": 1282685880,
+ "step": 29325,
+ "train_runtime": 226887.8468,
+ "train_tokens_per_second": 5653.392
+ },
+ {
+ "epoch": 1.9748182063021815,
+ "grad_norm": 0.7090978050641338,
+ "learning_rate": 3.932017907677943e-09,
+ "loss": 0.28195557594299314,
+ "num_input_tokens_seen": 1282915496,
+ "step": 29330,
+ "train_runtime": 226931.5302,
+ "train_tokens_per_second": 5653.315
+ },
+ {
+ "epoch": 1.9751548612981416,
+ "grad_norm": 0.7330650828079108,
+ "learning_rate": 3.827876413997067e-09,
+ "loss": 0.310672664642334,
+ "num_input_tokens_seen": 1283112440,
+ "step": 29335,
+ "train_runtime": 226962.5074,
+ "train_tokens_per_second": 5653.411
+ },
+ {
+ "epoch": 1.9754915162941018,
+ "grad_norm": 0.8645271417610324,
+ "learning_rate": 3.725132083918781e-09,
+ "loss": 0.2937079668045044,
+ "num_input_tokens_seen": 1283319208,
+ "step": 29340,
+ "train_runtime": 227004.0095,
+ "train_tokens_per_second": 5653.289
+ },
+ {
+ "epoch": 1.975828171290062,
+ "grad_norm": 0.7553494839592546,
+ "learning_rate": 3.623784946175102e-09,
+ "loss": 0.3103309631347656,
+ "num_input_tokens_seen": 1283519312,
+ "step": 29345,
+ "train_runtime": 227044.1149,
+ "train_tokens_per_second": 5653.171
+ },
+ {
+ "epoch": 1.976164826286022,
+ "grad_norm": 0.7711706358606312,
+ "learning_rate": 3.5238350291083577e-09,
+ "loss": 0.28435301780700684,
+ "num_input_tokens_seen": 1283726744,
+ "step": 29350,
+ "train_runtime": 227085.1855,
+ "train_tokens_per_second": 5653.062
+ },
+ {
+ "epoch": 1.976501481281982,
+ "grad_norm": 0.7233862194158842,
+ "learning_rate": 3.4252823606678588e-09,
+ "loss": 0.31540963649749754,
+ "num_input_tokens_seen": 1283952928,
+ "step": 29355,
+ "train_runtime": 227118.6155,
+ "train_tokens_per_second": 5653.226
+ },
+ {
+ "epoch": 1.9768381362779424,
+ "grad_norm": 0.7946045825578855,
+ "learning_rate": 3.328126968414336e-09,
+ "loss": 0.3130497455596924,
+ "num_input_tokens_seen": 1284147936,
+ "step": 29360,
+ "train_runtime": 227154.485,
+ "train_tokens_per_second": 5653.192
+ },
+ {
+ "epoch": 1.9771747912739026,
+ "grad_norm": 0.7860291343433327,
+ "learning_rate": 3.232368879517167e-09,
+ "loss": 0.31641981601715086,
+ "num_input_tokens_seen": 1284366488,
+ "step": 29365,
+ "train_runtime": 227200.0074,
+ "train_tokens_per_second": 5653.021
+ },
+ {
+ "epoch": 1.9775114462698626,
+ "grad_norm": 0.7901349246218392,
+ "learning_rate": 3.1380081207543766e-09,
+ "loss": 0.27493464946746826,
+ "num_input_tokens_seen": 1284580288,
+ "step": 29370,
+ "train_runtime": 227236.8347,
+ "train_tokens_per_second": 5653.046
+ },
+ {
+ "epoch": 1.9778481012658227,
+ "grad_norm": 0.7901988601043518,
+ "learning_rate": 3.0450447185137454e-09,
+ "loss": 0.300215482711792,
+ "num_input_tokens_seen": 1284790176,
+ "step": 29375,
+ "train_runtime": 227276.5524,
+ "train_tokens_per_second": 5652.982
+ },
+ {
+ "epoch": 1.978184756261783,
+ "grad_norm": 0.7905431031827841,
+ "learning_rate": 2.953478698792256e-09,
+ "loss": 0.29309883117675783,
+ "num_input_tokens_seen": 1285019536,
+ "step": 29380,
+ "train_runtime": 227311.7967,
+ "train_tokens_per_second": 5653.114
+ },
+ {
+ "epoch": 1.9785214112577432,
+ "grad_norm": 0.8196057690436338,
+ "learning_rate": 2.8633100871960917e-09,
+ "loss": 0.28349852561950684,
+ "num_input_tokens_seen": 1285225968,
+ "step": 29385,
+ "train_runtime": 227354.1237,
+ "train_tokens_per_second": 5652.97
+ },
+ {
+ "epoch": 1.9788580662537032,
+ "grad_norm": 0.8115546785952105,
+ "learning_rate": 2.774538908940638e-09,
+ "loss": 0.3020082950592041,
+ "num_input_tokens_seen": 1285433856,
+ "step": 29390,
+ "train_runtime": 227391.991,
+ "train_tokens_per_second": 5652.943
+ },
+ {
+ "epoch": 1.9791947212496632,
+ "grad_norm": 0.8333323595696321,
+ "learning_rate": 2.687165188849927e-09,
+ "loss": 0.2933856248855591,
+ "num_input_tokens_seen": 1285662136,
+ "step": 29395,
+ "train_runtime": 227423.7759,
+ "train_tokens_per_second": 5653.156
+ },
+ {
+ "epoch": 1.9795313762456235,
+ "grad_norm": 0.7761984521860247,
+ "learning_rate": 2.6011889513588574e-09,
+ "loss": 0.29178383350372317,
+ "num_input_tokens_seen": 1285879472,
+ "step": 29400,
+ "train_runtime": 227461.1494,
+ "train_tokens_per_second": 5653.183
+ },
+ {
+ "epoch": 1.9798680312415837,
+ "grad_norm": 0.7709375069582487,
+ "learning_rate": 2.5166102205093077e-09,
+ "loss": 0.31009225845336913,
+ "num_input_tokens_seen": 1286104184,
+ "step": 29405,
+ "train_runtime": 227495.7175,
+ "train_tokens_per_second": 5653.312
+ },
+ {
+ "epoch": 1.9802046862375438,
+ "grad_norm": 0.7529670231526651,
+ "learning_rate": 2.4334290199540254e-09,
+ "loss": 0.2966364860534668,
+ "num_input_tokens_seen": 1286326480,
+ "step": 29410,
+ "train_runtime": 227533.9984,
+ "train_tokens_per_second": 5653.337
+ },
+ {
+ "epoch": 1.9805413412335038,
+ "grad_norm": 0.7423799860118283,
+ "learning_rate": 2.351645372953848e-09,
+ "loss": 0.27869043350219724,
+ "num_input_tokens_seen": 1286539544,
+ "step": 29415,
+ "train_runtime": 227574.621,
+ "train_tokens_per_second": 5653.265
+ },
+ {
+ "epoch": 1.980877996229464,
+ "grad_norm": 0.830934035934882,
+ "learning_rate": 2.2712593023804795e-09,
+ "loss": 0.292054557800293,
+ "num_input_tokens_seen": 1286749296,
+ "step": 29420,
+ "train_runtime": 227603.0505,
+ "train_tokens_per_second": 5653.48
+ },
+ {
+ "epoch": 1.9812146512254243,
+ "grad_norm": 0.776340332148109,
+ "learning_rate": 2.1922708307120512e-09,
+ "loss": 0.3023597240447998,
+ "num_input_tokens_seen": 1286966040,
+ "step": 29425,
+ "train_runtime": 227646.316,
+ "train_tokens_per_second": 5653.358
+ },
+ {
+ "epoch": 1.9815513062213843,
+ "grad_norm": 0.7541577609267867,
+ "learning_rate": 2.1146799800386698e-09,
+ "loss": 0.3239266872406006,
+ "num_input_tokens_seen": 1287186672,
+ "step": 29430,
+ "train_runtime": 227678.08,
+ "train_tokens_per_second": 5653.538
+ },
+ {
+ "epoch": 1.9818879612173443,
+ "grad_norm": 0.8192940550401582,
+ "learning_rate": 2.0384867720579794e-09,
+ "loss": 0.278280234336853,
+ "num_input_tokens_seen": 1287391672,
+ "step": 29435,
+ "train_runtime": 227717.0952,
+ "train_tokens_per_second": 5653.47
+ },
+ {
+ "epoch": 1.9822246162133046,
+ "grad_norm": 0.7527540217103025,
+ "learning_rate": 1.96369122807738e-09,
+ "loss": 0.30411410331726074,
+ "num_input_tokens_seen": 1287620424,
+ "step": 29440,
+ "train_runtime": 227752.8072,
+ "train_tokens_per_second": 5653.588
+ },
+ {
+ "epoch": 1.9825612712092648,
+ "grad_norm": 0.8768248029758434,
+ "learning_rate": 1.8902933690123637e-09,
+ "loss": 0.2814857721328735,
+ "num_input_tokens_seen": 1287839296,
+ "step": 29445,
+ "train_runtime": 227783.5419,
+ "train_tokens_per_second": 5653.786
+ },
+ {
+ "epoch": 1.9828979262052249,
+ "grad_norm": 0.7931820825672209,
+ "learning_rate": 1.8182932153892884e-09,
+ "loss": 0.3011634826660156,
+ "num_input_tokens_seen": 1288057400,
+ "step": 29450,
+ "train_runtime": 227821.3297,
+ "train_tokens_per_second": 5653.805
+ },
+ {
+ "epoch": 1.983234581201185,
+ "grad_norm": 0.8278654551866558,
+ "learning_rate": 1.7476907873426041e-09,
+ "loss": 0.29094853401184084,
+ "num_input_tokens_seen": 1288283336,
+ "step": 29455,
+ "train_runtime": 227858.4256,
+ "train_tokens_per_second": 5653.876
+ },
+ {
+ "epoch": 1.9835712361971451,
+ "grad_norm": 0.7009932914918141,
+ "learning_rate": 1.6784861046159617e-09,
+ "loss": 0.29619736671447755,
+ "num_input_tokens_seen": 1288507000,
+ "step": 29460,
+ "train_runtime": 227904.983,
+ "train_tokens_per_second": 5653.703
+ },
+ {
+ "epoch": 1.9839078911931054,
+ "grad_norm": 0.696357225779694,
+ "learning_rate": 1.610679186561659e-09,
+ "loss": 0.3063995599746704,
+ "num_input_tokens_seen": 1288727264,
+ "step": 29465,
+ "train_runtime": 227944.9622,
+ "train_tokens_per_second": 5653.677
+ },
+ {
+ "epoch": 1.9842445461890654,
+ "grad_norm": 0.6754486614043711,
+ "learning_rate": 1.5442700521428599e-09,
+ "loss": 0.2689430952072144,
+ "num_input_tokens_seen": 1288937368,
+ "step": 29470,
+ "train_runtime": 227983.4782,
+ "train_tokens_per_second": 5653.644
+ },
+ {
+ "epoch": 1.9845812011850255,
+ "grad_norm": 0.7807038119601082,
+ "learning_rate": 1.4792587199297103e-09,
+ "loss": 0.29396581649780273,
+ "num_input_tokens_seen": 1289161136,
+ "step": 29475,
+ "train_runtime": 228025.749,
+ "train_tokens_per_second": 5653.577
+ },
+ {
+ "epoch": 1.9849178561809857,
+ "grad_norm": 0.8625253120569166,
+ "learning_rate": 1.4156452081026672e-09,
+ "loss": 0.310637903213501,
+ "num_input_tokens_seen": 1289388008,
+ "step": 29480,
+ "train_runtime": 228066.1418,
+ "train_tokens_per_second": 5653.57
+ },
+ {
+ "epoch": 1.985254511176946,
+ "grad_norm": 0.8386731246211869,
+ "learning_rate": 1.3534295344513893e-09,
+ "loss": 0.29576449394226073,
+ "num_input_tokens_seen": 1289614784,
+ "step": 29485,
+ "train_runtime": 228110.4974,
+ "train_tokens_per_second": 5653.465
+ },
+ {
+ "epoch": 1.985591166172906,
+ "grad_norm": 0.8448203849362906,
+ "learning_rate": 1.2926117163741813e-09,
+ "loss": 0.2955544710159302,
+ "num_input_tokens_seen": 1289834936,
+ "step": 29490,
+ "train_runtime": 228149.3127,
+ "train_tokens_per_second": 5653.468
+ },
+ {
+ "epoch": 1.985927821168866,
+ "grad_norm": 0.7551993663361088,
+ "learning_rate": 1.233191770877995e-09,
+ "loss": 0.3009219169616699,
+ "num_input_tokens_seen": 1290054672,
+ "step": 29495,
+ "train_runtime": 228188.1429,
+ "train_tokens_per_second": 5653.469
+ },
+ {
+ "epoch": 1.9862644761648263,
+ "grad_norm": 0.7385225987005458,
+ "learning_rate": 1.1751697145800934e-09,
+ "loss": 0.2817996978759766,
+ "num_input_tokens_seen": 1290281064,
+ "step": 29500,
+ "train_runtime": 228224.0221,
+ "train_tokens_per_second": 5653.573
+ },
+ {
+ "epoch": 1.9866011311607865,
+ "grad_norm": 0.8551674733846094,
+ "learning_rate": 1.1185455637052756e-09,
+ "loss": 0.3367082357406616,
+ "num_input_tokens_seen": 1290509624,
+ "step": 29505,
+ "train_runtime": 228265.6636,
+ "train_tokens_per_second": 5653.542
+ },
+ {
+ "epoch": 1.9869377861567465,
+ "grad_norm": 0.8607951663242355,
+ "learning_rate": 1.063319334089763e-09,
+ "loss": 0.3116638660430908,
+ "num_input_tokens_seen": 1290724632,
+ "step": 29510,
+ "train_runtime": 228308.5459,
+ "train_tokens_per_second": 5653.422
+ },
+ {
+ "epoch": 1.9872744411527066,
+ "grad_norm": 0.7586696985053731,
+ "learning_rate": 1.0094910411762027e-09,
+ "loss": 0.2789448261260986,
+ "num_input_tokens_seen": 1290947232,
+ "step": 29515,
+ "train_runtime": 228353.3345,
+ "train_tokens_per_second": 5653.288
+ },
+ {
+ "epoch": 1.9876110961486668,
+ "grad_norm": 0.8786362198012819,
+ "learning_rate": 9.570607000175536e-10,
+ "loss": 0.31690254211425783,
+ "num_input_tokens_seen": 1291165048,
+ "step": 29520,
+ "train_runtime": 228396.0938,
+ "train_tokens_per_second": 5653.184
+ },
+ {
+ "epoch": 1.987947751144627,
+ "grad_norm": 0.867401433919585,
+ "learning_rate": 9.060283252765311e-10,
+ "loss": 0.29549479484558105,
+ "num_input_tokens_seen": 1291376632,
+ "step": 29525,
+ "train_runtime": 228437.0857,
+ "train_tokens_per_second": 5653.095
+ },
+ {
+ "epoch": 1.988284406140587,
+ "grad_norm": 0.7283183527191147,
+ "learning_rate": 8.563939312233871e-10,
+ "loss": 0.2834463357925415,
+ "num_input_tokens_seen": 1291592160,
+ "step": 29530,
+ "train_runtime": 228475.8857,
+ "train_tokens_per_second": 5653.079
+ },
+ {
+ "epoch": 1.9886210611365471,
+ "grad_norm": 0.780933918989674,
+ "learning_rate": 8.081575317386847e-10,
+ "loss": 0.29933602809906007,
+ "num_input_tokens_seen": 1291812360,
+ "step": 29535,
+ "train_runtime": 228512.8712,
+ "train_tokens_per_second": 5653.127
+ },
+ {
+ "epoch": 1.9889577161325074,
+ "grad_norm": 0.7275658389704259,
+ "learning_rate": 7.613191403116338e-10,
+ "loss": 0.30489234924316405,
+ "num_input_tokens_seen": 1292039232,
+ "step": 29540,
+ "train_runtime": 228545.6355,
+ "train_tokens_per_second": 5653.31
+ },
+ {
+ "epoch": 1.9892943711284676,
+ "grad_norm": 0.7546810071907879,
+ "learning_rate": 7.158787700400905e-10,
+ "loss": 0.3050061225891113,
+ "num_input_tokens_seen": 1292268968,
+ "step": 29545,
+ "train_runtime": 228588.7605,
+ "train_tokens_per_second": 5653.248
+ },
+ {
+ "epoch": 1.9896310261244277,
+ "grad_norm": 0.7400073186881632,
+ "learning_rate": 6.718364336316674e-10,
+ "loss": 0.2916685104370117,
+ "num_input_tokens_seen": 1292491832,
+ "step": 29550,
+ "train_runtime": 228627.3323,
+ "train_tokens_per_second": 5653.269
+ },
+ {
+ "epoch": 1.9899676811203877,
+ "grad_norm": 0.8064188223143873,
+ "learning_rate": 6.291921434026238e-10,
+ "loss": 0.29037137031555177,
+ "num_input_tokens_seen": 1292713384,
+ "step": 29555,
+ "train_runtime": 228664.2829,
+ "train_tokens_per_second": 5653.324
+ },
+ {
+ "epoch": 1.990304336116348,
+ "grad_norm": 0.6299629051575274,
+ "learning_rate": 5.879459112778651e-10,
+ "loss": 0.26992483139038087,
+ "num_input_tokens_seen": 1292916816,
+ "step": 29560,
+ "train_runtime": 228707.5074,
+ "train_tokens_per_second": 5653.145
+ },
+ {
+ "epoch": 1.9906409911123082,
+ "grad_norm": 0.7057723847440374,
+ "learning_rate": 5.480977487920536e-10,
+ "loss": 0.3113460302352905,
+ "num_input_tokens_seen": 1293152760,
+ "step": 29565,
+ "train_runtime": 228744.4402,
+ "train_tokens_per_second": 5653.264
+ },
+ {
+ "epoch": 1.9909776461082682,
+ "grad_norm": 0.6985242431575329,
+ "learning_rate": 5.096476670890526e-10,
+ "loss": 0.29065916538238523,
+ "num_input_tokens_seen": 1293370744,
+ "step": 29570,
+ "train_runtime": 228786.0909,
+ "train_tokens_per_second": 5653.188
+ },
+ {
+ "epoch": 1.9913143011042282,
+ "grad_norm": 0.8726876861778484,
+ "learning_rate": 4.725956769202622e-10,
+ "loss": 0.3043833255767822,
+ "num_input_tokens_seen": 1293580568,
+ "step": 29575,
+ "train_runtime": 228821.7096,
+ "train_tokens_per_second": 5653.225
+ },
+ {
+ "epoch": 1.9916509561001885,
+ "grad_norm": 0.7789301695690733,
+ "learning_rate": 4.369417886485039e-10,
+ "loss": 0.3031582832336426,
+ "num_input_tokens_seen": 1293788160,
+ "step": 29580,
+ "train_runtime": 228859.1238,
+ "train_tokens_per_second": 5653.208
+ },
+ {
+ "epoch": 1.9919876110961487,
+ "grad_norm": 0.7445218059693559,
+ "learning_rate": 4.026860122430254e-10,
+ "loss": 0.293507981300354,
+ "num_input_tokens_seen": 1294011512,
+ "step": 29585,
+ "train_runtime": 228891.545,
+ "train_tokens_per_second": 5653.383
+ },
+ {
+ "epoch": 1.9923242660921088,
+ "grad_norm": 0.776822247685104,
+ "learning_rate": 3.698283572839412e-10,
+ "loss": 0.29284181594848635,
+ "num_input_tokens_seen": 1294233288,
+ "step": 29590,
+ "train_runtime": 228930.2438,
+ "train_tokens_per_second": 5653.396
+ },
+ {
+ "epoch": 1.9926609210880688,
+ "grad_norm": 0.8637476385715652,
+ "learning_rate": 3.38368832959457e-10,
+ "loss": 0.2951065540313721,
+ "num_input_tokens_seen": 1294439056,
+ "step": 29595,
+ "train_runtime": 228966.3788,
+ "train_tokens_per_second": 5653.402
+ },
+ {
+ "epoch": 1.992997576084029,
+ "grad_norm": 0.8185926706066555,
+ "learning_rate": 3.0830744806753523e-10,
+ "loss": 0.30010027885437013,
+ "num_input_tokens_seen": 1294656608,
+ "step": 29600,
+ "train_runtime": 229002.6257,
+ "train_tokens_per_second": 5653.457
+ },
+ {
+ "epoch": 1.9933342310799893,
+ "grad_norm": 0.7968229979488,
+ "learning_rate": 2.7964421101478457e-10,
+ "loss": 0.29233975410461427,
+ "num_input_tokens_seen": 1294866112,
+ "step": 29605,
+ "train_runtime": 229033.3425,
+ "train_tokens_per_second": 5653.614
+ },
+ {
+ "epoch": 1.9936708860759493,
+ "grad_norm": 0.6760975642489492,
+ "learning_rate": 2.5237912981646016e-10,
+ "loss": 0.3044607162475586,
+ "num_input_tokens_seen": 1295082752,
+ "step": 29610,
+ "train_runtime": 229068.0198,
+ "train_tokens_per_second": 5653.704
+ },
+ {
+ "epoch": 1.9940075410719094,
+ "grad_norm": 0.8833991167375352,
+ "learning_rate": 2.265122120975738e-10,
+ "loss": 0.2928770542144775,
+ "num_input_tokens_seen": 1295294464,
+ "step": 29615,
+ "train_runtime": 229108.5875,
+ "train_tokens_per_second": 5653.627
+ },
+ {
+ "epoch": 1.9943441960678696,
+ "grad_norm": 0.7026921457755877,
+ "learning_rate": 2.020434650906733e-10,
+ "loss": 0.2951038360595703,
+ "num_input_tokens_seen": 1295518440,
+ "step": 29620,
+ "train_runtime": 229144.7225,
+ "train_tokens_per_second": 5653.713
+ },
+ {
+ "epoch": 1.9946808510638299,
+ "grad_norm": 0.735480223574441,
+ "learning_rate": 1.7897289563972854e-10,
+ "loss": 0.28854942321777344,
+ "num_input_tokens_seen": 1295752200,
+ "step": 29625,
+ "train_runtime": 229181.9465,
+ "train_tokens_per_second": 5653.814
+ },
+ {
+ "epoch": 1.99501750605979,
+ "grad_norm": 0.7785852910887519,
+ "learning_rate": 1.5730051019569037e-10,
+ "loss": 0.28860130310058596,
+ "num_input_tokens_seen": 1295966672,
+ "step": 29630,
+ "train_runtime": 229217.582,
+ "train_tokens_per_second": 5653.871
+ },
+ {
+ "epoch": 1.99535416105575,
+ "grad_norm": 0.7483503409930179,
+ "learning_rate": 1.3702631481926633e-10,
+ "loss": 0.29309055805206297,
+ "num_input_tokens_seen": 1296186448,
+ "step": 29635,
+ "train_runtime": 229253.3432,
+ "train_tokens_per_second": 5653.948
+ },
+ {
+ "epoch": 1.9956908160517102,
+ "grad_norm": 0.8075255441882612,
+ "learning_rate": 1.1815031517981023e-10,
+ "loss": 0.28788046836853026,
+ "num_input_tokens_seen": 1296387216,
+ "step": 29640,
+ "train_runtime": 229284.8545,
+ "train_tokens_per_second": 5654.046
+ },
+ {
+ "epoch": 1.9960274710476704,
+ "grad_norm": 0.7216194041021678,
+ "learning_rate": 1.0067251655643262e-10,
+ "loss": 0.28290114402770994,
+ "num_input_tokens_seen": 1296614056,
+ "step": 29645,
+ "train_runtime": 229320.1711,
+ "train_tokens_per_second": 5654.165
+ },
+ {
+ "epoch": 1.9963641260436304,
+ "grad_norm": 0.8745531907030398,
+ "learning_rate": 8.459292383633522e-11,
+ "loss": 0.2950653076171875,
+ "num_input_tokens_seen": 1296815776,
+ "step": 29650,
+ "train_runtime": 229363.6026,
+ "train_tokens_per_second": 5653.974
+ },
+ {
+ "epoch": 1.9967007810395905,
+ "grad_norm": 0.8156583509555599,
+ "learning_rate": 6.991154151592128e-11,
+ "loss": 0.2833098888397217,
+ "num_input_tokens_seen": 1297021184,
+ "step": 29655,
+ "train_runtime": 229403.9752,
+ "train_tokens_per_second": 5653.874
+ },
+ {
+ "epoch": 1.9970374360355507,
+ "grad_norm": 0.734921335121816,
+ "learning_rate": 5.662837370135066e-11,
+ "loss": 0.3091259956359863,
+ "num_input_tokens_seen": 1297247832,
+ "step": 29660,
+ "train_runtime": 229441.4684,
+ "train_tokens_per_second": 5653.938
+ },
+ {
+ "epoch": 1.997374091031511,
+ "grad_norm": 0.8453937499810635,
+ "learning_rate": 4.47434241074296e-11,
+ "loss": 0.3012515068054199,
+ "num_input_tokens_seen": 1297458968,
+ "step": 29665,
+ "train_runtime": 229478.4365,
+ "train_tokens_per_second": 5653.947
+ },
+ {
+ "epoch": 1.997710746027471,
+ "grad_norm": 0.8579563343326696,
+ "learning_rate": 3.4256696057055615e-11,
+ "loss": 0.3033904552459717,
+ "num_input_tokens_seen": 1297673200,
+ "step": 29670,
+ "train_runtime": 229522.3122,
+ "train_tokens_per_second": 5653.8
+ },
+ {
+ "epoch": 1.998047401023431,
+ "grad_norm": 0.804937600894727,
+ "learning_rate": 2.5168192482882824e-11,
+ "loss": 0.30843775272369384,
+ "num_input_tokens_seen": 1297896616,
+ "step": 29675,
+ "train_runtime": 229556.9513,
+ "train_tokens_per_second": 5653.92
+ },
+ {
+ "epoch": 1.9983840560193913,
+ "grad_norm": 0.8014462539211099,
+ "learning_rate": 1.747791592676684e-11,
+ "loss": 0.27484722137451173,
+ "num_input_tokens_seen": 1298137256,
+ "step": 29680,
+ "train_runtime": 229589.775,
+ "train_tokens_per_second": 5654.16
+ },
+ {
+ "epoch": 1.9987207110153515,
+ "grad_norm": 0.7667553211393875,
+ "learning_rate": 1.1185868539209665e-11,
+ "loss": 0.3029062032699585,
+ "num_input_tokens_seen": 1298347688,
+ "step": 29685,
+ "train_runtime": 229623.1373,
+ "train_tokens_per_second": 5654.255
+ },
+ {
+ "epoch": 1.9990573660113116,
+ "grad_norm": 0.7997172260231348,
+ "learning_rate": 6.29205207991479e-12,
+ "loss": 0.30252552032470703,
+ "num_input_tokens_seen": 1298551880,
+ "step": 29690,
+ "train_runtime": 229659.802,
+ "train_tokens_per_second": 5654.241
+ },
+ {
+ "epoch": 1.9993940210072716,
+ "grad_norm": 0.8058607018219679,
+ "learning_rate": 2.7964679166769815e-12,
+ "loss": 0.2877742528915405,
+ "num_input_tokens_seen": 1298772440,
+ "step": 29695,
+ "train_runtime": 229699.7546,
+ "train_tokens_per_second": 5654.218
+ },
+ {
+ "epoch": 1.9997306760032318,
+ "grad_norm": 0.897980038082396,
+ "learning_rate": 6.991170281578364e-13,
+ "loss": 0.26116390228271485,
+ "num_input_tokens_seen": 1298981184,
+ "step": 29700,
+ "train_runtime": 229737.0922,
+ "train_tokens_per_second": 5654.207
+ }
+ ],
+ "logging_steps": 5,
+ "max_steps": 29704,
+ "num_input_tokens_seen": 1299153424,
+ "num_train_epochs": 2,
+ "save_steps": 500,
+ "stateful_callbacks": {
+ "TrainerControl": {
+ "args": {
+ "should_epoch_stop": false,
+ "should_evaluate": false,
+ "should_log": false,
+ "should_save": true,
+ "should_training_stop": true
+ },
+ "attributes": {}
+ }
+ },
+ "total_flos": 2402581590032384.0,
+ "train_batch_size": 1,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/checkpoint-29704/training_args.bin b/checkpoint-29704/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..9f35ce99940d0d7f4628bd149b2a2e946f1e8ec6
--- /dev/null
+++ b/checkpoint-29704/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c414b624070ab6682af9c82937e863fa17942c0203f8862756f767014d4fbaf7
+size 7761
diff --git a/checkpoint-29704/zero_to_fp32.py b/checkpoint-29704/zero_to_fp32.py
new file mode 100644
index 0000000000000000000000000000000000000000..5995d6e6f04e43b989587aa9022a3aef0c66d694
--- /dev/null
+++ b/checkpoint-29704/zero_to_fp32.py
@@ -0,0 +1,760 @@
+#!/usr/bin/env python
+
+# Copyright (c) Microsoft Corporation.
+# SPDX-License-Identifier: Apache-2.0
+
+# DeepSpeed Team
+
+# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets
+# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in
+# the future. Once extracted, the weights don't require DeepSpeed and can be used in any
+# application.
+#
+# example:
+# python zero_to_fp32.py . output_dir/
+# or
+# python zero_to_fp32.py . output_dir/ --safe_serialization
+
+import argparse
+import torch
+import glob
+import math
+import os
+import re
+import gc
+import json
+import numpy as np
+from tqdm import tqdm
+from collections import OrderedDict
+from dataclasses import dataclass
+
+# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with
+# DeepSpeed data structures it has to be available in the current python environment.
+from deepspeed.utils import logger
+from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS,
+ FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES,
+ FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS)
+
+
+@dataclass
+class zero_model_state:
+ buffers: dict()
+ param_shapes: dict()
+ shared_params: list
+ ds_version: int
+ frozen_param_shapes: dict()
+ frozen_param_fragments: dict()
+
+
+debug = 0
+
+# load to cpu
+device = torch.device('cpu')
+
+
+def atoi(text):
+ return int(text) if text.isdigit() else text
+
+
+def natural_keys(text):
+ '''
+ alist.sort(key=natural_keys) sorts in human order
+ http://nedbatchelder.com/blog/200712/human_sorting.html
+ (See Toothy's implementation in the comments)
+ '''
+ return [atoi(c) for c in re.split(r'(\d+)', text)]
+
+
+def get_model_state_file(checkpoint_dir, zero_stage):
+ if not os.path.isdir(checkpoint_dir):
+ raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist")
+
+ # there should be only one file
+ if zero_stage <= 2:
+ file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt")
+ elif zero_stage == 3:
+ file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt")
+
+ if not os.path.exists(file):
+ raise FileNotFoundError(f"can't find model states file at '{file}'")
+
+ return file
+
+
+def get_checkpoint_files(checkpoint_dir, glob_pattern):
+ # XXX: need to test that this simple glob rule works for multi-node setup too
+ ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys)
+
+ if len(ckpt_files) == 0:
+ raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'")
+
+ return ckpt_files
+
+
+def get_optim_files(checkpoint_dir):
+ return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt")
+
+
+def get_model_state_files(checkpoint_dir):
+ return get_checkpoint_files(checkpoint_dir, "*_model_states.pt")
+
+
+def parse_model_states(files):
+ zero_model_states = []
+ for file in files:
+ state_dict = torch.load(file, map_location=device, weights_only=False)
+
+ if BUFFER_NAMES not in state_dict:
+ raise ValueError(f"{file} is not a model state checkpoint")
+ buffer_names = state_dict[BUFFER_NAMES]
+ if debug:
+ print("Found buffers:", buffer_names)
+
+ # recover just the buffers while restoring them to fp32 if they were saved in fp16
+ buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names}
+ param_shapes = state_dict[PARAM_SHAPES]
+
+ # collect parameters that are included in param_shapes
+ param_names = []
+ for s in param_shapes:
+ for name in s.keys():
+ param_names.append(name)
+
+ # update with frozen parameters
+ frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None)
+ if frozen_param_shapes is not None:
+ if debug:
+ print(f"Found frozen_param_shapes: {frozen_param_shapes}")
+ param_names += list(frozen_param_shapes.keys())
+
+ # handle shared params
+ shared_params = [[k, v] for k, v in state_dict["shared_params"].items()]
+
+ ds_version = state_dict.get(DS_VERSION, None)
+
+ frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None)
+
+ z_model_state = zero_model_state(buffers=buffers,
+ param_shapes=param_shapes,
+ shared_params=shared_params,
+ ds_version=ds_version,
+ frozen_param_shapes=frozen_param_shapes,
+ frozen_param_fragments=frozen_param_fragments)
+ zero_model_states.append(z_model_state)
+
+ return zero_model_states
+
+
+def parse_optim_states(files, ds_checkpoint_dir):
+ total_files = len(files)
+ state_dicts = []
+ for f in tqdm(files, desc='Loading checkpoint shards'):
+ state_dict = torch.load(f, map_location=device, mmap=True, weights_only=False)
+ # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights
+ # and also handle the case where it was already removed by another helper script
+ state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None)
+ state_dicts.append(state_dict)
+
+ if ZERO_STAGE not in state_dicts[0][OPTIMIZER_STATE_DICT]:
+ raise ValueError(f"{files[0]} is not a zero checkpoint")
+ zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE]
+ world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT]
+
+ # For ZeRO-2 each param group can have different partition_count as data parallelism for expert
+ # parameters can be different from data parallelism for non-expert parameters. So we can just
+ # use the max of the partition_count to get the dp world_size.
+
+ if type(world_size) is list:
+ world_size = max(world_size)
+
+ if world_size != total_files:
+ raise ValueError(
+ f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. "
+ "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes."
+ )
+
+ # the groups are named differently in each stage
+ if zero_stage <= 2:
+ fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS
+ elif zero_stage == 3:
+ fp32_groups_key = FP32_FLAT_GROUPS
+ else:
+ raise ValueError(f"unknown zero stage {zero_stage}")
+
+ fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))]
+ return zero_stage, world_size, fp32_flat_groups
+
+
+def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters):
+ """
+ Returns fp32 state_dict reconstructed from ds checkpoint
+
+ Args:
+ - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are)
+
+ """
+ print(f"Processing zero checkpoint '{ds_checkpoint_dir}'")
+
+ optim_files = get_optim_files(ds_checkpoint_dir)
+ zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir)
+ print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}")
+
+ model_files = get_model_state_files(ds_checkpoint_dir)
+
+ zero_model_states = parse_model_states(model_files)
+ print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}')
+
+ if zero_stage <= 2:
+ return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+ exclude_frozen_parameters)
+ elif zero_stage == 3:
+ return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+ exclude_frozen_parameters)
+
+
+def _zero2_merge_frozen_params(state_dict, zero_model_states):
+ if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+ return
+
+ frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+ frozen_param_fragments = zero_model_states[0].frozen_param_fragments
+
+ if debug:
+ num_elem = sum(s.numel() for s in frozen_param_shapes.values())
+ print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+ wanted_params = len(frozen_param_shapes)
+ wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+ avail_numel = sum([p.numel() for p in frozen_param_fragments.values()])
+ print(f'Frozen params: Have {avail_numel} numels to process.')
+ print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+ total_params = 0
+ total_numel = 0
+ for name, shape in frozen_param_shapes.items():
+ total_params += 1
+ unpartitioned_numel = shape.numel()
+ total_numel += unpartitioned_numel
+
+ state_dict[name] = frozen_param_fragments[name]
+
+ if debug:
+ print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+
+ print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _has_callable(obj, fn):
+ attr = getattr(obj, fn, None)
+ return callable(attr)
+
+
+def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+ param_shapes = zero_model_states[0].param_shapes
+
+ # Reconstruction protocol:
+ #
+ # XXX: document this
+
+ if debug:
+ for i in range(world_size):
+ for j in range(len(fp32_flat_groups[0])):
+ print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}")
+
+ # XXX: memory usage doubles here (zero2)
+ num_param_groups = len(fp32_flat_groups[0])
+ merged_single_partition_of_fp32_groups = []
+ for i in range(num_param_groups):
+ merged_partitions = [sd[i] for sd in fp32_flat_groups]
+ full_single_fp32_vector = torch.cat(merged_partitions, 0)
+ merged_single_partition_of_fp32_groups.append(full_single_fp32_vector)
+ avail_numel = sum(
+ [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups])
+
+ if debug:
+ wanted_params = sum([len(shapes) for shapes in param_shapes])
+ wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes])
+ # not asserting if there is a mismatch due to possible padding
+ print(f"Have {avail_numel} numels to process.")
+ print(f"Need {wanted_numel} numels in {wanted_params} params.")
+
+ # params
+ # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+ # out-of-core computing solution
+ total_numel = 0
+ total_params = 0
+ for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups):
+ offset = 0
+ avail_numel = full_single_fp32_vector.numel()
+ for name, shape in shapes.items():
+
+ unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape)
+ total_numel += unpartitioned_numel
+ total_params += 1
+
+ if debug:
+ print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+ state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape)
+ offset += unpartitioned_numel
+
+ # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and
+ # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex
+ # paddings performed in the code it's almost impossible to predict the exact numbers w/o the
+ # live optimizer object, so we are checking that the numbers are within the right range
+ align_to = 2 * world_size
+
+ def zero2_align(x):
+ return align_to * math.ceil(x / align_to)
+
+ if debug:
+ print(f"original offset={offset}, avail_numel={avail_numel}")
+
+ offset = zero2_align(offset)
+ avail_numel = zero2_align(avail_numel)
+
+ if debug:
+ print(f"aligned offset={offset}, avail_numel={avail_numel}")
+
+ # Sanity check
+ if offset != avail_numel:
+ raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+ print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+ exclude_frozen_parameters):
+ state_dict = OrderedDict()
+
+ # buffers
+ buffers = zero_model_states[0].buffers
+ state_dict.update(buffers)
+ if debug:
+ print(f"added {len(buffers)} buffers")
+
+ if not exclude_frozen_parameters:
+ _zero2_merge_frozen_params(state_dict, zero_model_states)
+
+ _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+ # recover shared parameters
+ for pair in zero_model_states[0].shared_params:
+ if pair[1] in state_dict:
+ state_dict[pair[0]] = state_dict[pair[1]]
+
+ return state_dict
+
+
+def zero3_partitioned_param_info(unpartitioned_numel, world_size):
+ remainder = unpartitioned_numel % world_size
+ padding_numel = (world_size - remainder) if remainder else 0
+ partitioned_numel = math.ceil(unpartitioned_numel / world_size)
+ return partitioned_numel, padding_numel
+
+
+def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states):
+ if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+ return
+
+ if debug:
+ for i in range(world_size):
+ num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values())
+ print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+ frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+ wanted_params = len(frozen_param_shapes)
+ wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+ avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size
+ print(f'Frozen params: Have {avail_numel} numels to process.')
+ print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+ total_params = 0
+ total_numel = 0
+ for name, shape in zero_model_states[0].frozen_param_shapes.items():
+ total_params += 1
+ unpartitioned_numel = shape.numel()
+ total_numel += unpartitioned_numel
+
+ param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states)
+ state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape)
+
+ partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+ if debug:
+ print(
+ f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+ )
+
+ print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+class GatheredTensor:
+ """
+ A pseudo tensor that collects partitioned weights.
+ It is more memory efficient when there are multiple groups.
+ """
+
+ def __init__(self, flat_groups, flat_groups_offset, offset, partitioned_numel, shape):
+ self.flat_groups = flat_groups
+ self.flat_groups_offset = flat_groups_offset
+ self.offset = offset
+ self.partitioned_numel = partitioned_numel
+ self.shape = shape
+ self.dtype = self.flat_groups[0][0].dtype
+
+ def contiguous(self):
+ """
+ Merge partitioned weights from flat_groups into a single tensor.
+ """
+ end_idx = self.offset + self.partitioned_numel
+ world_size = len(self.flat_groups)
+ pad_flat_param_chunks = []
+
+ for rank_i in range(world_size):
+ # for each rank, we need to collect weights from related group/groups
+ flat_groups_at_rank_i = self.flat_groups[rank_i]
+ start_group_id = None
+ end_group_id = None
+ for group_id in range(len(self.flat_groups_offset)):
+ if self.flat_groups_offset[group_id] <= self.offset < self.flat_groups_offset[group_id + 1]:
+ start_group_id = group_id
+ if self.flat_groups_offset[group_id] < end_idx <= self.flat_groups_offset[group_id + 1]:
+ end_group_id = group_id
+ break
+ # collect weights from related group/groups
+ for group_id in range(start_group_id, end_group_id + 1):
+ flat_tensor = flat_groups_at_rank_i[group_id]
+ start_offset = self.offset - self.flat_groups_offset[group_id]
+ end_offset = min(end_idx, self.flat_groups_offset[group_id + 1]) - self.flat_groups_offset[group_id]
+ pad_flat_param_chunks.append(flat_tensor[start_offset:end_offset])
+
+ # collect weights from all ranks
+ pad_flat_param = torch.cat(pad_flat_param_chunks, dim=0)
+ param = pad_flat_param[:self.shape.numel()].view(self.shape).contiguous()
+ return param
+
+
+def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+ param_shapes = zero_model_states[0].param_shapes
+ avail_numel = sum([flat_group.numel() for flat_group in fp32_flat_groups[0]]) * world_size
+
+ # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each
+ # param, re-consolidating each param, while dealing with padding if any
+
+ # merge list of dicts, preserving order
+ param_shapes = {k: v for d in param_shapes for k, v in d.items()}
+
+ if debug:
+ for i in range(world_size):
+ print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}")
+
+ wanted_params = len(param_shapes)
+ wanted_numel = sum(shape.numel() for shape in param_shapes.values())
+ # not asserting if there is a mismatch due to possible padding
+ avail_numel = fp32_flat_groups[0].numel() * world_size
+ print(f"Trainable params: Have {avail_numel} numels to process.")
+ print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.")
+
+ # params
+ # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+ # out-of-core computing solution
+ offset = 0
+ total_numel = 0
+ total_params = 0
+ flat_groups_offset = [0] + list(np.cumsum([flat_tensor.numel() for flat_tensor in fp32_flat_groups[0]]))
+ for name, shape in tqdm(param_shapes.items(), desc='Gathering sharded weights'):
+ unpartitioned_numel = shape.numel()
+ total_numel += unpartitioned_numel
+ total_params += 1
+ partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+ if debug:
+ print(
+ f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+ )
+
+ # memory efficient tensor
+ tensor = GatheredTensor(fp32_flat_groups, flat_groups_offset, offset, partitioned_numel, shape)
+ state_dict[name] = tensor
+ offset += partitioned_numel
+
+ offset *= world_size
+
+ # Sanity check
+ if offset != avail_numel:
+ raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+ print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+ exclude_frozen_parameters):
+ state_dict = OrderedDict()
+
+ # buffers
+ buffers = zero_model_states[0].buffers
+ state_dict.update(buffers)
+ if debug:
+ print(f"added {len(buffers)} buffers")
+
+ if not exclude_frozen_parameters:
+ _zero3_merge_frozen_params(state_dict, world_size, zero_model_states)
+
+ _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+ # recover shared parameters
+ for pair in zero_model_states[0].shared_params:
+ if pair[1] in state_dict:
+ state_dict[pair[0]] = state_dict[pair[1]]
+
+ return state_dict
+
+
+def to_torch_tensor(state_dict, return_empty_tensor=False):
+ """
+ Convert state_dict of GatheredTensor to torch tensor
+ """
+ torch_state_dict = {}
+ converted_tensors = {}
+ for name, tensor in state_dict.items():
+ tensor_id = id(tensor)
+ if tensor_id in converted_tensors: # shared tensors
+ shared_tensor = torch_state_dict[converted_tensors[tensor_id]]
+ torch_state_dict[name] = shared_tensor
+ else:
+ converted_tensors[tensor_id] = name
+ if return_empty_tensor:
+ torch_state_dict[name] = torch.empty(tensor.shape, dtype=tensor.dtype)
+ else:
+ torch_state_dict[name] = tensor.contiguous()
+ return torch_state_dict
+
+
+def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir,
+ tag=None,
+ exclude_frozen_parameters=False,
+ lazy_mode=False):
+ """
+ Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with
+ ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example
+ via a model hub.
+
+ Args:
+ - ``checkpoint_dir``: path to the desired checkpoint folder
+ - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14``
+ - ``exclude_frozen_parameters``: exclude frozen parameters
+ - ``lazy_mode``: get state_dict in lazy mode. It returns a dict of pesduo tensor instead of torch tensor, which is more memory efficient.
+ Convert the pesduo tensor to torch tensor by ``.contiguous()``
+
+ Returns:
+ - pytorch ``state_dict``
+
+ A typical usage might be ::
+
+ from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+ # do the training and checkpoint saving
+ state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu
+ model = model.cpu() # move to cpu
+ model.load_state_dict(state_dict)
+ # submit to model hub or save the model to share with others
+
+ In this example the ``model`` will no longer be usable in the deepspeed context of the same
+ application. i.e. you will need to re-initialize the deepspeed engine, since
+ ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+ If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead.
+
+ Note: the above usage may not work if your application doesn't have sufficient free CPU memory.
+ You may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
+ the checkpoint. Or you can load state_dict in lazy mode ::
+
+ from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+ state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, lazy_mode=True) # not on cpu
+ for name, lazy_tensor in state_dict.item():
+ tensor = lazy_tensor.contiguous() # to cpu
+ print(name, tensor)
+ # del tensor to release memory if it no longer in use
+ """
+ if tag is None:
+ latest_path = os.path.join(checkpoint_dir, 'latest')
+ if os.path.isfile(latest_path):
+ with open(latest_path, 'r') as fd:
+ tag = fd.read().strip()
+ else:
+ raise ValueError(f"Unable to find 'latest' file at {latest_path}")
+
+ ds_checkpoint_dir = os.path.join(checkpoint_dir, tag)
+
+ if not os.path.isdir(ds_checkpoint_dir):
+ raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist")
+
+ state_dict = _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters)
+ if lazy_mode:
+ return state_dict
+ else:
+ return to_torch_tensor(state_dict)
+
+
+def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir,
+ output_dir,
+ max_shard_size="5GB",
+ safe_serialization=False,
+ tag=None,
+ exclude_frozen_parameters=False):
+ """
+ Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be
+ loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed.
+
+ Args:
+ - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+ - ``output_dir``: directory to the pytorch fp32 state_dict output files
+ - ``max_shard_size``: the maximum size for a checkpoint before being sharded, default value is 5GB
+ - ``safe_serialization``: whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).
+ - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+ - ``exclude_frozen_parameters``: exclude frozen parameters
+ """
+
+ # Dependency pre-check
+ if safe_serialization:
+ try:
+ from safetensors.torch import save_file
+ except ImportError:
+ print('If you want to use `safe_serialization`, please `pip install safetensors`')
+ raise
+ if max_shard_size is not None:
+ try:
+ from huggingface_hub import split_torch_state_dict_into_shards
+ except ImportError:
+ print('If you want to use `max_shard_size`, please `pip install huggingface_hub`')
+ raise
+
+ # Convert zero checkpoint to state_dict
+ state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir,
+ tag,
+ exclude_frozen_parameters,
+ lazy_mode=True)
+
+ # Shard the model if it is too big.
+ weights_name = "model.safetensors" if safe_serialization else "pytorch_model.bin"
+ if max_shard_size is not None:
+ filename_pattern = weights_name.replace(".bin", "{suffix}.bin").replace(".safetensors", "{suffix}.safetensors")
+ # an memory-efficient approach for sharding
+ empty_state_dict = to_torch_tensor(state_dict, return_empty_tensor=True)
+ state_dict_split = split_torch_state_dict_into_shards(empty_state_dict,
+ filename_pattern=filename_pattern,
+ max_shard_size=max_shard_size)
+ else:
+ from collections import namedtuple
+ StateDictSplit = namedtuple("StateDictSplit", ["is_sharded", "filename_to_tensors"])
+ state_dict_split = StateDictSplit(is_sharded=False,
+ filename_to_tensors={weights_name: list(state_dict.keys())})
+
+ # Save the model by shard
+ os.makedirs(output_dir, exist_ok=True)
+ filename_to_tensors = state_dict_split.filename_to_tensors.items()
+ for shard_file, tensors in tqdm(filename_to_tensors, desc="Saving checkpoint shards"):
+ shard_state_dict = {tensor_name: state_dict[tensor_name] for tensor_name in tensors}
+ shard_state_dict = to_torch_tensor(shard_state_dict)
+ output_path = os.path.join(output_dir, shard_file)
+ if safe_serialization:
+ save_file(shard_state_dict, output_path, metadata={"format": "pt"})
+ else:
+ torch.save(shard_state_dict, output_path)
+ # release the memory of current shard
+ for tensor_name in list(shard_state_dict.keys()):
+ del state_dict[tensor_name]
+ del shard_state_dict[tensor_name]
+ del shard_state_dict
+ gc.collect()
+
+ # Save index if sharded
+ if state_dict_split.is_sharded:
+ index = {
+ "metadata": state_dict_split.metadata,
+ "weight_map": state_dict_split.tensor_to_filename,
+ }
+ save_index_file = "model.safetensors.index.json" if safe_serialization else "pytorch_model.bin.index.json"
+ save_index_file = os.path.join(output_dir, save_index_file)
+ with open(save_index_file, "w", encoding="utf-8") as f:
+ content = json.dumps(index, indent=2, sort_keys=True) + "\n"
+ f.write(content)
+
+
+def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None):
+ """
+ 1. Put the provided model to cpu
+ 2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict``
+ 3. Load it into the provided model
+
+ Args:
+ - ``model``: the model object to update
+ - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+ - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+
+ Returns:
+ - ``model`: modified model
+
+ Make sure you have plenty of CPU memory available before you call this function. If you don't
+ have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it
+ conveniently placed for you in the checkpoint folder.
+
+ A typical usage might be ::
+
+ from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint
+ model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir)
+ # submit to model hub or save the model to share with others
+
+ Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context
+ of the same application. i.e. you will need to re-initialize the deepspeed engine, since
+ ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+ """
+ logger.info("Extracting fp32 weights")
+ state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
+
+ logger.info("Overwriting model with fp32 weights")
+ model = model.cpu()
+ model.load_state_dict(state_dict, strict=False)
+
+ return model
+
+
+if __name__ == "__main__":
+ parser = argparse.ArgumentParser()
+ parser.add_argument("checkpoint_dir",
+ type=str,
+ help="path to the desired checkpoint folder, e.g., path/checkpoint-12")
+ parser.add_argument("output_dir",
+ type=str,
+ help="directory to the pytorch fp32 state_dict output files"
+ "(e.g. path/checkpoint-12-output/)")
+ parser.add_argument(
+ "--max_shard_size",
+ type=str,
+ default="5GB",
+ help="The maximum size for a checkpoint before being sharded. Checkpoints shard will then be each of size"
+ "lower than this size. If expressed as a string, needs to be digits followed by a unit (like `5MB`"
+ "We default it to 5GB in order for models to be able to run easily on free-tier google colab instances"
+ "without CPU OOM issues.")
+ parser.add_argument(
+ "--safe_serialization",
+ default=False,
+ action='store_true',
+ help="Whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).")
+ parser.add_argument("-t",
+ "--tag",
+ type=str,
+ default=None,
+ help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1")
+ parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters")
+ parser.add_argument("-d", "--debug", action='store_true', help="enable debug")
+ args = parser.parse_args()
+
+ debug = args.debug
+
+ convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir,
+ args.output_dir,
+ max_shard_size=args.max_shard_size,
+ safe_serialization=args.safe_serialization,
+ tag=args.tag,
+ exclude_frozen_parameters=args.exclude_frozen_parameters)
diff --git a/config.json b/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..746f56efdea311c4c5e737f2e31814c30cc1e1bf
--- /dev/null
+++ b/config.json
@@ -0,0 +1,71 @@
+{
+ "architectures": [
+ "Qwen3ForCausalLM"
+ ],
+ "attention_bias": false,
+ "attention_dropout": 0.0,
+ "bos_token_id": null,
+ "dtype": "bfloat16",
+ "eos_token_id": 151645,
+ "head_dim": 128,
+ "hidden_act": "silu",
+ "hidden_size": 4096,
+ "initializer_range": 0.02,
+ "intermediate_size": 12288,
+ "layer_types": [
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention",
+ "full_attention"
+ ],
+ "max_position_embeddings": 40960,
+ "max_window_layers": 36,
+ "model_type": "qwen3",
+ "num_attention_heads": 32,
+ "num_hidden_layers": 36,
+ "num_key_value_heads": 8,
+ "pad_token_id": 151643,
+ "rms_norm_eps": 1e-06,
+ "rope_parameters": {
+ "rope_theta": 1000000,
+ "rope_type": "default"
+ },
+ "sliding_window": null,
+ "tie_word_embeddings": false,
+ "transformers_version": "5.6.0",
+ "use_cache": false,
+ "use_sliding_window": false,
+ "vocab_size": 151936
+}
diff --git a/generation_config.json b/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..cf5bf49b6941f1673361724230bb69080cd3f1e2
--- /dev/null
+++ b/generation_config.json
@@ -0,0 +1,12 @@
+{
+ "do_sample": true,
+ "eos_token_id": [
+ 151645,
+ 151643
+ ],
+ "pad_token_id": 151643,
+ "temperature": 0.6,
+ "top_k": 20,
+ "top_p": 0.95,
+ "transformers_version": "5.6.0"
+}
diff --git a/model.safetensors b/model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..e2403050ca19b0edeb2a4af3a391812ca8dad69c
--- /dev/null
+++ b/model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2d372672c4b3c93543df0eb6f735f1fb9c60753d3ca0b93a30cd40b192dfadcb
+size 16381517208
diff --git a/tokenizer.json b/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..c7afbed2efcdf019f88ab0572ec29d3bf595dfe2
--- /dev/null
+++ b/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be75606093db2094d7cd20f3c2f385c212750648bd6ea4fb2bf507a6a4c55506
+size 11422650
diff --git a/tokenizer_config.json b/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..6623da03457d79d2cd4af9557e9f31407cdb835c
--- /dev/null
+++ b/tokenizer_config.json
@@ -0,0 +1,31 @@
+{
+ "add_prefix_space": false,
+ "backend": "tokenizers",
+ "bos_token": null,
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "<|im_end|>",
+ "errors": "replace",
+ "extra_special_tokens": [
+ "<|im_start|>",
+ "<|im_end|>",
+ "<|object_ref_start|>",
+ "<|object_ref_end|>",
+ "<|box_start|>",
+ "<|box_end|>",
+ "<|quad_start|>",
+ "<|quad_end|>",
+ "<|vision_start|>",
+ "<|vision_end|>",
+ "<|vision_pad|>",
+ "<|image_pad|>",
+ "<|video_pad|>"
+ ],
+ "is_local": false,
+ "local_files_only": false,
+ "model_max_length": 131072,
+ "pad_token": "<|endoftext|>",
+ "padding_side": "right",
+ "split_special_tokens": false,
+ "tokenizer_class": "Qwen2Tokenizer",
+ "unk_token": null
+}
diff --git a/train_results.json b/train_results.json
new file mode 100644
index 0000000000000000000000000000000000000000..0fc254ffc9041351334a41c6a4b9ba8a72490a55
--- /dev/null
+++ b/train_results.json
@@ -0,0 +1,9 @@
+{
+ "epoch": 2.0,
+ "num_input_tokens_seen": 1299153424,
+ "total_flos": 2402581590032384.0,
+ "train_loss": 0.34685841939968143,
+ "train_runtime": 229818.4762,
+ "train_samples_per_second": 0.517,
+ "train_steps_per_second": 0.129
+}
\ No newline at end of file
diff --git a/trainer_log.jsonl b/trainer_log.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..941ef1686231625df60424af107ba4e2c3a4f83f
--- /dev/null
+++ b/trainer_log.jsonl
@@ -0,0 +1,5941 @@
+{"current_steps": 5, "total_steps": 29704, "loss": 0.8321187973022461, "lr": 9.999999552565099e-06, "epoch": 0.00033665499596014, "percentage": 0.02, "elapsed_time": "0:00:48", "remaining_time": "3 days, 8:05:04", "throughput": 4408.45, "total_tokens": 213976}
+{"current_steps": 10, "total_steps": 29704, "loss": 0.6714894294738769, "lr": 9.999997734860948e-06, "epoch": 0.00067330999192028, "percentage": 0.03, "elapsed_time": "0:01:19", "remaining_time": "2 days, 17:18:22", "throughput": 5281.87, "total_tokens": 418192}
+{"current_steps": 15, "total_steps": 29704, "loss": 0.6414656162261962, "lr": 9.999994518923374e-06, "epoch": 0.0010099649878804202, "percentage": 0.05, "elapsed_time": "0:01:56", "remaining_time": "2 days, 16:15:52", "throughput": 5404.84, "total_tokens": 631760}
+{"current_steps": 20, "total_steps": 29704, "loss": 0.6119038105010987, "lr": 9.999989904753278e-06, "epoch": 0.00134661998384056, "percentage": 0.07, "elapsed_time": "0:02:40", "remaining_time": "2 days, 18:10:27", "throughput": 5278.45, "total_tokens": 847240}
+{"current_steps": 25, "total_steps": 29704, "loss": 0.5918544292449951, "lr": 9.999983892351947e-06, "epoch": 0.0016832749798007002, "percentage": 0.08, "elapsed_time": "0:03:17", "remaining_time": "2 days, 17:07:31", "throughput": 5405.65, "total_tokens": 1067560}
+{"current_steps": 30, "total_steps": 29704, "loss": 0.5148830890655518, "lr": 9.999976481721066e-06, "epoch": 0.0020199299757608404, "percentage": 0.1, "elapsed_time": "0:03:51", "remaining_time": "2 days, 15:33:41", "throughput": 5570.97, "total_tokens": 1288760}
+{"current_steps": 35, "total_steps": 29704, "loss": 0.554470157623291, "lr": 9.999967672862704e-06, "epoch": 0.0023565849717209803, "percentage": 0.12, "elapsed_time": "0:04:31", "remaining_time": "2 days, 16:01:46", "throughput": 5523.2, "total_tokens": 1501896}
+{"current_steps": 40, "total_steps": 29704, "loss": 0.5239121437072753, "lr": 9.999957465779326e-06, "epoch": 0.00269323996768112, "percentage": 0.13, "elapsed_time": "0:05:19", "remaining_time": "2 days, 17:54:33", "throughput": 5365.84, "total_tokens": 1716792}
+{"current_steps": 45, "total_steps": 29704, "loss": 0.5400435924530029, "lr": 9.999945860473788e-06, "epoch": 0.0030298949636412605, "percentage": 0.15, "elapsed_time": "0:06:04", "remaining_time": "2 days, 18:40:12", "throughput": 5274.52, "total_tokens": 1920760}
+{"current_steps": 50, "total_steps": 29704, "loss": 0.5238595008850098, "lr": 9.999932856949333e-06, "epoch": 0.0033665499596014004, "percentage": 0.17, "elapsed_time": "0:06:42", "remaining_time": "2 days, 18:19:19", "throughput": 5316.62, "total_tokens": 2140344}
+{"current_steps": 55, "total_steps": 29704, "loss": 0.5059286117553711, "lr": 9.999918455209599e-06, "epoch": 0.0037032049555615404, "percentage": 0.19, "elapsed_time": "0:07:18", "remaining_time": "2 days, 17:35:36", "throughput": 5406.55, "total_tokens": 2368296}
+{"current_steps": 60, "total_steps": 29704, "loss": 0.5235630989074707, "lr": 9.99990265525861e-06, "epoch": 0.004039859951521681, "percentage": 0.2, "elapsed_time": "0:07:55", "remaining_time": "2 days, 17:12:55", "throughput": 5449.48, "total_tokens": 2589544}
+{"current_steps": 65, "total_steps": 29704, "loss": 0.5595571517944335, "lr": 9.999885457100789e-06, "epoch": 0.004376514947481821, "percentage": 0.22, "elapsed_time": "0:08:23", "remaining_time": "2 days, 15:49:24", "throughput": 5575.53, "total_tokens": 2809432}
+{"current_steps": 70, "total_steps": 29704, "loss": 0.5092310428619384, "lr": 9.999866860740943e-06, "epoch": 0.0047131699434419605, "percentage": 0.24, "elapsed_time": "0:09:03", "remaining_time": "2 days, 15:55:05", "throughput": 5556.58, "total_tokens": 3020248}
+{"current_steps": 75, "total_steps": 29704, "loss": 0.5114652633666992, "lr": 9.999846866184274e-06, "epoch": 0.005049824939402101, "percentage": 0.25, "elapsed_time": "0:09:45", "remaining_time": "2 days, 16:13:37", "throughput": 5552.19, "total_tokens": 3249600}
+{"current_steps": 80, "total_steps": 29704, "loss": 0.5040130615234375, "lr": 9.99982547343637e-06, "epoch": 0.00538647993536224, "percentage": 0.27, "elapsed_time": "0:10:26", "remaining_time": "2 days, 16:24:29", "throughput": 5499.91, "total_tokens": 3443864}
+{"current_steps": 85, "total_steps": 29704, "loss": 0.4742586612701416, "lr": 9.999802682503217e-06, "epoch": 0.005723134931322381, "percentage": 0.29, "elapsed_time": "0:11:03", "remaining_time": "2 days, 16:14:42", "throughput": 5515.84, "total_tokens": 3661032}
+{"current_steps": 90, "total_steps": 29704, "loss": 0.5044189453125, "lr": 9.999778493391186e-06, "epoch": 0.006059789927282521, "percentage": 0.3, "elapsed_time": "0:11:47", "remaining_time": "2 days, 16:40:27", "throughput": 5495.68, "total_tokens": 3888680}
+{"current_steps": 95, "total_steps": 29704, "loss": 0.5348274230957031, "lr": 9.999752906107043e-06, "epoch": 0.0063964449232426605, "percentage": 0.32, "elapsed_time": "0:12:24", "remaining_time": "2 days, 16:26:49", "throughput": 5526.94, "total_tokens": 4114248}
+{"current_steps": 100, "total_steps": 29704, "loss": 0.5229468822479248, "lr": 9.999725920657944e-06, "epoch": 0.006733099919202801, "percentage": 0.34, "elapsed_time": "0:13:10", "remaining_time": "2 days, 17:00:27", "throughput": 5476.42, "total_tokens": 4329256}
+{"current_steps": 105, "total_steps": 29704, "loss": 0.4990852355957031, "lr": 9.999697537051432e-06, "epoch": 0.007069754915162941, "percentage": 0.35, "elapsed_time": "0:13:48", "remaining_time": "2 days, 16:53:55", "throughput": 5470.34, "total_tokens": 4533832}
+{"current_steps": 110, "total_steps": 29704, "loss": 0.48107209205627444, "lr": 9.999667755295449e-06, "epoch": 0.007406409911123081, "percentage": 0.37, "elapsed_time": "0:14:24", "remaining_time": "2 days, 16:38:02", "throughput": 5491.08, "total_tokens": 4749088}
+{"current_steps": 115, "total_steps": 29704, "loss": 0.4919854164123535, "lr": 9.999636575398319e-06, "epoch": 0.007743064907083221, "percentage": 0.39, "elapsed_time": "0:15:04", "remaining_time": "2 days, 16:40:16", "throughput": 5482.35, "total_tokens": 4960752}
+{"current_steps": 120, "total_steps": 29704, "loss": 0.4375304222106934, "lr": 9.999603997368766e-06, "epoch": 0.008079719903043361, "percentage": 0.4, "elapsed_time": "0:15:39", "remaining_time": "2 days, 16:19:45", "throughput": 5501.43, "total_tokens": 5167864}
+{"current_steps": 125, "total_steps": 29704, "loss": 0.4794501304626465, "lr": 9.999570021215895e-06, "epoch": 0.008416374899003501, "percentage": 0.42, "elapsed_time": "0:16:13", "remaining_time": "2 days, 16:00:15", "throughput": 5529.92, "total_tokens": 5384648}
+{"current_steps": 130, "total_steps": 29704, "loss": 0.5004319667816162, "lr": 9.99953464694921e-06, "epoch": 0.008753029894963642, "percentage": 0.44, "elapsed_time": "0:16:53", "remaining_time": "2 days, 16:03:23", "throughput": 5514.97, "total_tokens": 5590392}
+{"current_steps": 135, "total_steps": 29704, "loss": 0.495189905166626, "lr": 9.999497874578606e-06, "epoch": 0.009089684890923782, "percentage": 0.45, "elapsed_time": "0:17:32", "remaining_time": "2 days, 16:02:49", "throughput": 5527.97, "total_tokens": 5819224}
+{"current_steps": 140, "total_steps": 29704, "loss": 0.5080117225646973, "lr": 9.999459704114361e-06, "epoch": 0.009426339886883921, "percentage": 0.47, "elapsed_time": "0:18:14", "remaining_time": "2 days, 16:12:21", "throughput": 5516.17, "total_tokens": 6037808}
+{"current_steps": 145, "total_steps": 29704, "loss": 0.526894474029541, "lr": 9.999420135567153e-06, "epoch": 0.00976299488284406, "percentage": 0.49, "elapsed_time": "0:18:55", "remaining_time": "2 days, 16:17:43", "throughput": 5506.95, "total_tokens": 6252752}
+{"current_steps": 150, "total_steps": 29704, "loss": 0.5105450630187989, "lr": 9.999379168948047e-06, "epoch": 0.010099649878804202, "percentage": 0.5, "elapsed_time": "0:19:32", "remaining_time": "2 days, 16:11:12", "throughput": 5522.45, "total_tokens": 6476720}
+{"current_steps": 155, "total_steps": 29704, "loss": 0.47139749526977537, "lr": 9.999336804268496e-06, "epoch": 0.010436304874764341, "percentage": 0.52, "elapsed_time": "0:20:17", "remaining_time": "2 days, 16:28:07", "throughput": 5501.83, "total_tokens": 6698024}
+{"current_steps": 160, "total_steps": 29704, "loss": 0.5010284900665283, "lr": 9.999293041540351e-06, "epoch": 0.01077295987072448, "percentage": 0.54, "elapsed_time": "0:20:58", "remaining_time": "2 days, 16:32:59", "throughput": 5491.35, "total_tokens": 6910784}
+{"current_steps": 165, "total_steps": 29704, "loss": 0.49982900619506837, "lr": 9.999247880775849e-06, "epoch": 0.011109614866684622, "percentage": 0.56, "elapsed_time": "0:21:35", "remaining_time": "2 days, 16:25:46", "throughput": 5492.03, "total_tokens": 7115568}
+{"current_steps": 170, "total_steps": 29704, "loss": 0.46059427261352537, "lr": 9.999201321987617e-06, "epoch": 0.011446269862644761, "percentage": 0.57, "elapsed_time": "0:22:17", "remaining_time": "2 days, 16:31:35", "throughput": 5481.6, "total_tokens": 7329488}
+{"current_steps": 175, "total_steps": 29704, "loss": 0.5039016723632812, "lr": 9.999153365188676e-06, "epoch": 0.011782924858604901, "percentage": 0.59, "elapsed_time": "0:23:03", "remaining_time": "2 days, 16:51:16", "throughput": 5454.96, "total_tokens": 7547880}
+{"current_steps": 180, "total_steps": 29704, "loss": 0.4546662330627441, "lr": 9.99910401039244e-06, "epoch": 0.012119579854565042, "percentage": 0.61, "elapsed_time": "0:23:44", "remaining_time": "2 days, 16:54:35", "throughput": 5445.32, "total_tokens": 7757728}
+{"current_steps": 185, "total_steps": 29704, "loss": 0.464589786529541, "lr": 9.999053257612708e-06, "epoch": 0.012456234850525182, "percentage": 0.62, "elapsed_time": "0:24:28", "remaining_time": "2 days, 17:04:02", "throughput": 5438.84, "total_tokens": 7984416}
+{"current_steps": 190, "total_steps": 29704, "loss": 0.5309792518615722, "lr": 9.999001106863672e-06, "epoch": 0.012792889846485321, "percentage": 0.64, "elapsed_time": "0:25:06", "remaining_time": "2 days, 16:59:32", "throughput": 5447.37, "total_tokens": 8204976}
+{"current_steps": 195, "total_steps": 29704, "loss": 0.4964266300201416, "lr": 9.998947558159916e-06, "epoch": 0.013129544842445462, "percentage": 0.66, "elapsed_time": "0:25:46", "remaining_time": "2 days, 16:59:55", "throughput": 5451.02, "total_tokens": 8428800}
+{"current_steps": 200, "total_steps": 29704, "loss": 0.44077634811401367, "lr": 9.99889261151642e-06, "epoch": 0.013466199838405602, "percentage": 0.67, "elapsed_time": "0:26:26", "remaining_time": "2 days, 16:59:39", "throughput": 5448.75, "total_tokens": 8642192}
+{"current_steps": 205, "total_steps": 29704, "loss": 0.47046546936035155, "lr": 9.998836266948542e-06, "epoch": 0.013802854834365741, "percentage": 0.69, "elapsed_time": "0:26:58", "remaining_time": "2 days, 16:42:49", "throughput": 5468.12, "total_tokens": 8852840}
+{"current_steps": 210, "total_steps": 29704, "loss": 0.4429631233215332, "lr": 9.998778524472043e-06, "epoch": 0.014139509830325882, "percentage": 0.71, "elapsed_time": "0:27:38", "remaining_time": "2 days, 16:41:43", "throughput": 5463.07, "total_tokens": 9059368}
+{"current_steps": 215, "total_steps": 29704, "loss": 0.46187715530395507, "lr": 9.99871938410307e-06, "epoch": 0.014476164826286022, "percentage": 0.72, "elapsed_time": "0:28:10", "remaining_time": "2 days, 16:25:12", "throughput": 5484.81, "total_tokens": 9273936}
+{"current_steps": 220, "total_steps": 29704, "loss": 0.5123334407806397, "lr": 9.99865884585816e-06, "epoch": 0.014812819822246161, "percentage": 0.74, "elapsed_time": "0:28:44", "remaining_time": "2 days, 16:12:27", "throughput": 5502.14, "total_tokens": 9489816}
+{"current_steps": 225, "total_steps": 29704, "loss": 0.4839674949645996, "lr": 9.998596909754245e-06, "epoch": 0.015149474818206303, "percentage": 0.76, "elapsed_time": "0:29:29", "remaining_time": "2 days, 16:22:58", "throughput": 5489.2, "total_tokens": 9710720}
+{"current_steps": 230, "total_steps": 29704, "loss": 0.455305290222168, "lr": 9.998533575808642e-06, "epoch": 0.015486129814166442, "percentage": 0.77, "elapsed_time": "0:30:04", "remaining_time": "2 days, 16:13:52", "throughput": 5498.82, "total_tokens": 9922160}
+{"current_steps": 235, "total_steps": 29704, "loss": 0.4588268280029297, "lr": 9.998468844039064e-06, "epoch": 0.015822784810126583, "percentage": 0.79, "elapsed_time": "0:30:43", "remaining_time": "2 days, 16:12:12", "throughput": 5497.34, "total_tokens": 10132480}
+{"current_steps": 240, "total_steps": 29704, "loss": 0.49381322860717775, "lr": 9.998402714463613e-06, "epoch": 0.016159439806086723, "percentage": 0.81, "elapsed_time": "0:31:26", "remaining_time": "2 days, 16:20:57", "throughput": 5488.16, "total_tokens": 10356032}
+{"current_steps": 245, "total_steps": 29704, "loss": 0.4929532051086426, "lr": 9.998335187100782e-06, "epoch": 0.016496094802046862, "percentage": 0.82, "elapsed_time": "0:32:11", "remaining_time": "2 days, 16:29:54", "throughput": 5474.51, "total_tokens": 10571688}
+{"current_steps": 250, "total_steps": 29704, "loss": 0.488722038269043, "lr": 9.998266261969453e-06, "epoch": 0.016832749798007002, "percentage": 0.84, "elapsed_time": "0:32:47", "remaining_time": "2 days, 16:23:07", "throughput": 5483.34, "total_tokens": 10787736}
+{"current_steps": 255, "total_steps": 29704, "loss": 0.46301860809326173, "lr": 9.998195939088905e-06, "epoch": 0.01716940479396714, "percentage": 0.86, "elapsed_time": "0:33:15", "remaining_time": "2 days, 16:00:32", "throughput": 5520.04, "total_tokens": 11014240}
+{"current_steps": 260, "total_steps": 29704, "loss": 0.4852250099182129, "lr": 9.998124218478797e-06, "epoch": 0.017506059789927284, "percentage": 0.88, "elapsed_time": "0:34:00", "remaining_time": "2 days, 16:11:56", "throughput": 5501.65, "total_tokens": 11227928}
+{"current_steps": 265, "total_steps": 29704, "loss": 0.4806126594543457, "lr": 9.998051100159192e-06, "epoch": 0.017842714785887424, "percentage": 0.89, "elapsed_time": "0:34:33", "remaining_time": "2 days, 15:58:46", "throughput": 5526.08, "total_tokens": 11457336}
+{"current_steps": 270, "total_steps": 29704, "loss": 0.4908778667449951, "lr": 9.997976584150533e-06, "epoch": 0.018179369781847563, "percentage": 0.91, "elapsed_time": "0:35:05", "remaining_time": "2 days, 15:45:43", "throughput": 5542.04, "total_tokens": 11669408}
+{"current_steps": 275, "total_steps": 29704, "loss": 0.4400303840637207, "lr": 9.997900670473662e-06, "epoch": 0.018516024777807703, "percentage": 0.93, "elapsed_time": "0:35:45", "remaining_time": "2 days, 15:47:18", "throughput": 5545.73, "total_tokens": 11900376}
+{"current_steps": 280, "total_steps": 29704, "loss": 0.47128896713256835, "lr": 9.997823359149804e-06, "epoch": 0.018852679773767842, "percentage": 0.94, "elapsed_time": "0:36:20", "remaining_time": "2 days, 15:38:14", "throughput": 5557.29, "total_tokens": 12115296}
+{"current_steps": 285, "total_steps": 29704, "loss": 0.5039300441741943, "lr": 9.997744650200579e-06, "epoch": 0.01918933476972798, "percentage": 0.96, "elapsed_time": "0:37:04", "remaining_time": "2 days, 15:47:30", "throughput": 5544.54, "total_tokens": 12335296}
+{"current_steps": 290, "total_steps": 29704, "loss": 0.47290778160095215, "lr": 9.997664543648e-06, "epoch": 0.01952598976568812, "percentage": 0.98, "elapsed_time": "0:37:49", "remaining_time": "2 days, 15:56:08", "throughput": 5530.37, "total_tokens": 12550016}
+{"current_steps": 295, "total_steps": 29704, "loss": 0.47549943923950194, "lr": 9.997583039514469e-06, "epoch": 0.019862644761648264, "percentage": 0.99, "elapsed_time": "0:38:30", "remaining_time": "2 days, 15:58:09", "throughput": 5528.07, "total_tokens": 12769944}
+{"current_steps": 300, "total_steps": 29704, "loss": 0.48865842819213867, "lr": 9.997500137822776e-06, "epoch": 0.020199299757608404, "percentage": 1.01, "elapsed_time": "0:39:07", "remaining_time": "2 days, 15:54:32", "throughput": 5535.64, "total_tokens": 12994152}
+{"current_steps": 305, "total_steps": 29704, "loss": 0.46552267074584963, "lr": 9.997415838596106e-06, "epoch": 0.020535954753568543, "percentage": 1.03, "elapsed_time": "0:39:50", "remaining_time": "2 days, 16:00:56", "throughput": 5523.56, "total_tokens": 13206104}
+{"current_steps": 310, "total_steps": 29704, "loss": 0.4945693492889404, "lr": 9.997330141858032e-06, "epoch": 0.020872609749528682, "percentage": 1.04, "elapsed_time": "0:40:29", "remaining_time": "2 days, 16:00:03", "throughput": 5524.27, "total_tokens": 13423544}
+{"current_steps": 315, "total_steps": 29704, "loss": 0.47294368743896487, "lr": 9.997243047632519e-06, "epoch": 0.021209264745488822, "percentage": 1.06, "elapsed_time": "0:41:03", "remaining_time": "2 days, 15:51:13", "throughput": 5536.28, "total_tokens": 13640584}
+{"current_steps": 320, "total_steps": 29704, "loss": 0.43665332794189454, "lr": 9.997154555943923e-06, "epoch": 0.02154591974144896, "percentage": 1.08, "elapsed_time": "0:41:46", "remaining_time": "2 days, 15:55:22", "throughput": 5530.06, "total_tokens": 13858888}
+{"current_steps": 325, "total_steps": 29704, "loss": 0.4625236034393311, "lr": 9.997064666816992e-06, "epoch": 0.021882574737409104, "percentage": 1.09, "elapsed_time": "0:42:24", "remaining_time": "2 days, 15:54:05", "throughput": 5531.45, "total_tokens": 14076648}
+{"current_steps": 330, "total_steps": 29704, "loss": 0.5440494060516358, "lr": 9.996973380276858e-06, "epoch": 0.022219229733369244, "percentage": 1.11, "elapsed_time": "0:42:55", "remaining_time": "2 days, 15:41:28", "throughput": 5551.36, "total_tokens": 14299864}
+{"current_steps": 335, "total_steps": 29704, "loss": 0.45033721923828124, "lr": 9.996880696349053e-06, "epoch": 0.022555884729329383, "percentage": 1.13, "elapsed_time": "0:43:34", "remaining_time": "2 days, 15:39:40", "throughput": 5555.33, "total_tokens": 14522560}
+{"current_steps": 340, "total_steps": 29704, "loss": 0.5329171180725097, "lr": 9.996786615059497e-06, "epoch": 0.022892539725289523, "percentage": 1.14, "elapsed_time": "0:44:15", "remaining_time": "2 days, 15:42:30", "throughput": 5553.04, "total_tokens": 14746648}
+{"current_steps": 345, "total_steps": 29704, "loss": 0.4342483997344971, "lr": 9.996691136434498e-06, "epoch": 0.023229194721249662, "percentage": 1.16, "elapsed_time": "0:44:50", "remaining_time": "2 days, 15:36:25", "throughput": 5562.99, "total_tokens": 14969080}
+{"current_steps": 350, "total_steps": 29704, "loss": 0.507387924194336, "lr": 9.996594260500753e-06, "epoch": 0.023565849717209802, "percentage": 1.18, "elapsed_time": "0:45:29", "remaining_time": "2 days, 15:35:17", "throughput": 5566.61, "total_tokens": 15193912}
+{"current_steps": 355, "total_steps": 29704, "loss": 0.4612261772155762, "lr": 9.996495987285358e-06, "epoch": 0.023902504713169945, "percentage": 1.2, "elapsed_time": "0:46:13", "remaining_time": "2 days, 15:40:55", "throughput": 5557.6, "total_tokens": 15411408}
+{"current_steps": 360, "total_steps": 29704, "loss": 0.4673755645751953, "lr": 9.996396316815793e-06, "epoch": 0.024239159709130084, "percentage": 1.21, "elapsed_time": "0:46:55", "remaining_time": "2 days, 15:45:02", "throughput": 5551.38, "total_tokens": 15630480}
+{"current_steps": 365, "total_steps": 29704, "loss": 0.4566653251647949, "lr": 9.99629524911993e-06, "epoch": 0.024575814705090224, "percentage": 1.23, "elapsed_time": "0:47:32", "remaining_time": "2 days, 15:42:04", "throughput": 5554.74, "total_tokens": 15847552}
+{"current_steps": 370, "total_steps": 29704, "loss": 0.440608549118042, "lr": 9.996192784226033e-06, "epoch": 0.024912469701050363, "percentage": 1.25, "elapsed_time": "0:48:07", "remaining_time": "2 days, 15:35:08", "throughput": 5558.63, "total_tokens": 16049488}
+{"current_steps": 375, "total_steps": 29704, "loss": 0.4601402759552002, "lr": 9.996088922162755e-06, "epoch": 0.025249124697010503, "percentage": 1.26, "elapsed_time": "0:48:40", "remaining_time": "2 days, 15:26:30", "throughput": 5576.18, "total_tokens": 16283544}
+{"current_steps": 380, "total_steps": 29704, "loss": 0.4678368091583252, "lr": 9.995983662959141e-06, "epoch": 0.025585779692970642, "percentage": 1.28, "elapsed_time": "0:49:19", "remaining_time": "2 days, 15:26:44", "throughput": 5577.09, "total_tokens": 16507176}
+{"current_steps": 385, "total_steps": 29704, "loss": 0.4893208026885986, "lr": 9.995877006644627e-06, "epoch": 0.025922434688930785, "percentage": 1.3, "elapsed_time": "0:50:03", "remaining_time": "2 days, 15:32:11", "throughput": 5573.27, "total_tokens": 16739664}
+{"current_steps": 390, "total_steps": 29704, "loss": 0.4782050132751465, "lr": 9.995768953249038e-06, "epoch": 0.026259089684890925, "percentage": 1.31, "elapsed_time": "0:50:42", "remaining_time": "2 days, 15:31:31", "throughput": 5576.25, "total_tokens": 16966112}
+{"current_steps": 395, "total_steps": 29704, "loss": 0.4591067314147949, "lr": 9.995659502802594e-06, "epoch": 0.026595744680851064, "percentage": 1.33, "elapsed_time": "0:51:21", "remaining_time": "2 days, 15:30:55", "throughput": 5576.36, "total_tokens": 17184144}
+{"current_steps": 400, "total_steps": 29704, "loss": 0.4847611427307129, "lr": 9.995548655335899e-06, "epoch": 0.026932399676811204, "percentage": 1.35, "elapsed_time": "0:51:59", "remaining_time": "2 days, 15:29:25", "throughput": 5574.99, "total_tokens": 17393504}
+{"current_steps": 405, "total_steps": 29704, "loss": 0.5011202812194824, "lr": 9.995436410879951e-06, "epoch": 0.027269054672771343, "percentage": 1.36, "elapsed_time": "0:52:38", "remaining_time": "2 days, 15:28:17", "throughput": 5578.04, "total_tokens": 17618352}
+{"current_steps": 410, "total_steps": 29704, "loss": 0.4390664577484131, "lr": 9.99532276946614e-06, "epoch": 0.027605709668731482, "percentage": 1.38, "elapsed_time": "0:53:17", "remaining_time": "2 days, 15:27:09", "throughput": 5578.0, "total_tokens": 17833504}
+{"current_steps": 415, "total_steps": 29704, "loss": 0.48116722106933596, "lr": 9.995207731126246e-06, "epoch": 0.027942364664691625, "percentage": 1.4, "elapsed_time": "0:53:53", "remaining_time": "2 days, 15:23:20", "throughput": 5585.45, "total_tokens": 18060040}
+{"current_steps": 420, "total_steps": 29704, "loss": 0.4614968776702881, "lr": 9.99509129589244e-06, "epoch": 0.028279019660651765, "percentage": 1.41, "elapsed_time": "0:54:36", "remaining_time": "2 days, 15:27:14", "throughput": 5577.53, "total_tokens": 18273536}
+{"current_steps": 425, "total_steps": 29704, "loss": 0.4571842193603516, "lr": 9.994973463797279e-06, "epoch": 0.028615674656611904, "percentage": 1.43, "elapsed_time": "0:55:12", "remaining_time": "2 days, 15:22:55", "throughput": 5585.98, "total_tokens": 18501288}
+{"current_steps": 430, "total_steps": 29704, "loss": 0.42873506546020507, "lr": 9.99485423487372e-06, "epoch": 0.028952329652572044, "percentage": 1.45, "elapsed_time": "0:55:54", "remaining_time": "2 days, 15:26:34", "throughput": 5578.02, "total_tokens": 18713400}
+{"current_steps": 435, "total_steps": 29704, "loss": 0.46686439514160155, "lr": 9.994733609155097e-06, "epoch": 0.029288984648532183, "percentage": 1.46, "elapsed_time": "0:56:30", "remaining_time": "2 days, 15:22:32", "throughput": 5581.93, "total_tokens": 18927424}
+{"current_steps": 440, "total_steps": 29704, "loss": 0.4384657382965088, "lr": 9.994611586675149e-06, "epoch": 0.029625639644492323, "percentage": 1.48, "elapsed_time": "0:57:07", "remaining_time": "2 days, 15:18:55", "throughput": 5588.88, "total_tokens": 19153848}
+{"current_steps": 445, "total_steps": 29704, "loss": 0.457914924621582, "lr": 9.994488167468e-06, "epoch": 0.029962294640452466, "percentage": 1.5, "elapsed_time": "0:57:43", "remaining_time": "2 days, 15:15:21", "throughput": 5592.09, "total_tokens": 19367752}
+{"current_steps": 450, "total_steps": 29704, "loss": 0.49051642417907715, "lr": 9.994363351568158e-06, "epoch": 0.030298949636412605, "percentage": 1.51, "elapsed_time": "0:58:27", "remaining_time": "2 days, 15:19:50", "throughput": 5583.41, "total_tokens": 19581376}
+{"current_steps": 455, "total_steps": 29704, "loss": 0.4277812480926514, "lr": 9.994237139010532e-06, "epoch": 0.030635604632372745, "percentage": 1.53, "elapsed_time": "0:59:06", "remaining_time": "2 days, 15:19:53", "throughput": 5579.2, "total_tokens": 19787728}
+{"current_steps": 460, "total_steps": 29704, "loss": 0.4721639633178711, "lr": 9.994109529830414e-06, "epoch": 0.030972259628332884, "percentage": 1.55, "elapsed_time": "0:59:45", "remaining_time": "2 days, 15:18:53", "throughput": 5585.16, "total_tokens": 20024672}
+{"current_steps": 465, "total_steps": 29704, "loss": 0.44678316116333006, "lr": 9.993980524063494e-06, "epoch": 0.03130891462429303, "percentage": 1.57, "elapsed_time": "1:00:22", "remaining_time": "2 days, 15:16:45", "throughput": 5588.86, "total_tokens": 20247816}
+{"current_steps": 470, "total_steps": 29704, "loss": 0.448548698425293, "lr": 9.993850121745842e-06, "epoch": 0.03164556962025317, "percentage": 1.58, "elapsed_time": "1:01:03", "remaining_time": "2 days, 15:18:14", "throughput": 5586.68, "total_tokens": 20469008}
+{"current_steps": 475, "total_steps": 29704, "loss": 0.44652185440063474, "lr": 9.99371832291393e-06, "epoch": 0.031982224616213306, "percentage": 1.6, "elapsed_time": "1:01:42", "remaining_time": "2 days, 15:17:03", "throughput": 5586.81, "total_tokens": 20684368}
+{"current_steps": 480, "total_steps": 29704, "loss": 0.4458156108856201, "lr": 9.993585127604612e-06, "epoch": 0.032318879612173446, "percentage": 1.62, "elapsed_time": "1:02:17", "remaining_time": "2 days, 15:12:54", "throughput": 5588.41, "total_tokens": 20888824}
+{"current_steps": 485, "total_steps": 29704, "loss": 0.4528946876525879, "lr": 9.993450535855136e-06, "epoch": 0.032655534608133585, "percentage": 1.63, "elapsed_time": "1:02:51", "remaining_time": "2 days, 15:06:28", "throughput": 5593.6, "total_tokens": 21093776}
+{"current_steps": 490, "total_steps": 29704, "loss": 0.43402805328369143, "lr": 9.993314547703142e-06, "epoch": 0.032992189604093725, "percentage": 1.65, "elapsed_time": "1:03:30", "remaining_time": "2 days, 15:06:41", "throughput": 5593.18, "total_tokens": 21314528}
+{"current_steps": 495, "total_steps": 29704, "loss": 0.44025378227233886, "lr": 9.993177163186655e-06, "epoch": 0.033328844600053864, "percentage": 1.67, "elapsed_time": "1:04:06", "remaining_time": "2 days, 15:02:43", "throughput": 5597.48, "total_tokens": 21529648}
+{"current_steps": 500, "total_steps": 29704, "loss": 0.43718814849853516, "lr": 9.993038382344097e-06, "epoch": 0.033665499596014004, "percentage": 1.68, "elapsed_time": "1:04:46", "remaining_time": "2 days, 15:03:46", "throughput": 5594.17, "total_tokens": 21744008}
+{"current_steps": 505, "total_steps": 29704, "loss": 0.44336519241333006, "lr": 9.992898205214279e-06, "epoch": 0.03400215459197414, "percentage": 1.7, "elapsed_time": "1:05:19", "remaining_time": "2 days, 14:57:31", "throughput": 5602.91, "total_tokens": 21963168}
+{"current_steps": 510, "total_steps": 29704, "loss": 0.4398432731628418, "lr": 9.992756631836397e-06, "epoch": 0.03433880958793428, "percentage": 1.72, "elapsed_time": "1:06:07", "remaining_time": "2 days, 15:05:01", "throughput": 5590.97, "total_tokens": 22181088}
+{"current_steps": 515, "total_steps": 29704, "loss": 0.4598443031311035, "lr": 9.992613662250044e-06, "epoch": 0.03467546458389442, "percentage": 1.73, "elapsed_time": "1:06:47", "remaining_time": "2 days, 15:05:33", "throughput": 5590.24, "total_tokens": 22402688}
+{"current_steps": 520, "total_steps": 29704, "loss": 0.4575651168823242, "lr": 9.992469296495201e-06, "epoch": 0.03501211957985457, "percentage": 1.75, "elapsed_time": "1:07:22", "remaining_time": "2 days, 15:01:31", "throughput": 5597.38, "total_tokens": 22628840}
+{"current_steps": 525, "total_steps": 29704, "loss": 0.414502477645874, "lr": 9.992323534612238e-06, "epoch": 0.03534877457581471, "percentage": 1.77, "elapsed_time": "1:08:02", "remaining_time": "2 days, 15:01:28", "throughput": 5597.02, "total_tokens": 22848504}
+{"current_steps": 530, "total_steps": 29704, "loss": 0.45860743522644043, "lr": 9.99217637664192e-06, "epoch": 0.03568542957177485, "percentage": 1.78, "elapsed_time": "1:08:39", "remaining_time": "2 days, 14:59:36", "throughput": 5600.31, "total_tokens": 23072192}
+{"current_steps": 535, "total_steps": 29704, "loss": 0.45654983520507814, "lr": 9.992027822625397e-06, "epoch": 0.03602208456773499, "percentage": 1.8, "elapsed_time": "1:09:21", "remaining_time": "2 days, 15:01:08", "throughput": 5595.17, "total_tokens": 23281952}
+{"current_steps": 540, "total_steps": 29704, "loss": 0.46324939727783204, "lr": 9.99187787260421e-06, "epoch": 0.036358739563695126, "percentage": 1.82, "elapsed_time": "1:09:59", "remaining_time": "2 days, 15:00:19", "throughput": 5593.75, "total_tokens": 23492592}
+{"current_steps": 545, "total_steps": 29704, "loss": 0.4310856819152832, "lr": 9.991726526620294e-06, "epoch": 0.036695394559655266, "percentage": 1.83, "elapsed_time": "1:10:35", "remaining_time": "2 days, 14:56:58", "throughput": 5601.21, "total_tokens": 23724784}
+{"current_steps": 550, "total_steps": 29704, "loss": 0.4721970081329346, "lr": 9.991573784715973e-06, "epoch": 0.037032049555615405, "percentage": 1.85, "elapsed_time": "1:11:18", "remaining_time": "2 days, 15:00:15", "throughput": 5595.25, "total_tokens": 23941816}
+{"current_steps": 555, "total_steps": 29704, "loss": 0.48006014823913573, "lr": 9.99141964693396e-06, "epoch": 0.037368704551575545, "percentage": 1.87, "elapsed_time": "1:11:55", "remaining_time": "2 days, 14:57:22", "throughput": 5599.89, "total_tokens": 24165232}
+{"current_steps": 560, "total_steps": 29704, "loss": 0.4417973518371582, "lr": 9.99126411331736e-06, "epoch": 0.037705359547535684, "percentage": 1.89, "elapsed_time": "1:12:39", "remaining_time": "2 days, 15:01:03", "throughput": 5595.07, "total_tokens": 24389896}
+{"current_steps": 565, "total_steps": 29704, "loss": 0.4680752754211426, "lr": 9.991107183909665e-06, "epoch": 0.038042014543495824, "percentage": 1.9, "elapsed_time": "1:13:11", "remaining_time": "2 days, 14:54:42", "throughput": 5605.21, "total_tokens": 24614992}
+{"current_steps": 570, "total_steps": 29704, "loss": 0.4640792369842529, "lr": 9.990948858754763e-06, "epoch": 0.03837866953945596, "percentage": 1.92, "elapsed_time": "1:13:54", "remaining_time": "2 days, 14:57:15", "throughput": 5598.26, "total_tokens": 24823080}
+{"current_steps": 575, "total_steps": 29704, "loss": 0.4609560966491699, "lr": 9.990789137896927e-06, "epoch": 0.0387153245354161, "percentage": 1.94, "elapsed_time": "1:14:31", "remaining_time": "2 days, 14:54:58", "throughput": 5601.58, "total_tokens": 25044784}
+{"current_steps": 580, "total_steps": 29704, "loss": 0.47663059234619143, "lr": 9.990628021380823e-06, "epoch": 0.03905197953137624, "percentage": 1.95, "elapsed_time": "1:15:09", "remaining_time": "2 days, 14:54:06", "throughput": 5601.81, "total_tokens": 25262200}
+{"current_steps": 585, "total_steps": 29704, "loss": 0.46056556701660156, "lr": 9.990465509251508e-06, "epoch": 0.03938863452733639, "percentage": 1.97, "elapsed_time": "1:15:41", "remaining_time": "2 days, 14:47:48", "throughput": 5608.71, "total_tokens": 25473136}
+{"current_steps": 590, "total_steps": 29704, "loss": 0.48001799583435056, "lr": 9.990301601554425e-06, "epoch": 0.03972528952329653, "percentage": 1.99, "elapsed_time": "1:16:21", "remaining_time": "2 days, 14:47:55", "throughput": 5603.26, "total_tokens": 25671088}
+{"current_steps": 595, "total_steps": 29704, "loss": 0.40839109420776365, "lr": 9.990136298335414e-06, "epoch": 0.04006194451925667, "percentage": 2.0, "elapsed_time": "1:16:58", "remaining_time": "2 days, 14:45:37", "throughput": 5605.45, "total_tokens": 25887472}
+{"current_steps": 600, "total_steps": 29704, "loss": 0.4223615646362305, "lr": 9.989969599640698e-06, "epoch": 0.04039859951521681, "percentage": 2.02, "elapsed_time": "1:17:30", "remaining_time": "2 days, 14:40:02", "throughput": 5615.49, "total_tokens": 26117456}
+{"current_steps": 605, "total_steps": 29704, "loss": 0.505678129196167, "lr": 9.989801505516895e-06, "epoch": 0.040735254511176947, "percentage": 2.04, "elapsed_time": "1:18:10", "remaining_time": "2 days, 14:39:39", "throughput": 5617.86, "total_tokens": 26348056}
+{"current_steps": 610, "total_steps": 29704, "loss": 0.47707581520080566, "lr": 9.989632016011012e-06, "epoch": 0.041071909507137086, "percentage": 2.05, "elapsed_time": "1:18:41", "remaining_time": "2 days, 14:33:20", "throughput": 5627.69, "total_tokens": 26572040}
+{"current_steps": 615, "total_steps": 29704, "loss": 0.4833499908447266, "lr": 9.989461131170448e-06, "epoch": 0.041408564503097225, "percentage": 2.07, "elapsed_time": "1:19:22", "remaining_time": "2 days, 14:34:09", "throughput": 5621.17, "total_tokens": 26769320}
+{"current_steps": 620, "total_steps": 29704, "loss": 0.4514059543609619, "lr": 9.98928885104299e-06, "epoch": 0.041745219499057365, "percentage": 2.09, "elapsed_time": "1:19:55", "remaining_time": "2 days, 14:29:08", "throughput": 5627.95, "total_tokens": 26988024}
+{"current_steps": 625, "total_steps": 29704, "loss": 0.4226998329162598, "lr": 9.989115175676812e-06, "epoch": 0.042081874495017504, "percentage": 2.1, "elapsed_time": "1:20:34", "remaining_time": "2 days, 14:28:49", "throughput": 5628.89, "total_tokens": 27212624}
+{"current_steps": 630, "total_steps": 29704, "loss": 0.47803831100463867, "lr": 9.988940105120483e-06, "epoch": 0.042418529490977644, "percentage": 2.12, "elapsed_time": "1:21:17", "remaining_time": "2 days, 14:31:37", "throughput": 5622.08, "total_tokens": 27422208}
+{"current_steps": 635, "total_steps": 29704, "loss": 0.4802694797515869, "lr": 9.988763639422965e-06, "epoch": 0.04275518448693778, "percentage": 2.14, "elapsed_time": "1:21:55", "remaining_time": "2 days, 14:30:42", "throughput": 5624.61, "total_tokens": 27650384}
+{"current_steps": 640, "total_steps": 29704, "loss": 0.4367190361022949, "lr": 9.988585778633604e-06, "epoch": 0.04309183948289792, "percentage": 2.15, "elapsed_time": "1:22:28", "remaining_time": "2 days, 14:25:40", "throughput": 5627.84, "total_tokens": 27851424}
+{"current_steps": 645, "total_steps": 29704, "loss": 0.45001683235168455, "lr": 9.988406522802134e-06, "epoch": 0.04342849447885807, "percentage": 2.17, "elapsed_time": "1:23:11", "remaining_time": "2 days, 14:27:48", "throughput": 5621.89, "total_tokens": 28060088}
+{"current_steps": 650, "total_steps": 29704, "loss": 0.41652770042419435, "lr": 9.98822587197869e-06, "epoch": 0.04376514947481821, "percentage": 2.19, "elapsed_time": "1:23:45", "remaining_time": "2 days, 14:24:02", "throughput": 5626.02, "total_tokens": 28274912}
+{"current_steps": 655, "total_steps": 29704, "loss": 0.4282545566558838, "lr": 9.988043826213785e-06, "epoch": 0.04410180447077835, "percentage": 2.21, "elapsed_time": "1:24:22", "remaining_time": "2 days, 14:21:44", "throughput": 5626.45, "total_tokens": 28481968}
+{"current_steps": 660, "total_steps": 29704, "loss": 0.4085527420043945, "lr": 9.98786038555833e-06, "epoch": 0.04443845946673849, "percentage": 2.22, "elapsed_time": "1:24:59", "remaining_time": "2 days, 14:20:23", "throughput": 5628.11, "total_tokens": 28702472}
+{"current_steps": 665, "total_steps": 29704, "loss": 0.44841432571411133, "lr": 9.987675550063625e-06, "epoch": 0.04477511446269863, "percentage": 2.24, "elapsed_time": "1:25:35", "remaining_time": "2 days, 14:17:24", "throughput": 5634.11, "total_tokens": 28932512}
+{"current_steps": 670, "total_steps": 29704, "loss": 0.4431906223297119, "lr": 9.987489319781355e-06, "epoch": 0.04511176945865877, "percentage": 2.26, "elapsed_time": "1:26:05", "remaining_time": "2 days, 14:11:01", "throughput": 5643.47, "total_tokens": 29153640}
+{"current_steps": 675, "total_steps": 29704, "loss": 0.4798222541809082, "lr": 9.987301694763602e-06, "epoch": 0.045448424454618906, "percentage": 2.27, "elapsed_time": "1:26:41", "remaining_time": "2 days, 14:08:08", "throughput": 5646.43, "total_tokens": 29368992}
+{"current_steps": 680, "total_steps": 29704, "loss": 0.4200302600860596, "lr": 9.987112675062834e-06, "epoch": 0.045785079450579046, "percentage": 2.29, "elapsed_time": "1:27:20", "remaining_time": "2 days, 14:08:10", "throughput": 5644.7, "total_tokens": 29582872}
+{"current_steps": 685, "total_steps": 29704, "loss": 0.42931652069091797, "lr": 9.986922260731908e-06, "epoch": 0.046121734446539185, "percentage": 2.31, "elapsed_time": "1:28:00", "remaining_time": "2 days, 14:08:17", "throughput": 5641.79, "total_tokens": 29791056}
+{"current_steps": 690, "total_steps": 29704, "loss": 0.4656373977661133, "lr": 9.986730451824073e-06, "epoch": 0.046458389442499325, "percentage": 2.32, "elapsed_time": "1:28:45", "remaining_time": "2 days, 14:12:16", "throughput": 5633.5, "total_tokens": 30001648}
+{"current_steps": 695, "total_steps": 29704, "loss": 0.4756736755371094, "lr": 9.986537248392969e-06, "epoch": 0.046795044438459464, "percentage": 2.34, "elapsed_time": "1:29:25", "remaining_time": "2 days, 14:12:49", "throughput": 5632.6, "total_tokens": 30223896}
+{"current_steps": 700, "total_steps": 29704, "loss": 0.4506352424621582, "lr": 9.986342650492626e-06, "epoch": 0.047131699434419604, "percentage": 2.36, "elapsed_time": "1:29:59", "remaining_time": "2 days, 14:08:42", "throughput": 5638.6, "total_tokens": 30445352}
+{"current_steps": 705, "total_steps": 29704, "loss": 0.42203187942504883, "lr": 9.98614665817746e-06, "epoch": 0.04746835443037975, "percentage": 2.37, "elapsed_time": "1:30:39", "remaining_time": "2 days, 14:08:58", "throughput": 5633.41, "total_tokens": 30642128}
+{"current_steps": 710, "total_steps": 29704, "loss": 0.427701473236084, "lr": 9.98594927150228e-06, "epoch": 0.04780500942633989, "percentage": 2.39, "elapsed_time": "1:31:18", "remaining_time": "2 days, 14:08:53", "throughput": 5635.09, "total_tokens": 30873224}
+{"current_steps": 715, "total_steps": 29704, "loss": 0.39564857482910154, "lr": 9.985750490522286e-06, "epoch": 0.04814166442230003, "percentage": 2.41, "elapsed_time": "1:32:01", "remaining_time": "2 days, 14:10:54", "throughput": 5630.5, "total_tokens": 31087512}
+{"current_steps": 720, "total_steps": 29704, "loss": 0.4636754035949707, "lr": 9.985550315293066e-06, "epoch": 0.04847831941826017, "percentage": 2.42, "elapsed_time": "1:32:37", "remaining_time": "2 days, 14:08:37", "throughput": 5635.5, "total_tokens": 31318912}
+{"current_steps": 725, "total_steps": 29704, "loss": 0.4597306728363037, "lr": 9.985348745870598e-06, "epoch": 0.04881497441422031, "percentage": 2.44, "elapsed_time": "1:33:12", "remaining_time": "2 days, 14:05:53", "throughput": 5640.39, "total_tokens": 31546096}
+{"current_steps": 730, "total_steps": 29704, "loss": 0.4194610595703125, "lr": 9.98514578231125e-06, "epoch": 0.04915162941018045, "percentage": 2.46, "elapsed_time": "1:33:50", "remaining_time": "2 days, 14:04:51", "throughput": 5642.99, "total_tokens": 31774968}
+{"current_steps": 735, "total_steps": 29704, "loss": 0.4705995559692383, "lr": 9.984941424671781e-06, "epoch": 0.04948828440614059, "percentage": 2.47, "elapsed_time": "1:34:32", "remaining_time": "2 days, 14:06:26", "throughput": 5639.4, "total_tokens": 31991384}
+{"current_steps": 740, "total_steps": 29704, "loss": 0.43364248275756834, "lr": 9.984735673009337e-06, "epoch": 0.049824939402100726, "percentage": 2.49, "elapsed_time": "1:35:17", "remaining_time": "2 days, 14:09:47", "throughput": 5633.77, "total_tokens": 32211232}
+{"current_steps": 745, "total_steps": 29704, "loss": 0.4496613502502441, "lr": 9.984528527381458e-06, "epoch": 0.050161594398060866, "percentage": 2.51, "elapsed_time": "1:35:58", "remaining_time": "2 days, 14:10:24", "throughput": 5632.49, "total_tokens": 32432504}
+{"current_steps": 750, "total_steps": 29704, "loss": 0.4476910591125488, "lr": 9.984319987846072e-06, "epoch": 0.050498249394021005, "percentage": 2.52, "elapsed_time": "1:36:35", "remaining_time": "2 days, 14:09:12", "throughput": 5632.29, "total_tokens": 32644184}
+{"current_steps": 755, "total_steps": 29704, "loss": 0.4494589328765869, "lr": 9.984110054461495e-06, "epoch": 0.050834904389981145, "percentage": 2.54, "elapsed_time": "1:37:11", "remaining_time": "2 days, 14:06:26", "throughput": 5639.15, "total_tokens": 32883080}
+{"current_steps": 760, "total_steps": 29704, "loss": 0.44231281280517576, "lr": 9.983898727286435e-06, "epoch": 0.051171559385941284, "percentage": 2.56, "elapsed_time": "1:37:44", "remaining_time": "2 days, 14:02:19", "throughput": 5644.71, "total_tokens": 33102688}
+{"current_steps": 765, "total_steps": 29704, "loss": 0.45505475997924805, "lr": 9.983686006379985e-06, "epoch": 0.05150821438190143, "percentage": 2.58, "elapsed_time": "1:38:21", "remaining_time": "2 days, 14:00:46", "throughput": 5645.92, "total_tokens": 33319336}
+{"current_steps": 770, "total_steps": 29704, "loss": 0.437741756439209, "lr": 9.983471891801638e-06, "epoch": 0.05184486937786157, "percentage": 2.59, "elapsed_time": "1:38:58", "remaining_time": "2 days, 13:59:19", "throughput": 5648.49, "total_tokens": 33545192}
+{"current_steps": 775, "total_steps": 29704, "loss": 0.441961669921875, "lr": 9.983256383611267e-06, "epoch": 0.05218152437382171, "percentage": 2.61, "elapsed_time": "1:39:34", "remaining_time": "2 days, 13:57:10", "throughput": 5651.48, "total_tokens": 33767136}
+{"current_steps": 780, "total_steps": 29704, "loss": 0.42367143630981446, "lr": 9.983039481869138e-06, "epoch": 0.05251817936978185, "percentage": 2.63, "elapsed_time": "1:40:09", "remaining_time": "2 days, 13:54:17", "throughput": 5655.0, "total_tokens": 33985712}
+{"current_steps": 785, "total_steps": 29704, "loss": 0.4347693920135498, "lr": 9.98282118663591e-06, "epoch": 0.05285483436574199, "percentage": 2.64, "elapsed_time": "1:40:46", "remaining_time": "2 days, 13:52:39", "throughput": 5657.24, "total_tokens": 34208000}
+{"current_steps": 790, "total_steps": 29704, "loss": 0.4524986267089844, "lr": 9.982601497972624e-06, "epoch": 0.05319148936170213, "percentage": 2.66, "elapsed_time": "1:41:25", "remaining_time": "2 days, 13:52:19", "throughput": 5656.64, "total_tokens": 34425040}
+{"current_steps": 795, "total_steps": 29704, "loss": 0.4259688377380371, "lr": 9.982380415940717e-06, "epoch": 0.05352814435766227, "percentage": 2.68, "elapsed_time": "1:42:00", "remaining_time": "2 days, 13:49:11", "throughput": 5659.09, "total_tokens": 34634712}
+{"current_steps": 800, "total_steps": 29704, "loss": 0.46526622772216797, "lr": 9.982157940602015e-06, "epoch": 0.05386479935362241, "percentage": 2.69, "elapsed_time": "1:42:36", "remaining_time": "2 days, 13:47:02", "throughput": 5663.48, "total_tokens": 34865376}
+{"current_steps": 805, "total_steps": 29704, "loss": 0.44102907180786133, "lr": 9.981934072018732e-06, "epoch": 0.05420145434958255, "percentage": 2.71, "elapsed_time": "1:43:09", "remaining_time": "2 days, 13:43:23", "throughput": 5670.54, "total_tokens": 35098448}
+{"current_steps": 810, "total_steps": 29704, "loss": 0.4654733657836914, "lr": 9.981708810253472e-06, "epoch": 0.054538109345542686, "percentage": 2.73, "elapsed_time": "1:43:43", "remaining_time": "2 days, 13:40:19", "throughput": 5674.38, "total_tokens": 35317184}
+{"current_steps": 815, "total_steps": 29704, "loss": 0.4581299781799316, "lr": 9.981482155369229e-06, "epoch": 0.054874764341502826, "percentage": 2.74, "elapsed_time": "1:44:24", "remaining_time": "2 days, 13:40:43", "throughput": 5674.01, "total_tokens": 35543024}
+{"current_steps": 820, "total_steps": 29704, "loss": 0.4073037147521973, "lr": 9.981254107429385e-06, "epoch": 0.055211419337462965, "percentage": 2.76, "elapsed_time": "1:45:01", "remaining_time": "2 days, 13:39:42", "throughput": 5673.31, "total_tokens": 35752976}
+{"current_steps": 825, "total_steps": 29704, "loss": 0.4087496757507324, "lr": 9.981024666497716e-06, "epoch": 0.05554807433342311, "percentage": 2.78, "elapsed_time": "1:45:36", "remaining_time": "2 days, 13:37:02", "throughput": 5676.39, "total_tokens": 35970808}
+{"current_steps": 830, "total_steps": 29704, "loss": 0.451265811920166, "lr": 9.98079383263838e-06, "epoch": 0.05588472932938325, "percentage": 2.79, "elapsed_time": "1:46:12", "remaining_time": "2 days, 13:34:34", "throughput": 5677.08, "total_tokens": 36175264}
+{"current_steps": 835, "total_steps": 29704, "loss": 0.43454971313476565, "lr": 9.980561605915932e-06, "epoch": 0.05622138432534339, "percentage": 2.81, "elapsed_time": "1:46:49", "remaining_time": "2 days, 13:33:20", "throughput": 5679.05, "total_tokens": 36399968}
+{"current_steps": 840, "total_steps": 29704, "loss": 0.44866161346435546, "lr": 9.980327986395316e-06, "epoch": 0.05655803932130353, "percentage": 2.83, "elapsed_time": "1:47:30", "remaining_time": "2 days, 13:34:23", "throughput": 5675.9, "total_tokens": 36614352}
+{"current_steps": 845, "total_steps": 29704, "loss": 0.43543548583984376, "lr": 9.980092974141855e-06, "epoch": 0.05689469431726367, "percentage": 2.84, "elapsed_time": "1:48:09", "remaining_time": "2 days, 13:34:06", "throughput": 5676.9, "total_tokens": 36842440}
+{"current_steps": 850, "total_steps": 29704, "loss": 0.434114933013916, "lr": 9.979856569221276e-06, "epoch": 0.05723134931322381, "percentage": 2.86, "elapsed_time": "1:48:55", "remaining_time": "2 days, 13:37:28", "throughput": 5671.45, "total_tokens": 37064928}
+{"current_steps": 855, "total_steps": 29704, "loss": 0.47480268478393556, "lr": 9.979618771699688e-06, "epoch": 0.05756800430918395, "percentage": 2.88, "elapsed_time": "1:49:31", "remaining_time": "2 days, 13:35:26", "throughput": 5674.16, "total_tokens": 37286696}
+{"current_steps": 860, "total_steps": 29704, "loss": 0.4271403312683105, "lr": 9.979379581643588e-06, "epoch": 0.05790465930514409, "percentage": 2.9, "elapsed_time": "1:50:12", "remaining_time": "2 days, 13:36:05", "throughput": 5672.21, "total_tokens": 37505104}
+{"current_steps": 865, "total_steps": 29704, "loss": 0.46260747909545896, "lr": 9.979138999119866e-06, "epoch": 0.05824131430110423, "percentage": 2.91, "elapsed_time": "1:50:57", "remaining_time": "2 days, 13:39:10", "throughput": 5666.21, "total_tokens": 37721104}
+{"current_steps": 870, "total_steps": 29704, "loss": 0.43666720390319824, "lr": 9.978897024195801e-06, "epoch": 0.05857796929706437, "percentage": 2.93, "elapsed_time": "1:51:32", "remaining_time": "2 days, 13:37:01", "throughput": 5671.62, "total_tokens": 37959848}
+{"current_steps": 875, "total_steps": 29704, "loss": 0.4272760391235352, "lr": 9.978653656939058e-06, "epoch": 0.058914624293024506, "percentage": 2.95, "elapsed_time": "1:52:16", "remaining_time": "2 days, 13:39:15", "throughput": 5667.44, "total_tokens": 38179560}
+{"current_steps": 880, "total_steps": 29704, "loss": 0.4613694190979004, "lr": 9.978408897417699e-06, "epoch": 0.059251279288984646, "percentage": 2.96, "elapsed_time": "1:52:56", "remaining_time": "2 days, 13:39:17", "throughput": 5667.49, "total_tokens": 38405040}
+{"current_steps": 885, "total_steps": 29704, "loss": 0.4417067527770996, "lr": 9.978162745700165e-06, "epoch": 0.059587934284944785, "percentage": 2.98, "elapsed_time": "1:53:34", "remaining_time": "2 days, 13:38:36", "throughput": 5668.72, "total_tokens": 38631232}
+{"current_steps": 890, "total_steps": 29704, "loss": 0.42413697242736814, "lr": 9.977915201855293e-06, "epoch": 0.05992458928090493, "percentage": 3.0, "elapsed_time": "1:54:10", "remaining_time": "2 days, 13:36:41", "throughput": 5670.71, "total_tokens": 38849712}
+{"current_steps": 895, "total_steps": 29704, "loss": 0.4329718589782715, "lr": 9.977666265952309e-06, "epoch": 0.06026124427686507, "percentage": 3.01, "elapsed_time": "1:54:50", "remaining_time": "2 days, 13:36:42", "throughput": 5669.6, "total_tokens": 39067416}
+{"current_steps": 900, "total_steps": 29704, "loss": 0.4434166431427002, "lr": 9.977415938060823e-06, "epoch": 0.06059789927282521, "percentage": 3.03, "elapsed_time": "1:55:26", "remaining_time": "2 days, 13:34:26", "throughput": 5672.52, "total_tokens": 39288624}
+{"current_steps": 905, "total_steps": 29704, "loss": 0.444110107421875, "lr": 9.977164218250846e-06, "epoch": 0.06093455426878535, "percentage": 3.05, "elapsed_time": "1:56:08", "remaining_time": "2 days, 13:36:03", "throughput": 5668.84, "total_tokens": 39505328}
+{"current_steps": 910, "total_steps": 29704, "loss": 0.4071072578430176, "lr": 9.976911106592765e-06, "epoch": 0.06127120926474549, "percentage": 3.06, "elapsed_time": "1:56:49", "remaining_time": "2 days, 13:36:21", "throughput": 5666.49, "total_tokens": 39717144}
+{"current_steps": 915, "total_steps": 29704, "loss": 0.4384052276611328, "lr": 9.976656603157363e-06, "epoch": 0.06160786426070563, "percentage": 3.08, "elapsed_time": "1:57:25", "remaining_time": "2 days, 13:34:43", "throughput": 5667.67, "total_tokens": 39933000}
+{"current_steps": 920, "total_steps": 29704, "loss": 0.42393884658813474, "lr": 9.976400708015811e-06, "epoch": 0.06194451925666577, "percentage": 3.1, "elapsed_time": "1:57:59", "remaining_time": "2 days, 13:31:33", "throughput": 5670.69, "total_tokens": 40145168}
+{"current_steps": 925, "total_steps": 29704, "loss": 0.4269254684448242, "lr": 9.97614342123967e-06, "epoch": 0.06228117425262591, "percentage": 3.11, "elapsed_time": "1:58:31", "remaining_time": "2 days, 13:27:24", "throughput": 5678.31, "total_tokens": 40379264}
+{"current_steps": 930, "total_steps": 29704, "loss": 0.4331661224365234, "lr": 9.97588474290089e-06, "epoch": 0.06261782924858605, "percentage": 3.13, "elapsed_time": "1:59:12", "remaining_time": "2 days, 13:28:24", "throughput": 5675.42, "total_tokens": 40594944}
+{"current_steps": 935, "total_steps": 29704, "loss": 0.4371979236602783, "lr": 9.975624673071806e-06, "epoch": 0.0629544842445462, "percentage": 3.15, "elapsed_time": "1:59:57", "remaining_time": "2 days, 13:31:06", "throughput": 5669.69, "total_tokens": 40808856}
+{"current_steps": 940, "total_steps": 29704, "loss": 0.4044677734375, "lr": 9.97536321182515e-06, "epoch": 0.06329113924050633, "percentage": 3.16, "elapsed_time": "2:00:27", "remaining_time": "2 days, 13:26:14", "throughput": 5674.98, "total_tokens": 41018224}
+{"current_steps": 945, "total_steps": 29704, "loss": 0.4564094066619873, "lr": 9.975100359234037e-06, "epoch": 0.06362779423646647, "percentage": 3.18, "elapsed_time": "2:01:05", "remaining_time": "2 days, 13:25:24", "throughput": 5675.68, "total_tokens": 41239392}
+{"current_steps": 950, "total_steps": 29704, "loss": 0.46511135101318357, "lr": 9.974836115371972e-06, "epoch": 0.06396444923242661, "percentage": 3.2, "elapsed_time": "2:01:44", "remaining_time": "2 days, 13:24:48", "throughput": 5675.16, "total_tokens": 41454352}
+{"current_steps": 955, "total_steps": 29704, "loss": 0.40421581268310547, "lr": 9.974570480312852e-06, "epoch": 0.06430110422838675, "percentage": 3.22, "elapsed_time": "2:02:26", "remaining_time": "2 days, 13:26:05", "throughput": 5673.31, "total_tokens": 41680576}
+{"current_steps": 960, "total_steps": 29704, "loss": 0.4464391231536865, "lr": 9.974303454130962e-06, "epoch": 0.06463775922434689, "percentage": 3.23, "elapsed_time": "2:03:00", "remaining_time": "2 days, 13:23:14", "throughput": 5674.47, "total_tokens": 41882360}
+{"current_steps": 965, "total_steps": 29704, "loss": 0.42637953758239744, "lr": 9.97403503690097e-06, "epoch": 0.06497441422030703, "percentage": 3.25, "elapsed_time": "2:03:35", "remaining_time": "2 days, 13:20:41", "throughput": 5677.77, "total_tokens": 42103128}
+{"current_steps": 970, "total_steps": 29704, "loss": 0.39764556884765623, "lr": 9.973765228697942e-06, "epoch": 0.06531106921626717, "percentage": 3.27, "elapsed_time": "2:04:13", "remaining_time": "2 days, 13:19:44", "throughput": 5679.17, "total_tokens": 42328192}
+{"current_steps": 975, "total_steps": 29704, "loss": 0.43375320434570314, "lr": 9.973494029597327e-06, "epoch": 0.06564772421222731, "percentage": 3.28, "elapsed_time": "2:04:47", "remaining_time": "2 days, 13:17:09", "throughput": 5680.22, "total_tokens": 42531664}
+{"current_steps": 980, "total_steps": 29704, "loss": 0.4309089660644531, "lr": 9.973221439674968e-06, "epoch": 0.06598437920818745, "percentage": 3.3, "elapsed_time": "2:05:27", "remaining_time": "2 days, 13:17:01", "throughput": 5678.25, "total_tokens": 42740856}
+{"current_steps": 985, "total_steps": 29704, "loss": 0.45548033714294434, "lr": 9.972947459007089e-06, "epoch": 0.06632103420414759, "percentage": 3.32, "elapsed_time": "2:06:02", "remaining_time": "2 days, 13:15:01", "throughput": 5680.01, "total_tokens": 42956288}
+{"current_steps": 990, "total_steps": 29704, "loss": 0.4385095119476318, "lr": 9.972672087670313e-06, "epoch": 0.06665768920010773, "percentage": 3.33, "elapsed_time": "2:06:39", "remaining_time": "2 days, 13:13:36", "throughput": 5681.6, "total_tokens": 43177272}
+{"current_steps": 995, "total_steps": 29704, "loss": 0.413985538482666, "lr": 9.972395325741644e-06, "epoch": 0.06699434419606787, "percentage": 3.35, "elapsed_time": "2:07:20", "remaining_time": "2 days, 13:14:01", "throughput": 5680.57, "total_tokens": 43400152}
+{"current_steps": 1000, "total_steps": 29704, "loss": 0.38621523380279543, "lr": 9.972117173298476e-06, "epoch": 0.06733099919202801, "percentage": 3.37, "elapsed_time": "2:07:54", "remaining_time": "2 days, 13:11:22", "throughput": 5683.27, "total_tokens": 43614968}
+{"current_steps": 1005, "total_steps": 29704, "loss": 0.42676420211791993, "lr": 9.971837630418597e-06, "epoch": 0.06766765418798815, "percentage": 3.38, "elapsed_time": "2:08:37", "remaining_time": "2 days, 13:12:51", "throughput": 5680.35, "total_tokens": 43835904}
+{"current_steps": 1010, "total_steps": 29704, "loss": 0.44001970291137693, "lr": 9.971556697180179e-06, "epoch": 0.06800430918394829, "percentage": 3.4, "elapsed_time": "2:09:16", "remaining_time": "2 days, 13:12:40", "throughput": 5678.93, "total_tokens": 44048512}
+{"current_steps": 1015, "total_steps": 29704, "loss": 0.4194046974182129, "lr": 9.971274373661784e-06, "epoch": 0.06834096417990843, "percentage": 3.42, "elapsed_time": "2:09:47", "remaining_time": "2 days, 13:08:31", "throughput": 5684.94, "total_tokens": 44271000}
+{"current_steps": 1020, "total_steps": 29704, "loss": 0.39097912311553956, "lr": 9.97099065994236e-06, "epoch": 0.06867761917586857, "percentage": 3.43, "elapsed_time": "2:10:22", "remaining_time": "2 days, 13:06:08", "throughput": 5688.99, "total_tokens": 44499576}
+{"current_steps": 1025, "total_steps": 29704, "loss": 0.43691391944885255, "lr": 9.970705556101253e-06, "epoch": 0.0690142741718287, "percentage": 3.45, "elapsed_time": "2:10:57", "remaining_time": "2 days, 13:04:17", "throughput": 5688.88, "total_tokens": 44702000}
+{"current_steps": 1030, "total_steps": 29704, "loss": 0.45786590576171876, "lr": 9.970419062218186e-06, "epoch": 0.06935092916778884, "percentage": 3.47, "elapsed_time": "2:11:39", "remaining_time": "2 days, 13:05:21", "throughput": 5688.34, "total_tokens": 44936832}
+{"current_steps": 1035, "total_steps": 29704, "loss": 0.4594112396240234, "lr": 9.970131178373276e-06, "epoch": 0.06968758416374898, "percentage": 3.48, "elapsed_time": "2:12:17", "remaining_time": "2 days, 13:04:28", "throughput": 5690.2, "total_tokens": 45166720}
+{"current_steps": 1040, "total_steps": 29704, "loss": 0.48036918640136717, "lr": 9.969841904647033e-06, "epoch": 0.07002423915970914, "percentage": 3.5, "elapsed_time": "2:12:55", "remaining_time": "2 days, 13:03:33", "throughput": 5690.77, "total_tokens": 45386112}
+{"current_steps": 1045, "total_steps": 29704, "loss": 0.4358417510986328, "lr": 9.969551241120349e-06, "epoch": 0.07036089415566928, "percentage": 3.52, "elapsed_time": "2:13:28", "remaining_time": "2 days, 13:00:28", "throughput": 5696.03, "total_tokens": 45615968}
+{"current_steps": 1050, "total_steps": 29704, "loss": 0.40404400825500486, "lr": 9.969259187874507e-06, "epoch": 0.07069754915162942, "percentage": 3.53, "elapsed_time": "2:14:04", "remaining_time": "2 days, 12:58:46", "throughput": 5699.25, "total_tokens": 45846704}
+{"current_steps": 1055, "total_steps": 29704, "loss": 0.43367910385131836, "lr": 9.968965744991178e-06, "epoch": 0.07103420414758956, "percentage": 3.55, "elapsed_time": "2:14:35", "remaining_time": "2 days, 12:54:50", "throughput": 5704.92, "total_tokens": 46069384}
+{"current_steps": 1060, "total_steps": 29704, "loss": 0.41605100631713865, "lr": 9.968670912552422e-06, "epoch": 0.0713708591435497, "percentage": 3.57, "elapsed_time": "2:15:10", "remaining_time": "2 days, 12:52:39", "throughput": 5707.52, "total_tokens": 46289280}
+{"current_steps": 1065, "total_steps": 29704, "loss": 0.4259674072265625, "lr": 9.968374690640692e-06, "epoch": 0.07170751413950983, "percentage": 3.59, "elapsed_time": "2:15:49", "remaining_time": "2 days, 12:52:18", "throughput": 5707.83, "total_tokens": 46513832}
+{"current_steps": 1070, "total_steps": 29704, "loss": 0.44624719619750974, "lr": 9.968077079338822e-06, "epoch": 0.07204416913546997, "percentage": 3.6, "elapsed_time": "2:16:23", "remaining_time": "2 days, 12:50:03", "throughput": 5710.4, "total_tokens": 46732528}
+{"current_steps": 1075, "total_steps": 29704, "loss": 0.4426886558532715, "lr": 9.967778078730038e-06, "epoch": 0.07238082413143011, "percentage": 3.62, "elapsed_time": "2:17:00", "remaining_time": "2 days, 12:48:38", "throughput": 5711.31, "total_tokens": 46948464}
+{"current_steps": 1080, "total_steps": 29704, "loss": 0.44336910247802735, "lr": 9.967477688897957e-06, "epoch": 0.07271747912739025, "percentage": 3.64, "elapsed_time": "2:17:39", "remaining_time": "2 days, 12:48:33", "throughput": 5710.96, "total_tokens": 47171032}
+{"current_steps": 1085, "total_steps": 29704, "loss": 0.46260600090026854, "lr": 9.967175909926578e-06, "epoch": 0.07305413412335039, "percentage": 3.65, "elapsed_time": "2:18:13", "remaining_time": "2 days, 12:46:02", "throughput": 5715.55, "total_tokens": 47402968}
+{"current_steps": 1090, "total_steps": 29704, "loss": 0.4760030746459961, "lr": 9.966872741900296e-06, "epoch": 0.07339078911931053, "percentage": 3.67, "elapsed_time": "2:19:02", "remaining_time": "2 days, 12:49:55", "throughput": 5709.97, "total_tokens": 47634040}
+{"current_steps": 1095, "total_steps": 29704, "loss": 0.4523357391357422, "lr": 9.96656818490389e-06, "epoch": 0.07372744411527067, "percentage": 3.69, "elapsed_time": "2:19:38", "remaining_time": "2 days, 12:48:22", "throughput": 5711.04, "total_tokens": 47849424}
+{"current_steps": 1100, "total_steps": 29704, "loss": 0.48172898292541505, "lr": 9.966262239022529e-06, "epoch": 0.07406409911123081, "percentage": 3.7, "elapsed_time": "2:20:14", "remaining_time": "2 days, 12:46:41", "throughput": 5713.45, "total_tokens": 48074544}
+{"current_steps": 1105, "total_steps": 29704, "loss": 0.4594748497009277, "lr": 9.965954904341768e-06, "epoch": 0.07440075410719095, "percentage": 3.72, "elapsed_time": "2:20:53", "remaining_time": "2 days, 12:46:22", "throughput": 5713.16, "total_tokens": 48294808}
+{"current_steps": 1110, "total_steps": 29704, "loss": 0.44074625968933107, "lr": 9.965646180947555e-06, "epoch": 0.07473740910315109, "percentage": 3.74, "elapsed_time": "2:21:36", "remaining_time": "2 days, 12:47:58", "throughput": 5709.19, "total_tokens": 48509312}
+{"current_steps": 1115, "total_steps": 29704, "loss": 0.42275848388671877, "lr": 9.96533606892622e-06, "epoch": 0.07507406409911123, "percentage": 3.75, "elapsed_time": "2:22:13", "remaining_time": "2 days, 12:46:42", "throughput": 5708.73, "total_tokens": 48715504}
+{"current_steps": 1120, "total_steps": 29704, "loss": 0.4448400497436523, "lr": 9.965024568364487e-06, "epoch": 0.07541071909507137, "percentage": 3.77, "elapsed_time": "2:22:51", "remaining_time": "2 days, 12:45:52", "throughput": 5710.25, "total_tokens": 48944392}
+{"current_steps": 1125, "total_steps": 29704, "loss": 0.41461782455444335, "lr": 9.964711679349467e-06, "epoch": 0.07574737409103151, "percentage": 3.79, "elapsed_time": "2:23:31", "remaining_time": "2 days, 12:45:52", "throughput": 5708.43, "total_tokens": 49155824}
+{"current_steps": 1130, "total_steps": 29704, "loss": 0.46135854721069336, "lr": 9.964397401968656e-06, "epoch": 0.07608402908699165, "percentage": 3.8, "elapsed_time": "2:24:10", "remaining_time": "2 days, 12:45:51", "throughput": 5708.23, "total_tokens": 49380912}
+{"current_steps": 1135, "total_steps": 29704, "loss": 0.41400885581970215, "lr": 9.964081736309942e-06, "epoch": 0.07642068408295179, "percentage": 3.82, "elapsed_time": "2:24:48", "remaining_time": "2 days, 12:44:56", "throughput": 5710.72, "total_tokens": 49617456}
+{"current_steps": 1140, "total_steps": 29704, "loss": 0.4346445083618164, "lr": 9.963764682461599e-06, "epoch": 0.07675733907891193, "percentage": 3.84, "elapsed_time": "2:25:22", "remaining_time": "2 days, 12:42:22", "throughput": 5715.19, "total_tokens": 49848584}
+{"current_steps": 1145, "total_steps": 29704, "loss": 0.4340538501739502, "lr": 9.96344624051229e-06, "epoch": 0.07709399407487207, "percentage": 3.85, "elapsed_time": "2:26:02", "remaining_time": "2 days, 12:42:25", "throughput": 5713.94, "total_tokens": 50065696}
+{"current_steps": 1150, "total_steps": 29704, "loss": 0.4169576644897461, "lr": 9.96312641055107e-06, "epoch": 0.0774306490708322, "percentage": 3.87, "elapsed_time": "2:26:41", "remaining_time": "2 days, 12:42:17", "throughput": 5713.11, "total_tokens": 50283968}
+{"current_steps": 1155, "total_steps": 29704, "loss": 0.41555070877075195, "lr": 9.962805192667375e-06, "epoch": 0.07776730406679234, "percentage": 3.89, "elapsed_time": "2:27:28", "remaining_time": "2 days, 12:45:09", "throughput": 5707.41, "total_tokens": 50500904}
+{"current_steps": 1160, "total_steps": 29704, "loss": 0.4202773094177246, "lr": 9.962482586951033e-06, "epoch": 0.07810395906275248, "percentage": 3.91, "elapsed_time": "2:28:09", "remaining_time": "2 days, 12:45:32", "throughput": 5707.38, "total_tokens": 50733408}
+{"current_steps": 1165, "total_steps": 29704, "loss": 0.442533016204834, "lr": 9.962158593492258e-06, "epoch": 0.07844061405871264, "percentage": 3.92, "elapsed_time": "2:28:50", "remaining_time": "2 days, 12:45:58", "throughput": 5704.28, "total_tokens": 50939416}
+{"current_steps": 1170, "total_steps": 29704, "loss": 0.4386699676513672, "lr": 9.961833212381657e-06, "epoch": 0.07877726905467278, "percentage": 3.94, "elapsed_time": "2:29:27", "remaining_time": "2 days, 12:45:06", "throughput": 5704.96, "total_tokens": 51160880}
+{"current_steps": 1175, "total_steps": 29704, "loss": 0.4202608108520508, "lr": 9.961506443710221e-06, "epoch": 0.07911392405063292, "percentage": 3.96, "elapsed_time": "2:30:09", "remaining_time": "2 days, 12:45:50", "throughput": 5703.13, "total_tokens": 51382192}
+{"current_steps": 1180, "total_steps": 29704, "loss": 0.4113266944885254, "lr": 9.961178287569326e-06, "epoch": 0.07945057904659306, "percentage": 3.97, "elapsed_time": "2:30:43", "remaining_time": "2 days, 12:43:27", "throughput": 5706.9, "total_tokens": 51610400}
+{"current_steps": 1185, "total_steps": 29704, "loss": 0.44871106147766116, "lr": 9.960848744050747e-06, "epoch": 0.0797872340425532, "percentage": 3.99, "elapsed_time": "2:31:25", "remaining_time": "2 days, 12:44:22", "throughput": 5704.42, "total_tokens": 51828672}
+{"current_steps": 1190, "total_steps": 29704, "loss": 0.44126176834106445, "lr": 9.960517813246633e-06, "epoch": 0.08012388903851334, "percentage": 4.01, "elapsed_time": "2:32:06", "remaining_time": "2 days, 12:44:40", "throughput": 5702.0, "total_tokens": 52038752}
+{"current_steps": 1195, "total_steps": 29704, "loss": 0.4168846607208252, "lr": 9.96018549524953e-06, "epoch": 0.08046054403447347, "percentage": 4.02, "elapsed_time": "2:32:45", "remaining_time": "2 days, 12:44:25", "throughput": 5701.39, "total_tokens": 52257272}
+{"current_steps": 1200, "total_steps": 29704, "loss": 0.41556696891784667, "lr": 9.959851790152371e-06, "epoch": 0.08079719903043361, "percentage": 4.04, "elapsed_time": "2:33:24", "remaining_time": "2 days, 12:44:05", "throughput": 5701.14, "total_tokens": 52478032}
+{"current_steps": 1205, "total_steps": 29704, "loss": 0.3835282325744629, "lr": 9.959516698048475e-06, "epoch": 0.08113385402639375, "percentage": 4.06, "elapsed_time": "2:34:05", "remaining_time": "2 days, 12:44:20", "throughput": 5698.43, "total_tokens": 52684592}
+{"current_steps": 1210, "total_steps": 29704, "loss": 0.4106024742126465, "lr": 9.959180219031551e-06, "epoch": 0.08147050902235389, "percentage": 4.07, "elapsed_time": "2:34:44", "remaining_time": "2 days, 12:44:01", "throughput": 5697.05, "total_tokens": 52895048}
+{"current_steps": 1215, "total_steps": 29704, "loss": 0.4241660118103027, "lr": 9.958842353195688e-06, "epoch": 0.08180716401831403, "percentage": 4.09, "elapsed_time": "2:35:23", "remaining_time": "2 days, 12:43:33", "throughput": 5697.73, "total_tokens": 53122496}
+{"current_steps": 1220, "total_steps": 29704, "loss": 0.43586392402648927, "lr": 9.958503100635377e-06, "epoch": 0.08214381901427417, "percentage": 4.11, "elapsed_time": "2:35:59", "remaining_time": "2 days, 12:41:59", "throughput": 5698.15, "total_tokens": 53331344}
+{"current_steps": 1225, "total_steps": 29704, "loss": 0.4607269287109375, "lr": 9.958162461445484e-06, "epoch": 0.08248047401023431, "percentage": 4.12, "elapsed_time": "2:36:38", "remaining_time": "2 days, 12:41:32", "throughput": 5698.94, "total_tokens": 53560160}
+{"current_steps": 1230, "total_steps": 29704, "loss": 0.4349203586578369, "lr": 9.957820435721271e-06, "epoch": 0.08281712900619445, "percentage": 4.14, "elapsed_time": "2:37:15", "remaining_time": "2 days, 12:40:25", "throughput": 5699.34, "total_tokens": 53775560}
+{"current_steps": 1235, "total_steps": 29704, "loss": 0.45670351982116697, "lr": 9.95747702355838e-06, "epoch": 0.08315378400215459, "percentage": 4.16, "elapsed_time": "2:37:50", "remaining_time": "2 days, 12:38:21", "throughput": 5702.28, "total_tokens": 54000808}
+{"current_steps": 1240, "total_steps": 29704, "loss": 0.4656820297241211, "lr": 9.957132225052848e-06, "epoch": 0.08349043899811473, "percentage": 4.17, "elapsed_time": "2:38:29", "remaining_time": "2 days, 12:38:07", "throughput": 5701.53, "total_tokens": 54218336}
+{"current_steps": 1245, "total_steps": 29704, "loss": 0.407243537902832, "lr": 9.956786040301098e-06, "epoch": 0.08382709399407487, "percentage": 4.19, "elapsed_time": "2:39:09", "remaining_time": "2 days, 12:37:57", "throughput": 5700.64, "total_tokens": 54435576}
+{"current_steps": 1250, "total_steps": 29704, "loss": 0.42780628204345705, "lr": 9.956438469399936e-06, "epoch": 0.08416374899003501, "percentage": 4.21, "elapsed_time": "2:39:44", "remaining_time": "2 days, 12:36:14", "throughput": 5701.52, "total_tokens": 54646448}
+{"current_steps": 1255, "total_steps": 29704, "loss": 0.4281005859375, "lr": 9.956089512446562e-06, "epoch": 0.08450040398599515, "percentage": 4.23, "elapsed_time": "2:40:24", "remaining_time": "2 days, 12:36:21", "throughput": 5698.7, "total_tokens": 54849384}
+{"current_steps": 1260, "total_steps": 29704, "loss": 0.43431816101074217, "lr": 9.955739169538559e-06, "epoch": 0.08483705898195529, "percentage": 4.24, "elapsed_time": "2:41:04", "remaining_time": "2 days, 12:36:08", "throughput": 5697.53, "total_tokens": 55062944}
+{"current_steps": 1265, "total_steps": 29704, "loss": 0.4328322887420654, "lr": 9.955387440773902e-06, "epoch": 0.08517371397791543, "percentage": 4.26, "elapsed_time": "2:41:44", "remaining_time": "2 days, 12:36:11", "throughput": 5694.98, "total_tokens": 55267032}
+{"current_steps": 1270, "total_steps": 29704, "loss": 0.42726478576660154, "lr": 9.955034326250947e-06, "epoch": 0.08551036897387557, "percentage": 4.28, "elapsed_time": "2:42:20", "remaining_time": "2 days, 12:34:35", "throughput": 5696.46, "total_tokens": 55485176}
+{"current_steps": 1275, "total_steps": 29704, "loss": 0.431905460357666, "lr": 9.954679826068444e-06, "epoch": 0.0858470239698357, "percentage": 4.29, "elapsed_time": "2:42:58", "remaining_time": "2 days, 12:33:48", "throughput": 5696.04, "total_tokens": 55697472}
+{"current_steps": 1280, "total_steps": 29704, "loss": 0.4367355823516846, "lr": 9.954323940325526e-06, "epoch": 0.08618367896579585, "percentage": 4.31, "elapsed_time": "2:43:35", "remaining_time": "2 days, 12:32:53", "throughput": 5695.81, "total_tokens": 55909208}
+{"current_steps": 1285, "total_steps": 29704, "loss": 0.4562966346740723, "lr": 9.953966669121715e-06, "epoch": 0.086520333961756, "percentage": 4.33, "elapsed_time": "2:44:15", "remaining_time": "2 days, 12:32:43", "throughput": 5696.77, "total_tokens": 56144336}
+{"current_steps": 1290, "total_steps": 29704, "loss": 0.4428553581237793, "lr": 9.953608012556924e-06, "epoch": 0.08685698895771614, "percentage": 4.34, "elapsed_time": "2:45:00", "remaining_time": "2 days, 12:34:32", "throughput": 5693.71, "total_tokens": 56370632}
+{"current_steps": 1295, "total_steps": 29704, "loss": 0.4518705368041992, "lr": 9.953247970731449e-06, "epoch": 0.08719364395367628, "percentage": 4.36, "elapsed_time": "2:45:41", "remaining_time": "2 days, 12:35:00", "throughput": 5691.23, "total_tokens": 56581824}
+{"current_steps": 1300, "total_steps": 29704, "loss": 0.45711684226989746, "lr": 9.952886543745973e-06, "epoch": 0.08753029894963642, "percentage": 4.38, "elapsed_time": "2:46:23", "remaining_time": "2 days, 12:35:39", "throughput": 5688.98, "total_tokens": 56797912}
+{"current_steps": 1305, "total_steps": 29704, "loss": 0.43997964859008787, "lr": 9.952523731701566e-06, "epoch": 0.08786695394559656, "percentage": 4.39, "elapsed_time": "2:47:03", "remaining_time": "2 days, 12:35:17", "throughput": 5689.74, "total_tokens": 57028400}
+{"current_steps": 1310, "total_steps": 29704, "loss": 0.39839365482330324, "lr": 9.952159534699693e-06, "epoch": 0.0882036089415567, "percentage": 4.41, "elapsed_time": "2:47:40", "remaining_time": "2 days, 12:34:09", "throughput": 5689.75, "total_tokens": 57239112}
+{"current_steps": 1315, "total_steps": 29704, "loss": 0.46162776947021483, "lr": 9.951793952842197e-06, "epoch": 0.08854026393751684, "percentage": 4.43, "elapsed_time": "2:48:10", "remaining_time": "2 days, 12:30:50", "throughput": 5693.39, "total_tokens": 57451840}
+{"current_steps": 1320, "total_steps": 29704, "loss": 0.4125502109527588, "lr": 9.95142698623131e-06, "epoch": 0.08887691893347698, "percentage": 4.44, "elapsed_time": "2:48:54", "remaining_time": "2 days, 12:32:00", "throughput": 5688.51, "total_tokens": 57649616}
+{"current_steps": 1325, "total_steps": 29704, "loss": 0.4594625473022461, "lr": 9.951058634969657e-06, "epoch": 0.08921357392943711, "percentage": 4.46, "elapsed_time": "2:49:33", "remaining_time": "2 days, 12:31:33", "throughput": 5687.36, "total_tokens": 57859504}
+{"current_steps": 1330, "total_steps": 29704, "loss": 0.42917823791503906, "lr": 9.950688899160244e-06, "epoch": 0.08955022892539725, "percentage": 4.48, "elapsed_time": "2:50:08", "remaining_time": "2 days, 12:29:47", "throughput": 5690.53, "total_tokens": 58092104}
+{"current_steps": 1335, "total_steps": 29704, "loss": 0.4345374584197998, "lr": 9.950317778906469e-06, "epoch": 0.0898868839213574, "percentage": 4.49, "elapsed_time": "2:50:49", "remaining_time": "2 days, 12:30:00", "throughput": 5690.05, "total_tokens": 58319256}
+{"current_steps": 1340, "total_steps": 29704, "loss": 0.40816512107849123, "lr": 9.949945274312109e-06, "epoch": 0.09022353891731753, "percentage": 4.51, "elapsed_time": "2:51:27", "remaining_time": "2 days, 12:29:20", "throughput": 5690.36, "total_tokens": 58540424}
+{"current_steps": 1345, "total_steps": 29704, "loss": 0.4445356369018555, "lr": 9.94957138548134e-06, "epoch": 0.09056019391327767, "percentage": 4.53, "elapsed_time": "2:52:06", "remaining_time": "2 days, 12:28:41", "throughput": 5688.23, "total_tokens": 58736800}
+{"current_steps": 1350, "total_steps": 29704, "loss": 0.41266183853149413, "lr": 9.949196112518715e-06, "epoch": 0.09089684890923781, "percentage": 4.54, "elapsed_time": "2:52:41", "remaining_time": "2 days, 12:26:58", "throughput": 5690.77, "total_tokens": 58963840}
+{"current_steps": 1355, "total_steps": 29704, "loss": 0.43543500900268556, "lr": 9.94881945552918e-06, "epoch": 0.09123350390519795, "percentage": 4.56, "elapsed_time": "2:53:26", "remaining_time": "2 days, 12:28:46", "throughput": 5687.31, "total_tokens": 59186072}
+{"current_steps": 1360, "total_steps": 29704, "loss": 0.45530238151550295, "lr": 9.948441414618064e-06, "epoch": 0.09157015890115809, "percentage": 4.58, "elapsed_time": "2:54:08", "remaining_time": "2 days, 12:29:20", "throughput": 5684.62, "total_tokens": 59396096}
+{"current_steps": 1365, "total_steps": 29704, "loss": 0.41725835800170896, "lr": 9.948061989891085e-06, "epoch": 0.09190681389711823, "percentage": 4.6, "elapsed_time": "2:54:44", "remaining_time": "2 days, 12:27:49", "throughput": 5686.72, "total_tokens": 59621960}
+{"current_steps": 1370, "total_steps": 29704, "loss": 0.4238771438598633, "lr": 9.94768118145435e-06, "epoch": 0.09224346889307837, "percentage": 4.61, "elapsed_time": "2:55:21", "remaining_time": "2 days, 12:26:49", "throughput": 5686.74, "total_tokens": 59834688}
+{"current_steps": 1375, "total_steps": 29704, "loss": 0.4197520732879639, "lr": 9.947298989414349e-06, "epoch": 0.09258012388903851, "percentage": 4.63, "elapsed_time": "2:56:02", "remaining_time": "2 days, 12:27:02", "throughput": 5686.2, "total_tokens": 60061512}
+{"current_steps": 1380, "total_steps": 29704, "loss": 0.4294243812561035, "lr": 9.946915413877963e-06, "epoch": 0.09291677888499865, "percentage": 4.65, "elapsed_time": "2:56:48", "remaining_time": "2 days, 12:29:00", "throughput": 5682.97, "total_tokens": 60289152}
+{"current_steps": 1385, "total_steps": 29704, "loss": 0.44629898071289065, "lr": 9.946530454952454e-06, "epoch": 0.09325343388095879, "percentage": 4.66, "elapsed_time": "2:57:29", "remaining_time": "2 days, 12:29:07", "throughput": 5682.62, "total_tokens": 60516576}
+{"current_steps": 1390, "total_steps": 29704, "loss": 0.4549723625183105, "lr": 9.946144112745477e-06, "epoch": 0.09359008887691893, "percentage": 4.68, "elapsed_time": "2:58:10", "remaining_time": "2 days, 12:29:26", "throughput": 5681.98, "total_tokens": 60744096}
+{"current_steps": 1395, "total_steps": 29704, "loss": 0.4265571594238281, "lr": 9.94575638736507e-06, "epoch": 0.09392674387287907, "percentage": 4.7, "elapsed_time": "2:58:50", "remaining_time": "2 days, 12:29:19", "throughput": 5681.89, "total_tokens": 60970464}
+{"current_steps": 1400, "total_steps": 29704, "loss": 0.469175910949707, "lr": 9.945367278919662e-06, "epoch": 0.09426339886883921, "percentage": 4.71, "elapsed_time": "2:59:31", "remaining_time": "2 days, 12:29:32", "throughput": 5681.15, "total_tokens": 61195536}
+{"current_steps": 1405, "total_steps": 29704, "loss": 0.43855986595153806, "lr": 9.944976787518064e-06, "epoch": 0.09460005386479936, "percentage": 4.73, "elapsed_time": "3:00:04", "remaining_time": "2 days, 12:27:09", "throughput": 5684.0, "total_tokens": 61415400}
+{"current_steps": 1410, "total_steps": 29704, "loss": 0.43834495544433594, "lr": 9.944584913269474e-06, "epoch": 0.0949367088607595, "percentage": 4.75, "elapsed_time": "3:00:46", "remaining_time": "2 days, 12:27:26", "throughput": 5683.61, "total_tokens": 61645440}
+{"current_steps": 1415, "total_steps": 29704, "loss": 0.41381168365478516, "lr": 9.944191656283481e-06, "epoch": 0.09527336385671964, "percentage": 4.76, "elapsed_time": "3:01:25", "remaining_time": "2 days, 12:27:07", "throughput": 5683.55, "total_tokens": 61868952}
+{"current_steps": 1420, "total_steps": 29704, "loss": 0.41778225898742677, "lr": 9.943797016670059e-06, "epoch": 0.09561001885267978, "percentage": 4.78, "elapsed_time": "3:02:04", "remaining_time": "2 days, 12:26:31", "throughput": 5682.96, "total_tokens": 62081744}
+{"current_steps": 1425, "total_steps": 29704, "loss": 0.41177783012390134, "lr": 9.943400994539565e-06, "epoch": 0.09594667384863992, "percentage": 4.8, "elapsed_time": "3:02:43", "remaining_time": "2 days, 12:26:03", "throughput": 5682.61, "total_tokens": 62299552}
+{"current_steps": 1430, "total_steps": 29704, "loss": 0.4324612617492676, "lr": 9.943003590002744e-06, "epoch": 0.09628332884460006, "percentage": 4.81, "elapsed_time": "3:03:23", "remaining_time": "2 days, 12:26:03", "throughput": 5682.18, "total_tokens": 62524424}
+{"current_steps": 1435, "total_steps": 29704, "loss": 0.4207320690155029, "lr": 9.942604803170733e-06, "epoch": 0.0966199838405602, "percentage": 4.83, "elapsed_time": "3:04:01", "remaining_time": "2 days, 12:25:03", "throughput": 5682.16, "total_tokens": 62736736}
+{"current_steps": 1440, "total_steps": 29704, "loss": 0.4587702751159668, "lr": 9.94220463415505e-06, "epoch": 0.09695663883652034, "percentage": 4.85, "elapsed_time": "3:04:42", "remaining_time": "2 days, 12:25:25", "throughput": 5680.17, "total_tokens": 62950552}
+{"current_steps": 1445, "total_steps": 29704, "loss": 0.4369199752807617, "lr": 9.9418030830676e-06, "epoch": 0.09729329383248048, "percentage": 4.86, "elapsed_time": "3:05:19", "remaining_time": "2 days, 12:24:22", "throughput": 5681.19, "total_tokens": 63173304}
+{"current_steps": 1450, "total_steps": 29704, "loss": 0.45250940322875977, "lr": 9.941400150020676e-06, "epoch": 0.09762994882844062, "percentage": 4.88, "elapsed_time": "3:05:53", "remaining_time": "2 days, 12:22:08", "throughput": 5682.28, "total_tokens": 63376456}
+{"current_steps": 1455, "total_steps": 29704, "loss": 0.4510367393493652, "lr": 9.940995835126957e-06, "epoch": 0.09796660382440076, "percentage": 4.9, "elapsed_time": "3:06:30", "remaining_time": "2 days, 12:21:03", "throughput": 5683.3, "total_tokens": 63598704}
+{"current_steps": 1460, "total_steps": 29704, "loss": 0.4070897102355957, "lr": 9.940590138499508e-06, "epoch": 0.0983032588203609, "percentage": 4.92, "elapsed_time": "3:07:08", "remaining_time": "2 days, 12:20:08", "throughput": 5684.43, "total_tokens": 63824984}
+{"current_steps": 1465, "total_steps": 29704, "loss": 0.424533748626709, "lr": 9.940183060251783e-06, "epoch": 0.09863991381632103, "percentage": 4.93, "elapsed_time": "3:07:49", "remaining_time": "2 days, 12:20:19", "throughput": 5681.08, "total_tokens": 64020440}
+{"current_steps": 1470, "total_steps": 29704, "loss": 0.3772727489471436, "lr": 9.939774600497616e-06, "epoch": 0.09897656881228117, "percentage": 4.95, "elapsed_time": "3:08:25", "remaining_time": "2 days, 12:18:56", "throughput": 5682.16, "total_tokens": 64237912}
+{"current_steps": 1475, "total_steps": 29704, "loss": 0.43614888191223145, "lr": 9.939364759351236e-06, "epoch": 0.09931322380824131, "percentage": 4.97, "elapsed_time": "3:09:02", "remaining_time": "2 days, 12:17:52", "throughput": 5682.5, "total_tokens": 64452808}
+{"current_steps": 1480, "total_steps": 29704, "loss": 0.4052250862121582, "lr": 9.938953536927249e-06, "epoch": 0.09964987880420145, "percentage": 4.98, "elapsed_time": "3:09:40", "remaining_time": "2 days, 12:17:16", "throughput": 5683.17, "total_tokens": 64679536}
+{"current_steps": 1485, "total_steps": 29704, "loss": 0.43981304168701174, "lr": 9.938540933340657e-06, "epoch": 0.09998653380016159, "percentage": 5.0, "elapsed_time": "3:10:14", "remaining_time": "2 days, 12:15:03", "throughput": 5686.5, "total_tokens": 64907672}
+{"current_steps": 1490, "total_steps": 29704, "loss": 0.4235081195831299, "lr": 9.938126948706839e-06, "epoch": 0.10032318879612173, "percentage": 5.02, "elapsed_time": "3:10:51", "remaining_time": "2 days, 12:13:54", "throughput": 5688.05, "total_tokens": 65134656}
+{"current_steps": 1495, "total_steps": 29704, "loss": 0.41340045928955077, "lr": 9.937711583141567e-06, "epoch": 0.10065984379208187, "percentage": 5.03, "elapsed_time": "3:11:28", "remaining_time": "2 days, 12:13:02", "throughput": 5688.4, "total_tokens": 65353344}
+{"current_steps": 1500, "total_steps": 29704, "loss": 0.43211612701416013, "lr": 9.937294836760995e-06, "epoch": 0.10099649878804201, "percentage": 5.05, "elapsed_time": "3:12:05", "remaining_time": "2 days, 12:11:54", "throughput": 5689.08, "total_tokens": 65570904}
+{"current_steps": 1505, "total_steps": 29704, "loss": 0.4765140533447266, "lr": 9.936876709681668e-06, "epoch": 0.10133315378400215, "percentage": 5.07, "elapsed_time": "3:12:47", "remaining_time": "2 days, 12:12:10", "throughput": 5687.19, "total_tokens": 65784200}
+{"current_steps": 1510, "total_steps": 29704, "loss": 0.3693789005279541, "lr": 9.93645720202051e-06, "epoch": 0.10166980877996229, "percentage": 5.08, "elapsed_time": "3:13:23", "remaining_time": "2 days, 12:10:52", "throughput": 5688.2, "total_tokens": 66002240}
+{"current_steps": 1515, "total_steps": 29704, "loss": 0.46529722213745117, "lr": 9.936036313894836e-06, "epoch": 0.10200646377592243, "percentage": 5.1, "elapsed_time": "3:13:57", "remaining_time": "2 days, 12:09:01", "throughput": 5689.79, "total_tokens": 66217160}
+{"current_steps": 1520, "total_steps": 29704, "loss": 0.4474299907684326, "lr": 9.935614045422349e-06, "epoch": 0.10234311877188257, "percentage": 5.12, "elapsed_time": "3:14:34", "remaining_time": "2 days, 12:07:47", "throughput": 5690.85, "total_tokens": 66437216}
+{"current_steps": 1525, "total_steps": 29704, "loss": 0.4093039035797119, "lr": 9.935190396721132e-06, "epoch": 0.10267977376784271, "percentage": 5.13, "elapsed_time": "3:15:14", "remaining_time": "2 days, 12:07:45", "throughput": 5690.24, "total_tokens": 66659632}
+{"current_steps": 1530, "total_steps": 29704, "loss": 0.442030143737793, "lr": 9.934765367909658e-06, "epoch": 0.10301642876380286, "percentage": 5.15, "elapsed_time": "3:15:51", "remaining_time": "2 days, 12:06:28", "throughput": 5690.9, "total_tokens": 66874200}
+{"current_steps": 1535, "total_steps": 29704, "loss": 0.4125958442687988, "lr": 9.934338959106783e-06, "epoch": 0.103353083759763, "percentage": 5.17, "elapsed_time": "3:16:27", "remaining_time": "2 days, 12:05:17", "throughput": 5692.09, "total_tokens": 67096584}
+{"current_steps": 1540, "total_steps": 29704, "loss": 0.45245895385742185, "lr": 9.933911170431753e-06, "epoch": 0.10368973875572314, "percentage": 5.18, "elapsed_time": "3:17:05", "remaining_time": "2 days, 12:04:27", "throughput": 5691.15, "total_tokens": 67300344}
+{"current_steps": 1545, "total_steps": 29704, "loss": 0.4369757652282715, "lr": 9.933482002004198e-06, "epoch": 0.10402639375168328, "percentage": 5.2, "elapsed_time": "3:17:48", "remaining_time": "2 days, 12:05:10", "throughput": 5689.79, "total_tokens": 67528208}
+{"current_steps": 1550, "total_steps": 29704, "loss": 0.47658519744873046, "lr": 9.933051453944134e-06, "epoch": 0.10436304874764342, "percentage": 5.22, "elapsed_time": "3:18:19", "remaining_time": "2 days, 12:02:18", "throughput": 5693.79, "total_tokens": 67752368}
+{"current_steps": 1555, "total_steps": 29704, "loss": 0.4100327014923096, "lr": 9.932619526371958e-06, "epoch": 0.10469970374360356, "percentage": 5.23, "elapsed_time": "3:18:53", "remaining_time": "2 days, 12:00:15", "throughput": 5696.13, "total_tokens": 67972272}
+{"current_steps": 1560, "total_steps": 29704, "loss": 0.40094866752624514, "lr": 9.932186219408463e-06, "epoch": 0.1050363587395637, "percentage": 5.25, "elapsed_time": "3:19:33", "remaining_time": "2 days, 12:00:14", "throughput": 5694.43, "total_tokens": 68182312}
+{"current_steps": 1565, "total_steps": 29704, "loss": 0.44454145431518555, "lr": 9.931751533174819e-06, "epoch": 0.10537301373552384, "percentage": 5.27, "elapsed_time": "3:20:14", "remaining_time": "2 days, 12:00:15", "throughput": 5692.44, "total_tokens": 68389480}
+{"current_steps": 1570, "total_steps": 29704, "loss": 0.45273237228393554, "lr": 9.931315467792584e-06, "epoch": 0.10570966873148398, "percentage": 5.29, "elapsed_time": "3:20:45", "remaining_time": "2 days, 11:57:39", "throughput": 5696.99, "total_tokens": 68625472}
+{"current_steps": 1575, "total_steps": 29704, "loss": 0.4475082874298096, "lr": 9.930878023383705e-06, "epoch": 0.10604632372744412, "percentage": 5.3, "elapsed_time": "3:21:21", "remaining_time": "2 days, 11:56:17", "throughput": 5698.0, "total_tokens": 68842256}
+{"current_steps": 1580, "total_steps": 29704, "loss": 0.39358346462249755, "lr": 9.93043920007051e-06, "epoch": 0.10638297872340426, "percentage": 5.32, "elapsed_time": "3:22:00", "remaining_time": "2 days, 11:55:52", "throughput": 5697.86, "total_tokens": 69063336}
+{"current_steps": 1585, "total_steps": 29704, "loss": 0.40912957191467286, "lr": 9.929998997975712e-06, "epoch": 0.1067196337193644, "percentage": 5.34, "elapsed_time": "3:22:37", "remaining_time": "2 days, 11:54:42", "throughput": 5699.96, "total_tokens": 69297392}
+{"current_steps": 1590, "total_steps": 29704, "loss": 0.4154947280883789, "lr": 9.929557417222417e-06, "epoch": 0.10705628871532454, "percentage": 5.35, "elapsed_time": "3:23:15", "remaining_time": "2 days, 11:53:49", "throughput": 5699.22, "total_tokens": 69502272}
+{"current_steps": 1595, "total_steps": 29704, "loss": 0.433414888381958, "lr": 9.929114457934109e-06, "epoch": 0.10739294371128467, "percentage": 5.37, "elapsed_time": "3:23:55", "remaining_time": "2 days, 11:53:54", "throughput": 5697.9, "total_tokens": 69718712}
+{"current_steps": 1600, "total_steps": 29704, "loss": 0.429976749420166, "lr": 9.92867012023466e-06, "epoch": 0.10772959870724481, "percentage": 5.39, "elapsed_time": "3:24:35", "remaining_time": "2 days, 11:53:32", "throughput": 5697.95, "total_tokens": 69942920}
+{"current_steps": 1605, "total_steps": 29704, "loss": 0.465771484375, "lr": 9.92822440424833e-06, "epoch": 0.10806625370320495, "percentage": 5.4, "elapsed_time": "3:25:11", "remaining_time": "2 days, 11:52:25", "throughput": 5699.17, "total_tokens": 70167408}
+{"current_steps": 1610, "total_steps": 29704, "loss": 0.4795114040374756, "lr": 9.927777310099759e-06, "epoch": 0.1084029086991651, "percentage": 5.42, "elapsed_time": "3:25:55", "remaining_time": "2 days, 11:53:17", "throughput": 5697.76, "total_tokens": 70398096}
+{"current_steps": 1615, "total_steps": 29704, "loss": 0.42298336029052735, "lr": 9.927328837913976e-06, "epoch": 0.10873956369512523, "percentage": 5.44, "elapsed_time": "3:26:32", "remaining_time": "2 days, 11:52:09", "throughput": 5698.92, "total_tokens": 70621368}
+{"current_steps": 1620, "total_steps": 29704, "loss": 0.4128897666931152, "lr": 9.926878987816397e-06, "epoch": 0.10907621869108537, "percentage": 5.45, "elapsed_time": "3:27:11", "remaining_time": "2 days, 11:51:58", "throughput": 5699.04, "total_tokens": 70850496}
+{"current_steps": 1625, "total_steps": 29704, "loss": 0.40384368896484374, "lr": 9.92642775993282e-06, "epoch": 0.10941287368704551, "percentage": 5.47, "elapsed_time": "3:27:52", "remaining_time": "2 days, 11:52:00", "throughput": 5698.77, "total_tokens": 71078912}
+{"current_steps": 1630, "total_steps": 29704, "loss": 0.39646482467651367, "lr": 9.925975154389428e-06, "epoch": 0.10974952868300565, "percentage": 5.49, "elapsed_time": "3:28:32", "remaining_time": "2 days, 11:51:51", "throughput": 5697.53, "total_tokens": 71291944}
+{"current_steps": 1635, "total_steps": 29704, "loss": 0.41344375610351564, "lr": 9.925521171312792e-06, "epoch": 0.11008618367896579, "percentage": 5.5, "elapsed_time": "3:29:12", "remaining_time": "2 days, 11:51:29", "throughput": 5697.3, "total_tokens": 71513056}
+{"current_steps": 1640, "total_steps": 29704, "loss": 0.43964757919311526, "lr": 9.925065810829868e-06, "epoch": 0.11042283867492593, "percentage": 5.52, "elapsed_time": "3:29:52", "remaining_time": "2 days, 11:51:30", "throughput": 5694.73, "total_tokens": 71712776}
+{"current_steps": 1645, "total_steps": 29704, "loss": 0.45009913444519045, "lr": 9.924609073067994e-06, "epoch": 0.11075949367088607, "percentage": 5.54, "elapsed_time": "3:30:37", "remaining_time": "2 days, 11:52:32", "throughput": 5691.09, "total_tokens": 71918888}
+{"current_steps": 1650, "total_steps": 29704, "loss": 0.43289871215820314, "lr": 9.924150958154897e-06, "epoch": 0.11109614866684622, "percentage": 5.55, "elapsed_time": "3:31:19", "remaining_time": "2 days, 11:53:00", "throughput": 5689.92, "total_tokens": 72144880}
+{"current_steps": 1655, "total_steps": 29704, "loss": 0.4033914566040039, "lr": 9.923691466218686e-06, "epoch": 0.11143280366280636, "percentage": 5.57, "elapsed_time": "3:32:02", "remaining_time": "2 days, 11:53:45", "throughput": 5688.11, "total_tokens": 72368424}
+{"current_steps": 1660, "total_steps": 29704, "loss": 0.4340330123901367, "lr": 9.923230597387856e-06, "epoch": 0.1117694586587665, "percentage": 5.59, "elapsed_time": "3:32:40", "remaining_time": "2 days, 11:52:53", "throughput": 5689.85, "total_tokens": 72604440}
+{"current_steps": 1665, "total_steps": 29704, "loss": 0.45328373908996583, "lr": 9.922768351791289e-06, "epoch": 0.11210611365472664, "percentage": 5.61, "elapsed_time": "3:33:24", "remaining_time": "2 days, 11:53:53", "throughput": 5687.2, "total_tokens": 72822440}
+{"current_steps": 1670, "total_steps": 29704, "loss": 0.41826558113098145, "lr": 9.92230472955825e-06, "epoch": 0.11244276865068678, "percentage": 5.62, "elapsed_time": "3:33:58", "remaining_time": "2 days, 11:51:58", "throughput": 5689.34, "total_tokens": 73042800}
+{"current_steps": 1675, "total_steps": 29704, "loss": 0.3946269750595093, "lr": 9.92183973081839e-06, "epoch": 0.11277942364664692, "percentage": 5.64, "elapsed_time": "3:34:42", "remaining_time": "2 days, 11:52:54", "throughput": 5687.34, "total_tokens": 73268056}
+{"current_steps": 1680, "total_steps": 29704, "loss": 0.46369400024414065, "lr": 9.921373355701743e-06, "epoch": 0.11311607864260706, "percentage": 5.66, "elapsed_time": "3:35:24", "remaining_time": "2 days, 11:53:09", "throughput": 5686.94, "total_tokens": 73499832}
+{"current_steps": 1685, "total_steps": 29704, "loss": 0.45543785095214845, "lr": 9.920905604338732e-06, "epoch": 0.1134527336385672, "percentage": 5.67, "elapsed_time": "3:36:03", "remaining_time": "2 days, 11:52:41", "throughput": 5686.8, "total_tokens": 73720264}
+{"current_steps": 1690, "total_steps": 29704, "loss": 0.38962454795837403, "lr": 9.920436476860158e-06, "epoch": 0.11378938863452734, "percentage": 5.69, "elapsed_time": "3:36:44", "remaining_time": "2 days, 11:52:41", "throughput": 5685.25, "total_tokens": 73931824}
+{"current_steps": 1695, "total_steps": 29704, "loss": 0.40491409301757814, "lr": 9.919965973397214e-06, "epoch": 0.11412604363048748, "percentage": 5.71, "elapsed_time": "3:37:17", "remaining_time": "2 days, 11:50:30", "throughput": 5687.7, "total_tokens": 74150736}
+{"current_steps": 1700, "total_steps": 29704, "loss": 0.45804147720336913, "lr": 9.919494094081475e-06, "epoch": 0.11446269862644762, "percentage": 5.72, "elapsed_time": "3:37:53", "remaining_time": "2 days, 11:49:21", "throughput": 5687.16, "total_tokens": 74352136}
+{"current_steps": 1705, "total_steps": 29704, "loss": 0.4247718334197998, "lr": 9.919020839044899e-06, "epoch": 0.11479935362240776, "percentage": 5.74, "elapsed_time": "3:38:28", "remaining_time": "2 days, 11:47:40", "throughput": 5690.16, "total_tokens": 74588328}
+{"current_steps": 1710, "total_steps": 29704, "loss": 0.4040073394775391, "lr": 9.918546208419832e-06, "epoch": 0.1151360086183679, "percentage": 5.76, "elapsed_time": "3:39:12", "remaining_time": "2 days, 11:48:38", "throughput": 5687.85, "total_tokens": 74810232}
+{"current_steps": 1715, "total_steps": 29704, "loss": 0.41643967628479006, "lr": 9.918070202339e-06, "epoch": 0.11547266361432804, "percentage": 5.77, "elapsed_time": "3:39:48", "remaining_time": "2 days, 11:47:09", "throughput": 5689.88, "total_tokens": 75038248}
+{"current_steps": 1720, "total_steps": 29704, "loss": 0.4160469532012939, "lr": 9.91759282093552e-06, "epoch": 0.11580931861028818, "percentage": 5.79, "elapsed_time": "3:40:21", "remaining_time": "2 days, 11:45:12", "throughput": 5692.2, "total_tokens": 75259984}
+{"current_steps": 1725, "total_steps": 29704, "loss": 0.38855648040771484, "lr": 9.917114064342888e-06, "epoch": 0.11614597360624832, "percentage": 5.81, "elapsed_time": "3:41:04", "remaining_time": "2 days, 11:45:52", "throughput": 5689.34, "total_tokens": 75468520}
+{"current_steps": 1730, "total_steps": 29704, "loss": 0.3962873458862305, "lr": 9.916633932694988e-06, "epoch": 0.11648262860220845, "percentage": 5.82, "elapsed_time": "3:41:41", "remaining_time": "2 days, 11:44:47", "throughput": 5690.29, "total_tokens": 75690392}
+{"current_steps": 1735, "total_steps": 29704, "loss": 0.3940844774246216, "lr": 9.916152426126086e-06, "epoch": 0.1168192835981686, "percentage": 5.84, "elapsed_time": "3:42:20", "remaining_time": "2 days, 11:44:12", "throughput": 5690.16, "total_tokens": 75908632}
+{"current_steps": 1740, "total_steps": 29704, "loss": 0.41963586807250974, "lr": 9.915669544770837e-06, "epoch": 0.11715593859412873, "percentage": 5.86, "elapsed_time": "3:42:53", "remaining_time": "2 days, 11:42:13", "throughput": 5693.52, "total_tokens": 76143760}
+{"current_steps": 1745, "total_steps": 29704, "loss": 0.4006998062133789, "lr": 9.915185288764272e-06, "epoch": 0.11749259359008887, "percentage": 5.87, "elapsed_time": "3:43:31", "remaining_time": "2 days, 11:41:26", "throughput": 5695.15, "total_tokens": 76381464}
+{"current_steps": 1750, "total_steps": 29704, "loss": 0.39071149826049806, "lr": 9.914699658241815e-06, "epoch": 0.11782924858604901, "percentage": 5.89, "elapsed_time": "3:44:15", "remaining_time": "2 days, 11:42:19", "throughput": 5693.28, "total_tokens": 76607736}
+{"current_steps": 1755, "total_steps": 29704, "loss": 0.44945535659790037, "lr": 9.91421265333927e-06, "epoch": 0.11816590358200915, "percentage": 5.91, "elapsed_time": "3:44:55", "remaining_time": "2 days, 11:42:04", "throughput": 5693.35, "total_tokens": 76836152}
+{"current_steps": 1760, "total_steps": 29704, "loss": 0.4147852897644043, "lr": 9.913724274192829e-06, "epoch": 0.11850255857796929, "percentage": 5.93, "elapsed_time": "3:45:37", "remaining_time": "2 days, 11:42:24", "throughput": 5692.89, "total_tokens": 77069784}
+{"current_steps": 1765, "total_steps": 29704, "loss": 0.41230435371398927, "lr": 9.91323452093906e-06, "epoch": 0.11883921357392943, "percentage": 5.94, "elapsed_time": "3:46:14", "remaining_time": "2 days, 11:41:23", "throughput": 5693.49, "total_tokens": 77288624}
+{"current_steps": 1770, "total_steps": 29704, "loss": 0.3945260286331177, "lr": 9.912743393714927e-06, "epoch": 0.11917586856988957, "percentage": 5.96, "elapsed_time": "3:46:45", "remaining_time": "2 days, 11:38:45", "throughput": 5697.52, "total_tokens": 77519384}
+{"current_steps": 1775, "total_steps": 29704, "loss": 0.40961732864379885, "lr": 9.912250892657769e-06, "epoch": 0.11951252356584972, "percentage": 5.98, "elapsed_time": "3:47:24", "remaining_time": "2 days, 11:38:05", "throughput": 5697.36, "total_tokens": 77735336}
+{"current_steps": 1780, "total_steps": 29704, "loss": 0.41979494094848635, "lr": 9.911757017905312e-06, "epoch": 0.11984917856180986, "percentage": 5.99, "elapsed_time": "3:47:55", "remaining_time": "2 days, 11:35:39", "throughput": 5699.44, "total_tokens": 77943944}
+{"current_steps": 1785, "total_steps": 29704, "loss": 0.4165688991546631, "lr": 9.911261769595668e-06, "epoch": 0.12018583355777, "percentage": 6.01, "elapsed_time": "3:48:40", "remaining_time": "2 days, 11:36:34", "throughput": 5697.96, "total_tokens": 78176688}
+{"current_steps": 1790, "total_steps": 29704, "loss": 0.4288275718688965, "lr": 9.910765147867332e-06, "epoch": 0.12052248855373014, "percentage": 6.03, "elapsed_time": "3:49:14", "remaining_time": "2 days, 11:34:54", "throughput": 5699.83, "total_tokens": 78398920}
+{"current_steps": 1795, "total_steps": 29704, "loss": 0.42678542137145997, "lr": 9.91026715285918e-06, "epoch": 0.12085914354969028, "percentage": 6.04, "elapsed_time": "3:49:55", "remaining_time": "2 days, 11:34:59", "throughput": 5698.66, "total_tokens": 78617400}
+{"current_steps": 1800, "total_steps": 29704, "loss": 0.4547886371612549, "lr": 9.909767784710476e-06, "epoch": 0.12119579854565042, "percentage": 6.06, "elapsed_time": "3:50:32", "remaining_time": "2 days, 11:33:48", "throughput": 5701.12, "total_tokens": 78858720}
+{"current_steps": 1805, "total_steps": 29704, "loss": 0.4295345783233643, "lr": 9.909267043560868e-06, "epoch": 0.12153245354161056, "percentage": 6.08, "elapsed_time": "3:51:07", "remaining_time": "2 days, 11:32:24", "throughput": 5702.6, "total_tokens": 79081264}
+{"current_steps": 1810, "total_steps": 29704, "loss": 0.4215728759765625, "lr": 9.908764929550388e-06, "epoch": 0.1218691085375707, "percentage": 6.09, "elapsed_time": "3:51:47", "remaining_time": "2 days, 11:32:03", "throughput": 5702.23, "total_tokens": 79301792}
+{"current_steps": 1815, "total_steps": 29704, "loss": 0.4215451717376709, "lr": 9.908261442819444e-06, "epoch": 0.12220576353353084, "percentage": 6.11, "elapsed_time": "3:52:35", "remaining_time": "2 days, 11:33:52", "throughput": 5697.1, "total_tokens": 79503840}
+{"current_steps": 1820, "total_steps": 29704, "loss": 0.3922400951385498, "lr": 9.907756583508842e-06, "epoch": 0.12254241852949098, "percentage": 6.13, "elapsed_time": "3:53:15", "remaining_time": "2 days, 11:33:45", "throughput": 5697.5, "total_tokens": 79740056}
+{"current_steps": 1825, "total_steps": 29704, "loss": 0.40973472595214844, "lr": 9.907250351759761e-06, "epoch": 0.12287907352545112, "percentage": 6.14, "elapsed_time": "3:53:55", "remaining_time": "2 days, 11:33:30", "throughput": 5697.05, "total_tokens": 79961848}
+{"current_steps": 1830, "total_steps": 29704, "loss": 0.4293796539306641, "lr": 9.906742747713766e-06, "epoch": 0.12321572852141126, "percentage": 6.16, "elapsed_time": "3:54:31", "remaining_time": "2 days, 11:32:11", "throughput": 5698.11, "total_tokens": 80180448}
+{"current_steps": 1835, "total_steps": 29704, "loss": 0.4379068374633789, "lr": 9.906233771512808e-06, "epoch": 0.1235523835173714, "percentage": 6.18, "elapsed_time": "3:55:08", "remaining_time": "2 days, 11:31:16", "throughput": 5697.95, "total_tokens": 80391288}
+{"current_steps": 1840, "total_steps": 29704, "loss": 0.44842100143432617, "lr": 9.905723423299222e-06, "epoch": 0.12388903851333154, "percentage": 6.19, "elapsed_time": "3:55:48", "remaining_time": "2 days, 11:30:51", "throughput": 5698.26, "total_tokens": 80619704}
+{"current_steps": 1845, "total_steps": 29704, "loss": 0.402877140045166, "lr": 9.905211703215723e-06, "epoch": 0.12422569350929168, "percentage": 6.21, "elapsed_time": "3:56:23", "remaining_time": "2 days, 11:29:33", "throughput": 5698.89, "total_tokens": 80832568}
+{"current_steps": 1850, "total_steps": 29704, "loss": 0.4264510631561279, "lr": 9.904698611405412e-06, "epoch": 0.12456234850525182, "percentage": 6.23, "elapsed_time": "3:57:03", "remaining_time": "2 days, 11:29:09", "throughput": 5698.97, "total_tokens": 81058296}
+{"current_steps": 1855, "total_steps": 29704, "loss": 0.438901424407959, "lr": 9.904184148011774e-06, "epoch": 0.12489900350121196, "percentage": 6.24, "elapsed_time": "3:57:43", "remaining_time": "2 days, 11:28:50", "throughput": 5698.28, "total_tokens": 81274840}
+{"current_steps": 1860, "total_steps": 29704, "loss": 0.44609222412109373, "lr": 9.903668313178681e-06, "epoch": 0.1252356584971721, "percentage": 6.26, "elapsed_time": "3:58:19", "remaining_time": "2 days, 11:27:44", "throughput": 5699.47, "total_tokens": 81500576}
+{"current_steps": 1865, "total_steps": 29704, "loss": 0.396915864944458, "lr": 9.903151107050377e-06, "epoch": 0.12557231349313225, "percentage": 6.28, "elapsed_time": "3:58:57", "remaining_time": "2 days, 11:27:01", "throughput": 5698.75, "total_tokens": 81707648}
+{"current_steps": 1870, "total_steps": 29704, "loss": 0.42331228256225584, "lr": 9.902632529771502e-06, "epoch": 0.1259089684890924, "percentage": 6.3, "elapsed_time": "3:59:36", "remaining_time": "2 days, 11:26:26", "throughput": 5698.66, "total_tokens": 81926600}
+{"current_steps": 1875, "total_steps": 29704, "loss": 0.42675061225891114, "lr": 9.902112581487074e-06, "epoch": 0.12624562348505253, "percentage": 6.31, "elapsed_time": "4:00:16", "remaining_time": "2 days, 11:26:11", "throughput": 5698.21, "total_tokens": 82148216}
+{"current_steps": 1880, "total_steps": 29704, "loss": 0.4384635925292969, "lr": 9.901591262342494e-06, "epoch": 0.12658227848101267, "percentage": 6.33, "elapsed_time": "4:00:59", "remaining_time": "2 days, 11:26:46", "throughput": 5695.68, "total_tokens": 82358848}
+{"current_steps": 1885, "total_steps": 29704, "loss": 0.4060489654541016, "lr": 9.901068572483548e-06, "epoch": 0.1269189334769728, "percentage": 6.35, "elapsed_time": "4:01:37", "remaining_time": "2 days, 11:25:58", "throughput": 5695.82, "total_tokens": 82576440}
+{"current_steps": 1890, "total_steps": 29704, "loss": 0.47153472900390625, "lr": 9.900544512056402e-06, "epoch": 0.12725558847293295, "percentage": 6.36, "elapsed_time": "4:02:14", "remaining_time": "2 days, 11:24:54", "throughput": 5697.34, "total_tokens": 82807688}
+{"current_steps": 1895, "total_steps": 29704, "loss": 0.4209029197692871, "lr": 9.900019081207612e-06, "epoch": 0.12759224346889309, "percentage": 6.38, "elapsed_time": "4:02:49", "remaining_time": "2 days, 11:23:22", "throughput": 5699.57, "total_tokens": 83038168}
+{"current_steps": 1900, "total_steps": 29704, "loss": 0.42281227111816405, "lr": 9.89949228008411e-06, "epoch": 0.12792889846485322, "percentage": 6.4, "elapsed_time": "4:03:26", "remaining_time": "2 days, 11:22:22", "throughput": 5700.84, "total_tokens": 83267720}
+{"current_steps": 1905, "total_steps": 29704, "loss": 0.42650370597839354, "lr": 9.898964108833216e-06, "epoch": 0.12826555346081336, "percentage": 6.41, "elapsed_time": "4:04:06", "remaining_time": "2 days, 11:22:15", "throughput": 5700.42, "total_tokens": 83492616}
+{"current_steps": 1910, "total_steps": 29704, "loss": 0.4016720771789551, "lr": 9.898434567602629e-06, "epoch": 0.1286022084567735, "percentage": 6.43, "elapsed_time": "4:04:46", "remaining_time": "2 days, 11:21:51", "throughput": 5700.28, "total_tokens": 83715840}
+{"current_steps": 1915, "total_steps": 29704, "loss": 0.4289149284362793, "lr": 9.897903656540437e-06, "epoch": 0.12893886345273364, "percentage": 6.45, "elapsed_time": "4:05:24", "remaining_time": "2 days, 11:21:17", "throughput": 5700.12, "total_tokens": 83934184}
+{"current_steps": 1920, "total_steps": 29704, "loss": 0.43784418106079104, "lr": 9.897371375795103e-06, "epoch": 0.12927551844869378, "percentage": 6.46, "elapsed_time": "4:05:59", "remaining_time": "2 days, 11:19:42", "throughput": 5701.43, "total_tokens": 84150312}
+{"current_steps": 1925, "total_steps": 29704, "loss": 0.44328908920288085, "lr": 9.896837725515484e-06, "epoch": 0.12961217344465392, "percentage": 6.48, "elapsed_time": "4:06:45", "remaining_time": "2 days, 11:20:58", "throughput": 5697.84, "total_tokens": 84361600}
+{"current_steps": 1930, "total_steps": 29704, "loss": 0.39231128692626954, "lr": 9.89630270585081e-06, "epoch": 0.12994882844061406, "percentage": 6.5, "elapsed_time": "4:07:22", "remaining_time": "2 days, 11:19:49", "throughput": 5698.43, "total_tokens": 84577168}
+{"current_steps": 1935, "total_steps": 29704, "loss": 0.4394209384918213, "lr": 9.895766316950694e-06, "epoch": 0.1302854834365742, "percentage": 6.51, "elapsed_time": "4:07:56", "remaining_time": "2 days, 11:18:15", "throughput": 5699.9, "total_tokens": 84796464}
+{"current_steps": 1940, "total_steps": 29704, "loss": 0.41849603652954104, "lr": 9.89522855896514e-06, "epoch": 0.13062213843253434, "percentage": 6.53, "elapsed_time": "4:08:39", "remaining_time": "2 days, 11:18:41", "throughput": 5698.84, "total_tokens": 85025304}
+{"current_steps": 1945, "total_steps": 29704, "loss": 0.4164949893951416, "lr": 9.89468943204453e-06, "epoch": 0.13095879342849448, "percentage": 6.55, "elapsed_time": "4:09:24", "remaining_time": "2 days, 11:19:33", "throughput": 5696.12, "total_tokens": 85239728}
+{"current_steps": 1950, "total_steps": 29704, "loss": 0.41168971061706544, "lr": 9.89414893633963e-06, "epoch": 0.13129544842445462, "percentage": 6.56, "elapsed_time": "4:09:58", "remaining_time": "2 days, 11:17:50", "throughput": 5697.34, "total_tokens": 85451488}
+{"current_steps": 1955, "total_steps": 29704, "loss": 0.42524333000183107, "lr": 9.893607072001584e-06, "epoch": 0.13163210342041476, "percentage": 6.58, "elapsed_time": "4:10:45", "remaining_time": "2 days, 11:19:15", "throughput": 5694.79, "total_tokens": 85681416}
+{"current_steps": 1960, "total_steps": 29704, "loss": 0.41834263801574706, "lr": 9.893063839181925e-06, "epoch": 0.1319687584163749, "percentage": 6.6, "elapsed_time": "4:11:26", "remaining_time": "2 days, 11:19:08", "throughput": 5693.75, "total_tokens": 85897864}
+{"current_steps": 1965, "total_steps": 29704, "loss": 0.4272817611694336, "lr": 9.892519238032566e-06, "epoch": 0.13230541341233504, "percentage": 6.62, "elapsed_time": "4:12:10", "remaining_time": "2 days, 11:19:44", "throughput": 5692.37, "total_tokens": 86126184}
+{"current_steps": 1970, "total_steps": 29704, "loss": 0.45201525688171384, "lr": 9.891973268705804e-06, "epoch": 0.13264206840829518, "percentage": 6.63, "elapsed_time": "4:12:46", "remaining_time": "2 days, 11:18:41", "throughput": 5692.69, "total_tokens": 86340160}
+{"current_steps": 1975, "total_steps": 29704, "loss": 0.4250628471374512, "lr": 9.891425931354316e-06, "epoch": 0.13297872340425532, "percentage": 6.65, "elapsed_time": "4:13:23", "remaining_time": "2 days, 11:17:36", "throughput": 5693.77, "total_tokens": 86565168}
+{"current_steps": 1980, "total_steps": 29704, "loss": 0.41506524085998536, "lr": 9.890877226131165e-06, "epoch": 0.13331537840021546, "percentage": 6.67, "elapsed_time": "4:14:07", "remaining_time": "2 days, 11:18:13", "throughput": 5692.58, "total_tokens": 86796552}
+{"current_steps": 1985, "total_steps": 29704, "loss": 0.4300036907196045, "lr": 9.890327153189793e-06, "epoch": 0.1336520333961756, "percentage": 6.68, "elapsed_time": "4:14:52", "remaining_time": "2 days, 11:19:12", "throughput": 5690.06, "total_tokens": 87016968}
+{"current_steps": 1990, "total_steps": 29704, "loss": 0.4278897285461426, "lr": 9.889775712684026e-06, "epoch": 0.13398868839213574, "percentage": 6.7, "elapsed_time": "4:15:35", "remaining_time": "2 days, 11:19:30", "throughput": 5688.84, "total_tokens": 87240744}
+{"current_steps": 1995, "total_steps": 29704, "loss": 0.42691898345947266, "lr": 9.889222904768074e-06, "epoch": 0.13432534338809587, "percentage": 6.72, "elapsed_time": "4:16:15", "remaining_time": "2 days, 11:19:09", "throughput": 5688.32, "total_tokens": 87458888}
+{"current_steps": 2000, "total_steps": 29704, "loss": 0.41673784255981444, "lr": 9.888668729596528e-06, "epoch": 0.13466199838405601, "percentage": 6.73, "elapsed_time": "4:16:57", "remaining_time": "2 days, 11:19:25", "throughput": 5685.76, "total_tokens": 87661224}
+{"current_steps": 2005, "total_steps": 29704, "loss": 0.4196599006652832, "lr": 9.88811318732436e-06, "epoch": 0.13499865338001615, "percentage": 6.75, "elapsed_time": "4:17:34", "remaining_time": "2 days, 11:18:26", "throughput": 5685.72, "total_tokens": 87871224}
+{"current_steps": 2010, "total_steps": 29704, "loss": 0.43319454193115237, "lr": 9.887556278106927e-06, "epoch": 0.1353353083759763, "percentage": 6.77, "elapsed_time": "4:18:12", "remaining_time": "2 days, 11:17:41", "throughput": 5686.61, "total_tokens": 88101424}
+{"current_steps": 2015, "total_steps": 29704, "loss": 0.42733011245727537, "lr": 9.886998002099964e-06, "epoch": 0.13567196337193643, "percentage": 6.78, "elapsed_time": "4:18:53", "remaining_time": "2 days, 11:17:30", "throughput": 5686.55, "total_tokens": 88331320}
+{"current_steps": 2020, "total_steps": 29704, "loss": 0.4501063346862793, "lr": 9.886438359459595e-06, "epoch": 0.13600861836789657, "percentage": 6.8, "elapsed_time": "4:19:32", "remaining_time": "2 days, 11:17:01", "throughput": 5686.05, "total_tokens": 88546592}
+{"current_steps": 2025, "total_steps": 29704, "loss": 0.3861879587173462, "lr": 9.885877350342319e-06, "epoch": 0.1363452733638567, "percentage": 6.82, "elapsed_time": "4:20:11", "remaining_time": "2 days, 11:16:23", "throughput": 5686.59, "total_tokens": 88774480}
+{"current_steps": 2030, "total_steps": 29704, "loss": 0.3875585556030273, "lr": 9.885314974905024e-06, "epoch": 0.13668192835981685, "percentage": 6.83, "elapsed_time": "4:20:52", "remaining_time": "2 days, 11:16:20", "throughput": 5686.8, "total_tokens": 89011576}
+{"current_steps": 2035, "total_steps": 29704, "loss": 0.39730267524719237, "lr": 9.884751233304973e-06, "epoch": 0.137018583355777, "percentage": 6.85, "elapsed_time": "4:21:26", "remaining_time": "2 days, 11:14:36", "throughput": 5687.69, "total_tokens": 89217520}
+{"current_steps": 2040, "total_steps": 29704, "loss": 0.42992243766784666, "lr": 9.884186125699816e-06, "epoch": 0.13735523835173713, "percentage": 6.87, "elapsed_time": "4:22:06", "remaining_time": "2 days, 11:14:25", "throughput": 5686.64, "total_tokens": 89431816}
+{"current_steps": 2045, "total_steps": 29704, "loss": 0.4123133659362793, "lr": 9.883619652247584e-06, "epoch": 0.13769189334769727, "percentage": 6.88, "elapsed_time": "4:22:48", "remaining_time": "2 days, 11:14:30", "throughput": 5685.54, "total_tokens": 89651968}
+{"current_steps": 2050, "total_steps": 29704, "loss": 0.3782381296157837, "lr": 9.883051813106687e-06, "epoch": 0.1380285483436574, "percentage": 6.9, "elapsed_time": "4:23:33", "remaining_time": "2 days, 11:15:18", "throughput": 5683.45, "total_tokens": 89874624}
+{"current_steps": 2055, "total_steps": 29704, "loss": 0.43795013427734375, "lr": 9.882482608435924e-06, "epoch": 0.13836520333961755, "percentage": 6.92, "elapsed_time": "4:24:11", "remaining_time": "2 days, 11:14:38", "throughput": 5683.71, "total_tokens": 90097216}
+{"current_steps": 2060, "total_steps": 29704, "loss": 0.42223529815673827, "lr": 9.881912038394468e-06, "epoch": 0.1387018583355777, "percentage": 6.94, "elapsed_time": "4:24:57", "remaining_time": "2 days, 11:15:31", "throughput": 5681.36, "total_tokens": 90317784}
+{"current_steps": 2065, "total_steps": 29704, "loss": 0.4242293357849121, "lr": 9.881340103141878e-06, "epoch": 0.13903851333153783, "percentage": 6.95, "elapsed_time": "4:25:40", "remaining_time": "2 days, 11:16:00", "throughput": 5680.18, "total_tokens": 90546968}
+{"current_steps": 2070, "total_steps": 29704, "loss": 0.4057013511657715, "lr": 9.880766802838092e-06, "epoch": 0.13937516832749797, "percentage": 6.97, "elapsed_time": "4:26:23", "remaining_time": "2 days, 11:16:18", "throughput": 5678.46, "total_tokens": 90762816}
+{"current_steps": 2075, "total_steps": 29704, "loss": 0.4075471878051758, "lr": 9.880192137643434e-06, "epoch": 0.1397118233234581, "percentage": 6.99, "elapsed_time": "4:26:57", "remaining_time": "2 days, 11:14:38", "throughput": 5679.93, "total_tokens": 90979408}
+{"current_steps": 2080, "total_steps": 29704, "loss": 0.4198787689208984, "lr": 9.879616107718608e-06, "epoch": 0.14004847831941827, "percentage": 7.0, "elapsed_time": "4:27:31", "remaining_time": "2 days, 11:12:50", "throughput": 5681.71, "total_tokens": 91197560}
+{"current_steps": 2085, "total_steps": 29704, "loss": 0.4208370685577393, "lr": 9.879038713224695e-06, "epoch": 0.1403851333153784, "percentage": 7.02, "elapsed_time": "4:28:13", "remaining_time": "2 days, 11:13:08", "throughput": 5680.51, "total_tokens": 91421872}
+{"current_steps": 2090, "total_steps": 29704, "loss": 0.4030628204345703, "lr": 9.878459954323165e-06, "epoch": 0.14072178831133855, "percentage": 7.04, "elapsed_time": "4:28:46", "remaining_time": "2 days, 11:11:09", "throughput": 5683.03, "total_tokens": 91647112}
+{"current_steps": 2095, "total_steps": 29704, "loss": 0.39506003856658933, "lr": 9.877879831175865e-06, "epoch": 0.1410584433072987, "percentage": 7.05, "elapsed_time": "4:29:20", "remaining_time": "2 days, 11:09:33", "throughput": 5684.91, "total_tokens": 91871896}
+{"current_steps": 2100, "total_steps": 29704, "loss": 0.41777982711791994, "lr": 9.877298343945025e-06, "epoch": 0.14139509830325883, "percentage": 7.07, "elapsed_time": "4:30:03", "remaining_time": "2 days, 11:09:48", "throughput": 5683.68, "total_tokens": 92094456}
+{"current_steps": 2105, "total_steps": 29704, "loss": 0.41480426788330077, "lr": 9.876715492793256e-06, "epoch": 0.14173175329921897, "percentage": 7.09, "elapsed_time": "4:30:42", "remaining_time": "2 days, 11:09:12", "throughput": 5683.72, "total_tokens": 92315440}
+{"current_steps": 2110, "total_steps": 29704, "loss": 0.389608097076416, "lr": 9.87613127788355e-06, "epoch": 0.1420684082951791, "percentage": 7.1, "elapsed_time": "4:31:18", "remaining_time": "2 days, 11:08:05", "throughput": 5684.03, "total_tokens": 92527368}
+{"current_steps": 2115, "total_steps": 29704, "loss": 0.4558421611785889, "lr": 9.87554569937928e-06, "epoch": 0.14240506329113925, "percentage": 7.12, "elapsed_time": "4:31:54", "remaining_time": "2 days, 11:06:59", "throughput": 5685.56, "total_tokens": 92759672}
+{"current_steps": 2120, "total_steps": 29704, "loss": 0.45552263259887693, "lr": 9.874958757444203e-06, "epoch": 0.1427417182870994, "percentage": 7.14, "elapsed_time": "4:32:30", "remaining_time": "2 days, 11:05:43", "throughput": 5687.56, "total_tokens": 92995096}
+{"current_steps": 2125, "total_steps": 29704, "loss": 0.43328495025634767, "lr": 9.874370452242454e-06, "epoch": 0.14307837328305953, "percentage": 7.15, "elapsed_time": "4:33:04", "remaining_time": "2 days, 11:03:58", "throughput": 5690.27, "total_tokens": 93229944}
+{"current_steps": 2130, "total_steps": 29704, "loss": 0.4462221622467041, "lr": 9.873780783938553e-06, "epoch": 0.14341502827901967, "percentage": 7.17, "elapsed_time": "4:33:37", "remaining_time": "2 days, 11:02:09", "throughput": 5692.24, "total_tokens": 93450480}
+{"current_steps": 2135, "total_steps": 29704, "loss": 0.3947939395904541, "lr": 9.873189752697396e-06, "epoch": 0.1437516832749798, "percentage": 7.19, "elapsed_time": "4:34:19", "remaining_time": "2 days, 11:02:17", "throughput": 5691.03, "total_tokens": 93670472}
+{"current_steps": 2140, "total_steps": 29704, "loss": 0.458558464050293, "lr": 9.872597358684265e-06, "epoch": 0.14408833827093995, "percentage": 7.2, "elapsed_time": "4:34:49", "remaining_time": "2 days, 10:59:51", "throughput": 5693.53, "total_tokens": 93883544}
+{"current_steps": 2145, "total_steps": 29704, "loss": 0.4299613475799561, "lr": 9.872003602064816e-06, "epoch": 0.1444249932669001, "percentage": 7.22, "elapsed_time": "4:35:23", "remaining_time": "2 days, 10:58:09", "throughput": 5696.07, "total_tokens": 94116824}
+{"current_steps": 2150, "total_steps": 29704, "loss": 0.4079564571380615, "lr": 9.8714084830051e-06, "epoch": 0.14476164826286023, "percentage": 7.24, "elapsed_time": "4:36:04", "remaining_time": "2 days, 10:58:10", "throughput": 5695.8, "total_tokens": 94349480}
+{"current_steps": 2155, "total_steps": 29704, "loss": 0.40776524543762205, "lr": 9.870812001671533e-06, "epoch": 0.14509830325882037, "percentage": 7.25, "elapsed_time": "4:36:40", "remaining_time": "2 days, 10:57:02", "throughput": 5697.13, "total_tokens": 94577624}
+{"current_steps": 2160, "total_steps": 29704, "loss": 0.4177376747131348, "lr": 9.870214158230922e-06, "epoch": 0.1454349582547805, "percentage": 7.27, "elapsed_time": "4:37:15", "remaining_time": "2 days, 10:55:29", "throughput": 5699.01, "total_tokens": 94804128}
+{"current_steps": 2165, "total_steps": 29704, "loss": 0.40900435447692873, "lr": 9.869614952850454e-06, "epoch": 0.14577161325074064, "percentage": 7.29, "elapsed_time": "4:37:49", "remaining_time": "2 days, 10:53:54", "throughput": 5700.69, "total_tokens": 95026192}
+{"current_steps": 2170, "total_steps": 29704, "loss": 0.42821083068847654, "lr": 9.86901438569769e-06, "epoch": 0.14610826824670078, "percentage": 7.31, "elapsed_time": "4:38:26", "remaining_time": "2 days, 10:53:03", "throughput": 5701.55, "total_tokens": 95254672}
+{"current_steps": 2175, "total_steps": 29704, "loss": 0.425125789642334, "lr": 9.86841245694058e-06, "epoch": 0.14644492324266092, "percentage": 7.32, "elapsed_time": "4:39:06", "remaining_time": "2 days, 10:52:36", "throughput": 5699.81, "total_tokens": 95450280}
+{"current_steps": 2180, "total_steps": 29704, "loss": 0.45731253623962403, "lr": 9.86780916674745e-06, "epoch": 0.14678157823862106, "percentage": 7.34, "elapsed_time": "4:39:43", "remaining_time": "2 days, 10:51:47", "throughput": 5699.0, "total_tokens": 95651304}
+{"current_steps": 2185, "total_steps": 29704, "loss": 0.40560040473937986, "lr": 9.86720451528701e-06, "epoch": 0.1471182332345812, "percentage": 7.36, "elapsed_time": "4:40:22", "remaining_time": "2 days, 10:51:12", "throughput": 5698.28, "total_tokens": 95859952}
+{"current_steps": 2190, "total_steps": 29704, "loss": 0.43126611709594725, "lr": 9.866598502728348e-06, "epoch": 0.14745488823054134, "percentage": 7.37, "elapsed_time": "4:41:02", "remaining_time": "2 days, 10:50:53", "throughput": 5697.14, "total_tokens": 96069168}
+{"current_steps": 2195, "total_steps": 29704, "loss": 0.41145782470703124, "lr": 9.865991129240931e-06, "epoch": 0.14779154322650148, "percentage": 7.39, "elapsed_time": "4:41:36", "remaining_time": "2 days, 10:49:13", "throughput": 5698.52, "total_tokens": 96283672}
+{"current_steps": 2200, "total_steps": 29704, "loss": 0.40999040603637693, "lr": 9.865382394994614e-06, "epoch": 0.14812819822246162, "percentage": 7.41, "elapsed_time": "4:42:19", "remaining_time": "2 days, 10:49:32", "throughput": 5696.95, "total_tokens": 96502528}
+{"current_steps": 2205, "total_steps": 29704, "loss": 0.3990957021713257, "lr": 9.864772300159622e-06, "epoch": 0.14846485321842176, "percentage": 7.42, "elapsed_time": "4:42:53", "remaining_time": "2 days, 10:47:59", "throughput": 5698.31, "total_tokens": 96720064}
+{"current_steps": 2210, "total_steps": 29704, "loss": 0.43683781623840334, "lr": 9.864160844906571e-06, "epoch": 0.1488015082143819, "percentage": 7.44, "elapsed_time": "4:43:29", "remaining_time": "2 days, 10:46:49", "throughput": 5699.27, "total_tokens": 96941160}
+{"current_steps": 2215, "total_steps": 29704, "loss": 0.419160795211792, "lr": 9.863548029406449e-06, "epoch": 0.14913816321034204, "percentage": 7.46, "elapsed_time": "4:44:10", "remaining_time": "2 days, 10:46:44", "throughput": 5698.43, "total_tokens": 97161584}
+{"current_steps": 2220, "total_steps": 29704, "loss": 0.3916934490203857, "lr": 9.86293385383063e-06, "epoch": 0.14947481820630218, "percentage": 7.47, "elapsed_time": "4:44:55", "remaining_time": "2 days, 10:47:28", "throughput": 5695.71, "total_tokens": 97372312}
+{"current_steps": 2225, "total_steps": 29704, "loss": 0.4371610641479492, "lr": 9.862318318350865e-06, "epoch": 0.14981147320226232, "percentage": 7.49, "elapsed_time": "4:45:32", "remaining_time": "2 days, 10:46:27", "throughput": 5696.45, "total_tokens": 97594144}
+{"current_steps": 2230, "total_steps": 29704, "loss": 0.39911181926727296, "lr": 9.861701423139287e-06, "epoch": 0.15014812819822246, "percentage": 7.51, "elapsed_time": "4:46:05", "remaining_time": "2 days, 10:44:47", "throughput": 5698.55, "total_tokens": 97820888}
+{"current_steps": 2235, "total_steps": 29704, "loss": 0.41806850433349607, "lr": 9.861083168368411e-06, "epoch": 0.1504847831941826, "percentage": 7.52, "elapsed_time": "4:46:44", "remaining_time": "2 days, 10:44:13", "throughput": 5698.76, "total_tokens": 98045776}
+{"current_steps": 2240, "total_steps": 29704, "loss": 0.3891399383544922, "lr": 9.860463554211125e-06, "epoch": 0.15082143819014274, "percentage": 7.54, "elapsed_time": "4:47:26", "remaining_time": "2 days, 10:44:09", "throughput": 5698.01, "total_tokens": 98268344}
+{"current_steps": 2245, "total_steps": 29704, "loss": 0.4640520095825195, "lr": 9.859842580840706e-06, "epoch": 0.15115809318610288, "percentage": 7.56, "elapsed_time": "4:48:09", "remaining_time": "2 days, 10:44:30", "throughput": 5696.71, "total_tokens": 98493000}
+{"current_steps": 2250, "total_steps": 29704, "loss": 0.41865386962890627, "lr": 9.859220248430806e-06, "epoch": 0.15149474818206302, "percentage": 7.57, "elapsed_time": "4:48:49", "remaining_time": "2 days, 10:44:07", "throughput": 5697.37, "total_tokens": 98731240}
+{"current_steps": 2255, "total_steps": 29704, "loss": 0.39298291206359864, "lr": 9.85859655715546e-06, "epoch": 0.15183140317802316, "percentage": 7.59, "elapsed_time": "4:49:35", "remaining_time": "2 days, 10:44:57", "throughput": 5695.14, "total_tokens": 98953224}
+{"current_steps": 2260, "total_steps": 29704, "loss": 0.41016359329223634, "lr": 9.857971507189077e-06, "epoch": 0.1521680581739833, "percentage": 7.61, "elapsed_time": "4:50:14", "remaining_time": "2 days, 10:44:25", "throughput": 5694.92, "total_tokens": 99171512}
+{"current_steps": 2265, "total_steps": 29704, "loss": 0.4255502700805664, "lr": 9.857345098706455e-06, "epoch": 0.15250471316994343, "percentage": 7.63, "elapsed_time": "4:50:51", "remaining_time": "2 days, 10:43:34", "throughput": 5695.88, "total_tokens": 99402080}
+{"current_steps": 2270, "total_steps": 29704, "loss": 0.4265317440032959, "lr": 9.856717331882765e-06, "epoch": 0.15284136816590357, "percentage": 7.64, "elapsed_time": "4:51:26", "remaining_time": "2 days, 10:42:14", "throughput": 5696.21, "total_tokens": 99607944}
+{"current_steps": 2275, "total_steps": 29704, "loss": 0.39440102577209474, "lr": 9.85608820689356e-06, "epoch": 0.1531780231618637, "percentage": 7.66, "elapsed_time": "4:52:08", "remaining_time": "2 days, 10:42:10", "throughput": 5695.47, "total_tokens": 99830656}
+{"current_steps": 2280, "total_steps": 29704, "loss": 0.4084943771362305, "lr": 9.855457723914772e-06, "epoch": 0.15351467815782385, "percentage": 7.68, "elapsed_time": "4:52:42", "remaining_time": "2 days, 10:40:40", "throughput": 5696.93, "total_tokens": 100050880}
+{"current_steps": 2285, "total_steps": 29704, "loss": 0.4020846366882324, "lr": 9.854825883122713e-06, "epoch": 0.153851333153784, "percentage": 7.69, "elapsed_time": "4:53:22", "remaining_time": "2 days, 10:40:18", "throughput": 5695.93, "total_tokens": 100260608}
+{"current_steps": 2290, "total_steps": 29704, "loss": 0.39870567321777345, "lr": 9.85419268469408e-06, "epoch": 0.15418798814974413, "percentage": 7.71, "elapsed_time": "4:53:58", "remaining_time": "2 days, 10:39:09", "throughput": 5695.98, "total_tokens": 100466752}
+{"current_steps": 2295, "total_steps": 29704, "loss": 0.3938899517059326, "lr": 9.85355812880594e-06, "epoch": 0.15452464314570427, "percentage": 7.73, "elapsed_time": "4:54:37", "remaining_time": "2 days, 10:38:41", "throughput": 5695.72, "total_tokens": 100686544}
+{"current_steps": 2300, "total_steps": 29704, "loss": 0.37209017276763917, "lr": 9.852922215635747e-06, "epoch": 0.1548612981416644, "percentage": 7.74, "elapsed_time": "4:55:19", "remaining_time": "2 days, 10:38:43", "throughput": 5694.81, "total_tokens": 100908792}
+{"current_steps": 2305, "total_steps": 29704, "loss": 0.4363281726837158, "lr": 9.85228494536133e-06, "epoch": 0.15519795313762455, "percentage": 7.76, "elapsed_time": "4:55:53", "remaining_time": "2 days, 10:37:05", "throughput": 5696.64, "total_tokens": 101132472}
+{"current_steps": 2310, "total_steps": 29704, "loss": 0.41831111907958984, "lr": 9.851646318160901e-06, "epoch": 0.1555346081335847, "percentage": 7.78, "elapsed_time": "4:56:33", "remaining_time": "2 days, 10:36:54", "throughput": 5695.75, "total_tokens": 101349024}
+{"current_steps": 2315, "total_steps": 29704, "loss": 0.39124910831451415, "lr": 9.85100633421305e-06, "epoch": 0.15587126312954483, "percentage": 7.79, "elapsed_time": "4:57:15", "remaining_time": "2 days, 10:36:58", "throughput": 5694.07, "total_tokens": 101558936}
+{"current_steps": 2320, "total_steps": 29704, "loss": 0.42600722312927247, "lr": 9.850364993696749e-06, "epoch": 0.15620791812550497, "percentage": 7.81, "elapsed_time": "4:57:55", "remaining_time": "2 days, 10:36:31", "throughput": 5693.41, "total_tokens": 101772184}
+{"current_steps": 2325, "total_steps": 29704, "loss": 0.4083821773529053, "lr": 9.849722296791339e-06, "epoch": 0.15654457312146514, "percentage": 7.83, "elapsed_time": "4:58:29", "remaining_time": "2 days, 10:34:58", "throughput": 5694.39, "total_tokens": 101982520}
+{"current_steps": 2330, "total_steps": 29704, "loss": 0.4323273658752441, "lr": 9.849078243676558e-06, "epoch": 0.15688122811742528, "percentage": 7.84, "elapsed_time": "4:59:10", "remaining_time": "2 days, 10:34:49", "throughput": 5693.31, "total_tokens": 102196704}
+{"current_steps": 2335, "total_steps": 29704, "loss": 0.3904409885406494, "lr": 9.848432834532506e-06, "epoch": 0.15721788311338541, "percentage": 7.86, "elapsed_time": "4:59:50", "remaining_time": "2 days, 10:34:26", "throughput": 5692.76, "total_tokens": 102413872}
+{"current_steps": 2340, "total_steps": 29704, "loss": 0.42157392501831054, "lr": 9.847786069539673e-06, "epoch": 0.15755453810934555, "percentage": 7.88, "elapsed_time": "5:00:27", "remaining_time": "2 days, 10:33:34", "throughput": 5693.33, "total_tokens": 102636848}
+{"current_steps": 2345, "total_steps": 29704, "loss": 0.4385071277618408, "lr": 9.847137948878926e-06, "epoch": 0.1578911931053057, "percentage": 7.89, "elapsed_time": "5:01:05", "remaining_time": "2 days, 10:32:47", "throughput": 5693.85, "total_tokens": 102861544}
+{"current_steps": 2350, "total_steps": 29704, "loss": 0.46036734580993655, "lr": 9.846488472731505e-06, "epoch": 0.15822784810126583, "percentage": 7.91, "elapsed_time": "5:01:40", "remaining_time": "2 days, 10:31:30", "throughput": 5695.54, "total_tokens": 103092600}
+{"current_steps": 2355, "total_steps": 29704, "loss": 0.4151791572570801, "lr": 9.845837641279038e-06, "epoch": 0.15856450309722597, "percentage": 7.93, "elapsed_time": "5:02:11", "remaining_time": "2 days, 10:29:19", "throughput": 5698.72, "total_tokens": 103323896}
+{"current_steps": 2360, "total_steps": 29704, "loss": 0.42508544921875, "lr": 9.845185454703528e-06, "epoch": 0.1589011580931861, "percentage": 7.95, "elapsed_time": "5:02:51", "remaining_time": "2 days, 10:28:57", "throughput": 5698.18, "total_tokens": 103541848}
+{"current_steps": 2365, "total_steps": 29704, "loss": 0.3977517604827881, "lr": 9.844531913187356e-06, "epoch": 0.15923781308914625, "percentage": 7.96, "elapsed_time": "5:03:29", "remaining_time": "2 days, 10:28:15", "throughput": 5698.16, "total_tokens": 103759048}
+{"current_steps": 2370, "total_steps": 29704, "loss": 0.41399388313293456, "lr": 9.84387701691328e-06, "epoch": 0.1595744680851064, "percentage": 7.98, "elapsed_time": "5:04:12", "remaining_time": "2 days, 10:28:32", "throughput": 5695.58, "total_tokens": 103958648}
+{"current_steps": 2375, "total_steps": 29704, "loss": 0.4396827697753906, "lr": 9.843220766064443e-06, "epoch": 0.15991112308106653, "percentage": 8.0, "elapsed_time": "5:04:47", "remaining_time": "2 days, 10:27:09", "throughput": 5696.02, "total_tokens": 104164280}
+{"current_steps": 2380, "total_steps": 29704, "loss": 0.4241004943847656, "lr": 9.842563160824364e-06, "epoch": 0.16024777807702667, "percentage": 8.01, "elapsed_time": "5:05:22", "remaining_time": "2 days, 10:25:54", "throughput": 5696.64, "total_tokens": 104376504}
+{"current_steps": 2385, "total_steps": 29704, "loss": 0.44525814056396484, "lr": 9.841904201376939e-06, "epoch": 0.1605844330729868, "percentage": 8.03, "elapsed_time": "5:06:01", "remaining_time": "2 days, 10:25:21", "throughput": 5697.13, "total_tokens": 104607616}
+{"current_steps": 2390, "total_steps": 29704, "loss": 0.3924037218093872, "lr": 9.841243887906443e-06, "epoch": 0.16092108806894695, "percentage": 8.05, "elapsed_time": "5:06:37", "remaining_time": "2 days, 10:24:16", "throughput": 5697.89, "total_tokens": 104827648}
+{"current_steps": 2395, "total_steps": 29704, "loss": 0.38776795864105223, "lr": 9.840582220597532e-06, "epoch": 0.1612577430649071, "percentage": 8.06, "elapsed_time": "5:07:17", "remaining_time": "2 days, 10:23:58", "throughput": 5697.43, "total_tokens": 105048616}
+{"current_steps": 2400, "total_steps": 29704, "loss": 0.44150123596191404, "lr": 9.839919199635238e-06, "epoch": 0.16159439806086723, "percentage": 8.08, "elapsed_time": "5:08:01", "remaining_time": "2 days, 10:24:12", "throughput": 5696.19, "total_tokens": 105271480}
+{"current_steps": 2405, "total_steps": 29704, "loss": 0.4216805934906006, "lr": 9.839254825204973e-06, "epoch": 0.16193105305682737, "percentage": 8.1, "elapsed_time": "5:08:42", "remaining_time": "2 days, 10:24:10", "throughput": 5695.58, "total_tokens": 105497528}
+{"current_steps": 2410, "total_steps": 29704, "loss": 0.4195371627807617, "lr": 9.838589097492527e-06, "epoch": 0.1622677080527875, "percentage": 8.11, "elapsed_time": "5:09:22", "remaining_time": "2 days, 10:23:43", "throughput": 5695.09, "total_tokens": 105713952}
+{"current_steps": 2415, "total_steps": 29704, "loss": 0.4166804313659668, "lr": 9.83792201668407e-06, "epoch": 0.16260436304874765, "percentage": 8.13, "elapsed_time": "5:09:58", "remaining_time": "2 days, 10:22:38", "throughput": 5696.23, "total_tokens": 105940976}
+{"current_steps": 2420, "total_steps": 29704, "loss": 0.4339381217956543, "lr": 9.837253582966146e-06, "epoch": 0.16294101804470779, "percentage": 8.15, "elapsed_time": "5:10:33", "remaining_time": "2 days, 10:21:16", "throughput": 5697.67, "total_tokens": 106165352}
+{"current_steps": 2425, "total_steps": 29704, "loss": 0.3936171531677246, "lr": 9.836583796525686e-06, "epoch": 0.16327767304066793, "percentage": 8.16, "elapsed_time": "5:11:15", "remaining_time": "2 days, 10:21:18", "throughput": 5695.5, "total_tokens": 106364536}
+{"current_steps": 2430, "total_steps": 29704, "loss": 0.4514013290405273, "lr": 9.835912657549986e-06, "epoch": 0.16361432803662807, "percentage": 8.18, "elapsed_time": "5:11:57", "remaining_time": "2 days, 10:21:26", "throughput": 5694.44, "total_tokens": 106587560}
+{"current_steps": 2435, "total_steps": 29704, "loss": 0.39427995681762695, "lr": 9.835240166226734e-06, "epoch": 0.1639509830325882, "percentage": 8.2, "elapsed_time": "5:12:41", "remaining_time": "2 days, 10:21:41", "throughput": 5692.6, "total_tokens": 106799528}
+{"current_steps": 2440, "total_steps": 29704, "loss": 0.43557186126708985, "lr": 9.834566322743987e-06, "epoch": 0.16428763802854834, "percentage": 8.21, "elapsed_time": "5:13:15", "remaining_time": "2 days, 10:20:21", "throughput": 5693.37, "total_tokens": 107012480}
+{"current_steps": 2445, "total_steps": 29704, "loss": 0.4289062976837158, "lr": 9.833891127290186e-06, "epoch": 0.16462429302450848, "percentage": 8.23, "elapsed_time": "5:13:47", "remaining_time": "2 days, 10:18:24", "throughput": 5696.5, "total_tokens": 107250496}
+{"current_steps": 2450, "total_steps": 29704, "loss": 0.3959768295288086, "lr": 9.833214580054145e-06, "epoch": 0.16496094802046862, "percentage": 8.25, "elapsed_time": "5:14:23", "remaining_time": "2 days, 10:17:15", "throughput": 5697.31, "total_tokens": 107469136}
+{"current_steps": 2455, "total_steps": 29704, "loss": 0.44450016021728517, "lr": 9.832536681225058e-06, "epoch": 0.16529760301642876, "percentage": 8.26, "elapsed_time": "5:14:58", "remaining_time": "2 days, 10:15:59", "throughput": 5698.48, "total_tokens": 107691864}
+{"current_steps": 2460, "total_steps": 29704, "loss": 0.3906950235366821, "lr": 9.831857430992497e-06, "epoch": 0.1656342580123889, "percentage": 8.28, "elapsed_time": "5:15:33", "remaining_time": "2 days, 10:14:45", "throughput": 5699.19, "total_tokens": 107905928}
+{"current_steps": 2465, "total_steps": 29704, "loss": 0.41658635139465333, "lr": 9.831176829546416e-06, "epoch": 0.16597091300834904, "percentage": 8.3, "elapsed_time": "5:16:10", "remaining_time": "2 days, 10:13:48", "throughput": 5699.74, "total_tokens": 108126176}
+{"current_steps": 2470, "total_steps": 29704, "loss": 0.46334171295166016, "lr": 9.830494877077137e-06, "epoch": 0.16630756800430918, "percentage": 8.32, "elapsed_time": "5:16:44", "remaining_time": "2 days, 10:12:26", "throughput": 5701.61, "total_tokens": 108358872}
+{"current_steps": 2475, "total_steps": 29704, "loss": 0.4094019889831543, "lr": 9.82981157377537e-06, "epoch": 0.16664422300026932, "percentage": 8.33, "elapsed_time": "5:17:27", "remaining_time": "2 days, 10:12:30", "throughput": 5700.16, "total_tokens": 108572432}
+{"current_steps": 2480, "total_steps": 29704, "loss": 0.3822948455810547, "lr": 9.829126919832198e-06, "epoch": 0.16698087799622946, "percentage": 8.35, "elapsed_time": "5:18:06", "remaining_time": "2 days, 10:11:57", "throughput": 5699.9, "total_tokens": 108789616}
+{"current_steps": 2485, "total_steps": 29704, "loss": 0.4369968414306641, "lr": 9.82844091543908e-06, "epoch": 0.1673175329921896, "percentage": 8.37, "elapsed_time": "5:18:45", "remaining_time": "2 days, 10:11:30", "throughput": 5700.37, "total_tokens": 109023616}
+{"current_steps": 2490, "total_steps": 29704, "loss": 0.4084890842437744, "lr": 9.82775356078786e-06, "epoch": 0.16765418798814974, "percentage": 8.38, "elapsed_time": "5:19:27", "remaining_time": "2 days, 10:11:25", "throughput": 5699.22, "total_tokens": 109238536}
+{"current_steps": 2495, "total_steps": 29704, "loss": 0.3479023456573486, "lr": 9.82706485607075e-06, "epoch": 0.16799084298410988, "percentage": 8.4, "elapsed_time": "5:20:02", "remaining_time": "2 days, 10:10:15", "throughput": 5699.56, "total_tokens": 109448064}
+{"current_steps": 2500, "total_steps": 29704, "loss": 0.4044170379638672, "lr": 9.826374801480346e-06, "epoch": 0.16832749798007002, "percentage": 8.42, "elapsed_time": "5:20:40", "remaining_time": "2 days, 10:09:26", "throughput": 5699.18, "total_tokens": 109654600}
+{"current_steps": 2505, "total_steps": 29704, "loss": 0.4128159999847412, "lr": 9.825683397209617e-06, "epoch": 0.16866415297603016, "percentage": 8.43, "elapsed_time": "5:21:17", "remaining_time": "2 days, 10:08:27", "throughput": 5698.23, "total_tokens": 109844880}
+{"current_steps": 2510, "total_steps": 29704, "loss": 0.40271682739257814, "lr": 9.824990643451915e-06, "epoch": 0.1690008079719903, "percentage": 8.45, "elapsed_time": "5:21:57", "remaining_time": "2 days, 10:08:05", "throughput": 5698.2, "total_tokens": 110072368}
+{"current_steps": 2515, "total_steps": 29704, "loss": 0.4032273769378662, "lr": 9.824296540400965e-06, "epoch": 0.16933746296795044, "percentage": 8.47, "elapsed_time": "5:22:39", "remaining_time": "2 days, 10:08:10", "throughput": 5697.79, "total_tokens": 110306168}
+{"current_steps": 2520, "total_steps": 29704, "loss": 0.43148112297058105, "lr": 9.823601088250872e-06, "epoch": 0.16967411796391058, "percentage": 8.48, "elapsed_time": "5:23:19", "remaining_time": "2 days, 10:07:49", "throughput": 5696.89, "total_tokens": 110517392}
+{"current_steps": 2525, "total_steps": 29704, "loss": 0.4280102729797363, "lr": 9.822904287196114e-06, "epoch": 0.17001077295987072, "percentage": 8.5, "elapsed_time": "5:23:57", "remaining_time": "2 days, 10:07:01", "throughput": 5696.76, "total_tokens": 110729000}
+{"current_steps": 2530, "total_steps": 29704, "loss": 0.4008312702178955, "lr": 9.822206137431553e-06, "epoch": 0.17034742795583085, "percentage": 8.52, "elapsed_time": "5:24:33", "remaining_time": "2 days, 10:05:55", "throughput": 5697.67, "total_tokens": 110951576}
+{"current_steps": 2535, "total_steps": 29704, "loss": 0.42267236709594724, "lr": 9.821506639152421e-06, "epoch": 0.170684082951791, "percentage": 8.53, "elapsed_time": "5:25:11", "remaining_time": "2 days, 10:05:20", "throughput": 5697.32, "total_tokens": 111165912}
+{"current_steps": 2540, "total_steps": 29704, "loss": 0.3978217124938965, "lr": 9.820805792554334e-06, "epoch": 0.17102073794775113, "percentage": 8.55, "elapsed_time": "5:25:48", "remaining_time": "2 days, 10:04:25", "throughput": 5698.04, "total_tokens": 111390648}
+{"current_steps": 2545, "total_steps": 29704, "loss": 0.40553808212280273, "lr": 9.820103597833278e-06, "epoch": 0.17135739294371127, "percentage": 8.57, "elapsed_time": "5:26:28", "remaining_time": "2 days, 10:04:02", "throughput": 5697.81, "total_tokens": 111613680}
+{"current_steps": 2550, "total_steps": 29704, "loss": 0.3968106031417847, "lr": 9.819400055185622e-06, "epoch": 0.1716940479396714, "percentage": 8.58, "elapsed_time": "5:27:04", "remaining_time": "2 days, 10:02:50", "throughput": 5698.63, "total_tokens": 111830672}
+{"current_steps": 2555, "total_steps": 29704, "loss": 0.40885286331176757, "lr": 9.818695164808108e-06, "epoch": 0.17203070293563155, "percentage": 8.6, "elapsed_time": "5:27:48", "remaining_time": "2 days, 10:03:16", "throughput": 5696.75, "total_tokens": 112047832}
+{"current_steps": 2560, "total_steps": 29704, "loss": 0.39391307830810546, "lr": 9.817988926897856e-06, "epoch": 0.1723673579315917, "percentage": 8.62, "elapsed_time": "5:28:30", "remaining_time": "2 days, 10:03:17", "throughput": 5695.67, "total_tokens": 112267160}
+{"current_steps": 2565, "total_steps": 29704, "loss": 0.4082347869873047, "lr": 9.817281341652367e-06, "epoch": 0.17270401292755186, "percentage": 8.64, "elapsed_time": "5:29:09", "remaining_time": "2 days, 10:02:37", "throughput": 5696.0, "total_tokens": 112491800}
+{"current_steps": 2570, "total_steps": 29704, "loss": 0.3824393033981323, "lr": 9.816572409269508e-06, "epoch": 0.173040667923512, "percentage": 8.65, "elapsed_time": "5:29:44", "remaining_time": "2 days, 10:01:22", "throughput": 5697.4, "total_tokens": 112719560}
+{"current_steps": 2575, "total_steps": 29704, "loss": 0.4002216339111328, "lr": 9.815862129947537e-06, "epoch": 0.17337732291947214, "percentage": 8.67, "elapsed_time": "5:30:21", "remaining_time": "2 days, 10:00:32", "throughput": 5697.8, "total_tokens": 112940512}
+{"current_steps": 2580, "total_steps": 29704, "loss": 0.37178959846496584, "lr": 9.815150503885078e-06, "epoch": 0.17371397791543228, "percentage": 8.69, "elapsed_time": "5:30:58", "remaining_time": "2 days, 9:59:32", "throughput": 5698.69, "total_tokens": 113165624}
+{"current_steps": 2585, "total_steps": 29704, "loss": 0.40114603042602537, "lr": 9.814437531281133e-06, "epoch": 0.17405063291139242, "percentage": 8.7, "elapsed_time": "5:31:43", "remaining_time": "2 days, 10:00:05", "throughput": 5696.83, "total_tokens": 113386744}
+{"current_steps": 2590, "total_steps": 29704, "loss": 0.43936843872070314, "lr": 9.813723212335085e-06, "epoch": 0.17438728790735256, "percentage": 8.72, "elapsed_time": "5:32:20", "remaining_time": "2 days, 9:59:10", "throughput": 5696.66, "total_tokens": 113593688}
+{"current_steps": 2595, "total_steps": 29704, "loss": 0.39149506092071534, "lr": 9.813007547246691e-06, "epoch": 0.1747239429033127, "percentage": 8.74, "elapsed_time": "5:32:56", "remaining_time": "2 days, 9:58:04", "throughput": 5697.28, "total_tokens": 113810536}
+{"current_steps": 2600, "total_steps": 29704, "loss": 0.4519649982452393, "lr": 9.812290536216083e-06, "epoch": 0.17506059789927284, "percentage": 8.75, "elapsed_time": "5:33:34", "remaining_time": "2 days, 9:57:22", "throughput": 5696.52, "total_tokens": 114012584}
+{"current_steps": 2605, "total_steps": 29704, "loss": 0.3978102445602417, "lr": 9.811572179443773e-06, "epoch": 0.17539725289523297, "percentage": 8.77, "elapsed_time": "5:34:09", "remaining_time": "2 days, 9:56:12", "throughput": 5697.79, "total_tokens": 114240040}
+{"current_steps": 2610, "total_steps": 29704, "loss": 0.40557031631469725, "lr": 9.810852477130645e-06, "epoch": 0.17573390789119311, "percentage": 8.79, "elapsed_time": "5:34:49", "remaining_time": "2 days, 9:55:46", "throughput": 5696.98, "total_tokens": 114449696}
+{"current_steps": 2615, "total_steps": 29704, "loss": 0.38374016284942625, "lr": 9.810131429477964e-06, "epoch": 0.17607056288715325, "percentage": 8.8, "elapsed_time": "5:35:21", "remaining_time": "2 days, 9:54:00", "throughput": 5699.26, "total_tokens": 114677776}
+{"current_steps": 2620, "total_steps": 29704, "loss": 0.396789026260376, "lr": 9.809409036687364e-06, "epoch": 0.1764072178831134, "percentage": 8.82, "elapsed_time": "5:35:58", "remaining_time": "2 days, 9:53:11", "throughput": 5699.52, "total_tokens": 114896336}
+{"current_steps": 2625, "total_steps": 29704, "loss": 0.4137075424194336, "lr": 9.808685298960867e-06, "epoch": 0.17674387287907353, "percentage": 8.84, "elapsed_time": "5:36:41", "remaining_time": "2 days, 9:53:15", "throughput": 5698.82, "total_tokens": 115125264}
+{"current_steps": 2630, "total_steps": 29704, "loss": 0.4165202617645264, "lr": 9.807960216500858e-06, "epoch": 0.17708052787503367, "percentage": 8.85, "elapsed_time": "5:37:16", "remaining_time": "2 days, 9:51:59", "throughput": 5699.8, "total_tokens": 115343304}
+{"current_steps": 2635, "total_steps": 29704, "loss": 0.38329482078552246, "lr": 9.807233789510106e-06, "epoch": 0.1774171828709938, "percentage": 8.87, "elapsed_time": "5:37:59", "remaining_time": "2 days, 9:52:12", "throughput": 5698.03, "total_tokens": 115555208}
+{"current_steps": 2640, "total_steps": 29704, "loss": 0.39183108806610106, "lr": 9.806506018191754e-06, "epoch": 0.17775383786695395, "percentage": 8.89, "elapsed_time": "5:38:37", "remaining_time": "2 days, 9:51:21", "throughput": 5698.51, "total_tokens": 115777312}
+{"current_steps": 2645, "total_steps": 29704, "loss": 0.39883904457092284, "lr": 9.80577690274932e-06, "epoch": 0.1780904928629141, "percentage": 8.9, "elapsed_time": "5:39:13", "remaining_time": "2 days, 9:50:25", "throughput": 5699.16, "total_tokens": 116000320}
+{"current_steps": 2650, "total_steps": 29704, "loss": 0.3788264751434326, "lr": 9.8050464433867e-06, "epoch": 0.17842714785887423, "percentage": 8.92, "elapsed_time": "5:39:48", "remaining_time": "2 days, 9:49:04", "throughput": 5700.76, "total_tokens": 116228264}
+{"current_steps": 2655, "total_steps": 29704, "loss": 0.43610692024230957, "lr": 9.804314640308165e-06, "epoch": 0.17876380285483437, "percentage": 8.94, "elapsed_time": "5:40:33", "remaining_time": "2 days, 9:49:37", "throughput": 5699.4, "total_tokens": 116459784}
+{"current_steps": 2660, "total_steps": 29704, "loss": 0.4152191162109375, "lr": 9.803581493718361e-06, "epoch": 0.1791004578507945, "percentage": 8.96, "elapsed_time": "5:41:11", "remaining_time": "2 days, 9:48:47", "throughput": 5699.65, "total_tokens": 116677976}
+{"current_steps": 2665, "total_steps": 29704, "loss": 0.4158869743347168, "lr": 9.80284700382231e-06, "epoch": 0.17943711284675465, "percentage": 8.97, "elapsed_time": "5:41:53", "remaining_time": "2 days, 9:48:54", "throughput": 5699.19, "total_tokens": 116913096}
+{"current_steps": 2670, "total_steps": 29704, "loss": 0.40905094146728516, "lr": 9.802111170825408e-06, "epoch": 0.1797737678427148, "percentage": 8.99, "elapsed_time": "5:42:35", "remaining_time": "2 days, 9:48:45", "throughput": 5698.7, "total_tokens": 117139040}
+{"current_steps": 2675, "total_steps": 29704, "loss": 0.4070711135864258, "lr": 9.80137399493343e-06, "epoch": 0.18011042283867493, "percentage": 9.01, "elapsed_time": "5:43:14", "remaining_time": "2 days, 9:48:09", "throughput": 5698.42, "total_tokens": 117354016}
+{"current_steps": 2680, "total_steps": 29704, "loss": 0.39305026531219484, "lr": 9.800635476352525e-06, "epoch": 0.18044707783463507, "percentage": 9.02, "elapsed_time": "5:43:56", "remaining_time": "2 days, 9:48:09", "throughput": 5696.36, "total_tokens": 117552920}
+{"current_steps": 2685, "total_steps": 29704, "loss": 0.38626153469085694, "lr": 9.799895615289218e-06, "epoch": 0.1807837328305952, "percentage": 9.04, "elapsed_time": "5:44:37", "remaining_time": "2 days, 9:47:56", "throughput": 5695.19, "total_tokens": 117762232}
+{"current_steps": 2690, "total_steps": 29704, "loss": 0.44461402893066404, "lr": 9.799154411950407e-06, "epoch": 0.18112038782655535, "percentage": 9.06, "elapsed_time": "5:45:16", "remaining_time": "2 days, 9:47:24", "throughput": 5695.8, "total_tokens": 117998240}
+{"current_steps": 2695, "total_steps": 29704, "loss": 0.4056252479553223, "lr": 9.79841186654337e-06, "epoch": 0.18145704282251549, "percentage": 9.07, "elapsed_time": "5:45:52", "remaining_time": "2 days, 9:46:23", "throughput": 5696.14, "total_tokens": 118211832}
+{"current_steps": 2700, "total_steps": 29704, "loss": 0.43903465270996095, "lr": 9.797667979275752e-06, "epoch": 0.18179369781847562, "percentage": 9.09, "elapsed_time": "5:46:25", "remaining_time": "2 days, 9:44:49", "throughput": 5697.21, "total_tokens": 118421416}
+{"current_steps": 2705, "total_steps": 29704, "loss": 0.3816461801528931, "lr": 9.796922750355584e-06, "epoch": 0.18213035281443576, "percentage": 9.11, "elapsed_time": "5:47:03", "remaining_time": "2 days, 9:44:05", "throughput": 5697.09, "total_tokens": 118635408}
+{"current_steps": 2710, "total_steps": 29704, "loss": 0.4243772029876709, "lr": 9.796176179991266e-06, "epoch": 0.1824670078103959, "percentage": 9.12, "elapsed_time": "5:47:41", "remaining_time": "2 days, 9:43:14", "throughput": 5697.38, "total_tokens": 118853520}
+{"current_steps": 2715, "total_steps": 29704, "loss": 0.39929049015045165, "lr": 9.795428268391572e-06, "epoch": 0.18280366280635604, "percentage": 9.14, "elapsed_time": "5:48:19", "remaining_time": "2 days, 9:42:32", "throughput": 5696.74, "total_tokens": 119057464}
+{"current_steps": 2720, "total_steps": 29704, "loss": 0.4025382995605469, "lr": 9.794679015765655e-06, "epoch": 0.18314031780231618, "percentage": 9.16, "elapsed_time": "5:48:56", "remaining_time": "2 days, 9:41:41", "throughput": 5696.79, "total_tokens": 119270504}
+{"current_steps": 2725, "total_steps": 29704, "loss": 0.40293426513671876, "lr": 9.793928422323041e-06, "epoch": 0.18347697279827632, "percentage": 9.17, "elapsed_time": "5:49:42", "remaining_time": "2 days, 9:42:15", "throughput": 5695.05, "total_tokens": 119494864}
+{"current_steps": 2730, "total_steps": 29704, "loss": 0.40497751235961915, "lr": 9.793176488273629e-06, "epoch": 0.18381362779423646, "percentage": 9.19, "elapsed_time": "5:50:18", "remaining_time": "2 days, 9:41:13", "throughput": 5695.03, "total_tokens": 119700000}
+{"current_steps": 2735, "total_steps": 29704, "loss": 0.3891148567199707, "lr": 9.792423213827697e-06, "epoch": 0.1841502827901966, "percentage": 9.21, "elapsed_time": "5:50:57", "remaining_time": "2 days, 9:40:45", "throughput": 5694.63, "total_tokens": 119916696}
+{"current_steps": 2740, "total_steps": 29704, "loss": 0.3831428289413452, "lr": 9.791668599195897e-06, "epoch": 0.18448693778615674, "percentage": 9.22, "elapsed_time": "5:51:41", "remaining_time": "2 days, 9:40:58", "throughput": 5693.49, "total_tokens": 120142080}
+{"current_steps": 2745, "total_steps": 29704, "loss": 0.42138309478759767, "lr": 9.790912644589248e-06, "epoch": 0.18482359278211688, "percentage": 9.24, "elapsed_time": "5:52:26", "remaining_time": "2 days, 9:41:17", "throughput": 5691.31, "total_tokens": 120348496}
+{"current_steps": 2750, "total_steps": 29704, "loss": 0.3965245246887207, "lr": 9.790155350219159e-06, "epoch": 0.18516024777807702, "percentage": 9.26, "elapsed_time": "5:53:06", "remaining_time": "2 days, 9:40:59", "throughput": 5690.86, "total_tokens": 120570160}
+{"current_steps": 2755, "total_steps": 29704, "loss": 0.4091955661773682, "lr": 9.789396716297399e-06, "epoch": 0.18549690277403716, "percentage": 9.27, "elapsed_time": "5:53:40", "remaining_time": "2 days, 9:39:31", "throughput": 5692.76, "total_tokens": 120800688}
+{"current_steps": 2760, "total_steps": 29704, "loss": 0.38339416980743407, "lr": 9.788636743036122e-06, "epoch": 0.1858335577699973, "percentage": 9.29, "elapsed_time": "5:54:17", "remaining_time": "2 days, 9:38:45", "throughput": 5692.6, "total_tokens": 121012328}
+{"current_steps": 2765, "total_steps": 29704, "loss": 0.3950460195541382, "lr": 9.787875430647847e-06, "epoch": 0.18617021276595744, "percentage": 9.31, "elapsed_time": "5:54:54", "remaining_time": "2 days, 9:37:49", "throughput": 5693.11, "total_tokens": 121232032}
+{"current_steps": 2770, "total_steps": 29704, "loss": 0.39361560344696045, "lr": 9.787112779345477e-06, "epoch": 0.18650686776191758, "percentage": 9.33, "elapsed_time": "5:55:31", "remaining_time": "2 days, 9:36:56", "throughput": 5692.8, "total_tokens": 121436200}
+{"current_steps": 2775, "total_steps": 29704, "loss": 0.4345618724822998, "lr": 9.786348789342283e-06, "epoch": 0.18684352275787772, "percentage": 9.34, "elapsed_time": "5:56:10", "remaining_time": "2 days, 9:36:20", "throughput": 5692.78, "total_tokens": 121656544}
+{"current_steps": 2780, "total_steps": 29704, "loss": 0.39746851921081544, "lr": 9.785583460851912e-06, "epoch": 0.18718017775383786, "percentage": 9.36, "elapsed_time": "5:56:52", "remaining_time": "2 days, 9:36:18", "throughput": 5692.06, "total_tokens": 121881368}
+{"current_steps": 2785, "total_steps": 29704, "loss": 0.4182771682739258, "lr": 9.784816794088386e-06, "epoch": 0.187516832749798, "percentage": 9.38, "elapsed_time": "5:57:30", "remaining_time": "2 days, 9:35:32", "throughput": 5692.99, "total_tokens": 122116272}
+{"current_steps": 2790, "total_steps": 29704, "loss": 0.4000124931335449, "lr": 9.784048789266101e-06, "epoch": 0.18785348774575814, "percentage": 9.39, "elapsed_time": "5:58:13", "remaining_time": "2 days, 9:35:42", "throughput": 5691.11, "total_tokens": 122324024}
+{"current_steps": 2795, "total_steps": 29704, "loss": 0.411564302444458, "lr": 9.78327944659983e-06, "epoch": 0.18819014274171827, "percentage": 9.41, "elapsed_time": "5:59:01", "remaining_time": "2 days, 9:36:32", "throughput": 5688.52, "total_tokens": 122539808}
+{"current_steps": 2800, "total_steps": 29704, "loss": 0.40423192977905276, "lr": 9.78250876630471e-06, "epoch": 0.18852679773767841, "percentage": 9.43, "elapsed_time": "5:59:39", "remaining_time": "2 days, 9:35:51", "throughput": 5688.06, "total_tokens": 122747320}
+{"current_steps": 2805, "total_steps": 29704, "loss": 0.39595718383789064, "lr": 9.781736748596267e-06, "epoch": 0.18886345273363855, "percentage": 9.44, "elapsed_time": "6:00:16", "remaining_time": "2 days, 9:34:55", "throughput": 5688.14, "total_tokens": 122957984}
+{"current_steps": 2810, "total_steps": 29704, "loss": 0.3889493942260742, "lr": 9.78096339369039e-06, "epoch": 0.18920010772959872, "percentage": 9.46, "elapsed_time": "6:00:51", "remaining_time": "2 days, 9:33:43", "throughput": 5688.61, "total_tokens": 123167712}
+{"current_steps": 2815, "total_steps": 29704, "loss": 0.42559328079223635, "lr": 9.780188701803345e-06, "epoch": 0.18953676272555886, "percentage": 9.48, "elapsed_time": "6:01:34", "remaining_time": "2 days, 9:33:43", "throughput": 5687.55, "total_tokens": 123386632}
+{"current_steps": 2820, "total_steps": 29704, "loss": 0.40504965782165525, "lr": 9.779412673151772e-06, "epoch": 0.189873417721519, "percentage": 9.49, "elapsed_time": "6:02:15", "remaining_time": "2 days, 9:33:32", "throughput": 5687.21, "total_tokens": 123614712}
+{"current_steps": 2825, "total_steps": 29704, "loss": 0.4043741703033447, "lr": 9.778635307952686e-06, "epoch": 0.19021007271747914, "percentage": 9.51, "elapsed_time": "6:02:57", "remaining_time": "2 days, 9:33:30", "throughput": 5686.47, "total_tokens": 123839728}
+{"current_steps": 2830, "total_steps": 29704, "loss": 0.403910493850708, "lr": 9.777856606423472e-06, "epoch": 0.19054672771343928, "percentage": 9.53, "elapsed_time": "6:03:37", "remaining_time": "2 days, 9:33:04", "throughput": 5686.41, "total_tokens": 124065288}
+{"current_steps": 2835, "total_steps": 29704, "loss": 0.43606057167053225, "lr": 9.777076568781897e-06, "epoch": 0.19088338270939942, "percentage": 9.54, "elapsed_time": "6:04:15", "remaining_time": "2 days, 9:32:16", "throughput": 5685.77, "total_tokens": 124264360}
+{"current_steps": 2840, "total_steps": 29704, "loss": 0.3660429954528809, "lr": 9.77629519524609e-06, "epoch": 0.19122003770535956, "percentage": 9.56, "elapsed_time": "6:04:54", "remaining_time": "2 days, 9:31:47", "throughput": 5685.31, "total_tokens": 124479288}
+{"current_steps": 2845, "total_steps": 29704, "loss": 0.4081143379211426, "lr": 9.775512486034564e-06, "epoch": 0.1915566927013197, "percentage": 9.58, "elapsed_time": "6:05:38", "remaining_time": "2 days, 9:31:55", "throughput": 5683.77, "total_tokens": 124692848}
+{"current_steps": 2850, "total_steps": 29704, "loss": 0.4261101245880127, "lr": 9.774728441366197e-06, "epoch": 0.19189334769727984, "percentage": 9.59, "elapsed_time": "6:06:15", "remaining_time": "2 days, 9:31:01", "throughput": 5683.63, "total_tokens": 124899520}
+{"current_steps": 2855, "total_steps": 29704, "loss": 0.39615979194641116, "lr": 9.773943061460249e-06, "epoch": 0.19223000269323998, "percentage": 9.61, "elapsed_time": "6:06:55", "remaining_time": "2 days, 9:30:34", "throughput": 5682.9, "total_tokens": 125109776}
+{"current_steps": 2860, "total_steps": 29704, "loss": 0.42858476638793946, "lr": 9.773156346536345e-06, "epoch": 0.19256665768920012, "percentage": 9.63, "elapsed_time": "6:07:34", "remaining_time": "2 days, 9:30:05", "throughput": 5682.55, "total_tokens": 125326624}
+{"current_steps": 2865, "total_steps": 29704, "loss": 0.3920739650726318, "lr": 9.772368296814489e-06, "epoch": 0.19290331268516026, "percentage": 9.65, "elapsed_time": "6:08:08", "remaining_time": "2 days, 9:28:41", "throughput": 5683.78, "total_tokens": 125545392}
+{"current_steps": 2870, "total_steps": 29704, "loss": 0.439731502532959, "lr": 9.771578912515058e-06, "epoch": 0.1932399676811204, "percentage": 9.66, "elapsed_time": "6:08:43", "remaining_time": "2 days, 9:27:34", "throughput": 5684.52, "total_tokens": 125763512}
+{"current_steps": 2875, "total_steps": 29704, "loss": 0.3899986743927002, "lr": 9.7707881938588e-06, "epoch": 0.19357662267708053, "percentage": 9.68, "elapsed_time": "6:09:26", "remaining_time": "2 days, 9:27:32", "throughput": 5683.53, "total_tokens": 125983312}
+{"current_steps": 2880, "total_steps": 29704, "loss": 0.4218841552734375, "lr": 9.769996141066836e-06, "epoch": 0.19391327767304067, "percentage": 9.7, "elapsed_time": "6:09:56", "remaining_time": "2 days, 9:25:36", "throughput": 5686.52, "total_tokens": 126221208}
+{"current_steps": 2885, "total_steps": 29704, "loss": 0.41857566833496096, "lr": 9.76920275436066e-06, "epoch": 0.1942499326690008, "percentage": 9.71, "elapsed_time": "6:10:32", "remaining_time": "2 days, 9:24:34", "throughput": 5688.03, "total_tokens": 126460024}
+{"current_steps": 2890, "total_steps": 29704, "loss": 0.4293510437011719, "lr": 9.768408033962145e-06, "epoch": 0.19458658766496095, "percentage": 9.73, "elapsed_time": "6:11:14", "remaining_time": "2 days, 9:24:31", "throughput": 5687.23, "total_tokens": 126682488}
+{"current_steps": 2895, "total_steps": 29704, "loss": 0.40836372375488283, "lr": 9.767611980093525e-06, "epoch": 0.1949232426609211, "percentage": 9.75, "elapsed_time": "6:11:55", "remaining_time": "2 days, 9:24:07", "throughput": 5686.9, "total_tokens": 126903552}
+{"current_steps": 2900, "total_steps": 29704, "loss": 0.41483469009399415, "lr": 9.76681459297742e-06, "epoch": 0.19525989765688123, "percentage": 9.76, "elapsed_time": "6:12:41", "remaining_time": "2 days, 9:24:38", "throughput": 5685.07, "total_tokens": 127124488}
+{"current_steps": 2905, "total_steps": 29704, "loss": 0.4586020469665527, "lr": 9.766015872836814e-06, "epoch": 0.19559655265284137, "percentage": 9.78, "elapsed_time": "6:13:23", "remaining_time": "2 days, 9:24:37", "throughput": 5684.37, "total_tokens": 127351272}
+{"current_steps": 2910, "total_steps": 29704, "loss": 0.4265925407409668, "lr": 9.765215819895066e-06, "epoch": 0.1959332076488015, "percentage": 9.8, "elapsed_time": "6:13:55", "remaining_time": "2 days, 9:22:56", "throughput": 5686.72, "total_tokens": 127584848}
+{"current_steps": 2915, "total_steps": 29704, "loss": 0.3775201320648193, "lr": 9.76441443437591e-06, "epoch": 0.19626986264476165, "percentage": 9.81, "elapsed_time": "6:14:28", "remaining_time": "2 days, 9:21:25", "throughput": 5688.69, "total_tokens": 127815392}
+{"current_steps": 2920, "total_steps": 29704, "loss": 0.4249995708465576, "lr": 9.76361171650345e-06, "epoch": 0.1966065176407218, "percentage": 9.83, "elapsed_time": "6:15:08", "remaining_time": "2 days, 9:20:58", "throughput": 5687.95, "total_tokens": 128025072}
+{"current_steps": 2925, "total_steps": 29704, "loss": 0.41265301704406737, "lr": 9.762807666502164e-06, "epoch": 0.19694317263668193, "percentage": 9.85, "elapsed_time": "6:15:48", "remaining_time": "2 days, 9:20:32", "throughput": 5687.25, "total_tokens": 128236720}
+{"current_steps": 2930, "total_steps": 29704, "loss": 0.4183157444000244, "lr": 9.762002284596898e-06, "epoch": 0.19727982763264207, "percentage": 9.86, "elapsed_time": "6:16:31", "remaining_time": "2 days, 9:20:38", "throughput": 5685.69, "total_tokens": 128448432}
+{"current_steps": 2935, "total_steps": 29704, "loss": 0.4257832050323486, "lr": 9.761195571012882e-06, "epoch": 0.1976164826286022, "percentage": 9.88, "elapsed_time": "6:17:09", "remaining_time": "2 days, 9:19:51", "throughput": 5685.51, "total_tokens": 128657896}
+{"current_steps": 2940, "total_steps": 29704, "loss": 0.395475172996521, "lr": 9.760387525975705e-06, "epoch": 0.19795313762456235, "percentage": 9.9, "elapsed_time": "6:17:42", "remaining_time": "2 days, 9:18:28", "throughput": 5687.01, "total_tokens": 128883360}
+{"current_steps": 2945, "total_steps": 29704, "loss": 0.35533039569854735, "lr": 9.759578149711337e-06, "epoch": 0.1982897926205225, "percentage": 9.91, "elapsed_time": "6:18:23", "remaining_time": "2 days, 9:18:06", "throughput": 5686.98, "total_tokens": 129112664}
+{"current_steps": 2950, "total_steps": 29704, "loss": 0.38588006496429444, "lr": 9.758767442446117e-06, "epoch": 0.19862644761648263, "percentage": 9.93, "elapsed_time": "6:19:08", "remaining_time": "2 days, 9:18:27", "throughput": 5685.82, "total_tokens": 129342864}
+{"current_steps": 2955, "total_steps": 29704, "loss": 0.38382346630096437, "lr": 9.757955404406755e-06, "epoch": 0.19896310261244277, "percentage": 9.95, "elapsed_time": "6:19:46", "remaining_time": "2 days, 9:17:44", "throughput": 5686.12, "total_tokens": 129566032}
+{"current_steps": 2960, "total_steps": 29704, "loss": 0.3917106628417969, "lr": 9.757142035820337e-06, "epoch": 0.1992997576084029, "percentage": 9.96, "elapsed_time": "6:20:24", "remaining_time": "2 days, 9:17:06", "throughput": 5685.27, "total_tokens": 129766056}
+{"current_steps": 2965, "total_steps": 29704, "loss": 0.4134823322296143, "lr": 9.756327336914318e-06, "epoch": 0.19963641260436304, "percentage": 9.98, "elapsed_time": "6:21:08", "remaining_time": "2 days, 9:17:11", "throughput": 5683.73, "total_tokens": 129977048}
+{"current_steps": 2970, "total_steps": 29704, "loss": 0.4055294990539551, "lr": 9.755511307916526e-06, "epoch": 0.19997306760032318, "percentage": 10.0, "elapsed_time": "6:21:41", "remaining_time": "2 days, 9:15:43", "throughput": 5685.0, "total_tokens": 130194224}
+{"current_steps": 2975, "total_steps": 29704, "loss": 0.4119466781616211, "lr": 9.754693949055161e-06, "epoch": 0.20030972259628332, "percentage": 10.02, "elapsed_time": "6:22:21", "remaining_time": "2 days, 9:15:16", "throughput": 5685.1, "total_tokens": 130423144}
+{"current_steps": 2980, "total_steps": 29704, "loss": 0.43303780555725097, "lr": 9.753875260558795e-06, "epoch": 0.20064637759224346, "percentage": 10.03, "elapsed_time": "6:22:57", "remaining_time": "2 days, 9:14:13", "throughput": 5684.92, "total_tokens": 130623168}
+{"current_steps": 2985, "total_steps": 29704, "loss": 0.39323015213012696, "lr": 9.75305524265637e-06, "epoch": 0.2009830325882036, "percentage": 10.05, "elapsed_time": "6:23:30", "remaining_time": "2 days, 9:12:52", "throughput": 5686.6, "total_tokens": 130853536}
+{"current_steps": 2990, "total_steps": 29704, "loss": 0.4042644023895264, "lr": 9.752233895577205e-06, "epoch": 0.20131968758416374, "percentage": 10.07, "elapsed_time": "6:24:07", "remaining_time": "2 days, 9:11:54", "throughput": 5686.56, "total_tokens": 131059592}
+{"current_steps": 2995, "total_steps": 29704, "loss": 0.40361328125, "lr": 9.751411219550984e-06, "epoch": 0.20165634258012388, "percentage": 10.08, "elapsed_time": "6:24:42", "remaining_time": "2 days, 9:10:49", "throughput": 5687.22, "total_tokens": 131276760}
+{"current_steps": 3000, "total_steps": 29704, "loss": 0.39671123027801514, "lr": 9.750587214807766e-06, "epoch": 0.20199299757608402, "percentage": 10.1, "elapsed_time": "6:25:24", "remaining_time": "2 days, 9:10:43", "throughput": 5685.89, "total_tokens": 131486248}
+{"current_steps": 3005, "total_steps": 29704, "loss": 0.42069306373596194, "lr": 9.749761881577981e-06, "epoch": 0.20232965257204416, "percentage": 10.12, "elapsed_time": "6:26:06", "remaining_time": "2 days, 9:10:29", "throughput": 5685.68, "total_tokens": 131716280}
+{"current_steps": 3010, "total_steps": 29704, "loss": 0.39777138233184817, "lr": 9.748935220092434e-06, "epoch": 0.2026663075680043, "percentage": 10.13, "elapsed_time": "6:26:42", "remaining_time": "2 days, 9:09:29", "throughput": 5686.28, "total_tokens": 131935832}
+{"current_steps": 3015, "total_steps": 29704, "loss": 0.39002056121826173, "lr": 9.748107230582295e-06, "epoch": 0.20300296256396444, "percentage": 10.15, "elapsed_time": "6:27:19", "remaining_time": "2 days, 9:08:40", "throughput": 5685.93, "total_tokens": 132140024}
+{"current_steps": 3020, "total_steps": 29704, "loss": 0.4608659267425537, "lr": 9.747277913279109e-06, "epoch": 0.20333961755992458, "percentage": 10.17, "elapsed_time": "6:28:05", "remaining_time": "2 days, 9:09:05", "throughput": 5684.15, "total_tokens": 132358504}
+{"current_steps": 3025, "total_steps": 29704, "loss": 0.44081912040710447, "lr": 9.746447268414794e-06, "epoch": 0.20367627255588472, "percentage": 10.18, "elapsed_time": "6:28:44", "remaining_time": "2 days, 9:08:29", "throughput": 5684.6, "total_tokens": 132589576}
+{"current_steps": 3030, "total_steps": 29704, "loss": 0.40747461318969724, "lr": 9.745615296221635e-06, "epoch": 0.20401292755184486, "percentage": 10.2, "elapsed_time": "6:29:18", "remaining_time": "2 days, 9:07:14", "throughput": 5686.05, "total_tokens": 132819496}
+{"current_steps": 3035, "total_steps": 29704, "loss": 0.42429242134094236, "lr": 9.74478199693229e-06, "epoch": 0.204349582547805, "percentage": 10.22, "elapsed_time": "6:30:01", "remaining_time": "2 days, 9:07:16", "throughput": 5685.24, "total_tokens": 133045880}
+{"current_steps": 3040, "total_steps": 29704, "loss": 0.4262086391448975, "lr": 9.743947370779791e-06, "epoch": 0.20468623754376514, "percentage": 10.23, "elapsed_time": "6:30:42", "remaining_time": "2 days, 9:06:56", "throughput": 5685.11, "total_tokens": 133273848}
+{"current_steps": 3045, "total_steps": 29704, "loss": 0.41955132484436036, "lr": 9.743111417997536e-06, "epoch": 0.20502289253972528, "percentage": 10.25, "elapsed_time": "6:31:21", "remaining_time": "2 days, 9:06:23", "throughput": 5685.28, "total_tokens": 133500640}
+{"current_steps": 3050, "total_steps": 29704, "loss": 0.416639518737793, "lr": 9.742274138819298e-06, "epoch": 0.20535954753568542, "percentage": 10.27, "elapsed_time": "6:31:58", "remaining_time": "2 days, 9:05:29", "throughput": 5685.72, "total_tokens": 133719848}
+{"current_steps": 3055, "total_steps": 29704, "loss": 0.45115389823913576, "lr": 9.74143553347922e-06, "epoch": 0.20569620253164558, "percentage": 10.28, "elapsed_time": "6:32:38", "remaining_time": "2 days, 9:04:58", "throughput": 5685.72, "total_tokens": 133944240}
+{"current_steps": 3060, "total_steps": 29704, "loss": 0.395220947265625, "lr": 9.740595602211814e-06, "epoch": 0.20603285752760572, "percentage": 10.3, "elapsed_time": "6:33:14", "remaining_time": "2 days, 9:03:58", "throughput": 5686.46, "total_tokens": 134166880}
+{"current_steps": 3065, "total_steps": 29704, "loss": 0.3915586471557617, "lr": 9.739754345251963e-06, "epoch": 0.20636951252356586, "percentage": 10.32, "elapsed_time": "6:33:53", "remaining_time": "2 days, 9:03:26", "throughput": 5685.73, "total_tokens": 134373184}
+{"current_steps": 3070, "total_steps": 29704, "loss": 0.36250555515289307, "lr": 9.738911762834924e-06, "epoch": 0.206706167519526, "percentage": 10.34, "elapsed_time": "6:34:29", "remaining_time": "2 days, 9:02:29", "throughput": 5686.34, "total_tokens": 134594840}
+{"current_steps": 3075, "total_steps": 29704, "loss": 0.40388903617858884, "lr": 9.738067855196323e-06, "epoch": 0.20704282251548614, "percentage": 10.35, "elapsed_time": "6:35:02", "remaining_time": "2 days, 9:00:59", "throughput": 5687.76, "total_tokens": 134814144}
+{"current_steps": 3080, "total_steps": 29704, "loss": 0.4489736557006836, "lr": 9.737222622572153e-06, "epoch": 0.20737947751144628, "percentage": 10.37, "elapsed_time": "6:35:47", "remaining_time": "2 days, 9:01:17", "throughput": 5685.85, "total_tokens": 135025088}
+{"current_steps": 3085, "total_steps": 29704, "loss": 0.4101541519165039, "lr": 9.736376065198784e-06, "epoch": 0.20771613250740642, "percentage": 10.39, "elapsed_time": "6:36:26", "remaining_time": "2 days, 9:00:43", "throughput": 5686.1, "total_tokens": 135253064}
+{"current_steps": 3090, "total_steps": 29704, "loss": 0.39399616718292235, "lr": 9.735528183312951e-06, "epoch": 0.20805278750336656, "percentage": 10.4, "elapsed_time": "6:37:07", "remaining_time": "2 days, 9:00:28", "throughput": 5686.06, "total_tokens": 135486752}
+{"current_steps": 3095, "total_steps": 29704, "loss": 0.3820789098739624, "lr": 9.734678977151762e-06, "epoch": 0.2083894424993267, "percentage": 10.42, "elapsed_time": "6:37:48", "remaining_time": "2 days, 9:00:10", "throughput": 5685.35, "total_tokens": 135702632}
+{"current_steps": 3100, "total_steps": 29704, "loss": 0.40059328079223633, "lr": 9.733828446952696e-06, "epoch": 0.20872609749528684, "percentage": 10.44, "elapsed_time": "6:38:24", "remaining_time": "2 days, 8:59:09", "throughput": 5686.13, "total_tokens": 135925976}
+{"current_steps": 3105, "total_steps": 29704, "loss": 0.3992907524108887, "lr": 9.732976592953598e-06, "epoch": 0.20906275249124698, "percentage": 10.45, "elapsed_time": "6:39:07", "remaining_time": "2 days, 8:59:02", "throughput": 5685.68, "total_tokens": 136155256}
+{"current_steps": 3110, "total_steps": 29704, "loss": 0.40617761611938474, "lr": 9.73212341539269e-06, "epoch": 0.20939940748720712, "percentage": 10.47, "elapsed_time": "6:39:55", "remaining_time": "2 days, 8:59:51", "throughput": 5683.4, "total_tokens": 136377968}
+{"current_steps": 3115, "total_steps": 29704, "loss": 0.39726805686950684, "lr": 9.731268914508556e-06, "epoch": 0.20973606248316726, "percentage": 10.49, "elapsed_time": "6:40:37", "remaining_time": "2 days, 8:59:42", "throughput": 5682.5, "total_tokens": 136595144}
+{"current_steps": 3120, "total_steps": 29704, "loss": 0.3819112777709961, "lr": 9.730413090540158e-06, "epoch": 0.2100727174791274, "percentage": 10.5, "elapsed_time": "6:41:15", "remaining_time": "2 days, 8:58:59", "throughput": 5682.77, "total_tokens": 136818256}
+{"current_steps": 3125, "total_steps": 29704, "loss": 0.40000243186950685, "lr": 9.729555943726825e-06, "epoch": 0.21040937247508754, "percentage": 10.52, "elapsed_time": "6:42:00", "remaining_time": "2 days, 8:59:10", "throughput": 5681.18, "total_tokens": 137031872}
+{"current_steps": 3130, "total_steps": 29704, "loss": 0.40857748985290526, "lr": 9.728697474308253e-06, "epoch": 0.21074602747104768, "percentage": 10.54, "elapsed_time": "6:42:39", "remaining_time": "2 days, 8:58:38", "throughput": 5680.79, "total_tokens": 137246200}
+{"current_steps": 3135, "total_steps": 29704, "loss": 0.38510382175445557, "lr": 9.72783768252451e-06, "epoch": 0.21108268246700782, "percentage": 10.55, "elapsed_time": "6:43:17", "remaining_time": "2 days, 8:57:53", "throughput": 5680.48, "total_tokens": 137454056}
+{"current_steps": 3140, "total_steps": 29704, "loss": 0.40147695541381834, "lr": 9.726976568616036e-06, "epoch": 0.21141933746296795, "percentage": 10.57, "elapsed_time": "6:44:01", "remaining_time": "2 days, 8:57:59", "throughput": 5678.87, "total_tokens": 137663648}
+{"current_steps": 3145, "total_steps": 29704, "loss": 0.389466381072998, "lr": 9.726114132823638e-06, "epoch": 0.2117559924589281, "percentage": 10.59, "elapsed_time": "6:44:40", "remaining_time": "2 days, 8:57:22", "throughput": 5678.79, "total_tokens": 137882432}
+{"current_steps": 3150, "total_steps": 29704, "loss": 0.4213107109069824, "lr": 9.725250375388493e-06, "epoch": 0.21209264745488823, "percentage": 10.6, "elapsed_time": "6:45:20", "remaining_time": "2 days, 8:56:57", "throughput": 5678.09, "total_tokens": 138093344}
+{"current_steps": 3155, "total_steps": 29704, "loss": 0.43596372604370115, "lr": 9.724385296552147e-06, "epoch": 0.21242930245084837, "percentage": 10.62, "elapsed_time": "6:46:02", "remaining_time": "2 days, 8:56:44", "throughput": 5677.53, "total_tokens": 138316336}
+{"current_steps": 3160, "total_steps": 29704, "loss": 0.4109373092651367, "lr": 9.72351889655652e-06, "epoch": 0.2127659574468085, "percentage": 10.64, "elapsed_time": "6:46:37", "remaining_time": "2 days, 8:55:37", "throughput": 5678.58, "total_tokens": 138541888}
+{"current_steps": 3165, "total_steps": 29704, "loss": 0.40399460792541503, "lr": 9.722651175643895e-06, "epoch": 0.21310261244276865, "percentage": 10.66, "elapsed_time": "6:47:16", "remaining_time": "2 days, 8:55:05", "throughput": 5678.39, "total_tokens": 138760928}
+{"current_steps": 3170, "total_steps": 29704, "loss": 0.36252574920654296, "lr": 9.721782134056928e-06, "epoch": 0.2134392674387288, "percentage": 10.67, "elapsed_time": "6:48:00", "remaining_time": "2 days, 8:55:09", "throughput": 5677.29, "total_tokens": 138982088}
+{"current_steps": 3175, "total_steps": 29704, "loss": 0.4204305648803711, "lr": 9.720911772038645e-06, "epoch": 0.21377592243468893, "percentage": 10.69, "elapsed_time": "6:48:34", "remaining_time": "2 days, 8:53:55", "throughput": 5678.58, "total_tokens": 139208968}
+{"current_steps": 3180, "total_steps": 29704, "loss": 0.4118339538574219, "lr": 9.720040089832436e-06, "epoch": 0.21411257743064907, "percentage": 10.71, "elapsed_time": "6:49:15", "remaining_time": "2 days, 8:53:32", "throughput": 5677.17, "total_tokens": 139404296}
+{"current_steps": 3185, "total_steps": 29704, "loss": 0.38251235485076907, "lr": 9.719167087682069e-06, "epoch": 0.2144492324266092, "percentage": 10.72, "elapsed_time": "6:49:56", "remaining_time": "2 days, 8:53:12", "throughput": 5676.78, "total_tokens": 139626528}
+{"current_steps": 3190, "total_steps": 29704, "loss": 0.39778890609741213, "lr": 9.718292765831673e-06, "epoch": 0.21478588742256935, "percentage": 10.74, "elapsed_time": "6:50:36", "remaining_time": "2 days, 8:52:49", "throughput": 5676.6, "total_tokens": 139852040}
+{"current_steps": 3195, "total_steps": 29704, "loss": 0.4149151802062988, "lr": 9.71741712452575e-06, "epoch": 0.2151225424185295, "percentage": 10.76, "elapsed_time": "6:51:09", "remaining_time": "2 days, 8:51:20", "throughput": 5678.12, "total_tokens": 140074344}
+{"current_steps": 3200, "total_steps": 29704, "loss": 0.3865767478942871, "lr": 9.716540164009174e-06, "epoch": 0.21545919741448963, "percentage": 10.77, "elapsed_time": "6:51:47", "remaining_time": "2 days, 8:50:38", "throughput": 5677.58, "total_tokens": 140277680}
+{"current_steps": 3205, "total_steps": 29704, "loss": 0.4173252582550049, "lr": 9.715661884527178e-06, "epoch": 0.21579585241044977, "percentage": 10.79, "elapsed_time": "6:52:30", "remaining_time": "2 days, 8:50:34", "throughput": 5676.91, "total_tokens": 140503936}
+{"current_steps": 3210, "total_steps": 29704, "loss": 0.4102613925933838, "lr": 9.714782286325374e-06, "epoch": 0.2161325074064099, "percentage": 10.81, "elapsed_time": "6:53:13", "remaining_time": "2 days, 8:50:32", "throughput": 5675.89, "total_tokens": 140723000}
+{"current_steps": 3215, "total_steps": 29704, "loss": 0.3924571990966797, "lr": 9.713901369649738e-06, "epoch": 0.21646916240237005, "percentage": 10.82, "elapsed_time": "6:53:51", "remaining_time": "2 days, 8:49:51", "throughput": 5675.98, "total_tokens": 140943200}
+{"current_steps": 3220, "total_steps": 29704, "loss": 0.38917341232299807, "lr": 9.713019134746618e-06, "epoch": 0.2168058173983302, "percentage": 10.84, "elapsed_time": "6:54:27", "remaining_time": "2 days, 8:48:51", "throughput": 5677.24, "total_tokens": 141178808}
+{"current_steps": 3225, "total_steps": 29704, "loss": 0.39778847694396974, "lr": 9.712135581862722e-06, "epoch": 0.21714247239429033, "percentage": 10.86, "elapsed_time": "6:55:09", "remaining_time": "2 days, 8:48:37", "throughput": 5677.05, "total_tokens": 141410600}
+{"current_steps": 3230, "total_steps": 29704, "loss": 0.41150903701782227, "lr": 9.711250711245139e-06, "epoch": 0.21747912739025047, "percentage": 10.87, "elapsed_time": "6:55:47", "remaining_time": "2 days, 8:47:55", "throughput": 5676.87, "total_tokens": 141622744}
+{"current_steps": 3235, "total_steps": 29704, "loss": 0.4173532009124756, "lr": 9.710364523141317e-06, "epoch": 0.2178157823862106, "percentage": 10.89, "elapsed_time": "6:56:30", "remaining_time": "2 days, 8:47:57", "throughput": 5675.86, "total_tokens": 141845232}
+{"current_steps": 3240, "total_steps": 29704, "loss": 0.4302227973937988, "lr": 9.709477017799076e-06, "epoch": 0.21815243738217074, "percentage": 10.91, "elapsed_time": "6:57:02", "remaining_time": "2 days, 8:46:23", "throughput": 5677.59, "total_tokens": 142068920}
+{"current_steps": 3245, "total_steps": 29704, "loss": 0.4126433849334717, "lr": 9.708588195466604e-06, "epoch": 0.21848909237813088, "percentage": 10.92, "elapsed_time": "6:57:34", "remaining_time": "2 days, 8:44:52", "throughput": 5679.09, "total_tokens": 142289600}
+{"current_steps": 3250, "total_steps": 29704, "loss": 0.4167231559753418, "lr": 9.707698056392458e-06, "epoch": 0.21882574737409102, "percentage": 10.94, "elapsed_time": "6:58:10", "remaining_time": "2 days, 8:43:47", "throughput": 5679.74, "total_tokens": 142506920}
+{"current_steps": 3255, "total_steps": 29704, "loss": 0.4131289005279541, "lr": 9.706806600825562e-06, "epoch": 0.21916240237005116, "percentage": 10.96, "elapsed_time": "6:58:53", "remaining_time": "2 days, 8:43:48", "throughput": 5678.41, "total_tokens": 142719992}
+{"current_steps": 3260, "total_steps": 29704, "loss": 0.42434329986572267, "lr": 9.70591382901521e-06, "epoch": 0.2194990573660113, "percentage": 10.97, "elapsed_time": "6:59:30", "remaining_time": "2 days, 8:42:51", "throughput": 5678.77, "total_tokens": 142935376}
+{"current_steps": 3265, "total_steps": 29704, "loss": 0.39461374282836914, "lr": 9.705019741211061e-06, "epoch": 0.21983571236197144, "percentage": 10.99, "elapsed_time": "7:00:08", "remaining_time": "2 days, 8:42:13", "throughput": 5678.79, "total_tokens": 143155712}
+{"current_steps": 3270, "total_steps": 29704, "loss": 0.38963048458099364, "lr": 9.704124337663145e-06, "epoch": 0.22017236735793158, "percentage": 11.01, "elapsed_time": "7:00:43", "remaining_time": "2 days, 8:41:02", "throughput": 5679.97, "total_tokens": 143381832}
+{"current_steps": 3275, "total_steps": 29704, "loss": 0.4044978618621826, "lr": 9.703227618621859e-06, "epoch": 0.22050902235389172, "percentage": 11.03, "elapsed_time": "7:01:24", "remaining_time": "2 days, 8:40:40", "throughput": 5679.77, "total_tokens": 143607328}
+{"current_steps": 3280, "total_steps": 29704, "loss": 0.41426868438720704, "lr": 9.702329584337965e-06, "epoch": 0.22084567734985186, "percentage": 11.04, "elapsed_time": "7:02:06", "remaining_time": "2 days, 8:40:33", "throughput": 5678.17, "total_tokens": 143809000}
+{"current_steps": 3285, "total_steps": 29704, "loss": 0.43207788467407227, "lr": 9.701430235062599e-06, "epoch": 0.221182332345812, "percentage": 11.06, "elapsed_time": "7:02:46", "remaining_time": "2 days, 8:40:02", "throughput": 5677.8, "total_tokens": 144023696}
+{"current_steps": 3290, "total_steps": 29704, "loss": 0.4370114803314209, "lr": 9.700529571047259e-06, "epoch": 0.22151898734177214, "percentage": 11.08, "elapsed_time": "7:03:23", "remaining_time": "2 days, 8:39:15", "throughput": 5677.78, "total_tokens": 144236616}
+{"current_steps": 3295, "total_steps": 29704, "loss": 0.3935758829116821, "lr": 9.699627592543814e-06, "epoch": 0.22185564233773228, "percentage": 11.09, "elapsed_time": "7:04:09", "remaining_time": "2 days, 8:39:34", "throughput": 5676.0, "total_tokens": 144451272}
+{"current_steps": 3300, "total_steps": 29704, "loss": 0.41451497077941896, "lr": 9.698724299804498e-06, "epoch": 0.22219229733369245, "percentage": 11.11, "elapsed_time": "7:04:49", "remaining_time": "2 days, 8:39:09", "throughput": 5675.79, "total_tokens": 144675240}
+{"current_steps": 3305, "total_steps": 29704, "loss": 0.40555124282836913, "lr": 9.697819693081915e-06, "epoch": 0.22252895232965259, "percentage": 11.13, "elapsed_time": "7:05:26", "remaining_time": "2 days, 8:38:17", "throughput": 5676.2, "total_tokens": 144894992}
+{"current_steps": 3310, "total_steps": 29704, "loss": 0.38943185806274416, "lr": 9.696913772629034e-06, "epoch": 0.22286560732561272, "percentage": 11.14, "elapsed_time": "7:06:03", "remaining_time": "2 days, 8:37:25", "throughput": 5676.77, "total_tokens": 145119152}
+{"current_steps": 3315, "total_steps": 29704, "loss": 0.42933268547058107, "lr": 9.696006538699195e-06, "epoch": 0.22320226232157286, "percentage": 11.16, "elapsed_time": "7:06:35", "remaining_time": "2 days, 8:35:52", "throughput": 5678.36, "total_tokens": 145340576}
+{"current_steps": 3320, "total_steps": 29704, "loss": 0.41726884841918943, "lr": 9.695097991546102e-06, "epoch": 0.223538917317533, "percentage": 11.18, "elapsed_time": "7:07:15", "remaining_time": "2 days, 8:35:22", "throughput": 5677.2, "total_tokens": 145536392}
+{"current_steps": 3325, "total_steps": 29704, "loss": 0.38661761283874513, "lr": 9.694188131423828e-06, "epoch": 0.22387557231349314, "percentage": 11.19, "elapsed_time": "7:07:55", "remaining_time": "2 days, 8:34:59", "throughput": 5676.67, "total_tokens": 145752592}
+{"current_steps": 3330, "total_steps": 29704, "loss": 0.4395936965942383, "lr": 9.693276958586809e-06, "epoch": 0.22421222730945328, "percentage": 11.21, "elapsed_time": "7:08:37", "remaining_time": "2 days, 8:34:47", "throughput": 5676.08, "total_tokens": 145975632}
+{"current_steps": 3335, "total_steps": 29704, "loss": 0.37718255519866944, "lr": 9.692364473289856e-06, "epoch": 0.22454888230541342, "percentage": 11.23, "elapsed_time": "7:09:18", "remaining_time": "2 days, 8:34:26", "throughput": 5675.39, "total_tokens": 146190016}
+{"current_steps": 3340, "total_steps": 29704, "loss": 0.38504073619842527, "lr": 9.69145067578814e-06, "epoch": 0.22488553730137356, "percentage": 11.24, "elapsed_time": "7:09:59", "remaining_time": "2 days, 8:34:05", "throughput": 5675.17, "total_tokens": 146415880}
+{"current_steps": 3345, "total_steps": 29704, "loss": 0.44390153884887695, "lr": 9.690535566337205e-06, "epoch": 0.2252221922973337, "percentage": 11.26, "elapsed_time": "7:10:41", "remaining_time": "2 days, 8:33:53", "throughput": 5674.49, "total_tokens": 146637056}
+{"current_steps": 3350, "total_steps": 29704, "loss": 0.4186522960662842, "lr": 9.689619145192955e-06, "epoch": 0.22555884729329384, "percentage": 11.28, "elapsed_time": "7:11:24", "remaining_time": "2 days, 8:33:49", "throughput": 5673.63, "total_tokens": 146858960}
+{"current_steps": 3355, "total_steps": 29704, "loss": 0.39981067180633545, "lr": 9.688701412611663e-06, "epoch": 0.22589550228925398, "percentage": 11.29, "elapsed_time": "7:12:04", "remaining_time": "2 days, 8:33:19", "throughput": 5673.81, "total_tokens": 147088760}
+{"current_steps": 3360, "total_steps": 29704, "loss": 0.40915794372558595, "lr": 9.687782368849973e-06, "epoch": 0.22623215728521412, "percentage": 11.31, "elapsed_time": "7:12:47", "remaining_time": "2 days, 8:33:14", "throughput": 5672.83, "total_tokens": 147306920}
+{"current_steps": 3365, "total_steps": 29704, "loss": 0.3879998207092285, "lr": 9.686862014164893e-06, "epoch": 0.22656881228117426, "percentage": 11.33, "elapsed_time": "7:13:24", "remaining_time": "2 days, 8:32:24", "throughput": 5673.44, "total_tokens": 147534144}
+{"current_steps": 3370, "total_steps": 29704, "loss": 0.3802268266677856, "lr": 9.685940348813795e-06, "epoch": 0.2269054672771344, "percentage": 11.35, "elapsed_time": "7:14:02", "remaining_time": "2 days, 8:31:42", "throughput": 5673.48, "total_tokens": 147751544}
+{"current_steps": 3375, "total_steps": 29704, "loss": 0.39569883346557616, "lr": 9.685017373054421e-06, "epoch": 0.22724212227309454, "percentage": 11.36, "elapsed_time": "7:14:41", "remaining_time": "2 days, 8:31:03", "throughput": 5673.5, "total_tokens": 147971208}
+{"current_steps": 3380, "total_steps": 29704, "loss": 0.3963681697845459, "lr": 9.68409308714488e-06, "epoch": 0.22757877726905468, "percentage": 11.38, "elapsed_time": "7:15:19", "remaining_time": "2 days, 8:30:23", "throughput": 5673.39, "total_tokens": 148186648}
+{"current_steps": 3385, "total_steps": 29704, "loss": 0.4038090705871582, "lr": 9.68316749134364e-06, "epoch": 0.22791543226501482, "percentage": 11.4, "elapsed_time": "7:16:02", "remaining_time": "2 days, 8:30:14", "throughput": 5673.02, "total_tokens": 148417752}
+{"current_steps": 3390, "total_steps": 29704, "loss": 0.38401403427124026, "lr": 9.682240585909545e-06, "epoch": 0.22825208726097496, "percentage": 11.41, "elapsed_time": "7:16:41", "remaining_time": "2 days, 8:29:45", "throughput": 5672.43, "total_tokens": 148628520}
+{"current_steps": 3395, "total_steps": 29704, "loss": 0.36235072612762453, "lr": 9.681312371101803e-06, "epoch": 0.2285887422569351, "percentage": 11.43, "elapsed_time": "7:17:15", "remaining_time": "2 days, 8:28:27", "throughput": 5674.0, "total_tokens": 148860120}
+{"current_steps": 3400, "total_steps": 29704, "loss": 0.4153308868408203, "lr": 9.680382847179984e-06, "epoch": 0.22892539725289524, "percentage": 11.45, "elapsed_time": "7:17:59", "remaining_time": "2 days, 8:28:26", "throughput": 5672.79, "total_tokens": 149075360}
+{"current_steps": 3405, "total_steps": 29704, "loss": 0.39875361919403074, "lr": 9.679452014404025e-06, "epoch": 0.22926205224885537, "percentage": 11.46, "elapsed_time": "7:18:38", "remaining_time": "2 days, 8:27:57", "throughput": 5672.13, "total_tokens": 149283904}
+{"current_steps": 3410, "total_steps": 29704, "loss": 0.435106086730957, "lr": 9.678519873034234e-06, "epoch": 0.22959870724481551, "percentage": 11.48, "elapsed_time": "7:19:17", "remaining_time": "2 days, 8:27:15", "throughput": 5672.53, "total_tokens": 149511344}
+{"current_steps": 3415, "total_steps": 29704, "loss": 0.42625818252563474, "lr": 9.677586423331277e-06, "epoch": 0.22993536224077565, "percentage": 11.5, "elapsed_time": "7:19:56", "remaining_time": "2 days, 8:26:39", "throughput": 5672.6, "total_tokens": 149734672}
+{"current_steps": 3420, "total_steps": 29704, "loss": 0.3853285312652588, "lr": 9.676651665556194e-06, "epoch": 0.2302720172367358, "percentage": 11.51, "elapsed_time": "7:20:37", "remaining_time": "2 days, 8:26:19", "throughput": 5672.58, "total_tokens": 149966984}
+{"current_steps": 3425, "total_steps": 29704, "loss": 0.39388408660888674, "lr": 9.675715599970383e-06, "epoch": 0.23060867223269593, "percentage": 11.53, "elapsed_time": "7:21:16", "remaining_time": "2 days, 8:25:48", "throughput": 5671.88, "total_tokens": 150173312}
+{"current_steps": 3430, "total_steps": 29704, "loss": 0.43226184844970705, "lr": 9.674778226835617e-06, "epoch": 0.23094532722865607, "percentage": 11.55, "elapsed_time": "7:21:54", "remaining_time": "2 days, 8:25:03", "throughput": 5671.54, "total_tokens": 150378728}
+{"current_steps": 3435, "total_steps": 29704, "loss": 0.4151242733001709, "lr": 9.673839546414026e-06, "epoch": 0.2312819822246162, "percentage": 11.56, "elapsed_time": "7:22:35", "remaining_time": "2 days, 8:24:42", "throughput": 5670.75, "total_tokens": 150589928}
+{"current_steps": 3440, "total_steps": 29704, "loss": 0.39006340503692627, "lr": 9.672899558968107e-06, "epoch": 0.23161863722057635, "percentage": 11.58, "elapsed_time": "7:23:19", "remaining_time": "2 days, 8:24:46", "throughput": 5669.32, "total_tokens": 150802664}
+{"current_steps": 3445, "total_steps": 29704, "loss": 0.44136791229248046, "lr": 9.67195826476073e-06, "epoch": 0.2319552922165365, "percentage": 11.6, "elapsed_time": "7:23:57", "remaining_time": "2 days, 8:24:01", "throughput": 5669.71, "total_tokens": 151027880}
+{"current_steps": 3450, "total_steps": 29704, "loss": 0.3967195749282837, "lr": 9.671015664055122e-06, "epoch": 0.23229194721249663, "percentage": 11.61, "elapsed_time": "7:24:31", "remaining_time": "2 days, 8:22:42", "throughput": 5670.56, "total_tokens": 151239928}
+{"current_steps": 3455, "total_steps": 29704, "loss": 0.4030034065246582, "lr": 9.670071757114877e-06, "epoch": 0.23262860220845677, "percentage": 11.63, "elapsed_time": "7:25:10", "remaining_time": "2 days, 8:22:10", "throughput": 5670.4, "total_tokens": 151459832}
+{"current_steps": 3460, "total_steps": 29704, "loss": 0.38636322021484376, "lr": 9.669126544203957e-06, "epoch": 0.2329652572044169, "percentage": 11.65, "elapsed_time": "7:25:49", "remaining_time": "2 days, 8:21:31", "throughput": 5670.01, "total_tokens": 151667896}
+{"current_steps": 3465, "total_steps": 29704, "loss": 0.4077507972717285, "lr": 9.668180025586689e-06, "epoch": 0.23330191220037705, "percentage": 11.67, "elapsed_time": "7:26:31", "remaining_time": "2 days, 8:21:19", "throughput": 5669.83, "total_tokens": 151902352}
+{"current_steps": 3470, "total_steps": 29704, "loss": 0.3916687250137329, "lr": 9.667232201527761e-06, "epoch": 0.2336385671963372, "percentage": 11.68, "elapsed_time": "7:27:07", "remaining_time": "2 days, 8:20:22", "throughput": 5669.65, "total_tokens": 152102680}
+{"current_steps": 3475, "total_steps": 29704, "loss": 0.41549954414367674, "lr": 9.666283072292234e-06, "epoch": 0.23397522219229733, "percentage": 11.7, "elapsed_time": "7:27:44", "remaining_time": "2 days, 8:19:32", "throughput": 5669.27, "total_tokens": 152303592}
+{"current_steps": 3480, "total_steps": 29704, "loss": 0.40798091888427734, "lr": 9.665332638145522e-06, "epoch": 0.23431187718825747, "percentage": 11.72, "elapsed_time": "7:28:24", "remaining_time": "2 days, 8:19:03", "throughput": 5669.14, "total_tokens": 152526320}
+{"current_steps": 3485, "total_steps": 29704, "loss": 0.44217610359191895, "lr": 9.664380899353414e-06, "epoch": 0.2346485321842176, "percentage": 11.73, "elapsed_time": "7:28:58", "remaining_time": "2 days, 8:17:50", "throughput": 5670.19, "total_tokens": 152747800}
+{"current_steps": 3490, "total_steps": 29704, "loss": 0.418949556350708, "lr": 9.663427856182062e-06, "epoch": 0.23498518718017775, "percentage": 11.75, "elapsed_time": "7:29:36", "remaining_time": "2 days, 8:17:07", "throughput": 5670.76, "total_tokens": 152978904}
+{"current_steps": 3495, "total_steps": 29704, "loss": 0.41005191802978513, "lr": 9.662473508897981e-06, "epoch": 0.23532184217613789, "percentage": 11.77, "elapsed_time": "7:30:21", "remaining_time": "2 days, 8:17:11", "throughput": 5669.65, "total_tokens": 153200600}
+{"current_steps": 3500, "total_steps": 29704, "loss": 0.43746633529663087, "lr": 9.661517857768052e-06, "epoch": 0.23565849717209802, "percentage": 11.78, "elapsed_time": "7:30:58", "remaining_time": "2 days, 8:16:21", "throughput": 5669.74, "total_tokens": 153413320}
+{"current_steps": 3505, "total_steps": 29704, "loss": 0.4074262619018555, "lr": 9.660560903059517e-06, "epoch": 0.23599515216805816, "percentage": 11.8, "elapsed_time": "7:31:39", "remaining_time": "2 days, 8:16:01", "throughput": 5668.74, "total_tokens": 153619928}
+{"current_steps": 3510, "total_steps": 29704, "loss": 0.4169961929321289, "lr": 9.659602645039988e-06, "epoch": 0.2363318071640183, "percentage": 11.82, "elapsed_time": "7:32:17", "remaining_time": "2 days, 8:15:17", "throughput": 5669.1, "total_tokens": 153844808}
+{"current_steps": 3515, "total_steps": 29704, "loss": 0.37222681045532224, "lr": 9.658643083977435e-06, "epoch": 0.23666846215997844, "percentage": 11.83, "elapsed_time": "7:32:52", "remaining_time": "2 days, 8:14:10", "throughput": 5669.65, "total_tokens": 154057008}
+{"current_steps": 3520, "total_steps": 29704, "loss": 0.3870618104934692, "lr": 9.657682220140202e-06, "epoch": 0.23700511715593858, "percentage": 11.85, "elapsed_time": "7:33:36", "remaining_time": "2 days, 8:14:10", "throughput": 5667.91, "total_tokens": 154257952}
+{"current_steps": 3525, "total_steps": 29704, "loss": 0.4269230365753174, "lr": 9.656720053796987e-06, "epoch": 0.23734177215189872, "percentage": 11.87, "elapsed_time": "7:34:13", "remaining_time": "2 days, 8:13:24", "throughput": 5667.8, "total_tokens": 154468968}
+{"current_steps": 3530, "total_steps": 29704, "loss": 0.39426770210266116, "lr": 9.655756585216859e-06, "epoch": 0.23767842714785886, "percentage": 11.88, "elapsed_time": "7:34:57", "remaining_time": "2 days, 8:13:26", "throughput": 5666.37, "total_tokens": 154679920}
+{"current_steps": 3535, "total_steps": 29704, "loss": 0.41727919578552247, "lr": 9.654791814669246e-06, "epoch": 0.238015082143819, "percentage": 11.9, "elapsed_time": "7:35:38", "remaining_time": "2 days, 8:13:05", "throughput": 5665.7, "total_tokens": 154894088}
+{"current_steps": 3540, "total_steps": 29704, "loss": 0.36776251792907716, "lr": 9.653825742423949e-06, "epoch": 0.23835173713977914, "percentage": 11.92, "elapsed_time": "7:36:20", "remaining_time": "2 days, 8:12:45", "throughput": 5665.36, "total_tokens": 155118416}
+{"current_steps": 3545, "total_steps": 29704, "loss": 0.38267481327056885, "lr": 9.652858368751118e-06, "epoch": 0.2386883921357393, "percentage": 11.93, "elapsed_time": "7:36:59", "remaining_time": "2 days, 8:12:09", "throughput": 5664.85, "total_tokens": 155325944}
+{"current_steps": 3550, "total_steps": 29704, "loss": 0.39499473571777344, "lr": 9.651889693921284e-06, "epoch": 0.23902504713169945, "percentage": 11.95, "elapsed_time": "7:37:33", "remaining_time": "2 days, 8:10:55", "throughput": 5666.19, "total_tokens": 155553936}
+{"current_steps": 3555, "total_steps": 29704, "loss": 0.46364564895629884, "lr": 9.650919718205331e-06, "epoch": 0.2393617021276596, "percentage": 11.97, "elapsed_time": "7:38:09", "remaining_time": "2 days, 8:10:02", "throughput": 5666.38, "total_tokens": 155767440}
+{"current_steps": 3560, "total_steps": 29704, "loss": 0.41176652908325195, "lr": 9.649948441874509e-06, "epoch": 0.23969835712361973, "percentage": 11.98, "elapsed_time": "7:38:53", "remaining_time": "2 days, 8:10:02", "throughput": 5665.47, "total_tokens": 155991432}
+{"current_steps": 3565, "total_steps": 29704, "loss": 0.3719303607940674, "lr": 9.648975865200434e-06, "epoch": 0.24003501211957987, "percentage": 12.0, "elapsed_time": "7:39:32", "remaining_time": "2 days, 8:09:23", "throughput": 5665.14, "total_tokens": 156200840}
+{"current_steps": 3570, "total_steps": 29704, "loss": 0.42696056365966795, "lr": 9.648001988455082e-06, "epoch": 0.24037166711554, "percentage": 12.02, "elapsed_time": "7:40:11", "remaining_time": "2 days, 8:08:44", "throughput": 5665.25, "total_tokens": 156423232}
+{"current_steps": 3575, "total_steps": 29704, "loss": 0.42109375, "lr": 9.647026811910794e-06, "epoch": 0.24070832211150014, "percentage": 12.04, "elapsed_time": "7:40:52", "remaining_time": "2 days, 8:08:23", "throughput": 5664.61, "total_tokens": 156638144}
+{"current_steps": 3580, "total_steps": 29704, "loss": 0.440277624130249, "lr": 9.646050335840279e-06, "epoch": 0.24104497710746028, "percentage": 12.05, "elapsed_time": "7:41:33", "remaining_time": "2 days, 8:08:06", "throughput": 5664.39, "total_tokens": 156867440}
+{"current_steps": 3585, "total_steps": 29704, "loss": 0.4013186454772949, "lr": 9.6450725605166e-06, "epoch": 0.24138163210342042, "percentage": 12.07, "elapsed_time": "7:42:06", "remaining_time": "2 days, 8:06:42", "throughput": 5665.41, "total_tokens": 157079728}
+{"current_steps": 3590, "total_steps": 29704, "loss": 0.4329217910766602, "lr": 9.644093486213192e-06, "epoch": 0.24171828709938056, "percentage": 12.09, "elapsed_time": "7:42:47", "remaining_time": "2 days, 8:06:20", "throughput": 5665.45, "total_tokens": 157312856}
+{"current_steps": 3595, "total_steps": 29704, "loss": 0.4125063896179199, "lr": 9.64311311320385e-06, "epoch": 0.2420549420953407, "percentage": 12.1, "elapsed_time": "7:43:21", "remaining_time": "2 days, 8:05:12", "throughput": 5666.45, "total_tokens": 157536824}
+{"current_steps": 3600, "total_steps": 29704, "loss": 0.39645483493804934, "lr": 9.642131441762732e-06, "epoch": 0.24239159709130084, "percentage": 12.12, "elapsed_time": "7:43:58", "remaining_time": "2 days, 8:04:22", "throughput": 5666.34, "total_tokens": 157744672}
+{"current_steps": 3605, "total_steps": 29704, "loss": 0.3727452516555786, "lr": 9.641148472164359e-06, "epoch": 0.24272825208726098, "percentage": 12.14, "elapsed_time": "7:44:35", "remaining_time": "2 days, 8:03:30", "throughput": 5666.48, "total_tokens": 157957176}
+{"current_steps": 3610, "total_steps": 29704, "loss": 0.39656991958618165, "lr": 9.640164204683616e-06, "epoch": 0.24306490708322112, "percentage": 12.15, "elapsed_time": "7:45:16", "remaining_time": "2 days, 8:03:04", "throughput": 5666.48, "total_tokens": 158185920}
+{"current_steps": 3615, "total_steps": 29704, "loss": 0.43434457778930663, "lr": 9.639178639595748e-06, "epoch": 0.24340156207918126, "percentage": 12.17, "elapsed_time": "7:45:44", "remaining_time": "2 days, 8:01:11", "throughput": 5668.75, "total_tokens": 158410048}
+{"current_steps": 3620, "total_steps": 29704, "loss": 0.40036301612854003, "lr": 9.638191777176367e-06, "epoch": 0.2437382170751414, "percentage": 12.19, "elapsed_time": "7:46:18", "remaining_time": "2 days, 8:00:01", "throughput": 5669.73, "total_tokens": 158632248}
+{"current_steps": 3625, "total_steps": 29704, "loss": 0.41867575645446775, "lr": 9.637203617701446e-06, "epoch": 0.24407487207110154, "percentage": 12.2, "elapsed_time": "7:47:01", "remaining_time": "2 days, 7:59:49", "throughput": 5669.38, "total_tokens": 158861832}
+{"current_steps": 3630, "total_steps": 29704, "loss": 0.3974162578582764, "lr": 9.63621416144732e-06, "epoch": 0.24441152706706168, "percentage": 12.22, "elapsed_time": "7:47:37", "remaining_time": "2 days, 7:58:52", "throughput": 5669.72, "total_tokens": 159076104}
+{"current_steps": 3635, "total_steps": 29704, "loss": 0.39289240837097167, "lr": 9.635223408690688e-06, "epoch": 0.24474818206302182, "percentage": 12.24, "elapsed_time": "7:48:12", "remaining_time": "2 days, 7:57:50", "throughput": 5669.68, "total_tokens": 159275688}
+{"current_steps": 3640, "total_steps": 29704, "loss": 0.4178638935089111, "lr": 9.63423135970861e-06, "epoch": 0.24508483705898196, "percentage": 12.25, "elapsed_time": "7:48:51", "remaining_time": "2 days, 7:57:17", "throughput": 5669.93, "total_tokens": 159506232}
+{"current_steps": 3645, "total_steps": 29704, "loss": 0.4228395462036133, "lr": 9.633238014778511e-06, "epoch": 0.2454214920549421, "percentage": 12.27, "elapsed_time": "7:49:24", "remaining_time": "2 days, 7:55:55", "throughput": 5671.5, "total_tokens": 159735712}
+{"current_steps": 3650, "total_steps": 29704, "loss": 0.3977930784225464, "lr": 9.632243374178173e-06, "epoch": 0.24575814705090224, "percentage": 12.29, "elapsed_time": "7:50:06", "remaining_time": "2 days, 7:55:39", "throughput": 5671.02, "total_tokens": 159959000}
+{"current_steps": 3655, "total_steps": 29704, "loss": 0.42597079277038574, "lr": 9.63124743818575e-06, "epoch": 0.24609480204686238, "percentage": 12.3, "elapsed_time": "7:50:42", "remaining_time": "2 days, 7:54:45", "throughput": 5671.64, "total_tokens": 160183304}
+{"current_steps": 3660, "total_steps": 29704, "loss": 0.385437536239624, "lr": 9.630250207079745e-06, "epoch": 0.24643145704282252, "percentage": 12.32, "elapsed_time": "7:51:19", "remaining_time": "2 days, 7:53:50", "throughput": 5671.88, "total_tokens": 160396208}
+{"current_steps": 3665, "total_steps": 29704, "loss": 0.41692662239074707, "lr": 9.629251681139038e-06, "epoch": 0.24676811203878266, "percentage": 12.34, "elapsed_time": "7:51:55", "remaining_time": "2 days, 7:52:57", "throughput": 5672.43, "total_tokens": 160619648}
+{"current_steps": 3670, "total_steps": 29704, "loss": 0.40501389503479, "lr": 9.628251860642857e-06, "epoch": 0.2471047670347428, "percentage": 12.36, "elapsed_time": "7:52:38", "remaining_time": "2 days, 7:52:46", "throughput": 5671.45, "total_tokens": 160832936}
+{"current_steps": 3675, "total_steps": 29704, "loss": 0.4110020637512207, "lr": 9.627250745870805e-06, "epoch": 0.24744142203070293, "percentage": 12.37, "elapsed_time": "7:53:15", "remaining_time": "2 days, 7:51:53", "throughput": 5671.61, "total_tokens": 161045560}
+{"current_steps": 3680, "total_steps": 29704, "loss": 0.4464590072631836, "lr": 9.626248337102834e-06, "epoch": 0.24777807702666307, "percentage": 12.39, "elapsed_time": "7:53:53", "remaining_time": "2 days, 7:51:17", "throughput": 5671.64, "total_tokens": 161267264}
+{"current_steps": 3685, "total_steps": 29704, "loss": 0.42216014862060547, "lr": 9.625244634619269e-06, "epoch": 0.2481147320226232, "percentage": 12.41, "elapsed_time": "7:54:30", "remaining_time": "2 days, 7:50:25", "throughput": 5671.97, "total_tokens": 161484928}
+{"current_steps": 3690, "total_steps": 29704, "loss": 0.4016411781311035, "lr": 9.62423963870079e-06, "epoch": 0.24845138701858335, "percentage": 12.42, "elapsed_time": "7:55:13", "remaining_time": "2 days, 7:50:19", "throughput": 5670.76, "total_tokens": 161695760}
+{"current_steps": 3695, "total_steps": 29704, "loss": 0.36588501930236816, "lr": 9.623233349628441e-06, "epoch": 0.2487880420145435, "percentage": 12.44, "elapsed_time": "7:55:54", "remaining_time": "2 days, 7:49:54", "throughput": 5670.09, "total_tokens": 161906344}
+{"current_steps": 3700, "total_steps": 29704, "loss": 0.37874155044555663, "lr": 9.62222576768363e-06, "epoch": 0.24912469701050363, "percentage": 12.46, "elapsed_time": "7:56:33", "remaining_time": "2 days, 7:49:19", "throughput": 5669.85, "total_tokens": 162121952}
+{"current_steps": 3705, "total_steps": 29704, "loss": 0.4508801460266113, "lr": 9.62121689314812e-06, "epoch": 0.24946135200646377, "percentage": 12.47, "elapsed_time": "7:57:17", "remaining_time": "2 days, 7:49:20", "throughput": 5668.69, "total_tokens": 162339592}
+{"current_steps": 3710, "total_steps": 29704, "loss": 0.4121876239776611, "lr": 9.620206726304045e-06, "epoch": 0.2497980070024239, "percentage": 12.49, "elapsed_time": "7:57:57", "remaining_time": "2 days, 7:48:46", "throughput": 5668.29, "total_tokens": 162551488}
+{"current_steps": 3715, "total_steps": 29704, "loss": 0.4007285118103027, "lr": 9.61919526743389e-06, "epoch": 0.2501346619983841, "percentage": 12.51, "elapsed_time": "7:58:37", "remaining_time": "2 days, 7:48:16", "throughput": 5668.23, "total_tokens": 162775296}
+{"current_steps": 3720, "total_steps": 29704, "loss": 0.4169614315032959, "lr": 9.618182516820507e-06, "epoch": 0.2504713169943442, "percentage": 12.52, "elapsed_time": "7:59:12", "remaining_time": "2 days, 7:47:16", "throughput": 5669.15, "total_tokens": 163004008}
+{"current_steps": 3725, "total_steps": 29704, "loss": 0.38545975685119627, "lr": 9.61716847474711e-06, "epoch": 0.25080797199030436, "percentage": 12.54, "elapsed_time": "7:59:48", "remaining_time": "2 days, 7:46:14", "throughput": 5669.58, "total_tokens": 163216408}
+{"current_steps": 3730, "total_steps": 29704, "loss": 0.41805257797241213, "lr": 9.616153141497273e-06, "epoch": 0.2511446269862645, "percentage": 12.56, "elapsed_time": "8:00:27", "remaining_time": "2 days, 7:45:41", "throughput": 5669.25, "total_tokens": 163430088}
+{"current_steps": 3735, "total_steps": 29704, "loss": 0.37319297790527345, "lr": 9.61513651735493e-06, "epoch": 0.25148128198222464, "percentage": 12.57, "elapsed_time": "8:01:05", "remaining_time": "2 days, 7:45:00", "throughput": 5669.06, "total_tokens": 163642328}
+{"current_steps": 3740, "total_steps": 29704, "loss": 0.3945745944976807, "lr": 9.614118602604376e-06, "epoch": 0.2518179369781848, "percentage": 12.59, "elapsed_time": "8:01:43", "remaining_time": "2 days, 7:44:18", "throughput": 5668.36, "total_tokens": 163838104}
+{"current_steps": 3745, "total_steps": 29704, "loss": 0.3722787380218506, "lr": 9.613099397530267e-06, "epoch": 0.2521545919741449, "percentage": 12.61, "elapsed_time": "8:02:27", "remaining_time": "2 days, 7:44:16", "throughput": 5667.57, "total_tokens": 164064200}
+{"current_steps": 3750, "total_steps": 29704, "loss": 0.4092113018035889, "lr": 9.612078902417623e-06, "epoch": 0.25249124697010505, "percentage": 12.62, "elapsed_time": "8:03:04", "remaining_time": "2 days, 7:43:24", "throughput": 5668.31, "total_tokens": 164293832}
+{"current_steps": 3755, "total_steps": 29704, "loss": 0.4049520492553711, "lr": 9.611057117551821e-06, "epoch": 0.2528279019660652, "percentage": 12.64, "elapsed_time": "8:03:42", "remaining_time": "2 days, 7:42:37", "throughput": 5667.89, "total_tokens": 164494120}
+{"current_steps": 3760, "total_steps": 29704, "loss": 0.41929073333740235, "lr": 9.610034043218598e-06, "epoch": 0.25316455696202533, "percentage": 12.66, "elapsed_time": "8:04:24", "remaining_time": "2 days, 7:42:26", "throughput": 5667.41, "total_tokens": 164721320}
+{"current_steps": 3765, "total_steps": 29704, "loss": 0.4089076995849609, "lr": 9.609009679704057e-06, "epoch": 0.2535012119579855, "percentage": 12.68, "elapsed_time": "8:05:00", "remaining_time": "2 days, 7:41:24", "throughput": 5668.08, "total_tokens": 164941328}
+{"current_steps": 3770, "total_steps": 29704, "loss": 0.4017925262451172, "lr": 9.607984027294654e-06, "epoch": 0.2538378669539456, "percentage": 12.69, "elapsed_time": "8:05:42", "remaining_time": "2 days, 7:41:14", "throughput": 5666.85, "total_tokens": 165148144}
+{"current_steps": 3775, "total_steps": 29704, "loss": 0.4037941455841064, "lr": 9.606957086277213e-06, "epoch": 0.25417452194990575, "percentage": 12.71, "elapsed_time": "8:06:23", "remaining_time": "2 days, 7:40:48", "throughput": 5666.69, "total_tokens": 165372064}
+{"current_steps": 3780, "total_steps": 29704, "loss": 0.4064283847808838, "lr": 9.605928856938913e-06, "epoch": 0.2545111769458659, "percentage": 12.73, "elapsed_time": "8:07:01", "remaining_time": "2 days, 7:40:06", "throughput": 5666.69, "total_tokens": 165589000}
+{"current_steps": 3785, "total_steps": 29704, "loss": 0.3694040060043335, "lr": 9.604899339567298e-06, "epoch": 0.25484783194182603, "percentage": 12.74, "elapsed_time": "8:07:42", "remaining_time": "2 days, 7:39:42", "throughput": 5666.22, "total_tokens": 165806496}
+{"current_steps": 3790, "total_steps": 29704, "loss": 0.370687198638916, "lr": 9.603868534450263e-06, "epoch": 0.25518448693778617, "percentage": 12.76, "elapsed_time": "8:08:17", "remaining_time": "2 days, 7:38:39", "throughput": 5667.4, "total_tokens": 166040152}
+{"current_steps": 3795, "total_steps": 29704, "loss": 0.3572968006134033, "lr": 9.602836441876074e-06, "epoch": 0.2555211419337463, "percentage": 12.78, "elapsed_time": "8:08:57", "remaining_time": "2 days, 7:38:13", "throughput": 5667.26, "total_tokens": 166264712}
+{"current_steps": 3800, "total_steps": 29704, "loss": 0.3995164394378662, "lr": 9.601803062133353e-06, "epoch": 0.25585779692970645, "percentage": 12.79, "elapsed_time": "8:09:41", "remaining_time": "2 days, 7:38:09", "throughput": 5666.36, "total_tokens": 166486544}
+{"current_steps": 3805, "total_steps": 29704, "loss": 0.3709728717803955, "lr": 9.600768395511078e-06, "epoch": 0.2561944519256666, "percentage": 12.81, "elapsed_time": "8:10:17", "remaining_time": "2 days, 7:37:10", "throughput": 5666.94, "total_tokens": 166705432}
+{"current_steps": 3810, "total_steps": 29704, "loss": 0.39358603954315186, "lr": 9.599732442298592e-06, "epoch": 0.25653110692162673, "percentage": 12.83, "elapsed_time": "8:10:53", "remaining_time": "2 days, 7:36:17", "throughput": 5667.46, "total_tokens": 166928072}
+{"current_steps": 3815, "total_steps": 29704, "loss": 0.3980567216873169, "lr": 9.598695202785596e-06, "epoch": 0.25686776191758687, "percentage": 12.84, "elapsed_time": "8:11:29", "remaining_time": "2 days, 7:35:21", "throughput": 5667.54, "total_tokens": 167135336}
+{"current_steps": 3820, "total_steps": 29704, "loss": 0.3841102600097656, "lr": 9.59765667726215e-06, "epoch": 0.257204416913547, "percentage": 12.86, "elapsed_time": "8:12:06", "remaining_time": "2 days, 7:34:30", "throughput": 5667.98, "total_tokens": 167356664}
+{"current_steps": 3825, "total_steps": 29704, "loss": 0.46039524078369143, "lr": 9.596616866018677e-06, "epoch": 0.25754107190950715, "percentage": 12.88, "elapsed_time": "8:12:48", "remaining_time": "2 days, 7:34:16", "throughput": 5667.04, "total_tokens": 167568600}
+{"current_steps": 3830, "total_steps": 29704, "loss": 0.3880587100982666, "lr": 9.595575769345955e-06, "epoch": 0.2578777269054673, "percentage": 12.89, "elapsed_time": "8:13:28", "remaining_time": "2 days, 7:33:43", "throughput": 5666.17, "total_tokens": 167766736}
+{"current_steps": 3835, "total_steps": 29704, "loss": 0.41129217147827146, "lr": 9.594533387535122e-06, "epoch": 0.2582143819014274, "percentage": 12.91, "elapsed_time": "8:14:07", "remaining_time": "2 days, 7:33:04", "throughput": 5666.51, "total_tokens": 167995600}
+{"current_steps": 3840, "total_steps": 29704, "loss": 0.4008028984069824, "lr": 9.593489720877678e-06, "epoch": 0.25855103689738757, "percentage": 12.93, "elapsed_time": "8:14:43", "remaining_time": "2 days, 7:32:11", "throughput": 5666.9, "total_tokens": 168214488}
+{"current_steps": 3845, "total_steps": 29704, "loss": 0.37677698135375975, "lr": 9.592444769665482e-06, "epoch": 0.2588876918933477, "percentage": 12.94, "elapsed_time": "8:15:26", "remaining_time": "2 days, 7:31:58", "throughput": 5666.64, "total_tokens": 168446672}
+{"current_steps": 3850, "total_steps": 29704, "loss": 0.3776779413223267, "lr": 9.59139853419075e-06, "epoch": 0.25922434688930784, "percentage": 12.96, "elapsed_time": "8:16:07", "remaining_time": "2 days, 7:31:40", "throughput": 5665.48, "total_tokens": 168649368}
+{"current_steps": 3855, "total_steps": 29704, "loss": 0.40406084060668945, "lr": 9.590351014746059e-06, "epoch": 0.259561001885268, "percentage": 12.98, "elapsed_time": "8:16:44", "remaining_time": "2 days, 7:30:46", "throughput": 5666.16, "total_tokens": 168875024}
+{"current_steps": 3860, "total_steps": 29704, "loss": 0.39687304496765136, "lr": 9.589302211624343e-06, "epoch": 0.2598976568812281, "percentage": 12.99, "elapsed_time": "8:17:20", "remaining_time": "2 days, 7:29:50", "throughput": 5666.54, "total_tokens": 169090984}
+{"current_steps": 3865, "total_steps": 29704, "loss": 0.3968550682067871, "lr": 9.5882521251189e-06, "epoch": 0.26023431187718826, "percentage": 13.01, "elapsed_time": "8:17:57", "remaining_time": "2 days, 7:29:05", "throughput": 5666.73, "total_tokens": 169310120}
+{"current_steps": 3870, "total_steps": 29704, "loss": 0.42056074142456057, "lr": 9.58720075552338e-06, "epoch": 0.2605709668731484, "percentage": 13.03, "elapsed_time": "8:18:29", "remaining_time": "2 days, 7:27:38", "throughput": 5667.93, "total_tokens": 169524504}
+{"current_steps": 3875, "total_steps": 29704, "loss": 0.40169625282287597, "lr": 9.586148103131797e-06, "epoch": 0.26090762186910854, "percentage": 13.05, "elapsed_time": "8:19:04", "remaining_time": "2 days, 7:26:37", "throughput": 5668.23, "total_tokens": 169733248}
+{"current_steps": 3880, "total_steps": 29704, "loss": 0.4272635459899902, "lr": 9.585094168238518e-06, "epoch": 0.2612442768650687, "percentage": 13.06, "elapsed_time": "8:19:43", "remaining_time": "2 days, 7:26:03", "throughput": 5667.82, "total_tokens": 169943480}
+{"current_steps": 3885, "total_steps": 29704, "loss": 0.39615912437438966, "lr": 9.58403895113828e-06, "epoch": 0.2615809318610288, "percentage": 13.08, "elapsed_time": "8:20:23", "remaining_time": "2 days, 7:25:29", "throughput": 5667.76, "total_tokens": 170164936}
+{"current_steps": 3890, "total_steps": 29704, "loss": 0.380422306060791, "lr": 9.582982452126164e-06, "epoch": 0.26191758685698896, "percentage": 13.1, "elapsed_time": "8:21:01", "remaining_time": "2 days, 7:24:46", "throughput": 5667.78, "total_tokens": 170380728}
+{"current_steps": 3895, "total_steps": 29704, "loss": 0.3720268726348877, "lr": 9.58192467149762e-06, "epoch": 0.2622542418529491, "percentage": 13.11, "elapsed_time": "8:21:42", "remaining_time": "2 days, 7:24:27", "throughput": 5666.95, "total_tokens": 170591568}
+{"current_steps": 3900, "total_steps": 29704, "loss": 0.4109041213989258, "lr": 9.580865609548452e-06, "epoch": 0.26259089684890924, "percentage": 13.13, "elapsed_time": "8:22:19", "remaining_time": "2 days, 7:23:32", "throughput": 5666.84, "total_tokens": 170793416}
+{"current_steps": 3905, "total_steps": 29704, "loss": 0.3842399835586548, "lr": 9.579805266574824e-06, "epoch": 0.2629275518448694, "percentage": 13.15, "elapsed_time": "8:22:58", "remaining_time": "2 days, 7:22:59", "throughput": 5666.48, "total_tokens": 171006328}
+{"current_steps": 3910, "total_steps": 29704, "loss": 0.40047216415405273, "lr": 9.578743642873258e-06, "epoch": 0.2632642068408295, "percentage": 13.16, "elapsed_time": "8:23:37", "remaining_time": "2 days, 7:22:24", "throughput": 5665.86, "total_tokens": 171209880}
+{"current_steps": 3915, "total_steps": 29704, "loss": 0.40533103942871096, "lr": 9.57768073874063e-06, "epoch": 0.26360086183678966, "percentage": 13.18, "elapsed_time": "8:24:10", "remaining_time": "2 days, 7:21:06", "throughput": 5667.18, "total_tokens": 171434584}
+{"current_steps": 3920, "total_steps": 29704, "loss": 0.41611337661743164, "lr": 9.576616554474182e-06, "epoch": 0.2639375168327498, "percentage": 13.2, "elapsed_time": "8:24:47", "remaining_time": "2 days, 7:20:19", "throughput": 5666.89, "total_tokens": 171637768}
+{"current_steps": 3925, "total_steps": 29704, "loss": 0.39514660835266113, "lr": 9.575551090371512e-06, "epoch": 0.26427417182870994, "percentage": 13.21, "elapsed_time": "8:25:28", "remaining_time": "2 days, 7:19:55", "throughput": 5666.09, "total_tokens": 171845072}
+{"current_steps": 3930, "total_steps": 29704, "loss": 0.4031528472900391, "lr": 9.574484346730564e-06, "epoch": 0.2646108268246701, "percentage": 13.23, "elapsed_time": "8:26:10", "remaining_time": "2 days, 7:19:35", "throughput": 5665.65, "total_tokens": 172066528}
+{"current_steps": 3935, "total_steps": 29704, "loss": 0.4126886367797852, "lr": 9.573416323849658e-06, "epoch": 0.2649474818206302, "percentage": 13.25, "elapsed_time": "8:26:51", "remaining_time": "2 days, 7:19:13", "throughput": 5665.19, "total_tokens": 172285504}
+{"current_steps": 3940, "total_steps": 29704, "loss": 0.3947685718536377, "lr": 9.572347022027461e-06, "epoch": 0.26528413681659035, "percentage": 13.26, "elapsed_time": "8:27:32", "remaining_time": "2 days, 7:18:54", "throughput": 5664.88, "total_tokens": 172512448}
+{"current_steps": 3945, "total_steps": 29704, "loss": 0.40018043518066404, "lr": 9.571276441563e-06, "epoch": 0.2656207918125505, "percentage": 13.28, "elapsed_time": "8:28:04", "remaining_time": "2 days, 7:17:30", "throughput": 5665.56, "total_tokens": 172712640}
+{"current_steps": 3950, "total_steps": 29704, "loss": 0.38366117477416994, "lr": 9.570204582755659e-06, "epoch": 0.26595744680851063, "percentage": 13.3, "elapsed_time": "8:28:44", "remaining_time": "2 days, 7:16:58", "throughput": 5665.48, "total_tokens": 172934512}
+{"current_steps": 3955, "total_steps": 29704, "loss": 0.4079860210418701, "lr": 9.569131445905178e-06, "epoch": 0.2662941018044708, "percentage": 13.31, "elapsed_time": "8:29:17", "remaining_time": "2 days, 7:15:43", "throughput": 5666.33, "total_tokens": 173148536}
+{"current_steps": 3960, "total_steps": 29704, "loss": 0.41454520225524905, "lr": 9.568057031311659e-06, "epoch": 0.2666307568004309, "percentage": 13.33, "elapsed_time": "8:29:52", "remaining_time": "2 days, 7:14:39", "throughput": 5666.73, "total_tokens": 173357456}
+{"current_steps": 3965, "total_steps": 29704, "loss": 0.43401298522949217, "lr": 9.566981339275557e-06, "epoch": 0.26696741179639105, "percentage": 13.35, "elapsed_time": "8:30:36", "remaining_time": "2 days, 7:14:40", "throughput": 5665.27, "total_tokens": 173565960}
+{"current_steps": 3970, "total_steps": 29704, "loss": 0.4027921199798584, "lr": 9.565904370097687e-06, "epoch": 0.2673040667923512, "percentage": 13.37, "elapsed_time": "8:31:15", "remaining_time": "2 days, 7:14:00", "throughput": 5664.67, "total_tokens": 173765184}
+{"current_steps": 3975, "total_steps": 29704, "loss": 0.3860454082489014, "lr": 9.564826124079219e-06, "epoch": 0.26764072178831133, "percentage": 13.38, "elapsed_time": "8:32:02", "remaining_time": "2 days, 7:14:18", "throughput": 5663.01, "total_tokens": 173982664}
+{"current_steps": 3980, "total_steps": 29704, "loss": 0.37142066955566405, "lr": 9.563746601521681e-06, "epoch": 0.26797737678427147, "percentage": 13.4, "elapsed_time": "8:32:44", "remaining_time": "2 days, 7:14:00", "throughput": 5662.64, "total_tokens": 174207720}
+{"current_steps": 3985, "total_steps": 29704, "loss": 0.39812779426574707, "lr": 9.562665802726961e-06, "epoch": 0.2683140317802316, "percentage": 13.42, "elapsed_time": "8:33:29", "remaining_time": "2 days, 7:14:04", "throughput": 5661.77, "total_tokens": 174437200}
+{"current_steps": 3990, "total_steps": 29704, "loss": 0.3973848819732666, "lr": 9.561583727997297e-06, "epoch": 0.26865068677619175, "percentage": 13.43, "elapsed_time": "8:34:12", "remaining_time": "2 days, 7:13:54", "throughput": 5660.71, "total_tokens": 174649000}
+{"current_steps": 3995, "total_steps": 29704, "loss": 0.4132582664489746, "lr": 9.560500377635288e-06, "epoch": 0.2689873417721519, "percentage": 13.45, "elapsed_time": "8:34:54", "remaining_time": "2 days, 7:13:37", "throughput": 5660.16, "total_tokens": 174869968}
+{"current_steps": 4000, "total_steps": 29704, "loss": 0.4121055603027344, "lr": 9.559415751943893e-06, "epoch": 0.26932399676811203, "percentage": 13.47, "elapsed_time": "8:35:28", "remaining_time": "2 days, 7:12:25", "throughput": 5661.07, "total_tokens": 175087088}
+{"current_steps": 4005, "total_steps": 29704, "loss": 0.3906848430633545, "lr": 9.558329851226422e-06, "epoch": 0.26966065176407217, "percentage": 13.48, "elapsed_time": "8:36:07", "remaining_time": "2 days, 7:11:48", "throughput": 5661.27, "total_tokens": 175314216}
+{"current_steps": 4010, "total_steps": 29704, "loss": 0.4193889141082764, "lr": 9.557242675786542e-06, "epoch": 0.2699973067600323, "percentage": 13.5, "elapsed_time": "8:36:53", "remaining_time": "2 days, 7:11:57", "throughput": 5660.17, "total_tokens": 175541064}
+{"current_steps": 4015, "total_steps": 29704, "loss": 0.4430998802185059, "lr": 9.55615422592828e-06, "epoch": 0.27033396175599245, "percentage": 13.52, "elapsed_time": "8:37:33", "remaining_time": "2 days, 7:11:25", "throughput": 5660.32, "total_tokens": 175770592}
+{"current_steps": 4020, "total_steps": 29704, "loss": 0.4185193061828613, "lr": 9.555064501956018e-06, "epoch": 0.2706706167519526, "percentage": 13.53, "elapsed_time": "8:38:10", "remaining_time": "2 days, 7:10:40", "throughput": 5660.45, "total_tokens": 175988096}
+{"current_steps": 4025, "total_steps": 29704, "loss": 0.3700920581817627, "lr": 9.553973504174493e-06, "epoch": 0.2710072717479127, "percentage": 13.55, "elapsed_time": "8:38:51", "remaining_time": "2 days, 7:10:14", "throughput": 5660.3, "total_tokens": 176213488}
+{"current_steps": 4030, "total_steps": 29704, "loss": 0.41269850730895996, "lr": 9.5528812328888e-06, "epoch": 0.27134392674387287, "percentage": 13.57, "elapsed_time": "8:39:33", "remaining_time": "2 days, 7:09:57", "throughput": 5659.29, "total_tokens": 176420144}
+{"current_steps": 4035, "total_steps": 29704, "loss": 0.3877572536468506, "lr": 9.551787688404386e-06, "epoch": 0.271680581739833, "percentage": 13.58, "elapsed_time": "8:40:18", "remaining_time": "2 days, 7:09:59", "throughput": 5658.03, "total_tokens": 176635200}
+{"current_steps": 4040, "total_steps": 29704, "loss": 0.3883471727371216, "lr": 9.550692871027061e-06, "epoch": 0.27201723673579314, "percentage": 13.6, "elapsed_time": "8:41:01", "remaining_time": "2 days, 7:09:46", "throughput": 5657.36, "total_tokens": 176856552}
+{"current_steps": 4045, "total_steps": 29704, "loss": 0.37834415435791013, "lr": 9.549596781062984e-06, "epoch": 0.2723538917317533, "percentage": 13.62, "elapsed_time": "8:41:39", "remaining_time": "2 days, 7:09:07", "throughput": 5657.72, "total_tokens": 177085864}
+{"current_steps": 4050, "total_steps": 29704, "loss": 0.3912312984466553, "lr": 9.548499418818677e-06, "epoch": 0.2726905467277134, "percentage": 13.63, "elapsed_time": "8:42:23", "remaining_time": "2 days, 7:09:00", "throughput": 5656.42, "total_tokens": 177292800}
+{"current_steps": 4055, "total_steps": 29704, "loss": 0.448984956741333, "lr": 9.547400784601011e-06, "epoch": 0.27302720172367356, "percentage": 13.65, "elapsed_time": "8:43:08", "remaining_time": "2 days, 7:08:58", "throughput": 5655.29, "total_tokens": 177509120}
+{"current_steps": 4060, "total_steps": 29704, "loss": 0.41724767684936526, "lr": 9.546300878717216e-06, "epoch": 0.2733638567196337, "percentage": 13.67, "elapsed_time": "8:43:39", "remaining_time": "2 days, 7:07:33", "throughput": 5656.91, "total_tokens": 177737192}
+{"current_steps": 4065, "total_steps": 29704, "loss": 0.3876931667327881, "lr": 9.545199701474877e-06, "epoch": 0.27370051171559384, "percentage": 13.69, "elapsed_time": "8:44:20", "remaining_time": "2 days, 7:07:08", "throughput": 5656.51, "total_tokens": 177955528}
+{"current_steps": 4070, "total_steps": 29704, "loss": 0.37918825149536134, "lr": 9.544097253181935e-06, "epoch": 0.274037166711554, "percentage": 13.7, "elapsed_time": "8:45:01", "remaining_time": "2 days, 7:06:43", "throughput": 5655.92, "total_tokens": 178168688}
+{"current_steps": 4075, "total_steps": 29704, "loss": 0.41271028518676756, "lr": 9.542993534146687e-06, "epoch": 0.2743738217075141, "percentage": 13.72, "elapsed_time": "8:45:48", "remaining_time": "2 days, 7:06:59", "throughput": 5654.19, "total_tokens": 178382216}
+{"current_steps": 4080, "total_steps": 29704, "loss": 0.4018410682678223, "lr": 9.541888544677784e-06, "epoch": 0.27471047670347426, "percentage": 13.74, "elapsed_time": "8:46:23", "remaining_time": "2 days, 7:05:57", "throughput": 5654.88, "total_tokens": 178601400}
+{"current_steps": 4085, "total_steps": 29704, "loss": 0.36272602081298827, "lr": 9.540782285084232e-06, "epoch": 0.2750471316994344, "percentage": 13.75, "elapsed_time": "8:47:04", "remaining_time": "2 days, 7:05:33", "throughput": 5654.32, "total_tokens": 178815704}
+{"current_steps": 4090, "total_steps": 29704, "loss": 0.39566807746887206, "lr": 9.539674755675392e-06, "epoch": 0.27538378669539454, "percentage": 13.77, "elapsed_time": "8:47:47", "remaining_time": "2 days, 7:05:20", "throughput": 5653.55, "total_tokens": 179033752}
+{"current_steps": 4095, "total_steps": 29704, "loss": 0.35824832916259763, "lr": 9.538565956760983e-06, "epoch": 0.2757204416913547, "percentage": 13.79, "elapsed_time": "8:48:21", "remaining_time": "2 days, 7:04:13", "throughput": 5654.9, "total_tokens": 179269312}
+{"current_steps": 4100, "total_steps": 29704, "loss": 0.3908623933792114, "lr": 9.537455888651078e-06, "epoch": 0.2760570966873148, "percentage": 13.8, "elapsed_time": "8:48:57", "remaining_time": "2 days, 7:03:17", "throughput": 5655.75, "total_tokens": 179500456}
+{"current_steps": 4105, "total_steps": 29704, "loss": 0.4113442420959473, "lr": 9.536344551656103e-06, "epoch": 0.27639375168327496, "percentage": 13.82, "elapsed_time": "8:49:38", "remaining_time": "2 days, 7:02:51", "throughput": 5655.65, "total_tokens": 179727568}
+{"current_steps": 4110, "total_steps": 29704, "loss": 0.392077112197876, "lr": 9.535231946086838e-06, "epoch": 0.2767304066792351, "percentage": 13.84, "elapsed_time": "8:50:10", "remaining_time": "2 days, 7:01:32", "throughput": 5656.86, "total_tokens": 179947920}
+{"current_steps": 4115, "total_steps": 29704, "loss": 0.41229848861694335, "lr": 9.534118072254421e-06, "epoch": 0.27706706167519524, "percentage": 13.85, "elapsed_time": "8:50:49", "remaining_time": "2 days, 7:00:55", "throughput": 5656.99, "total_tokens": 180172368}
+{"current_steps": 4120, "total_steps": 29704, "loss": 0.38343167304992676, "lr": 9.533002930470345e-06, "epoch": 0.2774037166711554, "percentage": 13.87, "elapsed_time": "8:51:33", "remaining_time": "2 days, 7:00:50", "throughput": 5656.07, "total_tokens": 180392712}
+{"current_steps": 4125, "total_steps": 29704, "loss": 0.38598432540893557, "lr": 9.531886521046452e-06, "epoch": 0.2777403716671155, "percentage": 13.89, "elapsed_time": "8:52:14", "remaining_time": "2 days, 7:00:27", "throughput": 5655.09, "total_tokens": 180594944}
+{"current_steps": 4130, "total_steps": 29704, "loss": 0.41448087692260743, "lr": 9.530768844294947e-06, "epoch": 0.27807702666307565, "percentage": 13.9, "elapsed_time": "8:52:55", "remaining_time": "2 days, 7:00:00", "throughput": 5654.83, "total_tokens": 180816024}
+{"current_steps": 4135, "total_steps": 29704, "loss": 0.41909160614013674, "lr": 9.52964990052838e-06, "epoch": 0.2784136816590358, "percentage": 13.92, "elapsed_time": "8:53:34", "remaining_time": "2 days, 6:59:22", "throughput": 5654.82, "total_tokens": 181035024}
+{"current_steps": 4140, "total_steps": 29704, "loss": 0.3741894245147705, "lr": 9.528529690059663e-06, "epoch": 0.27875033665499593, "percentage": 13.94, "elapsed_time": "8:54:15", "remaining_time": "2 days, 6:58:57", "throughput": 5654.58, "total_tokens": 181258672}
+{"current_steps": 4145, "total_steps": 29704, "loss": 0.4044219970703125, "lr": 9.52740821320206e-06, "epoch": 0.2790869916509561, "percentage": 13.95, "elapsed_time": "8:54:46", "remaining_time": "2 days, 6:57:34", "throughput": 5655.76, "total_tokens": 181475056}
+{"current_steps": 4150, "total_steps": 29704, "loss": 0.4395021915435791, "lr": 9.526285470269185e-06, "epoch": 0.2794236466469162, "percentage": 13.97, "elapsed_time": "8:55:21", "remaining_time": "2 days, 6:56:29", "throughput": 5656.74, "total_tokens": 181701504}
+{"current_steps": 4155, "total_steps": 29704, "loss": 0.4143718719482422, "lr": 9.525161461575012e-06, "epoch": 0.2797603016428764, "percentage": 13.99, "elapsed_time": "8:56:02", "remaining_time": "2 days, 6:56:03", "throughput": 5655.91, "total_tokens": 181905448}
+{"current_steps": 4160, "total_steps": 29704, "loss": 0.38724517822265625, "lr": 9.524036187433867e-06, "epoch": 0.28009695663883655, "percentage": 14.0, "elapsed_time": "8:56:39", "remaining_time": "2 days, 6:55:17", "throughput": 5655.72, "total_tokens": 182111832}
+{"current_steps": 4165, "total_steps": 29704, "loss": 0.38469388484954836, "lr": 9.522909648160427e-06, "epoch": 0.2804336116347967, "percentage": 14.02, "elapsed_time": "8:57:17", "remaining_time": "2 days, 6:54:35", "throughput": 5655.84, "total_tokens": 182331456}
+{"current_steps": 4170, "total_steps": 29704, "loss": 0.4097011566162109, "lr": 9.521781844069727e-06, "epoch": 0.2807702666307568, "percentage": 14.04, "elapsed_time": "8:58:02", "remaining_time": "2 days, 6:54:33", "throughput": 5654.94, "total_tokens": 182555536}
+{"current_steps": 4175, "total_steps": 29704, "loss": 0.38606297969818115, "lr": 9.520652775477154e-06, "epoch": 0.28110692162671697, "percentage": 14.06, "elapsed_time": "8:58:44", "remaining_time": "2 days, 6:54:13", "throughput": 5654.48, "total_tokens": 182776104}
+{"current_steps": 4180, "total_steps": 29704, "loss": 0.36697864532470703, "lr": 9.519522442698447e-06, "epoch": 0.2814435766226771, "percentage": 14.07, "elapsed_time": "8:59:14", "remaining_time": "2 days, 6:52:42", "throughput": 5655.72, "total_tokens": 182986640}
+{"current_steps": 4185, "total_steps": 29704, "loss": 0.4045860290527344, "lr": 9.518390846049699e-06, "epoch": 0.28178023161863724, "percentage": 14.09, "elapsed_time": "8:59:58", "remaining_time": "2 days, 6:52:35", "throughput": 5654.64, "total_tokens": 183200160}
+{"current_steps": 4190, "total_steps": 29704, "loss": 0.42084197998046874, "lr": 9.517257985847361e-06, "epoch": 0.2821168866145974, "percentage": 14.11, "elapsed_time": "9:00:41", "remaining_time": "2 days, 6:52:22", "throughput": 5653.68, "total_tokens": 183411672}
+{"current_steps": 4195, "total_steps": 29704, "loss": 0.4293764591217041, "lr": 9.516123862408232e-06, "epoch": 0.2824535416105575, "percentage": 14.12, "elapsed_time": "9:01:25", "remaining_time": "2 days, 6:52:16", "throughput": 5652.94, "total_tokens": 183636784}
+{"current_steps": 4200, "total_steps": 29704, "loss": 0.3744691848754883, "lr": 9.514988476049466e-06, "epoch": 0.28279019660651766, "percentage": 14.14, "elapsed_time": "9:02:01", "remaining_time": "2 days, 6:51:20", "throughput": 5653.26, "total_tokens": 183850024}
+{"current_steps": 4205, "total_steps": 29704, "loss": 0.39441556930541993, "lr": 9.513851827088567e-06, "epoch": 0.2831268516024778, "percentage": 14.16, "elapsed_time": "9:02:38", "remaining_time": "2 days, 6:50:31", "throughput": 5653.14, "total_tokens": 184055312}
+{"current_steps": 4210, "total_steps": 29704, "loss": 0.399395227432251, "lr": 9.512713915843402e-06, "epoch": 0.28346350659843794, "percentage": 14.17, "elapsed_time": "9:03:11", "remaining_time": "2 days, 6:49:18", "throughput": 5654.33, "total_tokens": 184280840}
+{"current_steps": 4215, "total_steps": 29704, "loss": 0.38000011444091797, "lr": 9.511574742632177e-06, "epoch": 0.2838001615943981, "percentage": 14.19, "elapsed_time": "9:03:55", "remaining_time": "2 days, 6:49:11", "throughput": 5653.42, "total_tokens": 184499800}
+{"current_steps": 4220, "total_steps": 29704, "loss": 0.45735921859741213, "lr": 9.510434307773464e-06, "epoch": 0.2841368165903582, "percentage": 14.21, "elapsed_time": "9:04:33", "remaining_time": "2 days, 6:48:32", "throughput": 5653.61, "total_tokens": 184725080}
+{"current_steps": 4225, "total_steps": 29704, "loss": 0.35365097522735595, "lr": 9.509292611586179e-06, "epoch": 0.28447347158631836, "percentage": 14.22, "elapsed_time": "9:05:14", "remaining_time": "2 days, 6:48:03", "throughput": 5652.39, "total_tokens": 184912768}
+{"current_steps": 4230, "total_steps": 29704, "loss": 0.40491819381713867, "lr": 9.508149654389592e-06, "epoch": 0.2848101265822785, "percentage": 14.24, "elapsed_time": "9:05:55", "remaining_time": "2 days, 6:47:40", "throughput": 5652.08, "total_tokens": 185136408}
+{"current_steps": 4235, "total_steps": 29704, "loss": 0.3746334552764893, "lr": 9.507005436503331e-06, "epoch": 0.28514678157823864, "percentage": 14.26, "elapsed_time": "9:06:38", "remaining_time": "2 days, 6:47:29", "throughput": 5651.65, "total_tokens": 185367152}
+{"current_steps": 4240, "total_steps": 29704, "loss": 0.4024499893188477, "lr": 9.505859958247373e-06, "epoch": 0.2854834365741988, "percentage": 14.27, "elapsed_time": "9:07:17", "remaining_time": "2 days, 6:46:49", "throughput": 5651.51, "total_tokens": 185580528}
+{"current_steps": 4245, "total_steps": 29704, "loss": 0.3693699359893799, "lr": 9.504713219942045e-06, "epoch": 0.2858200915701589, "percentage": 14.29, "elapsed_time": "9:07:53", "remaining_time": "2 days, 6:45:54", "throughput": 5652.02, "total_tokens": 185800152}
+{"current_steps": 4250, "total_steps": 29704, "loss": 0.3989689350128174, "lr": 9.503565221908029e-06, "epoch": 0.28615674656611906, "percentage": 14.31, "elapsed_time": "9:08:28", "remaining_time": "2 days, 6:44:55", "throughput": 5652.72, "total_tokens": 186023072}
+{"current_steps": 4255, "total_steps": 29704, "loss": 0.41822104454040526, "lr": 9.50241596446636e-06, "epoch": 0.2864934015620792, "percentage": 14.32, "elapsed_time": "9:09:08", "remaining_time": "2 days, 6:44:24", "throughput": 5652.77, "total_tokens": 186251344}
+{"current_steps": 4260, "total_steps": 29704, "loss": 0.40343165397644043, "lr": 9.501265447938423e-06, "epoch": 0.28683005655803934, "percentage": 14.34, "elapsed_time": "9:09:45", "remaining_time": "2 days, 6:43:37", "throughput": 5652.7, "total_tokens": 186459392}
+{"current_steps": 4265, "total_steps": 29704, "loss": 0.41559224128723143, "lr": 9.500113672645958e-06, "epoch": 0.2871667115539995, "percentage": 14.36, "elapsed_time": "9:10:20", "remaining_time": "2 days, 6:42:35", "throughput": 5653.24, "total_tokens": 186674536}
+{"current_steps": 4270, "total_steps": 29704, "loss": 0.3698106288909912, "lr": 9.498960638911054e-06, "epoch": 0.2875033665499596, "percentage": 14.38, "elapsed_time": "9:11:06", "remaining_time": "2 days, 6:42:39", "throughput": 5651.73, "total_tokens": 186883568}
+{"current_steps": 4275, "total_steps": 29704, "loss": 0.394811749458313, "lr": 9.497806347056153e-06, "epoch": 0.28784002154591976, "percentage": 14.39, "elapsed_time": "9:11:48", "remaining_time": "2 days, 6:42:17", "throughput": 5651.23, "total_tokens": 187101448}
+{"current_steps": 4280, "total_steps": 29704, "loss": 0.41505231857299807, "lr": 9.49665079740405e-06, "epoch": 0.2881766765418799, "percentage": 14.41, "elapsed_time": "9:12:24", "remaining_time": "2 days, 6:41:22", "throughput": 5651.79, "total_tokens": 187323568}
+{"current_steps": 4285, "total_steps": 29704, "loss": 0.41215314865112307, "lr": 9.495493990277889e-06, "epoch": 0.28851333153784003, "percentage": 14.43, "elapsed_time": "9:12:57", "remaining_time": "2 days, 6:40:09", "throughput": 5652.61, "total_tokens": 187536768}
+{"current_steps": 4290, "total_steps": 29704, "loss": 0.39543790817260743, "lr": 9.49433592600117e-06, "epoch": 0.2888499865338002, "percentage": 14.44, "elapsed_time": "9:13:38", "remaining_time": "2 days, 6:39:45", "throughput": 5652.35, "total_tokens": 187761936}
+{"current_steps": 4295, "total_steps": 29704, "loss": 0.38694300651550295, "lr": 9.49317660489774e-06, "epoch": 0.2891866415297603, "percentage": 14.46, "elapsed_time": "9:14:14", "remaining_time": "2 days, 6:38:53", "throughput": 5653.02, "total_tokens": 187989768}
+{"current_steps": 4300, "total_steps": 29704, "loss": 0.39570274353027346, "lr": 9.4920160272918e-06, "epoch": 0.28952329652572045, "percentage": 14.48, "elapsed_time": "9:14:52", "remaining_time": "2 days, 6:38:06", "throughput": 5653.63, "total_tokens": 188221328}
+{"current_steps": 4305, "total_steps": 29704, "loss": 0.40100975036621095, "lr": 9.490854193507904e-06, "epoch": 0.2898599515216806, "percentage": 14.49, "elapsed_time": "9:15:33", "remaining_time": "2 days, 6:37:43", "throughput": 5653.09, "total_tokens": 188436984}
+{"current_steps": 4310, "total_steps": 29704, "loss": 0.4165069580078125, "lr": 9.489691103870951e-06, "epoch": 0.29019660651764073, "percentage": 14.51, "elapsed_time": "9:16:10", "remaining_time": "2 days, 6:36:52", "throughput": 5653.78, "total_tokens": 188666800}
+{"current_steps": 4315, "total_steps": 29704, "loss": 0.3968646764755249, "lr": 9.488526758706198e-06, "epoch": 0.29053326151360087, "percentage": 14.53, "elapsed_time": "9:16:48", "remaining_time": "2 days, 6:36:14", "throughput": 5653.91, "total_tokens": 188890864}
+{"current_steps": 4320, "total_steps": 29704, "loss": 0.42871532440185545, "lr": 9.48736115833925e-06, "epoch": 0.290869916509561, "percentage": 14.54, "elapsed_time": "9:17:24", "remaining_time": "2 days, 6:35:18", "throughput": 5654.04, "total_tokens": 189097152}
+{"current_steps": 4325, "total_steps": 29704, "loss": 0.388767147064209, "lr": 9.486194303096062e-06, "epoch": 0.29120657150552115, "percentage": 14.56, "elapsed_time": "9:18:00", "remaining_time": "2 days, 6:34:21", "throughput": 5654.64, "total_tokens": 189319032}
+{"current_steps": 4330, "total_steps": 29704, "loss": 0.37795631885528563, "lr": 9.485026193302945e-06, "epoch": 0.2915432265014813, "percentage": 14.58, "elapsed_time": "9:18:35", "remaining_time": "2 days, 6:33:19", "throughput": 5655.37, "total_tokens": 189540056}
+{"current_steps": 4335, "total_steps": 29704, "loss": 0.43641042709350586, "lr": 9.483856829286552e-06, "epoch": 0.29187988149744143, "percentage": 14.59, "elapsed_time": "9:19:16", "remaining_time": "2 days, 6:32:58", "throughput": 5655.07, "total_tokens": 189765816}
+{"current_steps": 4340, "total_steps": 29704, "loss": 0.4038511276245117, "lr": 9.482686211373896e-06, "epoch": 0.29221653649340157, "percentage": 14.61, "elapsed_time": "9:19:57", "remaining_time": "2 days, 6:32:29", "throughput": 5655.02, "total_tokens": 189992032}
+{"current_steps": 4345, "total_steps": 29704, "loss": 0.4172847747802734, "lr": 9.481514339892335e-06, "epoch": 0.2925531914893617, "percentage": 14.63, "elapsed_time": "9:20:33", "remaining_time": "2 days, 6:31:36", "throughput": 5655.83, "total_tokens": 190224704}
+{"current_steps": 4350, "total_steps": 29704, "loss": 0.41391868591308595, "lr": 9.48034121516958e-06, "epoch": 0.29288984648532185, "percentage": 14.64, "elapsed_time": "9:21:11", "remaining_time": "2 days, 6:30:52", "throughput": 5655.83, "total_tokens": 190437840}
+{"current_steps": 4355, "total_steps": 29704, "loss": 0.38980252742767335, "lr": 9.47916683753369e-06, "epoch": 0.293226501481282, "percentage": 14.66, "elapsed_time": "9:21:53", "remaining_time": "2 days, 6:30:35", "throughput": 5655.18, "total_tokens": 190656112}
+{"current_steps": 4360, "total_steps": 29704, "loss": 0.4024055480957031, "lr": 9.477991207313077e-06, "epoch": 0.2935631564772421, "percentage": 14.68, "elapsed_time": "9:22:34", "remaining_time": "2 days, 6:30:09", "throughput": 5655.07, "total_tokens": 190883776}
+{"current_steps": 4365, "total_steps": 29704, "loss": 0.38675341606140134, "lr": 9.476814324836504e-06, "epoch": 0.29389981147320227, "percentage": 14.69, "elapsed_time": "9:23:09", "remaining_time": "2 days, 6:29:07", "throughput": 5656.0, "total_tokens": 191111960}
+{"current_steps": 4370, "total_steps": 29704, "loss": 0.44519591331481934, "lr": 9.475636190433078e-06, "epoch": 0.2942364664691624, "percentage": 14.71, "elapsed_time": "9:23:47", "remaining_time": "2 days, 6:28:26", "throughput": 5656.44, "total_tokens": 191343016}
+{"current_steps": 4375, "total_steps": 29704, "loss": 0.3989989995956421, "lr": 9.474456804432264e-06, "epoch": 0.29457312146512254, "percentage": 14.73, "elapsed_time": "9:24:24", "remaining_time": "2 days, 6:27:36", "throughput": 5657.07, "total_tokens": 191572600}
+{"current_steps": 4380, "total_steps": 29704, "loss": 0.41669330596923826, "lr": 9.473276167163872e-06, "epoch": 0.2949097764610827, "percentage": 14.75, "elapsed_time": "9:25:03", "remaining_time": "2 days, 6:27:01", "throughput": 5657.28, "total_tokens": 191802056}
+{"current_steps": 4385, "total_steps": 29704, "loss": 0.41675920486450196, "lr": 9.472094278958066e-06, "epoch": 0.2952464314570428, "percentage": 14.76, "elapsed_time": "9:25:47", "remaining_time": "2 days, 6:26:52", "throughput": 5656.79, "total_tokens": 192033856}
+{"current_steps": 4390, "total_steps": 29704, "loss": 0.3885148763656616, "lr": 9.470911140145353e-06, "epoch": 0.29558308645300296, "percentage": 14.78, "elapsed_time": "9:26:30", "remaining_time": "2 days, 6:26:41", "throughput": 5655.75, "total_tokens": 192244376}
+{"current_steps": 4395, "total_steps": 29704, "loss": 0.3994889736175537, "lr": 9.469726751056599e-06, "epoch": 0.2959197414489631, "percentage": 14.8, "elapsed_time": "9:27:11", "remaining_time": "2 days, 6:26:10", "throughput": 5655.62, "total_tokens": 192466432}
+{"current_steps": 4400, "total_steps": 29704, "loss": 0.42194247245788574, "lr": 9.468541112023009e-06, "epoch": 0.29625639644492324, "percentage": 14.81, "elapsed_time": "9:27:48", "remaining_time": "2 days, 6:25:27", "throughput": 5655.49, "total_tokens": 192676720}
+{"current_steps": 4405, "total_steps": 29704, "loss": 0.4138471603393555, "lr": 9.467354223376149e-06, "epoch": 0.2965930514408834, "percentage": 14.83, "elapsed_time": "9:28:21", "remaining_time": "2 days, 6:24:13", "throughput": 5655.69, "total_tokens": 192867824}
+{"current_steps": 4410, "total_steps": 29704, "loss": 0.4040337562561035, "lr": 9.466166085447923e-06, "epoch": 0.2969297064368435, "percentage": 14.85, "elapsed_time": "9:29:04", "remaining_time": "2 days, 6:23:57", "throughput": 5654.82, "total_tokens": 193079552}
+{"current_steps": 4415, "total_steps": 29704, "loss": 0.3852673053741455, "lr": 9.464976698570595e-06, "epoch": 0.29726636143280366, "percentage": 14.86, "elapsed_time": "9:29:39", "remaining_time": "2 days, 6:22:56", "throughput": 5654.77, "total_tokens": 193274704}
+{"current_steps": 4420, "total_steps": 29704, "loss": 0.42119736671447755, "lr": 9.463786063076769e-06, "epoch": 0.2976030164287638, "percentage": 14.88, "elapsed_time": "9:30:17", "remaining_time": "2 days, 6:22:16", "throughput": 5654.83, "total_tokens": 193494016}
+{"current_steps": 4425, "total_steps": 29704, "loss": 0.34813380241394043, "lr": 9.462594179299408e-06, "epoch": 0.29793967142472394, "percentage": 14.9, "elapsed_time": "9:30:55", "remaining_time": "2 days, 6:21:34", "throughput": 5654.81, "total_tokens": 193708944}
+{"current_steps": 4430, "total_steps": 29704, "loss": 0.39094252586364747, "lr": 9.461401047571811e-06, "epoch": 0.2982763264206841, "percentage": 14.91, "elapsed_time": "9:31:31", "remaining_time": "2 days, 6:20:37", "throughput": 5655.72, "total_tokens": 193940712}
+{"current_steps": 4435, "total_steps": 29704, "loss": 0.4155745983123779, "lr": 9.460206668227639e-06, "epoch": 0.2986129814166442, "percentage": 14.93, "elapsed_time": "9:32:10", "remaining_time": "2 days, 6:20:05", "throughput": 5655.79, "total_tokens": 194168936}
+{"current_steps": 4440, "total_steps": 29704, "loss": 0.39428043365478516, "lr": 9.459011041600895e-06, "epoch": 0.29894963641260436, "percentage": 14.95, "elapsed_time": "9:32:51", "remaining_time": "2 days, 6:19:35", "throughput": 5655.84, "total_tokens": 194397872}
+{"current_steps": 4445, "total_steps": 29704, "loss": 0.3819078207015991, "lr": 9.457814168025932e-06, "epoch": 0.2992862914085645, "percentage": 14.96, "elapsed_time": "9:33:29", "remaining_time": "2 days, 6:18:52", "throughput": 5656.0, "total_tokens": 194618856}
+{"current_steps": 4450, "total_steps": 29704, "loss": 0.4244089603424072, "lr": 9.456616047837452e-06, "epoch": 0.29962294640452464, "percentage": 14.98, "elapsed_time": "9:34:11", "remaining_time": "2 days, 6:18:33", "throughput": 5654.81, "total_tokens": 194815632}
+{"current_steps": 4455, "total_steps": 29704, "loss": 0.43105688095092776, "lr": 9.455416681370506e-06, "epoch": 0.2999596014004848, "percentage": 15.0, "elapsed_time": "9:34:51", "remaining_time": "2 days, 6:18:05", "throughput": 5654.76, "total_tokens": 195043320}
+{"current_steps": 4460, "total_steps": 29704, "loss": 0.4107870101928711, "lr": 9.45421606896049e-06, "epoch": 0.3002962563964449, "percentage": 15.01, "elapsed_time": "9:35:32", "remaining_time": "2 days, 6:17:38", "throughput": 5654.75, "total_tokens": 195273800}
+{"current_steps": 4465, "total_steps": 29704, "loss": 0.39834353923797605, "lr": 9.453014210943155e-06, "epoch": 0.30063291139240506, "percentage": 15.03, "elapsed_time": "9:36:14", "remaining_time": "2 days, 6:17:18", "throughput": 5654.19, "total_tokens": 195492912}
+{"current_steps": 4470, "total_steps": 29704, "loss": 0.39757986068725587, "lr": 9.451811107654596e-06, "epoch": 0.3009695663883652, "percentage": 15.05, "elapsed_time": "9:37:01", "remaining_time": "2 days, 6:17:26", "throughput": 5653.0, "total_tokens": 195716896}
+{"current_steps": 4475, "total_steps": 29704, "loss": 0.40091800689697266, "lr": 9.450606759431255e-06, "epoch": 0.30130622138432533, "percentage": 15.07, "elapsed_time": "9:37:36", "remaining_time": "2 days, 6:16:23", "throughput": 5653.97, "total_tokens": 195945064}
+{"current_steps": 4480, "total_steps": 29704, "loss": 0.42542014122009275, "lr": 9.449401166609928e-06, "epoch": 0.3016428763802855, "percentage": 15.08, "elapsed_time": "9:38:11", "remaining_time": "2 days, 6:15:23", "throughput": 5654.34, "total_tokens": 196155296}
+{"current_steps": 4485, "total_steps": 29704, "loss": 0.40885372161865235, "lr": 9.448194329527752e-06, "epoch": 0.3019795313762456, "percentage": 15.1, "elapsed_time": "9:38:45", "remaining_time": "2 days, 6:14:18", "throughput": 5654.89, "total_tokens": 196367168}
+{"current_steps": 4490, "total_steps": 29704, "loss": 0.38100175857543944, "lr": 9.446986248522216e-06, "epoch": 0.30231618637220575, "percentage": 15.12, "elapsed_time": "9:39:23", "remaining_time": "2 days, 6:13:38", "throughput": 5654.93, "total_tokens": 196585760}
+{"current_steps": 4495, "total_steps": 29704, "loss": 0.3696560859680176, "lr": 9.445776923931157e-06, "epoch": 0.3026528413681659, "percentage": 15.13, "elapsed_time": "9:40:01", "remaining_time": "2 days, 6:12:54", "throughput": 5654.93, "total_tokens": 196799296}
+{"current_steps": 4500, "total_steps": 29704, "loss": 0.3985780715942383, "lr": 9.444566356092754e-06, "epoch": 0.30298949636412603, "percentage": 15.15, "elapsed_time": "9:40:35", "remaining_time": "2 days, 6:11:48", "throughput": 5655.99, "total_tokens": 197027464}
+{"current_steps": 4505, "total_steps": 29704, "loss": 0.36351020336151124, "lr": 9.443354545345545e-06, "epoch": 0.30332615136008617, "percentage": 15.17, "elapsed_time": "9:41:12", "remaining_time": "2 days, 6:10:59", "throughput": 5656.27, "total_tokens": 197246488}
+{"current_steps": 4510, "total_steps": 29704, "loss": 0.3931256294250488, "lr": 9.442141492028406e-06, "epoch": 0.3036628063560463, "percentage": 15.18, "elapsed_time": "9:41:56", "remaining_time": "2 days, 6:10:54", "throughput": 5655.27, "total_tokens": 197464240}
+{"current_steps": 4515, "total_steps": 29704, "loss": 0.3859192609786987, "lr": 9.440927196480563e-06, "epoch": 0.30399946135200645, "percentage": 15.2, "elapsed_time": "9:42:32", "remaining_time": "2 days, 6:09:59", "throughput": 5656.03, "total_tokens": 197693416}
+{"current_steps": 4520, "total_steps": 29704, "loss": 0.3971871376037598, "lr": 9.439711659041593e-06, "epoch": 0.3043361163479666, "percentage": 15.22, "elapsed_time": "9:43:15", "remaining_time": "2 days, 6:09:43", "throughput": 5655.26, "total_tokens": 197908880}
+{"current_steps": 4525, "total_steps": 29704, "loss": 0.4050449371337891, "lr": 9.438494880051413e-06, "epoch": 0.30467277134392673, "percentage": 15.23, "elapsed_time": "9:43:51", "remaining_time": "2 days, 6:08:49", "throughput": 5655.79, "total_tokens": 198130240}
+{"current_steps": 4530, "total_steps": 29704, "loss": 0.39906351566314696, "lr": 9.437276859850293e-06, "epoch": 0.30500942633988687, "percentage": 15.25, "elapsed_time": "9:44:30", "remaining_time": "2 days, 6:08:11", "throughput": 5655.92, "total_tokens": 198354624}
+{"current_steps": 4535, "total_steps": 29704, "loss": 0.41982173919677734, "lr": 9.436057598778848e-06, "epoch": 0.305346081335847, "percentage": 15.27, "elapsed_time": "9:45:03", "remaining_time": "2 days, 6:07:02", "throughput": 5656.69, "total_tokens": 198569592}
+{"current_steps": 4540, "total_steps": 29704, "loss": 0.44964447021484377, "lr": 9.434837097178043e-06, "epoch": 0.30568273633180715, "percentage": 15.28, "elapsed_time": "9:45:36", "remaining_time": "2 days, 6:05:49", "throughput": 5657.75, "total_tokens": 198790848}
+{"current_steps": 4545, "total_steps": 29704, "loss": 0.4145000457763672, "lr": 9.433615355389183e-06, "epoch": 0.3060193913277673, "percentage": 15.3, "elapsed_time": "9:46:11", "remaining_time": "2 days, 6:04:50", "throughput": 5657.94, "total_tokens": 198995864}
+{"current_steps": 4550, "total_steps": 29704, "loss": 0.38867299556732177, "lr": 9.432392373753926e-06, "epoch": 0.3063560463237274, "percentage": 15.32, "elapsed_time": "9:46:44", "remaining_time": "2 days, 6:03:44", "throughput": 5659.12, "total_tokens": 199228112}
+{"current_steps": 4555, "total_steps": 29704, "loss": 0.398314380645752, "lr": 9.431168152614278e-06, "epoch": 0.30669270131968757, "percentage": 15.33, "elapsed_time": "9:47:26", "remaining_time": "2 days, 6:03:24", "throughput": 5658.5, "total_tokens": 199444448}
+{"current_steps": 4560, "total_steps": 29704, "loss": 0.36063923835754397, "lr": 9.429942692312585e-06, "epoch": 0.3070293563156477, "percentage": 15.35, "elapsed_time": "9:47:57", "remaining_time": "2 days, 6:02:01", "throughput": 5659.84, "total_tokens": 199665776}
+{"current_steps": 4565, "total_steps": 29704, "loss": 0.38531041145324707, "lr": 9.428715993191546e-06, "epoch": 0.30736601131160785, "percentage": 15.37, "elapsed_time": "9:48:38", "remaining_time": "2 days, 6:01:36", "throughput": 5659.57, "total_tokens": 199888952}
+{"current_steps": 4570, "total_steps": 29704, "loss": 0.4015589714050293, "lr": 9.427488055594199e-06, "epoch": 0.307702666307568, "percentage": 15.39, "elapsed_time": "9:49:16", "remaining_time": "2 days, 6:00:54", "throughput": 5659.31, "total_tokens": 200094448}
+{"current_steps": 4575, "total_steps": 29704, "loss": 0.40995278358459475, "lr": 9.426258879863937e-06, "epoch": 0.3080393213035281, "percentage": 15.4, "elapsed_time": "9:49:48", "remaining_time": "2 days, 5:59:40", "throughput": 5660.83, "total_tokens": 200330936}
+{"current_steps": 4580, "total_steps": 29704, "loss": 0.35400965213775637, "lr": 9.425028466344494e-06, "epoch": 0.30837597629948826, "percentage": 15.42, "elapsed_time": "9:50:23", "remaining_time": "2 days, 5:58:38", "throughput": 5661.51, "total_tokens": 200550336}
+{"current_steps": 4585, "total_steps": 29704, "loss": 0.40514564514160156, "lr": 9.42379681537995e-06, "epoch": 0.3087126312954484, "percentage": 15.44, "elapsed_time": "9:51:06", "remaining_time": "2 days, 5:58:25", "throughput": 5660.71, "total_tokens": 200767584}
+{"current_steps": 4590, "total_steps": 29704, "loss": 0.39676432609558104, "lr": 9.422563927314732e-06, "epoch": 0.30904928629140854, "percentage": 15.45, "elapsed_time": "9:51:43", "remaining_time": "2 days, 5:57:34", "throughput": 5661.18, "total_tokens": 200990368}
+{"current_steps": 4595, "total_steps": 29704, "loss": 0.3798659801483154, "lr": 9.421329802493615e-06, "epoch": 0.3093859412873687, "percentage": 15.47, "elapsed_time": "9:52:25", "remaining_time": "2 days, 5:57:16", "throughput": 5660.69, "total_tokens": 201213136}
+{"current_steps": 4600, "total_steps": 29704, "loss": 0.4128886222839355, "lr": 9.420094441261717e-06, "epoch": 0.3097225962833288, "percentage": 15.49, "elapsed_time": "9:53:11", "remaining_time": "2 days, 5:57:14", "throughput": 5659.74, "total_tokens": 201436008}
+{"current_steps": 4605, "total_steps": 29704, "loss": 0.3874027252197266, "lr": 9.418857843964506e-06, "epoch": 0.31005925127928896, "percentage": 15.5, "elapsed_time": "9:53:53", "remaining_time": "2 days, 5:56:54", "throughput": 5659.09, "total_tokens": 201651024}
+{"current_steps": 4610, "total_steps": 29704, "loss": 0.41231350898742675, "lr": 9.417620010947786e-06, "epoch": 0.3103959062752491, "percentage": 15.52, "elapsed_time": "9:54:32", "remaining_time": "2 days, 5:56:20", "throughput": 5658.72, "total_tokens": 201862568}
+{"current_steps": 4615, "total_steps": 29704, "loss": 0.3759624719619751, "lr": 9.416380942557719e-06, "epoch": 0.31073256127120924, "percentage": 15.54, "elapsed_time": "9:55:11", "remaining_time": "2 days, 5:55:43", "throughput": 5658.69, "total_tokens": 202081848}
+{"current_steps": 4620, "total_steps": 29704, "loss": 0.40959873199462893, "lr": 9.415140639140803e-06, "epoch": 0.3110692162671694, "percentage": 15.55, "elapsed_time": "9:55:48", "remaining_time": "2 days, 5:54:52", "throughput": 5658.7, "total_tokens": 202288416}
+{"current_steps": 4625, "total_steps": 29704, "loss": 0.4299426555633545, "lr": 9.413899101043887e-06, "epoch": 0.3114058712631295, "percentage": 15.57, "elapsed_time": "9:56:27", "remaining_time": "2 days, 5:54:15", "throughput": 5658.65, "total_tokens": 202507280}
+{"current_steps": 4630, "total_steps": 29704, "loss": 0.4051331043243408, "lr": 9.412656328614162e-06, "epoch": 0.31174252625908966, "percentage": 15.59, "elapsed_time": "9:57:06", "remaining_time": "2 days, 5:53:40", "throughput": 5658.07, "total_tokens": 202708520}
+{"current_steps": 4635, "total_steps": 29704, "loss": 0.40700321197509765, "lr": 9.411412322199165e-06, "epoch": 0.3120791812550498, "percentage": 15.6, "elapsed_time": "9:57:40", "remaining_time": "2 days, 5:52:34", "throughput": 5659.14, "total_tokens": 202938288}
+{"current_steps": 4640, "total_steps": 29704, "loss": 0.397929310798645, "lr": 9.41016708214678e-06, "epoch": 0.31241583625100994, "percentage": 15.62, "elapsed_time": "9:58:20", "remaining_time": "2 days, 5:52:06", "throughput": 5658.38, "total_tokens": 203140768}
+{"current_steps": 4645, "total_steps": 29704, "loss": 0.3729093074798584, "lr": 9.408920608805233e-06, "epoch": 0.31275249124697013, "percentage": 15.64, "elapsed_time": "9:59:03", "remaining_time": "2 days, 5:51:50", "throughput": 5657.55, "total_tokens": 203353200}
+{"current_steps": 4650, "total_steps": 29704, "loss": 0.42061967849731446, "lr": 9.407672902523097e-06, "epoch": 0.31308914624293027, "percentage": 15.65, "elapsed_time": "9:59:49", "remaining_time": "2 days, 5:51:52", "throughput": 5656.5, "total_tokens": 203576960}
+{"current_steps": 4655, "total_steps": 29704, "loss": 0.39535982608795167, "lr": 9.406423963649287e-06, "epoch": 0.3134258012388904, "percentage": 15.67, "elapsed_time": "10:00:33", "remaining_time": "2 days, 5:51:39", "throughput": 5655.33, "total_tokens": 203780664}
+{"current_steps": 4660, "total_steps": 29704, "loss": 0.3896958827972412, "lr": 9.405173792533069e-06, "epoch": 0.31376245623485055, "percentage": 15.69, "elapsed_time": "10:01:11", "remaining_time": "2 days, 5:50:59", "throughput": 5655.44, "total_tokens": 204002504}
+{"current_steps": 4665, "total_steps": 29704, "loss": 0.383258056640625, "lr": 9.403922389524045e-06, "epoch": 0.3140991112308107, "percentage": 15.7, "elapsed_time": "10:01:47", "remaining_time": "2 days, 5:50:02", "throughput": 5655.9, "total_tokens": 204219216}
+{"current_steps": 4670, "total_steps": 29704, "loss": 0.41370835304260256, "lr": 9.40266975497217e-06, "epoch": 0.31443576622677083, "percentage": 15.72, "elapsed_time": "10:02:26", "remaining_time": "2 days, 5:49:24", "throughput": 5655.79, "total_tokens": 204435080}
+{"current_steps": 4675, "total_steps": 29704, "loss": 0.4065183162689209, "lr": 9.401415889227736e-06, "epoch": 0.31477242122273097, "percentage": 15.74, "elapsed_time": "10:03:09", "remaining_time": "2 days, 5:49:13", "throughput": 5655.33, "total_tokens": 204665984}
+{"current_steps": 4680, "total_steps": 29704, "loss": 0.40553603172302244, "lr": 9.400160792641385e-06, "epoch": 0.3151090762186911, "percentage": 15.76, "elapsed_time": "10:03:50", "remaining_time": "2 days, 5:48:43", "throughput": 5655.5, "total_tokens": 204900584}
+{"current_steps": 4685, "total_steps": 29704, "loss": 0.3963156700134277, "lr": 9.398904465564097e-06, "epoch": 0.31544573121465125, "percentage": 15.77, "elapsed_time": "10:04:21", "remaining_time": "2 days, 5:47:24", "throughput": 5656.74, "total_tokens": 205121896}
+{"current_steps": 4690, "total_steps": 29704, "loss": 0.3701439142227173, "lr": 9.397646908347203e-06, "epoch": 0.3157823862106114, "percentage": 15.79, "elapsed_time": "10:04:59", "remaining_time": "2 days, 5:46:44", "throughput": 5656.71, "total_tokens": 205337424}
+{"current_steps": 4695, "total_steps": 29704, "loss": 0.37079999446868894, "lr": 9.396388121342375e-06, "epoch": 0.3161190412065715, "percentage": 15.81, "elapsed_time": "10:05:37", "remaining_time": "2 days, 5:45:58", "throughput": 5657.19, "total_tokens": 205566880}
+{"current_steps": 4700, "total_steps": 29704, "loss": 0.382697606086731, "lr": 9.395128104901628e-06, "epoch": 0.31645569620253167, "percentage": 15.82, "elapsed_time": "10:06:13", "remaining_time": "2 days, 5:45:04", "throughput": 5657.5, "total_tokens": 205780336}
+{"current_steps": 4705, "total_steps": 29704, "loss": 0.4276883125305176, "lr": 9.393866859377321e-06, "epoch": 0.3167923511984918, "percentage": 15.84, "elapsed_time": "10:06:49", "remaining_time": "2 days, 5:44:15", "throughput": 5657.95, "total_tokens": 206004992}
+{"current_steps": 4710, "total_steps": 29704, "loss": 0.39430766105651854, "lr": 9.392604385122157e-06, "epoch": 0.31712900619445195, "percentage": 15.86, "elapsed_time": "10:07:21", "remaining_time": "2 days, 5:42:58", "throughput": 5659.46, "total_tokens": 206237344}
+{"current_steps": 4715, "total_steps": 29704, "loss": 0.3808866024017334, "lr": 9.391340682489185e-06, "epoch": 0.3174656611904121, "percentage": 15.87, "elapsed_time": "10:08:00", "remaining_time": "2 days, 5:42:20", "throughput": 5659.23, "total_tokens": 206449496}
+{"current_steps": 4720, "total_steps": 29704, "loss": 0.39603471755981445, "lr": 9.390075751831794e-06, "epoch": 0.3178023161863722, "percentage": 15.89, "elapsed_time": "10:08:33", "remaining_time": "2 days, 5:41:14", "throughput": 5660.3, "total_tokens": 206677304}
+{"current_steps": 4725, "total_steps": 29704, "loss": 0.41028714179992676, "lr": 9.388809593503718e-06, "epoch": 0.31813897118233236, "percentage": 15.91, "elapsed_time": "10:09:14", "remaining_time": "2 days, 5:40:47", "throughput": 5660.11, "total_tokens": 206901904}
+{"current_steps": 4730, "total_steps": 29704, "loss": 0.428537654876709, "lr": 9.387542207859034e-06, "epoch": 0.3184756261782925, "percentage": 15.92, "elapsed_time": "10:09:44", "remaining_time": "2 days, 5:39:23", "throughput": 5661.46, "total_tokens": 207122320}
+{"current_steps": 4735, "total_steps": 29704, "loss": 0.41155662536621096, "lr": 9.386273595252161e-06, "epoch": 0.31881228117425264, "percentage": 15.94, "elapsed_time": "10:10:20", "remaining_time": "2 days, 5:38:30", "throughput": 5662.2, "total_tokens": 207352296}
+{"current_steps": 4740, "total_steps": 29704, "loss": 0.4223154067993164, "lr": 9.385003756037865e-06, "epoch": 0.3191489361702128, "percentage": 15.96, "elapsed_time": "10:11:02", "remaining_time": "2 days, 5:38:11", "throughput": 5661.72, "total_tokens": 207575072}
+{"current_steps": 4745, "total_steps": 29704, "loss": 0.40798358917236327, "lr": 9.383732690571253e-06, "epoch": 0.3194855911661729, "percentage": 15.97, "elapsed_time": "10:11:37", "remaining_time": "2 days, 5:37:09", "throughput": 5663.02, "total_tokens": 207817360}
+{"current_steps": 4750, "total_steps": 29704, "loss": 0.40032100677490234, "lr": 9.382460399207769e-06, "epoch": 0.31982224616213306, "percentage": 15.99, "elapsed_time": "10:12:19", "remaining_time": "2 days, 5:36:50", "throughput": 5662.7, "total_tokens": 208045464}
+{"current_steps": 4755, "total_steps": 29704, "loss": 0.3821758985519409, "lr": 9.381186882303212e-06, "epoch": 0.3201589011580932, "percentage": 16.01, "elapsed_time": "10:12:55", "remaining_time": "2 days, 5:35:56", "throughput": 5662.64, "total_tokens": 208245512}
+{"current_steps": 4760, "total_steps": 29704, "loss": 0.38662168979644773, "lr": 9.379912140213713e-06, "epoch": 0.32049555615405334, "percentage": 16.02, "elapsed_time": "10:13:30", "remaining_time": "2 days, 5:34:59", "throughput": 5663.02, "total_tokens": 208458376}
+{"current_steps": 4765, "total_steps": 29704, "loss": 0.37845721244812014, "lr": 9.37863617329575e-06, "epoch": 0.3208322111500135, "percentage": 16.04, "elapsed_time": "10:14:08", "remaining_time": "2 days, 5:34:18", "throughput": 5662.68, "total_tokens": 208662464}
+{"current_steps": 4770, "total_steps": 29704, "loss": 0.3855098485946655, "lr": 9.377358981906145e-06, "epoch": 0.3211688661459736, "percentage": 16.06, "elapsed_time": "10:14:45", "remaining_time": "2 days, 5:33:31", "throughput": 5662.98, "total_tokens": 208883752}
+{"current_steps": 4775, "total_steps": 29704, "loss": 0.41453986167907714, "lr": 9.376080566402059e-06, "epoch": 0.32150552114193376, "percentage": 16.08, "elapsed_time": "10:15:23", "remaining_time": "2 days, 5:32:49", "throughput": 5663.39, "total_tokens": 209114544}
+{"current_steps": 4780, "total_steps": 29704, "loss": 0.33689126968383787, "lr": 9.374800927140994e-06, "epoch": 0.3218421761378939, "percentage": 16.09, "elapsed_time": "10:15:59", "remaining_time": "2 days, 5:31:52", "throughput": 5664.18, "total_tokens": 209342880}
+{"current_steps": 4785, "total_steps": 29704, "loss": 0.3944025278091431, "lr": 9.373520064480804e-06, "epoch": 0.32217883113385404, "percentage": 16.11, "elapsed_time": "10:16:39", "remaining_time": "2 days, 5:31:23", "throughput": 5664.04, "total_tokens": 209567016}
+{"current_steps": 4790, "total_steps": 29704, "loss": 0.3935252666473389, "lr": 9.372237978779672e-06, "epoch": 0.3225154861298142, "percentage": 16.13, "elapsed_time": "10:17:15", "remaining_time": "2 days, 5:30:32", "throughput": 5664.74, "total_tokens": 209798224}
+{"current_steps": 4795, "total_steps": 29704, "loss": 0.3892039775848389, "lr": 9.37095467039613e-06, "epoch": 0.3228521411257743, "percentage": 16.14, "elapsed_time": "10:17:54", "remaining_time": "2 days, 5:29:51", "throughput": 5664.3, "total_tokens": 209998416}
+{"current_steps": 4800, "total_steps": 29704, "loss": 0.3860762119293213, "lr": 9.369670139689056e-06, "epoch": 0.32318879612173446, "percentage": 16.16, "elapsed_time": "10:18:33", "remaining_time": "2 days, 5:29:17", "throughput": 5664.04, "total_tokens": 210212136}
+{"current_steps": 4805, "total_steps": 29704, "loss": 0.377323055267334, "lr": 9.368384387017659e-06, "epoch": 0.3235254511176946, "percentage": 16.18, "elapsed_time": "10:19:10", "remaining_time": "2 days, 5:28:27", "throughput": 5663.87, "total_tokens": 210412912}
+{"current_steps": 4810, "total_steps": 29704, "loss": 0.3832876443862915, "lr": 9.367097412741497e-06, "epoch": 0.32386210611365474, "percentage": 16.19, "elapsed_time": "10:19:45", "remaining_time": "2 days, 5:27:32", "throughput": 5664.21, "total_tokens": 210626632}
+{"current_steps": 4815, "total_steps": 29704, "loss": 0.41832733154296875, "lr": 9.36580921722047e-06, "epoch": 0.3241987611096149, "percentage": 16.21, "elapsed_time": "10:20:26", "remaining_time": "2 days, 5:27:04", "throughput": 5663.93, "total_tokens": 210847304}
+{"current_steps": 4820, "total_steps": 29704, "loss": 0.3626012563705444, "lr": 9.364519800814818e-06, "epoch": 0.324535416105575, "percentage": 16.23, "elapsed_time": "10:21:09", "remaining_time": "2 days, 5:26:50", "throughput": 5663.36, "total_tokens": 211071840}
+{"current_steps": 4825, "total_steps": 29704, "loss": 0.3882687330245972, "lr": 9.36322916388512e-06, "epoch": 0.32487207110153515, "percentage": 16.24, "elapsed_time": "10:21:51", "remaining_time": "2 days, 5:26:27", "throughput": 5663.07, "total_tokens": 211297288}
+{"current_steps": 4830, "total_steps": 29704, "loss": 0.3612262487411499, "lr": 9.3619373067923e-06, "epoch": 0.3252087260974953, "percentage": 16.26, "elapsed_time": "10:22:28", "remaining_time": "2 days, 5:25:40", "throughput": 5663.69, "total_tokens": 211529896}
+{"current_steps": 4835, "total_steps": 29704, "loss": 0.39133119583129883, "lr": 9.360644229897622e-06, "epoch": 0.32554538109345543, "percentage": 16.28, "elapsed_time": "10:23:10", "remaining_time": "2 days, 5:25:19", "throughput": 5663.03, "total_tokens": 211744104}
+{"current_steps": 4840, "total_steps": 29704, "loss": 0.4070117950439453, "lr": 9.35934993356269e-06, "epoch": 0.32588203608941557, "percentage": 16.29, "elapsed_time": "10:23:54", "remaining_time": "2 days, 5:25:07", "throughput": 5662.2, "total_tokens": 211960664}
+{"current_steps": 4845, "total_steps": 29704, "loss": 0.37361793518066405, "lr": 9.35805441814945e-06, "epoch": 0.3262186910853757, "percentage": 16.31, "elapsed_time": "10:24:41", "remaining_time": "2 days, 5:25:11", "throughput": 5661.09, "total_tokens": 212184888}
+{"current_steps": 4850, "total_steps": 29704, "loss": 0.40865292549133303, "lr": 9.356757684020188e-06, "epoch": 0.32655534608133585, "percentage": 16.33, "elapsed_time": "10:25:21", "remaining_time": "2 days, 5:24:38", "throughput": 5660.46, "total_tokens": 212387000}
+{"current_steps": 4855, "total_steps": 29704, "loss": 0.37503221035003664, "lr": 9.355459731537533e-06, "epoch": 0.326892001077296, "percentage": 16.34, "elapsed_time": "10:26:03", "remaining_time": "2 days, 5:24:19", "throughput": 5659.78, "total_tokens": 212602240}
+{"current_steps": 4860, "total_steps": 29704, "loss": 0.4150728225708008, "lr": 9.354160561064451e-06, "epoch": 0.32722865607325613, "percentage": 16.36, "elapsed_time": "10:26:40", "remaining_time": "2 days, 5:23:32", "throughput": 5660.24, "total_tokens": 212829424}
+{"current_steps": 4865, "total_steps": 29704, "loss": 0.41104774475097655, "lr": 9.352860172964254e-06, "epoch": 0.32756531106921627, "percentage": 16.38, "elapsed_time": "10:27:19", "remaining_time": "2 days, 5:22:54", "throughput": 5660.51, "total_tokens": 213059616}
+{"current_steps": 4870, "total_steps": 29704, "loss": 0.37230610847473145, "lr": 9.35155856760059e-06, "epoch": 0.3279019660651764, "percentage": 16.4, "elapsed_time": "10:28:03", "remaining_time": "2 days, 5:22:43", "throughput": 5660.06, "total_tokens": 213291344}
+{"current_steps": 4875, "total_steps": 29704, "loss": 0.4028165817260742, "lr": 9.35025574533745e-06, "epoch": 0.32823862106113655, "percentage": 16.41, "elapsed_time": "10:28:36", "remaining_time": "2 days, 5:21:36", "throughput": 5661.11, "total_tokens": 213518768}
+{"current_steps": 4880, "total_steps": 29704, "loss": 0.37210307121276853, "lr": 9.34895170653916e-06, "epoch": 0.3285752760570967, "percentage": 16.43, "elapsed_time": "10:29:21", "remaining_time": "2 days, 5:21:26", "throughput": 5660.37, "total_tokens": 213742008}
+{"current_steps": 4885, "total_steps": 29704, "loss": 0.3780977725982666, "lr": 9.347646451570394e-06, "epoch": 0.3289119310530568, "percentage": 16.45, "elapsed_time": "10:29:55", "remaining_time": "2 days, 5:20:27", "throughput": 5661.15, "total_tokens": 213967120}
+{"current_steps": 4890, "total_steps": 29704, "loss": 0.44457039833068845, "lr": 9.346339980796162e-06, "epoch": 0.32924858604901697, "percentage": 16.46, "elapsed_time": "10:30:30", "remaining_time": "2 days, 5:19:25", "throughput": 5662.28, "total_tokens": 214204152}
+{"current_steps": 4895, "total_steps": 29704, "loss": 0.3991098403930664, "lr": 9.345032294581813e-06, "epoch": 0.3295852410449771, "percentage": 16.48, "elapsed_time": "10:31:01", "remaining_time": "2 days, 5:18:08", "throughput": 5663.51, "total_tokens": 214426328}
+{"current_steps": 4900, "total_steps": 29704, "loss": 0.37960023880004884, "lr": 9.343723393293038e-06, "epoch": 0.32992189604093725, "percentage": 16.5, "elapsed_time": "10:31:40", "remaining_time": "2 days, 5:17:34", "throughput": 5663.28, "total_tokens": 214641400}
+{"current_steps": 4905, "total_steps": 29704, "loss": 0.3947901725769043, "lr": 9.342413277295869e-06, "epoch": 0.3302585510368974, "percentage": 16.51, "elapsed_time": "10:32:21", "remaining_time": "2 days, 5:17:08", "throughput": 5663.02, "total_tokens": 214865368}
+{"current_steps": 4910, "total_steps": 29704, "loss": 0.39415407180786133, "lr": 9.341101946956672e-06, "epoch": 0.3305952060328575, "percentage": 16.53, "elapsed_time": "10:32:56", "remaining_time": "2 days, 5:16:08", "throughput": 5663.47, "total_tokens": 215077016}
+{"current_steps": 4915, "total_steps": 29704, "loss": 0.4069984436035156, "lr": 9.33978940264216e-06, "epoch": 0.33093186102881766, "percentage": 16.55, "elapsed_time": "10:33:40", "remaining_time": "2 days, 5:16:00", "throughput": 5662.18, "total_tokens": 215281168}
+{"current_steps": 4920, "total_steps": 29704, "loss": 0.396682071685791, "lr": 9.338475644719377e-06, "epoch": 0.3312685160247778, "percentage": 16.56, "elapsed_time": "10:34:20", "remaining_time": "2 days, 5:15:26", "throughput": 5662.25, "total_tokens": 215508928}
+{"current_steps": 4925, "total_steps": 29704, "loss": 0.3950789451599121, "lr": 9.337160673555719e-06, "epoch": 0.33160517102073794, "percentage": 16.58, "elapsed_time": "10:34:56", "remaining_time": "2 days, 5:14:34", "throughput": 5662.7, "total_tokens": 215730176}
+{"current_steps": 4930, "total_steps": 29704, "loss": 0.3960768222808838, "lr": 9.335844489518905e-06, "epoch": 0.3319418260166981, "percentage": 16.6, "elapsed_time": "10:35:35", "remaining_time": "2 days, 5:13:57", "throughput": 5662.47, "total_tokens": 215941544}
+{"current_steps": 4935, "total_steps": 29704, "loss": 0.3927053928375244, "lr": 9.334527092977008e-06, "epoch": 0.3322784810126582, "percentage": 16.61, "elapsed_time": "10:36:12", "remaining_time": "2 days, 5:13:10", "throughput": 5662.15, "total_tokens": 216139784}
+{"current_steps": 4940, "total_steps": 29704, "loss": 0.3905383825302124, "lr": 9.33320848429843e-06, "epoch": 0.33261513600861836, "percentage": 16.63, "elapsed_time": "10:36:53", "remaining_time": "2 days, 5:12:44", "throughput": 5662.27, "total_tokens": 216376872}
+{"current_steps": 4945, "total_steps": 29704, "loss": 0.41385631561279296, "lr": 9.331888663851917e-06, "epoch": 0.3329517910045785, "percentage": 16.65, "elapsed_time": "10:37:27", "remaining_time": "2 days, 5:11:41", "throughput": 5663.14, "total_tokens": 216602432}
+{"current_steps": 4950, "total_steps": 29704, "loss": 0.402661657333374, "lr": 9.330567632006552e-06, "epoch": 0.33328844600053864, "percentage": 16.66, "elapsed_time": "10:38:02", "remaining_time": "2 days, 5:10:43", "throughput": 5663.67, "total_tokens": 216820040}
+{"current_steps": 4955, "total_steps": 29704, "loss": 0.4041293144226074, "lr": 9.329245389131759e-06, "epoch": 0.3336251009964988, "percentage": 16.68, "elapsed_time": "10:38:41", "remaining_time": "2 days, 5:10:04", "throughput": 5663.74, "total_tokens": 217040648}
+{"current_steps": 4960, "total_steps": 29704, "loss": 0.373291015625, "lr": 9.327921935597298e-06, "epoch": 0.3339617559924589, "percentage": 16.7, "elapsed_time": "10:39:22", "remaining_time": "2 days, 5:09:41", "throughput": 5663.11, "total_tokens": 217253544}
+{"current_steps": 4965, "total_steps": 29704, "loss": 0.39936108589172364, "lr": 9.326597271773267e-06, "epoch": 0.33429841098841906, "percentage": 16.71, "elapsed_time": "10:40:04", "remaining_time": "2 days, 5:09:19", "throughput": 5662.1, "total_tokens": 217452896}
+{"current_steps": 4970, "total_steps": 29704, "loss": 0.3928532600402832, "lr": 9.325271398030107e-06, "epoch": 0.3346350659843792, "percentage": 16.73, "elapsed_time": "10:40:41", "remaining_time": "2 days, 5:08:32", "throughput": 5662.2, "total_tokens": 217666152}
+{"current_steps": 4975, "total_steps": 29704, "loss": 0.40926666259765626, "lr": 9.323944314738591e-06, "epoch": 0.33497172098033934, "percentage": 16.75, "elapsed_time": "10:41:26", "remaining_time": "2 days, 5:08:23", "throughput": 5661.36, "total_tokens": 217886336}
+{"current_steps": 4980, "total_steps": 29704, "loss": 0.3675239562988281, "lr": 9.322616022269838e-06, "epoch": 0.3353083759762995, "percentage": 16.77, "elapsed_time": "10:42:03", "remaining_time": "2 days, 5:07:33", "throughput": 5661.36, "total_tokens": 218092776}
+{"current_steps": 4985, "total_steps": 29704, "loss": 0.40448784828186035, "lr": 9.321286520995297e-06, "epoch": 0.3356450309722596, "percentage": 16.78, "elapsed_time": "10:42:43", "remaining_time": "2 days, 5:07:01", "throughput": 5660.94, "total_tokens": 218303136}
+{"current_steps": 4990, "total_steps": 29704, "loss": 0.38973236083984375, "lr": 9.319955811286758e-06, "epoch": 0.33598168596821976, "percentage": 16.8, "elapsed_time": "10:43:26", "remaining_time": "2 days, 5:06:48", "throughput": 5660.18, "total_tokens": 218521848}
+{"current_steps": 4995, "total_steps": 29704, "loss": 0.38318960666656493, "lr": 9.318623893516354e-06, "epoch": 0.3363183409641799, "percentage": 16.82, "elapsed_time": "10:44:03", "remaining_time": "2 days, 5:06:00", "throughput": 5660.27, "total_tokens": 218733712}
+{"current_steps": 5000, "total_steps": 29704, "loss": 0.3933337450027466, "lr": 9.317290768056548e-06, "epoch": 0.33665499596014004, "percentage": 16.83, "elapsed_time": "10:44:45", "remaining_time": "2 days, 5:05:38", "throughput": 5659.52, "total_tokens": 218942384}
+{"current_steps": 5005, "total_steps": 29704, "loss": 0.40426931381225584, "lr": 9.315956435280147e-06, "epoch": 0.3369916509561002, "percentage": 16.85, "elapsed_time": "10:45:26", "remaining_time": "2 days, 5:05:11", "throughput": 5658.82, "total_tokens": 219148584}
+{"current_steps": 5010, "total_steps": 29704, "loss": 0.3823359251022339, "lr": 9.314620895560292e-06, "epoch": 0.3373283059520603, "percentage": 16.87, "elapsed_time": "10:45:59", "remaining_time": "2 days, 5:04:05", "throughput": 5659.62, "total_tokens": 219366640}
+{"current_steps": 5015, "total_steps": 29704, "loss": 0.4273562431335449, "lr": 9.313284149270459e-06, "epoch": 0.33766496094802045, "percentage": 16.88, "elapsed_time": "10:46:37", "remaining_time": "2 days, 5:03:19", "throughput": 5659.71, "total_tokens": 219580136}
+{"current_steps": 5020, "total_steps": 29704, "loss": 0.396860933303833, "lr": 9.311946196784469e-06, "epoch": 0.3380016159439806, "percentage": 16.9, "elapsed_time": "10:47:18", "remaining_time": "2 days, 5:02:54", "throughput": 5659.19, "total_tokens": 219795256}
+{"current_steps": 5025, "total_steps": 29704, "loss": 0.40080647468566893, "lr": 9.310607038476476e-06, "epoch": 0.33833827093994073, "percentage": 16.92, "elapsed_time": "10:48:02", "remaining_time": "2 days, 5:02:40", "throughput": 5658.27, "total_tokens": 220007056}
+{"current_steps": 5030, "total_steps": 29704, "loss": 0.3821887016296387, "lr": 9.30926667472097e-06, "epoch": 0.3386749259359009, "percentage": 16.93, "elapsed_time": "10:48:49", "remaining_time": "2 days, 5:02:45", "throughput": 5657.08, "total_tokens": 220230040}
+{"current_steps": 5035, "total_steps": 29704, "loss": 0.4106111526489258, "lr": 9.307925105892779e-06, "epoch": 0.339011580931861, "percentage": 16.95, "elapsed_time": "10:49:24", "remaining_time": "2 days, 5:01:47", "throughput": 5658.02, "total_tokens": 220463304}
+{"current_steps": 5040, "total_steps": 29704, "loss": 0.38834619522094727, "lr": 9.30658233236707e-06, "epoch": 0.33934823592782115, "percentage": 16.97, "elapsed_time": "10:50:07", "remaining_time": "2 days, 5:01:27", "throughput": 5657.81, "total_tokens": 220695456}
+{"current_steps": 5045, "total_steps": 29704, "loss": 0.3923861265182495, "lr": 9.30523835451934e-06, "epoch": 0.3396848909237813, "percentage": 16.98, "elapsed_time": "10:50:53", "remaining_time": "2 days, 5:01:25", "throughput": 5656.74, "total_tokens": 220915344}
+{"current_steps": 5050, "total_steps": 29704, "loss": 0.4306797981262207, "lr": 9.303893172725437e-06, "epoch": 0.34002154591974143, "percentage": 17.0, "elapsed_time": "10:51:31", "remaining_time": "2 days, 5:00:42", "throughput": 5657.18, "total_tokens": 221146072}
+{"current_steps": 5055, "total_steps": 29704, "loss": 0.38288087844848634, "lr": 9.302546787361533e-06, "epoch": 0.34035820091570157, "percentage": 17.02, "elapsed_time": "10:52:13", "remaining_time": "2 days, 5:00:22", "throughput": 5656.79, "total_tokens": 221371344}
+{"current_steps": 5060, "total_steps": 29704, "loss": 0.37805628776550293, "lr": 9.301199198804139e-06, "epoch": 0.3406948559116617, "percentage": 17.03, "elapsed_time": "10:52:52", "remaining_time": "2 days, 4:59:45", "throughput": 5656.57, "total_tokens": 221583280}
+{"current_steps": 5065, "total_steps": 29704, "loss": 0.4092425346374512, "lr": 9.299850407430102e-06, "epoch": 0.34103151090762185, "percentage": 17.05, "elapsed_time": "10:53:35", "remaining_time": "2 days, 4:59:24", "throughput": 5656.22, "total_tokens": 221808808}
+{"current_steps": 5070, "total_steps": 29704, "loss": 0.3920405864715576, "lr": 9.298500413616613e-06, "epoch": 0.341368165903582, "percentage": 17.07, "elapsed_time": "10:54:11", "remaining_time": "2 days, 4:58:32", "throughput": 5656.66, "total_tokens": 222029984}
+{"current_steps": 5075, "total_steps": 29704, "loss": 0.3707792043685913, "lr": 9.297149217741188e-06, "epoch": 0.3417048208995421, "percentage": 17.09, "elapsed_time": "10:54:48", "remaining_time": "2 days, 4:57:48", "throughput": 5656.85, "total_tokens": 222251080}
+{"current_steps": 5080, "total_steps": 29704, "loss": 0.4238715171813965, "lr": 9.295796820181688e-06, "epoch": 0.34204147589550227, "percentage": 17.1, "elapsed_time": "10:55:26", "remaining_time": "2 days, 4:57:02", "throughput": 5657.2, "total_tokens": 222475224}
+{"current_steps": 5085, "total_steps": 29704, "loss": 0.3979609727859497, "lr": 9.294443221316305e-06, "epoch": 0.3423781308914624, "percentage": 17.12, "elapsed_time": "10:56:02", "remaining_time": "2 days, 4:56:15", "throughput": 5657.35, "total_tokens": 222689712}
+{"current_steps": 5090, "total_steps": 29704, "loss": 0.41632776260375975, "lr": 9.293088421523569e-06, "epoch": 0.34271478588742255, "percentage": 17.14, "elapsed_time": "10:56:39", "remaining_time": "2 days, 4:55:24", "throughput": 5657.57, "total_tokens": 222903112}
+{"current_steps": 5095, "total_steps": 29704, "loss": 0.3804980754852295, "lr": 9.291732421182346e-06, "epoch": 0.3430514408833827, "percentage": 17.15, "elapsed_time": "10:57:24", "remaining_time": "2 days, 4:55:17", "throughput": 5656.49, "total_tokens": 223116536}
+{"current_steps": 5100, "total_steps": 29704, "loss": 0.3959723949432373, "lr": 9.290375220671837e-06, "epoch": 0.3433880958793428, "percentage": 17.17, "elapsed_time": "10:58:03", "remaining_time": "2 days, 4:54:39", "throughput": 5656.35, "total_tokens": 223331328}
+{"current_steps": 5105, "total_steps": 29704, "loss": 0.3575240135192871, "lr": 9.28901682037158e-06, "epoch": 0.34372475087530296, "percentage": 17.19, "elapsed_time": "10:58:40", "remaining_time": "2 days, 4:53:53", "throughput": 5656.45, "total_tokens": 223545784}
+{"current_steps": 5110, "total_steps": 29704, "loss": 0.41797566413879395, "lr": 9.287657220661442e-06, "epoch": 0.3440614058712631, "percentage": 17.2, "elapsed_time": "10:59:16", "remaining_time": "2 days, 4:53:01", "throughput": 5657.04, "total_tokens": 223772168}
+{"current_steps": 5115, "total_steps": 29704, "loss": 0.3681803703308105, "lr": 9.286296421921636e-06, "epoch": 0.34439806086722324, "percentage": 17.22, "elapsed_time": "10:59:57", "remaining_time": "2 days, 4:52:33", "throughput": 5656.63, "total_tokens": 223987880}
+{"current_steps": 5120, "total_steps": 29704, "loss": 0.4501188278198242, "lr": 9.284934424532706e-06, "epoch": 0.3447347158631834, "percentage": 17.24, "elapsed_time": "11:00:37", "remaining_time": "2 days, 4:52:02", "throughput": 5656.19, "total_tokens": 224197912}
+{"current_steps": 5125, "total_steps": 29704, "loss": 0.35322113037109376, "lr": 9.283571228875525e-06, "epoch": 0.3450713708591435, "percentage": 17.25, "elapsed_time": "11:01:16", "remaining_time": "2 days, 4:51:23", "throughput": 5655.74, "total_tokens": 224398888}
+{"current_steps": 5130, "total_steps": 29704, "loss": 0.4238561153411865, "lr": 9.282206835331311e-06, "epoch": 0.3454080258551037, "percentage": 17.27, "elapsed_time": "11:01:54", "remaining_time": "2 days, 4:50:42", "throughput": 5655.99, "total_tokens": 224624840}
+{"current_steps": 5135, "total_steps": 29704, "loss": 0.3581863880157471, "lr": 9.28084124428161e-06, "epoch": 0.34574468085106386, "percentage": 17.29, "elapsed_time": "11:02:40", "remaining_time": "2 days, 4:50:36", "throughput": 5654.93, "total_tokens": 224840616}
+{"current_steps": 5140, "total_steps": 29704, "loss": 0.37324137687683107, "lr": 9.279474456108305e-06, "epoch": 0.346081335847024, "percentage": 17.3, "elapsed_time": "11:03:15", "remaining_time": "2 days, 4:49:42", "throughput": 5655.81, "total_tokens": 225076296}
+{"current_steps": 5145, "total_steps": 29704, "loss": 0.36896820068359376, "lr": 9.278106471193615e-06, "epoch": 0.34641799084298414, "percentage": 17.32, "elapsed_time": "11:03:53", "remaining_time": "2 days, 4:49:00", "throughput": 5655.88, "total_tokens": 225294528}
+{"current_steps": 5150, "total_steps": 29704, "loss": 0.4088866710662842, "lr": 9.276737289920093e-06, "epoch": 0.3467546458389443, "percentage": 17.34, "elapsed_time": "11:04:35", "remaining_time": "2 days, 4:48:38", "throughput": 5655.37, "total_tokens": 225512648}
+{"current_steps": 5155, "total_steps": 29704, "loss": 0.4070155143737793, "lr": 9.275366912670624e-06, "epoch": 0.3470913008349044, "percentage": 17.35, "elapsed_time": "11:05:10", "remaining_time": "2 days, 4:47:41", "throughput": 5655.72, "total_tokens": 225724000}
+{"current_steps": 5160, "total_steps": 29704, "loss": 0.3892085075378418, "lr": 9.273995339828432e-06, "epoch": 0.34742795583086455, "percentage": 17.37, "elapsed_time": "11:05:50", "remaining_time": "2 days, 4:47:05", "throughput": 5655.54, "total_tokens": 225939192}
+{"current_steps": 5165, "total_steps": 29704, "loss": 0.42869272232055666, "lr": 9.272622571777072e-06, "epoch": 0.3477646108268247, "percentage": 17.39, "elapsed_time": "11:06:26", "remaining_time": "2 days, 4:46:14", "throughput": 5656.01, "total_tokens": 226161472}
+{"current_steps": 5170, "total_steps": 29704, "loss": 0.3663766860961914, "lr": 9.271248608900434e-06, "epoch": 0.34810126582278483, "percentage": 17.41, "elapsed_time": "11:07:02", "remaining_time": "2 days, 4:45:26", "throughput": 5656.48, "total_tokens": 226388672}
+{"current_steps": 5175, "total_steps": 29704, "loss": 0.3990020275115967, "lr": 9.269873451582741e-06, "epoch": 0.348437920818745, "percentage": 17.42, "elapsed_time": "11:07:45", "remaining_time": "2 days, 4:45:08", "throughput": 5655.8, "total_tokens": 226604448}
+{"current_steps": 5180, "total_steps": 29704, "loss": 0.40553035736083987, "lr": 9.268497100208556e-06, "epoch": 0.3487745758147051, "percentage": 17.44, "elapsed_time": "11:08:28", "remaining_time": "2 days, 4:44:49", "throughput": 5654.99, "total_tokens": 226814192}
+{"current_steps": 5185, "total_steps": 29704, "loss": 0.40463762283325194, "lr": 9.267119555162765e-06, "epoch": 0.34911123081066525, "percentage": 17.46, "elapsed_time": "11:09:08", "remaining_time": "2 days, 4:44:16", "throughput": 5655.1, "total_tokens": 227044824}
+{"current_steps": 5190, "total_steps": 29704, "loss": 0.3876547336578369, "lr": 9.2657408168306e-06, "epoch": 0.3494478858066254, "percentage": 17.47, "elapsed_time": "11:09:41", "remaining_time": "2 days, 4:43:10", "throughput": 5655.91, "total_tokens": 227264048}
+{"current_steps": 5195, "total_steps": 29704, "loss": 0.3948791027069092, "lr": 9.264360885597617e-06, "epoch": 0.34978454080258553, "percentage": 17.49, "elapsed_time": "11:10:24", "remaining_time": "2 days, 4:42:53", "throughput": 5655.26, "total_tokens": 227482568}
+{"current_steps": 5200, "total_steps": 29704, "loss": 0.3999017238616943, "lr": 9.262979761849709e-06, "epoch": 0.35012119579854567, "percentage": 17.51, "elapsed_time": "11:11:05", "remaining_time": "2 days, 4:42:22", "throughput": 5655.04, "total_tokens": 227701672}
+{"current_steps": 5205, "total_steps": 29704, "loss": 0.40303988456726075, "lr": 9.261597445973106e-06, "epoch": 0.3504578507945058, "percentage": 17.52, "elapsed_time": "11:11:40", "remaining_time": "2 days, 4:41:27", "throughput": 5655.77, "total_tokens": 227930736}
+{"current_steps": 5210, "total_steps": 29704, "loss": 0.38801677227020265, "lr": 9.260213938354365e-06, "epoch": 0.35079450579046595, "percentage": 17.54, "elapsed_time": "11:12:19", "remaining_time": "2 days, 4:40:49", "throughput": 5655.56, "total_tokens": 228142248}
+{"current_steps": 5215, "total_steps": 29704, "loss": 0.40598435401916505, "lr": 9.25882923938038e-06, "epoch": 0.3511311607864261, "percentage": 17.56, "elapsed_time": "11:13:02", "remaining_time": "2 days, 4:40:32", "throughput": 5654.76, "total_tokens": 228355000}
+{"current_steps": 5220, "total_steps": 29704, "loss": 0.42745413780212405, "lr": 9.257443349438382e-06, "epoch": 0.35146781578238623, "percentage": 17.57, "elapsed_time": "11:13:39", "remaining_time": "2 days, 4:39:42", "throughput": 5655.24, "total_tokens": 228579256}
+{"current_steps": 5225, "total_steps": 29704, "loss": 0.40657625198364256, "lr": 9.256056268915924e-06, "epoch": 0.35180447077834637, "percentage": 17.59, "elapsed_time": "11:14:21", "remaining_time": "2 days, 4:39:20", "throughput": 5654.81, "total_tokens": 228801840}
+{"current_steps": 5230, "total_steps": 29704, "loss": 0.4108288764953613, "lr": 9.254667998200906e-06, "epoch": 0.3521411257743065, "percentage": 17.61, "elapsed_time": "11:15:02", "remaining_time": "2 days, 4:38:54", "throughput": 5654.5, "total_tokens": 229023432}
+{"current_steps": 5235, "total_steps": 29704, "loss": 0.4314168930053711, "lr": 9.253278537681548e-06, "epoch": 0.35247778077026665, "percentage": 17.62, "elapsed_time": "11:15:38", "remaining_time": "2 days, 4:37:59", "throughput": 5655.42, "total_tokens": 229259792}
+{"current_steps": 5240, "total_steps": 29704, "loss": 0.3931559085845947, "lr": 9.251887887746407e-06, "epoch": 0.3528144357662268, "percentage": 17.64, "elapsed_time": "11:16:17", "remaining_time": "2 days, 4:37:24", "throughput": 5655.49, "total_tokens": 229485656}
+{"current_steps": 5245, "total_steps": 29704, "loss": 0.38311750888824464, "lr": 9.25049604878438e-06, "epoch": 0.3531510907621869, "percentage": 17.66, "elapsed_time": "11:16:55", "remaining_time": "2 days, 4:36:42", "throughput": 5655.59, "total_tokens": 229704840}
+{"current_steps": 5250, "total_steps": 29704, "loss": 0.3960615634918213, "lr": 9.249103021184684e-06, "epoch": 0.35348774575814707, "percentage": 17.67, "elapsed_time": "11:17:33", "remaining_time": "2 days, 4:36:02", "throughput": 5655.54, "total_tokens": 229920256}
+{"current_steps": 5255, "total_steps": 29704, "loss": 0.4205930709838867, "lr": 9.24770880533688e-06, "epoch": 0.3538244007541072, "percentage": 17.69, "elapsed_time": "11:18:11", "remaining_time": "2 days, 4:35:20", "throughput": 5655.93, "total_tokens": 230150768}
+{"current_steps": 5260, "total_steps": 29704, "loss": 0.41542859077453614, "lr": 9.24631340163085e-06, "epoch": 0.35416105575006734, "percentage": 17.71, "elapsed_time": "11:18:47", "remaining_time": "2 days, 4:34:27", "throughput": 5655.58, "total_tokens": 230338192}
+{"current_steps": 5265, "total_steps": 29704, "loss": 0.3574668884277344, "lr": 9.244916810456822e-06, "epoch": 0.3544977107460275, "percentage": 17.72, "elapsed_time": "11:19:18", "remaining_time": "2 days, 4:33:11", "throughput": 5656.96, "total_tokens": 230568472}
+{"current_steps": 5270, "total_steps": 29704, "loss": 0.40234909057617185, "lr": 9.24351903220534e-06, "epoch": 0.3548343657419876, "percentage": 17.74, "elapsed_time": "11:19:58", "remaining_time": "2 days, 4:32:41", "throughput": 5656.84, "total_tokens": 230793504}
+{"current_steps": 5275, "total_steps": 29704, "loss": 0.3589898347854614, "lr": 9.242120067267295e-06, "epoch": 0.35517102073794776, "percentage": 17.76, "elapsed_time": "11:20:37", "remaining_time": "2 days, 4:32:01", "throughput": 5656.85, "total_tokens": 231011256}
+{"current_steps": 5280, "total_steps": 29704, "loss": 0.3916470050811768, "lr": 9.2407199160339e-06, "epoch": 0.3555076757339079, "percentage": 17.78, "elapsed_time": "11:21:19", "remaining_time": "2 days, 4:31:36", "throughput": 5656.3, "total_tokens": 231224304}
+{"current_steps": 5285, "total_steps": 29704, "loss": 0.3869880199432373, "lr": 9.239318578896701e-06, "epoch": 0.35584433072986804, "percentage": 17.79, "elapsed_time": "11:21:57", "remaining_time": "2 days, 4:30:58", "throughput": 5656.23, "total_tokens": 231440576}
+{"current_steps": 5290, "total_steps": 29704, "loss": 0.40935673713684084, "lr": 9.237916056247583e-06, "epoch": 0.3561809857258282, "percentage": 17.81, "elapsed_time": "11:22:35", "remaining_time": "2 days, 4:30:14", "throughput": 5656.23, "total_tokens": 231653280}
+{"current_steps": 5295, "total_steps": 29704, "loss": 0.4153427600860596, "lr": 9.23651234847875e-06, "epoch": 0.3565176407217883, "percentage": 17.83, "elapsed_time": "11:23:13", "remaining_time": "2 days, 4:29:32", "throughput": 5656.19, "total_tokens": 231867296}
+{"current_steps": 5300, "total_steps": 29704, "loss": 0.3877981185913086, "lr": 9.23510745598275e-06, "epoch": 0.35685429571774846, "percentage": 17.84, "elapsed_time": "11:23:52", "remaining_time": "2 days, 4:28:54", "throughput": 5655.87, "total_tokens": 232073536}
+{"current_steps": 5305, "total_steps": 29704, "loss": 0.3829216957092285, "lr": 9.233701379152456e-06, "epoch": 0.3571909507137086, "percentage": 17.86, "elapsed_time": "11:24:30", "remaining_time": "2 days, 4:28:14", "throughput": 5656.0, "total_tokens": 232295728}
+{"current_steps": 5310, "total_steps": 29704, "loss": 0.41249933242797854, "lr": 9.232294118381069e-06, "epoch": 0.35752760570966874, "percentage": 17.88, "elapsed_time": "11:25:07", "remaining_time": "2 days, 4:27:26", "throughput": 5656.41, "total_tokens": 232520456}
+{"current_steps": 5315, "total_steps": 29704, "loss": 0.3704015493392944, "lr": 9.230885674062129e-06, "epoch": 0.3578642607056289, "percentage": 17.89, "elapsed_time": "11:25:46", "remaining_time": "2 days, 4:26:49", "throughput": 5656.38, "total_tokens": 232740384}
+{"current_steps": 5320, "total_steps": 29704, "loss": 0.4049508094787598, "lr": 9.2294760465895e-06, "epoch": 0.358200915701589, "percentage": 17.91, "elapsed_time": "11:26:20", "remaining_time": "2 days, 4:25:47", "throughput": 5657.33, "total_tokens": 232969144}
+{"current_steps": 5325, "total_steps": 29704, "loss": 0.39788720607757566, "lr": 9.228065236357383e-06, "epoch": 0.35853757069754916, "percentage": 17.93, "elapsed_time": "11:26:57", "remaining_time": "2 days, 4:25:01", "throughput": 5657.73, "total_tokens": 233196464}
+{"current_steps": 5330, "total_steps": 29704, "loss": 0.3763267993927002, "lr": 9.226653243760304e-06, "epoch": 0.3588742256935093, "percentage": 17.94, "elapsed_time": "11:27:31", "remaining_time": "2 days, 4:24:01", "throughput": 5658.33, "total_tokens": 233413024}
+{"current_steps": 5335, "total_steps": 29704, "loss": 0.4120461463928223, "lr": 9.225240069193124e-06, "epoch": 0.35921088068946944, "percentage": 17.96, "elapsed_time": "11:28:15", "remaining_time": "2 days, 4:23:50", "throughput": 5657.44, "total_tokens": 233629584}
+{"current_steps": 5340, "total_steps": 29704, "loss": 0.4018904685974121, "lr": 9.223825713051034e-06, "epoch": 0.3595475356854296, "percentage": 17.98, "elapsed_time": "11:28:46", "remaining_time": "2 days, 4:22:35", "throughput": 5658.42, "total_tokens": 233844536}
+{"current_steps": 5345, "total_steps": 29704, "loss": 0.38492045402526853, "lr": 9.222410175729549e-06, "epoch": 0.3598841906813897, "percentage": 17.99, "elapsed_time": "11:29:27", "remaining_time": "2 days, 4:22:05", "throughput": 5658.14, "total_tokens": 234063096}
+{"current_steps": 5350, "total_steps": 29704, "loss": 0.3723012924194336, "lr": 9.220993457624523e-06, "epoch": 0.36022084567734985, "percentage": 18.01, "elapsed_time": "11:30:14", "remaining_time": "2 days, 4:22:06", "throughput": 5656.9, "total_tokens": 234279256}
+{"current_steps": 5355, "total_steps": 29704, "loss": 0.4064267635345459, "lr": 9.219575559132136e-06, "epoch": 0.36055750067331, "percentage": 18.03, "elapsed_time": "11:30:54", "remaining_time": "2 days, 4:21:33", "throughput": 5656.68, "total_tokens": 234496160}
+{"current_steps": 5360, "total_steps": 29704, "loss": 0.3711550235748291, "lr": 9.2181564806489e-06, "epoch": 0.36089415566927013, "percentage": 18.04, "elapsed_time": "11:31:28", "remaining_time": "2 days, 4:20:31", "throughput": 5657.15, "total_tokens": 234705872}
+{"current_steps": 5365, "total_steps": 29704, "loss": 0.36968867778778075, "lr": 9.216736222571655e-06, "epoch": 0.3612308106652303, "percentage": 18.06, "elapsed_time": "11:32:08", "remaining_time": "2 days, 4:19:59", "throughput": 5657.03, "total_tokens": 234928056}
+{"current_steps": 5370, "total_steps": 29704, "loss": 0.4384300708770752, "lr": 9.215314785297572e-06, "epoch": 0.3615674656611904, "percentage": 18.08, "elapsed_time": "11:32:46", "remaining_time": "2 days, 4:19:17", "throughput": 5656.65, "total_tokens": 235127224}
+{"current_steps": 5375, "total_steps": 29704, "loss": 0.3896491527557373, "lr": 9.21389216922415e-06, "epoch": 0.36190412065715055, "percentage": 18.1, "elapsed_time": "11:33:24", "remaining_time": "2 days, 4:18:35", "throughput": 5656.85, "total_tokens": 235350232}
+{"current_steps": 5380, "total_steps": 29704, "loss": 0.4335320949554443, "lr": 9.212468374749222e-06, "epoch": 0.3622407756531107, "percentage": 18.11, "elapsed_time": "11:34:02", "remaining_time": "2 days, 4:17:55", "throughput": 5656.93, "total_tokens": 235570704}
+{"current_steps": 5385, "total_steps": 29704, "loss": 0.39116473197937013, "lr": 9.211043402270942e-06, "epoch": 0.36257743064907083, "percentage": 18.13, "elapsed_time": "11:34:39", "remaining_time": "2 days, 4:17:08", "throughput": 5657.07, "total_tokens": 235785280}
+{"current_steps": 5390, "total_steps": 29704, "loss": 0.3792844533920288, "lr": 9.209617252187806e-06, "epoch": 0.36291408564503097, "percentage": 18.15, "elapsed_time": "11:35:17", "remaining_time": "2 days, 4:16:25", "throughput": 5657.08, "total_tokens": 235999696}
+{"current_steps": 5395, "total_steps": 29704, "loss": 0.3464975833892822, "lr": 9.208189924898627e-06, "epoch": 0.3632507406409911, "percentage": 18.16, "elapsed_time": "11:36:00", "remaining_time": "2 days, 4:16:06", "throughput": 5656.31, "total_tokens": 236210480}
+{"current_steps": 5400, "total_steps": 29704, "loss": 0.4210227966308594, "lr": 9.206761420802554e-06, "epoch": 0.36358739563695125, "percentage": 18.18, "elapsed_time": "11:36:42", "remaining_time": "2 days, 4:15:43", "throughput": 5655.91, "total_tokens": 236432304}
+{"current_steps": 5405, "total_steps": 29704, "loss": 0.3803566932678223, "lr": 9.205331740299065e-06, "epoch": 0.3639240506329114, "percentage": 18.2, "elapsed_time": "11:37:24", "remaining_time": "2 days, 4:15:19", "throughput": 5655.26, "total_tokens": 236642552}
+{"current_steps": 5410, "total_steps": 29704, "loss": 0.4222862243652344, "lr": 9.203900883787967e-06, "epoch": 0.36426070562887153, "percentage": 18.21, "elapsed_time": "11:38:00", "remaining_time": "2 days, 4:14:25", "throughput": 5655.81, "total_tokens": 236866080}
+{"current_steps": 5415, "total_steps": 29704, "loss": 0.42064590454101564, "lr": 9.202468851669388e-06, "epoch": 0.36459736062483167, "percentage": 18.23, "elapsed_time": "11:38:42", "remaining_time": "2 days, 4:14:03", "throughput": 5655.31, "total_tokens": 237084656}
+{"current_steps": 5420, "total_steps": 29704, "loss": 0.40491809844970705, "lr": 9.201035644343797e-06, "epoch": 0.3649340156207918, "percentage": 18.25, "elapsed_time": "11:39:23", "remaining_time": "2 days, 4:13:32", "throughput": 5654.95, "total_tokens": 237299088}
+{"current_steps": 5425, "total_steps": 29704, "loss": 0.3938578128814697, "lr": 9.199601262211985e-06, "epoch": 0.36527067061675195, "percentage": 18.26, "elapsed_time": "11:39:59", "remaining_time": "2 days, 4:12:44", "throughput": 5654.93, "total_tokens": 237505024}
+{"current_steps": 5430, "total_steps": 29704, "loss": 0.4014750480651855, "lr": 9.19816570567507e-06, "epoch": 0.3656073256127121, "percentage": 18.28, "elapsed_time": "11:40:38", "remaining_time": "2 days, 4:12:08", "throughput": 5655.0, "total_tokens": 237729312}
+{"current_steps": 5435, "total_steps": 29704, "loss": 0.41324663162231445, "lr": 9.196728975134503e-06, "epoch": 0.3659439806086722, "percentage": 18.3, "elapsed_time": "11:41:22", "remaining_time": "2 days, 4:11:50", "throughput": 5654.11, "total_tokens": 237937544}
+{"current_steps": 5440, "total_steps": 29704, "loss": 0.39972093105316164, "lr": 9.195291070992061e-06, "epoch": 0.36628063560463237, "percentage": 18.31, "elapsed_time": "11:41:59", "remaining_time": "2 days, 4:11:06", "throughput": 5653.98, "total_tokens": 238144328}
+{"current_steps": 5445, "total_steps": 29704, "loss": 0.36598644256591795, "lr": 9.193851993649849e-06, "epoch": 0.3666172906005925, "percentage": 18.33, "elapsed_time": "11:42:37", "remaining_time": "2 days, 4:10:21", "throughput": 5654.37, "total_tokens": 238371752}
+{"current_steps": 5450, "total_steps": 29704, "loss": 0.3939794063568115, "lr": 9.1924117435103e-06, "epoch": 0.36695394559655264, "percentage": 18.35, "elapsed_time": "11:43:18", "remaining_time": "2 days, 4:09:53", "throughput": 5653.94, "total_tokens": 238586128}
+{"current_steps": 5455, "total_steps": 29704, "loss": 0.3719717264175415, "lr": 9.190970320976176e-06, "epoch": 0.3672906005925128, "percentage": 18.36, "elapsed_time": "11:43:52", "remaining_time": "2 days, 4:08:56", "throughput": 5654.66, "total_tokens": 238811344}
+{"current_steps": 5460, "total_steps": 29704, "loss": 0.4350133895874023, "lr": 9.189527726450565e-06, "epoch": 0.3676272555884729, "percentage": 18.38, "elapsed_time": "11:44:26", "remaining_time": "2 days, 4:07:55", "throughput": 5654.67, "total_tokens": 239003208}
+{"current_steps": 5465, "total_steps": 29704, "loss": 0.3788605213165283, "lr": 9.188083960336885e-06, "epoch": 0.36796391058443306, "percentage": 18.4, "elapsed_time": "11:45:05", "remaining_time": "2 days, 4:07:19", "throughput": 5654.67, "total_tokens": 239225040}
+{"current_steps": 5470, "total_steps": 29704, "loss": 0.4101285934448242, "lr": 9.186639023038879e-06, "epoch": 0.3683005655803932, "percentage": 18.42, "elapsed_time": "11:45:45", "remaining_time": "2 days, 4:06:47", "throughput": 5654.51, "total_tokens": 239445688}
+{"current_steps": 5475, "total_steps": 29704, "loss": 0.43002777099609374, "lr": 9.185192914960622e-06, "epoch": 0.36863722057635334, "percentage": 18.43, "elapsed_time": "11:46:23", "remaining_time": "2 days, 4:06:05", "throughput": 5654.53, "total_tokens": 239660856}
+{"current_steps": 5480, "total_steps": 29704, "loss": 0.400468635559082, "lr": 9.183745636506511e-06, "epoch": 0.3689738755723135, "percentage": 18.45, "elapsed_time": "11:46:55", "remaining_time": "2 days, 4:04:55", "throughput": 5655.06, "total_tokens": 239863160}
+{"current_steps": 5485, "total_steps": 29704, "loss": 0.39147305488586426, "lr": 9.182297188081274e-06, "epoch": 0.3693105305682736, "percentage": 18.47, "elapsed_time": "11:47:25", "remaining_time": "2 days, 4:03:38", "throughput": 5656.38, "total_tokens": 240088504}
+{"current_steps": 5490, "total_steps": 29704, "loss": 0.38025264739990233, "lr": 9.180847570089964e-06, "epoch": 0.36964718556423376, "percentage": 18.48, "elapsed_time": "11:48:07", "remaining_time": "2 days, 4:03:13", "throughput": 5655.44, "total_tokens": 240285080}
+{"current_steps": 5495, "total_steps": 29704, "loss": 0.4285304546356201, "lr": 9.179396782937965e-06, "epoch": 0.3699838405601939, "percentage": 18.5, "elapsed_time": "11:48:42", "remaining_time": "2 days, 4:02:17", "throughput": 5656.31, "total_tokens": 240518464}
+{"current_steps": 5500, "total_steps": 29704, "loss": 0.41852726936340334, "lr": 9.17794482703098e-06, "epoch": 0.37032049555615404, "percentage": 18.52, "elapsed_time": "11:49:21", "remaining_time": "2 days, 4:01:39", "throughput": 5656.28, "total_tokens": 240737112}
+{"current_steps": 5505, "total_steps": 29704, "loss": 0.4076284408569336, "lr": 9.176491702775049e-06, "epoch": 0.3706571505521142, "percentage": 18.53, "elapsed_time": "11:50:05", "remaining_time": "2 days, 4:01:27", "throughput": 5655.11, "total_tokens": 240940632}
+{"current_steps": 5510, "total_steps": 29704, "loss": 0.42336077690124513, "lr": 9.17503741057653e-06, "epoch": 0.3709938055480743, "percentage": 18.55, "elapsed_time": "11:50:41", "remaining_time": "2 days, 4:00:34", "throughput": 5655.65, "total_tokens": 241163512}
+{"current_steps": 5515, "total_steps": 29704, "loss": 0.3858926773071289, "lr": 9.173581950842111e-06, "epoch": 0.37133046054403446, "percentage": 18.57, "elapsed_time": "11:51:17", "remaining_time": "2 days, 3:59:43", "throughput": 5655.97, "total_tokens": 241380280}
+{"current_steps": 5520, "total_steps": 29704, "loss": 0.39653089046478274, "lr": 9.172125323978811e-06, "epoch": 0.3716671155399946, "percentage": 18.58, "elapsed_time": "11:51:53", "remaining_time": "2 days, 3:58:53", "throughput": 5656.36, "total_tokens": 241601112}
+{"current_steps": 5525, "total_steps": 29704, "loss": 0.4098936080932617, "lr": 9.170667530393966e-06, "epoch": 0.37200377053595474, "percentage": 18.6, "elapsed_time": "11:52:27", "remaining_time": "2 days, 3:57:56", "throughput": 5657.09, "total_tokens": 241828432}
+{"current_steps": 5530, "total_steps": 29704, "loss": 0.3672953605651855, "lr": 9.169208570495247e-06, "epoch": 0.3723404255319149, "percentage": 18.62, "elapsed_time": "11:53:07", "remaining_time": "2 days, 3:57:23", "throughput": 5657.02, "total_tokens": 242051344}
+{"current_steps": 5535, "total_steps": 29704, "loss": 0.3897575855255127, "lr": 9.167748444690646e-06, "epoch": 0.372677080527875, "percentage": 18.63, "elapsed_time": "11:53:46", "remaining_time": "2 days, 3:56:44", "throughput": 5656.57, "total_tokens": 242250584}
+{"current_steps": 5540, "total_steps": 29704, "loss": 0.3871308326721191, "lr": 9.16628715338848e-06, "epoch": 0.37301373552383515, "percentage": 18.65, "elapsed_time": "11:54:19", "remaining_time": "2 days, 3:55:39", "throughput": 5657.14, "total_tokens": 242459936}
+{"current_steps": 5545, "total_steps": 29704, "loss": 0.37985944747924805, "lr": 9.164824696997398e-06, "epoch": 0.3733503905197953, "percentage": 18.67, "elapsed_time": "11:54:56", "remaining_time": "2 days, 3:54:54", "throughput": 5657.06, "total_tokens": 242665952}
+{"current_steps": 5550, "total_steps": 29704, "loss": 0.4154071807861328, "lr": 9.163361075926371e-06, "epoch": 0.37368704551575543, "percentage": 18.68, "elapsed_time": "11:55:32", "remaining_time": "2 days, 3:54:06", "throughput": 5657.36, "total_tokens": 242885960}
+{"current_steps": 5555, "total_steps": 29704, "loss": 0.37793383598327634, "lr": 9.161896290584695e-06, "epoch": 0.3740237005117156, "percentage": 18.7, "elapsed_time": "11:56:14", "remaining_time": "2 days, 3:53:40", "throughput": 5657.05, "total_tokens": 243108432}
+{"current_steps": 5560, "total_steps": 29704, "loss": 0.3929650545120239, "lr": 9.160430341381991e-06, "epoch": 0.3743603555076757, "percentage": 18.72, "elapsed_time": "11:57:00", "remaining_time": "2 days, 3:53:35", "throughput": 5655.9, "total_tokens": 243321200}
+{"current_steps": 5565, "total_steps": 29704, "loss": 0.39426612854003906, "lr": 9.15896322872821e-06, "epoch": 0.37469701050363585, "percentage": 18.73, "elapsed_time": "11:57:45", "remaining_time": "2 days, 3:53:23", "throughput": 5655.12, "total_tokens": 243541160}
+{"current_steps": 5570, "total_steps": 29704, "loss": 0.4226834297180176, "lr": 9.157494953033625e-06, "epoch": 0.375033665499596, "percentage": 18.75, "elapsed_time": "11:58:20", "remaining_time": "2 days, 3:52:28", "throughput": 5655.77, "total_tokens": 243766296}
+{"current_steps": 5575, "total_steps": 29704, "loss": 0.40192432403564454, "lr": 9.156025514708832e-06, "epoch": 0.37537032049555613, "percentage": 18.77, "elapsed_time": "11:59:06", "remaining_time": "2 days, 3:52:20", "throughput": 5655.24, "total_tokens": 244002616}
+{"current_steps": 5580, "total_steps": 29704, "loss": 0.38442695140838623, "lr": 9.154554914164757e-06, "epoch": 0.37570697549151627, "percentage": 18.79, "elapsed_time": "11:59:48", "remaining_time": "2 days, 3:51:56", "throughput": 5655.16, "total_tokens": 244237240}
+{"current_steps": 5585, "total_steps": 29704, "loss": 0.377176308631897, "lr": 9.153083151812649e-06, "epoch": 0.3760436304874764, "percentage": 18.8, "elapsed_time": "12:00:25", "remaining_time": "2 days, 3:51:10", "throughput": 5655.56, "total_tokens": 244463576}
+{"current_steps": 5590, "total_steps": 29704, "loss": 0.40382232666015627, "lr": 9.151610228064078e-06, "epoch": 0.37638028548343655, "percentage": 18.82, "elapsed_time": "12:01:04", "remaining_time": "2 days, 3:50:35", "throughput": 5655.52, "total_tokens": 244685816}
+{"current_steps": 5595, "total_steps": 29704, "loss": 0.42247867584228516, "lr": 9.150136143330949e-06, "epoch": 0.3767169404793967, "percentage": 18.84, "elapsed_time": "12:01:45", "remaining_time": "2 days, 3:50:06", "throughput": 5655.15, "total_tokens": 244901992}
+{"current_steps": 5600, "total_steps": 29704, "loss": 0.3760505676269531, "lr": 9.14866089802548e-06, "epoch": 0.37705359547535683, "percentage": 18.85, "elapsed_time": "12:02:28", "remaining_time": "2 days, 3:49:44", "throughput": 5654.82, "total_tokens": 245127560}
+{"current_steps": 5605, "total_steps": 29704, "loss": 0.3698588848114014, "lr": 9.147184492560219e-06, "epoch": 0.37739025047131697, "percentage": 18.87, "elapsed_time": "12:03:08", "remaining_time": "2 days, 3:49:12", "throughput": 5654.44, "total_tokens": 245339952}
+{"current_steps": 5610, "total_steps": 29704, "loss": 0.3787434816360474, "lr": 9.14570692734804e-06, "epoch": 0.3777269054672771, "percentage": 18.89, "elapsed_time": "12:03:45", "remaining_time": "2 days, 3:48:23", "throughput": 5654.91, "total_tokens": 245565672}
+{"current_steps": 5615, "total_steps": 29704, "loss": 0.41345486640930174, "lr": 9.144228202802137e-06, "epoch": 0.37806356046323725, "percentage": 18.9, "elapsed_time": "12:04:17", "remaining_time": "2 days, 3:47:15", "throughput": 5656.01, "total_tokens": 245793416}
+{"current_steps": 5620, "total_steps": 29704, "loss": 0.39602549076080323, "lr": 9.142748319336033e-06, "epoch": 0.37840021545919744, "percentage": 18.92, "elapsed_time": "12:04:55", "remaining_time": "2 days, 3:46:35", "throughput": 5656.12, "total_tokens": 246014600}
+{"current_steps": 5625, "total_steps": 29704, "loss": 0.3865189552307129, "lr": 9.141267277363573e-06, "epoch": 0.3787368704551576, "percentage": 18.94, "elapsed_time": "12:05:30", "remaining_time": "2 days, 3:45:40", "throughput": 5656.04, "total_tokens": 246209040}
+{"current_steps": 5630, "total_steps": 29704, "loss": 0.3669655084609985, "lr": 9.139785077298921e-06, "epoch": 0.3790735254511177, "percentage": 18.95, "elapsed_time": "12:06:10", "remaining_time": "2 days, 3:45:07", "throughput": 5656.04, "total_tokens": 246435304}
+{"current_steps": 5635, "total_steps": 29704, "loss": 0.3826982736587524, "lr": 9.138301719556577e-06, "epoch": 0.37941018044707786, "percentage": 18.97, "elapsed_time": "12:06:49", "remaining_time": "2 days, 3:44:29", "throughput": 5655.98, "total_tokens": 246652712}
+{"current_steps": 5640, "total_steps": 29704, "loss": 0.386492919921875, "lr": 9.136817204551352e-06, "epoch": 0.379746835443038, "percentage": 18.99, "elapsed_time": "12:07:26", "remaining_time": "2 days, 3:43:44", "throughput": 5656.43, "total_tokens": 246882568}
+{"current_steps": 5645, "total_steps": 29704, "loss": 0.3880929470062256, "lr": 9.135331532698385e-06, "epoch": 0.38008349043899814, "percentage": 19.0, "elapsed_time": "12:08:05", "remaining_time": "2 days, 3:43:07", "throughput": 5656.22, "total_tokens": 247094472}
+{"current_steps": 5650, "total_steps": 29704, "loss": 0.37475876808166503, "lr": 9.133844704413144e-06, "epoch": 0.3804201454349583, "percentage": 19.02, "elapsed_time": "12:08:44", "remaining_time": "2 days, 3:42:28", "throughput": 5656.04, "total_tokens": 247305056}
+{"current_steps": 5655, "total_steps": 29704, "loss": 0.40270652770996096, "lr": 9.132356720111413e-06, "epoch": 0.3807568004309184, "percentage": 19.04, "elapsed_time": "12:09:26", "remaining_time": "2 days, 3:42:07", "throughput": 5655.46, "total_tokens": 247521856}
+{"current_steps": 5660, "total_steps": 29704, "loss": 0.3874132394790649, "lr": 9.1308675802093e-06, "epoch": 0.38109345542687856, "percentage": 19.05, "elapsed_time": "12:10:01", "remaining_time": "2 days, 3:41:12", "throughput": 5655.81, "total_tokens": 247735120}
+{"current_steps": 5665, "total_steps": 29704, "loss": 0.3984661102294922, "lr": 9.129377285123241e-06, "epoch": 0.3814301104228387, "percentage": 19.07, "elapsed_time": "12:10:45", "remaining_time": "2 days, 3:40:54", "throughput": 5655.08, "total_tokens": 247949328}
+{"current_steps": 5670, "total_steps": 29704, "loss": 0.4090867519378662, "lr": 9.127885835269996e-06, "epoch": 0.38176676541879884, "percentage": 19.09, "elapsed_time": "12:11:19", "remaining_time": "2 days, 3:39:55", "throughput": 5655.83, "total_tokens": 248174032}
+{"current_steps": 5675, "total_steps": 29704, "loss": 0.3848613977432251, "lr": 9.126393231066636e-06, "epoch": 0.382103420414759, "percentage": 19.11, "elapsed_time": "12:11:57", "remaining_time": "2 days, 3:39:16", "throughput": 5655.65, "total_tokens": 248384368}
+{"current_steps": 5680, "total_steps": 29704, "loss": 0.4047283172607422, "lr": 9.124899472930568e-06, "epoch": 0.3824400754107191, "percentage": 19.12, "elapsed_time": "12:12:41", "remaining_time": "2 days, 3:38:59", "throughput": 5655.1, "total_tokens": 248607600}
+{"current_steps": 5685, "total_steps": 29704, "loss": 0.41053128242492676, "lr": 9.123404561279517e-06, "epoch": 0.38277673040667926, "percentage": 19.14, "elapsed_time": "12:13:26", "remaining_time": "2 days, 3:38:47", "throughput": 5653.94, "total_tokens": 248811928}
+{"current_steps": 5690, "total_steps": 29704, "loss": 0.3802008867263794, "lr": 9.121908496531527e-06, "epoch": 0.3831133854026394, "percentage": 19.16, "elapsed_time": "12:14:00", "remaining_time": "2 days, 3:37:46", "throughput": 5654.73, "total_tokens": 249035424}
+{"current_steps": 5695, "total_steps": 29704, "loss": 0.3904706954956055, "lr": 9.120411279104971e-06, "epoch": 0.38345004039859953, "percentage": 19.17, "elapsed_time": "12:14:43", "remaining_time": "2 days, 3:37:27", "throughput": 5654.15, "total_tokens": 249254456}
+{"current_steps": 5700, "total_steps": 29704, "loss": 0.39504170417785645, "lr": 9.118912909418538e-06, "epoch": 0.3837866953945597, "percentage": 19.19, "elapsed_time": "12:15:27", "remaining_time": "2 days, 3:37:09", "throughput": 5653.61, "total_tokens": 249477272}
+{"current_steps": 5705, "total_steps": 29704, "loss": 0.4095571041107178, "lr": 9.117413387891246e-06, "epoch": 0.3841233503905198, "percentage": 19.21, "elapsed_time": "12:16:10", "remaining_time": "2 days, 3:36:51", "throughput": 5652.92, "total_tokens": 249693520}
+{"current_steps": 5710, "total_steps": 29704, "loss": 0.39027183055877684, "lr": 9.11591271494243e-06, "epoch": 0.38446000538647995, "percentage": 19.22, "elapsed_time": "12:16:49", "remaining_time": "2 days, 3:36:11", "throughput": 5652.88, "total_tokens": 249909984}
+{"current_steps": 5715, "total_steps": 29704, "loss": 0.40212602615356446, "lr": 9.114410890991748e-06, "epoch": 0.3847966603824401, "percentage": 19.24, "elapsed_time": "12:17:30", "remaining_time": "2 days, 3:35:45", "throughput": 5652.41, "total_tokens": 250124664}
+{"current_steps": 5720, "total_steps": 29704, "loss": 0.37251665592193606, "lr": 9.112907916459177e-06, "epoch": 0.38513331537840023, "percentage": 19.26, "elapsed_time": "12:18:13", "remaining_time": "2 days, 3:35:23", "throughput": 5651.31, "total_tokens": 250317264}
+{"current_steps": 5725, "total_steps": 29704, "loss": 0.38798794746398924, "lr": 9.111403791765025e-06, "epoch": 0.38546997037436037, "percentage": 19.27, "elapsed_time": "12:18:49", "remaining_time": "2 days, 3:34:34", "throughput": 5651.81, "total_tokens": 250544576}
+{"current_steps": 5730, "total_steps": 29704, "loss": 0.3711621046066284, "lr": 9.109898517329914e-06, "epoch": 0.3858066253703205, "percentage": 19.29, "elapsed_time": "12:19:22", "remaining_time": "2 days, 3:33:28", "throughput": 5652.9, "total_tokens": 250774080}
+{"current_steps": 5735, "total_steps": 29704, "loss": 0.3934319972991943, "lr": 9.108392093574785e-06, "epoch": 0.38614328036628065, "percentage": 19.31, "elapsed_time": "12:19:56", "remaining_time": "2 days, 3:32:29", "throughput": 5653.48, "total_tokens": 250992112}
+{"current_steps": 5740, "total_steps": 29704, "loss": 0.36261420249938964, "lr": 9.10688452092091e-06, "epoch": 0.3864799353622408, "percentage": 19.32, "elapsed_time": "12:20:26", "remaining_time": "2 days, 3:31:16", "throughput": 5654.54, "total_tokens": 251210912}
+{"current_steps": 5745, "total_steps": 29704, "loss": 0.39951817989349364, "lr": 9.105375799789874e-06, "epoch": 0.38681659035820093, "percentage": 19.34, "elapsed_time": "12:21:15", "remaining_time": "2 days, 3:31:22", "throughput": 5653.1, "total_tokens": 251426584}
+{"current_steps": 5750, "total_steps": 29704, "loss": 0.3850820064544678, "lr": 9.103865930603586e-06, "epoch": 0.38715324535416107, "percentage": 19.36, "elapsed_time": "12:21:56", "remaining_time": "2 days, 3:30:52", "throughput": 5652.57, "total_tokens": 251633720}
+{"current_steps": 5755, "total_steps": 29704, "loss": 0.3853774070739746, "lr": 9.102354913784279e-06, "epoch": 0.3874899003501212, "percentage": 19.37, "elapsed_time": "12:22:39", "remaining_time": "2 days, 3:30:31", "throughput": 5652.1, "total_tokens": 251855296}
+{"current_steps": 5760, "total_steps": 29704, "loss": 0.3578346252441406, "lr": 9.100842749754499e-06, "epoch": 0.38782655534608135, "percentage": 19.39, "elapsed_time": "12:23:25", "remaining_time": "2 days, 3:30:21", "throughput": 5651.35, "total_tokens": 252080400}
+{"current_steps": 5765, "total_steps": 29704, "loss": 0.40854997634887696, "lr": 9.099329438937122e-06, "epoch": 0.3881632103420415, "percentage": 19.41, "elapsed_time": "12:24:09", "remaining_time": "2 days, 3:30:05", "throughput": 5650.6, "total_tokens": 252296312}
+{"current_steps": 5770, "total_steps": 29704, "loss": 0.3587602138519287, "lr": 9.09781498175534e-06, "epoch": 0.3884998653380016, "percentage": 19.42, "elapsed_time": "12:24:45", "remaining_time": "2 days, 3:29:15", "throughput": 5650.82, "total_tokens": 252509656}
+{"current_steps": 5775, "total_steps": 29704, "loss": 0.37611684799194334, "lr": 9.096299378632666e-06, "epoch": 0.38883652033396177, "percentage": 19.44, "elapsed_time": "12:25:27", "remaining_time": "2 days, 3:28:48", "throughput": 5650.72, "total_tokens": 252740072}
+{"current_steps": 5780, "total_steps": 29704, "loss": 0.4280585289001465, "lr": 9.09478262999293e-06, "epoch": 0.3891731753299219, "percentage": 19.46, "elapsed_time": "12:26:11", "remaining_time": "2 days, 3:28:32", "throughput": 5650.37, "total_tokens": 252973392}
+{"current_steps": 5785, "total_steps": 29704, "loss": 0.38141813278198244, "lr": 9.09326473626029e-06, "epoch": 0.38950983032588204, "percentage": 19.48, "elapsed_time": "12:26:49", "remaining_time": "2 days, 3:27:52", "throughput": 5650.22, "total_tokens": 253184320}
+{"current_steps": 5790, "total_steps": 29704, "loss": 0.41326370239257815, "lr": 9.091745697859218e-06, "epoch": 0.3898464853218422, "percentage": 19.49, "elapsed_time": "12:27:25", "remaining_time": "2 days, 3:27:01", "throughput": 5650.77, "total_tokens": 253411776}
+{"current_steps": 5795, "total_steps": 29704, "loss": 0.4117950439453125, "lr": 9.090225515214512e-06, "epoch": 0.3901831403178023, "percentage": 19.51, "elapsed_time": "12:28:05", "remaining_time": "2 days, 3:26:29", "throughput": 5650.51, "total_tokens": 253627512}
+{"current_steps": 5800, "total_steps": 29704, "loss": 0.4129201889038086, "lr": 9.088704188751281e-06, "epoch": 0.39051979531376246, "percentage": 19.53, "elapsed_time": "12:28:44", "remaining_time": "2 days, 3:25:49", "throughput": 5650.42, "total_tokens": 253840400}
+{"current_steps": 5805, "total_steps": 29704, "loss": 0.39529869556427, "lr": 9.087181718894963e-06, "epoch": 0.3908564503097226, "percentage": 19.54, "elapsed_time": "12:29:28", "remaining_time": "2 days, 3:25:33", "throughput": 5649.83, "total_tokens": 254064096}
+{"current_steps": 5810, "total_steps": 29704, "loss": 0.39313058853149413, "lr": 9.085658106071309e-06, "epoch": 0.39119310530568274, "percentage": 19.56, "elapsed_time": "12:30:05", "remaining_time": "2 days, 3:24:46", "throughput": 5650.07, "total_tokens": 254281632}
+{"current_steps": 5815, "total_steps": 29704, "loss": 0.3712653398513794, "lr": 9.084133350706394e-06, "epoch": 0.3915297603016429, "percentage": 19.58, "elapsed_time": "12:30:42", "remaining_time": "2 days, 3:24:00", "throughput": 5650.13, "total_tokens": 254493944}
+{"current_steps": 5820, "total_steps": 29704, "loss": 0.38478312492370603, "lr": 9.082607453226613e-06, "epoch": 0.391866415297603, "percentage": 19.59, "elapsed_time": "12:31:17", "remaining_time": "2 days, 3:23:07", "throughput": 5650.64, "total_tokens": 254715712}
+{"current_steps": 5825, "total_steps": 29704, "loss": 0.3994553565979004, "lr": 9.081080414058674e-06, "epoch": 0.39220307029356316, "percentage": 19.61, "elapsed_time": "12:31:53", "remaining_time": "2 days, 3:22:20", "throughput": 5650.74, "total_tokens": 254927592}
+{"current_steps": 5830, "total_steps": 29704, "loss": 0.3710631847381592, "lr": 9.079552233629612e-06, "epoch": 0.3925397252895233, "percentage": 19.63, "elapsed_time": "12:32:32", "remaining_time": "2 days, 3:21:38", "throughput": 5650.66, "total_tokens": 255138464}
+{"current_steps": 5835, "total_steps": 29704, "loss": 0.3914163827896118, "lr": 9.078022912366776e-06, "epoch": 0.39287638028548344, "percentage": 19.64, "elapsed_time": "12:33:12", "remaining_time": "2 days, 3:21:05", "throughput": 5650.69, "total_tokens": 255367048}
+{"current_steps": 5840, "total_steps": 29704, "loss": 0.40029544830322267, "lr": 9.076492450697835e-06, "epoch": 0.3932130352814436, "percentage": 19.66, "elapsed_time": "12:33:52", "remaining_time": "2 days, 3:20:35", "throughput": 5650.7, "total_tokens": 255598184}
+{"current_steps": 5845, "total_steps": 29704, "loss": 0.3852508068084717, "lr": 9.074960849050782e-06, "epoch": 0.3935496902774037, "percentage": 19.68, "elapsed_time": "12:34:32", "remaining_time": "2 days, 3:20:01", "throughput": 5650.62, "total_tokens": 255819944}
+{"current_steps": 5850, "total_steps": 29704, "loss": 0.37957100868225097, "lr": 9.07342810785392e-06, "epoch": 0.39388634527336386, "percentage": 19.69, "elapsed_time": "12:35:11", "remaining_time": "2 days, 3:19:22", "throughput": 5650.22, "total_tokens": 256019616}
+{"current_steps": 5855, "total_steps": 29704, "loss": 0.38977823257446287, "lr": 9.071894227535878e-06, "epoch": 0.394223000269324, "percentage": 19.71, "elapsed_time": "12:35:52", "remaining_time": "2 days, 3:18:53", "throughput": 5649.9, "total_tokens": 256237256}
+{"current_steps": 5860, "total_steps": 29704, "loss": 0.406312084197998, "lr": 9.070359208525598e-06, "epoch": 0.39455965526528414, "percentage": 19.73, "elapsed_time": "12:36:28", "remaining_time": "2 days, 3:18:02", "throughput": 5650.44, "total_tokens": 256464368}
+{"current_steps": 5865, "total_steps": 29704, "loss": 0.3981436491012573, "lr": 9.068823051252346e-06, "epoch": 0.3948963102612443, "percentage": 19.74, "elapsed_time": "12:37:09", "remaining_time": "2 days, 3:17:32", "throughput": 5650.28, "total_tokens": 256688032}
+{"current_steps": 5870, "total_steps": 29704, "loss": 0.3921481132507324, "lr": 9.067285756145702e-06, "epoch": 0.3952329652572044, "percentage": 19.76, "elapsed_time": "12:37:42", "remaining_time": "2 days, 3:16:31", "throughput": 5651.21, "total_tokens": 256917840}
+{"current_steps": 5875, "total_steps": 29704, "loss": 0.3327157020568848, "lr": 9.065747323635565e-06, "epoch": 0.39556962025316456, "percentage": 19.78, "elapsed_time": "12:38:21", "remaining_time": "2 days, 3:15:53", "throughput": 5650.41, "total_tokens": 257101736}
+{"current_steps": 5880, "total_steps": 29704, "loss": 0.3997276782989502, "lr": 9.064207754152154e-06, "epoch": 0.3959062752491247, "percentage": 19.8, "elapsed_time": "12:39:05", "remaining_time": "2 days, 3:15:35", "throughput": 5649.72, "total_tokens": 257317696}
+{"current_steps": 5885, "total_steps": 29704, "loss": 0.3864887714385986, "lr": 9.062667048126005e-06, "epoch": 0.39624293024508483, "percentage": 19.81, "elapsed_time": "12:39:49", "remaining_time": "2 days, 3:15:21", "throughput": 5649.21, "total_tokens": 257547392}
+{"current_steps": 5890, "total_steps": 29704, "loss": 0.4157553672790527, "lr": 9.061125205987971e-06, "epoch": 0.396579585241045, "percentage": 19.83, "elapsed_time": "12:40:30", "remaining_time": "2 days, 3:14:51", "throughput": 5648.84, "total_tokens": 257761408}
+{"current_steps": 5895, "total_steps": 29704, "loss": 0.37337889671325686, "lr": 9.059582228169222e-06, "epoch": 0.3969162402370051, "percentage": 19.85, "elapsed_time": "12:41:04", "remaining_time": "2 days, 3:13:52", "throughput": 5649.09, "total_tokens": 257963800}
+{"current_steps": 5900, "total_steps": 29704, "loss": 0.4037358283996582, "lr": 9.058038115101248e-06, "epoch": 0.39725289523296525, "percentage": 19.86, "elapsed_time": "12:41:42", "remaining_time": "2 days, 3:13:10", "throughput": 5649.48, "total_tokens": 258195648}
+{"current_steps": 5905, "total_steps": 29704, "loss": 0.3476010084152222, "lr": 9.056492867215857e-06, "epoch": 0.3975895502289254, "percentage": 19.88, "elapsed_time": "12:42:28", "remaining_time": "2 days, 3:13:01", "throughput": 5648.27, "total_tokens": 258400784}
+{"current_steps": 5910, "total_steps": 29704, "loss": 0.3933530330657959, "lr": 9.05494648494517e-06, "epoch": 0.39792620522488553, "percentage": 19.9, "elapsed_time": "12:43:06", "remaining_time": "2 days, 3:12:20", "throughput": 5648.07, "total_tokens": 258606968}
+{"current_steps": 5915, "total_steps": 29704, "loss": 0.37357821464538576, "lr": 9.053398968721629e-06, "epoch": 0.39826286022084567, "percentage": 19.91, "elapsed_time": "12:43:47", "remaining_time": "2 days, 3:11:47", "throughput": 5648.04, "total_tokens": 258832840}
+{"current_steps": 5920, "total_steps": 29704, "loss": 0.4264666080474854, "lr": 9.05185031897799e-06, "epoch": 0.3985995152168058, "percentage": 19.93, "elapsed_time": "12:44:20", "remaining_time": "2 days, 3:10:47", "throughput": 5648.66, "total_tokens": 259049920}
+{"current_steps": 5925, "total_steps": 29704, "loss": 0.42224712371826173, "lr": 9.050300536147331e-06, "epoch": 0.39893617021276595, "percentage": 19.95, "elapsed_time": "12:44:58", "remaining_time": "2 days, 3:10:07", "throughput": 5649.07, "total_tokens": 259285872}
+{"current_steps": 5930, "total_steps": 29704, "loss": 0.36397323608398435, "lr": 9.048749620663044e-06, "epoch": 0.3992728252087261, "percentage": 19.96, "elapsed_time": "12:45:35", "remaining_time": "2 days, 3:09:19", "throughput": 5649.7, "total_tokens": 259520536}
+{"current_steps": 5935, "total_steps": 29704, "loss": 0.37349615097045896, "lr": 9.047197572958834e-06, "epoch": 0.39960948020468623, "percentage": 19.98, "elapsed_time": "12:46:13", "remaining_time": "2 days, 3:08:38", "throughput": 5649.84, "total_tokens": 259742248}
+{"current_steps": 5940, "total_steps": 29704, "loss": 0.3917673110961914, "lr": 9.04564439346873e-06, "epoch": 0.39994613520064637, "percentage": 20.0, "elapsed_time": "12:46:48", "remaining_time": "2 days, 3:07:46", "throughput": 5650.03, "total_tokens": 259952080}
+{"current_steps": 5945, "total_steps": 29704, "loss": 0.39895708560943605, "lr": 9.044090082627071e-06, "epoch": 0.4002827901966065, "percentage": 20.01, "elapsed_time": "12:47:25", "remaining_time": "2 days, 3:07:01", "throughput": 5650.02, "total_tokens": 260160320}
+{"current_steps": 5950, "total_steps": 29704, "loss": 0.4021873950958252, "lr": 9.042534640868514e-06, "epoch": 0.40061944519256665, "percentage": 20.03, "elapsed_time": "12:48:01", "remaining_time": "2 days, 3:06:11", "throughput": 5650.11, "total_tokens": 260368008}
+{"current_steps": 5955, "total_steps": 29704, "loss": 0.40214033126831056, "lr": 9.04097806862804e-06, "epoch": 0.4009561001885268, "percentage": 20.05, "elapsed_time": "12:48:37", "remaining_time": "2 days, 3:05:18", "throughput": 5650.67, "total_tokens": 260592824}
+{"current_steps": 5960, "total_steps": 29704, "loss": 0.4055626392364502, "lr": 9.03942036634093e-06, "epoch": 0.4012927551844869, "percentage": 20.06, "elapsed_time": "12:49:16", "remaining_time": "2 days, 3:04:42", "throughput": 5650.57, "total_tokens": 260810288}
+{"current_steps": 5965, "total_steps": 29704, "loss": 0.3734895706176758, "lr": 9.037861534442797e-06, "epoch": 0.40162941018044707, "percentage": 20.08, "elapsed_time": "12:49:58", "remaining_time": "2 days, 3:04:15", "throughput": 5650.4, "total_tokens": 261038312}
+{"current_steps": 5970, "total_steps": 29704, "loss": 0.40189208984375, "lr": 9.036301573369563e-06, "epoch": 0.4019660651764072, "percentage": 20.1, "elapsed_time": "12:50:39", "remaining_time": "2 days, 3:03:47", "throughput": 5650.28, "total_tokens": 261266152}
+{"current_steps": 5975, "total_steps": 29704, "loss": 0.3900024175643921, "lr": 9.034740483557465e-06, "epoch": 0.40230272017236735, "percentage": 20.12, "elapsed_time": "12:51:14", "remaining_time": "2 days, 3:02:55", "throughput": 5650.65, "total_tokens": 261483152}
+{"current_steps": 5980, "total_steps": 29704, "loss": 0.38559999465942385, "lr": 9.033178265443055e-06, "epoch": 0.4026393751683275, "percentage": 20.13, "elapsed_time": "12:51:46", "remaining_time": "2 days, 3:01:50", "throughput": 5651.49, "total_tokens": 261703416}
+{"current_steps": 5985, "total_steps": 29704, "loss": 0.3772167682647705, "lr": 9.031614919463206e-06, "epoch": 0.4029760301642876, "percentage": 20.15, "elapsed_time": "12:52:22", "remaining_time": "2 days, 3:01:00", "throughput": 5651.93, "total_tokens": 261926896}
+{"current_steps": 5990, "total_steps": 29704, "loss": 0.3684276342391968, "lr": 9.030050446055099e-06, "epoch": 0.40331268516024776, "percentage": 20.17, "elapsed_time": "12:53:00", "remaining_time": "2 days, 3:00:16", "throughput": 5652.04, "total_tokens": 262143080}
+{"current_steps": 5995, "total_steps": 29704, "loss": 0.3965585231781006, "lr": 9.028484845656235e-06, "epoch": 0.4036493401562079, "percentage": 20.18, "elapsed_time": "12:53:33", "remaining_time": "2 days, 2:59:17", "throughput": 5653.04, "total_tokens": 262378952}
+{"current_steps": 6000, "total_steps": 29704, "loss": 0.42124166488647463, "lr": 9.026918118704431e-06, "epoch": 0.40398599515216804, "percentage": 20.2, "elapsed_time": "12:54:14", "remaining_time": "2 days, 2:58:45", "throughput": 5652.87, "total_tokens": 262599696}
+{"current_steps": 6005, "total_steps": 29704, "loss": 0.3700655221939087, "lr": 9.025350265637816e-06, "epoch": 0.4043226501481282, "percentage": 20.22, "elapsed_time": "12:54:50", "remaining_time": "2 days, 2:57:57", "throughput": 5653.4, "total_tokens": 262830992}
+{"current_steps": 6010, "total_steps": 29704, "loss": 0.38486456871032715, "lr": 9.023781286894834e-06, "epoch": 0.4046593051440883, "percentage": 20.23, "elapsed_time": "12:55:26", "remaining_time": "2 days, 2:57:06", "throughput": 5653.66, "total_tokens": 263044288}
+{"current_steps": 6015, "total_steps": 29704, "loss": 0.3843253135681152, "lr": 9.022211182914247e-06, "epoch": 0.40499596014004846, "percentage": 20.25, "elapsed_time": "12:56:04", "remaining_time": "2 days, 2:56:27", "throughput": 5653.71, "total_tokens": 263264800}
+{"current_steps": 6020, "total_steps": 29704, "loss": 0.3804231405258179, "lr": 9.020639954135128e-06, "epoch": 0.4053326151360086, "percentage": 20.27, "elapsed_time": "12:56:48", "remaining_time": "2 days, 2:56:06", "throughput": 5653.02, "total_tokens": 263476880}
+{"current_steps": 6025, "total_steps": 29704, "loss": 0.32609758377075193, "lr": 9.019067600996867e-06, "epoch": 0.40566927013196874, "percentage": 20.28, "elapsed_time": "12:57:27", "remaining_time": "2 days, 2:55:32", "throughput": 5652.48, "total_tokens": 263676440}
+{"current_steps": 6030, "total_steps": 29704, "loss": 0.3878431558609009, "lr": 9.017494123939169e-06, "epoch": 0.4060059251279289, "percentage": 20.3, "elapsed_time": "12:58:06", "remaining_time": "2 days, 2:54:52", "throughput": 5652.43, "total_tokens": 263891832}
+{"current_steps": 6035, "total_steps": 29704, "loss": 0.38746066093444825, "lr": 9.015919523402048e-06, "epoch": 0.406342580123889, "percentage": 20.32, "elapsed_time": "12:58:43", "remaining_time": "2 days, 2:54:09", "throughput": 5652.51, "total_tokens": 264108024}
+{"current_steps": 6040, "total_steps": 29704, "loss": 0.3903219699859619, "lr": 9.014343799825838e-06, "epoch": 0.40667923511984916, "percentage": 20.33, "elapsed_time": "12:59:18", "remaining_time": "2 days, 2:53:13", "throughput": 5652.7, "total_tokens": 264311560}
+{"current_steps": 6045, "total_steps": 29704, "loss": 0.42699203491210935, "lr": 9.012766953651186e-06, "epoch": 0.4070158901158093, "percentage": 20.35, "elapsed_time": "13:00:00", "remaining_time": "2 days, 2:52:48", "throughput": 5652.42, "total_tokens": 264536928}
+{"current_steps": 6050, "total_steps": 29704, "loss": 0.4040683746337891, "lr": 9.01118898531905e-06, "epoch": 0.40735254511176944, "percentage": 20.37, "elapsed_time": "13:00:41", "remaining_time": "2 days, 2:52:19", "throughput": 5651.9, "total_tokens": 264744560}
+{"current_steps": 6055, "total_steps": 29704, "loss": 0.3655755043029785, "lr": 9.009609895270708e-06, "epoch": 0.4076892001077296, "percentage": 20.38, "elapsed_time": "13:01:21", "remaining_time": "2 days, 2:51:44", "throughput": 5651.99, "total_tokens": 264973096}
+{"current_steps": 6060, "total_steps": 29704, "loss": 0.42575840950012206, "lr": 9.008029683947743e-06, "epoch": 0.4080258551036897, "percentage": 20.4, "elapsed_time": "13:01:57", "remaining_time": "2 days, 2:50:56", "throughput": 5652.39, "total_tokens": 265197544}
+{"current_steps": 6065, "total_steps": 29704, "loss": 0.3974879741668701, "lr": 9.006448351792057e-06, "epoch": 0.40836251009964986, "percentage": 20.42, "elapsed_time": "13:02:42", "remaining_time": "2 days, 2:50:42", "throughput": 5651.67, "total_tokens": 265418944}
+{"current_steps": 6070, "total_steps": 29704, "loss": 0.38395133018493655, "lr": 9.004865899245864e-06, "epoch": 0.40869916509561, "percentage": 20.43, "elapsed_time": "13:03:19", "remaining_time": "2 days, 2:49:57", "throughput": 5651.85, "total_tokens": 265636776}
+{"current_steps": 6075, "total_steps": 29704, "loss": 0.4008152961730957, "lr": 9.003282326751694e-06, "epoch": 0.40903582009157013, "percentage": 20.45, "elapsed_time": "13:03:59", "remaining_time": "2 days, 2:49:21", "throughput": 5651.82, "total_tokens": 265857800}
+{"current_steps": 6080, "total_steps": 29704, "loss": 0.40933728218078613, "lr": 9.001697634752387e-06, "epoch": 0.4093724750875303, "percentage": 20.47, "elapsed_time": "13:04:34", "remaining_time": "2 days, 2:48:29", "throughput": 5652.1, "total_tokens": 266070448}
+{"current_steps": 6085, "total_steps": 29704, "loss": 0.37639498710632324, "lr": 9.000111823691097e-06, "epoch": 0.4097091300834904, "percentage": 20.49, "elapsed_time": "13:05:12", "remaining_time": "2 days, 2:47:47", "throughput": 5652.27, "total_tokens": 266291816}
+{"current_steps": 6090, "total_steps": 29704, "loss": 0.40433249473571775, "lr": 8.99852489401129e-06, "epoch": 0.41004578507945055, "percentage": 20.5, "elapsed_time": "13:05:48", "remaining_time": "2 days, 2:46:57", "throughput": 5652.76, "total_tokens": 266517904}
+{"current_steps": 6095, "total_steps": 29704, "loss": 0.3635870456695557, "lr": 8.996936846156748e-06, "epoch": 0.4103824400754107, "percentage": 20.52, "elapsed_time": "13:06:23", "remaining_time": "2 days, 2:46:04", "throughput": 5653.2, "total_tokens": 266736992}
+{"current_steps": 6100, "total_steps": 29704, "loss": 0.41033496856689455, "lr": 8.995347680571563e-06, "epoch": 0.41071909507137083, "percentage": 20.54, "elapsed_time": "13:06:57", "remaining_time": "2 days, 2:45:08", "throughput": 5653.81, "total_tokens": 266959272}
+{"current_steps": 6105, "total_steps": 29704, "loss": 0.38593282699584963, "lr": 8.993757397700137e-06, "epoch": 0.41105575006733097, "percentage": 20.55, "elapsed_time": "13:07:34", "remaining_time": "2 days, 2:44:21", "throughput": 5654.08, "total_tokens": 267178696}
+{"current_steps": 6110, "total_steps": 29704, "loss": 0.42214293479919435, "lr": 8.99216599798719e-06, "epoch": 0.41139240506329117, "percentage": 20.57, "elapsed_time": "13:08:10", "remaining_time": "2 days, 2:43:34", "throughput": 5654.71, "total_tokens": 267415152}
+{"current_steps": 6115, "total_steps": 29704, "loss": 0.3639373779296875, "lr": 8.990573481877753e-06, "epoch": 0.4117290600592513, "percentage": 20.59, "elapsed_time": "13:08:45", "remaining_time": "2 days, 2:42:41", "throughput": 5654.83, "total_tokens": 267618552}
+{"current_steps": 6120, "total_steps": 29704, "loss": 0.36659536361694334, "lr": 8.988979849817167e-06, "epoch": 0.41206571505521145, "percentage": 20.6, "elapsed_time": "13:09:26", "remaining_time": "2 days, 2:42:10", "throughput": 5654.11, "total_tokens": 267814904}
+{"current_steps": 6125, "total_steps": 29704, "loss": 0.3810924291610718, "lr": 8.987385102251085e-06, "epoch": 0.4124023700511716, "percentage": 20.62, "elapsed_time": "13:10:05", "remaining_time": "2 days, 2:41:32", "throughput": 5653.73, "total_tokens": 268015432}
+{"current_steps": 6130, "total_steps": 29704, "loss": 0.3897581100463867, "lr": 8.985789239625475e-06, "epoch": 0.4127390250471317, "percentage": 20.64, "elapsed_time": "13:10:49", "remaining_time": "2 days, 2:41:15", "throughput": 5653.07, "total_tokens": 268235696}
+{"current_steps": 6135, "total_steps": 29704, "loss": 0.36377105712890623, "lr": 8.984192262386613e-06, "epoch": 0.41307568004309186, "percentage": 20.65, "elapsed_time": "13:11:23", "remaining_time": "2 days, 2:40:17", "throughput": 5653.67, "total_tokens": 268454176}
+{"current_steps": 6140, "total_steps": 29704, "loss": 0.3994285583496094, "lr": 8.98259417098109e-06, "epoch": 0.413412335039052, "percentage": 20.67, "elapsed_time": "13:12:03", "remaining_time": "2 days, 2:39:45", "throughput": 5653.73, "total_tokens": 268685424}
+{"current_steps": 6145, "total_steps": 29704, "loss": 0.39436161518096924, "lr": 8.980994965855808e-06, "epoch": 0.41374899003501214, "percentage": 20.69, "elapsed_time": "13:12:37", "remaining_time": "2 days, 2:38:47", "throughput": 5654.37, "total_tokens": 268906160}
+{"current_steps": 6150, "total_steps": 29704, "loss": 0.3927819490432739, "lr": 8.979394647457976e-06, "epoch": 0.4140856450309723, "percentage": 20.7, "elapsed_time": "13:13:20", "remaining_time": "2 days, 2:38:26", "throughput": 5653.89, "total_tokens": 269128016}
+{"current_steps": 6155, "total_steps": 29704, "loss": 0.35940072536468504, "lr": 8.97779321623512e-06, "epoch": 0.4144223000269324, "percentage": 20.72, "elapsed_time": "13:13:58", "remaining_time": "2 days, 2:37:45", "throughput": 5653.65, "total_tokens": 269333048}
+{"current_steps": 6160, "total_steps": 29704, "loss": 0.400145435333252, "lr": 8.976190672635077e-06, "epoch": 0.41475895502289256, "percentage": 20.74, "elapsed_time": "13:14:39", "remaining_time": "2 days, 2:37:14", "throughput": 5653.55, "total_tokens": 269557248}
+{"current_steps": 6165, "total_steps": 29704, "loss": 0.41222319602966306, "lr": 8.974587017105991e-06, "epoch": 0.4150956100188527, "percentage": 20.75, "elapsed_time": "13:15:13", "remaining_time": "2 days, 2:36:16", "throughput": 5653.97, "total_tokens": 269768408}
+{"current_steps": 6170, "total_steps": 29704, "loss": 0.3931986570358276, "lr": 8.97298225009632e-06, "epoch": 0.41543226501481284, "percentage": 20.77, "elapsed_time": "13:15:54", "remaining_time": "2 days, 2:35:47", "throughput": 5653.74, "total_tokens": 269990368}
+{"current_steps": 6175, "total_steps": 29704, "loss": 0.3943799495697021, "lr": 8.971376372054829e-06, "epoch": 0.415768920010773, "percentage": 20.79, "elapsed_time": "13:16:34", "remaining_time": "2 days, 2:35:15", "throughput": 5653.51, "total_tokens": 270208056}
+{"current_steps": 6180, "total_steps": 29704, "loss": 0.3858570337295532, "lr": 8.969769383430602e-06, "epoch": 0.4161055750067331, "percentage": 20.81, "elapsed_time": "13:17:09", "remaining_time": "2 days, 2:34:22", "throughput": 5654.06, "total_tokens": 270432152}
+{"current_steps": 6185, "total_steps": 29704, "loss": 0.37152729034423826, "lr": 8.968161284673027e-06, "epoch": 0.41644223000269326, "percentage": 20.82, "elapsed_time": "13:17:48", "remaining_time": "2 days, 2:33:43", "throughput": 5654.3, "total_tokens": 270661216}
+{"current_steps": 6190, "total_steps": 29704, "loss": 0.37863969802856445, "lr": 8.9665520762318e-06, "epoch": 0.4167788849986534, "percentage": 20.84, "elapsed_time": "13:18:24", "remaining_time": "2 days, 2:32:55", "throughput": 5654.49, "total_tokens": 270875256}
+{"current_steps": 6195, "total_steps": 29704, "loss": 0.3976884126663208, "lr": 8.964941758556934e-06, "epoch": 0.41711553999461354, "percentage": 20.86, "elapsed_time": "13:19:08", "remaining_time": "2 days, 2:32:37", "throughput": 5653.71, "total_tokens": 271088288}
+{"current_steps": 6200, "total_steps": 29704, "loss": 0.3814378261566162, "lr": 8.963330332098747e-06, "epoch": 0.4174521949905737, "percentage": 20.87, "elapsed_time": "13:19:45", "remaining_time": "2 days, 2:31:51", "throughput": 5653.7, "total_tokens": 271296528}
+{"current_steps": 6205, "total_steps": 29704, "loss": 0.37258970737457275, "lr": 8.961717797307872e-06, "epoch": 0.4177888499865338, "percentage": 20.89, "elapsed_time": "13:20:19", "remaining_time": "2 days, 2:30:54", "throughput": 5654.28, "total_tokens": 271515568}
+{"current_steps": 6210, "total_steps": 29704, "loss": 0.4039943695068359, "lr": 8.960104154635248e-06, "epoch": 0.41812550498249396, "percentage": 20.91, "elapsed_time": "13:20:57", "remaining_time": "2 days, 2:30:14", "throughput": 5654.32, "total_tokens": 271733704}
+{"current_steps": 6215, "total_steps": 29704, "loss": 0.4006314277648926, "lr": 8.958489404532125e-06, "epoch": 0.4184621599784541, "percentage": 20.92, "elapsed_time": "13:21:34", "remaining_time": "2 days, 2:29:30", "throughput": 5654.25, "total_tokens": 271941040}
+{"current_steps": 6220, "total_steps": 29704, "loss": 0.3996126651763916, "lr": 8.956873547450062e-06, "epoch": 0.41879881497441424, "percentage": 20.94, "elapsed_time": "13:22:15", "remaining_time": "2 days, 2:28:57", "throughput": 5653.99, "total_tokens": 272155560}
+{"current_steps": 6225, "total_steps": 29704, "loss": 0.4290035247802734, "lr": 8.95525658384093e-06, "epoch": 0.4191354699703744, "percentage": 20.96, "elapsed_time": "13:22:51", "remaining_time": "2 days, 2:28:11", "throughput": 5654.11, "total_tokens": 272369192}
+{"current_steps": 6230, "total_steps": 29704, "loss": 0.3810266494750977, "lr": 8.953638514156908e-06, "epoch": 0.4194721249663345, "percentage": 20.97, "elapsed_time": "13:23:28", "remaining_time": "2 days, 2:27:26", "throughput": 5654.37, "total_tokens": 272591248}
+{"current_steps": 6235, "total_steps": 29704, "loss": 0.3950758695602417, "lr": 8.952019338850481e-06, "epoch": 0.41980877996229465, "percentage": 20.99, "elapsed_time": "13:24:08", "remaining_time": "2 days, 2:26:50", "throughput": 5654.28, "total_tokens": 272810616}
+{"current_steps": 6240, "total_steps": 29704, "loss": 0.377358603477478, "lr": 8.950399058374447e-06, "epoch": 0.4201454349582548, "percentage": 21.01, "elapsed_time": "13:24:45", "remaining_time": "2 days, 2:26:06", "throughput": 5654.6, "total_tokens": 273036976}
+{"current_steps": 6245, "total_steps": 29704, "loss": 0.389499831199646, "lr": 8.948777673181913e-06, "epoch": 0.42048208995421493, "percentage": 21.02, "elapsed_time": "13:25:22", "remaining_time": "2 days, 2:25:20", "throughput": 5654.95, "total_tokens": 273260832}
+{"current_steps": 6250, "total_steps": 29704, "loss": 0.35937402248382566, "lr": 8.947155183726296e-06, "epoch": 0.42081874495017507, "percentage": 21.04, "elapsed_time": "13:26:01", "remaining_time": "2 days, 2:24:43", "throughput": 5655.09, "total_tokens": 273488360}
+{"current_steps": 6255, "total_steps": 29704, "loss": 0.3715068817138672, "lr": 8.94553159046132e-06, "epoch": 0.4211553999461352, "percentage": 21.06, "elapsed_time": "13:26:41", "remaining_time": "2 days, 2:24:08", "throughput": 5654.98, "total_tokens": 273707472}
+{"current_steps": 6260, "total_steps": 29704, "loss": 0.36748714447021485, "lr": 8.943906893841015e-06, "epoch": 0.42149205494209535, "percentage": 21.07, "elapsed_time": "13:27:13", "remaining_time": "2 days, 2:23:05", "throughput": 5655.78, "total_tokens": 273928544}
+{"current_steps": 6265, "total_steps": 29704, "loss": 0.4157464027404785, "lr": 8.942281094319721e-06, "epoch": 0.4218287099380555, "percentage": 21.09, "elapsed_time": "13:27:48", "remaining_time": "2 days, 2:22:14", "throughput": 5656.04, "total_tokens": 274141592}
+{"current_steps": 6270, "total_steps": 29704, "loss": 0.40111403465270995, "lr": 8.940654192352092e-06, "epoch": 0.42216536493401563, "percentage": 21.11, "elapsed_time": "13:28:30", "remaining_time": "2 days, 2:21:46", "throughput": 5655.49, "total_tokens": 274349792}
+{"current_steps": 6275, "total_steps": 29704, "loss": 0.3853888034820557, "lr": 8.939026188393083e-06, "epoch": 0.42250201992997577, "percentage": 21.13, "elapsed_time": "13:29:16", "remaining_time": "2 days, 2:21:35", "throughput": 5654.63, "total_tokens": 274569104}
+{"current_steps": 6280, "total_steps": 29704, "loss": 0.40306596755981444, "lr": 8.937397082897963e-06, "epoch": 0.4228386749259359, "percentage": 21.14, "elapsed_time": "13:29:56", "remaining_time": "2 days, 2:21:02", "throughput": 5654.64, "total_tokens": 274796616}
+{"current_steps": 6285, "total_steps": 29704, "loss": 0.4140591621398926, "lr": 8.935766876322303e-06, "epoch": 0.42317532992189605, "percentage": 21.16, "elapsed_time": "13:30:35", "remaining_time": "2 days, 2:20:23", "throughput": 5654.66, "total_tokens": 275015624}
+{"current_steps": 6290, "total_steps": 29704, "loss": 0.3959988832473755, "lr": 8.934135569121985e-06, "epoch": 0.4235119849178562, "percentage": 21.18, "elapsed_time": "13:31:13", "remaining_time": "2 days, 2:19:41", "throughput": 5654.94, "total_tokens": 275242840}
+{"current_steps": 6295, "total_steps": 29704, "loss": 0.38456072807312014, "lr": 8.932503161753202e-06, "epoch": 0.4238486399138163, "percentage": 21.19, "elapsed_time": "13:31:49", "remaining_time": "2 days, 2:18:53", "throughput": 5655.22, "total_tokens": 275462504}
+{"current_steps": 6300, "total_steps": 29704, "loss": 0.38416409492492676, "lr": 8.930869654672449e-06, "epoch": 0.42418529490977647, "percentage": 21.21, "elapsed_time": "13:32:27", "remaining_time": "2 days, 2:18:13", "throughput": 5655.28, "total_tokens": 275681808}
+{"current_steps": 6305, "total_steps": 29704, "loss": 0.37823896408081054, "lr": 8.929235048336531e-06, "epoch": 0.4245219499057366, "percentage": 21.23, "elapsed_time": "13:33:08", "remaining_time": "2 days, 2:17:43", "throughput": 5655.08, "total_tokens": 275903576}
+{"current_steps": 6310, "total_steps": 29704, "loss": 0.4197996139526367, "lr": 8.927599343202563e-06, "epoch": 0.42485860490169675, "percentage": 21.24, "elapsed_time": "13:33:41", "remaining_time": "2 days, 2:16:44", "throughput": 5655.94, "total_tokens": 276132792}
+{"current_steps": 6315, "total_steps": 29704, "loss": 0.39647321701049804, "lr": 8.92596253972796e-06, "epoch": 0.4251952598976569, "percentage": 21.26, "elapsed_time": "13:34:16", "remaining_time": "2 days, 2:15:49", "throughput": 5656.9, "total_tokens": 276373576}
+{"current_steps": 6320, "total_steps": 29704, "loss": 0.4159084320068359, "lr": 8.924324638370454e-06, "epoch": 0.425531914893617, "percentage": 21.28, "elapsed_time": "13:34:52", "remaining_time": "2 days, 2:15:01", "throughput": 5657.37, "total_tokens": 276601344}
+{"current_steps": 6325, "total_steps": 29704, "loss": 0.3394161701202393, "lr": 8.922685639588076e-06, "epoch": 0.42586856988957716, "percentage": 21.29, "elapsed_time": "13:35:29", "remaining_time": "2 days, 2:14:17", "throughput": 5657.15, "total_tokens": 276801216}
+{"current_steps": 6330, "total_steps": 29704, "loss": 0.412354850769043, "lr": 8.921045543839167e-06, "epoch": 0.4262052248855373, "percentage": 21.31, "elapsed_time": "13:36:04", "remaining_time": "2 days, 2:13:24", "throughput": 5657.7, "total_tokens": 277025984}
+{"current_steps": 6335, "total_steps": 29704, "loss": 0.37722196578979494, "lr": 8.919404351582376e-06, "epoch": 0.42654187988149744, "percentage": 21.33, "elapsed_time": "13:36:40", "remaining_time": "2 days, 2:12:36", "throughput": 5657.79, "total_tokens": 277234256}
+{"current_steps": 6340, "total_steps": 29704, "loss": 0.38721909523010256, "lr": 8.917762063276657e-06, "epoch": 0.4268785348774576, "percentage": 21.34, "elapsed_time": "13:37:19", "remaining_time": "2 days, 2:11:58", "throughput": 5657.68, "total_tokens": 277448936}
+{"current_steps": 6345, "total_steps": 29704, "loss": 0.3728447198867798, "lr": 8.916118679381268e-06, "epoch": 0.4272151898734177, "percentage": 21.36, "elapsed_time": "13:37:56", "remaining_time": "2 days, 2:11:15", "throughput": 5657.81, "total_tokens": 277667920}
+{"current_steps": 6350, "total_steps": 29704, "loss": 0.39159350395202636, "lr": 8.914474200355777e-06, "epoch": 0.42755184486937786, "percentage": 21.38, "elapsed_time": "13:38:41", "remaining_time": "2 days, 2:10:59", "throughput": 5657.09, "total_tokens": 277885288}
+{"current_steps": 6355, "total_steps": 29704, "loss": 0.4010343074798584, "lr": 8.912828626660059e-06, "epoch": 0.427888499865338, "percentage": 21.39, "elapsed_time": "13:39:21", "remaining_time": "2 days, 2:10:24", "throughput": 5657.14, "total_tokens": 278112192}
+{"current_steps": 6360, "total_steps": 29704, "loss": 0.4207916259765625, "lr": 8.911181958754294e-06, "epoch": 0.42822515486129814, "percentage": 21.41, "elapsed_time": "13:39:58", "remaining_time": "2 days, 2:09:39", "throughput": 5657.33, "total_tokens": 278332088}
+{"current_steps": 6365, "total_steps": 29704, "loss": 0.4089661121368408, "lr": 8.909534197098966e-06, "epoch": 0.4285618098572583, "percentage": 21.43, "elapsed_time": "13:40:40", "remaining_time": "2 days, 2:09:13", "throughput": 5657.16, "total_tokens": 278560320}
+{"current_steps": 6370, "total_steps": 29704, "loss": 0.3871757507324219, "lr": 8.907885342154865e-06, "epoch": 0.4288984648532184, "percentage": 21.44, "elapsed_time": "13:41:13", "remaining_time": "2 days, 2:08:14", "throughput": 5657.81, "total_tokens": 278780408}
+{"current_steps": 6375, "total_steps": 29704, "loss": 0.3723259925842285, "lr": 8.906235394383089e-06, "epoch": 0.42923511984917856, "percentage": 21.46, "elapsed_time": "13:41:54", "remaining_time": "2 days, 2:07:44", "throughput": 5657.11, "total_tokens": 278977576}
+{"current_steps": 6380, "total_steps": 29704, "loss": 0.38217949867248535, "lr": 8.904584354245042e-06, "epoch": 0.4295717748451387, "percentage": 21.48, "elapsed_time": "13:42:29", "remaining_time": "2 days, 2:06:50", "throughput": 5657.61, "total_tokens": 279198056}
+{"current_steps": 6385, "total_steps": 29704, "loss": 0.39810030460357665, "lr": 8.902932222202433e-06, "epoch": 0.42990842984109884, "percentage": 21.5, "elapsed_time": "13:43:10", "remaining_time": "2 days, 2:06:22", "throughput": 5657.56, "total_tokens": 279431128}
+{"current_steps": 6390, "total_steps": 29704, "loss": 0.37800168991088867, "lr": 8.901278998717272e-06, "epoch": 0.430245084837059, "percentage": 21.51, "elapsed_time": "13:43:40", "remaining_time": "2 days, 2:05:12", "throughput": 5658.65, "total_tokens": 279655568}
+{"current_steps": 6395, "total_steps": 29704, "loss": 0.38466432094573977, "lr": 8.899624684251878e-06, "epoch": 0.4305817398330191, "percentage": 21.53, "elapsed_time": "13:44:21", "remaining_time": "2 days, 2:04:39", "throughput": 5658.45, "total_tokens": 279872920}
+{"current_steps": 6400, "total_steps": 29704, "loss": 0.37961535453796386, "lr": 8.897969279268877e-06, "epoch": 0.43091839482897926, "percentage": 21.55, "elapsed_time": "13:44:54", "remaining_time": "2 days, 2:03:42", "throughput": 5659.07, "total_tokens": 280093688}
+{"current_steps": 6405, "total_steps": 29704, "loss": 0.3712193012237549, "lr": 8.896312784231196e-06, "epoch": 0.4312550498249394, "percentage": 21.56, "elapsed_time": "13:45:29", "remaining_time": "2 days, 2:02:49", "throughput": 5659.38, "total_tokens": 280305104}
+{"current_steps": 6410, "total_steps": 29704, "loss": 0.42307000160217284, "lr": 8.89465519960207e-06, "epoch": 0.43159170482089954, "percentage": 21.58, "elapsed_time": "13:46:02", "remaining_time": "2 days, 2:01:51", "throughput": 5660.2, "total_tokens": 280534896}
+{"current_steps": 6415, "total_steps": 29704, "loss": 0.39340283870697024, "lr": 8.892996525845037e-06, "epoch": 0.4319283598168597, "percentage": 21.6, "elapsed_time": "13:46:49", "remaining_time": "2 days, 2:01:40", "throughput": 5659.45, "total_tokens": 280760536}
+{"current_steps": 6420, "total_steps": 29704, "loss": 0.4034512996673584, "lr": 8.89133676342394e-06, "epoch": 0.4322650148128198, "percentage": 21.61, "elapsed_time": "13:47:27", "remaining_time": "2 days, 2:01:02", "throughput": 5659.06, "total_tokens": 280959600}
+{"current_steps": 6425, "total_steps": 29704, "loss": 0.38754520416259763, "lr": 8.889675912802923e-06, "epoch": 0.43260166980877995, "percentage": 21.63, "elapsed_time": "13:48:04", "remaining_time": "2 days, 2:00:16", "throughput": 5659.19, "total_tokens": 281173760}
+{"current_steps": 6430, "total_steps": 29704, "loss": 0.4189424514770508, "lr": 8.888013974446443e-06, "epoch": 0.4329383248047401, "percentage": 21.65, "elapsed_time": "13:48:45", "remaining_time": "2 days, 1:59:46", "throughput": 5658.97, "total_tokens": 281395808}
+{"current_steps": 6435, "total_steps": 29704, "loss": 0.38129727840423583, "lr": 8.886350948819253e-06, "epoch": 0.43327497980070023, "percentage": 21.66, "elapsed_time": "13:49:18", "remaining_time": "2 days, 1:58:46", "throughput": 5659.46, "total_tokens": 281606072}
+{"current_steps": 6440, "total_steps": 29704, "loss": 0.38291594982147215, "lr": 8.884686836386412e-06, "epoch": 0.4336116347966604, "percentage": 21.68, "elapsed_time": "13:49:59", "remaining_time": "2 days, 1:58:17", "throughput": 5659.11, "total_tokens": 281822424}
+{"current_steps": 6445, "total_steps": 29704, "loss": 0.4050453662872314, "lr": 8.883021637613286e-06, "epoch": 0.4339482897926205, "percentage": 21.7, "elapsed_time": "13:50:35", "remaining_time": "2 days, 1:57:26", "throughput": 5659.67, "total_tokens": 282049648}
+{"current_steps": 6450, "total_steps": 29704, "loss": 0.38397510051727296, "lr": 8.881355352965537e-06, "epoch": 0.43428494478858065, "percentage": 21.71, "elapsed_time": "13:51:07", "remaining_time": "2 days, 1:56:26", "throughput": 5660.28, "total_tokens": 282264512}
+{"current_steps": 6455, "total_steps": 29704, "loss": 0.3574419736862183, "lr": 8.879687982909145e-06, "epoch": 0.4346215997845408, "percentage": 21.73, "elapsed_time": "13:51:45", "remaining_time": "2 days, 1:55:45", "throughput": 5660.0, "total_tokens": 282465912}
+{"current_steps": 6460, "total_steps": 29704, "loss": 0.41092991828918457, "lr": 8.878019527910378e-06, "epoch": 0.43495825478050093, "percentage": 21.75, "elapsed_time": "13:52:25", "remaining_time": "2 days, 1:55:09", "throughput": 5659.63, "total_tokens": 282670240}
+{"current_steps": 6465, "total_steps": 29704, "loss": 0.44412593841552733, "lr": 8.876349988435815e-06, "epoch": 0.43529490977646107, "percentage": 21.76, "elapsed_time": "13:53:03", "remaining_time": "2 days, 1:54:28", "throughput": 5659.67, "total_tokens": 282887856}
+{"current_steps": 6470, "total_steps": 29704, "loss": 0.38706722259521487, "lr": 8.874679364952339e-06, "epoch": 0.4356315647724212, "percentage": 21.78, "elapsed_time": "13:53:41", "remaining_time": "2 days, 1:53:48", "throughput": 5660.12, "total_tokens": 283127040}
+{"current_steps": 6475, "total_steps": 29704, "loss": 0.3893885135650635, "lr": 8.873007657927135e-06, "epoch": 0.43596821976838135, "percentage": 21.8, "elapsed_time": "13:54:22", "remaining_time": "2 days, 1:53:18", "throughput": 5660.07, "total_tokens": 283357232}
+{"current_steps": 6480, "total_steps": 29704, "loss": 0.35058715343475344, "lr": 8.87133486782769e-06, "epoch": 0.4363048747643415, "percentage": 21.82, "elapsed_time": "13:54:56", "remaining_time": "2 days, 1:52:23", "throughput": 5660.24, "total_tokens": 283558232}
+{"current_steps": 6485, "total_steps": 29704, "loss": 0.3999923229217529, "lr": 8.869660995121792e-06, "epoch": 0.4366415297603016, "percentage": 21.83, "elapsed_time": "13:55:31", "remaining_time": "2 days, 1:51:32", "throughput": 5660.63, "total_tokens": 283777800}
+{"current_steps": 6490, "total_steps": 29704, "loss": 0.4016421318054199, "lr": 8.867986040277537e-06, "epoch": 0.43697818475626177, "percentage": 21.85, "elapsed_time": "13:56:12", "remaining_time": "2 days, 1:51:02", "throughput": 5660.14, "total_tokens": 283985824}
+{"current_steps": 6495, "total_steps": 29704, "loss": 0.4068891525268555, "lr": 8.86631000376332e-06, "epoch": 0.4373148397522219, "percentage": 21.87, "elapsed_time": "13:56:50", "remaining_time": "2 days, 1:50:19", "throughput": 5660.08, "total_tokens": 284193496}
+{"current_steps": 6500, "total_steps": 29704, "loss": 0.3719168186187744, "lr": 8.864632886047841e-06, "epoch": 0.43765149474818205, "percentage": 21.88, "elapsed_time": "13:57:26", "remaining_time": "2 days, 1:49:31", "throughput": 5660.26, "total_tokens": 284406672}
+{"current_steps": 6505, "total_steps": 29704, "loss": 0.4161073684692383, "lr": 8.862954687600097e-06, "epoch": 0.4379881497441422, "percentage": 21.9, "elapsed_time": "13:58:03", "remaining_time": "2 days, 1:48:49", "throughput": 5660.26, "total_tokens": 284620064}
+{"current_steps": 6510, "total_steps": 29704, "loss": 0.4044035911560059, "lr": 8.861275408889393e-06, "epoch": 0.4383248047401023, "percentage": 21.92, "elapsed_time": "13:58:39", "remaining_time": "2 days, 1:48:00", "throughput": 5660.36, "total_tokens": 284827680}
+{"current_steps": 6515, "total_steps": 29704, "loss": 0.40093269348144533, "lr": 8.859595050385332e-06, "epoch": 0.43866145973606246, "percentage": 21.93, "elapsed_time": "13:59:16", "remaining_time": "2 days, 1:47:15", "throughput": 5660.62, "total_tokens": 285050288}
+{"current_steps": 6520, "total_steps": 29704, "loss": 0.39177913665771485, "lr": 8.857913612557825e-06, "epoch": 0.4389981147320226, "percentage": 21.95, "elapsed_time": "13:59:59", "remaining_time": "2 days, 1:46:53", "throughput": 5660.25, "total_tokens": 285275776}
+{"current_steps": 6525, "total_steps": 29704, "loss": 0.37299246788024903, "lr": 8.856231095877077e-06, "epoch": 0.43933476972798274, "percentage": 21.97, "elapsed_time": "14:00:37", "remaining_time": "2 days, 1:46:11", "throughput": 5660.13, "total_tokens": 285483840}
+{"current_steps": 6530, "total_steps": 29704, "loss": 0.3728148937225342, "lr": 8.8545475008136e-06, "epoch": 0.4396714247239429, "percentage": 21.98, "elapsed_time": "14:01:18", "remaining_time": "2 days, 1:45:39", "throughput": 5660.04, "total_tokens": 285708616}
+{"current_steps": 6535, "total_steps": 29704, "loss": 0.37645254135131834, "lr": 8.852862827838207e-06, "epoch": 0.440008079719903, "percentage": 22.0, "elapsed_time": "14:01:53", "remaining_time": "2 days, 1:44:50", "throughput": 5660.12, "total_tokens": 285914952}
+{"current_steps": 6540, "total_steps": 29704, "loss": 0.4013407230377197, "lr": 8.851177077422008e-06, "epoch": 0.44034473471586316, "percentage": 22.02, "elapsed_time": "14:02:33", "remaining_time": "2 days, 1:44:15", "throughput": 5660.27, "total_tokens": 286146336}
+{"current_steps": 6545, "total_steps": 29704, "loss": 0.4249086380004883, "lr": 8.849490250036421e-06, "epoch": 0.4406813897118233, "percentage": 22.03, "elapsed_time": "14:03:10", "remaining_time": "2 days, 1:43:29", "throughput": 5660.43, "total_tokens": 286362088}
+{"current_steps": 6550, "total_steps": 29704, "loss": 0.3778714895248413, "lr": 8.847802346153159e-06, "epoch": 0.44101804470778344, "percentage": 22.05, "elapsed_time": "14:03:43", "remaining_time": "2 days, 1:42:30", "throughput": 5660.97, "total_tokens": 286576504}
+{"current_steps": 6555, "total_steps": 29704, "loss": 0.40006532669067385, "lr": 8.846113366244241e-06, "epoch": 0.4413546997037436, "percentage": 22.07, "elapsed_time": "14:04:13", "remaining_time": "2 days, 1:41:23", "throughput": 5662.02, "total_tokens": 286802512}
+{"current_steps": 6560, "total_steps": 29704, "loss": 0.3659277677536011, "lr": 8.844423310781986e-06, "epoch": 0.4416913546997037, "percentage": 22.08, "elapsed_time": "14:04:53", "remaining_time": "2 days, 1:40:49", "throughput": 5661.87, "total_tokens": 287019968}
+{"current_steps": 6565, "total_steps": 29704, "loss": 0.417068338394165, "lr": 8.842732180239011e-06, "epoch": 0.44202800969566386, "percentage": 22.1, "elapsed_time": "14:05:32", "remaining_time": "2 days, 1:40:13", "throughput": 5661.68, "total_tokens": 287233880}
+{"current_steps": 6570, "total_steps": 29704, "loss": 0.37861127853393556, "lr": 8.841039975088234e-06, "epoch": 0.442364664691624, "percentage": 22.12, "elapsed_time": "14:06:10", "remaining_time": "2 days, 1:39:32", "throughput": 5661.72, "total_tokens": 287450920}
+{"current_steps": 6575, "total_steps": 29704, "loss": 0.44512476921081545, "lr": 8.839346695802878e-06, "epoch": 0.44270131968758414, "percentage": 22.14, "elapsed_time": "14:06:48", "remaining_time": "2 days, 1:38:50", "throughput": 5662.01, "total_tokens": 287680008}
+{"current_steps": 6580, "total_steps": 29704, "loss": 0.4026740550994873, "lr": 8.837652342856459e-06, "epoch": 0.4430379746835443, "percentage": 22.15, "elapsed_time": "14:07:22", "remaining_time": "2 days, 1:37:55", "throughput": 5662.67, "total_tokens": 287904440}
+{"current_steps": 6585, "total_steps": 29704, "loss": 0.37171001434326173, "lr": 8.835956916722803e-06, "epoch": 0.4433746296795044, "percentage": 22.17, "elapsed_time": "14:08:04", "remaining_time": "2 days, 1:37:27", "throughput": 5662.31, "total_tokens": 288123072}
+{"current_steps": 6590, "total_steps": 29704, "loss": 0.39625725746154783, "lr": 8.834260417876024e-06, "epoch": 0.44371128467546456, "percentage": 22.19, "elapsed_time": "14:08:42", "remaining_time": "2 days, 1:36:47", "throughput": 5662.27, "total_tokens": 288337792}
+{"current_steps": 6595, "total_steps": 29704, "loss": 0.4032885074615479, "lr": 8.832562846790549e-06, "epoch": 0.4440479396714247, "percentage": 22.2, "elapsed_time": "14:09:18", "remaining_time": "2 days, 1:35:58", "throughput": 5662.73, "total_tokens": 288562504}
+{"current_steps": 6600, "total_steps": 29704, "loss": 0.4127056121826172, "lr": 8.830864203941092e-06, "epoch": 0.4443845946673849, "percentage": 22.22, "elapsed_time": "14:09:54", "remaining_time": "2 days, 1:35:13", "throughput": 5662.98, "total_tokens": 288783304}
+{"current_steps": 6605, "total_steps": 29704, "loss": 0.37214198112487795, "lr": 8.829164489802677e-06, "epoch": 0.44472124966334503, "percentage": 22.24, "elapsed_time": "14:10:30", "remaining_time": "2 days, 1:34:24", "throughput": 5663.17, "total_tokens": 288995032}
+{"current_steps": 6610, "total_steps": 29704, "loss": 0.3779730796813965, "lr": 8.827463704850622e-06, "epoch": 0.44505790465930517, "percentage": 22.25, "elapsed_time": "14:11:05", "remaining_time": "2 days, 1:33:31", "throughput": 5663.58, "total_tokens": 289211784}
+{"current_steps": 6615, "total_steps": 29704, "loss": 0.3779446125030518, "lr": 8.825761849560547e-06, "epoch": 0.4453945596552653, "percentage": 22.27, "elapsed_time": "14:11:44", "remaining_time": "2 days, 1:32:55", "throughput": 5663.42, "total_tokens": 289426496}
+{"current_steps": 6620, "total_steps": 29704, "loss": 0.3913204431533813, "lr": 8.824058924408371e-06, "epoch": 0.44573121465122545, "percentage": 22.29, "elapsed_time": "14:12:28", "remaining_time": "2 days, 1:32:34", "throughput": 5662.79, "total_tokens": 289642104}
+{"current_steps": 6625, "total_steps": 29704, "loss": 0.3713568687438965, "lr": 8.822354929870311e-06, "epoch": 0.4460678696471856, "percentage": 22.3, "elapsed_time": "14:13:06", "remaining_time": "2 days, 1:31:54", "throughput": 5662.73, "total_tokens": 289855080}
+{"current_steps": 6630, "total_steps": 29704, "loss": 0.3891900539398193, "lr": 8.820649866422884e-06, "epoch": 0.44640452464314573, "percentage": 22.32, "elapsed_time": "14:13:41", "remaining_time": "2 days, 1:31:01", "throughput": 5663.31, "total_tokens": 290080272}
+{"current_steps": 6635, "total_steps": 29704, "loss": 0.38543388843536375, "lr": 8.818943734542905e-06, "epoch": 0.44674117963910587, "percentage": 22.34, "elapsed_time": "14:14:20", "remaining_time": "2 days, 1:30:26", "throughput": 5663.42, "total_tokens": 290310296}
+{"current_steps": 6640, "total_steps": 29704, "loss": 0.38082070350646974, "lr": 8.817236534707486e-06, "epoch": 0.447077834635066, "percentage": 22.35, "elapsed_time": "14:14:58", "remaining_time": "2 days, 1:29:44", "throughput": 5663.56, "total_tokens": 290530408}
+{"current_steps": 6645, "total_steps": 29704, "loss": 0.38055622577667236, "lr": 8.815528267394045e-06, "epoch": 0.44741448963102615, "percentage": 22.37, "elapsed_time": "14:15:39", "remaining_time": "2 days, 1:29:14", "throughput": 5663.33, "total_tokens": 290752096}
+{"current_steps": 6650, "total_steps": 29704, "loss": 0.42238750457763674, "lr": 8.81381893308029e-06, "epoch": 0.4477511446269863, "percentage": 22.39, "elapsed_time": "14:16:16", "remaining_time": "2 days, 1:28:30", "throughput": 5663.59, "total_tokens": 290976696}
+{"current_steps": 6655, "total_steps": 29704, "loss": 0.397933292388916, "lr": 8.81210853224423e-06, "epoch": 0.4480877996229464, "percentage": 22.4, "elapsed_time": "14:16:56", "remaining_time": "2 days, 1:27:54", "throughput": 5663.5, "total_tokens": 291195128}
+{"current_steps": 6660, "total_steps": 29704, "loss": 0.3660165309906006, "lr": 8.810397065364177e-06, "epoch": 0.44842445461890656, "percentage": 22.42, "elapsed_time": "14:17:40", "remaining_time": "2 days, 1:27:35", "throughput": 5662.58, "total_tokens": 291397432}
+{"current_steps": 6665, "total_steps": 29704, "loss": 0.4035042762756348, "lr": 8.808684532918735e-06, "epoch": 0.4487611096148667, "percentage": 22.44, "elapsed_time": "14:18:17", "remaining_time": "2 days, 1:26:51", "throughput": 5662.67, "total_tokens": 291611984}
+{"current_steps": 6670, "total_steps": 29704, "loss": 0.37046058177948, "lr": 8.806970935386807e-06, "epoch": 0.44909776461082684, "percentage": 22.45, "elapsed_time": "14:18:59", "remaining_time": "2 days, 1:26:26", "throughput": 5661.99, "total_tokens": 291817952}
+{"current_steps": 6675, "total_steps": 29704, "loss": 0.3828955411911011, "lr": 8.805256273247597e-06, "epoch": 0.449434419606787, "percentage": 22.47, "elapsed_time": "14:19:45", "remaining_time": "2 days, 1:26:12", "throughput": 5661.07, "total_tokens": 292030296}
+{"current_steps": 6680, "total_steps": 29704, "loss": 0.3911006450653076, "lr": 8.803540546980605e-06, "epoch": 0.4497710746027471, "percentage": 22.49, "elapsed_time": "14:20:25", "remaining_time": "2 days, 1:25:38", "throughput": 5660.77, "total_tokens": 292240392}
+{"current_steps": 6685, "total_steps": 29704, "loss": 0.39244418144226073, "lr": 8.801823757065628e-06, "epoch": 0.45010772959870726, "percentage": 22.51, "elapsed_time": "14:21:01", "remaining_time": "2 days, 1:24:49", "throughput": 5661.04, "total_tokens": 292456784}
+{"current_steps": 6690, "total_steps": 29704, "loss": 0.37194869518280027, "lr": 8.800105903982758e-06, "epoch": 0.4504443845946674, "percentage": 22.52, "elapsed_time": "14:21:35", "remaining_time": "2 days, 1:23:56", "throughput": 5661.38, "total_tokens": 292669672}
+{"current_steps": 6695, "total_steps": 29704, "loss": 0.36289196014404296, "lr": 8.798386988212393e-06, "epoch": 0.45078103959062754, "percentage": 22.54, "elapsed_time": "14:22:17", "remaining_time": "2 days, 1:23:29", "throughput": 5660.6, "total_tokens": 292866896}
+{"current_steps": 6700, "total_steps": 29704, "loss": 0.4039313793182373, "lr": 8.796667010235216e-06, "epoch": 0.4511176945865877, "percentage": 22.56, "elapsed_time": "14:22:55", "remaining_time": "2 days, 1:22:46", "throughput": 5660.8, "total_tokens": 293088104}
+{"current_steps": 6705, "total_steps": 29704, "loss": 0.38548543453216555, "lr": 8.794945970532219e-06, "epoch": 0.4514543495825478, "percentage": 22.57, "elapsed_time": "14:23:39", "remaining_time": "2 days, 1:22:26", "throughput": 5660.32, "total_tokens": 293312720}
+{"current_steps": 6710, "total_steps": 29704, "loss": 0.42191410064697266, "lr": 8.79322386958468e-06, "epoch": 0.45179100457850796, "percentage": 22.59, "elapsed_time": "14:24:10", "remaining_time": "2 days, 1:21:22", "throughput": 5661.28, "total_tokens": 293540040}
+{"current_steps": 6715, "total_steps": 29704, "loss": 0.3590597152709961, "lr": 8.791500707874183e-06, "epoch": 0.4521276595744681, "percentage": 22.61, "elapsed_time": "14:24:47", "remaining_time": "2 days, 1:20:40", "throughput": 5661.3, "total_tokens": 293753504}
+{"current_steps": 6720, "total_steps": 29704, "loss": 0.40479679107666017, "lr": 8.789776485882601e-06, "epoch": 0.45246431457042824, "percentage": 22.62, "elapsed_time": "14:25:30", "remaining_time": "2 days, 1:20:15", "throughput": 5661.02, "total_tokens": 293980328}
+{"current_steps": 6725, "total_steps": 29704, "loss": 0.380597448348999, "lr": 8.788051204092112e-06, "epoch": 0.4528009695663884, "percentage": 22.64, "elapsed_time": "14:26:09", "remaining_time": "2 days, 1:19:36", "throughput": 5661.08, "total_tokens": 294201288}
+{"current_steps": 6730, "total_steps": 29704, "loss": 0.39116497039794923, "lr": 8.786324862985183e-06, "epoch": 0.4531376245623485, "percentage": 22.66, "elapsed_time": "14:26:47", "remaining_time": "2 days, 1:18:56", "throughput": 5661.01, "total_tokens": 294414792}
+{"current_steps": 6735, "total_steps": 29704, "loss": 0.37044734954833985, "lr": 8.78459746304458e-06, "epoch": 0.45347427955830866, "percentage": 22.67, "elapsed_time": "14:27:28", "remaining_time": "2 days, 1:18:25", "throughput": 5660.0, "total_tokens": 294593136}
+{"current_steps": 6740, "total_steps": 29704, "loss": 0.40142025947570803, "lr": 8.782869004753363e-06, "epoch": 0.4538109345542688, "percentage": 22.69, "elapsed_time": "14:28:04", "remaining_time": "2 days, 1:17:36", "throughput": 5660.24, "total_tokens": 294808392}
+{"current_steps": 6745, "total_steps": 29704, "loss": 0.39638261795043944, "lr": 8.781139488594892e-06, "epoch": 0.45414758955022894, "percentage": 22.71, "elapsed_time": "14:28:36", "remaining_time": "2 days, 1:16:37", "throughput": 5661.01, "total_tokens": 295032712}
+{"current_steps": 6750, "total_steps": 29704, "loss": 0.3619338274002075, "lr": 8.779408915052821e-06, "epoch": 0.4544842445461891, "percentage": 22.72, "elapsed_time": "14:29:18", "remaining_time": "2 days, 1:16:09", "throughput": 5660.64, "total_tokens": 295250848}
+{"current_steps": 6755, "total_steps": 29704, "loss": 0.36865906715393065, "lr": 8.777677284611096e-06, "epoch": 0.4548208995421492, "percentage": 22.74, "elapsed_time": "14:29:54", "remaining_time": "2 days, 1:15:23", "throughput": 5660.42, "total_tokens": 295444808}
+{"current_steps": 6760, "total_steps": 29704, "loss": 0.3817254304885864, "lr": 8.775944597753969e-06, "epoch": 0.45515755453810935, "percentage": 22.76, "elapsed_time": "14:30:35", "remaining_time": "2 days, 1:14:51", "throughput": 5659.97, "total_tokens": 295651640}
+{"current_steps": 6765, "total_steps": 29704, "loss": 0.4110917568206787, "lr": 8.774210854965972e-06, "epoch": 0.4554942095340695, "percentage": 22.77, "elapsed_time": "14:31:22", "remaining_time": "2 days, 1:14:41", "throughput": 5659.11, "total_tokens": 295872600}
+{"current_steps": 6770, "total_steps": 29704, "loss": 0.3540276050567627, "lr": 8.772476056731946e-06, "epoch": 0.45583086453002963, "percentage": 22.79, "elapsed_time": "14:32:01", "remaining_time": "2 days, 1:14:04", "throughput": 5659.25, "total_tokens": 296100760}
+{"current_steps": 6775, "total_steps": 29704, "loss": 0.37112271785736084, "lr": 8.77074020353702e-06, "epoch": 0.4561675195259898, "percentage": 22.81, "elapsed_time": "14:32:44", "remaining_time": "2 days, 1:13:40", "throughput": 5658.61, "total_tokens": 296310112}
+{"current_steps": 6780, "total_steps": 29704, "loss": 0.416597318649292, "lr": 8.76900329586662e-06, "epoch": 0.4565041745219499, "percentage": 22.83, "elapsed_time": "14:33:28", "remaining_time": "2 days, 1:13:19", "throughput": 5658.04, "total_tokens": 296529072}
+{"current_steps": 6785, "total_steps": 29704, "loss": 0.38317620754241943, "lr": 8.767265334206467e-06, "epoch": 0.45684082951791005, "percentage": 22.84, "elapsed_time": "14:34:13", "remaining_time": "2 days, 1:13:01", "throughput": 5657.43, "total_tokens": 296750896}
+{"current_steps": 6790, "total_steps": 29704, "loss": 0.38066811561584474, "lr": 8.765526319042577e-06, "epoch": 0.4571774845138702, "percentage": 22.86, "elapsed_time": "14:34:44", "remaining_time": "2 days, 1:11:58", "throughput": 5658.27, "total_tokens": 296973576}
+{"current_steps": 6795, "total_steps": 29704, "loss": 0.35711822509765623, "lr": 8.763786250861258e-06, "epoch": 0.45751413950983033, "percentage": 22.88, "elapsed_time": "14:35:25", "remaining_time": "2 days, 1:11:28", "throughput": 5658.13, "total_tokens": 297198488}
+{"current_steps": 6800, "total_steps": 29704, "loss": 0.37547504901885986, "lr": 8.762045130149114e-06, "epoch": 0.45785079450579047, "percentage": 22.89, "elapsed_time": "14:36:08", "remaining_time": "2 days, 1:11:01", "throughput": 5657.88, "total_tokens": 297424408}
+{"current_steps": 6805, "total_steps": 29704, "loss": 0.391666579246521, "lr": 8.76030295739305e-06, "epoch": 0.4581874495017506, "percentage": 22.91, "elapsed_time": "14:36:45", "remaining_time": "2 days, 1:10:20", "throughput": 5658.16, "total_tokens": 297652664}
+{"current_steps": 6810, "total_steps": 29704, "loss": 0.4111741065979004, "lr": 8.758559733080252e-06, "epoch": 0.45852410449771075, "percentage": 22.93, "elapsed_time": "14:37:17", "remaining_time": "2 days, 1:09:16", "throughput": 5659.03, "total_tokens": 297874936}
+{"current_steps": 6815, "total_steps": 29704, "loss": 0.38690164089202883, "lr": 8.75681545769821e-06, "epoch": 0.4588607594936709, "percentage": 22.94, "elapsed_time": "14:37:49", "remaining_time": "2 days, 1:08:17", "throughput": 5659.95, "total_tokens": 298107584}
+{"current_steps": 6820, "total_steps": 29704, "loss": 0.3867367744445801, "lr": 8.755070131734706e-06, "epoch": 0.45919741448963103, "percentage": 22.96, "elapsed_time": "14:38:27", "remaining_time": "2 days, 1:07:37", "throughput": 5659.88, "total_tokens": 298320208}
+{"current_steps": 6825, "total_steps": 29704, "loss": 0.4013802528381348, "lr": 8.753323755677812e-06, "epoch": 0.45953406948559117, "percentage": 22.98, "elapsed_time": "14:39:09", "remaining_time": "2 days, 1:07:09", "throughput": 5659.45, "total_tokens": 298533552}
+{"current_steps": 6830, "total_steps": 29704, "loss": 0.3463205575942993, "lr": 8.7515763300159e-06, "epoch": 0.4598707244815513, "percentage": 22.99, "elapsed_time": "14:39:49", "remaining_time": "2 days, 1:06:34", "throughput": 5659.22, "total_tokens": 298746696}
+{"current_steps": 6835, "total_steps": 29704, "loss": 0.40210418701171874, "lr": 8.74982785523763e-06, "epoch": 0.46020737947751145, "percentage": 23.01, "elapsed_time": "14:40:27", "remaining_time": "2 days, 1:05:52", "throughput": 5659.44, "total_tokens": 298972384}
+{"current_steps": 6840, "total_steps": 29704, "loss": 0.3934351682662964, "lr": 8.748078331831957e-06, "epoch": 0.4605440344734716, "percentage": 23.03, "elapsed_time": "14:41:02", "remaining_time": "2 days, 1:05:04", "throughput": 5659.92, "total_tokens": 299200112}
+{"current_steps": 6845, "total_steps": 29704, "loss": 0.3836989164352417, "lr": 8.746327760288129e-06, "epoch": 0.4608806894694317, "percentage": 23.04, "elapsed_time": "14:41:34", "remaining_time": "2 days, 1:04:02", "throughput": 5661.03, "total_tokens": 299437024}
+{"current_steps": 6850, "total_steps": 29704, "loss": 0.39553108215332033, "lr": 8.744576141095691e-06, "epoch": 0.46121734446539187, "percentage": 23.06, "elapsed_time": "14:42:12", "remaining_time": "2 days, 1:03:22", "throughput": 5660.79, "total_tokens": 299641752}
+{"current_steps": 6855, "total_steps": 29704, "loss": 0.39920735359191895, "lr": 8.742823474744476e-06, "epoch": 0.461553999461352, "percentage": 23.08, "elapsed_time": "14:42:50", "remaining_time": "2 days, 1:02:40", "throughput": 5660.76, "total_tokens": 299852504}
+{"current_steps": 6860, "total_steps": 29704, "loss": 0.40232462882995607, "lr": 8.74106976172461e-06, "epoch": 0.46189065445731214, "percentage": 23.09, "elapsed_time": "14:43:33", "remaining_time": "2 days, 1:02:16", "throughput": 5660.3, "total_tokens": 300073192}
+{"current_steps": 6865, "total_steps": 29704, "loss": 0.40238103866577146, "lr": 8.739315002526513e-06, "epoch": 0.4622273094532723, "percentage": 23.11, "elapsed_time": "14:44:15", "remaining_time": "2 days, 1:01:47", "throughput": 5659.93, "total_tokens": 300287552}
+{"current_steps": 6870, "total_steps": 29704, "loss": 0.3717548131942749, "lr": 8.737559197640902e-06, "epoch": 0.4625639644492324, "percentage": 23.13, "elapsed_time": "14:44:51", "remaining_time": "2 days, 1:01:02", "throughput": 5659.95, "total_tokens": 300496376}
+{"current_steps": 6875, "total_steps": 29704, "loss": 0.4027840614318848, "lr": 8.735802347558778e-06, "epoch": 0.46290061944519256, "percentage": 23.15, "elapsed_time": "14:45:35", "remaining_time": "2 days, 1:00:39", "throughput": 5659.33, "total_tokens": 300709416}
+{"current_steps": 6880, "total_steps": 29704, "loss": 0.38804852962493896, "lr": 8.734044452771442e-06, "epoch": 0.4632372744411527, "percentage": 23.16, "elapsed_time": "14:46:19", "remaining_time": "2 days, 1:00:18", "throughput": 5658.59, "total_tokens": 300918752}
+{"current_steps": 6885, "total_steps": 29704, "loss": 0.40845584869384766, "lr": 8.73228551377048e-06, "epoch": 0.46357392943711284, "percentage": 23.18, "elapsed_time": "14:46:48", "remaining_time": "2 days, 0:59:10", "throughput": 5659.57, "total_tokens": 301139720}
+{"current_steps": 6890, "total_steps": 29704, "loss": 0.4137433052062988, "lr": 8.730525531047776e-06, "epoch": 0.463910584433073, "percentage": 23.2, "elapsed_time": "14:47:34", "remaining_time": "2 days, 0:58:53", "throughput": 5658.95, "total_tokens": 301362752}
+{"current_steps": 6895, "total_steps": 29704, "loss": 0.4120682716369629, "lr": 8.728764505095503e-06, "epoch": 0.4642472394290331, "percentage": 23.21, "elapsed_time": "14:48:11", "remaining_time": "2 days, 0:58:10", "throughput": 5659.21, "total_tokens": 301587944}
+{"current_steps": 6900, "total_steps": 29704, "loss": 0.3724491834640503, "lr": 8.727002436406126e-06, "epoch": 0.46458389442499326, "percentage": 23.23, "elapsed_time": "14:48:51", "remaining_time": "2 days, 0:57:37", "throughput": 5659.35, "total_tokens": 301823392}
+{"current_steps": 6905, "total_steps": 29704, "loss": 0.4057363510131836, "lr": 8.725239325472404e-06, "epoch": 0.4649205494209534, "percentage": 23.25, "elapsed_time": "14:49:30", "remaining_time": "2 days, 0:56:59", "throughput": 5659.43, "total_tokens": 302046392}
+{"current_steps": 6910, "total_steps": 29704, "loss": 0.3984755277633667, "lr": 8.723475172787381e-06, "epoch": 0.46525720441691354, "percentage": 23.26, "elapsed_time": "14:50:13", "remaining_time": "2 days, 0:56:34", "throughput": 5659.1, "total_tokens": 302270848}
+{"current_steps": 6915, "total_steps": 29704, "loss": 0.38761351108551023, "lr": 8.7217099788444e-06, "epoch": 0.4655938594128737, "percentage": 23.28, "elapsed_time": "14:50:52", "remaining_time": "2 days, 0:55:57", "throughput": 5658.92, "total_tokens": 302484424}
+{"current_steps": 6920, "total_steps": 29704, "loss": 0.3843079566955566, "lr": 8.719943744137091e-06, "epoch": 0.4659305144088338, "percentage": 23.3, "elapsed_time": "14:51:34", "remaining_time": "2 days, 0:55:29", "throughput": 5658.67, "total_tokens": 302706608}
+{"current_steps": 6925, "total_steps": 29704, "loss": 0.40040907859802244, "lr": 8.718176469159378e-06, "epoch": 0.46626716940479396, "percentage": 23.31, "elapsed_time": "14:52:07", "remaining_time": "2 days, 0:54:33", "throughput": 5659.6, "total_tokens": 302946080}
+{"current_steps": 6930, "total_steps": 29704, "loss": 0.370440411567688, "lr": 8.716408154405469e-06, "epoch": 0.4666038244007541, "percentage": 23.33, "elapsed_time": "14:52:38", "remaining_time": "2 days, 0:53:29", "throughput": 5660.58, "total_tokens": 303172472}
+{"current_steps": 6935, "total_steps": 29704, "loss": 0.4163699150085449, "lr": 8.714638800369872e-06, "epoch": 0.46694047939671424, "percentage": 23.35, "elapsed_time": "14:53:16", "remaining_time": "2 days, 0:52:49", "throughput": 5660.3, "total_tokens": 303374536}
+{"current_steps": 6940, "total_steps": 29704, "loss": 0.41262283325195315, "lr": 8.712868407547378e-06, "epoch": 0.4672771343926744, "percentage": 23.36, "elapsed_time": "14:53:54", "remaining_time": "2 days, 0:52:07", "throughput": 5660.63, "total_tokens": 303604704}
+{"current_steps": 6945, "total_steps": 29704, "loss": 0.40007991790771485, "lr": 8.711096976433076e-06, "epoch": 0.4676137893886345, "percentage": 23.38, "elapsed_time": "14:54:35", "remaining_time": "2 days, 0:51:37", "throughput": 5660.47, "total_tokens": 303830584}
+{"current_steps": 6950, "total_steps": 29704, "loss": 0.3609914779663086, "lr": 8.709324507522337e-06, "epoch": 0.46795044438459465, "percentage": 23.4, "elapsed_time": "14:55:10", "remaining_time": "2 days, 0:50:44", "throughput": 5660.99, "total_tokens": 304051832}
+{"current_steps": 6955, "total_steps": 29704, "loss": 0.3842365264892578, "lr": 8.707551001310828e-06, "epoch": 0.4682870993805548, "percentage": 23.41, "elapsed_time": "14:55:49", "remaining_time": "2 days, 0:50:09", "throughput": 5661.03, "total_tokens": 304279352}
+{"current_steps": 6960, "total_steps": 29704, "loss": 0.387627911567688, "lr": 8.705776458294503e-06, "epoch": 0.46862375437651493, "percentage": 23.43, "elapsed_time": "14:56:25", "remaining_time": "2 days, 0:49:19", "throughput": 5661.18, "total_tokens": 304487200}
+{"current_steps": 6965, "total_steps": 29704, "loss": 0.37026023864746094, "lr": 8.704000878969608e-06, "epoch": 0.4689604093724751, "percentage": 23.45, "elapsed_time": "14:57:02", "remaining_time": "2 days, 0:48:38", "throughput": 5661.04, "total_tokens": 304693840}
+{"current_steps": 6970, "total_steps": 29704, "loss": 0.3628101825714111, "lr": 8.702224263832679e-06, "epoch": 0.4692970643684352, "percentage": 23.46, "elapsed_time": "14:57:41", "remaining_time": "2 days, 0:47:59", "throughput": 5661.21, "total_tokens": 304921392}
+{"current_steps": 6975, "total_steps": 29704, "loss": 0.4275547504425049, "lr": 8.700446613380542e-06, "epoch": 0.46963371936439535, "percentage": 23.48, "elapsed_time": "14:58:19", "remaining_time": "2 days, 0:47:17", "throughput": 5661.49, "total_tokens": 305149136}
+{"current_steps": 6980, "total_steps": 29704, "loss": 0.373789119720459, "lr": 8.698667928110307e-06, "epoch": 0.4699703743603555, "percentage": 23.5, "elapsed_time": "14:58:56", "remaining_time": "2 days, 0:46:36", "throughput": 5661.61, "total_tokens": 305369336}
+{"current_steps": 6985, "total_steps": 29704, "loss": 0.3973226070404053, "lr": 8.696888208519381e-06, "epoch": 0.47030702935631563, "percentage": 23.52, "elapsed_time": "14:59:33", "remaining_time": "2 days, 0:45:50", "throughput": 5661.7, "total_tokens": 305581504}
+{"current_steps": 6990, "total_steps": 29704, "loss": 0.34683785438537595, "lr": 8.695107455105454e-06, "epoch": 0.47064368435227577, "percentage": 23.53, "elapsed_time": "15:00:06", "remaining_time": "2 days, 0:44:52", "throughput": 5662.15, "total_tokens": 305790632}
+{"current_steps": 6995, "total_steps": 29704, "loss": 0.39075660705566406, "lr": 8.693325668366513e-06, "epoch": 0.4709803393482359, "percentage": 23.55, "elapsed_time": "15:00:42", "remaining_time": "2 days, 0:44:06", "throughput": 5662.3, "total_tokens": 306004232}
+{"current_steps": 7000, "total_steps": 29704, "loss": 0.39600386619567873, "lr": 8.691542848800825e-06, "epoch": 0.47131699434419605, "percentage": 23.57, "elapsed_time": "15:01:24", "remaining_time": "2 days, 0:43:39", "throughput": 5661.73, "total_tokens": 306212208}
+{"current_steps": 7005, "total_steps": 29704, "loss": 0.38110928535461425, "lr": 8.68975899690695e-06, "epoch": 0.4716536493401562, "percentage": 23.58, "elapsed_time": "15:02:05", "remaining_time": "2 days, 0:43:06", "throughput": 5661.66, "total_tokens": 306437744}
+{"current_steps": 7010, "total_steps": 29704, "loss": 0.37776660919189453, "lr": 8.687974113183738e-06, "epoch": 0.47199030433611633, "percentage": 23.6, "elapsed_time": "15:02:40", "remaining_time": "2 days, 0:42:18", "throughput": 5662.19, "total_tokens": 306667400}
+{"current_steps": 7015, "total_steps": 29704, "loss": 0.3893289089202881, "lr": 8.686188198130326e-06, "epoch": 0.47232695933207647, "percentage": 23.62, "elapsed_time": "15:03:15", "remaining_time": "2 days, 0:41:26", "throughput": 5662.7, "total_tokens": 306891232}
+{"current_steps": 7020, "total_steps": 29704, "loss": 0.3991250038146973, "lr": 8.684401252246138e-06, "epoch": 0.4726636143280366, "percentage": 23.63, "elapsed_time": "15:03:53", "remaining_time": "2 days, 0:40:47", "throughput": 5662.82, "total_tokens": 307116336}
+{"current_steps": 7025, "total_steps": 29704, "loss": 0.42037010192871094, "lr": 8.68261327603089e-06, "epoch": 0.47300026932399675, "percentage": 23.65, "elapsed_time": "15:04:33", "remaining_time": "2 days, 0:40:12", "throughput": 5662.59, "total_tokens": 307328112}
+{"current_steps": 7030, "total_steps": 29704, "loss": 0.42117109298706057, "lr": 8.68082426998458e-06, "epoch": 0.4733369243199569, "percentage": 23.67, "elapsed_time": "15:05:15", "remaining_time": "2 days, 0:39:43", "throughput": 5662.48, "total_tokens": 307558384}
+{"current_steps": 7035, "total_steps": 29704, "loss": 0.355437707901001, "lr": 8.6790342346075e-06, "epoch": 0.473673579315917, "percentage": 23.68, "elapsed_time": "15:05:55", "remaining_time": "2 days, 0:39:09", "throughput": 5661.99, "total_tokens": 307757832}
+{"current_steps": 7040, "total_steps": 29704, "loss": 0.3744319438934326, "lr": 8.677243170400228e-06, "epoch": 0.47401023431187717, "percentage": 23.7, "elapsed_time": "15:06:32", "remaining_time": "2 days, 0:38:26", "throughput": 5661.9, "total_tokens": 307964848}
+{"current_steps": 7045, "total_steps": 29704, "loss": 0.40117859840393066, "lr": 8.675451077863632e-06, "epoch": 0.4743468893078373, "percentage": 23.72, "elapsed_time": "15:07:09", "remaining_time": "2 days, 0:37:44", "throughput": 5661.97, "total_tokens": 308180176}
+{"current_steps": 7050, "total_steps": 29704, "loss": 0.37619283199310305, "lr": 8.67365795749886e-06, "epoch": 0.47468354430379744, "percentage": 23.73, "elapsed_time": "15:07:55", "remaining_time": "2 days, 0:37:28", "throughput": 5661.12, "total_tokens": 308393256}
+{"current_steps": 7055, "total_steps": 29704, "loss": 0.36246466636657715, "lr": 8.671863809807355e-06, "epoch": 0.4750201992997576, "percentage": 23.75, "elapsed_time": "15:08:32", "remaining_time": "2 days, 0:36:44", "throughput": 5661.03, "total_tokens": 308598120}
+{"current_steps": 7060, "total_steps": 29704, "loss": 0.3815620899200439, "lr": 8.670068635290844e-06, "epoch": 0.4753568542957177, "percentage": 23.77, "elapsed_time": "15:09:13", "remaining_time": "2 days, 0:36:13", "throughput": 5660.75, "total_tokens": 308814312}
+{"current_steps": 7065, "total_steps": 29704, "loss": 0.3547971725463867, "lr": 8.668272434451343e-06, "epoch": 0.47569350929167786, "percentage": 23.78, "elapsed_time": "15:09:49", "remaining_time": "2 days, 0:35:24", "throughput": 5661.13, "total_tokens": 309036648}
+{"current_steps": 7070, "total_steps": 29704, "loss": 0.3938549518585205, "lr": 8.666475207791154e-06, "epoch": 0.476030164287638, "percentage": 23.8, "elapsed_time": "15:10:26", "remaining_time": "2 days, 0:34:40", "throughput": 5660.99, "total_tokens": 309237312}
+{"current_steps": 7075, "total_steps": 29704, "loss": 0.3805175065994263, "lr": 8.664676955812863e-06, "epoch": 0.47636681928359814, "percentage": 23.82, "elapsed_time": "15:11:05", "remaining_time": "2 days, 0:34:04", "throughput": 5660.99, "total_tokens": 309461072}
+{"current_steps": 7080, "total_steps": 29704, "loss": 0.38736543655395506, "lr": 8.662877679019348e-06, "epoch": 0.4767034742795583, "percentage": 23.84, "elapsed_time": "15:11:42", "remaining_time": "2 days, 0:33:21", "throughput": 5661.09, "total_tokens": 309676264}
+{"current_steps": 7085, "total_steps": 29704, "loss": 0.40035252571105956, "lr": 8.66107737791377e-06, "epoch": 0.4770401292755185, "percentage": 23.85, "elapsed_time": "15:12:19", "remaining_time": "2 days, 0:32:38", "throughput": 5661.48, "total_tokens": 309908584}
+{"current_steps": 7090, "total_steps": 29704, "loss": 0.37158200740814207, "lr": 8.659276052999574e-06, "epoch": 0.4773767842714786, "percentage": 23.87, "elapsed_time": "15:13:04", "remaining_time": "2 days, 0:32:18", "throughput": 5660.55, "total_tokens": 310109536}
+{"current_steps": 7095, "total_steps": 29704, "loss": 0.3958035707473755, "lr": 8.657473704780504e-06, "epoch": 0.47771343926743876, "percentage": 23.89, "elapsed_time": "15:13:43", "remaining_time": "2 days, 0:31:39", "throughput": 5660.33, "total_tokens": 310316416}
+{"current_steps": 7100, "total_steps": 29704, "loss": 0.3861806392669678, "lr": 8.65567033376057e-06, "epoch": 0.4780500942633989, "percentage": 23.9, "elapsed_time": "15:14:18", "remaining_time": "2 days, 0:30:49", "throughput": 5660.84, "total_tokens": 310543280}
+{"current_steps": 7105, "total_steps": 29704, "loss": 0.38321571350097655, "lr": 8.65386594044409e-06, "epoch": 0.47838674925935903, "percentage": 23.92, "elapsed_time": "15:14:59", "remaining_time": "2 days, 0:30:19", "throughput": 5660.68, "total_tokens": 310767656}
+{"current_steps": 7110, "total_steps": 29704, "loss": 0.35958447456359866, "lr": 8.652060525335647e-06, "epoch": 0.4787234042553192, "percentage": 23.94, "elapsed_time": "15:15:35", "remaining_time": "2 days, 0:29:31", "throughput": 5660.82, "total_tokens": 310977648}
+{"current_steps": 7115, "total_steps": 29704, "loss": 0.37933621406555174, "lr": 8.650254088940125e-06, "epoch": 0.4790600592512793, "percentage": 23.95, "elapsed_time": "15:16:10", "remaining_time": "2 days, 0:28:43", "throughput": 5660.9, "total_tokens": 311184472}
+{"current_steps": 7120, "total_steps": 29704, "loss": 0.3870699882507324, "lr": 8.648446631762686e-06, "epoch": 0.47939671424723945, "percentage": 23.97, "elapsed_time": "15:16:47", "remaining_time": "2 days, 0:27:58", "throughput": 5660.81, "total_tokens": 311386608}
+{"current_steps": 7125, "total_steps": 29704, "loss": 0.3672234296798706, "lr": 8.646638154308781e-06, "epoch": 0.4797333692431996, "percentage": 23.99, "elapsed_time": "15:17:25", "remaining_time": "2 days, 0:27:19", "throughput": 5660.74, "total_tokens": 311600392}
+{"current_steps": 7130, "total_steps": 29704, "loss": 0.3536520004272461, "lr": 8.644828657084144e-06, "epoch": 0.48007002423915973, "percentage": 24.0, "elapsed_time": "15:18:06", "remaining_time": "2 days, 0:26:46", "throughput": 5660.68, "total_tokens": 311826176}
+{"current_steps": 7135, "total_steps": 29704, "loss": 0.35062260627746583, "lr": 8.643018140594797e-06, "epoch": 0.48040667923511987, "percentage": 24.02, "elapsed_time": "15:18:38", "remaining_time": "2 days, 0:25:48", "throughput": 5661.24, "total_tokens": 312040616}
+{"current_steps": 7140, "total_steps": 29704, "loss": 0.3680727958679199, "lr": 8.641206605347042e-06, "epoch": 0.48074333423108, "percentage": 24.04, "elapsed_time": "15:19:14", "remaining_time": "2 days, 0:25:01", "throughput": 5661.37, "total_tokens": 312250776}
+{"current_steps": 7145, "total_steps": 29704, "loss": 0.3820206642150879, "lr": 8.639394051847472e-06, "epoch": 0.48107998922704015, "percentage": 24.05, "elapsed_time": "15:19:54", "remaining_time": "2 days, 0:24:25", "throughput": 5661.28, "total_tokens": 312469768}
+{"current_steps": 7150, "total_steps": 29704, "loss": 0.3749295473098755, "lr": 8.63758048060296e-06, "epoch": 0.4814166442230003, "percentage": 24.07, "elapsed_time": "15:20:34", "remaining_time": "2 days, 0:23:52", "throughput": 5661.05, "total_tokens": 312686816}
+{"current_steps": 7155, "total_steps": 29704, "loss": 0.35863730907440183, "lr": 8.635765892120666e-06, "epoch": 0.48175329921896043, "percentage": 24.09, "elapsed_time": "15:21:16", "remaining_time": "2 days, 0:23:23", "throughput": 5661.16, "total_tokens": 312928240}
+{"current_steps": 7160, "total_steps": 29704, "loss": 0.3957623243331909, "lr": 8.633950286908032e-06, "epoch": 0.48208995421492057, "percentage": 24.1, "elapsed_time": "15:21:58", "remaining_time": "2 days, 0:22:57", "throughput": 5660.75, "total_tokens": 313146400}
+{"current_steps": 7165, "total_steps": 29704, "loss": 0.41290960311889646, "lr": 8.63213366547279e-06, "epoch": 0.4824266092108807, "percentage": 24.12, "elapsed_time": "15:22:34", "remaining_time": "2 days, 0:22:08", "throughput": 5660.86, "total_tokens": 313353080}
+{"current_steps": 7170, "total_steps": 29704, "loss": 0.3599478960037231, "lr": 8.630316028322952e-06, "epoch": 0.48276326420684085, "percentage": 24.14, "elapsed_time": "15:23:13", "remaining_time": "2 days, 0:21:30", "throughput": 5660.56, "total_tokens": 313556520}
+{"current_steps": 7175, "total_steps": 29704, "loss": 0.3688513278961182, "lr": 8.628497375966811e-06, "epoch": 0.483099919202801, "percentage": 24.15, "elapsed_time": "15:23:53", "remaining_time": "2 days, 0:20:58", "throughput": 5660.63, "total_tokens": 313791304}
+{"current_steps": 7180, "total_steps": 29704, "loss": 0.4179924488067627, "lr": 8.62667770891295e-06, "epoch": 0.4834365741987611, "percentage": 24.17, "elapsed_time": "15:24:39", "remaining_time": "2 days, 0:20:41", "throughput": 5660.23, "total_tokens": 314025432}
+{"current_steps": 7185, "total_steps": 29704, "loss": 0.3718327522277832, "lr": 8.624857027670232e-06, "epoch": 0.48377322919472127, "percentage": 24.19, "elapsed_time": "15:25:18", "remaining_time": "2 days, 0:20:03", "throughput": 5660.37, "total_tokens": 314253744}
+{"current_steps": 7190, "total_steps": 29704, "loss": 0.4232837677001953, "lr": 8.623035332747804e-06, "epoch": 0.4841098841906814, "percentage": 24.21, "elapsed_time": "15:25:55", "remaining_time": "2 days, 0:19:21", "throughput": 5660.21, "total_tokens": 314457360}
+{"current_steps": 7195, "total_steps": 29704, "loss": 0.35295858383178713, "lr": 8.6212126246551e-06, "epoch": 0.48444653918664154, "percentage": 24.22, "elapsed_time": "15:26:30", "remaining_time": "2 days, 0:18:31", "throughput": 5660.26, "total_tokens": 314659296}
+{"current_steps": 7200, "total_steps": 29704, "loss": 0.397071385383606, "lr": 8.619388903901833e-06, "epoch": 0.4847831941826017, "percentage": 24.24, "elapsed_time": "15:27:12", "remaining_time": "2 days, 0:18:03", "throughput": 5660.09, "total_tokens": 314886400}
+{"current_steps": 7205, "total_steps": 29704, "loss": 0.393186354637146, "lr": 8.617564170997998e-06, "epoch": 0.4851198491785618, "percentage": 24.26, "elapsed_time": "15:27:45", "remaining_time": "2 days, 0:17:06", "throughput": 5660.82, "total_tokens": 315113560}
+{"current_steps": 7210, "total_steps": 29704, "loss": 0.38975393772125244, "lr": 8.61573842645388e-06, "epoch": 0.48545650417452196, "percentage": 24.27, "elapsed_time": "15:28:21", "remaining_time": "2 days, 0:16:19", "throughput": 5661.23, "total_tokens": 315339760}
+{"current_steps": 7215, "total_steps": 29704, "loss": 0.3773904085159302, "lr": 8.61391167078004e-06, "epoch": 0.4857931591704821, "percentage": 24.29, "elapsed_time": "15:28:59", "remaining_time": "2 days, 0:15:40", "throughput": 5661.37, "total_tokens": 315564560}
+{"current_steps": 7220, "total_steps": 29704, "loss": 0.4223905563354492, "lr": 8.612083904487324e-06, "epoch": 0.48612981416644224, "percentage": 24.31, "elapsed_time": "15:29:42", "remaining_time": "2 days, 0:15:12", "throughput": 5660.96, "total_tokens": 315780664}
+{"current_steps": 7225, "total_steps": 29704, "loss": 0.39981255531311033, "lr": 8.610255128086867e-06, "epoch": 0.4864664691624024, "percentage": 24.32, "elapsed_time": "15:30:17", "remaining_time": "2 days, 0:14:24", "throughput": 5661.03, "total_tokens": 315985224}
+{"current_steps": 7230, "total_steps": 29704, "loss": 0.35159389972686766, "lr": 8.60842534209007e-06, "epoch": 0.4868031241583625, "percentage": 24.34, "elapsed_time": "15:30:56", "remaining_time": "2 days, 0:13:45", "throughput": 5661.08, "total_tokens": 316206120}
+{"current_steps": 7235, "total_steps": 29704, "loss": 0.33089404106140136, "lr": 8.606594547008636e-06, "epoch": 0.48713977915432266, "percentage": 24.36, "elapsed_time": "15:31:29", "remaining_time": "2 days, 0:12:50", "throughput": 5661.67, "total_tokens": 316428112}
+{"current_steps": 7240, "total_steps": 29704, "loss": 0.39783239364624023, "lr": 8.604762743354536e-06, "epoch": 0.4874764341502828, "percentage": 24.37, "elapsed_time": "15:32:02", "remaining_time": "2 days, 0:11:55", "throughput": 5662.56, "total_tokens": 316666376}
+{"current_steps": 7245, "total_steps": 29704, "loss": 0.3995728254318237, "lr": 8.602929931640029e-06, "epoch": 0.48781308914624294, "percentage": 24.39, "elapsed_time": "15:32:39", "remaining_time": "2 days, 0:11:10", "throughput": 5662.46, "total_tokens": 316868832}
+{"current_steps": 7250, "total_steps": 29704, "loss": 0.4125559329986572, "lr": 8.601096112377659e-06, "epoch": 0.4881497441422031, "percentage": 24.41, "elapsed_time": "15:33:22", "remaining_time": "2 days, 0:10:46", "throughput": 5661.91, "total_tokens": 317082248}
+{"current_steps": 7255, "total_steps": 29704, "loss": 0.3738250255584717, "lr": 8.59926128608024e-06, "epoch": 0.4884863991381632, "percentage": 24.42, "elapsed_time": "15:33:59", "remaining_time": "2 days, 0:10:02", "throughput": 5662.11, "total_tokens": 317303120}
+{"current_steps": 7260, "total_steps": 29704, "loss": 0.3914709806442261, "lr": 8.597425453260881e-06, "epoch": 0.48882305413412336, "percentage": 24.44, "elapsed_time": "15:34:42", "remaining_time": "2 days, 0:09:36", "throughput": 5661.6, "total_tokens": 317515672}
+{"current_steps": 7265, "total_steps": 29704, "loss": 0.3669420719146729, "lr": 8.595588614432967e-06, "epoch": 0.4891597091300835, "percentage": 24.46, "elapsed_time": "15:35:28", "remaining_time": "2 days, 0:09:19", "throughput": 5660.63, "total_tokens": 317720352}
+{"current_steps": 7270, "total_steps": 29704, "loss": 0.36806163787841795, "lr": 8.59375077011016e-06, "epoch": 0.48949636412604364, "percentage": 24.47, "elapsed_time": "15:36:07", "remaining_time": "2 days, 0:08:43", "throughput": 5660.7, "total_tokens": 317948384}
+{"current_steps": 7275, "total_steps": 29704, "loss": 0.39014444351196287, "lr": 8.591911920806412e-06, "epoch": 0.4898330191220038, "percentage": 24.49, "elapsed_time": "15:36:53", "remaining_time": "2 days, 0:08:28", "throughput": 5660.14, "total_tokens": 318179072}
+{"current_steps": 7280, "total_steps": 29704, "loss": 0.38969347476959226, "lr": 8.590072067035947e-06, "epoch": 0.4901696741179639, "percentage": 24.51, "elapsed_time": "15:37:35", "remaining_time": "2 days, 0:07:59", "throughput": 5660.08, "total_tokens": 318410480}
+{"current_steps": 7285, "total_steps": 29704, "loss": 0.37391541004180906, "lr": 8.588231209313278e-06, "epoch": 0.49050632911392406, "percentage": 24.53, "elapsed_time": "15:38:11", "remaining_time": "2 days, 0:07:13", "throughput": 5660.78, "total_tokens": 318654632}
+{"current_steps": 7290, "total_steps": 29704, "loss": 0.3896606206893921, "lr": 8.586389348153192e-06, "epoch": 0.4908429841098842, "percentage": 24.54, "elapsed_time": "15:38:46", "remaining_time": "2 days, 0:06:22", "throughput": 5661.19, "total_tokens": 318874824}
+{"current_steps": 7295, "total_steps": 29704, "loss": 0.40265254974365233, "lr": 8.584546484070762e-06, "epoch": 0.49117963910584433, "percentage": 24.56, "elapsed_time": "15:39:25", "remaining_time": "2 days, 0:05:44", "throughput": 5661.0, "total_tokens": 319084200}
+{"current_steps": 7300, "total_steps": 29704, "loss": 0.37858309745788576, "lr": 8.582702617581338e-06, "epoch": 0.4915162941018045, "percentage": 24.58, "elapsed_time": "15:40:03", "remaining_time": "2 days, 0:05:04", "throughput": 5661.13, "total_tokens": 319307784}
+{"current_steps": 7305, "total_steps": 29704, "loss": 0.38918490409851075, "lr": 8.58085774920055e-06, "epoch": 0.4918529490977646, "percentage": 24.59, "elapsed_time": "15:40:38", "remaining_time": "2 days, 0:04:15", "throughput": 5661.26, "total_tokens": 319513304}
+{"current_steps": 7310, "total_steps": 29704, "loss": 0.414321231842041, "lr": 8.579011879444313e-06, "epoch": 0.49218960409372475, "percentage": 24.61, "elapsed_time": "15:41:23", "remaining_time": "2 days, 0:03:55", "throughput": 5660.84, "total_tokens": 319743560}
+{"current_steps": 7315, "total_steps": 29704, "loss": 0.34825422763824465, "lr": 8.577165008828817e-06, "epoch": 0.4925262590896849, "percentage": 24.63, "elapsed_time": "15:42:01", "remaining_time": "2 days, 0:03:15", "throughput": 5660.62, "total_tokens": 319947112}
+{"current_steps": 7320, "total_steps": 29704, "loss": 0.3858429193496704, "lr": 8.57531713787053e-06, "epoch": 0.49286291408564503, "percentage": 24.64, "elapsed_time": "15:42:38", "remaining_time": "2 days, 0:02:30", "throughput": 5660.65, "total_tokens": 320155920}
+{"current_steps": 7325, "total_steps": 29704, "loss": 0.40245862007141114, "lr": 8.573468267086208e-06, "epoch": 0.49319956908160517, "percentage": 24.66, "elapsed_time": "15:43:12", "remaining_time": "2 days, 0:01:40", "throughput": 5661.0, "total_tokens": 320372792}
+{"current_steps": 7330, "total_steps": 29704, "loss": 0.36266283988952636, "lr": 8.57161839699288e-06, "epoch": 0.4935362240775653, "percentage": 24.68, "elapsed_time": "15:43:51", "remaining_time": "2 days, 0:01:01", "throughput": 5661.01, "total_tokens": 320592160}
+{"current_steps": 7335, "total_steps": 29704, "loss": 0.3856183052062988, "lr": 8.569767528107857e-06, "epoch": 0.49387287907352545, "percentage": 24.69, "elapsed_time": "15:44:29", "remaining_time": "2 days, 0:00:19", "throughput": 5661.06, "total_tokens": 320807512}
+{"current_steps": 7340, "total_steps": 29704, "loss": 0.3747032880783081, "lr": 8.567915660948727e-06, "epoch": 0.4942095340694856, "percentage": 24.71, "elapsed_time": "15:45:09", "remaining_time": "1 day, 23:59:46", "throughput": 5660.98, "total_tokens": 321031440}
+{"current_steps": 7345, "total_steps": 29704, "loss": 0.4063011646270752, "lr": 8.566062796033358e-06, "epoch": 0.49454618906544573, "percentage": 24.73, "elapsed_time": "15:45:46", "remaining_time": "1 day, 23:59:03", "throughput": 5661.0, "total_tokens": 321244008}
+{"current_steps": 7350, "total_steps": 29704, "loss": 0.37364883422851564, "lr": 8.564208933879903e-06, "epoch": 0.49488284406140587, "percentage": 24.74, "elapsed_time": "15:46:27", "remaining_time": "1 day, 23:58:30", "throughput": 5660.73, "total_tokens": 321456992}
+{"current_steps": 7355, "total_steps": 29704, "loss": 0.3439052104949951, "lr": 8.562354075006782e-06, "epoch": 0.495219499057366, "percentage": 24.76, "elapsed_time": "15:47:05", "remaining_time": "1 day, 23:57:49", "throughput": 5660.79, "total_tokens": 321675512}
+{"current_steps": 7360, "total_steps": 29704, "loss": 0.4257340431213379, "lr": 8.560498219932706e-06, "epoch": 0.49555615405332615, "percentage": 24.78, "elapsed_time": "15:47:39", "remaining_time": "1 day, 23:56:57", "throughput": 5661.32, "total_tokens": 321898448}
+{"current_steps": 7365, "total_steps": 29704, "loss": 0.33791351318359375, "lr": 8.558641369176654e-06, "epoch": 0.4958928090492863, "percentage": 24.79, "elapsed_time": "15:48:12", "remaining_time": "1 day, 23:56:01", "throughput": 5661.58, "total_tokens": 322099904}
+{"current_steps": 7370, "total_steps": 29704, "loss": 0.3972716569900513, "lr": 8.556783523257892e-06, "epoch": 0.4962294640452464, "percentage": 24.81, "elapsed_time": "15:48:55", "remaining_time": "1 day, 23:55:35", "throughput": 5661.01, "total_tokens": 322310456}
+{"current_steps": 7375, "total_steps": 29704, "loss": 0.3530308246612549, "lr": 8.554924682695959e-06, "epoch": 0.49656611904120657, "percentage": 24.83, "elapsed_time": "15:49:31", "remaining_time": "1 day, 23:54:51", "throughput": 5661.01, "total_tokens": 322518744}
+{"current_steps": 7380, "total_steps": 29704, "loss": 0.3955545902252197, "lr": 8.553064848010677e-06, "epoch": 0.4969027740371667, "percentage": 24.85, "elapsed_time": "15:50:07", "remaining_time": "1 day, 23:54:04", "throughput": 5661.09, "total_tokens": 322725520}
+{"current_steps": 7385, "total_steps": 29704, "loss": 0.37126519680023196, "lr": 8.551204019722137e-06, "epoch": 0.49723942903312685, "percentage": 24.86, "elapsed_time": "15:50:46", "remaining_time": "1 day, 23:53:25", "throughput": 5660.99, "total_tokens": 322938184}
+{"current_steps": 7390, "total_steps": 29704, "loss": 0.40738329887390134, "lr": 8.549342198350718e-06, "epoch": 0.497576084029087, "percentage": 24.88, "elapsed_time": "15:51:26", "remaining_time": "1 day, 23:52:50", "throughput": 5660.94, "total_tokens": 323160704}
+{"current_steps": 7395, "total_steps": 29704, "loss": 0.36683225631713867, "lr": 8.54747938441707e-06, "epoch": 0.4979127390250471, "percentage": 24.9, "elapsed_time": "15:52:08", "remaining_time": "1 day, 23:52:23", "throughput": 5660.61, "total_tokens": 323382024}
+{"current_steps": 7400, "total_steps": 29704, "loss": 0.35772018432617186, "lr": 8.545615578442126e-06, "epoch": 0.49824939402100726, "percentage": 24.91, "elapsed_time": "15:52:46", "remaining_time": "1 day, 23:51:42", "throughput": 5660.68, "total_tokens": 323601056}
+{"current_steps": 7405, "total_steps": 29704, "loss": 0.3746786117553711, "lr": 8.54375078094709e-06, "epoch": 0.4985860490169674, "percentage": 24.93, "elapsed_time": "15:53:36", "remaining_time": "1 day, 23:51:37", "throughput": 5659.67, "total_tokens": 323824464}
+{"current_steps": 7410, "total_steps": 29704, "loss": 0.3711742401123047, "lr": 8.541884992453449e-06, "epoch": 0.49892270401292754, "percentage": 24.95, "elapsed_time": "15:54:14", "remaining_time": "1 day, 23:50:57", "throughput": 5659.8, "total_tokens": 324047472}
+{"current_steps": 7415, "total_steps": 29704, "loss": 0.37660138607025145, "lr": 8.540018213482966e-06, "epoch": 0.4992593590088877, "percentage": 24.96, "elapsed_time": "15:54:54", "remaining_time": "1 day, 23:50:23", "throughput": 5659.9, "total_tokens": 324280688}
+{"current_steps": 7420, "total_steps": 29704, "loss": 0.3920082330703735, "lr": 8.538150444557676e-06, "epoch": 0.4995960140048478, "percentage": 24.98, "elapsed_time": "15:55:33", "remaining_time": "1 day, 23:49:47", "throughput": 5659.62, "total_tokens": 324488104}
+{"current_steps": 7425, "total_steps": 29704, "loss": 0.385677433013916, "lr": 8.536281686199896e-06, "epoch": 0.49993266900080796, "percentage": 25.0, "elapsed_time": "15:56:06", "remaining_time": "1 day, 23:48:50", "throughput": 5660.28, "total_tokens": 324711216}
+{"current_steps": 7430, "total_steps": 29704, "loss": 0.3787369728088379, "lr": 8.534411938932218e-06, "epoch": 0.5002693239967682, "percentage": 25.01, "elapsed_time": "15:56:40", "remaining_time": "1 day, 23:47:58", "throughput": 5660.88, "total_tokens": 324939096}
+{"current_steps": 7435, "total_steps": 29704, "loss": 0.3697648525238037, "lr": 8.532541203277515e-06, "epoch": 0.5006059789927283, "percentage": 25.03, "elapsed_time": "15:57:17", "remaining_time": "1 day, 23:47:13", "throughput": 5661.2, "total_tokens": 325163640}
+{"current_steps": 7440, "total_steps": 29704, "loss": 0.37598912715911864, "lr": 8.530669479758926e-06, "epoch": 0.5009426339886884, "percentage": 25.05, "elapsed_time": "15:57:55", "remaining_time": "1 day, 23:46:35", "throughput": 5661.21, "total_tokens": 325383192}
+{"current_steps": 7445, "total_steps": 29704, "loss": 0.39255619049072266, "lr": 8.528796768899878e-06, "epoch": 0.5012792889846486, "percentage": 25.06, "elapsed_time": "15:58:36", "remaining_time": "1 day, 23:46:02", "throughput": 5660.73, "total_tokens": 325585216}
+{"current_steps": 7450, "total_steps": 29704, "loss": 0.39842636585235597, "lr": 8.526923071224064e-06, "epoch": 0.5016159439806087, "percentage": 25.08, "elapsed_time": "15:59:09", "remaining_time": "1 day, 23:45:06", "throughput": 5661.4, "total_tokens": 325810472}
+{"current_steps": 7455, "total_steps": 29704, "loss": 0.3887434482574463, "lr": 8.525048387255461e-06, "epoch": 0.5019525989765689, "percentage": 25.1, "elapsed_time": "15:59:50", "remaining_time": "1 day, 23:44:34", "throughput": 5661.3, "total_tokens": 326034504}
+{"current_steps": 7460, "total_steps": 29704, "loss": 0.3696585178375244, "lr": 8.523172717518315e-06, "epoch": 0.502289253972529, "percentage": 25.11, "elapsed_time": "16:00:27", "remaining_time": "1 day, 23:43:51", "throughput": 5661.31, "total_tokens": 326245696}
+{"current_steps": 7465, "total_steps": 29704, "loss": 0.37876601219177247, "lr": 8.521296062537156e-06, "epoch": 0.5026259089684891, "percentage": 25.13, "elapsed_time": "16:01:04", "remaining_time": "1 day, 23:43:09", "throughput": 5661.33, "total_tokens": 326460104}
+{"current_steps": 7470, "total_steps": 29704, "loss": 0.36045174598693847, "lr": 8.51941842283678e-06, "epoch": 0.5029625639644493, "percentage": 25.15, "elapsed_time": "16:01:44", "remaining_time": "1 day, 23:42:34", "throughput": 5661.11, "total_tokens": 326673104}
+{"current_steps": 7475, "total_steps": 29704, "loss": 0.3735339641571045, "lr": 8.517539798942265e-06, "epoch": 0.5032992189604094, "percentage": 25.16, "elapsed_time": "16:02:17", "remaining_time": "1 day, 23:41:39", "throughput": 5661.93, "total_tokens": 326906960}
+{"current_steps": 7480, "total_steps": 29704, "loss": 0.39676947593688966, "lr": 8.515660191378962e-06, "epoch": 0.5036358739563696, "percentage": 25.18, "elapsed_time": "16:02:56", "remaining_time": "1 day, 23:41:02", "throughput": 5661.93, "total_tokens": 327128416}
+{"current_steps": 7485, "total_steps": 29704, "loss": 0.3535287380218506, "lr": 8.513779600672495e-06, "epoch": 0.5039725289523297, "percentage": 25.2, "elapsed_time": "16:03:31", "remaining_time": "1 day, 23:40:11", "throughput": 5662.63, "total_tokens": 327364152}
+{"current_steps": 7490, "total_steps": 29704, "loss": 0.37773268222808837, "lr": 8.511898027348771e-06, "epoch": 0.5043091839482898, "percentage": 25.22, "elapsed_time": "16:04:13", "remaining_time": "1 day, 23:39:42", "throughput": 5662.06, "total_tokens": 327568048}
+{"current_steps": 7495, "total_steps": 29704, "loss": 0.3704256296157837, "lr": 8.51001547193396e-06, "epoch": 0.50464583894425, "percentage": 25.23, "elapsed_time": "16:04:53", "remaining_time": "1 day, 23:39:08", "throughput": 5662.0, "total_tokens": 327792152}
+{"current_steps": 7500, "total_steps": 29704, "loss": 0.36852757930755614, "lr": 8.508131934954515e-06, "epoch": 0.5049824939402101, "percentage": 25.25, "elapsed_time": "16:05:32", "remaining_time": "1 day, 23:38:32", "throughput": 5661.63, "total_tokens": 327994400}
+{"current_steps": 7505, "total_steps": 29704, "loss": 0.3951726913452148, "lr": 8.50624741693716e-06, "epoch": 0.5053191489361702, "percentage": 25.27, "elapsed_time": "16:06:11", "remaining_time": "1 day, 23:37:54", "throughput": 5661.39, "total_tokens": 328200448}
+{"current_steps": 7510, "total_steps": 29704, "loss": 0.4119250297546387, "lr": 8.5043619184089e-06, "epoch": 0.5056558039321304, "percentage": 25.28, "elapsed_time": "16:06:45", "remaining_time": "1 day, 23:37:02", "throughput": 5661.86, "total_tokens": 328421224}
+{"current_steps": 7515, "total_steps": 29704, "loss": 0.4078369140625, "lr": 8.502475439897e-06, "epoch": 0.5059924589280905, "percentage": 25.3, "elapsed_time": "16:07:23", "remaining_time": "1 day, 23:36:22", "throughput": 5662.03, "total_tokens": 328646192}
+{"current_steps": 7520, "total_steps": 29704, "loss": 0.3898489475250244, "lr": 8.500587981929013e-06, "epoch": 0.5063291139240507, "percentage": 25.32, "elapsed_time": "16:08:08", "remaining_time": "1 day, 23:35:59", "throughput": 5661.53, "total_tokens": 328867064}
+{"current_steps": 7525, "total_steps": 29704, "loss": 0.38454551696777345, "lr": 8.498699545032758e-06, "epoch": 0.5066657689200108, "percentage": 25.33, "elapsed_time": "16:08:44", "remaining_time": "1 day, 23:35:13", "throughput": 5662.08, "total_tokens": 329103168}
+{"current_steps": 7530, "total_steps": 29704, "loss": 0.3850215435028076, "lr": 8.496810129736334e-06, "epoch": 0.507002423915971, "percentage": 25.35, "elapsed_time": "16:09:26", "remaining_time": "1 day, 23:34:47", "throughput": 5661.68, "total_tokens": 329322920}
+{"current_steps": 7535, "total_steps": 29704, "loss": 0.3788848638534546, "lr": 8.494919736568105e-06, "epoch": 0.5073390789119311, "percentage": 25.37, "elapsed_time": "16:10:10", "remaining_time": "1 day, 23:34:21", "throughput": 5661.12, "total_tokens": 329533896}
+{"current_steps": 7540, "total_steps": 29704, "loss": 0.37194099426269533, "lr": 8.49302836605672e-06, "epoch": 0.5076757339078912, "percentage": 25.38, "elapsed_time": "16:10:46", "remaining_time": "1 day, 23:33:36", "throughput": 5661.48, "total_tokens": 329760624}
+{"current_steps": 7545, "total_steps": 29704, "loss": 0.3646608114242554, "lr": 8.491136018731088e-06, "epoch": 0.5080123889038514, "percentage": 25.4, "elapsed_time": "16:11:27", "remaining_time": "1 day, 23:33:05", "throughput": 5661.31, "total_tokens": 329984728}
+{"current_steps": 7550, "total_steps": 29704, "loss": 0.3987517118453979, "lr": 8.4892426951204e-06, "epoch": 0.5083490438998115, "percentage": 25.42, "elapsed_time": "16:11:56", "remaining_time": "1 day, 23:31:57", "throughput": 5662.34, "total_tokens": 330206096}
+{"current_steps": 7555, "total_steps": 29704, "loss": 0.3717711687088013, "lr": 8.48734839575412e-06, "epoch": 0.5086856988957716, "percentage": 25.43, "elapsed_time": "16:12:31", "remaining_time": "1 day, 23:31:08", "throughput": 5662.89, "total_tokens": 330437336}
+{"current_steps": 7560, "total_steps": 29704, "loss": 0.37380030155181887, "lr": 8.485453121161983e-06, "epoch": 0.5090223538917318, "percentage": 25.45, "elapsed_time": "16:13:12", "remaining_time": "1 day, 23:30:36", "throughput": 5662.64, "total_tokens": 330653616}
+{"current_steps": 7565, "total_steps": 29704, "loss": 0.42113819122314455, "lr": 8.483556871873993e-06, "epoch": 0.5093590088876919, "percentage": 25.47, "elapsed_time": "16:13:53", "remaining_time": "1 day, 23:30:05", "throughput": 5662.46, "total_tokens": 330875672}
+{"current_steps": 7570, "total_steps": 29704, "loss": 0.3786670207977295, "lr": 8.481659648420433e-06, "epoch": 0.5096956638836521, "percentage": 25.48, "elapsed_time": "16:14:35", "remaining_time": "1 day, 23:29:35", "throughput": 5662.18, "total_tokens": 331096488}
+{"current_steps": 7575, "total_steps": 29704, "loss": 0.3809337615966797, "lr": 8.479761451331855e-06, "epoch": 0.5100323188796122, "percentage": 25.5, "elapsed_time": "16:15:16", "remaining_time": "1 day, 23:29:04", "throughput": 5661.44, "total_tokens": 331285184}
+{"current_steps": 7580, "total_steps": 29704, "loss": 0.3540458917617798, "lr": 8.477862281139082e-06, "epoch": 0.5103689738755723, "percentage": 25.52, "elapsed_time": "16:15:58", "remaining_time": "1 day, 23:28:36", "throughput": 5661.04, "total_tokens": 331501760}
+{"current_steps": 7585, "total_steps": 29704, "loss": 0.3986123323440552, "lr": 8.475962138373212e-06, "epoch": 0.5107056288715325, "percentage": 25.54, "elapsed_time": "16:16:31", "remaining_time": "1 day, 23:27:42", "throughput": 5661.24, "total_tokens": 331702152}
+{"current_steps": 7590, "total_steps": 29704, "loss": 0.3940688371658325, "lr": 8.474061023565614e-06, "epoch": 0.5110422838674926, "percentage": 25.55, "elapsed_time": "16:17:11", "remaining_time": "1 day, 23:27:06", "throughput": 5661.29, "total_tokens": 331929992}
+{"current_steps": 7595, "total_steps": 29704, "loss": 0.40605816841125486, "lr": 8.472158937247931e-06, "epoch": 0.5113789388634528, "percentage": 25.57, "elapsed_time": "16:17:46", "remaining_time": "1 day, 23:26:17", "throughput": 5661.65, "total_tokens": 332148472}
+{"current_steps": 7600, "total_steps": 29704, "loss": 0.38570313453674315, "lr": 8.47025587995207e-06, "epoch": 0.5117155938594129, "percentage": 25.59, "elapsed_time": "16:18:24", "remaining_time": "1 day, 23:25:37", "throughput": 5661.63, "total_tokens": 332362728}
+{"current_steps": 7605, "total_steps": 29704, "loss": 0.39829018115997317, "lr": 8.46835185221022e-06, "epoch": 0.512052248855373, "percentage": 25.6, "elapsed_time": "16:19:03", "remaining_time": "1 day, 23:25:01", "throughput": 5661.34, "total_tokens": 332569544}
+{"current_steps": 7610, "total_steps": 29704, "loss": 0.3549684762954712, "lr": 8.466446854554835e-06, "epoch": 0.5123889038513332, "percentage": 25.62, "elapsed_time": "16:19:40", "remaining_time": "1 day, 23:24:17", "throughput": 5661.35, "total_tokens": 332778920}
+{"current_steps": 7615, "total_steps": 29704, "loss": 0.37906813621520996, "lr": 8.464540887518638e-06, "epoch": 0.5127255588472933, "percentage": 25.64, "elapsed_time": "16:20:23", "remaining_time": "1 day, 23:23:49", "throughput": 5660.76, "total_tokens": 332984080}
+{"current_steps": 7620, "total_steps": 29704, "loss": 0.3685762405395508, "lr": 8.462633951634631e-06, "epoch": 0.5130622138432535, "percentage": 25.65, "elapsed_time": "16:21:08", "remaining_time": "1 day, 23:23:29", "throughput": 5660.14, "total_tokens": 333202640}
+{"current_steps": 7625, "total_steps": 29704, "loss": 0.35547614097595215, "lr": 8.46072604743608e-06, "epoch": 0.5133988688392136, "percentage": 25.67, "elapsed_time": "16:21:45", "remaining_time": "1 day, 23:22:45", "throughput": 5660.2, "total_tokens": 333413904}
+{"current_steps": 7630, "total_steps": 29704, "loss": 0.3930670261383057, "lr": 8.458817175456528e-06, "epoch": 0.5137355238351737, "percentage": 25.69, "elapsed_time": "16:22:27", "remaining_time": "1 day, 23:22:17", "throughput": 5659.82, "total_tokens": 333630616}
+{"current_steps": 7635, "total_steps": 29704, "loss": 0.35927329063415525, "lr": 8.45690733622978e-06, "epoch": 0.5140721788311339, "percentage": 25.7, "elapsed_time": "16:23:06", "remaining_time": "1 day, 23:21:41", "throughput": 5659.83, "total_tokens": 333854600}
+{"current_steps": 7640, "total_steps": 29704, "loss": 0.36787919998168944, "lr": 8.45499653028992e-06, "epoch": 0.514408833827094, "percentage": 25.72, "elapsed_time": "16:23:42", "remaining_time": "1 day, 23:20:53", "throughput": 5660.09, "total_tokens": 334071192}
+{"current_steps": 7645, "total_steps": 29704, "loss": 0.37957746982574464, "lr": 8.453084758171297e-06, "epoch": 0.5147454888230542, "percentage": 25.74, "elapsed_time": "16:24:27", "remaining_time": "1 day, 23:20:33", "throughput": 5659.52, "total_tokens": 334291888}
+{"current_steps": 7650, "total_steps": 29704, "loss": 0.38945965766906737, "lr": 8.451172020408532e-06, "epoch": 0.5150821438190143, "percentage": 25.75, "elapsed_time": "16:25:01", "remaining_time": "1 day, 23:19:42", "throughput": 5660.04, "total_tokens": 334516424}
+{"current_steps": 7655, "total_steps": 29704, "loss": 0.35926315784454343, "lr": 8.449258317536517e-06, "epoch": 0.5154187988149744, "percentage": 25.77, "elapsed_time": "16:25:34", "remaining_time": "1 day, 23:18:48", "throughput": 5660.51, "total_tokens": 334732808}
+{"current_steps": 7660, "total_steps": 29704, "loss": 0.35725347995758056, "lr": 8.447343650090412e-06, "epoch": 0.5157554538109346, "percentage": 25.79, "elapsed_time": "16:26:12", "remaining_time": "1 day, 23:18:08", "throughput": 5660.6, "total_tokens": 334954280}
+{"current_steps": 7665, "total_steps": 29704, "loss": 0.406490421295166, "lr": 8.445428018605647e-06, "epoch": 0.5160921088068947, "percentage": 25.8, "elapsed_time": "16:26:50", "remaining_time": "1 day, 23:17:25", "throughput": 5660.82, "total_tokens": 335178568}
+{"current_steps": 7670, "total_steps": 29704, "loss": 0.34410724639892576, "lr": 8.443511423617924e-06, "epoch": 0.5164287638028549, "percentage": 25.82, "elapsed_time": "16:27:26", "remaining_time": "1 day, 23:16:40", "throughput": 5661.06, "total_tokens": 335398608}
+{"current_steps": 7675, "total_steps": 29704, "loss": 0.39874303340911865, "lr": 8.441593865663213e-06, "epoch": 0.516765418798815, "percentage": 25.84, "elapsed_time": "16:28:05", "remaining_time": "1 day, 23:16:01", "throughput": 5661.15, "total_tokens": 335621696}
+{"current_steps": 7680, "total_steps": 29704, "loss": 0.40140485763549805, "lr": 8.439675345277751e-06, "epoch": 0.5171020737947751, "percentage": 25.86, "elapsed_time": "16:28:47", "remaining_time": "1 day, 23:15:33", "throughput": 5660.79, "total_tokens": 335839504}
+{"current_steps": 7685, "total_steps": 29704, "loss": 0.3619039297103882, "lr": 8.437755862998047e-06, "epoch": 0.5174387287907353, "percentage": 25.87, "elapsed_time": "16:29:22", "remaining_time": "1 day, 23:14:44", "throughput": 5661.21, "total_tokens": 336062080}
+{"current_steps": 7690, "total_steps": 29704, "loss": 0.3882591247558594, "lr": 8.435835419360878e-06, "epoch": 0.5177753837866954, "percentage": 25.89, "elapsed_time": "16:29:58", "remaining_time": "1 day, 23:13:59", "throughput": 5661.74, "total_tokens": 336299944}
+{"current_steps": 7695, "total_steps": 29704, "loss": 0.42388019561767576, "lr": 8.433914014903291e-06, "epoch": 0.5181120387826555, "percentage": 25.91, "elapsed_time": "16:30:37", "remaining_time": "1 day, 23:13:22", "throughput": 5661.63, "total_tokens": 336514968}
+{"current_steps": 7700, "total_steps": 29704, "loss": 0.37606000900268555, "lr": 8.431991650162599e-06, "epoch": 0.5184486937786157, "percentage": 25.92, "elapsed_time": "16:31:23", "remaining_time": "1 day, 23:13:03", "throughput": 5660.95, "total_tokens": 336732600}
+{"current_steps": 7705, "total_steps": 29704, "loss": 0.39625885486602785, "lr": 8.430068325676386e-06, "epoch": 0.5187853487745758, "percentage": 25.94, "elapsed_time": "16:32:01", "remaining_time": "1 day, 23:12:22", "throughput": 5661.0, "total_tokens": 336948776}
+{"current_steps": 7710, "total_steps": 29704, "loss": 0.39479496479034426, "lr": 8.428144041982502e-06, "epoch": 0.519122003770536, "percentage": 25.96, "elapsed_time": "16:32:36", "remaining_time": "1 day, 23:11:34", "throughput": 5661.25, "total_tokens": 337163520}
+{"current_steps": 7715, "total_steps": 29704, "loss": 0.38160171508789065, "lr": 8.42621879961907e-06, "epoch": 0.5194586587664961, "percentage": 25.97, "elapsed_time": "16:33:17", "remaining_time": "1 day, 23:11:01", "throughput": 5661.05, "total_tokens": 337382992}
+{"current_steps": 7720, "total_steps": 29704, "loss": 0.3532627820968628, "lr": 8.424292599124476e-06, "epoch": 0.5197953137624562, "percentage": 25.99, "elapsed_time": "16:33:56", "remaining_time": "1 day, 23:10:24", "throughput": 5661.12, "total_tokens": 337609304}
+{"current_steps": 7725, "total_steps": 29704, "loss": 0.39478967189788816, "lr": 8.422365441037374e-06, "epoch": 0.5201319687584164, "percentage": 26.01, "elapsed_time": "16:34:29", "remaining_time": "1 day, 23:09:29", "throughput": 5661.86, "total_tokens": 337838072}
+{"current_steps": 7730, "total_steps": 29704, "loss": 0.3906223297119141, "lr": 8.420437325896692e-06, "epoch": 0.5204686237543765, "percentage": 26.02, "elapsed_time": "16:35:05", "remaining_time": "1 day, 23:08:43", "throughput": 5662.28, "total_tokens": 338067512}
+{"current_steps": 7735, "total_steps": 29704, "loss": 0.3925325870513916, "lr": 8.418508254241617e-06, "epoch": 0.5208052787503367, "percentage": 26.04, "elapsed_time": "16:35:48", "remaining_time": "1 day, 23:08:16", "throughput": 5661.85, "total_tokens": 338284504}
+{"current_steps": 7740, "total_steps": 29704, "loss": 0.35642185211181643, "lr": 8.416578226611612e-06, "epoch": 0.5211419337462968, "percentage": 26.06, "elapsed_time": "16:36:23", "remaining_time": "1 day, 23:07:28", "throughput": 5662.04, "total_tokens": 338495440}
+{"current_steps": 7745, "total_steps": 29704, "loss": 0.38446238040924074, "lr": 8.414647243546398e-06, "epoch": 0.5214785887422569, "percentage": 26.07, "elapsed_time": "16:37:01", "remaining_time": "1 day, 23:06:48", "throughput": 5662.14, "total_tokens": 338717192}
+{"current_steps": 7750, "total_steps": 29704, "loss": 0.367059588432312, "lr": 8.412715305585973e-06, "epoch": 0.5218152437382171, "percentage": 26.09, "elapsed_time": "16:37:39", "remaining_time": "1 day, 23:06:09", "throughput": 5662.33, "total_tokens": 338946888}
+{"current_steps": 7755, "total_steps": 29704, "loss": 0.3799006462097168, "lr": 8.410782413270594e-06, "epoch": 0.5221518987341772, "percentage": 26.11, "elapsed_time": "16:38:20", "remaining_time": "1 day, 23:05:35", "throughput": 5662.29, "total_tokens": 339172224}
+{"current_steps": 7760, "total_steps": 29704, "loss": 0.37287948131561277, "lr": 8.408848567140787e-06, "epoch": 0.5224885537301374, "percentage": 26.12, "elapsed_time": "16:39:01", "remaining_time": "1 day, 23:05:03", "throughput": 5661.64, "total_tokens": 339364672}
+{"current_steps": 7765, "total_steps": 29704, "loss": 0.37642457485198977, "lr": 8.406913767737353e-06, "epoch": 0.5228252087260975, "percentage": 26.14, "elapsed_time": "16:39:42", "remaining_time": "1 day, 23:04:32", "throughput": 5661.32, "total_tokens": 339579064}
+{"current_steps": 7770, "total_steps": 29704, "loss": 0.3964075565338135, "lr": 8.404978015601344e-06, "epoch": 0.5231618637220576, "percentage": 26.16, "elapsed_time": "16:40:29", "remaining_time": "1 day, 23:04:18", "throughput": 5660.57, "total_tokens": 339801840}
+{"current_steps": 7775, "total_steps": 29704, "loss": 0.39160215854644775, "lr": 8.403041311274091e-06, "epoch": 0.5234985187180178, "percentage": 26.17, "elapsed_time": "16:41:11", "remaining_time": "1 day, 23:03:48", "throughput": 5660.37, "total_tokens": 340025832}
+{"current_steps": 7780, "total_steps": 29704, "loss": 0.3895624876022339, "lr": 8.401103655297188e-06, "epoch": 0.5238351737139779, "percentage": 26.19, "elapsed_time": "16:41:45", "remaining_time": "1 day, 23:02:58", "throughput": 5660.67, "total_tokens": 340239616}
+{"current_steps": 7785, "total_steps": 29704, "loss": 0.3589164257049561, "lr": 8.399165048212494e-06, "epoch": 0.5241718287099381, "percentage": 26.21, "elapsed_time": "16:42:25", "remaining_time": "1 day, 23:02:21", "throughput": 5660.53, "total_tokens": 340454984}
+{"current_steps": 7790, "total_steps": 29704, "loss": 0.4032485008239746, "lr": 8.397225490562131e-06, "epoch": 0.5245084837058982, "percentage": 26.23, "elapsed_time": "16:43:00", "remaining_time": "1 day, 23:01:32", "throughput": 5661.08, "total_tokens": 340684288}
+{"current_steps": 7795, "total_steps": 29704, "loss": 0.3588679313659668, "lr": 8.395284982888494e-06, "epoch": 0.5248451387018583, "percentage": 26.24, "elapsed_time": "16:43:36", "remaining_time": "1 day, 23:00:47", "throughput": 5661.3, "total_tokens": 340904512}
+{"current_steps": 7800, "total_steps": 29704, "loss": 0.39098711013793946, "lr": 8.393343525734239e-06, "epoch": 0.5251817936978185, "percentage": 26.26, "elapsed_time": "16:44:14", "remaining_time": "1 day, 23:00:08", "throughput": 5661.43, "total_tokens": 341128688}
+{"current_steps": 7805, "total_steps": 29704, "loss": 0.39932818412780763, "lr": 8.391401119642286e-06, "epoch": 0.5255184486937786, "percentage": 26.28, "elapsed_time": "16:45:02", "remaining_time": "1 day, 22:59:54", "throughput": 5660.23, "total_tokens": 341326256}
+{"current_steps": 7810, "total_steps": 29704, "loss": 0.40273380279541016, "lr": 8.389457765155824e-06, "epoch": 0.5258551036897388, "percentage": 26.29, "elapsed_time": "16:45:47", "remaining_time": "1 day, 22:59:32", "throughput": 5659.37, "total_tokens": 341526840}
+{"current_steps": 7815, "total_steps": 29704, "loss": 0.36842942237854004, "lr": 8.387513462818309e-06, "epoch": 0.5261917586856989, "percentage": 26.31, "elapsed_time": "16:46:28", "remaining_time": "1 day, 22:59:00", "throughput": 5659.19, "total_tokens": 341747720}
+{"current_steps": 7820, "total_steps": 29704, "loss": 0.3687218427658081, "lr": 8.385568213173453e-06, "epoch": 0.526528413681659, "percentage": 26.33, "elapsed_time": "16:47:10", "remaining_time": "1 day, 22:58:33", "throughput": 5658.74, "total_tokens": 341961736}
+{"current_steps": 7825, "total_steps": 29704, "loss": 0.3595340013504028, "lr": 8.383622016765242e-06, "epoch": 0.5268650686776192, "percentage": 26.34, "elapsed_time": "16:47:54", "remaining_time": "1 day, 22:58:08", "throughput": 5658.28, "total_tokens": 342180688}
+{"current_steps": 7830, "total_steps": 29704, "loss": 0.4014129638671875, "lr": 8.381674874137926e-06, "epoch": 0.5272017236735793, "percentage": 26.36, "elapsed_time": "16:48:27", "remaining_time": "1 day, 22:57:15", "throughput": 5658.62, "total_tokens": 342390504}
+{"current_steps": 7835, "total_steps": 29704, "loss": 0.3654167652130127, "lr": 8.379726785836013e-06, "epoch": 0.5275383786695395, "percentage": 26.38, "elapsed_time": "16:49:10", "remaining_time": "1 day, 22:56:48", "throughput": 5658.12, "total_tokens": 342602080}
+{"current_steps": 7840, "total_steps": 29704, "loss": 0.36977086067199705, "lr": 8.377777752404283e-06, "epoch": 0.5278750336654996, "percentage": 26.39, "elapsed_time": "16:49:51", "remaining_time": "1 day, 22:56:15", "throughput": 5658.12, "total_tokens": 342831800}
+{"current_steps": 7845, "total_steps": 29704, "loss": 0.34728164672851564, "lr": 8.375827774387774e-06, "epoch": 0.5282116886614597, "percentage": 26.41, "elapsed_time": "16:50:33", "remaining_time": "1 day, 22:55:47", "throughput": 5657.42, "total_tokens": 343029568}
+{"current_steps": 7850, "total_steps": 29704, "loss": 0.3438518762588501, "lr": 8.373876852331793e-06, "epoch": 0.5285483436574199, "percentage": 26.43, "elapsed_time": "16:51:11", "remaining_time": "1 day, 22:55:07", "throughput": 5657.55, "total_tokens": 343254472}
+{"current_steps": 7855, "total_steps": 29704, "loss": 0.38301682472229004, "lr": 8.371924986781909e-06, "epoch": 0.52888499865338, "percentage": 26.44, "elapsed_time": "16:51:49", "remaining_time": "1 day, 22:54:25", "throughput": 5657.61, "total_tokens": 343470008}
+{"current_steps": 7860, "total_steps": 29704, "loss": 0.36738762855529783, "lr": 8.369972178283955e-06, "epoch": 0.5292216536493402, "percentage": 26.46, "elapsed_time": "16:52:27", "remaining_time": "1 day, 22:53:45", "throughput": 5657.81, "total_tokens": 343698072}
+{"current_steps": 7865, "total_steps": 29704, "loss": 0.38115482330322265, "lr": 8.368018427384027e-06, "epoch": 0.5295583086453003, "percentage": 26.48, "elapsed_time": "16:53:11", "remaining_time": "1 day, 22:53:20", "throughput": 5657.41, "total_tokens": 343919960}
+{"current_steps": 7870, "total_steps": 29704, "loss": 0.38156135082244874, "lr": 8.366063734628485e-06, "epoch": 0.5298949636412604, "percentage": 26.49, "elapsed_time": "16:53:47", "remaining_time": "1 day, 22:52:36", "throughput": 5657.36, "total_tokens": 344123568}
+{"current_steps": 7875, "total_steps": 29704, "loss": 0.4007874011993408, "lr": 8.364108100563955e-06, "epoch": 0.5302316186372206, "percentage": 26.51, "elapsed_time": "16:54:24", "remaining_time": "1 day, 22:51:52", "throughput": 5657.52, "total_tokens": 344341784}
+{"current_steps": 7880, "total_steps": 29704, "loss": 0.4212652206420898, "lr": 8.362151525737321e-06, "epoch": 0.5305682736331807, "percentage": 26.53, "elapsed_time": "16:55:01", "remaining_time": "1 day, 22:51:09", "throughput": 5657.69, "total_tokens": 344562552}
+{"current_steps": 7885, "total_steps": 29704, "loss": 0.3570136547088623, "lr": 8.360194010695733e-06, "epoch": 0.5309049286291408, "percentage": 26.55, "elapsed_time": "16:55:42", "remaining_time": "1 day, 22:50:36", "throughput": 5657.51, "total_tokens": 344782336}
+{"current_steps": 7890, "total_steps": 29704, "loss": 0.39661598205566406, "lr": 8.358235555986607e-06, "epoch": 0.531241583625101, "percentage": 26.56, "elapsed_time": "16:56:16", "remaining_time": "1 day, 22:49:46", "throughput": 5658.03, "total_tokens": 345008672}
+{"current_steps": 7895, "total_steps": 29704, "loss": 0.3597188234329224, "lr": 8.356276162157613e-06, "epoch": 0.5315782386210611, "percentage": 26.58, "elapsed_time": "16:57:01", "remaining_time": "1 day, 22:49:25", "throughput": 5657.25, "total_tokens": 345216360}
+{"current_steps": 7900, "total_steps": 29704, "loss": 0.37782256603240966, "lr": 8.354315829756696e-06, "epoch": 0.5319148936170213, "percentage": 26.6, "elapsed_time": "16:57:36", "remaining_time": "1 day, 22:48:37", "throughput": 5657.72, "total_tokens": 345442968}
+{"current_steps": 7905, "total_steps": 29704, "loss": 0.36695542335510256, "lr": 8.352354559332053e-06, "epoch": 0.5322515486129814, "percentage": 26.61, "elapsed_time": "16:58:13", "remaining_time": "1 day, 22:47:53", "throughput": 5657.93, "total_tokens": 345663248}
+{"current_steps": 7910, "total_steps": 29704, "loss": 0.3877138376235962, "lr": 8.350392351432146e-06, "epoch": 0.5325882036089415, "percentage": 26.63, "elapsed_time": "16:58:45", "remaining_time": "1 day, 22:46:56", "throughput": 5658.46, "total_tokens": 345877640}
+{"current_steps": 7915, "total_steps": 29704, "loss": 0.3623414754867554, "lr": 8.348429206605702e-06, "epoch": 0.5329248586049017, "percentage": 26.65, "elapsed_time": "16:59:30", "remaining_time": "1 day, 22:46:33", "throughput": 5658.2, "total_tokens": 346113080}
+{"current_steps": 7920, "total_steps": 29704, "loss": 0.41824941635131835, "lr": 8.346465125401706e-06, "epoch": 0.5332615136008618, "percentage": 26.66, "elapsed_time": "17:00:13", "remaining_time": "1 day, 22:46:08", "throughput": 5657.74, "total_tokens": 346331632}
+{"current_steps": 7925, "total_steps": 29704, "loss": 0.3689335107803345, "lr": 8.344500108369412e-06, "epoch": 0.533598168596822, "percentage": 26.68, "elapsed_time": "17:00:54", "remaining_time": "1 day, 22:45:36", "throughput": 5657.36, "total_tokens": 346540104}
+{"current_steps": 7930, "total_steps": 29704, "loss": 0.3848552942276001, "lr": 8.342534156058326e-06, "epoch": 0.5339348235927821, "percentage": 26.7, "elapsed_time": "17:01:33", "remaining_time": "1 day, 22:44:56", "throughput": 5657.4, "total_tokens": 346759048}
+{"current_steps": 7935, "total_steps": 29704, "loss": 0.39329664707183837, "lr": 8.34056726901822e-06, "epoch": 0.5342714785887422, "percentage": 26.71, "elapsed_time": "17:02:13", "remaining_time": "1 day, 22:44:24", "throughput": 5657.09, "total_tokens": 346971992}
+{"current_steps": 7940, "total_steps": 29704, "loss": 0.37961349487304685, "lr": 8.338599447799132e-06, "epoch": 0.5346081335847024, "percentage": 26.73, "elapsed_time": "17:02:47", "remaining_time": "1 day, 22:43:30", "throughput": 5657.38, "total_tokens": 347176784}
+{"current_steps": 7945, "total_steps": 29704, "loss": 0.3706023931503296, "lr": 8.336630692951353e-06, "epoch": 0.5349447885806625, "percentage": 26.75, "elapsed_time": "17:03:25", "remaining_time": "1 day, 22:42:51", "throughput": 5657.42, "total_tokens": 347396336}
+{"current_steps": 7950, "total_steps": 29704, "loss": 0.3996375560760498, "lr": 8.334661005025441e-06, "epoch": 0.5352814435766227, "percentage": 26.76, "elapsed_time": "17:04:03", "remaining_time": "1 day, 22:42:10", "throughput": 5657.45, "total_tokens": 347612392}
+{"current_steps": 7955, "total_steps": 29704, "loss": 0.4039102077484131, "lr": 8.332690384572213e-06, "epoch": 0.5356180985725828, "percentage": 26.78, "elapsed_time": "17:04:41", "remaining_time": "1 day, 22:41:31", "throughput": 5657.41, "total_tokens": 347827352}
+{"current_steps": 7960, "total_steps": 29704, "loss": 0.3981315612792969, "lr": 8.330718832142746e-06, "epoch": 0.5359547535685429, "percentage": 26.8, "elapsed_time": "17:05:20", "remaining_time": "1 day, 22:40:52", "throughput": 5657.66, "total_tokens": 348060432}
+{"current_steps": 7965, "total_steps": 29704, "loss": 0.3804043769836426, "lr": 8.328746348288379e-06, "epoch": 0.5362914085645031, "percentage": 26.81, "elapsed_time": "17:06:02", "remaining_time": "1 day, 22:40:23", "throughput": 5657.12, "total_tokens": 348266616}
+{"current_steps": 7970, "total_steps": 29704, "loss": 0.42568235397338866, "lr": 8.32677293356071e-06, "epoch": 0.5366280635604632, "percentage": 26.83, "elapsed_time": "17:06:38", "remaining_time": "1 day, 22:39:36", "throughput": 5657.61, "total_tokens": 348498272}
+{"current_steps": 7975, "total_steps": 29704, "loss": 0.3702836275100708, "lr": 8.324798588511598e-06, "epoch": 0.5369647185564234, "percentage": 26.85, "elapsed_time": "17:07:15", "remaining_time": "1 day, 22:38:54", "throughput": 5657.4, "total_tokens": 348696416}
+{"current_steps": 7980, "total_steps": 29704, "loss": 0.38841800689697265, "lr": 8.322823313693162e-06, "epoch": 0.5373013735523835, "percentage": 26.87, "elapsed_time": "17:07:51", "remaining_time": "1 day, 22:38:08", "throughput": 5657.57, "total_tokens": 348910592}
+{"current_steps": 7985, "total_steps": 29704, "loss": 0.3783905506134033, "lr": 8.320847109657782e-06, "epoch": 0.5376380285483436, "percentage": 26.88, "elapsed_time": "17:08:32", "remaining_time": "1 day, 22:37:37", "throughput": 5657.16, "total_tokens": 349120256}
+{"current_steps": 7990, "total_steps": 29704, "loss": 0.3449576377868652, "lr": 8.318869976958099e-06, "epoch": 0.5379746835443038, "percentage": 26.9, "elapsed_time": "17:09:15", "remaining_time": "1 day, 22:37:10", "throughput": 5656.79, "total_tokens": 349340688}
+{"current_steps": 7995, "total_steps": 29704, "loss": 0.40518803596496583, "lr": 8.316891916147007e-06, "epoch": 0.5383113385402639, "percentage": 26.92, "elapsed_time": "17:09:55", "remaining_time": "1 day, 22:36:33", "throughput": 5656.72, "total_tokens": 349557408}
+{"current_steps": 8000, "total_steps": 29704, "loss": 0.3914914608001709, "lr": 8.314912927777671e-06, "epoch": 0.5386479935362241, "percentage": 26.93, "elapsed_time": "17:10:29", "remaining_time": "1 day, 22:35:43", "throughput": 5657.11, "total_tokens": 349776936}
+{"current_steps": 8005, "total_steps": 29704, "loss": 0.3830781698226929, "lr": 8.312933012403503e-06, "epoch": 0.5389846485321842, "percentage": 26.95, "elapsed_time": "17:11:10", "remaining_time": "1 day, 22:35:10", "throughput": 5656.72, "total_tokens": 349981944}
+{"current_steps": 8010, "total_steps": 29704, "loss": 0.3917949676513672, "lr": 8.310952170578184e-06, "epoch": 0.5393213035281443, "percentage": 26.97, "elapsed_time": "17:11:50", "remaining_time": "1 day, 22:34:37", "throughput": 5656.46, "total_tokens": 350196560}
+{"current_steps": 8015, "total_steps": 29704, "loss": 0.35423126220703127, "lr": 8.308970402855647e-06, "epoch": 0.5396579585241045, "percentage": 26.98, "elapsed_time": "17:12:22", "remaining_time": "1 day, 22:33:40", "throughput": 5656.94, "total_tokens": 350407320}
+{"current_steps": 8020, "total_steps": 29704, "loss": 0.36304192543029784, "lr": 8.306987709790087e-06, "epoch": 0.5399946135200646, "percentage": 27.0, "elapsed_time": "17:12:54", "remaining_time": "1 day, 22:32:43", "throughput": 5657.52, "total_tokens": 350621920}
+{"current_steps": 8025, "total_steps": 29704, "loss": 0.3827538013458252, "lr": 8.305004091935962e-06, "epoch": 0.5403312685160248, "percentage": 27.02, "elapsed_time": "17:13:30", "remaining_time": "1 day, 22:31:57", "throughput": 5657.75, "total_tokens": 350840936}
+{"current_steps": 8030, "total_steps": 29704, "loss": 0.40552120208740233, "lr": 8.303019549847979e-06, "epoch": 0.5406679235119849, "percentage": 27.03, "elapsed_time": "17:14:11", "remaining_time": "1 day, 22:31:25", "throughput": 5657.49, "total_tokens": 351057864}
+{"current_steps": 8035, "total_steps": 29704, "loss": 0.38002500534057615, "lr": 8.301034084081114e-06, "epoch": 0.541004578507945, "percentage": 27.05, "elapsed_time": "17:14:50", "remaining_time": "1 day, 22:30:47", "throughput": 5657.63, "total_tokens": 351285720}
+{"current_steps": 8040, "total_steps": 29704, "loss": 0.408983850479126, "lr": 8.299047695190592e-06, "epoch": 0.5413412335039052, "percentage": 27.07, "elapsed_time": "17:15:27", "remaining_time": "1 day, 22:30:03", "throughput": 5657.62, "total_tokens": 351491576}
+{"current_steps": 8045, "total_steps": 29704, "loss": 0.38953261375427245, "lr": 8.297060383731903e-06, "epoch": 0.5416778884998653, "percentage": 27.08, "elapsed_time": "17:16:02", "remaining_time": "1 day, 22:29:17", "throughput": 5657.81, "total_tokens": 351706376}
+{"current_steps": 8050, "total_steps": 29704, "loss": 0.3989473819732666, "lr": 8.29507215026079e-06, "epoch": 0.5420145434958255, "percentage": 27.1, "elapsed_time": "17:16:34", "remaining_time": "1 day, 22:28:18", "throughput": 5658.48, "total_tokens": 351925296}
+{"current_steps": 8055, "total_steps": 29704, "loss": 0.3805055618286133, "lr": 8.293082995333257e-06, "epoch": 0.5423511984917856, "percentage": 27.12, "elapsed_time": "17:17:15", "remaining_time": "1 day, 22:27:46", "throughput": 5658.41, "total_tokens": 352152728}
+{"current_steps": 8060, "total_steps": 29704, "loss": 0.359727144241333, "lr": 8.291092919505568e-06, "epoch": 0.5426878534877457, "percentage": 27.13, "elapsed_time": "17:17:52", "remaining_time": "1 day, 22:27:04", "throughput": 5658.48, "total_tokens": 352368040}
+{"current_steps": 8065, "total_steps": 29704, "loss": 0.3927903175354004, "lr": 8.289101923334237e-06, "epoch": 0.5430245084837059, "percentage": 27.15, "elapsed_time": "17:18:34", "remaining_time": "1 day, 22:26:33", "throughput": 5658.31, "total_tokens": 352592456}
+{"current_steps": 8070, "total_steps": 29704, "loss": 0.37551441192626955, "lr": 8.28711000737604e-06, "epoch": 0.543361163479666, "percentage": 27.17, "elapsed_time": "17:19:12", "remaining_time": "1 day, 22:25:53", "throughput": 5658.21, "total_tokens": 352801680}
+{"current_steps": 8075, "total_steps": 29704, "loss": 0.3938951253890991, "lr": 8.285117172188013e-06, "epoch": 0.5436978184756261, "percentage": 27.18, "elapsed_time": "17:19:46", "remaining_time": "1 day, 22:25:01", "throughput": 5658.73, "total_tokens": 353025760}
+{"current_steps": 8080, "total_steps": 29704, "loss": 0.4004195213317871, "lr": 8.283123418327443e-06, "epoch": 0.5440344734715863, "percentage": 27.2, "elapsed_time": "17:20:19", "remaining_time": "1 day, 22:24:08", "throughput": 5659.23, "total_tokens": 353244352}
+{"current_steps": 8085, "total_steps": 29704, "loss": 0.3595161437988281, "lr": 8.281128746351878e-06, "epoch": 0.5443711284675464, "percentage": 27.22, "elapsed_time": "17:20:53", "remaining_time": "1 day, 22:23:17", "throughput": 5659.83, "total_tokens": 353475168}
+{"current_steps": 8090, "total_steps": 29704, "loss": 0.40300402641296384, "lr": 8.279133156819123e-06, "epoch": 0.5447077834635066, "percentage": 27.24, "elapsed_time": "17:21:36", "remaining_time": "1 day, 22:22:52", "throughput": 5659.35, "total_tokens": 353691088}
+{"current_steps": 8095, "total_steps": 29704, "loss": 0.3694662570953369, "lr": 8.277136650287237e-06, "epoch": 0.5450444384594667, "percentage": 27.25, "elapsed_time": "17:22:16", "remaining_time": "1 day, 22:22:16", "throughput": 5659.4, "total_tokens": 353917984}
+{"current_steps": 8100, "total_steps": 29704, "loss": 0.3757363557815552, "lr": 8.275139227314537e-06, "epoch": 0.5453810934554268, "percentage": 27.27, "elapsed_time": "17:22:55", "remaining_time": "1 day, 22:21:38", "throughput": 5659.19, "total_tokens": 354125952}
+{"current_steps": 8105, "total_steps": 29704, "loss": 0.4242837905883789, "lr": 8.273140888459594e-06, "epoch": 0.545717748451387, "percentage": 27.29, "elapsed_time": "17:23:26", "remaining_time": "1 day, 22:20:39", "throughput": 5660.09, "total_tokens": 354358280}
+{"current_steps": 8110, "total_steps": 29704, "loss": 0.3934792518615723, "lr": 8.27114163428124e-06, "epoch": 0.5460544034473471, "percentage": 27.3, "elapsed_time": "17:24:02", "remaining_time": "1 day, 22:19:53", "throughput": 5660.34, "total_tokens": 354575664}
+{"current_steps": 8115, "total_steps": 29704, "loss": 0.38090806007385253, "lr": 8.26914146533856e-06, "epoch": 0.5463910584433073, "percentage": 27.32, "elapsed_time": "17:24:41", "remaining_time": "1 day, 22:19:17", "throughput": 5660.47, "total_tokens": 354808384}
+{"current_steps": 8120, "total_steps": 29704, "loss": 0.3828258037567139, "lr": 8.267140382190892e-06, "epoch": 0.5467277134392674, "percentage": 27.34, "elapsed_time": "17:25:20", "remaining_time": "1 day, 22:18:39", "throughput": 5660.46, "total_tokens": 355026976}
+{"current_steps": 8125, "total_steps": 29704, "loss": 0.3777566432952881, "lr": 8.265138385397835e-06, "epoch": 0.5470643684352275, "percentage": 27.35, "elapsed_time": "17:25:58", "remaining_time": "1 day, 22:17:57", "throughput": 5660.53, "total_tokens": 355244064}
+{"current_steps": 8130, "total_steps": 29704, "loss": 0.34886605739593507, "lr": 8.263135475519241e-06, "epoch": 0.5474010234311877, "percentage": 27.37, "elapsed_time": "17:26:34", "remaining_time": "1 day, 22:17:12", "throughput": 5661.07, "total_tokens": 355483272}
+{"current_steps": 8135, "total_steps": 29704, "loss": 0.42696261405944824, "lr": 8.261131653115216e-06, "epoch": 0.5477376784271478, "percentage": 27.39, "elapsed_time": "17:27:14", "remaining_time": "1 day, 22:16:37", "throughput": 5661.1, "total_tokens": 355711240}
+{"current_steps": 8140, "total_steps": 29704, "loss": 0.36356501579284667, "lr": 8.259126918746122e-06, "epoch": 0.548074333423108, "percentage": 27.4, "elapsed_time": "17:27:48", "remaining_time": "1 day, 22:15:47", "throughput": 5661.49, "total_tokens": 355930576}
+{"current_steps": 8145, "total_steps": 29704, "loss": 0.3778836250305176, "lr": 8.25712127297258e-06, "epoch": 0.5484109884190681, "percentage": 27.42, "elapsed_time": "17:28:24", "remaining_time": "1 day, 22:15:02", "throughput": 5661.77, "total_tokens": 356150816}
+{"current_steps": 8150, "total_steps": 29704, "loss": 0.4094264507293701, "lr": 8.255114716355459e-06, "epoch": 0.5487476434150282, "percentage": 27.44, "elapsed_time": "17:29:02", "remaining_time": "1 day, 22:14:20", "throughput": 5662.07, "total_tokens": 356382088}
+{"current_steps": 8155, "total_steps": 29704, "loss": 0.3917287826538086, "lr": 8.253107249455887e-06, "epoch": 0.5490842984109884, "percentage": 27.45, "elapsed_time": "17:29:35", "remaining_time": "1 day, 22:13:27", "throughput": 5662.57, "total_tokens": 356602352}
+{"current_steps": 8160, "total_steps": 29704, "loss": 0.39372997283935546, "lr": 8.251098872835247e-06, "epoch": 0.5494209534069485, "percentage": 27.47, "elapsed_time": "17:30:17", "remaining_time": "1 day, 22:12:59", "throughput": 5662.28, "total_tokens": 356825280}
+{"current_steps": 8165, "total_steps": 29704, "loss": 0.3526461601257324, "lr": 8.249089587055175e-06, "epoch": 0.5497576084029087, "percentage": 27.49, "elapsed_time": "17:30:59", "remaining_time": "1 day, 22:12:29", "throughput": 5661.91, "total_tokens": 357037664}
+{"current_steps": 8170, "total_steps": 29704, "loss": 0.3740200996398926, "lr": 8.247079392677558e-06, "epoch": 0.5500942633988688, "percentage": 27.5, "elapsed_time": "17:31:41", "remaining_time": "1 day, 22:11:58", "throughput": 5661.66, "total_tokens": 357257424}
+{"current_steps": 8175, "total_steps": 29704, "loss": 0.3697060108184814, "lr": 8.245068290264544e-06, "epoch": 0.5504309183948289, "percentage": 27.52, "elapsed_time": "17:32:20", "remaining_time": "1 day, 22:11:22", "throughput": 5661.63, "total_tokens": 357480144}
+{"current_steps": 8180, "total_steps": 29704, "loss": 0.37436251640319823, "lr": 8.243056280378531e-06, "epoch": 0.5507675733907891, "percentage": 27.54, "elapsed_time": "17:33:01", "remaining_time": "1 day, 22:10:50", "throughput": 5661.43, "total_tokens": 357699472}
+{"current_steps": 8185, "total_steps": 29704, "loss": 0.36437788009643557, "lr": 8.241043363582171e-06, "epoch": 0.5511042283867492, "percentage": 27.56, "elapsed_time": "17:33:38", "remaining_time": "1 day, 22:10:05", "throughput": 5661.76, "total_tokens": 357926808}
+{"current_steps": 8190, "total_steps": 29704, "loss": 0.3505790948867798, "lr": 8.239029540438369e-06, "epoch": 0.5514408833827094, "percentage": 27.57, "elapsed_time": "17:34:17", "remaining_time": "1 day, 22:09:28", "throughput": 5661.71, "total_tokens": 358145064}
+{"current_steps": 8195, "total_steps": 29704, "loss": 0.3656752109527588, "lr": 8.237014811510284e-06, "epoch": 0.5517775383786695, "percentage": 27.59, "elapsed_time": "17:35:04", "remaining_time": "1 day, 22:09:12", "throughput": 5661.02, "total_tokens": 358369824}
+{"current_steps": 8200, "total_steps": 29704, "loss": 0.3714550018310547, "lr": 8.234999177361329e-06, "epoch": 0.5521141933746296, "percentage": 27.61, "elapsed_time": "17:35:40", "remaining_time": "1 day, 22:08:27", "throughput": 5661.39, "total_tokens": 358596784}
+{"current_steps": 8205, "total_steps": 29704, "loss": 0.40174274444580077, "lr": 8.232982638555169e-06, "epoch": 0.5524508483705898, "percentage": 27.62, "elapsed_time": "17:36:12", "remaining_time": "1 day, 22:07:31", "throughput": 5662.04, "total_tokens": 358820312}
+{"current_steps": 8210, "total_steps": 29704, "loss": 0.39043726921081545, "lr": 8.230965195655724e-06, "epoch": 0.5527875033665499, "percentage": 27.64, "elapsed_time": "17:36:52", "remaining_time": "1 day, 22:06:55", "throughput": 5662.22, "total_tokens": 359055496}
+{"current_steps": 8215, "total_steps": 29704, "loss": 0.3990142345428467, "lr": 8.228946849227167e-06, "epoch": 0.55312415836251, "percentage": 27.66, "elapsed_time": "17:37:29", "remaining_time": "1 day, 22:06:12", "throughput": 5662.59, "total_tokens": 359287376}
+{"current_steps": 8220, "total_steps": 29704, "loss": 0.386507511138916, "lr": 8.226927599833916e-06, "epoch": 0.5534608133584702, "percentage": 27.67, "elapsed_time": "17:38:01", "remaining_time": "1 day, 22:05:16", "throughput": 5663.31, "total_tokens": 359514576}
+{"current_steps": 8225, "total_steps": 29704, "loss": 0.37993340492248534, "lr": 8.224907448040655e-06, "epoch": 0.5537974683544303, "percentage": 27.69, "elapsed_time": "17:38:33", "remaining_time": "1 day, 22:04:19", "throughput": 5664.12, "total_tokens": 359745160}
+{"current_steps": 8230, "total_steps": 29704, "loss": 0.4058523178100586, "lr": 8.222886394412309e-06, "epoch": 0.5541341233503905, "percentage": 27.71, "elapsed_time": "17:39:13", "remaining_time": "1 day, 22:03:46", "throughput": 5663.95, "total_tokens": 359965008}
+{"current_steps": 8235, "total_steps": 29704, "loss": 0.384306526184082, "lr": 8.220864439514057e-06, "epoch": 0.5544707783463506, "percentage": 27.72, "elapsed_time": "17:39:52", "remaining_time": "1 day, 22:03:07", "throughput": 5664.12, "total_tokens": 360194040}
+{"current_steps": 8240, "total_steps": 29704, "loss": 0.3680997848510742, "lr": 8.218841583911338e-06, "epoch": 0.5548074333423108, "percentage": 27.74, "elapsed_time": "17:40:34", "remaining_time": "1 day, 22:02:39", "throughput": 5663.72, "total_tokens": 360409248}
+{"current_steps": 8245, "total_steps": 29704, "loss": 0.413088321685791, "lr": 8.216817828169834e-06, "epoch": 0.5551440883382709, "percentage": 27.76, "elapsed_time": "17:41:10", "remaining_time": "1 day, 22:01:53", "throughput": 5663.92, "total_tokens": 360624480}
+{"current_steps": 8250, "total_steps": 29704, "loss": 0.36465370655059814, "lr": 8.214793172855481e-06, "epoch": 0.555480743334231, "percentage": 27.77, "elapsed_time": "17:41:51", "remaining_time": "1 day, 22:01:19", "throughput": 5663.64, "total_tokens": 360837048}
+{"current_steps": 8255, "total_steps": 29704, "loss": 0.36249728202819825, "lr": 8.21276761853447e-06, "epoch": 0.5558173983301912, "percentage": 27.79, "elapsed_time": "17:42:34", "remaining_time": "1 day, 22:00:54", "throughput": 5663.28, "total_tokens": 361060272}
+{"current_steps": 8260, "total_steps": 29704, "loss": 0.39040985107421877, "lr": 8.210741165773238e-06, "epoch": 0.5561540533261513, "percentage": 27.81, "elapsed_time": "17:43:05", "remaining_time": "1 day, 21:59:55", "throughput": 5664.21, "total_tokens": 361294192}
+{"current_steps": 8265, "total_steps": 29704, "loss": 0.36567642688751223, "lr": 8.208713815138476e-06, "epoch": 0.5564907083221114, "percentage": 27.82, "elapsed_time": "17:43:46", "remaining_time": "1 day, 21:59:22", "throughput": 5664.18, "total_tokens": 361524240}
+{"current_steps": 8270, "total_steps": 29704, "loss": 0.35714569091796877, "lr": 8.20668556719713e-06, "epoch": 0.5568273633180716, "percentage": 27.84, "elapsed_time": "17:44:20", "remaining_time": "1 day, 21:58:31", "throughput": 5664.45, "total_tokens": 361732392}
+{"current_steps": 8275, "total_steps": 29704, "loss": 0.3372737646102905, "lr": 8.204656422516388e-06, "epoch": 0.5571640183140317, "percentage": 27.86, "elapsed_time": "17:45:01", "remaining_time": "1 day, 21:57:59", "throughput": 5664.15, "total_tokens": 361946008}
+{"current_steps": 8280, "total_steps": 29704, "loss": 0.38366384506225587, "lr": 8.2026263816637e-06, "epoch": 0.5575006733099919, "percentage": 27.88, "elapsed_time": "17:45:38", "remaining_time": "1 day, 21:57:16", "throughput": 5664.08, "total_tokens": 362152248}
+{"current_steps": 8285, "total_steps": 29704, "loss": 0.37342875003814696, "lr": 8.200595445206753e-06, "epoch": 0.557837328305952, "percentage": 27.89, "elapsed_time": "17:46:19", "remaining_time": "1 day, 21:56:45", "throughput": 5663.84, "total_tokens": 362371096}
+{"current_steps": 8290, "total_steps": 29704, "loss": 0.37401297092437746, "lr": 8.198563613713497e-06, "epoch": 0.5581739833019121, "percentage": 27.91, "elapsed_time": "17:47:00", "remaining_time": "1 day, 21:56:11", "throughput": 5663.52, "total_tokens": 362579576}
+{"current_steps": 8295, "total_steps": 29704, "loss": 0.3654287338256836, "lr": 8.196530887752127e-06, "epoch": 0.5585106382978723, "percentage": 27.93, "elapsed_time": "17:47:34", "remaining_time": "1 day, 21:55:20", "throughput": 5664.03, "total_tokens": 362803720}
+{"current_steps": 8300, "total_steps": 29704, "loss": 0.37125024795532224, "lr": 8.194497267891086e-06, "epoch": 0.5588472932938324, "percentage": 27.94, "elapsed_time": "17:48:12", "remaining_time": "1 day, 21:54:41", "throughput": 5664.13, "total_tokens": 363027432}
+{"current_steps": 8305, "total_steps": 29704, "loss": 0.3796513795852661, "lr": 8.19246275469907e-06, "epoch": 0.5591839482897926, "percentage": 27.96, "elapsed_time": "17:48:56", "remaining_time": "1 day, 21:54:16", "throughput": 5663.79, "total_tokens": 363254432}
+{"current_steps": 8310, "total_steps": 29704, "loss": 0.36992511749267576, "lr": 8.190427348745027e-06, "epoch": 0.5595206032857528, "percentage": 27.98, "elapsed_time": "17:49:31", "remaining_time": "1 day, 21:53:28", "throughput": 5664.2, "total_tokens": 363479976}
+{"current_steps": 8315, "total_steps": 29704, "loss": 0.3632178544998169, "lr": 8.188391050598146e-06, "epoch": 0.559857258281713, "percentage": 27.99, "elapsed_time": "17:50:16", "remaining_time": "1 day, 21:53:06", "throughput": 5663.85, "total_tokens": 363712504}
+{"current_steps": 8320, "total_steps": 29704, "loss": 0.37043955326080324, "lr": 8.186353860827878e-06, "epoch": 0.5601939132776731, "percentage": 28.01, "elapsed_time": "17:50:56", "remaining_time": "1 day, 21:52:32", "throughput": 5663.52, "total_tokens": 363919784}
+{"current_steps": 8325, "total_steps": 29704, "loss": 0.3709959745407104, "lr": 8.184315780003911e-06, "epoch": 0.5605305682736332, "percentage": 28.03, "elapsed_time": "17:51:42", "remaining_time": "1 day, 21:52:12", "throughput": 5662.92, "total_tokens": 364140736}
+{"current_steps": 8330, "total_steps": 29704, "loss": 0.40601282119750975, "lr": 8.182276808696193e-06, "epoch": 0.5608672232695934, "percentage": 28.04, "elapsed_time": "17:52:27", "remaining_time": "1 day, 21:51:48", "throughput": 5662.23, "total_tokens": 364348280}
+{"current_steps": 8335, "total_steps": 29704, "loss": 0.3827899694442749, "lr": 8.18023694747491e-06, "epoch": 0.5612038782655535, "percentage": 28.06, "elapsed_time": "17:53:01", "remaining_time": "1 day, 21:51:00", "throughput": 5662.7, "total_tokens": 364574968}
+{"current_steps": 8340, "total_steps": 29704, "loss": 0.3656409740447998, "lr": 8.178196196910508e-06, "epoch": 0.5615405332615137, "percentage": 28.08, "elapsed_time": "17:53:41", "remaining_time": "1 day, 21:50:24", "throughput": 5662.74, "total_tokens": 364803736}
+{"current_steps": 8345, "total_steps": 29704, "loss": 0.4000543594360352, "lr": 8.176154557573672e-06, "epoch": 0.5618771882574738, "percentage": 28.09, "elapsed_time": "17:54:19", "remaining_time": "1 day, 21:49:43", "throughput": 5663.06, "total_tokens": 365037240}
+{"current_steps": 8350, "total_steps": 29704, "loss": 0.3629467010498047, "lr": 8.174112030035344e-06, "epoch": 0.5622138432534339, "percentage": 28.11, "elapsed_time": "17:54:53", "remaining_time": "1 day, 21:48:52", "throughput": 5663.68, "total_tokens": 365268200}
+{"current_steps": 8355, "total_steps": 29704, "loss": 0.36340205669403075, "lr": 8.172068614866707e-06, "epoch": 0.5625504982493941, "percentage": 28.13, "elapsed_time": "17:55:31", "remaining_time": "1 day, 21:48:14", "throughput": 5663.52, "total_tokens": 365477184}
+{"current_steps": 8360, "total_steps": 29704, "loss": 0.39545679092407227, "lr": 8.170024312639197e-06, "epoch": 0.5628871532453542, "percentage": 28.14, "elapsed_time": "17:56:10", "remaining_time": "1 day, 21:47:36", "throughput": 5663.66, "total_tokens": 365707680}
+{"current_steps": 8365, "total_steps": 29704, "loss": 0.3859426498413086, "lr": 8.167979123924496e-06, "epoch": 0.5632238082413143, "percentage": 28.16, "elapsed_time": "17:56:44", "remaining_time": "1 day, 21:46:45", "throughput": 5664.14, "total_tokens": 365930048}
+{"current_steps": 8370, "total_steps": 29704, "loss": 0.4087996482849121, "lr": 8.165933049294537e-06, "epoch": 0.5635604632372745, "percentage": 28.18, "elapsed_time": "17:57:26", "remaining_time": "1 day, 21:46:14", "throughput": 5663.85, "total_tokens": 366147352}
+{"current_steps": 8375, "total_steps": 29704, "loss": 0.37345223426818847, "lr": 8.163886089321493e-06, "epoch": 0.5638971182332346, "percentage": 28.19, "elapsed_time": "17:58:08", "remaining_time": "1 day, 21:45:45", "throughput": 5663.7, "total_tokens": 366377160}
+{"current_steps": 8380, "total_steps": 29704, "loss": 0.3573953151702881, "lr": 8.161838244577795e-06, "epoch": 0.5642337732291948, "percentage": 28.21, "elapsed_time": "17:58:46", "remaining_time": "1 day, 21:45:04", "throughput": 5663.8, "total_tokens": 366596264}
+{"current_steps": 8385, "total_steps": 29704, "loss": 0.3621669769287109, "lr": 8.159789515636114e-06, "epoch": 0.5645704282251549, "percentage": 28.23, "elapsed_time": "17:59:28", "remaining_time": "1 day, 21:44:35", "throughput": 5663.46, "total_tokens": 366814672}
+{"current_steps": 8390, "total_steps": 29704, "loss": 0.39387807846069334, "lr": 8.15773990306937e-06, "epoch": 0.564907083221115, "percentage": 28.25, "elapsed_time": "18:00:03", "remaining_time": "1 day, 21:43:47", "throughput": 5663.43, "total_tokens": 367010840}
+{"current_steps": 8395, "total_steps": 29704, "loss": 0.37447366714477537, "lr": 8.155689407450732e-06, "epoch": 0.5652437382170752, "percentage": 28.26, "elapsed_time": "18:00:40", "remaining_time": "1 day, 21:43:04", "throughput": 5663.69, "total_tokens": 367235664}
+{"current_steps": 8400, "total_steps": 29704, "loss": 0.3721097707748413, "lr": 8.153638029353615e-06, "epoch": 0.5655803932130353, "percentage": 28.28, "elapsed_time": "18:01:20", "remaining_time": "1 day, 21:42:28", "throughput": 5663.34, "total_tokens": 367439472}
+{"current_steps": 8405, "total_steps": 29704, "loss": 0.416213321685791, "lr": 8.151585769351678e-06, "epoch": 0.5659170482089955, "percentage": 28.3, "elapsed_time": "18:01:55", "remaining_time": "1 day, 21:41:42", "throughput": 5663.72, "total_tokens": 367665000}
+{"current_steps": 8410, "total_steps": 29704, "loss": 0.3906120538711548, "lr": 8.149532628018833e-06, "epoch": 0.5662537032049556, "percentage": 28.31, "elapsed_time": "18:02:34", "remaining_time": "1 day, 21:41:04", "throughput": 5663.53, "total_tokens": 367873544}
+{"current_steps": 8415, "total_steps": 29704, "loss": 0.36734464168548586, "lr": 8.147478605929228e-06, "epoch": 0.5665903582009157, "percentage": 28.33, "elapsed_time": "18:03:08", "remaining_time": "1 day, 21:40:14", "throughput": 5664.15, "total_tokens": 368106424}
+{"current_steps": 8420, "total_steps": 29704, "loss": 0.3696259021759033, "lr": 8.145423703657269e-06, "epoch": 0.5669270131968759, "percentage": 28.35, "elapsed_time": "18:03:55", "remaining_time": "1 day, 21:39:55", "throughput": 5663.54, "total_tokens": 368329816}
+{"current_steps": 8425, "total_steps": 29704, "loss": 0.3901944637298584, "lr": 8.143367921777601e-06, "epoch": 0.567263668192836, "percentage": 28.36, "elapsed_time": "18:04:33", "remaining_time": "1 day, 21:39:15", "throughput": 5663.72, "total_tokens": 368556080}
+{"current_steps": 8430, "total_steps": 29704, "loss": 0.42763843536376955, "lr": 8.141311260865117e-06, "epoch": 0.5676003231887962, "percentage": 28.38, "elapsed_time": "18:05:07", "remaining_time": "1 day, 21:38:24", "throughput": 5664.31, "total_tokens": 368786216}
+{"current_steps": 8435, "total_steps": 29704, "loss": 0.4124480724334717, "lr": 8.139253721494956e-06, "epoch": 0.5679369781847563, "percentage": 28.4, "elapsed_time": "18:05:43", "remaining_time": "1 day, 21:37:40", "throughput": 5664.38, "total_tokens": 368998320}
+{"current_steps": 8440, "total_steps": 29704, "loss": 0.3694611549377441, "lr": 8.137195304242502e-06, "epoch": 0.5682736331807164, "percentage": 28.41, "elapsed_time": "18:06:20", "remaining_time": "1 day, 21:36:56", "throughput": 5664.64, "total_tokens": 369221808}
+{"current_steps": 8445, "total_steps": 29704, "loss": 0.3659370899200439, "lr": 8.135136009683387e-06, "epoch": 0.5686102881766766, "percentage": 28.43, "elapsed_time": "18:06:54", "remaining_time": "1 day, 21:36:06", "throughput": 5665.03, "total_tokens": 369439952}
+{"current_steps": 8450, "total_steps": 29704, "loss": 0.37821600437164304, "lr": 8.13307583839348e-06, "epoch": 0.5689469431726367, "percentage": 28.45, "elapsed_time": "18:07:33", "remaining_time": "1 day, 21:35:31", "throughput": 5665.03, "total_tokens": 369665384}
+{"current_steps": 8455, "total_steps": 29704, "loss": 0.3932258367538452, "lr": 8.131014790948908e-06, "epoch": 0.5692835981685969, "percentage": 28.46, "elapsed_time": "18:08:12", "remaining_time": "1 day, 21:34:52", "throughput": 5665.04, "total_tokens": 369885352}
+{"current_steps": 8460, "total_steps": 29704, "loss": 0.42728986740112307, "lr": 8.128952867926035e-06, "epoch": 0.569620253164557, "percentage": 28.48, "elapsed_time": "18:08:46", "remaining_time": "1 day, 21:34:02", "throughput": 5665.42, "total_tokens": 370102168}
+{"current_steps": 8465, "total_steps": 29704, "loss": 0.3966690540313721, "lr": 8.126890069901468e-06, "epoch": 0.5699569081605171, "percentage": 28.5, "elapsed_time": "18:09:25", "remaining_time": "1 day, 21:33:24", "throughput": 5665.42, "total_tokens": 370323328}
+{"current_steps": 8470, "total_steps": 29704, "loss": 0.3729779005050659, "lr": 8.124826397452064e-06, "epoch": 0.5702935631564773, "percentage": 28.51, "elapsed_time": "18:10:03", "remaining_time": "1 day, 21:32:44", "throughput": 5665.64, "total_tokens": 370551792}
+{"current_steps": 8475, "total_steps": 29704, "loss": 0.37224202156066893, "lr": 8.122761851154923e-06, "epoch": 0.5706302181524374, "percentage": 28.53, "elapsed_time": "18:10:37", "remaining_time": "1 day, 21:31:53", "throughput": 5666.04, "total_tokens": 370769584}
+{"current_steps": 8480, "total_steps": 29704, "loss": 0.3793643474578857, "lr": 8.120696431587387e-06, "epoch": 0.5709668731483976, "percentage": 28.55, "elapsed_time": "18:11:17", "remaining_time": "1 day, 21:31:18", "throughput": 5665.87, "total_tokens": 370986856}
+{"current_steps": 8485, "total_steps": 29704, "loss": 0.36144418716430665, "lr": 8.118630139327046e-06, "epoch": 0.5713035281443577, "percentage": 28.57, "elapsed_time": "18:12:00", "remaining_time": "1 day, 21:30:51", "throughput": 5665.45, "total_tokens": 371203736}
+{"current_steps": 8490, "total_steps": 29704, "loss": 0.3969851016998291, "lr": 8.116562974951731e-06, "epoch": 0.5716401831403178, "percentage": 28.58, "elapsed_time": "18:12:42", "remaining_time": "1 day, 21:30:20", "throughput": 5665.37, "total_tokens": 371433168}
+{"current_steps": 8495, "total_steps": 29704, "loss": 0.37904407978057864, "lr": 8.11449493903952e-06, "epoch": 0.571976838136278, "percentage": 28.6, "elapsed_time": "18:13:20", "remaining_time": "1 day, 21:29:42", "throughput": 5665.49, "total_tokens": 371661928}
+{"current_steps": 8500, "total_steps": 29704, "loss": 0.39436898231506345, "lr": 8.112426032168727e-06, "epoch": 0.5723134931322381, "percentage": 28.62, "elapsed_time": "18:13:59", "remaining_time": "1 day, 21:29:04", "throughput": 5665.42, "total_tokens": 371877904}
+{"current_steps": 8505, "total_steps": 29704, "loss": 0.35653374195098875, "lr": 8.11035625491792e-06, "epoch": 0.5726501481281983, "percentage": 28.63, "elapsed_time": "18:14:33", "remaining_time": "1 day, 21:28:13", "throughput": 5665.66, "total_tokens": 372083344}
+{"current_steps": 8510, "total_steps": 29704, "loss": 0.3748607158660889, "lr": 8.108285607865907e-06, "epoch": 0.5729868031241584, "percentage": 28.65, "elapsed_time": "18:15:09", "remaining_time": "1 day, 21:27:29", "throughput": 5665.76, "total_tokens": 372296008}
+{"current_steps": 8515, "total_steps": 29704, "loss": 0.36768136024475095, "lr": 8.106214091591731e-06, "epoch": 0.5733234581201185, "percentage": 28.67, "elapsed_time": "18:15:47", "remaining_time": "1 day, 21:26:47", "throughput": 5665.92, "total_tokens": 372517376}
+{"current_steps": 8520, "total_steps": 29704, "loss": 0.3505872249603271, "lr": 8.104141706674693e-06, "epoch": 0.5736601131160787, "percentage": 28.68, "elapsed_time": "18:16:26", "remaining_time": "1 day, 21:26:10", "throughput": 5665.97, "total_tokens": 372744072}
+{"current_steps": 8525, "total_steps": 29704, "loss": 0.39815869331359866, "lr": 8.102068453694324e-06, "epoch": 0.5739967681120388, "percentage": 28.7, "elapsed_time": "18:17:07", "remaining_time": "1 day, 21:25:36", "throughput": 5665.83, "total_tokens": 372965496}
+{"current_steps": 8530, "total_steps": 29704, "loss": 0.39165263175964354, "lr": 8.099994333230404e-06, "epoch": 0.574333423107999, "percentage": 28.72, "elapsed_time": "18:17:49", "remaining_time": "1 day, 21:25:08", "throughput": 5665.35, "total_tokens": 373176000}
+{"current_steps": 8535, "total_steps": 29704, "loss": 0.387595534324646, "lr": 8.097919345862952e-06, "epoch": 0.5746700781039591, "percentage": 28.73, "elapsed_time": "18:18:21", "remaining_time": "1 day, 21:24:12", "throughput": 5665.65, "total_tokens": 373375616}
+{"current_steps": 8540, "total_steps": 29704, "loss": 0.3876967906951904, "lr": 8.095843492172235e-06, "epoch": 0.5750067330999192, "percentage": 28.75, "elapsed_time": "18:19:01", "remaining_time": "1 day, 21:23:38", "throughput": 5665.48, "total_tokens": 373591328}
+{"current_steps": 8545, "total_steps": 29704, "loss": 0.3301082611083984, "lr": 8.093766772738758e-06, "epoch": 0.5753433880958794, "percentage": 28.77, "elapsed_time": "18:19:44", "remaining_time": "1 day, 21:23:09", "throughput": 5665.07, "total_tokens": 373805712}
+{"current_steps": 8550, "total_steps": 29704, "loss": 0.3910914659500122, "lr": 8.091689188143267e-06, "epoch": 0.5756800430918395, "percentage": 28.78, "elapsed_time": "18:20:30", "remaining_time": "1 day, 21:22:49", "throughput": 5664.34, "total_tokens": 374017824}
+{"current_steps": 8555, "total_steps": 29704, "loss": 0.3839992046356201, "lr": 8.089610738966754e-06, "epoch": 0.5760166980877997, "percentage": 28.8, "elapsed_time": "18:21:07", "remaining_time": "1 day, 21:22:06", "throughput": 5664.25, "total_tokens": 374221088}
+{"current_steps": 8560, "total_steps": 29704, "loss": 0.41312179565429685, "lr": 8.08753142579045e-06, "epoch": 0.5763533530837598, "percentage": 28.82, "elapsed_time": "18:21:43", "remaining_time": "1 day, 21:21:20", "throughput": 5664.47, "total_tokens": 374438784}
+{"current_steps": 8565, "total_steps": 29704, "loss": 0.369646692276001, "lr": 8.085451249195828e-06, "epoch": 0.5766900080797199, "percentage": 28.83, "elapsed_time": "18:22:24", "remaining_time": "1 day, 21:20:49", "throughput": 5664.48, "total_tokens": 374674400}
+{"current_steps": 8570, "total_steps": 29704, "loss": 0.38304858207702636, "lr": 8.083370209764603e-06, "epoch": 0.5770266630756801, "percentage": 28.85, "elapsed_time": "18:23:07", "remaining_time": "1 day, 21:20:21", "throughput": 5664.14, "total_tokens": 374895048}
+{"current_steps": 8575, "total_steps": 29704, "loss": 0.39395363330841066, "lr": 8.081288308078731e-06, "epoch": 0.5773633180716402, "percentage": 28.87, "elapsed_time": "18:23:40", "remaining_time": "1 day, 21:19:29", "throughput": 5664.86, "total_tokens": 375130432}
+{"current_steps": 8580, "total_steps": 29704, "loss": 0.35710582733154295, "lr": 8.07920554472041e-06, "epoch": 0.5776999730676003, "percentage": 28.88, "elapsed_time": "18:24:26", "remaining_time": "1 day, 21:19:09", "throughput": 5664.31, "total_tokens": 375356416}
+{"current_steps": 8585, "total_steps": 29704, "loss": 0.3807799816131592, "lr": 8.077121920272075e-06, "epoch": 0.5780366280635605, "percentage": 28.9, "elapsed_time": "18:25:03", "remaining_time": "1 day, 21:18:25", "throughput": 5664.67, "total_tokens": 375586768}
+{"current_steps": 8590, "total_steps": 29704, "loss": 0.3493949890136719, "lr": 8.07503743531641e-06, "epoch": 0.5783732830595206, "percentage": 28.92, "elapsed_time": "18:25:47", "remaining_time": "1 day, 21:18:00", "throughput": 5664.05, "total_tokens": 375796816}
+{"current_steps": 8595, "total_steps": 29704, "loss": 0.34188032150268555, "lr": 8.072952090436328e-06, "epoch": 0.5787099380554808, "percentage": 28.94, "elapsed_time": "18:26:19", "remaining_time": "1 day, 21:17:04", "throughput": 5664.83, "total_tokens": 376025856}
+{"current_steps": 8600, "total_steps": 29704, "loss": 0.378603458404541, "lr": 8.070865886214996e-06, "epoch": 0.5790465930514409, "percentage": 28.95, "elapsed_time": "18:26:52", "remaining_time": "1 day, 21:16:14", "throughput": 5665.28, "total_tokens": 376248416}
+{"current_steps": 8605, "total_steps": 29704, "loss": 0.3776103496551514, "lr": 8.06877882323581e-06, "epoch": 0.579383248047401, "percentage": 28.97, "elapsed_time": "18:27:25", "remaining_time": "1 day, 21:15:19", "throughput": 5665.87, "total_tokens": 376468952}
+{"current_steps": 8610, "total_steps": 29704, "loss": 0.35461578369140623, "lr": 8.066690902082413e-06, "epoch": 0.5797199030433612, "percentage": 28.99, "elapsed_time": "18:28:02", "remaining_time": "1 day, 21:14:37", "throughput": 5666.03, "total_tokens": 376690720}
+{"current_steps": 8615, "total_steps": 29704, "loss": 0.4189821720123291, "lr": 8.064602123338683e-06, "epoch": 0.5800565580393213, "percentage": 29.0, "elapsed_time": "18:28:43", "remaining_time": "1 day, 21:14:05", "throughput": 5666.06, "total_tokens": 376926952}
+{"current_steps": 8620, "total_steps": 29704, "loss": 0.3706305980682373, "lr": 8.06251248758874e-06, "epoch": 0.5803932130352815, "percentage": 29.02, "elapsed_time": "18:29:20", "remaining_time": "1 day, 21:13:24", "throughput": 5666.05, "total_tokens": 377137424}
+{"current_steps": 8625, "total_steps": 29704, "loss": 0.4162487030029297, "lr": 8.060421995416949e-06, "epoch": 0.5807298680312416, "percentage": 29.04, "elapsed_time": "18:30:01", "remaining_time": "1 day, 21:12:51", "throughput": 5665.63, "total_tokens": 377341792}
+{"current_steps": 8630, "total_steps": 29704, "loss": 0.3742199420928955, "lr": 8.058330647407904e-06, "epoch": 0.5810665230272017, "percentage": 29.05, "elapsed_time": "18:30:40", "remaining_time": "1 day, 21:12:13", "throughput": 5665.51, "total_tokens": 377553480}
+{"current_steps": 8635, "total_steps": 29704, "loss": 0.4032620429992676, "lr": 8.056238444146445e-06, "epoch": 0.5814031780231619, "percentage": 29.07, "elapsed_time": "18:31:16", "remaining_time": "1 day, 21:11:27", "throughput": 5665.69, "total_tokens": 377768576}
+{"current_steps": 8640, "total_steps": 29704, "loss": 0.35233349800109864, "lr": 8.054145386217652e-06, "epoch": 0.581739833019122, "percentage": 29.09, "elapsed_time": "18:31:49", "remaining_time": "1 day, 21:10:34", "throughput": 5666.24, "total_tokens": 377990912}
+{"current_steps": 8645, "total_steps": 29704, "loss": 0.36639404296875, "lr": 8.05205147420684e-06, "epoch": 0.5820764880150822, "percentage": 29.1, "elapsed_time": "18:32:22", "remaining_time": "1 day, 21:09:42", "throughput": 5666.65, "total_tokens": 378205832}
+{"current_steps": 8650, "total_steps": 29704, "loss": 0.41215500831604, "lr": 8.049956708699567e-06, "epoch": 0.5824131430110423, "percentage": 29.12, "elapsed_time": "18:32:59", "remaining_time": "1 day, 21:09:00", "throughput": 5666.76, "total_tokens": 378422632}
+{"current_steps": 8655, "total_steps": 29704, "loss": 0.4017993450164795, "lr": 8.047861090281623e-06, "epoch": 0.5827497980070024, "percentage": 29.14, "elapsed_time": "18:33:37", "remaining_time": "1 day, 21:08:19", "throughput": 5666.76, "total_tokens": 378636312}
+{"current_steps": 8660, "total_steps": 29704, "loss": 0.36281633377075195, "lr": 8.045764619539047e-06, "epoch": 0.5830864530029626, "percentage": 29.15, "elapsed_time": "18:34:19", "remaining_time": "1 day, 21:07:50", "throughput": 5666.54, "total_tokens": 378862480}
+{"current_steps": 8665, "total_steps": 29704, "loss": 0.3568310022354126, "lr": 8.043667297058106e-06, "epoch": 0.5834231079989227, "percentage": 29.17, "elapsed_time": "18:34:54", "remaining_time": "1 day, 21:07:03", "throughput": 5666.72, "total_tokens": 379073648}
+{"current_steps": 8670, "total_steps": 29704, "loss": 0.3924489259719849, "lr": 8.04156912342531e-06, "epoch": 0.5837597629948829, "percentage": 29.19, "elapsed_time": "18:35:30", "remaining_time": "1 day, 21:06:18", "throughput": 5667.0, "total_tokens": 379296048}
+{"current_steps": 8675, "total_steps": 29704, "loss": 0.3730125904083252, "lr": 8.03947009922741e-06, "epoch": 0.584096417990843, "percentage": 29.2, "elapsed_time": "18:36:11", "remaining_time": "1 day, 21:05:44", "throughput": 5666.84, "total_tokens": 379514216}
+{"current_steps": 8680, "total_steps": 29704, "loss": 0.4212483406066895, "lr": 8.037370225051387e-06, "epoch": 0.5844330729868031, "percentage": 29.22, "elapsed_time": "18:36:51", "remaining_time": "1 day, 21:05:10", "throughput": 5666.58, "total_tokens": 379726496}
+{"current_steps": 8685, "total_steps": 29704, "loss": 0.3820772886276245, "lr": 8.035269501484467e-06, "epoch": 0.5847697279827633, "percentage": 29.24, "elapsed_time": "18:37:25", "remaining_time": "1 day, 21:04:21", "throughput": 5667.05, "total_tokens": 379952664}
+{"current_steps": 8690, "total_steps": 29704, "loss": 0.41135897636413576, "lr": 8.03316792911411e-06, "epoch": 0.5851063829787234, "percentage": 29.26, "elapsed_time": "18:38:06", "remaining_time": "1 day, 21:03:46", "throughput": 5666.86, "total_tokens": 380168720}
+{"current_steps": 8695, "total_steps": 29704, "loss": 0.35835254192352295, "lr": 8.031065508528011e-06, "epoch": 0.5854430379746836, "percentage": 29.27, "elapsed_time": "18:38:44", "remaining_time": "1 day, 21:03:07", "throughput": 5667.02, "total_tokens": 380396160}
+{"current_steps": 8700, "total_steps": 29704, "loss": 0.3686984539031982, "lr": 8.028962240314111e-06, "epoch": 0.5857796929706437, "percentage": 29.29, "elapsed_time": "18:39:21", "remaining_time": "1 day, 21:02:25", "throughput": 5667.23, "total_tokens": 380620472}
+{"current_steps": 8705, "total_steps": 29704, "loss": 0.3879531383514404, "lr": 8.026858125060577e-06, "epoch": 0.5861163479666038, "percentage": 29.31, "elapsed_time": "18:40:01", "remaining_time": "1 day, 21:01:48", "throughput": 5667.14, "total_tokens": 380837592}
+{"current_steps": 8710, "total_steps": 29704, "loss": 0.4247321128845215, "lr": 8.024753163355821e-06, "epoch": 0.586453002962564, "percentage": 29.32, "elapsed_time": "18:40:42", "remaining_time": "1 day, 21:01:17", "throughput": 5666.94, "total_tokens": 381061216}
+{"current_steps": 8715, "total_steps": 29704, "loss": 0.3679481029510498, "lr": 8.022647355788489e-06, "epoch": 0.5867896579585241, "percentage": 29.34, "elapsed_time": "18:41:21", "remaining_time": "1 day, 21:00:38", "throughput": 5666.95, "total_tokens": 381280288}
+{"current_steps": 8720, "total_steps": 29704, "loss": 0.40247626304626466, "lr": 8.02054070294746e-06, "epoch": 0.5871263129544843, "percentage": 29.36, "elapsed_time": "18:41:59", "remaining_time": "1 day, 20:59:59", "throughput": 5666.96, "total_tokens": 381496992}
+{"current_steps": 8725, "total_steps": 29704, "loss": 0.36869523525238035, "lr": 8.018433205421859e-06, "epoch": 0.5874629679504444, "percentage": 29.37, "elapsed_time": "18:42:39", "remaining_time": "1 day, 20:59:24", "throughput": 5666.98, "total_tokens": 381727008}
+{"current_steps": 8730, "total_steps": 29704, "loss": 0.3896385669708252, "lr": 8.016324863801036e-06, "epoch": 0.5877996229464045, "percentage": 29.39, "elapsed_time": "18:43:15", "remaining_time": "1 day, 20:58:38", "throughput": 5667.32, "total_tokens": 381950608}
+{"current_steps": 8735, "total_steps": 29704, "loss": 0.397304368019104, "lr": 8.014215678674581e-06, "epoch": 0.5881362779423647, "percentage": 29.41, "elapsed_time": "18:43:56", "remaining_time": "1 day, 20:58:06", "throughput": 5667.15, "total_tokens": 382173664}
+{"current_steps": 8740, "total_steps": 29704, "loss": 0.37210235595703123, "lr": 8.012105650632325e-06, "epoch": 0.5884729329383248, "percentage": 29.42, "elapsed_time": "18:44:36", "remaining_time": "1 day, 20:57:31", "throughput": 5666.97, "total_tokens": 382388568}
+{"current_steps": 8745, "total_steps": 29704, "loss": 0.38750429153442384, "lr": 8.009994780264328e-06, "epoch": 0.588809587934285, "percentage": 29.44, "elapsed_time": "18:45:17", "remaining_time": "1 day, 20:56:58", "throughput": 5666.43, "total_tokens": 382584944}
+{"current_steps": 8750, "total_steps": 29704, "loss": 0.38128883838653566, "lr": 8.00788306816089e-06, "epoch": 0.5891462429302451, "percentage": 29.46, "elapsed_time": "18:45:54", "remaining_time": "1 day, 20:56:15", "throughput": 5666.56, "total_tokens": 382801808}
+{"current_steps": 8755, "total_steps": 29704, "loss": 0.39379191398620605, "lr": 8.005770514912543e-06, "epoch": 0.5894828979262052, "percentage": 29.47, "elapsed_time": "18:46:33", "remaining_time": "1 day, 20:55:36", "throughput": 5666.65, "total_tokens": 383025832}
+{"current_steps": 8760, "total_steps": 29704, "loss": 0.3845071792602539, "lr": 8.003657121110055e-06, "epoch": 0.5898195529221654, "percentage": 29.49, "elapsed_time": "18:47:08", "remaining_time": "1 day, 20:54:50", "throughput": 5667.16, "total_tokens": 383260728}
+{"current_steps": 8765, "total_steps": 29704, "loss": 0.36870722770690917, "lr": 8.001542887344432e-06, "epoch": 0.5901562079181255, "percentage": 29.51, "elapsed_time": "18:47:43", "remaining_time": "1 day, 20:54:03", "throughput": 5667.48, "total_tokens": 383481016}
+{"current_steps": 8770, "total_steps": 29704, "loss": 0.38972065448760984, "lr": 7.999427814206911e-06, "epoch": 0.5904928629140856, "percentage": 29.52, "elapsed_time": "18:48:18", "remaining_time": "1 day, 20:53:15", "throughput": 5667.81, "total_tokens": 383699928}
+{"current_steps": 8775, "total_steps": 29704, "loss": 0.3479362964630127, "lr": 7.997311902288968e-06, "epoch": 0.5908295179100458, "percentage": 29.54, "elapsed_time": "18:48:55", "remaining_time": "1 day, 20:52:34", "throughput": 5667.97, "total_tokens": 383923208}
+{"current_steps": 8780, "total_steps": 29704, "loss": 0.35241219997406004, "lr": 7.995195152182307e-06, "epoch": 0.5911661729060059, "percentage": 29.56, "elapsed_time": "18:49:35", "remaining_time": "1 day, 20:51:57", "throughput": 5667.64, "total_tokens": 384124912}
+{"current_steps": 8785, "total_steps": 29704, "loss": 0.367362380027771, "lr": 7.993077564478875e-06, "epoch": 0.5915028279019661, "percentage": 29.58, "elapsed_time": "18:50:13", "remaining_time": "1 day, 20:51:19", "throughput": 5667.76, "total_tokens": 384352936}
+{"current_steps": 8790, "total_steps": 29704, "loss": 0.3980868339538574, "lr": 7.990959139770844e-06, "epoch": 0.5918394828979262, "percentage": 29.59, "elapsed_time": "18:50:55", "remaining_time": "1 day, 20:50:48", "throughput": 5667.73, "total_tokens": 384586952}
+{"current_steps": 8795, "total_steps": 29704, "loss": 0.3829834461212158, "lr": 7.988839878650628e-06, "epoch": 0.5921761378938863, "percentage": 29.61, "elapsed_time": "18:51:39", "remaining_time": "1 day, 20:50:22", "throughput": 5667.41, "total_tokens": 384813912}
+{"current_steps": 8800, "total_steps": 29704, "loss": 0.3936153411865234, "lr": 7.986719781710872e-06, "epoch": 0.5925127928898465, "percentage": 29.63, "elapsed_time": "18:52:21", "remaining_time": "1 day, 20:49:51", "throughput": 5667.26, "total_tokens": 385040600}
+{"current_steps": 8805, "total_steps": 29704, "loss": 0.37601919174194337, "lr": 7.984598849544452e-06, "epoch": 0.5928494478858066, "percentage": 29.64, "elapsed_time": "18:53:00", "remaining_time": "1 day, 20:49:13", "throughput": 5667.19, "total_tokens": 385256736}
+{"current_steps": 8810, "total_steps": 29704, "loss": 0.3897037744522095, "lr": 7.98247708274448e-06, "epoch": 0.5931861028817668, "percentage": 29.66, "elapsed_time": "18:53:35", "remaining_time": "1 day, 20:48:26", "throughput": 5667.42, "total_tokens": 385471192}
+{"current_steps": 8815, "total_steps": 29704, "loss": 0.4111155033111572, "lr": 7.980354481904303e-06, "epoch": 0.5935227578777269, "percentage": 29.68, "elapsed_time": "18:54:13", "remaining_time": "1 day, 20:47:47", "throughput": 5667.52, "total_tokens": 385695968}
+{"current_steps": 8820, "total_steps": 29704, "loss": 0.41021156311035156, "lr": 7.978231047617498e-06, "epoch": 0.593859412873687, "percentage": 29.69, "elapsed_time": "18:54:51", "remaining_time": "1 day, 20:47:07", "throughput": 5667.5, "total_tokens": 385910576}
+{"current_steps": 8825, "total_steps": 29704, "loss": 0.4051943778991699, "lr": 7.976106780477878e-06, "epoch": 0.5941960678696472, "percentage": 29.71, "elapsed_time": "18:55:34", "remaining_time": "1 day, 20:46:39", "throughput": 5667.23, "total_tokens": 386135568}
+{"current_steps": 8830, "total_steps": 29704, "loss": 0.3718585729598999, "lr": 7.973981681079487e-06, "epoch": 0.5945327228656073, "percentage": 29.73, "elapsed_time": "18:56:15", "remaining_time": "1 day, 20:46:07", "throughput": 5666.88, "total_tokens": 386344944}
+{"current_steps": 8835, "total_steps": 29704, "loss": 0.37234504222869874, "lr": 7.971855750016603e-06, "epoch": 0.5948693778615675, "percentage": 29.74, "elapsed_time": "18:56:50", "remaining_time": "1 day, 20:45:18", "throughput": 5667.18, "total_tokens": 386561056}
+{"current_steps": 8840, "total_steps": 29704, "loss": 0.386487340927124, "lr": 7.969728987883734e-06, "epoch": 0.5952060328575276, "percentage": 29.76, "elapsed_time": "18:57:30", "remaining_time": "1 day, 20:44:42", "throughput": 5667.13, "total_tokens": 386782376}
+{"current_steps": 8845, "total_steps": 29704, "loss": 0.38187594413757325, "lr": 7.967601395275624e-06, "epoch": 0.5955426878534877, "percentage": 29.78, "elapsed_time": "18:58:11", "remaining_time": "1 day, 20:44:10", "throughput": 5666.89, "total_tokens": 386999696}
+{"current_steps": 8850, "total_steps": 29704, "loss": 0.35875325202941893, "lr": 7.965472972787247e-06, "epoch": 0.5958793428494479, "percentage": 29.79, "elapsed_time": "18:58:48", "remaining_time": "1 day, 20:43:27", "throughput": 5667.08, "total_tokens": 387220408}
+{"current_steps": 8855, "total_steps": 29704, "loss": 0.3718303680419922, "lr": 7.963343721013808e-06, "epoch": 0.596215997845408, "percentage": 29.81, "elapsed_time": "18:59:26", "remaining_time": "1 day, 20:42:49", "throughput": 5667.3, "total_tokens": 387455840}
+{"current_steps": 8860, "total_steps": 29704, "loss": 0.3902780771255493, "lr": 7.961213640550748e-06, "epoch": 0.5965526528413682, "percentage": 29.83, "elapsed_time": "19:00:10", "remaining_time": "1 day, 20:42:21", "throughput": 5666.96, "total_tokens": 387678720}
+{"current_steps": 8865, "total_steps": 29704, "loss": 0.3638461589813232, "lr": 7.959082731993736e-06, "epoch": 0.5968893078373283, "percentage": 29.84, "elapsed_time": "19:00:43", "remaining_time": "1 day, 20:41:29", "throughput": 5667.45, "total_tokens": 387897464}
+{"current_steps": 8870, "total_steps": 29704, "loss": 0.3705005168914795, "lr": 7.956950995938673e-06, "epoch": 0.5972259628332884, "percentage": 29.86, "elapsed_time": "19:01:26", "remaining_time": "1 day, 20:41:02", "throughput": 5666.93, "total_tokens": 388107624}
+{"current_steps": 8875, "total_steps": 29704, "loss": 0.3910086631774902, "lr": 7.954818432981694e-06, "epoch": 0.5975626178292486, "percentage": 29.88, "elapsed_time": "19:02:05", "remaining_time": "1 day, 20:40:24", "throughput": 5666.66, "total_tokens": 388310680}
+{"current_steps": 8880, "total_steps": 29704, "loss": 0.3656719923019409, "lr": 7.952685043719162e-06, "epoch": 0.5978992728252087, "percentage": 29.89, "elapsed_time": "19:02:48", "remaining_time": "1 day, 20:39:56", "throughput": 5666.35, "total_tokens": 388533832}
+{"current_steps": 8885, "total_steps": 29704, "loss": 0.37486002445220945, "lr": 7.950550828747672e-06, "epoch": 0.5982359278211689, "percentage": 29.91, "elapsed_time": "19:03:28", "remaining_time": "1 day, 20:39:21", "throughput": 5666.24, "total_tokens": 388754728}
+{"current_steps": 8890, "total_steps": 29704, "loss": 0.40519165992736816, "lr": 7.948415788664052e-06, "epoch": 0.598572582817129, "percentage": 29.93, "elapsed_time": "19:04:01", "remaining_time": "1 day, 20:38:30", "throughput": 5666.75, "total_tokens": 388977056}
+{"current_steps": 8895, "total_steps": 29704, "loss": 0.3696116924285889, "lr": 7.946279924065359e-06, "epoch": 0.5989092378130891, "percentage": 29.95, "elapsed_time": "19:04:36", "remaining_time": "1 day, 20:37:41", "throughput": 5667.11, "total_tokens": 389195368}
+{"current_steps": 8900, "total_steps": 29704, "loss": 0.3721637725830078, "lr": 7.94414323554888e-06, "epoch": 0.5992458928090493, "percentage": 29.96, "elapsed_time": "19:05:07", "remaining_time": "1 day, 20:36:46", "throughput": 5667.82, "total_tokens": 389424256}
+{"current_steps": 8905, "total_steps": 29704, "loss": 0.3689994812011719, "lr": 7.942005723712135e-06, "epoch": 0.5995825478050094, "percentage": 29.98, "elapsed_time": "19:05:42", "remaining_time": "1 day, 20:35:59", "throughput": 5668.37, "total_tokens": 389659456}
+{"current_steps": 8910, "total_steps": 29704, "loss": 0.36181044578552246, "lr": 7.939867389152868e-06, "epoch": 0.5999192028009696, "percentage": 30.0, "elapsed_time": "19:06:22", "remaining_time": "1 day, 20:35:24", "throughput": 5668.24, "total_tokens": 389878008}
+{"current_steps": 8915, "total_steps": 29704, "loss": 0.3904815435409546, "lr": 7.937728232469062e-06, "epoch": 0.6002558577969297, "percentage": 30.01, "elapsed_time": "19:07:03", "remaining_time": "1 day, 20:34:50", "throughput": 5667.98, "total_tokens": 390090680}
+{"current_steps": 8920, "total_steps": 29704, "loss": 0.407269811630249, "lr": 7.935588254258921e-06, "epoch": 0.6005925127928898, "percentage": 30.03, "elapsed_time": "19:07:39", "remaining_time": "1 day, 20:34:06", "throughput": 5668.02, "total_tokens": 390298200}
+{"current_steps": 8925, "total_steps": 29704, "loss": 0.3707296848297119, "lr": 7.933447455120889e-06, "epoch": 0.60092916778885, "percentage": 30.05, "elapsed_time": "19:08:26", "remaining_time": "1 day, 20:33:47", "throughput": 5667.26, "total_tokens": 390513112}
+{"current_steps": 8930, "total_steps": 29704, "loss": 0.34468765258789064, "lr": 7.931305835653629e-06, "epoch": 0.6012658227848101, "percentage": 30.06, "elapsed_time": "19:09:05", "remaining_time": "1 day, 20:33:09", "throughput": 5667.4, "total_tokens": 390741944}
+{"current_steps": 8935, "total_steps": 29704, "loss": 0.39664325714111326, "lr": 7.929163396456038e-06, "epoch": 0.6016024777807703, "percentage": 30.08, "elapsed_time": "19:09:42", "remaining_time": "1 day, 20:32:26", "throughput": 5667.4, "total_tokens": 390950400}
+{"current_steps": 8940, "total_steps": 29704, "loss": 0.369353199005127, "lr": 7.927020138127245e-06, "epoch": 0.6019391327767304, "percentage": 30.1, "elapsed_time": "19:10:22", "remaining_time": "1 day, 20:31:50", "throughput": 5667.35, "total_tokens": 391172832}
+{"current_steps": 8945, "total_steps": 29704, "loss": 0.3477231502532959, "lr": 7.924876061266603e-06, "epoch": 0.6022757877726905, "percentage": 30.11, "elapsed_time": "19:11:00", "remaining_time": "1 day, 20:31:11", "throughput": 5667.52, "total_tokens": 391403392}
+{"current_steps": 8950, "total_steps": 29704, "loss": 0.40044050216674804, "lr": 7.922731166473698e-06, "epoch": 0.6026124427686507, "percentage": 30.13, "elapsed_time": "19:11:35", "remaining_time": "1 day, 20:30:24", "throughput": 5667.81, "total_tokens": 391619536}
+{"current_steps": 8955, "total_steps": 29704, "loss": 0.35779447555541993, "lr": 7.920585454348341e-06, "epoch": 0.6029490977646108, "percentage": 30.15, "elapsed_time": "19:12:10", "remaining_time": "1 day, 20:29:38", "throughput": 5667.98, "total_tokens": 391832424}
+{"current_steps": 8960, "total_steps": 29704, "loss": 0.35099799633026124, "lr": 7.918438925490578e-06, "epoch": 0.603285752760571, "percentage": 30.16, "elapsed_time": "19:12:48", "remaining_time": "1 day, 20:28:57", "throughput": 5668.15, "total_tokens": 392057736}
+{"current_steps": 8965, "total_steps": 29704, "loss": 0.37104010581970215, "lr": 7.916291580500672e-06, "epoch": 0.6036224077565311, "percentage": 30.18, "elapsed_time": "19:13:29", "remaining_time": "1 day, 20:28:23", "throughput": 5667.96, "total_tokens": 392274152}
+{"current_steps": 8970, "total_steps": 29704, "loss": 0.36745285987854004, "lr": 7.914143419979127e-06, "epoch": 0.6039590627524912, "percentage": 30.2, "elapsed_time": "19:14:07", "remaining_time": "1 day, 20:27:43", "throughput": 5668.09, "total_tokens": 392499120}
+{"current_steps": 8975, "total_steps": 29704, "loss": 0.34250531196594236, "lr": 7.911994444526667e-06, "epoch": 0.6042957177484514, "percentage": 30.21, "elapsed_time": "19:14:41", "remaining_time": "1 day, 20:26:55", "throughput": 5668.46, "total_tokens": 392720016}
+{"current_steps": 8980, "total_steps": 29704, "loss": 0.3526078939437866, "lr": 7.909844654744246e-06, "epoch": 0.6046323727444115, "percentage": 30.23, "elapsed_time": "19:15:18", "remaining_time": "1 day, 20:26:12", "throughput": 5668.69, "total_tokens": 392945160}
+{"current_steps": 8985, "total_steps": 29704, "loss": 0.38707795143127444, "lr": 7.907694051233044e-06, "epoch": 0.6049690277403716, "percentage": 30.25, "elapsed_time": "19:15:55", "remaining_time": "1 day, 20:25:31", "throughput": 5668.8, "total_tokens": 393164952}
+{"current_steps": 8990, "total_steps": 29704, "loss": 0.3759605884552002, "lr": 7.905542634594476e-06, "epoch": 0.6053056827363318, "percentage": 30.27, "elapsed_time": "19:16:43", "remaining_time": "1 day, 20:25:12", "throughput": 5668.23, "total_tokens": 393392000}
+{"current_steps": 8995, "total_steps": 29704, "loss": 0.36977381706237794, "lr": 7.903390405430174e-06, "epoch": 0.6056423377322919, "percentage": 30.28, "elapsed_time": "19:17:23", "remaining_time": "1 day, 20:24:38", "throughput": 5668.06, "total_tokens": 393611632}
+{"current_steps": 9000, "total_steps": 29704, "loss": 0.3545074939727783, "lr": 7.901237364342003e-06, "epoch": 0.6059789927282521, "percentage": 30.3, "elapsed_time": "19:17:58", "remaining_time": "1 day, 20:23:50", "throughput": 5668.15, "total_tokens": 393813184}
+{"current_steps": 9005, "total_steps": 29704, "loss": 0.3470455646514893, "lr": 7.899083511932053e-06, "epoch": 0.6063156477242122, "percentage": 30.32, "elapsed_time": "19:18:38", "remaining_time": "1 day, 20:23:16", "throughput": 5667.93, "total_tokens": 394027272}
+{"current_steps": 9010, "total_steps": 29704, "loss": 0.372318959236145, "lr": 7.896928848802645e-06, "epoch": 0.6066523027201723, "percentage": 30.33, "elapsed_time": "19:19:13", "remaining_time": "1 day, 20:22:29", "throughput": 5668.32, "total_tokens": 394252480}
+{"current_steps": 9015, "total_steps": 29704, "loss": 0.39665966033935546, "lr": 7.894773375556322e-06, "epoch": 0.6069889577161325, "percentage": 30.35, "elapsed_time": "19:19:53", "remaining_time": "1 day, 20:21:54", "throughput": 5667.88, "total_tokens": 394449024}
+{"current_steps": 9020, "total_steps": 29704, "loss": 0.38990225791931155, "lr": 7.892617092795855e-06, "epoch": 0.6073256127120926, "percentage": 30.37, "elapsed_time": "19:20:32", "remaining_time": "1 day, 20:21:16", "throughput": 5667.93, "total_tokens": 394672680}
+{"current_steps": 9025, "total_steps": 29704, "loss": 0.3693479061126709, "lr": 7.890460001124242e-06, "epoch": 0.6076622677080528, "percentage": 30.38, "elapsed_time": "19:21:08", "remaining_time": "1 day, 20:20:30", "throughput": 5668.05, "total_tokens": 394882720}
+{"current_steps": 9030, "total_steps": 29704, "loss": 0.3727553367614746, "lr": 7.888302101144707e-06, "epoch": 0.6079989227040129, "percentage": 30.4, "elapsed_time": "19:21:50", "remaining_time": "1 day, 20:20:00", "throughput": 5667.67, "total_tokens": 395096336}
+{"current_steps": 9035, "total_steps": 29704, "loss": 0.38645596504211427, "lr": 7.886143393460699e-06, "epoch": 0.608335577699973, "percentage": 30.42, "elapsed_time": "19:22:26", "remaining_time": "1 day, 20:19:16", "throughput": 5668.04, "total_tokens": 395327104}
+{"current_steps": 9040, "total_steps": 29704, "loss": 0.3641960620880127, "lr": 7.883983878675895e-06, "epoch": 0.6086722326959332, "percentage": 30.43, "elapsed_time": "19:23:14", "remaining_time": "1 day, 20:18:58", "throughput": 5667.52, "total_tokens": 395561000}
+{"current_steps": 9045, "total_steps": 29704, "loss": 0.3701021194458008, "lr": 7.881823557394194e-06, "epoch": 0.6090088876918933, "percentage": 30.45, "elapsed_time": "19:23:52", "remaining_time": "1 day, 20:18:19", "throughput": 5667.69, "total_tokens": 395790544}
+{"current_steps": 9050, "total_steps": 29704, "loss": 0.35156404972076416, "lr": 7.879662430219728e-06, "epoch": 0.6093455426878535, "percentage": 30.47, "elapsed_time": "19:24:33", "remaining_time": "1 day, 20:17:45", "throughput": 5667.23, "total_tokens": 395987680}
+{"current_steps": 9055, "total_steps": 29704, "loss": 0.4006368637084961, "lr": 7.877500497756843e-06, "epoch": 0.6096821976838136, "percentage": 30.48, "elapsed_time": "19:25:14", "remaining_time": "1 day, 20:17:13", "throughput": 5667.0, "total_tokens": 396207224}
+{"current_steps": 9060, "total_steps": 29704, "loss": 0.3919075012207031, "lr": 7.87533776061012e-06, "epoch": 0.6100188526797737, "percentage": 30.5, "elapsed_time": "19:25:54", "remaining_time": "1 day, 20:16:38", "throughput": 5666.96, "total_tokens": 396431144}
+{"current_steps": 9065, "total_steps": 29704, "loss": 0.3918348789215088, "lr": 7.873174219384362e-06, "epoch": 0.6103555076757339, "percentage": 30.52, "elapsed_time": "19:26:36", "remaining_time": "1 day, 20:16:06", "throughput": 5666.69, "total_tokens": 396647760}
+{"current_steps": 9070, "total_steps": 29704, "loss": 0.3941289186477661, "lr": 7.871009874684595e-06, "epoch": 0.610692162671694, "percentage": 30.53, "elapsed_time": "19:27:08", "remaining_time": "1 day, 20:15:13", "throughput": 5667.39, "total_tokens": 396880640}
+{"current_steps": 9075, "total_steps": 29704, "loss": 0.3926485300064087, "lr": 7.868844727116072e-06, "epoch": 0.6110288176676542, "percentage": 30.55, "elapsed_time": "19:27:48", "remaining_time": "1 day, 20:14:37", "throughput": 5667.38, "total_tokens": 397103976}
+{"current_steps": 9080, "total_steps": 29704, "loss": 0.3840303897857666, "lr": 7.866678777284267e-06, "epoch": 0.6113654726636143, "percentage": 30.57, "elapsed_time": "19:28:22", "remaining_time": "1 day, 20:13:49", "throughput": 5667.82, "total_tokens": 397331160}
+{"current_steps": 9085, "total_steps": 29704, "loss": 0.3659891843795776, "lr": 7.864512025794886e-06, "epoch": 0.6117021276595744, "percentage": 30.59, "elapsed_time": "19:29:02", "remaining_time": "1 day, 20:13:14", "throughput": 5667.8, "total_tokens": 397555992}
+{"current_steps": 9090, "total_steps": 29704, "loss": 0.36196537017822267, "lr": 7.86234447325385e-06, "epoch": 0.6120387826555346, "percentage": 30.6, "elapsed_time": "19:29:45", "remaining_time": "1 day, 20:12:44", "throughput": 5667.41, "total_tokens": 397769720}
+{"current_steps": 9095, "total_steps": 29704, "loss": 0.4144433498382568, "lr": 7.860176120267309e-06, "epoch": 0.6123754376514947, "percentage": 30.62, "elapsed_time": "19:30:27", "remaining_time": "1 day, 20:12:12", "throughput": 5667.42, "total_tokens": 398006352}
+{"current_steps": 9100, "total_steps": 29704, "loss": 0.3602044105529785, "lr": 7.858006967441637e-06, "epoch": 0.6127120926474549, "percentage": 30.64, "elapsed_time": "19:31:08", "remaining_time": "1 day, 20:11:39", "throughput": 5667.09, "total_tokens": 398216992}
+{"current_steps": 9105, "total_steps": 29704, "loss": 0.40687074661254885, "lr": 7.85583701538343e-06, "epoch": 0.613048747643415, "percentage": 30.65, "elapsed_time": "19:31:53", "remaining_time": "1 day, 20:11:16", "throughput": 5666.71, "total_tokens": 398447240}
+{"current_steps": 9110, "total_steps": 29704, "loss": 0.3638491153717041, "lr": 7.853666264699506e-06, "epoch": 0.6133854026393751, "percentage": 30.67, "elapsed_time": "19:32:35", "remaining_time": "1 day, 20:10:45", "throughput": 5666.29, "total_tokens": 398656928}
+{"current_steps": 9115, "total_steps": 29704, "loss": 0.41129069328308104, "lr": 7.851494715996914e-06, "epoch": 0.6137220576353353, "percentage": 30.69, "elapsed_time": "19:33:15", "remaining_time": "1 day, 20:10:09", "throughput": 5666.13, "total_tokens": 398870368}
+{"current_steps": 9120, "total_steps": 29704, "loss": 0.3685478687286377, "lr": 7.849322369882915e-06, "epoch": 0.6140587126312954, "percentage": 30.7, "elapsed_time": "19:34:04", "remaining_time": "1 day, 20:09:55", "throughput": 5665.26, "total_tokens": 399088536}
+{"current_steps": 9125, "total_steps": 29704, "loss": 0.34872984886169434, "lr": 7.847149226965003e-06, "epoch": 0.6143953676272556, "percentage": 30.72, "elapsed_time": "19:34:45", "remaining_time": "1 day, 20:09:20", "throughput": 5665.05, "total_tokens": 399301232}
+{"current_steps": 9130, "total_steps": 29704, "loss": 0.35025968551635744, "lr": 7.844975287850886e-06, "epoch": 0.6147320226232157, "percentage": 30.74, "elapsed_time": "19:35:22", "remaining_time": "1 day, 20:08:38", "throughput": 5665.19, "total_tokens": 399523504}
+{"current_steps": 9135, "total_steps": 29704, "loss": 0.3998849630355835, "lr": 7.842800553148506e-06, "epoch": 0.6150686776191758, "percentage": 30.75, "elapsed_time": "19:35:58", "remaining_time": "1 day, 20:07:54", "throughput": 5665.28, "total_tokens": 399735256}
+{"current_steps": 9140, "total_steps": 29704, "loss": 0.3812243938446045, "lr": 7.840625023466014e-06, "epoch": 0.615405332615136, "percentage": 30.77, "elapsed_time": "19:36:38", "remaining_time": "1 day, 20:07:18", "throughput": 5665.28, "total_tokens": 399958824}
+{"current_steps": 9145, "total_steps": 29704, "loss": 0.3413673400878906, "lr": 7.838448699411792e-06, "epoch": 0.6157419876110961, "percentage": 30.79, "elapsed_time": "19:37:21", "remaining_time": "1 day, 20:06:49", "throughput": 5664.98, "total_tokens": 400181528}
+{"current_steps": 9150, "total_steps": 29704, "loss": 0.4210306167602539, "lr": 7.836271581594442e-06, "epoch": 0.6160786426070562, "percentage": 30.8, "elapsed_time": "19:37:57", "remaining_time": "1 day, 20:06:05", "throughput": 5665.23, "total_tokens": 400404728}
+{"current_steps": 9155, "total_steps": 29704, "loss": 0.3944708824157715, "lr": 7.834093670622789e-06, "epoch": 0.6164152976030164, "percentage": 30.82, "elapsed_time": "19:38:38", "remaining_time": "1 day, 20:05:33", "throughput": 5664.97, "total_tokens": 400620128}
+{"current_steps": 9160, "total_steps": 29704, "loss": 0.4071969985961914, "lr": 7.831914967105878e-06, "epoch": 0.6167519525989765, "percentage": 30.84, "elapsed_time": "19:39:19", "remaining_time": "1 day, 20:04:58", "throughput": 5665.01, "total_tokens": 400851832}
+{"current_steps": 9165, "total_steps": 29704, "loss": 0.3579000234603882, "lr": 7.829735471652978e-06, "epoch": 0.6170886075949367, "percentage": 30.85, "elapsed_time": "19:39:59", "remaining_time": "1 day, 20:04:23", "throughput": 5664.89, "total_tokens": 401070744}
+{"current_steps": 9170, "total_steps": 29704, "loss": 0.38968710899353026, "lr": 7.827555184873575e-06, "epoch": 0.6174252625908968, "percentage": 30.87, "elapsed_time": "19:40:31", "remaining_time": "1 day, 20:03:30", "throughput": 5665.49, "total_tokens": 401294816}
+{"current_steps": 9175, "total_steps": 29704, "loss": 0.35830237865448, "lr": 7.825374107377381e-06, "epoch": 0.617761917586857, "percentage": 30.89, "elapsed_time": "19:41:13", "remaining_time": "1 day, 20:02:59", "throughput": 5665.14, "total_tokens": 401509200}
+{"current_steps": 9180, "total_steps": 29704, "loss": 0.3784201145172119, "lr": 7.823192239774327e-06, "epoch": 0.6180985725828171, "percentage": 30.9, "elapsed_time": "19:41:53", "remaining_time": "1 day, 20:02:23", "throughput": 5664.95, "total_tokens": 401720584}
+{"current_steps": 9185, "total_steps": 29704, "loss": 0.3932128667831421, "lr": 7.821009582674567e-06, "epoch": 0.6184352275787772, "percentage": 30.92, "elapsed_time": "19:42:30", "remaining_time": "1 day, 20:01:41", "throughput": 5665.03, "total_tokens": 401937000}
+{"current_steps": 9190, "total_steps": 29704, "loss": 0.41967573165893557, "lr": 7.818826136688473e-06, "epoch": 0.6187718825747374, "percentage": 30.94, "elapsed_time": "19:43:08", "remaining_time": "1 day, 20:01:00", "throughput": 5665.2, "total_tokens": 402163016}
+{"current_steps": 9195, "total_steps": 29704, "loss": 0.3871551513671875, "lr": 7.816641902426638e-06, "epoch": 0.6191085375706975, "percentage": 30.96, "elapsed_time": "19:43:46", "remaining_time": "1 day, 20:00:20", "throughput": 5665.38, "total_tokens": 402390608}
+{"current_steps": 9200, "total_steps": 29704, "loss": 0.39989705085754396, "lr": 7.814456880499877e-06, "epoch": 0.6194451925666576, "percentage": 30.97, "elapsed_time": "19:44:19", "remaining_time": "1 day, 19:59:29", "throughput": 5665.83, "total_tokens": 402608440}
+{"current_steps": 9205, "total_steps": 29704, "loss": 0.40456953048706057, "lr": 7.812271071519222e-06, "epoch": 0.6197818475626178, "percentage": 30.99, "elapsed_time": "19:44:52", "remaining_time": "1 day, 19:58:39", "throughput": 5666.28, "total_tokens": 402830072}
+{"current_steps": 9210, "total_steps": 29704, "loss": 0.3537304401397705, "lr": 7.810084476095932e-06, "epoch": 0.6201185025585779, "percentage": 31.01, "elapsed_time": "19:45:32", "remaining_time": "1 day, 19:58:03", "throughput": 5666.21, "total_tokens": 403051544}
+{"current_steps": 9215, "total_steps": 29704, "loss": 0.3726407527923584, "lr": 7.807897094841476e-06, "epoch": 0.6204551575545381, "percentage": 31.02, "elapsed_time": "19:46:11", "remaining_time": "1 day, 19:57:25", "throughput": 5666.03, "total_tokens": 403260152}
+{"current_steps": 9220, "total_steps": 29704, "loss": 0.3734896183013916, "lr": 7.805708928367555e-06, "epoch": 0.6207918125504982, "percentage": 31.04, "elapsed_time": "19:46:51", "remaining_time": "1 day, 19:56:49", "throughput": 5665.8, "total_tokens": 403468824}
+{"current_steps": 9225, "total_steps": 29704, "loss": 0.4021895885467529, "lr": 7.803519977286075e-06, "epoch": 0.6211284675464583, "percentage": 31.06, "elapsed_time": "19:47:29", "remaining_time": "1 day, 19:56:09", "throughput": 5665.7, "total_tokens": 403677704}
+{"current_steps": 9230, "total_steps": 29704, "loss": 0.38091113567352297, "lr": 7.801330242209176e-06, "epoch": 0.6214651225424185, "percentage": 31.07, "elapsed_time": "19:48:10", "remaining_time": "1 day, 19:55:35", "throughput": 5665.34, "total_tokens": 403883200}
+{"current_steps": 9235, "total_steps": 29704, "loss": 0.3860699415206909, "lr": 7.799139723749207e-06, "epoch": 0.6218017775383786, "percentage": 31.09, "elapsed_time": "19:48:48", "remaining_time": "1 day, 19:54:56", "throughput": 5665.24, "total_tokens": 404093128}
+{"current_steps": 9240, "total_steps": 29704, "loss": 0.3864139080047607, "lr": 7.79694842251874e-06, "epoch": 0.6221384325343388, "percentage": 31.11, "elapsed_time": "19:49:28", "remaining_time": "1 day, 19:54:20", "throughput": 5665.27, "total_tokens": 404319728}
+{"current_steps": 9245, "total_steps": 29704, "loss": 0.3734030485153198, "lr": 7.794756339130565e-06, "epoch": 0.6224750875302989, "percentage": 31.12, "elapsed_time": "19:50:11", "remaining_time": "1 day, 19:53:51", "throughput": 5664.96, "total_tokens": 404541968}
+{"current_steps": 9250, "total_steps": 29704, "loss": 0.3814718008041382, "lr": 7.792563474197692e-06, "epoch": 0.622811742526259, "percentage": 31.14, "elapsed_time": "19:50:48", "remaining_time": "1 day, 19:53:10", "throughput": 5664.97, "total_tokens": 404754312}
+{"current_steps": 9255, "total_steps": 29704, "loss": 0.4030300617218018, "lr": 7.79036982833335e-06, "epoch": 0.6231483975222192, "percentage": 31.16, "elapsed_time": "19:51:19", "remaining_time": "1 day, 19:52:13", "throughput": 5665.79, "total_tokens": 404985704}
+{"current_steps": 9260, "total_steps": 29704, "loss": 0.3688653469085693, "lr": 7.78817540215098e-06, "epoch": 0.6234850525181793, "percentage": 31.17, "elapsed_time": "19:52:03", "remaining_time": "1 day, 19:51:47", "throughput": 5665.39, "total_tokens": 405208848}
+{"current_steps": 9265, "total_steps": 29704, "loss": 0.37583611011505125, "lr": 7.78598019626425e-06, "epoch": 0.6238217075141395, "percentage": 31.19, "elapsed_time": "19:52:35", "remaining_time": "1 day, 19:50:55", "throughput": 5666.17, "total_tokens": 405447552}
+{"current_steps": 9270, "total_steps": 29704, "loss": 0.4006516456604004, "lr": 7.783784211287044e-06, "epoch": 0.6241583625100996, "percentage": 31.21, "elapsed_time": "19:53:17", "remaining_time": "1 day, 19:50:23", "throughput": 5665.71, "total_tokens": 405651112}
+{"current_steps": 9275, "total_steps": 29704, "loss": 0.408467960357666, "lr": 7.78158744783346e-06, "epoch": 0.6244950175060597, "percentage": 31.22, "elapsed_time": "19:53:56", "remaining_time": "1 day, 19:49:46", "throughput": 5665.71, "total_tokens": 405873152}
+{"current_steps": 9280, "total_steps": 29704, "loss": 0.35638930797576907, "lr": 7.779389906517813e-06, "epoch": 0.6248316725020199, "percentage": 31.24, "elapsed_time": "19:54:30", "remaining_time": "1 day, 19:48:56", "throughput": 5666.04, "total_tokens": 406086632}
+{"current_steps": 9285, "total_steps": 29704, "loss": 0.3749531745910645, "lr": 7.777191587954645e-06, "epoch": 0.6251683274979801, "percentage": 31.26, "elapsed_time": "19:55:12", "remaining_time": "1 day, 19:48:26", "throughput": 5665.64, "total_tokens": 406298704}
+{"current_steps": 9290, "total_steps": 29704, "loss": 0.36041698455810545, "lr": 7.774992492758704e-06, "epoch": 0.6255049824939403, "percentage": 31.28, "elapsed_time": "19:55:50", "remaining_time": "1 day, 19:47:45", "throughput": 5665.51, "total_tokens": 406502360}
+{"current_steps": 9295, "total_steps": 29704, "loss": 0.38366947174072263, "lr": 7.77279262154496e-06, "epoch": 0.6258416374899004, "percentage": 31.29, "elapsed_time": "19:56:29", "remaining_time": "1 day, 19:47:08", "throughput": 5665.54, "total_tokens": 406727224}
+{"current_steps": 9300, "total_steps": 29704, "loss": 0.3997042655944824, "lr": 7.770591974928604e-06, "epoch": 0.6261782924858605, "percentage": 31.31, "elapsed_time": "19:57:04", "remaining_time": "1 day, 19:46:22", "throughput": 5665.99, "total_tokens": 406958480}
+{"current_steps": 9305, "total_steps": 29704, "loss": 0.3705620527267456, "lr": 7.768390553525034e-06, "epoch": 0.6265149474818207, "percentage": 31.33, "elapsed_time": "19:57:45", "remaining_time": "1 day, 19:45:46", "throughput": 5665.99, "total_tokens": 407186440}
+{"current_steps": 9310, "total_steps": 29704, "loss": 0.3714077711105347, "lr": 7.766188357949875e-06, "epoch": 0.6268516024777808, "percentage": 31.34, "elapsed_time": "19:58:29", "remaining_time": "1 day, 19:45:20", "throughput": 5665.58, "total_tokens": 407407560}
+{"current_steps": 9315, "total_steps": 29704, "loss": 0.36980538368225097, "lr": 7.763985388818963e-06, "epoch": 0.627188257473741, "percentage": 31.36, "elapsed_time": "19:59:09", "remaining_time": "1 day, 19:44:44", "throughput": 5665.51, "total_tokens": 407628232}
+{"current_steps": 9320, "total_steps": 29704, "loss": 0.3548977613449097, "lr": 7.76178164674835e-06, "epoch": 0.6275249124697011, "percentage": 31.38, "elapsed_time": "19:59:39", "remaining_time": "1 day, 19:43:47", "throughput": 5666.07, "total_tokens": 407838184}
+{"current_steps": 9325, "total_steps": 29704, "loss": 0.3763836622238159, "lr": 7.759577132354305e-06, "epoch": 0.6278615674656612, "percentage": 31.39, "elapsed_time": "20:00:13", "remaining_time": "1 day, 19:42:58", "throughput": 5666.54, "total_tokens": 408065528}
+{"current_steps": 9330, "total_steps": 29704, "loss": 0.3501241683959961, "lr": 7.757371846253315e-06, "epoch": 0.6281982224616214, "percentage": 31.41, "elapsed_time": "20:00:52", "remaining_time": "1 day, 19:42:20", "throughput": 5666.4, "total_tokens": 408275816}
+{"current_steps": 9335, "total_steps": 29704, "loss": 0.4022618293762207, "lr": 7.755165789062083e-06, "epoch": 0.6285348774575815, "percentage": 31.43, "elapsed_time": "20:01:30", "remaining_time": "1 day, 19:41:42", "throughput": 5666.45, "total_tokens": 408498408}
+{"current_steps": 9340, "total_steps": 29704, "loss": 0.3877957105636597, "lr": 7.752958961397522e-06, "epoch": 0.6288715324535417, "percentage": 31.44, "elapsed_time": "20:02:00", "remaining_time": "1 day, 19:40:44", "throughput": 5667.37, "total_tokens": 408733352}
+{"current_steps": 9345, "total_steps": 29704, "loss": 0.39035401344299314, "lr": 7.750751363876765e-06, "epoch": 0.6292081874495018, "percentage": 31.46, "elapsed_time": "20:02:44", "remaining_time": "1 day, 19:40:18", "throughput": 5666.78, "total_tokens": 408941920}
+{"current_steps": 9350, "total_steps": 29704, "loss": 0.3737621307373047, "lr": 7.748542997117162e-06, "epoch": 0.6295448424454619, "percentage": 31.48, "elapsed_time": "20:03:16", "remaining_time": "1 day, 19:39:23", "throughput": 5667.37, "total_tokens": 409162768}
+{"current_steps": 9355, "total_steps": 29704, "loss": 0.37030327320098877, "lr": 7.746333861736275e-06, "epoch": 0.6298814974414221, "percentage": 31.49, "elapsed_time": "20:03:50", "remaining_time": "1 day, 19:38:36", "throughput": 5667.57, "total_tokens": 409372800}
+{"current_steps": 9360, "total_steps": 29704, "loss": 0.3631394386291504, "lr": 7.74412395835188e-06, "epoch": 0.6302181524373822, "percentage": 31.51, "elapsed_time": "20:04:32", "remaining_time": "1 day, 19:38:03", "throughput": 5667.23, "total_tokens": 409583128}
+{"current_steps": 9365, "total_steps": 29704, "loss": 0.37232241630554197, "lr": 7.74191328758197e-06, "epoch": 0.6305548074333424, "percentage": 31.53, "elapsed_time": "20:05:13", "remaining_time": "1 day, 19:37:31", "throughput": 5666.9, "total_tokens": 409794192}
+{"current_steps": 9370, "total_steps": 29704, "loss": 0.4075340270996094, "lr": 7.739701850044752e-06, "epoch": 0.6308914624293025, "percentage": 31.54, "elapsed_time": "20:05:50", "remaining_time": "1 day, 19:36:48", "throughput": 5667.17, "total_tokens": 410020536}
+{"current_steps": 9375, "total_steps": 29704, "loss": 0.3718372583389282, "lr": 7.73748964635865e-06, "epoch": 0.6312281174252626, "percentage": 31.56, "elapsed_time": "20:06:27", "remaining_time": "1 day, 19:36:06", "throughput": 5667.31, "total_tokens": 410241072}
+{"current_steps": 9380, "total_steps": 29704, "loss": 0.35228054523468016, "lr": 7.735276677142297e-06, "epoch": 0.6315647724212228, "percentage": 31.58, "elapsed_time": "20:07:03", "remaining_time": "1 day, 19:35:22", "throughput": 5667.28, "total_tokens": 410443760}
+{"current_steps": 9385, "total_steps": 29704, "loss": 0.3992206573486328, "lr": 7.733062943014542e-06, "epoch": 0.6319014274171829, "percentage": 31.6, "elapsed_time": "20:07:43", "remaining_time": "1 day, 19:34:46", "throughput": 5667.3, "total_tokens": 410670656}
+{"current_steps": 9390, "total_steps": 29704, "loss": 0.3573267698287964, "lr": 7.730848444594452e-06, "epoch": 0.632238082413143, "percentage": 31.61, "elapsed_time": "20:08:17", "remaining_time": "1 day, 19:33:58", "throughput": 5667.85, "total_tokens": 410903448}
+{"current_steps": 9395, "total_steps": 29704, "loss": 0.37146763801574706, "lr": 7.7286331825013e-06, "epoch": 0.6325747374091032, "percentage": 31.63, "elapsed_time": "20:08:51", "remaining_time": "1 day, 19:33:10", "throughput": 5668.24, "total_tokens": 411126128}
+{"current_steps": 9400, "total_steps": 29704, "loss": 0.3655902624130249, "lr": 7.72641715735458e-06, "epoch": 0.6329113924050633, "percentage": 31.65, "elapsed_time": "20:09:38", "remaining_time": "1 day, 19:32:49", "throughput": 5667.48, "total_tokens": 411336720}
+{"current_steps": 9405, "total_steps": 29704, "loss": 0.38719449043273924, "lr": 7.724200369773998e-06, "epoch": 0.6332480474010235, "percentage": 31.66, "elapsed_time": "20:10:18", "remaining_time": "1 day, 19:32:13", "throughput": 5667.51, "total_tokens": 411563872}
+{"current_steps": 9410, "total_steps": 29704, "loss": 0.37484476566314695, "lr": 7.721982820379467e-06, "epoch": 0.6335847023969836, "percentage": 31.68, "elapsed_time": "20:10:52", "remaining_time": "1 day, 19:31:24", "throughput": 5667.94, "total_tokens": 411787952}
+{"current_steps": 9415, "total_steps": 29704, "loss": 0.3741923332214355, "lr": 7.719764509791123e-06, "epoch": 0.6339213573929438, "percentage": 31.7, "elapsed_time": "20:11:25", "remaining_time": "1 day, 19:30:34", "throughput": 5667.94, "total_tokens": 411975928}
+{"current_steps": 9420, "total_steps": 29704, "loss": 0.43533172607421877, "lr": 7.717545438629303e-06, "epoch": 0.6342580123889039, "percentage": 31.71, "elapsed_time": "20:12:01", "remaining_time": "1 day, 19:29:51", "throughput": 5668.36, "total_tokens": 412214368}
+{"current_steps": 9425, "total_steps": 29704, "loss": 0.3732184410095215, "lr": 7.71532560751457e-06, "epoch": 0.634594667384864, "percentage": 31.73, "elapsed_time": "20:12:37", "remaining_time": "1 day, 19:29:07", "throughput": 5668.58, "total_tokens": 412434712}
+{"current_steps": 9430, "total_steps": 29704, "loss": 0.38740081787109376, "lr": 7.713105017067686e-06, "epoch": 0.6349313223808242, "percentage": 31.75, "elapsed_time": "20:13:11", "remaining_time": "1 day, 19:28:18", "throughput": 5669.18, "total_tokens": 412668752}
+{"current_steps": 9435, "total_steps": 29704, "loss": 0.37178239822387693, "lr": 7.710883667909639e-06, "epoch": 0.6352679773767843, "percentage": 31.76, "elapsed_time": "20:13:47", "remaining_time": "1 day, 19:27:33", "throughput": 5669.38, "total_tokens": 412886888}
+{"current_steps": 9440, "total_steps": 29704, "loss": 0.3967950105667114, "lr": 7.708661560661617e-06, "epoch": 0.6356046323727444, "percentage": 31.78, "elapsed_time": "20:14:23", "remaining_time": "1 day, 19:26:49", "throughput": 5669.67, "total_tokens": 413111704}
+{"current_steps": 9445, "total_steps": 29704, "loss": 0.39545307159423826, "lr": 7.706438695945026e-06, "epoch": 0.6359412873687046, "percentage": 31.8, "elapsed_time": "20:15:01", "remaining_time": "1 day, 19:26:09", "throughput": 5669.71, "total_tokens": 413330656}
+{"current_steps": 9450, "total_steps": 29704, "loss": 0.3966789960861206, "lr": 7.704215074381484e-06, "epoch": 0.6362779423646647, "percentage": 31.81, "elapsed_time": "20:15:37", "remaining_time": "1 day, 19:25:26", "throughput": 5669.73, "total_tokens": 413537792}
+{"current_steps": 9455, "total_steps": 29704, "loss": 0.38933987617492677, "lr": 7.70199069659282e-06, "epoch": 0.6366145973606249, "percentage": 31.83, "elapsed_time": "20:16:13", "remaining_time": "1 day, 19:24:41", "throughput": 5669.88, "total_tokens": 413751760}
+{"current_steps": 9460, "total_steps": 29704, "loss": 0.3942718982696533, "lr": 7.699765563201075e-06, "epoch": 0.636951252356585, "percentage": 31.85, "elapsed_time": "20:16:43", "remaining_time": "1 day, 19:23:45", "throughput": 5670.63, "total_tokens": 413978016}
+{"current_steps": 9465, "total_steps": 29704, "loss": 0.36391506195068357, "lr": 7.697539674828498e-06, "epoch": 0.6372879073525451, "percentage": 31.86, "elapsed_time": "20:17:23", "remaining_time": "1 day, 19:23:07", "throughput": 5670.78, "total_tokens": 414211120}
+{"current_steps": 9470, "total_steps": 29704, "loss": 0.35442147254943845, "lr": 7.695313032097554e-06, "epoch": 0.6376245623485053, "percentage": 31.88, "elapsed_time": "20:17:59", "remaining_time": "1 day, 19:22:25", "throughput": 5670.92, "total_tokens": 414430048}
+{"current_steps": 9475, "total_steps": 29704, "loss": 0.37603862285614015, "lr": 7.693085635630913e-06, "epoch": 0.6379612173444654, "percentage": 31.9, "elapsed_time": "20:18:46", "remaining_time": "1 day, 19:22:03", "throughput": 5670.08, "total_tokens": 414632320}
+{"current_steps": 9480, "total_steps": 29704, "loss": 0.3694300174713135, "lr": 7.690857486051463e-06, "epoch": 0.6382978723404256, "percentage": 31.91, "elapsed_time": "20:19:26", "remaining_time": "1 day, 19:21:28", "throughput": 5670.08, "total_tokens": 414859816}
+{"current_steps": 9485, "total_steps": 29704, "loss": 0.3734642267227173, "lr": 7.688628583982298e-06, "epoch": 0.6386345273363857, "percentage": 31.93, "elapsed_time": "20:20:08", "remaining_time": "1 day, 19:20:56", "throughput": 5669.74, "total_tokens": 415071856}
+{"current_steps": 9490, "total_steps": 29704, "loss": 0.3597568035125732, "lr": 7.68639893004672e-06, "epoch": 0.6389711823323458, "percentage": 31.95, "elapsed_time": "20:20:49", "remaining_time": "1 day, 19:20:22", "throughput": 5669.63, "total_tokens": 415294888}
+{"current_steps": 9495, "total_steps": 29704, "loss": 0.3566474437713623, "lr": 7.684168524868253e-06, "epoch": 0.639307837328306, "percentage": 31.97, "elapsed_time": "20:21:28", "remaining_time": "1 day, 19:19:46", "throughput": 5669.72, "total_tokens": 415526408}
+{"current_steps": 9500, "total_steps": 29704, "loss": 0.37108328342437746, "lr": 7.681937369070612e-06, "epoch": 0.6396444923242661, "percentage": 31.98, "elapsed_time": "20:22:01", "remaining_time": "1 day, 19:18:55", "throughput": 5670.13, "total_tokens": 415741912}
+{"current_steps": 9505, "total_steps": 29704, "loss": 0.3800150156021118, "lr": 7.67970546327774e-06, "epoch": 0.6399811473202263, "percentage": 32.0, "elapsed_time": "20:22:39", "remaining_time": "1 day, 19:18:15", "throughput": 5670.21, "total_tokens": 415962440}
+{"current_steps": 9510, "total_steps": 29704, "loss": 0.3659864902496338, "lr": 7.67747280811378e-06, "epoch": 0.6403178023161864, "percentage": 32.02, "elapsed_time": "20:23:19", "remaining_time": "1 day, 19:17:39", "throughput": 5670.07, "total_tokens": 416179208}
+{"current_steps": 9515, "total_steps": 29704, "loss": 0.35662126541137695, "lr": 7.675239404203088e-06, "epoch": 0.6406544573121465, "percentage": 32.03, "elapsed_time": "20:23:59", "remaining_time": "1 day, 19:17:04", "throughput": 5669.9, "total_tokens": 416394640}
+{"current_steps": 9520, "total_steps": 29704, "loss": 0.3727959156036377, "lr": 7.673005252170226e-06, "epoch": 0.6409911123081067, "percentage": 32.05, "elapsed_time": "20:24:44", "remaining_time": "1 day, 19:16:38", "throughput": 5669.54, "total_tokens": 416620784}
+{"current_steps": 9525, "total_steps": 29704, "loss": 0.3673388957977295, "lr": 7.670770352639968e-06, "epoch": 0.6413277673040668, "percentage": 32.07, "elapsed_time": "20:25:25", "remaining_time": "1 day, 19:16:05", "throughput": 5669.16, "total_tokens": 416826168}
+{"current_steps": 9530, "total_steps": 29704, "loss": 0.37580122947692873, "lr": 7.668534706237299e-06, "epoch": 0.641664422300027, "percentage": 32.08, "elapsed_time": "20:25:56", "remaining_time": "1 day, 19:15:12", "throughput": 5669.72, "total_tokens": 417047480}
+{"current_steps": 9535, "total_steps": 29704, "loss": 0.4170052528381348, "lr": 7.666298313587409e-06, "epoch": 0.6420010772959871, "percentage": 32.1, "elapsed_time": "20:26:37", "remaining_time": "1 day, 19:14:37", "throughput": 5669.6, "total_tokens": 417268504}
+{"current_steps": 9540, "total_steps": 29704, "loss": 0.3715662956237793, "lr": 7.664061175315696e-06, "epoch": 0.6423377322919472, "percentage": 32.12, "elapsed_time": "20:27:11", "remaining_time": "1 day, 19:13:50", "throughput": 5669.62, "total_tokens": 417464000}
+{"current_steps": 9545, "total_steps": 29704, "loss": 0.3764479160308838, "lr": 7.661823292047771e-06, "epoch": 0.6426743872879074, "percentage": 32.13, "elapsed_time": "20:27:52", "remaining_time": "1 day, 19:13:16", "throughput": 5669.45, "total_tokens": 417683920}
+{"current_steps": 9550, "total_steps": 29704, "loss": 0.3806633949279785, "lr": 7.659584664409452e-06, "epoch": 0.6430110422838675, "percentage": 32.15, "elapsed_time": "20:28:34", "remaining_time": "1 day, 19:12:44", "throughput": 5669.32, "total_tokens": 417910080}
+{"current_steps": 9555, "total_steps": 29704, "loss": 0.33327679634094237, "lr": 7.65734529302676e-06, "epoch": 0.6433476972798277, "percentage": 32.17, "elapsed_time": "20:29:11", "remaining_time": "1 day, 19:12:02", "throughput": 5669.24, "total_tokens": 418114864}
+{"current_steps": 9560, "total_steps": 29704, "loss": 0.36725008487701416, "lr": 7.655105178525932e-06, "epoch": 0.6436843522757878, "percentage": 32.18, "elapsed_time": "20:29:49", "remaining_time": "1 day, 19:11:22", "throughput": 5669.26, "total_tokens": 418331608}
+{"current_steps": 9565, "total_steps": 29704, "loss": 0.37874999046325686, "lr": 7.652864321533406e-06, "epoch": 0.6440210072717479, "percentage": 32.2, "elapsed_time": "20:30:27", "remaining_time": "1 day, 19:10:42", "throughput": 5669.22, "total_tokens": 418542736}
+{"current_steps": 9570, "total_steps": 29704, "loss": 0.39217414855957033, "lr": 7.650622722675833e-06, "epoch": 0.6443576622677081, "percentage": 32.22, "elapsed_time": "20:31:09", "remaining_time": "1 day, 19:10:11", "throughput": 5669.03, "total_tokens": 418769104}
+{"current_steps": 9575, "total_steps": 29704, "loss": 0.3911130428314209, "lr": 7.648380382580067e-06, "epoch": 0.6446943172636682, "percentage": 32.23, "elapsed_time": "20:31:51", "remaining_time": "1 day, 19:09:41", "throughput": 5668.68, "total_tokens": 418983272}
+{"current_steps": 9580, "total_steps": 29704, "loss": 0.3877583026885986, "lr": 7.646137301873173e-06, "epoch": 0.6450309722596284, "percentage": 32.25, "elapsed_time": "20:32:32", "remaining_time": "1 day, 19:09:07", "throughput": 5668.39, "total_tokens": 419193152}
+{"current_steps": 9585, "total_steps": 29704, "loss": 0.35512332916259765, "lr": 7.643893481182421e-06, "epoch": 0.6453676272555885, "percentage": 32.27, "elapsed_time": "20:33:10", "remaining_time": "1 day, 19:08:25", "throughput": 5668.53, "total_tokens": 419415144}
+{"current_steps": 9590, "total_steps": 29704, "loss": 0.4030734062194824, "lr": 7.641648921135287e-06, "epoch": 0.6457042822515486, "percentage": 32.29, "elapsed_time": "20:33:52", "remaining_time": "1 day, 19:07:55", "throughput": 5668.35, "total_tokens": 419642840}
+{"current_steps": 9595, "total_steps": 29704, "loss": 0.38401873111724855, "lr": 7.639403622359455e-06, "epoch": 0.6460409372475088, "percentage": 32.3, "elapsed_time": "20:34:28", "remaining_time": "1 day, 19:07:12", "throughput": 5668.39, "total_tokens": 419851960}
+{"current_steps": 9600, "total_steps": 29704, "loss": 0.3698650598526001, "lr": 7.637157585482817e-06, "epoch": 0.6463775922434689, "percentage": 32.32, "elapsed_time": "20:35:13", "remaining_time": "1 day, 19:06:45", "throughput": 5667.94, "total_tokens": 420069368}
+{"current_steps": 9605, "total_steps": 29704, "loss": 0.3979948043823242, "lr": 7.63491081113347e-06, "epoch": 0.646714247239429, "percentage": 32.34, "elapsed_time": "20:35:49", "remaining_time": "1 day, 19:06:02", "throughput": 5668.28, "total_tokens": 420301264}
+{"current_steps": 9610, "total_steps": 29704, "loss": 0.4009084701538086, "lr": 7.632663299939718e-06, "epoch": 0.6470509022353892, "percentage": 32.35, "elapsed_time": "20:36:31", "remaining_time": "1 day, 19:05:30", "throughput": 5668.13, "total_tokens": 420528480}
+{"current_steps": 9615, "total_steps": 29704, "loss": 0.37017059326171875, "lr": 7.630415052530066e-06, "epoch": 0.6473875572313493, "percentage": 32.37, "elapsed_time": "20:37:08", "remaining_time": "1 day, 19:04:49", "throughput": 5668.2, "total_tokens": 420744352}
+{"current_steps": 9620, "total_steps": 29704, "loss": 0.3800983905792236, "lr": 7.628166069533235e-06, "epoch": 0.6477242122273095, "percentage": 32.39, "elapsed_time": "20:37:44", "remaining_time": "1 day, 19:04:05", "throughput": 5668.24, "total_tokens": 420951552}
+{"current_steps": 9625, "total_steps": 29704, "loss": 0.4101721286773682, "lr": 7.625916351578141e-06, "epoch": 0.6480608672232696, "percentage": 32.4, "elapsed_time": "20:38:22", "remaining_time": "1 day, 19:03:23", "throughput": 5668.34, "total_tokens": 421168904}
+{"current_steps": 9630, "total_steps": 29704, "loss": 0.35078365802764894, "lr": 7.623665899293914e-06, "epoch": 0.6483975222192297, "percentage": 32.42, "elapsed_time": "20:39:03", "remaining_time": "1 day, 19:02:51", "throughput": 5668.09, "total_tokens": 421386640}
+{"current_steps": 9635, "total_steps": 29704, "loss": 0.3464890241622925, "lr": 7.621414713309885e-06, "epoch": 0.6487341772151899, "percentage": 32.44, "elapsed_time": "20:39:40", "remaining_time": "1 day, 19:02:09", "throughput": 5668.19, "total_tokens": 421604504}
+{"current_steps": 9640, "total_steps": 29704, "loss": 0.38260612487792967, "lr": 7.61916279425559e-06, "epoch": 0.64907083221115, "percentage": 32.45, "elapsed_time": "20:40:22", "remaining_time": "1 day, 19:01:37", "throughput": 5668.1, "total_tokens": 421833672}
+{"current_steps": 9645, "total_steps": 29704, "loss": 0.38509573936462405, "lr": 7.6169101427607716e-06, "epoch": 0.6494074872071102, "percentage": 32.47, "elapsed_time": "20:40:56", "remaining_time": "1 day, 19:00:49", "throughput": 5668.64, "total_tokens": 422067496}
+{"current_steps": 9650, "total_steps": 29704, "loss": 0.38222179412841795, "lr": 7.614656759455375e-06, "epoch": 0.6497441422030703, "percentage": 32.49, "elapsed_time": "20:41:35", "remaining_time": "1 day, 19:00:12", "throughput": 5668.62, "total_tokens": 422289144}
+{"current_steps": 9655, "total_steps": 29704, "loss": 0.38720729351043703, "lr": 7.612402644969555e-06, "epoch": 0.6500807971990304, "percentage": 32.5, "elapsed_time": "20:42:17", "remaining_time": "1 day, 18:59:39", "throughput": 5668.33, "total_tokens": 422500848}
+{"current_steps": 9660, "total_steps": 29704, "loss": 0.37707133293151857, "lr": 7.6101477999336646e-06, "epoch": 0.6504174521949906, "percentage": 32.52, "elapsed_time": "20:43:04", "remaining_time": "1 day, 18:59:19", "throughput": 5667.63, "total_tokens": 422718672}
+{"current_steps": 9665, "total_steps": 29704, "loss": 0.3739673376083374, "lr": 7.607892224978266e-06, "epoch": 0.6507541071909507, "percentage": 32.54, "elapsed_time": "20:43:40", "remaining_time": "1 day, 18:58:35", "throughput": 5667.73, "total_tokens": 422929480}
+{"current_steps": 9670, "total_steps": 29704, "loss": 0.3418617010116577, "lr": 7.605635920734122e-06, "epoch": 0.6510907621869109, "percentage": 32.55, "elapsed_time": "20:44:15", "remaining_time": "1 day, 18:57:49", "throughput": 5667.89, "total_tokens": 423142128}
+{"current_steps": 9675, "total_steps": 29704, "loss": 0.3855189561843872, "lr": 7.603378887832202e-06, "epoch": 0.651427417182871, "percentage": 32.57, "elapsed_time": "20:44:55", "remaining_time": "1 day, 18:57:13", "throughput": 5667.67, "total_tokens": 423349056}
+{"current_steps": 9680, "total_steps": 29704, "loss": 0.3803261280059814, "lr": 7.601121126903676e-06, "epoch": 0.6517640721788311, "percentage": 32.59, "elapsed_time": "20:45:38", "remaining_time": "1 day, 18:56:43", "throughput": 5667.43, "total_tokens": 423573240}
+{"current_steps": 9685, "total_steps": 29704, "loss": 0.3738381385803223, "lr": 7.598862638579922e-06, "epoch": 0.6521007271747913, "percentage": 32.61, "elapsed_time": "20:46:21", "remaining_time": "1 day, 18:56:14", "throughput": 5667.28, "total_tokens": 423809440}
+{"current_steps": 9690, "total_steps": 29704, "loss": 0.3761345624923706, "lr": 7.5966034234925205e-06, "epoch": 0.6524373821707514, "percentage": 32.62, "elapsed_time": "20:47:01", "remaining_time": "1 day, 18:55:38", "throughput": 5667.39, "total_tokens": 424041528}
+{"current_steps": 9695, "total_steps": 29704, "loss": 0.43105735778808596, "lr": 7.59434348227325e-06, "epoch": 0.6527740371667116, "percentage": 32.64, "elapsed_time": "20:47:35", "remaining_time": "1 day, 18:54:49", "throughput": 5667.84, "total_tokens": 424268184}
+{"current_steps": 9700, "total_steps": 29704, "loss": 0.40675368309021, "lr": 7.592082815554098e-06, "epoch": 0.6531106921626717, "percentage": 32.66, "elapsed_time": "20:48:14", "remaining_time": "1 day, 18:54:11", "throughput": 5667.91, "total_tokens": 424492616}
+{"current_steps": 9705, "total_steps": 29704, "loss": 0.419390869140625, "lr": 7.58982142396725e-06, "epoch": 0.6534473471586318, "percentage": 32.67, "elapsed_time": "20:48:59", "remaining_time": "1 day, 18:53:46", "throughput": 5667.07, "total_tokens": 424686848}
+{"current_steps": 9710, "total_steps": 29704, "loss": 0.3695266008377075, "lr": 7.5875593081451e-06, "epoch": 0.653784002154592, "percentage": 32.69, "elapsed_time": "20:49:45", "remaining_time": "1 day, 18:53:24", "throughput": 5666.59, "total_tokens": 424913632}
+{"current_steps": 9715, "total_steps": 29704, "loss": 0.3619475126266479, "lr": 7.585296468720243e-06, "epoch": 0.6541206571505521, "percentage": 32.71, "elapsed_time": "20:50:26", "remaining_time": "1 day, 18:52:50", "throughput": 5666.51, "total_tokens": 425139040}
+{"current_steps": 9720, "total_steps": 29704, "loss": 0.38277277946472166, "lr": 7.5830329063254705e-06, "epoch": 0.6544573121465123, "percentage": 32.72, "elapsed_time": "20:51:05", "remaining_time": "1 day, 18:52:11", "throughput": 5666.44, "total_tokens": 425352824}
+{"current_steps": 9725, "total_steps": 29704, "loss": 0.38765199184417726, "lr": 7.5807686215937835e-06, "epoch": 0.6547939671424724, "percentage": 32.74, "elapsed_time": "20:51:45", "remaining_time": "1 day, 18:51:36", "throughput": 5666.25, "total_tokens": 425565784}
+{"current_steps": 9730, "total_steps": 29704, "loss": 0.3398409366607666, "lr": 7.578503615158379e-06, "epoch": 0.6551306221384325, "percentage": 32.76, "elapsed_time": "20:52:29", "remaining_time": "1 day, 18:51:08", "throughput": 5665.77, "total_tokens": 425778128}
+{"current_steps": 9735, "total_steps": 29704, "loss": 0.38698830604553225, "lr": 7.576237887652662e-06, "epoch": 0.6554672771343927, "percentage": 32.77, "elapsed_time": "20:53:08", "remaining_time": "1 day, 18:50:30", "throughput": 5665.87, "total_tokens": 426007040}
+{"current_steps": 9740, "total_steps": 29704, "loss": 0.34668645858764646, "lr": 7.573971439710235e-06, "epoch": 0.6558039321303528, "percentage": 32.79, "elapsed_time": "20:53:45", "remaining_time": "1 day, 18:49:48", "throughput": 5665.97, "total_tokens": 426223488}
+{"current_steps": 9745, "total_steps": 29704, "loss": 0.3675506591796875, "lr": 7.571704271964902e-06, "epoch": 0.656140587126313, "percentage": 32.81, "elapsed_time": "20:54:20", "remaining_time": "1 day, 18:49:03", "throughput": 5666.09, "total_tokens": 426433528}
+{"current_steps": 9750, "total_steps": 29704, "loss": 0.3636334419250488, "lr": 7.56943638505067e-06, "epoch": 0.6564772421222731, "percentage": 32.82, "elapsed_time": "20:54:56", "remaining_time": "1 day, 18:48:19", "throughput": 5666.38, "total_tokens": 426658696}
+{"current_steps": 9755, "total_steps": 29704, "loss": 0.39123508930206297, "lr": 7.567167779601746e-06, "epoch": 0.6568138971182332, "percentage": 32.84, "elapsed_time": "20:55:38", "remaining_time": "1 day, 18:47:46", "throughput": 5666.28, "total_tokens": 426886792}
+{"current_steps": 9760, "total_steps": 29704, "loss": 0.413257360458374, "lr": 7.56489845625254e-06, "epoch": 0.6571505521141934, "percentage": 32.86, "elapsed_time": "20:56:26", "remaining_time": "1 day, 18:47:27", "throughput": 5665.55, "total_tokens": 427104192}
+{"current_steps": 9765, "total_steps": 29704, "loss": 0.369996976852417, "lr": 7.562628415637659e-06, "epoch": 0.6574872071101535, "percentage": 32.87, "elapsed_time": "20:57:06", "remaining_time": "1 day, 18:46:52", "throughput": 5665.28, "total_tokens": 427312504}
+{"current_steps": 9770, "total_steps": 29704, "loss": 0.4071043491363525, "lr": 7.560357658391913e-06, "epoch": 0.6578238621061137, "percentage": 32.89, "elapsed_time": "20:57:45", "remaining_time": "1 day, 18:46:14", "throughput": 5665.29, "total_tokens": 427533792}
+{"current_steps": 9775, "total_steps": 29704, "loss": 0.3804465770721436, "lr": 7.558086185150311e-06, "epoch": 0.6581605171020738, "percentage": 32.91, "elapsed_time": "20:58:27", "remaining_time": "1 day, 18:45:41", "throughput": 5665.0, "total_tokens": 427747608}
+{"current_steps": 9780, "total_steps": 29704, "loss": 0.40397090911865235, "lr": 7.555813996548065e-06, "epoch": 0.6584971720980339, "percentage": 32.92, "elapsed_time": "20:59:03", "remaining_time": "1 day, 18:44:59", "throughput": 5665.22, "total_tokens": 427972008}
+{"current_steps": 9785, "total_steps": 29704, "loss": 0.3842537641525269, "lr": 7.553541093220586e-06, "epoch": 0.6588338270939941, "percentage": 32.94, "elapsed_time": "20:59:39", "remaining_time": "1 day, 18:44:14", "throughput": 5665.38, "total_tokens": 428186312}
+{"current_steps": 9790, "total_steps": 29704, "loss": 0.37559702396392824, "lr": 7.551267475803481e-06, "epoch": 0.6591704820899542, "percentage": 32.96, "elapsed_time": "21:00:19", "remaining_time": "1 day, 18:43:37", "throughput": 5665.43, "total_tokens": 428413952}
+{"current_steps": 9795, "total_steps": 29704, "loss": 0.3736457347869873, "lr": 7.5489931449325635e-06, "epoch": 0.6595071370859144, "percentage": 32.98, "elapsed_time": "21:00:58", "remaining_time": "1 day, 18:43:01", "throughput": 5665.33, "total_tokens": 428630728}
+{"current_steps": 9800, "total_steps": 29704, "loss": 0.39138045310974123, "lr": 7.546718101243842e-06, "epoch": 0.6598437920818745, "percentage": 32.99, "elapsed_time": "21:01:33", "remaining_time": "1 day, 18:42:15", "throughput": 5665.49, "total_tokens": 428842760}
+{"current_steps": 9805, "total_steps": 29704, "loss": 0.3698113918304443, "lr": 7.544442345373522e-06, "epoch": 0.6601804470778346, "percentage": 33.01, "elapsed_time": "21:02:14", "remaining_time": "1 day, 18:41:41", "throughput": 5665.19, "total_tokens": 429051072}
+{"current_steps": 9810, "total_steps": 29704, "loss": 0.3717064380645752, "lr": 7.542165877958014e-06, "epoch": 0.6605171020737948, "percentage": 33.03, "elapsed_time": "21:02:51", "remaining_time": "1 day, 18:40:59", "throughput": 5665.25, "total_tokens": 429266232}
+{"current_steps": 9815, "total_steps": 29704, "loss": 0.38160529136657717, "lr": 7.539888699633926e-06, "epoch": 0.6608537570697549, "percentage": 33.04, "elapsed_time": "21:03:28", "remaining_time": "1 day, 18:40:18", "throughput": 5665.5, "total_tokens": 429495032}
+{"current_steps": 9820, "total_steps": 29704, "loss": 0.34899187088012695, "lr": 7.5376108110380606e-06, "epoch": 0.661190412065715, "percentage": 33.06, "elapsed_time": "21:04:07", "remaining_time": "1 day, 18:39:39", "throughput": 5665.63, "total_tokens": 429723384}
+{"current_steps": 9825, "total_steps": 29704, "loss": 0.3428369998931885, "lr": 7.535332212807426e-06, "epoch": 0.6615270670616752, "percentage": 33.08, "elapsed_time": "21:04:46", "remaining_time": "1 day, 18:39:01", "throughput": 5665.37, "total_tokens": 429924744}
+{"current_steps": 9830, "total_steps": 29704, "loss": 0.39725618362426757, "lr": 7.53305290557922e-06, "epoch": 0.6618637220576353, "percentage": 33.09, "elapsed_time": "21:05:25", "remaining_time": "1 day, 18:38:23", "throughput": 5665.5, "total_tokens": 430154928}
+{"current_steps": 9835, "total_steps": 29704, "loss": 0.3740149974822998, "lr": 7.530772889990848e-06, "epoch": 0.6622003770535955, "percentage": 33.11, "elapsed_time": "21:05:59", "remaining_time": "1 day, 18:37:36", "throughput": 5666.0, "total_tokens": 430386800}
+{"current_steps": 9840, "total_steps": 29704, "loss": 0.35007429122924805, "lr": 7.528492166679907e-06, "epoch": 0.6625370320495556, "percentage": 33.13, "elapsed_time": "21:06:41", "remaining_time": "1 day, 18:37:04", "throughput": 5665.89, "total_tokens": 430616592}
+{"current_steps": 9845, "total_steps": 29704, "loss": 0.38547792434692385, "lr": 7.526210736284194e-06, "epoch": 0.6628736870455157, "percentage": 33.14, "elapsed_time": "21:07:16", "remaining_time": "1 day, 18:36:17", "throughput": 5666.0, "total_tokens": 430820368}
+{"current_steps": 9850, "total_steps": 29704, "loss": 0.3851293087005615, "lr": 7.523928599441706e-06, "epoch": 0.6632103420414759, "percentage": 33.16, "elapsed_time": "21:07:56", "remaining_time": "1 day, 18:35:42", "throughput": 5665.93, "total_tokens": 431045232}
+{"current_steps": 9855, "total_steps": 29704, "loss": 0.3921393632888794, "lr": 7.5216457567906295e-06, "epoch": 0.663546997037436, "percentage": 33.18, "elapsed_time": "21:08:39", "remaining_time": "1 day, 18:35:11", "throughput": 5665.76, "total_tokens": 431272728}
+{"current_steps": 9860, "total_steps": 29704, "loss": 0.3589025974273682, "lr": 7.519362208969362e-06, "epoch": 0.6638836520333962, "percentage": 33.19, "elapsed_time": "21:09:24", "remaining_time": "1 day, 18:34:47", "throughput": 5665.31, "total_tokens": 431497576}
+{"current_steps": 9865, "total_steps": 29704, "loss": 0.39649462699890137, "lr": 7.5170779566164855e-06, "epoch": 0.6642203070293563, "percentage": 33.21, "elapsed_time": "21:10:07", "remaining_time": "1 day, 18:34:17", "throughput": 5665.0, "total_tokens": 431716768}
+{"current_steps": 9870, "total_steps": 29704, "loss": 0.39783761501312254, "lr": 7.514793000370783e-06, "epoch": 0.6645569620253164, "percentage": 33.23, "elapsed_time": "21:10:52", "remaining_time": "1 day, 18:33:50", "throughput": 5664.45, "total_tokens": 431927680}
+{"current_steps": 9875, "total_steps": 29704, "loss": 0.3664602756500244, "lr": 7.512507340871239e-06, "epoch": 0.6648936170212766, "percentage": 33.24, "elapsed_time": "21:11:33", "remaining_time": "1 day, 18:33:16", "throughput": 5664.26, "total_tokens": 432144784}
+{"current_steps": 9880, "total_steps": 29704, "loss": 0.39647390842437746, "lr": 7.510220978757028e-06, "epoch": 0.6652302720172367, "percentage": 33.26, "elapsed_time": "21:12:15", "remaining_time": "1 day, 18:32:44", "throughput": 5663.99, "total_tokens": 432361504}
+{"current_steps": 9885, "total_steps": 29704, "loss": 0.4201164722442627, "lr": 7.507933914667524e-06, "epoch": 0.6655669270131969, "percentage": 33.28, "elapsed_time": "21:12:55", "remaining_time": "1 day, 18:32:10", "throughput": 5663.93, "total_tokens": 432586992}
+{"current_steps": 9890, "total_steps": 29704, "loss": 0.37460851669311523, "lr": 7.505646149242299e-06, "epoch": 0.665903582009157, "percentage": 33.3, "elapsed_time": "21:13:35", "remaining_time": "1 day, 18:31:34", "throughput": 5663.86, "total_tokens": 432809080}
+{"current_steps": 9895, "total_steps": 29704, "loss": 0.34871246814727785, "lr": 7.503357683121119e-06, "epoch": 0.6662402370051171, "percentage": 33.31, "elapsed_time": "21:14:22", "remaining_time": "1 day, 18:31:12", "throughput": 5663.35, "total_tokens": 433035120}
+{"current_steps": 9900, "total_steps": 29704, "loss": 0.3231820583343506, "lr": 7.501068516943944e-06, "epoch": 0.6665768920010773, "percentage": 33.33, "elapsed_time": "21:15:02", "remaining_time": "1 day, 18:30:35", "throughput": 5663.21, "total_tokens": 433248272}
+{"current_steps": 9905, "total_steps": 29704, "loss": 0.3865694284439087, "lr": 7.498778651350934e-06, "epoch": 0.6669135469970374, "percentage": 33.35, "elapsed_time": "21:15:39", "remaining_time": "1 day, 18:29:53", "throughput": 5663.42, "total_tokens": 433472920}
+{"current_steps": 9910, "total_steps": 29704, "loss": 0.3751338720321655, "lr": 7.496488086982443e-06, "epoch": 0.6672502019929976, "percentage": 33.36, "elapsed_time": "21:16:22", "remaining_time": "1 day, 18:29:23", "throughput": 5663.13, "total_tokens": 433694440}
+{"current_steps": 9915, "total_steps": 29704, "loss": 0.37354316711425783, "lr": 7.494196824479019e-06, "epoch": 0.6675868569889577, "percentage": 33.38, "elapsed_time": "21:16:59", "remaining_time": "1 day, 18:28:42", "throughput": 5663.16, "total_tokens": 433909064}
+{"current_steps": 9920, "total_steps": 29704, "loss": 0.3818557024002075, "lr": 7.491904864481406e-06, "epoch": 0.6679235119849178, "percentage": 33.4, "elapsed_time": "21:17:41", "remaining_time": "1 day, 18:28:10", "throughput": 5662.84, "total_tokens": 434122768}
+{"current_steps": 9925, "total_steps": 29704, "loss": 0.34175124168396, "lr": 7.489612207630543e-06, "epoch": 0.668260166980878, "percentage": 33.41, "elapsed_time": "21:18:15", "remaining_time": "1 day, 18:27:22", "throughput": 5663.36, "total_tokens": 434354528}
+{"current_steps": 9930, "total_steps": 29704, "loss": 0.37988905906677245, "lr": 7.487318854567566e-06, "epoch": 0.6685968219768381, "percentage": 33.43, "elapsed_time": "21:18:49", "remaining_time": "1 day, 18:26:34", "throughput": 5663.59, "total_tokens": 434564096}
+{"current_steps": 9935, "total_steps": 29704, "loss": 0.3599559307098389, "lr": 7.485024805933803e-06, "epoch": 0.6689334769727983, "percentage": 33.45, "elapsed_time": "21:19:25", "remaining_time": "1 day, 18:25:51", "throughput": 5663.55, "total_tokens": 434767584}
+{"current_steps": 9940, "total_steps": 29704, "loss": 0.3699958086013794, "lr": 7.482730062370776e-06, "epoch": 0.6692701319687584, "percentage": 33.46, "elapsed_time": "21:20:10", "remaining_time": "1 day, 18:25:24", "throughput": 5663.06, "total_tokens": 434981392}
+{"current_steps": 9945, "total_steps": 29704, "loss": 0.3593663215637207, "lr": 7.480434624520205e-06, "epoch": 0.6696067869647185, "percentage": 33.48, "elapsed_time": "21:20:43", "remaining_time": "1 day, 18:24:33", "throughput": 5663.72, "total_tokens": 435217648}
+{"current_steps": 9950, "total_steps": 29704, "loss": 0.3879536151885986, "lr": 7.478138493024001e-06, "epoch": 0.6699434419606787, "percentage": 33.5, "elapsed_time": "21:21:22", "remaining_time": "1 day, 18:23:56", "throughput": 5663.76, "total_tokens": 435444608}
+{"current_steps": 9955, "total_steps": 29704, "loss": 0.34809188842773436, "lr": 7.475841668524268e-06, "epoch": 0.6702800969566388, "percentage": 33.51, "elapsed_time": "21:22:03", "remaining_time": "1 day, 18:23:23", "throughput": 5663.47, "total_tokens": 435656280}
+{"current_steps": 9960, "total_steps": 29704, "loss": 0.4048673152923584, "lr": 7.473544151663309e-06, "epoch": 0.670616751952599, "percentage": 33.53, "elapsed_time": "21:22:38", "remaining_time": "1 day, 18:22:37", "throughput": 5663.75, "total_tokens": 435875456}
+{"current_steps": 9965, "total_steps": 29704, "loss": 0.3677277326583862, "lr": 7.471245943083615e-06, "epoch": 0.6709534069485591, "percentage": 33.55, "elapsed_time": "21:23:18", "remaining_time": "1 day, 18:22:00", "throughput": 5663.73, "total_tokens": 436097568}
+{"current_steps": 9970, "total_steps": 29704, "loss": 0.3423744201660156, "lr": 7.468947043427873e-06, "epoch": 0.6712900619445192, "percentage": 33.56, "elapsed_time": "21:23:52", "remaining_time": "1 day, 18:21:12", "throughput": 5664.07, "total_tokens": 436314872}
+{"current_steps": 9975, "total_steps": 29704, "loss": 0.38165833950042727, "lr": 7.466647453338965e-06, "epoch": 0.6716267169404794, "percentage": 33.58, "elapsed_time": "21:24:30", "remaining_time": "1 day, 18:20:32", "throughput": 5663.99, "total_tokens": 436524088}
+{"current_steps": 9980, "total_steps": 29704, "loss": 0.3900636911392212, "lr": 7.464347173459958e-06, "epoch": 0.6719633719364395, "percentage": 33.6, "elapsed_time": "21:25:08", "remaining_time": "1 day, 18:19:53", "throughput": 5664.28, "total_tokens": 436763792}
+{"current_steps": 9985, "total_steps": 29704, "loss": 0.35174126625061036, "lr": 7.462046204434126e-06, "epoch": 0.6723000269323997, "percentage": 33.62, "elapsed_time": "21:25:43", "remaining_time": "1 day, 18:19:07", "throughput": 5664.79, "total_tokens": 437001592}
+{"current_steps": 9990, "total_steps": 29704, "loss": 0.37253947257995607, "lr": 7.459744546904922e-06, "epoch": 0.6726366819283598, "percentage": 33.63, "elapsed_time": "21:26:17", "remaining_time": "1 day, 18:18:19", "throughput": 5665.09, "total_tokens": 437217472}
+{"current_steps": 9995, "total_steps": 29704, "loss": 0.34635467529296876, "lr": 7.457442201516001e-06, "epoch": 0.6729733369243199, "percentage": 33.65, "elapsed_time": "21:27:02", "remaining_time": "1 day, 18:17:54", "throughput": 5664.55, "total_tokens": 437432256}
+{"current_steps": 10000, "total_steps": 29704, "loss": 0.3714813232421875, "lr": 7.455139168911203e-06, "epoch": 0.6733099919202801, "percentage": 33.67, "elapsed_time": "21:27:39", "remaining_time": "1 day, 18:17:12", "throughput": 5664.43, "total_tokens": 437632184}
+{"current_steps": 10005, "total_steps": 29704, "loss": 0.37085671424865724, "lr": 7.452835449734564e-06, "epoch": 0.6736466469162402, "percentage": 33.68, "elapsed_time": "21:28:17", "remaining_time": "1 day, 18:16:32", "throughput": 5664.51, "total_tokens": 437852448}
+{"current_steps": 10010, "total_steps": 29704, "loss": 0.35476658344268797, "lr": 7.450531044630315e-06, "epoch": 0.6739833019122003, "percentage": 33.7, "elapsed_time": "21:28:50", "remaining_time": "1 day, 18:15:43", "throughput": 5664.92, "total_tokens": 438072720}
+{"current_steps": 10015, "total_steps": 29704, "loss": 0.3909281253814697, "lr": 7.448225954242874e-06, "epoch": 0.6743199569081605, "percentage": 33.72, "elapsed_time": "21:29:32", "remaining_time": "1 day, 18:15:11", "throughput": 5664.78, "total_tokens": 438300752}
+{"current_steps": 10020, "total_steps": 29704, "loss": 0.37700977325439455, "lr": 7.44592017921685e-06, "epoch": 0.6746566119041206, "percentage": 33.73, "elapsed_time": "21:30:08", "remaining_time": "1 day, 18:14:26", "throughput": 5665.06, "total_tokens": 438521952}
+{"current_steps": 10025, "total_steps": 29704, "loss": 0.39572868347167967, "lr": 7.443613720197048e-06, "epoch": 0.6749932669000808, "percentage": 33.75, "elapsed_time": "21:30:51", "remaining_time": "1 day, 18:13:56", "throughput": 5664.72, "total_tokens": 438742144}
+{"current_steps": 10030, "total_steps": 29704, "loss": 0.35796802043914794, "lr": 7.44130657782846e-06, "epoch": 0.6753299218960409, "percentage": 33.77, "elapsed_time": "21:31:27", "remaining_time": "1 day, 18:13:12", "throughput": 5664.74, "total_tokens": 438946944}
+{"current_steps": 10035, "total_steps": 29704, "loss": 0.3807486057281494, "lr": 7.438998752756274e-06, "epoch": 0.675666576892001, "percentage": 33.78, "elapsed_time": "21:32:10", "remaining_time": "1 day, 18:12:42", "throughput": 5664.44, "total_tokens": 439164968}
+{"current_steps": 10040, "total_steps": 29704, "loss": 0.36953206062316896, "lr": 7.436690245625864e-06, "epoch": 0.6760032318879612, "percentage": 33.8, "elapsed_time": "21:32:48", "remaining_time": "1 day, 18:12:03", "throughput": 5664.43, "total_tokens": 439382032}
+{"current_steps": 10045, "total_steps": 29704, "loss": 0.369028639793396, "lr": 7.434381057082797e-06, "epoch": 0.6763398868839213, "percentage": 33.82, "elapsed_time": "21:33:28", "remaining_time": "1 day, 18:11:27", "throughput": 5664.17, "total_tokens": 439590072}
+{"current_steps": 10050, "total_steps": 29704, "loss": 0.3970766067504883, "lr": 7.4320711877728305e-06, "epoch": 0.6766765418798815, "percentage": 33.83, "elapsed_time": "21:34:15", "remaining_time": "1 day, 18:11:04", "throughput": 5663.65, "total_tokens": 439814224}
+{"current_steps": 10055, "total_steps": 29704, "loss": 0.34726481437683104, "lr": 7.429760638341909e-06, "epoch": 0.6770131968758416, "percentage": 33.85, "elapsed_time": "21:34:54", "remaining_time": "1 day, 18:10:26", "throughput": 5663.39, "total_tokens": 440013360}
+{"current_steps": 10060, "total_steps": 29704, "loss": 0.42118000984191895, "lr": 7.427449409436176e-06, "epoch": 0.6773498518718017, "percentage": 33.87, "elapsed_time": "21:35:35", "remaining_time": "1 day, 18:09:53", "throughput": 5663.18, "total_tokens": 440232216}
+{"current_steps": 10065, "total_steps": 29704, "loss": 0.35333929061889646, "lr": 7.425137501701955e-06, "epoch": 0.6776865068677619, "percentage": 33.88, "elapsed_time": "21:36:18", "remaining_time": "1 day, 18:09:22", "throughput": 5663.1, "total_tokens": 440465120}
+{"current_steps": 10070, "total_steps": 29704, "loss": 0.3732592582702637, "lr": 7.422824915785765e-06, "epoch": 0.678023161863722, "percentage": 33.9, "elapsed_time": "21:36:53", "remaining_time": "1 day, 18:08:37", "throughput": 5663.33, "total_tokens": 440684520}
+{"current_steps": 10075, "total_steps": 29704, "loss": 0.3922509431838989, "lr": 7.420511652334313e-06, "epoch": 0.6783598168596822, "percentage": 33.92, "elapsed_time": "21:37:36", "remaining_time": "1 day, 18:08:06", "throughput": 5662.83, "total_tokens": 440888096}
+{"current_steps": 10080, "total_steps": 29704, "loss": 0.3609792947769165, "lr": 7.418197711994495e-06, "epoch": 0.6786964718556423, "percentage": 33.93, "elapsed_time": "21:38:16", "remaining_time": "1 day, 18:07:30", "throughput": 5662.63, "total_tokens": 441099168}
+{"current_steps": 10085, "total_steps": 29704, "loss": 0.37356295585632326, "lr": 7.4158830954133984e-06, "epoch": 0.6790331268516024, "percentage": 33.95, "elapsed_time": "21:38:56", "remaining_time": "1 day, 18:06:54", "throughput": 5662.69, "total_tokens": 441329664}
+{"current_steps": 10090, "total_steps": 29704, "loss": 0.36446294784545896, "lr": 7.413567803238298e-06, "epoch": 0.6793697818475626, "percentage": 33.97, "elapsed_time": "21:39:33", "remaining_time": "1 day, 18:06:13", "throughput": 5662.74, "total_tokens": 441545160}
+{"current_steps": 10095, "total_steps": 29704, "loss": 0.3654582977294922, "lr": 7.411251836116658e-06, "epoch": 0.6797064368435227, "percentage": 33.99, "elapsed_time": "21:40:21", "remaining_time": "1 day, 18:05:52", "throughput": 5662.18, "total_tokens": 441770600}
+{"current_steps": 10100, "total_steps": 29704, "loss": 0.38489809036254885, "lr": 7.408935194696132e-06, "epoch": 0.6800430918394829, "percentage": 34.0, "elapsed_time": "21:41:02", "remaining_time": "1 day, 18:05:19", "throughput": 5661.98, "total_tokens": 441989608}
+{"current_steps": 10105, "total_steps": 29704, "loss": 0.39519965648651123, "lr": 7.406617879624558e-06, "epoch": 0.680379746835443, "percentage": 34.02, "elapsed_time": "21:41:44", "remaining_time": "1 day, 18:04:45", "throughput": 5661.93, "total_tokens": 442219464}
+{"current_steps": 10110, "total_steps": 29704, "loss": 0.35456171035766604, "lr": 7.40429989154997e-06, "epoch": 0.6807164018314031, "percentage": 34.04, "elapsed_time": "21:42:26", "remaining_time": "1 day, 18:04:15", "throughput": 5661.57, "total_tokens": 442434504}
+{"current_steps": 10115, "total_steps": 29704, "loss": 0.396740984916687, "lr": 7.4019812311205826e-06, "epoch": 0.6810530568273633, "percentage": 34.05, "elapsed_time": "21:43:06", "remaining_time": "1 day, 18:03:39", "throughput": 5661.61, "total_tokens": 442663960}
+{"current_steps": 10120, "total_steps": 29704, "loss": 0.39820547103881837, "lr": 7.399661898984805e-06, "epoch": 0.6813897118233234, "percentage": 34.07, "elapsed_time": "21:43:48", "remaining_time": "1 day, 18:03:06", "throughput": 5661.65, "total_tokens": 442904544}
+{"current_steps": 10125, "total_steps": 29704, "loss": 0.3785873889923096, "lr": 7.397341895791228e-06, "epoch": 0.6817263668192836, "percentage": 34.09, "elapsed_time": "21:44:31", "remaining_time": "1 day, 18:02:35", "throughput": 5661.33, "total_tokens": 443119656}
+{"current_steps": 10130, "total_steps": 29704, "loss": 0.37588326930999755, "lr": 7.395021222188634e-06, "epoch": 0.6820630218152437, "percentage": 34.1, "elapsed_time": "21:45:07", "remaining_time": "1 day, 18:01:52", "throughput": 5661.58, "total_tokens": 443345048}
+{"current_steps": 10135, "total_steps": 29704, "loss": 0.4026289939880371, "lr": 7.392699878825993e-06, "epoch": 0.6823996768112038, "percentage": 34.12, "elapsed_time": "21:45:48", "remaining_time": "1 day, 18:01:18", "throughput": 5661.23, "total_tokens": 443550416}
+{"current_steps": 10140, "total_steps": 29704, "loss": 0.3739679574966431, "lr": 7.390377866352462e-06, "epoch": 0.682736331807164, "percentage": 34.14, "elapsed_time": "21:46:31", "remaining_time": "1 day, 18:00:46", "throughput": 5660.85, "total_tokens": 443760728}
+{"current_steps": 10145, "total_steps": 29704, "loss": 0.36473128795623777, "lr": 7.388055185417382e-06, "epoch": 0.6830729868031241, "percentage": 34.15, "elapsed_time": "21:47:09", "remaining_time": "1 day, 18:00:08", "throughput": 5660.94, "total_tokens": 443986040}
+{"current_steps": 10150, "total_steps": 29704, "loss": 0.35809712409973143, "lr": 7.3857318366702844e-06, "epoch": 0.6834096417990843, "percentage": 34.17, "elapsed_time": "21:47:52", "remaining_time": "1 day, 17:59:38", "throughput": 5660.64, "total_tokens": 444206480}
+{"current_steps": 10155, "total_steps": 29704, "loss": 0.3713981151580811, "lr": 7.383407820760885e-06, "epoch": 0.6837462967950444, "percentage": 34.19, "elapsed_time": "21:48:35", "remaining_time": "1 day, 17:59:07", "throughput": 5660.43, "total_tokens": 444432096}
+{"current_steps": 10160, "total_steps": 29704, "loss": 0.38779325485229493, "lr": 7.381083138339089e-06, "epoch": 0.6840829517910045, "percentage": 34.2, "elapsed_time": "21:49:15", "remaining_time": "1 day, 17:58:31", "throughput": 5660.21, "total_tokens": 444643008}
+{"current_steps": 10165, "total_steps": 29704, "loss": 0.39658608436584475, "lr": 7.378757790054988e-06, "epoch": 0.6844196067869647, "percentage": 34.22, "elapsed_time": "21:49:52", "remaining_time": "1 day, 17:57:49", "throughput": 5660.54, "total_tokens": 444875528}
+{"current_steps": 10170, "total_steps": 29704, "loss": 0.3956130504608154, "lr": 7.376431776558855e-06, "epoch": 0.6847562617829248, "percentage": 34.24, "elapsed_time": "21:50:30", "remaining_time": "1 day, 17:57:09", "throughput": 5660.6, "total_tokens": 445097312}
+{"current_steps": 10175, "total_steps": 29704, "loss": 0.3585235118865967, "lr": 7.3741050985011546e-06, "epoch": 0.685092916778885, "percentage": 34.25, "elapsed_time": "21:51:10", "remaining_time": "1 day, 17:56:33", "throughput": 5660.57, "total_tokens": 445319152}
+{"current_steps": 10180, "total_steps": 29704, "loss": 0.34885153770446775, "lr": 7.371777756532531e-06, "epoch": 0.6854295717748451, "percentage": 34.27, "elapsed_time": "21:51:44", "remaining_time": "1 day, 17:55:45", "throughput": 5660.97, "total_tokens": 445544848}
+{"current_steps": 10185, "total_steps": 29704, "loss": 0.3933322191238403, "lr": 7.369449751303823e-06, "epoch": 0.6857662267708052, "percentage": 34.29, "elapsed_time": "21:52:21", "remaining_time": "1 day, 17:55:03", "throughput": 5661.12, "total_tokens": 445764768}
+{"current_steps": 10190, "total_steps": 29704, "loss": 0.3657109498977661, "lr": 7.367121083466048e-06, "epoch": 0.6861028817667654, "percentage": 34.31, "elapsed_time": "21:52:56", "remaining_time": "1 day, 17:54:18", "throughput": 5661.42, "total_tokens": 445988536}
+{"current_steps": 10195, "total_steps": 29704, "loss": 0.3847865104675293, "lr": 7.364791753670409e-06, "epoch": 0.6864395367627255, "percentage": 34.32, "elapsed_time": "21:53:30", "remaining_time": "1 day, 17:53:29", "throughput": 5661.89, "total_tokens": 446213600}
+{"current_steps": 10200, "total_steps": 29704, "loss": 0.372448205947876, "lr": 7.362461762568298e-06, "epoch": 0.6867761917586857, "percentage": 34.34, "elapsed_time": "21:54:08", "remaining_time": "1 day, 17:52:50", "throughput": 5661.92, "total_tokens": 446432912}
+{"current_steps": 10205, "total_steps": 29704, "loss": 0.3532183408737183, "lr": 7.360131110811287e-06, "epoch": 0.6871128467546458, "percentage": 34.36, "elapsed_time": "21:54:48", "remaining_time": "1 day, 17:52:14", "throughput": 5661.88, "total_tokens": 446656416}
+{"current_steps": 10210, "total_steps": 29704, "loss": 0.3632460355758667, "lr": 7.357799799051138e-06, "epoch": 0.6874495017506059, "percentage": 34.37, "elapsed_time": "21:55:27", "remaining_time": "1 day, 17:51:36", "throughput": 5661.8, "total_tokens": 446871712}
+{"current_steps": 10215, "total_steps": 29704, "loss": 0.38349783420562744, "lr": 7.355467827939794e-06, "epoch": 0.6877861567465661, "percentage": 34.39, "elapsed_time": "21:56:07", "remaining_time": "1 day, 17:51:01", "throughput": 5661.57, "total_tokens": 447082192}
+{"current_steps": 10220, "total_steps": 29704, "loss": 0.3912900686264038, "lr": 7.353135198129382e-06, "epoch": 0.6881228117425262, "percentage": 34.41, "elapsed_time": "21:56:42", "remaining_time": "1 day, 17:50:15", "throughput": 5661.71, "total_tokens": 447291168}
+{"current_steps": 10225, "total_steps": 29704, "loss": 0.38165950775146484, "lr": 7.350801910272216e-06, "epoch": 0.6884594667384863, "percentage": 34.42, "elapsed_time": "21:57:26", "remaining_time": "1 day, 17:49:46", "throughput": 5661.37, "total_tokens": 447512392}
+{"current_steps": 10230, "total_steps": 29704, "loss": 0.34427103996276853, "lr": 7.348467965020791e-06, "epoch": 0.6887961217344465, "percentage": 34.44, "elapsed_time": "21:58:02", "remaining_time": "1 day, 17:49:02", "throughput": 5661.59, "total_tokens": 447731304}
+{"current_steps": 10235, "total_steps": 29704, "loss": 0.3613192081451416, "lr": 7.3461333630277885e-06, "epoch": 0.6891327767304066, "percentage": 34.46, "elapsed_time": "21:58:37", "remaining_time": "1 day, 17:48:17", "throughput": 5661.49, "total_tokens": 447922416}
+{"current_steps": 10240, "total_steps": 29704, "loss": 0.3512859344482422, "lr": 7.343798104946074e-06, "epoch": 0.6894694317263668, "percentage": 34.47, "elapsed_time": "21:59:16", "remaining_time": "1 day, 17:47:38", "throughput": 5661.58, "total_tokens": 448148520}
+{"current_steps": 10245, "total_steps": 29704, "loss": 0.381636905670166, "lr": 7.341462191428692e-06, "epoch": 0.6898060867223269, "percentage": 34.49, "elapsed_time": "21:59:59", "remaining_time": "1 day, 17:47:08", "throughput": 5661.23, "total_tokens": 448364664}
+{"current_steps": 10250, "total_steps": 29704, "loss": 0.3913971662521362, "lr": 7.339125623128874e-06, "epoch": 0.690142741718287, "percentage": 34.51, "elapsed_time": "22:00:48", "remaining_time": "1 day, 17:46:50", "throughput": 5660.48, "total_tokens": 448587016}
+{"current_steps": 10255, "total_steps": 29704, "loss": 0.36610395908355714, "lr": 7.336788400700035e-06, "epoch": 0.6904793967142472, "percentage": 34.52, "elapsed_time": "22:01:32", "remaining_time": "1 day, 17:46:21", "throughput": 5660.03, "total_tokens": 448798728}
+{"current_steps": 10260, "total_steps": 29704, "loss": 0.408566951751709, "lr": 7.334450524795771e-06, "epoch": 0.6908160517102074, "percentage": 34.54, "elapsed_time": "22:02:13", "remaining_time": "1 day, 17:45:47", "throughput": 5659.93, "total_tokens": 449023728}
+{"current_steps": 10265, "total_steps": 29704, "loss": 0.3746421575546265, "lr": 7.332111996069863e-06, "epoch": 0.6911527067061676, "percentage": 34.56, "elapsed_time": "22:02:50", "remaining_time": "1 day, 17:45:05", "throughput": 5660.11, "total_tokens": 449245560}
+{"current_steps": 10270, "total_steps": 29704, "loss": 0.39997851848602295, "lr": 7.32977281517627e-06, "epoch": 0.6914893617021277, "percentage": 34.57, "elapsed_time": "22:03:26", "remaining_time": "1 day, 17:44:22", "throughput": 5660.05, "total_tokens": 449446632}
+{"current_steps": 10275, "total_steps": 29704, "loss": 0.3859497308731079, "lr": 7.3274329827691406e-06, "epoch": 0.6918260166980879, "percentage": 34.59, "elapsed_time": "22:04:09", "remaining_time": "1 day, 17:43:51", "throughput": 5659.68, "total_tokens": 449659944}
+{"current_steps": 10280, "total_steps": 29704, "loss": 0.35598082542419435, "lr": 7.325092499502796e-06, "epoch": 0.692162671694048, "percentage": 34.61, "elapsed_time": "22:04:49", "remaining_time": "1 day, 17:43:14", "throughput": 5659.77, "total_tokens": 449892656}
+{"current_steps": 10285, "total_steps": 29704, "loss": 0.3739547967910767, "lr": 7.322751366031751e-06, "epoch": 0.6924993266900081, "percentage": 34.62, "elapsed_time": "22:05:28", "remaining_time": "1 day, 17:42:37", "throughput": 5659.78, "total_tokens": 450115160}
+{"current_steps": 10290, "total_steps": 29704, "loss": 0.3593981027603149, "lr": 7.320409583010691e-06, "epoch": 0.6928359816859683, "percentage": 34.64, "elapsed_time": "22:06:07", "remaining_time": "1 day, 17:41:59", "throughput": 5659.83, "total_tokens": 450339952}
+{"current_steps": 10295, "total_steps": 29704, "loss": 0.4208518981933594, "lr": 7.318067151094493e-06, "epoch": 0.6931726366819284, "percentage": 34.66, "elapsed_time": "22:06:43", "remaining_time": "1 day, 17:41:14", "throughput": 5660.1, "total_tokens": 450562568}
+{"current_steps": 10300, "total_steps": 29704, "loss": 0.3863611936569214, "lr": 7.3157240709382046e-06, "epoch": 0.6935092916778886, "percentage": 34.68, "elapsed_time": "22:07:14", "remaining_time": "1 day, 17:40:22", "throughput": 5660.79, "total_tokens": 450794024}
+{"current_steps": 10305, "total_steps": 29704, "loss": 0.3639280080795288, "lr": 7.313380343197066e-06, "epoch": 0.6938459466738487, "percentage": 34.69, "elapsed_time": "22:07:54", "remaining_time": "1 day, 17:39:45", "throughput": 5660.91, "total_tokens": 451027856}
+{"current_steps": 10310, "total_steps": 29704, "loss": 0.3794968605041504, "lr": 7.311035968526492e-06, "epoch": 0.6941826016698088, "percentage": 34.71, "elapsed_time": "22:08:35", "remaining_time": "1 day, 17:39:11", "throughput": 5660.74, "total_tokens": 451247432}
+{"current_steps": 10315, "total_steps": 29704, "loss": 0.38078649044036866, "lr": 7.308690947582078e-06, "epoch": 0.694519256665769, "percentage": 34.73, "elapsed_time": "22:09:10", "remaining_time": "1 day, 17:38:26", "throughput": 5660.87, "total_tokens": 451456696}
+{"current_steps": 10320, "total_steps": 29704, "loss": 0.37413511276245115, "lr": 7.306345281019602e-06, "epoch": 0.6948559116617291, "percentage": 34.74, "elapsed_time": "22:09:50", "remaining_time": "1 day, 17:37:49", "throughput": 5660.82, "total_tokens": 451677704}
+{"current_steps": 10325, "total_steps": 29704, "loss": 0.40517454147338866, "lr": 7.303998969495023e-06, "epoch": 0.6951925666576892, "percentage": 34.76, "elapsed_time": "22:10:32", "remaining_time": "1 day, 17:37:16", "throughput": 5660.54, "total_tokens": 451892384}
+{"current_steps": 10330, "total_steps": 29704, "loss": 0.39863314628601076, "lr": 7.301652013664479e-06, "epoch": 0.6955292216536494, "percentage": 34.78, "elapsed_time": "22:11:08", "remaining_time": "1 day, 17:36:34", "throughput": 5660.77, "total_tokens": 452118752}
+{"current_steps": 10335, "total_steps": 29704, "loss": 0.40618147850036623, "lr": 7.299304414184288e-06, "epoch": 0.6958658766496095, "percentage": 34.79, "elapsed_time": "22:11:49", "remaining_time": "1 day, 17:36:00", "throughput": 5660.62, "total_tokens": 452338552}
+{"current_steps": 10340, "total_steps": 29704, "loss": 0.3879892587661743, "lr": 7.296956171710951e-06, "epoch": 0.6962025316455697, "percentage": 34.81, "elapsed_time": "22:12:26", "remaining_time": "1 day, 17:35:17", "throughput": 5660.97, "total_tokens": 452573592}
+{"current_steps": 10345, "total_steps": 29704, "loss": 0.3926121711730957, "lr": 7.294607286901146e-06, "epoch": 0.6965391866415298, "percentage": 34.83, "elapsed_time": "22:13:07", "remaining_time": "1 day, 17:34:44", "throughput": 5660.82, "total_tokens": 452795696}
+{"current_steps": 10350, "total_steps": 29704, "loss": 0.3886493444442749, "lr": 7.292257760411726e-06, "epoch": 0.69687584163749, "percentage": 34.84, "elapsed_time": "22:13:43", "remaining_time": "1 day, 17:34:00", "throughput": 5660.94, "total_tokens": 453009328}
+{"current_steps": 10355, "total_steps": 29704, "loss": 0.37845487594604493, "lr": 7.289907592899737e-06, "epoch": 0.6972124966334501, "percentage": 34.86, "elapsed_time": "22:14:22", "remaining_time": "1 day, 17:33:22", "throughput": 5660.73, "total_tokens": 453214464}
+{"current_steps": 10360, "total_steps": 29704, "loss": 0.3894767761230469, "lr": 7.287556785022389e-06, "epoch": 0.6975491516294102, "percentage": 34.88, "elapsed_time": "22:15:02", "remaining_time": "1 day, 17:32:46", "throughput": 5660.68, "total_tokens": 453437368}
+{"current_steps": 10365, "total_steps": 29704, "loss": 0.3892651557922363, "lr": 7.285205337437082e-06, "epoch": 0.6978858066253704, "percentage": 34.89, "elapsed_time": "22:15:41", "remaining_time": "1 day, 17:32:07", "throughput": 5660.61, "total_tokens": 453647680}
+{"current_steps": 10370, "total_steps": 29704, "loss": 0.399712347984314, "lr": 7.2828532508013895e-06, "epoch": 0.6982224616213305, "percentage": 34.91, "elapsed_time": "22:16:18", "remaining_time": "1 day, 17:31:25", "throughput": 5660.74, "total_tokens": 453867592}
+{"current_steps": 10375, "total_steps": 29704, "loss": 0.3772284030914307, "lr": 7.2805005257730625e-06, "epoch": 0.6985591166172906, "percentage": 34.93, "elapsed_time": "22:16:51", "remaining_time": "1 day, 17:30:36", "throughput": 5661.14, "total_tokens": 454087008}
+{"current_steps": 10380, "total_steps": 29704, "loss": 0.36341052055358886, "lr": 7.278147163010037e-06, "epoch": 0.6988957716132508, "percentage": 34.94, "elapsed_time": "22:17:30", "remaining_time": "1 day, 17:29:59", "throughput": 5661.17, "total_tokens": 454313224}
+{"current_steps": 10385, "total_steps": 29704, "loss": 0.36755235195159913, "lr": 7.275793163170423e-06, "epoch": 0.6992324266092109, "percentage": 34.96, "elapsed_time": "22:18:08", "remaining_time": "1 day, 17:29:18", "throughput": 5661.36, "total_tokens": 454539992}
+{"current_steps": 10390, "total_steps": 29704, "loss": 0.3956979990005493, "lr": 7.273438526912506e-06, "epoch": 0.6995690816051711, "percentage": 34.98, "elapsed_time": "22:18:45", "remaining_time": "1 day, 17:28:36", "throughput": 5661.36, "total_tokens": 454748824}
+{"current_steps": 10395, "total_steps": 29704, "loss": 0.36125593185424804, "lr": 7.271083254894757e-06, "epoch": 0.6999057366011312, "percentage": 35.0, "elapsed_time": "22:19:24", "remaining_time": "1 day, 17:27:59", "throughput": 5661.23, "total_tokens": 454962456}
+{"current_steps": 10400, "total_steps": 29704, "loss": 0.3766735076904297, "lr": 7.268727347775815e-06, "epoch": 0.7002423915970913, "percentage": 35.01, "elapsed_time": "22:19:58", "remaining_time": "1 day, 17:27:11", "throughput": 5661.68, "total_tokens": 455189544}
+{"current_steps": 10405, "total_steps": 29704, "loss": 0.32742795944213865, "lr": 7.266370806214506e-06, "epoch": 0.7005790465930515, "percentage": 35.03, "elapsed_time": "22:20:46", "remaining_time": "1 day, 17:26:50", "throughput": 5661.03, "total_tokens": 455409928}
+{"current_steps": 10410, "total_steps": 29704, "loss": 0.37685844898223875, "lr": 7.264013630869827e-06, "epoch": 0.7009157015890116, "percentage": 35.05, "elapsed_time": "22:21:31", "remaining_time": "1 day, 17:26:23", "throughput": 5660.57, "total_tokens": 455626104}
+{"current_steps": 10415, "total_steps": 29704, "loss": 0.39275636672973635, "lr": 7.261655822400957e-06, "epoch": 0.7012523565849718, "percentage": 35.06, "elapsed_time": "22:22:06", "remaining_time": "1 day, 17:25:37", "throughput": 5660.99, "total_tokens": 455857232}
+{"current_steps": 10420, "total_steps": 29704, "loss": 0.38013458251953125, "lr": 7.259297381467248e-06, "epoch": 0.7015890115809319, "percentage": 35.08, "elapsed_time": "22:22:49", "remaining_time": "1 day, 17:25:07", "throughput": 5660.68, "total_tokens": 456077832}
+{"current_steps": 10425, "total_steps": 29704, "loss": 0.34854826927185056, "lr": 7.2569383087282285e-06, "epoch": 0.701925666576892, "percentage": 35.1, "elapsed_time": "22:23:31", "remaining_time": "1 day, 17:24:34", "throughput": 5660.17, "total_tokens": 456272960}
+{"current_steps": 10430, "total_steps": 29704, "loss": 0.39771556854248047, "lr": 7.25457860484361e-06, "epoch": 0.7022623215728522, "percentage": 35.11, "elapsed_time": "22:24:11", "remaining_time": "1 day, 17:23:59", "throughput": 5660.05, "total_tokens": 456493600}
+{"current_steps": 10435, "total_steps": 29704, "loss": 0.3760495662689209, "lr": 7.252218270473274e-06, "epoch": 0.7025989765688123, "percentage": 35.13, "elapsed_time": "22:24:52", "remaining_time": "1 day, 17:23:23", "throughput": 5659.85, "total_tokens": 456704912}
+{"current_steps": 10440, "total_steps": 29704, "loss": 0.369310188293457, "lr": 7.249857306277279e-06, "epoch": 0.7029356315647725, "percentage": 35.15, "elapsed_time": "22:25:26", "remaining_time": "1 day, 17:22:37", "throughput": 5660.11, "total_tokens": 456921864}
+{"current_steps": 10445, "total_steps": 29704, "loss": 0.35996944904327394, "lr": 7.247495712915864e-06, "epoch": 0.7032722865607326, "percentage": 35.16, "elapsed_time": "22:26:00", "remaining_time": "1 day, 17:21:50", "throughput": 5660.47, "total_tokens": 457143032}
+{"current_steps": 10450, "total_steps": 29704, "loss": 0.3736717939376831, "lr": 7.245133491049439e-06, "epoch": 0.7036089415566927, "percentage": 35.18, "elapsed_time": "22:26:39", "remaining_time": "1 day, 17:21:12", "throughput": 5660.45, "total_tokens": 457363400}
+{"current_steps": 10455, "total_steps": 29704, "loss": 0.4129630088806152, "lr": 7.242770641338592e-06, "epoch": 0.7039455965526529, "percentage": 35.2, "elapsed_time": "22:27:21", "remaining_time": "1 day, 17:20:38", "throughput": 5660.29, "total_tokens": 457584008}
+{"current_steps": 10460, "total_steps": 29704, "loss": 0.36982312202453616, "lr": 7.240407164444088e-06, "epoch": 0.704282251548613, "percentage": 35.21, "elapsed_time": "22:27:59", "remaining_time": "1 day, 17:19:59", "throughput": 5660.25, "total_tokens": 457798528}
+{"current_steps": 10465, "total_steps": 29704, "loss": 0.34938392639160154, "lr": 7.238043061026862e-06, "epoch": 0.7046189065445732, "percentage": 35.23, "elapsed_time": "22:28:41", "remaining_time": "1 day, 17:19:26", "throughput": 5659.9, "total_tokens": 458006080}
+{"current_steps": 10470, "total_steps": 29704, "loss": 0.3546873092651367, "lr": 7.235678331748034e-06, "epoch": 0.7049555615405333, "percentage": 35.25, "elapsed_time": "22:29:16", "remaining_time": "1 day, 17:18:42", "throughput": 5659.86, "total_tokens": 458203208}
+{"current_steps": 10475, "total_steps": 29704, "loss": 0.40183773040771487, "lr": 7.233312977268887e-06, "epoch": 0.7052922165364934, "percentage": 35.26, "elapsed_time": "22:29:53", "remaining_time": "1 day, 17:18:00", "throughput": 5660.06, "total_tokens": 458427952}
+{"current_steps": 10480, "total_steps": 29704, "loss": 0.36978516578674314, "lr": 7.230946998250889e-06, "epoch": 0.7056288715324536, "percentage": 35.28, "elapsed_time": "22:30:28", "remaining_time": "1 day, 17:17:15", "throughput": 5660.11, "total_tokens": 458631504}
+{"current_steps": 10485, "total_steps": 29704, "loss": 0.34577772617340086, "lr": 7.228580395355676e-06, "epoch": 0.7059655265284137, "percentage": 35.3, "elapsed_time": "22:31:12", "remaining_time": "1 day, 17:16:45", "throughput": 5659.71, "total_tokens": 458845056}
+{"current_steps": 10490, "total_steps": 29704, "loss": 0.38037862777709963, "lr": 7.226213169245062e-06, "epoch": 0.7063021815243739, "percentage": 35.32, "elapsed_time": "22:31:53", "remaining_time": "1 day, 17:16:10", "throughput": 5659.56, "total_tokens": 459063872}
+{"current_steps": 10495, "total_steps": 29704, "loss": 0.3649828672409058, "lr": 7.223845320581035e-06, "epoch": 0.706638836520334, "percentage": 35.33, "elapsed_time": "22:32:38", "remaining_time": "1 day, 17:15:43", "throughput": 5659.04, "total_tokens": 459276768}
+{"current_steps": 10500, "total_steps": 29704, "loss": 0.3691579818725586, "lr": 7.221476850025754e-06, "epoch": 0.7069754915162941, "percentage": 35.35, "elapsed_time": "22:33:17", "remaining_time": "1 day, 17:15:06", "throughput": 5659.1, "total_tokens": 459504320}
+{"current_steps": 10505, "total_steps": 29704, "loss": 0.381423807144165, "lr": 7.219107758241557e-06, "epoch": 0.7073121465122543, "percentage": 35.37, "elapsed_time": "22:33:59", "remaining_time": "1 day, 17:14:33", "throughput": 5658.92, "total_tokens": 459726544}
+{"current_steps": 10510, "total_steps": 29704, "loss": 0.3745572090148926, "lr": 7.2167380458909505e-06, "epoch": 0.7076488015082144, "percentage": 35.38, "elapsed_time": "22:34:34", "remaining_time": "1 day, 17:13:48", "throughput": 5659.22, "total_tokens": 459949656}
+{"current_steps": 10515, "total_steps": 29704, "loss": 0.375703763961792, "lr": 7.214367713636619e-06, "epoch": 0.7079854565041745, "percentage": 35.4, "elapsed_time": "22:35:12", "remaining_time": "1 day, 17:13:09", "throughput": 5659.28, "total_tokens": 460172128}
+{"current_steps": 10520, "total_steps": 29704, "loss": 0.37557039260864256, "lr": 7.211996762141418e-06, "epoch": 0.7083221115001347, "percentage": 35.42, "elapsed_time": "22:35:45", "remaining_time": "1 day, 17:12:20", "throughput": 5659.66, "total_tokens": 460390848}
+{"current_steps": 10525, "total_steps": 29704, "loss": 0.4028931617736816, "lr": 7.209625192068374e-06, "epoch": 0.7086587664960948, "percentage": 35.43, "elapsed_time": "22:36:19", "remaining_time": "1 day, 17:11:33", "throughput": 5660.02, "total_tokens": 460611832}
+{"current_steps": 10530, "total_steps": 29704, "loss": 0.37696285247802735, "lr": 7.207253004080692e-06, "epoch": 0.708995421492055, "percentage": 35.45, "elapsed_time": "22:36:58", "remaining_time": "1 day, 17:10:54", "throughput": 5660.19, "total_tokens": 460844344}
+{"current_steps": 10535, "total_steps": 29704, "loss": 0.34694790840148926, "lr": 7.204880198841746e-06, "epoch": 0.7093320764880151, "percentage": 35.47, "elapsed_time": "22:37:33", "remaining_time": "1 day, 17:10:09", "throughput": 5660.37, "total_tokens": 461058504}
+{"current_steps": 10540, "total_steps": 29704, "loss": 0.4015054225921631, "lr": 7.2025067770150835e-06, "epoch": 0.7096687314839752, "percentage": 35.48, "elapsed_time": "22:38:13", "remaining_time": "1 day, 17:09:33", "throughput": 5660.38, "total_tokens": 461286472}
+{"current_steps": 10545, "total_steps": 29704, "loss": 0.3497929573059082, "lr": 7.2001327392644235e-06, "epoch": 0.7100053864799354, "percentage": 35.5, "elapsed_time": "22:38:48", "remaining_time": "1 day, 17:08:47", "throughput": 5660.8, "total_tokens": 461515496}
+{"current_steps": 10550, "total_steps": 29704, "loss": 0.39971365928649905, "lr": 7.197758086253659e-06, "epoch": 0.7103420414758955, "percentage": 35.52, "elapsed_time": "22:39:28", "remaining_time": "1 day, 17:08:12", "throughput": 5660.72, "total_tokens": 461738504}
+{"current_steps": 10555, "total_steps": 29704, "loss": 0.39182677268981936, "lr": 7.195382818646854e-06, "epoch": 0.7106786964718557, "percentage": 35.53, "elapsed_time": "22:40:08", "remaining_time": "1 day, 17:07:34", "throughput": 5660.53, "total_tokens": 461947016}
+{"current_steps": 10560, "total_steps": 29704, "loss": 0.36522927284240725, "lr": 7.1930069371082445e-06, "epoch": 0.7110153514678158, "percentage": 35.55, "elapsed_time": "22:40:52", "remaining_time": "1 day, 17:07:06", "throughput": 5659.99, "total_tokens": 462152784}
+{"current_steps": 10565, "total_steps": 29704, "loss": 0.3833275079727173, "lr": 7.190630442302238e-06, "epoch": 0.7113520064637759, "percentage": 35.57, "elapsed_time": "22:41:34", "remaining_time": "1 day, 17:06:33", "throughput": 5659.87, "total_tokens": 462379128}
+{"current_steps": 10570, "total_steps": 29704, "loss": 0.3999313831329346, "lr": 7.188253334893414e-06, "epoch": 0.7116886614597361, "percentage": 35.58, "elapsed_time": "22:42:10", "remaining_time": "1 day, 17:05:49", "throughput": 5660.06, "total_tokens": 462598648}
+{"current_steps": 10575, "total_steps": 29704, "loss": 0.3910887956619263, "lr": 7.18587561554652e-06, "epoch": 0.7120253164556962, "percentage": 35.6, "elapsed_time": "22:42:51", "remaining_time": "1 day, 17:05:15", "throughput": 5659.77, "total_tokens": 462807208}
+{"current_steps": 10580, "total_steps": 29704, "loss": 0.36777448654174805, "lr": 7.183497284926483e-06, "epoch": 0.7123619714516564, "percentage": 35.62, "elapsed_time": "22:43:27", "remaining_time": "1 day, 17:04:31", "throughput": 5660.16, "total_tokens": 463042928}
+{"current_steps": 10585, "total_steps": 29704, "loss": 0.36260690689086916, "lr": 7.181118343698392e-06, "epoch": 0.7126986264476165, "percentage": 35.63, "elapsed_time": "22:44:11", "remaining_time": "1 day, 17:04:02", "throughput": 5659.69, "total_tokens": 463252424}
+{"current_steps": 10590, "total_steps": 29704, "loss": 0.39423255920410155, "lr": 7.178738792527512e-06, "epoch": 0.7130352814435766, "percentage": 35.65, "elapsed_time": "22:44:51", "remaining_time": "1 day, 17:03:25", "throughput": 5659.59, "total_tokens": 463469984}
+{"current_steps": 10595, "total_steps": 29704, "loss": 0.37251434326171873, "lr": 7.176358632079277e-06, "epoch": 0.7133719364395368, "percentage": 35.67, "elapsed_time": "22:45:29", "remaining_time": "1 day, 17:02:47", "throughput": 5659.52, "total_tokens": 463682208}
+{"current_steps": 10600, "total_steps": 29704, "loss": 0.37181549072265624, "lr": 7.173977863019288e-06, "epoch": 0.7137085914354969, "percentage": 35.69, "elapsed_time": "22:46:10", "remaining_time": "1 day, 17:02:11", "throughput": 5659.47, "total_tokens": 463907464}
+{"current_steps": 10605, "total_steps": 29704, "loss": 0.34790878295898436, "lr": 7.171596486013323e-06, "epoch": 0.7140452464314571, "percentage": 35.7, "elapsed_time": "22:46:47", "remaining_time": "1 day, 17:01:30", "throughput": 5659.71, "total_tokens": 464138256}
+{"current_steps": 10610, "total_steps": 29704, "loss": 0.37795867919921877, "lr": 7.169214501727326e-06, "epoch": 0.7143819014274172, "percentage": 35.72, "elapsed_time": "22:47:23", "remaining_time": "1 day, 17:00:46", "throughput": 5659.68, "total_tokens": 464338936}
+{"current_steps": 10615, "total_steps": 29704, "loss": 0.36943345069885253, "lr": 7.16683191082741e-06, "epoch": 0.7147185564233773, "percentage": 35.74, "elapsed_time": "22:47:59", "remaining_time": "1 day, 17:00:03", "throughput": 5659.91, "total_tokens": 464560144}
+{"current_steps": 10620, "total_steps": 29704, "loss": 0.368468713760376, "lr": 7.16444871397986e-06, "epoch": 0.7150552114193375, "percentage": 35.75, "elapsed_time": "22:48:37", "remaining_time": "1 day, 16:59:23", "throughput": 5659.77, "total_tokens": 464765496}
+{"current_steps": 10625, "total_steps": 29704, "loss": 0.36743860244750975, "lr": 7.162064911851129e-06, "epoch": 0.7153918664152976, "percentage": 35.77, "elapsed_time": "22:49:20", "remaining_time": "1 day, 16:58:53", "throughput": 5659.38, "total_tokens": 464978408}
+{"current_steps": 10630, "total_steps": 29704, "loss": 0.40234060287475587, "lr": 7.15968050510784e-06, "epoch": 0.7157285214112578, "percentage": 35.79, "elapsed_time": "22:49:58", "remaining_time": "1 day, 16:58:12", "throughput": 5659.41, "total_tokens": 465192704}
+{"current_steps": 10635, "total_steps": 29704, "loss": 0.3755232810974121, "lr": 7.1572954944167834e-06, "epoch": 0.7160651764072179, "percentage": 35.8, "elapsed_time": "22:50:39", "remaining_time": "1 day, 16:57:39", "throughput": 5659.18, "total_tokens": 465408616}
+{"current_steps": 10640, "total_steps": 29704, "loss": 0.3969583988189697, "lr": 7.154909880444922e-06, "epoch": 0.716401831403178, "percentage": 35.82, "elapsed_time": "22:51:22", "remaining_time": "1 day, 16:57:08", "throughput": 5658.94, "total_tokens": 465633232}
+{"current_steps": 10645, "total_steps": 29704, "loss": 0.37885403633117676, "lr": 7.152523663859384e-06, "epoch": 0.7167384863991382, "percentage": 35.84, "elapsed_time": "22:52:04", "remaining_time": "1 day, 16:56:36", "throughput": 5658.61, "total_tokens": 465845256}
+{"current_steps": 10650, "total_steps": 29704, "loss": 0.3639181613922119, "lr": 7.150136845327466e-06, "epoch": 0.7170751413950983, "percentage": 35.85, "elapsed_time": "22:52:49", "remaining_time": "1 day, 16:56:08", "throughput": 5658.1, "total_tokens": 466056584}
+{"current_steps": 10655, "total_steps": 29704, "loss": 0.3720954418182373, "lr": 7.147749425516637e-06, "epoch": 0.7174117963910585, "percentage": 35.87, "elapsed_time": "22:53:27", "remaining_time": "1 day, 16:55:27", "throughput": 5658.35, "total_tokens": 466287992}
+{"current_steps": 10660, "total_steps": 29704, "loss": 0.3578562021255493, "lr": 7.145361405094529e-06, "epoch": 0.7177484513870186, "percentage": 35.89, "elapsed_time": "22:53:59", "remaining_time": "1 day, 16:54:37", "throughput": 5658.99, "total_tokens": 466523176}
+{"current_steps": 10665, "total_steps": 29704, "loss": 0.3409886360168457, "lr": 7.142972784728945e-06, "epoch": 0.7180851063829787, "percentage": 35.9, "elapsed_time": "22:54:43", "remaining_time": "1 day, 16:54:07", "throughput": 5658.55, "total_tokens": 466734112}
+{"current_steps": 10670, "total_steps": 29704, "loss": 0.35494747161865237, "lr": 7.140583565087856e-06, "epoch": 0.7184217613789389, "percentage": 35.92, "elapsed_time": "22:55:19", "remaining_time": "1 day, 16:53:24", "throughput": 5658.71, "total_tokens": 466952312}
+{"current_steps": 10675, "total_steps": 29704, "loss": 0.3933510065078735, "lr": 7.138193746839399e-06, "epoch": 0.718758416374899, "percentage": 35.94, "elapsed_time": "22:56:05", "remaining_time": "1 day, 16:52:58", "throughput": 5658.38, "total_tokens": 467185248}
+{"current_steps": 10680, "total_steps": 29704, "loss": 0.37902536392211916, "lr": 7.135803330651878e-06, "epoch": 0.7190950713708592, "percentage": 35.95, "elapsed_time": "22:56:39", "remaining_time": "1 day, 16:52:12", "throughput": 5658.9, "total_tokens": 467421448}
+{"current_steps": 10685, "total_steps": 29704, "loss": 0.3649748802185059, "lr": 7.133412317193766e-06, "epoch": 0.7194317263668193, "percentage": 35.97, "elapsed_time": "22:57:27", "remaining_time": "1 day, 16:51:50", "throughput": 5658.32, "total_tokens": 467648088}
+{"current_steps": 10690, "total_steps": 29704, "loss": 0.39012324810028076, "lr": 7.131020707133703e-06, "epoch": 0.7197683813627794, "percentage": 35.99, "elapsed_time": "22:58:05", "remaining_time": "1 day, 16:51:09", "throughput": 5658.24, "total_tokens": 467852784}
+{"current_steps": 10695, "total_steps": 29704, "loss": 0.3994007587432861, "lr": 7.128628501140495e-06, "epoch": 0.7201050363587396, "percentage": 36.01, "elapsed_time": "22:58:39", "remaining_time": "1 day, 16:50:23", "throughput": 5658.43, "total_tokens": 468062360}
+{"current_steps": 10700, "total_steps": 29704, "loss": 0.3543556213378906, "lr": 7.126235699883115e-06, "epoch": 0.7204416913546997, "percentage": 36.02, "elapsed_time": "22:59:22", "remaining_time": "1 day, 16:49:51", "throughput": 5657.92, "total_tokens": 468260752}
+{"current_steps": 10705, "total_steps": 29704, "loss": 0.3659778594970703, "lr": 7.1238423040306995e-06, "epoch": 0.7207783463506598, "percentage": 36.04, "elapsed_time": "23:00:07", "remaining_time": "1 day, 16:49:24", "throughput": 5657.47, "total_tokens": 468480296}
+{"current_steps": 10710, "total_steps": 29704, "loss": 0.33471217155456545, "lr": 7.121448314252557e-06, "epoch": 0.72111500134662, "percentage": 36.06, "elapsed_time": "23:00:49", "remaining_time": "1 day, 16:48:51", "throughput": 5657.46, "total_tokens": 468714640}
+{"current_steps": 10715, "total_steps": 29704, "loss": 0.3590662479400635, "lr": 7.1190537312181586e-06, "epoch": 0.7214516563425801, "percentage": 36.07, "elapsed_time": "23:01:23", "remaining_time": "1 day, 16:48:05", "throughput": 5657.57, "total_tokens": 468920200}
+{"current_steps": 10720, "total_steps": 29704, "loss": 0.35546276569366453, "lr": 7.116658555597141e-06, "epoch": 0.7217883113385403, "percentage": 36.09, "elapsed_time": "23:02:00", "remaining_time": "1 day, 16:47:23", "throughput": 5657.28, "total_tokens": 469105088}
+{"current_steps": 10725, "total_steps": 29704, "loss": 0.36028232574462893, "lr": 7.114262788059308e-06, "epoch": 0.7221249663345004, "percentage": 36.11, "elapsed_time": "23:02:36", "remaining_time": "1 day, 16:46:39", "throughput": 5657.39, "total_tokens": 469316144}
+{"current_steps": 10730, "total_steps": 29704, "loss": 0.3734895706176758, "lr": 7.111866429274627e-06, "epoch": 0.7224616213304605, "percentage": 36.12, "elapsed_time": "23:03:15", "remaining_time": "1 day, 16:46:02", "throughput": 5657.3, "total_tokens": 469533344}
+{"current_steps": 10735, "total_steps": 29704, "loss": 0.38394627571105955, "lr": 7.1094694799132356e-06, "epoch": 0.7227982763264207, "percentage": 36.14, "elapsed_time": "23:03:59", "remaining_time": "1 day, 16:45:32", "throughput": 5657.04, "total_tokens": 469757776}
+{"current_steps": 10740, "total_steps": 29704, "loss": 0.355570125579834, "lr": 7.10707194064543e-06, "epoch": 0.7231349313223808, "percentage": 36.16, "elapsed_time": "23:04:42", "remaining_time": "1 day, 16:45:01", "throughput": 5656.7, "total_tokens": 469973048}
+{"current_steps": 10745, "total_steps": 29704, "loss": 0.4048286437988281, "lr": 7.104673812141676e-06, "epoch": 0.723471586318341, "percentage": 36.17, "elapsed_time": "23:05:23", "remaining_time": "1 day, 16:44:27", "throughput": 5656.64, "total_tokens": 470200352}
+{"current_steps": 10750, "total_steps": 29704, "loss": 0.3921640157699585, "lr": 7.102275095072601e-06, "epoch": 0.7238082413143011, "percentage": 36.19, "elapsed_time": "23:06:10", "remaining_time": "1 day, 16:44:02", "throughput": 5656.29, "total_tokens": 470434416}
+{"current_steps": 10755, "total_steps": 29704, "loss": 0.36566712856292727, "lr": 7.099875790108998e-06, "epoch": 0.7241448963102612, "percentage": 36.21, "elapsed_time": "23:06:59", "remaining_time": "1 day, 16:43:43", "throughput": 5655.65, "total_tokens": 470662208}
+{"current_steps": 10760, "total_steps": 29704, "loss": 0.37137765884399415, "lr": 7.097475897921827e-06, "epoch": 0.7244815513062214, "percentage": 36.22, "elapsed_time": "23:07:39", "remaining_time": "1 day, 16:43:06", "throughput": 5655.52, "total_tokens": 470876192}
+{"current_steps": 10765, "total_steps": 29704, "loss": 0.3979093790054321, "lr": 7.09507541918221e-06, "epoch": 0.7248182063021815, "percentage": 36.24, "elapsed_time": "23:08:18", "remaining_time": "1 day, 16:42:27", "throughput": 5655.49, "total_tokens": 471091920}
+{"current_steps": 10770, "total_steps": 29704, "loss": 0.3879514217376709, "lr": 7.092674354561433e-06, "epoch": 0.7251548612981417, "percentage": 36.26, "elapsed_time": "23:08:59", "remaining_time": "1 day, 16:41:52", "throughput": 5655.41, "total_tokens": 471316336}
+{"current_steps": 10775, "total_steps": 29704, "loss": 0.3511503219604492, "lr": 7.090272704730945e-06, "epoch": 0.7254915162941018, "percentage": 36.27, "elapsed_time": "23:09:35", "remaining_time": "1 day, 16:41:09", "throughput": 5655.37, "total_tokens": 471518800}
+{"current_steps": 10780, "total_steps": 29704, "loss": 0.35472354888916013, "lr": 7.08787047036236e-06, "epoch": 0.7258281712900619, "percentage": 36.29, "elapsed_time": "23:10:18", "remaining_time": "1 day, 16:40:38", "throughput": 5655.31, "total_tokens": 471754616}
+{"current_steps": 10785, "total_steps": 29704, "loss": 0.3795314788818359, "lr": 7.085467652127457e-06, "epoch": 0.7261648262860221, "percentage": 36.31, "elapsed_time": "23:10:50", "remaining_time": "1 day, 16:39:47", "throughput": 5655.86, "total_tokens": 471981872}
+{"current_steps": 10790, "total_steps": 29704, "loss": 0.3529500961303711, "lr": 7.083064250698175e-06, "epoch": 0.7265014812819822, "percentage": 36.33, "elapsed_time": "23:11:31", "remaining_time": "1 day, 16:39:14", "throughput": 5655.8, "total_tokens": 472213144}
+{"current_steps": 10795, "total_steps": 29704, "loss": 0.37914767265319826, "lr": 7.0806602667466175e-06, "epoch": 0.7268381362779424, "percentage": 36.34, "elapsed_time": "23:12:14", "remaining_time": "1 day, 16:38:43", "throughput": 5655.66, "total_tokens": 472443584}
+{"current_steps": 10800, "total_steps": 29704, "loss": 0.38105590343475343, "lr": 7.078255700945051e-06, "epoch": 0.7271747912739025, "percentage": 36.36, "elapsed_time": "23:12:46", "remaining_time": "1 day, 16:37:53", "throughput": 5656.01, "total_tokens": 472655528}
+{"current_steps": 10805, "total_steps": 29704, "loss": 0.366654634475708, "lr": 7.075850553965904e-06, "epoch": 0.7275114462698626, "percentage": 36.38, "elapsed_time": "23:13:25", "remaining_time": "1 day, 16:37:15", "throughput": 5655.86, "total_tokens": 472863848}
+{"current_steps": 10810, "total_steps": 29704, "loss": 0.3818976879119873, "lr": 7.07344482648177e-06, "epoch": 0.7278481012658228, "percentage": 36.39, "elapsed_time": "23:13:57", "remaining_time": "1 day, 16:36:24", "throughput": 5656.37, "total_tokens": 473087584}
+{"current_steps": 10815, "total_steps": 29704, "loss": 0.3804640293121338, "lr": 7.071038519165402e-06, "epoch": 0.7281847562617829, "percentage": 36.41, "elapsed_time": "23:14:40", "remaining_time": "1 day, 16:35:52", "throughput": 5656.05, "total_tokens": 473302440}
+{"current_steps": 10820, "total_steps": 29704, "loss": 0.3819848299026489, "lr": 7.068631632689716e-06, "epoch": 0.7285214112577431, "percentage": 36.43, "elapsed_time": "23:15:19", "remaining_time": "1 day, 16:35:14", "throughput": 5656.14, "total_tokens": 473528888}
+{"current_steps": 10825, "total_steps": 29704, "loss": 0.3868781566619873, "lr": 7.066224167727791e-06, "epoch": 0.7288580662537032, "percentage": 36.44, "elapsed_time": "23:15:53", "remaining_time": "1 day, 16:34:28", "throughput": 5656.45, "total_tokens": 473749232}
+{"current_steps": 10830, "total_steps": 29704, "loss": 0.36863956451416013, "lr": 7.063816124952865e-06, "epoch": 0.7291947212496633, "percentage": 36.46, "elapsed_time": "23:16:31", "remaining_time": "1 day, 16:33:48", "throughput": 5656.48, "total_tokens": 473966688}
+{"current_steps": 10835, "total_steps": 29704, "loss": 0.3608518362045288, "lr": 7.061407505038341e-06, "epoch": 0.7295313762456235, "percentage": 36.48, "elapsed_time": "23:17:10", "remaining_time": "1 day, 16:33:10", "throughput": 5656.44, "total_tokens": 474183856}
+{"current_steps": 10840, "total_steps": 29704, "loss": 0.3866151809692383, "lr": 7.058998308657782e-06, "epoch": 0.7298680312415836, "percentage": 36.49, "elapsed_time": "23:17:43", "remaining_time": "1 day, 16:32:20", "throughput": 5657.01, "total_tokens": 474414672}
+{"current_steps": 10845, "total_steps": 29704, "loss": 0.35731284618377684, "lr": 7.056588536484912e-06, "epoch": 0.7302046862375438, "percentage": 36.51, "elapsed_time": "23:18:12", "remaining_time": "1 day, 16:31:26", "throughput": 5657.49, "total_tokens": 474622840}
+{"current_steps": 10850, "total_steps": 29704, "loss": 0.35231194496154783, "lr": 7.0541781891936145e-06, "epoch": 0.7305413412335039, "percentage": 36.53, "elapsed_time": "23:18:51", "remaining_time": "1 day, 16:30:46", "throughput": 5657.53, "total_tokens": 474842808}
+{"current_steps": 10855, "total_steps": 29704, "loss": 0.3739452838897705, "lr": 7.0517672674579364e-06, "epoch": 0.730877996229464, "percentage": 36.54, "elapsed_time": "23:19:31", "remaining_time": "1 day, 16:30:10", "throughput": 5657.41, "total_tokens": 475060040}
+{"current_steps": 10860, "total_steps": 29704, "loss": 0.367911171913147, "lr": 7.049355771952087e-06, "epoch": 0.7312146512254242, "percentage": 36.56, "elapsed_time": "23:20:12", "remaining_time": "1 day, 16:29:36", "throughput": 5657.14, "total_tokens": 475270960}
+{"current_steps": 10865, "total_steps": 29704, "loss": 0.3706895112991333, "lr": 7.046943703350429e-06, "epoch": 0.7315513062213843, "percentage": 36.58, "elapsed_time": "23:20:56", "remaining_time": "1 day, 16:29:06", "throughput": 5656.9, "total_tokens": 475496232}
+{"current_steps": 10870, "total_steps": 29704, "loss": 0.356654691696167, "lr": 7.044531062327492e-06, "epoch": 0.7318879612173445, "percentage": 36.59, "elapsed_time": "23:21:29", "remaining_time": "1 day, 16:28:18", "throughput": 5657.18, "total_tokens": 475709832}
+{"current_steps": 10875, "total_steps": 29704, "loss": 0.3708888053894043, "lr": 7.042117849557963e-06, "epoch": 0.7322246162133046, "percentage": 36.61, "elapsed_time": "23:22:13", "remaining_time": "1 day, 16:27:49", "throughput": 5656.86, "total_tokens": 475932072}
+{"current_steps": 10880, "total_steps": 29704, "loss": 0.36887059211730955, "lr": 7.0397040657166885e-06, "epoch": 0.7325612712092647, "percentage": 36.63, "elapsed_time": "23:22:49", "remaining_time": "1 day, 16:27:05", "throughput": 5657.21, "total_tokens": 476162776}
+{"current_steps": 10885, "total_steps": 29704, "loss": 0.34908170700073243, "lr": 7.037289711478676e-06, "epoch": 0.7328979262052249, "percentage": 36.64, "elapsed_time": "23:23:24", "remaining_time": "1 day, 16:26:20", "throughput": 5657.38, "total_tokens": 476376384}
+{"current_steps": 10890, "total_steps": 29704, "loss": 0.3759069204330444, "lr": 7.034874787519095e-06, "epoch": 0.733234581201185, "percentage": 36.66, "elapsed_time": "23:24:02", "remaining_time": "1 day, 16:25:40", "throughput": 5657.49, "total_tokens": 476599400}
+{"current_steps": 10895, "total_steps": 29704, "loss": 0.4069540023803711, "lr": 7.0324592945132655e-06, "epoch": 0.7335712361971451, "percentage": 36.68, "elapsed_time": "23:24:36", "remaining_time": "1 day, 16:24:54", "throughput": 5657.81, "total_tokens": 476821304}
+{"current_steps": 10900, "total_steps": 29704, "loss": 0.3802814245223999, "lr": 7.0300432331366764e-06, "epoch": 0.7339078911931053, "percentage": 36.7, "elapsed_time": "23:25:09", "remaining_time": "1 day, 16:24:05", "throughput": 5658.14, "total_tokens": 477036424}
+{"current_steps": 10905, "total_steps": 29704, "loss": 0.3741856813430786, "lr": 7.02762660406497e-06, "epoch": 0.7342445461890654, "percentage": 36.71, "elapsed_time": "23:25:48", "remaining_time": "1 day, 16:23:26", "throughput": 5658.19, "total_tokens": 477257288}
+{"current_steps": 10910, "total_steps": 29704, "loss": 0.3721340179443359, "lr": 7.02520940797395e-06, "epoch": 0.7345812011850256, "percentage": 36.73, "elapsed_time": "23:26:25", "remaining_time": "1 day, 16:22:46", "throughput": 5658.12, "total_tokens": 477464672}
+{"current_steps": 10915, "total_steps": 29704, "loss": 0.37077898979187013, "lr": 7.022791645539576e-06, "epoch": 0.7349178561809857, "percentage": 36.75, "elapsed_time": "23:27:03", "remaining_time": "1 day, 16:22:05", "throughput": 5658.04, "total_tokens": 477670600}
+{"current_steps": 10920, "total_steps": 29704, "loss": 0.4068927764892578, "lr": 7.02037331743797e-06, "epoch": 0.7352545111769458, "percentage": 36.76, "elapsed_time": "23:27:41", "remaining_time": "1 day, 16:21:25", "throughput": 5658.1, "total_tokens": 477891112}
+{"current_steps": 10925, "total_steps": 29704, "loss": 0.3525757551193237, "lr": 7.0179544243454076e-06, "epoch": 0.735591166172906, "percentage": 36.78, "elapsed_time": "23:28:13", "remaining_time": "1 day, 16:20:35", "throughput": 5658.68, "total_tokens": 478120960}
+{"current_steps": 10930, "total_steps": 29704, "loss": 0.36997184753417967, "lr": 7.0155349669383245e-06, "epoch": 0.7359278211688661, "percentage": 36.8, "elapsed_time": "23:28:54", "remaining_time": "1 day, 16:20:01", "throughput": 5658.57, "total_tokens": 478344240}
+{"current_steps": 10935, "total_steps": 29704, "loss": 0.36500933170318606, "lr": 7.013114945893316e-06, "epoch": 0.7362644761648263, "percentage": 36.81, "elapsed_time": "23:29:39", "remaining_time": "1 day, 16:19:33", "throughput": 5658.18, "total_tokens": 478567880}
+{"current_steps": 10940, "total_steps": 29704, "loss": 0.3656963109970093, "lr": 7.010694361887131e-06, "epoch": 0.7366011311607864, "percentage": 36.83, "elapsed_time": "23:30:24", "remaining_time": "1 day, 16:19:05", "throughput": 5657.91, "total_tokens": 478795792}
+{"current_steps": 10945, "total_steps": 29704, "loss": 0.3773782253265381, "lr": 7.008273215596681e-06, "epoch": 0.7369377861567465, "percentage": 36.85, "elapsed_time": "23:31:01", "remaining_time": "1 day, 16:18:24", "throughput": 5657.98, "total_tokens": 479014424}
+{"current_steps": 10950, "total_steps": 29704, "loss": 0.3759654760360718, "lr": 7.005851507699029e-06, "epoch": 0.7372744411527067, "percentage": 36.86, "elapsed_time": "23:31:34", "remaining_time": "1 day, 16:17:36", "throughput": 5658.48, "total_tokens": 479243632}
+{"current_steps": 10955, "total_steps": 29704, "loss": 0.3657572031021118, "lr": 7.003429238871399e-06, "epoch": 0.7376110961486668, "percentage": 36.88, "elapsed_time": "23:32:18", "remaining_time": "1 day, 16:17:05", "throughput": 5658.29, "total_tokens": 479473192}
+{"current_steps": 10960, "total_steps": 29704, "loss": 0.40654420852661133, "lr": 7.001006409791171e-06, "epoch": 0.737947751144627, "percentage": 36.9, "elapsed_time": "23:32:56", "remaining_time": "1 day, 16:16:25", "throughput": 5658.3, "total_tokens": 479687920}
+{"current_steps": 10965, "total_steps": 29704, "loss": 0.3780509948730469, "lr": 6.998583021135882e-06, "epoch": 0.7382844061405871, "percentage": 36.91, "elapsed_time": "23:33:35", "remaining_time": "1 day, 16:15:47", "throughput": 5658.25, "total_tokens": 479906096}
+{"current_steps": 10970, "total_steps": 29704, "loss": 0.36737675666809083, "lr": 6.996159073583224e-06, "epoch": 0.7386210611365472, "percentage": 36.93, "elapsed_time": "23:34:12", "remaining_time": "1 day, 16:15:07", "throughput": 5658.32, "total_tokens": 480125696}
+{"current_steps": 10975, "total_steps": 29704, "loss": 0.34843399524688723, "lr": 6.993734567811046e-06, "epoch": 0.7389577161325074, "percentage": 36.95, "elapsed_time": "23:34:56", "remaining_time": "1 day, 16:14:36", "throughput": 5658.01, "total_tokens": 480343296}
+{"current_steps": 10980, "total_steps": 29704, "loss": 0.37664797306060793, "lr": 6.991309504497352e-06, "epoch": 0.7392943711284675, "percentage": 36.96, "elapsed_time": "23:35:34", "remaining_time": "1 day, 16:13:56", "throughput": 5658.12, "total_tokens": 480566544}
+{"current_steps": 10985, "total_steps": 29704, "loss": 0.3457742214202881, "lr": 6.988883884320307e-06, "epoch": 0.7396310261244277, "percentage": 36.98, "elapsed_time": "23:36:10", "remaining_time": "1 day, 16:13:13", "throughput": 5658.13, "total_tokens": 480774232}
+{"current_steps": 10990, "total_steps": 29704, "loss": 0.39705471992492675, "lr": 6.9864577079582254e-06, "epoch": 0.7399676811203878, "percentage": 37.0, "elapsed_time": "23:36:48", "remaining_time": "1 day, 16:12:34", "throughput": 5658.25, "total_tokens": 481000200}
+{"current_steps": 10995, "total_steps": 29704, "loss": 0.3594042301177979, "lr": 6.984030976089581e-06, "epoch": 0.7403043361163479, "percentage": 37.02, "elapsed_time": "23:37:23", "remaining_time": "1 day, 16:11:50", "throughput": 5658.48, "total_tokens": 481219544}
+{"current_steps": 11000, "total_steps": 29704, "loss": 0.3729727745056152, "lr": 6.981603689392999e-06, "epoch": 0.7406409911123081, "percentage": 37.03, "elapsed_time": "23:38:11", "remaining_time": "1 day, 16:11:27", "throughput": 5657.8, "total_tokens": 481433184}
+{"current_steps": 11005, "total_steps": 29704, "loss": 0.3587386131286621, "lr": 6.979175848547263e-06, "epoch": 0.7409776461082682, "percentage": 37.05, "elapsed_time": "23:38:53", "remaining_time": "1 day, 16:10:53", "throughput": 5657.57, "total_tokens": 481648384}
+{"current_steps": 11010, "total_steps": 29704, "loss": 0.36256041526794436, "lr": 6.976747454231316e-06, "epoch": 0.7413143011042284, "percentage": 37.07, "elapsed_time": "23:39:35", "remaining_time": "1 day, 16:10:20", "throughput": 5657.19, "total_tokens": 481853600}
+{"current_steps": 11015, "total_steps": 29704, "loss": 0.37796127796173096, "lr": 6.974318507124245e-06, "epoch": 0.7416509561001885, "percentage": 37.08, "elapsed_time": "23:40:11", "remaining_time": "1 day, 16:09:36", "throughput": 5657.32, "total_tokens": 482067848}
+{"current_steps": 11020, "total_steps": 29704, "loss": 0.38568458557128904, "lr": 6.971889007905299e-06, "epoch": 0.7419876110961486, "percentage": 37.1, "elapsed_time": "23:40:42", "remaining_time": "1 day, 16:08:45", "throughput": 5657.64, "total_tokens": 482272608}
+{"current_steps": 11025, "total_steps": 29704, "loss": 0.4131626605987549, "lr": 6.9694589572538795e-06, "epoch": 0.7423242660921088, "percentage": 37.12, "elapsed_time": "23:41:19", "remaining_time": "1 day, 16:08:03", "throughput": 5657.9, "total_tokens": 482500568}
+{"current_steps": 11030, "total_steps": 29704, "loss": 0.3677205085754395, "lr": 6.9670283558495436e-06, "epoch": 0.7426609210880689, "percentage": 37.13, "elapsed_time": "23:41:55", "remaining_time": "1 day, 16:07:21", "throughput": 5658.12, "total_tokens": 482727520}
+{"current_steps": 11035, "total_steps": 29704, "loss": 0.3734050035476685, "lr": 6.964597204371999e-06, "epoch": 0.742997576084029, "percentage": 37.15, "elapsed_time": "23:42:29", "remaining_time": "1 day, 16:06:35", "throughput": 5658.39, "total_tokens": 482943600}
+{"current_steps": 11040, "total_steps": 29704, "loss": 0.3419958591461182, "lr": 6.962165503501111e-06, "epoch": 0.7433342310799892, "percentage": 37.17, "elapsed_time": "23:43:12", "remaining_time": "1 day, 16:06:03", "throughput": 5658.13, "total_tokens": 483163040}
+{"current_steps": 11045, "total_steps": 29704, "loss": 0.39629015922546384, "lr": 6.959733253916899e-06, "epoch": 0.7436708860759493, "percentage": 37.18, "elapsed_time": "23:43:50", "remaining_time": "1 day, 16:05:22", "throughput": 5658.27, "total_tokens": 483387400}
+{"current_steps": 11050, "total_steps": 29704, "loss": 0.3637987613677979, "lr": 6.95730045629953e-06, "epoch": 0.7440075410719095, "percentage": 37.2, "elapsed_time": "23:44:29", "remaining_time": "1 day, 16:04:44", "throughput": 5658.01, "total_tokens": 483586168}
+{"current_steps": 11055, "total_steps": 29704, "loss": 0.3997496604919434, "lr": 6.954867111329329e-06, "epoch": 0.7443441960678696, "percentage": 37.22, "elapsed_time": "23:45:01", "remaining_time": "1 day, 16:03:54", "throughput": 5658.55, "total_tokens": 483812040}
+{"current_steps": 11060, "total_steps": 29704, "loss": 0.36485040187835693, "lr": 6.952433219686775e-06, "epoch": 0.7446808510638298, "percentage": 37.23, "elapsed_time": "23:45:44", "remaining_time": "1 day, 16:03:23", "throughput": 5658.27, "total_tokens": 484033200}
+{"current_steps": 11065, "total_steps": 29704, "loss": 0.38755149841308595, "lr": 6.949998782052497e-06, "epoch": 0.7450175060597899, "percentage": 37.25, "elapsed_time": "23:46:24", "remaining_time": "1 day, 16:02:46", "throughput": 5658.18, "total_tokens": 484251832}
+{"current_steps": 11070, "total_steps": 29704, "loss": 0.36263265609741213, "lr": 6.947563799107277e-06, "epoch": 0.74535416105575, "percentage": 37.27, "elapsed_time": "23:46:57", "remaining_time": "1 day, 16:01:59", "throughput": 5658.4, "total_tokens": 484460952}
+{"current_steps": 11075, "total_steps": 29704, "loss": 0.37139124870300294, "lr": 6.945128271532052e-06, "epoch": 0.7456908160517102, "percentage": 37.28, "elapsed_time": "23:47:38", "remaining_time": "1 day, 16:01:23", "throughput": 5658.49, "total_tokens": 484696168}
+{"current_steps": 11080, "total_steps": 29704, "loss": 0.37421250343322754, "lr": 6.942692200007908e-06, "epoch": 0.7460274710476703, "percentage": 37.3, "elapsed_time": "23:48:13", "remaining_time": "1 day, 16:00:39", "throughput": 5658.69, "total_tokens": 484914304}
+{"current_steps": 11085, "total_steps": 29704, "loss": 0.3868663787841797, "lr": 6.9402555852160845e-06, "epoch": 0.7463641260436304, "percentage": 37.32, "elapsed_time": "23:48:53", "remaining_time": "1 day, 16:00:03", "throughput": 5658.68, "total_tokens": 485140952}
+{"current_steps": 11090, "total_steps": 29704, "loss": 0.35773367881774903, "lr": 6.937818427837975e-06, "epoch": 0.7467007810395906, "percentage": 37.34, "elapsed_time": "23:49:35", "remaining_time": "1 day, 15:59:30", "throughput": 5658.53, "total_tokens": 485364088}
+{"current_steps": 11095, "total_steps": 29704, "loss": 0.40668773651123047, "lr": 6.93538072855512e-06, "epoch": 0.7470374360355507, "percentage": 37.35, "elapsed_time": "23:50:11", "remaining_time": "1 day, 15:58:46", "throughput": 5658.76, "total_tokens": 485585456}
+{"current_steps": 11100, "total_steps": 29704, "loss": 0.3412586212158203, "lr": 6.932942488049217e-06, "epoch": 0.7473740910315109, "percentage": 37.37, "elapsed_time": "23:50:48", "remaining_time": "1 day, 15:58:05", "throughput": 5658.78, "total_tokens": 485799760}
+{"current_steps": 11105, "total_steps": 29704, "loss": 0.39055523872375486, "lr": 6.930503707002108e-06, "epoch": 0.747710746027471, "percentage": 37.39, "elapsed_time": "23:51:33", "remaining_time": "1 day, 15:57:37", "throughput": 5658.36, "total_tokens": 486017000}
+{"current_steps": 11110, "total_steps": 29704, "loss": 0.3696824789047241, "lr": 6.928064386095796e-06, "epoch": 0.7480474010234311, "percentage": 37.4, "elapsed_time": "23:52:14", "remaining_time": "1 day, 15:57:02", "throughput": 5658.3, "total_tokens": 486243416}
+{"current_steps": 11115, "total_steps": 29704, "loss": 0.369446325302124, "lr": 6.9256245260124264e-06, "epoch": 0.7483840560193913, "percentage": 37.42, "elapsed_time": "23:52:46", "remaining_time": "1 day, 15:56:13", "throughput": 5658.8, "total_tokens": 486469728}
+{"current_steps": 11120, "total_steps": 29704, "loss": 0.3864800453186035, "lr": 6.923184127434298e-06, "epoch": 0.7487207110153514, "percentage": 37.44, "elapsed_time": "23:53:21", "remaining_time": "1 day, 15:55:28", "throughput": 5658.85, "total_tokens": 486670960}
+{"current_steps": 11125, "total_steps": 29704, "loss": 0.35923840999603274, "lr": 6.920743191043861e-06, "epoch": 0.7490573660113116, "percentage": 37.45, "elapsed_time": "23:53:59", "remaining_time": "1 day, 15:54:47", "throughput": 5658.84, "total_tokens": 486881136}
+{"current_steps": 11130, "total_steps": 29704, "loss": 0.3482653617858887, "lr": 6.918301717523714e-06, "epoch": 0.7493940210072717, "percentage": 37.47, "elapsed_time": "23:54:37", "remaining_time": "1 day, 15:54:08", "throughput": 5658.79, "total_tokens": 487095072}
+{"current_steps": 11135, "total_steps": 29704, "loss": 0.3809635639190674, "lr": 6.915859707556611e-06, "epoch": 0.7497306760032318, "percentage": 37.49, "elapsed_time": "23:55:12", "remaining_time": "1 day, 15:53:22", "throughput": 5659.05, "total_tokens": 487312640}
+{"current_steps": 11140, "total_steps": 29704, "loss": 0.40050134658813474, "lr": 6.913417161825449e-06, "epoch": 0.750067330999192, "percentage": 37.5, "elapsed_time": "23:55:45", "remaining_time": "1 day, 15:52:35", "throughput": 5659.33, "total_tokens": 487525824}
+{"current_steps": 11145, "total_steps": 29704, "loss": 0.3789609432220459, "lr": 6.91097408101328e-06, "epoch": 0.7504039859951521, "percentage": 37.52, "elapsed_time": "23:56:26", "remaining_time": "1 day, 15:52:00", "throughput": 5659.15, "total_tokens": 487742328}
+{"current_steps": 11150, "total_steps": 29704, "loss": 0.370347785949707, "lr": 6.908530465803302e-06, "epoch": 0.7507406409911123, "percentage": 37.54, "elapsed_time": "23:56:57", "remaining_time": "1 day, 15:51:08", "throughput": 5659.52, "total_tokens": 487949632}
+{"current_steps": 11155, "total_steps": 29704, "loss": 0.3664633989334106, "lr": 6.906086316878865e-06, "epoch": 0.7510772959870724, "percentage": 37.55, "elapsed_time": "23:57:38", "remaining_time": "1 day, 15:50:34", "throughput": 5659.18, "total_tokens": 488152136}
+{"current_steps": 11160, "total_steps": 29704, "loss": 0.3477164268493652, "lr": 6.903641634923468e-06, "epoch": 0.7514139509830325, "percentage": 37.57, "elapsed_time": "23:58:17", "remaining_time": "1 day, 15:49:56", "throughput": 5659.09, "total_tokens": 488367048}
+{"current_steps": 11165, "total_steps": 29704, "loss": 0.3994622230529785, "lr": 6.9011964206207585e-06, "epoch": 0.7517506059789927, "percentage": 37.59, "elapsed_time": "23:58:50", "remaining_time": "1 day, 15:49:07", "throughput": 5659.27, "total_tokens": 488567560}
+{"current_steps": 11170, "total_steps": 29704, "loss": 0.36271369457244873, "lr": 6.898750674654531e-06, "epoch": 0.7520872609749528, "percentage": 37.6, "elapsed_time": "23:59:29", "remaining_time": "1 day, 15:48:30", "throughput": 5659.24, "total_tokens": 488788328}
+{"current_steps": 11175, "total_steps": 29704, "loss": 0.35087008476257325, "lr": 6.896304397708733e-06, "epoch": 0.752423915970913, "percentage": 37.62, "elapsed_time": "1 day, 0:00:08", "remaining_time": "1 day, 15:47:52", "throughput": 5659.09, "total_tokens": 488995192}
+{"current_steps": 11180, "total_steps": 29704, "loss": 0.3524162292480469, "lr": 6.893857590467456e-06, "epoch": 0.7527605709668731, "percentage": 37.64, "elapsed_time": "1 day, 0:00:45", "remaining_time": "1 day, 15:47:09", "throughput": 5659.23, "total_tokens": 489213016}
+{"current_steps": 11185, "total_steps": 29704, "loss": 0.34860227108001707, "lr": 6.891410253614944e-06, "epoch": 0.7530972259628332, "percentage": 37.65, "elapsed_time": "1 day, 0:01:24", "remaining_time": "1 day, 15:46:32", "throughput": 5659.04, "total_tokens": 489418272}
+{"current_steps": 11190, "total_steps": 29704, "loss": 0.384684419631958, "lr": 6.888962387835586e-06, "epoch": 0.7534338809587934, "percentage": 37.67, "elapsed_time": "1 day, 0:01:57", "remaining_time": "1 day, 15:45:43", "throughput": 5659.59, "total_tokens": 489652568}
+{"current_steps": 11195, "total_steps": 29704, "loss": 0.38306422233581544, "lr": 6.886513993813919e-06, "epoch": 0.7537705359547535, "percentage": 37.69, "elapsed_time": "1 day, 0:02:37", "remaining_time": "1 day, 15:45:07", "throughput": 5659.41, "total_tokens": 489864096}
+{"current_steps": 11200, "total_steps": 29704, "loss": 0.36569690704345703, "lr": 6.88406507223463e-06, "epoch": 0.7541071909507137, "percentage": 37.71, "elapsed_time": "1 day, 0:03:14", "remaining_time": "1 day, 15:44:26", "throughput": 5659.63, "total_tokens": 490092880}
+{"current_steps": 11205, "total_steps": 29704, "loss": 0.35335478782653806, "lr": 6.881615623782551e-06, "epoch": 0.7544438459466738, "percentage": 37.72, "elapsed_time": "1 day, 0:03:51", "remaining_time": "1 day, 15:43:45", "throughput": 5659.61, "total_tokens": 490301440}
+{"current_steps": 11210, "total_steps": 29704, "loss": 0.3659096956253052, "lr": 6.879165649142663e-06, "epoch": 0.7547805009426339, "percentage": 37.74, "elapsed_time": "1 day, 0:04:29", "remaining_time": "1 day, 15:43:04", "throughput": 5659.46, "total_tokens": 490500792}
+{"current_steps": 11215, "total_steps": 29704, "loss": 0.36233794689178467, "lr": 6.876715149000093e-06, "epoch": 0.7551171559385941, "percentage": 37.76, "elapsed_time": "1 day, 0:05:08", "remaining_time": "1 day, 15:42:26", "throughput": 5659.27, "total_tokens": 490705136}
+{"current_steps": 11220, "total_steps": 29704, "loss": 0.3791798114776611, "lr": 6.874264124040117e-06, "epoch": 0.7554538109345542, "percentage": 37.77, "elapsed_time": "1 day, 0:05:44", "remaining_time": "1 day, 15:41:43", "throughput": 5659.52, "total_tokens": 490930904}
+{"current_steps": 11225, "total_steps": 29704, "loss": 0.394960880279541, "lr": 6.871812574948154e-06, "epoch": 0.7557904659305144, "percentage": 37.79, "elapsed_time": "1 day, 0:06:17", "remaining_time": "1 day, 15:40:55", "throughput": 5660.0, "total_tokens": 491158544}
+{"current_steps": 11230, "total_steps": 29704, "loss": 0.3845766544342041, "lr": 6.869360502409773e-06, "epoch": 0.7561271209264745, "percentage": 37.81, "elapsed_time": "1 day, 0:06:54", "remaining_time": "1 day, 15:40:15", "throughput": 5660.15, "total_tokens": 491385232}
+{"current_steps": 11235, "total_steps": 29704, "loss": 0.351468563079834, "lr": 6.866907907110689e-06, "epoch": 0.7564637759224346, "percentage": 37.82, "elapsed_time": "1 day, 0:07:32", "remaining_time": "1 day, 15:39:35", "throughput": 5660.09, "total_tokens": 491593976}
+{"current_steps": 11240, "total_steps": 29704, "loss": 0.3472340822219849, "lr": 6.864454789736763e-06, "epoch": 0.7568004309183949, "percentage": 37.84, "elapsed_time": "1 day, 0:08:10", "remaining_time": "1 day, 15:38:54", "throughput": 5660.34, "total_tokens": 491827504}
+{"current_steps": 11245, "total_steps": 29704, "loss": 0.3707188844680786, "lr": 6.8620011509740005e-06, "epoch": 0.757137085914355, "percentage": 37.86, "elapsed_time": "1 day, 0:08:49", "remaining_time": "1 day, 15:38:16", "throughput": 5660.26, "total_tokens": 492042264}
+{"current_steps": 11250, "total_steps": 29704, "loss": 0.3541130542755127, "lr": 6.859546991508552e-06, "epoch": 0.7574737409103152, "percentage": 37.87, "elapsed_time": "1 day, 0:09:31", "remaining_time": "1 day, 15:37:44", "throughput": 5660.03, "total_tokens": 492261592}
+{"current_steps": 11255, "total_steps": 29704, "loss": 0.3816833972930908, "lr": 6.857092312026716e-06, "epoch": 0.7578103959062753, "percentage": 37.89, "elapsed_time": "1 day, 0:10:06", "remaining_time": "1 day, 15:36:59", "throughput": 5660.43, "total_tokens": 492495240}
+{"current_steps": 11260, "total_steps": 29704, "loss": 0.39476380348205564, "lr": 6.8546371132149395e-06, "epoch": 0.7581470509022354, "percentage": 37.91, "elapsed_time": "1 day, 0:10:46", "remaining_time": "1 day, 15:36:22", "throughput": 5660.35, "total_tokens": 492712328}
+{"current_steps": 11265, "total_steps": 29704, "loss": 0.39330875873565674, "lr": 6.852181395759807e-06, "epoch": 0.7584837058981956, "percentage": 37.92, "elapsed_time": "1 day, 0:11:27", "remaining_time": "1 day, 15:35:47", "throughput": 5660.34, "total_tokens": 492943248}
+{"current_steps": 11270, "total_steps": 29704, "loss": 0.3455686330795288, "lr": 6.849725160348054e-06, "epoch": 0.7588203608941557, "percentage": 37.94, "elapsed_time": "1 day, 0:12:07", "remaining_time": "1 day, 15:35:11", "throughput": 5660.13, "total_tokens": 493152344}
+{"current_steps": 11275, "total_steps": 29704, "loss": 0.35138664245605467, "lr": 6.847268407666557e-06, "epoch": 0.7591570158901159, "percentage": 37.96, "elapsed_time": "1 day, 0:12:51", "remaining_time": "1 day, 15:34:41", "throughput": 5659.84, "total_tokens": 493375584}
+{"current_steps": 11280, "total_steps": 29704, "loss": 0.37623722553253175, "lr": 6.844811138402341e-06, "epoch": 0.759493670886076, "percentage": 37.97, "elapsed_time": "1 day, 0:13:32", "remaining_time": "1 day, 15:34:07", "throughput": 5659.67, "total_tokens": 493594016}
+{"current_steps": 11285, "total_steps": 29704, "loss": 0.36553401947021485, "lr": 6.842353353242574e-06, "epoch": 0.7598303258820361, "percentage": 37.99, "elapsed_time": "1 day, 0:14:14", "remaining_time": "1 day, 15:33:34", "throughput": 5659.32, "total_tokens": 493803080}
+{"current_steps": 11290, "total_steps": 29704, "loss": 0.3759953737258911, "lr": 6.839895052874565e-06, "epoch": 0.7601669808779963, "percentage": 38.01, "elapsed_time": "1 day, 0:14:50", "remaining_time": "1 day, 15:32:50", "throughput": 5659.58, "total_tokens": 494027672}
+{"current_steps": 11295, "total_steps": 29704, "loss": 0.3641194343566895, "lr": 6.837436237985772e-06, "epoch": 0.7605036358739564, "percentage": 38.03, "elapsed_time": "1 day, 0:15:33", "remaining_time": "1 day, 15:32:20", "throughput": 5659.18, "total_tokens": 494238208}
+{"current_steps": 11300, "total_steps": 29704, "loss": 0.37904021739959715, "lr": 6.834976909263794e-06, "epoch": 0.7608402908699166, "percentage": 38.04, "elapsed_time": "1 day, 0:16:12", "remaining_time": "1 day, 15:31:41", "throughput": 5659.08, "total_tokens": 494450104}
+{"current_steps": 11305, "total_steps": 29704, "loss": 0.3586588859558105, "lr": 6.832517067396374e-06, "epoch": 0.7611769458658767, "percentage": 38.06, "elapsed_time": "1 day, 0:16:56", "remaining_time": "1 day, 15:31:11", "throughput": 5658.92, "total_tokens": 494683264}
+{"current_steps": 11310, "total_steps": 29704, "loss": 0.39732561111450193, "lr": 6.8300567130714e-06, "epoch": 0.7615136008618368, "percentage": 38.08, "elapsed_time": "1 day, 0:17:35", "remaining_time": "1 day, 15:30:33", "throughput": 5659.0, "total_tokens": 494912080}
+{"current_steps": 11315, "total_steps": 29704, "loss": 0.38528354167938234, "lr": 6.8275958469769035e-06, "epoch": 0.761850255857797, "percentage": 38.09, "elapsed_time": "1 day, 0:18:16", "remaining_time": "1 day, 15:29:57", "throughput": 5659.03, "total_tokens": 495143768}
+{"current_steps": 11320, "total_steps": 29704, "loss": 0.3580173969268799, "lr": 6.825134469801055e-06, "epoch": 0.7621869108537571, "percentage": 38.11, "elapsed_time": "1 day, 0:18:56", "remaining_time": "1 day, 15:29:22", "throughput": 5659.04, "total_tokens": 495373776}
+{"current_steps": 11325, "total_steps": 29704, "loss": 0.4001301765441895, "lr": 6.8226725822321715e-06, "epoch": 0.7625235658497173, "percentage": 38.13, "elapsed_time": "1 day, 0:19:34", "remaining_time": "1 day, 15:28:42", "throughput": 5658.96, "total_tokens": 495581360}
+{"current_steps": 11330, "total_steps": 29704, "loss": 0.38148252964019774, "lr": 6.820210184958712e-06, "epoch": 0.7628602208456774, "percentage": 38.14, "elapsed_time": "1 day, 0:20:11", "remaining_time": "1 day, 15:28:00", "throughput": 5659.15, "total_tokens": 495805664}
+{"current_steps": 11335, "total_steps": 29704, "loss": 0.38228158950805663, "lr": 6.81774727866928e-06, "epoch": 0.7631968758416375, "percentage": 38.16, "elapsed_time": "1 day, 0:20:47", "remaining_time": "1 day, 15:27:18", "throughput": 5659.18, "total_tokens": 496015888}
+{"current_steps": 11340, "total_steps": 29704, "loss": 0.3840935468673706, "lr": 6.815283864052617e-06, "epoch": 0.7635335308375977, "percentage": 38.18, "elapsed_time": "1 day, 0:21:27", "remaining_time": "1 day, 15:26:41", "throughput": 5659.14, "total_tokens": 496237176}
+{"current_steps": 11345, "total_steps": 29704, "loss": 0.38556504249572754, "lr": 6.812819941797611e-06, "epoch": 0.7638701858335578, "percentage": 38.19, "elapsed_time": "1 day, 0:22:10", "remaining_time": "1 day, 15:26:08", "throughput": 5658.75, "total_tokens": 496442520}
+{"current_steps": 11350, "total_steps": 29704, "loss": 0.3861824035644531, "lr": 6.810355512593288e-06, "epoch": 0.764206840829518, "percentage": 38.21, "elapsed_time": "1 day, 0:22:47", "remaining_time": "1 day, 15:25:27", "throughput": 5658.95, "total_tokens": 496669864}
+{"current_steps": 11355, "total_steps": 29704, "loss": 0.33897886276245115, "lr": 6.807890577128817e-06, "epoch": 0.7645434958254781, "percentage": 38.23, "elapsed_time": "1 day, 0:23:33", "remaining_time": "1 day, 15:25:01", "throughput": 5658.51, "total_tokens": 496892112}
+{"current_steps": 11360, "total_steps": 29704, "loss": 0.3582730531692505, "lr": 6.805425136093513e-06, "epoch": 0.7648801508214382, "percentage": 38.24, "elapsed_time": "1 day, 0:24:07", "remaining_time": "1 day, 15:24:15", "throughput": 5658.79, "total_tokens": 497111304}
+{"current_steps": 11365, "total_steps": 29704, "loss": 0.38888885974884035, "lr": 6.802959190176827e-06, "epoch": 0.7652168058173984, "percentage": 38.26, "elapsed_time": "1 day, 0:24:46", "remaining_time": "1 day, 15:23:37", "throughput": 5658.64, "total_tokens": 497318752}
+{"current_steps": 11370, "total_steps": 29704, "loss": 0.3655484676361084, "lr": 6.800492740068352e-06, "epoch": 0.7655534608133585, "percentage": 38.28, "elapsed_time": "1 day, 0:25:23", "remaining_time": "1 day, 15:22:55", "throughput": 5658.61, "total_tokens": 497523256}
+{"current_steps": 11375, "total_steps": 29704, "loss": 0.35955286026000977, "lr": 6.798025786457825e-06, "epoch": 0.7658901158093187, "percentage": 38.29, "elapsed_time": "1 day, 0:26:03", "remaining_time": "1 day, 15:22:19", "throughput": 5658.56, "total_tokens": 497749392}
+{"current_steps": 11380, "total_steps": 29704, "loss": 0.3689566612243652, "lr": 6.795558330035118e-06, "epoch": 0.7662267708052788, "percentage": 38.31, "elapsed_time": "1 day, 0:26:50", "remaining_time": "1 day, 15:21:53", "throughput": 5658.17, "total_tokens": 497978632}
+{"current_steps": 11385, "total_steps": 29704, "loss": 0.37029447555541994, "lr": 6.793090371490251e-06, "epoch": 0.7665634258012389, "percentage": 38.33, "elapsed_time": "1 day, 0:27:23", "remaining_time": "1 day, 15:21:05", "throughput": 5658.71, "total_tokens": 498211416}
+{"current_steps": 11390, "total_steps": 29704, "loss": 0.36082067489624026, "lr": 6.790621911513378e-06, "epoch": 0.7669000807971991, "percentage": 38.35, "elapsed_time": "1 day, 0:27:56", "remaining_time": "1 day, 15:20:19", "throughput": 5659.03, "total_tokens": 498430376}
+{"current_steps": 11395, "total_steps": 29704, "loss": 0.3912408113479614, "lr": 6.788152950794799e-06, "epoch": 0.7672367357931592, "percentage": 38.36, "elapsed_time": "1 day, 0:28:33", "remaining_time": "1 day, 15:19:36", "throughput": 5659.17, "total_tokens": 498646776}
+{"current_steps": 11400, "total_steps": 29704, "loss": 0.3919992208480835, "lr": 6.7856834900249495e-06, "epoch": 0.7675733907891193, "percentage": 38.38, "elapsed_time": "1 day, 0:29:17", "remaining_time": "1 day, 15:19:07", "throughput": 5658.76, "total_tokens": 498865144}
+{"current_steps": 11405, "total_steps": 29704, "loss": 0.3473628044128418, "lr": 6.783213529894405e-06, "epoch": 0.7679100457850795, "percentage": 38.4, "elapsed_time": "1 day, 0:30:02", "remaining_time": "1 day, 15:18:38", "throughput": 5658.36, "total_tokens": 499082328}
+{"current_steps": 11410, "total_steps": 29704, "loss": 0.35947589874267577, "lr": 6.780743071093883e-06, "epoch": 0.7682467007810396, "percentage": 38.41, "elapsed_time": "1 day, 0:30:42", "remaining_time": "1 day, 15:18:01", "throughput": 5658.36, "total_tokens": 499306216}
+{"current_steps": 11415, "total_steps": 29704, "loss": 0.36936905384063723, "lr": 6.77827211431424e-06, "epoch": 0.7685833557769998, "percentage": 38.43, "elapsed_time": "1 day, 0:31:20", "remaining_time": "1 day, 15:17:21", "throughput": 5658.49, "total_tokens": 499532072}
+{"current_steps": 11420, "total_steps": 29704, "loss": 0.36505095958709716, "lr": 6.775800660246471e-06, "epoch": 0.7689200107729599, "percentage": 38.45, "elapsed_time": "1 day, 0:31:56", "remaining_time": "1 day, 15:16:39", "throughput": 5658.41, "total_tokens": 499732272}
+{"current_steps": 11425, "total_steps": 29704, "loss": 0.3719770908355713, "lr": 6.77332870958171e-06, "epoch": 0.76925666576892, "percentage": 38.46, "elapsed_time": "1 day, 0:32:41", "remaining_time": "1 day, 15:16:10", "throughput": 5658.18, "total_tokens": 499963848}
+{"current_steps": 11430, "total_steps": 29704, "loss": 0.3673600196838379, "lr": 6.770856263011229e-06, "epoch": 0.7695933207648802, "percentage": 38.48, "elapsed_time": "1 day, 0:33:14", "remaining_time": "1 day, 15:15:22", "throughput": 5658.46, "total_tokens": 500175216}
+{"current_steps": 11435, "total_steps": 29704, "loss": 0.38931114673614503, "lr": 6.768383321226442e-06, "epoch": 0.7699299757608403, "percentage": 38.5, "elapsed_time": "1 day, 0:33:52", "remaining_time": "1 day, 15:14:43", "throughput": 5658.32, "total_tokens": 500380304}
+{"current_steps": 11440, "total_steps": 29704, "loss": 0.3583209753036499, "lr": 6.765909884918898e-06, "epoch": 0.7702666307568005, "percentage": 38.51, "elapsed_time": "1 day, 0:34:29", "remaining_time": "1 day, 15:14:01", "throughput": 5658.35, "total_tokens": 500590256}
+{"current_steps": 11445, "total_steps": 29704, "loss": 0.3920794725418091, "lr": 6.763435954780286e-06, "epoch": 0.7706032857527606, "percentage": 38.53, "elapsed_time": "1 day, 0:35:04", "remaining_time": "1 day, 15:13:18", "throughput": 5658.56, "total_tokens": 500810896}
+{"current_steps": 11450, "total_steps": 29704, "loss": 0.37000212669372556, "lr": 6.7609615315024325e-06, "epoch": 0.7709399407487207, "percentage": 38.55, "elapsed_time": "1 day, 0:35:43", "remaining_time": "1 day, 15:12:38", "throughput": 5658.56, "total_tokens": 501027440}
+{"current_steps": 11455, "total_steps": 29704, "loss": 0.37416849136352537, "lr": 6.758486615777301e-06, "epoch": 0.7712765957446809, "percentage": 38.56, "elapsed_time": "1 day, 0:36:23", "remaining_time": "1 day, 15:12:03", "throughput": 5658.39, "total_tokens": 501241576}
+{"current_steps": 11460, "total_steps": 29704, "loss": 0.3626854419708252, "lr": 6.756011208296995e-06, "epoch": 0.771613250740641, "percentage": 38.58, "elapsed_time": "1 day, 0:37:01", "remaining_time": "1 day, 15:11:22", "throughput": 5658.39, "total_tokens": 501454696}
+{"current_steps": 11465, "total_steps": 29704, "loss": 0.35544500350952146, "lr": 6.753535309753755e-06, "epoch": 0.7719499057366012, "percentage": 38.6, "elapsed_time": "1 day, 0:37:42", "remaining_time": "1 day, 15:10:48", "throughput": 5658.07, "total_tokens": 501658528}
+{"current_steps": 11470, "total_steps": 29704, "loss": 0.3747244119644165, "lr": 6.7510589208399565e-06, "epoch": 0.7722865607325613, "percentage": 38.61, "elapsed_time": "1 day, 0:38:23", "remaining_time": "1 day, 15:10:13", "throughput": 5657.91, "total_tokens": 501877728}
+{"current_steps": 11475, "total_steps": 29704, "loss": 0.35962686538696287, "lr": 6.748582042248115e-06, "epoch": 0.7726232157285214, "percentage": 38.63, "elapsed_time": "1 day, 0:39:00", "remaining_time": "1 day, 15:09:31", "throughput": 5657.94, "total_tokens": 502089032}
+{"current_steps": 11480, "total_steps": 29704, "loss": 0.38360817432403566, "lr": 6.746104674670882e-06, "epoch": 0.7729598707244816, "percentage": 38.65, "elapsed_time": "1 day, 0:39:37", "remaining_time": "1 day, 15:08:50", "throughput": 5658.12, "total_tokens": 502313408}
+{"current_steps": 11485, "total_steps": 29704, "loss": 0.39516823291778563, "lr": 6.743626818801042e-06, "epoch": 0.7732965257204417, "percentage": 38.66, "elapsed_time": "1 day, 0:40:18", "remaining_time": "1 day, 15:08:15", "throughput": 5658.0, "total_tokens": 502534360}
+{"current_steps": 11490, "total_steps": 29704, "loss": 0.37048795223236086, "lr": 6.741148475331525e-06, "epoch": 0.7736331807164019, "percentage": 38.68, "elapsed_time": "1 day, 0:40:54", "remaining_time": "1 day, 15:07:32", "throughput": 5658.22, "total_tokens": 502757200}
+{"current_steps": 11495, "total_steps": 29704, "loss": 0.3468164443969727, "lr": 6.738669644955388e-06, "epoch": 0.773969835712362, "percentage": 38.7, "elapsed_time": "1 day, 0:41:37", "remaining_time": "1 day, 15:07:00", "throughput": 5657.9, "total_tokens": 502970848}
+{"current_steps": 11500, "total_steps": 29704, "loss": 0.3672001838684082, "lr": 6.736190328365828e-06, "epoch": 0.7743064907083221, "percentage": 38.72, "elapsed_time": "1 day, 0:42:18", "remaining_time": "1 day, 15:06:25", "throughput": 5657.79, "total_tokens": 503195432}
+{"current_steps": 11505, "total_steps": 29704, "loss": 0.37738971710205077, "lr": 6.73371052625618e-06, "epoch": 0.7746431457042823, "percentage": 38.73, "elapsed_time": "1 day, 0:42:59", "remaining_time": "1 day, 15:05:51", "throughput": 5657.67, "total_tokens": 503417816}
+{"current_steps": 11510, "total_steps": 29704, "loss": 0.37387096881866455, "lr": 6.731230239319912e-06, "epoch": 0.7749798007002424, "percentage": 38.75, "elapsed_time": "1 day, 0:43:41", "remaining_time": "1 day, 15:05:17", "throughput": 5657.53, "total_tokens": 503642040}
+{"current_steps": 11515, "total_steps": 29704, "loss": 0.38176569938659666, "lr": 6.728749468250627e-06, "epoch": 0.7753164556962026, "percentage": 38.77, "elapsed_time": "1 day, 0:44:18", "remaining_time": "1 day, 15:04:35", "throughput": 5657.76, "total_tokens": 503870440}
+{"current_steps": 11520, "total_steps": 29704, "loss": 0.375903582572937, "lr": 6.726268213742066e-06, "epoch": 0.7756531106921627, "percentage": 38.78, "elapsed_time": "1 day, 0:45:00", "remaining_time": "1 day, 15:04:01", "throughput": 5657.42, "total_tokens": 504076280}
+{"current_steps": 11525, "total_steps": 29704, "loss": 0.38536639213562013, "lr": 6.723786476488105e-06, "epoch": 0.7759897656881228, "percentage": 38.8, "elapsed_time": "1 day, 0:45:34", "remaining_time": "1 day, 15:03:15", "throughput": 5657.81, "total_tokens": 504303624}
+{"current_steps": 11530, "total_steps": 29704, "loss": 0.3504000186920166, "lr": 6.72130425718275e-06, "epoch": 0.776326420684083, "percentage": 38.82, "elapsed_time": "1 day, 0:46:10", "remaining_time": "1 day, 15:02:33", "throughput": 5657.83, "total_tokens": 504511120}
+{"current_steps": 11535, "total_steps": 29704, "loss": 0.3902331829071045, "lr": 6.718821556520151e-06, "epoch": 0.7766630756800431, "percentage": 38.83, "elapsed_time": "1 day, 0:46:43", "remaining_time": "1 day, 15:01:46", "throughput": 5658.06, "total_tokens": 504720744}
+{"current_steps": 11540, "total_steps": 29704, "loss": 0.3558895826339722, "lr": 6.716338375194583e-06, "epoch": 0.7769997306760033, "percentage": 38.85, "elapsed_time": "1 day, 0:47:19", "remaining_time": "1 day, 15:01:02", "throughput": 5658.35, "total_tokens": 504945672}
+{"current_steps": 11545, "total_steps": 29704, "loss": 0.39481263160705565, "lr": 6.713854713900463e-06, "epoch": 0.7773363856719634, "percentage": 38.87, "elapsed_time": "1 day, 0:48:00", "remaining_time": "1 day, 15:00:27", "throughput": 5658.24, "total_tokens": 505168112}
+{"current_steps": 11550, "total_steps": 29704, "loss": 0.3915176153182983, "lr": 6.7113705733323365e-06, "epoch": 0.7776730406679235, "percentage": 38.88, "elapsed_time": "1 day, 0:48:41", "remaining_time": "1 day, 14:59:53", "throughput": 5657.89, "total_tokens": 505373904}
+{"current_steps": 11555, "total_steps": 29704, "loss": 0.36997599601745607, "lr": 6.708885954184885e-06, "epoch": 0.7780096956638837, "percentage": 38.9, "elapsed_time": "1 day, 0:49:22", "remaining_time": "1 day, 14:59:17", "throughput": 5657.8, "total_tokens": 505592960}
+{"current_steps": 11560, "total_steps": 29704, "loss": 0.3574796199798584, "lr": 6.7064008571529275e-06, "epoch": 0.7783463506598438, "percentage": 38.92, "elapsed_time": "1 day, 0:50:07", "remaining_time": "1 day, 14:58:49", "throughput": 5657.42, "total_tokens": 505815992}
+{"current_steps": 11565, "total_steps": 29704, "loss": 0.37357444763183595, "lr": 6.703915282931411e-06, "epoch": 0.778683005655804, "percentage": 38.93, "elapsed_time": "1 day, 0:50:52", "remaining_time": "1 day, 14:58:20", "throughput": 5657.13, "total_tokens": 506041792}
+{"current_steps": 11570, "total_steps": 29704, "loss": 0.3822374105453491, "lr": 6.701429232215418e-06, "epoch": 0.7790196606517641, "percentage": 38.95, "elapsed_time": "1 day, 0:51:31", "remaining_time": "1 day, 14:57:42", "throughput": 5656.91, "total_tokens": 506244224}
+{"current_steps": 11575, "total_steps": 29704, "loss": 0.3804741144180298, "lr": 6.6989427057001665e-06, "epoch": 0.7793563156477242, "percentage": 38.97, "elapsed_time": "1 day, 0:52:12", "remaining_time": "1 day, 14:57:08", "throughput": 5656.81, "total_tokens": 506470096}
+{"current_steps": 11580, "total_steps": 29704, "loss": 0.3787699699401855, "lr": 6.696455704081006e-06, "epoch": 0.7796929706436844, "percentage": 38.98, "elapsed_time": "1 day, 0:52:54", "remaining_time": "1 day, 14:56:33", "throughput": 5656.62, "total_tokens": 506688376}
+{"current_steps": 11585, "total_steps": 29704, "loss": 0.38877334594726565, "lr": 6.693968228053416e-06, "epoch": 0.7800296256396445, "percentage": 39.0, "elapsed_time": "1 day, 0:53:29", "remaining_time": "1 day, 14:55:50", "throughput": 5656.84, "total_tokens": 506908432}
+{"current_steps": 11590, "total_steps": 29704, "loss": 0.37210679054260254, "lr": 6.691480278313013e-06, "epoch": 0.7803662806356046, "percentage": 39.02, "elapsed_time": "1 day, 0:54:10", "remaining_time": "1 day, 14:55:15", "throughput": 5656.64, "total_tokens": 507123272}
+{"current_steps": 11595, "total_steps": 29704, "loss": 0.35350847244262695, "lr": 6.688991855555545e-06, "epoch": 0.7807029356315648, "percentage": 39.04, "elapsed_time": "1 day, 0:54:54", "remaining_time": "1 day, 14:54:43", "throughput": 5656.33, "total_tokens": 507339456}
+{"current_steps": 11600, "total_steps": 29704, "loss": 0.3558246612548828, "lr": 6.686502960476888e-06, "epoch": 0.7810395906275249, "percentage": 39.05, "elapsed_time": "1 day, 0:55:31", "remaining_time": "1 day, 14:54:02", "throughput": 5656.63, "total_tokens": 507577136}
+{"current_steps": 11605, "total_steps": 29704, "loss": 0.3307861089706421, "lr": 6.684013593773058e-06, "epoch": 0.7813762456234851, "percentage": 39.07, "elapsed_time": "1 day, 0:56:06", "remaining_time": "1 day, 14:53:18", "throughput": 5657.03, "total_tokens": 507811288}
+{"current_steps": 11610, "total_steps": 29704, "loss": 0.33051750659942625, "lr": 6.6815237561401955e-06, "epoch": 0.7817129006194452, "percentage": 39.09, "elapsed_time": "1 day, 0:56:42", "remaining_time": "1 day, 14:52:35", "throughput": 5656.94, "total_tokens": 508006304}
+{"current_steps": 11615, "total_steps": 29704, "loss": 0.35503997802734377, "lr": 6.679033448274577e-06, "epoch": 0.7820495556154053, "percentage": 39.1, "elapsed_time": "1 day, 0:57:30", "remaining_time": "1 day, 14:52:11", "throughput": 5656.35, "total_tokens": 508223832}
+{"current_steps": 11620, "total_steps": 29704, "loss": 0.3701477527618408, "lr": 6.676542670872609e-06, "epoch": 0.7823862106113655, "percentage": 39.12, "elapsed_time": "1 day, 0:58:10", "remaining_time": "1 day, 14:51:35", "throughput": 5656.21, "total_tokens": 508440024}
+{"current_steps": 11625, "total_steps": 29704, "loss": 0.36185245513916015, "lr": 6.674051424630829e-06, "epoch": 0.7827228656073256, "percentage": 39.14, "elapsed_time": "1 day, 0:58:44", "remaining_time": "1 day, 14:50:48", "throughput": 5656.6, "total_tokens": 508666440}
+{"current_steps": 11630, "total_steps": 29704, "loss": 0.36898825168609617, "lr": 6.671559710245905e-06, "epoch": 0.7830595206032858, "percentage": 39.15, "elapsed_time": "1 day, 0:59:30", "remaining_time": "1 day, 14:50:21", "throughput": 5656.24, "total_tokens": 508892304}
+{"current_steps": 11635, "total_steps": 29704, "loss": 0.36539783477783205, "lr": 6.669067528414639e-06, "epoch": 0.7833961755992459, "percentage": 39.17, "elapsed_time": "1 day, 1:00:12", "remaining_time": "1 day, 14:49:47", "throughput": 5656.1, "total_tokens": 509118040}
+{"current_steps": 11640, "total_steps": 29704, "loss": 0.39262051582336427, "lr": 6.6665748798339606e-06, "epoch": 0.783732830595206, "percentage": 39.19, "elapsed_time": "1 day, 1:00:50", "remaining_time": "1 day, 14:49:07", "throughput": 5656.25, "total_tokens": 509346248}
+{"current_steps": 11645, "total_steps": 29704, "loss": 0.38922972679138185, "lr": 6.664081765200932e-06, "epoch": 0.7840694855911662, "percentage": 39.2, "elapsed_time": "1 day, 1:01:22", "remaining_time": "1 day, 14:48:19", "throughput": 5656.45, "total_tokens": 509547192}
+{"current_steps": 11650, "total_steps": 29704, "loss": 0.38661224842071534, "lr": 6.661588185212745e-06, "epoch": 0.7844061405871263, "percentage": 39.22, "elapsed_time": "1 day, 1:02:04", "remaining_time": "1 day, 14:47:45", "throughput": 5656.28, "total_tokens": 509767456}
+{"current_steps": 11655, "total_steps": 29704, "loss": 0.37398757934570315, "lr": 6.65909414056672e-06, "epoch": 0.7847427955830865, "percentage": 39.24, "elapsed_time": "1 day, 1:02:42", "remaining_time": "1 day, 14:47:06", "throughput": 5656.19, "total_tokens": 509978544}
+{"current_steps": 11660, "total_steps": 29704, "loss": 0.36649136543273925, "lr": 6.6565996319603075e-06, "epoch": 0.7850794505790466, "percentage": 39.25, "elapsed_time": "1 day, 1:03:25", "remaining_time": "1 day, 14:46:33", "throughput": 5655.93, "total_tokens": 510194488}
+{"current_steps": 11665, "total_steps": 29704, "loss": 0.35945792198181153, "lr": 6.6541046600910915e-06, "epoch": 0.7854161055750067, "percentage": 39.27, "elapsed_time": "1 day, 1:04:02", "remaining_time": "1 day, 14:45:53", "throughput": 5655.96, "total_tokens": 510408368}
+{"current_steps": 11670, "total_steps": 29704, "loss": 0.36875619888305666, "lr": 6.6516092256567824e-06, "epoch": 0.7857527605709669, "percentage": 39.29, "elapsed_time": "1 day, 1:04:39", "remaining_time": "1 day, 14:45:11", "throughput": 5656.15, "total_tokens": 510635536}
+{"current_steps": 11675, "total_steps": 29704, "loss": 0.39651241302490237, "lr": 6.64911332935522e-06, "epoch": 0.786089415566927, "percentage": 39.3, "elapsed_time": "1 day, 1:05:20", "remaining_time": "1 day, 14:44:36", "throughput": 5656.09, "total_tokens": 510860904}
+{"current_steps": 11680, "total_steps": 29704, "loss": 0.3647106885910034, "lr": 6.646616971884373e-06, "epoch": 0.7864260705628872, "percentage": 39.32, "elapsed_time": "1 day, 1:05:59", "remaining_time": "1 day, 14:43:58", "throughput": 5656.12, "total_tokens": 511085760}
+{"current_steps": 11685, "total_steps": 29704, "loss": 0.35194456577301025, "lr": 6.64412015394234e-06, "epoch": 0.7867627255588473, "percentage": 39.34, "elapsed_time": "1 day, 1:06:37", "remaining_time": "1 day, 14:43:18", "throughput": 5655.97, "total_tokens": 511285624}
+{"current_steps": 11690, "total_steps": 29704, "loss": 0.3812446594238281, "lr": 6.641622876227349e-06, "epoch": 0.7870993805548074, "percentage": 39.35, "elapsed_time": "1 day, 1:07:19", "remaining_time": "1 day, 14:42:45", "throughput": 5655.61, "total_tokens": 511491128}
+{"current_steps": 11695, "total_steps": 29704, "loss": 0.3623605012893677, "lr": 6.639125139437756e-06, "epoch": 0.7874360355507676, "percentage": 39.37, "elapsed_time": "1 day, 1:08:02", "remaining_time": "1 day, 14:42:12", "throughput": 5655.33, "total_tokens": 511707944}
+{"current_steps": 11700, "total_steps": 29704, "loss": 0.36579647064208987, "lr": 6.6366269442720446e-06, "epoch": 0.7877726905467277, "percentage": 39.39, "elapsed_time": "1 day, 1:08:42", "remaining_time": "1 day, 14:41:36", "throughput": 5655.14, "total_tokens": 511918704}
+{"current_steps": 11705, "total_steps": 29704, "loss": 0.4204265594482422, "lr": 6.634128291428827e-06, "epoch": 0.7881093455426879, "percentage": 39.41, "elapsed_time": "1 day, 1:09:21", "remaining_time": "1 day, 14:40:57", "throughput": 5655.08, "total_tokens": 512130360}
+{"current_steps": 11710, "total_steps": 29704, "loss": 0.37847514152526857, "lr": 6.631629181606842e-06, "epoch": 0.788446000538648, "percentage": 39.42, "elapsed_time": "1 day, 1:10:04", "remaining_time": "1 day, 14:40:26", "throughput": 5654.81, "total_tokens": 512353072}
+{"current_steps": 11715, "total_steps": 29704, "loss": 0.3778711795806885, "lr": 6.629129615504961e-06, "epoch": 0.7887826555346081, "percentage": 39.44, "elapsed_time": "1 day, 1:10:44", "remaining_time": "1 day, 14:39:49", "throughput": 5654.81, "total_tokens": 512578040}
+{"current_steps": 11720, "total_steps": 29704, "loss": 0.349456787109375, "lr": 6.626629593822178e-06, "epoch": 0.7891193105305683, "percentage": 39.46, "elapsed_time": "1 day, 1:11:23", "remaining_time": "1 day, 14:39:11", "throughput": 5654.76, "total_tokens": 512795568}
+{"current_steps": 11725, "total_steps": 29704, "loss": 0.3877539157867432, "lr": 6.624129117257616e-06, "epoch": 0.7894559655265284, "percentage": 39.47, "elapsed_time": "1 day, 1:12:01", "remaining_time": "1 day, 14:38:31", "throughput": 5654.89, "total_tokens": 513020704}
+{"current_steps": 11730, "total_steps": 29704, "loss": 0.37906270027160643, "lr": 6.621628186510524e-06, "epoch": 0.7897926205224886, "percentage": 39.49, "elapsed_time": "1 day, 1:12:43", "remaining_time": "1 day, 14:37:58", "throughput": 5654.67, "total_tokens": 513238728}
+{"current_steps": 11735, "total_steps": 29704, "loss": 0.35970292091369627, "lr": 6.619126802280281e-06, "epoch": 0.7901292755184487, "percentage": 39.51, "elapsed_time": "1 day, 1:13:26", "remaining_time": "1 day, 14:37:26", "throughput": 5654.36, "total_tokens": 513454240}
+{"current_steps": 11740, "total_steps": 29704, "loss": 0.3722025632858276, "lr": 6.616624965266392e-06, "epoch": 0.7904659305144088, "percentage": 39.52, "elapsed_time": "1 day, 1:14:05", "remaining_time": "1 day, 14:36:46", "throughput": 5654.4, "total_tokens": 513674120}
+{"current_steps": 11745, "total_steps": 29704, "loss": 0.39596946239471437, "lr": 6.614122676168486e-06, "epoch": 0.790802585510369, "percentage": 39.54, "elapsed_time": "1 day, 1:14:49", "remaining_time": "1 day, 14:36:16", "throughput": 5654.02, "total_tokens": 513889520}
+{"current_steps": 11750, "total_steps": 29704, "loss": 0.34925665855407717, "lr": 6.611619935686321e-06, "epoch": 0.7911392405063291, "percentage": 39.56, "elapsed_time": "1 day, 1:15:23", "remaining_time": "1 day, 14:35:31", "throughput": 5654.13, "total_tokens": 514094568}
+{"current_steps": 11755, "total_steps": 29704, "loss": 0.3882757663726807, "lr": 6.609116744519779e-06, "epoch": 0.7914758955022893, "percentage": 39.57, "elapsed_time": "1 day, 1:16:01", "remaining_time": "1 day, 14:34:51", "throughput": 5654.24, "total_tokens": 514319376}
+{"current_steps": 11760, "total_steps": 29704, "loss": 0.3384627103805542, "lr": 6.60661310336887e-06, "epoch": 0.7918125504982494, "percentage": 39.59, "elapsed_time": "1 day, 1:16:48", "remaining_time": "1 day, 14:34:25", "throughput": 5653.6, "total_tokens": 514527584}
+{"current_steps": 11765, "total_steps": 29704, "loss": 0.37465848922729494, "lr": 6.604109012933732e-06, "epoch": 0.7921492054942095, "percentage": 39.61, "elapsed_time": "1 day, 1:17:20", "remaining_time": "1 day, 14:33:37", "throughput": 5653.83, "total_tokens": 514729464}
+{"current_steps": 11770, "total_steps": 29704, "loss": 0.39129223823547366, "lr": 6.601604473914622e-06, "epoch": 0.7924858604901697, "percentage": 39.62, "elapsed_time": "1 day, 1:18:00", "remaining_time": "1 day, 14:33:00", "throughput": 5653.76, "total_tokens": 514948712}
+{"current_steps": 11775, "total_steps": 29704, "loss": 0.3703568458557129, "lr": 6.599099487011928e-06, "epoch": 0.7928225154861298, "percentage": 39.64, "elapsed_time": "1 day, 1:18:43", "remaining_time": "1 day, 14:32:27", "throughput": 5653.57, "total_tokens": 515174152}
+{"current_steps": 11780, "total_steps": 29704, "loss": 0.37824716567993166, "lr": 6.596594052926163e-06, "epoch": 0.79315917048209, "percentage": 39.66, "elapsed_time": "1 day, 1:19:21", "remaining_time": "1 day, 14:31:48", "throughput": 5653.66, "total_tokens": 515398616}
+{"current_steps": 11785, "total_steps": 29704, "loss": 0.36647372245788573, "lr": 6.5940881723579595e-06, "epoch": 0.7934958254780501, "percentage": 39.67, "elapsed_time": "1 day, 1:19:59", "remaining_time": "1 day, 14:31:08", "throughput": 5653.73, "total_tokens": 515618208}
+{"current_steps": 11790, "total_steps": 29704, "loss": 0.34111249446868896, "lr": 6.591581846008085e-06, "epoch": 0.7938324804740102, "percentage": 39.69, "elapsed_time": "1 day, 1:20:42", "remaining_time": "1 day, 14:30:36", "throughput": 5653.46, "total_tokens": 515838000}
+{"current_steps": 11795, "total_steps": 29704, "loss": 0.3614863634109497, "lr": 6.589075074577421e-06, "epoch": 0.7941691354699704, "percentage": 39.71, "elapsed_time": "1 day, 1:21:28", "remaining_time": "1 day, 14:30:07", "throughput": 5653.08, "total_tokens": 516058872}
+{"current_steps": 11800, "total_steps": 29704, "loss": 0.35341219902038573, "lr": 6.5865678587669805e-06, "epoch": 0.7945057904659305, "percentage": 39.73, "elapsed_time": "1 day, 1:22:04", "remaining_time": "1 day, 14:29:25", "throughput": 5653.33, "total_tokens": 516288480}
+{"current_steps": 11805, "total_steps": 29704, "loss": 0.3922135829925537, "lr": 6.584060199277897e-06, "epoch": 0.7948424454618906, "percentage": 39.74, "elapsed_time": "1 day, 1:22:36", "remaining_time": "1 day, 14:28:36", "throughput": 5653.68, "total_tokens": 516500456}
+{"current_steps": 11810, "total_steps": 29704, "loss": 0.3736285209655762, "lr": 6.581552096811428e-06, "epoch": 0.7951791004578508, "percentage": 39.76, "elapsed_time": "1 day, 1:23:11", "remaining_time": "1 day, 14:27:52", "throughput": 5653.9, "total_tokens": 516717184}
+{"current_steps": 11815, "total_steps": 29704, "loss": 0.3753064632415771, "lr": 6.57904355206896e-06, "epoch": 0.7955157554538109, "percentage": 39.78, "elapsed_time": "1 day, 1:23:53", "remaining_time": "1 day, 14:27:18", "throughput": 5653.7, "total_tokens": 516937256}
+{"current_steps": 11820, "total_steps": 29704, "loss": 0.35328757762908936, "lr": 6.576534565751997e-06, "epoch": 0.7958524104497711, "percentage": 39.79, "elapsed_time": "1 day, 1:24:32", "remaining_time": "1 day, 14:26:40", "throughput": 5653.68, "total_tokens": 517155784}
+{"current_steps": 11825, "total_steps": 29704, "loss": 0.3843714475631714, "lr": 6.574025138562172e-06, "epoch": 0.7961890654457312, "percentage": 39.81, "elapsed_time": "1 day, 1:25:07", "remaining_time": "1 day, 14:25:55", "throughput": 5653.79, "total_tokens": 517361632}
+{"current_steps": 11830, "total_steps": 29704, "loss": 0.3463775634765625, "lr": 6.571515271201234e-06, "epoch": 0.7965257204416913, "percentage": 39.83, "elapsed_time": "1 day, 1:25:47", "remaining_time": "1 day, 14:25:19", "throughput": 5653.74, "total_tokens": 517585680}
+{"current_steps": 11835, "total_steps": 29704, "loss": 0.3861451864242554, "lr": 6.569004964371062e-06, "epoch": 0.7968623754376515, "percentage": 39.84, "elapsed_time": "1 day, 1:26:20", "remaining_time": "1 day, 14:24:32", "throughput": 5654.15, "total_tokens": 517810808}
+{"current_steps": 11840, "total_steps": 29704, "loss": 0.3636291265487671, "lr": 6.566494218773656e-06, "epoch": 0.7971990304336116, "percentage": 39.86, "elapsed_time": "1 day, 1:27:03", "remaining_time": "1 day, 14:24:00", "throughput": 5653.96, "total_tokens": 518036944}
+{"current_steps": 11845, "total_steps": 29704, "loss": 0.376003098487854, "lr": 6.563983035111136e-06, "epoch": 0.7975356854295718, "percentage": 39.88, "elapsed_time": "1 day, 1:27:40", "remaining_time": "1 day, 14:23:18", "throughput": 5654.21, "total_tokens": 518266216}
+{"current_steps": 11850, "total_steps": 29704, "loss": 0.36989893913269045, "lr": 6.561471414085748e-06, "epoch": 0.7978723404255319, "percentage": 39.89, "elapsed_time": "1 day, 1:28:16", "remaining_time": "1 day, 14:22:35", "throughput": 5654.28, "total_tokens": 518477336}
+{"current_steps": 11855, "total_steps": 29704, "loss": 0.3683455944061279, "lr": 6.558959356399859e-06, "epoch": 0.798208995421492, "percentage": 39.91, "elapsed_time": "1 day, 1:28:55", "remaining_time": "1 day, 14:21:57", "throughput": 5654.22, "total_tokens": 518691120}
+{"current_steps": 11860, "total_steps": 29704, "loss": 0.3579908609390259, "lr": 6.5564468627559544e-06, "epoch": 0.7985456504174522, "percentage": 39.93, "elapsed_time": "1 day, 1:29:35", "remaining_time": "1 day, 14:21:21", "throughput": 5653.97, "total_tokens": 518897696}
+{"current_steps": 11865, "total_steps": 29704, "loss": 0.38613746166229246, "lr": 6.5539339338566495e-06, "epoch": 0.7988823054134123, "percentage": 39.94, "elapsed_time": "1 day, 1:30:16", "remaining_time": "1 day, 14:20:45", "throughput": 5653.83, "total_tokens": 519112712}
+{"current_steps": 11870, "total_steps": 29704, "loss": 0.34462246894836424, "lr": 6.5514205704046754e-06, "epoch": 0.7992189604093725, "percentage": 39.96, "elapsed_time": "1 day, 1:30:57", "remaining_time": "1 day, 14:20:11", "throughput": 5653.58, "total_tokens": 519326360}
+{"current_steps": 11875, "total_steps": 29704, "loss": 0.34732234477996826, "lr": 6.548906773102886e-06, "epoch": 0.7995556154053326, "percentage": 39.98, "elapsed_time": "1 day, 1:31:32", "remaining_time": "1 day, 14:19:26", "throughput": 5653.75, "total_tokens": 519538040}
+{"current_steps": 11880, "total_steps": 29704, "loss": 0.3555797576904297, "lr": 6.546392542654257e-06, "epoch": 0.7998922704012927, "percentage": 39.99, "elapsed_time": "1 day, 1:32:11", "remaining_time": "1 day, 14:18:48", "throughput": 5653.69, "total_tokens": 519754440}
+{"current_steps": 11885, "total_steps": 29704, "loss": 0.39305408000946046, "lr": 6.543877879761882e-06, "epoch": 0.8002289253972529, "percentage": 40.01, "elapsed_time": "1 day, 1:32:52", "remaining_time": "1 day, 14:18:12", "throughput": 5653.56, "total_tokens": 519971256}
+{"current_steps": 11890, "total_steps": 29704, "loss": 0.3501169681549072, "lr": 6.5413627851289825e-06, "epoch": 0.800565580393213, "percentage": 40.03, "elapsed_time": "1 day, 1:33:30", "remaining_time": "1 day, 14:17:32", "throughput": 5653.7, "total_tokens": 520197464}
+{"current_steps": 11895, "total_steps": 29704, "loss": 0.3551483631134033, "lr": 6.538847259458896e-06, "epoch": 0.8009022353891732, "percentage": 40.05, "elapsed_time": "1 day, 1:34:02", "remaining_time": "1 day, 14:16:45", "throughput": 5654.13, "total_tokens": 520422616}
+{"current_steps": 11900, "total_steps": 29704, "loss": 0.34134459495544434, "lr": 6.5363313034550805e-06, "epoch": 0.8012388903851333, "percentage": 40.06, "elapsed_time": "1 day, 1:34:47", "remaining_time": "1 day, 14:16:15", "throughput": 5653.67, "total_tokens": 520633392}
+{"current_steps": 11905, "total_steps": 29704, "loss": 0.3754891395568848, "lr": 6.533814917821114e-06, "epoch": 0.8015755453810934, "percentage": 40.08, "elapsed_time": "1 day, 1:35:27", "remaining_time": "1 day, 14:15:39", "throughput": 5653.62, "total_tokens": 520855784}
+{"current_steps": 11910, "total_steps": 29704, "loss": 0.3585964202880859, "lr": 6.531298103260696e-06, "epoch": 0.8019122003770536, "percentage": 40.1, "elapsed_time": "1 day, 1:36:03", "remaining_time": "1 day, 14:14:55", "throughput": 5653.72, "total_tokens": 521066736}
+{"current_steps": 11915, "total_steps": 29704, "loss": 0.3797733783721924, "lr": 6.528780860477648e-06, "epoch": 0.8022488553730137, "percentage": 40.11, "elapsed_time": "1 day, 1:36:40", "remaining_time": "1 day, 14:14:14", "throughput": 5654.0, "total_tokens": 521300416}
+{"current_steps": 11920, "total_steps": 29704, "loss": 0.3375946044921875, "lr": 6.526263190175907e-06, "epoch": 0.8025855103689739, "percentage": 40.13, "elapsed_time": "1 day, 1:37:14", "remaining_time": "1 day, 14:13:28", "throughput": 5654.42, "total_tokens": 521532280}
+{"current_steps": 11925, "total_steps": 29704, "loss": 0.3637096643447876, "lr": 6.523745093059531e-06, "epoch": 0.802922165364934, "percentage": 40.15, "elapsed_time": "1 day, 1:37:56", "remaining_time": "1 day, 14:12:55", "throughput": 5654.22, "total_tokens": 521751944}
+{"current_steps": 11930, "total_steps": 29704, "loss": 0.3677360057830811, "lr": 6.521226569832699e-06, "epoch": 0.8032588203608941, "percentage": 40.16, "elapsed_time": "1 day, 1:38:31", "remaining_time": "1 day, 14:12:10", "throughput": 5654.5, "total_tokens": 521974016}
+{"current_steps": 11935, "total_steps": 29704, "loss": 0.355226993560791, "lr": 6.518707621199707e-06, "epoch": 0.8035954753568543, "percentage": 40.18, "elapsed_time": "1 day, 1:39:13", "remaining_time": "1 day, 14:11:36", "throughput": 5654.27, "total_tokens": 522190040}
+{"current_steps": 11940, "total_steps": 29704, "loss": 0.3752752304077148, "lr": 6.516188247864972e-06, "epoch": 0.8039321303528144, "percentage": 40.2, "elapsed_time": "1 day, 1:39:52", "remaining_time": "1 day, 14:10:59", "throughput": 5654.31, "total_tokens": 522418072}
+{"current_steps": 11945, "total_steps": 29704, "loss": 0.3633857250213623, "lr": 6.513668450533028e-06, "epoch": 0.8042687853487746, "percentage": 40.21, "elapsed_time": "1 day, 1:40:28", "remaining_time": "1 day, 14:10:15", "throughput": 5654.61, "total_tokens": 522645296}
+{"current_steps": 11950, "total_steps": 29704, "loss": 0.39172661304473877, "lr": 6.511148229908531e-06, "epoch": 0.8046054403447347, "percentage": 40.23, "elapsed_time": "1 day, 1:41:01", "remaining_time": "1 day, 14:09:29", "throughput": 5654.98, "total_tokens": 522868752}
+{"current_steps": 11955, "total_steps": 29704, "loss": 0.3619298696517944, "lr": 6.508627586696248e-06, "epoch": 0.8049420953406948, "percentage": 40.25, "elapsed_time": "1 day, 1:41:39", "remaining_time": "1 day, 14:08:49", "throughput": 5654.96, "total_tokens": 523082504}
+{"current_steps": 11960, "total_steps": 29704, "loss": 0.3667879581451416, "lr": 6.5061065216010695e-06, "epoch": 0.805278750336655, "percentage": 40.26, "elapsed_time": "1 day, 1:42:22", "remaining_time": "1 day, 14:08:17", "throughput": 5654.7, "total_tokens": 523299872}
+{"current_steps": 11965, "total_steps": 29704, "loss": 0.3360764503479004, "lr": 6.5035850353280075e-06, "epoch": 0.8056154053326151, "percentage": 40.28, "elapsed_time": "1 day, 1:43:01", "remaining_time": "1 day, 14:07:38", "throughput": 5654.67, "total_tokens": 523516992}
+{"current_steps": 11970, "total_steps": 29704, "loss": 0.35983996391296386, "lr": 6.501063128582183e-06, "epoch": 0.8059520603285752, "percentage": 40.3, "elapsed_time": "1 day, 1:43:35", "remaining_time": "1 day, 14:06:53", "throughput": 5654.82, "total_tokens": 523724536}
+{"current_steps": 11975, "total_steps": 29704, "loss": 0.36988580226898193, "lr": 6.498540802068843e-06, "epoch": 0.8062887153245354, "percentage": 40.31, "elapsed_time": "1 day, 1:44:10", "remaining_time": "1 day, 14:06:09", "throughput": 5655.02, "total_tokens": 523941648}
+{"current_steps": 11980, "total_steps": 29704, "loss": 0.3595852851867676, "lr": 6.4960180564933454e-06, "epoch": 0.8066253703204955, "percentage": 40.33, "elapsed_time": "1 day, 1:44:51", "remaining_time": "1 day, 14:05:33", "throughput": 5654.94, "total_tokens": 524162560}
+{"current_steps": 11985, "total_steps": 29704, "loss": 0.3594093084335327, "lr": 6.493494892561167e-06, "epoch": 0.8069620253164557, "percentage": 40.35, "elapsed_time": "1 day, 1:45:28", "remaining_time": "1 day, 14:04:52", "throughput": 5654.95, "total_tokens": 524374312}
+{"current_steps": 11990, "total_steps": 29704, "loss": 0.3831813812255859, "lr": 6.490971310977907e-06, "epoch": 0.8072986803124158, "percentage": 40.36, "elapsed_time": "1 day, 1:46:05", "remaining_time": "1 day, 14:04:12", "throughput": 5655.06, "total_tokens": 524596088}
+{"current_steps": 11995, "total_steps": 29704, "loss": 0.384428596496582, "lr": 6.488447312449273e-06, "epoch": 0.807635335308376, "percentage": 40.38, "elapsed_time": "1 day, 1:46:36", "remaining_time": "1 day, 14:03:21", "throughput": 5655.62, "total_tokens": 524821680}
+{"current_steps": 12000, "total_steps": 29704, "loss": 0.3725404739379883, "lr": 6.485922897681095e-06, "epoch": 0.8079719903043361, "percentage": 40.4, "elapsed_time": "1 day, 1:47:12", "remaining_time": "1 day, 14:02:38", "throughput": 5655.85, "total_tokens": 525045696}
+{"current_steps": 12005, "total_steps": 29704, "loss": 0.4056462287902832, "lr": 6.483398067379316e-06, "epoch": 0.8083086453002962, "percentage": 40.42, "elapsed_time": "1 day, 1:47:42", "remaining_time": "1 day, 14:01:47", "throughput": 5656.54, "total_tokens": 525279952}
+{"current_steps": 12010, "total_steps": 29704, "loss": 0.3556082248687744, "lr": 6.480872822249998e-06, "epoch": 0.8086453002962564, "percentage": 40.43, "elapsed_time": "1 day, 1:48:24", "remaining_time": "1 day, 14:01:13", "throughput": 5656.42, "total_tokens": 525506864}
+{"current_steps": 12015, "total_steps": 29704, "loss": 0.3839968919754028, "lr": 6.478347162999318e-06, "epoch": 0.8089819552922165, "percentage": 40.45, "elapsed_time": "1 day, 1:49:07", "remaining_time": "1 day, 14:00:41", "throughput": 5656.1, "total_tokens": 525719016}
+{"current_steps": 12020, "total_steps": 29704, "loss": 0.3664535999298096, "lr": 6.475821090333568e-06, "epoch": 0.8093186102881766, "percentage": 40.47, "elapsed_time": "1 day, 1:49:54", "remaining_time": "1 day, 14:00:15", "throughput": 5655.58, "total_tokens": 525939808}
+{"current_steps": 12025, "total_steps": 29704, "loss": 0.38412063121795653, "lr": 6.473294604959156e-06, "epoch": 0.8096552652841368, "percentage": 40.48, "elapsed_time": "1 day, 1:50:31", "remaining_time": "1 day, 13:59:33", "throughput": 5655.62, "total_tokens": 526150200}
+{"current_steps": 12030, "total_steps": 29704, "loss": 0.3502906322479248, "lr": 6.470767707582605e-06, "epoch": 0.8099919202800969, "percentage": 40.5, "elapsed_time": "1 day, 1:51:13", "remaining_time": "1 day, 13:59:00", "throughput": 5655.43, "total_tokens": 526371776}
+{"current_steps": 12035, "total_steps": 29704, "loss": 0.36304702758789065, "lr": 6.468240398910555e-06, "epoch": 0.8103285752760571, "percentage": 40.52, "elapsed_time": "1 day, 1:51:53", "remaining_time": "1 day, 13:58:23", "throughput": 5655.1, "total_tokens": 526566912}
+{"current_steps": 12040, "total_steps": 29704, "loss": 0.3545745849609375, "lr": 6.4657126796497595e-06, "epoch": 0.8106652302720172, "percentage": 40.53, "elapsed_time": "1 day, 1:52:36", "remaining_time": "1 day, 13:57:50", "throughput": 5655.0, "total_tokens": 526798856}
+{"current_steps": 12045, "total_steps": 29704, "loss": 0.37404961585998536, "lr": 6.463184550507087e-06, "epoch": 0.8110018852679773, "percentage": 40.55, "elapsed_time": "1 day, 1:53:21", "remaining_time": "1 day, 13:57:21", "throughput": 5654.55, "total_tokens": 527012296}
+{"current_steps": 12050, "total_steps": 29704, "loss": 0.34106624126434326, "lr": 6.46065601218952e-06, "epoch": 0.8113385402639375, "percentage": 40.57, "elapsed_time": "1 day, 1:53:58", "remaining_time": "1 day, 13:56:40", "throughput": 5654.63, "total_tokens": 527230672}
+{"current_steps": 12055, "total_steps": 29704, "loss": 0.3812570095062256, "lr": 6.458127065404156e-06, "epoch": 0.8116751952598976, "percentage": 40.58, "elapsed_time": "1 day, 1:54:38", "remaining_time": "1 day, 13:56:03", "throughput": 5654.63, "total_tokens": 527453272}
+{"current_steps": 12060, "total_steps": 29704, "loss": 0.35934345722198485, "lr": 6.45559771085821e-06, "epoch": 0.8120118502558578, "percentage": 40.6, "elapsed_time": "1 day, 1:55:16", "remaining_time": "1 day, 13:55:23", "throughput": 5654.76, "total_tokens": 527679752}
+{"current_steps": 12065, "total_steps": 29704, "loss": 0.35833373069763186, "lr": 6.453067949259005e-06, "epoch": 0.8123485052518179, "percentage": 40.62, "elapsed_time": "1 day, 1:55:50", "remaining_time": "1 day, 13:54:38", "throughput": 5655.07, "total_tokens": 527905280}
+{"current_steps": 12070, "total_steps": 29704, "loss": 0.33934841156005857, "lr": 6.450537781313982e-06, "epoch": 0.812685160247778, "percentage": 40.63, "elapsed_time": "1 day, 1:56:38", "remaining_time": "1 day, 13:54:13", "throughput": 5654.46, "total_tokens": 528118504}
+{"current_steps": 12075, "total_steps": 29704, "loss": 0.3756300926208496, "lr": 6.448007207730693e-06, "epoch": 0.8130218152437382, "percentage": 40.65, "elapsed_time": "1 day, 1:57:16", "remaining_time": "1 day, 13:53:33", "throughput": 5654.54, "total_tokens": 528340504}
+{"current_steps": 12080, "total_steps": 29704, "loss": 0.35174586772918703, "lr": 6.445476229216805e-06, "epoch": 0.8133584702396983, "percentage": 40.67, "elapsed_time": "1 day, 1:57:54", "remaining_time": "1 day, 13:52:53", "throughput": 5654.76, "total_tokens": 528575776}
+{"current_steps": 12085, "total_steps": 29704, "loss": 0.39254226684570315, "lr": 6.442944846480099e-06, "epoch": 0.8136951252356585, "percentage": 40.68, "elapsed_time": "1 day, 1:58:35", "remaining_time": "1 day, 13:52:18", "throughput": 5654.54, "total_tokens": 528786904}
+{"current_steps": 12090, "total_steps": 29704, "loss": 0.39618897438049316, "lr": 6.44041306022847e-06, "epoch": 0.8140317802316186, "percentage": 40.7, "elapsed_time": "1 day, 1:59:14", "remaining_time": "1 day, 13:51:40", "throughput": 5654.63, "total_tokens": 529015944}
+{"current_steps": 12095, "total_steps": 29704, "loss": 0.39859035015106203, "lr": 6.43788087116992e-06, "epoch": 0.8143684352275787, "percentage": 40.72, "elapsed_time": "1 day, 1:59:55", "remaining_time": "1 day, 13:51:04", "throughput": 5654.63, "total_tokens": 529246224}
+{"current_steps": 12100, "total_steps": 29704, "loss": 0.3658564567565918, "lr": 6.43534828001257e-06, "epoch": 0.8147050902235389, "percentage": 40.74, "elapsed_time": "1 day, 2:00:26", "remaining_time": "1 day, 13:50:14", "throughput": 5655.21, "total_tokens": 529476248}
+{"current_steps": 12105, "total_steps": 29704, "loss": 0.372475266456604, "lr": 6.43281528746465e-06, "epoch": 0.815041745219499, "percentage": 40.75, "elapsed_time": "1 day, 2:01:00", "remaining_time": "1 day, 13:49:29", "throughput": 5655.34, "total_tokens": 529683080}
+{"current_steps": 12110, "total_steps": 29704, "loss": 0.36925411224365234, "lr": 6.430281894234504e-06, "epoch": 0.8153784002154592, "percentage": 40.77, "elapsed_time": "1 day, 2:01:49", "remaining_time": "1 day, 13:49:05", "throughput": 5654.79, "total_tokens": 529906632}
+{"current_steps": 12115, "total_steps": 29704, "loss": 0.3794065237045288, "lr": 6.427748101030587e-06, "epoch": 0.8157150552114193, "percentage": 40.79, "elapsed_time": "1 day, 2:02:26", "remaining_time": "1 day, 13:48:24", "throughput": 5654.85, "total_tokens": 530122608}
+{"current_steps": 12120, "total_steps": 29704, "loss": 0.3644280433654785, "lr": 6.425213908561468e-06, "epoch": 0.8160517102073794, "percentage": 40.8, "elapsed_time": "1 day, 2:03:02", "remaining_time": "1 day, 13:47:42", "throughput": 5655.04, "total_tokens": 530344760}
+{"current_steps": 12125, "total_steps": 29704, "loss": 0.3801795721054077, "lr": 6.422679317535823e-06, "epoch": 0.8163883652033396, "percentage": 40.82, "elapsed_time": "1 day, 2:03:38", "remaining_time": "1 day, 13:46:59", "throughput": 5655.11, "total_tokens": 530554320}
+{"current_steps": 12130, "total_steps": 29704, "loss": 0.4110711097717285, "lr": 6.420144328662441e-06, "epoch": 0.8167250201992997, "percentage": 40.84, "elapsed_time": "1 day, 2:04:18", "remaining_time": "1 day, 13:46:23", "throughput": 5655.12, "total_tokens": 530784216}
+{"current_steps": 12135, "total_steps": 29704, "loss": 0.36923315525054934, "lr": 6.417608942650229e-06, "epoch": 0.8170616751952599, "percentage": 40.85, "elapsed_time": "1 day, 2:05:00", "remaining_time": "1 day, 13:45:48", "throughput": 5654.78, "total_tokens": 530985400}
+{"current_steps": 12140, "total_steps": 29704, "loss": 0.368359637260437, "lr": 6.415073160208196e-06, "epoch": 0.81739833019122, "percentage": 40.87, "elapsed_time": "1 day, 2:05:38", "remaining_time": "1 day, 13:45:09", "throughput": 5654.97, "total_tokens": 531219992}
+{"current_steps": 12145, "total_steps": 29704, "loss": 0.38031327724456787, "lr": 6.412536982045465e-06, "epoch": 0.8177349851871801, "percentage": 40.89, "elapsed_time": "1 day, 2:06:15", "remaining_time": "1 day, 13:44:27", "throughput": 5654.87, "total_tokens": 531416856}
+{"current_steps": 12150, "total_steps": 29704, "loss": 0.3821054458618164, "lr": 6.410000408871273e-06, "epoch": 0.8180716401831403, "percentage": 40.9, "elapsed_time": "1 day, 2:06:50", "remaining_time": "1 day, 13:43:43", "throughput": 5655.05, "total_tokens": 531631960}
+{"current_steps": 12155, "total_steps": 29704, "loss": 0.37601168155670167, "lr": 6.407463441394961e-06, "epoch": 0.8184082951791004, "percentage": 40.92, "elapsed_time": "1 day, 2:07:26", "remaining_time": "1 day, 13:43:00", "throughput": 5655.3, "total_tokens": 531859288}
+{"current_steps": 12160, "total_steps": 29704, "loss": 0.38929071426391604, "lr": 6.4049260803259874e-06, "epoch": 0.8187449501750605, "percentage": 40.94, "elapsed_time": "1 day, 2:08:08", "remaining_time": "1 day, 13:42:27", "throughput": 5655.14, "total_tokens": 532082152}
+{"current_steps": 12165, "total_steps": 29704, "loss": 0.3606018304824829, "lr": 6.402388326373914e-06, "epoch": 0.8190816051710207, "percentage": 40.95, "elapsed_time": "1 day, 2:08:48", "remaining_time": "1 day, 13:41:51", "throughput": 5655.06, "total_tokens": 532304432}
+{"current_steps": 12170, "total_steps": 29704, "loss": 0.4054881572723389, "lr": 6.399850180248418e-06, "epoch": 0.8194182601669808, "percentage": 40.97, "elapsed_time": "1 day, 2:09:27", "remaining_time": "1 day, 13:41:11", "throughput": 5655.24, "total_tokens": 532538000}
+{"current_steps": 12175, "total_steps": 29704, "loss": 0.3806436538696289, "lr": 6.397311642659283e-06, "epoch": 0.819754915162941, "percentage": 40.99, "elapsed_time": "1 day, 2:10:03", "remaining_time": "1 day, 13:40:29", "throughput": 5655.33, "total_tokens": 532749856}
+{"current_steps": 12180, "total_steps": 29704, "loss": 0.3875831842422485, "lr": 6.394772714316402e-06, "epoch": 0.8200915701589011, "percentage": 41.0, "elapsed_time": "1 day, 2:10:38", "remaining_time": "1 day, 13:39:45", "throughput": 5655.66, "total_tokens": 532981096}
+{"current_steps": 12185, "total_steps": 29704, "loss": 0.34835638999938967, "lr": 6.392233395929779e-06, "epoch": 0.8204282251548612, "percentage": 41.02, "elapsed_time": "1 day, 2:11:07", "remaining_time": "1 day, 13:38:53", "throughput": 5656.16, "total_tokens": 533193808}
+{"current_steps": 12190, "total_steps": 29704, "loss": 0.362264609336853, "lr": 6.389693688209526e-06, "epoch": 0.8207648801508214, "percentage": 41.04, "elapsed_time": "1 day, 2:11:47", "remaining_time": "1 day, 13:38:16", "throughput": 5656.11, "total_tokens": 533413896}
+{"current_steps": 12195, "total_steps": 29704, "loss": 0.3703956604003906, "lr": 6.387153591865867e-06, "epoch": 0.8211015351467815, "percentage": 41.06, "elapsed_time": "1 day, 2:12:31", "remaining_time": "1 day, 13:37:45", "throughput": 5655.91, "total_tokens": 533642648}
+{"current_steps": 12200, "total_steps": 29704, "loss": 0.38781538009643557, "lr": 6.384613107609127e-06, "epoch": 0.8214381901427417, "percentage": 41.07, "elapsed_time": "1 day, 2:13:08", "remaining_time": "1 day, 13:37:03", "throughput": 5655.97, "total_tokens": 533855672}
+{"current_steps": 12205, "total_steps": 29704, "loss": 0.3636336326599121, "lr": 6.382072236149747e-06, "epoch": 0.8217748451387018, "percentage": 41.09, "elapsed_time": "1 day, 2:13:38", "remaining_time": "1 day, 13:36:12", "throughput": 5656.61, "total_tokens": 534086688}
+{"current_steps": 12210, "total_steps": 29704, "loss": 0.33962430953979494, "lr": 6.379530978198273e-06, "epoch": 0.8221115001346619, "percentage": 41.11, "elapsed_time": "1 day, 2:14:12", "remaining_time": "1 day, 13:35:28", "throughput": 5656.75, "total_tokens": 534295064}
+{"current_steps": 12215, "total_steps": 29704, "loss": 0.34957318305969237, "lr": 6.376989334465361e-06, "epoch": 0.8224481551306222, "percentage": 41.12, "elapsed_time": "1 day, 2:14:47", "remaining_time": "1 day, 13:34:43", "throughput": 5657.03, "total_tokens": 534518376}
+{"current_steps": 12220, "total_steps": 29704, "loss": 0.359330415725708, "lr": 6.374447305661772e-06, "epoch": 0.8227848101265823, "percentage": 41.14, "elapsed_time": "1 day, 2:15:26", "remaining_time": "1 day, 13:34:05", "throughput": 5657.16, "total_tokens": 534753352}
+{"current_steps": 12225, "total_steps": 29704, "loss": 0.3536032199859619, "lr": 6.3719048924983776e-06, "epoch": 0.8231214651225425, "percentage": 41.16, "elapsed_time": "1 day, 2:16:05", "remaining_time": "1 day, 13:33:28", "throughput": 5657.0, "total_tokens": 534959200}
+{"current_steps": 12230, "total_steps": 29704, "loss": 0.3803597688674927, "lr": 6.369362095686152e-06, "epoch": 0.8234581201185026, "percentage": 41.17, "elapsed_time": "1 day, 2:16:41", "remaining_time": "1 day, 13:32:44", "throughput": 5657.17, "total_tokens": 535175808}
+{"current_steps": 12235, "total_steps": 29704, "loss": 0.34799120426177976, "lr": 6.366818915936185e-06, "epoch": 0.8237947751144628, "percentage": 41.19, "elapsed_time": "1 day, 2:17:17", "remaining_time": "1 day, 13:32:02", "throughput": 5657.29, "total_tokens": 535394008}
+{"current_steps": 12240, "total_steps": 29704, "loss": 0.3779747486114502, "lr": 6.364275353959667e-06, "epoch": 0.8241314301104229, "percentage": 41.21, "elapsed_time": "1 day, 2:17:59", "remaining_time": "1 day, 13:31:28", "throughput": 5657.02, "total_tokens": 535603560}
+{"current_steps": 12245, "total_steps": 29704, "loss": 0.36225066184997556, "lr": 6.3617314104678966e-06, "epoch": 0.824468085106383, "percentage": 41.22, "elapsed_time": "1 day, 2:18:40", "remaining_time": "1 day, 13:30:52", "throughput": 5657.07, "total_tokens": 535838080}
+{"current_steps": 12250, "total_steps": 29704, "loss": 0.3977229118347168, "lr": 6.359187086172278e-06, "epoch": 0.8248047401023432, "percentage": 41.24, "elapsed_time": "1 day, 2:19:15", "remaining_time": "1 day, 13:30:09", "throughput": 5657.11, "total_tokens": 536041408}
+{"current_steps": 12255, "total_steps": 29704, "loss": 0.3828271389007568, "lr": 6.356642381784326e-06, "epoch": 0.8251413950983033, "percentage": 41.26, "elapsed_time": "1 day, 2:19:49", "remaining_time": "1 day, 13:29:23", "throughput": 5657.33, "total_tokens": 536255712}
+{"current_steps": 12260, "total_steps": 29704, "loss": 0.3526939392089844, "lr": 6.354097298015658e-06, "epoch": 0.8254780500942634, "percentage": 41.27, "elapsed_time": "1 day, 2:20:28", "remaining_time": "1 day, 13:28:45", "throughput": 5657.08, "total_tokens": 536450456}
+{"current_steps": 12265, "total_steps": 29704, "loss": 0.3932209014892578, "lr": 6.3515518355779985e-06, "epoch": 0.8258147050902236, "percentage": 41.29, "elapsed_time": "1 day, 2:21:05", "remaining_time": "1 day, 13:28:05", "throughput": 5656.98, "total_tokens": 536655032}
+{"current_steps": 12270, "total_steps": 29704, "loss": 0.3485422134399414, "lr": 6.349005995183177e-06, "epoch": 0.8261513600861837, "percentage": 41.31, "elapsed_time": "1 day, 2:21:39", "remaining_time": "1 day, 13:27:19", "throughput": 5657.38, "total_tokens": 536883576}
+{"current_steps": 12275, "total_steps": 29704, "loss": 0.3577850341796875, "lr": 6.346459777543131e-06, "epoch": 0.8264880150821439, "percentage": 41.32, "elapsed_time": "1 day, 2:22:13", "remaining_time": "1 day, 13:26:33", "throughput": 5657.67, "total_tokens": 537102112}
+{"current_steps": 12280, "total_steps": 29704, "loss": 0.3586315155029297, "lr": 6.343913183369902e-06, "epoch": 0.826824670078104, "percentage": 41.34, "elapsed_time": "1 day, 2:22:50", "remaining_time": "1 day, 13:25:53", "throughput": 5657.6, "total_tokens": 537306360}
+{"current_steps": 12285, "total_steps": 29704, "loss": 0.3537776231765747, "lr": 6.341366213375637e-06, "epoch": 0.8271613250740641, "percentage": 41.36, "elapsed_time": "1 day, 2:23:23", "remaining_time": "1 day, 13:25:05", "throughput": 5658.06, "total_tokens": 537534592}
+{"current_steps": 12290, "total_steps": 29704, "loss": 0.3713988780975342, "lr": 6.338818868272587e-06, "epoch": 0.8274979800700243, "percentage": 41.37, "elapsed_time": "1 day, 2:24:01", "remaining_time": "1 day, 13:24:26", "throughput": 5658.05, "total_tokens": 537747240}
+{"current_steps": 12295, "total_steps": 29704, "loss": 0.3642918109893799, "lr": 6.336271148773111e-06, "epoch": 0.8278346350659844, "percentage": 41.39, "elapsed_time": "1 day, 2:24:45", "remaining_time": "1 day, 13:23:55", "throughput": 5657.88, "total_tokens": 537982080}
+{"current_steps": 12300, "total_steps": 29704, "loss": 0.4124462127685547, "lr": 6.333723055589669e-06, "epoch": 0.8281712900619446, "percentage": 41.41, "elapsed_time": "1 day, 2:25:23", "remaining_time": "1 day, 13:23:16", "throughput": 5657.91, "total_tokens": 538202248}
+{"current_steps": 12305, "total_steps": 29704, "loss": 0.3763313293457031, "lr": 6.331174589434826e-06, "epoch": 0.8285079450579047, "percentage": 41.43, "elapsed_time": "1 day, 2:26:10", "remaining_time": "1 day, 13:22:49", "throughput": 5657.41, "total_tokens": 538420056}
+{"current_steps": 12310, "total_steps": 29704, "loss": 0.35441150665283205, "lr": 6.328625751021254e-06, "epoch": 0.8288446000538648, "percentage": 41.44, "elapsed_time": "1 day, 2:26:54", "remaining_time": "1 day, 13:22:17", "throughput": 5657.01, "total_tokens": 538627360}
+{"current_steps": 12315, "total_steps": 29704, "loss": 0.334460711479187, "lr": 6.326076541061729e-06, "epoch": 0.829181255049825, "percentage": 41.46, "elapsed_time": "1 day, 2:27:35", "remaining_time": "1 day, 13:21:42", "throughput": 5656.84, "total_tokens": 538844328}
+{"current_steps": 12320, "total_steps": 29704, "loss": 0.3949697971343994, "lr": 6.323526960269127e-06, "epoch": 0.8295179100457851, "percentage": 41.48, "elapsed_time": "1 day, 2:28:11", "remaining_time": "1 day, 13:20:59", "throughput": 5656.91, "total_tokens": 539054992}
+{"current_steps": 12325, "total_steps": 29704, "loss": 0.3561330556869507, "lr": 6.3209770093564315e-06, "epoch": 0.8298545650417453, "percentage": 41.49, "elapsed_time": "1 day, 2:28:52", "remaining_time": "1 day, 13:20:24", "throughput": 5656.85, "total_tokens": 539280216}
+{"current_steps": 12330, "total_steps": 29704, "loss": 0.3534719705581665, "lr": 6.318426689036727e-06, "epoch": 0.8301912200377054, "percentage": 41.51, "elapsed_time": "1 day, 2:29:36", "remaining_time": "1 day, 13:19:52", "throughput": 5656.49, "total_tokens": 539494040}
+{"current_steps": 12335, "total_steps": 29704, "loss": 0.37946810722351076, "lr": 6.315876000023202e-06, "epoch": 0.8305278750336655, "percentage": 41.53, "elapsed_time": "1 day, 2:30:16", "remaining_time": "1 day, 13:19:16", "throughput": 5656.31, "total_tokens": 539706008}
+{"current_steps": 12340, "total_steps": 29704, "loss": 0.3778734922409058, "lr": 6.313324943029151e-06, "epoch": 0.8308645300296257, "percentage": 41.54, "elapsed_time": "1 day, 2:31:00", "remaining_time": "1 day, 13:18:45", "throughput": 5655.94, "total_tokens": 539917272}
+{"current_steps": 12345, "total_steps": 29704, "loss": 0.35590362548828125, "lr": 6.310773518767967e-06, "epoch": 0.8312011850255858, "percentage": 41.56, "elapsed_time": "1 day, 2:31:33", "remaining_time": "1 day, 13:17:58", "throughput": 5656.19, "total_tokens": 540130136}
+{"current_steps": 12350, "total_steps": 29704, "loss": 0.36078743934631347, "lr": 6.30822172795315e-06, "epoch": 0.831537840021546, "percentage": 41.58, "elapsed_time": "1 day, 2:32:08", "remaining_time": "1 day, 13:17:14", "throughput": 5656.3, "total_tokens": 540336200}
+{"current_steps": 12355, "total_steps": 29704, "loss": 0.4154792785644531, "lr": 6.3056695712982965e-06, "epoch": 0.8318744950175061, "percentage": 41.59, "elapsed_time": "1 day, 2:32:39", "remaining_time": "1 day, 13:16:25", "throughput": 5656.78, "total_tokens": 540561200}
+{"current_steps": 12360, "total_steps": 29704, "loss": 0.3399226427078247, "lr": 6.303117049517111e-06, "epoch": 0.8322111500134662, "percentage": 41.61, "elapsed_time": "1 day, 2:33:20", "remaining_time": "1 day, 13:15:49", "throughput": 5656.64, "total_tokens": 540776592}
+{"current_steps": 12365, "total_steps": 29704, "loss": 0.35541400909423826, "lr": 6.300564163323398e-06, "epoch": 0.8325478050094264, "percentage": 41.63, "elapsed_time": "1 day, 2:34:00", "remaining_time": "1 day, 13:15:13", "throughput": 5656.48, "total_tokens": 540989568}
+{"current_steps": 12370, "total_steps": 29704, "loss": 0.36992754936218264, "lr": 6.298010913431065e-06, "epoch": 0.8328844600053865, "percentage": 41.64, "elapsed_time": "1 day, 2:34:39", "remaining_time": "1 day, 13:14:34", "throughput": 5656.47, "total_tokens": 541205456}
+{"current_steps": 12375, "total_steps": 29704, "loss": 0.3586021184921265, "lr": 6.295457300554119e-06, "epoch": 0.8332211150013467, "percentage": 41.66, "elapsed_time": "1 day, 2:35:17", "remaining_time": "1 day, 13:13:55", "throughput": 5656.59, "total_tokens": 541434640}
+{"current_steps": 12380, "total_steps": 29704, "loss": 0.38977057933807374, "lr": 6.292903325406668e-06, "epoch": 0.8335577699973068, "percentage": 41.68, "elapsed_time": "1 day, 2:35:51", "remaining_time": "1 day, 13:13:10", "throughput": 5656.68, "total_tokens": 541635304}
+{"current_steps": 12385, "total_steps": 29704, "loss": 0.34544246196746825, "lr": 6.2903489887029255e-06, "epoch": 0.8338944249932669, "percentage": 41.69, "elapsed_time": "1 day, 2:36:26", "remaining_time": "1 day, 13:12:26", "throughput": 5656.92, "total_tokens": 541858656}
+{"current_steps": 12390, "total_steps": 29704, "loss": 0.36299178600311277, "lr": 6.287794291157204e-06, "epoch": 0.8342310799892271, "percentage": 41.71, "elapsed_time": "1 day, 2:37:00", "remaining_time": "1 day, 13:11:41", "throughput": 5657.19, "total_tokens": 542076592}
+{"current_steps": 12395, "total_steps": 29704, "loss": 0.36720116138458253, "lr": 6.285239233483915e-06, "epoch": 0.8345677349851872, "percentage": 41.73, "elapsed_time": "1 day, 2:37:39", "remaining_time": "1 day, 13:11:03", "throughput": 5657.3, "total_tokens": 542309048}
+{"current_steps": 12400, "total_steps": 29704, "loss": 0.3783855438232422, "lr": 6.2826838163975724e-06, "epoch": 0.8349043899811474, "percentage": 41.75, "elapsed_time": "1 day, 2:38:18", "remaining_time": "1 day, 13:10:24", "throughput": 5657.3, "total_tokens": 542526280}
+{"current_steps": 12405, "total_steps": 29704, "loss": 0.35349900722503663, "lr": 6.2801280406127905e-06, "epoch": 0.8352410449771075, "percentage": 41.76, "elapsed_time": "1 day, 2:38:54", "remaining_time": "1 day, 13:09:42", "throughput": 5657.49, "total_tokens": 542750064}
+{"current_steps": 12410, "total_steps": 29704, "loss": 0.36704421043395996, "lr": 6.2775719068442845e-06, "epoch": 0.8355776999730676, "percentage": 41.78, "elapsed_time": "1 day, 2:39:32", "remaining_time": "1 day, 13:09:02", "throughput": 5657.55, "total_tokens": 542968240}
+{"current_steps": 12415, "total_steps": 29704, "loss": 0.3532416343688965, "lr": 6.275015415806869e-06, "epoch": 0.8359143549690278, "percentage": 41.8, "elapsed_time": "1 day, 2:40:16", "remaining_time": "1 day, 13:08:31", "throughput": 5657.11, "total_tokens": 543175408}
+{"current_steps": 12420, "total_steps": 29704, "loss": 0.3773714303970337, "lr": 6.272458568215458e-06, "epoch": 0.8362510099649879, "percentage": 41.81, "elapsed_time": "1 day, 2:40:57", "remaining_time": "1 day, 13:07:55", "throughput": 5656.95, "total_tokens": 543389992}
+{"current_steps": 12425, "total_steps": 29704, "loss": 0.37525081634521484, "lr": 6.269901364785066e-06, "epoch": 0.836587664960948, "percentage": 41.83, "elapsed_time": "1 day, 2:41:35", "remaining_time": "1 day, 13:07:16", "throughput": 5657.04, "total_tokens": 543616944}
+{"current_steps": 12430, "total_steps": 29704, "loss": 0.37595961093902586, "lr": 6.2673438062308055e-06, "epoch": 0.8369243199569082, "percentage": 41.85, "elapsed_time": "1 day, 2:42:11", "remaining_time": "1 day, 13:06:33", "throughput": 5657.47, "total_tokens": 543859984}
+{"current_steps": 12435, "total_steps": 29704, "loss": 0.3472005367279053, "lr": 6.264785893267892e-06, "epoch": 0.8372609749528683, "percentage": 41.86, "elapsed_time": "1 day, 2:42:42", "remaining_time": "1 day, 13:05:45", "throughput": 5657.73, "total_tokens": 544063976}
+{"current_steps": 12440, "total_steps": 29704, "loss": 0.36200590133666993, "lr": 6.2622276266116366e-06, "epoch": 0.8375976299488285, "percentage": 41.88, "elapsed_time": "1 day, 2:43:19", "remaining_time": "1 day, 13:05:04", "throughput": 5657.9, "total_tokens": 544288640}
+{"current_steps": 12445, "total_steps": 29704, "loss": 0.37799394130706787, "lr": 6.25966900697745e-06, "epoch": 0.8379342849447886, "percentage": 41.9, "elapsed_time": "1 day, 2:43:57", "remaining_time": "1 day, 13:04:24", "throughput": 5658.03, "total_tokens": 544514152}
+{"current_steps": 12450, "total_steps": 29704, "loss": 0.3491935729980469, "lr": 6.257110035080843e-06, "epoch": 0.8382709399407487, "percentage": 41.91, "elapsed_time": "1 day, 2:44:34", "remaining_time": "1 day, 13:03:43", "throughput": 5658.2, "total_tokens": 544741824}
+{"current_steps": 12455, "total_steps": 29704, "loss": 0.40814781188964844, "lr": 6.254550711637422e-06, "epoch": 0.8386075949367089, "percentage": 41.93, "elapsed_time": "1 day, 2:45:14", "remaining_time": "1 day, 13:03:06", "throughput": 5658.03, "total_tokens": 544950448}
+{"current_steps": 12460, "total_steps": 29704, "loss": 0.38295414447784426, "lr": 6.251991037362897e-06, "epoch": 0.838944249932669, "percentage": 41.95, "elapsed_time": "1 day, 2:45:49", "remaining_time": "1 day, 13:02:22", "throughput": 5658.15, "total_tokens": 545159072}
+{"current_steps": 12465, "total_steps": 29704, "loss": 0.37713871002197263, "lr": 6.249431012973068e-06, "epoch": 0.8392809049286292, "percentage": 41.96, "elapsed_time": "1 day, 2:46:28", "remaining_time": "1 day, 13:01:44", "throughput": 5658.0, "total_tokens": 545367120}
+{"current_steps": 12470, "total_steps": 29704, "loss": 0.37398107051849366, "lr": 6.246870639183843e-06, "epoch": 0.8396175599245893, "percentage": 41.98, "elapsed_time": "1 day, 2:47:06", "remaining_time": "1 day, 13:01:05", "throughput": 5657.97, "total_tokens": 545578664}
+{"current_steps": 12475, "total_steps": 29704, "loss": 0.39701476097106936, "lr": 6.2443099167112185e-06, "epoch": 0.8399542149205494, "percentage": 42.0, "elapsed_time": "1 day, 2:47:42", "remaining_time": "1 day, 13:00:22", "throughput": 5658.19, "total_tokens": 545803912}
+{"current_steps": 12480, "total_steps": 29704, "loss": 0.3778507709503174, "lr": 6.241748846271293e-06, "epoch": 0.8402908699165096, "percentage": 42.01, "elapsed_time": "1 day, 2:48:16", "remaining_time": "1 day, 12:59:37", "throughput": 5658.55, "total_tokens": 546030536}
+{"current_steps": 12485, "total_steps": 29704, "loss": 0.37636547088623046, "lr": 6.239187428580263e-06, "epoch": 0.8406275249124697, "percentage": 42.03, "elapsed_time": "1 day, 2:48:58", "remaining_time": "1 day, 12:59:03", "throughput": 5658.22, "total_tokens": 546236768}
+{"current_steps": 12490, "total_steps": 29704, "loss": 0.3784951686859131, "lr": 6.236625664354421e-06, "epoch": 0.8409641799084299, "percentage": 42.05, "elapsed_time": "1 day, 2:49:34", "remaining_time": "1 day, 12:58:21", "throughput": 5658.44, "total_tokens": 546462096}
+{"current_steps": 12495, "total_steps": 29704, "loss": 0.3676713466644287, "lr": 6.2340635543101545e-06, "epoch": 0.84130083490439, "percentage": 42.07, "elapsed_time": "1 day, 2:50:14", "remaining_time": "1 day, 12:57:44", "throughput": 5658.13, "total_tokens": 546657168}
+{"current_steps": 12500, "total_steps": 29704, "loss": 0.33987085819244384, "lr": 6.23150109916395e-06, "epoch": 0.8416374899003501, "percentage": 42.08, "elapsed_time": "1 day, 2:50:55", "remaining_time": "1 day, 12:57:08", "throughput": 5658.06, "total_tokens": 546881208}
+{"current_steps": 12505, "total_steps": 29704, "loss": 0.3859384059906006, "lr": 6.22893829963239e-06, "epoch": 0.8419741448963103, "percentage": 42.1, "elapsed_time": "1 day, 2:51:31", "remaining_time": "1 day, 12:56:26", "throughput": 5658.14, "total_tokens": 547092816}
+{"current_steps": 12510, "total_steps": 29704, "loss": 0.357137393951416, "lr": 6.226375156432153e-06, "epoch": 0.8423107998922704, "percentage": 42.12, "elapsed_time": "1 day, 2:52:11", "remaining_time": "1 day, 12:55:49", "throughput": 5658.16, "total_tokens": 547322656}
+{"current_steps": 12515, "total_steps": 29704, "loss": 0.3710299491882324, "lr": 6.2238116702800145e-06, "epoch": 0.8426474548882306, "percentage": 42.13, "elapsed_time": "1 day, 2:52:42", "remaining_time": "1 day, 12:55:00", "throughput": 5658.58, "total_tokens": 547537944}
+{"current_steps": 12520, "total_steps": 29704, "loss": 0.3879734992980957, "lr": 6.221247841892846e-06, "epoch": 0.8429841098841907, "percentage": 42.15, "elapsed_time": "1 day, 2:53:25", "remaining_time": "1 day, 12:54:28", "throughput": 5658.21, "total_tokens": 547746624}
+{"current_steps": 12525, "total_steps": 29704, "loss": 0.3691586971282959, "lr": 6.218683671987611e-06, "epoch": 0.8433207648801508, "percentage": 42.17, "elapsed_time": "1 day, 2:54:00", "remaining_time": "1 day, 12:53:43", "throughput": 5658.63, "total_tokens": 547983928}
+{"current_steps": 12530, "total_steps": 29704, "loss": 0.37589943408966064, "lr": 6.216119161281373e-06, "epoch": 0.843657419876111, "percentage": 42.18, "elapsed_time": "1 day, 2:54:39", "remaining_time": "1 day, 12:53:05", "throughput": 5658.7, "total_tokens": 548209600}
+{"current_steps": 12535, "total_steps": 29704, "loss": 0.3672966003417969, "lr": 6.213554310491291e-06, "epoch": 0.8439940748720711, "percentage": 42.2, "elapsed_time": "1 day, 2:55:28", "remaining_time": "1 day, 12:52:41", "throughput": 5658.0, "total_tokens": 548421688}
+{"current_steps": 12540, "total_steps": 29704, "loss": 0.362609338760376, "lr": 6.210989120334616e-06, "epoch": 0.8443307298680313, "percentage": 42.22, "elapsed_time": "1 day, 2:56:11", "remaining_time": "1 day, 12:52:08", "throughput": 5657.78, "total_tokens": 548644072}
+{"current_steps": 12545, "total_steps": 29704, "loss": 0.38643317222595214, "lr": 6.208423591528694e-06, "epoch": 0.8446673848639914, "percentage": 42.23, "elapsed_time": "1 day, 2:56:49", "remaining_time": "1 day, 12:51:29", "throughput": 5657.72, "total_tokens": 548854336}
+{"current_steps": 12550, "total_steps": 29704, "loss": 0.3612468481063843, "lr": 6.205857724790969e-06, "epoch": 0.8450040398599515, "percentage": 42.25, "elapsed_time": "1 day, 2:57:28", "remaining_time": "1 day, 12:50:51", "throughput": 5657.71, "total_tokens": 549072600}
+{"current_steps": 12555, "total_steps": 29704, "loss": 0.409790563583374, "lr": 6.203291520838974e-06, "epoch": 0.8453406948559117, "percentage": 42.27, "elapsed_time": "1 day, 2:58:06", "remaining_time": "1 day, 12:50:11", "throughput": 5657.9, "total_tokens": 549303984}
+{"current_steps": 12560, "total_steps": 29704, "loss": 0.34305951595306394, "lr": 6.200724980390344e-06, "epoch": 0.8456773498518718, "percentage": 42.28, "elapsed_time": "1 day, 2:58:45", "remaining_time": "1 day, 12:49:33", "throughput": 5657.61, "total_tokens": 549497696}
+{"current_steps": 12565, "total_steps": 29704, "loss": 0.3178701400756836, "lr": 6.198158104162804e-06, "epoch": 0.846014004847832, "percentage": 42.3, "elapsed_time": "1 day, 2:59:18", "remaining_time": "1 day, 12:48:47", "throughput": 5657.88, "total_tokens": 549713600}
+{"current_steps": 12570, "total_steps": 29704, "loss": 0.37602977752685546, "lr": 6.1955908928741685e-06, "epoch": 0.8463506598437921, "percentage": 42.32, "elapsed_time": "1 day, 2:59:55", "remaining_time": "1 day, 12:48:06", "throughput": 5658.2, "total_tokens": 549953184}
+{"current_steps": 12575, "total_steps": 29704, "loss": 0.38083457946777344, "lr": 6.193023347242353e-06, "epoch": 0.8466873148397522, "percentage": 42.33, "elapsed_time": "1 day, 3:00:32", "remaining_time": "1 day, 12:47:24", "throughput": 5658.45, "total_tokens": 550184352}
+{"current_steps": 12580, "total_steps": 29704, "loss": 0.3633177042007446, "lr": 6.190455467985361e-06, "epoch": 0.8470239698357124, "percentage": 42.35, "elapsed_time": "1 day, 3:01:12", "remaining_time": "1 day, 12:46:48", "throughput": 5658.37, "total_tokens": 550405552}
+{"current_steps": 12585, "total_steps": 29704, "loss": 0.3954836368560791, "lr": 6.187887255821295e-06, "epoch": 0.8473606248316725, "percentage": 42.37, "elapsed_time": "1 day, 3:01:45", "remaining_time": "1 day, 12:46:02", "throughput": 5658.71, "total_tokens": 550625896}
+{"current_steps": 12590, "total_steps": 29704, "loss": 0.3501850128173828, "lr": 6.185318711468345e-06, "epoch": 0.8476972798276327, "percentage": 42.38, "elapsed_time": "1 day, 3:02:26", "remaining_time": "1 day, 12:45:25", "throughput": 5658.44, "total_tokens": 550827088}
+{"current_steps": 12595, "total_steps": 29704, "loss": 0.3587897777557373, "lr": 6.182749835644799e-06, "epoch": 0.8480339348235928, "percentage": 42.4, "elapsed_time": "1 day, 3:02:58", "remaining_time": "1 day, 12:44:39", "throughput": 5658.69, "total_tokens": 551036712}
+{"current_steps": 12600, "total_steps": 29704, "loss": 0.3534966468811035, "lr": 6.180180629069031e-06, "epoch": 0.8483705898195529, "percentage": 42.42, "elapsed_time": "1 day, 3:03:36", "remaining_time": "1 day, 12:43:59", "throughput": 5658.86, "total_tokens": 551269008}
+{"current_steps": 12605, "total_steps": 29704, "loss": 0.3404214859008789, "lr": 6.177611092459514e-06, "epoch": 0.8487072448155131, "percentage": 42.44, "elapsed_time": "1 day, 3:04:14", "remaining_time": "1 day, 12:43:20", "throughput": 5658.89, "total_tokens": 551486584}
+{"current_steps": 12610, "total_steps": 29704, "loss": 0.3584805965423584, "lr": 6.175041226534809e-06, "epoch": 0.8490438998114732, "percentage": 42.45, "elapsed_time": "1 day, 3:04:47", "remaining_time": "1 day, 12:42:32", "throughput": 5659.28, "total_tokens": 551707952}
+{"current_steps": 12615, "total_steps": 29704, "loss": 0.327656888961792, "lr": 6.172471032013574e-06, "epoch": 0.8493805548074334, "percentage": 42.47, "elapsed_time": "1 day, 3:05:26", "remaining_time": "1 day, 12:41:55", "throughput": 5659.28, "total_tokens": 551932560}
+{"current_steps": 12620, "total_steps": 29704, "loss": 0.3601629972457886, "lr": 6.169900509614553e-06, "epoch": 0.8497172098033935, "percentage": 42.49, "elapsed_time": "1 day, 3:06:17", "remaining_time": "1 day, 12:41:33", "throughput": 5658.71, "total_tokens": 552163800}
+{"current_steps": 12625, "total_steps": 29704, "loss": 0.36720144748687744, "lr": 6.167329660056585e-06, "epoch": 0.8500538647993536, "percentage": 42.5, "elapsed_time": "1 day, 3:06:57", "remaining_time": "1 day, 12:40:56", "throughput": 5658.56, "total_tokens": 552375472}
+{"current_steps": 12630, "total_steps": 29704, "loss": 0.36503446102142334, "lr": 6.1647584840586e-06, "epoch": 0.8503905197953138, "percentage": 42.52, "elapsed_time": "1 day, 3:07:45", "remaining_time": "1 day, 12:40:30", "throughput": 5658.07, "total_tokens": 552599576}
+{"current_steps": 12635, "total_steps": 29704, "loss": 0.37128496170043945, "lr": 6.1621869823396175e-06, "epoch": 0.8507271747912739, "percentage": 42.54, "elapsed_time": "1 day, 3:08:23", "remaining_time": "1 day, 12:39:51", "throughput": 5658.24, "total_tokens": 552832696}
+{"current_steps": 12640, "total_steps": 29704, "loss": 0.37408173084259033, "lr": 6.159615155618752e-06, "epoch": 0.851063829787234, "percentage": 42.55, "elapsed_time": "1 day, 3:08:57", "remaining_time": "1 day, 12:39:04", "throughput": 5658.65, "total_tokens": 553059920}
+{"current_steps": 12645, "total_steps": 29704, "loss": 0.38596835136413576, "lr": 6.157043004615207e-06, "epoch": 0.8514004847831942, "percentage": 42.57, "elapsed_time": "1 day, 3:09:40", "remaining_time": "1 day, 12:38:32", "throughput": 5658.61, "total_tokens": 553300640}
+{"current_steps": 12650, "total_steps": 29704, "loss": 0.3616677761077881, "lr": 6.1544705300482755e-06, "epoch": 0.8517371397791543, "percentage": 42.59, "elapsed_time": "1 day, 3:10:20", "remaining_time": "1 day, 12:37:56", "throughput": 5658.62, "total_tokens": 553532032}
+{"current_steps": 12655, "total_steps": 29704, "loss": 0.3330955982208252, "lr": 6.151897732637341e-06, "epoch": 0.8520737947751145, "percentage": 42.6, "elapsed_time": "1 day, 3:10:57", "remaining_time": "1 day, 12:37:15", "throughput": 5658.62, "total_tokens": 553739416}
+{"current_steps": 12660, "total_steps": 29704, "loss": 0.37527806758880616, "lr": 6.149324613101877e-06, "epoch": 0.8524104497710746, "percentage": 42.62, "elapsed_time": "1 day, 3:11:37", "remaining_time": "1 day, 12:36:37", "throughput": 5658.6, "total_tokens": 553959976}
+{"current_steps": 12665, "total_steps": 29704, "loss": 0.33863201141357424, "lr": 6.146751172161453e-06, "epoch": 0.8527471047670347, "percentage": 42.64, "elapsed_time": "1 day, 3:12:20", "remaining_time": "1 day, 12:36:05", "throughput": 5658.44, "total_tokens": 554189776}
+{"current_steps": 12670, "total_steps": 29704, "loss": 0.3652313470840454, "lr": 6.144177410535719e-06, "epoch": 0.8530837597629949, "percentage": 42.65, "elapsed_time": "1 day, 3:12:59", "remaining_time": "1 day, 12:35:26", "throughput": 5658.49, "total_tokens": 554414232}
+{"current_steps": 12675, "total_steps": 29704, "loss": 0.3651700973510742, "lr": 6.14160332894442e-06, "epoch": 0.853420414758955, "percentage": 42.67, "elapsed_time": "1 day, 3:13:35", "remaining_time": "1 day, 12:34:44", "throughput": 5658.55, "total_tokens": 554625192}
+{"current_steps": 12680, "total_steps": 29704, "loss": 0.36736459732055665, "lr": 6.139028928107391e-06, "epoch": 0.8537570697549152, "percentage": 42.69, "elapsed_time": "1 day, 3:14:16", "remaining_time": "1 day, 12:34:09", "throughput": 5658.41, "total_tokens": 554842240}
+{"current_steps": 12685, "total_steps": 29704, "loss": 0.3479247808456421, "lr": 6.136454208744553e-06, "epoch": 0.8540937247508753, "percentage": 42.7, "elapsed_time": "1 day, 3:14:57", "remaining_time": "1 day, 12:33:34", "throughput": 5658.23, "total_tokens": 555060144}
+{"current_steps": 12690, "total_steps": 29704, "loss": 0.3790309906005859, "lr": 6.133879171575921e-06, "epoch": 0.8544303797468354, "percentage": 42.72, "elapsed_time": "1 day, 3:15:35", "remaining_time": "1 day, 12:32:54", "throughput": 5658.27, "total_tokens": 555276648}
+{"current_steps": 12695, "total_steps": 29704, "loss": 0.3976943254470825, "lr": 6.131303817321595e-06, "epoch": 0.8547670347427956, "percentage": 42.74, "elapsed_time": "1 day, 3:16:13", "remaining_time": "1 day, 12:32:14", "throughput": 5658.45, "total_tokens": 555506600}
+{"current_steps": 12700, "total_steps": 29704, "loss": 0.3907043933868408, "lr": 6.128728146701763e-06, "epoch": 0.8551036897387557, "percentage": 42.76, "elapsed_time": "1 day, 3:16:48", "remaining_time": "1 day, 12:31:31", "throughput": 5658.73, "total_tokens": 555735640}
+{"current_steps": 12705, "total_steps": 29704, "loss": 0.3401200294494629, "lr": 6.126152160436705e-06, "epoch": 0.8554403447347159, "percentage": 42.77, "elapsed_time": "1 day, 3:17:27", "remaining_time": "1 day, 12:30:52", "throughput": 5658.72, "total_tokens": 555954936}
+{"current_steps": 12710, "total_steps": 29704, "loss": 0.3820429801940918, "lr": 6.123575859246784e-06, "epoch": 0.855776999730676, "percentage": 42.79, "elapsed_time": "1 day, 3:18:05", "remaining_time": "1 day, 12:30:13", "throughput": 5658.63, "total_tokens": 556162568}
+{"current_steps": 12715, "total_steps": 29704, "loss": 0.3488877773284912, "lr": 6.120999243852459e-06, "epoch": 0.8561136547266361, "percentage": 42.81, "elapsed_time": "1 day, 3:18:45", "remaining_time": "1 day, 12:29:36", "throughput": 5658.5, "total_tokens": 556374088}
+{"current_steps": 12720, "total_steps": 29704, "loss": 0.3801596164703369, "lr": 6.118422314974269e-06, "epoch": 0.8564503097225963, "percentage": 42.82, "elapsed_time": "1 day, 3:19:28", "remaining_time": "1 day, 12:29:02", "throughput": 5658.3, "total_tokens": 556596000}
+{"current_steps": 12725, "total_steps": 29704, "loss": 0.36686029434204104, "lr": 6.115845073332845e-06, "epoch": 0.8567869647185564, "percentage": 42.84, "elapsed_time": "1 day, 3:20:13", "remaining_time": "1 day, 12:28:33", "throughput": 5658.06, "total_tokens": 556828456}
+{"current_steps": 12730, "total_steps": 29704, "loss": 0.39199421405792234, "lr": 6.113267519648905e-06, "epoch": 0.8571236197145166, "percentage": 42.86, "elapsed_time": "1 day, 3:20:57", "remaining_time": "1 day, 12:28:01", "throughput": 5657.95, "total_tokens": 557067040}
+{"current_steps": 12735, "total_steps": 29704, "loss": 0.36382412910461426, "lr": 6.110689654643253e-06, "epoch": 0.8574602747104767, "percentage": 42.87, "elapsed_time": "1 day, 3:21:37", "remaining_time": "1 day, 12:27:25", "throughput": 5657.73, "total_tokens": 557274264}
+{"current_steps": 12740, "total_steps": 29704, "loss": 0.37306723594665525, "lr": 6.1081114790367805e-06, "epoch": 0.8577969297064368, "percentage": 42.89, "elapsed_time": "1 day, 3:22:11", "remaining_time": "1 day, 12:26:40", "throughput": 5657.9, "total_tokens": 557484064}
+{"current_steps": 12745, "total_steps": 29704, "loss": 0.37079343795776365, "lr": 6.105532993550467e-06, "epoch": 0.858133584702397, "percentage": 42.91, "elapsed_time": "1 day, 3:22:50", "remaining_time": "1 day, 12:26:02", "throughput": 5657.94, "total_tokens": 557707168}
+{"current_steps": 12750, "total_steps": 29704, "loss": 0.3675063610076904, "lr": 6.1029541989053765e-06, "epoch": 0.8584702396983571, "percentage": 42.92, "elapsed_time": "1 day, 3:23:31", "remaining_time": "1 day, 12:25:26", "throughput": 5657.77, "total_tokens": 557923328}
+{"current_steps": 12755, "total_steps": 29704, "loss": 0.3660135746002197, "lr": 6.100375095822661e-06, "epoch": 0.8588068946943173, "percentage": 42.94, "elapsed_time": "1 day, 3:24:13", "remaining_time": "1 day, 12:24:52", "throughput": 5657.59, "total_tokens": 558141912}
+{"current_steps": 12760, "total_steps": 29704, "loss": 0.3654344081878662, "lr": 6.097795685023558e-06, "epoch": 0.8591435496902774, "percentage": 42.96, "elapsed_time": "1 day, 3:24:49", "remaining_time": "1 day, 12:24:09", "throughput": 5657.79, "total_tokens": 558363664}
+{"current_steps": 12765, "total_steps": 29704, "loss": 0.3468952655792236, "lr": 6.095215967229396e-06, "epoch": 0.8594802046862375, "percentage": 42.97, "elapsed_time": "1 day, 3:25:26", "remaining_time": "1 day, 12:23:29", "throughput": 5657.72, "total_tokens": 558569128}
+{"current_steps": 12770, "total_steps": 29704, "loss": 0.37009525299072266, "lr": 6.092635943161578e-06, "epoch": 0.8598168596821977, "percentage": 42.99, "elapsed_time": "1 day, 3:26:10", "remaining_time": "1 day, 12:22:57", "throughput": 5657.54, "total_tokens": 558798144}
+{"current_steps": 12775, "total_steps": 29704, "loss": 0.34846246242523193, "lr": 6.090055613541603e-06, "epoch": 0.8601535146781578, "percentage": 43.01, "elapsed_time": "1 day, 3:26:53", "remaining_time": "1 day, 12:22:24", "throughput": 5657.39, "total_tokens": 559026000}
+{"current_steps": 12780, "total_steps": 29704, "loss": 0.3586388111114502, "lr": 6.087474979091052e-06, "epoch": 0.860490169674118, "percentage": 43.02, "elapsed_time": "1 day, 3:27:40", "remaining_time": "1 day, 12:21:56", "throughput": 5656.79, "total_tokens": 559233040}
+{"current_steps": 12785, "total_steps": 29704, "loss": 0.3598465919494629, "lr": 6.084894040531591e-06, "epoch": 0.8608268246700781, "percentage": 43.04, "elapsed_time": "1 day, 3:28:23", "remaining_time": "1 day, 12:21:23", "throughput": 5656.58, "total_tokens": 559456240}
+{"current_steps": 12790, "total_steps": 29704, "loss": 0.35327887535095215, "lr": 6.0823127985849705e-06, "epoch": 0.8611634796660382, "percentage": 43.06, "elapsed_time": "1 day, 3:29:02", "remaining_time": "1 day, 12:20:45", "throughput": 5656.45, "total_tokens": 559662696}
+{"current_steps": 12795, "total_steps": 29704, "loss": 0.3576143741607666, "lr": 6.079731253973028e-06, "epoch": 0.8615001346619984, "percentage": 43.08, "elapsed_time": "1 day, 3:29:48", "remaining_time": "1 day, 12:20:16", "throughput": 5656.07, "total_tokens": 559886624}
+{"current_steps": 12800, "total_steps": 29704, "loss": 0.36043851375579833, "lr": 6.077149407417683e-06, "epoch": 0.8618367896579585, "percentage": 43.09, "elapsed_time": "1 day, 3:30:24", "remaining_time": "1 day, 12:19:33", "throughput": 5656.21, "total_tokens": 560101432}
+{"current_steps": 12805, "total_steps": 29704, "loss": 0.3793959617614746, "lr": 6.07456725964094e-06, "epoch": 0.8621734446539187, "percentage": 43.11, "elapsed_time": "1 day, 3:30:59", "remaining_time": "1 day, 12:18:50", "throughput": 5656.3, "total_tokens": 560309472}
+{"current_steps": 12810, "total_steps": 29704, "loss": 0.3436178684234619, "lr": 6.07198481136489e-06, "epoch": 0.8625100996498788, "percentage": 43.13, "elapsed_time": "1 day, 3:31:32", "remaining_time": "1 day, 12:18:04", "throughput": 5656.53, "total_tokens": 560520408}
+{"current_steps": 12815, "total_steps": 29704, "loss": 0.3549325704574585, "lr": 6.069402063311706e-06, "epoch": 0.8628467546458389, "percentage": 43.14, "elapsed_time": "1 day, 3:32:15", "remaining_time": "1 day, 12:17:31", "throughput": 5656.2, "total_tokens": 560731064}
+{"current_steps": 12820, "total_steps": 29704, "loss": 0.36306018829345704, "lr": 6.066819016203646e-06, "epoch": 0.8631834096417991, "percentage": 43.16, "elapsed_time": "1 day, 3:32:53", "remaining_time": "1 day, 12:16:51", "throughput": 5656.22, "total_tokens": 560944360}
+{"current_steps": 12825, "total_steps": 29704, "loss": 0.36905884742736816, "lr": 6.064235670763048e-06, "epoch": 0.8635200646377592, "percentage": 43.18, "elapsed_time": "1 day, 3:33:29", "remaining_time": "1 day, 12:16:09", "throughput": 5656.38, "total_tokens": 561164928}
+{"current_steps": 12830, "total_steps": 29704, "loss": 0.3765472412109375, "lr": 6.06165202771234e-06, "epoch": 0.8638567196337193, "percentage": 43.19, "elapsed_time": "1 day, 3:34:10", "remaining_time": "1 day, 12:15:34", "throughput": 5656.34, "total_tokens": 561394056}
+{"current_steps": 12835, "total_steps": 29704, "loss": 0.34474341869354247, "lr": 6.059068087774026e-06, "epoch": 0.8641933746296795, "percentage": 43.21, "elapsed_time": "1 day, 3:34:51", "remaining_time": "1 day, 12:14:57", "throughput": 5656.03, "total_tokens": 561593424}
+{"current_steps": 12840, "total_steps": 29704, "loss": 0.38959674835205077, "lr": 6.0564838516707005e-06, "epoch": 0.8645300296256396, "percentage": 43.23, "elapsed_time": "1 day, 3:35:28", "remaining_time": "1 day, 12:14:17", "throughput": 5656.22, "total_tokens": 561821784}
+{"current_steps": 12845, "total_steps": 29704, "loss": 0.34748311042785646, "lr": 6.053899320125033e-06, "epoch": 0.8648666846215998, "percentage": 43.24, "elapsed_time": "1 day, 3:36:08", "remaining_time": "1 day, 12:13:40", "throughput": 5656.15, "total_tokens": 562042168}
+{"current_steps": 12850, "total_steps": 29704, "loss": 0.37701342105865476, "lr": 6.051314493859782e-06, "epoch": 0.8652033396175599, "percentage": 43.26, "elapsed_time": "1 day, 3:36:45", "remaining_time": "1 day, 12:12:59", "throughput": 5656.31, "total_tokens": 562268616}
+{"current_steps": 12855, "total_steps": 29704, "loss": 0.36400351524353025, "lr": 6.048729373597786e-06, "epoch": 0.86553999461352, "percentage": 43.28, "elapsed_time": "1 day, 3:37:21", "remaining_time": "1 day, 12:12:17", "throughput": 5656.5, "total_tokens": 562488728}
+{"current_steps": 12860, "total_steps": 29704, "loss": 0.38647964000701907, "lr": 6.046143960061963e-06, "epoch": 0.8658766496094802, "percentage": 43.29, "elapsed_time": "1 day, 3:37:59", "remaining_time": "1 day, 12:11:38", "throughput": 5656.59, "total_tokens": 562715632}
+{"current_steps": 12865, "total_steps": 29704, "loss": 0.3893282413482666, "lr": 6.043558253975319e-06, "epoch": 0.8662133046054403, "percentage": 43.31, "elapsed_time": "1 day, 3:38:42", "remaining_time": "1 day, 12:11:04", "throughput": 5656.53, "total_tokens": 562949544}
+{"current_steps": 12870, "total_steps": 29704, "loss": 0.35498225688934326, "lr": 6.040972256060936e-06, "epoch": 0.8665499596014005, "percentage": 43.33, "elapsed_time": "1 day, 3:39:27", "remaining_time": "1 day, 12:10:34", "throughput": 5656.09, "total_tokens": 563161168}
+{"current_steps": 12875, "total_steps": 29704, "loss": 0.3538069248199463, "lr": 6.038385967041982e-06, "epoch": 0.8668866145973606, "percentage": 43.34, "elapsed_time": "1 day, 3:40:13", "remaining_time": "1 day, 12:10:05", "throughput": 5655.54, "total_tokens": 563369656}
+{"current_steps": 12880, "total_steps": 29704, "loss": 0.36878552436828616, "lr": 6.035799387641703e-06, "epoch": 0.8672232695933207, "percentage": 43.36, "elapsed_time": "1 day, 3:40:55", "remaining_time": "1 day, 12:09:31", "throughput": 5655.43, "total_tokens": 563594552}
+{"current_steps": 12885, "total_steps": 29704, "loss": 0.3786178112030029, "lr": 6.033212518583427e-06, "epoch": 0.8675599245892809, "percentage": 43.38, "elapsed_time": "1 day, 3:41:35", "remaining_time": "1 day, 12:08:53", "throughput": 5655.27, "total_tokens": 563803512}
+{"current_steps": 12890, "total_steps": 29704, "loss": 0.35876643657684326, "lr": 6.030625360590567e-06, "epoch": 0.867896579585241, "percentage": 43.39, "elapsed_time": "1 day, 3:42:14", "remaining_time": "1 day, 12:08:15", "throughput": 5655.33, "total_tokens": 564030520}
+{"current_steps": 12895, "total_steps": 29704, "loss": 0.3620233297348022, "lr": 6.028037914386609e-06, "epoch": 0.8682332345812012, "percentage": 43.41, "elapsed_time": "1 day, 3:42:53", "remaining_time": "1 day, 12:07:38", "throughput": 5655.31, "total_tokens": 564252688}
+{"current_steps": 12900, "total_steps": 29704, "loss": 0.38210210800170896, "lr": 6.025450180695128e-06, "epoch": 0.8685698895771613, "percentage": 43.43, "elapsed_time": "1 day, 3:43:28", "remaining_time": "1 day, 12:06:54", "throughput": 5655.52, "total_tokens": 564471472}
+{"current_steps": 12905, "total_steps": 29704, "loss": 0.3630260467529297, "lr": 6.022862160239774e-06, "epoch": 0.8689065445731214, "percentage": 43.45, "elapsed_time": "1 day, 3:44:10", "remaining_time": "1 day, 12:06:20", "throughput": 5655.26, "total_tokens": 564683208}
+{"current_steps": 12910, "total_steps": 29704, "loss": 0.3535295009613037, "lr": 6.020273853744277e-06, "epoch": 0.8692431995690816, "percentage": 43.46, "elapsed_time": "1 day, 3:44:48", "remaining_time": "1 day, 12:05:40", "throughput": 5655.34, "total_tokens": 564903760}
+{"current_steps": 12915, "total_steps": 29704, "loss": 0.35518722534179686, "lr": 6.017685261932452e-06, "epoch": 0.8695798545650417, "percentage": 43.48, "elapsed_time": "1 day, 3:45:23", "remaining_time": "1 day, 12:04:56", "throughput": 5655.38, "total_tokens": 565105760}
+{"current_steps": 12920, "total_steps": 29704, "loss": 0.37440195083618166, "lr": 6.015096385528189e-06, "epoch": 0.8699165095610019, "percentage": 43.5, "elapsed_time": "1 day, 3:46:00", "remaining_time": "1 day, 12:04:15", "throughput": 5655.42, "total_tokens": 565318768}
+{"current_steps": 12925, "total_steps": 29704, "loss": 0.3945824146270752, "lr": 6.012507225255457e-06, "epoch": 0.870253164556962, "percentage": 43.51, "elapsed_time": "1 day, 3:46:35", "remaining_time": "1 day, 12:03:32", "throughput": 5655.4, "total_tokens": 565515560}
+{"current_steps": 12930, "total_steps": 29704, "loss": 0.3706216812133789, "lr": 6.009917781838309e-06, "epoch": 0.8705898195529221, "percentage": 43.53, "elapsed_time": "1 day, 3:47:07", "remaining_time": "1 day, 12:02:45", "throughput": 5655.79, "total_tokens": 565736648}
+{"current_steps": 12935, "total_steps": 29704, "loss": 0.37667386531829833, "lr": 6.007328056000873e-06, "epoch": 0.8709264745488823, "percentage": 43.55, "elapsed_time": "1 day, 3:47:43", "remaining_time": "1 day, 12:02:03", "throughput": 5656.0, "total_tokens": 565960800}
+{"current_steps": 12940, "total_steps": 29704, "loss": 0.379984712600708, "lr": 6.004738048467359e-06, "epoch": 0.8712631295448424, "percentage": 43.56, "elapsed_time": "1 day, 3:48:21", "remaining_time": "1 day, 12:01:23", "throughput": 5656.08, "total_tokens": 566180672}
+{"current_steps": 12945, "total_steps": 29704, "loss": 0.37138075828552247, "lr": 6.002147759962054e-06, "epoch": 0.8715997845408026, "percentage": 43.58, "elapsed_time": "1 day, 3:48:59", "remaining_time": "1 day, 12:00:43", "throughput": 5656.14, "total_tokens": 566401496}
+{"current_steps": 12950, "total_steps": 29704, "loss": 0.39212737083435056, "lr": 5.999557191209323e-06, "epoch": 0.8719364395367627, "percentage": 43.6, "elapsed_time": "1 day, 3:49:35", "remaining_time": "1 day, 12:00:01", "throughput": 5656.39, "total_tokens": 566630216}
+{"current_steps": 12955, "total_steps": 29704, "loss": 0.33372251987457274, "lr": 5.996966342933611e-06, "epoch": 0.8722730945327228, "percentage": 43.61, "elapsed_time": "1 day, 3:50:12", "remaining_time": "1 day, 11:59:21", "throughput": 5656.44, "total_tokens": 566848176}
+{"current_steps": 12960, "total_steps": 29704, "loss": 0.37457475662231443, "lr": 5.99437521585944e-06, "epoch": 0.872609749528683, "percentage": 43.63, "elapsed_time": "1 day, 3:50:50", "remaining_time": "1 day, 11:58:40", "throughput": 5656.42, "total_tokens": 567056408}
+{"current_steps": 12965, "total_steps": 29704, "loss": 0.3829956531524658, "lr": 5.99178381071141e-06, "epoch": 0.8729464045246431, "percentage": 43.65, "elapsed_time": "1 day, 3:51:27", "remaining_time": "1 day, 11:57:59", "throughput": 5656.51, "total_tokens": 567275168}
+{"current_steps": 12970, "total_steps": 29704, "loss": 0.36388020515441893, "lr": 5.989192128214201e-06, "epoch": 0.8732830595206033, "percentage": 43.66, "elapsed_time": "1 day, 3:52:07", "remaining_time": "1 day, 11:57:23", "throughput": 5656.42, "total_tokens": 567493544}
+{"current_steps": 12975, "total_steps": 29704, "loss": 0.350065803527832, "lr": 5.986600169092568e-06, "epoch": 0.8736197145165634, "percentage": 43.68, "elapsed_time": "1 day, 3:52:44", "remaining_time": "1 day, 11:56:42", "throughput": 5656.49, "total_tokens": 567708176}
+{"current_steps": 12980, "total_steps": 29704, "loss": 0.3742049217224121, "lr": 5.984007934071343e-06, "epoch": 0.8739563695125235, "percentage": 43.7, "elapsed_time": "1 day, 3:53:27", "remaining_time": "1 day, 11:56:08", "throughput": 5656.39, "total_tokens": 567941160}
+{"current_steps": 12985, "total_steps": 29704, "loss": 0.36548290252685545, "lr": 5.981415423875436e-06, "epoch": 0.8742930245084837, "percentage": 43.71, "elapsed_time": "1 day, 3:54:02", "remaining_time": "1 day, 11:55:26", "throughput": 5656.63, "total_tokens": 568166040}
+{"current_steps": 12990, "total_steps": 29704, "loss": 0.3484132528305054, "lr": 5.978822639229836e-06, "epoch": 0.8746296795044438, "percentage": 43.73, "elapsed_time": "1 day, 3:54:39", "remaining_time": "1 day, 11:54:45", "throughput": 5656.56, "total_tokens": 568367984}
+{"current_steps": 12995, "total_steps": 29704, "loss": 0.36139798164367676, "lr": 5.976229580859607e-06, "epoch": 0.874966334500404, "percentage": 43.75, "elapsed_time": "1 day, 3:55:16", "remaining_time": "1 day, 11:54:04", "throughput": 5656.74, "total_tokens": 568596000}
+{"current_steps": 13000, "total_steps": 29704, "loss": 0.3539880275726318, "lr": 5.9736362494898895e-06, "epoch": 0.8753029894963641, "percentage": 43.77, "elapsed_time": "1 day, 3:55:55", "remaining_time": "1 day, 11:53:25", "throughput": 5656.74, "total_tokens": 568814440}
+{"current_steps": 13005, "total_steps": 29704, "loss": 0.3666597843170166, "lr": 5.971042645845897e-06, "epoch": 0.8756396444923242, "percentage": 43.78, "elapsed_time": "1 day, 3:56:35", "remaining_time": "1 day, 11:52:48", "throughput": 5656.62, "total_tokens": 569029824}
+{"current_steps": 13010, "total_steps": 29704, "loss": 0.36510615348815917, "lr": 5.968448770652926e-06, "epoch": 0.8759762994882844, "percentage": 43.8, "elapsed_time": "1 day, 3:57:07", "remaining_time": "1 day, 11:52:02", "throughput": 5657.14, "total_tokens": 569265096}
+{"current_steps": 13015, "total_steps": 29704, "loss": 0.3825402736663818, "lr": 5.965854624636345e-06, "epoch": 0.8763129544842445, "percentage": 43.82, "elapsed_time": "1 day, 3:57:49", "remaining_time": "1 day, 11:51:27", "throughput": 5656.93, "total_tokens": 569479936}
+{"current_steps": 13020, "total_steps": 29704, "loss": 0.3698084354400635, "lr": 5.963260208521598e-06, "epoch": 0.8766496094802047, "percentage": 43.83, "elapsed_time": "1 day, 3:58:28", "remaining_time": "1 day, 11:50:49", "throughput": 5656.84, "total_tokens": 569691888}
+{"current_steps": 13025, "total_steps": 29704, "loss": 0.38090927600860597, "lr": 5.960665523034205e-06, "epoch": 0.8769862644761648, "percentage": 43.85, "elapsed_time": "1 day, 3:59:09", "remaining_time": "1 day, 11:50:13", "throughput": 5656.63, "total_tokens": 569901576}
+{"current_steps": 13030, "total_steps": 29704, "loss": 0.33312511444091797, "lr": 5.958070568899762e-06, "epoch": 0.8773229194721249, "percentage": 43.87, "elapsed_time": "1 day, 3:59:50", "remaining_time": "1 day, 11:49:37", "throughput": 5656.52, "total_tokens": 570123424}
+{"current_steps": 13035, "total_steps": 29704, "loss": 0.3555187225341797, "lr": 5.955475346843939e-06, "epoch": 0.8776595744680851, "percentage": 43.88, "elapsed_time": "1 day, 4:00:24", "remaining_time": "1 day, 11:48:52", "throughput": 5656.76, "total_tokens": 570337248}
+{"current_steps": 13040, "total_steps": 29704, "loss": 0.3440871000289917, "lr": 5.952879857592482e-06, "epoch": 0.8779962294640452, "percentage": 43.9, "elapsed_time": "1 day, 4:01:07", "remaining_time": "1 day, 11:48:20", "throughput": 5656.4, "total_tokens": 570549824}
+{"current_steps": 13045, "total_steps": 29704, "loss": 0.38202319145202634, "lr": 5.950284101871212e-06, "epoch": 0.8783328844600053, "percentage": 43.92, "elapsed_time": "1 day, 4:01:46", "remaining_time": "1 day, 11:47:41", "throughput": 5656.31, "total_tokens": 570757152}
+{"current_steps": 13050, "total_steps": 29704, "loss": 0.3463796615600586, "lr": 5.947688080406023e-06, "epoch": 0.8786695394559655, "percentage": 43.93, "elapsed_time": "1 day, 4:02:20", "remaining_time": "1 day, 11:46:57", "throughput": 5656.41, "total_tokens": 570960928}
+{"current_steps": 13055, "total_steps": 29704, "loss": 0.3679834842681885, "lr": 5.945091793922885e-06, "epoch": 0.8790061944519256, "percentage": 43.95, "elapsed_time": "1 day, 4:02:54", "remaining_time": "1 day, 11:46:11", "throughput": 5656.67, "total_tokens": 571177360}
+{"current_steps": 13060, "total_steps": 29704, "loss": 0.3929899215698242, "lr": 5.942495243147839e-06, "epoch": 0.8793428494478858, "percentage": 43.97, "elapsed_time": "1 day, 4:03:25", "remaining_time": "1 day, 11:45:24", "throughput": 5657.1, "total_tokens": 571399192}
+{"current_steps": 13065, "total_steps": 29704, "loss": 0.3815101146697998, "lr": 5.939898428807004e-06, "epoch": 0.8796795044438459, "percentage": 43.98, "elapsed_time": "1 day, 4:04:03", "remaining_time": "1 day, 11:44:44", "throughput": 5657.1, "total_tokens": 571612808}
+{"current_steps": 13070, "total_steps": 29704, "loss": 0.379590368270874, "lr": 5.937301351626571e-06, "epoch": 0.880016159439806, "percentage": 44.0, "elapsed_time": "1 day, 4:04:41", "remaining_time": "1 day, 11:44:05", "throughput": 5657.04, "total_tokens": 571824200}
+{"current_steps": 13075, "total_steps": 29704, "loss": 0.35865001678466796, "lr": 5.934704012332803e-06, "epoch": 0.8803528144357662, "percentage": 44.02, "elapsed_time": "1 day, 4:05:22", "remaining_time": "1 day, 11:43:29", "throughput": 5657.0, "total_tokens": 572052552}
+{"current_steps": 13080, "total_steps": 29704, "loss": 0.4059410095214844, "lr": 5.93210641165204e-06, "epoch": 0.8806894694317263, "percentage": 44.03, "elapsed_time": "1 day, 4:06:05", "remaining_time": "1 day, 11:42:56", "throughput": 5656.8, "total_tokens": 572274280}
+{"current_steps": 13085, "total_steps": 29704, "loss": 0.36967930793762205, "lr": 5.92950855031069e-06, "epoch": 0.8810261244276865, "percentage": 44.05, "elapsed_time": "1 day, 4:06:42", "remaining_time": "1 day, 11:42:15", "throughput": 5656.92, "total_tokens": 572493808}
+{"current_steps": 13090, "total_steps": 29704, "loss": 0.36150131225585935, "lr": 5.926910429035239e-06, "epoch": 0.8813627794236466, "percentage": 44.07, "elapsed_time": "1 day, 4:07:19", "remaining_time": "1 day, 11:41:34", "throughput": 5656.95, "total_tokens": 572706072}
+{"current_steps": 13095, "total_steps": 29704, "loss": 0.3525059223175049, "lr": 5.924312048552241e-06, "epoch": 0.8816994344196067, "percentage": 44.08, "elapsed_time": "1 day, 4:07:53", "remaining_time": "1 day, 11:40:49", "throughput": 5657.28, "total_tokens": 572932096}
+{"current_steps": 13100, "total_steps": 29704, "loss": 0.3641670227050781, "lr": 5.9217134095883265e-06, "epoch": 0.8820360894155669, "percentage": 44.1, "elapsed_time": "1 day, 4:08:34", "remaining_time": "1 day, 11:40:14", "throughput": 5657.14, "total_tokens": 573150384}
+{"current_steps": 13105, "total_steps": 29704, "loss": 0.3603955268859863, "lr": 5.919114512870197e-06, "epoch": 0.882372744411527, "percentage": 44.12, "elapsed_time": "1 day, 4:09:19", "remaining_time": "1 day, 11:39:43", "throughput": 5656.92, "total_tokens": 573381144}
+{"current_steps": 13110, "total_steps": 29704, "loss": 0.3425842046737671, "lr": 5.916515359124623e-06, "epoch": 0.8827093994074872, "percentage": 44.14, "elapsed_time": "1 day, 4:09:59", "remaining_time": "1 day, 11:39:06", "throughput": 5656.66, "total_tokens": 573581832}
+{"current_steps": 13115, "total_steps": 29704, "loss": 0.36979331970214846, "lr": 5.913915949078453e-06, "epoch": 0.8830460544034473, "percentage": 44.15, "elapsed_time": "1 day, 4:10:38", "remaining_time": "1 day, 11:38:28", "throughput": 5656.63, "total_tokens": 573801512}
+{"current_steps": 13120, "total_steps": 29704, "loss": 0.3616600513458252, "lr": 5.911316283458601e-06, "epoch": 0.8833827093994074, "percentage": 44.17, "elapsed_time": "1 day, 4:11:18", "remaining_time": "1 day, 11:37:51", "throughput": 5656.45, "total_tokens": 574007544}
+{"current_steps": 13125, "total_steps": 29704, "loss": 0.36803760528564455, "lr": 5.908716362992057e-06, "epoch": 0.8837193643953676, "percentage": 44.19, "elapsed_time": "1 day, 4:11:54", "remaining_time": "1 day, 11:37:09", "throughput": 5656.47, "total_tokens": 574215752}
+{"current_steps": 13130, "total_steps": 29704, "loss": 0.379679536819458, "lr": 5.90611618840588e-06, "epoch": 0.8840560193913277, "percentage": 44.2, "elapsed_time": "1 day, 4:12:34", "remaining_time": "1 day, 11:36:32", "throughput": 5656.39, "total_tokens": 574434272}
+{"current_steps": 13135, "total_steps": 29704, "loss": 0.4000962257385254, "lr": 5.903515760427198e-06, "epoch": 0.8843926743872879, "percentage": 44.22, "elapsed_time": "1 day, 4:13:08", "remaining_time": "1 day, 11:35:47", "throughput": 5656.7, "total_tokens": 574653864}
+{"current_steps": 13140, "total_steps": 29704, "loss": 0.3558906316757202, "lr": 5.900915079783218e-06, "epoch": 0.884729329383248, "percentage": 44.24, "elapsed_time": "1 day, 4:13:42", "remaining_time": "1 day, 11:35:02", "throughput": 5657.1, "total_tokens": 574887384}
+{"current_steps": 13145, "total_steps": 29704, "loss": 0.3416097640991211, "lr": 5.898314147201207e-06, "epoch": 0.8850659843792081, "percentage": 44.25, "elapsed_time": "1 day, 4:14:24", "remaining_time": "1 day, 11:34:28", "throughput": 5656.85, "total_tokens": 575101544}
+{"current_steps": 13150, "total_steps": 29704, "loss": 0.3921904325485229, "lr": 5.895712963408511e-06, "epoch": 0.8854026393751683, "percentage": 44.27, "elapsed_time": "1 day, 4:15:04", "remaining_time": "1 day, 11:33:51", "throughput": 5656.66, "total_tokens": 575309032}
+{"current_steps": 13155, "total_steps": 29704, "loss": 0.38542513847351073, "lr": 5.893111529132539e-06, "epoch": 0.8857392943711284, "percentage": 44.29, "elapsed_time": "1 day, 4:15:42", "remaining_time": "1 day, 11:33:11", "throughput": 5656.63, "total_tokens": 575518184}
+{"current_steps": 13160, "total_steps": 29704, "loss": 0.37914443016052246, "lr": 5.8905098451007755e-06, "epoch": 0.8860759493670886, "percentage": 44.3, "elapsed_time": "1 day, 4:16:21", "remaining_time": "1 day, 11:32:34", "throughput": 5656.42, "total_tokens": 575721848}
+{"current_steps": 13165, "total_steps": 29704, "loss": 0.3751526355743408, "lr": 5.887907912040774e-06, "epoch": 0.8864126043630487, "percentage": 44.32, "elapsed_time": "1 day, 4:16:54", "remaining_time": "1 day, 11:31:48", "throughput": 5656.81, "total_tokens": 575946104}
+{"current_steps": 13170, "total_steps": 29704, "loss": 0.3788419723510742, "lr": 5.885305730680158e-06, "epoch": 0.8867492593590088, "percentage": 44.34, "elapsed_time": "1 day, 4:17:27", "remaining_time": "1 day, 11:31:02", "throughput": 5657.23, "total_tokens": 576173784}
+{"current_steps": 13175, "total_steps": 29704, "loss": 0.36290099620819094, "lr": 5.882703301746616e-06, "epoch": 0.887085914354969, "percentage": 44.35, "elapsed_time": "1 day, 4:18:09", "remaining_time": "1 day, 11:30:28", "throughput": 5656.98, "total_tokens": 576389120}
+{"current_steps": 13180, "total_steps": 29704, "loss": 0.35980806350708006, "lr": 5.88010062596791e-06, "epoch": 0.8874225693509291, "percentage": 44.37, "elapsed_time": "1 day, 4:18:50", "remaining_time": "1 day, 11:29:52", "throughput": 5656.76, "total_tokens": 576598016}
+{"current_steps": 13185, "total_steps": 29704, "loss": 0.3475220203399658, "lr": 5.877497704071871e-06, "epoch": 0.8877592243468893, "percentage": 44.39, "elapsed_time": "1 day, 4:19:29", "remaining_time": "1 day, 11:29:13", "throughput": 5656.93, "total_tokens": 576831392}
+{"current_steps": 13190, "total_steps": 29704, "loss": 0.3340578079223633, "lr": 5.874894536786396e-06, "epoch": 0.8880958793428494, "percentage": 44.4, "elapsed_time": "1 day, 4:20:09", "remaining_time": "1 day, 11:28:36", "throughput": 5656.78, "total_tokens": 577044568}
+{"current_steps": 13195, "total_steps": 29704, "loss": 0.33777599334716796, "lr": 5.8722911248394555e-06, "epoch": 0.8884325343388096, "percentage": 44.42, "elapsed_time": "1 day, 4:20:50", "remaining_time": "1 day, 11:28:01", "throughput": 5656.44, "total_tokens": 577243936}
+{"current_steps": 13200, "total_steps": 29704, "loss": 0.3437623977661133, "lr": 5.8696874689590824e-06, "epoch": 0.8887691893347698, "percentage": 44.44, "elapsed_time": "1 day, 4:21:23", "remaining_time": "1 day, 11:27:15", "throughput": 5656.69, "total_tokens": 577453888}
+{"current_steps": 13205, "total_steps": 29704, "loss": 0.3634350776672363, "lr": 5.867083569873383e-06, "epoch": 0.8891058443307299, "percentage": 44.46, "elapsed_time": "1 day, 4:21:57", "remaining_time": "1 day, 11:26:30", "throughput": 5657.01, "total_tokens": 577679688}
+{"current_steps": 13210, "total_steps": 29704, "loss": 0.36774277687072754, "lr": 5.8644794283105255e-06, "epoch": 0.8894424993266901, "percentage": 44.47, "elapsed_time": "1 day, 4:22:35", "remaining_time": "1 day, 11:25:51", "throughput": 5657.05, "total_tokens": 577900960}
+{"current_steps": 13215, "total_steps": 29704, "loss": 0.40239391326904295, "lr": 5.861875044998755e-06, "epoch": 0.8897791543226502, "percentage": 44.49, "elapsed_time": "1 day, 4:23:10", "remaining_time": "1 day, 11:25:08", "throughput": 5657.34, "total_tokens": 578125704}
+{"current_steps": 13220, "total_steps": 29704, "loss": 0.36860146522521975, "lr": 5.859270420666377e-06, "epoch": 0.8901158093186103, "percentage": 44.51, "elapsed_time": "1 day, 4:23:48", "remaining_time": "1 day, 11:24:28", "throughput": 5657.36, "total_tokens": 578343816}
+{"current_steps": 13225, "total_steps": 29704, "loss": 0.3481595993041992, "lr": 5.856665556041764e-06, "epoch": 0.8904524643145705, "percentage": 44.52, "elapsed_time": "1 day, 4:24:25", "remaining_time": "1 day, 11:23:48", "throughput": 5657.29, "total_tokens": 578548144}
+{"current_steps": 13230, "total_steps": 29704, "loss": 0.3859154224395752, "lr": 5.8540604518533605e-06, "epoch": 0.8907891193105306, "percentage": 44.54, "elapsed_time": "1 day, 4:24:59", "remaining_time": "1 day, 11:23:03", "throughput": 5657.62, "total_tokens": 578772776}
+{"current_steps": 13235, "total_steps": 29704, "loss": 0.34519567489624026, "lr": 5.851455108829675e-06, "epoch": 0.8911257743064908, "percentage": 44.56, "elapsed_time": "1 day, 4:25:40", "remaining_time": "1 day, 11:22:28", "throughput": 5657.38, "total_tokens": 578981024}
+{"current_steps": 13240, "total_steps": 29704, "loss": 0.3698782205581665, "lr": 5.848849527699283e-06, "epoch": 0.8914624293024509, "percentage": 44.57, "elapsed_time": "1 day, 4:26:19", "remaining_time": "1 day, 11:21:49", "throughput": 5657.41, "total_tokens": 579201768}
+{"current_steps": 13245, "total_steps": 29704, "loss": 0.35757324695587156, "lr": 5.846243709190827e-06, "epoch": 0.891799084298411, "percentage": 44.59, "elapsed_time": "1 day, 4:26:57", "remaining_time": "1 day, 11:21:10", "throughput": 5657.42, "total_tokens": 579419880}
+{"current_steps": 13250, "total_steps": 29704, "loss": 0.37553791999816893, "lr": 5.843637654033019e-06, "epoch": 0.8921357392943712, "percentage": 44.61, "elapsed_time": "1 day, 4:27:39", "remaining_time": "1 day, 11:20:35", "throughput": 5657.05, "total_tokens": 579620760}
+{"current_steps": 13255, "total_steps": 29704, "loss": 0.39575533866882323, "lr": 5.841031362954629e-06, "epoch": 0.8924723942903313, "percentage": 44.62, "elapsed_time": "1 day, 4:28:15", "remaining_time": "1 day, 11:19:52", "throughput": 5657.34, "total_tokens": 579849048}
+{"current_steps": 13260, "total_steps": 29704, "loss": 0.37047085762023924, "lr": 5.8384248366845e-06, "epoch": 0.8928090492862915, "percentage": 44.64, "elapsed_time": "1 day, 4:29:00", "remaining_time": "1 day, 11:19:22", "throughput": 5657.07, "total_tokens": 580077352}
+{"current_steps": 13265, "total_steps": 29704, "loss": 0.37846667766571046, "lr": 5.8358180759515396e-06, "epoch": 0.8931457042822516, "percentage": 44.66, "elapsed_time": "1 day, 4:29:37", "remaining_time": "1 day, 11:18:41", "throughput": 5657.09, "total_tokens": 580288200}
+{"current_steps": 13270, "total_steps": 29704, "loss": 0.37510037422180176, "lr": 5.833211081484719e-06, "epoch": 0.8934823592782117, "percentage": 44.67, "elapsed_time": "1 day, 4:30:18", "remaining_time": "1 day, 11:18:06", "throughput": 5656.93, "total_tokens": 580506904}
+{"current_steps": 13275, "total_steps": 29704, "loss": 0.3673901081085205, "lr": 5.830603854013077e-06, "epoch": 0.8938190142741719, "percentage": 44.69, "elapsed_time": "1 day, 4:31:00", "remaining_time": "1 day, 11:17:31", "throughput": 5656.75, "total_tokens": 580723408}
+{"current_steps": 13280, "total_steps": 29704, "loss": 0.35067110061645507, "lr": 5.827996394265716e-06, "epoch": 0.894155669270132, "percentage": 44.71, "elapsed_time": "1 day, 4:31:41", "remaining_time": "1 day, 11:16:55", "throughput": 5656.53, "total_tokens": 580935112}
+{"current_steps": 13285, "total_steps": 29704, "loss": 0.35003342628479006, "lr": 5.825388702971802e-06, "epoch": 0.8944923242660922, "percentage": 44.72, "elapsed_time": "1 day, 4:32:13", "remaining_time": "1 day, 11:16:08", "throughput": 5656.89, "total_tokens": 581151304}
+{"current_steps": 13290, "total_steps": 29704, "loss": 0.36042060852050783, "lr": 5.822780780860571e-06, "epoch": 0.8948289792620523, "percentage": 44.74, "elapsed_time": "1 day, 4:32:50", "remaining_time": "1 day, 11:15:28", "throughput": 5656.93, "total_tokens": 581367952}
+{"current_steps": 13295, "total_steps": 29704, "loss": 0.3688685178756714, "lr": 5.820172628661317e-06, "epoch": 0.8951656342580124, "percentage": 44.76, "elapsed_time": "1 day, 4:33:24", "remaining_time": "1 day, 11:14:43", "throughput": 5657.38, "total_tokens": 581603552}
+{"current_steps": 13300, "total_steps": 29704, "loss": 0.34978251457214354, "lr": 5.8175642471034045e-06, "epoch": 0.8955022892539726, "percentage": 44.78, "elapsed_time": "1 day, 4:33:58", "remaining_time": "1 day, 11:13:58", "throughput": 5657.67, "total_tokens": 581823928}
+{"current_steps": 13305, "total_steps": 29704, "loss": 0.3728696346282959, "lr": 5.8149556369162565e-06, "epoch": 0.8958389442499327, "percentage": 44.79, "elapsed_time": "1 day, 4:34:37", "remaining_time": "1 day, 11:13:20", "throughput": 5657.55, "total_tokens": 582031800}
+{"current_steps": 13310, "total_steps": 29704, "loss": 0.37227132320404055, "lr": 5.812346798829361e-06, "epoch": 0.8961755992458929, "percentage": 44.81, "elapsed_time": "1 day, 4:35:18", "remaining_time": "1 day, 11:12:45", "throughput": 5657.45, "total_tokens": 582257072}
+{"current_steps": 13315, "total_steps": 29704, "loss": 0.35695157051086424, "lr": 5.809737733572276e-06, "epoch": 0.896512254241853, "percentage": 44.83, "elapsed_time": "1 day, 4:35:57", "remaining_time": "1 day, 11:12:06", "throughput": 5657.41, "total_tokens": 582470280}
+{"current_steps": 13320, "total_steps": 29704, "loss": 0.3608556747436523, "lr": 5.8071284418746166e-06, "epoch": 0.8968489092378131, "percentage": 44.84, "elapsed_time": "1 day, 4:36:37", "remaining_time": "1 day, 11:11:30", "throughput": 5657.3, "total_tokens": 582689968}
+{"current_steps": 13325, "total_steps": 29704, "loss": 0.38565864562988283, "lr": 5.80451892446606e-06, "epoch": 0.8971855642337733, "percentage": 44.86, "elapsed_time": "1 day, 4:37:16", "remaining_time": "1 day, 11:10:51", "throughput": 5657.38, "total_tokens": 582917944}
+{"current_steps": 13330, "total_steps": 29704, "loss": 0.3657494068145752, "lr": 5.801909182076354e-06, "epoch": 0.8975222192297334, "percentage": 44.88, "elapsed_time": "1 day, 4:37:51", "remaining_time": "1 day, 11:10:08", "throughput": 5657.67, "total_tokens": 583143528}
+{"current_steps": 13335, "total_steps": 29704, "loss": 0.3965157985687256, "lr": 5.7992992154353e-06, "epoch": 0.8978588742256935, "percentage": 44.89, "elapsed_time": "1 day, 4:38:26", "remaining_time": "1 day, 11:09:25", "throughput": 5657.85, "total_tokens": 583360120}
+{"current_steps": 13340, "total_steps": 29704, "loss": 0.3728632926940918, "lr": 5.79668902527277e-06, "epoch": 0.8981955292216537, "percentage": 44.91, "elapsed_time": "1 day, 4:39:03", "remaining_time": "1 day, 11:08:44", "throughput": 5657.91, "total_tokens": 583576392}
+{"current_steps": 13345, "total_steps": 29704, "loss": 0.36152513027191163, "lr": 5.794078612318695e-06, "epoch": 0.8985321842176138, "percentage": 44.93, "elapsed_time": "1 day, 4:39:47", "remaining_time": "1 day, 11:08:12", "throughput": 5657.59, "total_tokens": 583793888}
+{"current_steps": 13350, "total_steps": 29704, "loss": 0.3539185285568237, "lr": 5.791467977303068e-06, "epoch": 0.898868839213574, "percentage": 44.94, "elapsed_time": "1 day, 4:40:21", "remaining_time": "1 day, 11:07:28", "throughput": 5657.8, "total_tokens": 584007880}
+{"current_steps": 13355, "total_steps": 29704, "loss": 0.3753284692764282, "lr": 5.788857120955945e-06, "epoch": 0.8992054942095341, "percentage": 44.96, "elapsed_time": "1 day, 4:40:59", "remaining_time": "1 day, 11:06:48", "throughput": 5657.82, "total_tokens": 584222720}
+{"current_steps": 13360, "total_steps": 29704, "loss": 0.3408811092376709, "lr": 5.786246044007443e-06, "epoch": 0.8995421492054942, "percentage": 44.98, "elapsed_time": "1 day, 4:41:43", "remaining_time": "1 day, 11:06:16", "throughput": 5657.42, "total_tokens": 584430032}
+{"current_steps": 13365, "total_steps": 29704, "loss": 0.38433132171630857, "lr": 5.783634747187743e-06, "epoch": 0.8998788042014544, "percentage": 44.99, "elapsed_time": "1 day, 4:42:16", "remaining_time": "1 day, 11:05:31", "throughput": 5657.8, "total_tokens": 584657272}
+{"current_steps": 13370, "total_steps": 29704, "loss": 0.3663713693618774, "lr": 5.781023231227084e-06, "epoch": 0.9002154591974145, "percentage": 45.01, "elapsed_time": "1 day, 4:43:03", "remaining_time": "1 day, 11:05:02", "throughput": 5657.31, "total_tokens": 584873064}
+{"current_steps": 13375, "total_steps": 29704, "loss": 0.3779282093048096, "lr": 5.778411496855769e-06, "epoch": 0.9005521141933747, "percentage": 45.03, "elapsed_time": "1 day, 4:43:42", "remaining_time": "1 day, 11:04:23", "throughput": 5657.35, "total_tokens": 585095664}
+{"current_steps": 13380, "total_steps": 29704, "loss": 0.3911623477935791, "lr": 5.775799544804161e-06, "epoch": 0.9008887691893348, "percentage": 45.04, "elapsed_time": "1 day, 4:44:18", "remaining_time": "1 day, 11:03:42", "throughput": 5657.58, "total_tokens": 585323224}
+{"current_steps": 13385, "total_steps": 29704, "loss": 0.3567417860031128, "lr": 5.7731873758026845e-06, "epoch": 0.9012254241852949, "percentage": 45.06, "elapsed_time": "1 day, 4:44:59", "remaining_time": "1 day, 11:03:06", "throughput": 5657.41, "total_tokens": 585539800}
+{"current_steps": 13390, "total_steps": 29704, "loss": 0.3617391109466553, "lr": 5.770574990581823e-06, "epoch": 0.9015620791812551, "percentage": 45.08, "elapsed_time": "1 day, 4:45:38", "remaining_time": "1 day, 11:02:28", "throughput": 5657.37, "total_tokens": 585756160}
+{"current_steps": 13395, "total_steps": 29704, "loss": 0.34462101459503175, "lr": 5.767962389872124e-06, "epoch": 0.9018987341772152, "percentage": 45.09, "elapsed_time": "1 day, 4:46:18", "remaining_time": "1 day, 11:01:51", "throughput": 5657.27, "total_tokens": 585972312}
+{"current_steps": 13400, "total_steps": 29704, "loss": 0.3628375053405762, "lr": 5.765349574404191e-06, "epoch": 0.9022353891731754, "percentage": 45.11, "elapsed_time": "1 day, 4:46:51", "remaining_time": "1 day, 11:01:06", "throughput": 5657.42, "total_tokens": 586175624}
+{"current_steps": 13405, "total_steps": 29704, "loss": 0.364549446105957, "lr": 5.7627365449086905e-06, "epoch": 0.9025720441691355, "percentage": 45.13, "elapsed_time": "1 day, 4:47:25", "remaining_time": "1 day, 11:00:21", "throughput": 5657.45, "total_tokens": 586370072}
+{"current_steps": 13410, "total_steps": 29704, "loss": 0.3901118040084839, "lr": 5.760123302116345e-06, "epoch": 0.9029086991650956, "percentage": 45.15, "elapsed_time": "1 day, 4:47:57", "remaining_time": "1 day, 10:59:35", "throughput": 5657.84, "total_tokens": 586593560}
+{"current_steps": 13415, "total_steps": 29704, "loss": 0.36999359130859377, "lr": 5.757509846757945e-06, "epoch": 0.9032453541610558, "percentage": 45.16, "elapsed_time": "1 day, 4:48:36", "remaining_time": "1 day, 10:58:56", "throughput": 5657.92, "total_tokens": 586819408}
+{"current_steps": 13420, "total_steps": 29704, "loss": 0.3971453428268433, "lr": 5.75489617956433e-06, "epoch": 0.9035820091570159, "percentage": 45.18, "elapsed_time": "1 day, 4:49:22", "remaining_time": "1 day, 10:58:26", "throughput": 5657.55, "total_tokens": 587040712}
+{"current_steps": 13425, "total_steps": 29704, "loss": 0.37274601459503176, "lr": 5.752282301266406e-06, "epoch": 0.9039186641529761, "percentage": 45.2, "elapsed_time": "1 day, 4:49:59", "remaining_time": "1 day, 10:57:45", "throughput": 5657.64, "total_tokens": 587260024}
+{"current_steps": 13430, "total_steps": 29704, "loss": 0.36958420276641846, "lr": 5.749668212595136e-06, "epoch": 0.9042553191489362, "percentage": 45.21, "elapsed_time": "1 day, 4:50:38", "remaining_time": "1 day, 10:57:07", "throughput": 5657.6, "total_tokens": 587475152}
+{"current_steps": 13435, "total_steps": 29704, "loss": 0.36184403896331785, "lr": 5.747053914281539e-06, "epoch": 0.9045919741448963, "percentage": 45.23, "elapsed_time": "1 day, 4:51:21", "remaining_time": "1 day, 10:56:33", "throughput": 5657.23, "total_tokens": 587679656}
+{"current_steps": 13440, "total_steps": 29704, "loss": 0.3511162757873535, "lr": 5.744439407056699e-06, "epoch": 0.9049286291408565, "percentage": 45.25, "elapsed_time": "1 day, 4:51:58", "remaining_time": "1 day, 10:55:54", "throughput": 5657.19, "total_tokens": 587888648}
+{"current_steps": 13445, "total_steps": 29704, "loss": 0.3408759832382202, "lr": 5.741824691651752e-06, "epoch": 0.9052652841368166, "percentage": 45.26, "elapsed_time": "1 day, 4:52:34", "remaining_time": "1 day, 10:55:12", "throughput": 5657.4, "total_tokens": 588114264}
+{"current_steps": 13450, "total_steps": 29704, "loss": 0.36322770118713377, "lr": 5.739209768797896e-06, "epoch": 0.9056019391327768, "percentage": 45.28, "elapsed_time": "1 day, 4:53:11", "remaining_time": "1 day, 10:54:31", "throughput": 5657.41, "total_tokens": 588322904}
+{"current_steps": 13455, "total_steps": 29704, "loss": 0.32691378593444825, "lr": 5.736594639226385e-06, "epoch": 0.9059385941287369, "percentage": 45.3, "elapsed_time": "1 day, 4:53:46", "remaining_time": "1 day, 10:53:48", "throughput": 5657.6, "total_tokens": 588542624}
+{"current_steps": 13460, "total_steps": 29704, "loss": 0.3772242546081543, "lr": 5.7339793036685306e-06, "epoch": 0.906275249124697, "percentage": 45.31, "elapsed_time": "1 day, 4:54:28", "remaining_time": "1 day, 10:53:12", "throughput": 5657.51, "total_tokens": 588766448}
+{"current_steps": 13465, "total_steps": 29704, "loss": 0.3471216678619385, "lr": 5.731363762855706e-06, "epoch": 0.9066119041206572, "percentage": 45.33, "elapsed_time": "1 day, 4:55:09", "remaining_time": "1 day, 10:52:37", "throughput": 5657.35, "total_tokens": 588981896}
+{"current_steps": 13470, "total_steps": 29704, "loss": 0.3998070001602173, "lr": 5.728748017519337e-06, "epoch": 0.9069485591166173, "percentage": 45.35, "elapsed_time": "1 day, 4:55:46", "remaining_time": "1 day, 10:51:56", "throughput": 5657.41, "total_tokens": 589199016}
+{"current_steps": 13475, "total_steps": 29704, "loss": 0.3687425136566162, "lr": 5.726132068390908e-06, "epoch": 0.9072852141125775, "percentage": 45.36, "elapsed_time": "1 day, 4:56:20", "remaining_time": "1 day, 10:51:12", "throughput": 5657.52, "total_tokens": 589402952}
+{"current_steps": 13480, "total_steps": 29704, "loss": 0.4100666046142578, "lr": 5.723515916201962e-06, "epoch": 0.9076218691085376, "percentage": 45.38, "elapsed_time": "1 day, 4:56:58", "remaining_time": "1 day, 10:50:33", "throughput": 5657.62, "total_tokens": 589630440}
+{"current_steps": 13485, "total_steps": 29704, "loss": 0.3434640645980835, "lr": 5.720899561684098e-06, "epoch": 0.9079585241044977, "percentage": 45.4, "elapsed_time": "1 day, 4:57:42", "remaining_time": "1 day, 10:50:00", "throughput": 5657.37, "total_tokens": 589849368}
+{"current_steps": 13490, "total_steps": 29704, "loss": 0.36190190315246584, "lr": 5.71828300556897e-06, "epoch": 0.9082951791004579, "percentage": 45.41, "elapsed_time": "1 day, 4:58:24", "remaining_time": "1 day, 10:49:26", "throughput": 5657.26, "total_tokens": 590075552}
+{"current_steps": 13495, "total_steps": 29704, "loss": 0.3661900520324707, "lr": 5.715666248588289e-06, "epoch": 0.908631834096418, "percentage": 45.43, "elapsed_time": "1 day, 4:59:10", "remaining_time": "1 day, 10:48:56", "throughput": 5656.93, "total_tokens": 590300848}
+{"current_steps": 13500, "total_steps": 29704, "loss": 0.39053096771240237, "lr": 5.713049291473825e-06, "epoch": 0.9089684890923782, "percentage": 45.45, "elapsed_time": "1 day, 4:59:50", "remaining_time": "1 day, 10:48:19", "throughput": 5656.93, "total_tokens": 590529208}
+{"current_steps": 13505, "total_steps": 29704, "loss": 0.3421627998352051, "lr": 5.710432134957399e-06, "epoch": 0.9093051440883383, "percentage": 45.47, "elapsed_time": "1 day, 5:00:37", "remaining_time": "1 day, 10:47:51", "throughput": 5656.28, "total_tokens": 590729424}
+{"current_steps": 13510, "total_steps": 29704, "loss": 0.34261083602905273, "lr": 5.707814779770892e-06, "epoch": 0.9096417990842984, "percentage": 45.48, "elapsed_time": "1 day, 5:01:24", "remaining_time": "1 day, 10:47:22", "throughput": 5655.77, "total_tokens": 590940928}
+{"current_steps": 13515, "total_steps": 29704, "loss": 0.35787391662597656, "lr": 5.7051972266462395e-06, "epoch": 0.9099784540802586, "percentage": 45.5, "elapsed_time": "1 day, 5:02:01", "remaining_time": "1 day, 10:46:41", "throughput": 5655.83, "total_tokens": 591158192}
+{"current_steps": 13520, "total_steps": 29704, "loss": 0.3310979127883911, "lr": 5.702579476315431e-06, "epoch": 0.9103151090762187, "percentage": 45.52, "elapsed_time": "1 day, 5:02:46", "remaining_time": "1 day, 10:46:10", "throughput": 5655.5, "total_tokens": 591376656}
+{"current_steps": 13525, "total_steps": 29704, "loss": 0.3578958988189697, "lr": 5.699961529510513e-06, "epoch": 0.9106517640721788, "percentage": 45.53, "elapsed_time": "1 day, 5:03:27", "remaining_time": "1 day, 10:45:34", "throughput": 5655.27, "total_tokens": 591581336}
+{"current_steps": 13530, "total_steps": 29704, "loss": 0.3766601800918579, "lr": 5.697343386963586e-06, "epoch": 0.910988419068139, "percentage": 45.55, "elapsed_time": "1 day, 5:04:11", "remaining_time": "1 day, 10:45:02", "throughput": 5654.91, "total_tokens": 591797696}
+{"current_steps": 13535, "total_steps": 29704, "loss": 0.3710474967956543, "lr": 5.694725049406801e-06, "epoch": 0.9113250740640991, "percentage": 45.57, "elapsed_time": "1 day, 5:04:57", "remaining_time": "1 day, 10:44:32", "throughput": 5654.54, "total_tokens": 592014888}
+{"current_steps": 13540, "total_steps": 29704, "loss": 0.38451037406921384, "lr": 5.692106517572373e-06, "epoch": 0.9116617290600593, "percentage": 45.58, "elapsed_time": "1 day, 5:05:31", "remaining_time": "1 day, 10:43:48", "throughput": 5654.72, "total_tokens": 592229672}
+{"current_steps": 13545, "total_steps": 29704, "loss": 0.3990262746810913, "lr": 5.689487792192566e-06, "epoch": 0.9119983840560194, "percentage": 45.6, "elapsed_time": "1 day, 5:06:13", "remaining_time": "1 day, 10:43:13", "throughput": 5654.59, "total_tokens": 592451696}
+{"current_steps": 13550, "total_steps": 29704, "loss": 0.3797611236572266, "lr": 5.686868873999696e-06, "epoch": 0.9123350390519795, "percentage": 45.62, "elapsed_time": "1 day, 5:06:56", "remaining_time": "1 day, 10:42:39", "throughput": 5654.35, "total_tokens": 592668960}
+{"current_steps": 13555, "total_steps": 29704, "loss": 0.3764026641845703, "lr": 5.684249763726135e-06, "epoch": 0.9126716940479397, "percentage": 45.63, "elapsed_time": "1 day, 5:07:33", "remaining_time": "1 day, 10:41:58", "throughput": 5654.49, "total_tokens": 592890144}
+{"current_steps": 13560, "total_steps": 29704, "loss": 0.3508880615234375, "lr": 5.681630462104308e-06, "epoch": 0.9130083490438998, "percentage": 45.65, "elapsed_time": "1 day, 5:08:10", "remaining_time": "1 day, 10:41:18", "throughput": 5654.39, "total_tokens": 593092200}
+{"current_steps": 13565, "total_steps": 29704, "loss": 0.3628666877746582, "lr": 5.679010969866697e-06, "epoch": 0.91334500403986, "percentage": 45.67, "elapsed_time": "1 day, 5:08:47", "remaining_time": "1 day, 10:40:38", "throughput": 5654.4, "total_tokens": 593304400}
+{"current_steps": 13570, "total_steps": 29704, "loss": 0.35071396827697754, "lr": 5.676391287745835e-06, "epoch": 0.9136816590358201, "percentage": 45.68, "elapsed_time": "1 day, 5:09:19", "remaining_time": "1 day, 10:39:50", "throughput": 5654.95, "total_tokens": 593537816}
+{"current_steps": 13575, "total_steps": 29704, "loss": 0.3439845085144043, "lr": 5.673771416474305e-06, "epoch": 0.9140183140317802, "percentage": 45.7, "elapsed_time": "1 day, 5:09:57", "remaining_time": "1 day, 10:39:11", "throughput": 5655.03, "total_tokens": 593761520}
+{"current_steps": 13580, "total_steps": 29704, "loss": 0.3460673809051514, "lr": 5.671151356784747e-06, "epoch": 0.9143549690277404, "percentage": 45.72, "elapsed_time": "1 day, 5:10:34", "remaining_time": "1 day, 10:38:31", "throughput": 5655.01, "total_tokens": 593970768}
+{"current_steps": 13585, "total_steps": 29704, "loss": 0.3715926170349121, "lr": 5.668531109409851e-06, "epoch": 0.9146916240237005, "percentage": 45.73, "elapsed_time": "1 day, 5:11:11", "remaining_time": "1 day, 10:37:50", "throughput": 5655.12, "total_tokens": 594193064}
+{"current_steps": 13590, "total_steps": 29704, "loss": 0.36374545097351074, "lr": 5.665910675082362e-06, "epoch": 0.9150282790196607, "percentage": 45.75, "elapsed_time": "1 day, 5:11:49", "remaining_time": "1 day, 10:37:11", "throughput": 5655.13, "total_tokens": 594409288}
+{"current_steps": 13595, "total_steps": 29704, "loss": 0.34845700263977053, "lr": 5.663290054535078e-06, "epoch": 0.9153649340156208, "percentage": 45.77, "elapsed_time": "1 day, 5:12:30", "remaining_time": "1 day, 10:36:35", "throughput": 5654.93, "total_tokens": 594620336}
+{"current_steps": 13600, "total_steps": 29704, "loss": 0.39510331153869627, "lr": 5.660669248500846e-06, "epoch": 0.9157015890115809, "percentage": 45.79, "elapsed_time": "1 day, 5:13:09", "remaining_time": "1 day, 10:35:56", "throughput": 5654.95, "total_tokens": 594838312}
+{"current_steps": 13605, "total_steps": 29704, "loss": 0.3466895818710327, "lr": 5.658048257712563e-06, "epoch": 0.9160382440075411, "percentage": 45.8, "elapsed_time": "1 day, 5:13:54", "remaining_time": "1 day, 10:35:25", "throughput": 5654.62, "total_tokens": 595060816}
+{"current_steps": 13610, "total_steps": 29704, "loss": 0.3633090019226074, "lr": 5.655427082903184e-06, "epoch": 0.9163748990035012, "percentage": 45.82, "elapsed_time": "1 day, 5:14:37", "remaining_time": "1 day, 10:34:51", "throughput": 5654.4, "total_tokens": 595279584}
+{"current_steps": 13615, "total_steps": 29704, "loss": 0.35727887153625487, "lr": 5.652805724805711e-06, "epoch": 0.9167115539994614, "percentage": 45.84, "elapsed_time": "1 day, 5:15:14", "remaining_time": "1 day, 10:34:11", "throughput": 5654.48, "total_tokens": 595499192}
+{"current_steps": 13620, "total_steps": 29704, "loss": 0.3687946319580078, "lr": 5.650184184153199e-06, "epoch": 0.9170482089954215, "percentage": 45.85, "elapsed_time": "1 day, 5:15:52", "remaining_time": "1 day, 10:33:31", "throughput": 5654.59, "total_tokens": 595723264}
+{"current_steps": 13625, "total_steps": 29704, "loss": 0.39588251113891604, "lr": 5.647562461678754e-06, "epoch": 0.9173848639913816, "percentage": 45.87, "elapsed_time": "1 day, 5:16:28", "remaining_time": "1 day, 10:32:50", "throughput": 5654.75, "total_tokens": 595947688}
+{"current_steps": 13630, "total_steps": 29704, "loss": 0.35439510345458985, "lr": 5.64494055811553e-06, "epoch": 0.9177215189873418, "percentage": 45.89, "elapsed_time": "1 day, 5:17:12", "remaining_time": "1 day, 10:32:17", "throughput": 5654.51, "total_tokens": 596167280}
+{"current_steps": 13635, "total_steps": 29704, "loss": 0.3450098991394043, "lr": 5.642318474196737e-06, "epoch": 0.9180581739833019, "percentage": 45.9, "elapsed_time": "1 day, 5:17:50", "remaining_time": "1 day, 10:31:37", "throughput": 5654.57, "total_tokens": 596389280}
+{"current_steps": 13640, "total_steps": 29704, "loss": 0.36347208023071287, "lr": 5.63969621065563e-06, "epoch": 0.9183948289792621, "percentage": 45.92, "elapsed_time": "1 day, 5:18:30", "remaining_time": "1 day, 10:31:00", "throughput": 5654.47, "total_tokens": 596603992}
+{"current_steps": 13645, "total_steps": 29704, "loss": 0.3404127836227417, "lr": 5.637073768225517e-06, "epoch": 0.9187314839752222, "percentage": 45.94, "elapsed_time": "1 day, 5:19:15", "remaining_time": "1 day, 10:30:29", "throughput": 5654.03, "total_tokens": 596814560}
+{"current_steps": 13650, "total_steps": 29704, "loss": 0.3368164300918579, "lr": 5.634451147639758e-06, "epoch": 0.9190681389711823, "percentage": 45.95, "elapsed_time": "1 day, 5:19:56", "remaining_time": "1 day, 10:29:53", "throughput": 5653.77, "total_tokens": 597016352}
+{"current_steps": 13655, "total_steps": 29704, "loss": 0.34926304817199705, "lr": 5.631828349631757e-06, "epoch": 0.9194047939671425, "percentage": 45.97, "elapsed_time": "1 day, 5:20:33", "remaining_time": "1 day, 10:29:12", "throughput": 5653.92, "total_tokens": 597241424}
+{"current_steps": 13660, "total_steps": 29704, "loss": 0.3418566703796387, "lr": 5.629205374934974e-06, "epoch": 0.9197414489631026, "percentage": 45.99, "elapsed_time": "1 day, 5:21:10", "remaining_time": "1 day, 10:28:32", "throughput": 5653.88, "total_tokens": 597449976}
+{"current_steps": 13665, "total_steps": 29704, "loss": 0.36661295890808104, "lr": 5.626582224282913e-06, "epoch": 0.9200781039590628, "percentage": 46.0, "elapsed_time": "1 day, 5:21:54", "remaining_time": "1 day, 10:27:59", "throughput": 5653.7, "total_tokens": 597675264}
+{"current_steps": 13670, "total_steps": 29704, "loss": 0.39702858924865725, "lr": 5.623958898409132e-06, "epoch": 0.9204147589550229, "percentage": 46.02, "elapsed_time": "1 day, 5:22:23", "remaining_time": "1 day, 10:27:10", "throughput": 5654.23, "total_tokens": 597900752}
+{"current_steps": 13675, "total_steps": 29704, "loss": 0.3848641157150269, "lr": 5.621335398047234e-06, "epoch": 0.920751413950983, "percentage": 46.04, "elapsed_time": "1 day, 5:23:07", "remaining_time": "1 day, 10:26:37", "throughput": 5653.96, "total_tokens": 598115680}
+{"current_steps": 13680, "total_steps": 29704, "loss": 0.36865854263305664, "lr": 5.618711723930874e-06, "epoch": 0.9210880689469432, "percentage": 46.05, "elapsed_time": "1 day, 5:23:37", "remaining_time": "1 day, 10:25:49", "throughput": 5654.37, "total_tokens": 598333664}
+{"current_steps": 13685, "total_steps": 29704, "loss": 0.3818777561187744, "lr": 5.616087876793753e-06, "epoch": 0.9214247239429033, "percentage": 46.07, "elapsed_time": "1 day, 5:24:13", "remaining_time": "1 day, 10:25:06", "throughput": 5654.63, "total_tokens": 598560984}
+{"current_steps": 13690, "total_steps": 29704, "loss": 0.3648639678955078, "lr": 5.613463857369621e-06, "epoch": 0.9217613789388635, "percentage": 46.09, "elapsed_time": "1 day, 5:24:55", "remaining_time": "1 day, 10:24:32", "throughput": 5654.49, "total_tokens": 598784192}
+{"current_steps": 13695, "total_steps": 29704, "loss": 0.35527966022491453, "lr": 5.610839666392278e-06, "epoch": 0.9220980339348236, "percentage": 46.1, "elapsed_time": "1 day, 5:25:41", "remaining_time": "1 day, 10:24:01", "throughput": 5654.14, "total_tokens": 599005456}
+{"current_steps": 13700, "total_steps": 29704, "loss": 0.3818879842758179, "lr": 5.608215304595571e-06, "epoch": 0.9224346889307837, "percentage": 46.12, "elapsed_time": "1 day, 5:26:27", "remaining_time": "1 day, 10:23:32", "throughput": 5653.7, "total_tokens": 599222856}
+{"current_steps": 13705, "total_steps": 29704, "loss": 0.35137481689453126, "lr": 5.605590772713393e-06, "epoch": 0.9227713439267439, "percentage": 46.14, "elapsed_time": "1 day, 5:27:05", "remaining_time": "1 day, 10:22:52", "throughput": 5653.58, "total_tokens": 599421784}
+{"current_steps": 13710, "total_steps": 29704, "loss": 0.3719889402389526, "lr": 5.602966071479687e-06, "epoch": 0.923107998922704, "percentage": 46.16, "elapsed_time": "1 day, 5:27:37", "remaining_time": "1 day, 10:22:05", "throughput": 5653.8, "total_tokens": 599626912}
+{"current_steps": 13715, "total_steps": 29704, "loss": 0.37358889579772947, "lr": 5.600341201628439e-06, "epoch": 0.9234446539186641, "percentage": 46.17, "elapsed_time": "1 day, 5:28:10", "remaining_time": "1 day, 10:21:20", "throughput": 5654.17, "total_tokens": 599854192}
+{"current_steps": 13720, "total_steps": 29704, "loss": 0.31069388389587405, "lr": 5.597716163893692e-06, "epoch": 0.9237813089146243, "percentage": 46.19, "elapsed_time": "1 day, 5:28:54", "remaining_time": "1 day, 10:20:48", "throughput": 5653.99, "total_tokens": 600085080}
+{"current_steps": 13725, "total_steps": 29704, "loss": 0.34327847957611085, "lr": 5.595090959009525e-06, "epoch": 0.9241179639105844, "percentage": 46.21, "elapsed_time": "1 day, 5:29:40", "remaining_time": "1 day, 10:20:17", "throughput": 5653.69, "total_tokens": 600310096}
+{"current_steps": 13730, "total_steps": 29704, "loss": 0.34142630100250243, "lr": 5.592465587710068e-06, "epoch": 0.9244546189065446, "percentage": 46.22, "elapsed_time": "1 day, 5:30:18", "remaining_time": "1 day, 10:19:38", "throughput": 5653.86, "total_tokens": 600545032}
+{"current_steps": 13735, "total_steps": 29704, "loss": 0.3286783218383789, "lr": 5.589840050729498e-06, "epoch": 0.9247912739025047, "percentage": 46.24, "elapsed_time": "1 day, 5:30:58", "remaining_time": "1 day, 10:19:00", "throughput": 5653.86, "total_tokens": 600768064}
+{"current_steps": 13740, "total_steps": 29704, "loss": 0.3909108638763428, "lr": 5.587214348802039e-06, "epoch": 0.9251279288984648, "percentage": 46.26, "elapsed_time": "1 day, 5:31:38", "remaining_time": "1 day, 10:18:24", "throughput": 5653.71, "total_tokens": 600980632}
+{"current_steps": 13745, "total_steps": 29704, "loss": 0.36343495845794677, "lr": 5.5845884826619615e-06, "epoch": 0.925464583894425, "percentage": 46.27, "elapsed_time": "1 day, 5:32:13", "remaining_time": "1 day, 10:17:41", "throughput": 5653.97, "total_tokens": 601207448}
+{"current_steps": 13750, "total_steps": 29704, "loss": 0.3585641622543335, "lr": 5.5819624530435775e-06, "epoch": 0.9258012388903851, "percentage": 46.29, "elapsed_time": "1 day, 5:32:53", "remaining_time": "1 day, 10:17:04", "throughput": 5653.83, "total_tokens": 601419704}
+{"current_steps": 13755, "total_steps": 29704, "loss": 0.3933718681335449, "lr": 5.57933626068125e-06, "epoch": 0.9261378938863453, "percentage": 46.31, "elapsed_time": "1 day, 5:33:34", "remaining_time": "1 day, 10:16:28", "throughput": 5653.73, "total_tokens": 601639824}
+{"current_steps": 13760, "total_steps": 29704, "loss": 0.3709197998046875, "lr": 5.5767099063093835e-06, "epoch": 0.9264745488823054, "percentage": 46.32, "elapsed_time": "1 day, 5:34:12", "remaining_time": "1 day, 10:15:48", "throughput": 5653.88, "total_tokens": 601868536}
+{"current_steps": 13765, "total_steps": 29704, "loss": 0.373383617401123, "lr": 5.574083390662431e-06, "epoch": 0.9268112038782655, "percentage": 46.34, "elapsed_time": "1 day, 5:34:49", "remaining_time": "1 day, 10:15:07", "throughput": 5654.09, "total_tokens": 602098696}
+{"current_steps": 13770, "total_steps": 29704, "loss": 0.3655261993408203, "lr": 5.57145671447489e-06, "epoch": 0.9271478588742257, "percentage": 46.36, "elapsed_time": "1 day, 5:35:25", "remaining_time": "1 day, 10:14:26", "throughput": 5654.05, "total_tokens": 602302480}
+{"current_steps": 13775, "total_steps": 29704, "loss": 0.3984712600708008, "lr": 5.5688298784813e-06, "epoch": 0.9274845138701858, "percentage": 46.37, "elapsed_time": "1 day, 5:36:04", "remaining_time": "1 day, 10:13:47", "throughput": 5654.07, "total_tokens": 602522208}
+{"current_steps": 13780, "total_steps": 29704, "loss": 0.3970597743988037, "lr": 5.566202883416249e-06, "epoch": 0.927821168866146, "percentage": 46.39, "elapsed_time": "1 day, 5:36:46", "remaining_time": "1 day, 10:13:13", "throughput": 5653.96, "total_tokens": 602750424}
+{"current_steps": 13785, "total_steps": 29704, "loss": 0.39171276092529295, "lr": 5.563575730014366e-06, "epoch": 0.9281578238621061, "percentage": 46.41, "elapsed_time": "1 day, 5:37:23", "remaining_time": "1 day, 10:12:32", "throughput": 5654.03, "total_tokens": 602965320}
+{"current_steps": 13790, "total_steps": 29704, "loss": 0.3737471342086792, "lr": 5.560948419010328e-06, "epoch": 0.9284944788580662, "percentage": 46.42, "elapsed_time": "1 day, 5:38:01", "remaining_time": "1 day, 10:11:52", "throughput": 5654.09, "total_tokens": 603186136}
+{"current_steps": 13795, "total_steps": 29704, "loss": 0.34050416946411133, "lr": 5.558320951138853e-06, "epoch": 0.9288311338540264, "percentage": 46.44, "elapsed_time": "1 day, 5:38:42", "remaining_time": "1 day, 10:11:17", "throughput": 5653.89, "total_tokens": 603397288}
+{"current_steps": 13800, "total_steps": 29704, "loss": 0.3723226308822632, "lr": 5.5556933271347046e-06, "epoch": 0.9291677888499865, "percentage": 46.46, "elapsed_time": "1 day, 5:39:23", "remaining_time": "1 day, 10:10:41", "throughput": 5653.67, "total_tokens": 603606088}
+{"current_steps": 13805, "total_steps": 29704, "loss": 0.3821061611175537, "lr": 5.55306554773269e-06, "epoch": 0.9295044438459467, "percentage": 46.48, "elapsed_time": "1 day, 5:40:05", "remaining_time": "1 day, 10:10:06", "throughput": 5653.4, "total_tokens": 603816144}
+{"current_steps": 13810, "total_steps": 29704, "loss": 0.3593113660812378, "lr": 5.550437613667658e-06, "epoch": 0.9298410988419068, "percentage": 46.49, "elapsed_time": "1 day, 5:40:36", "remaining_time": "1 day, 10:09:19", "throughput": 5653.52, "total_tokens": 604004552}
+{"current_steps": 13815, "total_steps": 29704, "loss": 0.35454933643341063, "lr": 5.547809525674501e-06, "epoch": 0.9301777538378669, "percentage": 46.51, "elapsed_time": "1 day, 5:41:15", "remaining_time": "1 day, 10:08:39", "throughput": 5653.44, "total_tokens": 604212440}
+{"current_steps": 13820, "total_steps": 29704, "loss": 0.38926100730895996, "lr": 5.545181284488159e-06, "epoch": 0.9305144088338271, "percentage": 46.53, "elapsed_time": "1 day, 5:41:57", "remaining_time": "1 day, 10:08:05", "throughput": 5653.33, "total_tokens": 604439792}
+{"current_steps": 13825, "total_steps": 29704, "loss": 0.35577256679534913, "lr": 5.5425528908436085e-06, "epoch": 0.9308510638297872, "percentage": 46.54, "elapsed_time": "1 day, 5:42:33", "remaining_time": "1 day, 10:07:23", "throughput": 5653.34, "total_tokens": 604643600}
+{"current_steps": 13830, "total_steps": 29704, "loss": 0.3907893180847168, "lr": 5.539924345475873e-06, "epoch": 0.9311877188257474, "percentage": 46.56, "elapsed_time": "1 day, 5:43:08", "remaining_time": "1 day, 10:06:41", "throughput": 5653.29, "total_tokens": 604839736}
+{"current_steps": 13835, "total_steps": 29704, "loss": 0.37813870906829833, "lr": 5.5372956491200145e-06, "epoch": 0.9315243738217075, "percentage": 46.58, "elapsed_time": "1 day, 5:43:45", "remaining_time": "1 day, 10:06:00", "throughput": 5653.33, "total_tokens": 605051728}
+{"current_steps": 13840, "total_steps": 29704, "loss": 0.36364994049072263, "lr": 5.534666802511143e-06, "epoch": 0.9318610288176676, "percentage": 46.59, "elapsed_time": "1 day, 5:44:22", "remaining_time": "1 day, 10:05:20", "throughput": 5653.23, "total_tokens": 605251488}
+{"current_steps": 13845, "total_steps": 29704, "loss": 0.372878885269165, "lr": 5.532037806384404e-06, "epoch": 0.9321976838136278, "percentage": 46.61, "elapsed_time": "1 day, 5:45:01", "remaining_time": "1 day, 10:04:41", "throughput": 5653.3, "total_tokens": 605478256}
+{"current_steps": 13850, "total_steps": 29704, "loss": 0.33782191276550294, "lr": 5.529408661474989e-06, "epoch": 0.9325343388095879, "percentage": 46.63, "elapsed_time": "1 day, 5:45:40", "remaining_time": "1 day, 10:04:02", "throughput": 5653.12, "total_tokens": 605678152}
+{"current_steps": 13855, "total_steps": 29704, "loss": 0.3900485038757324, "lr": 5.526779368518129e-06, "epoch": 0.932870993805548, "percentage": 46.64, "elapsed_time": "1 day, 5:46:14", "remaining_time": "1 day, 10:03:18", "throughput": 5653.44, "total_tokens": 605902976}
+{"current_steps": 13860, "total_steps": 29704, "loss": 0.35591723918914797, "lr": 5.5241499282491e-06, "epoch": 0.9332076488015082, "percentage": 46.66, "elapsed_time": "1 day, 5:46:52", "remaining_time": "1 day, 10:02:39", "throughput": 5653.28, "total_tokens": 606103560}
+{"current_steps": 13865, "total_steps": 29704, "loss": 0.3359647274017334, "lr": 5.5215203414032135e-06, "epoch": 0.9335443037974683, "percentage": 46.68, "elapsed_time": "1 day, 5:47:41", "remaining_time": "1 day, 10:02:12", "throughput": 5652.86, "total_tokens": 606332784}
+{"current_steps": 13870, "total_steps": 29704, "loss": 0.3636192798614502, "lr": 5.518890608715828e-06, "epoch": 0.9338809587934285, "percentage": 46.69, "elapsed_time": "1 day, 5:48:26", "remaining_time": "1 day, 10:01:41", "throughput": 5652.47, "total_tokens": 606549080}
+{"current_steps": 13875, "total_steps": 29704, "loss": 0.3732717275619507, "lr": 5.516260730922337e-06, "epoch": 0.9342176137893886, "percentage": 46.71, "elapsed_time": "1 day, 5:49:04", "remaining_time": "1 day, 10:01:01", "throughput": 5652.41, "total_tokens": 606754664}
+{"current_steps": 13880, "total_steps": 29704, "loss": 0.3665364503860474, "lr": 5.51363070875818e-06, "epoch": 0.9345542687853488, "percentage": 46.73, "elapsed_time": "1 day, 5:49:40", "remaining_time": "1 day, 10:00:20", "throughput": 5652.51, "total_tokens": 606971336}
+{"current_steps": 13885, "total_steps": 29704, "loss": 0.3829598903656006, "lr": 5.511000542958832e-06, "epoch": 0.9348909237813089, "percentage": 46.74, "elapsed_time": "1 day, 5:50:25", "remaining_time": "1 day, 9:59:48", "throughput": 5652.13, "total_tokens": 607182888}
+{"current_steps": 13890, "total_steps": 29704, "loss": 0.3617416858673096, "lr": 5.508370234259812e-06, "epoch": 0.935227578777269, "percentage": 46.76, "elapsed_time": "1 day, 5:51:07", "remaining_time": "1 day, 9:59:13", "throughput": 5651.92, "total_tokens": 607395376}
+{"current_steps": 13895, "total_steps": 29704, "loss": 0.4197434425354004, "lr": 5.505739783396678e-06, "epoch": 0.9355642337732292, "percentage": 46.78, "elapsed_time": "1 day, 5:51:41", "remaining_time": "1 day, 9:58:29", "throughput": 5652.08, "total_tokens": 607606560}
+{"current_steps": 13900, "total_steps": 29704, "loss": 0.368869423866272, "lr": 5.503109191105026e-06, "epoch": 0.9359008887691893, "percentage": 46.8, "elapsed_time": "1 day, 5:52:22", "remaining_time": "1 day, 9:57:53", "throughput": 5651.9, "total_tokens": 607818184}
+{"current_steps": 13905, "total_steps": 29704, "loss": 0.3878501892089844, "lr": 5.500478458120493e-06, "epoch": 0.9362375437651494, "percentage": 46.81, "elapsed_time": "1 day, 5:52:58", "remaining_time": "1 day, 9:57:12", "throughput": 5652.09, "total_tokens": 608046144}
+{"current_steps": 13910, "total_steps": 29704, "loss": 0.37720532417297364, "lr": 5.497847585178754e-06, "epoch": 0.9365741987611096, "percentage": 46.83, "elapsed_time": "1 day, 5:53:34", "remaining_time": "1 day, 9:56:29", "throughput": 5652.36, "total_tokens": 608274496}
+{"current_steps": 13915, "total_steps": 29704, "loss": 0.34202401638031005, "lr": 5.495216573015526e-06, "epoch": 0.9369108537570697, "percentage": 46.85, "elapsed_time": "1 day, 5:54:09", "remaining_time": "1 day, 9:55:47", "throughput": 5652.47, "total_tokens": 608487072}
+{"current_steps": 13920, "total_steps": 29704, "loss": 0.3395219802856445, "lr": 5.492585422366562e-06, "epoch": 0.9372475087530299, "percentage": 46.86, "elapsed_time": "1 day, 5:54:50", "remaining_time": "1 day, 9:55:11", "throughput": 5652.33, "total_tokens": 608702136}
+{"current_steps": 13925, "total_steps": 29704, "loss": 0.3353360652923584, "lr": 5.489954133967657e-06, "epoch": 0.93758416374899, "percentage": 46.88, "elapsed_time": "1 day, 5:55:30", "remaining_time": "1 day, 9:54:34", "throughput": 5652.13, "total_tokens": 608907648}
+{"current_steps": 13930, "total_steps": 29704, "loss": 0.3447211027145386, "lr": 5.48732270855464e-06, "epoch": 0.9379208187449501, "percentage": 46.9, "elapsed_time": "1 day, 5:56:10", "remaining_time": "1 day, 9:53:56", "throughput": 5652.19, "total_tokens": 609137704}
+{"current_steps": 13935, "total_steps": 29704, "loss": 0.36331336498260497, "lr": 5.48469114686338e-06, "epoch": 0.9382574737409103, "percentage": 46.91, "elapsed_time": "1 day, 5:56:48", "remaining_time": "1 day, 9:53:17", "throughput": 5652.06, "total_tokens": 609342520}
+{"current_steps": 13940, "total_steps": 29704, "loss": 0.3539914131164551, "lr": 5.482059449629788e-06, "epoch": 0.9385941287368704, "percentage": 46.93, "elapsed_time": "1 day, 5:57:27", "remaining_time": "1 day, 9:52:39", "throughput": 5652.15, "total_tokens": 609571440}
+{"current_steps": 13945, "total_steps": 29704, "loss": 0.3414954662322998, "lr": 5.479427617589808e-06, "epoch": 0.9389307837328306, "percentage": 46.95, "elapsed_time": "1 day, 5:58:05", "remaining_time": "1 day, 9:51:59", "throughput": 5652.17, "total_tokens": 609787136}
+{"current_steps": 13950, "total_steps": 29704, "loss": 0.3689284801483154, "lr": 5.4767956514794226e-06, "epoch": 0.9392674387287907, "percentage": 46.96, "elapsed_time": "1 day, 5:58:38", "remaining_time": "1 day, 9:51:14", "throughput": 5652.58, "total_tokens": 610015912}
+{"current_steps": 13955, "total_steps": 29704, "loss": 0.3568282127380371, "lr": 5.474163552034655e-06, "epoch": 0.9396040937247508, "percentage": 46.98, "elapsed_time": "1 day, 5:59:18", "remaining_time": "1 day, 9:50:37", "throughput": 5652.38, "total_tokens": 610224872}
+{"current_steps": 13960, "total_steps": 29704, "loss": 0.33748788833618165, "lr": 5.4715313199915595e-06, "epoch": 0.939940748720711, "percentage": 47.0, "elapsed_time": "1 day, 5:59:58", "remaining_time": "1 day, 9:50:00", "throughput": 5652.22, "total_tokens": 610433896}
+{"current_steps": 13965, "total_steps": 29704, "loss": 0.36934537887573243, "lr": 5.468898956086236e-06, "epoch": 0.9402774037166711, "percentage": 47.01, "elapsed_time": "1 day, 6:00:37", "remaining_time": "1 day, 9:49:21", "throughput": 5652.3, "total_tokens": 610657928}
+{"current_steps": 13970, "total_steps": 29704, "loss": 0.3693251132965088, "lr": 5.466266461054815e-06, "epoch": 0.9406140587126313, "percentage": 47.03, "elapsed_time": "1 day, 6:01:18", "remaining_time": "1 day, 9:48:46", "throughput": 5652.16, "total_tokens": 610879072}
+{"current_steps": 13975, "total_steps": 29704, "loss": 0.38323063850402833, "lr": 5.463633835633464e-06, "epoch": 0.9409507137085914, "percentage": 47.05, "elapsed_time": "1 day, 6:01:56", "remaining_time": "1 day, 9:48:06", "throughput": 5652.33, "total_tokens": 611111488}
+{"current_steps": 13980, "total_steps": 29704, "loss": 0.37401399612426756, "lr": 5.46100108055839e-06, "epoch": 0.9412873687045515, "percentage": 47.06, "elapsed_time": "1 day, 6:02:33", "remaining_time": "1 day, 9:47:25", "throughput": 5652.5, "total_tokens": 611338984}
+{"current_steps": 13985, "total_steps": 29704, "loss": 0.34747042655944826, "lr": 5.458368196565833e-06, "epoch": 0.9416240237005117, "percentage": 47.08, "elapsed_time": "1 day, 6:03:07", "remaining_time": "1 day, 9:46:41", "throughput": 5652.76, "total_tokens": 611557384}
+{"current_steps": 13990, "total_steps": 29704, "loss": 0.39068779945373533, "lr": 5.455735184392073e-06, "epoch": 0.9419606786964718, "percentage": 47.1, "elapsed_time": "1 day, 6:03:42", "remaining_time": "1 day, 9:45:58", "throughput": 5653.05, "total_tokens": 611785288}
+{"current_steps": 13995, "total_steps": 29704, "loss": 0.3776577949523926, "lr": 5.4531020447734204e-06, "epoch": 0.942297333692432, "percentage": 47.11, "elapsed_time": "1 day, 6:04:15", "remaining_time": "1 day, 9:45:13", "throughput": 5653.3, "total_tokens": 611999096}
+{"current_steps": 14000, "total_steps": 29704, "loss": 0.38014535903930663, "lr": 5.450468778446226e-06, "epoch": 0.9426339886883921, "percentage": 47.13, "elapsed_time": "1 day, 6:05:02", "remaining_time": "1 day, 9:44:44", "throughput": 5652.97, "total_tokens": 612229616}
+{"current_steps": 14005, "total_steps": 29704, "loss": 0.3620244264602661, "lr": 5.447835386146875e-06, "epoch": 0.9429706436843522, "percentage": 47.15, "elapsed_time": "1 day, 6:05:39", "remaining_time": "1 day, 9:44:04", "throughput": 5653.12, "total_tokens": 612457288}
+{"current_steps": 14010, "total_steps": 29704, "loss": 0.4034558296203613, "lr": 5.445201868611784e-06, "epoch": 0.9433072986803124, "percentage": 47.17, "elapsed_time": "1 day, 6:06:16", "remaining_time": "1 day, 9:43:23", "throughput": 5653.13, "total_tokens": 612668352}
+{"current_steps": 14015, "total_steps": 29704, "loss": 0.37315356731414795, "lr": 5.442568226577413e-06, "epoch": 0.9436439536762725, "percentage": 47.18, "elapsed_time": "1 day, 6:06:55", "remaining_time": "1 day, 9:42:45", "throughput": 5653.11, "total_tokens": 612885272}
+{"current_steps": 14020, "total_steps": 29704, "loss": 0.3591346740722656, "lr": 5.439934460780247e-06, "epoch": 0.9439806086722327, "percentage": 47.2, "elapsed_time": "1 day, 6:07:33", "remaining_time": "1 day, 9:42:06", "throughput": 5653.17, "total_tokens": 613107944}
+{"current_steps": 14025, "total_steps": 29704, "loss": 0.3745762586593628, "lr": 5.437300571956813e-06, "epoch": 0.9443172636681928, "percentage": 47.22, "elapsed_time": "1 day, 6:08:09", "remaining_time": "1 day, 9:41:23", "throughput": 5653.15, "total_tokens": 613308168}
+{"current_steps": 14030, "total_steps": 29704, "loss": 0.3551823616027832, "lr": 5.434666560843667e-06, "epoch": 0.9446539186641529, "percentage": 47.23, "elapsed_time": "1 day, 6:08:45", "remaining_time": "1 day, 9:40:41", "throughput": 5653.21, "total_tokens": 613516376}
+{"current_steps": 14035, "total_steps": 29704, "loss": 0.3733973979949951, "lr": 5.432032428177405e-06, "epoch": 0.9449905736601131, "percentage": 47.25, "elapsed_time": "1 day, 6:09:25", "remaining_time": "1 day, 9:40:04", "throughput": 5653.23, "total_tokens": 613743992}
+{"current_steps": 14040, "total_steps": 29704, "loss": 0.35695390701293944, "lr": 5.429398174694651e-06, "epoch": 0.9453272286560732, "percentage": 47.27, "elapsed_time": "1 day, 6:10:00", "remaining_time": "1 day, 9:39:22", "throughput": 5653.51, "total_tokens": 613975032}
+{"current_steps": 14045, "total_steps": 29704, "loss": 0.3617269039154053, "lr": 5.426763801132067e-06, "epoch": 0.9456638836520334, "percentage": 47.28, "elapsed_time": "1 day, 6:10:37", "remaining_time": "1 day, 9:38:41", "throughput": 5653.32, "total_tokens": 614162568}
+{"current_steps": 14050, "total_steps": 29704, "loss": 0.3843861103057861, "lr": 5.424129308226347e-06, "epoch": 0.9460005386479935, "percentage": 47.3, "elapsed_time": "1 day, 6:11:15", "remaining_time": "1 day, 9:38:02", "throughput": 5653.25, "total_tokens": 614370656}
+{"current_steps": 14055, "total_steps": 29704, "loss": 0.3902265548706055, "lr": 5.421494696714219e-06, "epoch": 0.9463371936439536, "percentage": 47.32, "elapsed_time": "1 day, 6:11:55", "remaining_time": "1 day, 9:37:24", "throughput": 5653.38, "total_tokens": 614608864}
+{"current_steps": 14060, "total_steps": 29704, "loss": 0.3707080841064453, "lr": 5.418859967332442e-06, "epoch": 0.9466738486399138, "percentage": 47.33, "elapsed_time": "1 day, 6:12:35", "remaining_time": "1 day, 9:36:47", "throughput": 5653.4, "total_tokens": 614836608}
+{"current_steps": 14065, "total_steps": 29704, "loss": 0.3681238412857056, "lr": 5.416225120817811e-06, "epoch": 0.9470105036358739, "percentage": 47.35, "elapsed_time": "1 day, 6:13:10", "remaining_time": "1 day, 9:36:05", "throughput": 5653.29, "total_tokens": 615027064}
+{"current_steps": 14070, "total_steps": 29704, "loss": 0.3346075534820557, "lr": 5.413590157907153e-06, "epoch": 0.947347158631834, "percentage": 47.37, "elapsed_time": "1 day, 6:13:46", "remaining_time": "1 day, 9:35:23", "throughput": 5653.68, "total_tokens": 615270632}
+{"current_steps": 14075, "total_steps": 29704, "loss": 0.355800724029541, "lr": 5.4109550793373255e-06, "epoch": 0.9476838136277942, "percentage": 47.38, "elapsed_time": "1 day, 6:14:28", "remaining_time": "1 day, 9:34:48", "throughput": 5653.57, "total_tokens": 615493608}
+{"current_steps": 14080, "total_steps": 29704, "loss": 0.38836262226104734, "lr": 5.40831988584522e-06, "epoch": 0.9480204686237543, "percentage": 47.4, "elapsed_time": "1 day, 6:15:04", "remaining_time": "1 day, 9:34:06", "throughput": 5653.78, "total_tokens": 615719216}
+{"current_steps": 14085, "total_steps": 29704, "loss": 0.37587385177612304, "lr": 5.40568457816776e-06, "epoch": 0.9483571236197145, "percentage": 47.42, "elapsed_time": "1 day, 6:15:47", "remaining_time": "1 day, 9:33:32", "throughput": 5653.68, "total_tokens": 615951848}
+{"current_steps": 14090, "total_steps": 29704, "loss": 0.35560946464538573, "lr": 5.403049157041903e-06, "epoch": 0.9486937786156746, "percentage": 47.43, "elapsed_time": "1 day, 6:16:24", "remaining_time": "1 day, 9:32:52", "throughput": 5653.76, "total_tokens": 616172928}
+{"current_steps": 14095, "total_steps": 29704, "loss": 0.35712380409240724, "lr": 5.400413623204635e-06, "epoch": 0.9490304336116347, "percentage": 47.45, "elapsed_time": "1 day, 6:17:03", "remaining_time": "1 day, 9:32:13", "throughput": 5653.75, "total_tokens": 616390224}
+{"current_steps": 14100, "total_steps": 29704, "loss": 0.39092726707458497, "lr": 5.397777977392973e-06, "epoch": 0.9493670886075949, "percentage": 47.47, "elapsed_time": "1 day, 6:17:51", "remaining_time": "1 day, 9:31:45", "throughput": 5653.3, "total_tokens": 616612064}
+{"current_steps": 14105, "total_steps": 29704, "loss": 0.37581348419189453, "lr": 5.395142220343967e-06, "epoch": 0.949703743603555, "percentage": 47.49, "elapsed_time": "1 day, 6:18:34", "remaining_time": "1 day, 9:31:12", "throughput": 5653.1, "total_tokens": 616835560}
+{"current_steps": 14110, "total_steps": 29704, "loss": 0.34862990379333497, "lr": 5.3925063527947e-06, "epoch": 0.9500403985995152, "percentage": 47.5, "elapsed_time": "1 day, 6:19:21", "remaining_time": "1 day, 9:30:42", "throughput": 5652.63, "total_tokens": 617051200}
+{"current_steps": 14115, "total_steps": 29704, "loss": 0.37904350757598876, "lr": 5.389870375482283e-06, "epoch": 0.9503770535954753, "percentage": 47.52, "elapsed_time": "1 day, 6:19:54", "remaining_time": "1 day, 9:29:57", "throughput": 5652.74, "total_tokens": 617248976}
+{"current_steps": 14120, "total_steps": 29704, "loss": 0.331813383102417, "lr": 5.387234289143859e-06, "epoch": 0.9507137085914354, "percentage": 47.54, "elapsed_time": "1 day, 6:20:26", "remaining_time": "1 day, 9:29:11", "throughput": 5653.08, "total_tokens": 617469104}
+{"current_steps": 14125, "total_steps": 29704, "loss": 0.3729023218154907, "lr": 5.384598094516601e-06, "epoch": 0.9510503635873956, "percentage": 47.55, "elapsed_time": "1 day, 6:21:11", "remaining_time": "1 day, 9:28:39", "throughput": 5652.82, "total_tokens": 617692328}
+{"current_steps": 14130, "total_steps": 29704, "loss": 0.3759523153305054, "lr": 5.381961792337712e-06, "epoch": 0.9513870185833557, "percentage": 47.57, "elapsed_time": "1 day, 6:21:47", "remaining_time": "1 day, 9:27:58", "throughput": 5653.06, "total_tokens": 617921872}
+{"current_steps": 14135, "total_steps": 29704, "loss": 0.3401124715805054, "lr": 5.3793253833444245e-06, "epoch": 0.9517236735793159, "percentage": 47.59, "elapsed_time": "1 day, 6:22:22", "remaining_time": "1 day, 9:27:15", "throughput": 5653.29, "total_tokens": 618147624}
+{"current_steps": 14140, "total_steps": 29704, "loss": 0.3542716264724731, "lr": 5.3766888682740045e-06, "epoch": 0.952060328575276, "percentage": 47.6, "elapsed_time": "1 day, 6:22:58", "remaining_time": "1 day, 9:26:33", "throughput": 5653.53, "total_tokens": 618375336}
+{"current_steps": 14145, "total_steps": 29704, "loss": 0.3719951152801514, "lr": 5.374052247863745e-06, "epoch": 0.9523969835712361, "percentage": 47.62, "elapsed_time": "1 day, 6:23:32", "remaining_time": "1 day, 9:25:50", "throughput": 5653.71, "total_tokens": 618588568}
+{"current_steps": 14150, "total_steps": 29704, "loss": 0.34739990234375, "lr": 5.371415522850966e-06, "epoch": 0.9527336385671963, "percentage": 47.64, "elapsed_time": "1 day, 6:24:15", "remaining_time": "1 day, 9:25:16", "throughput": 5653.5, "total_tokens": 618807136}
+{"current_steps": 14155, "total_steps": 29704, "loss": 0.35079226493835447, "lr": 5.368778693973019e-06, "epoch": 0.9530702935631564, "percentage": 47.65, "elapsed_time": "1 day, 6:24:57", "remaining_time": "1 day, 9:24:40", "throughput": 5653.3, "total_tokens": 619020192}
+{"current_steps": 14160, "total_steps": 29704, "loss": 0.3731173753738403, "lr": 5.366141761967287e-06, "epoch": 0.9534069485591166, "percentage": 47.67, "elapsed_time": "1 day, 6:25:35", "remaining_time": "1 day, 9:24:01", "throughput": 5653.26, "total_tokens": 619234312}
+{"current_steps": 14165, "total_steps": 29704, "loss": 0.38480224609375, "lr": 5.363504727571179e-06, "epoch": 0.9537436035550767, "percentage": 47.69, "elapsed_time": "1 day, 6:26:14", "remaining_time": "1 day, 9:23:23", "throughput": 5653.29, "total_tokens": 619455112}
+{"current_steps": 14170, "total_steps": 29704, "loss": 0.4258725166320801, "lr": 5.360867591522133e-06, "epoch": 0.954080258551037, "percentage": 47.7, "elapsed_time": "1 day, 6:26:54", "remaining_time": "1 day, 9:22:45", "throughput": 5653.33, "total_tokens": 619686368}
+{"current_steps": 14175, "total_steps": 29704, "loss": 0.34602572917938235, "lr": 5.358230354557617e-06, "epoch": 0.9544169135469971, "percentage": 47.72, "elapsed_time": "1 day, 6:27:27", "remaining_time": "1 day, 9:22:01", "throughput": 5653.68, "total_tokens": 619912184}
+{"current_steps": 14180, "total_steps": 29704, "loss": 0.39844369888305664, "lr": 5.355593017415123e-06, "epoch": 0.9547535685429572, "percentage": 47.74, "elapsed_time": "1 day, 6:28:01", "remaining_time": "1 day, 9:21:16", "throughput": 5654.02, "total_tokens": 620138400}
+{"current_steps": 14185, "total_steps": 29704, "loss": 0.36531069278717043, "lr": 5.352955580832176e-06, "epoch": 0.9550902235389174, "percentage": 47.75, "elapsed_time": "1 day, 6:28:40", "remaining_time": "1 day, 9:20:39", "throughput": 5654.01, "total_tokens": 620363336}
+{"current_steps": 14190, "total_steps": 29704, "loss": 0.378409218788147, "lr": 5.350318045546326e-06, "epoch": 0.9554268785348775, "percentage": 47.77, "elapsed_time": "1 day, 6:29:22", "remaining_time": "1 day, 9:20:04", "throughput": 5653.89, "total_tokens": 620586608}
+{"current_steps": 14195, "total_steps": 29704, "loss": 0.3777376413345337, "lr": 5.347680412295152e-06, "epoch": 0.9557635335308377, "percentage": 47.79, "elapsed_time": "1 day, 6:30:03", "remaining_time": "1 day, 9:19:27", "throughput": 5653.79, "total_tokens": 620803016}
+{"current_steps": 14200, "total_steps": 29704, "loss": 0.37516131401062014, "lr": 5.34504268181626e-06, "epoch": 0.9561001885267978, "percentage": 47.81, "elapsed_time": "1 day, 6:30:42", "remaining_time": "1 day, 9:18:49", "throughput": 5653.65, "total_tokens": 621009712}
+{"current_steps": 14205, "total_steps": 29704, "loss": 0.36751456260681153, "lr": 5.342404854847282e-06, "epoch": 0.9564368435227579, "percentage": 47.82, "elapsed_time": "1 day, 6:31:18", "remaining_time": "1 day, 9:18:08", "throughput": 5653.6, "total_tokens": 621209912}
+{"current_steps": 14210, "total_steps": 29704, "loss": 0.3533796310424805, "lr": 5.3397669321258776e-06, "epoch": 0.9567734985187181, "percentage": 47.84, "elapsed_time": "1 day, 6:31:53", "remaining_time": "1 day, 9:17:25", "throughput": 5653.89, "total_tokens": 621439136}
+{"current_steps": 14215, "total_steps": 29704, "loss": 0.3940083980560303, "lr": 5.337128914389734e-06, "epoch": 0.9571101535146782, "percentage": 47.86, "elapsed_time": "1 day, 6:32:28", "remaining_time": "1 day, 9:16:42", "throughput": 5654.19, "total_tokens": 621670256}
+{"current_steps": 14220, "total_steps": 29704, "loss": 0.3630239009857178, "lr": 5.334490802376565e-06, "epoch": 0.9574468085106383, "percentage": 47.87, "elapsed_time": "1 day, 6:33:12", "remaining_time": "1 day, 9:16:09", "throughput": 5653.99, "total_tokens": 621896344}
+{"current_steps": 14225, "total_steps": 29704, "loss": 0.38904123306274413, "lr": 5.33185259682411e-06, "epoch": 0.9577834635065985, "percentage": 47.89, "elapsed_time": "1 day, 6:33:48", "remaining_time": "1 day, 9:15:27", "throughput": 5654.1, "total_tokens": 622111488}
+{"current_steps": 14230, "total_steps": 29704, "loss": 0.3952740907669067, "lr": 5.329214298470134e-06, "epoch": 0.9581201185025586, "percentage": 47.91, "elapsed_time": "1 day, 6:34:27", "remaining_time": "1 day, 9:14:49", "throughput": 5654.14, "total_tokens": 622335752}
+{"current_steps": 14235, "total_steps": 29704, "loss": 0.34938879013061525, "lr": 5.326575908052428e-06, "epoch": 0.9584567734985188, "percentage": 47.92, "elapsed_time": "1 day, 6:35:03", "remaining_time": "1 day, 9:14:07", "throughput": 5654.37, "total_tokens": 622563512}
+{"current_steps": 14240, "total_steps": 29704, "loss": 0.4064329147338867, "lr": 5.323937426308813e-06, "epoch": 0.9587934284944789, "percentage": 47.94, "elapsed_time": "1 day, 6:35:40", "remaining_time": "1 day, 9:13:27", "throughput": 5654.51, "total_tokens": 622789680}
+{"current_steps": 14245, "total_steps": 29704, "loss": 0.3482068538665771, "lr": 5.321298853977128e-06, "epoch": 0.959130083490439, "percentage": 47.96, "elapsed_time": "1 day, 6:36:12", "remaining_time": "1 day, 9:12:41", "throughput": 5655.03, "total_tokens": 623030288}
+{"current_steps": 14250, "total_steps": 29704, "loss": 0.35120339393615724, "lr": 5.318660191795244e-06, "epoch": 0.9594667384863992, "percentage": 47.97, "elapsed_time": "1 day, 6:36:50", "remaining_time": "1 day, 9:12:02", "throughput": 5655.1, "total_tokens": 623250248}
+{"current_steps": 14255, "total_steps": 29704, "loss": 0.32753448486328124, "lr": 5.316021440501053e-06, "epoch": 0.9598033934823593, "percentage": 47.99, "elapsed_time": "1 day, 6:37:34", "remaining_time": "1 day, 9:11:29", "throughput": 5654.8, "total_tokens": 623468032}
+{"current_steps": 14260, "total_steps": 29704, "loss": 0.32216849327087405, "lr": 5.313382600832474e-06, "epoch": 0.9601400484783195, "percentage": 48.01, "elapsed_time": "1 day, 6:38:15", "remaining_time": "1 day, 9:10:52", "throughput": 5654.72, "total_tokens": 623687536}
+{"current_steps": 14265, "total_steps": 29704, "loss": 0.3820059776306152, "lr": 5.310743673527451e-06, "epoch": 0.9604767034742796, "percentage": 48.02, "elapsed_time": "1 day, 6:38:57", "remaining_time": "1 day, 9:10:18", "throughput": 5654.62, "total_tokens": 623916888}
+{"current_steps": 14270, "total_steps": 29704, "loss": 0.38223018646240237, "lr": 5.3081046593239496e-06, "epoch": 0.9608133584702397, "percentage": 48.04, "elapsed_time": "1 day, 6:39:38", "remaining_time": "1 day, 9:09:42", "throughput": 5654.5, "total_tokens": 624136416}
+{"current_steps": 14275, "total_steps": 29704, "loss": 0.366811728477478, "lr": 5.305465558959964e-06, "epoch": 0.9611500134661999, "percentage": 48.06, "elapsed_time": "1 day, 6:40:21", "remaining_time": "1 day, 9:09:07", "throughput": 5654.24, "total_tokens": 624349152}
+{"current_steps": 14280, "total_steps": 29704, "loss": 0.3593759536743164, "lr": 5.302826373173506e-06, "epoch": 0.96148666846216, "percentage": 48.07, "elapsed_time": "1 day, 6:41:00", "remaining_time": "1 day, 9:08:29", "throughput": 5654.32, "total_tokens": 624577352}
+{"current_steps": 14285, "total_steps": 29704, "loss": 0.3493657112121582, "lr": 5.3001871027026194e-06, "epoch": 0.9618233234581202, "percentage": 48.09, "elapsed_time": "1 day, 6:41:42", "remaining_time": "1 day, 9:07:54", "throughput": 5654.09, "total_tokens": 624788664}
+{"current_steps": 14290, "total_steps": 29704, "loss": 0.3475620746612549, "lr": 5.2975477482853685e-06, "epoch": 0.9621599784540803, "percentage": 48.11, "elapsed_time": "1 day, 6:42:23", "remaining_time": "1 day, 9:07:18", "throughput": 5654.02, "total_tokens": 625013976}
+{"current_steps": 14295, "total_steps": 29704, "loss": 0.3636843204498291, "lr": 5.294908310659838e-06, "epoch": 0.9624966334500404, "percentage": 48.12, "elapsed_time": "1 day, 6:42:59", "remaining_time": "1 day, 9:06:37", "throughput": 5654.22, "total_tokens": 625243416}
+{"current_steps": 14300, "total_steps": 29704, "loss": 0.3458188772201538, "lr": 5.292268790564138e-06, "epoch": 0.9628332884460006, "percentage": 48.14, "elapsed_time": "1 day, 6:43:41", "remaining_time": "1 day, 9:06:02", "throughput": 5654.0, "total_tokens": 625456472}
+{"current_steps": 14305, "total_steps": 29704, "loss": 0.39026923179626466, "lr": 5.289629188736403e-06, "epoch": 0.9631699434419607, "percentage": 48.16, "elapsed_time": "1 day, 6:44:27", "remaining_time": "1 day, 9:05:31", "throughput": 5653.64, "total_tokens": 625676128}
+{"current_steps": 14310, "total_steps": 29704, "loss": 0.38088412284851075, "lr": 5.286989505914788e-06, "epoch": 0.9635065984379209, "percentage": 48.18, "elapsed_time": "1 day, 6:45:02", "remaining_time": "1 day, 9:04:48", "throughput": 5653.92, "total_tokens": 625904384}
+{"current_steps": 14315, "total_steps": 29704, "loss": 0.3758919954299927, "lr": 5.284349742837475e-06, "epoch": 0.963843253433881, "percentage": 48.19, "elapsed_time": "1 day, 6:45:44", "remaining_time": "1 day, 9:04:13", "throughput": 5653.87, "total_tokens": 626134784}
+{"current_steps": 14320, "total_steps": 29704, "loss": 0.35522916316986086, "lr": 5.281709900242662e-06, "epoch": 0.9641799084298411, "percentage": 48.21, "elapsed_time": "1 day, 6:46:29", "remaining_time": "1 day, 9:03:40", "throughput": 5653.6, "total_tokens": 626357296}
+{"current_steps": 14325, "total_steps": 29704, "loss": 0.3454322099685669, "lr": 5.279069978868574e-06, "epoch": 0.9645165634258013, "percentage": 48.23, "elapsed_time": "1 day, 6:47:03", "remaining_time": "1 day, 9:02:57", "throughput": 5653.84, "total_tokens": 626576680}
+{"current_steps": 14330, "total_steps": 29704, "loss": 0.36722531318664553, "lr": 5.276429979453455e-06, "epoch": 0.9648532184217614, "percentage": 48.24, "elapsed_time": "1 day, 6:47:35", "remaining_time": "1 day, 9:02:11", "throughput": 5654.18, "total_tokens": 626795800}
+{"current_steps": 14335, "total_steps": 29704, "loss": 0.36404809951782224, "lr": 5.273789902735575e-06, "epoch": 0.9651898734177216, "percentage": 48.26, "elapsed_time": "1 day, 6:48:10", "remaining_time": "1 day, 9:01:28", "throughput": 5654.35, "total_tokens": 627011600}
+{"current_steps": 14340, "total_steps": 29704, "loss": 0.33545889854431155, "lr": 5.27114974945322e-06, "epoch": 0.9655265284136817, "percentage": 48.28, "elapsed_time": "1 day, 6:48:57", "remaining_time": "1 day, 9:00:59", "throughput": 5653.88, "total_tokens": 627226888}
+{"current_steps": 14345, "total_steps": 29704, "loss": 0.3596933841705322, "lr": 5.2685095203447046e-06, "epoch": 0.9658631834096418, "percentage": 48.29, "elapsed_time": "1 day, 6:49:36", "remaining_time": "1 day, 9:00:21", "throughput": 5653.93, "total_tokens": 627453768}
+{"current_steps": 14350, "total_steps": 29704, "loss": 0.352811861038208, "lr": 5.265869216148357e-06, "epoch": 0.966199838405602, "percentage": 48.31, "elapsed_time": "1 day, 6:50:14", "remaining_time": "1 day, 8:59:42", "throughput": 5653.95, "total_tokens": 627673216}
+{"current_steps": 14355, "total_steps": 29704, "loss": 0.42299542427062986, "lr": 5.263228837602532e-06, "epoch": 0.9665364934015621, "percentage": 48.33, "elapsed_time": "1 day, 6:50:50", "remaining_time": "1 day, 8:59:00", "throughput": 5654.05, "total_tokens": 627887552}
+{"current_steps": 14360, "total_steps": 29704, "loss": 0.3737421989440918, "lr": 5.2605883854456e-06, "epoch": 0.9668731483975223, "percentage": 48.34, "elapsed_time": "1 day, 6:51:29", "remaining_time": "1 day, 8:58:21", "throughput": 5654.07, "total_tokens": 628108328}
+{"current_steps": 14365, "total_steps": 29704, "loss": 0.3357997894287109, "lr": 5.257947860415957e-06, "epoch": 0.9672098033934824, "percentage": 48.36, "elapsed_time": "1 day, 6:52:02", "remaining_time": "1 day, 8:57:37", "throughput": 5654.43, "total_tokens": 628335808}
+{"current_steps": 14370, "total_steps": 29704, "loss": 0.3541123867034912, "lr": 5.255307263252019e-06, "epoch": 0.9675464583894425, "percentage": 48.38, "elapsed_time": "1 day, 6:52:36", "remaining_time": "1 day, 8:56:53", "throughput": 5654.77, "total_tokens": 628566792}
+{"current_steps": 14375, "total_steps": 29704, "loss": 0.3823538303375244, "lr": 5.2526665946922185e-06, "epoch": 0.9678831133854027, "percentage": 48.39, "elapsed_time": "1 day, 6:53:16", "remaining_time": "1 day, 8:56:16", "throughput": 5654.71, "total_tokens": 628784336}
+{"current_steps": 14380, "total_steps": 29704, "loss": 0.35442066192626953, "lr": 5.25002585547501e-06, "epoch": 0.9682197683813628, "percentage": 48.41, "elapsed_time": "1 day, 6:53:54", "remaining_time": "1 day, 8:55:36", "throughput": 5654.81, "total_tokens": 629008320}
+{"current_steps": 14385, "total_steps": 29704, "loss": 0.3943564176559448, "lr": 5.2473850463388675e-06, "epoch": 0.968556423377323, "percentage": 48.43, "elapsed_time": "1 day, 6:54:34", "remaining_time": "1 day, 8:54:59", "throughput": 5654.76, "total_tokens": 629232256}
+{"current_steps": 14390, "total_steps": 29704, "loss": 0.3620326995849609, "lr": 5.244744168022288e-06, "epoch": 0.9688930783732831, "percentage": 48.44, "elapsed_time": "1 day, 6:55:14", "remaining_time": "1 day, 8:54:22", "throughput": 5654.73, "total_tokens": 629455456}
+{"current_steps": 14395, "total_steps": 29704, "loss": 0.36865873336791993, "lr": 5.24210322126378e-06, "epoch": 0.9692297333692432, "percentage": 48.46, "elapsed_time": "1 day, 6:55:53", "remaining_time": "1 day, 8:53:43", "throughput": 5654.9, "total_tokens": 629691848}
+{"current_steps": 14400, "total_steps": 29704, "loss": 0.3680101871490479, "lr": 5.23946220680188e-06, "epoch": 0.9695663883652034, "percentage": 48.48, "elapsed_time": "1 day, 6:56:31", "remaining_time": "1 day, 8:53:04", "throughput": 5655.09, "total_tokens": 629928368}
+{"current_steps": 14405, "total_steps": 29704, "loss": 0.3584733486175537, "lr": 5.236821125375136e-06, "epoch": 0.9699030433611635, "percentage": 48.5, "elapsed_time": "1 day, 6:57:09", "remaining_time": "1 day, 8:52:25", "throughput": 5655.14, "total_tokens": 630150304}
+{"current_steps": 14410, "total_steps": 29704, "loss": 0.31973896026611326, "lr": 5.234179977722119e-06, "epoch": 0.9702396983571236, "percentage": 48.51, "elapsed_time": "1 day, 6:57:45", "remaining_time": "1 day, 8:51:43", "throughput": 5655.48, "total_tokens": 630389640}
+{"current_steps": 14415, "total_steps": 29704, "loss": 0.36993358135223386, "lr": 5.23153876458142e-06, "epoch": 0.9705763533530838, "percentage": 48.53, "elapsed_time": "1 day, 6:58:29", "remaining_time": "1 day, 8:51:10", "throughput": 5655.13, "total_tokens": 630599464}
+{"current_steps": 14420, "total_steps": 29704, "loss": 0.36161553859710693, "lr": 5.2288974866916435e-06, "epoch": 0.9709130083490439, "percentage": 48.55, "elapsed_time": "1 day, 6:59:07", "remaining_time": "1 day, 8:50:31", "throughput": 5654.92, "total_tokens": 630794912}
+{"current_steps": 14425, "total_steps": 29704, "loss": 0.36026706695556643, "lr": 5.226256144791414e-06, "epoch": 0.9712496633450041, "percentage": 48.56, "elapsed_time": "1 day, 6:59:42", "remaining_time": "1 day, 8:49:48", "throughput": 5655.02, "total_tokens": 631000992}
+{"current_steps": 14430, "total_steps": 29704, "loss": 0.37513973712921145, "lr": 5.223614739619375e-06, "epoch": 0.9715863183409642, "percentage": 48.58, "elapsed_time": "1 day, 7:00:20", "remaining_time": "1 day, 8:49:09", "throughput": 5654.89, "total_tokens": 631202192}
+{"current_steps": 14435, "total_steps": 29704, "loss": 0.33454012870788574, "lr": 5.2209732719141855e-06, "epoch": 0.9719229733369243, "percentage": 48.6, "elapsed_time": "1 day, 7:01:00", "remaining_time": "1 day, 8:48:31", "throughput": 5654.75, "total_tokens": 631411712}
+{"current_steps": 14440, "total_steps": 29704, "loss": 0.3455739736557007, "lr": 5.218331742414528e-06, "epoch": 0.9722596283328845, "percentage": 48.61, "elapsed_time": "1 day, 7:01:34", "remaining_time": "1 day, 8:47:48", "throughput": 5655.08, "total_tokens": 631641904}
+{"current_steps": 14445, "total_steps": 29704, "loss": 0.3380829572677612, "lr": 5.215690151859094e-06, "epoch": 0.9725962833288446, "percentage": 48.63, "elapsed_time": "1 day, 7:02:13", "remaining_time": "1 day, 8:47:10", "throughput": 5654.94, "total_tokens": 631848056}
+{"current_steps": 14450, "total_steps": 29704, "loss": 0.3667490243911743, "lr": 5.213048500986597e-06, "epoch": 0.9729329383248048, "percentage": 48.65, "elapsed_time": "1 day, 7:03:02", "remaining_time": "1 day, 8:46:41", "throughput": 5654.47, "total_tokens": 632068664}
+{"current_steps": 14455, "total_steps": 29704, "loss": 0.3663896083831787, "lr": 5.210406790535766e-06, "epoch": 0.9732695933207649, "percentage": 48.66, "elapsed_time": "1 day, 7:03:37", "remaining_time": "1 day, 8:45:59", "throughput": 5654.58, "total_tokens": 632279544}
+{"current_steps": 14460, "total_steps": 29704, "loss": 0.3660223722457886, "lr": 5.207765021245346e-06, "epoch": 0.973606248316725, "percentage": 48.68, "elapsed_time": "1 day, 7:04:18", "remaining_time": "1 day, 8:45:23", "throughput": 5654.4, "total_tokens": 632492984}
+{"current_steps": 14465, "total_steps": 29704, "loss": 0.3708993911743164, "lr": 5.2051231938541015e-06, "epoch": 0.9739429033126852, "percentage": 48.7, "elapsed_time": "1 day, 7:04:54", "remaining_time": "1 day, 8:44:41", "throughput": 5654.41, "total_tokens": 632694568}
+{"current_steps": 14470, "total_steps": 29704, "loss": 0.3687734127044678, "lr": 5.202481309100808e-06, "epoch": 0.9742795583086453, "percentage": 48.71, "elapsed_time": "1 day, 7:05:35", "remaining_time": "1 day, 8:44:05", "throughput": 5654.2, "total_tokens": 632904360}
+{"current_steps": 14475, "total_steps": 29704, "loss": 0.3596921920776367, "lr": 5.199839367724266e-06, "epoch": 0.9746162133046055, "percentage": 48.73, "elapsed_time": "1 day, 7:06:16", "remaining_time": "1 day, 8:43:28", "throughput": 5654.18, "total_tokens": 633132800}
+{"current_steps": 14480, "total_steps": 29704, "loss": 0.343227481842041, "lr": 5.197197370463278e-06, "epoch": 0.9749528683005656, "percentage": 48.75, "elapsed_time": "1 day, 7:06:57", "remaining_time": "1 day, 8:42:53", "throughput": 5654.21, "total_tokens": 633370648}
+{"current_steps": 14485, "total_steps": 29704, "loss": 0.39154577255249023, "lr": 5.194555318056675e-06, "epoch": 0.9752895232965257, "percentage": 48.76, "elapsed_time": "1 day, 7:07:30", "remaining_time": "1 day, 8:42:08", "throughput": 5654.66, "total_tokens": 633605816}
+{"current_steps": 14490, "total_steps": 29704, "loss": 0.37716875076293943, "lr": 5.191913211243299e-06, "epoch": 0.9756261782924859, "percentage": 48.78, "elapsed_time": "1 day, 7:08:15", "remaining_time": "1 day, 8:41:36", "throughput": 5654.39, "total_tokens": 633832456}
+{"current_steps": 14495, "total_steps": 29704, "loss": 0.36735081672668457, "lr": 5.1892710507620055e-06, "epoch": 0.975962833288446, "percentage": 48.8, "elapsed_time": "1 day, 7:08:52", "remaining_time": "1 day, 8:40:55", "throughput": 5654.51, "total_tokens": 634052208}
+{"current_steps": 14500, "total_steps": 29704, "loss": 0.3544626235961914, "lr": 5.186628837351666e-06, "epoch": 0.9762994882844062, "percentage": 48.81, "elapsed_time": "1 day, 7:09:28", "remaining_time": "1 day, 8:40:14", "throughput": 5654.72, "total_tokens": 634282120}
+{"current_steps": 14505, "total_steps": 29704, "loss": 0.3681643962860107, "lr": 5.183986571751167e-06, "epoch": 0.9766361432803663, "percentage": 48.83, "elapsed_time": "1 day, 7:10:08", "remaining_time": "1 day, 8:39:37", "throughput": 5654.67, "total_tokens": 634503040}
+{"current_steps": 14510, "total_steps": 29704, "loss": 0.384121036529541, "lr": 5.181344254699408e-06, "epoch": 0.9769727982763264, "percentage": 48.85, "elapsed_time": "1 day, 7:10:46", "remaining_time": "1 day, 8:38:58", "throughput": 5654.66, "total_tokens": 634718224}
+{"current_steps": 14515, "total_steps": 29704, "loss": 0.3466785907745361, "lr": 5.178701886935309e-06, "epoch": 0.9773094532722866, "percentage": 48.87, "elapsed_time": "1 day, 7:11:23", "remaining_time": "1 day, 8:38:17", "throughput": 5654.69, "total_tokens": 634926976}
+{"current_steps": 14520, "total_steps": 29704, "loss": 0.3563254356384277, "lr": 5.176059469197796e-06, "epoch": 0.9776461082682467, "percentage": 48.88, "elapsed_time": "1 day, 7:11:58", "remaining_time": "1 day, 8:37:34", "throughput": 5654.88, "total_tokens": 635147104}
+{"current_steps": 14525, "total_steps": 29704, "loss": 0.3569037914276123, "lr": 5.173417002225815e-06, "epoch": 0.9779827632642069, "percentage": 48.9, "elapsed_time": "1 day, 7:12:37", "remaining_time": "1 day, 8:36:56", "throughput": 5654.86, "total_tokens": 635366792}
+{"current_steps": 14530, "total_steps": 29704, "loss": 0.36590094566345216, "lr": 5.170774486758322e-06, "epoch": 0.978319418260167, "percentage": 48.92, "elapsed_time": "1 day, 7:13:13", "remaining_time": "1 day, 8:36:15", "throughput": 5654.76, "total_tokens": 635560568}
+{"current_steps": 14535, "total_steps": 29704, "loss": 0.38362579345703124, "lr": 5.168131923534287e-06, "epoch": 0.9786560732561271, "percentage": 48.93, "elapsed_time": "1 day, 7:13:48", "remaining_time": "1 day, 8:35:32", "throughput": 5655.02, "total_tokens": 635787144}
+{"current_steps": 14540, "total_steps": 29704, "loss": 0.35508084297180176, "lr": 5.165489313292696e-06, "epoch": 0.9789927282520873, "percentage": 48.95, "elapsed_time": "1 day, 7:14:29", "remaining_time": "1 day, 8:34:56", "throughput": 5654.85, "total_tokens": 636000120}
+{"current_steps": 14545, "total_steps": 29704, "loss": 0.35072224140167235, "lr": 5.162846656772548e-06, "epoch": 0.9793293832480474, "percentage": 48.97, "elapsed_time": "1 day, 7:15:12", "remaining_time": "1 day, 8:34:22", "throughput": 5654.59, "total_tokens": 636214280}
+{"current_steps": 14550, "total_steps": 29704, "loss": 0.3438509464263916, "lr": 5.16020395471285e-06, "epoch": 0.9796660382440076, "percentage": 48.98, "elapsed_time": "1 day, 7:15:54", "remaining_time": "1 day, 8:33:46", "throughput": 5654.48, "total_tokens": 636435816}
+{"current_steps": 14555, "total_steps": 29704, "loss": 0.3831623077392578, "lr": 5.157561207852628e-06, "epoch": 0.9800026932399677, "percentage": 49.0, "elapsed_time": "1 day, 7:16:31", "remaining_time": "1 day, 8:33:06", "throughput": 5654.44, "total_tokens": 636640224}
+{"current_steps": 14560, "total_steps": 29704, "loss": 0.3411484479904175, "lr": 5.154918416930915e-06, "epoch": 0.9803393482359278, "percentage": 49.02, "elapsed_time": "1 day, 7:17:09", "remaining_time": "1 day, 8:32:27", "throughput": 5654.45, "total_tokens": 636858520}
+{"current_steps": 14565, "total_steps": 29704, "loss": 0.3839298486709595, "lr": 5.152275582686761e-06, "epoch": 0.980676003231888, "percentage": 49.03, "elapsed_time": "1 day, 7:17:46", "remaining_time": "1 day, 8:31:46", "throughput": 5654.52, "total_tokens": 637073856}
+{"current_steps": 14570, "total_steps": 29704, "loss": 0.36935174465179443, "lr": 5.149632705859227e-06, "epoch": 0.9810126582278481, "percentage": 49.05, "elapsed_time": "1 day, 7:18:27", "remaining_time": "1 day, 8:31:10", "throughput": 5654.43, "total_tokens": 637299440}
+{"current_steps": 14575, "total_steps": 29704, "loss": 0.37570605278015134, "lr": 5.146989787187383e-06, "epoch": 0.9813493132238083, "percentage": 49.07, "elapsed_time": "1 day, 7:19:03", "remaining_time": "1 day, 8:30:28", "throughput": 5654.69, "total_tokens": 637529440}
+{"current_steps": 14580, "total_steps": 29704, "loss": 0.3703890562057495, "lr": 5.144346827410314e-06, "epoch": 0.9816859682197684, "percentage": 49.08, "elapsed_time": "1 day, 7:19:33", "remaining_time": "1 day, 8:29:40", "throughput": 5655.2, "total_tokens": 637754408}
+{"current_steps": 14585, "total_steps": 29704, "loss": 0.3702056646347046, "lr": 5.141703827267117e-06, "epoch": 0.9820226232157285, "percentage": 49.1, "elapsed_time": "1 day, 7:20:04", "remaining_time": "1 day, 8:28:54", "throughput": 5655.64, "total_tokens": 637982336}
+{"current_steps": 14590, "total_steps": 29704, "loss": 0.3754692077636719, "lr": 5.139060787496893e-06, "epoch": 0.9823592782116887, "percentage": 49.12, "elapsed_time": "1 day, 7:20:41", "remaining_time": "1 day, 8:28:14", "throughput": 5655.74, "total_tokens": 638203872}
+{"current_steps": 14595, "total_steps": 29704, "loss": 0.3775701284408569, "lr": 5.136417708838766e-06, "epoch": 0.9826959332076488, "percentage": 49.13, "elapsed_time": "1 day, 7:21:25", "remaining_time": "1 day, 8:27:40", "throughput": 5655.38, "total_tokens": 638407968}
+{"current_steps": 14600, "total_steps": 29704, "loss": 0.3628222465515137, "lr": 5.13377459203186e-06, "epoch": 0.983032588203609, "percentage": 49.15, "elapsed_time": "1 day, 7:22:03", "remaining_time": "1 day, 8:27:01", "throughput": 5655.32, "total_tokens": 638616040}
+{"current_steps": 14605, "total_steps": 29704, "loss": 0.3371574401855469, "lr": 5.131131437815318e-06, "epoch": 0.9833692431995691, "percentage": 49.17, "elapsed_time": "1 day, 7:22:43", "remaining_time": "1 day, 8:26:24", "throughput": 5655.27, "total_tokens": 638837176}
+{"current_steps": 14610, "total_steps": 29704, "loss": 0.36636881828308104, "lr": 5.128488246928287e-06, "epoch": 0.9837058981955292, "percentage": 49.19, "elapsed_time": "1 day, 7:23:20", "remaining_time": "1 day, 8:25:43", "throughput": 5655.44, "total_tokens": 639065552}
+{"current_steps": 14615, "total_steps": 29704, "loss": 0.37039480209350584, "lr": 5.125845020109925e-06, "epoch": 0.9840425531914894, "percentage": 49.2, "elapsed_time": "1 day, 7:24:07", "remaining_time": "1 day, 8:25:13", "throughput": 5655.02, "total_tokens": 639285272}
+{"current_steps": 14620, "total_steps": 29704, "loss": 0.36545164585113527, "lr": 5.123201758099406e-06, "epoch": 0.9843792081874495, "percentage": 49.22, "elapsed_time": "1 day, 7:24:45", "remaining_time": "1 day, 8:24:34", "throughput": 5655.0, "total_tokens": 639499208}
+{"current_steps": 14625, "total_steps": 29704, "loss": 0.37010979652404785, "lr": 5.12055846163591e-06, "epoch": 0.9847158631834096, "percentage": 49.24, "elapsed_time": "1 day, 7:25:23", "remaining_time": "1 day, 8:23:55", "throughput": 5655.0, "total_tokens": 639714312}
+{"current_steps": 14630, "total_steps": 29704, "loss": 0.3291551828384399, "lr": 5.117915131458623e-06, "epoch": 0.9850525181793698, "percentage": 49.25, "elapsed_time": "1 day, 7:26:02", "remaining_time": "1 day, 8:23:16", "throughput": 5655.06, "total_tokens": 639938488}
+{"current_steps": 14635, "total_steps": 29704, "loss": 0.36716470718383787, "lr": 5.115271768306745e-06, "epoch": 0.9853891731753299, "percentage": 49.27, "elapsed_time": "1 day, 7:26:39", "remaining_time": "1 day, 8:22:36", "throughput": 5655.08, "total_tokens": 640152840}
+{"current_steps": 14640, "total_steps": 29704, "loss": 0.3815732717514038, "lr": 5.112628372919485e-06, "epoch": 0.9857258281712901, "percentage": 49.29, "elapsed_time": "1 day, 7:27:14", "remaining_time": "1 day, 8:21:53", "throughput": 5655.27, "total_tokens": 640369952}
+{"current_steps": 14645, "total_steps": 29704, "loss": 0.34613332748413084, "lr": 5.1099849460360595e-06, "epoch": 0.9860624831672502, "percentage": 49.3, "elapsed_time": "1 day, 7:27:51", "remaining_time": "1 day, 8:21:13", "throughput": 5655.14, "total_tokens": 640567240}
+{"current_steps": 14650, "total_steps": 29704, "loss": 0.3379997730255127, "lr": 5.107341488395695e-06, "epoch": 0.9863991381632103, "percentage": 49.32, "elapsed_time": "1 day, 7:28:30", "remaining_time": "1 day, 8:20:35", "throughput": 5655.16, "total_tokens": 640788384}
+{"current_steps": 14655, "total_steps": 29704, "loss": 0.37780165672302246, "lr": 5.104698000737626e-06, "epoch": 0.9867357931591705, "percentage": 49.34, "elapsed_time": "1 day, 7:29:11", "remaining_time": "1 day, 8:19:58", "throughput": 5655.08, "total_tokens": 641011792}
+{"current_steps": 14660, "total_steps": 29704, "loss": 0.38533506393432615, "lr": 5.102054483801094e-06, "epoch": 0.9870724481551306, "percentage": 49.35, "elapsed_time": "1 day, 7:29:50", "remaining_time": "1 day, 8:19:20", "throughput": 5655.15, "total_tokens": 641239984}
+{"current_steps": 14665, "total_steps": 29704, "loss": 0.3746023178100586, "lr": 5.099410938325351e-06, "epoch": 0.9874091031510908, "percentage": 49.37, "elapsed_time": "1 day, 7:30:24", "remaining_time": "1 day, 8:18:37", "throughput": 5655.45, "total_tokens": 641468200}
+{"current_steps": 14670, "total_steps": 29704, "loss": 0.3502329111099243, "lr": 5.096767365049657e-06, "epoch": 0.9877457581470509, "percentage": 49.39, "elapsed_time": "1 day, 7:31:04", "remaining_time": "1 day, 8:18:00", "throughput": 5655.25, "total_tokens": 641672792}
+{"current_steps": 14675, "total_steps": 29704, "loss": 0.3421436309814453, "lr": 5.094123764713277e-06, "epoch": 0.988082413143011, "percentage": 49.4, "elapsed_time": "1 day, 7:31:40", "remaining_time": "1 day, 8:17:18", "throughput": 5655.5, "total_tokens": 641904736}
+{"current_steps": 14680, "total_steps": 29704, "loss": 0.36416258811950686, "lr": 5.091480138055487e-06, "epoch": 0.9884190681389712, "percentage": 49.42, "elapsed_time": "1 day, 7:32:11", "remaining_time": "1 day, 8:16:32", "throughput": 5656.0, "total_tokens": 642136592}
+{"current_steps": 14685, "total_steps": 29704, "loss": 0.35428576469421386, "lr": 5.088836485815568e-06, "epoch": 0.9887557231349313, "percentage": 49.44, "elapsed_time": "1 day, 7:32:46", "remaining_time": "1 day, 8:15:49", "throughput": 5656.0, "total_tokens": 642334648}
+{"current_steps": 14690, "total_steps": 29704, "loss": 0.35656602382659913, "lr": 5.086192808732809e-06, "epoch": 0.9890923781308915, "percentage": 49.45, "elapsed_time": "1 day, 7:33:28", "remaining_time": "1 day, 8:15:14", "throughput": 5655.72, "total_tokens": 642538480}
+{"current_steps": 14695, "total_steps": 29704, "loss": 0.36766040325164795, "lr": 5.083549107546505e-06, "epoch": 0.9894290331268516, "percentage": 49.47, "elapsed_time": "1 day, 7:34:07", "remaining_time": "1 day, 8:14:36", "throughput": 5655.59, "total_tokens": 642744792}
+{"current_steps": 14700, "total_steps": 29704, "loss": 0.38895292282104493, "lr": 5.080905382995961e-06, "epoch": 0.9897656881228117, "percentage": 49.49, "elapsed_time": "1 day, 7:34:43", "remaining_time": "1 day, 8:13:54", "throughput": 5655.81, "total_tokens": 642974608}
+{"current_steps": 14705, "total_steps": 29704, "loss": 0.37378993034362795, "lr": 5.0782616358204835e-06, "epoch": 0.9901023431187719, "percentage": 49.51, "elapsed_time": "1 day, 7:35:23", "remaining_time": "1 day, 8:13:17", "throughput": 5655.75, "total_tokens": 643191928}
+{"current_steps": 14710, "total_steps": 29704, "loss": 0.33492867946624755, "lr": 5.075617866759389e-06, "epoch": 0.990438998114732, "percentage": 49.52, "elapsed_time": "1 day, 7:36:06", "remaining_time": "1 day, 8:12:42", "throughput": 5655.49, "total_tokens": 643403376}
+{"current_steps": 14715, "total_steps": 29704, "loss": 0.4053798675537109, "lr": 5.0729740765519995e-06, "epoch": 0.9907756531106922, "percentage": 49.54, "elapsed_time": "1 day, 7:36:42", "remaining_time": "1 day, 8:12:01", "throughput": 5655.56, "total_tokens": 643617760}
+{"current_steps": 14720, "total_steps": 29704, "loss": 0.3686873435974121, "lr": 5.070330265937643e-06, "epoch": 0.9911123081066523, "percentage": 49.56, "elapsed_time": "1 day, 7:37:18", "remaining_time": "1 day, 8:11:20", "throughput": 5655.62, "total_tokens": 643827376}
+{"current_steps": 14725, "total_steps": 29704, "loss": 0.3935819625854492, "lr": 5.067686435655651e-06, "epoch": 0.9914489631026124, "percentage": 49.57, "elapsed_time": "1 day, 7:37:55", "remaining_time": "1 day, 8:10:39", "throughput": 5655.78, "total_tokens": 644051992}
+{"current_steps": 14730, "total_steps": 29704, "loss": 0.39838461875915526, "lr": 5.065042586445363e-06, "epoch": 0.9917856180985726, "percentage": 49.59, "elapsed_time": "1 day, 7:38:27", "remaining_time": "1 day, 8:09:54", "throughput": 5655.96, "total_tokens": 644255512}
+{"current_steps": 14735, "total_steps": 29704, "loss": 0.30464193820953367, "lr": 5.0623987190461245e-06, "epoch": 0.9921222730945327, "percentage": 49.61, "elapsed_time": "1 day, 7:39:04", "remaining_time": "1 day, 8:09:14", "throughput": 5655.9, "total_tokens": 644460824}
+{"current_steps": 14740, "total_steps": 29704, "loss": 0.3790989875793457, "lr": 5.059754834197281e-06, "epoch": 0.9924589280904929, "percentage": 49.62, "elapsed_time": "1 day, 7:39:48", "remaining_time": "1 day, 8:08:40", "throughput": 5655.61, "total_tokens": 644675152}
+{"current_steps": 14745, "total_steps": 29704, "loss": 0.32619643211364746, "lr": 5.05711093263819e-06, "epoch": 0.992795583086453, "percentage": 49.64, "elapsed_time": "1 day, 7:40:26", "remaining_time": "1 day, 8:08:01", "throughput": 5655.63, "total_tokens": 644890856}
+{"current_steps": 14750, "total_steps": 29704, "loss": 0.37743263244628905, "lr": 5.054467015108208e-06, "epoch": 0.9931322380824131, "percentage": 49.66, "elapsed_time": "1 day, 7:41:01", "remaining_time": "1 day, 8:07:19", "throughput": 5655.91, "total_tokens": 645124328}
+{"current_steps": 14755, "total_steps": 29704, "loss": 0.37684605121612547, "lr": 5.051823082346701e-06, "epoch": 0.9934688930783733, "percentage": 49.67, "elapsed_time": "1 day, 7:41:35", "remaining_time": "1 day, 8:06:35", "throughput": 5656.18, "total_tokens": 645342896}
+{"current_steps": 14760, "total_steps": 29704, "loss": 0.3667085409164429, "lr": 5.049179135093033e-06, "epoch": 0.9938055480743334, "percentage": 49.69, "elapsed_time": "1 day, 7:42:08", "remaining_time": "1 day, 8:05:51", "throughput": 5656.52, "total_tokens": 645572616}
+{"current_steps": 14765, "total_steps": 29704, "loss": 0.37410502433776854, "lr": 5.046535174086575e-06, "epoch": 0.9941422030702936, "percentage": 49.71, "elapsed_time": "1 day, 7:42:51", "remaining_time": "1 day, 8:05:16", "throughput": 5656.27, "total_tokens": 645784112}
+{"current_steps": 14770, "total_steps": 29704, "loss": 0.35855536460876464, "lr": 5.043891200066706e-06, "epoch": 0.9944788580662537, "percentage": 49.72, "elapsed_time": "1 day, 7:43:31", "remaining_time": "1 day, 8:04:39", "throughput": 5656.17, "total_tokens": 646000480}
+{"current_steps": 14775, "total_steps": 29704, "loss": 0.3647798538208008, "lr": 5.041247213772802e-06, "epoch": 0.9948155130622138, "percentage": 49.74, "elapsed_time": "1 day, 7:44:08", "remaining_time": "1 day, 8:03:59", "throughput": 5656.33, "total_tokens": 646228928}
+{"current_steps": 14780, "total_steps": 29704, "loss": 0.3697342872619629, "lr": 5.038603215944247e-06, "epoch": 0.995152168058174, "percentage": 49.76, "elapsed_time": "1 day, 7:44:43", "remaining_time": "1 day, 8:03:17", "throughput": 5656.33, "total_tokens": 646427096}
+{"current_steps": 14785, "total_steps": 29704, "loss": 0.34227933883666994, "lr": 5.0359592073204254e-06, "epoch": 0.9954888230541341, "percentage": 49.77, "elapsed_time": "1 day, 7:45:27", "remaining_time": "1 day, 8:02:43", "throughput": 5656.07, "total_tokens": 646643272}
+{"current_steps": 14790, "total_steps": 29704, "loss": 0.3627659797668457, "lr": 5.033315188640725e-06, "epoch": 0.9958254780500942, "percentage": 49.79, "elapsed_time": "1 day, 7:46:06", "remaining_time": "1 day, 8:02:05", "throughput": 5656.12, "total_tokens": 646872200}
+{"current_steps": 14795, "total_steps": 29704, "loss": 0.3450484752655029, "lr": 5.03067116064454e-06, "epoch": 0.9961621330460544, "percentage": 49.81, "elapsed_time": "1 day, 7:46:48", "remaining_time": "1 day, 8:01:30", "throughput": 5655.98, "total_tokens": 647094832}
+{"current_steps": 14800, "total_steps": 29704, "loss": 0.36185712814331056, "lr": 5.028027124071262e-06, "epoch": 0.9964987880420145, "percentage": 49.82, "elapsed_time": "1 day, 7:47:29", "remaining_time": "1 day, 8:00:54", "throughput": 5655.76, "total_tokens": 647301568}
+{"current_steps": 14805, "total_steps": 29704, "loss": 0.3468491554260254, "lr": 5.025383079660289e-06, "epoch": 0.9968354430379747, "percentage": 49.84, "elapsed_time": "1 day, 7:48:08", "remaining_time": "1 day, 8:00:15", "throughput": 5655.86, "total_tokens": 647531264}
+{"current_steps": 14810, "total_steps": 29704, "loss": 0.3663314342498779, "lr": 5.0227390281510166e-06, "epoch": 0.9971720980339348, "percentage": 49.86, "elapsed_time": "1 day, 7:48:46", "remaining_time": "1 day, 7:59:35", "throughput": 5655.95, "total_tokens": 647753976}
+{"current_steps": 14815, "total_steps": 29704, "loss": 0.3663675308227539, "lr": 5.020094970282849e-06, "epoch": 0.997508753029895, "percentage": 49.88, "elapsed_time": "1 day, 7:49:26", "remaining_time": "1 day, 7:58:58", "throughput": 5655.71, "total_tokens": 647955768}
+{"current_steps": 14820, "total_steps": 29704, "loss": 0.36784412860870364, "lr": 5.017450906795187e-06, "epoch": 0.9978454080258551, "percentage": 49.89, "elapsed_time": "1 day, 7:50:01", "remaining_time": "1 day, 7:58:16", "throughput": 5655.96, "total_tokens": 648181384}
+{"current_steps": 14825, "total_steps": 29704, "loss": 0.36796035766601565, "lr": 5.014806838427434e-06, "epoch": 0.9981820630218152, "percentage": 49.91, "elapsed_time": "1 day, 7:50:46", "remaining_time": "1 day, 7:57:44", "throughput": 5655.47, "total_tokens": 648382672}
+{"current_steps": 14830, "total_steps": 29704, "loss": 0.3591431140899658, "lr": 5.0121627659189964e-06, "epoch": 0.9985187180177754, "percentage": 49.93, "elapsed_time": "1 day, 7:51:22", "remaining_time": "1 day, 7:57:02", "throughput": 5655.73, "total_tokens": 648612168}
+{"current_steps": 14835, "total_steps": 29704, "loss": 0.3566739082336426, "lr": 5.009518690009281e-06, "epoch": 0.9988553730137355, "percentage": 49.94, "elapsed_time": "1 day, 7:52:04", "remaining_time": "1 day, 7:56:27", "throughput": 5655.48, "total_tokens": 648821208}
+{"current_steps": 14840, "total_steps": 29704, "loss": 0.3776280879974365, "lr": 5.006874611437692e-06, "epoch": 0.9991920280096956, "percentage": 49.96, "elapsed_time": "1 day, 7:52:44", "remaining_time": "1 day, 7:55:50", "throughput": 5655.45, "total_tokens": 649046288}
+{"current_steps": 14845, "total_steps": 29704, "loss": 0.35755281448364257, "lr": 5.004230530943641e-06, "epoch": 0.9995286830056558, "percentage": 49.98, "elapsed_time": "1 day, 7:53:24", "remaining_time": "1 day, 7:55:12", "throughput": 5655.37, "total_tokens": 649262232}
+{"current_steps": 14850, "total_steps": 29704, "loss": 0.3260646343231201, "lr": 5.001586449266536e-06, "epoch": 0.9998653380016159, "percentage": 49.99, "elapsed_time": "1 day, 7:54:05", "remaining_time": "1 day, 7:54:36", "throughput": 5655.31, "total_tokens": 649487024}
+{"current_steps": 14855, "total_steps": 29704, "loss": 0.35132179260253904, "lr": 4.998942367145787e-06, "epoch": 1.000201992997576, "percentage": 50.01, "elapsed_time": "1 day, 7:55:29", "remaining_time": "1 day, 7:54:42", "throughput": 5653.1, "total_tokens": 649706888}
+{"current_steps": 14860, "total_steps": 29704, "loss": 0.30704555511474607, "lr": 4.996298285320798e-06, "epoch": 1.0005386479935363, "percentage": 50.03, "elapsed_time": "1 day, 7:56:05", "remaining_time": "1 day, 7:54:01", "throughput": 5653.11, "total_tokens": 649911248}
+{"current_steps": 14865, "total_steps": 29704, "loss": 0.2908186435699463, "lr": 4.993654204530983e-06, "epoch": 1.0008753029894963, "percentage": 50.04, "elapsed_time": "1 day, 7:56:49", "remaining_time": "1 day, 7:53:27", "throughput": 5652.91, "total_tokens": 650136336}
+{"current_steps": 14870, "total_steps": 29704, "loss": 0.3061246633529663, "lr": 4.99101012551575e-06, "epoch": 1.0012119579854566, "percentage": 50.06, "elapsed_time": "1 day, 7:57:27", "remaining_time": "1 day, 7:52:49", "throughput": 5652.65, "total_tokens": 650324944}
+{"current_steps": 14875, "total_steps": 29704, "loss": 0.31706483364105226, "lr": 4.9883660490145075e-06, "epoch": 1.0015486129814166, "percentage": 50.08, "elapsed_time": "1 day, 7:58:03", "remaining_time": "1 day, 7:52:07", "throughput": 5652.92, "total_tokens": 650556192}
+{"current_steps": 14880, "total_steps": 29704, "loss": 0.32414817810058594, "lr": 4.985721975766661e-06, "epoch": 1.0018852679773769, "percentage": 50.09, "elapsed_time": "1 day, 7:58:38", "remaining_time": "1 day, 7:51:25", "throughput": 5653.1, "total_tokens": 650775960}
+{"current_steps": 14885, "total_steps": 29704, "loss": 0.32347331047058103, "lr": 4.983077906511619e-06, "epoch": 1.002221922973337, "percentage": 50.11, "elapsed_time": "1 day, 7:59:21", "remaining_time": "1 day, 7:50:50", "throughput": 5652.79, "total_tokens": 650983240}
+{"current_steps": 14890, "total_steps": 29704, "loss": 0.30762290954589844, "lr": 4.980433841988786e-06, "epoch": 1.0025585779692971, "percentage": 50.13, "elapsed_time": "1 day, 7:59:57", "remaining_time": "1 day, 7:50:09", "throughput": 5652.93, "total_tokens": 651204568}
+{"current_steps": 14895, "total_steps": 29704, "loss": 0.3130316257476807, "lr": 4.977789782937568e-06, "epoch": 1.0028952329652572, "percentage": 50.14, "elapsed_time": "1 day, 8:00:31", "remaining_time": "1 day, 7:49:26", "throughput": 5653.21, "total_tokens": 651431064}
+{"current_steps": 14900, "total_steps": 29704, "loss": 0.3236842155456543, "lr": 4.975145730097364e-06, "epoch": 1.0032318879612174, "percentage": 50.16, "elapsed_time": "1 day, 8:01:09", "remaining_time": "1 day, 7:48:46", "throughput": 5653.18, "total_tokens": 651636984}
+{"current_steps": 14905, "total_steps": 29704, "loss": 0.3103510618209839, "lr": 4.97250168420758e-06, "epoch": 1.0035685429571775, "percentage": 50.18, "elapsed_time": "1 day, 8:01:44", "remaining_time": "1 day, 7:48:04", "throughput": 5653.31, "total_tokens": 651849816}
+{"current_steps": 14910, "total_steps": 29704, "loss": 0.3144733428955078, "lr": 4.96985764600761e-06, "epoch": 1.0039051979531377, "percentage": 50.2, "elapsed_time": "1 day, 8:02:23", "remaining_time": "1 day, 7:47:26", "throughput": 5653.35, "total_tokens": 652078544}
+{"current_steps": 14915, "total_steps": 29704, "loss": 0.32990431785583496, "lr": 4.967213616236855e-06, "epoch": 1.0042418529490977, "percentage": 50.21, "elapsed_time": "1 day, 8:02:59", "remaining_time": "1 day, 7:46:45", "throughput": 5653.41, "total_tokens": 652288816}
+{"current_steps": 14920, "total_steps": 29704, "loss": 0.3154884338378906, "lr": 4.964569595634706e-06, "epoch": 1.004578507945058, "percentage": 50.23, "elapsed_time": "1 day, 8:03:42", "remaining_time": "1 day, 7:46:10", "throughput": 5653.23, "total_tokens": 652509832}
+{"current_steps": 14925, "total_steps": 29704, "loss": 0.3344886779785156, "lr": 4.961925584940559e-06, "epoch": 1.004915162941018, "percentage": 50.25, "elapsed_time": "1 day, 8:04:29", "remaining_time": "1 day, 7:45:40", "throughput": 5652.88, "total_tokens": 652737176}
+{"current_steps": 14930, "total_steps": 29704, "loss": 0.28301913738250734, "lr": 4.959281584893799e-06, "epoch": 1.0052518179369783, "percentage": 50.26, "elapsed_time": "1 day, 8:05:07", "remaining_time": "1 day, 7:45:01", "throughput": 5652.85, "total_tokens": 652949264}
+{"current_steps": 14935, "total_steps": 29704, "loss": 0.3222982883453369, "lr": 4.956637596233814e-06, "epoch": 1.0055884729329383, "percentage": 50.28, "elapsed_time": "1 day, 8:05:42", "remaining_time": "1 day, 7:44:18", "throughput": 5652.89, "total_tokens": 653149432}
+{"current_steps": 14940, "total_steps": 29704, "loss": 0.29294121265411377, "lr": 4.953993619699986e-06, "epoch": 1.0059251279288985, "percentage": 50.3, "elapsed_time": "1 day, 8:06:27", "remaining_time": "1 day, 7:43:45", "throughput": 5652.63, "total_tokens": 653371256}
+{"current_steps": 14945, "total_steps": 29704, "loss": 0.30971436500549315, "lr": 4.951349656031698e-06, "epoch": 1.0062617829248586, "percentage": 50.31, "elapsed_time": "1 day, 8:07:08", "remaining_time": "1 day, 7:43:09", "throughput": 5652.44, "total_tokens": 653582728}
+{"current_steps": 14950, "total_steps": 29704, "loss": 0.3157053470611572, "lr": 4.948705705968322e-06, "epoch": 1.0065984379208188, "percentage": 50.33, "elapsed_time": "1 day, 8:07:47", "remaining_time": "1 day, 7:42:30", "throughput": 5652.49, "total_tokens": 653808728}
+{"current_steps": 14955, "total_steps": 29704, "loss": 0.31903715133666993, "lr": 4.946061770249233e-06, "epoch": 1.0069350929167789, "percentage": 50.35, "elapsed_time": "1 day, 8:08:29", "remaining_time": "1 day, 7:41:55", "throughput": 5652.34, "total_tokens": 654028536}
+{"current_steps": 14960, "total_steps": 29704, "loss": 0.3309792518615723, "lr": 4.9434178496137955e-06, "epoch": 1.007271747912739, "percentage": 50.36, "elapsed_time": "1 day, 8:09:06", "remaining_time": "1 day, 7:41:15", "throughput": 5652.39, "total_tokens": 654245296}
+{"current_steps": 14965, "total_steps": 29704, "loss": 0.31335494518280027, "lr": 4.940773944801378e-06, "epoch": 1.0076084029086991, "percentage": 50.38, "elapsed_time": "1 day, 8:09:39", "remaining_time": "1 day, 7:40:31", "throughput": 5652.66, "total_tokens": 654464232}
+{"current_steps": 14970, "total_steps": 29704, "loss": 0.3319870948791504, "lr": 4.938130056551337e-06, "epoch": 1.0079450579046594, "percentage": 50.4, "elapsed_time": "1 day, 8:10:19", "remaining_time": "1 day, 7:39:53", "throughput": 5652.55, "total_tokens": 654675824}
+{"current_steps": 14975, "total_steps": 29704, "loss": 0.3344868183135986, "lr": 4.9354861856030305e-06, "epoch": 1.0082817129006194, "percentage": 50.41, "elapsed_time": "1 day, 8:10:59", "remaining_time": "1 day, 7:39:16", "throughput": 5652.49, "total_tokens": 654897216}
+{"current_steps": 14980, "total_steps": 29704, "loss": 0.28916707038879397, "lr": 4.932842332695803e-06, "epoch": 1.0086183678965797, "percentage": 50.43, "elapsed_time": "1 day, 8:11:34", "remaining_time": "1 day, 7:38:33", "throughput": 5652.57, "total_tokens": 655099488}
+{"current_steps": 14985, "total_steps": 29704, "loss": 0.31049747467041017, "lr": 4.9301984985690055e-06, "epoch": 1.0089550228925397, "percentage": 50.45, "elapsed_time": "1 day, 8:12:15", "remaining_time": "1 day, 7:37:57", "throughput": 5652.56, "total_tokens": 655334416}
+{"current_steps": 14990, "total_steps": 29704, "loss": 0.3354443073272705, "lr": 4.927554683961974e-06, "epoch": 1.0092916778885, "percentage": 50.46, "elapsed_time": "1 day, 8:12:50", "remaining_time": "1 day, 7:37:15", "throughput": 5652.65, "total_tokens": 655543112}
+{"current_steps": 14995, "total_steps": 29704, "loss": 0.3087354898452759, "lr": 4.924910889614046e-06, "epoch": 1.00962833288446, "percentage": 50.48, "elapsed_time": "1 day, 8:13:29", "remaining_time": "1 day, 7:36:36", "throughput": 5652.67, "total_tokens": 655763808}
+{"current_steps": 15000, "total_steps": 29704, "loss": 0.33311076164245607, "lr": 4.922267116264546e-06, "epoch": 1.0099649878804202, "percentage": 50.5, "elapsed_time": "1 day, 8:14:06", "remaining_time": "1 day, 7:35:56", "throughput": 5652.66, "total_tokens": 655973824}
+{"current_steps": 15005, "total_steps": 29704, "loss": 0.3059119462966919, "lr": 4.919623364652801e-06, "epoch": 1.0103016428763802, "percentage": 50.52, "elapsed_time": "1 day, 8:14:47", "remaining_time": "1 day, 7:35:20", "throughput": 5652.52, "total_tokens": 656189544}
+{"current_steps": 15010, "total_steps": 29704, "loss": 0.31961765289306643, "lr": 4.916979635518124e-06, "epoch": 1.0106382978723405, "percentage": 50.53, "elapsed_time": "1 day, 8:15:27", "remaining_time": "1 day, 7:34:43", "throughput": 5652.34, "total_tokens": 656393600}
+{"current_steps": 15015, "total_steps": 29704, "loss": 0.3200997829437256, "lr": 4.914335929599828e-06, "epoch": 1.0109749528683005, "percentage": 50.55, "elapsed_time": "1 day, 8:16:05", "remaining_time": "1 day, 7:34:03", "throughput": 5652.45, "total_tokens": 656618032}
+{"current_steps": 15020, "total_steps": 29704, "loss": 0.32773504257202146, "lr": 4.9116922476372155e-06, "epoch": 1.0113116078642608, "percentage": 50.57, "elapsed_time": "1 day, 8:16:47", "remaining_time": "1 day, 7:33:28", "throughput": 5652.29, "total_tokens": 656838496}
+{"current_steps": 15025, "total_steps": 29704, "loss": 0.31574647426605223, "lr": 4.909048590369586e-06, "epoch": 1.0116482628602208, "percentage": 50.58, "elapsed_time": "1 day, 8:17:20", "remaining_time": "1 day, 7:32:43", "throughput": 5652.56, "total_tokens": 657057672}
+{"current_steps": 15030, "total_steps": 29704, "loss": 0.3148768901824951, "lr": 4.906404958536227e-06, "epoch": 1.011984917856181, "percentage": 50.6, "elapsed_time": "1 day, 8:18:06", "remaining_time": "1 day, 7:32:11", "throughput": 5652.29, "total_tokens": 657282040}
+{"current_steps": 15035, "total_steps": 29704, "loss": 0.31647629737854005, "lr": 4.903761352876422e-06, "epoch": 1.012321572852141, "percentage": 50.62, "elapsed_time": "1 day, 8:18:49", "remaining_time": "1 day, 7:31:37", "throughput": 5651.85, "total_tokens": 657477448}
+{"current_steps": 15040, "total_steps": 29704, "loss": 0.3354557275772095, "lr": 4.9011177741294485e-06, "epoch": 1.0126582278481013, "percentage": 50.63, "elapsed_time": "1 day, 8:19:28", "remaining_time": "1 day, 7:30:59", "throughput": 5651.76, "total_tokens": 657686648}
+{"current_steps": 15045, "total_steps": 29704, "loss": 0.3180086612701416, "lr": 4.898474223034576e-06, "epoch": 1.0129948828440614, "percentage": 50.65, "elapsed_time": "1 day, 8:20:10", "remaining_time": "1 day, 7:30:23", "throughput": 5651.7, "total_tokens": 657917808}
+{"current_steps": 15050, "total_steps": 29704, "loss": 0.31407153606414795, "lr": 4.895830700331061e-06, "epoch": 1.0133315378400216, "percentage": 50.67, "elapsed_time": "1 day, 8:20:51", "remaining_time": "1 day, 7:29:47", "throughput": 5651.48, "total_tokens": 658124800}
+{"current_steps": 15055, "total_steps": 29704, "loss": 0.3087860822677612, "lr": 4.893187206758158e-06, "epoch": 1.0136681928359816, "percentage": 50.68, "elapsed_time": "1 day, 8:21:24", "remaining_time": "1 day, 7:29:03", "throughput": 5651.83, "total_tokens": 658352392}
+{"current_steps": 15060, "total_steps": 29704, "loss": 0.3047621250152588, "lr": 4.890543743055115e-06, "epoch": 1.014004847831942, "percentage": 50.7, "elapsed_time": "1 day, 8:22:03", "remaining_time": "1 day, 7:28:25", "throughput": 5651.81, "total_tokens": 658570240}
+{"current_steps": 15065, "total_steps": 29704, "loss": 0.310034441947937, "lr": 4.887900309961163e-06, "epoch": 1.014341502827902, "percentage": 50.72, "elapsed_time": "1 day, 8:22:42", "remaining_time": "1 day, 7:27:46", "throughput": 5651.85, "total_tokens": 658791472}
+{"current_steps": 15070, "total_steps": 29704, "loss": 0.32690412998199464, "lr": 4.885256908215532e-06, "epoch": 1.0146781578238622, "percentage": 50.73, "elapsed_time": "1 day, 8:23:10", "remaining_time": "1 day, 7:26:57", "throughput": 5652.37, "total_tokens": 659014128}
+{"current_steps": 15075, "total_steps": 29704, "loss": 0.3006293296813965, "lr": 4.88261353855744e-06, "epoch": 1.0150148128198222, "percentage": 50.75, "elapsed_time": "1 day, 8:23:52", "remaining_time": "1 day, 7:26:21", "throughput": 5652.29, "total_tokens": 659239712}
+{"current_steps": 15080, "total_steps": 29704, "loss": 0.28703746795654295, "lr": 4.8799702017261e-06, "epoch": 1.0153514678157824, "percentage": 50.77, "elapsed_time": "1 day, 8:24:28", "remaining_time": "1 day, 7:25:40", "throughput": 5652.39, "total_tokens": 659455440}
+{"current_steps": 15085, "total_steps": 29704, "loss": 0.31946167945861814, "lr": 4.877326898460709e-06, "epoch": 1.0156881228117425, "percentage": 50.78, "elapsed_time": "1 day, 8:25:10", "remaining_time": "1 day, 7:25:04", "throughput": 5652.18, "total_tokens": 659666248}
+{"current_steps": 15090, "total_steps": 29704, "loss": 0.28933143615722656, "lr": 4.87468362950046e-06, "epoch": 1.0160247778077027, "percentage": 50.8, "elapsed_time": "1 day, 8:25:56", "remaining_time": "1 day, 7:24:33", "throughput": 5651.87, "total_tokens": 659891120}
+{"current_steps": 15095, "total_steps": 29704, "loss": 0.3063974380493164, "lr": 4.872040395584533e-06, "epoch": 1.0163614328036628, "percentage": 50.82, "elapsed_time": "1 day, 8:26:32", "remaining_time": "1 day, 7:23:51", "throughput": 5651.93, "total_tokens": 660099960}
+{"current_steps": 15100, "total_steps": 29704, "loss": 0.2840407848358154, "lr": 4.869397197452104e-06, "epoch": 1.016698087799623, "percentage": 50.83, "elapsed_time": "1 day, 8:27:07", "remaining_time": "1 day, 7:23:10", "throughput": 5652.08, "total_tokens": 660319352}
+{"current_steps": 15105, "total_steps": 29704, "loss": 0.3019397258758545, "lr": 4.866754035842329e-06, "epoch": 1.017034742795583, "percentage": 50.85, "elapsed_time": "1 day, 8:27:44", "remaining_time": "1 day, 7:22:30", "throughput": 5652.17, "total_tokens": 660540440}
+{"current_steps": 15110, "total_steps": 29704, "loss": 0.30378923416137693, "lr": 4.8641109114943666e-06, "epoch": 1.0173713977915433, "percentage": 50.87, "elapsed_time": "1 day, 8:28:20", "remaining_time": "1 day, 7:21:48", "throughput": 5652.28, "total_tokens": 660755584}
+{"current_steps": 15115, "total_steps": 29704, "loss": 0.3153241157531738, "lr": 4.861467825147352e-06, "epoch": 1.0177080527875033, "percentage": 50.89, "elapsed_time": "1 day, 8:28:53", "remaining_time": "1 day, 7:21:04", "throughput": 5652.57, "total_tokens": 660976736}
+{"current_steps": 15120, "total_steps": 29704, "loss": 0.3002683401107788, "lr": 4.858824777540419e-06, "epoch": 1.0180447077834636, "percentage": 50.9, "elapsed_time": "1 day, 8:29:36", "remaining_time": "1 day, 7:20:29", "throughput": 5652.37, "total_tokens": 661194120}
+{"current_steps": 15125, "total_steps": 29704, "loss": 0.3039540767669678, "lr": 4.856181769412686e-06, "epoch": 1.0183813627794236, "percentage": 50.92, "elapsed_time": "1 day, 8:30:21", "remaining_time": "1 day, 7:19:56", "throughput": 5652.02, "total_tokens": 661406256}
+{"current_steps": 15130, "total_steps": 29704, "loss": 0.30543341636657717, "lr": 4.8535388015032644e-06, "epoch": 1.0187180177753838, "percentage": 50.94, "elapsed_time": "1 day, 8:30:59", "remaining_time": "1 day, 7:19:17", "throughput": 5651.97, "total_tokens": 661615064}
+{"current_steps": 15135, "total_steps": 29704, "loss": 0.32107832431793215, "lr": 4.850895874551248e-06, "epoch": 1.0190546727713439, "percentage": 50.95, "elapsed_time": "1 day, 8:31:38", "remaining_time": "1 day, 7:18:39", "throughput": 5652.02, "total_tokens": 661843768}
+{"current_steps": 15140, "total_steps": 29704, "loss": 0.3267232894897461, "lr": 4.848252989295726e-06, "epoch": 1.0193913277673041, "percentage": 50.97, "elapsed_time": "1 day, 8:32:15", "remaining_time": "1 day, 7:17:58", "throughput": 5651.97, "total_tokens": 662043496}
+{"current_steps": 15145, "total_steps": 29704, "loss": 0.2995804786682129, "lr": 4.845610146475771e-06, "epoch": 1.0197279827632642, "percentage": 50.99, "elapsed_time": "1 day, 8:32:49", "remaining_time": "1 day, 7:17:15", "throughput": 5652.19, "total_tokens": 662264176}
+{"current_steps": 15150, "total_steps": 29704, "loss": 0.2979107141494751, "lr": 4.842967346830447e-06, "epoch": 1.0200646377592244, "percentage": 51.0, "elapsed_time": "1 day, 8:33:25", "remaining_time": "1 day, 7:16:34", "throughput": 5652.33, "total_tokens": 662484464}
+{"current_steps": 15155, "total_steps": 29704, "loss": 0.3103074073791504, "lr": 4.840324591098803e-06, "epoch": 1.0204012927551844, "percentage": 51.02, "elapsed_time": "1 day, 8:34:05", "remaining_time": "1 day, 7:15:57", "throughput": 5652.17, "total_tokens": 662691688}
+{"current_steps": 15160, "total_steps": 29704, "loss": 0.3106090068817139, "lr": 4.83768188001988e-06, "epoch": 1.0207379477511447, "percentage": 51.04, "elapsed_time": "1 day, 8:34:50", "remaining_time": "1 day, 7:15:24", "throughput": 5651.73, "total_tokens": 662895200}
+{"current_steps": 15165, "total_steps": 29704, "loss": 0.30320937633514405, "lr": 4.835039214332699e-06, "epoch": 1.0210746027471047, "percentage": 51.05, "elapsed_time": "1 day, 8:35:31", "remaining_time": "1 day, 7:14:48", "throughput": 5651.48, "total_tokens": 663097200}
+{"current_steps": 15170, "total_steps": 29704, "loss": 0.3088703155517578, "lr": 4.832396594776277e-06, "epoch": 1.021411257743065, "percentage": 51.07, "elapsed_time": "1 day, 8:36:15", "remaining_time": "1 day, 7:14:14", "throughput": 5651.28, "total_tokens": 663319712}
+{"current_steps": 15175, "total_steps": 29704, "loss": 0.2921560764312744, "lr": 4.82975402208961e-06, "epoch": 1.021747912739025, "percentage": 51.09, "elapsed_time": "1 day, 8:36:53", "remaining_time": "1 day, 7:13:35", "throughput": 5651.27, "total_tokens": 663537384}
+{"current_steps": 15180, "total_steps": 29704, "loss": 0.284089994430542, "lr": 4.827111497011692e-06, "epoch": 1.0220845677349852, "percentage": 51.1, "elapsed_time": "1 day, 8:37:32", "remaining_time": "1 day, 7:12:57", "throughput": 5651.43, "total_tokens": 663777040}
+{"current_steps": 15185, "total_steps": 29704, "loss": 0.3103042125701904, "lr": 4.82446902028149e-06, "epoch": 1.0224212227309453, "percentage": 51.12, "elapsed_time": "1 day, 8:38:10", "remaining_time": "1 day, 7:12:17", "throughput": 5651.51, "total_tokens": 663998952}
+{"current_steps": 15190, "total_steps": 29704, "loss": 0.3256609201431274, "lr": 4.821826592637968e-06, "epoch": 1.0227578777269055, "percentage": 51.14, "elapsed_time": "1 day, 8:38:45", "remaining_time": "1 day, 7:11:35", "throughput": 5651.69, "total_tokens": 664216976}
+{"current_steps": 15195, "total_steps": 29704, "loss": 0.30333666801452636, "lr": 4.8191842148200694e-06, "epoch": 1.0230945327228655, "percentage": 51.15, "elapsed_time": "1 day, 8:39:23", "remaining_time": "1 day, 7:10:55", "throughput": 5651.82, "total_tokens": 664447704}
+{"current_steps": 15200, "total_steps": 29704, "loss": 0.3009442090988159, "lr": 4.8165418875667315e-06, "epoch": 1.0234311877188258, "percentage": 51.17, "elapsed_time": "1 day, 8:40:09", "remaining_time": "1 day, 7:10:23", "throughput": 5651.59, "total_tokens": 664678616}
+{"current_steps": 15205, "total_steps": 29704, "loss": 0.3210268020629883, "lr": 4.813899611616867e-06, "epoch": 1.0237678427147858, "percentage": 51.19, "elapsed_time": "1 day, 8:40:45", "remaining_time": "1 day, 7:09:42", "throughput": 5651.72, "total_tokens": 664897488}
+{"current_steps": 15210, "total_steps": 29704, "loss": 0.3092465400695801, "lr": 4.8112573877093855e-06, "epoch": 1.024104497710746, "percentage": 51.21, "elapsed_time": "1 day, 8:41:26", "remaining_time": "1 day, 7:09:06", "throughput": 5651.7, "total_tokens": 665131288}
+{"current_steps": 15215, "total_steps": 29704, "loss": 0.3192662000656128, "lr": 4.808615216583171e-06, "epoch": 1.024441152706706, "percentage": 51.22, "elapsed_time": "1 day, 8:42:00", "remaining_time": "1 day, 7:08:23", "throughput": 5651.88, "total_tokens": 665344872}
+{"current_steps": 15220, "total_steps": 29704, "loss": 0.32068495750427245, "lr": 4.805973098977102e-06, "epoch": 1.0247778077026664, "percentage": 51.24, "elapsed_time": "1 day, 8:42:35", "remaining_time": "1 day, 7:07:40", "throughput": 5652.03, "total_tokens": 665556568}
+{"current_steps": 15225, "total_steps": 29704, "loss": 0.32355175018310545, "lr": 4.803331035630037e-06, "epoch": 1.0251144626986264, "percentage": 51.26, "elapsed_time": "1 day, 8:43:08", "remaining_time": "1 day, 7:06:56", "throughput": 5652.4, "total_tokens": 665786432}
+{"current_steps": 15230, "total_steps": 29704, "loss": 0.29448685646057127, "lr": 4.800689027280822e-06, "epoch": 1.0254511176945866, "percentage": 51.27, "elapsed_time": "1 day, 8:43:48", "remaining_time": "1 day, 7:06:19", "throughput": 5652.3, "total_tokens": 666001856}
+{"current_steps": 15235, "total_steps": 29704, "loss": 0.2994450330734253, "lr": 4.798047074668283e-06, "epoch": 1.0257877726905467, "percentage": 51.29, "elapsed_time": "1 day, 8:44:26", "remaining_time": "1 day, 7:05:39", "throughput": 5652.35, "total_tokens": 666220128}
+{"current_steps": 15240, "total_steps": 29704, "loss": 0.29120640754699706, "lr": 4.795405178531236e-06, "epoch": 1.026124427686507, "percentage": 51.31, "elapsed_time": "1 day, 8:44:59", "remaining_time": "1 day, 7:04:56", "throughput": 5652.51, "total_tokens": 666430824}
+{"current_steps": 15245, "total_steps": 29704, "loss": 0.31427597999572754, "lr": 4.792763339608478e-06, "epoch": 1.026461082682467, "percentage": 51.32, "elapsed_time": "1 day, 8:45:41", "remaining_time": "1 day, 7:04:20", "throughput": 5652.28, "total_tokens": 666637408}
+{"current_steps": 15250, "total_steps": 29704, "loss": 0.2697123050689697, "lr": 4.790121558638793e-06, "epoch": 1.0267977376784272, "percentage": 51.34, "elapsed_time": "1 day, 8:46:23", "remaining_time": "1 day, 7:03:44", "throughput": 5652.13, "total_tokens": 666856712}
+{"current_steps": 15255, "total_steps": 29704, "loss": 0.33616783618927004, "lr": 4.787479836360944e-06, "epoch": 1.0271343926743872, "percentage": 51.36, "elapsed_time": "1 day, 8:46:55", "remaining_time": "1 day, 7:03:00", "throughput": 5652.41, "total_tokens": 667072880}
+{"current_steps": 15260, "total_steps": 29704, "loss": 0.308072566986084, "lr": 4.784838173513682e-06, "epoch": 1.0274710476703475, "percentage": 51.37, "elapsed_time": "1 day, 8:47:35", "remaining_time": "1 day, 7:02:22", "throughput": 5652.34, "total_tokens": 667290960}
+{"current_steps": 15265, "total_steps": 29704, "loss": 0.3036174297332764, "lr": 4.782196570835737e-06, "epoch": 1.0278077026663075, "percentage": 51.39, "elapsed_time": "1 day, 8:48:12", "remaining_time": "1 day, 7:01:42", "throughput": 5652.52, "total_tokens": 667517912}
+{"current_steps": 15270, "total_steps": 29704, "loss": 0.29191629886627196, "lr": 4.779555029065827e-06, "epoch": 1.0281443576622677, "percentage": 51.41, "elapsed_time": "1 day, 8:48:51", "remaining_time": "1 day, 7:01:04", "throughput": 5652.3, "total_tokens": 667717400}
+{"current_steps": 15275, "total_steps": 29704, "loss": 0.31882758140563966, "lr": 4.776913548942649e-06, "epoch": 1.0284810126582278, "percentage": 51.42, "elapsed_time": "1 day, 8:49:31", "remaining_time": "1 day, 7:00:26", "throughput": 5652.32, "total_tokens": 667943704}
+{"current_steps": 15280, "total_steps": 29704, "loss": 0.31909172534942626, "lr": 4.774272131204889e-06, "epoch": 1.028817667654188, "percentage": 51.44, "elapsed_time": "1 day, 8:50:14", "remaining_time": "1 day, 6:59:52", "throughput": 5652.17, "total_tokens": 668171240}
+{"current_steps": 15285, "total_steps": 29704, "loss": 0.3175906419754028, "lr": 4.7716307765912045e-06, "epoch": 1.029154322650148, "percentage": 51.46, "elapsed_time": "1 day, 8:50:59", "remaining_time": "1 day, 6:59:19", "throughput": 5651.89, "total_tokens": 668389056}
+{"current_steps": 15290, "total_steps": 29704, "loss": 0.3248094081878662, "lr": 4.768989485840246e-06, "epoch": 1.0294909776461083, "percentage": 51.47, "elapsed_time": "1 day, 8:51:35", "remaining_time": "1 day, 6:58:37", "throughput": 5652.13, "total_tokens": 668619360}
+{"current_steps": 15295, "total_steps": 29704, "loss": 0.2917296886444092, "lr": 4.766348259690641e-06, "epoch": 1.0298276326420683, "percentage": 51.49, "elapsed_time": "1 day, 8:52:07", "remaining_time": "1 day, 6:57:53", "throughput": 5652.49, "total_tokens": 668845224}
+{"current_steps": 15300, "total_steps": 29704, "loss": 0.3045883893966675, "lr": 4.763707098881002e-06, "epoch": 1.0301642876380286, "percentage": 51.51, "elapsed_time": "1 day, 8:52:46", "remaining_time": "1 day, 6:57:14", "throughput": 5652.43, "total_tokens": 669059584}
+{"current_steps": 15305, "total_steps": 29704, "loss": 0.3286751270294189, "lr": 4.761066004149917e-06, "epoch": 1.0305009426339886, "percentage": 51.53, "elapsed_time": "1 day, 8:53:25", "remaining_time": "1 day, 6:56:36", "throughput": 5652.32, "total_tokens": 669264992}
+{"current_steps": 15310, "total_steps": 29704, "loss": 0.3184011936187744, "lr": 4.758424976235964e-06, "epoch": 1.0308375976299489, "percentage": 51.54, "elapsed_time": "1 day, 8:54:05", "remaining_time": "1 day, 6:55:58", "throughput": 5652.29, "total_tokens": 669485880}
+{"current_steps": 15315, "total_steps": 29704, "loss": 0.33234100341796874, "lr": 4.755784015877694e-06, "epoch": 1.031174252625909, "percentage": 51.56, "elapsed_time": "1 day, 8:54:39", "remaining_time": "1 day, 6:55:15", "throughput": 5652.61, "total_tokens": 669718792}
+{"current_steps": 15320, "total_steps": 29704, "loss": 0.3180215358734131, "lr": 4.753143123813646e-06, "epoch": 1.0315109076218691, "percentage": 51.58, "elapsed_time": "1 day, 8:55:16", "remaining_time": "1 day, 6:54:35", "throughput": 5652.66, "total_tokens": 669934720}
+{"current_steps": 15325, "total_steps": 29704, "loss": 0.3165398597717285, "lr": 4.7505023007823365e-06, "epoch": 1.0318475626178292, "percentage": 51.59, "elapsed_time": "1 day, 8:55:58", "remaining_time": "1 day, 6:53:59", "throughput": 5652.56, "total_tokens": 670156656}
+{"current_steps": 15330, "total_steps": 29704, "loss": 0.28763623237609864, "lr": 4.747861547522264e-06, "epoch": 1.0321842176137894, "percentage": 51.61, "elapsed_time": "1 day, 8:56:32", "remaining_time": "1 day, 6:53:16", "throughput": 5652.78, "total_tokens": 670377432}
+{"current_steps": 15335, "total_steps": 29704, "loss": 0.29185004234313966, "lr": 4.745220864771904e-06, "epoch": 1.0325208726097495, "percentage": 51.63, "elapsed_time": "1 day, 8:57:15", "remaining_time": "1 day, 6:52:41", "throughput": 5652.47, "total_tokens": 670582320}
+{"current_steps": 15340, "total_steps": 29704, "loss": 0.31283178329467776, "lr": 4.742580253269718e-06, "epoch": 1.0328575276057097, "percentage": 51.64, "elapsed_time": "1 day, 8:57:52", "remaining_time": "1 day, 6:52:02", "throughput": 5652.41, "total_tokens": 670787368}
+{"current_steps": 15345, "total_steps": 29704, "loss": 0.28745934963226316, "lr": 4.739939713754141e-06, "epoch": 1.0331941826016697, "percentage": 51.66, "elapsed_time": "1 day, 8:58:38", "remaining_time": "1 day, 6:51:30", "throughput": 5652.02, "total_tokens": 670998288}
+{"current_steps": 15350, "total_steps": 29704, "loss": 0.3006887912750244, "lr": 4.737299246963597e-06, "epoch": 1.03353083759763, "percentage": 51.68, "elapsed_time": "1 day, 8:59:16", "remaining_time": "1 day, 6:50:50", "throughput": 5651.97, "total_tokens": 671207256}
+{"current_steps": 15355, "total_steps": 29704, "loss": 0.3070893526077271, "lr": 4.734658853636479e-06, "epoch": 1.03386749259359, "percentage": 51.69, "elapsed_time": "1 day, 8:59:53", "remaining_time": "1 day, 6:50:10", "throughput": 5652.07, "total_tokens": 671429312}
+{"current_steps": 15360, "total_steps": 29704, "loss": 0.3325775146484375, "lr": 4.732018534511167e-06, "epoch": 1.0342041475895503, "percentage": 51.71, "elapsed_time": "1 day, 9:00:28", "remaining_time": "1 day, 6:49:28", "throughput": 5652.24, "total_tokens": 671648736}
+{"current_steps": 15365, "total_steps": 29704, "loss": 0.2876759052276611, "lr": 4.729378290326017e-06, "epoch": 1.0345408025855103, "percentage": 51.73, "elapsed_time": "1 day, 9:01:01", "remaining_time": "1 day, 6:48:44", "throughput": 5652.39, "total_tokens": 671854064}
+{"current_steps": 15370, "total_steps": 29704, "loss": 0.3200536727905273, "lr": 4.726738121819365e-06, "epoch": 1.0348774575814705, "percentage": 51.74, "elapsed_time": "1 day, 9:01:35", "remaining_time": "1 day, 6:48:01", "throughput": 5652.61, "total_tokens": 672072384}
+{"current_steps": 15375, "total_steps": 29704, "loss": 0.2775277137756348, "lr": 4.7240980297295255e-06, "epoch": 1.0352141125774306, "percentage": 51.76, "elapsed_time": "1 day, 9:02:12", "remaining_time": "1 day, 6:47:21", "throughput": 5652.64, "total_tokens": 672283880}
+{"current_steps": 15380, "total_steps": 29704, "loss": 0.30312108993530273, "lr": 4.721458014794795e-06, "epoch": 1.0355507675733908, "percentage": 51.78, "elapsed_time": "1 day, 9:02:45", "remaining_time": "1 day, 6:46:37", "throughput": 5652.98, "total_tokens": 672508584}
+{"current_steps": 15385, "total_steps": 29704, "loss": 0.32550725936889646, "lr": 4.718818077753439e-06, "epoch": 1.0358874225693508, "percentage": 51.79, "elapsed_time": "1 day, 9:03:21", "remaining_time": "1 day, 6:45:56", "throughput": 5653.18, "total_tokens": 672737840}
+{"current_steps": 15390, "total_steps": 29704, "loss": 0.310680890083313, "lr": 4.716178219343712e-06, "epoch": 1.036224077565311, "percentage": 51.81, "elapsed_time": "1 day, 9:03:57", "remaining_time": "1 day, 6:45:15", "throughput": 5653.27, "total_tokens": 672951456}
+{"current_steps": 15395, "total_steps": 29704, "loss": 0.3095111846923828, "lr": 4.713538440303841e-06, "epoch": 1.0365607325612711, "percentage": 51.83, "elapsed_time": "1 day, 9:04:29", "remaining_time": "1 day, 6:44:30", "throughput": 5653.35, "total_tokens": 673143664}
+{"current_steps": 15400, "total_steps": 29704, "loss": 0.3282665491104126, "lr": 4.7108987413720335e-06, "epoch": 1.0368973875572314, "percentage": 51.84, "elapsed_time": "1 day, 9:05:08", "remaining_time": "1 day, 6:43:51", "throughput": 5653.31, "total_tokens": 673355776}
+{"current_steps": 15405, "total_steps": 29704, "loss": 0.31505331993103025, "lr": 4.708259123286469e-06, "epoch": 1.0372340425531914, "percentage": 51.86, "elapsed_time": "1 day, 9:05:52", "remaining_time": "1 day, 6:43:17", "throughput": 5652.99, "total_tokens": 673568088}
+{"current_steps": 15410, "total_steps": 29704, "loss": 0.30636959075927733, "lr": 4.7056195867853105e-06, "epoch": 1.0375706975491517, "percentage": 51.88, "elapsed_time": "1 day, 9:06:35", "remaining_time": "1 day, 6:42:42", "throughput": 5652.87, "total_tokens": 673795064}
+{"current_steps": 15415, "total_steps": 29704, "loss": 0.2860846996307373, "lr": 4.702980132606696e-06, "epoch": 1.0379073525451117, "percentage": 51.9, "elapsed_time": "1 day, 9:07:09", "remaining_time": "1 day, 6:42:00", "throughput": 5653.13, "total_tokens": 674021712}
+{"current_steps": 15420, "total_steps": 29704, "loss": 0.297334623336792, "lr": 4.700340761488741e-06, "epoch": 1.038244007541072, "percentage": 51.91, "elapsed_time": "1 day, 9:07:39", "remaining_time": "1 day, 6:41:13", "throughput": 5653.58, "total_tokens": 674242240}
+{"current_steps": 15425, "total_steps": 29704, "loss": 0.3014695167541504, "lr": 4.6977014741695345e-06, "epoch": 1.038580662537032, "percentage": 51.93, "elapsed_time": "1 day, 9:08:21", "remaining_time": "1 day, 6:40:38", "throughput": 5653.44, "total_tokens": 674464624}
+{"current_steps": 15430, "total_steps": 29704, "loss": 0.31081855297088623, "lr": 4.69506227138715e-06, "epoch": 1.0389173175329922, "percentage": 51.95, "elapsed_time": "1 day, 9:09:04", "remaining_time": "1 day, 6:40:03", "throughput": 5653.35, "total_tokens": 674697456}
+{"current_steps": 15435, "total_steps": 29704, "loss": 0.3378600120544434, "lr": 4.6924231538796265e-06, "epoch": 1.0392539725289522, "percentage": 51.96, "elapsed_time": "1 day, 9:09:39", "remaining_time": "1 day, 6:39:20", "throughput": 5653.59, "total_tokens": 674921088}
+{"current_steps": 15440, "total_steps": 29704, "loss": 0.3055275917053223, "lr": 4.6897841223849885e-06, "epoch": 1.0395906275249125, "percentage": 51.98, "elapsed_time": "1 day, 9:10:23", "remaining_time": "1 day, 6:38:47", "throughput": 5653.31, "total_tokens": 675136400}
+{"current_steps": 15445, "total_steps": 29704, "loss": 0.2911995887756348, "lr": 4.68714517764123e-06, "epoch": 1.0399272825208725, "percentage": 52.0, "elapsed_time": "1 day, 9:11:03", "remaining_time": "1 day, 6:38:10", "throughput": 5653.23, "total_tokens": 675354608}
+{"current_steps": 15450, "total_steps": 29704, "loss": 0.2882180690765381, "lr": 4.684506320386327e-06, "epoch": 1.0402639375168328, "percentage": 52.01, "elapsed_time": "1 day, 9:11:42", "remaining_time": "1 day, 6:37:31", "throughput": 5653.09, "total_tokens": 675560352}
+{"current_steps": 15455, "total_steps": 29704, "loss": 0.2950881481170654, "lr": 4.6818675513582235e-06, "epoch": 1.0406005925127928, "percentage": 52.03, "elapsed_time": "1 day, 9:12:29", "remaining_time": "1 day, 6:37:00", "throughput": 5652.72, "total_tokens": 675780176}
+{"current_steps": 15460, "total_steps": 29704, "loss": 0.2878568649291992, "lr": 4.679228871294846e-06, "epoch": 1.040937247508753, "percentage": 52.05, "elapsed_time": "1 day, 9:13:10", "remaining_time": "1 day, 6:36:23", "throughput": 5652.66, "total_tokens": 676001504}
+{"current_steps": 15465, "total_steps": 29704, "loss": 0.31134207248687745, "lr": 4.67659028093409e-06, "epoch": 1.041273902504713, "percentage": 52.06, "elapsed_time": "1 day, 9:13:46", "remaining_time": "1 day, 6:35:43", "throughput": 5652.78, "total_tokens": 676224760}
+{"current_steps": 15470, "total_steps": 29704, "loss": 0.3257226228713989, "lr": 4.673951781013834e-06, "epoch": 1.0416105575006733, "percentage": 52.08, "elapsed_time": "1 day, 9:14:23", "remaining_time": "1 day, 6:35:03", "throughput": 5652.81, "total_tokens": 676437448}
+{"current_steps": 15475, "total_steps": 29704, "loss": 0.32440500259399413, "lr": 4.67131337227192e-06, "epoch": 1.0419472124966334, "percentage": 52.1, "elapsed_time": "1 day, 9:14:56", "remaining_time": "1 day, 6:34:18", "throughput": 5653.14, "total_tokens": 676661152}
+{"current_steps": 15480, "total_steps": 29704, "loss": 0.3229750394821167, "lr": 4.668675055446175e-06, "epoch": 1.0422838674925936, "percentage": 52.11, "elapsed_time": "1 day, 9:15:31", "remaining_time": "1 day, 6:33:36", "throughput": 5653.3, "total_tokens": 676878032}
+{"current_steps": 15485, "total_steps": 29704, "loss": 0.27671065330505373, "lr": 4.666036831274392e-06, "epoch": 1.0426205224885536, "percentage": 52.13, "elapsed_time": "1 day, 9:16:07", "remaining_time": "1 day, 6:32:55", "throughput": 5653.4, "total_tokens": 677092008}
+{"current_steps": 15490, "total_steps": 29704, "loss": 0.30718288421630857, "lr": 4.663398700494346e-06, "epoch": 1.0429571774845139, "percentage": 52.15, "elapsed_time": "1 day, 9:16:48", "remaining_time": "1 day, 6:32:18", "throughput": 5653.22, "total_tokens": 677302040}
+{"current_steps": 15495, "total_steps": 29704, "loss": 0.3258528232574463, "lr": 4.66076066384378e-06, "epoch": 1.043293832480474, "percentage": 52.16, "elapsed_time": "1 day, 9:17:21", "remaining_time": "1 day, 6:31:35", "throughput": 5653.53, "total_tokens": 677525784}
+{"current_steps": 15500, "total_steps": 29704, "loss": 0.3136504888534546, "lr": 4.658122722060413e-06, "epoch": 1.0436304874764342, "percentage": 52.18, "elapsed_time": "1 day, 9:17:53", "remaining_time": "1 day, 6:30:50", "throughput": 5653.84, "total_tokens": 677745832}
+{"current_steps": 15505, "total_steps": 29704, "loss": 0.3184205055236816, "lr": 4.655484875881935e-06, "epoch": 1.0439671424723942, "percentage": 52.2, "elapsed_time": "1 day, 9:18:39", "remaining_time": "1 day, 6:30:18", "throughput": 5653.45, "total_tokens": 677956888}
+{"current_steps": 15510, "total_steps": 29704, "loss": 0.335153865814209, "lr": 4.652847126046014e-06, "epoch": 1.0443037974683544, "percentage": 52.22, "elapsed_time": "1 day, 9:19:15", "remaining_time": "1 day, 6:29:37", "throughput": 5653.35, "total_tokens": 678150664}
+{"current_steps": 15515, "total_steps": 29704, "loss": 0.33000261783599855, "lr": 4.650209473290286e-06, "epoch": 1.0446404524643145, "percentage": 52.23, "elapsed_time": "1 day, 9:19:50", "remaining_time": "1 day, 6:28:55", "throughput": 5653.61, "total_tokens": 678377032}
+{"current_steps": 15520, "total_steps": 29704, "loss": 0.31568212509155275, "lr": 4.647571918352366e-06, "epoch": 1.0449771074602747, "percentage": 52.25, "elapsed_time": "1 day, 9:20:30", "remaining_time": "1 day, 6:28:18", "throughput": 5653.51, "total_tokens": 678595080}
+{"current_steps": 15525, "total_steps": 29704, "loss": 0.305867600440979, "lr": 4.644934461969833e-06, "epoch": 1.0453137624562348, "percentage": 52.27, "elapsed_time": "1 day, 9:21:18", "remaining_time": "1 day, 6:27:48", "throughput": 5653.11, "total_tokens": 678818928}
+{"current_steps": 15530, "total_steps": 29704, "loss": 0.2917011260986328, "lr": 4.642297104880247e-06, "epoch": 1.045650417452195, "percentage": 52.28, "elapsed_time": "1 day, 9:21:57", "remaining_time": "1 day, 6:27:09", "throughput": 5653.15, "total_tokens": 679044840}
+{"current_steps": 15535, "total_steps": 29704, "loss": 0.29921350479125974, "lr": 4.639659847821133e-06, "epoch": 1.045987072448155, "percentage": 52.3, "elapsed_time": "1 day, 9:22:35", "remaining_time": "1 day, 6:26:30", "throughput": 5653.18, "total_tokens": 679260504}
+{"current_steps": 15540, "total_steps": 29704, "loss": 0.2979047060012817, "lr": 4.6370226915299945e-06, "epoch": 1.0463237274441153, "percentage": 52.32, "elapsed_time": "1 day, 9:23:08", "remaining_time": "1 day, 6:25:46", "throughput": 5653.42, "total_tokens": 679478208}
+{"current_steps": 15545, "total_steps": 29704, "loss": 0.3120341539382935, "lr": 4.634385636744302e-06, "epoch": 1.0466603824400753, "percentage": 52.33, "elapsed_time": "1 day, 9:23:46", "remaining_time": "1 day, 6:25:07", "throughput": 5653.57, "total_tokens": 679708344}
+{"current_steps": 15550, "total_steps": 29704, "loss": 0.3280055522918701, "lr": 4.631748684201503e-06, "epoch": 1.0469970374360356, "percentage": 52.35, "elapsed_time": "1 day, 9:24:27", "remaining_time": "1 day, 6:24:30", "throughput": 5653.56, "total_tokens": 679937752}
+{"current_steps": 15555, "total_steps": 29704, "loss": 0.3046785593032837, "lr": 4.629111834639008e-06, "epoch": 1.0473336924319956, "percentage": 52.37, "elapsed_time": "1 day, 9:25:05", "remaining_time": "1 day, 6:23:50", "throughput": 5653.66, "total_tokens": 680165064}
+{"current_steps": 15560, "total_steps": 29704, "loss": 0.3018321990966797, "lr": 4.626475088794206e-06, "epoch": 1.0476703474279558, "percentage": 52.38, "elapsed_time": "1 day, 9:25:42", "remaining_time": "1 day, 6:23:10", "throughput": 5653.78, "total_tokens": 680389272}
+{"current_steps": 15565, "total_steps": 29704, "loss": 0.3129916191101074, "lr": 4.623838447404454e-06, "epoch": 1.0480070024239159, "percentage": 52.4, "elapsed_time": "1 day, 9:26:25", "remaining_time": "1 day, 6:22:36", "throughput": 5653.68, "total_tokens": 680620144}
+{"current_steps": 15570, "total_steps": 29704, "loss": 0.29700002670288084, "lr": 4.621201911207082e-06, "epoch": 1.0483436574198761, "percentage": 52.42, "elapsed_time": "1 day, 9:27:00", "remaining_time": "1 day, 6:21:54", "throughput": 5653.77, "total_tokens": 680832464}
+{"current_steps": 15575, "total_steps": 29704, "loss": 0.3175528526306152, "lr": 4.618565480939387e-06, "epoch": 1.0486803124158361, "percentage": 52.43, "elapsed_time": "1 day, 9:27:34", "remaining_time": "1 day, 6:21:11", "throughput": 5654.07, "total_tokens": 681059856}
+{"current_steps": 15580, "total_steps": 29704, "loss": 0.3175390958786011, "lr": 4.6159291573386395e-06, "epoch": 1.0490169674117964, "percentage": 52.45, "elapsed_time": "1 day, 9:28:11", "remaining_time": "1 day, 6:20:31", "throughput": 5654.17, "total_tokens": 681279128}
+{"current_steps": 15585, "total_steps": 29704, "loss": 0.30297222137451174, "lr": 4.613292941142077e-06, "epoch": 1.0493536224077564, "percentage": 52.47, "elapsed_time": "1 day, 9:28:52", "remaining_time": "1 day, 6:19:54", "throughput": 5654.08, "total_tokens": 681498080}
+{"current_steps": 15590, "total_steps": 29704, "loss": 0.31702475547790526, "lr": 4.610656833086912e-06, "epoch": 1.0496902774037167, "percentage": 52.48, "elapsed_time": "1 day, 9:29:31", "remaining_time": "1 day, 6:19:15", "throughput": 5653.99, "total_tokens": 681707736}
+{"current_steps": 15595, "total_steps": 29704, "loss": 0.3515254259109497, "lr": 4.60802083391032e-06, "epoch": 1.0500269323996767, "percentage": 52.5, "elapsed_time": "1 day, 9:30:07", "remaining_time": "1 day, 6:18:35", "throughput": 5654.11, "total_tokens": 681929096}
+{"current_steps": 15600, "total_steps": 29704, "loss": 0.29188697338104247, "lr": 4.605384944349454e-06, "epoch": 1.050363587395637, "percentage": 52.52, "elapsed_time": "1 day, 9:30:43", "remaining_time": "1 day, 6:17:54", "throughput": 5654.15, "total_tokens": 682138848}
+{"current_steps": 15605, "total_steps": 29704, "loss": 0.2967026948928833, "lr": 4.602749165141429e-06, "epoch": 1.050700242391597, "percentage": 52.54, "elapsed_time": "1 day, 9:31:20", "remaining_time": "1 day, 6:17:14", "throughput": 5654.1, "total_tokens": 682341504}
+{"current_steps": 15610, "total_steps": 29704, "loss": 0.3200972080230713, "lr": 4.600113497023333e-06, "epoch": 1.0510368973875572, "percentage": 52.55, "elapsed_time": "1 day, 9:31:58", "remaining_time": "1 day, 6:16:34", "throughput": 5654.15, "total_tokens": 682558744}
+{"current_steps": 15615, "total_steps": 29704, "loss": 0.2950571537017822, "lr": 4.597477940732222e-06, "epoch": 1.0513735523835175, "percentage": 52.57, "elapsed_time": "1 day, 9:32:38", "remaining_time": "1 day, 6:15:57", "throughput": 5653.95, "total_tokens": 682762744}
+{"current_steps": 15620, "total_steps": 29704, "loss": 0.2987983226776123, "lr": 4.594842497005124e-06, "epoch": 1.0517102073794775, "percentage": 52.59, "elapsed_time": "1 day, 9:33:11", "remaining_time": "1 day, 6:15:13", "throughput": 5654.34, "total_tokens": 682994336}
+{"current_steps": 15625, "total_steps": 29704, "loss": 0.3123783111572266, "lr": 4.5922071665790264e-06, "epoch": 1.0520468623754375, "percentage": 52.6, "elapsed_time": "1 day, 9:33:50", "remaining_time": "1 day, 6:14:35", "throughput": 5654.25, "total_tokens": 683207456}
+{"current_steps": 15630, "total_steps": 29704, "loss": 0.3103508949279785, "lr": 4.589571950190897e-06, "epoch": 1.0523835173713978, "percentage": 52.62, "elapsed_time": "1 day, 9:34:33", "remaining_time": "1 day, 6:13:59", "throughput": 5654.17, "total_tokens": 683437384}
+{"current_steps": 15635, "total_steps": 29704, "loss": 0.3118293285369873, "lr": 4.586936848577661e-06, "epoch": 1.052720172367358, "percentage": 52.64, "elapsed_time": "1 day, 9:35:07", "remaining_time": "1 day, 6:13:17", "throughput": 5654.38, "total_tokens": 683657552}
+{"current_steps": 15640, "total_steps": 29704, "loss": 0.30995869636535645, "lr": 4.584301862476219e-06, "epoch": 1.053056827363318, "percentage": 52.65, "elapsed_time": "1 day, 9:35:44", "remaining_time": "1 day, 6:12:37", "throughput": 5654.45, "total_tokens": 683876008}
+{"current_steps": 15645, "total_steps": 29704, "loss": 0.2860646963119507, "lr": 4.581666992623435e-06, "epoch": 1.053393482359278, "percentage": 52.67, "elapsed_time": "1 day, 9:36:24", "remaining_time": "1 day, 6:11:59", "throughput": 5654.48, "total_tokens": 684104824}
+{"current_steps": 15650, "total_steps": 29704, "loss": 0.3144806146621704, "lr": 4.579032239756144e-06, "epoch": 1.0537301373552383, "percentage": 52.69, "elapsed_time": "1 day, 9:37:06", "remaining_time": "1 day, 6:11:23", "throughput": 5654.37, "total_tokens": 684326216}
+{"current_steps": 15655, "total_steps": 29704, "loss": 0.29763078689575195, "lr": 4.576397604611143e-06, "epoch": 1.0540667923511986, "percentage": 52.7, "elapsed_time": "1 day, 9:37:40", "remaining_time": "1 day, 6:10:41", "throughput": 5654.58, "total_tokens": 684548424}
+{"current_steps": 15660, "total_steps": 29704, "loss": 0.30714902877807615, "lr": 4.573763087925202e-06, "epoch": 1.0544034473471586, "percentage": 52.72, "elapsed_time": "1 day, 9:38:25", "remaining_time": "1 day, 6:10:08", "throughput": 5654.33, "total_tokens": 684772760}
+{"current_steps": 15665, "total_steps": 29704, "loss": 0.3167744159698486, "lr": 4.571128690435053e-06, "epoch": 1.0547401023431189, "percentage": 52.74, "elapsed_time": "1 day, 9:39:09", "remaining_time": "1 day, 6:09:33", "throughput": 5654.15, "total_tokens": 684995472}
+{"current_steps": 15670, "total_steps": 29704, "loss": 0.31170904636383057, "lr": 4.568494412877401e-06, "epoch": 1.055076757339079, "percentage": 52.75, "elapsed_time": "1 day, 9:39:38", "remaining_time": "1 day, 6:08:47", "throughput": 5654.46, "total_tokens": 685198200}
+{"current_steps": 15675, "total_steps": 29704, "loss": 0.31058855056762696, "lr": 4.565860255988908e-06, "epoch": 1.0554134123350392, "percentage": 52.77, "elapsed_time": "1 day, 9:40:10", "remaining_time": "1 day, 6:08:02", "throughput": 5654.79, "total_tokens": 685420416}
+{"current_steps": 15680, "total_steps": 29704, "loss": 0.31565635204315184, "lr": 4.563226220506212e-06, "epoch": 1.0557500673309992, "percentage": 52.79, "elapsed_time": "1 day, 9:40:47", "remaining_time": "1 day, 6:07:22", "throughput": 5655.04, "total_tokens": 685659776}
+{"current_steps": 15685, "total_steps": 29704, "loss": 0.3112238645553589, "lr": 4.5605923071659085e-06, "epoch": 1.0560867223269594, "percentage": 52.8, "elapsed_time": "1 day, 9:41:23", "remaining_time": "1 day, 6:06:41", "throughput": 5655.09, "total_tokens": 685871672}
+{"current_steps": 15690, "total_steps": 29704, "loss": 0.2988229990005493, "lr": 4.557958516704566e-06, "epoch": 1.0564233773229195, "percentage": 52.82, "elapsed_time": "1 day, 9:42:10", "remaining_time": "1 day, 6:06:09", "throughput": 5654.75, "total_tokens": 686092808}
+{"current_steps": 15695, "total_steps": 29704, "loss": 0.3125143051147461, "lr": 4.5553248498587144e-06, "epoch": 1.0567600323188797, "percentage": 52.84, "elapsed_time": "1 day, 9:42:40", "remaining_time": "1 day, 6:05:23", "throughput": 5655.12, "total_tokens": 686309592}
+{"current_steps": 15700, "total_steps": 29704, "loss": 0.31246707439422605, "lr": 4.552691307364853e-06, "epoch": 1.0570966873148397, "percentage": 52.85, "elapsed_time": "1 day, 9:43:21", "remaining_time": "1 day, 6:04:47", "throughput": 5654.96, "total_tokens": 686520592}
+{"current_steps": 15705, "total_steps": 29704, "loss": 0.33534066677093505, "lr": 4.5500578899594375e-06, "epoch": 1.0574333423108, "percentage": 52.87, "elapsed_time": "1 day, 9:44:03", "remaining_time": "1 day, 6:04:11", "throughput": 5654.84, "total_tokens": 686743480}
+{"current_steps": 15710, "total_steps": 29704, "loss": 0.3143591403961182, "lr": 4.5474245983789e-06, "epoch": 1.05776999730676, "percentage": 52.89, "elapsed_time": "1 day, 9:44:42", "remaining_time": "1 day, 6:03:33", "throughput": 5654.82, "total_tokens": 686964184}
+{"current_steps": 15715, "total_steps": 29704, "loss": 0.29776525497436523, "lr": 4.54479143335963e-06, "epoch": 1.0581066523027203, "percentage": 52.91, "elapsed_time": "1 day, 9:45:28", "remaining_time": "1 day, 6:03:00", "throughput": 5654.5, "total_tokens": 687183192}
+{"current_steps": 15720, "total_steps": 29704, "loss": 0.32465856075286864, "lr": 4.5421583956379855e-06, "epoch": 1.0584433072986803, "percentage": 52.92, "elapsed_time": "1 day, 9:46:03", "remaining_time": "1 day, 6:02:19", "throughput": 5654.59, "total_tokens": 687394088}
+{"current_steps": 15725, "total_steps": 29704, "loss": 0.29742043018341063, "lr": 4.539525485950284e-06, "epoch": 1.0587799622946406, "percentage": 52.94, "elapsed_time": "1 day, 9:46:43", "remaining_time": "1 day, 6:01:41", "throughput": 5654.62, "total_tokens": 687621296}
+{"current_steps": 15730, "total_steps": 29704, "loss": 0.3188284397125244, "lr": 4.536892705032815e-06, "epoch": 1.0591166172906006, "percentage": 52.96, "elapsed_time": "1 day, 9:47:17", "remaining_time": "1 day, 6:00:58", "throughput": 5654.89, "total_tokens": 687846312}
+{"current_steps": 15735, "total_steps": 29704, "loss": 0.29900522232055665, "lr": 4.534260053621821e-06, "epoch": 1.0594532722865608, "percentage": 52.97, "elapsed_time": "1 day, 9:47:51", "remaining_time": "1 day, 6:00:15", "throughput": 5655.1, "total_tokens": 688062864}
+{"current_steps": 15740, "total_steps": 29704, "loss": 0.2740945816040039, "lr": 4.531627532453519e-06, "epoch": 1.0597899272825209, "percentage": 52.99, "elapsed_time": "1 day, 9:48:39", "remaining_time": "1 day, 5:59:45", "throughput": 5654.65, "total_tokens": 688281760}
+{"current_steps": 15745, "total_steps": 29704, "loss": 0.3090967655181885, "lr": 4.5289951422640826e-06, "epoch": 1.0601265822784811, "percentage": 53.01, "elapsed_time": "1 day, 9:49:17", "remaining_time": "1 day, 5:59:06", "throughput": 5654.54, "total_tokens": 688482288}
+{"current_steps": 15750, "total_steps": 29704, "loss": 0.29933021068572996, "lr": 4.526362883789655e-06, "epoch": 1.0604632372744411, "percentage": 53.02, "elapsed_time": "1 day, 9:49:59", "remaining_time": "1 day, 5:58:30", "throughput": 5654.43, "total_tokens": 688706904}
+{"current_steps": 15755, "total_steps": 29704, "loss": 0.32481019496917723, "lr": 4.5237307577663345e-06, "epoch": 1.0607998922704014, "percentage": 53.04, "elapsed_time": "1 day, 9:50:38", "remaining_time": "1 day, 5:57:52", "throughput": 5654.47, "total_tokens": 688934432}
+{"current_steps": 15760, "total_steps": 29704, "loss": 0.3126739501953125, "lr": 4.52109876493019e-06, "epoch": 1.0611365472663614, "percentage": 53.06, "elapsed_time": "1 day, 9:51:21", "remaining_time": "1 day, 5:57:17", "throughput": 5654.29, "total_tokens": 689154176}
+{"current_steps": 15765, "total_steps": 29704, "loss": 0.3287156343460083, "lr": 4.518466906017249e-06, "epoch": 1.0614732022623217, "percentage": 53.07, "elapsed_time": "1 day, 9:52:00", "remaining_time": "1 day, 5:56:39", "throughput": 5654.21, "total_tokens": 689365536}
+{"current_steps": 15770, "total_steps": 29704, "loss": 0.29747960567474363, "lr": 4.515835181763503e-06, "epoch": 1.0618098572582817, "percentage": 53.09, "elapsed_time": "1 day, 9:52:40", "remaining_time": "1 day, 5:56:01", "throughput": 5654.14, "total_tokens": 689584480}
+{"current_steps": 15775, "total_steps": 29704, "loss": 0.3129979848861694, "lr": 4.513203592904903e-06, "epoch": 1.062146512254242, "percentage": 53.11, "elapsed_time": "1 day, 9:53:16", "remaining_time": "1 day, 5:55:20", "throughput": 5654.26, "total_tokens": 689802248}
+{"current_steps": 15780, "total_steps": 29704, "loss": 0.3110002040863037, "lr": 4.510572140177368e-06, "epoch": 1.062483167250202, "percentage": 53.12, "elapsed_time": "1 day, 9:54:01", "remaining_time": "1 day, 5:54:47", "throughput": 5654.04, "total_tokens": 690027296}
+{"current_steps": 15785, "total_steps": 29704, "loss": 0.31113996505737307, "lr": 4.507940824316771e-06, "epoch": 1.0628198222461622, "percentage": 53.14, "elapsed_time": "1 day, 9:54:42", "remaining_time": "1 day, 5:54:11", "throughput": 5653.9, "total_tokens": 690244872}
+{"current_steps": 15790, "total_steps": 29704, "loss": 0.306329345703125, "lr": 4.505309646058954e-06, "epoch": 1.0631564772421223, "percentage": 53.16, "elapsed_time": "1 day, 9:55:23", "remaining_time": "1 day, 5:53:34", "throughput": 5653.76, "total_tokens": 690457904}
+{"current_steps": 15795, "total_steps": 29704, "loss": 0.290619421005249, "lr": 4.502678606139716e-06, "epoch": 1.0634931322380825, "percentage": 53.17, "elapsed_time": "1 day, 9:56:01", "remaining_time": "1 day, 5:52:55", "throughput": 5653.73, "total_tokens": 690669592}
+{"current_steps": 15800, "total_steps": 29704, "loss": 0.28872241973876955, "lr": 4.500047705294823e-06, "epoch": 1.0638297872340425, "percentage": 53.19, "elapsed_time": "1 day, 9:56:34", "remaining_time": "1 day, 5:52:11", "throughput": 5654.04, "total_tokens": 690892416}
+{"current_steps": 15805, "total_steps": 29704, "loss": 0.3109187364578247, "lr": 4.497416944259993e-06, "epoch": 1.0641664422300028, "percentage": 53.21, "elapsed_time": "1 day, 9:57:07", "remaining_time": "1 day, 5:51:27", "throughput": 5654.3, "total_tokens": 691111312}
+{"current_steps": 15810, "total_steps": 29704, "loss": 0.3253108263015747, "lr": 4.494786323770912e-06, "epoch": 1.0645030972259628, "percentage": 53.23, "elapsed_time": "1 day, 9:57:51", "remaining_time": "1 day, 5:50:53", "throughput": 5654.0, "total_tokens": 691322056}
+{"current_steps": 15815, "total_steps": 29704, "loss": 0.30454020500183104, "lr": 4.492155844563224e-06, "epoch": 1.064839752221923, "percentage": 53.24, "elapsed_time": "1 day, 9:58:27", "remaining_time": "1 day, 5:50:12", "throughput": 5654.19, "total_tokens": 691548176}
+{"current_steps": 15820, "total_steps": 29704, "loss": 0.3255009651184082, "lr": 4.489525507372537e-06, "epoch": 1.065176407217883, "percentage": 53.26, "elapsed_time": "1 day, 9:59:09", "remaining_time": "1 day, 5:49:37", "throughput": 5654.0, "total_tokens": 691767320}
+{"current_steps": 15825, "total_steps": 29704, "loss": 0.3073612689971924, "lr": 4.486895312934411e-06, "epoch": 1.0655130622138433, "percentage": 53.28, "elapsed_time": "1 day, 9:59:51", "remaining_time": "1 day, 5:49:01", "throughput": 5653.83, "total_tokens": 691981536}
+{"current_steps": 15830, "total_steps": 29704, "loss": 0.3512305974960327, "lr": 4.484265261984378e-06, "epoch": 1.0658497172098034, "percentage": 53.29, "elapsed_time": "1 day, 10:00:34", "remaining_time": "1 day, 5:48:25", "throughput": 5653.64, "total_tokens": 692199872}
+{"current_steps": 15835, "total_steps": 29704, "loss": 0.292151141166687, "lr": 4.4816353552579165e-06, "epoch": 1.0661863722057636, "percentage": 53.31, "elapsed_time": "1 day, 10:01:16", "remaining_time": "1 day, 5:47:50", "throughput": 5653.44, "total_tokens": 692412728}
+{"current_steps": 15840, "total_steps": 29704, "loss": 0.3161210536956787, "lr": 4.479005593490477e-06, "epoch": 1.0665230272017237, "percentage": 53.33, "elapsed_time": "1 day, 10:01:53", "remaining_time": "1 day, 5:47:10", "throughput": 5653.5, "total_tokens": 692630960}
+{"current_steps": 15845, "total_steps": 29704, "loss": 0.28519272804260254, "lr": 4.476375977417461e-06, "epoch": 1.066859682197684, "percentage": 53.34, "elapsed_time": "1 day, 10:02:35", "remaining_time": "1 day, 5:46:34", "throughput": 5653.4, "total_tokens": 692854048}
+{"current_steps": 15850, "total_steps": 29704, "loss": 0.3064600944519043, "lr": 4.473746507774235e-06, "epoch": 1.067196337193644, "percentage": 53.36, "elapsed_time": "1 day, 10:03:11", "remaining_time": "1 day, 5:45:53", "throughput": 5653.58, "total_tokens": 693081408}
+{"current_steps": 15855, "total_steps": 29704, "loss": 0.28726582527160643, "lr": 4.471117185296117e-06, "epoch": 1.0675329921896042, "percentage": 53.38, "elapsed_time": "1 day, 10:03:53", "remaining_time": "1 day, 5:45:17", "throughput": 5653.43, "total_tokens": 693299008}
+{"current_steps": 15860, "total_steps": 29704, "loss": 0.2859110593795776, "lr": 4.468488010718394e-06, "epoch": 1.0678696471855642, "percentage": 53.39, "elapsed_time": "1 day, 10:04:29", "remaining_time": "1 day, 5:44:36", "throughput": 5653.45, "total_tokens": 693503552}
+{"current_steps": 15865, "total_steps": 29704, "loss": 0.31541666984558103, "lr": 4.465858984776302e-06, "epoch": 1.0682063021815245, "percentage": 53.41, "elapsed_time": "1 day, 10:05:05", "remaining_time": "1 day, 5:43:55", "throughput": 5653.53, "total_tokens": 693719200}
+{"current_steps": 15870, "total_steps": 29704, "loss": 0.30719547271728515, "lr": 4.463230108205044e-06, "epoch": 1.0685429571774845, "percentage": 53.43, "elapsed_time": "1 day, 10:05:45", "remaining_time": "1 day, 5:43:18", "throughput": 5653.23, "total_tokens": 693910832}
+{"current_steps": 15875, "total_steps": 29704, "loss": 0.3158994674682617, "lr": 4.460601381739772e-06, "epoch": 1.0688796121734447, "percentage": 53.44, "elapsed_time": "1 day, 10:06:24", "remaining_time": "1 day, 5:42:39", "throughput": 5653.25, "total_tokens": 694130912}
+{"current_steps": 15880, "total_steps": 29704, "loss": 0.33564248085021975, "lr": 4.457972806115607e-06, "epoch": 1.0692162671694048, "percentage": 53.46, "elapsed_time": "1 day, 10:07:06", "remaining_time": "1 day, 5:42:04", "throughput": 5653.05, "total_tokens": 694345664}
+{"current_steps": 15885, "total_steps": 29704, "loss": 0.31251869201660154, "lr": 4.455344382067616e-06, "epoch": 1.069552922165365, "percentage": 53.48, "elapsed_time": "1 day, 10:07:49", "remaining_time": "1 day, 5:41:29", "throughput": 5652.83, "total_tokens": 694563184}
+{"current_steps": 15890, "total_steps": 29704, "loss": 0.3079527378082275, "lr": 4.452716110330832e-06, "epoch": 1.069889577161325, "percentage": 53.49, "elapsed_time": "1 day, 10:08:24", "remaining_time": "1 day, 5:40:46", "throughput": 5653.07, "total_tokens": 694784912}
+{"current_steps": 15895, "total_steps": 29704, "loss": 0.31774344444274905, "lr": 4.450087991640242e-06, "epoch": 1.0702262321572853, "percentage": 53.51, "elapsed_time": "1 day, 10:08:59", "remaining_time": "1 day, 5:40:05", "throughput": 5653.09, "total_tokens": 694988912}
+{"current_steps": 15900, "total_steps": 29704, "loss": 0.3184868812561035, "lr": 4.447460026730794e-06, "epoch": 1.0705628871532453, "percentage": 53.53, "elapsed_time": "1 day, 10:09:40", "remaining_time": "1 day, 5:39:28", "throughput": 5652.93, "total_tokens": 695201352}
+{"current_steps": 15905, "total_steps": 29704, "loss": 0.30800135135650636, "lr": 4.444832216337384e-06, "epoch": 1.0708995421492056, "percentage": 53.54, "elapsed_time": "1 day, 10:10:19", "remaining_time": "1 day, 5:38:50", "throughput": 5652.97, "total_tokens": 695423576}
+{"current_steps": 15910, "total_steps": 29704, "loss": 0.29862637519836427, "lr": 4.442204561194877e-06, "epoch": 1.0712361971451656, "percentage": 53.56, "elapsed_time": "1 day, 10:10:54", "remaining_time": "1 day, 5:38:08", "throughput": 5653.1, "total_tokens": 695642552}
+{"current_steps": 15915, "total_steps": 29704, "loss": 0.29393310546875, "lr": 4.439577062038085e-06, "epoch": 1.0715728521411259, "percentage": 53.58, "elapsed_time": "1 day, 10:11:35", "remaining_time": "1 day, 5:37:32", "throughput": 5653.14, "total_tokens": 695878256}
+{"current_steps": 15920, "total_steps": 29704, "loss": 0.3315741062164307, "lr": 4.436949719601782e-06, "epoch": 1.0719095071370859, "percentage": 53.6, "elapsed_time": "1 day, 10:12:08", "remaining_time": "1 day, 5:36:48", "throughput": 5653.41, "total_tokens": 696096832}
+{"current_steps": 15925, "total_steps": 29704, "loss": 0.3238795757293701, "lr": 4.434322534620692e-06, "epoch": 1.0722461621330461, "percentage": 53.61, "elapsed_time": "1 day, 10:12:45", "remaining_time": "1 day, 5:36:08", "throughput": 5653.36, "total_tokens": 696300144}
+{"current_steps": 15930, "total_steps": 29704, "loss": 0.2765275716781616, "lr": 4.431695507829504e-06, "epoch": 1.0725828171290062, "percentage": 53.63, "elapsed_time": "1 day, 10:13:25", "remaining_time": "1 day, 5:35:30", "throughput": 5653.35, "total_tokens": 696521264}
+{"current_steps": 15935, "total_steps": 29704, "loss": 0.3107874870300293, "lr": 4.429068639962853e-06, "epoch": 1.0729194721249664, "percentage": 53.65, "elapsed_time": "1 day, 10:14:05", "remaining_time": "1 day, 5:34:53", "throughput": 5653.25, "total_tokens": 696737432}
+{"current_steps": 15940, "total_steps": 29704, "loss": 0.3024940490722656, "lr": 4.426441931755337e-06, "epoch": 1.0732561271209264, "percentage": 53.66, "elapsed_time": "1 day, 10:14:46", "remaining_time": "1 day, 5:34:16", "throughput": 5653.18, "total_tokens": 696961616}
+{"current_steps": 15945, "total_steps": 29704, "loss": 0.32619800567626955, "lr": 4.4238153839415045e-06, "epoch": 1.0735927821168867, "percentage": 53.68, "elapsed_time": "1 day, 10:15:22", "remaining_time": "1 day, 5:33:35", "throughput": 5653.41, "total_tokens": 697193784}
+{"current_steps": 15950, "total_steps": 29704, "loss": 0.3372673749923706, "lr": 4.421188997255865e-06, "epoch": 1.0739294371128467, "percentage": 53.7, "elapsed_time": "1 day, 10:15:58", "remaining_time": "1 day, 5:32:54", "throughput": 5653.44, "total_tokens": 697403080}
+{"current_steps": 15955, "total_steps": 29704, "loss": 0.310252046585083, "lr": 4.418562772432874e-06, "epoch": 1.074266092108807, "percentage": 53.71, "elapsed_time": "1 day, 10:16:36", "remaining_time": "1 day, 5:32:15", "throughput": 5653.49, "total_tokens": 697621144}
+{"current_steps": 15960, "total_steps": 29704, "loss": 0.31459197998046873, "lr": 4.415936710206952e-06, "epoch": 1.074602747104767, "percentage": 53.73, "elapsed_time": "1 day, 10:17:14", "remaining_time": "1 day, 5:31:36", "throughput": 5653.65, "total_tokens": 697855736}
+{"current_steps": 15965, "total_steps": 29704, "loss": 0.31325018405914307, "lr": 4.413310811312465e-06, "epoch": 1.0749394021007272, "percentage": 53.75, "elapsed_time": "1 day, 10:17:48", "remaining_time": "1 day, 5:30:53", "throughput": 5653.84, "total_tokens": 698071128}
+{"current_steps": 15970, "total_steps": 29704, "loss": 0.2819394826889038, "lr": 4.41068507648374e-06, "epoch": 1.0752760570966873, "percentage": 53.76, "elapsed_time": "1 day, 10:18:23", "remaining_time": "1 day, 5:30:11", "throughput": 5654.08, "total_tokens": 698297384}
+{"current_steps": 15975, "total_steps": 29704, "loss": 0.29783015251159667, "lr": 4.408059506455053e-06, "epoch": 1.0756127120926475, "percentage": 53.78, "elapsed_time": "1 day, 10:19:01", "remaining_time": "1 day, 5:29:32", "throughput": 5654.25, "total_tokens": 698536768}
+{"current_steps": 15980, "total_steps": 29704, "loss": 0.3107433795928955, "lr": 4.40543410196064e-06, "epoch": 1.0759493670886076, "percentage": 53.8, "elapsed_time": "1 day, 10:19:39", "remaining_time": "1 day, 5:28:52", "throughput": 5654.35, "total_tokens": 698759848}
+{"current_steps": 15985, "total_steps": 29704, "loss": 0.32532343864440916, "lr": 4.402808863734681e-06, "epoch": 1.0762860220845678, "percentage": 53.81, "elapsed_time": "1 day, 10:20:25", "remaining_time": "1 day, 5:28:20", "throughput": 5654.02, "total_tokens": 698978376}
+{"current_steps": 15990, "total_steps": 29704, "loss": 0.3170433044433594, "lr": 4.40018379251132e-06, "epoch": 1.0766226770805278, "percentage": 53.83, "elapsed_time": "1 day, 10:21:08", "remaining_time": "1 day, 5:27:45", "throughput": 5653.81, "total_tokens": 699198368}
+{"current_steps": 15995, "total_steps": 29704, "loss": 0.350922966003418, "lr": 4.397558889024649e-06, "epoch": 1.076959332076488, "percentage": 53.85, "elapsed_time": "1 day, 10:21:44", "remaining_time": "1 day, 5:27:04", "throughput": 5653.78, "total_tokens": 699398432}
+{"current_steps": 16000, "total_steps": 29704, "loss": 0.3178732395172119, "lr": 4.394934154008715e-06, "epoch": 1.0772959870724481, "percentage": 53.86, "elapsed_time": "1 day, 10:22:31", "remaining_time": "1 day, 5:26:33", "throughput": 5653.42, "total_tokens": 699622136}
+{"current_steps": 16005, "total_steps": 29704, "loss": 0.3029392957687378, "lr": 4.392309588197514e-06, "epoch": 1.0776326420684084, "percentage": 53.88, "elapsed_time": "1 day, 10:23:09", "remaining_time": "1 day, 5:25:53", "throughput": 5653.52, "total_tokens": 699844832}
+{"current_steps": 16010, "total_steps": 29704, "loss": 0.30281388759613037, "lr": 4.389685192324999e-06, "epoch": 1.0779692970643684, "percentage": 53.9, "elapsed_time": "1 day, 10:23:51", "remaining_time": "1 day, 5:25:18", "throughput": 5653.33, "total_tokens": 700062776}
+{"current_steps": 16015, "total_steps": 29704, "loss": 0.3248000144958496, "lr": 4.387060967125073e-06, "epoch": 1.0783059520603286, "percentage": 53.92, "elapsed_time": "1 day, 10:24:26", "remaining_time": "1 day, 5:24:36", "throughput": 5653.43, "total_tokens": 700270472}
+{"current_steps": 16020, "total_steps": 29704, "loss": 0.3387366533279419, "lr": 4.384436913331596e-06, "epoch": 1.0786426070562887, "percentage": 53.93, "elapsed_time": "1 day, 10:25:08", "remaining_time": "1 day, 5:24:00", "throughput": 5653.28, "total_tokens": 700490600}
+{"current_steps": 16025, "total_steps": 29704, "loss": 0.3053700923919678, "lr": 4.38181303167837e-06, "epoch": 1.078979262052249, "percentage": 53.95, "elapsed_time": "1 day, 10:25:49", "remaining_time": "1 day, 5:23:23", "throughput": 5653.2, "total_tokens": 700708944}
+{"current_steps": 16030, "total_steps": 29704, "loss": 0.29577248096466063, "lr": 4.379189322899159e-06, "epoch": 1.079315917048209, "percentage": 53.97, "elapsed_time": "1 day, 10:26:28", "remaining_time": "1 day, 5:22:45", "throughput": 5653.07, "total_tokens": 700917296}
+{"current_steps": 16035, "total_steps": 29704, "loss": 0.30431132316589354, "lr": 4.376565787727676e-06, "epoch": 1.0796525720441692, "percentage": 53.98, "elapsed_time": "1 day, 10:27:00", "remaining_time": "1 day, 5:22:00", "throughput": 5653.44, "total_tokens": 701140832}
+{"current_steps": 16040, "total_steps": 29704, "loss": 0.30581619739532473, "lr": 4.373942426897581e-06, "epoch": 1.0799892270401292, "percentage": 54.0, "elapsed_time": "1 day, 10:27:36", "remaining_time": "1 day, 5:21:20", "throughput": 5653.42, "total_tokens": 701344984}
+{"current_steps": 16045, "total_steps": 29704, "loss": 0.32230820655822756, "lr": 4.37131924114249e-06, "epoch": 1.0803258820360895, "percentage": 54.02, "elapsed_time": "1 day, 10:28:16", "remaining_time": "1 day, 5:20:42", "throughput": 5653.27, "total_tokens": 701550320}
+{"current_steps": 16050, "total_steps": 29704, "loss": 0.3164224624633789, "lr": 4.368696231195968e-06, "epoch": 1.0806625370320495, "percentage": 54.03, "elapsed_time": "1 day, 10:28:53", "remaining_time": "1 day, 5:20:02", "throughput": 5653.36, "total_tokens": 701773408}
+{"current_steps": 16055, "total_steps": 29704, "loss": 0.29906353950500486, "lr": 4.366073397791534e-06, "epoch": 1.0809991920280098, "percentage": 54.05, "elapsed_time": "1 day, 10:29:31", "remaining_time": "1 day, 5:19:23", "throughput": 5653.36, "total_tokens": 701989056}
+{"current_steps": 16060, "total_steps": 29704, "loss": 0.28748531341552735, "lr": 4.363450741662651e-06, "epoch": 1.0813358470239698, "percentage": 54.07, "elapsed_time": "1 day, 10:30:08", "remaining_time": "1 day, 5:18:43", "throughput": 5653.44, "total_tokens": 702206288}
+{"current_steps": 16065, "total_steps": 29704, "loss": 0.2912368059158325, "lr": 4.360828263542739e-06, "epoch": 1.08167250201993, "percentage": 54.08, "elapsed_time": "1 day, 10:30:48", "remaining_time": "1 day, 5:18:05", "throughput": 5653.41, "total_tokens": 702427720}
+{"current_steps": 16070, "total_steps": 29704, "loss": 0.3195831298828125, "lr": 4.358205964165164e-06, "epoch": 1.08200915701589, "percentage": 54.1, "elapsed_time": "1 day, 10:31:28", "remaining_time": "1 day, 5:17:28", "throughput": 5653.35, "total_tokens": 702647704}
+{"current_steps": 16075, "total_steps": 29704, "loss": 0.32871217727661134, "lr": 4.355583844263247e-06, "epoch": 1.0823458120118503, "percentage": 54.12, "elapsed_time": "1 day, 10:32:14", "remaining_time": "1 day, 5:16:55", "throughput": 5653.09, "total_tokens": 702872568}
+{"current_steps": 16080, "total_steps": 29704, "loss": 0.3125025272369385, "lr": 4.352961904570251e-06, "epoch": 1.0826824670078103, "percentage": 54.13, "elapsed_time": "1 day, 10:32:54", "remaining_time": "1 day, 5:16:17", "throughput": 5653.01, "total_tokens": 703089672}
+{"current_steps": 16085, "total_steps": 29704, "loss": 0.31962432861328127, "lr": 4.350340145819397e-06, "epoch": 1.0830191220037706, "percentage": 54.15, "elapsed_time": "1 day, 10:33:27", "remaining_time": "1 day, 5:15:34", "throughput": 5653.23, "total_tokens": 703303976}
+{"current_steps": 16090, "total_steps": 29704, "loss": 0.31935825347900393, "lr": 4.347718568743847e-06, "epoch": 1.0833557769997306, "percentage": 54.17, "elapsed_time": "1 day, 10:34:07", "remaining_time": "1 day, 5:14:57", "throughput": 5653.17, "total_tokens": 703525192}
+{"current_steps": 16095, "total_steps": 29704, "loss": 0.32219245433807375, "lr": 4.34509717407672e-06, "epoch": 1.0836924319956909, "percentage": 54.18, "elapsed_time": "1 day, 10:34:45", "remaining_time": "1 day, 5:14:17", "throughput": 5653.36, "total_tokens": 703758184}
+{"current_steps": 16100, "total_steps": 29704, "loss": 0.3197646141052246, "lr": 4.342475962551081e-06, "epoch": 1.084029086991651, "percentage": 54.2, "elapsed_time": "1 day, 10:35:20", "remaining_time": "1 day, 5:13:35", "throughput": 5653.5, "total_tokens": 703976600}
+{"current_steps": 16105, "total_steps": 29704, "loss": 0.2914067268371582, "lr": 4.339854934899943e-06, "epoch": 1.0843657419876112, "percentage": 54.22, "elapsed_time": "1 day, 10:35:58", "remaining_time": "1 day, 5:12:56", "throughput": 5653.5, "total_tokens": 704189712}
+{"current_steps": 16110, "total_steps": 29704, "loss": 0.2758800983428955, "lr": 4.337234091856265e-06, "epoch": 1.0847023969835712, "percentage": 54.24, "elapsed_time": "1 day, 10:36:38", "remaining_time": "1 day, 5:12:19", "throughput": 5653.51, "total_tokens": 704420104}
+{"current_steps": 16115, "total_steps": 29704, "loss": 0.3056293249130249, "lr": 4.334613434152961e-06, "epoch": 1.0850390519795314, "percentage": 54.25, "elapsed_time": "1 day, 10:37:13", "remaining_time": "1 day, 5:11:37", "throughput": 5653.59, "total_tokens": 704628384}
+{"current_steps": 16120, "total_steps": 29704, "loss": 0.3311766147613525, "lr": 4.331992962522888e-06, "epoch": 1.0853757069754915, "percentage": 54.27, "elapsed_time": "1 day, 10:37:47", "remaining_time": "1 day, 5:10:54", "throughput": 5653.82, "total_tokens": 704848816}
+{"current_steps": 16125, "total_steps": 29704, "loss": 0.3087399244308472, "lr": 4.329372677698854e-06, "epoch": 1.0857123619714517, "percentage": 54.29, "elapsed_time": "1 day, 10:38:25", "remaining_time": "1 day, 5:10:15", "throughput": 5653.94, "total_tokens": 705076168}
+{"current_steps": 16130, "total_steps": 29704, "loss": 0.3147927761077881, "lr": 4.3267525804136114e-06, "epoch": 1.0860490169674117, "percentage": 54.3, "elapsed_time": "1 day, 10:39:03", "remaining_time": "1 day, 5:09:36", "throughput": 5653.97, "total_tokens": 705296784}
+{"current_steps": 16135, "total_steps": 29704, "loss": 0.3137171268463135, "lr": 4.324132671399864e-06, "epoch": 1.086385671963372, "percentage": 54.32, "elapsed_time": "1 day, 10:39:45", "remaining_time": "1 day, 5:09:00", "throughput": 5653.69, "total_tokens": 705500048}
+{"current_steps": 16140, "total_steps": 29704, "loss": 0.3282973289489746, "lr": 4.321512951390258e-06, "epoch": 1.086722326959332, "percentage": 54.34, "elapsed_time": "1 day, 10:40:23", "remaining_time": "1 day, 5:08:21", "throughput": 5653.73, "total_tokens": 705720120}
+{"current_steps": 16145, "total_steps": 29704, "loss": 0.3061366081237793, "lr": 4.318893421117393e-06, "epoch": 1.0870589819552923, "percentage": 54.35, "elapsed_time": "1 day, 10:41:02", "remaining_time": "1 day, 5:07:42", "throughput": 5653.76, "total_tokens": 705941352}
+{"current_steps": 16150, "total_steps": 29704, "loss": 0.2968632698059082, "lr": 4.31627408131381e-06, "epoch": 1.0873956369512523, "percentage": 54.37, "elapsed_time": "1 day, 10:41:48", "remaining_time": "1 day, 5:07:09", "throughput": 5653.36, "total_tokens": 706150776}
+{"current_steps": 16155, "total_steps": 29704, "loss": 0.2813676357269287, "lr": 4.313654932712001e-06, "epoch": 1.0877322919472125, "percentage": 54.39, "elapsed_time": "1 day, 10:42:25", "remaining_time": "1 day, 5:06:30", "throughput": 5653.42, "total_tokens": 706367696}
+{"current_steps": 16160, "total_steps": 29704, "loss": 0.3084303617477417, "lr": 4.3110359760444e-06, "epoch": 1.0880689469431726, "percentage": 54.4, "elapsed_time": "1 day, 10:42:57", "remaining_time": "1 day, 5:05:45", "throughput": 5653.82, "total_tokens": 706598168}
+{"current_steps": 16165, "total_steps": 29704, "loss": 0.2846923589706421, "lr": 4.308417212043392e-06, "epoch": 1.0884056019391328, "percentage": 54.42, "elapsed_time": "1 day, 10:43:35", "remaining_time": "1 day, 5:05:06", "throughput": 5653.79, "total_tokens": 706810256}
+{"current_steps": 16170, "total_steps": 29704, "loss": 0.2945555210113525, "lr": 4.305798641441304e-06, "epoch": 1.0887422569350929, "percentage": 54.44, "elapsed_time": "1 day, 10:44:17", "remaining_time": "1 day, 5:04:30", "throughput": 5653.72, "total_tokens": 707037952}
+{"current_steps": 16175, "total_steps": 29704, "loss": 0.3075662851333618, "lr": 4.303180264970416e-06, "epoch": 1.089078911931053, "percentage": 54.45, "elapsed_time": "1 day, 10:45:03", "remaining_time": "1 day, 5:03:57", "throughput": 5653.35, "total_tokens": 707251272}
+{"current_steps": 16180, "total_steps": 29704, "loss": 0.2843637466430664, "lr": 4.3005620833629415e-06, "epoch": 1.0894155669270131, "percentage": 54.47, "elapsed_time": "1 day, 10:45:44", "remaining_time": "1 day, 5:03:21", "throughput": 5653.22, "total_tokens": 707466896}
+{"current_steps": 16185, "total_steps": 29704, "loss": 0.3128535270690918, "lr": 4.297944097351053e-06, "epoch": 1.0897522219229734, "percentage": 54.49, "elapsed_time": "1 day, 10:46:27", "remaining_time": "1 day, 5:02:46", "throughput": 5652.89, "total_tokens": 707669312}
+{"current_steps": 16190, "total_steps": 29704, "loss": 0.32403197288513186, "lr": 4.295326307666856e-06, "epoch": 1.0900888769189334, "percentage": 54.5, "elapsed_time": "1 day, 10:47:03", "remaining_time": "1 day, 5:02:05", "throughput": 5653.11, "total_tokens": 707901856}
+{"current_steps": 16195, "total_steps": 29704, "loss": 0.3054851531982422, "lr": 4.292708715042411e-06, "epoch": 1.0904255319148937, "percentage": 54.52, "elapsed_time": "1 day, 10:47:42", "remaining_time": "1 day, 5:01:27", "throughput": 5653.1, "total_tokens": 708121472}
+{"current_steps": 16200, "total_steps": 29704, "loss": 0.32215137481689454, "lr": 4.290091320209718e-06, "epoch": 1.0907621869108537, "percentage": 54.54, "elapsed_time": "1 day, 10:48:17", "remaining_time": "1 day, 5:00:45", "throughput": 5653.35, "total_tokens": 708352952}
+{"current_steps": 16205, "total_steps": 29704, "loss": 0.3153803825378418, "lr": 4.287474123900726e-06, "epoch": 1.091098841906814, "percentage": 54.55, "elapsed_time": "1 day, 10:48:53", "remaining_time": "1 day, 5:00:05", "throughput": 5653.37, "total_tokens": 708559152}
+{"current_steps": 16210, "total_steps": 29704, "loss": 0.3124639272689819, "lr": 4.284857126847321e-06, "epoch": 1.091435496902774, "percentage": 54.57, "elapsed_time": "1 day, 10:49:35", "remaining_time": "1 day, 4:59:28", "throughput": 5653.21, "total_tokens": 708773424}
+{"current_steps": 16215, "total_steps": 29704, "loss": 0.32616453170776366, "lr": 4.28224032978134e-06, "epoch": 1.0917721518987342, "percentage": 54.59, "elapsed_time": "1 day, 10:50:15", "remaining_time": "1 day, 4:58:51", "throughput": 5653.21, "total_tokens": 709001032}
+{"current_steps": 16220, "total_steps": 29704, "loss": 0.29432287216186526, "lr": 4.279623733434561e-06, "epoch": 1.0921088068946943, "percentage": 54.61, "elapsed_time": "1 day, 10:50:53", "remaining_time": "1 day, 4:58:11", "throughput": 5653.27, "total_tokens": 709220648}
+{"current_steps": 16225, "total_steps": 29704, "loss": 0.28021812438964844, "lr": 4.277007338538711e-06, "epoch": 1.0924454618906545, "percentage": 54.62, "elapsed_time": "1 day, 10:51:29", "remaining_time": "1 day, 4:57:30", "throughput": 5653.54, "total_tokens": 709458880}
+{"current_steps": 16230, "total_steps": 29704, "loss": 0.278457498550415, "lr": 4.27439114582545e-06, "epoch": 1.0927821168866145, "percentage": 54.64, "elapsed_time": "1 day, 10:52:08", "remaining_time": "1 day, 4:56:52", "throughput": 5653.54, "total_tokens": 709678888}
+{"current_steps": 16235, "total_steps": 29704, "loss": 0.3240532636642456, "lr": 4.271775156026395e-06, "epoch": 1.0931187718825748, "percentage": 54.66, "elapsed_time": "1 day, 10:52:50", "remaining_time": "1 day, 4:56:16", "throughput": 5653.25, "total_tokens": 709881120}
+{"current_steps": 16240, "total_steps": 29704, "loss": 0.30025343894958495, "lr": 4.269159369873093e-06, "epoch": 1.0934554268785348, "percentage": 54.67, "elapsed_time": "1 day, 10:53:28", "remaining_time": "1 day, 4:55:37", "throughput": 5653.18, "total_tokens": 710087584}
+{"current_steps": 16245, "total_steps": 29704, "loss": 0.30265159606933595, "lr": 4.266543788097043e-06, "epoch": 1.093792081874495, "percentage": 54.69, "elapsed_time": "1 day, 10:54:02", "remaining_time": "1 day, 4:54:55", "throughput": 5653.52, "total_tokens": 710323016}
+{"current_steps": 16250, "total_steps": 29704, "loss": 0.34309751987457277, "lr": 4.263928411429683e-06, "epoch": 1.094128736870455, "percentage": 54.71, "elapsed_time": "1 day, 10:54:45", "remaining_time": "1 day, 4:54:20", "throughput": 5653.26, "total_tokens": 710535008}
+{"current_steps": 16255, "total_steps": 29704, "loss": 0.32427158355712893, "lr": 4.261313240602398e-06, "epoch": 1.0944653918664153, "percentage": 54.72, "elapsed_time": "1 day, 10:55:21", "remaining_time": "1 day, 4:53:38", "throughput": 5653.35, "total_tokens": 710748160}
+{"current_steps": 16260, "total_steps": 29704, "loss": 0.3176891326904297, "lr": 4.2586982763465064e-06, "epoch": 1.0948020468623754, "percentage": 54.74, "elapsed_time": "1 day, 10:55:55", "remaining_time": "1 day, 4:52:56", "throughput": 5653.65, "total_tokens": 710976808}
+{"current_steps": 16265, "total_steps": 29704, "loss": 0.30221972465515134, "lr": 4.256083519393281e-06, "epoch": 1.0951387018583356, "percentage": 54.76, "elapsed_time": "1 day, 10:56:31", "remaining_time": "1 day, 4:52:15", "throughput": 5653.76, "total_tokens": 711195160}
+{"current_steps": 16270, "total_steps": 29704, "loss": 0.32169513702392577, "lr": 4.253468970473925e-06, "epoch": 1.0954753568542956, "percentage": 54.77, "elapsed_time": "1 day, 10:57:03", "remaining_time": "1 day, 4:51:31", "throughput": 5654.1, "total_tokens": 711418152}
+{"current_steps": 16275, "total_steps": 29704, "loss": 0.345171594619751, "lr": 4.250854630319593e-06, "epoch": 1.095812011850256, "percentage": 54.79, "elapsed_time": "1 day, 10:57:44", "remaining_time": "1 day, 4:50:54", "throughput": 5654.02, "total_tokens": 711640424}
+{"current_steps": 16280, "total_steps": 29704, "loss": 0.3070409059524536, "lr": 4.248240499661373e-06, "epoch": 1.096148666846216, "percentage": 54.81, "elapsed_time": "1 day, 10:58:23", "remaining_time": "1 day, 4:50:16", "throughput": 5653.92, "total_tokens": 711848088}
+{"current_steps": 16285, "total_steps": 29704, "loss": 0.2884035587310791, "lr": 4.245626579230302e-06, "epoch": 1.0964853218421762, "percentage": 54.82, "elapsed_time": "1 day, 10:58:55", "remaining_time": "1 day, 4:49:31", "throughput": 5654.26, "total_tokens": 712069664}
+{"current_steps": 16290, "total_steps": 29704, "loss": 0.296952223777771, "lr": 4.243012869757351e-06, "epoch": 1.0968219768381362, "percentage": 54.84, "elapsed_time": "1 day, 10:59:35", "remaining_time": "1 day, 4:48:54", "throughput": 5654.32, "total_tokens": 712306208}
+{"current_steps": 16295, "total_steps": 29704, "loss": 0.31108913421630857, "lr": 4.240399371973438e-06, "epoch": 1.0971586318340965, "percentage": 54.86, "elapsed_time": "1 day, 11:00:15", "remaining_time": "1 day, 4:48:16", "throughput": 5654.36, "total_tokens": 712536912}
+{"current_steps": 16300, "total_steps": 29704, "loss": 0.3221109390258789, "lr": 4.2377860866094165e-06, "epoch": 1.0974952868300565, "percentage": 54.87, "elapsed_time": "1 day, 11:00:50", "remaining_time": "1 day, 4:47:35", "throughput": 5654.53, "total_tokens": 712758800}
+{"current_steps": 16305, "total_steps": 29704, "loss": 0.3079149007797241, "lr": 4.235173014396088e-06, "epoch": 1.0978319418260167, "percentage": 54.89, "elapsed_time": "1 day, 11:01:26", "remaining_time": "1 day, 4:46:54", "throughput": 5654.66, "total_tokens": 712977904}
+{"current_steps": 16310, "total_steps": 29704, "loss": 0.3280055046081543, "lr": 4.2325601560641854e-06, "epoch": 1.0981685968219768, "percentage": 54.91, "elapsed_time": "1 day, 11:02:07", "remaining_time": "1 day, 4:46:17", "throughput": 5654.63, "total_tokens": 713204864}
+{"current_steps": 16315, "total_steps": 29704, "loss": 0.3023355484008789, "lr": 4.229947512344389e-06, "epoch": 1.098505251817937, "percentage": 54.93, "elapsed_time": "1 day, 11:02:48", "remaining_time": "1 day, 4:45:40", "throughput": 5654.54, "total_tokens": 713422520}
+{"current_steps": 16320, "total_steps": 29704, "loss": 0.3024130344390869, "lr": 4.227335083967314e-06, "epoch": 1.098841906813897, "percentage": 54.94, "elapsed_time": "1 day, 11:03:28", "remaining_time": "1 day, 4:45:03", "throughput": 5654.45, "total_tokens": 713640344}
+{"current_steps": 16325, "total_steps": 29704, "loss": 0.3010653018951416, "lr": 4.224722871663522e-06, "epoch": 1.0991785618098573, "percentage": 54.96, "elapsed_time": "1 day, 11:04:01", "remaining_time": "1 day, 4:44:19", "throughput": 5654.74, "total_tokens": 713862120}
+{"current_steps": 16330, "total_steps": 29704, "loss": 0.3363258600234985, "lr": 4.222110876163503e-06, "epoch": 1.0995152168058173, "percentage": 54.98, "elapsed_time": "1 day, 11:04:39", "remaining_time": "1 day, 4:43:40", "throughput": 5654.86, "total_tokens": 714091328}
+{"current_steps": 16335, "total_steps": 29704, "loss": 0.2941534757614136, "lr": 4.2194990981977e-06, "epoch": 1.0998518718017776, "percentage": 54.99, "elapsed_time": "1 day, 11:05:15", "remaining_time": "1 day, 4:43:00", "throughput": 5655.01, "total_tokens": 714318104}
+{"current_steps": 16340, "total_steps": 29704, "loss": 0.2995609760284424, "lr": 4.216887538496482e-06, "epoch": 1.1001885267977376, "percentage": 55.01, "elapsed_time": "1 day, 11:05:56", "remaining_time": "1 day, 4:42:23", "throughput": 5654.91, "total_tokens": 714533544}
+{"current_steps": 16345, "total_steps": 29704, "loss": 0.3597081184387207, "lr": 4.2142761977901695e-06, "epoch": 1.1005251817936978, "percentage": 55.03, "elapsed_time": "1 day, 11:06:32", "remaining_time": "1 day, 4:41:42", "throughput": 5655.0, "total_tokens": 714751432}
+{"current_steps": 16350, "total_steps": 29704, "loss": 0.3166736364364624, "lr": 4.21166507680901e-06, "epoch": 1.1008618367896579, "percentage": 55.04, "elapsed_time": "1 day, 11:07:02", "remaining_time": "1 day, 4:40:56", "throughput": 5655.35, "total_tokens": 714962360}
+{"current_steps": 16355, "total_steps": 29704, "loss": 0.31155028343200686, "lr": 4.209054176283202e-06, "epoch": 1.1011984917856181, "percentage": 55.06, "elapsed_time": "1 day, 11:07:36", "remaining_time": "1 day, 4:40:13", "throughput": 5655.5, "total_tokens": 715172784}
+{"current_steps": 16360, "total_steps": 29704, "loss": 0.3161408185958862, "lr": 4.206443496942868e-06, "epoch": 1.1015351467815782, "percentage": 55.08, "elapsed_time": "1 day, 11:08:15", "remaining_time": "1 day, 4:39:35", "throughput": 5655.47, "total_tokens": 715390344}
+{"current_steps": 16365, "total_steps": 29704, "loss": 0.2889822006225586, "lr": 4.203833039518081e-06, "epoch": 1.1018718017775384, "percentage": 55.09, "elapsed_time": "1 day, 11:08:53", "remaining_time": "1 day, 4:38:56", "throughput": 5655.44, "total_tokens": 715601072}
+{"current_steps": 16370, "total_steps": 29704, "loss": 0.29911801815032957, "lr": 4.201222804738845e-06, "epoch": 1.1022084567734984, "percentage": 55.11, "elapsed_time": "1 day, 11:09:27", "remaining_time": "1 day, 4:38:14", "throughput": 5655.74, "total_tokens": 715832864}
+{"current_steps": 16375, "total_steps": 29704, "loss": 0.28797073364257814, "lr": 4.198612793335107e-06, "epoch": 1.1025451117694587, "percentage": 55.13, "elapsed_time": "1 day, 11:10:11", "remaining_time": "1 day, 4:37:39", "throughput": 5655.54, "total_tokens": 716055080}
+{"current_steps": 16380, "total_steps": 29704, "loss": 0.3057976722717285, "lr": 4.196003006036743e-06, "epoch": 1.1028817667654187, "percentage": 55.14, "elapsed_time": "1 day, 11:10:48", "remaining_time": "1 day, 4:36:59", "throughput": 5655.55, "total_tokens": 716266872}
+{"current_steps": 16385, "total_steps": 29704, "loss": 0.3109589576721191, "lr": 4.193393443573576e-06, "epoch": 1.103218421761379, "percentage": 55.16, "elapsed_time": "1 day, 11:11:24", "remaining_time": "1 day, 4:36:18", "throughput": 5655.71, "total_tokens": 716490328}
+{"current_steps": 16390, "total_steps": 29704, "loss": 0.3338963985443115, "lr": 4.19078410667536e-06, "epoch": 1.103555076757339, "percentage": 55.18, "elapsed_time": "1 day, 11:12:04", "remaining_time": "1 day, 4:35:41", "throughput": 5655.58, "total_tokens": 716702856}
+{"current_steps": 16395, "total_steps": 29704, "loss": 0.32840120792388916, "lr": 4.188174996071788e-06, "epoch": 1.1038917317532992, "percentage": 55.19, "elapsed_time": "1 day, 11:12:38", "remaining_time": "1 day, 4:34:58", "throughput": 5655.59, "total_tokens": 716893944}
+{"current_steps": 16400, "total_steps": 29704, "loss": 0.30612549781799314, "lr": 4.18556611249249e-06, "epoch": 1.1042283867492593, "percentage": 55.21, "elapsed_time": "1 day, 11:13:12", "remaining_time": "1 day, 4:34:16", "throughput": 5655.78, "total_tokens": 717111816}
+{"current_steps": 16405, "total_steps": 29704, "loss": 0.3163015127182007, "lr": 4.182957456667032e-06, "epoch": 1.1045650417452195, "percentage": 55.23, "elapsed_time": "1 day, 11:13:56", "remaining_time": "1 day, 4:33:42", "throughput": 5655.59, "total_tokens": 717335736}
+{"current_steps": 16410, "total_steps": 29704, "loss": 0.3357135057449341, "lr": 4.1803490293249145e-06, "epoch": 1.1049016967411796, "percentage": 55.25, "elapsed_time": "1 day, 11:14:33", "remaining_time": "1 day, 4:33:01", "throughput": 5655.7, "total_tokens": 717557144}
+{"current_steps": 16415, "total_steps": 29704, "loss": 0.3057410717010498, "lr": 4.1777408311955785e-06, "epoch": 1.1052383517371398, "percentage": 55.26, "elapsed_time": "1 day, 11:15:09", "remaining_time": "1 day, 4:32:21", "throughput": 5655.7, "total_tokens": 717761472}
+{"current_steps": 16420, "total_steps": 29704, "loss": 0.33652937412261963, "lr": 4.175132863008395e-06, "epoch": 1.1055750067330998, "percentage": 55.28, "elapsed_time": "1 day, 11:15:49", "remaining_time": "1 day, 4:31:43", "throughput": 5655.69, "total_tokens": 717986048}
+{"current_steps": 16425, "total_steps": 29704, "loss": 0.3076220273971558, "lr": 4.17252512549268e-06, "epoch": 1.10591166172906, "percentage": 55.3, "elapsed_time": "1 day, 11:16:27", "remaining_time": "1 day, 4:31:04", "throughput": 5655.5, "total_tokens": 718179376}
+{"current_steps": 16430, "total_steps": 29704, "loss": 0.3038554906845093, "lr": 4.169917619377671e-06, "epoch": 1.10624831672502, "percentage": 55.31, "elapsed_time": "1 day, 11:17:10", "remaining_time": "1 day, 4:30:29", "throughput": 5655.26, "total_tokens": 718392416}
+{"current_steps": 16435, "total_steps": 29704, "loss": 0.34373929500579836, "lr": 4.167310345392557e-06, "epoch": 1.1065849717209804, "percentage": 55.33, "elapsed_time": "1 day, 11:17:46", "remaining_time": "1 day, 4:29:49", "throughput": 5655.34, "total_tokens": 718606440}
+{"current_steps": 16440, "total_steps": 29704, "loss": 0.3425747871398926, "lr": 4.1647033042664455e-06, "epoch": 1.1069216267169404, "percentage": 55.35, "elapsed_time": "1 day, 11:18:25", "remaining_time": "1 day, 4:29:10", "throughput": 5655.31, "total_tokens": 718818792}
+{"current_steps": 16445, "total_steps": 29704, "loss": 0.3010833263397217, "lr": 4.162096496728394e-06, "epoch": 1.1072582817129006, "percentage": 55.36, "elapsed_time": "1 day, 11:19:06", "remaining_time": "1 day, 4:28:33", "throughput": 5655.21, "total_tokens": 719039416}
+{"current_steps": 16450, "total_steps": 29704, "loss": 0.2992356300354004, "lr": 4.1594899235073835e-06, "epoch": 1.1075949367088607, "percentage": 55.38, "elapsed_time": "1 day, 11:19:50", "remaining_time": "1 day, 4:27:59", "throughput": 5655.05, "total_tokens": 719268528}
+{"current_steps": 16455, "total_steps": 29704, "loss": 0.33159723281860354, "lr": 4.156883585332338e-06, "epoch": 1.107931591704821, "percentage": 55.4, "elapsed_time": "1 day, 11:20:29", "remaining_time": "1 day, 4:27:20", "throughput": 5655.13, "total_tokens": 719497592}
+{"current_steps": 16460, "total_steps": 29704, "loss": 0.31698899269104003, "lr": 4.154277482932107e-06, "epoch": 1.108268246700781, "percentage": 55.41, "elapsed_time": "1 day, 11:21:03", "remaining_time": "1 day, 4:26:38", "throughput": 5655.22, "total_tokens": 719703328}
+{"current_steps": 16465, "total_steps": 29704, "loss": 0.30119948387145995, "lr": 4.151671617035482e-06, "epoch": 1.1086049016967412, "percentage": 55.43, "elapsed_time": "1 day, 11:21:43", "remaining_time": "1 day, 4:26:00", "throughput": 5655.12, "total_tokens": 719915352}
+{"current_steps": 16470, "total_steps": 29704, "loss": 0.2912993669509888, "lr": 4.149065988371184e-06, "epoch": 1.1089415566927012, "percentage": 55.45, "elapsed_time": "1 day, 11:22:16", "remaining_time": "1 day, 4:25:17", "throughput": 5655.37, "total_tokens": 720134040}
+{"current_steps": 16475, "total_steps": 29704, "loss": 0.30213978290557864, "lr": 4.14646059766787e-06, "epoch": 1.1092782116886615, "percentage": 55.46, "elapsed_time": "1 day, 11:22:49", "remaining_time": "1 day, 4:24:34", "throughput": 5655.49, "total_tokens": 720337824}
+{"current_steps": 16480, "total_steps": 29704, "loss": 0.29933414459228513, "lr": 4.143855445654127e-06, "epoch": 1.1096148666846215, "percentage": 55.48, "elapsed_time": "1 day, 11:23:29", "remaining_time": "1 day, 4:23:56", "throughput": 5655.52, "total_tokens": 720567936}
+{"current_steps": 16485, "total_steps": 29704, "loss": 0.2931232213973999, "lr": 4.14125053305848e-06, "epoch": 1.1099515216805818, "percentage": 55.5, "elapsed_time": "1 day, 11:24:06", "remaining_time": "1 day, 4:23:17", "throughput": 5655.46, "total_tokens": 720770472}
+{"current_steps": 16490, "total_steps": 29704, "loss": 0.30602593421936036, "lr": 4.1386458606093805e-06, "epoch": 1.1102881766765418, "percentage": 55.51, "elapsed_time": "1 day, 11:24:47", "remaining_time": "1 day, 4:22:39", "throughput": 5655.39, "total_tokens": 720990928}
+{"current_steps": 16495, "total_steps": 29704, "loss": 0.3390794038772583, "lr": 4.1360414290352214e-06, "epoch": 1.110624831672502, "percentage": 55.53, "elapsed_time": "1 day, 11:25:31", "remaining_time": "1 day, 4:22:06", "throughput": 5655.23, "total_tokens": 721222304}
+{"current_steps": 16500, "total_steps": 29704, "loss": 0.30769715309143064, "lr": 4.1334372390643205e-06, "epoch": 1.110961486668462, "percentage": 55.55, "elapsed_time": "1 day, 11:26:12", "remaining_time": "1 day, 4:21:28", "throughput": 5655.23, "total_tokens": 721451296}
+{"current_steps": 16505, "total_steps": 29704, "loss": 0.32526416778564454, "lr": 4.130833291424934e-06, "epoch": 1.1112981416644223, "percentage": 55.56, "elapsed_time": "1 day, 11:26:49", "remaining_time": "1 day, 4:20:49", "throughput": 5655.31, "total_tokens": 721672688}
+{"current_steps": 16510, "total_steps": 29704, "loss": 0.30047011375427246, "lr": 4.128229586845244e-06, "epoch": 1.1116347966603823, "percentage": 55.58, "elapsed_time": "1 day, 11:27:32", "remaining_time": "1 day, 4:20:13", "throughput": 5655.15, "total_tokens": 721893296}
+{"current_steps": 16515, "total_steps": 29704, "loss": 0.2854645013809204, "lr": 4.125626126053372e-06, "epoch": 1.1119714516563426, "percentage": 55.6, "elapsed_time": "1 day, 11:28:13", "remaining_time": "1 day, 4:19:36", "throughput": 5655.08, "total_tokens": 722117224}
+{"current_steps": 16520, "total_steps": 29704, "loss": 0.31157827377319336, "lr": 4.123022909777364e-06, "epoch": 1.1123081066523026, "percentage": 55.62, "elapsed_time": "1 day, 11:28:44", "remaining_time": "1 day, 4:18:51", "throughput": 5655.46, "total_tokens": 722338424}
+{"current_steps": 16525, "total_steps": 29704, "loss": 0.31059443950653076, "lr": 4.120419938745205e-06, "epoch": 1.1126447616482629, "percentage": 55.63, "elapsed_time": "1 day, 11:29:28", "remaining_time": "1 day, 4:18:17", "throughput": 5655.32, "total_tokens": 722572192}
+{"current_steps": 16530, "total_steps": 29704, "loss": 0.30442132949829104, "lr": 4.1178172136848035e-06, "epoch": 1.112981416644223, "percentage": 55.65, "elapsed_time": "1 day, 11:30:05", "remaining_time": "1 day, 4:17:38", "throughput": 5655.28, "total_tokens": 722777960}
+{"current_steps": 16535, "total_steps": 29704, "loss": 0.29233555793762206, "lr": 4.115214735324007e-06, "epoch": 1.1133180716401831, "percentage": 55.67, "elapsed_time": "1 day, 11:30:45", "remaining_time": "1 day, 4:17:00", "throughput": 5655.36, "total_tokens": 723011784}
+{"current_steps": 16540, "total_steps": 29704, "loss": 0.2731541872024536, "lr": 4.112612504390586e-06, "epoch": 1.1136547266361432, "percentage": 55.68, "elapsed_time": "1 day, 11:31:27", "remaining_time": "1 day, 4:16:24", "throughput": 5655.23, "total_tokens": 723233976}
+{"current_steps": 16545, "total_steps": 29704, "loss": 0.3273204803466797, "lr": 4.11001052161225e-06, "epoch": 1.1139913816321034, "percentage": 55.7, "elapsed_time": "1 day, 11:32:04", "remaining_time": "1 day, 4:15:44", "throughput": 5655.3, "total_tokens": 723450040}
+{"current_steps": 16550, "total_steps": 29704, "loss": 0.3083867788314819, "lr": 4.107408787716632e-06, "epoch": 1.1143280366280637, "percentage": 55.72, "elapsed_time": "1 day, 11:32:46", "remaining_time": "1 day, 4:15:08", "throughput": 5655.17, "total_tokens": 723674424}
+{"current_steps": 16555, "total_steps": 29704, "loss": 0.30321598052978516, "lr": 4.104807303431303e-06, "epoch": 1.1146646916240237, "percentage": 55.73, "elapsed_time": "1 day, 11:33:33", "remaining_time": "1 day, 4:14:36", "throughput": 5654.85, "total_tokens": 723898176}
+{"current_steps": 16560, "total_steps": 29704, "loss": 0.32450571060180666, "lr": 4.102206069483754e-06, "epoch": 1.1150013466199837, "percentage": 55.75, "elapsed_time": "1 day, 11:34:10", "remaining_time": "1 day, 4:13:56", "throughput": 5654.87, "total_tokens": 724110632}
+{"current_steps": 16565, "total_steps": 29704, "loss": 0.28465938568115234, "lr": 4.099605086601417e-06, "epoch": 1.115338001615944, "percentage": 55.77, "elapsed_time": "1 day, 11:34:58", "remaining_time": "1 day, 4:13:25", "throughput": 5654.33, "total_tokens": 724312128}
+{"current_steps": 16570, "total_steps": 29704, "loss": 0.2872473239898682, "lr": 4.097004355511645e-06, "epoch": 1.1156746566119042, "percentage": 55.78, "elapsed_time": "1 day, 11:35:35", "remaining_time": "1 day, 4:12:44", "throughput": 5654.44, "total_tokens": 724532224}
+{"current_steps": 16575, "total_steps": 29704, "loss": 0.2786348819732666, "lr": 4.094403876941728e-06, "epoch": 1.1160113116078643, "percentage": 55.8, "elapsed_time": "1 day, 11:36:12", "remaining_time": "1 day, 4:12:04", "throughput": 5654.5, "total_tokens": 724749712}
+{"current_steps": 16580, "total_steps": 29704, "loss": 0.3147793531417847, "lr": 4.0918036516188755e-06, "epoch": 1.1163479666038243, "percentage": 55.82, "elapsed_time": "1 day, 11:36:55", "remaining_time": "1 day, 4:11:30", "throughput": 5654.17, "total_tokens": 724953832}
+{"current_steps": 16585, "total_steps": 29704, "loss": 0.3022450923919678, "lr": 4.0892036802702395e-06, "epoch": 1.1166846215997845, "percentage": 55.83, "elapsed_time": "1 day, 11:37:38", "remaining_time": "1 day, 4:10:54", "throughput": 5654.06, "total_tokens": 725178424}
+{"current_steps": 16590, "total_steps": 29704, "loss": 0.2855158090591431, "lr": 4.086603963622887e-06, "epoch": 1.1170212765957448, "percentage": 55.85, "elapsed_time": "1 day, 11:38:19", "remaining_time": "1 day, 4:10:17", "throughput": 5653.97, "total_tokens": 725402000}
+{"current_steps": 16595, "total_steps": 29704, "loss": 0.32064228057861327, "lr": 4.084004502403825e-06, "epoch": 1.1173579315917048, "percentage": 55.87, "elapsed_time": "1 day, 11:39:01", "remaining_time": "1 day, 4:09:41", "throughput": 5653.83, "total_tokens": 725622480}
+{"current_steps": 16600, "total_steps": 29704, "loss": 0.30630931854248045, "lr": 4.081405297339982e-06, "epoch": 1.1176945865876649, "percentage": 55.88, "elapsed_time": "1 day, 11:39:35", "remaining_time": "1 day, 4:08:59", "throughput": 5654.18, "total_tokens": 725856392}
+{"current_steps": 16605, "total_steps": 29704, "loss": 0.3028933048248291, "lr": 4.0788063491582205e-06, "epoch": 1.118031241583625, "percentage": 55.9, "elapsed_time": "1 day, 11:40:11", "remaining_time": "1 day, 4:08:18", "throughput": 5654.2, "total_tokens": 726063792}
+{"current_steps": 16610, "total_steps": 29704, "loss": 0.3076032638549805, "lr": 4.076207658585323e-06, "epoch": 1.1183678965795854, "percentage": 55.92, "elapsed_time": "1 day, 11:40:53", "remaining_time": "1 day, 4:07:42", "throughput": 5654.16, "total_tokens": 726293720}
+{"current_steps": 16615, "total_steps": 29704, "loss": 0.31174705028533933, "lr": 4.07360922634801e-06, "epoch": 1.1187045515755454, "percentage": 55.94, "elapsed_time": "1 day, 11:41:30", "remaining_time": "1 day, 4:07:02", "throughput": 5654.3, "total_tokens": 726522048}
+{"current_steps": 16620, "total_steps": 29704, "loss": 0.3052211284637451, "lr": 4.071011053172922e-06, "epoch": 1.1190412065715054, "percentage": 55.95, "elapsed_time": "1 day, 11:42:08", "remaining_time": "1 day, 4:06:23", "throughput": 5654.18, "total_tokens": 726722760}
+{"current_steps": 16625, "total_steps": 29704, "loss": 0.29850754737854, "lr": 4.068413139786631e-06, "epoch": 1.1193778615674657, "percentage": 55.97, "elapsed_time": "1 day, 11:42:47", "remaining_time": "1 day, 4:05:44", "throughput": 5654.15, "total_tokens": 726939496}
+{"current_steps": 16630, "total_steps": 29704, "loss": 0.30380845069885254, "lr": 4.065815486915635e-06, "epoch": 1.119714516563426, "percentage": 55.99, "elapsed_time": "1 day, 11:43:25", "remaining_time": "1 day, 4:05:05", "throughput": 5654.25, "total_tokens": 727169648}
+{"current_steps": 16635, "total_steps": 29704, "loss": 0.31917145252227785, "lr": 4.063218095286359e-06, "epoch": 1.120051171559386, "percentage": 56.0, "elapsed_time": "1 day, 11:44:02", "remaining_time": "1 day, 4:04:26", "throughput": 5654.41, "total_tokens": 727399600}
+{"current_steps": 16640, "total_steps": 29704, "loss": 0.30777583122253416, "lr": 4.0606209656251555e-06, "epoch": 1.120387826555346, "percentage": 56.02, "elapsed_time": "1 day, 11:44:39", "remaining_time": "1 day, 4:03:45", "throughput": 5654.55, "total_tokens": 727623584}
+{"current_steps": 16645, "total_steps": 29704, "loss": 0.32628979682922366, "lr": 4.058024098658304e-06, "epoch": 1.1207244815513062, "percentage": 56.04, "elapsed_time": "1 day, 11:45:21", "remaining_time": "1 day, 4:03:09", "throughput": 5654.47, "total_tokens": 727849768}
+{"current_steps": 16650, "total_steps": 29704, "loss": 0.33713221549987793, "lr": 4.055427495112008e-06, "epoch": 1.1210611365472665, "percentage": 56.05, "elapsed_time": "1 day, 11:46:04", "remaining_time": "1 day, 4:02:34", "throughput": 5654.3, "total_tokens": 728072752}
+{"current_steps": 16655, "total_steps": 29704, "loss": 0.28829092979431153, "lr": 4.052831155712403e-06, "epoch": 1.1213977915432265, "percentage": 56.07, "elapsed_time": "1 day, 11:46:38", "remaining_time": "1 day, 4:01:52", "throughput": 5654.53, "total_tokens": 728297032}
+{"current_steps": 16660, "total_steps": 29704, "loss": 0.27930502891540526, "lr": 4.050235081185544e-06, "epoch": 1.1217344465391865, "percentage": 56.09, "elapsed_time": "1 day, 11:47:17", "remaining_time": "1 day, 4:01:14", "throughput": 5654.53, "total_tokens": 728518648}
+{"current_steps": 16665, "total_steps": 29704, "loss": 0.31568381786346433, "lr": 4.047639272257415e-06, "epoch": 1.1220711015351468, "percentage": 56.1, "elapsed_time": "1 day, 11:48:01", "remaining_time": "1 day, 4:00:39", "throughput": 5654.4, "total_tokens": 728747776}
+{"current_steps": 16670, "total_steps": 29704, "loss": 0.2794900894165039, "lr": 4.045043729653927e-06, "epoch": 1.122407756531107, "percentage": 56.12, "elapsed_time": "1 day, 11:48:40", "remaining_time": "1 day, 4:00:00", "throughput": 5654.48, "total_tokens": 728975240}
+{"current_steps": 16675, "total_steps": 29704, "loss": 0.3211031436920166, "lr": 4.0424484541009175e-06, "epoch": 1.122744411527067, "percentage": 56.14, "elapsed_time": "1 day, 11:49:21", "remaining_time": "1 day, 3:59:24", "throughput": 5654.44, "total_tokens": 729207240}
+{"current_steps": 16680, "total_steps": 29704, "loss": 0.2974057197570801, "lr": 4.039853446324141e-06, "epoch": 1.1230810665230273, "percentage": 56.15, "elapsed_time": "1 day, 11:50:04", "remaining_time": "1 day, 3:58:48", "throughput": 5654.21, "total_tokens": 729416088}
+{"current_steps": 16685, "total_steps": 29704, "loss": 0.2974997043609619, "lr": 4.037258707049289e-06, "epoch": 1.1234177215189873, "percentage": 56.17, "elapsed_time": "1 day, 11:50:48", "remaining_time": "1 day, 3:58:14", "throughput": 5653.93, "total_tokens": 729633568}
+{"current_steps": 16690, "total_steps": 29704, "loss": 0.29579739570617675, "lr": 4.034664237001969e-06, "epoch": 1.1237543765149476, "percentage": 56.19, "elapsed_time": "1 day, 11:51:27", "remaining_time": "1 day, 3:57:35", "throughput": 5653.95, "total_tokens": 729855424}
+{"current_steps": 16695, "total_steps": 29704, "loss": 0.3193542003631592, "lr": 4.032070036907716e-06, "epoch": 1.1240910315109076, "percentage": 56.2, "elapsed_time": "1 day, 11:52:06", "remaining_time": "1 day, 3:56:57", "throughput": 5653.84, "total_tokens": 730062656}
+{"current_steps": 16700, "total_steps": 29704, "loss": 0.3203211784362793, "lr": 4.029476107491991e-06, "epoch": 1.1244276865068679, "percentage": 56.22, "elapsed_time": "1 day, 11:52:51", "remaining_time": "1 day, 3:56:23", "throughput": 5653.59, "total_tokens": 730282736}
+{"current_steps": 16705, "total_steps": 29704, "loss": 0.30711615085601807, "lr": 4.0268824494801784e-06, "epoch": 1.124764341502828, "percentage": 56.24, "elapsed_time": "1 day, 11:53:29", "remaining_time": "1 day, 3:55:44", "throughput": 5653.64, "total_tokens": 730503648}
+{"current_steps": 16710, "total_steps": 29704, "loss": 0.31215999126434324, "lr": 4.024289063597585e-06, "epoch": 1.1251009964987881, "percentage": 56.26, "elapsed_time": "1 day, 11:54:11", "remaining_time": "1 day, 3:55:08", "throughput": 5653.39, "total_tokens": 730706704}
+{"current_steps": 16715, "total_steps": 29704, "loss": 0.29782519340515134, "lr": 4.021695950569443e-06, "epoch": 1.1254376514947482, "percentage": 56.27, "elapsed_time": "1 day, 11:54:46", "remaining_time": "1 day, 3:54:26", "throughput": 5653.57, "total_tokens": 730929536}
+{"current_steps": 16720, "total_steps": 29704, "loss": 0.3227576017379761, "lr": 4.01910311112091e-06, "epoch": 1.1257743064907082, "percentage": 56.29, "elapsed_time": "1 day, 11:55:25", "remaining_time": "1 day, 3:53:48", "throughput": 5653.54, "total_tokens": 731146120}
+{"current_steps": 16725, "total_steps": 29704, "loss": 0.30303683280944826, "lr": 4.016510545977064e-06, "epoch": 1.1261109614866684, "percentage": 56.31, "elapsed_time": "1 day, 11:56:09", "remaining_time": "1 day, 3:53:13", "throughput": 5653.38, "total_tokens": 731374192}
+{"current_steps": 16730, "total_steps": 29704, "loss": 0.3152627944946289, "lr": 4.013918255862908e-06, "epoch": 1.1264476164826287, "percentage": 56.32, "elapsed_time": "1 day, 11:56:53", "remaining_time": "1 day, 3:52:39", "throughput": 5653.05, "total_tokens": 731580072}
+{"current_steps": 16735, "total_steps": 29704, "loss": 0.2968592643737793, "lr": 4.011326241503369e-06, "epoch": 1.1267842714785887, "percentage": 56.34, "elapsed_time": "1 day, 11:57:38", "remaining_time": "1 day, 3:52:05", "throughput": 5652.71, "total_tokens": 731789896}
+{"current_steps": 16740, "total_steps": 29704, "loss": 0.32601251602172854, "lr": 4.00873450362329e-06, "epoch": 1.127120926474549, "percentage": 56.36, "elapsed_time": "1 day, 11:58:14", "remaining_time": "1 day, 3:51:24", "throughput": 5652.87, "total_tokens": 732012656}
+{"current_steps": 16745, "total_steps": 29704, "loss": 0.31506369113922117, "lr": 4.0061430429474476e-06, "epoch": 1.127457581470509, "percentage": 56.37, "elapsed_time": "1 day, 11:58:50", "remaining_time": "1 day, 3:50:43", "throughput": 5652.86, "total_tokens": 732216280}
+{"current_steps": 16750, "total_steps": 29704, "loss": 0.3079033851623535, "lr": 4.003551860200534e-06, "epoch": 1.1277942364664693, "percentage": 56.39, "elapsed_time": "1 day, 11:59:34", "remaining_time": "1 day, 3:50:09", "throughput": 5652.53, "total_tokens": 732421912}
+{"current_steps": 16755, "total_steps": 29704, "loss": 0.3259746551513672, "lr": 4.000960956107167e-06, "epoch": 1.1281308914624293, "percentage": 56.41, "elapsed_time": "1 day, 12:00:16", "remaining_time": "1 day, 3:49:32", "throughput": 5652.29, "total_tokens": 732628200}
+{"current_steps": 16760, "total_steps": 29704, "loss": 0.3024310111999512, "lr": 3.998370331391881e-06, "epoch": 1.1284675464583895, "percentage": 56.42, "elapsed_time": "1 day, 12:00:53", "remaining_time": "1 day, 3:48:53", "throughput": 5652.19, "total_tokens": 732828608}
+{"current_steps": 16765, "total_steps": 29704, "loss": 0.3032145738601685, "lr": 3.995779986779139e-06, "epoch": 1.1288042014543496, "percentage": 56.44, "elapsed_time": "1 day, 12:01:36", "remaining_time": "1 day, 3:48:17", "throughput": 5651.9, "total_tokens": 733030240}
+{"current_steps": 16770, "total_steps": 29704, "loss": 0.32830450534820554, "lr": 3.9931899229933204e-06, "epoch": 1.1291408564503098, "percentage": 56.46, "elapsed_time": "1 day, 12:02:09", "remaining_time": "1 day, 3:47:34", "throughput": 5652.16, "total_tokens": 733249296}
+{"current_steps": 16775, "total_steps": 29704, "loss": 0.3006105422973633, "lr": 3.990600140758731e-06, "epoch": 1.1294775114462698, "percentage": 56.47, "elapsed_time": "1 day, 12:02:43", "remaining_time": "1 day, 3:46:52", "throughput": 5652.42, "total_tokens": 733474720}
+{"current_steps": 16780, "total_steps": 29704, "loss": 0.2877049922943115, "lr": 3.988010640799591e-06, "epoch": 1.12981416644223, "percentage": 56.49, "elapsed_time": "1 day, 12:03:26", "remaining_time": "1 day, 3:46:17", "throughput": 5652.21, "total_tokens": 733694216}
+{"current_steps": 16785, "total_steps": 29704, "loss": 0.33561320304870607, "lr": 3.985421423840051e-06, "epoch": 1.1301508214381901, "percentage": 56.51, "elapsed_time": "1 day, 12:04:04", "remaining_time": "1 day, 3:45:38", "throughput": 5652.2, "total_tokens": 733906352}
+{"current_steps": 16790, "total_steps": 29704, "loss": 0.28998355865478515, "lr": 3.982832490604173e-06, "epoch": 1.1304874764341504, "percentage": 56.52, "elapsed_time": "1 day, 12:04:48", "remaining_time": "1 day, 3:45:03", "throughput": 5652.01, "total_tokens": 734133280}
+{"current_steps": 16795, "total_steps": 29704, "loss": 0.3185850143432617, "lr": 3.9802438418159465e-06, "epoch": 1.1308241314301104, "percentage": 56.54, "elapsed_time": "1 day, 12:05:28", "remaining_time": "1 day, 3:44:26", "throughput": 5652.11, "total_tokens": 734371888}
+{"current_steps": 16800, "total_steps": 29704, "loss": 0.32431769371032715, "lr": 3.977655478199275e-06, "epoch": 1.1311607864260707, "percentage": 56.56, "elapsed_time": "1 day, 12:06:09", "remaining_time": "1 day, 3:43:48", "throughput": 5652.1, "total_tokens": 734599608}
+{"current_steps": 16805, "total_steps": 29704, "loss": 0.3372927188873291, "lr": 3.975067400477993e-06, "epoch": 1.1314974414220307, "percentage": 56.57, "elapsed_time": "1 day, 12:06:43", "remaining_time": "1 day, 3:43:06", "throughput": 5652.11, "total_tokens": 734795752}
+{"current_steps": 16810, "total_steps": 29704, "loss": 0.31949615478515625, "lr": 3.972479609375842e-06, "epoch": 1.131834096417991, "percentage": 56.59, "elapsed_time": "1 day, 12:07:22", "remaining_time": "1 day, 3:42:28", "throughput": 5652.15, "total_tokens": 735021664}
+{"current_steps": 16815, "total_steps": 29704, "loss": 0.31243975162506105, "lr": 3.969892105616492e-06, "epoch": 1.132170751413951, "percentage": 56.61, "elapsed_time": "1 day, 12:07:59", "remaining_time": "1 day, 3:41:47", "throughput": 5652.3, "total_tokens": 735246192}
+{"current_steps": 16820, "total_steps": 29704, "loss": 0.30039591789245607, "lr": 3.967304889923529e-06, "epoch": 1.1325074064099112, "percentage": 56.63, "elapsed_time": "1 day, 12:08:38", "remaining_time": "1 day, 3:41:09", "throughput": 5652.25, "total_tokens": 735462560}
+{"current_steps": 16825, "total_steps": 29704, "loss": 0.3070765256881714, "lr": 3.964717963020463e-06, "epoch": 1.1328440614058712, "percentage": 56.64, "elapsed_time": "1 day, 12:09:15", "remaining_time": "1 day, 3:40:29", "throughput": 5652.33, "total_tokens": 735678824}
+{"current_steps": 16830, "total_steps": 29704, "loss": 0.29083225727081297, "lr": 3.962131325630715e-06, "epoch": 1.1331807164018315, "percentage": 56.66, "elapsed_time": "1 day, 12:10:03", "remaining_time": "1 day, 3:39:58", "throughput": 5651.99, "total_tokens": 735911416}
+{"current_steps": 16835, "total_steps": 29704, "loss": 0.33116788864135743, "lr": 3.959544978477634e-06, "epoch": 1.1335173713977915, "percentage": 56.68, "elapsed_time": "1 day, 12:10:42", "remaining_time": "1 day, 3:39:20", "throughput": 5651.89, "total_tokens": 736118352}
+{"current_steps": 16840, "total_steps": 29704, "loss": 0.31282644271850585, "lr": 3.9569589222844805e-06, "epoch": 1.1338540263937518, "percentage": 56.69, "elapsed_time": "1 day, 12:11:27", "remaining_time": "1 day, 3:38:45", "throughput": 5651.61, "total_tokens": 736333424}
+{"current_steps": 16845, "total_steps": 29704, "loss": 0.3130516052246094, "lr": 3.954373157774439e-06, "epoch": 1.1341906813897118, "percentage": 56.71, "elapsed_time": "1 day, 12:12:04", "remaining_time": "1 day, 3:38:05", "throughput": 5651.74, "total_tokens": 736558888}
+{"current_steps": 16850, "total_steps": 29704, "loss": 0.3079373359680176, "lr": 3.951787685670609e-06, "epoch": 1.134527336385672, "percentage": 56.73, "elapsed_time": "1 day, 12:12:43", "remaining_time": "1 day, 3:37:27", "throughput": 5651.76, "total_tokens": 736782672}
+{"current_steps": 16855, "total_steps": 29704, "loss": 0.315033483505249, "lr": 3.949202506696012e-06, "epoch": 1.134863991381632, "percentage": 56.74, "elapsed_time": "1 day, 12:13:20", "remaining_time": "1 day, 3:36:47", "throughput": 5651.79, "total_tokens": 736994032}
+{"current_steps": 16860, "total_steps": 29704, "loss": 0.3076453685760498, "lr": 3.946617621573581e-06, "epoch": 1.1352006463775923, "percentage": 56.76, "elapsed_time": "1 day, 12:14:01", "remaining_time": "1 day, 3:36:10", "throughput": 5651.8, "total_tokens": 737227128}
+{"current_steps": 16865, "total_steps": 29704, "loss": 0.2819002628326416, "lr": 3.944033031026175e-06, "epoch": 1.1355373013735524, "percentage": 56.78, "elapsed_time": "1 day, 12:14:40", "remaining_time": "1 day, 3:35:32", "throughput": 5651.61, "total_tokens": 737425512}
+{"current_steps": 16870, "total_steps": 29704, "loss": 0.3037100791931152, "lr": 3.941448735776563e-06, "epoch": 1.1358739563695126, "percentage": 56.79, "elapsed_time": "1 day, 12:15:15", "remaining_time": "1 day, 3:34:50", "throughput": 5651.78, "total_tokens": 737645856}
+{"current_steps": 16875, "total_steps": 29704, "loss": 0.31839513778686523, "lr": 3.938864736547439e-06, "epoch": 1.1362106113654726, "percentage": 56.81, "elapsed_time": "1 day, 12:15:52", "remaining_time": "1 day, 3:34:11", "throughput": 5651.92, "total_tokens": 737874128}
+{"current_steps": 16880, "total_steps": 29704, "loss": 0.2992375135421753, "lr": 3.936281034061405e-06, "epoch": 1.1365472663614329, "percentage": 56.83, "elapsed_time": "1 day, 12:16:33", "remaining_time": "1 day, 3:33:34", "throughput": 5651.87, "total_tokens": 738099488}
+{"current_steps": 16885, "total_steps": 29704, "loss": 0.3111462354660034, "lr": 3.93369762904099e-06, "epoch": 1.136883921357393, "percentage": 56.84, "elapsed_time": "1 day, 12:17:19", "remaining_time": "1 day, 3:33:00", "throughput": 5651.49, "total_tokens": 738305920}
+{"current_steps": 16890, "total_steps": 29704, "loss": 0.3166337251663208, "lr": 3.931114522208631e-06, "epoch": 1.1372205763533532, "percentage": 56.86, "elapsed_time": "1 day, 12:18:02", "remaining_time": "1 day, 3:32:25", "throughput": 5651.16, "total_tokens": 738508624}
+{"current_steps": 16895, "total_steps": 29704, "loss": 0.30945444107055664, "lr": 3.928531714286689e-06, "epoch": 1.1375572313493132, "percentage": 56.88, "elapsed_time": "1 day, 12:18:39", "remaining_time": "1 day, 3:31:45", "throughput": 5651.29, "total_tokens": 738732992}
+{"current_steps": 16900, "total_steps": 29704, "loss": 0.32941608428955077, "lr": 3.925949205997434e-06, "epoch": 1.1378938863452734, "percentage": 56.89, "elapsed_time": "1 day, 12:19:21", "remaining_time": "1 day, 3:31:09", "throughput": 5651.06, "total_tokens": 738942840}
+{"current_steps": 16905, "total_steps": 29704, "loss": 0.2874737739562988, "lr": 3.923366998063062e-06, "epoch": 1.1382305413412335, "percentage": 56.91, "elapsed_time": "1 day, 12:20:01", "remaining_time": "1 day, 3:30:31", "throughput": 5650.96, "total_tokens": 739155240}
+{"current_steps": 16910, "total_steps": 29704, "loss": 0.27861971855163575, "lr": 3.920785091205674e-06, "epoch": 1.1385671963371937, "percentage": 56.93, "elapsed_time": "1 day, 12:20:35", "remaining_time": "1 day, 3:29:49", "throughput": 5651.26, "total_tokens": 739385256}
+{"current_steps": 16915, "total_steps": 29704, "loss": 0.34842791557312014, "lr": 3.918203486147294e-06, "epoch": 1.1389038513331537, "percentage": 56.95, "elapsed_time": "1 day, 12:21:16", "remaining_time": "1 day, 3:29:12", "throughput": 5651.11, "total_tokens": 739594936}
+{"current_steps": 16920, "total_steps": 29704, "loss": 0.3238081455230713, "lr": 3.9156221836098594e-06, "epoch": 1.139240506329114, "percentage": 56.96, "elapsed_time": "1 day, 12:21:52", "remaining_time": "1 day, 3:28:31", "throughput": 5651.26, "total_tokens": 739819304}
+{"current_steps": 16925, "total_steps": 29704, "loss": 0.3074013710021973, "lr": 3.913041184315224e-06, "epoch": 1.139577161325074, "percentage": 56.98, "elapsed_time": "1 day, 12:22:34", "remaining_time": "1 day, 3:27:55", "throughput": 5651.06, "total_tokens": 740033032}
+{"current_steps": 16930, "total_steps": 29704, "loss": 0.3118597030639648, "lr": 3.910460488985154e-06, "epoch": 1.1399138163210343, "percentage": 57.0, "elapsed_time": "1 day, 12:23:23", "remaining_time": "1 day, 3:27:24", "throughput": 5650.61, "total_tokens": 740249656}
+{"current_steps": 16935, "total_steps": 29704, "loss": 0.3097679138183594, "lr": 3.9078800983413355e-06, "epoch": 1.1402504713169943, "percentage": 57.01, "elapsed_time": "1 day, 12:23:59", "remaining_time": "1 day, 3:26:44", "throughput": 5650.76, "total_tokens": 740474224}
+{"current_steps": 16940, "total_steps": 29704, "loss": 0.2923974752426147, "lr": 3.905300013105365e-06, "epoch": 1.1405871263129546, "percentage": 57.03, "elapsed_time": "1 day, 12:24:39", "remaining_time": "1 day, 3:26:06", "throughput": 5650.74, "total_tokens": 740697112}
+{"current_steps": 16945, "total_steps": 29704, "loss": 0.30002388954162595, "lr": 3.902720233998754e-06, "epoch": 1.1409237813089146, "percentage": 57.05, "elapsed_time": "1 day, 12:25:16", "remaining_time": "1 day, 3:25:26", "throughput": 5650.78, "total_tokens": 740910376}
+{"current_steps": 16950, "total_steps": 29704, "loss": 0.31175391674041747, "lr": 3.90014076174293e-06, "epoch": 1.1412604363048748, "percentage": 57.06, "elapsed_time": "1 day, 12:25:56", "remaining_time": "1 day, 3:24:48", "throughput": 5650.76, "total_tokens": 741133216}
+{"current_steps": 16955, "total_steps": 29704, "loss": 0.2938354253768921, "lr": 3.897561597059237e-06, "epoch": 1.1415970913008349, "percentage": 57.08, "elapsed_time": "1 day, 12:26:33", "remaining_time": "1 day, 3:24:08", "throughput": 5650.88, "total_tokens": 741357712}
+{"current_steps": 16960, "total_steps": 29704, "loss": 0.29590883255004885, "lr": 3.894982740668927e-06, "epoch": 1.1419337462967951, "percentage": 57.1, "elapsed_time": "1 day, 12:27:07", "remaining_time": "1 day, 3:23:26", "throughput": 5650.99, "total_tokens": 741562560}
+{"current_steps": 16965, "total_steps": 29704, "loss": 0.3243727207183838, "lr": 3.89240419329317e-06, "epoch": 1.1422704012927551, "percentage": 57.11, "elapsed_time": "1 day, 12:27:43", "remaining_time": "1 day, 3:22:45", "throughput": 5651.26, "total_tokens": 741804536}
+{"current_steps": 16970, "total_steps": 29704, "loss": 0.31747913360595703, "lr": 3.889825955653046e-06, "epoch": 1.1426070562887154, "percentage": 57.13, "elapsed_time": "1 day, 12:28:19", "remaining_time": "1 day, 3:22:04", "throughput": 5651.39, "total_tokens": 742023480}
+{"current_steps": 16975, "total_steps": 29704, "loss": 0.2832490921020508, "lr": 3.887248028469558e-06, "epoch": 1.1429437112846754, "percentage": 57.15, "elapsed_time": "1 day, 12:28:53", "remaining_time": "1 day, 3:21:22", "throughput": 5651.73, "total_tokens": 742258408}
+{"current_steps": 16980, "total_steps": 29704, "loss": 0.2930821657180786, "lr": 3.884670412463607e-06, "epoch": 1.1432803662806357, "percentage": 57.16, "elapsed_time": "1 day, 12:29:31", "remaining_time": "1 day, 3:20:43", "throughput": 5651.71, "total_tokens": 742474896}
+{"current_steps": 16985, "total_steps": 29704, "loss": 0.31524713039398194, "lr": 3.8820931083560194e-06, "epoch": 1.1436170212765957, "percentage": 57.18, "elapsed_time": "1 day, 12:30:10", "remaining_time": "1 day, 3:20:05", "throughput": 5651.59, "total_tokens": 742680248}
+{"current_steps": 16990, "total_steps": 29704, "loss": 0.29150216579437255, "lr": 3.879516116867531e-06, "epoch": 1.143953676272556, "percentage": 57.2, "elapsed_time": "1 day, 12:30:45", "remaining_time": "1 day, 3:19:23", "throughput": 5651.71, "total_tokens": 742894640}
+{"current_steps": 16995, "total_steps": 29704, "loss": 0.30789361000061033, "lr": 3.876939438718786e-06, "epoch": 1.144290331268516, "percentage": 57.21, "elapsed_time": "1 day, 12:31:27", "remaining_time": "1 day, 3:18:47", "throughput": 5651.59, "total_tokens": 743112224}
+{"current_steps": 17000, "total_steps": 29704, "loss": 0.28911523818969725, "lr": 3.874363074630345e-06, "epoch": 1.1446269862644762, "percentage": 57.23, "elapsed_time": "1 day, 12:32:04", "remaining_time": "1 day, 3:18:07", "throughput": 5651.55, "total_tokens": 743319176}
+{"current_steps": 17005, "total_steps": 29704, "loss": 0.3129148006439209, "lr": 3.871787025322681e-06, "epoch": 1.1449636412604363, "percentage": 57.25, "elapsed_time": "1 day, 12:32:41", "remaining_time": "1 day, 3:17:27", "throughput": 5651.68, "total_tokens": 743545736}
+{"current_steps": 17010, "total_steps": 29704, "loss": 0.3087345361709595, "lr": 3.86921129151618e-06, "epoch": 1.1453002962563965, "percentage": 57.27, "elapsed_time": "1 day, 12:33:16", "remaining_time": "1 day, 3:16:46", "throughput": 5651.78, "total_tokens": 743756824}
+{"current_steps": 17015, "total_steps": 29704, "loss": 0.3323315143585205, "lr": 3.866635873931133e-06, "epoch": 1.1456369512523565, "percentage": 57.28, "elapsed_time": "1 day, 12:33:55", "remaining_time": "1 day, 3:16:07", "throughput": 5651.74, "total_tokens": 743968848}
+{"current_steps": 17020, "total_steps": 29704, "loss": 0.3157732725143433, "lr": 3.864060773287751e-06, "epoch": 1.1459736062483168, "percentage": 57.3, "elapsed_time": "1 day, 12:34:37", "remaining_time": "1 day, 3:15:31", "throughput": 5651.6, "total_tokens": 744191936}
+{"current_steps": 17025, "total_steps": 29704, "loss": 0.3161588191986084, "lr": 3.861485990306151e-06, "epoch": 1.1463102612442768, "percentage": 57.32, "elapsed_time": "1 day, 12:35:22", "remaining_time": "1 day, 3:14:57", "throughput": 5651.4, "total_tokens": 744416472}
+{"current_steps": 17030, "total_steps": 29704, "loss": 0.3021253108978271, "lr": 3.858911525706365e-06, "epoch": 1.146646916240237, "percentage": 57.33, "elapsed_time": "1 day, 12:35:59", "remaining_time": "1 day, 3:14:17", "throughput": 5651.54, "total_tokens": 744645048}
+{"current_steps": 17035, "total_steps": 29704, "loss": 0.34988272190093994, "lr": 3.85633738020833e-06, "epoch": 1.146983571236197, "percentage": 57.35, "elapsed_time": "1 day, 12:36:34", "remaining_time": "1 day, 3:13:35", "throughput": 5651.84, "total_tokens": 744878048}
+{"current_steps": 17040, "total_steps": 29704, "loss": 0.3174929618835449, "lr": 3.853763554531902e-06, "epoch": 1.1473202262321573, "percentage": 57.37, "elapsed_time": "1 day, 12:37:09", "remaining_time": "1 day, 3:12:54", "throughput": 5651.88, "total_tokens": 745082920}
+{"current_steps": 17045, "total_steps": 29704, "loss": 0.2886343479156494, "lr": 3.851190049396839e-06, "epoch": 1.1476568812281174, "percentage": 57.38, "elapsed_time": "1 day, 12:37:45", "remaining_time": "1 day, 3:12:14", "throughput": 5651.98, "total_tokens": 745302104}
+{"current_steps": 17050, "total_steps": 29704, "loss": 0.31344852447509763, "lr": 3.848616865522815e-06, "epoch": 1.1479935362240776, "percentage": 57.4, "elapsed_time": "1 day, 12:38:27", "remaining_time": "1 day, 3:11:38", "throughput": 5651.85, "total_tokens": 745522704}
+{"current_steps": 17055, "total_steps": 29704, "loss": 0.30931525230407714, "lr": 3.846044003629414e-06, "epoch": 1.1483301912200377, "percentage": 57.42, "elapsed_time": "1 day, 12:39:05", "remaining_time": "1 day, 3:10:58", "throughput": 5651.9, "total_tokens": 745742992}
+{"current_steps": 17060, "total_steps": 29704, "loss": 0.2803271055221558, "lr": 3.843471464436128e-06, "epoch": 1.148666846215998, "percentage": 57.43, "elapsed_time": "1 day, 12:39:49", "remaining_time": "1 day, 3:10:23", "throughput": 5651.54, "total_tokens": 745942912}
+{"current_steps": 17065, "total_steps": 29704, "loss": 0.33543803691864016, "lr": 3.840899248662358e-06, "epoch": 1.149003501211958, "percentage": 57.45, "elapsed_time": "1 day, 12:40:25", "remaining_time": "1 day, 3:09:43", "throughput": 5651.58, "total_tokens": 746154376}
+{"current_steps": 17070, "total_steps": 29704, "loss": 0.32558441162109375, "lr": 3.838327357027419e-06, "epoch": 1.1493401562079182, "percentage": 57.47, "elapsed_time": "1 day, 12:41:04", "remaining_time": "1 day, 3:09:04", "throughput": 5651.55, "total_tokens": 746367240}
+{"current_steps": 17075, "total_steps": 29704, "loss": 0.3117891550064087, "lr": 3.835755790250528e-06, "epoch": 1.1496768112038782, "percentage": 57.48, "elapsed_time": "1 day, 12:41:41", "remaining_time": "1 day, 3:08:24", "throughput": 5651.67, "total_tokens": 746592480}
+{"current_steps": 17080, "total_steps": 29704, "loss": 0.2934911251068115, "lr": 3.83318454905082e-06, "epoch": 1.1500134661998385, "percentage": 57.5, "elapsed_time": "1 day, 12:42:23", "remaining_time": "1 day, 3:07:48", "throughput": 5651.56, "total_tokens": 746818640}
+{"current_steps": 17085, "total_steps": 29704, "loss": 0.32155680656433105, "lr": 3.830613634147331e-06, "epoch": 1.1503501211957985, "percentage": 57.52, "elapsed_time": "1 day, 12:43:04", "remaining_time": "1 day, 3:07:11", "throughput": 5651.48, "total_tokens": 747039496}
+{"current_steps": 17090, "total_steps": 29704, "loss": 0.3073138236999512, "lr": 3.82804304625901e-06, "epoch": 1.1506867761917587, "percentage": 57.53, "elapsed_time": "1 day, 12:43:40", "remaining_time": "1 day, 3:06:30", "throughput": 5651.72, "total_tokens": 747272792}
+{"current_steps": 17095, "total_steps": 29704, "loss": 0.2988899707794189, "lr": 3.825472786104715e-06, "epoch": 1.1510234311877188, "percentage": 57.55, "elapsed_time": "1 day, 12:44:16", "remaining_time": "1 day, 3:05:50", "throughput": 5651.94, "total_tokens": 747507048}
+{"current_steps": 17100, "total_steps": 29704, "loss": 0.3167643308639526, "lr": 3.8229028544032095e-06, "epoch": 1.151360086183679, "percentage": 57.57, "elapsed_time": "1 day, 12:44:58", "remaining_time": "1 day, 3:05:13", "throughput": 5651.68, "total_tokens": 747707608}
+{"current_steps": 17105, "total_steps": 29704, "loss": 0.3144956111907959, "lr": 3.8203332518731666e-06, "epoch": 1.151696741179639, "percentage": 57.58, "elapsed_time": "1 day, 12:45:33", "remaining_time": "1 day, 3:04:32", "throughput": 5651.85, "total_tokens": 747928560}
+{"current_steps": 17110, "total_steps": 29704, "loss": 0.3034801483154297, "lr": 3.8177639792331695e-06, "epoch": 1.1520333961755993, "percentage": 57.6, "elapsed_time": "1 day, 12:46:12", "remaining_time": "1 day, 3:03:54", "throughput": 5651.9, "total_tokens": 748157856}
+{"current_steps": 17115, "total_steps": 29704, "loss": 0.34812791347503663, "lr": 3.815195037201704e-06, "epoch": 1.1523700511715593, "percentage": 57.62, "elapsed_time": "1 day, 12:46:52", "remaining_time": "1 day, 3:03:16", "throughput": 5651.75, "total_tokens": 748364016}
+{"current_steps": 17120, "total_steps": 29704, "loss": 0.321405029296875, "lr": 3.8126264264971694e-06, "epoch": 1.1527067061675196, "percentage": 57.64, "elapsed_time": "1 day, 12:47:34", "remaining_time": "1 day, 3:02:40", "throughput": 5651.64, "total_tokens": 748586232}
+{"current_steps": 17125, "total_steps": 29704, "loss": 0.3137218952178955, "lr": 3.8100581478378673e-06, "epoch": 1.1530433611634796, "percentage": 57.65, "elapsed_time": "1 day, 12:48:12", "remaining_time": "1 day, 3:02:01", "throughput": 5651.66, "total_tokens": 748805128}
+{"current_steps": 17130, "total_steps": 29704, "loss": 0.3160563468933105, "lr": 3.807490201942012e-06, "epoch": 1.1533800161594399, "percentage": 57.67, "elapsed_time": "1 day, 12:48:47", "remaining_time": "1 day, 3:01:19", "throughput": 5651.88, "total_tokens": 749027264}
+{"current_steps": 17135, "total_steps": 29704, "loss": 0.29862165451049805, "lr": 3.804922589527717e-06, "epoch": 1.1537166711553999, "percentage": 57.69, "elapsed_time": "1 day, 12:49:27", "remaining_time": "1 day, 3:00:41", "throughput": 5651.8, "total_tokens": 749243080}
+{"current_steps": 17140, "total_steps": 29704, "loss": 0.3242902040481567, "lr": 3.8023553113130096e-06, "epoch": 1.1540533261513601, "percentage": 57.7, "elapsed_time": "1 day, 12:50:02", "remaining_time": "1 day, 3:00:00", "throughput": 5652.06, "total_tokens": 749478816}
+{"current_steps": 17145, "total_steps": 29704, "loss": 0.28427910804748535, "lr": 3.799788368015821e-06, "epoch": 1.1543899811473202, "percentage": 57.72, "elapsed_time": "1 day, 12:50:42", "remaining_time": "1 day, 2:59:22", "throughput": 5652.05, "total_tokens": 749699336}
+{"current_steps": 17150, "total_steps": 29704, "loss": 0.30901274681091306, "lr": 3.7972217603539873e-06, "epoch": 1.1547266361432804, "percentage": 57.74, "elapsed_time": "1 day, 12:51:21", "remaining_time": "1 day, 2:58:44", "throughput": 5651.92, "total_tokens": 749905168}
+{"current_steps": 17155, "total_steps": 29704, "loss": 0.3213947772979736, "lr": 3.7946554890452524e-06, "epoch": 1.1550632911392404, "percentage": 57.75, "elapsed_time": "1 day, 12:51:59", "remaining_time": "1 day, 2:58:05", "throughput": 5652.03, "total_tokens": 750135464}
+{"current_steps": 17160, "total_steps": 29704, "loss": 0.3054837226867676, "lr": 3.792089554807269e-06, "epoch": 1.1553999461352007, "percentage": 57.77, "elapsed_time": "1 day, 12:52:41", "remaining_time": "1 day, 2:57:29", "throughput": 5651.94, "total_tokens": 750362056}
+{"current_steps": 17165, "total_steps": 29704, "loss": 0.3161867618560791, "lr": 3.789523958357587e-06, "epoch": 1.1557366011311607, "percentage": 57.79, "elapsed_time": "1 day, 12:53:22", "remaining_time": "1 day, 2:56:51", "throughput": 5651.81, "total_tokens": 750572160}
+{"current_steps": 17170, "total_steps": 29704, "loss": 0.2760681390762329, "lr": 3.7869587004136714e-06, "epoch": 1.156073256127121, "percentage": 57.8, "elapsed_time": "1 day, 12:53:59", "remaining_time": "1 day, 2:56:12", "throughput": 5651.93, "total_tokens": 750802552}
+{"current_steps": 17175, "total_steps": 29704, "loss": 0.3197911262512207, "lr": 3.7843937816928864e-06, "epoch": 1.156409911123081, "percentage": 57.82, "elapsed_time": "1 day, 12:54:44", "remaining_time": "1 day, 2:55:38", "throughput": 5651.76, "total_tokens": 751031672}
+{"current_steps": 17180, "total_steps": 29704, "loss": 0.3260814666748047, "lr": 3.781829202912506e-06, "epoch": 1.1567465661190413, "percentage": 57.84, "elapsed_time": "1 day, 12:55:21", "remaining_time": "1 day, 2:54:57", "throughput": 5651.8, "total_tokens": 751243800}
+{"current_steps": 17185, "total_steps": 29704, "loss": 0.3070995807647705, "lr": 3.7792649647897023e-06, "epoch": 1.1570832211150013, "percentage": 57.85, "elapsed_time": "1 day, 12:56:03", "remaining_time": "1 day, 2:54:21", "throughput": 5651.7, "total_tokens": 751468240}
+{"current_steps": 17190, "total_steps": 29704, "loss": 0.30820441246032715, "lr": 3.7767010680415595e-06, "epoch": 1.1574198761109615, "percentage": 57.87, "elapsed_time": "1 day, 12:56:42", "remaining_time": "1 day, 2:53:43", "throughput": 5651.62, "total_tokens": 751680248}
+{"current_steps": 17195, "total_steps": 29704, "loss": 0.3007169008255005, "lr": 3.774137513385061e-06, "epoch": 1.1577565311069216, "percentage": 57.89, "elapsed_time": "1 day, 12:57:30", "remaining_time": "1 day, 2:53:11", "throughput": 5651.21, "total_tokens": 751897104}
+{"current_steps": 17200, "total_steps": 29704, "loss": 0.316827392578125, "lr": 3.7715743015371005e-06, "epoch": 1.1580931861028818, "percentage": 57.9, "elapsed_time": "1 day, 12:58:09", "remaining_time": "1 day, 2:52:32", "throughput": 5651.31, "total_tokens": 752126600}
+{"current_steps": 17205, "total_steps": 29704, "loss": 0.28647899627685547, "lr": 3.769011433214467e-06, "epoch": 1.1584298410988418, "percentage": 57.92, "elapsed_time": "1 day, 12:58:50", "remaining_time": "1 day, 2:51:55", "throughput": 5651.22, "total_tokens": 752350528}
+{"current_steps": 17210, "total_steps": 29704, "loss": 0.32767200469970703, "lr": 3.766448909133863e-06, "epoch": 1.158766496094802, "percentage": 57.94, "elapsed_time": "1 day, 12:59:21", "remaining_time": "1 day, 2:51:12", "throughput": 5651.64, "total_tokens": 752583080}
+{"current_steps": 17215, "total_steps": 29704, "loss": 0.29516313076019285, "lr": 3.7638867300118864e-06, "epoch": 1.1591031510907621, "percentage": 57.96, "elapsed_time": "1 day, 12:59:58", "remaining_time": "1 day, 2:50:32", "throughput": 5651.66, "total_tokens": 752794792}
+{"current_steps": 17220, "total_steps": 29704, "loss": 0.2927243709564209, "lr": 3.7613248965650447e-06, "epoch": 1.1594398060867224, "percentage": 57.97, "elapsed_time": "1 day, 13:00:36", "remaining_time": "1 day, 2:49:52", "throughput": 5651.73, "total_tokens": 753018152}
+{"current_steps": 17225, "total_steps": 29704, "loss": 0.3014249324798584, "lr": 3.758763409509745e-06, "epoch": 1.1597764610826824, "percentage": 57.99, "elapsed_time": "1 day, 13:01:16", "remaining_time": "1 day, 2:49:15", "throughput": 5651.74, "total_tokens": 753245648}
+{"current_steps": 17230, "total_steps": 29704, "loss": 0.3143294334411621, "lr": 3.7562022695623013e-06, "epoch": 1.1601131160786426, "percentage": 58.01, "elapsed_time": "1 day, 13:01:55", "remaining_time": "1 day, 2:48:36", "throughput": 5651.74, "total_tokens": 753463144}
+{"current_steps": 17235, "total_steps": 29704, "loss": 0.3114176273345947, "lr": 3.7536414774389252e-06, "epoch": 1.1604497710746027, "percentage": 58.02, "elapsed_time": "1 day, 13:02:34", "remaining_time": "1 day, 2:47:58", "throughput": 5651.75, "total_tokens": 753687504}
+{"current_steps": 17240, "total_steps": 29704, "loss": 0.2993874788284302, "lr": 3.7510810338557363e-06, "epoch": 1.160786426070563, "percentage": 58.04, "elapsed_time": "1 day, 13:03:19", "remaining_time": "1 day, 2:47:23", "throughput": 5651.41, "total_tokens": 753893032}
+{"current_steps": 17245, "total_steps": 29704, "loss": 0.30895147323608396, "lr": 3.748520939528753e-06, "epoch": 1.161123081066523, "percentage": 58.06, "elapsed_time": "1 day, 13:04:00", "remaining_time": "1 day, 2:46:46", "throughput": 5651.31, "total_tokens": 754112432}
+{"current_steps": 17250, "total_steps": 29704, "loss": 0.3002096891403198, "lr": 3.7459611951739007e-06, "epoch": 1.1614597360624832, "percentage": 58.07, "elapsed_time": "1 day, 13:04:41", "remaining_time": "1 day, 2:46:10", "throughput": 5651.24, "total_tokens": 754338736}
+{"current_steps": 17255, "total_steps": 29704, "loss": 0.3243889093399048, "lr": 3.7434018015069984e-06, "epoch": 1.1617963910584432, "percentage": 58.09, "elapsed_time": "1 day, 13:05:25", "remaining_time": "1 day, 2:45:35", "throughput": 5651.04, "total_tokens": 754559056}
+{"current_steps": 17260, "total_steps": 29704, "loss": 0.3150045871734619, "lr": 3.740842759243778e-06, "epoch": 1.1621330460544035, "percentage": 58.11, "elapsed_time": "1 day, 13:06:05", "remaining_time": "1 day, 2:44:57", "throughput": 5650.99, "total_tokens": 754778800}
+{"current_steps": 17265, "total_steps": 29704, "loss": 0.28803367614746095, "lr": 3.738284069099862e-06, "epoch": 1.1624697010503635, "percentage": 58.12, "elapsed_time": "1 day, 13:06:45", "remaining_time": "1 day, 2:44:19", "throughput": 5651.07, "total_tokens": 755011760}
+{"current_steps": 17270, "total_steps": 29704, "loss": 0.31182174682617186, "lr": 3.735725731790785e-06, "epoch": 1.1628063560463238, "percentage": 58.14, "elapsed_time": "1 day, 13:07:18", "remaining_time": "1 day, 2:43:36", "throughput": 5651.35, "total_tokens": 755236808}
+{"current_steps": 17275, "total_steps": 29704, "loss": 0.3179009914398193, "lr": 3.733167748031974e-06, "epoch": 1.1631430110422838, "percentage": 58.16, "elapsed_time": "1 day, 13:07:56", "remaining_time": "1 day, 2:42:57", "throughput": 5651.44, "total_tokens": 755464248}
+{"current_steps": 17280, "total_steps": 29704, "loss": 0.3087700366973877, "lr": 3.7306101185387646e-06, "epoch": 1.163479666038244, "percentage": 58.17, "elapsed_time": "1 day, 13:08:37", "remaining_time": "1 day, 2:42:20", "throughput": 5651.39, "total_tokens": 755687464}
+{"current_steps": 17285, "total_steps": 29704, "loss": 0.3141305923461914, "lr": 3.728052844026386e-06, "epoch": 1.163816321034204, "percentage": 58.19, "elapsed_time": "1 day, 13:09:20", "remaining_time": "1 day, 2:41:44", "throughput": 5651.22, "total_tokens": 755907672}
+{"current_steps": 17290, "total_steps": 29704, "loss": 0.3007349014282227, "lr": 3.7254959252099745e-06, "epoch": 1.1641529760301643, "percentage": 58.21, "elapsed_time": "1 day, 13:10:00", "remaining_time": "1 day, 2:41:07", "throughput": 5651.06, "total_tokens": 756114320}
+{"current_steps": 17295, "total_steps": 29704, "loss": 0.2961156368255615, "lr": 3.7229393628045617e-06, "epoch": 1.1644896310261243, "percentage": 58.22, "elapsed_time": "1 day, 13:10:36", "remaining_time": "1 day, 2:40:26", "throughput": 5651.21, "total_tokens": 756338136}
+{"current_steps": 17300, "total_steps": 29704, "loss": 0.2917820453643799, "lr": 3.720383157525087e-06, "epoch": 1.1648262860220846, "percentage": 58.24, "elapsed_time": "1 day, 13:11:16", "remaining_time": "1 day, 2:39:48", "throughput": 5651.16, "total_tokens": 756556624}
+{"current_steps": 17305, "total_steps": 29704, "loss": 0.3004648685455322, "lr": 3.717827310086379e-06, "epoch": 1.1651629410180446, "percentage": 58.26, "elapsed_time": "1 day, 13:11:54", "remaining_time": "1 day, 2:39:09", "throughput": 5651.16, "total_tokens": 756770512}
+{"current_steps": 17310, "total_steps": 29704, "loss": 0.31314473152160643, "lr": 3.7152718212031778e-06, "epoch": 1.1654995960140049, "percentage": 58.27, "elapsed_time": "1 day, 13:12:37", "remaining_time": "1 day, 2:38:34", "throughput": 5650.94, "total_tokens": 756987488}
+{"current_steps": 17315, "total_steps": 29704, "loss": 0.33815641403198243, "lr": 3.7127166915901123e-06, "epoch": 1.165836251009965, "percentage": 58.29, "elapsed_time": "1 day, 13:13:15", "remaining_time": "1 day, 2:37:54", "throughput": 5650.89, "total_tokens": 757190888}
+{"current_steps": 17320, "total_steps": 29704, "loss": 0.3213828086853027, "lr": 3.710161921961721e-06, "epoch": 1.1661729060059252, "percentage": 58.31, "elapsed_time": "1 day, 13:13:53", "remaining_time": "1 day, 2:37:15", "throughput": 5650.9, "total_tokens": 757406736}
+{"current_steps": 17325, "total_steps": 29704, "loss": 0.29859418869018556, "lr": 3.707607513032434e-06, "epoch": 1.1665095610018852, "percentage": 58.33, "elapsed_time": "1 day, 13:14:30", "remaining_time": "1 day, 2:36:35", "throughput": 5650.91, "total_tokens": 757621448}
+{"current_steps": 17330, "total_steps": 29704, "loss": 0.2971299171447754, "lr": 3.705053465516587e-06, "epoch": 1.1668462159978454, "percentage": 58.34, "elapsed_time": "1 day, 13:15:06", "remaining_time": "1 day, 2:35:54", "throughput": 5651.06, "total_tokens": 757842136}
+{"current_steps": 17335, "total_steps": 29704, "loss": 0.2950276851654053, "lr": 3.702499780128407e-06, "epoch": 1.1671828709938055, "percentage": 58.36, "elapsed_time": "1 day, 13:15:46", "remaining_time": "1 day, 2:35:17", "throughput": 5650.89, "total_tokens": 758048560}
+{"current_steps": 17340, "total_steps": 29704, "loss": 0.3086616277694702, "lr": 3.6999464575820266e-06, "epoch": 1.1675195259897657, "percentage": 58.38, "elapsed_time": "1 day, 13:16:25", "remaining_time": "1 day, 2:34:39", "throughput": 5650.86, "total_tokens": 758266576}
+{"current_steps": 17345, "total_steps": 29704, "loss": 0.3014980792999268, "lr": 3.697393498591473e-06, "epoch": 1.1678561809857257, "percentage": 58.39, "elapsed_time": "1 day, 13:17:10", "remaining_time": "1 day, 2:34:04", "throughput": 5650.62, "total_tokens": 758487384}
+{"current_steps": 17350, "total_steps": 29704, "loss": 0.32812952995300293, "lr": 3.694840903870676e-06, "epoch": 1.168192835981686, "percentage": 58.41, "elapsed_time": "1 day, 13:17:46", "remaining_time": "1 day, 2:33:23", "throughput": 5650.74, "total_tokens": 758704816}
+{"current_steps": 17355, "total_steps": 29704, "loss": 0.3267998695373535, "lr": 3.6922886741334564e-06, "epoch": 1.168529490977646, "percentage": 58.43, "elapsed_time": "1 day, 13:18:25", "remaining_time": "1 day, 2:32:45", "throughput": 5650.74, "total_tokens": 758924344}
+{"current_steps": 17360, "total_steps": 29704, "loss": 0.2991864442825317, "lr": 3.689736810093541e-06, "epoch": 1.1688661459736063, "percentage": 58.44, "elapsed_time": "1 day, 13:19:02", "remaining_time": "1 day, 2:32:05", "throughput": 5650.85, "total_tokens": 759152032}
+{"current_steps": 17365, "total_steps": 29704, "loss": 0.31711387634277344, "lr": 3.6871853124645464e-06, "epoch": 1.1692028009695663, "percentage": 58.46, "elapsed_time": "1 day, 13:19:42", "remaining_time": "1 day, 2:31:27", "throughput": 5650.88, "total_tokens": 759378848}
+{"current_steps": 17370, "total_steps": 29704, "loss": 0.30228276252746583, "lr": 3.684634181959994e-06, "epoch": 1.1695394559655266, "percentage": 58.48, "elapsed_time": "1 day, 13:20:22", "remaining_time": "1 day, 2:30:49", "throughput": 5650.96, "total_tokens": 759614152}
+{"current_steps": 17375, "total_steps": 29704, "loss": 0.3172401666641235, "lr": 3.6820834192932974e-06, "epoch": 1.1698761109614866, "percentage": 58.49, "elapsed_time": "1 day, 13:21:00", "remaining_time": "1 day, 2:30:11", "throughput": 5650.94, "total_tokens": 759830040}
+{"current_steps": 17380, "total_steps": 29704, "loss": 0.3354950904846191, "lr": 3.679533025177772e-06, "epoch": 1.1702127659574468, "percentage": 58.51, "elapsed_time": "1 day, 13:21:36", "remaining_time": "1 day, 2:29:30", "throughput": 5651.24, "total_tokens": 760073280}
+{"current_steps": 17385, "total_steps": 29704, "loss": 0.31007990837097166, "lr": 3.6769830003266234e-06, "epoch": 1.1705494209534069, "percentage": 58.53, "elapsed_time": "1 day, 13:22:08", "remaining_time": "1 day, 2:28:46", "throughput": 5651.54, "total_tokens": 760294432}
+{"current_steps": 17390, "total_steps": 29704, "loss": 0.3161858081817627, "lr": 3.6744333454529607e-06, "epoch": 1.1708860759493671, "percentage": 58.54, "elapsed_time": "1 day, 13:22:51", "remaining_time": "1 day, 2:28:11", "throughput": 5651.43, "total_tokens": 760522824}
+{"current_steps": 17395, "total_steps": 29704, "loss": 0.3075212001800537, "lr": 3.6718840612697847e-06, "epoch": 1.1712227309453271, "percentage": 58.56, "elapsed_time": "1 day, 13:23:31", "remaining_time": "1 day, 2:27:33", "throughput": 5651.53, "total_tokens": 760757728}
+{"current_steps": 17400, "total_steps": 29704, "loss": 0.3011133670806885, "lr": 3.6693351484899986e-06, "epoch": 1.1715593859412874, "percentage": 58.58, "elapsed_time": "1 day, 13:24:03", "remaining_time": "1 day, 2:26:49", "throughput": 5651.74, "total_tokens": 760969040}
+{"current_steps": 17405, "total_steps": 29704, "loss": 0.30281815528869627, "lr": 3.6667866078263937e-06, "epoch": 1.1718960409372474, "percentage": 58.59, "elapsed_time": "1 day, 13:24:40", "remaining_time": "1 day, 2:26:09", "throughput": 5651.76, "total_tokens": 761178944}
+{"current_steps": 17410, "total_steps": 29704, "loss": 0.34009361267089844, "lr": 3.664238439991663e-06, "epoch": 1.1722326959332077, "percentage": 58.61, "elapsed_time": "1 day, 13:25:17", "remaining_time": "1 day, 2:25:30", "throughput": 5651.8, "total_tokens": 761396128}
+{"current_steps": 17415, "total_steps": 29704, "loss": 0.3160113334655762, "lr": 3.6616906456983925e-06, "epoch": 1.1725693509291677, "percentage": 58.63, "elapsed_time": "1 day, 13:25:52", "remaining_time": "1 day, 2:24:48", "throughput": 5651.99, "total_tokens": 761617984}
+{"current_steps": 17420, "total_steps": 29704, "loss": 0.32563726902008056, "lr": 3.659143225659065e-06, "epoch": 1.172906005925128, "percentage": 58.65, "elapsed_time": "1 day, 13:26:29", "remaining_time": "1 day, 2:24:08", "throughput": 5652.05, "total_tokens": 761835752}
+{"current_steps": 17425, "total_steps": 29704, "loss": 0.30316720008850095, "lr": 3.6565961805860595e-06, "epoch": 1.173242660921088, "percentage": 58.66, "elapsed_time": "1 day, 13:27:13", "remaining_time": "1 day, 2:23:34", "throughput": 5651.92, "total_tokens": 762067720}
+{"current_steps": 17430, "total_steps": 29704, "loss": 0.3012946605682373, "lr": 3.654049511191649e-06, "epoch": 1.1735793159170482, "percentage": 58.68, "elapsed_time": "1 day, 13:27:54", "remaining_time": "1 day, 2:22:56", "throughput": 5651.76, "total_tokens": 762275792}
+{"current_steps": 17435, "total_steps": 29704, "loss": 0.3098176956176758, "lr": 3.6515032181879995e-06, "epoch": 1.1739159709130083, "percentage": 58.7, "elapsed_time": "1 day, 13:28:34", "remaining_time": "1 day, 2:22:19", "throughput": 5651.65, "total_tokens": 762491920}
+{"current_steps": 17440, "total_steps": 29704, "loss": 0.28149569034576416, "lr": 3.6489573022871755e-06, "epoch": 1.1742526259089685, "percentage": 58.71, "elapsed_time": "1 day, 13:29:12", "remaining_time": "1 day, 2:21:39", "throughput": 5651.73, "total_tokens": 762712840}
+{"current_steps": 17445, "total_steps": 29704, "loss": 0.31446964740753175, "lr": 3.6464117642011333e-06, "epoch": 1.1745892809049285, "percentage": 58.73, "elapsed_time": "1 day, 13:29:51", "remaining_time": "1 day, 2:21:01", "throughput": 5651.62, "total_tokens": 762919536}
+{"current_steps": 17450, "total_steps": 29704, "loss": 0.31410171985626223, "lr": 3.643866604641726e-06, "epoch": 1.1749259359008888, "percentage": 58.75, "elapsed_time": "1 day, 13:30:25", "remaining_time": "1 day, 2:20:19", "throughput": 5651.91, "total_tokens": 763152304}
+{"current_steps": 17455, "total_steps": 29704, "loss": 0.29862222671508787, "lr": 3.641321824320697e-06, "epoch": 1.1752625908968488, "percentage": 58.76, "elapsed_time": "1 day, 13:31:11", "remaining_time": "1 day, 2:19:45", "throughput": 5651.62, "total_tokens": 763371656}
+{"current_steps": 17460, "total_steps": 29704, "loss": 0.2896681785583496, "lr": 3.6387774239496893e-06, "epoch": 1.175599245892809, "percentage": 58.78, "elapsed_time": "1 day, 13:31:50", "remaining_time": "1 day, 2:19:07", "throughput": 5651.73, "total_tokens": 763607368}
+{"current_steps": 17465, "total_steps": 29704, "loss": 0.3099272012710571, "lr": 3.6362334042402326e-06, "epoch": 1.175935900888769, "percentage": 58.8, "elapsed_time": "1 day, 13:32:36", "remaining_time": "1 day, 2:18:34", "throughput": 5651.41, "total_tokens": 763825752}
+{"current_steps": 17470, "total_steps": 29704, "loss": 0.2802994966506958, "lr": 3.6336897659037563e-06, "epoch": 1.1762725558847293, "percentage": 58.81, "elapsed_time": "1 day, 13:33:08", "remaining_time": "1 day, 2:17:50", "throughput": 5651.61, "total_tokens": 764035624}
+{"current_steps": 17475, "total_steps": 29704, "loss": 0.29577035903930665, "lr": 3.6311465096515787e-06, "epoch": 1.1766092108806894, "percentage": 58.83, "elapsed_time": "1 day, 13:33:44", "remaining_time": "1 day, 2:17:10", "throughput": 5651.8, "total_tokens": 764261488}
+{"current_steps": 17480, "total_steps": 29704, "loss": 0.2771427154541016, "lr": 3.6286036361949175e-06, "epoch": 1.1769458658766496, "percentage": 58.85, "elapsed_time": "1 day, 13:34:22", "remaining_time": "1 day, 2:16:31", "throughput": 5651.87, "total_tokens": 764486984}
+{"current_steps": 17485, "total_steps": 29704, "loss": 0.2587808847427368, "lr": 3.6260611462448736e-06, "epoch": 1.1772825208726099, "percentage": 58.86, "elapsed_time": "1 day, 13:35:02", "remaining_time": "1 day, 2:15:52", "throughput": 5651.88, "total_tokens": 764711328}
+{"current_steps": 17490, "total_steps": 29704, "loss": 0.3395826816558838, "lr": 3.623519040512451e-06, "epoch": 1.17761917586857, "percentage": 58.88, "elapsed_time": "1 day, 13:35:38", "remaining_time": "1 day, 2:15:12", "throughput": 5651.86, "total_tokens": 764916768}
+{"current_steps": 17495, "total_steps": 29704, "loss": 0.2713226556777954, "lr": 3.6209773197085373e-06, "epoch": 1.17795583086453, "percentage": 58.9, "elapsed_time": "1 day, 13:36:20", "remaining_time": "1 day, 2:14:36", "throughput": 5651.76, "total_tokens": 765139336}
+{"current_steps": 17500, "total_steps": 29704, "loss": 0.31320526599884035, "lr": 3.6184359845439214e-06, "epoch": 1.1782924858604902, "percentage": 58.91, "elapsed_time": "1 day, 13:36:53", "remaining_time": "1 day, 2:13:53", "throughput": 5652.14, "total_tokens": 765373584}
+{"current_steps": 17505, "total_steps": 29704, "loss": 0.32944774627685547, "lr": 3.615895035729275e-06, "epoch": 1.1786291408564504, "percentage": 58.93, "elapsed_time": "1 day, 13:37:39", "remaining_time": "1 day, 2:13:19", "throughput": 5651.9, "total_tokens": 765601768}
+{"current_steps": 17510, "total_steps": 29704, "loss": 0.2919255256652832, "lr": 3.6133544739751703e-06, "epoch": 1.1789657958524105, "percentage": 58.95, "elapsed_time": "1 day, 13:38:19", "remaining_time": "1 day, 2:12:41", "throughput": 5651.8, "total_tokens": 765813632}
+{"current_steps": 17515, "total_steps": 29704, "loss": 0.3174884796142578, "lr": 3.6108142999920633e-06, "epoch": 1.1793024508483705, "percentage": 58.97, "elapsed_time": "1 day, 13:39:00", "remaining_time": "1 day, 2:12:05", "throughput": 5651.67, "total_tokens": 766030160}
+{"current_steps": 17520, "total_steps": 29704, "loss": 0.30716781616210936, "lr": 3.608274514490308e-06, "epoch": 1.1796391058443307, "percentage": 58.98, "elapsed_time": "1 day, 13:39:40", "remaining_time": "1 day, 2:11:27", "throughput": 5651.64, "total_tokens": 766254576}
+{"current_steps": 17525, "total_steps": 29704, "loss": 0.2879929542541504, "lr": 3.605735118180146e-06, "epoch": 1.179975760840291, "percentage": 59.0, "elapsed_time": "1 day, 13:40:26", "remaining_time": "1 day, 2:10:53", "throughput": 5651.38, "total_tokens": 766474424}
+{"current_steps": 17530, "total_steps": 29704, "loss": 0.2810915470123291, "lr": 3.6031961117717144e-06, "epoch": 1.180312415836251, "percentage": 59.02, "elapsed_time": "1 day, 13:41:05", "remaining_time": "1 day, 2:10:14", "throughput": 5651.31, "total_tokens": 766686368}
+{"current_steps": 17535, "total_steps": 29704, "loss": 0.29715049266815186, "lr": 3.600657495975034e-06, "epoch": 1.180649070832211, "percentage": 59.03, "elapsed_time": "1 day, 13:41:46", "remaining_time": "1 day, 2:09:37", "throughput": 5651.27, "total_tokens": 766912752}
+{"current_steps": 17540, "total_steps": 29704, "loss": 0.30086421966552734, "lr": 3.598119271500024e-06, "epoch": 1.1809857258281713, "percentage": 59.05, "elapsed_time": "1 day, 13:42:18", "remaining_time": "1 day, 2:08:54", "throughput": 5651.53, "total_tokens": 767131128}
+{"current_steps": 17545, "total_steps": 29704, "loss": 0.30273914337158203, "lr": 3.5955814390564882e-06, "epoch": 1.1813223808241315, "percentage": 59.07, "elapsed_time": "1 day, 13:42:58", "remaining_time": "1 day, 2:08:16", "throughput": 5651.51, "total_tokens": 767352664}
+{"current_steps": 17550, "total_steps": 29704, "loss": 0.3133152961730957, "lr": 3.5930439993541265e-06, "epoch": 1.1816590358200916, "percentage": 59.08, "elapsed_time": "1 day, 13:43:32", "remaining_time": "1 day, 2:07:35", "throughput": 5651.63, "total_tokens": 767564224}
+{"current_steps": 17555, "total_steps": 29704, "loss": 0.3256534576416016, "lr": 3.590506953102522e-06, "epoch": 1.1819956908160516, "percentage": 59.1, "elapsed_time": "1 day, 13:44:13", "remaining_time": "1 day, 2:06:57", "throughput": 5651.54, "total_tokens": 767781400}
+{"current_steps": 17560, "total_steps": 29704, "loss": 0.3131730318069458, "lr": 3.5879703010111554e-06, "epoch": 1.1823323458120119, "percentage": 59.12, "elapsed_time": "1 day, 13:44:57", "remaining_time": "1 day, 2:06:22", "throughput": 5651.29, "total_tokens": 767996072}
+{"current_steps": 17565, "total_steps": 29704, "loss": 0.29679055213928224, "lr": 3.5854340437893898e-06, "epoch": 1.182669000807972, "percentage": 59.13, "elapsed_time": "1 day, 13:45:33", "remaining_time": "1 day, 2:05:42", "throughput": 5651.43, "total_tokens": 768218592}
+{"current_steps": 17570, "total_steps": 29704, "loss": 0.31975440979003905, "lr": 3.5828981821464838e-06, "epoch": 1.1830056558039321, "percentage": 59.15, "elapsed_time": "1 day, 13:46:12", "remaining_time": "1 day, 2:05:03", "throughput": 5651.34, "total_tokens": 768427128}
+{"current_steps": 17575, "total_steps": 29704, "loss": 0.31680521965026853, "lr": 3.580362716791582e-06, "epoch": 1.1833423107998922, "percentage": 59.17, "elapsed_time": "1 day, 13:46:44", "remaining_time": "1 day, 2:04:20", "throughput": 5651.74, "total_tokens": 768660704}
+{"current_steps": 17580, "total_steps": 29704, "loss": 0.3270276069641113, "lr": 3.5778276484337205e-06, "epoch": 1.1836789657958524, "percentage": 59.18, "elapsed_time": "1 day, 13:47:19", "remaining_time": "1 day, 2:03:38", "throughput": 5651.84, "total_tokens": 768871496}
+{"current_steps": 17585, "total_steps": 29704, "loss": 0.2917544603347778, "lr": 3.575292977781821e-06, "epoch": 1.1840156207918127, "percentage": 59.2, "elapsed_time": "1 day, 13:47:58", "remaining_time": "1 day, 2:03:00", "throughput": 5651.76, "total_tokens": 769081176}
+{"current_steps": 17590, "total_steps": 29704, "loss": 0.30277276039123535, "lr": 3.5727587055446976e-06, "epoch": 1.1843522757877727, "percentage": 59.22, "elapsed_time": "1 day, 13:48:35", "remaining_time": "1 day, 2:02:20", "throughput": 5651.9, "total_tokens": 769310368}
+{"current_steps": 17595, "total_steps": 29704, "loss": 0.3234860420227051, "lr": 3.570224832431051e-06, "epoch": 1.1846889307837327, "percentage": 59.23, "elapsed_time": "1 day, 13:49:10", "remaining_time": "1 day, 2:01:39", "throughput": 5652.08, "total_tokens": 769534008}
+{"current_steps": 17600, "total_steps": 29704, "loss": 0.3072806358337402, "lr": 3.5676913591494715e-06, "epoch": 1.185025585779693, "percentage": 59.25, "elapsed_time": "1 day, 13:49:48", "remaining_time": "1 day, 2:01:00", "throughput": 5652.14, "total_tokens": 769757384}
+{"current_steps": 17605, "total_steps": 29704, "loss": 0.32632970809936523, "lr": 3.5651582864084345e-06, "epoch": 1.1853622407756532, "percentage": 59.27, "elapsed_time": "1 day, 13:50:30", "remaining_time": "1 day, 2:00:24", "throughput": 5651.99, "total_tokens": 769975064}
+{"current_steps": 17610, "total_steps": 29704, "loss": 0.3052994966506958, "lr": 3.5626256149163096e-06, "epoch": 1.1856988957716132, "percentage": 59.28, "elapsed_time": "1 day, 13:51:06", "remaining_time": "1 day, 1:59:43", "throughput": 5652.13, "total_tokens": 770193168}
+{"current_steps": 17615, "total_steps": 29704, "loss": 0.3075841188430786, "lr": 3.5600933453813464e-06, "epoch": 1.1860355507675733, "percentage": 59.3, "elapsed_time": "1 day, 13:51:40", "remaining_time": "1 day, 1:59:01", "throughput": 5652.27, "total_tokens": 770408384}
+{"current_steps": 17620, "total_steps": 29704, "loss": 0.28684291839599607, "lr": 3.5575614785116886e-06, "epoch": 1.1863722057635335, "percentage": 59.32, "elapsed_time": "1 day, 13:52:25", "remaining_time": "1 day, 1:58:27", "throughput": 5651.95, "total_tokens": 770620728}
+{"current_steps": 17625, "total_steps": 29704, "loss": 0.3115447998046875, "lr": 3.5550300150153618e-06, "epoch": 1.1867088607594938, "percentage": 59.34, "elapsed_time": "1 day, 13:53:03", "remaining_time": "1 day, 1:57:48", "throughput": 5652.02, "total_tokens": 770842752}
+{"current_steps": 17630, "total_steps": 29704, "loss": 0.29783077239990235, "lr": 3.5524989556002863e-06, "epoch": 1.1870455157554538, "percentage": 59.35, "elapsed_time": "1 day, 13:53:48", "remaining_time": "1 day, 1:57:13", "throughput": 5651.76, "total_tokens": 771060848}
+{"current_steps": 17635, "total_steps": 29704, "loss": 0.30118565559387206, "lr": 3.5499683009742604e-06, "epoch": 1.1873821707514138, "percentage": 59.37, "elapsed_time": "1 day, 13:54:22", "remaining_time": "1 day, 1:56:31", "throughput": 5651.85, "total_tokens": 771263296}
+{"current_steps": 17640, "total_steps": 29704, "loss": 0.2976966381072998, "lr": 3.5474380518449757e-06, "epoch": 1.187718825747374, "percentage": 59.39, "elapsed_time": "1 day, 13:54:56", "remaining_time": "1 day, 1:55:50", "throughput": 5652.03, "total_tokens": 771484736}
+{"current_steps": 17645, "total_steps": 29704, "loss": 0.301554536819458, "lr": 3.5449082089200072e-06, "epoch": 1.1880554807433343, "percentage": 59.4, "elapsed_time": "1 day, 13:55:36", "remaining_time": "1 day, 1:55:12", "throughput": 5651.99, "total_tokens": 771704632}
+{"current_steps": 17650, "total_steps": 29704, "loss": 0.32130672931671145, "lr": 3.5423787729068206e-06, "epoch": 1.1883921357392944, "percentage": 59.42, "elapsed_time": "1 day, 13:56:13", "remaining_time": "1 day, 1:54:32", "throughput": 5652.12, "total_tokens": 771927880}
+{"current_steps": 17655, "total_steps": 29704, "loss": 0.29790964126586916, "lr": 3.539849744512759e-06, "epoch": 1.1887287907352544, "percentage": 59.44, "elapsed_time": "1 day, 13:57:01", "remaining_time": "1 day, 1:53:59", "throughput": 5651.67, "total_tokens": 772137016}
+{"current_steps": 17660, "total_steps": 29704, "loss": 0.29845280647277833, "lr": 3.5373211244450633e-06, "epoch": 1.1890654457312146, "percentage": 59.45, "elapsed_time": "1 day, 13:57:42", "remaining_time": "1 day, 1:53:22", "throughput": 5651.6, "total_tokens": 772361904}
+{"current_steps": 17665, "total_steps": 29704, "loss": 0.31483676433563235, "lr": 3.5347929134108483e-06, "epoch": 1.189402100727175, "percentage": 59.47, "elapsed_time": "1 day, 13:58:19", "remaining_time": "1 day, 1:52:42", "throughput": 5651.63, "total_tokens": 772575080}
+{"current_steps": 17670, "total_steps": 29704, "loss": 0.33406968116760255, "lr": 3.532265112117125e-06, "epoch": 1.189738755723135, "percentage": 59.49, "elapsed_time": "1 day, 13:59:00", "remaining_time": "1 day, 1:52:06", "throughput": 5651.6, "total_tokens": 772803264}
+{"current_steps": 17675, "total_steps": 29704, "loss": 0.2623422861099243, "lr": 3.5297377212707805e-06, "epoch": 1.190075410719095, "percentage": 59.5, "elapsed_time": "1 day, 13:59:35", "remaining_time": "1 day, 1:51:24", "throughput": 5651.69, "total_tokens": 773015632}
+{"current_steps": 17680, "total_steps": 29704, "loss": 0.3026540279388428, "lr": 3.527210741578596e-06, "epoch": 1.1904120657150552, "percentage": 59.52, "elapsed_time": "1 day, 14:00:08", "remaining_time": "1 day, 1:50:42", "throughput": 5651.99, "total_tokens": 773239976}
+{"current_steps": 17685, "total_steps": 29704, "loss": 0.3219871520996094, "lr": 3.5246841737472304e-06, "epoch": 1.1907487207110155, "percentage": 59.54, "elapsed_time": "1 day, 14:00:55", "remaining_time": "1 day, 1:50:08", "throughput": 5651.59, "total_tokens": 773448912}
+{"current_steps": 17690, "total_steps": 29704, "loss": 0.2920974254608154, "lr": 3.522158018483231e-06, "epoch": 1.1910853757069755, "percentage": 59.55, "elapsed_time": "1 day, 14:01:31", "remaining_time": "1 day, 1:49:28", "throughput": 5651.76, "total_tokens": 773675944}
+{"current_steps": 17695, "total_steps": 29704, "loss": 0.31039018630981446, "lr": 3.519632276493029e-06, "epoch": 1.1914220307029355, "percentage": 59.57, "elapsed_time": "1 day, 14:02:17", "remaining_time": "1 day, 1:48:54", "throughput": 5651.46, "total_tokens": 773894272}
+{"current_steps": 17700, "total_steps": 29704, "loss": 0.32564287185668944, "lr": 3.5171069484829417e-06, "epoch": 1.1917586856988958, "percentage": 59.59, "elapsed_time": "1 day, 14:03:01", "remaining_time": "1 day, 1:48:20", "throughput": 5651.19, "total_tokens": 774110520}
+{"current_steps": 17705, "total_steps": 29704, "loss": 0.28891453742980955, "lr": 3.5145820351591663e-06, "epoch": 1.192095340694856, "percentage": 59.6, "elapsed_time": "1 day, 14:03:42", "remaining_time": "1 day, 1:47:42", "throughput": 5651.06, "total_tokens": 774319632}
+{"current_steps": 17710, "total_steps": 29704, "loss": 0.3185939073562622, "lr": 3.51205753722779e-06, "epoch": 1.192431995690816, "percentage": 59.62, "elapsed_time": "1 day, 14:04:16", "remaining_time": "1 day, 1:47:00", "throughput": 5651.33, "total_tokens": 774550784}
+{"current_steps": 17715, "total_steps": 29704, "loss": 0.33892183303833007, "lr": 3.5095334553947757e-06, "epoch": 1.1927686506867763, "percentage": 59.64, "elapsed_time": "1 day, 14:04:54", "remaining_time": "1 day, 1:46:21", "throughput": 5651.43, "total_tokens": 774781848}
+{"current_steps": 17720, "total_steps": 29704, "loss": 0.2988421440124512, "lr": 3.50700979036598e-06, "epoch": 1.1931053056827363, "percentage": 59.66, "elapsed_time": "1 day, 14:05:34", "remaining_time": "1 day, 1:45:43", "throughput": 5651.38, "total_tokens": 774997592}
+{"current_steps": 17725, "total_steps": 29704, "loss": 0.28873891830444337, "lr": 3.5044865428471343e-06, "epoch": 1.1934419606786966, "percentage": 59.67, "elapsed_time": "1 day, 14:06:10", "remaining_time": "1 day, 1:45:03", "throughput": 5651.39, "total_tokens": 775203648}
+{"current_steps": 17730, "total_steps": 29704, "loss": 0.30784637928009034, "lr": 3.50196371354386e-06, "epoch": 1.1937786156746566, "percentage": 59.69, "elapsed_time": "1 day, 14:06:47", "remaining_time": "1 day, 1:44:23", "throughput": 5651.42, "total_tokens": 775418528}
+{"current_steps": 17735, "total_steps": 29704, "loss": 0.3373323202133179, "lr": 3.499441303161655e-06, "epoch": 1.1941152706706168, "percentage": 59.71, "elapsed_time": "1 day, 14:07:23", "remaining_time": "1 day, 1:43:43", "throughput": 5651.58, "total_tokens": 775644000}
+{"current_steps": 17740, "total_steps": 29704, "loss": 0.29674830436706545, "lr": 3.4969193124059042e-06, "epoch": 1.1944519256665769, "percentage": 59.72, "elapsed_time": "1 day, 14:08:00", "remaining_time": "1 day, 1:43:03", "throughput": 5651.72, "total_tokens": 775872464}
+{"current_steps": 17745, "total_steps": 29704, "loss": 0.2963244438171387, "lr": 3.4943977419818748e-06, "epoch": 1.1947885806625371, "percentage": 59.74, "elapsed_time": "1 day, 14:08:37", "remaining_time": "1 day, 1:42:23", "throughput": 5651.91, "total_tokens": 776107304}
+{"current_steps": 17750, "total_steps": 29704, "loss": 0.2947703838348389, "lr": 3.4918765925947173e-06, "epoch": 1.1951252356584972, "percentage": 59.76, "elapsed_time": "1 day, 14:09:21", "remaining_time": "1 day, 1:41:47", "throughput": 5651.77, "total_tokens": 776333360}
+{"current_steps": 17755, "total_steps": 29704, "loss": 0.32445640563964845, "lr": 3.4893558649494595e-06, "epoch": 1.1954618906544574, "percentage": 59.77, "elapsed_time": "1 day, 14:10:07", "remaining_time": "1 day, 1:41:14", "throughput": 5651.55, "total_tokens": 776565352}
+{"current_steps": 17760, "total_steps": 29704, "loss": 0.29216394424438474, "lr": 3.4868355597510194e-06, "epoch": 1.1957985456504174, "percentage": 59.79, "elapsed_time": "1 day, 14:10:51", "remaining_time": "1 day, 1:40:39", "throughput": 5651.3, "total_tokens": 776782064}
+{"current_steps": 17765, "total_steps": 29704, "loss": 0.27714879512786866, "lr": 3.4843156777041887e-06, "epoch": 1.1961352006463777, "percentage": 59.81, "elapsed_time": "1 day, 14:11:27", "remaining_time": "1 day, 1:39:58", "throughput": 5651.48, "total_tokens": 777006896}
+{"current_steps": 17770, "total_steps": 29704, "loss": 0.3087809085845947, "lr": 3.4817962195136457e-06, "epoch": 1.1964718556423377, "percentage": 59.82, "elapsed_time": "1 day, 14:12:10", "remaining_time": "1 day, 1:39:22", "throughput": 5651.29, "total_tokens": 777222560}
+{"current_steps": 17775, "total_steps": 29704, "loss": 0.3070957183837891, "lr": 3.479277185883947e-06, "epoch": 1.196808510638298, "percentage": 59.84, "elapsed_time": "1 day, 14:12:51", "remaining_time": "1 day, 1:38:45", "throughput": 5651.16, "total_tokens": 777436752}
+{"current_steps": 17780, "total_steps": 29704, "loss": 0.29088459014892576, "lr": 3.476758577519537e-06, "epoch": 1.197145165634258, "percentage": 59.86, "elapsed_time": "1 day, 14:13:25", "remaining_time": "1 day, 1:38:03", "throughput": 5651.35, "total_tokens": 777654880}
+{"current_steps": 17785, "total_steps": 29704, "loss": 0.3081394672393799, "lr": 3.4742403951247318e-06, "epoch": 1.1974818206302182, "percentage": 59.87, "elapsed_time": "1 day, 14:14:06", "remaining_time": "1 day, 1:37:26", "throughput": 5651.25, "total_tokens": 777876192}
+{"current_steps": 17790, "total_steps": 29704, "loss": 0.29077887535095215, "lr": 3.471722639403735e-06, "epoch": 1.1978184756261783, "percentage": 59.89, "elapsed_time": "1 day, 14:14:48", "remaining_time": "1 day, 1:36:50", "throughput": 5651.11, "total_tokens": 778090376}
+{"current_steps": 17795, "total_steps": 29704, "loss": 0.2993917942047119, "lr": 3.4692053110606277e-06, "epoch": 1.1981551306221385, "percentage": 59.91, "elapsed_time": "1 day, 14:15:28", "remaining_time": "1 day, 1:36:12", "throughput": 5650.91, "total_tokens": 778289376}
+{"current_steps": 17800, "total_steps": 29704, "loss": 0.2942534923553467, "lr": 3.4666884107993766e-06, "epoch": 1.1984917856180985, "percentage": 59.92, "elapsed_time": "1 day, 14:16:04", "remaining_time": "1 day, 1:35:31", "throughput": 5650.97, "total_tokens": 778502056}
+{"current_steps": 17805, "total_steps": 29704, "loss": 0.30054686069488523, "lr": 3.464171939323819e-06, "epoch": 1.1988284406140588, "percentage": 59.94, "elapsed_time": "1 day, 14:16:44", "remaining_time": "1 day, 1:34:54", "throughput": 5650.87, "total_tokens": 778716688}
+{"current_steps": 17810, "total_steps": 29704, "loss": 0.3028707981109619, "lr": 3.4616558973376848e-06, "epoch": 1.1991650956100188, "percentage": 59.96, "elapsed_time": "1 day, 14:17:20", "remaining_time": "1 day, 1:34:13", "throughput": 5650.94, "total_tokens": 778927512}
+{"current_steps": 17815, "total_steps": 29704, "loss": 0.30633254051208497, "lr": 3.4591402855445697e-06, "epoch": 1.199501750605979, "percentage": 59.98, "elapsed_time": "1 day, 14:17:53", "remaining_time": "1 day, 1:33:30", "throughput": 5651.22, "total_tokens": 779150112}
+{"current_steps": 17820, "total_steps": 29704, "loss": 0.3079712152481079, "lr": 3.4566251046479625e-06, "epoch": 1.199838405601939, "percentage": 59.99, "elapsed_time": "1 day, 14:18:35", "remaining_time": "1 day, 1:32:54", "throughput": 5651.08, "total_tokens": 779369912}
+{"current_steps": 17825, "total_steps": 29704, "loss": 0.3142574310302734, "lr": 3.454110355351221e-06, "epoch": 1.2001750605978994, "percentage": 60.01, "elapsed_time": "1 day, 14:19:14", "remaining_time": "1 day, 1:32:16", "throughput": 5651.07, "total_tokens": 779590624}
+{"current_steps": 17830, "total_steps": 29704, "loss": 0.3034769773483276, "lr": 3.451596038357593e-06, "epoch": 1.2005117155938594, "percentage": 60.03, "elapsed_time": "1 day, 14:19:53", "remaining_time": "1 day, 1:31:37", "throughput": 5651.15, "total_tokens": 779821712}
+{"current_steps": 17835, "total_steps": 29704, "loss": 0.32932286262512206, "lr": 3.4490821543701924e-06, "epoch": 1.2008483705898196, "percentage": 60.04, "elapsed_time": "1 day, 14:20:29", "remaining_time": "1 day, 1:30:57", "throughput": 5651.28, "total_tokens": 780045176}
+{"current_steps": 17840, "total_steps": 29704, "loss": 0.30574448108673097, "lr": 3.4465687040920227e-06, "epoch": 1.2011850255857797, "percentage": 60.06, "elapsed_time": "1 day, 14:21:12", "remaining_time": "1 day, 1:30:21", "throughput": 5651.17, "total_tokens": 780268064}
+{"current_steps": 17845, "total_steps": 29704, "loss": 0.278488302230835, "lr": 3.44405568822596e-06, "epoch": 1.20152168058174, "percentage": 60.08, "elapsed_time": "1 day, 14:21:53", "remaining_time": "1 day, 1:29:43", "throughput": 5651.06, "total_tokens": 780486632}
+{"current_steps": 17850, "total_steps": 29704, "loss": 0.3007037162780762, "lr": 3.441543107474765e-06, "epoch": 1.2018583355777, "percentage": 60.09, "elapsed_time": "1 day, 14:22:28", "remaining_time": "1 day, 1:29:03", "throughput": 5651.12, "total_tokens": 780694864}
+{"current_steps": 17855, "total_steps": 29704, "loss": 0.293853759765625, "lr": 3.439030962541069e-06, "epoch": 1.2021949905736602, "percentage": 60.11, "elapsed_time": "1 day, 14:23:07", "remaining_time": "1 day, 1:28:24", "throughput": 5651.1, "total_tokens": 780910776}
+{"current_steps": 17860, "total_steps": 29704, "loss": 0.30453169345855713, "lr": 3.4365192541273885e-06, "epoch": 1.2025316455696202, "percentage": 60.13, "elapsed_time": "1 day, 14:23:43", "remaining_time": "1 day, 1:27:43", "throughput": 5651.09, "total_tokens": 781111576}
+{"current_steps": 17865, "total_steps": 29704, "loss": 0.3177401304244995, "lr": 3.4340079829361108e-06, "epoch": 1.2028683005655805, "percentage": 60.14, "elapsed_time": "1 day, 14:24:16", "remaining_time": "1 day, 1:27:01", "throughput": 5651.23, "total_tokens": 781317608}
+{"current_steps": 17870, "total_steps": 29704, "loss": 0.31711881160736083, "lr": 3.4314971496695078e-06, "epoch": 1.2032049555615405, "percentage": 60.16, "elapsed_time": "1 day, 14:24:50", "remaining_time": "1 day, 1:26:19", "throughput": 5651.48, "total_tokens": 781545368}
+{"current_steps": 17875, "total_steps": 29704, "loss": 0.2896330118179321, "lr": 3.428986755029724e-06, "epoch": 1.2035416105575008, "percentage": 60.18, "elapsed_time": "1 day, 14:25:36", "remaining_time": "1 day, 1:25:45", "throughput": 5651.25, "total_tokens": 781774192}
+{"current_steps": 17880, "total_steps": 29704, "loss": 0.2978813171386719, "lr": 3.4264767997187874e-06, "epoch": 1.2038782655534608, "percentage": 60.19, "elapsed_time": "1 day, 14:26:12", "remaining_time": "1 day, 1:25:05", "throughput": 5651.43, "total_tokens": 782004920}
+{"current_steps": 17885, "total_steps": 29704, "loss": 0.3156701326370239, "lr": 3.423967284438594e-06, "epoch": 1.204214920549421, "percentage": 60.21, "elapsed_time": "1 day, 14:26:48", "remaining_time": "1 day, 1:24:25", "throughput": 5651.5, "total_tokens": 782217384}
+{"current_steps": 17890, "total_steps": 29704, "loss": 0.2962948799133301, "lr": 3.421458209890925e-06, "epoch": 1.204551575545381, "percentage": 60.23, "elapsed_time": "1 day, 14:27:23", "remaining_time": "1 day, 1:23:43", "throughput": 5651.78, "total_tokens": 782450248}
+{"current_steps": 17895, "total_steps": 29704, "loss": 0.28663947582244875, "lr": 3.418949576777433e-06, "epoch": 1.2048882305413413, "percentage": 60.24, "elapsed_time": "1 day, 14:28:00", "remaining_time": "1 day, 1:23:04", "throughput": 5651.92, "total_tokens": 782681456}
+{"current_steps": 17900, "total_steps": 29704, "loss": 0.30020842552185056, "lr": 3.4164413857996513e-06, "epoch": 1.2052248855373013, "percentage": 60.26, "elapsed_time": "1 day, 14:28:42", "remaining_time": "1 day, 1:22:27", "throughput": 5651.71, "total_tokens": 782891192}
+{"current_steps": 17905, "total_steps": 29704, "loss": 0.291434121131897, "lr": 3.4139336376589853e-06, "epoch": 1.2055615405332616, "percentage": 60.28, "elapsed_time": "1 day, 14:29:14", "remaining_time": "1 day, 1:21:44", "throughput": 5652.06, "total_tokens": 783119112}
+{"current_steps": 17910, "total_steps": 29704, "loss": 0.30569398403167725, "lr": 3.411426333056722e-06, "epoch": 1.2058981955292216, "percentage": 60.29, "elapsed_time": "1 day, 14:29:57", "remaining_time": "1 day, 1:21:08", "throughput": 5651.79, "total_tokens": 783321392}
+{"current_steps": 17915, "total_steps": 29704, "loss": 0.321109676361084, "lr": 3.408919472694017e-06, "epoch": 1.2062348505251819, "percentage": 60.31, "elapsed_time": "1 day, 14:30:37", "remaining_time": "1 day, 1:20:30", "throughput": 5651.75, "total_tokens": 783543848}
+{"current_steps": 17920, "total_steps": 29704, "loss": 0.30394299030303956, "lr": 3.406413057271909e-06, "epoch": 1.206571505521142, "percentage": 60.33, "elapsed_time": "1 day, 14:31:16", "remaining_time": "1 day, 1:19:52", "throughput": 5651.7, "total_tokens": 783758592}
+{"current_steps": 17925, "total_steps": 29704, "loss": 0.303491997718811, "lr": 3.403907087491306e-06, "epoch": 1.2069081605171021, "percentage": 60.35, "elapsed_time": "1 day, 14:31:52", "remaining_time": "1 day, 1:19:11", "throughput": 5651.81, "total_tokens": 783976568}
+{"current_steps": 17930, "total_steps": 29704, "loss": 0.30534958839416504, "lr": 3.4014015640529996e-06, "epoch": 1.2072448155130622, "percentage": 60.36, "elapsed_time": "1 day, 14:32:28", "remaining_time": "1 day, 1:18:31", "throughput": 5651.91, "total_tokens": 784193960}
+{"current_steps": 17935, "total_steps": 29704, "loss": 0.31456284523010253, "lr": 3.398896487657646e-06, "epoch": 1.2075814705090224, "percentage": 60.38, "elapsed_time": "1 day, 14:33:11", "remaining_time": "1 day, 1:17:55", "throughput": 5651.76, "total_tokens": 784414136}
+{"current_steps": 17940, "total_steps": 29704, "loss": 0.2981652975082397, "lr": 3.396391859005785e-06, "epoch": 1.2079181255049825, "percentage": 60.4, "elapsed_time": "1 day, 14:33:45", "remaining_time": "1 day, 1:17:13", "throughput": 5651.99, "total_tokens": 784638336}
+{"current_steps": 17945, "total_steps": 29704, "loss": 0.3018413305282593, "lr": 3.393887678797826e-06, "epoch": 1.2082547805009427, "percentage": 60.41, "elapsed_time": "1 day, 14:34:23", "remaining_time": "1 day, 1:16:34", "throughput": 5651.91, "total_tokens": 784844856}
+{"current_steps": 17950, "total_steps": 29704, "loss": 0.30618746280670167, "lr": 3.3913839477340575e-06, "epoch": 1.2085914354969027, "percentage": 60.43, "elapsed_time": "1 day, 14:35:02", "remaining_time": "1 day, 1:15:56", "throughput": 5651.87, "total_tokens": 785058936}
+{"current_steps": 17955, "total_steps": 29704, "loss": 0.28571882247924807, "lr": 3.388880666514637e-06, "epoch": 1.208928090492863, "percentage": 60.45, "elapsed_time": "1 day, 14:35:37", "remaining_time": "1 day, 1:15:14", "throughput": 5652.09, "total_tokens": 785286832}
+{"current_steps": 17960, "total_steps": 29704, "loss": 0.3155225276947021, "lr": 3.3863778358396033e-06, "epoch": 1.209264745488823, "percentage": 60.46, "elapsed_time": "1 day, 14:36:12", "remaining_time": "1 day, 1:14:34", "throughput": 5652.21, "total_tokens": 785504144}
+{"current_steps": 17965, "total_steps": 29704, "loss": 0.3005061149597168, "lr": 3.3838754564088592e-06, "epoch": 1.2096014004847833, "percentage": 60.48, "elapsed_time": "1 day, 14:36:51", "remaining_time": "1 day, 1:13:55", "throughput": 5652.26, "total_tokens": 785729392}
+{"current_steps": 17970, "total_steps": 29704, "loss": 0.29288544654846194, "lr": 3.3813735289221923e-06, "epoch": 1.2099380554807433, "percentage": 60.5, "elapsed_time": "1 day, 14:37:29", "remaining_time": "1 day, 1:13:16", "throughput": 5652.28, "total_tokens": 785948848}
+{"current_steps": 17975, "total_steps": 29704, "loss": 0.29325079917907715, "lr": 3.378872054079255e-06, "epoch": 1.2102747104767035, "percentage": 60.51, "elapsed_time": "1 day, 14:38:09", "remaining_time": "1 day, 1:12:38", "throughput": 5652.24, "total_tokens": 786168672}
+{"current_steps": 17980, "total_steps": 29704, "loss": 0.3322885513305664, "lr": 3.3763710325795784e-06, "epoch": 1.2106113654726636, "percentage": 60.53, "elapsed_time": "1 day, 14:38:46", "remaining_time": "1 day, 1:11:58", "throughput": 5652.2, "total_tokens": 786371976}
+{"current_steps": 17985, "total_steps": 29704, "loss": 0.3040419340133667, "lr": 3.3738704651225673e-06, "epoch": 1.2109480204686238, "percentage": 60.55, "elapsed_time": "1 day, 14:39:22", "remaining_time": "1 day, 1:11:18", "throughput": 5652.39, "total_tokens": 786600776}
+{"current_steps": 17990, "total_steps": 29704, "loss": 0.2779910326004028, "lr": 3.3713703524074936e-06, "epoch": 1.2112846754645838, "percentage": 60.56, "elapsed_time": "1 day, 14:39:59", "remaining_time": "1 day, 1:10:38", "throughput": 5652.35, "total_tokens": 786804904}
+{"current_steps": 17995, "total_steps": 29704, "loss": 0.31528596878051757, "lr": 3.3688706951335082e-06, "epoch": 1.211621330460544, "percentage": 60.58, "elapsed_time": "1 day, 14:40:35", "remaining_time": "1 day, 1:09:58", "throughput": 5652.34, "total_tokens": 787008024}
+{"current_steps": 18000, "total_steps": 29704, "loss": 0.30718555450439455, "lr": 3.3663714939996305e-06, "epoch": 1.2119579854565041, "percentage": 60.6, "elapsed_time": "1 day, 14:41:12", "remaining_time": "1 day, 1:09:17", "throughput": 5652.57, "total_tokens": 787245096}
+{"current_steps": 18005, "total_steps": 29704, "loss": 0.2900678634643555, "lr": 3.3638727497047575e-06, "epoch": 1.2122946404524644, "percentage": 60.61, "elapsed_time": "1 day, 14:41:55", "remaining_time": "1 day, 1:08:41", "throughput": 5652.29, "total_tokens": 787449376}
+{"current_steps": 18010, "total_steps": 29704, "loss": 0.29748926162719724, "lr": 3.3613744629476507e-06, "epoch": 1.2126312954484244, "percentage": 60.63, "elapsed_time": "1 day, 14:42:36", "remaining_time": "1 day, 1:08:04", "throughput": 5652.15, "total_tokens": 787661624}
+{"current_steps": 18015, "total_steps": 29704, "loss": 0.2816985845565796, "lr": 3.358876634426952e-06, "epoch": 1.2129679504443847, "percentage": 60.65, "elapsed_time": "1 day, 14:43:13", "remaining_time": "1 day, 1:07:25", "throughput": 5652.27, "total_tokens": 787888360}
+{"current_steps": 18020, "total_steps": 29704, "loss": 0.32491137981414797, "lr": 3.3563792648411676e-06, "epoch": 1.2133046054403447, "percentage": 60.67, "elapsed_time": "1 day, 14:43:49", "remaining_time": "1 day, 1:06:45", "throughput": 5652.24, "total_tokens": 788091136}
+{"current_steps": 18025, "total_steps": 29704, "loss": 0.288041353225708, "lr": 3.353882354888681e-06, "epoch": 1.213641260436305, "percentage": 60.68, "elapsed_time": "1 day, 14:44:30", "remaining_time": "1 day, 1:06:07", "throughput": 5652.18, "total_tokens": 788312048}
+{"current_steps": 18030, "total_steps": 29704, "loss": 0.2752676486968994, "lr": 3.3513859052677444e-06, "epoch": 1.213977915432265, "percentage": 60.7, "elapsed_time": "1 day, 14:45:08", "remaining_time": "1 day, 1:05:28", "throughput": 5652.05, "total_tokens": 788512224}
+{"current_steps": 18035, "total_steps": 29704, "loss": 0.3178478479385376, "lr": 3.3488899166764842e-06, "epoch": 1.2143145704282252, "percentage": 60.72, "elapsed_time": "1 day, 14:45:42", "remaining_time": "1 day, 1:04:46", "throughput": 5652.3, "total_tokens": 788737240}
+{"current_steps": 18040, "total_steps": 29704, "loss": 0.3054813385009766, "lr": 3.346394389812892e-06, "epoch": 1.2146512254241852, "percentage": 60.73, "elapsed_time": "1 day, 14:46:17", "remaining_time": "1 day, 1:04:05", "throughput": 5652.5, "total_tokens": 788959120}
+{"current_steps": 18045, "total_steps": 29704, "loss": 0.2954627752304077, "lr": 3.3438993253748365e-06, "epoch": 1.2149878804201455, "percentage": 60.75, "elapsed_time": "1 day, 14:47:00", "remaining_time": "1 day, 1:03:29", "throughput": 5652.25, "total_tokens": 789168384}
+{"current_steps": 18050, "total_steps": 29704, "loss": 0.3028736591339111, "lr": 3.3414047240600526e-06, "epoch": 1.2153245354161055, "percentage": 60.77, "elapsed_time": "1 day, 14:47:40", "remaining_time": "1 day, 1:02:51", "throughput": 5652.22, "total_tokens": 789390496}
+{"current_steps": 18055, "total_steps": 29704, "loss": 0.3098297119140625, "lr": 3.338910586566151e-06, "epoch": 1.2156611904120658, "percentage": 60.78, "elapsed_time": "1 day, 14:48:15", "remaining_time": "1 day, 1:02:11", "throughput": 5652.29, "total_tokens": 789600216}
+{"current_steps": 18060, "total_steps": 29704, "loss": 0.3021399974822998, "lr": 3.336416913590605e-06, "epoch": 1.2159978454080258, "percentage": 60.8, "elapsed_time": "1 day, 14:48:51", "remaining_time": "1 day, 1:01:30", "throughput": 5652.34, "total_tokens": 789807928}
+{"current_steps": 18065, "total_steps": 29704, "loss": 0.3216712474822998, "lr": 3.333923705830766e-06, "epoch": 1.216334500403986, "percentage": 60.82, "elapsed_time": "1 day, 14:49:30", "remaining_time": "1 day, 1:00:52", "throughput": 5652.33, "total_tokens": 790032400}
+{"current_steps": 18070, "total_steps": 29704, "loss": 0.3249736547470093, "lr": 3.3314309639838484e-06, "epoch": 1.216671155399946, "percentage": 60.83, "elapsed_time": "1 day, 14:50:01", "remaining_time": "1 day, 1:00:08", "throughput": 5652.67, "total_tokens": 790253040}
+{"current_steps": 18075, "total_steps": 29704, "loss": 0.30673856735229493, "lr": 3.328938688746942e-06, "epoch": 1.2170078103959063, "percentage": 60.85, "elapsed_time": "1 day, 14:50:48", "remaining_time": "1 day, 0:59:34", "throughput": 5652.43, "total_tokens": 790481600}
+{"current_steps": 18080, "total_steps": 29704, "loss": 0.3439304828643799, "lr": 3.3264468808170012e-06, "epoch": 1.2173444653918664, "percentage": 60.87, "elapsed_time": "1 day, 14:51:24", "remaining_time": "1 day, 0:58:54", "throughput": 5652.57, "total_tokens": 790707832}
+{"current_steps": 18085, "total_steps": 29704, "loss": 0.3145643711090088, "lr": 3.3239555408908562e-06, "epoch": 1.2176811203878266, "percentage": 60.88, "elapsed_time": "1 day, 14:52:07", "remaining_time": "1 day, 0:58:18", "throughput": 5652.22, "total_tokens": 790904072}
+{"current_steps": 18090, "total_steps": 29704, "loss": 0.2987529277801514, "lr": 3.3214646696651974e-06, "epoch": 1.2180177753837866, "percentage": 60.9, "elapsed_time": "1 day, 14:52:47", "remaining_time": "1 day, 0:57:40", "throughput": 5652.17, "total_tokens": 791119712}
+{"current_steps": 18095, "total_steps": 29704, "loss": 0.31772985458374026, "lr": 3.3189742678365923e-06, "epoch": 1.2183544303797469, "percentage": 60.92, "elapsed_time": "1 day, 14:53:22", "remaining_time": "1 day, 0:57:00", "throughput": 5652.31, "total_tokens": 791340456}
+{"current_steps": 18100, "total_steps": 29704, "loss": 0.28296403884887694, "lr": 3.316484336101472e-06, "epoch": 1.218691085375707, "percentage": 60.93, "elapsed_time": "1 day, 14:53:58", "remaining_time": "1 day, 0:56:19", "throughput": 5652.37, "total_tokens": 791546784}
+{"current_steps": 18105, "total_steps": 29704, "loss": 0.31280903816223143, "lr": 3.313994875156141e-06, "epoch": 1.2190277403716672, "percentage": 60.95, "elapsed_time": "1 day, 14:54:41", "remaining_time": "1 day, 0:55:43", "throughput": 5652.17, "total_tokens": 791764024}
+{"current_steps": 18110, "total_steps": 29704, "loss": 0.31598572731018065, "lr": 3.3115058856967643e-06, "epoch": 1.2193643953676272, "percentage": 60.97, "elapsed_time": "1 day, 14:55:22", "remaining_time": "1 day, 0:55:06", "throughput": 5652.15, "total_tokens": 791993616}
+{"current_steps": 18115, "total_steps": 29704, "loss": 0.27434990406036375, "lr": 3.3090173684193845e-06, "epoch": 1.2197010503635874, "percentage": 60.99, "elapsed_time": "1 day, 14:55:59", "remaining_time": "1 day, 0:54:26", "throughput": 5652.17, "total_tokens": 792207856}
+{"current_steps": 18120, "total_steps": 29704, "loss": 0.31843407154083253, "lr": 3.3065293240199037e-06, "epoch": 1.2200377053595475, "percentage": 61.0, "elapsed_time": "1 day, 14:56:42", "remaining_time": "1 day, 0:53:50", "throughput": 5652.04, "total_tokens": 792431928}
+{"current_steps": 18125, "total_steps": 29704, "loss": 0.3226888179779053, "lr": 3.3040417531940982e-06, "epoch": 1.2203743603555077, "percentage": 61.02, "elapsed_time": "1 day, 14:57:18", "remaining_time": "1 day, 0:53:10", "throughput": 5652.2, "total_tokens": 792655832}
+{"current_steps": 18130, "total_steps": 29704, "loss": 0.2980408906936646, "lr": 3.301554656637607e-06, "epoch": 1.2207110153514678, "percentage": 61.04, "elapsed_time": "1 day, 14:57:51", "remaining_time": "1 day, 0:52:27", "throughput": 5652.4, "total_tokens": 792868664}
+{"current_steps": 18135, "total_steps": 29704, "loss": 0.29863677024841306, "lr": 3.2990680350459427e-06, "epoch": 1.221047670347428, "percentage": 61.05, "elapsed_time": "1 day, 14:58:30", "remaining_time": "1 day, 0:51:49", "throughput": 5652.3, "total_tokens": 793080312}
+{"current_steps": 18140, "total_steps": 29704, "loss": 0.3160408020019531, "lr": 3.2965818891144763e-06, "epoch": 1.221384325343388, "percentage": 61.07, "elapsed_time": "1 day, 14:59:00", "remaining_time": "1 day, 0:51:05", "throughput": 5652.69, "total_tokens": 793303920}
+{"current_steps": 18145, "total_steps": 29704, "loss": 0.322130012512207, "lr": 3.294096219538454e-06, "epoch": 1.2217209803393483, "percentage": 61.09, "elapsed_time": "1 day, 14:59:47", "remaining_time": "1 day, 0:50:31", "throughput": 5652.49, "total_tokens": 793536920}
+{"current_steps": 18150, "total_steps": 29704, "loss": 0.3116767406463623, "lr": 3.291611027012983e-06, "epoch": 1.2220576353353083, "percentage": 61.1, "elapsed_time": "1 day, 15:00:29", "remaining_time": "1 day, 0:49:54", "throughput": 5652.4, "total_tokens": 793761088}
+{"current_steps": 18155, "total_steps": 29704, "loss": 0.29644060134887695, "lr": 3.289126312233043e-06, "epoch": 1.2223942903312686, "percentage": 61.12, "elapsed_time": "1 day, 15:01:09", "remaining_time": "1 day, 0:49:17", "throughput": 5652.42, "total_tokens": 793993376}
+{"current_steps": 18160, "total_steps": 29704, "loss": 0.3092994213104248, "lr": 3.2866420758934714e-06, "epoch": 1.2227309453272286, "percentage": 61.14, "elapsed_time": "1 day, 15:01:43", "remaining_time": "1 day, 0:48:35", "throughput": 5652.65, "total_tokens": 794216288}
+{"current_steps": 18165, "total_steps": 29704, "loss": 0.30644276142120364, "lr": 3.2841583186889826e-06, "epoch": 1.2230676003231888, "percentage": 61.15, "elapsed_time": "1 day, 15:02:20", "remaining_time": "1 day, 0:47:56", "throughput": 5652.75, "total_tokens": 794442800}
+{"current_steps": 18170, "total_steps": 29704, "loss": 0.2917419195175171, "lr": 3.281675041314146e-06, "epoch": 1.2234042553191489, "percentage": 61.17, "elapsed_time": "1 day, 15:03:03", "remaining_time": "1 day, 0:47:19", "throughput": 5652.57, "total_tokens": 794656160}
+{"current_steps": 18175, "total_steps": 29704, "loss": 0.29529433250427245, "lr": 3.279192244463406e-06, "epoch": 1.2237409103151091, "percentage": 61.19, "elapsed_time": "1 day, 15:03:44", "remaining_time": "1 day, 0:46:42", "throughput": 5652.43, "total_tokens": 794869448}
+{"current_steps": 18180, "total_steps": 29704, "loss": 0.29960105419158933, "lr": 3.2767099288310667e-06, "epoch": 1.2240775653110691, "percentage": 61.2, "elapsed_time": "1 day, 15:04:24", "remaining_time": "1 day, 0:46:04", "throughput": 5652.41, "total_tokens": 795093624}
+{"current_steps": 18185, "total_steps": 29704, "loss": 0.2962838888168335, "lr": 3.2742280951113025e-06, "epoch": 1.2244142203070294, "percentage": 61.22, "elapsed_time": "1 day, 15:05:02", "remaining_time": "1 day, 0:45:25", "throughput": 5652.5, "total_tokens": 795322792}
+{"current_steps": 18190, "total_steps": 29704, "loss": 0.30564584732055666, "lr": 3.2717467439981464e-06, "epoch": 1.2247508753029894, "percentage": 61.24, "elapsed_time": "1 day, 15:05:45", "remaining_time": "1 day, 0:44:50", "throughput": 5652.26, "total_tokens": 795532552}
+{"current_steps": 18195, "total_steps": 29704, "loss": 0.27375977039337157, "lr": 3.2692658761855033e-06, "epoch": 1.2250875302989497, "percentage": 61.25, "elapsed_time": "1 day, 15:06:21", "remaining_time": "1 day, 0:44:09", "throughput": 5652.29, "total_tokens": 795735568}
+{"current_steps": 18200, "total_steps": 29704, "loss": 0.2808785915374756, "lr": 3.266785492367138e-06, "epoch": 1.2254241852949097, "percentage": 61.27, "elapsed_time": "1 day, 15:06:55", "remaining_time": "1 day, 0:43:28", "throughput": 5652.46, "total_tokens": 795955960}
+{"current_steps": 18205, "total_steps": 29704, "loss": 0.28676323890686034, "lr": 3.2643055932366852e-06, "epoch": 1.22576084029087, "percentage": 61.29, "elapsed_time": "1 day, 15:07:34", "remaining_time": "1 day, 0:42:49", "throughput": 5652.41, "total_tokens": 796165624}
+{"current_steps": 18210, "total_steps": 29704, "loss": 0.3010567665100098, "lr": 3.2618261794876365e-06, "epoch": 1.22609749528683, "percentage": 61.3, "elapsed_time": "1 day, 15:08:09", "remaining_time": "1 day, 0:42:07", "throughput": 5652.64, "total_tokens": 796395392}
+{"current_steps": 18215, "total_steps": 29704, "loss": 0.2966732978820801, "lr": 3.259347251813356e-06, "epoch": 1.2264341502827902, "percentage": 61.32, "elapsed_time": "1 day, 15:08:47", "remaining_time": "1 day, 0:41:29", "throughput": 5652.66, "total_tokens": 796617856}
+{"current_steps": 18220, "total_steps": 29704, "loss": 0.3215049982070923, "lr": 3.256868810907064e-06, "epoch": 1.2267708052787503, "percentage": 61.34, "elapsed_time": "1 day, 15:09:27", "remaining_time": "1 day, 0:40:51", "throughput": 5652.72, "total_tokens": 796852648}
+{"current_steps": 18225, "total_steps": 29704, "loss": 0.3135005712509155, "lr": 3.2543908574618514e-06, "epoch": 1.2271074602747105, "percentage": 61.36, "elapsed_time": "1 day, 15:10:12", "remaining_time": "1 day, 0:40:16", "throughput": 5652.49, "total_tokens": 797070736}
+{"current_steps": 18230, "total_steps": 29704, "loss": 0.2944124460220337, "lr": 3.251913392170668e-06, "epoch": 1.2274441152706705, "percentage": 61.37, "elapsed_time": "1 day, 15:10:57", "remaining_time": "1 day, 0:39:41", "throughput": 5652.24, "total_tokens": 797290688}
+{"current_steps": 18235, "total_steps": 29704, "loss": 0.29227581024169924, "lr": 3.249436415726333e-06, "epoch": 1.2277807702666308, "percentage": 61.39, "elapsed_time": "1 day, 15:11:35", "remaining_time": "1 day, 0:39:02", "throughput": 5652.33, "total_tokens": 797516256}
+{"current_steps": 18240, "total_steps": 29704, "loss": 0.3139004707336426, "lr": 3.24695992882152e-06, "epoch": 1.2281174252625908, "percentage": 61.41, "elapsed_time": "1 day, 15:12:14", "remaining_time": "1 day, 0:38:24", "throughput": 5652.22, "total_tokens": 797725232}
+{"current_steps": 18245, "total_steps": 29704, "loss": 0.3086080551147461, "lr": 3.244483932148773e-06, "epoch": 1.228454080258551, "percentage": 61.42, "elapsed_time": "1 day, 15:12:46", "remaining_time": "1 day, 0:37:41", "throughput": 5652.52, "total_tokens": 797943968}
+{"current_steps": 18250, "total_steps": 29704, "loss": 0.30695390701293945, "lr": 3.2420084264004966e-06, "epoch": 1.228790735254511, "percentage": 61.44, "elapsed_time": "1 day, 15:13:21", "remaining_time": "1 day, 0:37:00", "throughput": 5652.64, "total_tokens": 798161800}
+{"current_steps": 18255, "total_steps": 29704, "loss": 0.30524253845214844, "lr": 3.2395334122689594e-06, "epoch": 1.2291273902504714, "percentage": 61.46, "elapsed_time": "1 day, 15:14:00", "remaining_time": "1 day, 0:36:22", "throughput": 5652.61, "total_tokens": 798379968}
+{"current_steps": 18260, "total_steps": 29704, "loss": 0.3120718955993652, "lr": 3.2370588904462873e-06, "epoch": 1.2294640452464314, "percentage": 61.47, "elapsed_time": "1 day, 15:14:36", "remaining_time": "1 day, 0:35:41", "throughput": 5652.82, "total_tokens": 798609312}
+{"current_steps": 18265, "total_steps": 29704, "loss": 0.3038747549057007, "lr": 3.2345848616244775e-06, "epoch": 1.2298007002423916, "percentage": 61.49, "elapsed_time": "1 day, 15:15:15", "remaining_time": "1 day, 0:35:03", "throughput": 5652.62, "total_tokens": 798804840}
+{"current_steps": 18270, "total_steps": 29704, "loss": 0.307041072845459, "lr": 3.2321113264953797e-06, "epoch": 1.2301373552383517, "percentage": 61.51, "elapsed_time": "1 day, 15:15:47", "remaining_time": "1 day, 0:34:20", "throughput": 5652.93, "total_tokens": 799027080}
+{"current_steps": 18275, "total_steps": 29704, "loss": 0.2948000431060791, "lr": 3.2296382857507124e-06, "epoch": 1.230474010234312, "percentage": 61.52, "elapsed_time": "1 day, 15:16:26", "remaining_time": "1 day, 0:33:41", "throughput": 5652.86, "total_tokens": 799238584}
+{"current_steps": 18280, "total_steps": 29704, "loss": 0.31221818923950195, "lr": 3.227165740082053e-06, "epoch": 1.230810665230272, "percentage": 61.54, "elapsed_time": "1 day, 15:17:08", "remaining_time": "1 day, 0:33:05", "throughput": 5652.69, "total_tokens": 799452544}
+{"current_steps": 18285, "total_steps": 29704, "loss": 0.2914170980453491, "lr": 3.2246936901808424e-06, "epoch": 1.2311473202262322, "percentage": 61.56, "elapsed_time": "1 day, 15:17:47", "remaining_time": "1 day, 0:32:26", "throughput": 5652.71, "total_tokens": 799675552}
+{"current_steps": 18290, "total_steps": 29704, "loss": 0.32509727478027345, "lr": 3.2222221367383786e-06, "epoch": 1.2314839752221922, "percentage": 61.57, "elapsed_time": "1 day, 15:18:35", "remaining_time": "1 day, 0:31:53", "throughput": 5652.38, "total_tokens": 799898272}
+{"current_steps": 18295, "total_steps": 29704, "loss": 0.30308525562286376, "lr": 3.2197510804458265e-06, "epoch": 1.2318206302181525, "percentage": 61.59, "elapsed_time": "1 day, 15:19:14", "remaining_time": "1 day, 0:31:15", "throughput": 5652.28, "total_tokens": 800105152}
+{"current_steps": 18300, "total_steps": 29704, "loss": 0.2986292362213135, "lr": 3.217280521994207e-06, "epoch": 1.2321572852141125, "percentage": 61.61, "elapsed_time": "1 day, 15:19:50", "remaining_time": "1 day, 0:30:34", "throughput": 5652.45, "total_tokens": 800331312}
+{"current_steps": 18305, "total_steps": 29704, "loss": 0.3016390562057495, "lr": 3.214810462074407e-06, "epoch": 1.2324939402100727, "percentage": 61.62, "elapsed_time": "1 day, 15:20:23", "remaining_time": "1 day, 0:29:52", "throughput": 5652.68, "total_tokens": 800551688}
+{"current_steps": 18310, "total_steps": 29704, "loss": 0.30267770290374757, "lr": 3.2123409013771666e-06, "epoch": 1.2328305952060328, "percentage": 61.64, "elapsed_time": "1 day, 15:21:00", "remaining_time": "1 day, 0:29:13", "throughput": 5652.78, "total_tokens": 800777344}
+{"current_steps": 18315, "total_steps": 29704, "loss": 0.31267609596252444, "lr": 3.209871840593095e-06, "epoch": 1.233167250201993, "percentage": 61.66, "elapsed_time": "1 day, 15:21:46", "remaining_time": "1 day, 0:28:38", "throughput": 5652.61, "total_tokens": 801009256}
+{"current_steps": 18320, "total_steps": 29704, "loss": 0.30893242359161377, "lr": 3.207403280412652e-06, "epoch": 1.233503905197953, "percentage": 61.68, "elapsed_time": "1 day, 15:22:23", "remaining_time": "1 day, 0:27:58", "throughput": 5652.68, "total_tokens": 801228256}
+{"current_steps": 18325, "total_steps": 29704, "loss": 0.28535709381103513, "lr": 3.2049352215261677e-06, "epoch": 1.2338405601939133, "percentage": 61.69, "elapsed_time": "1 day, 15:23:00", "remaining_time": "1 day, 0:27:19", "throughput": 5652.81, "total_tokens": 801458800}
+{"current_steps": 18330, "total_steps": 29704, "loss": 0.29358530044555664, "lr": 3.2024676646238222e-06, "epoch": 1.2341772151898733, "percentage": 61.71, "elapsed_time": "1 day, 15:23:36", "remaining_time": "1 day, 0:26:39", "throughput": 5652.92, "total_tokens": 801679448}
+{"current_steps": 18335, "total_steps": 29704, "loss": 0.3098346948623657, "lr": 3.2000006103956653e-06, "epoch": 1.2345138701858336, "percentage": 61.73, "elapsed_time": "1 day, 15:24:17", "remaining_time": "1 day, 0:26:01", "throughput": 5652.88, "total_tokens": 801900976}
+{"current_steps": 18340, "total_steps": 29704, "loss": 0.28231058120727537, "lr": 3.1975340595315956e-06, "epoch": 1.2348505251817936, "percentage": 61.74, "elapsed_time": "1 day, 15:24:53", "remaining_time": "1 day, 0:25:21", "throughput": 5652.92, "total_tokens": 802111128}
+{"current_steps": 18345, "total_steps": 29704, "loss": 0.2976810455322266, "lr": 3.1950680127213796e-06, "epoch": 1.2351871801777539, "percentage": 61.76, "elapsed_time": "1 day, 15:25:32", "remaining_time": "1 day, 0:24:43", "throughput": 5652.9, "total_tokens": 802330912}
+{"current_steps": 18350, "total_steps": 29704, "loss": 0.32010388374328613, "lr": 3.192602470654638e-06, "epoch": 1.235523835173714, "percentage": 61.78, "elapsed_time": "1 day, 15:26:14", "remaining_time": "1 day, 0:24:06", "throughput": 5652.75, "total_tokens": 802544448}
+{"current_steps": 18355, "total_steps": 29704, "loss": 0.2869425296783447, "lr": 3.190137434020853e-06, "epoch": 1.2358604901696741, "percentage": 61.79, "elapsed_time": "1 day, 15:27:01", "remaining_time": "1 day, 0:23:32", "throughput": 5652.44, "total_tokens": 802766184}
+{"current_steps": 18360, "total_steps": 29704, "loss": 0.29856894016265867, "lr": 3.1876729035093616e-06, "epoch": 1.2361971451656342, "percentage": 61.81, "elapsed_time": "1 day, 15:27:36", "remaining_time": "1 day, 0:22:51", "throughput": 5652.59, "total_tokens": 802986112}
+{"current_steps": 18365, "total_steps": 29704, "loss": 0.29802355766296384, "lr": 3.185208879809366e-06, "epoch": 1.2365338001615944, "percentage": 61.83, "elapsed_time": "1 day, 15:28:20", "remaining_time": "1 day, 0:22:16", "throughput": 5652.43, "total_tokens": 803214056}
+{"current_steps": 18370, "total_steps": 29704, "loss": 0.3028829574584961, "lr": 3.1827453636099183e-06, "epoch": 1.2368704551575544, "percentage": 61.84, "elapsed_time": "1 day, 15:28:56", "remaining_time": "1 day, 0:21:36", "throughput": 5652.66, "total_tokens": 803449464}
+{"current_steps": 18375, "total_steps": 29704, "loss": 0.2902355194091797, "lr": 3.1802823555999356e-06, "epoch": 1.2372071101535147, "percentage": 61.86, "elapsed_time": "1 day, 15:29:33", "remaining_time": "1 day, 0:20:56", "throughput": 5652.76, "total_tokens": 803675088}
+{"current_steps": 18380, "total_steps": 29704, "loss": 0.32343668937683107, "lr": 3.1778198564681885e-06, "epoch": 1.2375437651494747, "percentage": 61.88, "elapsed_time": "1 day, 15:30:05", "remaining_time": "1 day, 0:20:13", "throughput": 5653.1, "total_tokens": 803903288}
+{"current_steps": 18385, "total_steps": 29704, "loss": 0.30882773399353025, "lr": 3.17535786690331e-06, "epoch": 1.237880420145435, "percentage": 61.89, "elapsed_time": "1 day, 15:30:46", "remaining_time": "1 day, 0:19:36", "throughput": 5653.02, "total_tokens": 804121616}
+{"current_steps": 18390, "total_steps": 29704, "loss": 0.304299521446228, "lr": 3.172896387593784e-06, "epoch": 1.238217075141395, "percentage": 61.91, "elapsed_time": "1 day, 15:31:28", "remaining_time": "1 day, 0:18:59", "throughput": 5652.94, "total_tokens": 804348264}
+{"current_steps": 18395, "total_steps": 29704, "loss": 0.31928482055664065, "lr": 3.1704354192279573e-06, "epoch": 1.2385537301373553, "percentage": 61.93, "elapsed_time": "1 day, 15:32:03", "remaining_time": "1 day, 0:18:18", "throughput": 5653.05, "total_tokens": 804564656}
+{"current_steps": 18400, "total_steps": 29704, "loss": 0.318212103843689, "lr": 3.1679749624940316e-06, "epoch": 1.2388903851333153, "percentage": 61.94, "elapsed_time": "1 day, 15:32:40", "remaining_time": "1 day, 0:17:38", "throughput": 5653.17, "total_tokens": 804786400}
+{"current_steps": 18405, "total_steps": 29704, "loss": 0.3026642084121704, "lr": 3.165515018080067e-06, "epoch": 1.2392270401292755, "percentage": 61.96, "elapsed_time": "1 day, 15:33:14", "remaining_time": "1 day, 0:16:57", "throughput": 5653.36, "total_tokens": 805005768}
+{"current_steps": 18410, "total_steps": 29704, "loss": 0.29436650276184084, "lr": 3.163055586673976e-06, "epoch": 1.2395636951252356, "percentage": 61.98, "elapsed_time": "1 day, 15:33:47", "remaining_time": "1 day, 0:16:15", "throughput": 5653.58, "total_tokens": 805226016}
+{"current_steps": 18415, "total_steps": 29704, "loss": 0.32002964019775393, "lr": 3.1605966689635348e-06, "epoch": 1.2399003501211958, "percentage": 62.0, "elapsed_time": "1 day, 15:34:24", "remaining_time": "1 day, 0:15:35", "throughput": 5653.71, "total_tokens": 805453752}
+{"current_steps": 18420, "total_steps": 29704, "loss": 0.30423684120178224, "lr": 3.1581382656363666e-06, "epoch": 1.2402370051171558, "percentage": 62.01, "elapsed_time": "1 day, 15:34:54", "remaining_time": "1 day, 0:14:51", "throughput": 5653.95, "total_tokens": 805658856}
+{"current_steps": 18425, "total_steps": 29704, "loss": 0.3185387372970581, "lr": 3.1556803773799616e-06, "epoch": 1.240573660113116, "percentage": 62.03, "elapsed_time": "1 day, 15:35:36", "remaining_time": "1 day, 0:14:15", "throughput": 5653.84, "total_tokens": 805880136}
+{"current_steps": 18430, "total_steps": 29704, "loss": 0.3022879123687744, "lr": 3.1532230048816564e-06, "epoch": 1.2409103151090761, "percentage": 62.05, "elapsed_time": "1 day, 15:36:15", "remaining_time": "1 day, 0:13:36", "throughput": 5653.81, "total_tokens": 806097056}
+{"current_steps": 18435, "total_steps": 29704, "loss": 0.3036804676055908, "lr": 3.150766148828651e-06, "epoch": 1.2412469701050364, "percentage": 62.06, "elapsed_time": "1 day, 15:36:53", "remaining_time": "1 day, 0:12:56", "throughput": 5653.87, "total_tokens": 806315480}
+{"current_steps": 18440, "total_steps": 29704, "loss": 0.2874735116958618, "lr": 3.1483098099079934e-06, "epoch": 1.2415836251009964, "percentage": 62.08, "elapsed_time": "1 day, 15:37:41", "remaining_time": "1 day, 0:12:24", "throughput": 5653.51, "total_tokens": 806538456}
+{"current_steps": 18445, "total_steps": 29704, "loss": 0.3029060363769531, "lr": 3.1458539888065937e-06, "epoch": 1.2419202800969567, "percentage": 62.1, "elapsed_time": "1 day, 15:38:22", "remaining_time": "1 day, 0:11:47", "throughput": 5653.36, "total_tokens": 806750424}
+{"current_steps": 18450, "total_steps": 29704, "loss": 0.29645981788635256, "lr": 3.143398686211212e-06, "epoch": 1.2422569350929167, "percentage": 62.11, "elapsed_time": "1 day, 15:39:00", "remaining_time": "1 day, 0:11:07", "throughput": 5653.34, "total_tokens": 806958552}
+{"current_steps": 18455, "total_steps": 29704, "loss": 0.3129459857940674, "lr": 3.1409439028084706e-06, "epoch": 1.242593590088877, "percentage": 62.13, "elapsed_time": "1 day, 15:39:39", "remaining_time": "1 day, 0:10:29", "throughput": 5653.3, "total_tokens": 807178464}
+{"current_steps": 18460, "total_steps": 29704, "loss": 0.28752105236053466, "lr": 3.1384896392848363e-06, "epoch": 1.2429302450848372, "percentage": 62.15, "elapsed_time": "1 day, 15:40:21", "remaining_time": "1 day, 0:09:52", "throughput": 5653.21, "total_tokens": 807397576}
+{"current_steps": 18465, "total_steps": 29704, "loss": 0.32807331085205077, "lr": 3.1360358963266404e-06, "epoch": 1.2432669000807972, "percentage": 62.16, "elapsed_time": "1 day, 15:40:52", "remaining_time": "1 day, 0:09:09", "throughput": 5653.56, "total_tokens": 807624936}
+{"current_steps": 18470, "total_steps": 29704, "loss": 0.3099084377288818, "lr": 3.1335826746200594e-06, "epoch": 1.2436035550767572, "percentage": 62.18, "elapsed_time": "1 day, 15:41:30", "remaining_time": "1 day, 0:08:30", "throughput": 5653.43, "total_tokens": 807820280}
+{"current_steps": 18475, "total_steps": 29704, "loss": 0.32409205436706545, "lr": 3.1311299748511335e-06, "epoch": 1.2439402100727175, "percentage": 62.2, "elapsed_time": "1 day, 15:42:11", "remaining_time": "1 day, 0:07:53", "throughput": 5653.34, "total_tokens": 808042832}
+{"current_steps": 18480, "total_steps": 29704, "loss": 0.29155697822570803, "lr": 3.1286777977057504e-06, "epoch": 1.2442768650686777, "percentage": 62.21, "elapsed_time": "1 day, 15:42:50", "remaining_time": "1 day, 0:07:14", "throughput": 5653.3, "total_tokens": 808252592}
+{"current_steps": 18485, "total_steps": 29704, "loss": 0.3249779224395752, "lr": 3.126226143869655e-06, "epoch": 1.2446135200646378, "percentage": 62.23, "elapsed_time": "1 day, 15:43:29", "remaining_time": "1 day, 0:06:35", "throughput": 5653.24, "total_tokens": 808465720}
+{"current_steps": 18490, "total_steps": 29704, "loss": 0.27743706703186033, "lr": 3.1237750140284424e-06, "epoch": 1.2449501750605978, "percentage": 62.25, "elapsed_time": "1 day, 15:44:03", "remaining_time": "1 day, 0:05:54", "throughput": 5653.39, "total_tokens": 808678680}
+{"current_steps": 18495, "total_steps": 29704, "loss": 0.30376696586608887, "lr": 3.1213244088675647e-06, "epoch": 1.245286830056558, "percentage": 62.26, "elapsed_time": "1 day, 15:44:41", "remaining_time": "1 day, 0:05:15", "throughput": 5653.38, "total_tokens": 808892656}
+{"current_steps": 18500, "total_steps": 29704, "loss": 0.317877197265625, "lr": 3.1188743290723246e-06, "epoch": 1.2456234850525183, "percentage": 62.28, "elapsed_time": "1 day, 15:45:25", "remaining_time": "1 day, 0:04:39", "throughput": 5653.18, "total_tokens": 809111464}
+{"current_steps": 18505, "total_steps": 29704, "loss": 0.3008237361907959, "lr": 3.116424775327882e-06, "epoch": 1.2459601400484783, "percentage": 62.3, "elapsed_time": "1 day, 15:46:13", "remaining_time": "1 day, 0:04:06", "throughput": 5652.8, "total_tokens": 809328360}
+{"current_steps": 18510, "total_steps": 29704, "loss": 0.3034308433532715, "lr": 3.1139757483192423e-06, "epoch": 1.2462967950444384, "percentage": 62.31, "elapsed_time": "1 day, 15:46:47", "remaining_time": "1 day, 0:03:25", "throughput": 5653.03, "total_tokens": 809557648}
+{"current_steps": 18515, "total_steps": 29704, "loss": 0.29793879985809324, "lr": 3.1115272487312727e-06, "epoch": 1.2466334500403986, "percentage": 62.33, "elapsed_time": "1 day, 15:47:19", "remaining_time": "1 day, 0:02:42", "throughput": 5653.25, "total_tokens": 809766040}
+{"current_steps": 18520, "total_steps": 29704, "loss": 0.29328839778900145, "lr": 3.109079277248684e-06, "epoch": 1.2469701050363589, "percentage": 62.35, "elapsed_time": "1 day, 15:47:58", "remaining_time": "1 day, 0:02:04", "throughput": 5653.14, "total_tokens": 809972912}
+{"current_steps": 18525, "total_steps": 29704, "loss": 0.32238657474517823, "lr": 3.1066318345560463e-06, "epoch": 1.2473067600323189, "percentage": 62.37, "elapsed_time": "1 day, 15:48:32", "remaining_time": "1 day, 0:01:22", "throughput": 5653.3, "total_tokens": 810187728}
+{"current_steps": 18530, "total_steps": 29704, "loss": 0.3177293300628662, "lr": 3.1041849213377782e-06, "epoch": 1.247643415028279, "percentage": 62.38, "elapsed_time": "1 day, 15:49:07", "remaining_time": "1 day, 0:00:41", "throughput": 5653.53, "total_tokens": 810419880}
+{"current_steps": 18535, "total_steps": 29704, "loss": 0.29890563488006594, "lr": 3.101738538278154e-06, "epoch": 1.2479800700242392, "percentage": 62.4, "elapsed_time": "1 day, 15:49:42", "remaining_time": "1 day, 0:00:01", "throughput": 5653.69, "total_tokens": 810642576}
+{"current_steps": 18540, "total_steps": 29704, "loss": 0.30640869140625, "lr": 3.0992926860612914e-06, "epoch": 1.2483167250201994, "percentage": 62.42, "elapsed_time": "1 day, 15:50:22", "remaining_time": "23:59:23", "throughput": 5653.6, "total_tokens": 810855296}
+{"current_steps": 18545, "total_steps": 29704, "loss": 0.2990746259689331, "lr": 3.0968473653711693e-06, "epoch": 1.2486533800161594, "percentage": 62.43, "elapsed_time": "1 day, 15:51:02", "remaining_time": "23:58:44", "throughput": 5653.72, "total_tokens": 811094136}
+{"current_steps": 18550, "total_steps": 29704, "loss": 0.3187594413757324, "lr": 3.094402576891613e-06, "epoch": 1.2489900350121195, "percentage": 62.45, "elapsed_time": "1 day, 15:51:37", "remaining_time": "23:58:04", "throughput": 5653.86, "total_tokens": 811315008}
+{"current_steps": 18555, "total_steps": 29704, "loss": 0.3195340156555176, "lr": 3.0919583213063004e-06, "epoch": 1.2493266900080797, "percentage": 62.47, "elapsed_time": "1 day, 15:52:11", "remaining_time": "23:57:22", "throughput": 5654.11, "total_tokens": 811546128}
+{"current_steps": 18560, "total_steps": 29704, "loss": 0.30277261734008787, "lr": 3.0895145992987574e-06, "epoch": 1.24966334500404, "percentage": 62.48, "elapsed_time": "1 day, 15:52:49", "remaining_time": "23:56:43", "throughput": 5654.24, "total_tokens": 811776888}
+{"current_steps": 18565, "total_steps": 29704, "loss": 0.2947868824005127, "lr": 3.087071411552366e-06, "epoch": 1.25, "percentage": 62.5, "elapsed_time": "1 day, 15:53:25", "remaining_time": "23:56:03", "throughput": 5654.32, "total_tokens": 811992848}
+{"current_steps": 18570, "total_steps": 29704, "loss": 0.31966290473937986, "lr": 3.084628758750353e-06, "epoch": 1.25033665499596, "percentage": 62.52, "elapsed_time": "1 day, 15:54:04", "remaining_time": "23:55:24", "throughput": 5654.32, "total_tokens": 812211480}
+{"current_steps": 18575, "total_steps": 29704, "loss": 0.3240236759185791, "lr": 3.082186641575801e-06, "epoch": 1.2506733099919203, "percentage": 62.53, "elapsed_time": "1 day, 15:54:46", "remaining_time": "23:54:48", "throughput": 5654.17, "total_tokens": 812427616}
+{"current_steps": 18580, "total_steps": 29704, "loss": 0.30055992603302, "lr": 3.0797450607116384e-06, "epoch": 1.2510099649878805, "percentage": 62.55, "elapsed_time": "1 day, 15:55:27", "remaining_time": "23:54:11", "throughput": 5654.06, "total_tokens": 812646144}
+{"current_steps": 18585, "total_steps": 29704, "loss": 0.31271209716796877, "lr": 3.0773040168406475e-06, "epoch": 1.2513466199838406, "percentage": 62.57, "elapsed_time": "1 day, 15:56:01", "remaining_time": "23:53:29", "throughput": 5654.34, "total_tokens": 812878472}
+{"current_steps": 18590, "total_steps": 29704, "loss": 0.31382598876953127, "lr": 3.074863510645456e-06, "epoch": 1.2516832749798006, "percentage": 62.58, "elapsed_time": "1 day, 15:56:41", "remaining_time": "23:52:51", "throughput": 5654.3, "total_tokens": 813099832}
+{"current_steps": 18595, "total_steps": 29704, "loss": 0.30039525032043457, "lr": 3.072423542808546e-06, "epoch": 1.2520199299757608, "percentage": 62.6, "elapsed_time": "1 day, 15:57:15", "remaining_time": "23:52:10", "throughput": 5654.48, "total_tokens": 813316560}
+{"current_steps": 18600, "total_steps": 29704, "loss": 0.3078362464904785, "lr": 3.0699841140122443e-06, "epoch": 1.252356584971721, "percentage": 62.62, "elapsed_time": "1 day, 15:57:57", "remaining_time": "23:51:33", "throughput": 5654.38, "total_tokens": 813538248}
+{"current_steps": 18605, "total_steps": 29704, "loss": 0.2891314744949341, "lr": 3.0675452249387327e-06, "epoch": 1.2526932399676811, "percentage": 62.63, "elapsed_time": "1 day, 15:58:36", "remaining_time": "23:50:55", "throughput": 5654.41, "total_tokens": 813765928}
+{"current_steps": 18610, "total_steps": 29704, "loss": 0.3373257637023926, "lr": 3.0651068762700354e-06, "epoch": 1.2530298949636411, "percentage": 62.65, "elapsed_time": "1 day, 15:59:16", "remaining_time": "23:50:16", "throughput": 5654.31, "total_tokens": 813974344}
+{"current_steps": 18615, "total_steps": 29704, "loss": 0.32206950187683103, "lr": 3.0626690686880313e-06, "epoch": 1.2533665499596014, "percentage": 62.67, "elapsed_time": "1 day, 15:59:54", "remaining_time": "23:49:38", "throughput": 5654.45, "total_tokens": 814211744}
+{"current_steps": 18620, "total_steps": 29704, "loss": 0.3022738218307495, "lr": 3.060231802874443e-06, "epoch": 1.2537032049555616, "percentage": 62.69, "elapsed_time": "1 day, 16:00:31", "remaining_time": "23:48:58", "throughput": 5654.6, "total_tokens": 814438872}
+{"current_steps": 18625, "total_steps": 29704, "loss": 0.30109267234802245, "lr": 3.0577950795108462e-06, "epoch": 1.2540398599515217, "percentage": 62.7, "elapsed_time": "1 day, 16:01:11", "remaining_time": "23:48:20", "throughput": 5654.58, "total_tokens": 814661808}
+{"current_steps": 18630, "total_steps": 29704, "loss": 0.2987940788269043, "lr": 3.055358899278662e-06, "epoch": 1.2543765149474817, "percentage": 62.72, "elapsed_time": "1 day, 16:01:49", "remaining_time": "23:47:41", "throughput": 5654.53, "total_tokens": 814871016}
+{"current_steps": 18635, "total_steps": 29704, "loss": 0.3024166107177734, "lr": 3.052923262859162e-06, "epoch": 1.254713169943442, "percentage": 62.74, "elapsed_time": "1 day, 16:02:29", "remaining_time": "23:47:03", "throughput": 5654.53, "total_tokens": 815098200}
+{"current_steps": 18640, "total_steps": 29704, "loss": 0.29818005561828614, "lr": 3.0504881709334603e-06, "epoch": 1.2550498249394022, "percentage": 62.75, "elapsed_time": "1 day, 16:03:09", "remaining_time": "23:46:25", "throughput": 5654.51, "total_tokens": 815319256}
+{"current_steps": 18645, "total_steps": 29704, "loss": 0.28137338161468506, "lr": 3.0480536241825263e-06, "epoch": 1.2553864799353622, "percentage": 62.77, "elapsed_time": "1 day, 16:03:50", "remaining_time": "23:45:48", "throughput": 5654.41, "total_tokens": 815537448}
+{"current_steps": 18650, "total_steps": 29704, "loss": 0.31012353897094724, "lr": 3.0456196232871715e-06, "epoch": 1.2557231349313223, "percentage": 62.79, "elapsed_time": "1 day, 16:04:28", "remaining_time": "23:45:09", "throughput": 5654.37, "total_tokens": 815749224}
+{"current_steps": 18655, "total_steps": 29704, "loss": 0.30425825119018557, "lr": 3.043186168928059e-06, "epoch": 1.2560597899272825, "percentage": 62.8, "elapsed_time": "1 day, 16:05:16", "remaining_time": "23:44:35", "throughput": 5653.93, "total_tokens": 815954440}
+{"current_steps": 18660, "total_steps": 29704, "loss": 0.2835689067840576, "lr": 3.040753261785692e-06, "epoch": 1.2563964449232428, "percentage": 62.82, "elapsed_time": "1 day, 16:05:53", "remaining_time": "23:43:56", "throughput": 5653.93, "total_tokens": 816164248}
+{"current_steps": 18665, "total_steps": 29704, "loss": 0.28765387535095216, "lr": 3.038320902540429e-06, "epoch": 1.2567330999192028, "percentage": 62.84, "elapsed_time": "1 day, 16:06:34", "remaining_time": "23:43:18", "throughput": 5653.82, "total_tokens": 816379936}
+{"current_steps": 18670, "total_steps": 29704, "loss": 0.303347110748291, "lr": 3.0358890918724706e-06, "epoch": 1.2570697549151628, "percentage": 62.85, "elapsed_time": "1 day, 16:07:21", "remaining_time": "23:42:45", "throughput": 5653.58, "total_tokens": 816611832}
+{"current_steps": 18675, "total_steps": 29704, "loss": 0.30634298324584963, "lr": 3.0334578304618646e-06, "epoch": 1.257406409911123, "percentage": 62.87, "elapsed_time": "1 day, 16:08:03", "remaining_time": "23:42:08", "throughput": 5653.33, "total_tokens": 816814520}
+{"current_steps": 18680, "total_steps": 29704, "loss": 0.3063030481338501, "lr": 3.0310271189885037e-06, "epoch": 1.2577430649070833, "percentage": 62.89, "elapsed_time": "1 day, 16:08:37", "remaining_time": "23:41:26", "throughput": 5653.47, "total_tokens": 817024696}
+{"current_steps": 18685, "total_steps": 29704, "loss": 0.296706485748291, "lr": 3.0285969581321328e-06, "epoch": 1.2580797199030433, "percentage": 62.9, "elapsed_time": "1 day, 16:09:14", "remaining_time": "23:40:47", "throughput": 5653.54, "total_tokens": 817244872}
+{"current_steps": 18690, "total_steps": 29704, "loss": 0.3513062000274658, "lr": 3.0261673485723343e-06, "epoch": 1.2584163748990034, "percentage": 62.92, "elapsed_time": "1 day, 16:09:52", "remaining_time": "23:40:08", "throughput": 5653.51, "total_tokens": 817456408}
+{"current_steps": 18695, "total_steps": 29704, "loss": 0.28501205444335936, "lr": 3.023738290988544e-06, "epoch": 1.2587530298949636, "percentage": 62.94, "elapsed_time": "1 day, 16:10:34", "remaining_time": "23:39:31", "throughput": 5653.44, "total_tokens": 817681112}
+{"current_steps": 18700, "total_steps": 29704, "loss": 0.30846569538116453, "lr": 3.021309786060037e-06, "epoch": 1.2590896848909239, "percentage": 62.95, "elapsed_time": "1 day, 16:11:14", "remaining_time": "23:38:53", "throughput": 5653.3, "total_tokens": 817885488}
+{"current_steps": 18705, "total_steps": 29704, "loss": 0.290647029876709, "lr": 3.018881834465941e-06, "epoch": 1.259426339886884, "percentage": 62.97, "elapsed_time": "1 day, 16:11:52", "remaining_time": "23:38:14", "throughput": 5653.28, "total_tokens": 818098808}
+{"current_steps": 18710, "total_steps": 29704, "loss": 0.3127418041229248, "lr": 3.016454436885221e-06, "epoch": 1.259762994882844, "percentage": 62.99, "elapsed_time": "1 day, 16:12:28", "remaining_time": "23:37:34", "throughput": 5653.38, "total_tokens": 818317480}
+{"current_steps": 18715, "total_steps": 29704, "loss": 0.3062695026397705, "lr": 3.0140275939966934e-06, "epoch": 1.2600996498788042, "percentage": 63.0, "elapsed_time": "1 day, 16:13:09", "remaining_time": "23:36:57", "throughput": 5653.29, "total_tokens": 818539128}
+{"current_steps": 18720, "total_steps": 29704, "loss": 0.3510124921798706, "lr": 3.0116013064790166e-06, "epoch": 1.2604363048747644, "percentage": 63.02, "elapsed_time": "1 day, 16:13:42", "remaining_time": "23:36:14", "throughput": 5653.61, "total_tokens": 818770136}
+{"current_steps": 18725, "total_steps": 29704, "loss": 0.31281888484954834, "lr": 3.0091755750106937e-06, "epoch": 1.2607729598707245, "percentage": 63.04, "elapsed_time": "1 day, 16:14:17", "remaining_time": "23:35:34", "throughput": 5653.72, "total_tokens": 818987192}
+{"current_steps": 18730, "total_steps": 29704, "loss": 0.2828061580657959, "lr": 3.006750400270072e-06, "epoch": 1.2611096148666845, "percentage": 63.06, "elapsed_time": "1 day, 16:14:53", "remaining_time": "23:34:53", "throughput": 5653.84, "total_tokens": 819201568}
+{"current_steps": 18735, "total_steps": 29704, "loss": 0.2915013790130615, "lr": 3.0043257829353466e-06, "epoch": 1.2614462698626447, "percentage": 63.07, "elapsed_time": "1 day, 16:15:32", "remaining_time": "23:34:15", "throughput": 5653.84, "total_tokens": 819423736}
+{"current_steps": 18740, "total_steps": 29704, "loss": 0.3179666042327881, "lr": 3.0019017236845504e-06, "epoch": 1.261782924858605, "percentage": 63.09, "elapsed_time": "1 day, 16:16:09", "remaining_time": "23:33:35", "throughput": 5653.89, "total_tokens": 819640880}
+{"current_steps": 18745, "total_steps": 29704, "loss": 0.3084989547729492, "lr": 2.9994782231955673e-06, "epoch": 1.262119579854565, "percentage": 63.11, "elapsed_time": "1 day, 16:16:46", "remaining_time": "23:32:55", "throughput": 5653.96, "total_tokens": 819858960}
+{"current_steps": 18750, "total_steps": 29704, "loss": 0.28854634761810305, "lr": 2.9970552821461185e-06, "epoch": 1.262456234850525, "percentage": 63.12, "elapsed_time": "1 day, 16:17:25", "remaining_time": "23:32:17", "throughput": 5653.87, "total_tokens": 820070184}
+{"current_steps": 18755, "total_steps": 29704, "loss": 0.2994524955749512, "lr": 2.994632901213775e-06, "epoch": 1.2627928898464853, "percentage": 63.14, "elapsed_time": "1 day, 16:18:03", "remaining_time": "23:31:38", "throughput": 5653.85, "total_tokens": 820280600}
+{"current_steps": 18760, "total_steps": 29704, "loss": 0.3167592763900757, "lr": 2.9922110810759443e-06, "epoch": 1.2631295448424456, "percentage": 63.16, "elapsed_time": "1 day, 16:18:47", "remaining_time": "23:31:03", "throughput": 5653.58, "total_tokens": 820492832}
+{"current_steps": 18765, "total_steps": 29704, "loss": 0.3189209461212158, "lr": 2.989789822409883e-06, "epoch": 1.2634661998384056, "percentage": 63.17, "elapsed_time": "1 day, 16:19:22", "remaining_time": "23:30:21", "throughput": 5653.8, "total_tokens": 820717400}
+{"current_steps": 18770, "total_steps": 29704, "loss": 0.3269504547119141, "lr": 2.9873691258926864e-06, "epoch": 1.2638028548343656, "percentage": 63.19, "elapsed_time": "1 day, 16:19:58", "remaining_time": "23:29:41", "throughput": 5654.04, "total_tokens": 820956960}
+{"current_steps": 18775, "total_steps": 29704, "loss": 0.3042433261871338, "lr": 2.9849489922012985e-06, "epoch": 1.2641395098303259, "percentage": 63.21, "elapsed_time": "1 day, 16:20:39", "remaining_time": "23:29:04", "throughput": 5653.89, "total_tokens": 821165096}
+{"current_steps": 18780, "total_steps": 29704, "loss": 0.338847017288208, "lr": 2.982529422012498e-06, "epoch": 1.264476164826286, "percentage": 63.22, "elapsed_time": "1 day, 16:21:17", "remaining_time": "23:28:25", "throughput": 5653.98, "total_tokens": 821398992}
+{"current_steps": 18785, "total_steps": 29704, "loss": 0.2665702819824219, "lr": 2.9801104160029127e-06, "epoch": 1.2648128198222461, "percentage": 63.24, "elapsed_time": "1 day, 16:21:59", "remaining_time": "23:27:48", "throughput": 5653.79, "total_tokens": 821606024}
+{"current_steps": 18790, "total_steps": 29704, "loss": 0.31261477470397947, "lr": 2.9776919748490063e-06, "epoch": 1.2651494748182062, "percentage": 63.26, "elapsed_time": "1 day, 16:22:39", "remaining_time": "23:27:10", "throughput": 5653.71, "total_tokens": 821822040}
+{"current_steps": 18795, "total_steps": 29704, "loss": 0.32703323364257814, "lr": 2.975274099227092e-06, "epoch": 1.2654861298141664, "percentage": 63.27, "elapsed_time": "1 day, 16:23:23", "remaining_time": "23:26:35", "throughput": 5653.42, "total_tokens": 822027760}
+{"current_steps": 18800, "total_steps": 29704, "loss": 0.31384882926940916, "lr": 2.9728567898133183e-06, "epoch": 1.2658227848101267, "percentage": 63.29, "elapsed_time": "1 day, 16:24:07", "remaining_time": "23:25:59", "throughput": 5653.23, "total_tokens": 822249072}
+{"current_steps": 18805, "total_steps": 29704, "loss": 0.3009427309036255, "lr": 2.9704400472836816e-06, "epoch": 1.2661594398060867, "percentage": 63.31, "elapsed_time": "1 day, 16:24:51", "remaining_time": "23:25:24", "throughput": 5653.02, "total_tokens": 822468088}
+{"current_steps": 18810, "total_steps": 29704, "loss": 0.33100879192352295, "lr": 2.9680238723140108e-06, "epoch": 1.266496094802047, "percentage": 63.32, "elapsed_time": "1 day, 16:25:31", "remaining_time": "23:24:46", "throughput": 5652.95, "total_tokens": 822682520}
+{"current_steps": 18815, "total_steps": 29704, "loss": 0.3179032325744629, "lr": 2.9656082655799866e-06, "epoch": 1.266832749798007, "percentage": 63.34, "elapsed_time": "1 day, 16:26:09", "remaining_time": "23:24:06", "throughput": 5652.93, "total_tokens": 822894336}
+{"current_steps": 18820, "total_steps": 29704, "loss": 0.31571259498596194, "lr": 2.9631932277571225e-06, "epoch": 1.2671694047939672, "percentage": 63.36, "elapsed_time": "1 day, 16:26:42", "remaining_time": "23:23:24", "throughput": 5653.18, "total_tokens": 823116688}
+{"current_steps": 18825, "total_steps": 29704, "loss": 0.2941970109939575, "lr": 2.9607787595207794e-06, "epoch": 1.2675060597899273, "percentage": 63.38, "elapsed_time": "1 day, 16:27:22", "remaining_time": "23:22:47", "throughput": 5653.11, "total_tokens": 823334680}
+{"current_steps": 18830, "total_steps": 29704, "loss": 0.2951462507247925, "lr": 2.9583648615461514e-06, "epoch": 1.2678427147858875, "percentage": 63.39, "elapsed_time": "1 day, 16:28:04", "remaining_time": "23:22:10", "throughput": 5652.95, "total_tokens": 823548976}
+{"current_steps": 18835, "total_steps": 29704, "loss": 0.32911660671234133, "lr": 2.955951534508281e-06, "epoch": 1.2681793697818475, "percentage": 63.41, "elapsed_time": "1 day, 16:28:40", "remaining_time": "23:21:29", "throughput": 5653.02, "total_tokens": 823759784}
+{"current_steps": 18840, "total_steps": 29704, "loss": 0.3123172283172607, "lr": 2.9535387790820452e-06, "epoch": 1.2685160247778078, "percentage": 63.43, "elapsed_time": "1 day, 16:29:23", "remaining_time": "23:20:54", "throughput": 5652.91, "total_tokens": 823989856}
+{"current_steps": 18845, "total_steps": 29704, "loss": 0.29312863349914553, "lr": 2.951126595942165e-06, "epoch": 1.2688526797737678, "percentage": 63.44, "elapsed_time": "1 day, 16:30:00", "remaining_time": "23:20:14", "throughput": 5653.05, "total_tokens": 824217960}
+{"current_steps": 18850, "total_steps": 29704, "loss": 0.2757392883300781, "lr": 2.948714985763197e-06, "epoch": 1.269189334769728, "percentage": 63.46, "elapsed_time": "1 day, 16:30:41", "remaining_time": "23:19:37", "throughput": 5652.85, "total_tokens": 824421568}
+{"current_steps": 18855, "total_steps": 29704, "loss": 0.3131323575973511, "lr": 2.9463039492195445e-06, "epoch": 1.269525989765688, "percentage": 63.48, "elapsed_time": "1 day, 16:31:20", "remaining_time": "23:18:58", "throughput": 5652.95, "total_tokens": 824655760}
+{"current_steps": 18860, "total_steps": 29704, "loss": 0.2902811527252197, "lr": 2.9438934869854417e-06, "epoch": 1.2698626447616483, "percentage": 63.49, "elapsed_time": "1 day, 16:32:03", "remaining_time": "23:18:22", "throughput": 5652.81, "total_tokens": 824879032}
+{"current_steps": 18865, "total_steps": 29704, "loss": 0.2894813060760498, "lr": 2.9414835997349705e-06, "epoch": 1.2701992997576084, "percentage": 63.51, "elapsed_time": "1 day, 16:32:40", "remaining_time": "23:17:42", "throughput": 5652.89, "total_tokens": 825095288}
+{"current_steps": 18870, "total_steps": 29704, "loss": 0.32269885540008547, "lr": 2.939074288142045e-06, "epoch": 1.2705359547535686, "percentage": 63.53, "elapsed_time": "1 day, 16:33:18", "remaining_time": "23:17:03", "throughput": 5652.97, "total_tokens": 825326376}
+{"current_steps": 18875, "total_steps": 29704, "loss": 0.29350924491882324, "lr": 2.9366655528804257e-06, "epoch": 1.2708726097495286, "percentage": 63.54, "elapsed_time": "1 day, 16:34:02", "remaining_time": "23:16:27", "throughput": 5652.79, "total_tokens": 825544816}
+{"current_steps": 18880, "total_steps": 29704, "loss": 0.2879796504974365, "lr": 2.934257394623702e-06, "epoch": 1.271209264745489, "percentage": 63.56, "elapsed_time": "1 day, 16:34:39", "remaining_time": "23:15:48", "throughput": 5652.88, "total_tokens": 825768752}
+{"current_steps": 18885, "total_steps": 29704, "loss": 0.32280969619750977, "lr": 2.9318498140453134e-06, "epoch": 1.271545919741449, "percentage": 63.58, "elapsed_time": "1 day, 16:35:22", "remaining_time": "23:15:11", "throughput": 5652.73, "total_tokens": 825987712}
+{"current_steps": 18890, "total_steps": 29704, "loss": 0.3143021106719971, "lr": 2.929442811818527e-06, "epoch": 1.2718825747374092, "percentage": 63.59, "elapsed_time": "1 day, 16:36:01", "remaining_time": "23:14:33", "throughput": 5652.62, "total_tokens": 826196048}
+{"current_steps": 18895, "total_steps": 29704, "loss": 0.29908573627471924, "lr": 2.927036388616457e-06, "epoch": 1.2722192297333692, "percentage": 63.61, "elapsed_time": "1 day, 16:36:35", "remaining_time": "23:13:52", "throughput": 5652.68, "total_tokens": 826399384}
+{"current_steps": 18900, "total_steps": 29704, "loss": 0.3228297233581543, "lr": 2.9246305451120494e-06, "epoch": 1.2725558847293295, "percentage": 63.63, "elapsed_time": "1 day, 16:37:14", "remaining_time": "23:13:13", "throughput": 5652.61, "total_tokens": 826603896}
+{"current_steps": 18905, "total_steps": 29704, "loss": 0.3007101058959961, "lr": 2.922225281978095e-06, "epoch": 1.2728925397252895, "percentage": 63.64, "elapsed_time": "1 day, 16:37:59", "remaining_time": "23:12:38", "throughput": 5652.33, "total_tokens": 826821328}
+{"current_steps": 18910, "total_steps": 29704, "loss": 0.3024411916732788, "lr": 2.919820599887212e-06, "epoch": 1.2732291947212497, "percentage": 63.66, "elapsed_time": "1 day, 16:38:41", "remaining_time": "23:12:01", "throughput": 5652.21, "total_tokens": 827039792}
+{"current_steps": 18915, "total_steps": 29704, "loss": 0.30447540283203123, "lr": 2.917416499511866e-06, "epoch": 1.2735658497172098, "percentage": 63.68, "elapsed_time": "1 day, 16:39:23", "remaining_time": "23:11:24", "throughput": 5651.98, "total_tokens": 827242576}
+{"current_steps": 18920, "total_steps": 29704, "loss": 0.288972806930542, "lr": 2.9150129815243553e-06, "epoch": 1.27390250471317, "percentage": 63.7, "elapsed_time": "1 day, 16:40:04", "remaining_time": "23:10:47", "throughput": 5651.87, "total_tokens": 827457896}
+{"current_steps": 18925, "total_steps": 29704, "loss": 0.2999333620071411, "lr": 2.912610046596816e-06, "epoch": 1.27423915970913, "percentage": 63.71, "elapsed_time": "1 day, 16:40:46", "remaining_time": "23:10:10", "throughput": 5651.68, "total_tokens": 827666848}
+{"current_steps": 18930, "total_steps": 29704, "loss": 0.3152765274047852, "lr": 2.91020769540122e-06, "epoch": 1.2745758147050903, "percentage": 63.73, "elapsed_time": "1 day, 16:41:25", "remaining_time": "23:09:32", "throughput": 5651.52, "total_tokens": 827866000}
+{"current_steps": 18935, "total_steps": 29704, "loss": 0.2963433027267456, "lr": 2.9078059286093795e-06, "epoch": 1.2749124697010503, "percentage": 63.75, "elapsed_time": "1 day, 16:42:04", "remaining_time": "23:08:53", "throughput": 5651.44, "total_tokens": 828072488}
+{"current_steps": 18940, "total_steps": 29704, "loss": 0.28579273223876955, "lr": 2.905404746892937e-06, "epoch": 1.2752491246970106, "percentage": 63.76, "elapsed_time": "1 day, 16:42:40", "remaining_time": "23:08:13", "throughput": 5651.49, "total_tokens": 828287296}
+{"current_steps": 18945, "total_steps": 29704, "loss": 0.3039486169815063, "lr": 2.903004150923382e-06, "epoch": 1.2755857796929706, "percentage": 63.78, "elapsed_time": "1 day, 16:43:18", "remaining_time": "23:07:34", "throughput": 5651.57, "total_tokens": 828510696}
+{"current_steps": 18950, "total_steps": 29704, "loss": 0.3282867431640625, "lr": 2.9006041413720252e-06, "epoch": 1.2759224346889309, "percentage": 63.8, "elapsed_time": "1 day, 16:44:05", "remaining_time": "23:07:00", "throughput": 5651.28, "total_tokens": 828734304}
+{"current_steps": 18955, "total_steps": 29704, "loss": 0.30024127960205077, "lr": 2.8982047189100276e-06, "epoch": 1.2762590896848909, "percentage": 63.81, "elapsed_time": "1 day, 16:44:40", "remaining_time": "23:06:19", "throughput": 5651.54, "total_tokens": 828967864}
+{"current_steps": 18960, "total_steps": 29704, "loss": 0.3082295894622803, "lr": 2.895805884208378e-06, "epoch": 1.2765957446808511, "percentage": 63.83, "elapsed_time": "1 day, 16:45:17", "remaining_time": "23:05:40", "throughput": 5651.45, "total_tokens": 829167224}
+{"current_steps": 18965, "total_steps": 29704, "loss": 0.31248860359191893, "lr": 2.8934076379379016e-06, "epoch": 1.2769323996768112, "percentage": 63.85, "elapsed_time": "1 day, 16:45:55", "remaining_time": "23:05:00", "throughput": 5651.56, "total_tokens": 829396808}
+{"current_steps": 18970, "total_steps": 29704, "loss": 0.3060382127761841, "lr": 2.891009980769266e-06, "epoch": 1.2772690546727714, "percentage": 63.86, "elapsed_time": "1 day, 16:46:38", "remaining_time": "23:04:24", "throughput": 5651.43, "total_tokens": 829619304}
+{"current_steps": 18975, "total_steps": 29704, "loss": 0.29961209297180175, "lr": 2.88861291337296e-06, "epoch": 1.2776057096687314, "percentage": 63.88, "elapsed_time": "1 day, 16:47:16", "remaining_time": "23:03:45", "throughput": 5651.43, "total_tokens": 829837784}
+{"current_steps": 18980, "total_steps": 29704, "loss": 0.2904377460479736, "lr": 2.8862164364193236e-06, "epoch": 1.2779423646646917, "percentage": 63.9, "elapsed_time": "1 day, 16:47:53", "remaining_time": "23:03:05", "throughput": 5651.53, "total_tokens": 830058208}
+{"current_steps": 18985, "total_steps": 29704, "loss": 0.3134331703186035, "lr": 2.8838205505785168e-06, "epoch": 1.2782790196606517, "percentage": 63.91, "elapsed_time": "1 day, 16:48:34", "remaining_time": "23:02:28", "throughput": 5651.48, "total_tokens": 830283008}
+{"current_steps": 18990, "total_steps": 29704, "loss": 0.2921187400817871, "lr": 2.881425256520547e-06, "epoch": 1.278615674656612, "percentage": 63.93, "elapsed_time": "1 day, 16:49:11", "remaining_time": "23:01:48", "throughput": 5651.54, "total_tokens": 830500192}
+{"current_steps": 18995, "total_steps": 29704, "loss": 0.26155910491943357, "lr": 2.8790305549152487e-06, "epoch": 1.278952329652572, "percentage": 63.95, "elapsed_time": "1 day, 16:49:50", "remaining_time": "23:01:10", "throughput": 5651.54, "total_tokens": 830725544}
+{"current_steps": 19000, "total_steps": 29704, "loss": 0.3240873098373413, "lr": 2.876636446432292e-06, "epoch": 1.2792889846485322, "percentage": 63.96, "elapsed_time": "1 day, 16:50:25", "remaining_time": "23:00:29", "throughput": 5651.73, "total_tokens": 830948824}
+{"current_steps": 19005, "total_steps": 29704, "loss": 0.30750136375427245, "lr": 2.8742429317411826e-06, "epoch": 1.2796256396444923, "percentage": 63.98, "elapsed_time": "1 day, 16:51:02", "remaining_time": "22:59:49", "throughput": 5651.84, "total_tokens": 831173768}
+{"current_steps": 19010, "total_steps": 29704, "loss": 0.30974340438842773, "lr": 2.8718500115112585e-06, "epoch": 1.2799622946404525, "percentage": 64.0, "elapsed_time": "1 day, 16:51:36", "remaining_time": "22:59:08", "throughput": 5651.93, "total_tokens": 831380944}
+{"current_steps": 19015, "total_steps": 29704, "loss": 0.33095221519470214, "lr": 2.869457686411691e-06, "epoch": 1.2802989496364126, "percentage": 64.01, "elapsed_time": "1 day, 16:52:17", "remaining_time": "22:58:31", "throughput": 5651.77, "total_tokens": 831589040}
+{"current_steps": 19020, "total_steps": 29704, "loss": 0.30541186332702636, "lr": 2.8670659571114912e-06, "epoch": 1.2806356046323728, "percentage": 64.03, "elapsed_time": "1 day, 16:52:56", "remaining_time": "22:57:52", "throughput": 5651.85, "total_tokens": 831816448}
+{"current_steps": 19025, "total_steps": 29704, "loss": 0.2939921855926514, "lr": 2.864674824279491e-06, "epoch": 1.2809722596283328, "percentage": 64.05, "elapsed_time": "1 day, 16:53:33", "remaining_time": "22:57:12", "throughput": 5651.9, "total_tokens": 832035736}
+{"current_steps": 19030, "total_steps": 29704, "loss": 0.2885020971298218, "lr": 2.862284288584371e-06, "epoch": 1.281308914624293, "percentage": 64.07, "elapsed_time": "1 day, 16:54:11", "remaining_time": "22:56:33", "throughput": 5652.01, "total_tokens": 832264232}
+{"current_steps": 19035, "total_steps": 29704, "loss": 0.300329327583313, "lr": 2.8598943506946297e-06, "epoch": 1.2816455696202531, "percentage": 64.08, "elapsed_time": "1 day, 16:54:48", "remaining_time": "22:55:54", "throughput": 5652.05, "total_tokens": 832482744}
+{"current_steps": 19040, "total_steps": 29704, "loss": 0.277694034576416, "lr": 2.85750501127861e-06, "epoch": 1.2819822246162134, "percentage": 64.1, "elapsed_time": "1 day, 16:55:29", "remaining_time": "22:55:17", "throughput": 5651.93, "total_tokens": 832697464}
+{"current_steps": 19045, "total_steps": 29704, "loss": 0.30247793197631834, "lr": 2.8551162710044827e-06, "epoch": 1.2823188796121734, "percentage": 64.12, "elapsed_time": "1 day, 16:56:05", "remaining_time": "22:54:36", "throughput": 5652.0, "total_tokens": 832912640}
+{"current_steps": 19050, "total_steps": 29704, "loss": 0.32385740280151365, "lr": 2.8527281305402503e-06, "epoch": 1.2826555346081336, "percentage": 64.13, "elapsed_time": "1 day, 16:56:43", "remaining_time": "22:53:57", "throughput": 5652.07, "total_tokens": 833133224}
+{"current_steps": 19055, "total_steps": 29704, "loss": 0.29269657135009763, "lr": 2.8503405905537496e-06, "epoch": 1.2829921896040937, "percentage": 64.15, "elapsed_time": "1 day, 16:57:22", "remaining_time": "22:53:18", "throughput": 5652.03, "total_tokens": 833346936}
+{"current_steps": 19060, "total_steps": 29704, "loss": 0.3061253309249878, "lr": 2.847953651712647e-06, "epoch": 1.283328844600054, "percentage": 64.17, "elapsed_time": "1 day, 16:57:57", "remaining_time": "22:52:38", "throughput": 5652.26, "total_tokens": 833581464}
+{"current_steps": 19065, "total_steps": 29704, "loss": 0.30112695693969727, "lr": 2.8455673146844426e-06, "epoch": 1.283665499596014, "percentage": 64.18, "elapsed_time": "1 day, 16:58:41", "remaining_time": "22:52:02", "throughput": 5651.95, "total_tokens": 833784496}
+{"current_steps": 19070, "total_steps": 29704, "loss": 0.3190593719482422, "lr": 2.8431815801364717e-06, "epoch": 1.2840021545919742, "percentage": 64.2, "elapsed_time": "1 day, 16:59:23", "remaining_time": "22:51:25", "throughput": 5651.81, "total_tokens": 834001896}
+{"current_steps": 19075, "total_steps": 29704, "loss": 0.3083981513977051, "lr": 2.8407964487358906e-06, "epoch": 1.2843388095879342, "percentage": 64.22, "elapsed_time": "1 day, 17:00:06", "remaining_time": "22:50:49", "throughput": 5651.71, "total_tokens": 834226528}
+{"current_steps": 19080, "total_steps": 29704, "loss": 0.30108425617218015, "lr": 2.838411921149702e-06, "epoch": 1.2846754645838945, "percentage": 64.23, "elapsed_time": "1 day, 17:00:43", "remaining_time": "22:50:09", "throughput": 5651.65, "total_tokens": 834428112}
+{"current_steps": 19085, "total_steps": 29704, "loss": 0.3017698287963867, "lr": 2.836027998044722e-06, "epoch": 1.2850121195798545, "percentage": 64.25, "elapsed_time": "1 day, 17:01:27", "remaining_time": "22:49:34", "throughput": 5651.39, "total_tokens": 834638200}
+{"current_steps": 19090, "total_steps": 29704, "loss": 0.3067366600036621, "lr": 2.8336446800876143e-06, "epoch": 1.2853487745758148, "percentage": 64.27, "elapsed_time": "1 day, 17:02:02", "remaining_time": "22:48:53", "throughput": 5651.36, "total_tokens": 834833144}
+{"current_steps": 19095, "total_steps": 29704, "loss": 0.2865460872650146, "lr": 2.8312619679448637e-06, "epoch": 1.2856854295717748, "percentage": 64.28, "elapsed_time": "1 day, 17:02:42", "remaining_time": "22:48:15", "throughput": 5651.35, "total_tokens": 835055128}
+{"current_steps": 19100, "total_steps": 29704, "loss": 0.31798462867736815, "lr": 2.8288798622827884e-06, "epoch": 1.286022084567735, "percentage": 64.3, "elapsed_time": "1 day, 17:03:22", "remaining_time": "22:47:37", "throughput": 5651.14, "total_tokens": 835252248}
+{"current_steps": 19105, "total_steps": 29704, "loss": 0.316135311126709, "lr": 2.826498363767536e-06, "epoch": 1.286358739563695, "percentage": 64.32, "elapsed_time": "1 day, 17:04:00", "remaining_time": "22:46:58", "throughput": 5651.04, "total_tokens": 835451680}
+{"current_steps": 19110, "total_steps": 29704, "loss": 0.3065540313720703, "lr": 2.8241174730650855e-06, "epoch": 1.2866953945596553, "percentage": 64.33, "elapsed_time": "1 day, 17:04:40", "remaining_time": "22:46:20", "throughput": 5651.02, "total_tokens": 835675080}
+{"current_steps": 19115, "total_steps": 29704, "loss": 0.29743661880493166, "lr": 2.821737190841244e-06, "epoch": 1.2870320495556153, "percentage": 64.35, "elapsed_time": "1 day, 17:05:15", "remaining_time": "22:45:39", "throughput": 5651.14, "total_tokens": 835891040}
+{"current_steps": 19120, "total_steps": 29704, "loss": 0.2965965986251831, "lr": 2.8193575177616537e-06, "epoch": 1.2873687045515756, "percentage": 64.37, "elapsed_time": "1 day, 17:05:59", "remaining_time": "22:45:04", "throughput": 5650.94, "total_tokens": 836111592}
+{"current_steps": 19125, "total_steps": 29704, "loss": 0.2831472396850586, "lr": 2.816978454491777e-06, "epoch": 1.2877053595475356, "percentage": 64.39, "elapsed_time": "1 day, 17:06:34", "remaining_time": "22:44:23", "throughput": 5651.18, "total_tokens": 836342224}
+{"current_steps": 19130, "total_steps": 29704, "loss": 0.31551055908203124, "lr": 2.814600001696919e-06, "epoch": 1.2880420145434959, "percentage": 64.4, "elapsed_time": "1 day, 17:07:10", "remaining_time": "22:43:43", "throughput": 5651.22, "total_tokens": 836555312}
+{"current_steps": 19135, "total_steps": 29704, "loss": 0.3081847906112671, "lr": 2.812222160042196e-06, "epoch": 1.288378669539456, "percentage": 64.42, "elapsed_time": "1 day, 17:07:51", "remaining_time": "22:43:05", "throughput": 5651.18, "total_tokens": 836781384}
+{"current_steps": 19140, "total_steps": 29704, "loss": 0.31632413864135744, "lr": 2.809844930192573e-06, "epoch": 1.2887153245354162, "percentage": 64.44, "elapsed_time": "1 day, 17:08:30", "remaining_time": "22:42:27", "throughput": 5651.17, "total_tokens": 836996448}
+{"current_steps": 19145, "total_steps": 29704, "loss": 0.3241159677505493, "lr": 2.807468312812831e-06, "epoch": 1.2890519795313762, "percentage": 64.45, "elapsed_time": "1 day, 17:09:05", "remaining_time": "22:41:46", "throughput": 5651.27, "total_tokens": 837211800}
+{"current_steps": 19150, "total_steps": 29704, "loss": 0.2902689933776855, "lr": 2.805092308567584e-06, "epoch": 1.2893886345273364, "percentage": 64.47, "elapsed_time": "1 day, 17:09:37", "remaining_time": "22:41:04", "throughput": 5651.58, "total_tokens": 837439120}
+{"current_steps": 19155, "total_steps": 29704, "loss": 0.2991164207458496, "lr": 2.802716918121273e-06, "epoch": 1.2897252895232965, "percentage": 64.49, "elapsed_time": "1 day, 17:10:22", "remaining_time": "22:40:28", "throughput": 5651.39, "total_tokens": 837665040}
+{"current_steps": 19160, "total_steps": 29704, "loss": 0.29270741939544676, "lr": 2.8003421421381704e-06, "epoch": 1.2900619445192567, "percentage": 64.5, "elapsed_time": "1 day, 17:10:55", "remaining_time": "22:39:46", "throughput": 5651.69, "total_tokens": 837893216}
+{"current_steps": 19165, "total_steps": 29704, "loss": 0.3194019317626953, "lr": 2.797967981282371e-06, "epoch": 1.2903985995152167, "percentage": 64.52, "elapsed_time": "1 day, 17:11:37", "remaining_time": "22:39:10", "throughput": 5651.57, "total_tokens": 838115864}
+{"current_steps": 19170, "total_steps": 29704, "loss": 0.29573659896850585, "lr": 2.795594436217808e-06, "epoch": 1.290735254511177, "percentage": 64.54, "elapsed_time": "1 day, 17:12:18", "remaining_time": "22:38:32", "throughput": 5651.48, "total_tokens": 838332696}
+{"current_steps": 19175, "total_steps": 29704, "loss": 0.29705004692077636, "lr": 2.7932215076082277e-06, "epoch": 1.291071909507137, "percentage": 64.55, "elapsed_time": "1 day, 17:13:02", "remaining_time": "22:37:57", "throughput": 5651.2, "total_tokens": 838541408}
+{"current_steps": 19180, "total_steps": 29704, "loss": 0.30868139266967776, "lr": 2.790849196117219e-06, "epoch": 1.2914085645030973, "percentage": 64.57, "elapsed_time": "1 day, 17:13:45", "remaining_time": "22:37:20", "throughput": 5651.18, "total_tokens": 838777616}
+{"current_steps": 19185, "total_steps": 29704, "loss": 0.30340185165405276, "lr": 2.7884775024081856e-06, "epoch": 1.2917452194990573, "percentage": 64.59, "elapsed_time": "1 day, 17:14:21", "remaining_time": "22:36:40", "throughput": 5651.24, "total_tokens": 838994528}
+{"current_steps": 19190, "total_steps": 29704, "loss": 0.292337703704834, "lr": 2.786106427144367e-06, "epoch": 1.2920818744950175, "percentage": 64.6, "elapsed_time": "1 day, 17:14:57", "remaining_time": "22:36:00", "throughput": 5651.37, "total_tokens": 839211816}
+{"current_steps": 19195, "total_steps": 29704, "loss": 0.27770419120788575, "lr": 2.783735970988828e-06, "epoch": 1.2924185294909776, "percentage": 64.62, "elapsed_time": "1 day, 17:15:36", "remaining_time": "22:35:21", "throughput": 5651.31, "total_tokens": 839425992}
+{"current_steps": 19200, "total_steps": 29704, "loss": 0.31344926357269287, "lr": 2.7813661346044566e-06, "epoch": 1.2927551844869378, "percentage": 64.64, "elapsed_time": "1 day, 17:16:10", "remaining_time": "22:34:40", "throughput": 5651.46, "total_tokens": 839639736}
+{"current_steps": 19205, "total_steps": 29704, "loss": 0.30696806907653806, "lr": 2.7789969186539713e-06, "epoch": 1.2930918394828979, "percentage": 64.65, "elapsed_time": "1 day, 17:16:51", "remaining_time": "22:34:02", "throughput": 5651.28, "total_tokens": 839843624}
+{"current_steps": 19210, "total_steps": 29704, "loss": 0.32203989028930663, "lr": 2.776628323799915e-06, "epoch": 1.293428494478858, "percentage": 64.67, "elapsed_time": "1 day, 17:17:24", "remaining_time": "22:33:21", "throughput": 5651.48, "total_tokens": 840061080}
+{"current_steps": 19215, "total_steps": 29704, "loss": 0.3181305408477783, "lr": 2.774260350704657e-06, "epoch": 1.2937651494748181, "percentage": 64.69, "elapsed_time": "1 day, 17:17:59", "remaining_time": "22:32:40", "throughput": 5651.6, "total_tokens": 840276920}
+{"current_steps": 19220, "total_steps": 29704, "loss": 0.3186459541320801, "lr": 2.771893000030398e-06, "epoch": 1.2941018044707784, "percentage": 64.71, "elapsed_time": "1 day, 17:18:36", "remaining_time": "22:32:01", "throughput": 5651.61, "total_tokens": 840489952}
+{"current_steps": 19225, "total_steps": 29704, "loss": 0.29137458801269533, "lr": 2.7695262724391526e-06, "epoch": 1.2944384594667384, "percentage": 64.72, "elapsed_time": "1 day, 17:19:19", "remaining_time": "22:31:24", "throughput": 5651.42, "total_tokens": 840702144}
+{"current_steps": 19230, "total_steps": 29704, "loss": 0.33965349197387695, "lr": 2.767160168592777e-06, "epoch": 1.2947751144626987, "percentage": 64.74, "elapsed_time": "1 day, 17:20:06", "remaining_time": "22:30:50", "throughput": 5651.18, "total_tokens": 840931688}
+{"current_steps": 19235, "total_steps": 29704, "loss": 0.3225222110748291, "lr": 2.7647946891529355e-06, "epoch": 1.2951117694586587, "percentage": 64.76, "elapsed_time": "1 day, 17:20:42", "remaining_time": "22:30:10", "throughput": 5651.13, "total_tokens": 841129968}
+{"current_steps": 19240, "total_steps": 29704, "loss": 0.3056331157684326, "lr": 2.7624298347811342e-06, "epoch": 1.295448424454619, "percentage": 64.77, "elapsed_time": "1 day, 17:21:24", "remaining_time": "22:29:33", "throughput": 5650.91, "total_tokens": 841331904}
+{"current_steps": 19245, "total_steps": 29704, "loss": 0.31122641563415526, "lr": 2.7600656061386945e-06, "epoch": 1.295785079450579, "percentage": 64.79, "elapsed_time": "1 day, 17:21:59", "remaining_time": "22:28:52", "throughput": 5651.13, "total_tokens": 841562064}
+{"current_steps": 19250, "total_steps": 29704, "loss": 0.2915956974029541, "lr": 2.757702003886765e-06, "epoch": 1.2961217344465392, "percentage": 64.81, "elapsed_time": "1 day, 17:22:33", "remaining_time": "22:28:11", "throughput": 5651.3, "total_tokens": 841780328}
+{"current_steps": 19255, "total_steps": 29704, "loss": 0.29277448654174804, "lr": 2.7553390286863205e-06, "epoch": 1.2964583894424992, "percentage": 64.82, "elapsed_time": "1 day, 17:23:10", "remaining_time": "22:27:31", "throughput": 5651.25, "total_tokens": 841982320}
+{"current_steps": 19260, "total_steps": 29704, "loss": 0.3195603609085083, "lr": 2.752976681198159e-06, "epoch": 1.2967950444384595, "percentage": 64.84, "elapsed_time": "1 day, 17:23:48", "remaining_time": "22:26:52", "throughput": 5651.35, "total_tokens": 842210920}
+{"current_steps": 19265, "total_steps": 29704, "loss": 0.28080661296844484, "lr": 2.750614962082901e-06, "epoch": 1.2971316994344195, "percentage": 64.86, "elapsed_time": "1 day, 17:24:26", "remaining_time": "22:26:13", "throughput": 5651.4, "total_tokens": 842433856}
+{"current_steps": 19270, "total_steps": 29704, "loss": 0.3025611162185669, "lr": 2.748253872001e-06, "epoch": 1.2974683544303798, "percentage": 64.87, "elapsed_time": "1 day, 17:25:03", "remaining_time": "22:25:34", "throughput": 5651.45, "total_tokens": 842651984}
+{"current_steps": 19275, "total_steps": 29704, "loss": 0.30447611808776853, "lr": 2.74589341161272e-06, "epoch": 1.2978050094263398, "percentage": 64.89, "elapsed_time": "1 day, 17:25:43", "remaining_time": "22:24:56", "throughput": 5651.27, "total_tokens": 842849968}
+{"current_steps": 19280, "total_steps": 29704, "loss": 0.31205015182495116, "lr": 2.743533581578163e-06, "epoch": 1.2981416644223, "percentage": 64.91, "elapsed_time": "1 day, 17:26:20", "remaining_time": "22:24:16", "throughput": 5651.28, "total_tokens": 843059320}
+{"current_steps": 19285, "total_steps": 29704, "loss": 0.3455479145050049, "lr": 2.7411743825572407e-06, "epoch": 1.29847831941826, "percentage": 64.92, "elapsed_time": "1 day, 17:26:55", "remaining_time": "22:23:35", "throughput": 5651.51, "total_tokens": 843291856}
+{"current_steps": 19290, "total_steps": 29704, "loss": 0.30530009269714353, "lr": 2.7388158152097015e-06, "epoch": 1.2988149744142203, "percentage": 64.94, "elapsed_time": "1 day, 17:27:25", "remaining_time": "22:22:52", "throughput": 5651.89, "total_tokens": 843521128}
+{"current_steps": 19295, "total_steps": 29704, "loss": 0.3188295841217041, "lr": 2.736457880195109e-06, "epoch": 1.2991516294101804, "percentage": 64.96, "elapsed_time": "1 day, 17:28:03", "remaining_time": "22:22:13", "throughput": 5651.89, "total_tokens": 843733872}
+{"current_steps": 19300, "total_steps": 29704, "loss": 0.3281911849975586, "lr": 2.7341005781728523e-06, "epoch": 1.2994882844061406, "percentage": 64.97, "elapsed_time": "1 day, 17:28:49", "remaining_time": "22:21:38", "throughput": 5651.7, "total_tokens": 843964944}
+{"current_steps": 19305, "total_steps": 29704, "loss": 0.3072381496429443, "lr": 2.7317439098021444e-06, "epoch": 1.2998249394021006, "percentage": 64.99, "elapsed_time": "1 day, 17:29:33", "remaining_time": "22:21:02", "throughput": 5651.5, "total_tokens": 844183152}
+{"current_steps": 19310, "total_steps": 29704, "loss": 0.32221331596374514, "lr": 2.7293878757420185e-06, "epoch": 1.300161594398061, "percentage": 65.01, "elapsed_time": "1 day, 17:30:15", "remaining_time": "22:20:25", "throughput": 5651.35, "total_tokens": 844400152}
+{"current_steps": 19315, "total_steps": 29704, "loss": 0.3104707717895508, "lr": 2.7270324766513306e-06, "epoch": 1.3004982493940211, "percentage": 65.02, "elapsed_time": "1 day, 17:30:54", "remaining_time": "22:19:47", "throughput": 5651.43, "total_tokens": 844631680}
+{"current_steps": 19320, "total_steps": 29704, "loss": 0.3056204795837402, "lr": 2.724677713188767e-06, "epoch": 1.3008349043899812, "percentage": 65.04, "elapsed_time": "1 day, 17:31:36", "remaining_time": "22:19:10", "throughput": 5651.22, "total_tokens": 844836448}
+{"current_steps": 19325, "total_steps": 29704, "loss": 0.313628101348877, "lr": 2.722323586012821e-06, "epoch": 1.3011715593859412, "percentage": 65.06, "elapsed_time": "1 day, 17:32:13", "remaining_time": "22:18:31", "throughput": 5651.26, "total_tokens": 845053136}
+{"current_steps": 19330, "total_steps": 29704, "loss": 0.3013652801513672, "lr": 2.7199700957818252e-06, "epoch": 1.3015082143819015, "percentage": 65.08, "elapsed_time": "1 day, 17:32:55", "remaining_time": "22:17:54", "throughput": 5651.15, "total_tokens": 845276504}
+{"current_steps": 19335, "total_steps": 29704, "loss": 0.3019449710845947, "lr": 2.717617243153918e-06, "epoch": 1.3018448693778617, "percentage": 65.09, "elapsed_time": "1 day, 17:33:42", "remaining_time": "22:17:19", "throughput": 5650.81, "total_tokens": 845488208}
+{"current_steps": 19340, "total_steps": 29704, "loss": 0.31660406589508056, "lr": 2.715265028787073e-06, "epoch": 1.3021815243738217, "percentage": 65.11, "elapsed_time": "1 day, 17:34:20", "remaining_time": "22:16:40", "throughput": 5650.74, "total_tokens": 845695240}
+{"current_steps": 19345, "total_steps": 29704, "loss": 0.33323040008544924, "lr": 2.7129134533390765e-06, "epoch": 1.3025181793697818, "percentage": 65.13, "elapsed_time": "1 day, 17:34:57", "remaining_time": "22:16:01", "throughput": 5650.89, "total_tokens": 845925768}
+{"current_steps": 19350, "total_steps": 29704, "loss": 0.2952426433563232, "lr": 2.7105625174675404e-06, "epoch": 1.302854834365742, "percentage": 65.14, "elapsed_time": "1 day, 17:35:36", "remaining_time": "22:15:22", "throughput": 5650.96, "total_tokens": 846155368}
+{"current_steps": 19355, "total_steps": 29704, "loss": 0.2969479560852051, "lr": 2.7082122218298957e-06, "epoch": 1.3031914893617023, "percentage": 65.16, "elapsed_time": "1 day, 17:36:13", "remaining_time": "22:14:43", "throughput": 5650.97, "total_tokens": 846367080}
+{"current_steps": 19360, "total_steps": 29704, "loss": 0.3046135902404785, "lr": 2.7058625670833947e-06, "epoch": 1.3035281443576623, "percentage": 65.18, "elapsed_time": "1 day, 17:36:50", "remaining_time": "22:14:03", "throughput": 5651.05, "total_tokens": 846584096}
+{"current_steps": 19365, "total_steps": 29704, "loss": 0.306339168548584, "lr": 2.70351355388511e-06, "epoch": 1.3038647993536223, "percentage": 65.19, "elapsed_time": "1 day, 17:37:24", "remaining_time": "22:13:22", "throughput": 5651.19, "total_tokens": 846800416}
+{"current_steps": 19370, "total_steps": 29704, "loss": 0.30881786346435547, "lr": 2.7011651828919407e-06, "epoch": 1.3042014543495826, "percentage": 65.21, "elapsed_time": "1 day, 17:38:06", "remaining_time": "22:12:45", "throughput": 5651.11, "total_tokens": 847024880}
+{"current_steps": 19375, "total_steps": 29704, "loss": 0.305006217956543, "lr": 2.698817454760593e-06, "epoch": 1.3045381093455428, "percentage": 65.23, "elapsed_time": "1 day, 17:38:39", "remaining_time": "22:12:03", "throughput": 5651.33, "total_tokens": 847244184}
+{"current_steps": 19380, "total_steps": 29704, "loss": 0.3149790048599243, "lr": 2.696470370147611e-06, "epoch": 1.3048747643415028, "percentage": 65.24, "elapsed_time": "1 day, 17:39:15", "remaining_time": "22:11:23", "throughput": 5651.43, "total_tokens": 847461096}
+{"current_steps": 19385, "total_steps": 29704, "loss": 0.3325794219970703, "lr": 2.69412392970934e-06, "epoch": 1.3052114193374629, "percentage": 65.26, "elapsed_time": "1 day, 17:39:53", "remaining_time": "22:10:44", "throughput": 5651.46, "total_tokens": 847680648}
+{"current_steps": 19390, "total_steps": 29704, "loss": 0.2991641044616699, "lr": 2.6917781341019615e-06, "epoch": 1.3055480743334231, "percentage": 65.28, "elapsed_time": "1 day, 17:40:31", "remaining_time": "22:10:05", "throughput": 5651.49, "total_tokens": 847901040}
+{"current_steps": 19395, "total_steps": 29704, "loss": 0.2986927032470703, "lr": 2.689432983981467e-06, "epoch": 1.3058847293293834, "percentage": 65.29, "elapsed_time": "1 day, 17:41:20", "remaining_time": "22:09:31", "throughput": 5651.17, "total_tokens": 848128768}
+{"current_steps": 19400, "total_steps": 29704, "loss": 0.3080791234970093, "lr": 2.6870884800036713e-06, "epoch": 1.3062213843253434, "percentage": 65.31, "elapsed_time": "1 day, 17:41:53", "remaining_time": "22:08:50", "throughput": 5651.32, "total_tokens": 848340072}
+{"current_steps": 19405, "total_steps": 29704, "loss": 0.32254228591918943, "lr": 2.684744622824207e-06, "epoch": 1.3065580393213034, "percentage": 65.33, "elapsed_time": "1 day, 17:42:32", "remaining_time": "22:08:12", "throughput": 5651.23, "total_tokens": 848548544}
+{"current_steps": 19410, "total_steps": 29704, "loss": 0.31671667098999023, "lr": 2.6824014130985255e-06, "epoch": 1.3068946943172637, "percentage": 65.34, "elapsed_time": "1 day, 17:43:10", "remaining_time": "22:07:32", "throughput": 5651.28, "total_tokens": 848765608}
+{"current_steps": 19415, "total_steps": 29704, "loss": 0.2992823600769043, "lr": 2.680058851481896e-06, "epoch": 1.307231349313224, "percentage": 65.36, "elapsed_time": "1 day, 17:43:45", "remaining_time": "22:06:51", "throughput": 5651.41, "total_tokens": 848984336}
+{"current_steps": 19420, "total_steps": 29704, "loss": 0.3124098777770996, "lr": 2.6777169386294154e-06, "epoch": 1.307568004309184, "percentage": 65.38, "elapsed_time": "1 day, 17:44:23", "remaining_time": "22:06:13", "throughput": 5651.42, "total_tokens": 849202320}
+{"current_steps": 19425, "total_steps": 29704, "loss": 0.30565834045410156, "lr": 2.6753756751959846e-06, "epoch": 1.307904659305144, "percentage": 65.4, "elapsed_time": "1 day, 17:45:04", "remaining_time": "22:05:35", "throughput": 5651.39, "total_tokens": 849432016}
+{"current_steps": 19430, "total_steps": 29704, "loss": 0.3007196426391602, "lr": 2.6730350618363367e-06, "epoch": 1.3082413143011042, "percentage": 65.41, "elapsed_time": "1 day, 17:45:39", "remaining_time": "22:04:54", "throughput": 5651.56, "total_tokens": 849652096}
+{"current_steps": 19435, "total_steps": 29704, "loss": 0.2976806640625, "lr": 2.6706950992050097e-06, "epoch": 1.3085779692970645, "percentage": 65.43, "elapsed_time": "1 day, 17:46:17", "remaining_time": "22:04:16", "throughput": 5651.59, "total_tokens": 849873624}
+{"current_steps": 19440, "total_steps": 29704, "loss": 0.3227806091308594, "lr": 2.6683557879563726e-06, "epoch": 1.3089146242930245, "percentage": 65.45, "elapsed_time": "1 day, 17:46:51", "remaining_time": "22:03:34", "throughput": 5651.79, "total_tokens": 850094824}
+{"current_steps": 19445, "total_steps": 29704, "loss": 0.33275818824768066, "lr": 2.6660171287446045e-06, "epoch": 1.3092512792889845, "percentage": 65.46, "elapsed_time": "1 day, 17:47:28", "remaining_time": "22:02:55", "throughput": 5651.9, "total_tokens": 850318184}
+{"current_steps": 19450, "total_steps": 29704, "loss": 0.3194304943084717, "lr": 2.6636791222237034e-06, "epoch": 1.3095879342849448, "percentage": 65.48, "elapsed_time": "1 day, 17:48:05", "remaining_time": "22:02:15", "throughput": 5651.97, "total_tokens": 850537512}
+{"current_steps": 19455, "total_steps": 29704, "loss": 0.3084381580352783, "lr": 2.6613417690474864e-06, "epoch": 1.309924589280905, "percentage": 65.5, "elapsed_time": "1 day, 17:48:36", "remaining_time": "22:01:32", "throughput": 5652.31, "total_tokens": 850762520}
+{"current_steps": 19460, "total_steps": 29704, "loss": 0.30487637519836425, "lr": 2.6590050698695856e-06, "epoch": 1.310261244276865, "percentage": 65.51, "elapsed_time": "1 day, 17:49:12", "remaining_time": "22:00:53", "throughput": 5652.37, "total_tokens": 850980888}
+{"current_steps": 19465, "total_steps": 29704, "loss": 0.31617035865783694, "lr": 2.6566690253434508e-06, "epoch": 1.310597899272825, "percentage": 65.53, "elapsed_time": "1 day, 17:49:51", "remaining_time": "22:00:14", "throughput": 5652.48, "total_tokens": 851212384}
+{"current_steps": 19470, "total_steps": 29704, "loss": 0.29657797813415526, "lr": 2.654333636122354e-06, "epoch": 1.3109345542687854, "percentage": 65.55, "elapsed_time": "1 day, 17:50:31", "remaining_time": "21:59:36", "throughput": 5652.46, "total_tokens": 851438520}
+{"current_steps": 19475, "total_steps": 29704, "loss": 0.3158660173416138, "lr": 2.6519989028593728e-06, "epoch": 1.3112712092647456, "percentage": 65.56, "elapsed_time": "1 day, 17:51:11", "remaining_time": "21:58:58", "throughput": 5652.48, "total_tokens": 851666592}
+{"current_steps": 19480, "total_steps": 29704, "loss": 0.31950523853302004, "lr": 2.649664826207414e-06, "epoch": 1.3116078642607056, "percentage": 65.58, "elapsed_time": "1 day, 17:51:50", "remaining_time": "21:58:19", "throughput": 5652.46, "total_tokens": 851881728}
+{"current_steps": 19485, "total_steps": 29704, "loss": 0.3350534915924072, "lr": 2.647331406819189e-06, "epoch": 1.3119445192566657, "percentage": 65.6, "elapsed_time": "1 day, 17:52:27", "remaining_time": "21:57:40", "throughput": 5652.44, "total_tokens": 852092288}
+{"current_steps": 19490, "total_steps": 29704, "loss": 0.29921770095825195, "lr": 2.644998645347236e-06, "epoch": 1.312281174252626, "percentage": 65.61, "elapsed_time": "1 day, 17:53:11", "remaining_time": "21:57:04", "throughput": 5652.2, "total_tokens": 852306504}
+{"current_steps": 19495, "total_steps": 29704, "loss": 0.316993260383606, "lr": 2.6426665424439025e-06, "epoch": 1.3126178292485862, "percentage": 65.63, "elapsed_time": "1 day, 17:53:56", "remaining_time": "21:56:28", "throughput": 5652.03, "total_tokens": 852532944}
+{"current_steps": 19500, "total_steps": 29704, "loss": 0.31356039047241213, "lr": 2.6403350987613526e-06, "epoch": 1.3129544842445462, "percentage": 65.65, "elapsed_time": "1 day, 17:54:39", "remaining_time": "21:55:52", "throughput": 5651.91, "total_tokens": 852756160}
+{"current_steps": 19505, "total_steps": 29704, "loss": 0.3076605796813965, "lr": 2.6380043149515676e-06, "epoch": 1.3132911392405062, "percentage": 65.66, "elapsed_time": "1 day, 17:55:13", "remaining_time": "21:55:11", "throughput": 5652.06, "total_tokens": 852971624}
+{"current_steps": 19510, "total_steps": 29704, "loss": 0.3078343868255615, "lr": 2.6356741916663453e-06, "epoch": 1.3136277942364665, "percentage": 65.68, "elapsed_time": "1 day, 17:55:52", "remaining_time": "21:54:33", "throughput": 5652.06, "total_tokens": 853194624}
+{"current_steps": 19515, "total_steps": 29704, "loss": 0.310660195350647, "lr": 2.6333447295572933e-06, "epoch": 1.3139644492324267, "percentage": 65.7, "elapsed_time": "1 day, 17:56:25", "remaining_time": "21:53:51", "throughput": 5652.28, "total_tokens": 853414920}
+{"current_steps": 19520, "total_steps": 29704, "loss": 0.3059165239334106, "lr": 2.631015929275845e-06, "epoch": 1.3143011042283868, "percentage": 65.72, "elapsed_time": "1 day, 17:57:06", "remaining_time": "21:53:13", "throughput": 5652.13, "total_tokens": 853622712}
+{"current_steps": 19525, "total_steps": 29704, "loss": 0.29071602821350095, "lr": 2.628687791473234e-06, "epoch": 1.3146377592243468, "percentage": 65.73, "elapsed_time": "1 day, 17:57:37", "remaining_time": "21:52:30", "throughput": 5652.52, "total_tokens": 853853296}
+{"current_steps": 19530, "total_steps": 29704, "loss": 0.2854055643081665, "lr": 2.6263603168005224e-06, "epoch": 1.314974414220307, "percentage": 65.75, "elapsed_time": "1 day, 17:58:22", "remaining_time": "21:51:55", "throughput": 5652.27, "total_tokens": 854070296}
+{"current_steps": 19535, "total_steps": 29704, "loss": 0.30446481704711914, "lr": 2.6240335059085797e-06, "epoch": 1.3153110692162673, "percentage": 65.77, "elapsed_time": "1 day, 17:59:03", "remaining_time": "21:51:18", "throughput": 5652.24, "total_tokens": 854298216}
+{"current_steps": 19540, "total_steps": 29704, "loss": 0.3120380163192749, "lr": 2.62170735944809e-06, "epoch": 1.3156477242122273, "percentage": 65.78, "elapsed_time": "1 day, 17:59:40", "remaining_time": "21:50:38", "throughput": 5652.29, "total_tokens": 854517360}
+{"current_steps": 19545, "total_steps": 29704, "loss": 0.298751163482666, "lr": 2.6193818780695537e-06, "epoch": 1.3159843792081873, "percentage": 65.8, "elapsed_time": "1 day, 18:00:14", "remaining_time": "21:49:57", "throughput": 5652.55, "total_tokens": 854745560}
+{"current_steps": 19550, "total_steps": 29704, "loss": 0.3143803119659424, "lr": 2.6170570624232837e-06, "epoch": 1.3163210342041476, "percentage": 65.82, "elapsed_time": "1 day, 18:00:57", "remaining_time": "21:49:21", "throughput": 5652.22, "total_tokens": 854941592}
+{"current_steps": 19555, "total_steps": 29704, "loss": 0.2849498510360718, "lr": 2.6147329131594078e-06, "epoch": 1.3166576892001078, "percentage": 65.83, "elapsed_time": "1 day, 18:01:38", "remaining_time": "21:48:43", "throughput": 5652.21, "total_tokens": 855170576}
+{"current_steps": 19560, "total_steps": 29704, "loss": 0.2712059020996094, "lr": 2.612409430927867e-06, "epoch": 1.3169943441960679, "percentage": 65.85, "elapsed_time": "1 day, 18:02:15", "remaining_time": "21:48:04", "throughput": 5652.26, "total_tokens": 855390328}
+{"current_steps": 19565, "total_steps": 29704, "loss": 0.29099502563476565, "lr": 2.610086616378413e-06, "epoch": 1.317330999192028, "percentage": 65.87, "elapsed_time": "1 day, 18:02:49", "remaining_time": "21:47:22", "throughput": 5652.4, "total_tokens": 855599680}
+{"current_steps": 19570, "total_steps": 29704, "loss": 0.3147557497024536, "lr": 2.60776447016062e-06, "epoch": 1.3176676541879881, "percentage": 65.88, "elapsed_time": "1 day, 18:03:26", "remaining_time": "21:46:43", "throughput": 5652.46, "total_tokens": 855821216}
+{"current_steps": 19575, "total_steps": 29704, "loss": 0.31384851932525637, "lr": 2.6054429929238607e-06, "epoch": 1.3180043091839484, "percentage": 65.9, "elapsed_time": "1 day, 18:04:00", "remaining_time": "21:46:02", "throughput": 5652.64, "total_tokens": 856037064}
+{"current_steps": 19580, "total_steps": 29704, "loss": 0.29209444522857664, "lr": 2.603122185317335e-06, "epoch": 1.3183409641799084, "percentage": 65.92, "elapsed_time": "1 day, 18:04:35", "remaining_time": "21:45:21", "throughput": 5652.66, "total_tokens": 856240856}
+{"current_steps": 19585, "total_steps": 29704, "loss": 0.2708132743835449, "lr": 2.6008020479900465e-06, "epoch": 1.3186776191758685, "percentage": 65.93, "elapsed_time": "1 day, 18:05:15", "remaining_time": "21:44:43", "throughput": 5652.59, "total_tokens": 856453280}
+{"current_steps": 19590, "total_steps": 29704, "loss": 0.30856051445007326, "lr": 2.598482581590815e-06, "epoch": 1.3190142741718287, "percentage": 65.95, "elapsed_time": "1 day, 18:05:50", "remaining_time": "21:44:03", "throughput": 5652.8, "total_tokens": 856685160}
+{"current_steps": 19595, "total_steps": 29704, "loss": 0.280717658996582, "lr": 2.5961637867682715e-06, "epoch": 1.319350929167789, "percentage": 65.97, "elapsed_time": "1 day, 18:06:31", "remaining_time": "21:43:25", "throughput": 5652.81, "total_tokens": 856916440}
+{"current_steps": 19600, "total_steps": 29704, "loss": 0.28789021968841555, "lr": 2.5938456641708598e-06, "epoch": 1.319687584163749, "percentage": 65.98, "elapsed_time": "1 day, 18:07:00", "remaining_time": "21:42:42", "throughput": 5653.15, "total_tokens": 857134672}
+{"current_steps": 19605, "total_steps": 29704, "loss": 0.3085463523864746, "lr": 2.591528214446836e-06, "epoch": 1.320024239159709, "percentage": 66.0, "elapsed_time": "1 day, 18:07:36", "remaining_time": "21:42:01", "throughput": 5653.34, "total_tokens": 857366576}
+{"current_steps": 19610, "total_steps": 29704, "loss": 0.3224550724029541, "lr": 2.5892114382442664e-06, "epoch": 1.3203608941556693, "percentage": 66.02, "elapsed_time": "1 day, 18:08:13", "remaining_time": "21:41:22", "throughput": 5653.39, "total_tokens": 857583784}
+{"current_steps": 19615, "total_steps": 29704, "loss": 0.29941411018371583, "lr": 2.5868953362110287e-06, "epoch": 1.3206975491516295, "percentage": 66.03, "elapsed_time": "1 day, 18:08:49", "remaining_time": "21:40:42", "throughput": 5653.51, "total_tokens": 857806240}
+{"current_steps": 19620, "total_steps": 29704, "loss": 0.29595515727996824, "lr": 2.5845799089948196e-06, "epoch": 1.3210342041475895, "percentage": 66.05, "elapsed_time": "1 day, 18:09:27", "remaining_time": "21:40:03", "throughput": 5653.48, "total_tokens": 858015600}
+{"current_steps": 19625, "total_steps": 29704, "loss": 0.30276174545288087, "lr": 2.5822651572431323e-06, "epoch": 1.3213708591435496, "percentage": 66.07, "elapsed_time": "1 day, 18:10:07", "remaining_time": "21:39:25", "throughput": 5653.47, "total_tokens": 858241536}
+{"current_steps": 19630, "total_steps": 29704, "loss": 0.3044028043746948, "lr": 2.5799510816032867e-06, "epoch": 1.3217075141395098, "percentage": 66.09, "elapsed_time": "1 day, 18:10:49", "remaining_time": "21:38:48", "throughput": 5653.18, "total_tokens": 858431176}
+{"current_steps": 19635, "total_steps": 29704, "loss": 0.31405105590820315, "lr": 2.577637682722404e-06, "epoch": 1.32204416913547, "percentage": 66.1, "elapsed_time": "1 day, 18:11:33", "remaining_time": "21:38:12", "throughput": 5652.98, "total_tokens": 858649160}
+{"current_steps": 19640, "total_steps": 29704, "loss": 0.3185321807861328, "lr": 2.5753249612474195e-06, "epoch": 1.32238082413143, "percentage": 66.12, "elapsed_time": "1 day, 18:12:15", "remaining_time": "21:37:35", "throughput": 5652.83, "total_tokens": 858864688}
+{"current_steps": 19645, "total_steps": 29704, "loss": 0.3150054693222046, "lr": 2.5730129178250774e-06, "epoch": 1.3227174791273901, "percentage": 66.14, "elapsed_time": "1 day, 18:12:53", "remaining_time": "21:36:56", "throughput": 5652.91, "total_tokens": 859090984}
+{"current_steps": 19650, "total_steps": 29704, "loss": 0.3087807893753052, "lr": 2.570701553101933e-06, "epoch": 1.3230541341233504, "percentage": 66.15, "elapsed_time": "1 day, 18:13:36", "remaining_time": "21:36:20", "throughput": 5652.69, "total_tokens": 859303816}
+{"current_steps": 19655, "total_steps": 29704, "loss": 0.3278617858886719, "lr": 2.568390867724354e-06, "epoch": 1.3233907891193106, "percentage": 66.17, "elapsed_time": "1 day, 18:14:19", "remaining_time": "21:35:43", "throughput": 5652.57, "total_tokens": 859526840}
+{"current_steps": 19660, "total_steps": 29704, "loss": 0.3053846836090088, "lr": 2.5660808623385143e-06, "epoch": 1.3237274441152707, "percentage": 66.19, "elapsed_time": "1 day, 18:14:57", "remaining_time": "21:35:04", "throughput": 5652.53, "total_tokens": 859736800}
+{"current_steps": 19665, "total_steps": 29704, "loss": 0.2824453830718994, "lr": 2.5637715375903984e-06, "epoch": 1.3240640991112307, "percentage": 66.2, "elapsed_time": "1 day, 18:15:39", "remaining_time": "21:34:27", "throughput": 5652.4, "total_tokens": 859955360}
+{"current_steps": 19670, "total_steps": 29704, "loss": 0.3014561176300049, "lr": 2.5614628941258075e-06, "epoch": 1.324400754107191, "percentage": 66.22, "elapsed_time": "1 day, 18:16:19", "remaining_time": "21:33:49", "throughput": 5652.29, "total_tokens": 860163408}
+{"current_steps": 19675, "total_steps": 29704, "loss": 0.3199615478515625, "lr": 2.5591549325903388e-06, "epoch": 1.3247374091031512, "percentage": 66.24, "elapsed_time": "1 day, 18:17:00", "remaining_time": "21:33:12", "throughput": 5652.33, "total_tokens": 860403016}
+{"current_steps": 19680, "total_steps": 29704, "loss": 0.3173452615737915, "lr": 2.5568476536294105e-06, "epoch": 1.3250740640991112, "percentage": 66.25, "elapsed_time": "1 day, 18:17:33", "remaining_time": "21:32:30", "throughput": 5652.59, "total_tokens": 860626712}
+{"current_steps": 19685, "total_steps": 29704, "loss": 0.2950143814086914, "lr": 2.5545410578882466e-06, "epoch": 1.3254107190950712, "percentage": 66.27, "elapsed_time": "1 day, 18:18:09", "remaining_time": "21:31:50", "throughput": 5652.78, "total_tokens": 860857936}
+{"current_steps": 19690, "total_steps": 29704, "loss": 0.2903186321258545, "lr": 2.5522351460118776e-06, "epoch": 1.3257473740910315, "percentage": 66.29, "elapsed_time": "1 day, 18:18:39", "remaining_time": "21:31:07", "throughput": 5653.01, "total_tokens": 861063800}
+{"current_steps": 19695, "total_steps": 29704, "loss": 0.31392974853515626, "lr": 2.549929918645144e-06, "epoch": 1.3260840290869917, "percentage": 66.3, "elapsed_time": "1 day, 18:19:15", "remaining_time": "21:30:27", "throughput": 5653.21, "total_tokens": 861297432}
+{"current_steps": 19700, "total_steps": 29704, "loss": 0.2875689506530762, "lr": 2.5476253764326963e-06, "epoch": 1.3264206840829518, "percentage": 66.32, "elapsed_time": "1 day, 18:19:51", "remaining_time": "21:29:47", "throughput": 5653.22, "total_tokens": 861503656}
+{"current_steps": 19705, "total_steps": 29704, "loss": 0.2943393707275391, "lr": 2.5453215200189927e-06, "epoch": 1.3267573390789118, "percentage": 66.34, "elapsed_time": "1 day, 18:20:29", "remaining_time": "21:29:08", "throughput": 5653.2, "total_tokens": 861717016}
+{"current_steps": 19710, "total_steps": 29704, "loss": 0.29305644035339357, "lr": 2.543018350048298e-06, "epoch": 1.327093994074872, "percentage": 66.35, "elapsed_time": "1 day, 18:21:08", "remaining_time": "21:28:29", "throughput": 5653.23, "total_tokens": 861937640}
+{"current_steps": 19715, "total_steps": 29704, "loss": 0.3149756908416748, "lr": 2.540715867164685e-06, "epoch": 1.3274306490708323, "percentage": 66.37, "elapsed_time": "1 day, 18:21:46", "remaining_time": "21:27:50", "throughput": 5653.25, "total_tokens": 862156848}
+{"current_steps": 19720, "total_steps": 29704, "loss": 0.30890326499938964, "lr": 2.538414072012042e-06, "epoch": 1.3277673040667923, "percentage": 66.39, "elapsed_time": "1 day, 18:22:27", "remaining_time": "21:27:13", "throughput": 5653.06, "total_tokens": 862362672}
+{"current_steps": 19725, "total_steps": 29704, "loss": 0.3164846420288086, "lr": 2.53611296523405e-06, "epoch": 1.3281039590627524, "percentage": 66.41, "elapsed_time": "1 day, 18:23:00", "remaining_time": "21:26:31", "throughput": 5653.24, "total_tokens": 862573904}
+{"current_steps": 19730, "total_steps": 29704, "loss": 0.2981910228729248, "lr": 2.5338125474742127e-06, "epoch": 1.3284406140587126, "percentage": 66.42, "elapsed_time": "1 day, 18:23:38", "remaining_time": "21:25:52", "throughput": 5653.24, "total_tokens": 862787064}
+{"current_steps": 19735, "total_steps": 29704, "loss": 0.2908968448638916, "lr": 2.531512819375832e-06, "epoch": 1.3287772690546729, "percentage": 66.44, "elapsed_time": "1 day, 18:24:22", "remaining_time": "21:25:16", "throughput": 5653.05, "total_tokens": 863007040}
+{"current_steps": 19740, "total_steps": 29704, "loss": 0.2935811996459961, "lr": 2.52921378158202e-06, "epoch": 1.3291139240506329, "percentage": 66.46, "elapsed_time": "1 day, 18:24:57", "remaining_time": "21:24:35", "throughput": 5653.13, "total_tokens": 863219112}
+{"current_steps": 19745, "total_steps": 29704, "loss": 0.2813416957855225, "lr": 2.5269154347356957e-06, "epoch": 1.329450579046593, "percentage": 66.47, "elapsed_time": "1 day, 18:25:36", "remaining_time": "21:23:57", "throughput": 5653.14, "total_tokens": 863441536}
+{"current_steps": 19750, "total_steps": 29704, "loss": 0.30994884967803954, "lr": 2.5246177794795834e-06, "epoch": 1.3297872340425532, "percentage": 66.49, "elapsed_time": "1 day, 18:26:23", "remaining_time": "21:23:23", "throughput": 5652.71, "total_tokens": 863641736}
+{"current_steps": 19755, "total_steps": 29704, "loss": 0.30829753875732424, "lr": 2.5223208164562163e-06, "epoch": 1.3301238890385134, "percentage": 66.51, "elapsed_time": "1 day, 18:27:03", "remaining_time": "21:22:44", "throughput": 5652.66, "total_tokens": 863857152}
+{"current_steps": 19760, "total_steps": 29704, "loss": 0.31928205490112305, "lr": 2.5200245463079315e-06, "epoch": 1.3304605440344734, "percentage": 66.52, "elapsed_time": "1 day, 18:27:39", "remaining_time": "21:22:04", "throughput": 5652.71, "total_tokens": 864070800}
+{"current_steps": 19765, "total_steps": 29704, "loss": 0.32152862548828126, "lr": 2.5177289696768725e-06, "epoch": 1.3307971990304335, "percentage": 66.54, "elapsed_time": "1 day, 18:28:13", "remaining_time": "21:21:23", "throughput": 5652.92, "total_tokens": 864291928}
+{"current_steps": 19770, "total_steps": 29704, "loss": 0.29552688598632815, "lr": 2.515434087204996e-06, "epoch": 1.3311338540263937, "percentage": 66.56, "elapsed_time": "1 day, 18:28:53", "remaining_time": "21:20:45", "throughput": 5652.97, "total_tokens": 864526144}
+{"current_steps": 19775, "total_steps": 29704, "loss": 0.3004995107650757, "lr": 2.51313989953405e-06, "epoch": 1.331470509022354, "percentage": 66.57, "elapsed_time": "1 day, 18:29:27", "remaining_time": "21:20:04", "throughput": 5653.22, "total_tokens": 864756000}
+{"current_steps": 19780, "total_steps": 29704, "loss": 0.298743748664856, "lr": 2.510846407305603e-06, "epoch": 1.331807164018314, "percentage": 66.59, "elapsed_time": "1 day, 18:30:02", "remaining_time": "21:19:24", "throughput": 5653.33, "total_tokens": 864976528}
+{"current_steps": 19785, "total_steps": 29704, "loss": 0.3048709392547607, "lr": 2.508553611161021e-06, "epoch": 1.332143819014274, "percentage": 66.61, "elapsed_time": "1 day, 18:30:45", "remaining_time": "21:18:47", "throughput": 5653.2, "total_tokens": 865197968}
+{"current_steps": 19790, "total_steps": 29704, "loss": 0.3029000759124756, "lr": 2.5062615117414773e-06, "epoch": 1.3324804740102343, "percentage": 66.62, "elapsed_time": "1 day, 18:31:29", "remaining_time": "21:18:11", "throughput": 5653.1, "total_tokens": 865431584}
+{"current_steps": 19795, "total_steps": 29704, "loss": 0.2859670400619507, "lr": 2.50397010968795e-06, "epoch": 1.3328171290061945, "percentage": 66.64, "elapsed_time": "1 day, 18:32:13", "remaining_time": "21:17:35", "throughput": 5652.89, "total_tokens": 865646552}
+{"current_steps": 19800, "total_steps": 29704, "loss": 0.32290475368499755, "lr": 2.501679405641222e-06, "epoch": 1.3331537840021546, "percentage": 66.66, "elapsed_time": "1 day, 18:32:48", "remaining_time": "21:16:55", "throughput": 5653.04, "total_tokens": 865868968}
+{"current_steps": 19805, "total_steps": 29704, "loss": 0.30891990661621094, "lr": 2.499389400241881e-06, "epoch": 1.3334904389981148, "percentage": 66.67, "elapsed_time": "1 day, 18:33:25", "remaining_time": "21:16:15", "throughput": 5653.0, "total_tokens": 866072568}
+{"current_steps": 19810, "total_steps": 29704, "loss": 0.27096853256225584, "lr": 2.497100094130321e-06, "epoch": 1.3338270939940748, "percentage": 66.69, "elapsed_time": "1 day, 18:34:02", "remaining_time": "21:15:36", "throughput": 5653.06, "total_tokens": 866291048}
+{"current_steps": 19815, "total_steps": 29704, "loss": 0.2911532402038574, "lr": 2.494811487946737e-06, "epoch": 1.334163748990035, "percentage": 66.71, "elapsed_time": "1 day, 18:34:37", "remaining_time": "21:14:55", "throughput": 5653.26, "total_tokens": 866515272}
+{"current_steps": 19820, "total_steps": 29704, "loss": 0.3089088201522827, "lr": 2.492523582331135e-06, "epoch": 1.3345004039859951, "percentage": 66.73, "elapsed_time": "1 day, 18:35:18", "remaining_time": "21:14:18", "throughput": 5653.18, "total_tokens": 866738320}
+{"current_steps": 19825, "total_steps": 29704, "loss": 0.29610767364501955, "lr": 2.4902363779233134e-06, "epoch": 1.3348370589819554, "percentage": 66.74, "elapsed_time": "1 day, 18:35:58", "remaining_time": "21:13:40", "throughput": 5653.14, "total_tokens": 866959032}
+{"current_steps": 19830, "total_steps": 29704, "loss": 0.31332998275756835, "lr": 2.4879498753628885e-06, "epoch": 1.3351737139779154, "percentage": 66.76, "elapsed_time": "1 day, 18:36:40", "remaining_time": "21:13:03", "throughput": 5652.88, "total_tokens": 867155816}
+{"current_steps": 19835, "total_steps": 29704, "loss": 0.27421958446502687, "lr": 2.4856640752892702e-06, "epoch": 1.3355103689738757, "percentage": 66.78, "elapsed_time": "1 day, 18:37:18", "remaining_time": "21:12:23", "throughput": 5652.98, "total_tokens": 867382880}
+{"current_steps": 19840, "total_steps": 29704, "loss": 0.28227849006652833, "lr": 2.4833789783416757e-06, "epoch": 1.3358470239698357, "percentage": 66.79, "elapsed_time": "1 day, 18:37:58", "remaining_time": "21:11:45", "throughput": 5652.96, "total_tokens": 867604504}
+{"current_steps": 19845, "total_steps": 29704, "loss": 0.3176878452301025, "lr": 2.4810945851591245e-06, "epoch": 1.336183678965796, "percentage": 66.81, "elapsed_time": "1 day, 18:38:34", "remaining_time": "21:11:06", "throughput": 5652.96, "total_tokens": 867810880}
+{"current_steps": 19850, "total_steps": 29704, "loss": 0.2944268941879272, "lr": 2.4788108963804406e-06, "epoch": 1.336520333961756, "percentage": 66.83, "elapsed_time": "1 day, 18:39:15", "remaining_time": "21:10:28", "throughput": 5652.9, "total_tokens": 868031240}
+{"current_steps": 19855, "total_steps": 29704, "loss": 0.30596926212310793, "lr": 2.4765279126442502e-06, "epoch": 1.3368569889577162, "percentage": 66.84, "elapsed_time": "1 day, 18:39:54", "remaining_time": "21:09:49", "throughput": 5652.9, "total_tokens": 868253728}
+{"current_steps": 19860, "total_steps": 29704, "loss": 0.3069468975067139, "lr": 2.4742456345889828e-06, "epoch": 1.3371936439536762, "percentage": 66.86, "elapsed_time": "1 day, 18:40:33", "remaining_time": "21:09:11", "throughput": 5652.84, "total_tokens": 868466312}
+{"current_steps": 19865, "total_steps": 29704, "loss": 0.30750603675842286, "lr": 2.471964062852867e-06, "epoch": 1.3375302989496365, "percentage": 66.88, "elapsed_time": "1 day, 18:41:05", "remaining_time": "21:08:29", "throughput": 5653.12, "total_tokens": 868689096}
+{"current_steps": 19870, "total_steps": 29704, "loss": 0.29093983173370364, "lr": 2.469683198073943e-06, "epoch": 1.3378669539455965, "percentage": 66.89, "elapsed_time": "1 day, 18:41:36", "remaining_time": "21:07:46", "throughput": 5653.47, "total_tokens": 868917088}
+{"current_steps": 19875, "total_steps": 29704, "loss": 0.2891664505004883, "lr": 2.4674030408900414e-06, "epoch": 1.3382036089415568, "percentage": 66.91, "elapsed_time": "1 day, 18:42:15", "remaining_time": "21:07:08", "throughput": 5653.26, "total_tokens": 869105376}
+{"current_steps": 19880, "total_steps": 29704, "loss": 0.3123000144958496, "lr": 2.465123591938805e-06, "epoch": 1.3385402639375168, "percentage": 66.93, "elapsed_time": "1 day, 18:42:51", "remaining_time": "21:06:28", "throughput": 5653.31, "total_tokens": 869319584}
+{"current_steps": 19885, "total_steps": 29704, "loss": 0.31596615314483645, "lr": 2.462844851857673e-06, "epoch": 1.338876918933477, "percentage": 66.94, "elapsed_time": "1 day, 18:43:35", "remaining_time": "21:05:52", "throughput": 5653.06, "total_tokens": 869526576}
+{"current_steps": 19890, "total_steps": 29704, "loss": 0.31440134048461915, "lr": 2.4605668212838878e-06, "epoch": 1.339213573929437, "percentage": 66.96, "elapsed_time": "1 day, 18:44:15", "remaining_time": "21:05:14", "throughput": 5652.98, "total_tokens": 869742096}
+{"current_steps": 19895, "total_steps": 29704, "loss": 0.34726858139038086, "lr": 2.458289500854493e-06, "epoch": 1.3395502289253973, "percentage": 66.98, "elapsed_time": "1 day, 18:44:50", "remaining_time": "21:04:34", "throughput": 5653.11, "total_tokens": 869961536}
+{"current_steps": 19900, "total_steps": 29704, "loss": 0.3165955305099487, "lr": 2.4560128912063353e-06, "epoch": 1.3398868839213574, "percentage": 66.99, "elapsed_time": "1 day, 18:45:32", "remaining_time": "21:03:56", "throughput": 5652.9, "total_tokens": 870165224}
+{"current_steps": 19905, "total_steps": 29704, "loss": 0.34606313705444336, "lr": 2.4537369929760596e-06, "epoch": 1.3402235389173176, "percentage": 67.01, "elapsed_time": "1 day, 18:46:09", "remaining_time": "21:03:17", "throughput": 5653.03, "total_tokens": 870396672}
+{"current_steps": 19910, "total_steps": 29704, "loss": 0.305745005607605, "lr": 2.4514618068001144e-06, "epoch": 1.3405601939132776, "percentage": 67.03, "elapsed_time": "1 day, 18:46:48", "remaining_time": "21:02:38", "throughput": 5653.0, "total_tokens": 870610560}
+{"current_steps": 19915, "total_steps": 29704, "loss": 0.29233665466308595, "lr": 2.4491873333147463e-06, "epoch": 1.3408968489092379, "percentage": 67.04, "elapsed_time": "1 day, 18:47:30", "remaining_time": "21:02:01", "throughput": 5652.85, "total_tokens": 870825304}
+{"current_steps": 19920, "total_steps": 29704, "loss": 0.2776491641998291, "lr": 2.4469135731560106e-06, "epoch": 1.341233503905198, "percentage": 67.06, "elapsed_time": "1 day, 18:48:08", "remaining_time": "21:01:23", "throughput": 5652.8, "total_tokens": 871033624}
+{"current_steps": 19925, "total_steps": 29704, "loss": 0.31336369514465334, "lr": 2.4446405269597494e-06, "epoch": 1.3415701589011582, "percentage": 67.08, "elapsed_time": "1 day, 18:48:58", "remaining_time": "21:00:49", "throughput": 5652.47, "total_tokens": 871261480}
+{"current_steps": 19930, "total_steps": 29704, "loss": 0.27323174476623535, "lr": 2.442368195361618e-06, "epoch": 1.3419068138971182, "percentage": 67.1, "elapsed_time": "1 day, 18:49:35", "remaining_time": "21:00:10", "throughput": 5652.55, "total_tokens": 871484784}
+{"current_steps": 19935, "total_steps": 29704, "loss": 0.2858338594436646, "lr": 2.4400965789970647e-06, "epoch": 1.3422434688930784, "percentage": 67.11, "elapsed_time": "1 day, 18:50:14", "remaining_time": "20:59:31", "throughput": 5652.49, "total_tokens": 871694432}
+{"current_steps": 19940, "total_steps": 29704, "loss": 0.3141868352890015, "lr": 2.437825678501339e-06, "epoch": 1.3425801238890385, "percentage": 67.13, "elapsed_time": "1 day, 18:50:56", "remaining_time": "20:58:54", "throughput": 5652.4, "total_tokens": 871916776}
+{"current_steps": 19945, "total_steps": 29704, "loss": 0.3040180683135986, "lr": 2.4355554945094956e-06, "epoch": 1.3429167788849987, "percentage": 67.15, "elapsed_time": "1 day, 18:51:38", "remaining_time": "20:58:17", "throughput": 5652.25, "total_tokens": 872136592}
+{"current_steps": 19950, "total_steps": 29704, "loss": 0.2974998950958252, "lr": 2.4332860276563763e-06, "epoch": 1.3432534338809587, "percentage": 67.16, "elapsed_time": "1 day, 18:52:15", "remaining_time": "20:57:38", "throughput": 5652.33, "total_tokens": 872357128}
+{"current_steps": 19955, "total_steps": 29704, "loss": 0.30943219661712645, "lr": 2.4310172785766396e-06, "epoch": 1.343590088876919, "percentage": 67.18, "elapsed_time": "1 day, 18:52:58", "remaining_time": "20:57:01", "throughput": 5652.18, "total_tokens": 872573624}
+{"current_steps": 19960, "total_steps": 29704, "loss": 0.31881489753723147, "lr": 2.428749247904724e-06, "epoch": 1.343926743872879, "percentage": 67.2, "elapsed_time": "1 day, 18:53:35", "remaining_time": "20:56:22", "throughput": 5652.21, "total_tokens": 872789824}
+{"current_steps": 19965, "total_steps": 29704, "loss": 0.3125936031341553, "lr": 2.4264819362748856e-06, "epoch": 1.3442633988688393, "percentage": 67.21, "elapsed_time": "1 day, 18:54:12", "remaining_time": "20:55:42", "throughput": 5652.3, "total_tokens": 873009864}
+{"current_steps": 19970, "total_steps": 29704, "loss": 0.3280599594116211, "lr": 2.4242153443211663e-06, "epoch": 1.3446000538647993, "percentage": 67.23, "elapsed_time": "1 day, 18:54:43", "remaining_time": "20:55:00", "throughput": 5652.57, "total_tokens": 873230264}
+{"current_steps": 19975, "total_steps": 29704, "loss": 0.2943082809448242, "lr": 2.4219494726774124e-06, "epoch": 1.3449367088607596, "percentage": 67.25, "elapsed_time": "1 day, 18:55:19", "remaining_time": "20:54:19", "throughput": 5652.72, "total_tokens": 873453312}
+{"current_steps": 19980, "total_steps": 29704, "loss": 0.2926133632659912, "lr": 2.419684321977268e-06, "epoch": 1.3452733638567196, "percentage": 67.26, "elapsed_time": "1 day, 18:55:53", "remaining_time": "20:53:39", "throughput": 5652.8, "total_tokens": 873660984}
+{"current_steps": 19985, "total_steps": 29704, "loss": 0.29560866355896, "lr": 2.417419892854175e-06, "epoch": 1.3456100188526798, "percentage": 67.28, "elapsed_time": "1 day, 18:56:38", "remaining_time": "20:53:03", "throughput": 5652.67, "total_tokens": 873892712}
+{"current_steps": 19990, "total_steps": 29704, "loss": 0.29713964462280273, "lr": 2.415156185941371e-06, "epoch": 1.3459466738486399, "percentage": 67.3, "elapsed_time": "1 day, 18:57:13", "remaining_time": "20:52:23", "throughput": 5652.74, "total_tokens": 874102984}
+{"current_steps": 19995, "total_steps": 29704, "loss": 0.30781164169311526, "lr": 2.412893201871902e-06, "epoch": 1.3462833288446001, "percentage": 67.31, "elapsed_time": "1 day, 18:57:54", "remaining_time": "20:51:45", "throughput": 5652.63, "total_tokens": 874317048}
+{"current_steps": 20000, "total_steps": 29704, "loss": 0.31746225357055663, "lr": 2.410630941278595e-06, "epoch": 1.3466199838405601, "percentage": 67.33, "elapsed_time": "1 day, 18:58:37", "remaining_time": "20:51:08", "throughput": 5652.49, "total_tokens": 874537536}
+{"current_steps": 20005, "total_steps": 29704, "loss": 0.3095715045928955, "lr": 2.408369404794092e-06, "epoch": 1.3469566388365204, "percentage": 67.35, "elapsed_time": "1 day, 18:59:12", "remaining_time": "20:50:28", "throughput": 5652.58, "total_tokens": 874752264}
+{"current_steps": 20010, "total_steps": 29704, "loss": 0.30551714897155763, "lr": 2.4061085930508176e-06, "epoch": 1.3472932938324804, "percentage": 67.36, "elapsed_time": "1 day, 18:59:49", "remaining_time": "20:49:49", "throughput": 5652.6, "total_tokens": 874964968}
+{"current_steps": 20015, "total_steps": 29704, "loss": 0.3143286228179932, "lr": 2.4038485066810058e-06, "epoch": 1.3476299488284407, "percentage": 67.38, "elapsed_time": "1 day, 19:00:36", "remaining_time": "20:49:14", "throughput": 5652.26, "total_tokens": 875176368}
+{"current_steps": 20020, "total_steps": 29704, "loss": 0.276894736289978, "lr": 2.40158914631668e-06, "epoch": 1.3479666038244007, "percentage": 67.4, "elapsed_time": "1 day, 19:01:10", "remaining_time": "20:48:33", "throughput": 5652.38, "total_tokens": 875385288}
+{"current_steps": 20025, "total_steps": 29704, "loss": 0.29750776290893555, "lr": 2.399330512589664e-06, "epoch": 1.348303258820361, "percentage": 67.42, "elapsed_time": "1 day, 19:01:51", "remaining_time": "20:47:56", "throughput": 5652.31, "total_tokens": 875609880}
+{"current_steps": 20030, "total_steps": 29704, "loss": 0.30971436500549315, "lr": 2.3970726061315775e-06, "epoch": 1.348639913816321, "percentage": 67.43, "elapsed_time": "1 day, 19:02:28", "remaining_time": "20:47:16", "throughput": 5652.39, "total_tokens": 875831888}
+{"current_steps": 20035, "total_steps": 29704, "loss": 0.30228476524353026, "lr": 2.394815427573836e-06, "epoch": 1.3489765688122812, "percentage": 67.45, "elapsed_time": "1 day, 19:03:12", "remaining_time": "20:46:40", "throughput": 5652.27, "total_tokens": 876061376}
+{"current_steps": 20040, "total_steps": 29704, "loss": 0.2862797021865845, "lr": 2.3925589775476516e-06, "epoch": 1.3493132238082413, "percentage": 67.47, "elapsed_time": "1 day, 19:03:52", "remaining_time": "20:46:01", "throughput": 5652.2, "total_tokens": 876272016}
+{"current_steps": 20045, "total_steps": 29704, "loss": 0.31173529624938967, "lr": 2.3903032566840377e-06, "epoch": 1.3496498788042015, "percentage": 67.48, "elapsed_time": "1 day, 19:04:33", "remaining_time": "20:45:24", "throughput": 5652.1, "total_tokens": 876494336}
+{"current_steps": 20050, "total_steps": 29704, "loss": 0.31239540576934816, "lr": 2.3880482656137926e-06, "epoch": 1.3499865338001615, "percentage": 67.5, "elapsed_time": "1 day, 19:05:14", "remaining_time": "20:44:46", "throughput": 5652.14, "total_tokens": 876727136}
+{"current_steps": 20055, "total_steps": 29704, "loss": 0.3017667055130005, "lr": 2.3857940049675254e-06, "epoch": 1.3503231887961218, "percentage": 67.52, "elapsed_time": "1 day, 19:05:54", "remaining_time": "20:44:08", "throughput": 5652.13, "total_tokens": 876951616}
+{"current_steps": 20060, "total_steps": 29704, "loss": 0.2906331062316895, "lr": 2.383540475375625e-06, "epoch": 1.3506598437920818, "percentage": 67.53, "elapsed_time": "1 day, 19:06:35", "remaining_time": "20:43:31", "throughput": 5652.05, "total_tokens": 877172960}
+{"current_steps": 20065, "total_steps": 29704, "loss": 0.30210440158843993, "lr": 2.3812876774682886e-06, "epoch": 1.350996498788042, "percentage": 67.55, "elapsed_time": "1 day, 19:07:09", "remaining_time": "20:42:50", "throughput": 5652.33, "total_tokens": 877405296}
+{"current_steps": 20070, "total_steps": 29704, "loss": 0.3308712959289551, "lr": 2.379035611875503e-06, "epoch": 1.351333153784002, "percentage": 67.57, "elapsed_time": "1 day, 19:07:41", "remaining_time": "20:42:08", "throughput": 5652.64, "total_tokens": 877636392}
+{"current_steps": 20075, "total_steps": 29704, "loss": 0.2974379062652588, "lr": 2.376784279227051e-06, "epoch": 1.3516698087799623, "percentage": 67.58, "elapsed_time": "1 day, 19:08:21", "remaining_time": "20:41:30", "throughput": 5652.7, "total_tokens": 877871256}
+{"current_steps": 20080, "total_steps": 29704, "loss": 0.3024725437164307, "lr": 2.3745336801525105e-06, "epoch": 1.3520064637759224, "percentage": 67.6, "elapsed_time": "1 day, 19:09:05", "remaining_time": "20:40:54", "throughput": 5652.59, "total_tokens": 878104688}
+{"current_steps": 20085, "total_steps": 29704, "loss": 0.3230156898498535, "lr": 2.3722838152812543e-06, "epoch": 1.3523431187718826, "percentage": 67.62, "elapsed_time": "1 day, 19:09:51", "remaining_time": "20:40:19", "throughput": 5652.32, "total_tokens": 878325104}
+{"current_steps": 20090, "total_steps": 29704, "loss": 0.2935077428817749, "lr": 2.3700346852424485e-06, "epoch": 1.3526797737678427, "percentage": 67.63, "elapsed_time": "1 day, 19:10:23", "remaining_time": "20:39:37", "throughput": 5652.55, "total_tokens": 878541320}
+{"current_steps": 20095, "total_steps": 29704, "loss": 0.3263228416442871, "lr": 2.3677862906650594e-06, "epoch": 1.353016428763803, "percentage": 67.65, "elapsed_time": "1 day, 19:10:57", "remaining_time": "20:38:56", "throughput": 5652.71, "total_tokens": 878755824}
+{"current_steps": 20100, "total_steps": 29704, "loss": 0.3379992961883545, "lr": 2.365538632177837e-06, "epoch": 1.353353083759763, "percentage": 67.67, "elapsed_time": "1 day, 19:11:39", "remaining_time": "20:38:19", "throughput": 5652.49, "total_tokens": 878961448}
+{"current_steps": 20105, "total_steps": 29704, "loss": 0.3040963649749756, "lr": 2.36329171040934e-06, "epoch": 1.3536897387557232, "percentage": 67.68, "elapsed_time": "1 day, 19:12:11", "remaining_time": "20:37:37", "throughput": 5652.82, "total_tokens": 879189696}
+{"current_steps": 20110, "total_steps": 29704, "loss": 0.2924042224884033, "lr": 2.3610455259879045e-06, "epoch": 1.3540263937516832, "percentage": 67.7, "elapsed_time": "1 day, 19:12:47", "remaining_time": "20:36:57", "throughput": 5652.86, "total_tokens": 879399600}
+{"current_steps": 20115, "total_steps": 29704, "loss": 0.3243632078170776, "lr": 2.358800079541674e-06, "epoch": 1.3543630487476435, "percentage": 67.72, "elapsed_time": "1 day, 19:13:27", "remaining_time": "20:36:19", "throughput": 5652.75, "total_tokens": 879611920}
+{"current_steps": 20120, "total_steps": 29704, "loss": 0.3272987365722656, "lr": 2.3565553716985794e-06, "epoch": 1.3546997037436035, "percentage": 67.73, "elapsed_time": "1 day, 19:14:01", "remaining_time": "20:35:38", "throughput": 5652.95, "total_tokens": 879834920}
+{"current_steps": 20125, "total_steps": 29704, "loss": 0.2879811763763428, "lr": 2.3543114030863457e-06, "epoch": 1.3550363587395637, "percentage": 67.75, "elapsed_time": "1 day, 19:14:39", "remaining_time": "20:34:59", "throughput": 5652.88, "total_tokens": 880035976}
+{"current_steps": 20130, "total_steps": 29704, "loss": 0.2902420997619629, "lr": 2.352068174332491e-06, "epoch": 1.3553730137355238, "percentage": 67.77, "elapsed_time": "1 day, 19:15:21", "remaining_time": "20:34:22", "throughput": 5652.64, "total_tokens": 880240280}
+{"current_steps": 20135, "total_steps": 29704, "loss": 0.2633873462677002, "lr": 2.349825686064329e-06, "epoch": 1.355709668731484, "percentage": 67.79, "elapsed_time": "1 day, 19:16:00", "remaining_time": "20:33:44", "throughput": 5652.62, "total_tokens": 880457816}
+{"current_steps": 20140, "total_steps": 29704, "loss": 0.31766154766082766, "lr": 2.3475839389089605e-06, "epoch": 1.356046323727444, "percentage": 67.8, "elapsed_time": "1 day, 19:16:42", "remaining_time": "20:33:06", "throughput": 5652.61, "total_tokens": 880688632}
+{"current_steps": 20145, "total_steps": 29704, "loss": 0.3041571617126465, "lr": 2.345342933493289e-06, "epoch": 1.3563829787234043, "percentage": 67.82, "elapsed_time": "1 day, 19:17:22", "remaining_time": "20:32:28", "throughput": 5652.5, "total_tokens": 880900704}
+{"current_steps": 20150, "total_steps": 29704, "loss": 0.2988900661468506, "lr": 2.343102670443998e-06, "epoch": 1.3567196337193643, "percentage": 67.84, "elapsed_time": "1 day, 19:18:05", "remaining_time": "20:31:51", "throughput": 5652.35, "total_tokens": 881116800}
+{"current_steps": 20155, "total_steps": 29704, "loss": 0.32421116828918456, "lr": 2.3408631503875763e-06, "epoch": 1.3570562887153246, "percentage": 67.85, "elapsed_time": "1 day, 19:18:45", "remaining_time": "20:31:13", "throughput": 5652.39, "total_tokens": 881349184}
+{"current_steps": 20160, "total_steps": 29704, "loss": 0.2701118946075439, "lr": 2.338624373950291e-06, "epoch": 1.3573929437112846, "percentage": 67.87, "elapsed_time": "1 day, 19:19:17", "remaining_time": "20:30:32", "throughput": 5652.62, "total_tokens": 881566872}
+{"current_steps": 20165, "total_steps": 29704, "loss": 0.30777089595794677, "lr": 2.336386341758216e-06, "epoch": 1.3577295987072449, "percentage": 67.89, "elapsed_time": "1 day, 19:19:54", "remaining_time": "20:29:52", "throughput": 5652.79, "total_tokens": 881802424}
+{"current_steps": 20170, "total_steps": 29704, "loss": 0.3082592964172363, "lr": 2.3341490544372063e-06, "epoch": 1.3580662537032049, "percentage": 67.9, "elapsed_time": "1 day, 19:20:33", "remaining_time": "20:29:14", "throughput": 5652.77, "total_tokens": 882020264}
+{"current_steps": 20175, "total_steps": 29704, "loss": 0.3048335790634155, "lr": 2.3319125126129128e-06, "epoch": 1.3584029086991651, "percentage": 67.92, "elapsed_time": "1 day, 19:21:10", "remaining_time": "20:28:34", "throughput": 5652.84, "total_tokens": 882240304}
+{"current_steps": 20180, "total_steps": 29704, "loss": 0.32310731410980226, "lr": 2.329676716910777e-06, "epoch": 1.3587395636951252, "percentage": 67.94, "elapsed_time": "1 day, 19:21:47", "remaining_time": "20:27:55", "throughput": 5652.86, "total_tokens": 882456760}
+{"current_steps": 20185, "total_steps": 29704, "loss": 0.3185749053955078, "lr": 2.3274416679560325e-06, "epoch": 1.3590762186910854, "percentage": 67.95, "elapsed_time": "1 day, 19:22:19", "remaining_time": "20:27:13", "throughput": 5653.12, "total_tokens": 882677128}
+{"current_steps": 20190, "total_steps": 29704, "loss": 0.31891803741455077, "lr": 2.3252073663737016e-06, "epoch": 1.3594128736870454, "percentage": 67.97, "elapsed_time": "1 day, 19:22:56", "remaining_time": "20:26:34", "throughput": 5653.28, "total_tokens": 882909016}
+{"current_steps": 20195, "total_steps": 29704, "loss": 0.312581205368042, "lr": 2.3229738127886054e-06, "epoch": 1.3597495286830057, "percentage": 67.99, "elapsed_time": "1 day, 19:23:41", "remaining_time": "20:25:58", "throughput": 5652.98, "total_tokens": 883114832}
+{"current_steps": 20200, "total_steps": 29704, "loss": 0.30325491428375245, "lr": 2.3207410078253416e-06, "epoch": 1.3600861836789657, "percentage": 68.0, "elapsed_time": "1 day, 19:24:28", "remaining_time": "20:25:23", "throughput": 5652.71, "total_tokens": 883339080}
+{"current_steps": 20205, "total_steps": 29704, "loss": 0.3062901496887207, "lr": 2.3185089521083154e-06, "epoch": 1.360422838674926, "percentage": 68.02, "elapsed_time": "1 day, 19:25:05", "remaining_time": "20:24:43", "throughput": 5652.74, "total_tokens": 883551600}
+{"current_steps": 20210, "total_steps": 29704, "loss": 0.3058837175369263, "lr": 2.3162776462617067e-06, "epoch": 1.360759493670886, "percentage": 68.04, "elapsed_time": "1 day, 19:25:51", "remaining_time": "20:24:08", "throughput": 5652.47, "total_tokens": 883769552}
+{"current_steps": 20215, "total_steps": 29704, "loss": 0.28980307579040526, "lr": 2.314047090909498e-06, "epoch": 1.3610961486668463, "percentage": 68.05, "elapsed_time": "1 day, 19:26:28", "remaining_time": "20:23:29", "throughput": 5652.51, "total_tokens": 883987024}
+{"current_steps": 20220, "total_steps": 29704, "loss": 0.289400577545166, "lr": 2.3118172866754557e-06, "epoch": 1.3614328036628063, "percentage": 68.07, "elapsed_time": "1 day, 19:27:16", "remaining_time": "20:22:55", "throughput": 5652.16, "total_tokens": 884204584}
+{"current_steps": 20225, "total_steps": 29704, "loss": 0.29900097846984863, "lr": 2.309588234183137e-06, "epoch": 1.3617694586587665, "percentage": 68.09, "elapsed_time": "1 day, 19:27:59", "remaining_time": "20:22:18", "throughput": 5652.06, "total_tokens": 884430104}
+{"current_steps": 20230, "total_steps": 29704, "loss": 0.3193103313446045, "lr": 2.3073599340558896e-06, "epoch": 1.3621061136547266, "percentage": 68.11, "elapsed_time": "1 day, 19:28:33", "remaining_time": "20:21:37", "throughput": 5652.15, "total_tokens": 884638136}
+{"current_steps": 20235, "total_steps": 29704, "loss": 0.31695847511291503, "lr": 2.3051323869168506e-06, "epoch": 1.3624427686506868, "percentage": 68.12, "elapsed_time": "1 day, 19:29:12", "remaining_time": "20:20:59", "throughput": 5652.19, "total_tokens": 884867312}
+{"current_steps": 20240, "total_steps": 29704, "loss": 0.299372124671936, "lr": 2.3029055933889443e-06, "epoch": 1.3627794236466468, "percentage": 68.14, "elapsed_time": "1 day, 19:29:51", "remaining_time": "20:20:20", "throughput": 5652.15, "total_tokens": 885078648}
+{"current_steps": 20245, "total_steps": 29704, "loss": 0.27194240093231203, "lr": 2.300679554094892e-06, "epoch": 1.363116078642607, "percentage": 68.16, "elapsed_time": "1 day, 19:30:24", "remaining_time": "20:19:39", "throughput": 5652.29, "total_tokens": 885288824}
+{"current_steps": 20250, "total_steps": 29704, "loss": 0.2977342367172241, "lr": 2.298454269657191e-06, "epoch": 1.3634527336385671, "percentage": 68.17, "elapsed_time": "1 day, 19:31:05", "remaining_time": "20:19:01", "throughput": 5652.22, "total_tokens": 885509192}
+{"current_steps": 20255, "total_steps": 29704, "loss": 0.30438592433929446, "lr": 2.2962297406981432e-06, "epoch": 1.3637893886345274, "percentage": 68.19, "elapsed_time": "1 day, 19:31:44", "remaining_time": "20:18:23", "throughput": 5652.21, "total_tokens": 885728368}
+{"current_steps": 20260, "total_steps": 29704, "loss": 0.2986233949661255, "lr": 2.2940059678398223e-06, "epoch": 1.3641260436304874, "percentage": 68.21, "elapsed_time": "1 day, 19:32:24", "remaining_time": "20:17:45", "throughput": 5652.16, "total_tokens": 885946024}
+{"current_steps": 20265, "total_steps": 29704, "loss": 0.30042304992675783, "lr": 2.291782951704106e-06, "epoch": 1.3644626986264476, "percentage": 68.22, "elapsed_time": "1 day, 19:33:01", "remaining_time": "20:17:05", "throughput": 5652.21, "total_tokens": 886159456}
+{"current_steps": 20270, "total_steps": 29704, "loss": 0.30829734802246095, "lr": 2.289560692912651e-06, "epoch": 1.3647993536224077, "percentage": 68.24, "elapsed_time": "1 day, 19:33:42", "remaining_time": "20:16:27", "throughput": 5652.11, "total_tokens": 886378184}
+{"current_steps": 20275, "total_steps": 29704, "loss": 0.30705411434173585, "lr": 2.287339192086906e-06, "epoch": 1.365136008618368, "percentage": 68.26, "elapsed_time": "1 day, 19:34:22", "remaining_time": "20:15:49", "throughput": 5652.05, "total_tokens": 886596912}
+{"current_steps": 20280, "total_steps": 29704, "loss": 0.30739502906799315, "lr": 2.2851184498481056e-06, "epoch": 1.365472663614328, "percentage": 68.27, "elapsed_time": "1 day, 19:35:00", "remaining_time": "20:15:10", "throughput": 5652.05, "total_tokens": 886810320}
+{"current_steps": 20285, "total_steps": 29704, "loss": 0.29115149974822996, "lr": 2.2828984668172732e-06, "epoch": 1.3658093186102882, "percentage": 68.29, "elapsed_time": "1 day, 19:35:38", "remaining_time": "20:14:31", "throughput": 5652.01, "total_tokens": 887018752}
+{"current_steps": 20290, "total_steps": 29704, "loss": 0.3051044464111328, "lr": 2.2806792436152185e-06, "epoch": 1.3661459736062485, "percentage": 68.31, "elapsed_time": "1 day, 19:36:15", "remaining_time": "20:13:52", "throughput": 5652.07, "total_tokens": 887236112}
+{"current_steps": 20295, "total_steps": 29704, "loss": 0.30942864418029786, "lr": 2.2784607808625463e-06, "epoch": 1.3664826286022085, "percentage": 68.32, "elapsed_time": "1 day, 19:36:54", "remaining_time": "20:13:13", "throughput": 5652.02, "total_tokens": 887448744}
+{"current_steps": 20300, "total_steps": 29704, "loss": 0.2999575138092041, "lr": 2.276243079179634e-06, "epoch": 1.3668192835981685, "percentage": 68.34, "elapsed_time": "1 day, 19:37:33", "remaining_time": "20:12:35", "throughput": 5651.98, "total_tokens": 887665496}
+{"current_steps": 20305, "total_steps": 29704, "loss": 0.31462917327880857, "lr": 2.2740261391866634e-06, "epoch": 1.3671559385941288, "percentage": 68.36, "elapsed_time": "1 day, 19:38:05", "remaining_time": "20:11:53", "throughput": 5652.25, "total_tokens": 887885544}
+{"current_steps": 20310, "total_steps": 29704, "loss": 0.2609941720962524, "lr": 2.2718099615035865e-06, "epoch": 1.367492593590089, "percentage": 68.37, "elapsed_time": "1 day, 19:38:46", "remaining_time": "20:11:16", "throughput": 5652.16, "total_tokens": 888107128}
+{"current_steps": 20315, "total_steps": 29704, "loss": 0.3140725612640381, "lr": 2.2695945467501567e-06, "epoch": 1.367829248586049, "percentage": 68.39, "elapsed_time": "1 day, 19:39:25", "remaining_time": "20:10:37", "throughput": 5652.19, "total_tokens": 888330336}
+{"current_steps": 20320, "total_steps": 29704, "loss": 0.2964684009552002, "lr": 2.267379895545905e-06, "epoch": 1.368165903582009, "percentage": 68.41, "elapsed_time": "1 day, 19:40:05", "remaining_time": "20:09:59", "throughput": 5652.16, "total_tokens": 888549184}
+{"current_steps": 20325, "total_steps": 29704, "loss": 0.3142958641052246, "lr": 2.2651660085101513e-06, "epoch": 1.3685025585779693, "percentage": 68.43, "elapsed_time": "1 day, 19:40:43", "remaining_time": "20:09:20", "throughput": 5652.21, "total_tokens": 888773632}
+{"current_steps": 20330, "total_steps": 29704, "loss": 0.30502166748046877, "lr": 2.262952886262003e-06, "epoch": 1.3688392135739296, "percentage": 68.44, "elapsed_time": "1 day, 19:41:25", "remaining_time": "20:08:43", "throughput": 5652.17, "total_tokens": 889003680}
+{"current_steps": 20335, "total_steps": 29704, "loss": 0.29352798461914065, "lr": 2.260740529420352e-06, "epoch": 1.3691758685698896, "percentage": 68.46, "elapsed_time": "1 day, 19:42:09", "remaining_time": "20:08:07", "throughput": 5651.97, "total_tokens": 889224368}
+{"current_steps": 20340, "total_steps": 29704, "loss": 0.29394705295562745, "lr": 2.2585289386038754e-06, "epoch": 1.3695125235658496, "percentage": 68.48, "elapsed_time": "1 day, 19:42:49", "remaining_time": "20:07:28", "throughput": 5652.02, "total_tokens": 889456760}
+{"current_steps": 20345, "total_steps": 29704, "loss": 0.3354970932006836, "lr": 2.256318114431043e-06, "epoch": 1.3698491785618099, "percentage": 68.49, "elapsed_time": "1 day, 19:43:24", "remaining_time": "20:06:48", "throughput": 5652.26, "total_tokens": 889693056}
+{"current_steps": 20350, "total_steps": 29704, "loss": 0.2972100734710693, "lr": 2.2541080575200973e-06, "epoch": 1.3701858335577701, "percentage": 68.51, "elapsed_time": "1 day, 19:44:07", "remaining_time": "20:06:11", "throughput": 5652.05, "total_tokens": 889902256}
+{"current_steps": 20355, "total_steps": 29704, "loss": 0.3170339822769165, "lr": 2.251898768489081e-06, "epoch": 1.3705224885537302, "percentage": 68.53, "elapsed_time": "1 day, 19:44:41", "remaining_time": "20:05:30", "throughput": 5652.32, "total_tokens": 890133696}
+{"current_steps": 20360, "total_steps": 29704, "loss": 0.31004133224487307, "lr": 2.249690247955807e-06, "epoch": 1.3708591435496902, "percentage": 68.54, "elapsed_time": "1 day, 19:45:25", "remaining_time": "20:04:54", "throughput": 5652.11, "total_tokens": 890349704}
+{"current_steps": 20365, "total_steps": 29704, "loss": 0.32703423500061035, "lr": 2.247482496537887e-06, "epoch": 1.3711957985456504, "percentage": 68.56, "elapsed_time": "1 day, 19:46:01", "remaining_time": "20:04:14", "throughput": 5652.21, "total_tokens": 890570904}
+{"current_steps": 20370, "total_steps": 29704, "loss": 0.3082599639892578, "lr": 2.24527551485271e-06, "epoch": 1.3715324535416107, "percentage": 68.58, "elapsed_time": "1 day, 19:46:38", "remaining_time": "20:03:35", "throughput": 5652.26, "total_tokens": 890789232}
+{"current_steps": 20375, "total_steps": 29704, "loss": 0.3278525352478027, "lr": 2.243069303517451e-06, "epoch": 1.3718691085375707, "percentage": 68.59, "elapsed_time": "1 day, 19:47:11", "remaining_time": "20:02:53", "throughput": 5652.44, "total_tokens": 891003792}
+{"current_steps": 20380, "total_steps": 29704, "loss": 0.2803332328796387, "lr": 2.2408638631490696e-06, "epoch": 1.3722057635335307, "percentage": 68.61, "elapsed_time": "1 day, 19:47:51", "remaining_time": "20:02:15", "throughput": 5652.4, "total_tokens": 891222136}
+{"current_steps": 20385, "total_steps": 29704, "loss": 0.2959303617477417, "lr": 2.2386591943643117e-06, "epoch": 1.372542418529491, "percentage": 68.63, "elapsed_time": "1 day, 19:48:29", "remaining_time": "20:01:36", "throughput": 5652.46, "total_tokens": 891447208}
+{"current_steps": 20390, "total_steps": 29704, "loss": 0.3103938579559326, "lr": 2.2364552977797023e-06, "epoch": 1.3728790735254512, "percentage": 68.64, "elapsed_time": "1 day, 19:49:11", "remaining_time": "20:00:59", "throughput": 5652.35, "total_tokens": 891668296}
+{"current_steps": 20395, "total_steps": 29704, "loss": 0.3227031707763672, "lr": 2.23425217401156e-06, "epoch": 1.3732157285214113, "percentage": 68.66, "elapsed_time": "1 day, 19:49:47", "remaining_time": "20:00:19", "throughput": 5652.47, "total_tokens": 891888096}
+{"current_steps": 20400, "total_steps": 29704, "loss": 0.2954218864440918, "lr": 2.2320498236759746e-06, "epoch": 1.3735523835173713, "percentage": 68.68, "elapsed_time": "1 day, 19:50:17", "remaining_time": "19:59:37", "throughput": 5652.78, "total_tokens": 892110048}
+{"current_steps": 20405, "total_steps": 29704, "loss": 0.2996779441833496, "lr": 2.2298482473888335e-06, "epoch": 1.3738890385133316, "percentage": 68.69, "elapsed_time": "1 day, 19:50:59", "remaining_time": "19:59:00", "throughput": 5652.66, "total_tokens": 892325640}
+{"current_steps": 20410, "total_steps": 29704, "loss": 0.3093862533569336, "lr": 2.227647445765792e-06, "epoch": 1.3742256935092918, "percentage": 68.71, "elapsed_time": "1 day, 19:51:39", "remaining_time": "19:58:21", "throughput": 5652.65, "total_tokens": 892551144}
+{"current_steps": 20415, "total_steps": 29704, "loss": 0.2831476926803589, "lr": 2.225447419422305e-06, "epoch": 1.3745623485052518, "percentage": 68.73, "elapsed_time": "1 day, 19:52:17", "remaining_time": "19:57:43", "throughput": 5652.62, "total_tokens": 892763912}
+{"current_steps": 20420, "total_steps": 29704, "loss": 0.29417757987976073, "lr": 2.223248168973599e-06, "epoch": 1.3748990035012119, "percentage": 68.74, "elapsed_time": "1 day, 19:52:56", "remaining_time": "19:57:04", "throughput": 5652.64, "total_tokens": 892984176}
+{"current_steps": 20425, "total_steps": 29704, "loss": 0.2913064479827881, "lr": 2.221049695034688e-06, "epoch": 1.375235658497172, "percentage": 68.76, "elapsed_time": "1 day, 19:53:29", "remaining_time": "19:56:23", "throughput": 5652.89, "total_tokens": 893210312}
+{"current_steps": 20430, "total_steps": 29704, "loss": 0.3103857755661011, "lr": 2.218851998220369e-06, "epoch": 1.3755723134931324, "percentage": 68.78, "elapsed_time": "1 day, 19:54:03", "remaining_time": "19:55:42", "throughput": 5653.08, "total_tokens": 893433344}
+{"current_steps": 20435, "total_steps": 29704, "loss": 0.2799984931945801, "lr": 2.21665507914522e-06, "epoch": 1.3759089684890924, "percentage": 68.8, "elapsed_time": "1 day, 19:54:42", "remaining_time": "19:55:03", "throughput": 5652.93, "total_tokens": 893630528}
+{"current_steps": 20440, "total_steps": 29704, "loss": 0.311299204826355, "lr": 2.214458938423601e-06, "epoch": 1.3762456234850524, "percentage": 68.81, "elapsed_time": "1 day, 19:55:25", "remaining_time": "19:54:26", "throughput": 5652.83, "total_tokens": 893854672}
+{"current_steps": 20445, "total_steps": 29704, "loss": 0.2979740142822266, "lr": 2.212263576669661e-06, "epoch": 1.3765822784810127, "percentage": 68.83, "elapsed_time": "1 day, 19:56:09", "remaining_time": "19:53:50", "throughput": 5652.7, "total_tokens": 894084224}
+{"current_steps": 20450, "total_steps": 29704, "loss": 0.3165121555328369, "lr": 2.2100689944973182e-06, "epoch": 1.376918933476973, "percentage": 68.85, "elapsed_time": "1 day, 19:56:55", "remaining_time": "19:53:15", "throughput": 5652.49, "total_tokens": 894311056}
+{"current_steps": 20455, "total_steps": 29704, "loss": 0.34031267166137696, "lr": 2.2078751925202886e-06, "epoch": 1.377255588472933, "percentage": 68.86, "elapsed_time": "1 day, 19:57:43", "remaining_time": "19:52:40", "throughput": 5652.26, "total_tokens": 894543864}
+{"current_steps": 20460, "total_steps": 29704, "loss": 0.30671300888061526, "lr": 2.205682171352054e-06, "epoch": 1.377592243468893, "percentage": 68.88, "elapsed_time": "1 day, 19:58:22", "remaining_time": "19:52:02", "throughput": 5652.26, "total_tokens": 894767288}
+{"current_steps": 20465, "total_steps": 29704, "loss": 0.287479043006897, "lr": 2.2034899316058926e-06, "epoch": 1.3779288984648532, "percentage": 68.9, "elapsed_time": "1 day, 19:59:03", "remaining_time": "19:51:24", "throughput": 5652.13, "total_tokens": 894979648}
+{"current_steps": 20470, "total_steps": 29704, "loss": 0.29344854354858396, "lr": 2.201298473894854e-06, "epoch": 1.3782655534608135, "percentage": 68.91, "elapsed_time": "1 day, 19:59:40", "remaining_time": "19:50:45", "throughput": 5652.11, "total_tokens": 895182680}
+{"current_steps": 20475, "total_steps": 29704, "loss": 0.30328798294067383, "lr": 2.1991077988317723e-06, "epoch": 1.3786022084567735, "percentage": 68.93, "elapsed_time": "1 day, 20:00:22", "remaining_time": "19:50:08", "throughput": 5652.05, "total_tokens": 895414024}
+{"current_steps": 20480, "total_steps": 29704, "loss": 0.2803157329559326, "lr": 2.1969179070292637e-06, "epoch": 1.3789388634527335, "percentage": 68.95, "elapsed_time": "1 day, 20:01:03", "remaining_time": "19:49:30", "throughput": 5651.95, "total_tokens": 895624392}
+{"current_steps": 20485, "total_steps": 29704, "loss": 0.3004035711288452, "lr": 2.1947287990997236e-06, "epoch": 1.3792755184486938, "percentage": 68.96, "elapsed_time": "1 day, 20:01:34", "remaining_time": "19:48:48", "throughput": 5652.19, "total_tokens": 895840912}
+{"current_steps": 20490, "total_steps": 29704, "loss": 0.299238920211792, "lr": 2.192540475655328e-06, "epoch": 1.379612173444654, "percentage": 68.98, "elapsed_time": "1 day, 20:02:12", "remaining_time": "19:48:09", "throughput": 5652.17, "total_tokens": 896053216}
+{"current_steps": 20495, "total_steps": 29704, "loss": 0.3175162076950073, "lr": 2.1903529373080397e-06, "epoch": 1.379948828440614, "percentage": 69.0, "elapsed_time": "1 day, 20:02:51", "remaining_time": "19:47:30", "throughput": 5652.14, "total_tokens": 896270408}
+{"current_steps": 20500, "total_steps": 29704, "loss": 0.2798121452331543, "lr": 2.1881661846695895e-06, "epoch": 1.380285483436574, "percentage": 69.01, "elapsed_time": "1 day, 20:03:24", "remaining_time": "19:46:49", "throughput": 5652.21, "total_tokens": 896464640}
+{"current_steps": 20505, "total_steps": 29704, "loss": 0.31658909320831297, "lr": 2.185980218351503e-06, "epoch": 1.3806221384325343, "percentage": 69.03, "elapsed_time": "1 day, 20:04:03", "remaining_time": "19:46:11", "throughput": 5652.15, "total_tokens": 896678032}
+{"current_steps": 20510, "total_steps": 29704, "loss": 0.3018453598022461, "lr": 2.1837950389650714e-06, "epoch": 1.3809587934284946, "percentage": 69.05, "elapsed_time": "1 day, 20:04:44", "remaining_time": "19:45:33", "throughput": 5652.15, "total_tokens": 896910472}
+{"current_steps": 20515, "total_steps": 29704, "loss": 0.30421650409698486, "lr": 2.1816106471213787e-06, "epoch": 1.3812954484244546, "percentage": 69.06, "elapsed_time": "1 day, 20:05:27", "remaining_time": "19:44:56", "throughput": 5652.08, "total_tokens": 897142376}
+{"current_steps": 20520, "total_steps": 29704, "loss": 0.32655086517333987, "lr": 2.179427043431282e-06, "epoch": 1.3816321034204146, "percentage": 69.08, "elapsed_time": "1 day, 20:06:00", "remaining_time": "19:44:15", "throughput": 5652.35, "total_tokens": 897368312}
+{"current_steps": 20525, "total_steps": 29704, "loss": 0.32046022415161135, "lr": 2.177244228505418e-06, "epoch": 1.381968758416375, "percentage": 69.1, "elapsed_time": "1 day, 20:06:40", "remaining_time": "19:43:37", "throughput": 5652.18, "total_tokens": 897570856}
+{"current_steps": 20530, "total_steps": 29704, "loss": 0.2898330926895142, "lr": 2.1750622029542047e-06, "epoch": 1.3823054134123351, "percentage": 69.12, "elapsed_time": "1 day, 20:07:16", "remaining_time": "19:42:57", "throughput": 5652.35, "total_tokens": 897801520}
+{"current_steps": 20535, "total_steps": 29704, "loss": 0.31139211654663085, "lr": 2.1728809673878374e-06, "epoch": 1.3826420684082952, "percentage": 69.13, "elapsed_time": "1 day, 20:07:57", "remaining_time": "19:42:19", "throughput": 5652.24, "total_tokens": 898013200}
+{"current_steps": 20540, "total_steps": 29704, "loss": 0.2919374227523804, "lr": 2.1707005224162915e-06, "epoch": 1.3829787234042552, "percentage": 69.15, "elapsed_time": "1 day, 20:08:36", "remaining_time": "19:41:41", "throughput": 5652.23, "total_tokens": 898234880}
+{"current_steps": 20545, "total_steps": 29704, "loss": 0.30305407047271726, "lr": 2.1685208686493257e-06, "epoch": 1.3833153784002155, "percentage": 69.17, "elapsed_time": "1 day, 20:09:18", "remaining_time": "19:41:04", "throughput": 5652.01, "total_tokens": 898436912}
+{"current_steps": 20550, "total_steps": 29704, "loss": 0.30848112106323244, "lr": 2.166342006696467e-06, "epoch": 1.3836520333961757, "percentage": 69.18, "elapsed_time": "1 day, 20:09:51", "remaining_time": "19:40:22", "throughput": 5652.2, "total_tokens": 898651112}
+{"current_steps": 20555, "total_steps": 29704, "loss": 0.3011596202850342, "lr": 2.1641639371670335e-06, "epoch": 1.3839886883921357, "percentage": 69.2, "elapsed_time": "1 day, 20:10:24", "remaining_time": "19:39:41", "throughput": 5652.39, "total_tokens": 898866064}
+{"current_steps": 20560, "total_steps": 29704, "loss": 0.3262584686279297, "lr": 2.161986660670109e-06, "epoch": 1.3843253433880958, "percentage": 69.22, "elapsed_time": "1 day, 20:10:55", "remaining_time": "19:38:59", "throughput": 5652.7, "total_tokens": 899090152}
+{"current_steps": 20565, "total_steps": 29704, "loss": 0.2925774574279785, "lr": 2.159810177814566e-06, "epoch": 1.384661998384056, "percentage": 69.23, "elapsed_time": "1 day, 20:11:28", "remaining_time": "19:38:18", "throughput": 5652.82, "total_tokens": 899299560}
+{"current_steps": 20570, "total_steps": 29704, "loss": 0.3194908618927002, "lr": 2.1576344892090507e-06, "epoch": 1.3849986533800163, "percentage": 69.25, "elapsed_time": "1 day, 20:12:07", "remaining_time": "19:37:39", "throughput": 5652.81, "total_tokens": 899514728}
+{"current_steps": 20575, "total_steps": 29704, "loss": 0.30469086170196535, "lr": 2.1554595954619866e-06, "epoch": 1.3853353083759763, "percentage": 69.27, "elapsed_time": "1 day, 20:12:48", "remaining_time": "19:37:02", "throughput": 5652.7, "total_tokens": 899730168}
+{"current_steps": 20580, "total_steps": 29704, "loss": 0.30314340591430666, "lr": 2.1532854971815763e-06, "epoch": 1.3856719633719363, "percentage": 69.28, "elapsed_time": "1 day, 20:13:30", "remaining_time": "19:36:24", "throughput": 5652.52, "total_tokens": 899940496}
+{"current_steps": 20585, "total_steps": 29704, "loss": 0.2792366504669189, "lr": 2.151112194975799e-06, "epoch": 1.3860086183678966, "percentage": 69.3, "elapsed_time": "1 day, 20:14:16", "remaining_time": "19:35:49", "throughput": 5652.3, "total_tokens": 900163032}
+{"current_steps": 20590, "total_steps": 29704, "loss": 0.2858273506164551, "lr": 2.1489396894524106e-06, "epoch": 1.3863452733638568, "percentage": 69.32, "elapsed_time": "1 day, 20:15:03", "remaining_time": "19:35:14", "throughput": 5652.06, "total_tokens": 900392200}
+{"current_steps": 20595, "total_steps": 29704, "loss": 0.3015708923339844, "lr": 2.1467679812189497e-06, "epoch": 1.3866819283598169, "percentage": 69.33, "elapsed_time": "1 day, 20:15:40", "remaining_time": "19:34:35", "throughput": 5652.11, "total_tokens": 900610928}
+{"current_steps": 20600, "total_steps": 29704, "loss": 0.30586059093475343, "lr": 2.1445970708827213e-06, "epoch": 1.3870185833557769, "percentage": 69.35, "elapsed_time": "1 day, 20:16:20", "remaining_time": "19:33:56", "throughput": 5651.97, "total_tokens": 900814704}
+{"current_steps": 20605, "total_steps": 29704, "loss": 0.289659309387207, "lr": 2.14242695905082e-06, "epoch": 1.3873552383517371, "percentage": 69.37, "elapsed_time": "1 day, 20:16:52", "remaining_time": "19:33:15", "throughput": 5652.21, "total_tokens": 901036328}
+{"current_steps": 20610, "total_steps": 29704, "loss": 0.3204467535018921, "lr": 2.1402576463301044e-06, "epoch": 1.3876918933476974, "percentage": 69.38, "elapsed_time": "1 day, 20:17:36", "remaining_time": "19:32:38", "throughput": 5652.15, "total_tokens": 901271056}
+{"current_steps": 20615, "total_steps": 29704, "loss": 0.28775959014892577, "lr": 2.13808913332722e-06, "epoch": 1.3880285483436574, "percentage": 69.4, "elapsed_time": "1 day, 20:18:14", "remaining_time": "19:32:00", "throughput": 5652.13, "total_tokens": 901485976}
+{"current_steps": 20620, "total_steps": 29704, "loss": 0.32880635261535646, "lr": 2.1359214206485845e-06, "epoch": 1.3883652033396174, "percentage": 69.42, "elapsed_time": "1 day, 20:18:53", "remaining_time": "19:31:21", "throughput": 5652.13, "total_tokens": 901703104}
+{"current_steps": 20625, "total_steps": 29704, "loss": 0.2626586675643921, "lr": 2.1337545089003905e-06, "epoch": 1.3887018583355777, "percentage": 69.44, "elapsed_time": "1 day, 20:19:26", "remaining_time": "19:30:40", "throughput": 5652.2, "total_tokens": 901902224}
+{"current_steps": 20630, "total_steps": 29704, "loss": 0.31684534549713134, "lr": 2.131588398688608e-06, "epoch": 1.389038513331538, "percentage": 69.45, "elapsed_time": "1 day, 20:20:11", "remaining_time": "19:30:04", "throughput": 5652.01, "total_tokens": 902125584}
+{"current_steps": 20635, "total_steps": 29704, "loss": 0.28316431045532225, "lr": 2.1294230906189833e-06, "epoch": 1.389375168327498, "percentage": 69.47, "elapsed_time": "1 day, 20:20:53", "remaining_time": "19:29:27", "throughput": 5651.82, "total_tokens": 902333832}
+{"current_steps": 20640, "total_steps": 29704, "loss": 0.29244768619537354, "lr": 2.1272585852970358e-06, "epoch": 1.389711823323458, "percentage": 69.49, "elapsed_time": "1 day, 20:21:32", "remaining_time": "19:28:48", "throughput": 5651.74, "total_tokens": 902541064}
+{"current_steps": 20645, "total_steps": 29704, "loss": 0.2933189392089844, "lr": 2.125094883328068e-06, "epoch": 1.3900484783194182, "percentage": 69.5, "elapsed_time": "1 day, 20:22:13", "remaining_time": "19:28:10", "throughput": 5651.58, "total_tokens": 902745264}
+{"current_steps": 20650, "total_steps": 29704, "loss": 0.2766158103942871, "lr": 2.1229319853171447e-06, "epoch": 1.3903851333153785, "percentage": 69.52, "elapsed_time": "1 day, 20:22:47", "remaining_time": "19:27:30", "throughput": 5651.7, "total_tokens": 902957784}
+{"current_steps": 20655, "total_steps": 29704, "loss": 0.3175971508026123, "lr": 2.1207698918691217e-06, "epoch": 1.3907217883113385, "percentage": 69.54, "elapsed_time": "1 day, 20:23:29", "remaining_time": "19:26:52", "throughput": 5651.6, "total_tokens": 903176368}
+{"current_steps": 20660, "total_steps": 29704, "loss": 0.28777422904968264, "lr": 2.1186086035886133e-06, "epoch": 1.3910584433072986, "percentage": 69.55, "elapsed_time": "1 day, 20:24:07", "remaining_time": "19:26:13", "throughput": 5651.66, "total_tokens": 903403424}
+{"current_steps": 20665, "total_steps": 29704, "loss": 0.309741735458374, "lr": 2.116448121080023e-06, "epoch": 1.3913950983032588, "percentage": 69.57, "elapsed_time": "1 day, 20:24:48", "remaining_time": "19:25:36", "throughput": 5651.72, "total_tokens": 903643016}
+{"current_steps": 20670, "total_steps": 29704, "loss": 0.30854334831237795, "lr": 2.1142884449475205e-06, "epoch": 1.391731753299219, "percentage": 69.59, "elapsed_time": "1 day, 20:25:25", "remaining_time": "19:24:56", "throughput": 5651.75, "total_tokens": 903861496}
+{"current_steps": 20675, "total_steps": 29704, "loss": 0.2980179309844971, "lr": 2.112129575795052e-06, "epoch": 1.392068408295179, "percentage": 69.6, "elapsed_time": "1 day, 20:26:10", "remaining_time": "19:24:21", "throughput": 5651.55, "total_tokens": 904082696}
+{"current_steps": 20680, "total_steps": 29704, "loss": 0.2740116357803345, "lr": 2.1099715142263394e-06, "epoch": 1.3924050632911391, "percentage": 69.62, "elapsed_time": "1 day, 20:26:52", "remaining_time": "19:23:43", "throughput": 5651.4, "total_tokens": 904295832}
+{"current_steps": 20685, "total_steps": 29704, "loss": 0.28600759506225587, "lr": 2.1078142608448773e-06, "epoch": 1.3927417182870994, "percentage": 69.64, "elapsed_time": "1 day, 20:27:26", "remaining_time": "19:23:02", "throughput": 5651.58, "total_tokens": 904516064}
+{"current_steps": 20690, "total_steps": 29704, "loss": 0.32352604866027834, "lr": 2.1056578162539327e-06, "epoch": 1.3930783732830596, "percentage": 69.65, "elapsed_time": "1 day, 20:28:08", "remaining_time": "19:22:25", "throughput": 5651.48, "total_tokens": 904737152}
+{"current_steps": 20695, "total_steps": 29704, "loss": 0.30927181243896484, "lr": 2.103502181056554e-06, "epoch": 1.3934150282790196, "percentage": 69.67, "elapsed_time": "1 day, 20:28:50", "remaining_time": "19:21:48", "throughput": 5651.34, "total_tokens": 904951984}
+{"current_steps": 20700, "total_steps": 29704, "loss": 0.2932461738586426, "lr": 2.1013473558555503e-06, "epoch": 1.3937516832749797, "percentage": 69.69, "elapsed_time": "1 day, 20:29:37", "remaining_time": "19:21:13", "throughput": 5651.15, "total_tokens": 905188432}
+{"current_steps": 20705, "total_steps": 29704, "loss": 0.29349417686462403, "lr": 2.099193341253519e-06, "epoch": 1.39408833827094, "percentage": 69.7, "elapsed_time": "1 day, 20:30:14", "remaining_time": "19:20:33", "throughput": 5651.16, "total_tokens": 905395280}
+{"current_steps": 20710, "total_steps": 29704, "loss": 0.286403751373291, "lr": 2.0970401378528167e-06, "epoch": 1.3944249932669002, "percentage": 69.72, "elapsed_time": "1 day, 20:31:01", "remaining_time": "19:19:58", "throughput": 5650.77, "total_tokens": 905601488}
+{"current_steps": 20715, "total_steps": 29704, "loss": 0.2749554872512817, "lr": 2.0948877462555844e-06, "epoch": 1.3947616482628602, "percentage": 69.74, "elapsed_time": "1 day, 20:31:44", "remaining_time": "19:19:22", "throughput": 5650.67, "total_tokens": 905827264}
+{"current_steps": 20720, "total_steps": 29704, "loss": 0.32062878608703616, "lr": 2.09273616706373e-06, "epoch": 1.3950983032588202, "percentage": 69.75, "elapsed_time": "1 day, 20:32:17", "remaining_time": "19:18:40", "throughput": 5650.95, "total_tokens": 906057568}
+{"current_steps": 20725, "total_steps": 29704, "loss": 0.3149780035018921, "lr": 2.090585400878936e-06, "epoch": 1.3954349582547805, "percentage": 69.77, "elapsed_time": "1 day, 20:33:00", "remaining_time": "19:18:04", "throughput": 5650.79, "total_tokens": 906277360}
+{"current_steps": 20730, "total_steps": 29704, "loss": 0.3101172924041748, "lr": 2.0884354483026576e-06, "epoch": 1.3957716132507407, "percentage": 69.79, "elapsed_time": "1 day, 20:33:33", "remaining_time": "19:17:22", "throughput": 5651.02, "total_tokens": 906499640}
+{"current_steps": 20735, "total_steps": 29704, "loss": 0.30629117488861085, "lr": 2.0862863099361214e-06, "epoch": 1.3961082682467008, "percentage": 69.81, "elapsed_time": "1 day, 20:34:05", "remaining_time": "19:16:41", "throughput": 5651.35, "total_tokens": 906734072}
+{"current_steps": 20740, "total_steps": 29704, "loss": 0.3239948272705078, "lr": 2.084137986380326e-06, "epoch": 1.3964449232426608, "percentage": 69.82, "elapsed_time": "1 day, 20:34:46", "remaining_time": "19:16:03", "throughput": 5651.24, "total_tokens": 906945840}
+{"current_steps": 20745, "total_steps": 29704, "loss": 0.28044352531433103, "lr": 2.081990478236048e-06, "epoch": 1.396781578238621, "percentage": 69.84, "elapsed_time": "1 day, 20:35:22", "remaining_time": "19:15:23", "throughput": 5651.23, "total_tokens": 907146456}
+{"current_steps": 20750, "total_steps": 29704, "loss": 0.2951197624206543, "lr": 2.079843786103824e-06, "epoch": 1.3971182332345813, "percentage": 69.86, "elapsed_time": "1 day, 20:36:05", "remaining_time": "19:14:46", "throughput": 5651.02, "total_tokens": 907356536}
+{"current_steps": 20755, "total_steps": 29704, "loss": 0.31067094802856443, "lr": 2.0776979105839774e-06, "epoch": 1.3974548882305413, "percentage": 69.87, "elapsed_time": "1 day, 20:36:45", "remaining_time": "19:14:09", "throughput": 5651.01, "total_tokens": 907585320}
+{"current_steps": 20760, "total_steps": 29704, "loss": 0.33663692474365237, "lr": 2.075552852276588e-06, "epoch": 1.3977915432265013, "percentage": 69.89, "elapsed_time": "1 day, 20:37:27", "remaining_time": "19:13:31", "throughput": 5650.96, "total_tokens": 907812376}
+{"current_steps": 20765, "total_steps": 29704, "loss": 0.3083709716796875, "lr": 2.07340861178152e-06, "epoch": 1.3981281982224616, "percentage": 69.91, "elapsed_time": "1 day, 20:38:04", "remaining_time": "19:12:51", "throughput": 5651.02, "total_tokens": 908030480}
+{"current_steps": 20770, "total_steps": 29704, "loss": 0.30435686111450194, "lr": 2.0712651896984013e-06, "epoch": 1.3984648532184218, "percentage": 69.92, "elapsed_time": "1 day, 20:38:38", "remaining_time": "19:12:11", "throughput": 5651.21, "total_tokens": 908257064}
+{"current_steps": 20775, "total_steps": 29704, "loss": 0.3061100244522095, "lr": 2.0691225866266335e-06, "epoch": 1.3988015082143819, "percentage": 69.94, "elapsed_time": "1 day, 20:39:18", "remaining_time": "19:11:33", "throughput": 5651.14, "total_tokens": 908466856}
+{"current_steps": 20780, "total_steps": 29704, "loss": 0.3086406707763672, "lr": 2.066980803165388e-06, "epoch": 1.3991381632103421, "percentage": 69.96, "elapsed_time": "1 day, 20:39:49", "remaining_time": "19:10:51", "throughput": 5651.43, "total_tokens": 908690664}
+{"current_steps": 20785, "total_steps": 29704, "loss": 0.30843343734741213, "lr": 2.0648398399136088e-06, "epoch": 1.3994748182063022, "percentage": 69.97, "elapsed_time": "1 day, 20:40:33", "remaining_time": "19:10:14", "throughput": 5651.3, "total_tokens": 908916400}
+{"current_steps": 20790, "total_steps": 29704, "loss": 0.295151948928833, "lr": 2.062699697470007e-06, "epoch": 1.3998114732022624, "percentage": 69.99, "elapsed_time": "1 day, 20:41:17", "remaining_time": "19:09:38", "throughput": 5651.06, "total_tokens": 909128960}
+{"current_steps": 20795, "total_steps": 29704, "loss": 0.33457579612731936, "lr": 2.060560376433072e-06, "epoch": 1.4001481281982224, "percentage": 70.01, "elapsed_time": "1 day, 20:42:00", "remaining_time": "19:09:01", "throughput": 5650.98, "total_tokens": 909357768}
+{"current_steps": 20800, "total_steps": 29704, "loss": 0.3098001956939697, "lr": 2.05842187740105e-06, "epoch": 1.4004847831941827, "percentage": 70.02, "elapsed_time": "1 day, 20:42:43", "remaining_time": "19:08:24", "throughput": 5650.8, "total_tokens": 909572480}
+{"current_steps": 20805, "total_steps": 29704, "loss": 0.2820146560668945, "lr": 2.0562842009719745e-06, "epoch": 1.4008214381901427, "percentage": 70.04, "elapsed_time": "1 day, 20:43:28", "remaining_time": "19:07:48", "throughput": 5650.35, "total_tokens": 909757296}
+{"current_steps": 20810, "total_steps": 29704, "loss": 0.2844945669174194, "lr": 2.0541473477436317e-06, "epoch": 1.401158093186103, "percentage": 70.06, "elapsed_time": "1 day, 20:44:03", "remaining_time": "19:07:08", "throughput": 5650.5, "total_tokens": 909975984}
+{"current_steps": 20815, "total_steps": 29704, "loss": 0.28799991607666015, "lr": 2.052011318313591e-06, "epoch": 1.401494748182063, "percentage": 70.07, "elapsed_time": "1 day, 20:44:40", "remaining_time": "19:06:29", "throughput": 5650.57, "total_tokens": 910199280}
+{"current_steps": 20820, "total_steps": 29704, "loss": 0.32419145107269287, "lr": 2.049876113279185e-06, "epoch": 1.4018314031780232, "percentage": 70.09, "elapsed_time": "1 day, 20:45:15", "remaining_time": "19:05:48", "throughput": 5650.74, "total_tokens": 910419968}
+{"current_steps": 20825, "total_steps": 29704, "loss": 0.28672337532043457, "lr": 2.0477417332375165e-06, "epoch": 1.4021680581739833, "percentage": 70.11, "elapsed_time": "1 day, 20:45:46", "remaining_time": "19:05:06", "throughput": 5651.07, "total_tokens": 910648216}
+{"current_steps": 20830, "total_steps": 29704, "loss": 0.29271349906921384, "lr": 2.045608178785458e-06, "epoch": 1.4025047131699435, "percentage": 70.13, "elapsed_time": "1 day, 20:46:25", "remaining_time": "19:04:28", "throughput": 5651.08, "total_tokens": 910871288}
+{"current_steps": 20835, "total_steps": 29704, "loss": 0.29650259017944336, "lr": 2.043475450519652e-06, "epoch": 1.4028413681659035, "percentage": 70.14, "elapsed_time": "1 day, 20:46:57", "remaining_time": "19:03:46", "throughput": 5651.31, "total_tokens": 911087016}
+{"current_steps": 20840, "total_steps": 29704, "loss": 0.31277635097503664, "lr": 2.0413435490365064e-06, "epoch": 1.4031780231618638, "percentage": 70.16, "elapsed_time": "1 day, 20:47:35", "remaining_time": "19:03:07", "throughput": 5651.24, "total_tokens": 911295624}
+{"current_steps": 20845, "total_steps": 29704, "loss": 0.31907684803009034, "lr": 2.0392124749322064e-06, "epoch": 1.4035146781578238, "percentage": 70.18, "elapsed_time": "1 day, 20:48:18", "remaining_time": "19:02:30", "throughput": 5651.16, "total_tokens": 911522672}
+{"current_steps": 20850, "total_steps": 29704, "loss": 0.28636276721954346, "lr": 2.037082228802693e-06, "epoch": 1.403851333153784, "percentage": 70.19, "elapsed_time": "1 day, 20:48:52", "remaining_time": "19:01:50", "throughput": 5651.33, "total_tokens": 911742024}
+{"current_steps": 20855, "total_steps": 29704, "loss": 0.279879093170166, "lr": 2.03495281124369e-06, "epoch": 1.404187988149744, "percentage": 70.21, "elapsed_time": "1 day, 20:49:31", "remaining_time": "19:01:11", "throughput": 5651.3, "total_tokens": 911958464}
+{"current_steps": 20860, "total_steps": 29704, "loss": 0.28923535346984863, "lr": 2.0328242228506746e-06, "epoch": 1.4045246431457044, "percentage": 70.23, "elapsed_time": "1 day, 20:50:10", "remaining_time": "19:00:33", "throughput": 5651.26, "total_tokens": 912172576}
+{"current_steps": 20865, "total_steps": 29704, "loss": 0.3036557674407959, "lr": 2.0306964642189053e-06, "epoch": 1.4048612981416644, "percentage": 70.24, "elapsed_time": "1 day, 20:50:49", "remaining_time": "18:59:54", "throughput": 5651.24, "total_tokens": 912389440}
+{"current_steps": 20870, "total_steps": 29704, "loss": 0.2902818202972412, "lr": 2.028569535943402e-06, "epoch": 1.4051979531376246, "percentage": 70.26, "elapsed_time": "1 day, 20:51:27", "remaining_time": "18:59:15", "throughput": 5651.21, "total_tokens": 912599400}
+{"current_steps": 20875, "total_steps": 29704, "loss": 0.30182933807373047, "lr": 2.0264434386189524e-06, "epoch": 1.4055346081335847, "percentage": 70.28, "elapsed_time": "1 day, 20:52:02", "remaining_time": "18:58:35", "throughput": 5651.32, "total_tokens": 912816512}
+{"current_steps": 20880, "total_steps": 29704, "loss": 0.29143590927124025, "lr": 2.0243181728401136e-06, "epoch": 1.405871263129545, "percentage": 70.29, "elapsed_time": "1 day, 20:52:45", "remaining_time": "18:57:58", "throughput": 5651.18, "total_tokens": 913038160}
+{"current_steps": 20885, "total_steps": 29704, "loss": 0.31396236419677737, "lr": 2.022193739201208e-06, "epoch": 1.406207918125505, "percentage": 70.31, "elapsed_time": "1 day, 20:53:19", "remaining_time": "18:57:17", "throughput": 5651.23, "total_tokens": 913238344}
+{"current_steps": 20890, "total_steps": 29704, "loss": 0.2820024013519287, "lr": 2.0200701382963267e-06, "epoch": 1.4065445731214652, "percentage": 70.33, "elapsed_time": "1 day, 20:54:00", "remaining_time": "18:56:40", "throughput": 5651.14, "total_tokens": 913456008}
+{"current_steps": 20895, "total_steps": 29704, "loss": 0.2967421531677246, "lr": 2.0179473707193324e-06, "epoch": 1.4068812281174252, "percentage": 70.34, "elapsed_time": "1 day, 20:54:41", "remaining_time": "18:56:02", "throughput": 5651.13, "total_tokens": 913682016}
+{"current_steps": 20900, "total_steps": 29704, "loss": 0.30292630195617676, "lr": 2.0158254370638435e-06, "epoch": 1.4072178831133855, "percentage": 70.36, "elapsed_time": "1 day, 20:55:23", "remaining_time": "18:55:25", "throughput": 5651.02, "total_tokens": 913903464}
+{"current_steps": 20905, "total_steps": 29704, "loss": 0.3011344909667969, "lr": 2.0137043379232585e-06, "epoch": 1.4075545381093455, "percentage": 70.38, "elapsed_time": "1 day, 20:56:05", "remaining_time": "18:54:47", "throughput": 5650.94, "total_tokens": 914128144}
+{"current_steps": 20910, "total_steps": 29704, "loss": 0.2961131572723389, "lr": 2.0115840738907293e-06, "epoch": 1.4078911931053057, "percentage": 70.39, "elapsed_time": "1 day, 20:56:43", "remaining_time": "18:54:08", "throughput": 5650.88, "total_tokens": 914332560}
+{"current_steps": 20915, "total_steps": 29704, "loss": 0.31174521446228026, "lr": 2.0094646455591865e-06, "epoch": 1.4082278481012658, "percentage": 70.41, "elapsed_time": "1 day, 20:57:26", "remaining_time": "18:53:31", "throughput": 5650.75, "total_tokens": 914554936}
+{"current_steps": 20920, "total_steps": 29704, "loss": 0.3245561599731445, "lr": 2.007346053521319e-06, "epoch": 1.408564503097226, "percentage": 70.43, "elapsed_time": "1 day, 20:58:05", "remaining_time": "18:52:53", "throughput": 5650.76, "total_tokens": 914776376}
+{"current_steps": 20925, "total_steps": 29704, "loss": 0.3163960695266724, "lr": 2.005228298369583e-06, "epoch": 1.408901158093186, "percentage": 70.45, "elapsed_time": "1 day, 20:58:44", "remaining_time": "18:52:14", "throughput": 5650.64, "total_tokens": 914979152}
+{"current_steps": 20930, "total_steps": 29704, "loss": 0.282145619392395, "lr": 2.003111380696208e-06, "epoch": 1.4092378130891463, "percentage": 70.46, "elapsed_time": "1 day, 20:59:32", "remaining_time": "18:51:40", "throughput": 5650.35, "total_tokens": 915204416}
+{"current_steps": 20935, "total_steps": 29704, "loss": 0.2879758834838867, "lr": 2.0009953010931747e-06, "epoch": 1.4095744680851063, "percentage": 70.48, "elapsed_time": "1 day, 21:00:18", "remaining_time": "18:51:04", "throughput": 5650.12, "total_tokens": 915424616}
+{"current_steps": 20940, "total_steps": 29704, "loss": 0.32476277351379396, "lr": 1.998880060152244e-06, "epoch": 1.4099111230810666, "percentage": 70.5, "elapsed_time": "1 day, 21:00:57", "remaining_time": "18:50:25", "throughput": 5650.16, "total_tokens": 915651288}
+{"current_steps": 20945, "total_steps": 29704, "loss": 0.3187078952789307, "lr": 1.9967656584649354e-06, "epoch": 1.4102477780770266, "percentage": 70.51, "elapsed_time": "1 day, 21:01:34", "remaining_time": "18:49:46", "throughput": 5650.21, "total_tokens": 915870040}
+{"current_steps": 20950, "total_steps": 29704, "loss": 0.3237470149993896, "lr": 1.994652096622533e-06, "epoch": 1.4105844330729869, "percentage": 70.53, "elapsed_time": "1 day, 21:02:15", "remaining_time": "18:49:08", "throughput": 5650.16, "total_tokens": 916090120}
+{"current_steps": 20955, "total_steps": 29704, "loss": 0.3227944612503052, "lr": 1.992539375216089e-06, "epoch": 1.410921088068947, "percentage": 70.55, "elapsed_time": "1 day, 21:02:52", "remaining_time": "18:48:29", "throughput": 5650.23, "total_tokens": 916315064}
+{"current_steps": 20960, "total_steps": 29704, "loss": 0.30767481327056884, "lr": 1.9904274948364184e-06, "epoch": 1.4112577430649071, "percentage": 70.56, "elapsed_time": "1 day, 21:03:33", "remaining_time": "18:47:51", "throughput": 5650.22, "total_tokens": 916539152}
+{"current_steps": 20965, "total_steps": 29704, "loss": 0.30903284549713134, "lr": 1.988316456074102e-06, "epoch": 1.4115943980608672, "percentage": 70.58, "elapsed_time": "1 day, 21:04:10", "remaining_time": "18:47:11", "throughput": 5650.33, "total_tokens": 916766256}
+{"current_steps": 20970, "total_steps": 29704, "loss": 0.31888184547424314, "lr": 1.9862062595194854e-06, "epoch": 1.4119310530568274, "percentage": 70.6, "elapsed_time": "1 day, 21:04:48", "remaining_time": "18:46:33", "throughput": 5650.29, "total_tokens": 916978512}
+{"current_steps": 20975, "total_steps": 29704, "loss": 0.2966637372970581, "lr": 1.984096905762676e-06, "epoch": 1.4122677080527875, "percentage": 70.61, "elapsed_time": "1 day, 21:05:31", "remaining_time": "18:45:56", "throughput": 5650.1, "total_tokens": 917190424}
+{"current_steps": 20980, "total_steps": 29704, "loss": 0.3042705535888672, "lr": 1.9819883953935533e-06, "epoch": 1.4126043630487477, "percentage": 70.63, "elapsed_time": "1 day, 21:06:08", "remaining_time": "18:45:16", "throughput": 5650.25, "total_tokens": 917420304}
+{"current_steps": 20985, "total_steps": 29704, "loss": 0.2959493637084961, "lr": 1.9798807290017485e-06, "epoch": 1.4129410180447077, "percentage": 70.65, "elapsed_time": "1 day, 21:06:43", "remaining_time": "18:44:36", "throughput": 5650.32, "total_tokens": 917628664}
+{"current_steps": 20990, "total_steps": 29704, "loss": 0.30003719329833983, "lr": 1.9777739071766693e-06, "epoch": 1.413277673040668, "percentage": 70.66, "elapsed_time": "1 day, 21:07:25", "remaining_time": "18:43:59", "throughput": 5650.15, "total_tokens": 917841064}
+{"current_steps": 20995, "total_steps": 29704, "loss": 0.2987996578216553, "lr": 1.97566793050748e-06, "epoch": 1.413614328036628, "percentage": 70.68, "elapsed_time": "1 day, 21:08:00", "remaining_time": "18:43:19", "throughput": 5650.16, "total_tokens": 918043080}
+{"current_steps": 21000, "total_steps": 29704, "loss": 0.2807154178619385, "lr": 1.9735627995831096e-06, "epoch": 1.4139509830325883, "percentage": 70.7, "elapsed_time": "1 day, 21:08:31", "remaining_time": "18:42:37", "throughput": 5650.48, "total_tokens": 918268936}
+{"current_steps": 21005, "total_steps": 29704, "loss": 0.322419261932373, "lr": 1.9714585149922523e-06, "epoch": 1.4142876380285483, "percentage": 70.71, "elapsed_time": "1 day, 21:09:08", "remaining_time": "18:41:57", "throughput": 5650.6, "total_tokens": 918495704}
+{"current_steps": 21010, "total_steps": 29704, "loss": 0.2878579139709473, "lr": 1.9693550773233634e-06, "epoch": 1.4146242930245085, "percentage": 70.73, "elapsed_time": "1 day, 21:09:44", "remaining_time": "18:41:17", "throughput": 5650.74, "total_tokens": 918721536}
+{"current_steps": 21015, "total_steps": 29704, "loss": 0.27617802619934084, "lr": 1.967252487164663e-06, "epoch": 1.4149609480204686, "percentage": 70.75, "elapsed_time": "1 day, 21:10:25", "remaining_time": "18:40:40", "throughput": 5650.7, "total_tokens": 918949576}
+{"current_steps": 21020, "total_steps": 29704, "loss": 0.31108310222625735, "lr": 1.9651507451041344e-06, "epoch": 1.4152976030164288, "percentage": 70.76, "elapsed_time": "1 day, 21:11:04", "remaining_time": "18:40:01", "throughput": 5650.66, "total_tokens": 919161544}
+{"current_steps": 21025, "total_steps": 29704, "loss": 0.3056536436080933, "lr": 1.9630498517295204e-06, "epoch": 1.4156342580123888, "percentage": 70.78, "elapsed_time": "1 day, 21:11:37", "remaining_time": "18:39:20", "throughput": 5650.94, "total_tokens": 919390784}
+{"current_steps": 21030, "total_steps": 29704, "loss": 0.31153440475463867, "lr": 1.960949807628335e-06, "epoch": 1.415970913008349, "percentage": 70.8, "elapsed_time": "1 day, 21:12:11", "remaining_time": "18:38:40", "throughput": 5651.08, "total_tokens": 919610680}
+{"current_steps": 21035, "total_steps": 29704, "loss": 0.33806021213531495, "lr": 1.9588506133878415e-06, "epoch": 1.4163075680043091, "percentage": 70.82, "elapsed_time": "1 day, 21:12:48", "remaining_time": "18:38:00", "throughput": 5651.17, "total_tokens": 919830696}
+{"current_steps": 21040, "total_steps": 29704, "loss": 0.31885838508605957, "lr": 1.956752269595078e-06, "epoch": 1.4166442230002694, "percentage": 70.83, "elapsed_time": "1 day, 21:13:31", "remaining_time": "18:37:23", "throughput": 5651.06, "total_tokens": 920055416}
+{"current_steps": 21045, "total_steps": 29704, "loss": 0.3016183853149414, "lr": 1.9546547768368384e-06, "epoch": 1.4169808779962294, "percentage": 70.85, "elapsed_time": "1 day, 21:14:06", "remaining_time": "18:36:43", "throughput": 5651.26, "total_tokens": 920286312}
+{"current_steps": 21050, "total_steps": 29704, "loss": 0.28722732067108153, "lr": 1.9525581356996793e-06, "epoch": 1.4173175329921897, "percentage": 70.87, "elapsed_time": "1 day, 21:14:45", "remaining_time": "18:36:04", "throughput": 5651.21, "total_tokens": 920500576}
+{"current_steps": 21055, "total_steps": 29704, "loss": 0.30141215324401854, "lr": 1.95046234676992e-06, "epoch": 1.4176541879881497, "percentage": 70.88, "elapsed_time": "1 day, 21:15:31", "remaining_time": "18:35:29", "throughput": 5651.08, "total_tokens": 920738648}
+{"current_steps": 21060, "total_steps": 29704, "loss": 0.2801323413848877, "lr": 1.9483674106336416e-06, "epoch": 1.41799084298411, "percentage": 70.9, "elapsed_time": "1 day, 21:16:13", "remaining_time": "18:34:51", "throughput": 5650.93, "total_tokens": 920949800}
+{"current_steps": 21065, "total_steps": 29704, "loss": 0.2972491025924683, "lr": 1.9462733278766853e-06, "epoch": 1.41832749798007, "percentage": 70.92, "elapsed_time": "1 day, 21:16:53", "remaining_time": "18:34:13", "throughput": 5650.82, "total_tokens": 921159152}
+{"current_steps": 21070, "total_steps": 29704, "loss": 0.29517397880554197, "lr": 1.9441800990846553e-06, "epoch": 1.4186641529760302, "percentage": 70.93, "elapsed_time": "1 day, 21:17:24", "remaining_time": "18:33:32", "throughput": 5651.15, "total_tokens": 921390072}
+{"current_steps": 21075, "total_steps": 29704, "loss": 0.3248338460922241, "lr": 1.942087724842915e-06, "epoch": 1.4190008079719902, "percentage": 70.95, "elapsed_time": "1 day, 21:17:59", "remaining_time": "18:32:51", "throughput": 5651.21, "total_tokens": 921598928}
+{"current_steps": 21080, "total_steps": 29704, "loss": 0.30252177715301515, "lr": 1.9399962057365944e-06, "epoch": 1.4193374629679505, "percentage": 70.97, "elapsed_time": "1 day, 21:18:34", "remaining_time": "18:32:11", "throughput": 5651.26, "total_tokens": 921803544}
+{"current_steps": 21085, "total_steps": 29704, "loss": 0.32372136116027833, "lr": 1.937905542350574e-06, "epoch": 1.4196741179639105, "percentage": 70.98, "elapsed_time": "1 day, 21:19:16", "remaining_time": "18:31:34", "throughput": 5651.19, "total_tokens": 922027736}
+{"current_steps": 21090, "total_steps": 29704, "loss": 0.30223875045776366, "lr": 1.935815735269506e-06, "epoch": 1.4200107729598708, "percentage": 71.0, "elapsed_time": "1 day, 21:19:50", "remaining_time": "18:30:53", "throughput": 5651.37, "total_tokens": 922247592}
+{"current_steps": 21095, "total_steps": 29704, "loss": 0.3228907108306885, "lr": 1.933726785077797e-06, "epoch": 1.4203474279558308, "percentage": 71.02, "elapsed_time": "1 day, 21:20:30", "remaining_time": "18:30:15", "throughput": 5651.28, "total_tokens": 922462256}
+{"current_steps": 21100, "total_steps": 29704, "loss": 0.3193122386932373, "lr": 1.9316386923596146e-06, "epoch": 1.420684082951791, "percentage": 71.03, "elapsed_time": "1 day, 21:21:04", "remaining_time": "18:29:34", "throughput": 5651.45, "total_tokens": 922683024}
+{"current_steps": 21105, "total_steps": 29704, "loss": 0.31147422790527346, "lr": 1.929551457698888e-06, "epoch": 1.421020737947751, "percentage": 71.05, "elapsed_time": "1 day, 21:21:34", "remaining_time": "18:28:52", "throughput": 5651.79, "total_tokens": 922907808}
+{"current_steps": 21110, "total_steps": 29704, "loss": 0.3035153388977051, "lr": 1.9274650816793044e-06, "epoch": 1.4213573929437113, "percentage": 71.07, "elapsed_time": "1 day, 21:22:07", "remaining_time": "18:28:11", "throughput": 5652.07, "total_tokens": 923140672}
+{"current_steps": 21115, "total_steps": 29704, "loss": 0.3144373893737793, "lr": 1.9253795648843136e-06, "epoch": 1.4216940479396714, "percentage": 71.08, "elapsed_time": "1 day, 21:22:42", "remaining_time": "18:27:31", "throughput": 5652.26, "total_tokens": 923369128}
+{"current_steps": 21120, "total_steps": 29704, "loss": 0.3086548805236816, "lr": 1.9232949078971236e-06, "epoch": 1.4220307029356316, "percentage": 71.1, "elapsed_time": "1 day, 21:23:21", "remaining_time": "18:26:52", "throughput": 5652.25, "total_tokens": 923585136}
+{"current_steps": 21125, "total_steps": 29704, "loss": 0.28868756294250486, "lr": 1.9212111113006998e-06, "epoch": 1.4223673579315916, "percentage": 71.12, "elapsed_time": "1 day, 21:24:04", "remaining_time": "18:26:15", "throughput": 5652.19, "total_tokens": 923819592}
+{"current_steps": 21130, "total_steps": 29704, "loss": 0.2978044986724854, "lr": 1.919128175677775e-06, "epoch": 1.4227040129275519, "percentage": 71.14, "elapsed_time": "1 day, 21:24:42", "remaining_time": "18:25:36", "throughput": 5652.13, "total_tokens": 924024304}
+{"current_steps": 21135, "total_steps": 29704, "loss": 0.30063571929931643, "lr": 1.917046101610827e-06, "epoch": 1.423040667923512, "percentage": 71.15, "elapsed_time": "1 day, 21:25:17", "remaining_time": "18:24:56", "throughput": 5652.31, "total_tokens": 924252928}
+{"current_steps": 21140, "total_steps": 29704, "loss": 0.3125055551528931, "lr": 1.9149648896821087e-06, "epoch": 1.4233773229194722, "percentage": 71.17, "elapsed_time": "1 day, 21:26:02", "remaining_time": "18:24:20", "throughput": 5652.09, "total_tokens": 924467024}
+{"current_steps": 21145, "total_steps": 29704, "loss": 0.2941263675689697, "lr": 1.912884540473621e-06, "epoch": 1.4237139779154322, "percentage": 71.19, "elapsed_time": "1 day, 21:26:40", "remaining_time": "18:23:41", "throughput": 5652.0, "total_tokens": 924669616}
+{"current_steps": 21150, "total_steps": 29704, "loss": 0.27542452812194823, "lr": 1.9108050545671283e-06, "epoch": 1.4240506329113924, "percentage": 71.2, "elapsed_time": "1 day, 21:27:15", "remaining_time": "18:23:01", "throughput": 5652.05, "total_tokens": 924875728}
+{"current_steps": 21155, "total_steps": 29704, "loss": 0.27084195613861084, "lr": 1.9087264325441507e-06, "epoch": 1.4243872879073525, "percentage": 71.22, "elapsed_time": "1 day, 21:27:56", "remaining_time": "18:22:23", "throughput": 5651.9, "total_tokens": 925083640}
+{"current_steps": 21160, "total_steps": 29704, "loss": 0.3030634880065918, "lr": 1.9066486749859686e-06, "epoch": 1.4247239429033127, "percentage": 71.24, "elapsed_time": "1 day, 21:28:30", "remaining_time": "18:21:43", "throughput": 5652.02, "total_tokens": 925295656}
+{"current_steps": 21165, "total_steps": 29704, "loss": 0.3159012794494629, "lr": 1.9045717824736209e-06, "epoch": 1.4250605978992728, "percentage": 71.25, "elapsed_time": "1 day, 21:29:08", "remaining_time": "18:21:04", "throughput": 5652.09, "total_tokens": 925523192}
+{"current_steps": 21170, "total_steps": 29704, "loss": 0.30693175792694094, "lr": 1.9024957555879037e-06, "epoch": 1.425397252895233, "percentage": 71.27, "elapsed_time": "1 day, 21:29:45", "remaining_time": "18:20:24", "throughput": 5652.1, "total_tokens": 925733280}
+{"current_steps": 21175, "total_steps": 29704, "loss": 0.31174449920654296, "lr": 1.9004205949093695e-06, "epoch": 1.425733907891193, "percentage": 71.29, "elapsed_time": "1 day, 21:30:19", "remaining_time": "18:19:44", "throughput": 5652.26, "total_tokens": 925952408}
+{"current_steps": 21180, "total_steps": 29704, "loss": 0.3182375907897949, "lr": 1.8983463010183355e-06, "epoch": 1.4260705628871533, "percentage": 71.3, "elapsed_time": "1 day, 21:30:52", "remaining_time": "18:19:03", "throughput": 5652.43, "total_tokens": 926166584}
+{"current_steps": 21185, "total_steps": 29704, "loss": 0.30806000232696534, "lr": 1.8962728744948644e-06, "epoch": 1.4264072178831133, "percentage": 71.32, "elapsed_time": "1 day, 21:31:32", "remaining_time": "18:18:25", "throughput": 5652.46, "total_tokens": 926394392}
+{"current_steps": 21190, "total_steps": 29704, "loss": 0.3054067850112915, "lr": 1.894200315918789e-06, "epoch": 1.4267438728790736, "percentage": 71.34, "elapsed_time": "1 day, 21:32:06", "remaining_time": "18:17:44", "throughput": 5652.73, "total_tokens": 926630992}
+{"current_steps": 21195, "total_steps": 29704, "loss": 0.2958783388137817, "lr": 1.8921286258696913e-06, "epoch": 1.4270805278750336, "percentage": 71.35, "elapsed_time": "1 day, 21:32:45", "remaining_time": "18:17:05", "throughput": 5652.63, "total_tokens": 926834656}
+{"current_steps": 21200, "total_steps": 29704, "loss": 0.2869250297546387, "lr": 1.890057804926913e-06, "epoch": 1.4274171828709938, "percentage": 71.37, "elapsed_time": "1 day, 21:33:23", "remaining_time": "18:16:26", "throughput": 5652.71, "total_tokens": 927061800}
+{"current_steps": 21205, "total_steps": 29704, "loss": 0.2620058536529541, "lr": 1.887987853669553e-06, "epoch": 1.4277538378669539, "percentage": 71.39, "elapsed_time": "1 day, 21:34:06", "remaining_time": "18:15:50", "throughput": 5652.58, "total_tokens": 927286432}
+{"current_steps": 21210, "total_steps": 29704, "loss": 0.3177152633666992, "lr": 1.8859187726764654e-06, "epoch": 1.4280904928629141, "percentage": 71.4, "elapsed_time": "1 day, 21:34:44", "remaining_time": "18:15:11", "throughput": 5652.64, "total_tokens": 927512824}
+{"current_steps": 21215, "total_steps": 29704, "loss": 0.29890553951263427, "lr": 1.8838505625262638e-06, "epoch": 1.4284271478588741, "percentage": 71.42, "elapsed_time": "1 day, 21:35:25", "remaining_time": "18:14:33", "throughput": 5652.65, "total_tokens": 927741616}
+{"current_steps": 21220, "total_steps": 29704, "loss": 0.2987877368927002, "lr": 1.8817832237973154e-06, "epoch": 1.4287638028548344, "percentage": 71.44, "elapsed_time": "1 day, 21:36:04", "remaining_time": "18:13:54", "throughput": 5652.63, "total_tokens": 927959432}
+{"current_steps": 21225, "total_steps": 29704, "loss": 0.28604328632354736, "lr": 1.8797167570677439e-06, "epoch": 1.4291004578507944, "percentage": 71.46, "elapsed_time": "1 day, 21:36:53", "remaining_time": "18:13:20", "throughput": 5652.29, "total_tokens": 928181056}
+{"current_steps": 21230, "total_steps": 29704, "loss": 0.3128670215606689, "lr": 1.8776511629154342e-06, "epoch": 1.4294371128467547, "percentage": 71.47, "elapsed_time": "1 day, 21:37:37", "remaining_time": "18:12:43", "throughput": 5652.19, "total_tokens": 928412792}
+{"current_steps": 21235, "total_steps": 29704, "loss": 0.30298452377319335, "lr": 1.8755864419180176e-06, "epoch": 1.4297737678427147, "percentage": 71.49, "elapsed_time": "1 day, 21:38:24", "remaining_time": "18:12:08", "throughput": 5651.94, "total_tokens": 928637368}
+{"current_steps": 21240, "total_steps": 29704, "loss": 0.3099356174468994, "lr": 1.8735225946528906e-06, "epoch": 1.430110422838675, "percentage": 71.51, "elapsed_time": "1 day, 21:39:00", "remaining_time": "18:11:28", "throughput": 5652.08, "total_tokens": 928865136}
+{"current_steps": 21245, "total_steps": 29704, "loss": 0.3137706518173218, "lr": 1.8714596216972008e-06, "epoch": 1.430447077834635, "percentage": 71.52, "elapsed_time": "1 day, 21:39:31", "remaining_time": "18:10:46", "throughput": 5652.32, "total_tokens": 929082920}
+{"current_steps": 21250, "total_steps": 29704, "loss": 0.29973363876342773, "lr": 1.8693975236278512e-06, "epoch": 1.4307837328305952, "percentage": 71.54, "elapsed_time": "1 day, 21:40:13", "remaining_time": "18:10:09", "throughput": 5652.18, "total_tokens": 929295744}
+{"current_steps": 21255, "total_steps": 29704, "loss": 0.32357001304626465, "lr": 1.8673363010215017e-06, "epoch": 1.4311203878265553, "percentage": 71.56, "elapsed_time": "1 day, 21:40:54", "remaining_time": "18:09:31", "throughput": 5652.08, "total_tokens": 929509608}
+{"current_steps": 21260, "total_steps": 29704, "loss": 0.3077038288116455, "lr": 1.8652759544545661e-06, "epoch": 1.4314570428225155, "percentage": 71.57, "elapsed_time": "1 day, 21:41:31", "remaining_time": "18:08:52", "throughput": 5652.16, "total_tokens": 929730176}
+{"current_steps": 21265, "total_steps": 29704, "loss": 0.31038436889648435, "lr": 1.863216484503214e-06, "epoch": 1.4317936978184758, "percentage": 71.59, "elapsed_time": "1 day, 21:42:14", "remaining_time": "18:08:15", "throughput": 5652.0, "total_tokens": 929946784}
+{"current_steps": 21270, "total_steps": 29704, "loss": 0.3216505527496338, "lr": 1.8611578917433698e-06, "epoch": 1.4321303528144358, "percentage": 71.61, "elapsed_time": "1 day, 21:42:47", "remaining_time": "18:07:34", "throughput": 5652.23, "total_tokens": 930172032}
+{"current_steps": 21275, "total_steps": 29704, "loss": 0.32173566818237304, "lr": 1.85910017675071e-06, "epoch": 1.4324670078103958, "percentage": 71.62, "elapsed_time": "1 day, 21:43:29", "remaining_time": "18:06:56", "throughput": 5652.11, "total_tokens": 930388552}
+{"current_steps": 21280, "total_steps": 29704, "loss": 0.2714304208755493, "lr": 1.8570433401006737e-06, "epoch": 1.432803662806356, "percentage": 71.64, "elapsed_time": "1 day, 21:44:03", "remaining_time": "18:06:16", "throughput": 5652.29, "total_tokens": 930614664}
+{"current_steps": 21285, "total_steps": 29704, "loss": 0.2765933990478516, "lr": 1.8549873823684418e-06, "epoch": 1.4331403178023163, "percentage": 71.66, "elapsed_time": "1 day, 21:44:44", "remaining_time": "18:05:38", "throughput": 5652.18, "total_tokens": 930827792}
+{"current_steps": 21290, "total_steps": 29704, "loss": 0.32932586669921876, "lr": 1.8529323041289616e-06, "epoch": 1.4334769727982763, "percentage": 71.67, "elapsed_time": "1 day, 21:45:19", "remaining_time": "18:04:58", "throughput": 5652.34, "total_tokens": 931052808}
+{"current_steps": 21295, "total_steps": 29704, "loss": 0.30021870136260986, "lr": 1.850878105956927e-06, "epoch": 1.4338136277942364, "percentage": 71.69, "elapsed_time": "1 day, 21:45:48", "remaining_time": "18:04:16", "throughput": 5652.73, "total_tokens": 931280240}
+{"current_steps": 21300, "total_steps": 29704, "loss": 0.29010815620422364, "lr": 1.8488247884267885e-06, "epoch": 1.4341502827901966, "percentage": 71.71, "elapsed_time": "1 day, 21:46:26", "remaining_time": "18:03:36", "throughput": 5652.85, "total_tokens": 931510352}
+{"current_steps": 21305, "total_steps": 29704, "loss": 0.31166844367980956, "lr": 1.8467723521127495e-06, "epoch": 1.4344869377861569, "percentage": 71.72, "elapsed_time": "1 day, 21:47:08", "remaining_time": "18:02:59", "throughput": 5652.64, "total_tokens": 931718464}
+{"current_steps": 21310, "total_steps": 29704, "loss": 0.28693480491638185, "lr": 1.8447207975887671e-06, "epoch": 1.434823592782117, "percentage": 71.74, "elapsed_time": "1 day, 21:47:51", "remaining_time": "18:02:22", "throughput": 5652.49, "total_tokens": 931935064}
+{"current_steps": 21315, "total_steps": 29704, "loss": 0.3131574630737305, "lr": 1.8426701254285522e-06, "epoch": 1.435160247778077, "percentage": 71.76, "elapsed_time": "1 day, 21:48:39", "remaining_time": "18:01:47", "throughput": 5652.24, "total_tokens": 932161648}
+{"current_steps": 21320, "total_steps": 29704, "loss": 0.3200539588928223, "lr": 1.840620336205569e-06, "epoch": 1.4354969027740372, "percentage": 71.77, "elapsed_time": "1 day, 21:49:22", "remaining_time": "18:01:10", "throughput": 5652.12, "total_tokens": 932386520}
+{"current_steps": 21325, "total_steps": 29704, "loss": 0.2745990753173828, "lr": 1.8385714304930324e-06, "epoch": 1.4358335577699974, "percentage": 71.79, "elapsed_time": "1 day, 21:49:59", "remaining_time": "18:00:31", "throughput": 5652.23, "total_tokens": 932614272}
+{"current_steps": 21330, "total_steps": 29704, "loss": 0.2942768096923828, "lr": 1.8365234088639178e-06, "epoch": 1.4361702127659575, "percentage": 71.81, "elapsed_time": "1 day, 21:50:36", "remaining_time": "17:59:52", "throughput": 5652.28, "total_tokens": 932835368}
+{"current_steps": 21335, "total_steps": 29704, "loss": 0.2779007196426392, "lr": 1.834476271890941e-06, "epoch": 1.4365068677619175, "percentage": 71.83, "elapsed_time": "1 day, 21:51:15", "remaining_time": "17:59:13", "throughput": 5652.16, "total_tokens": 933035248}
+{"current_steps": 21340, "total_steps": 29704, "loss": 0.28595268726348877, "lr": 1.8324300201465823e-06, "epoch": 1.4368435227578777, "percentage": 71.84, "elapsed_time": "1 day, 21:51:50", "remaining_time": "17:58:33", "throughput": 5652.32, "total_tokens": 933258792}
+{"current_steps": 21345, "total_steps": 29704, "loss": 0.3224499225616455, "lr": 1.8303846542030674e-06, "epoch": 1.437180177753838, "percentage": 71.86, "elapsed_time": "1 day, 21:52:29", "remaining_time": "17:57:54", "throughput": 5652.28, "total_tokens": 933469416}
+{"current_steps": 21350, "total_steps": 29704, "loss": 0.28315880298614504, "lr": 1.828340174632377e-06, "epoch": 1.437516832749798, "percentage": 71.88, "elapsed_time": "1 day, 21:53:02", "remaining_time": "17:57:13", "throughput": 5652.5, "total_tokens": 933691024}
+{"current_steps": 21355, "total_steps": 29704, "loss": 0.29636235237121583, "lr": 1.8262965820062434e-06, "epoch": 1.437853487745758, "percentage": 71.89, "elapsed_time": "1 day, 21:53:36", "remaining_time": "17:56:33", "throughput": 5652.66, "total_tokens": 933911248}
+{"current_steps": 21360, "total_steps": 29704, "loss": 0.2795313835144043, "lr": 1.8242538768961499e-06, "epoch": 1.4381901427417183, "percentage": 71.91, "elapsed_time": "1 day, 21:54:09", "remaining_time": "17:55:52", "throughput": 5652.85, "total_tokens": 934130456}
+{"current_steps": 21365, "total_steps": 29704, "loss": 0.28819353580474855, "lr": 1.8222120598733317e-06, "epoch": 1.4385267977376786, "percentage": 71.93, "elapsed_time": "1 day, 21:54:49", "remaining_time": "17:55:14", "throughput": 5652.85, "total_tokens": 934359280}
+{"current_steps": 21370, "total_steps": 29704, "loss": 0.32363121509552, "lr": 1.820171131508781e-06, "epoch": 1.4388634527336386, "percentage": 71.94, "elapsed_time": "1 day, 21:55:24", "remaining_time": "17:54:34", "throughput": 5652.93, "total_tokens": 934570728}
+{"current_steps": 21375, "total_steps": 29704, "loss": 0.2775664091110229, "lr": 1.8181310923732304e-06, "epoch": 1.4392001077295986, "percentage": 71.96, "elapsed_time": "1 day, 21:56:07", "remaining_time": "17:53:57", "throughput": 5652.8, "total_tokens": 934787104}
+{"current_steps": 21380, "total_steps": 29704, "loss": 0.3209282875061035, "lr": 1.8160919430371764e-06, "epoch": 1.4395367627255589, "percentage": 71.98, "elapsed_time": "1 day, 21:56:47", "remaining_time": "17:53:19", "throughput": 5652.66, "total_tokens": 934994512}
+{"current_steps": 21385, "total_steps": 29704, "loss": 0.29017982482910154, "lr": 1.814053684070855e-06, "epoch": 1.439873417721519, "percentage": 71.99, "elapsed_time": "1 day, 21:57:24", "remaining_time": "17:52:39", "throughput": 5652.77, "total_tokens": 935221480}
+{"current_steps": 21390, "total_steps": 29704, "loss": 0.28193049430847167, "lr": 1.8120163160442633e-06, "epoch": 1.4402100727174791, "percentage": 72.01, "elapsed_time": "1 day, 21:58:01", "remaining_time": "17:52:00", "throughput": 5652.9, "total_tokens": 935448880}
+{"current_steps": 21395, "total_steps": 29704, "loss": 0.31296658515930176, "lr": 1.8099798395271428e-06, "epoch": 1.4405467277134392, "percentage": 72.03, "elapsed_time": "1 day, 21:58:39", "remaining_time": "17:51:21", "throughput": 5653.03, "total_tokens": 935684528}
+{"current_steps": 21400, "total_steps": 29704, "loss": 0.28637051582336426, "lr": 1.807944255088988e-06, "epoch": 1.4408833827093994, "percentage": 72.04, "elapsed_time": "1 day, 21:59:15", "remaining_time": "17:50:41", "throughput": 5653.07, "total_tokens": 935896728}
+{"current_steps": 21405, "total_steps": 29704, "loss": 0.29571921825408937, "lr": 1.8059095632990437e-06, "epoch": 1.4412200377053597, "percentage": 72.06, "elapsed_time": "1 day, 21:59:47", "remaining_time": "17:50:00", "throughput": 5653.28, "total_tokens": 936114728}
+{"current_steps": 21410, "total_steps": 29704, "loss": 0.312378978729248, "lr": 1.8038757647263045e-06, "epoch": 1.4415566927013197, "percentage": 72.08, "elapsed_time": "1 day, 22:00:25", "remaining_time": "17:49:21", "throughput": 5653.32, "total_tokens": 936331896}
+{"current_steps": 21415, "total_steps": 29704, "loss": 0.28726487159729003, "lr": 1.8018428599395143e-06, "epoch": 1.4418933476972797, "percentage": 72.09, "elapsed_time": "1 day, 22:01:07", "remaining_time": "17:48:44", "throughput": 5653.27, "total_tokens": 936563512}
+{"current_steps": 21420, "total_steps": 29704, "loss": 0.2904759645462036, "lr": 1.7998108495071743e-06, "epoch": 1.44223000269324, "percentage": 72.11, "elapsed_time": "1 day, 22:01:49", "remaining_time": "17:48:06", "throughput": 5653.18, "total_tokens": 936786872}
+{"current_steps": 21425, "total_steps": 29704, "loss": 0.2962934970855713, "lr": 1.797779733997521e-06, "epoch": 1.4425666576892002, "percentage": 72.13, "elapsed_time": "1 day, 22:02:29", "remaining_time": "17:47:28", "throughput": 5653.14, "total_tokens": 937007176}
+{"current_steps": 21430, "total_steps": 29704, "loss": 0.3217646837234497, "lr": 1.7957495139785585e-06, "epoch": 1.4429033126851603, "percentage": 72.15, "elapsed_time": "1 day, 22:03:08", "remaining_time": "17:46:50", "throughput": 5653.15, "total_tokens": 937228952}
+{"current_steps": 21435, "total_steps": 29704, "loss": 0.3023351192474365, "lr": 1.7937201900180228e-06, "epoch": 1.4432399676811203, "percentage": 72.16, "elapsed_time": "1 day, 22:03:49", "remaining_time": "17:46:12", "throughput": 5652.99, "total_tokens": 937432152}
+{"current_steps": 21440, "total_steps": 29704, "loss": 0.339884877204895, "lr": 1.7916917626834134e-06, "epoch": 1.4435766226770805, "percentage": 72.18, "elapsed_time": "1 day, 22:04:25", "remaining_time": "17:45:32", "throughput": 5653.08, "total_tokens": 937653160}
+{"current_steps": 21445, "total_steps": 29704, "loss": 0.31574015617370604, "lr": 1.7896642325419722e-06, "epoch": 1.4439132776730408, "percentage": 72.2, "elapsed_time": "1 day, 22:05:02", "remaining_time": "17:44:53", "throughput": 5653.13, "total_tokens": 937871808}
+{"current_steps": 21450, "total_steps": 29704, "loss": 0.30422613620758054, "lr": 1.7876376001606915e-06, "epoch": 1.4442499326690008, "percentage": 72.21, "elapsed_time": "1 day, 22:05:37", "remaining_time": "17:44:13", "throughput": 5653.19, "total_tokens": 938077432}
+{"current_steps": 21455, "total_steps": 29704, "loss": 0.2780045747756958, "lr": 1.7856118661063128e-06, "epoch": 1.4445865876649608, "percentage": 72.23, "elapsed_time": "1 day, 22:06:19", "remaining_time": "17:43:35", "throughput": 5653.16, "total_tokens": 938306192}
+{"current_steps": 21460, "total_steps": 29704, "loss": 0.3040392637252808, "lr": 1.7835870309453251e-06, "epoch": 1.444923242660921, "percentage": 72.25, "elapsed_time": "1 day, 22:06:56", "remaining_time": "17:42:56", "throughput": 5653.18, "total_tokens": 938518824}
+{"current_steps": 21465, "total_steps": 29704, "loss": 0.29809937477111814, "lr": 1.7815630952439667e-06, "epoch": 1.4452598976568813, "percentage": 72.26, "elapsed_time": "1 day, 22:07:35", "remaining_time": "17:42:17", "throughput": 5653.11, "total_tokens": 938731144}
+{"current_steps": 21470, "total_steps": 29704, "loss": 0.2990201473236084, "lr": 1.7795400595682289e-06, "epoch": 1.4455965526528414, "percentage": 72.28, "elapsed_time": "1 day, 22:08:13", "remaining_time": "17:41:38", "throughput": 5653.09, "total_tokens": 938939520}
+{"current_steps": 21475, "total_steps": 29704, "loss": 0.342014741897583, "lr": 1.7775179244838408e-06, "epoch": 1.4459332076488014, "percentage": 72.3, "elapsed_time": "1 day, 22:08:48", "remaining_time": "17:40:58", "throughput": 5653.15, "total_tokens": 939148712}
+{"current_steps": 21480, "total_steps": 29704, "loss": 0.29918684959411623, "lr": 1.7754966905562925e-06, "epoch": 1.4462698626447617, "percentage": 72.31, "elapsed_time": "1 day, 22:09:22", "remaining_time": "17:40:18", "throughput": 5653.22, "total_tokens": 939352016}
+{"current_steps": 21485, "total_steps": 29704, "loss": 0.29963269233703616, "lr": 1.7734763583508098e-06, "epoch": 1.446606517640722, "percentage": 72.33, "elapsed_time": "1 day, 22:10:05", "remaining_time": "17:39:41", "throughput": 5653.07, "total_tokens": 939569112}
+{"current_steps": 21490, "total_steps": 29704, "loss": 0.31328535079956055, "lr": 1.7714569284323757e-06, "epoch": 1.446943172636682, "percentage": 72.35, "elapsed_time": "1 day, 22:10:43", "remaining_time": "17:39:02", "throughput": 5653.14, "total_tokens": 939800608}
+{"current_steps": 21495, "total_steps": 29704, "loss": 0.29224376678466796, "lr": 1.7694384013657173e-06, "epoch": 1.447279827632642, "percentage": 72.36, "elapsed_time": "1 day, 22:11:26", "remaining_time": "17:38:25", "throughput": 5652.99, "total_tokens": 940016152}
+{"current_steps": 21500, "total_steps": 29704, "loss": 0.31572024822235106, "lr": 1.767420777715308e-06, "epoch": 1.4476164826286022, "percentage": 72.38, "elapsed_time": "1 day, 22:12:04", "remaining_time": "17:37:46", "throughput": 5653.1, "total_tokens": 940248920}
+{"current_steps": 21505, "total_steps": 29704, "loss": 0.2916224241256714, "lr": 1.7654040580453702e-06, "epoch": 1.4479531376245625, "percentage": 72.4, "elapsed_time": "1 day, 22:12:39", "remaining_time": "17:37:06", "throughput": 5653.28, "total_tokens": 940476808}
+{"current_steps": 21510, "total_steps": 29704, "loss": 0.3200731039047241, "lr": 1.7633882429198734e-06, "epoch": 1.4482897926205225, "percentage": 72.41, "elapsed_time": "1 day, 22:13:20", "remaining_time": "17:36:28", "throughput": 5653.26, "total_tokens": 940704272}
+{"current_steps": 21515, "total_steps": 29704, "loss": 0.29833736419677737, "lr": 1.7613733329025318e-06, "epoch": 1.4486264476164825, "percentage": 72.43, "elapsed_time": "1 day, 22:14:00", "remaining_time": "17:35:50", "throughput": 5653.3, "total_tokens": 940937904}
+{"current_steps": 21520, "total_steps": 29704, "loss": 0.3069751739501953, "lr": 1.759359328556814e-06, "epoch": 1.4489631026124428, "percentage": 72.45, "elapsed_time": "1 day, 22:14:42", "remaining_time": "17:35:12", "throughput": 5653.22, "total_tokens": 941164048}
+{"current_steps": 21525, "total_steps": 29704, "loss": 0.3017618179321289, "lr": 1.7573462304459227e-06, "epoch": 1.449299757608403, "percentage": 72.46, "elapsed_time": "1 day, 22:15:15", "remaining_time": "17:34:32", "throughput": 5653.44, "total_tokens": 941384664}
+{"current_steps": 21530, "total_steps": 29704, "loss": 0.28963470458984375, "lr": 1.7553340391328216e-06, "epoch": 1.449636412604363, "percentage": 72.48, "elapsed_time": "1 day, 22:15:53", "remaining_time": "17:33:53", "throughput": 5653.46, "total_tokens": 941603416}
+{"current_steps": 21535, "total_steps": 29704, "loss": 0.3050259590148926, "lr": 1.7533227551802068e-06, "epoch": 1.449973067600323, "percentage": 72.5, "elapsed_time": "1 day, 22:16:30", "remaining_time": "17:33:13", "throughput": 5653.58, "total_tokens": 941831232}
+{"current_steps": 21540, "total_steps": 29704, "loss": 0.32262425422668456, "lr": 1.7513123791505321e-06, "epoch": 1.4503097225962833, "percentage": 72.52, "elapsed_time": "1 day, 22:17:15", "remaining_time": "17:32:37", "throughput": 5653.31, "total_tokens": 942039264}
+{"current_steps": 21545, "total_steps": 29704, "loss": 0.3137768268585205, "lr": 1.7493029116059913e-06, "epoch": 1.4506463775922436, "percentage": 72.53, "elapsed_time": "1 day, 22:17:55", "remaining_time": "17:31:59", "throughput": 5653.22, "total_tokens": 942254168}
+{"current_steps": 21550, "total_steps": 29704, "loss": 0.29879701137542725, "lr": 1.7472943531085252e-06, "epoch": 1.4509830325882036, "percentage": 72.55, "elapsed_time": "1 day, 22:18:34", "remaining_time": "17:31:20", "throughput": 5653.26, "total_tokens": 942479744}
+{"current_steps": 21555, "total_steps": 29704, "loss": 0.28659663200378416, "lr": 1.7452867042198212e-06, "epoch": 1.4513196875841636, "percentage": 72.57, "elapsed_time": "1 day, 22:19:11", "remaining_time": "17:30:41", "throughput": 5653.26, "total_tokens": 942691224}
+{"current_steps": 21560, "total_steps": 29704, "loss": 0.2767528533935547, "lr": 1.7432799655013122e-06, "epoch": 1.4516563425801239, "percentage": 72.58, "elapsed_time": "1 day, 22:19:50", "remaining_time": "17:30:02", "throughput": 5653.29, "total_tokens": 942915144}
+{"current_steps": 21565, "total_steps": 29704, "loss": 0.31976196765899656, "lr": 1.741274137514174e-06, "epoch": 1.4519929975760841, "percentage": 72.6, "elapsed_time": "1 day, 22:20:27", "remaining_time": "17:29:23", "throughput": 5653.28, "total_tokens": 943124816}
+{"current_steps": 21570, "total_steps": 29704, "loss": 0.28882317543029784, "lr": 1.7392692208193351e-06, "epoch": 1.4523296525720442, "percentage": 72.62, "elapsed_time": "1 day, 22:21:14", "remaining_time": "17:28:48", "throughput": 5653.0, "total_tokens": 943341696}
+{"current_steps": 21575, "total_steps": 29704, "loss": 0.27709264755249025, "lr": 1.7372652159774578e-06, "epoch": 1.4526663075680042, "percentage": 72.63, "elapsed_time": "1 day, 22:22:01", "remaining_time": "17:28:12", "throughput": 5652.77, "total_tokens": 943566704}
+{"current_steps": 21580, "total_steps": 29704, "loss": 0.2953044414520264, "lr": 1.7352621235489625e-06, "epoch": 1.4530029625639644, "percentage": 72.65, "elapsed_time": "1 day, 22:22:43", "remaining_time": "17:27:34", "throughput": 5652.64, "total_tokens": 943783040}
+{"current_steps": 21585, "total_steps": 29704, "loss": 0.3123426914215088, "lr": 1.733259944094e-06, "epoch": 1.4533396175599247, "percentage": 72.67, "elapsed_time": "1 day, 22:23:24", "remaining_time": "17:26:57", "throughput": 5652.56, "total_tokens": 944002144}
+{"current_steps": 21590, "total_steps": 29704, "loss": 0.2972878456115723, "lr": 1.7312586781724789e-06, "epoch": 1.4536762725558847, "percentage": 72.68, "elapsed_time": "1 day, 22:23:55", "remaining_time": "17:26:15", "throughput": 5652.83, "total_tokens": 944220536}
+{"current_steps": 21595, "total_steps": 29704, "loss": 0.31235198974609374, "lr": 1.7292583263440454e-06, "epoch": 1.4540129275518447, "percentage": 72.7, "elapsed_time": "1 day, 22:24:34", "remaining_time": "17:25:37", "throughput": 5652.76, "total_tokens": 944434048}
+{"current_steps": 21600, "total_steps": 29704, "loss": 0.28976993560791015, "lr": 1.7272588891680914e-06, "epoch": 1.454349582547805, "percentage": 72.72, "elapsed_time": "1 day, 22:25:14", "remaining_time": "17:24:58", "throughput": 5652.64, "total_tokens": 944637656}
+{"current_steps": 21605, "total_steps": 29704, "loss": 0.29179136753082274, "lr": 1.7252603672037527e-06, "epoch": 1.4546862375437652, "percentage": 72.73, "elapsed_time": "1 day, 22:25:51", "remaining_time": "17:24:19", "throughput": 5652.64, "total_tokens": 944849080}
+{"current_steps": 21610, "total_steps": 29704, "loss": 0.2792097568511963, "lr": 1.72326276100991e-06, "epoch": 1.4550228925397253, "percentage": 72.75, "elapsed_time": "1 day, 22:26:28", "remaining_time": "17:23:40", "throughput": 5652.78, "total_tokens": 945079208}
+{"current_steps": 21615, "total_steps": 29704, "loss": 0.3040107250213623, "lr": 1.721266071145186e-06, "epoch": 1.4553595475356853, "percentage": 72.77, "elapsed_time": "1 day, 22:27:07", "remaining_time": "17:23:01", "throughput": 5652.8, "total_tokens": 945305216}
+{"current_steps": 21620, "total_steps": 29704, "loss": 0.30785455703735354, "lr": 1.7192702981679531e-06, "epoch": 1.4556962025316456, "percentage": 72.78, "elapsed_time": "1 day, 22:27:42", "remaining_time": "17:22:21", "throughput": 5652.98, "total_tokens": 945532656}
+{"current_steps": 21625, "total_steps": 29704, "loss": 0.27479963302612304, "lr": 1.717275442636317e-06, "epoch": 1.4560328575276058, "percentage": 72.8, "elapsed_time": "1 day, 22:28:18", "remaining_time": "17:21:42", "throughput": 5653.08, "total_tokens": 945753800}
+{"current_steps": 21630, "total_steps": 29704, "loss": 0.30873837471008303, "lr": 1.7152815051081384e-06, "epoch": 1.4563695125235658, "percentage": 72.82, "elapsed_time": "1 day, 22:28:56", "remaining_time": "17:21:02", "throughput": 5653.12, "total_tokens": 945972456}
+{"current_steps": 21635, "total_steps": 29704, "loss": 0.3021131753921509, "lr": 1.7132884861410093e-06, "epoch": 1.4567061675195259, "percentage": 72.84, "elapsed_time": "1 day, 22:29:32", "remaining_time": "17:20:23", "throughput": 5653.16, "total_tokens": 946181664}
+{"current_steps": 21640, "total_steps": 29704, "loss": 0.28493921756744384, "lr": 1.7112963862922766e-06, "epoch": 1.4570428225154861, "percentage": 72.85, "elapsed_time": "1 day, 22:30:06", "remaining_time": "17:19:42", "throughput": 5653.26, "total_tokens": 946392928}
+{"current_steps": 21645, "total_steps": 29704, "loss": 0.30887959003448484, "lr": 1.7093052061190224e-06, "epoch": 1.4573794775114464, "percentage": 72.87, "elapsed_time": "1 day, 22:30:42", "remaining_time": "17:19:03", "throughput": 5653.43, "total_tokens": 946623584}
+{"current_steps": 21650, "total_steps": 29704, "loss": 0.3057863712310791, "lr": 1.7073149461780742e-06, "epoch": 1.4577161325074064, "percentage": 72.89, "elapsed_time": "1 day, 22:31:20", "remaining_time": "17:18:24", "throughput": 5653.55, "total_tokens": 946857512}
+{"current_steps": 21655, "total_steps": 29704, "loss": 0.31071517467498777, "lr": 1.705325607026001e-06, "epoch": 1.4580527875033664, "percentage": 72.9, "elapsed_time": "1 day, 22:31:55", "remaining_time": "17:17:44", "throughput": 5653.59, "total_tokens": 947062648}
+{"current_steps": 21660, "total_steps": 29704, "loss": 0.30583949089050294, "lr": 1.7033371892191164e-06, "epoch": 1.4583894424993267, "percentage": 72.92, "elapsed_time": "1 day, 22:32:31", "remaining_time": "17:17:04", "throughput": 5653.69, "total_tokens": 947284464}
+{"current_steps": 21665, "total_steps": 29704, "loss": 0.3118964433670044, "lr": 1.7013496933134726e-06, "epoch": 1.458726097495287, "percentage": 72.94, "elapsed_time": "1 day, 22:33:11", "remaining_time": "17:16:26", "throughput": 5653.59, "total_tokens": 947494920}
+{"current_steps": 21670, "total_steps": 29704, "loss": 0.34239089488983154, "lr": 1.6993631198648724e-06, "epoch": 1.459062752491247, "percentage": 72.95, "elapsed_time": "1 day, 22:33:45", "remaining_time": "17:15:46", "throughput": 5653.76, "total_tokens": 947716632}
+{"current_steps": 21675, "total_steps": 29704, "loss": 0.29693849086761476, "lr": 1.697377469428847e-06, "epoch": 1.459399407487207, "percentage": 72.97, "elapsed_time": "1 day, 22:34:22", "remaining_time": "17:15:06", "throughput": 5653.88, "total_tokens": 947941544}
+{"current_steps": 21680, "total_steps": 29704, "loss": 0.32032740116119385, "lr": 1.6953927425606842e-06, "epoch": 1.4597360624831672, "percentage": 72.99, "elapsed_time": "1 day, 22:35:06", "remaining_time": "17:14:29", "throughput": 5653.69, "total_tokens": 948159704}
+{"current_steps": 21685, "total_steps": 29704, "loss": 0.2828591585159302, "lr": 1.6934089398154002e-06, "epoch": 1.4600727174791275, "percentage": 73.0, "elapsed_time": "1 day, 22:35:45", "remaining_time": "17:13:51", "throughput": 5653.67, "total_tokens": 948375840}
+{"current_steps": 21690, "total_steps": 29704, "loss": 0.284234356880188, "lr": 1.6914260617477634e-06, "epoch": 1.4604093724750875, "percentage": 73.02, "elapsed_time": "1 day, 22:36:18", "remaining_time": "17:13:10", "throughput": 5653.91, "total_tokens": 948601288}
+{"current_steps": 21695, "total_steps": 29704, "loss": 0.2686470985412598, "lr": 1.6894441089122788e-06, "epoch": 1.4607460274710475, "percentage": 73.04, "elapsed_time": "1 day, 22:36:53", "remaining_time": "17:12:30", "throughput": 5653.98, "total_tokens": 948816664}
+{"current_steps": 21700, "total_steps": 29704, "loss": 0.31239607334136965, "lr": 1.6874630818631926e-06, "epoch": 1.4610826824670078, "percentage": 73.05, "elapsed_time": "1 day, 22:37:36", "remaining_time": "17:11:53", "throughput": 5653.87, "total_tokens": 949036440}
+{"current_steps": 21705, "total_steps": 29704, "loss": 0.31434266567230223, "lr": 1.685482981154492e-06, "epoch": 1.461419337462968, "percentage": 73.07, "elapsed_time": "1 day, 22:38:06", "remaining_time": "17:11:11", "throughput": 5654.15, "total_tokens": 949257872}
+{"current_steps": 21710, "total_steps": 29704, "loss": 0.29442682266235354, "lr": 1.683503807339907e-06, "epoch": 1.461755992458928, "percentage": 73.09, "elapsed_time": "1 day, 22:38:50", "remaining_time": "17:10:35", "throughput": 5654.03, "total_tokens": 949486984}
+{"current_steps": 21715, "total_steps": 29704, "loss": 0.3005738496780396, "lr": 1.6815255609729047e-06, "epoch": 1.462092647454888, "percentage": 73.1, "elapsed_time": "1 day, 22:39:26", "remaining_time": "17:09:55", "throughput": 5654.18, "total_tokens": 949713336}
+{"current_steps": 21720, "total_steps": 29704, "loss": 0.30016207695007324, "lr": 1.6795482426067012e-06, "epoch": 1.4624293024508483, "percentage": 73.12, "elapsed_time": "1 day, 22:40:01", "remaining_time": "17:09:15", "throughput": 5654.23, "total_tokens": 949920736}
+{"current_steps": 21725, "total_steps": 29704, "loss": 0.2928857564926147, "lr": 1.6775718527942403e-06, "epoch": 1.4627659574468086, "percentage": 73.14, "elapsed_time": "1 day, 22:40:39", "remaining_time": "17:08:36", "throughput": 5654.13, "total_tokens": 950115240}
+{"current_steps": 21730, "total_steps": 29704, "loss": 0.3051966667175293, "lr": 1.675596392088219e-06, "epoch": 1.4631026124427686, "percentage": 73.16, "elapsed_time": "1 day, 22:41:13", "remaining_time": "17:07:56", "throughput": 5654.32, "total_tokens": 950343040}
+{"current_steps": 21735, "total_steps": 29704, "loss": 0.28461332321166993, "lr": 1.6736218610410633e-06, "epoch": 1.4634392674387287, "percentage": 73.17, "elapsed_time": "1 day, 22:41:52", "remaining_time": "17:07:17", "throughput": 5654.4, "total_tokens": 950576064}
+{"current_steps": 21740, "total_steps": 29704, "loss": 0.30020885467529296, "lr": 1.6716482602049483e-06, "epoch": 1.463775922434689, "percentage": 73.19, "elapsed_time": "1 day, 22:42:29", "remaining_time": "17:06:38", "throughput": 5654.48, "total_tokens": 950795848}
+{"current_steps": 21745, "total_steps": 29704, "loss": 0.3058682203292847, "lr": 1.669675590131784e-06, "epoch": 1.4641125774306492, "percentage": 73.21, "elapsed_time": "1 day, 22:43:03", "remaining_time": "17:05:57", "throughput": 5654.55, "total_tokens": 950999432}
+{"current_steps": 21750, "total_steps": 29704, "loss": 0.29998371601104734, "lr": 1.6677038513732214e-06, "epoch": 1.4644492324266092, "percentage": 73.22, "elapsed_time": "1 day, 22:43:41", "remaining_time": "17:05:18", "throughput": 5654.6, "total_tokens": 951224688}
+{"current_steps": 21755, "total_steps": 29704, "loss": 0.3094784259796143, "lr": 1.6657330444806508e-06, "epoch": 1.4647858874225694, "percentage": 73.24, "elapsed_time": "1 day, 22:44:16", "remaining_time": "17:04:38", "throughput": 5654.68, "total_tokens": 951438456}
+{"current_steps": 21760, "total_steps": 29704, "loss": 0.29812400341033934, "lr": 1.6637631700052021e-06, "epoch": 1.4651225424185295, "percentage": 73.26, "elapsed_time": "1 day, 22:44:54", "remaining_time": "17:03:59", "throughput": 5654.79, "total_tokens": 951670352}
+{"current_steps": 21765, "total_steps": 29704, "loss": 0.2911461591720581, "lr": 1.6617942284977428e-06, "epoch": 1.4654591974144897, "percentage": 73.27, "elapsed_time": "1 day, 22:45:29", "remaining_time": "17:03:19", "throughput": 5654.99, "total_tokens": 951904024}
+{"current_steps": 21770, "total_steps": 29704, "loss": 0.3206942081451416, "lr": 1.6598262205088862e-06, "epoch": 1.4657958524104497, "percentage": 73.29, "elapsed_time": "1 day, 22:46:07", "remaining_time": "17:02:41", "throughput": 5655.05, "total_tokens": 952128864}
+{"current_steps": 21775, "total_steps": 29704, "loss": 0.306864070892334, "lr": 1.6578591465889722e-06, "epoch": 1.46613250740641, "percentage": 73.31, "elapsed_time": "1 day, 22:46:47", "remaining_time": "17:02:02", "throughput": 5655.09, "total_tokens": 952357128}
+{"current_steps": 21780, "total_steps": 29704, "loss": 0.3149855136871338, "lr": 1.6558930072880942e-06, "epoch": 1.46646916240237, "percentage": 73.32, "elapsed_time": "1 day, 22:47:23", "remaining_time": "17:01:23", "throughput": 5655.21, "total_tokens": 952581712}
+{"current_steps": 21785, "total_steps": 29704, "loss": 0.3216851234436035, "lr": 1.6539278031560695e-06, "epoch": 1.4668058173983303, "percentage": 73.34, "elapsed_time": "1 day, 22:48:02", "remaining_time": "17:00:44", "throughput": 5655.15, "total_tokens": 952795432}
+{"current_steps": 21790, "total_steps": 29704, "loss": 0.29906625747680665, "lr": 1.6519635347424672e-06, "epoch": 1.4671424723942903, "percentage": 73.36, "elapsed_time": "1 day, 22:48:39", "remaining_time": "17:00:05", "throughput": 5655.19, "total_tokens": 953010928}
+{"current_steps": 21795, "total_steps": 29704, "loss": 0.3279500246047974, "lr": 1.650000202596586e-06, "epoch": 1.4674791273902505, "percentage": 73.37, "elapsed_time": "1 day, 22:49:22", "remaining_time": "16:59:28", "throughput": 5655.14, "total_tokens": 953243200}
+{"current_steps": 21800, "total_steps": 29704, "loss": 0.3130605220794678, "lr": 1.6480378072674659e-06, "epoch": 1.4678157823862106, "percentage": 73.39, "elapsed_time": "1 day, 22:49:58", "remaining_time": "16:58:48", "throughput": 5655.26, "total_tokens": 953466792}
+{"current_steps": 21805, "total_steps": 29704, "loss": 0.29395854473114014, "lr": 1.646076349303884e-06, "epoch": 1.4681524373821708, "percentage": 73.41, "elapsed_time": "1 day, 22:50:38", "remaining_time": "16:58:10", "throughput": 5655.24, "total_tokens": 953688232}
+{"current_steps": 21810, "total_steps": 29704, "loss": 0.29840645790100095, "lr": 1.6441158292543557e-06, "epoch": 1.4684890923781309, "percentage": 73.42, "elapsed_time": "1 day, 22:51:16", "remaining_time": "16:57:31", "throughput": 5655.19, "total_tokens": 953899232}
+{"current_steps": 21815, "total_steps": 29704, "loss": 0.325870156288147, "lr": 1.6421562476671339e-06, "epoch": 1.468825747374091, "percentage": 73.44, "elapsed_time": "1 day, 22:51:54", "remaining_time": "16:56:52", "throughput": 5655.2, "total_tokens": 954116752}
+{"current_steps": 21820, "total_steps": 29704, "loss": 0.27675824165344237, "lr": 1.640197605090213e-06, "epoch": 1.4691624023700511, "percentage": 73.46, "elapsed_time": "1 day, 22:52:31", "remaining_time": "16:56:13", "throughput": 5655.3, "total_tokens": 954338024}
+{"current_steps": 21825, "total_steps": 29704, "loss": 0.29128522872924806, "lr": 1.6382399020713146e-06, "epoch": 1.4694990573660114, "percentage": 73.47, "elapsed_time": "1 day, 22:53:09", "remaining_time": "16:55:34", "throughput": 5655.2, "total_tokens": 954538472}
+{"current_steps": 21830, "total_steps": 29704, "loss": 0.2947737455368042, "lr": 1.6362831391579115e-06, "epoch": 1.4698357123619714, "percentage": 73.49, "elapsed_time": "1 day, 22:53:48", "remaining_time": "16:54:55", "throughput": 5655.15, "total_tokens": 954753480}
+{"current_steps": 21835, "total_steps": 29704, "loss": 0.3025893926620483, "lr": 1.634327316897199e-06, "epoch": 1.4701723673579317, "percentage": 73.51, "elapsed_time": "1 day, 22:54:26", "remaining_time": "16:54:16", "throughput": 5655.15, "total_tokens": 954963128}
+{"current_steps": 21840, "total_steps": 29704, "loss": 0.2903714656829834, "lr": 1.6323724358361216e-06, "epoch": 1.4705090223538917, "percentage": 73.53, "elapsed_time": "1 day, 22:55:04", "remaining_time": "16:53:38", "throughput": 5655.08, "total_tokens": 955171016}
+{"current_steps": 21845, "total_steps": 29704, "loss": 0.29531302452087405, "lr": 1.6304184965213537e-06, "epoch": 1.470845677349852, "percentage": 73.54, "elapsed_time": "1 day, 22:55:43", "remaining_time": "16:52:59", "throughput": 5655.03, "total_tokens": 955381568}
+{"current_steps": 21850, "total_steps": 29704, "loss": 0.30909888744354247, "lr": 1.6284654994993088e-06, "epoch": 1.471182332345812, "percentage": 73.56, "elapsed_time": "1 day, 22:56:22", "remaining_time": "16:52:20", "throughput": 5655.1, "total_tokens": 955612384}
+{"current_steps": 21855, "total_steps": 29704, "loss": 0.29757180213928225, "lr": 1.6265134453161358e-06, "epoch": 1.4715189873417722, "percentage": 73.58, "elapsed_time": "1 day, 22:57:01", "remaining_time": "16:51:42", "throughput": 5655.05, "total_tokens": 955824680}
+{"current_steps": 21860, "total_steps": 29704, "loss": 0.3056344985961914, "lr": 1.6245623345177203e-06, "epoch": 1.4718556423377323, "percentage": 73.59, "elapsed_time": "1 day, 22:57:40", "remaining_time": "16:51:03", "throughput": 5654.97, "total_tokens": 956030528}
+{"current_steps": 21865, "total_steps": 29704, "loss": 0.28513193130493164, "lr": 1.622612167649683e-06, "epoch": 1.4721922973336925, "percentage": 73.61, "elapsed_time": "1 day, 22:58:14", "remaining_time": "16:50:23", "throughput": 5655.05, "total_tokens": 956238824}
+{"current_steps": 21870, "total_steps": 29704, "loss": 0.2704298496246338, "lr": 1.6206629452573868e-06, "epoch": 1.4725289523296525, "percentage": 73.63, "elapsed_time": "1 day, 22:58:48", "remaining_time": "16:49:43", "throughput": 5655.23, "total_tokens": 956461272}
+{"current_steps": 21875, "total_steps": 29704, "loss": 0.2946579933166504, "lr": 1.618714667885919e-06, "epoch": 1.4728656073256128, "percentage": 73.64, "elapsed_time": "1 day, 22:59:29", "remaining_time": "16:49:05", "throughput": 5655.16, "total_tokens": 956679360}
+{"current_steps": 21880, "total_steps": 29704, "loss": 0.29600162506103517, "lr": 1.6167673360801151e-06, "epoch": 1.4732022623215728, "percentage": 73.66, "elapsed_time": "1 day, 23:00:11", "remaining_time": "16:48:27", "throughput": 5655.12, "total_tokens": 956909072}
+{"current_steps": 21885, "total_steps": 29704, "loss": 0.33778862953186034, "lr": 1.6148209503845336e-06, "epoch": 1.473538917317533, "percentage": 73.68, "elapsed_time": "1 day, 23:00:47", "remaining_time": "16:47:48", "throughput": 5655.2, "total_tokens": 957127864}
+{"current_steps": 21890, "total_steps": 29704, "loss": 0.2987597703933716, "lr": 1.6128755113434808e-06, "epoch": 1.473875572313493, "percentage": 73.69, "elapsed_time": "1 day, 23:01:26", "remaining_time": "16:47:09", "throughput": 5655.17, "total_tokens": 957342536}
+{"current_steps": 21895, "total_steps": 29704, "loss": 0.303321099281311, "lr": 1.6109310195009908e-06, "epoch": 1.4742122273094533, "percentage": 73.71, "elapsed_time": "1 day, 23:02:05", "remaining_time": "16:46:30", "throughput": 5655.24, "total_tokens": 957573936}
+{"current_steps": 21900, "total_steps": 29704, "loss": 0.29223430156707764, "lr": 1.6089874754008323e-06, "epoch": 1.4745488823054134, "percentage": 73.73, "elapsed_time": "1 day, 23:02:40", "remaining_time": "16:45:51", "throughput": 5655.26, "total_tokens": 957779504}
+{"current_steps": 21905, "total_steps": 29704, "loss": 0.3116729974746704, "lr": 1.6070448795865162e-06, "epoch": 1.4748855373013736, "percentage": 73.74, "elapsed_time": "1 day, 23:03:16", "remaining_time": "16:45:11", "throughput": 5655.35, "total_tokens": 957995792}
+{"current_steps": 21910, "total_steps": 29704, "loss": 0.29828710556030275, "lr": 1.605103232601277e-06, "epoch": 1.4752221922973336, "percentage": 73.76, "elapsed_time": "1 day, 23:03:59", "remaining_time": "16:44:34", "throughput": 5655.19, "total_tokens": 958213080}
+{"current_steps": 21915, "total_steps": 29704, "loss": 0.2677397966384888, "lr": 1.603162534988096e-06, "epoch": 1.475558847293294, "percentage": 73.78, "elapsed_time": "1 day, 23:04:37", "remaining_time": "16:43:55", "throughput": 5655.22, "total_tokens": 958431216}
+{"current_steps": 21920, "total_steps": 29704, "loss": 0.28536086082458495, "lr": 1.6012227872896791e-06, "epoch": 1.475895502289254, "percentage": 73.79, "elapsed_time": "1 day, 23:05:10", "remaining_time": "16:43:14", "throughput": 5655.35, "total_tokens": 958642264}
+{"current_steps": 21925, "total_steps": 29704, "loss": 0.2980980396270752, "lr": 1.599283990048472e-06, "epoch": 1.4762321572852142, "percentage": 73.81, "elapsed_time": "1 day, 23:05:45", "remaining_time": "16:42:34", "throughput": 5655.4, "total_tokens": 958850528}
+{"current_steps": 21930, "total_steps": 29704, "loss": 0.2987915754318237, "lr": 1.5973461438066533e-06, "epoch": 1.4765688122811742, "percentage": 73.83, "elapsed_time": "1 day, 23:06:31", "remaining_time": "16:41:58", "throughput": 5655.18, "total_tokens": 959072368}
+{"current_steps": 21935, "total_steps": 29704, "loss": 0.28679566383361815, "lr": 1.5954092491061351e-06, "epoch": 1.4769054672771345, "percentage": 73.85, "elapsed_time": "1 day, 23:07:13", "remaining_time": "16:41:21", "throughput": 5655.05, "total_tokens": 959287640}
+{"current_steps": 21940, "total_steps": 29704, "loss": 0.30909249782562254, "lr": 1.5934733064885637e-06, "epoch": 1.4772421222730945, "percentage": 73.86, "elapsed_time": "1 day, 23:07:54", "remaining_time": "16:40:43", "throughput": 5655.03, "total_tokens": 959515464}
+{"current_steps": 21945, "total_steps": 29704, "loss": 0.3021413803100586, "lr": 1.5915383164953203e-06, "epoch": 1.4775787772690547, "percentage": 73.88, "elapsed_time": "1 day, 23:08:32", "remaining_time": "16:40:04", "throughput": 5654.98, "total_tokens": 959722168}
+{"current_steps": 21950, "total_steps": 29704, "loss": 0.2926707983016968, "lr": 1.5896042796675155e-06, "epoch": 1.4779154322650148, "percentage": 73.9, "elapsed_time": "1 day, 23:09:09", "remaining_time": "16:39:25", "throughput": 5654.97, "total_tokens": 959926384}
+{"current_steps": 21955, "total_steps": 29704, "loss": 0.3000811576843262, "lr": 1.5876711965460023e-06, "epoch": 1.478252087260975, "percentage": 73.91, "elapsed_time": "1 day, 23:09:53", "remaining_time": "16:38:48", "throughput": 5654.78, "total_tokens": 960147848}
+{"current_steps": 21960, "total_steps": 29704, "loss": 0.3030038118362427, "lr": 1.5857390676713546e-06, "epoch": 1.478588742256935, "percentage": 73.93, "elapsed_time": "1 day, 23:10:26", "remaining_time": "16:38:07", "throughput": 5654.91, "total_tokens": 960354248}
+{"current_steps": 21965, "total_steps": 29704, "loss": 0.2610928058624268, "lr": 1.5838078935838913e-06, "epoch": 1.4789253972528953, "percentage": 73.95, "elapsed_time": "1 day, 23:11:05", "remaining_time": "16:37:29", "throughput": 5654.83, "total_tokens": 960559904}
+{"current_steps": 21970, "total_steps": 29704, "loss": 0.286395788192749, "lr": 1.581877674823657e-06, "epoch": 1.4792620522488553, "percentage": 73.96, "elapsed_time": "1 day, 23:11:44", "remaining_time": "16:36:50", "throughput": 5654.72, "total_tokens": 960763880}
+{"current_steps": 21975, "total_steps": 29704, "loss": 0.3111164093017578, "lr": 1.5799484119304308e-06, "epoch": 1.4795987072448156, "percentage": 73.98, "elapsed_time": "1 day, 23:12:28", "remaining_time": "16:36:13", "throughput": 5654.57, "total_tokens": 960983256}
+{"current_steps": 21980, "total_steps": 29704, "loss": 0.2867883205413818, "lr": 1.578020105443726e-06, "epoch": 1.4799353622407756, "percentage": 74.0, "elapsed_time": "1 day, 23:13:09", "remaining_time": "16:35:36", "throughput": 5654.49, "total_tokens": 961204296}
+{"current_steps": 21985, "total_steps": 29704, "loss": 0.3115290641784668, "lr": 1.576092755902787e-06, "epoch": 1.4802720172367358, "percentage": 74.01, "elapsed_time": "1 day, 23:13:45", "remaining_time": "16:34:56", "throughput": 5654.55, "total_tokens": 961419104}
+{"current_steps": 21990, "total_steps": 29704, "loss": 0.32115159034729, "lr": 1.5741663638465903e-06, "epoch": 1.4806086722326959, "percentage": 74.03, "elapsed_time": "1 day, 23:14:19", "remaining_time": "16:34:16", "throughput": 5654.65, "total_tokens": 961624552}
+{"current_steps": 21995, "total_steps": 29704, "loss": 0.2831075668334961, "lr": 1.5722409298138463e-06, "epoch": 1.4809453272286561, "percentage": 74.05, "elapsed_time": "1 day, 23:14:57", "remaining_time": "16:33:37", "throughput": 5654.57, "total_tokens": 961828112}
+{"current_steps": 22000, "total_steps": 29704, "loss": 0.3216114044189453, "lr": 1.570316454342994e-06, "epoch": 1.4812819822246162, "percentage": 74.06, "elapsed_time": "1 day, 23:15:35", "remaining_time": "16:32:58", "throughput": 5654.57, "total_tokens": 962043280}
+{"current_steps": 22005, "total_steps": 29704, "loss": 0.30264816284179685, "lr": 1.568392937972212e-06, "epoch": 1.4816186372205764, "percentage": 74.08, "elapsed_time": "1 day, 23:16:17", "remaining_time": "16:32:20", "throughput": 5654.56, "total_tokens": 962277624}
+{"current_steps": 22010, "total_steps": 29704, "loss": 0.30075869560241697, "lr": 1.5664703812393988e-06, "epoch": 1.4819552922165364, "percentage": 74.1, "elapsed_time": "1 day, 23:16:55", "remaining_time": "16:31:41", "throughput": 5654.67, "total_tokens": 962510352}
+{"current_steps": 22015, "total_steps": 29704, "loss": 0.2980914831161499, "lr": 1.5645487846821967e-06, "epoch": 1.4822919472124967, "percentage": 74.11, "elapsed_time": "1 day, 23:17:36", "remaining_time": "16:31:04", "throughput": 5654.63, "total_tokens": 962736264}
+{"current_steps": 22020, "total_steps": 29704, "loss": 0.30644855499267576, "lr": 1.5626281488379724e-06, "epoch": 1.4826286022084567, "percentage": 74.13, "elapsed_time": "1 day, 23:18:10", "remaining_time": "16:30:23", "throughput": 5654.76, "total_tokens": 962954856}
+{"current_steps": 22025, "total_steps": 29704, "loss": 0.28392848968505857, "lr": 1.560708474243825e-06, "epoch": 1.482965257204417, "percentage": 74.15, "elapsed_time": "1 day, 23:18:50", "remaining_time": "16:29:45", "throughput": 5654.72, "total_tokens": 963172840}
+{"current_steps": 22030, "total_steps": 29704, "loss": 0.3137376070022583, "lr": 1.5587897614365854e-06, "epoch": 1.483301912200377, "percentage": 74.17, "elapsed_time": "1 day, 23:19:23", "remaining_time": "16:29:05", "throughput": 5654.97, "total_tokens": 963402704}
+{"current_steps": 22035, "total_steps": 29704, "loss": 0.29897422790527345, "lr": 1.5568720109528163e-06, "epoch": 1.4836385671963372, "percentage": 74.18, "elapsed_time": "1 day, 23:20:07", "remaining_time": "16:28:28", "throughput": 5654.8, "total_tokens": 963622656}
+{"current_steps": 22040, "total_steps": 29704, "loss": 0.32768850326538085, "lr": 1.5549552233288095e-06, "epoch": 1.4839752221922973, "percentage": 74.2, "elapsed_time": "1 day, 23:20:44", "remaining_time": "16:27:48", "throughput": 5654.87, "total_tokens": 963839112}
+{"current_steps": 22045, "total_steps": 29704, "loss": 0.2966951370239258, "lr": 1.5530393991005893e-06, "epoch": 1.4843118771882575, "percentage": 74.22, "elapsed_time": "1 day, 23:21:35", "remaining_time": "16:27:14", "throughput": 5654.45, "total_tokens": 964056312}
+{"current_steps": 22050, "total_steps": 29704, "loss": 0.3010308265686035, "lr": 1.551124538803908e-06, "epoch": 1.4846485321842176, "percentage": 74.23, "elapsed_time": "1 day, 23:22:14", "remaining_time": "16:26:35", "throughput": 5654.48, "total_tokens": 964282488}
+{"current_steps": 22055, "total_steps": 29704, "loss": 0.30379014015197753, "lr": 1.5492106429742547e-06, "epoch": 1.4849851871801778, "percentage": 74.25, "elapsed_time": "1 day, 23:22:54", "remaining_time": "16:25:57", "throughput": 5654.46, "total_tokens": 964505592}
+{"current_steps": 22060, "total_steps": 29704, "loss": 0.3085505962371826, "lr": 1.5472977121468386e-06, "epoch": 1.4853218421761378, "percentage": 74.27, "elapsed_time": "1 day, 23:23:25", "remaining_time": "16:25:16", "throughput": 5654.77, "total_tokens": 964734320}
+{"current_steps": 22065, "total_steps": 29704, "loss": 0.2914187669754028, "lr": 1.545385746856609e-06, "epoch": 1.485658497172098, "percentage": 74.28, "elapsed_time": "1 day, 23:24:01", "remaining_time": "16:24:36", "throughput": 5654.83, "total_tokens": 964946888}
+{"current_steps": 22070, "total_steps": 29704, "loss": 0.2726666212081909, "lr": 1.5434747476382388e-06, "epoch": 1.485995152168058, "percentage": 74.3, "elapsed_time": "1 day, 23:24:36", "remaining_time": "16:23:56", "throughput": 5655.03, "total_tokens": 965178856}
+{"current_steps": 22075, "total_steps": 29704, "loss": 0.2993777751922607, "lr": 1.5415647150261337e-06, "epoch": 1.4863318071640184, "percentage": 74.32, "elapsed_time": "1 day, 23:25:14", "remaining_time": "16:23:18", "throughput": 5654.94, "total_tokens": 965382264}
+{"current_steps": 22080, "total_steps": 29704, "loss": 0.2940176010131836, "lr": 1.539655649554428e-06, "epoch": 1.4866684621599784, "percentage": 74.33, "elapsed_time": "1 day, 23:25:55", "remaining_time": "16:22:40", "throughput": 5654.86, "total_tokens": 965597136}
+{"current_steps": 22085, "total_steps": 29704, "loss": 0.29267857074737547, "lr": 1.5377475517569856e-06, "epoch": 1.4870051171559386, "percentage": 74.35, "elapsed_time": "1 day, 23:26:31", "remaining_time": "16:22:00", "throughput": 5654.85, "total_tokens": 965803184}
+{"current_steps": 22090, "total_steps": 29704, "loss": 0.30087296962738036, "lr": 1.5358404221674e-06, "epoch": 1.4873417721518987, "percentage": 74.37, "elapsed_time": "1 day, 23:27:11", "remaining_time": "16:21:22", "throughput": 5654.75, "total_tokens": 966009728}
+{"current_steps": 22095, "total_steps": 29704, "loss": 0.30022130012512205, "lr": 1.533934261318994e-06, "epoch": 1.487678427147859, "percentage": 74.38, "elapsed_time": "1 day, 23:27:50", "remaining_time": "16:20:43", "throughput": 5654.77, "total_tokens": 966233096}
+{"current_steps": 22100, "total_steps": 29704, "loss": 0.2852438449859619, "lr": 1.5320290697448177e-06, "epoch": 1.488015082143819, "percentage": 74.4, "elapsed_time": "1 day, 23:28:31", "remaining_time": "16:20:05", "throughput": 5654.66, "total_tokens": 966445288}
+{"current_steps": 22105, "total_steps": 29704, "loss": 0.320376992225647, "lr": 1.530124847977657e-06, "epoch": 1.4883517371397792, "percentage": 74.42, "elapsed_time": "1 day, 23:29:08", "remaining_time": "16:19:26", "throughput": 5654.67, "total_tokens": 966659272}
+{"current_steps": 22110, "total_steps": 29704, "loss": 0.2888146638870239, "lr": 1.528221596550014e-06, "epoch": 1.4886883921357392, "percentage": 74.43, "elapsed_time": "1 day, 23:29:48", "remaining_time": "16:18:48", "throughput": 5654.6, "total_tokens": 966874112}
+{"current_steps": 22115, "total_steps": 29704, "loss": 0.2921769618988037, "lr": 1.5263193159941331e-06, "epoch": 1.4890250471316995, "percentage": 74.45, "elapsed_time": "1 day, 23:30:28", "remaining_time": "16:18:10", "throughput": 5654.55, "total_tokens": 967092032}
+{"current_steps": 22120, "total_steps": 29704, "loss": 0.28116235733032224, "lr": 1.5244180068419783e-06, "epoch": 1.4893617021276595, "percentage": 74.47, "elapsed_time": "1 day, 23:31:17", "remaining_time": "16:17:34", "throughput": 5654.23, "total_tokens": 967308280}
+{"current_steps": 22125, "total_steps": 29704, "loss": 0.32404913902282717, "lr": 1.5225176696252453e-06, "epoch": 1.4896983571236198, "percentage": 74.48, "elapsed_time": "1 day, 23:32:04", "remaining_time": "16:16:59", "throughput": 5653.89, "total_tokens": 967517656}
+{"current_steps": 22130, "total_steps": 29704, "loss": 0.30272302627563474, "lr": 1.5206183048753575e-06, "epoch": 1.4900350121195798, "percentage": 74.5, "elapsed_time": "1 day, 23:32:41", "remaining_time": "16:16:20", "throughput": 5653.89, "total_tokens": 967731128}
+{"current_steps": 22135, "total_steps": 29704, "loss": 0.34148373603820803, "lr": 1.5187199131234653e-06, "epoch": 1.49037166711554, "percentage": 74.52, "elapsed_time": "1 day, 23:33:17", "remaining_time": "16:15:40", "throughput": 5654.0, "total_tokens": 967953912}
+{"current_steps": 22140, "total_steps": 29704, "loss": 0.2946127653121948, "lr": 1.5168224949004484e-06, "epoch": 1.4907083221115, "percentage": 74.54, "elapsed_time": "1 day, 23:33:59", "remaining_time": "16:15:02", "throughput": 5653.87, "total_tokens": 968163840}
+{"current_steps": 22145, "total_steps": 29704, "loss": 0.2730919361114502, "lr": 1.514926050736914e-06, "epoch": 1.4910449771074603, "percentage": 74.55, "elapsed_time": "1 day, 23:34:48", "remaining_time": "16:14:27", "throughput": 5653.56, "total_tokens": 968387800}
+{"current_steps": 22150, "total_steps": 29704, "loss": 0.3045715570449829, "lr": 1.5130305811631945e-06, "epoch": 1.4913816321034203, "percentage": 74.57, "elapsed_time": "1 day, 23:35:26", "remaining_time": "16:13:48", "throughput": 5653.61, "total_tokens": 968613216}
+{"current_steps": 22155, "total_steps": 29704, "loss": 0.3100686550140381, "lr": 1.5111360867093571e-06, "epoch": 1.4917182870993806, "percentage": 74.59, "elapsed_time": "1 day, 23:36:00", "remaining_time": "16:13:08", "throughput": 5653.8, "total_tokens": 968837880}
+{"current_steps": 22160, "total_steps": 29704, "loss": 0.2817687511444092, "lr": 1.5092425679051848e-06, "epoch": 1.4920549420953406, "percentage": 74.6, "elapsed_time": "1 day, 23:36:34", "remaining_time": "16:12:28", "throughput": 5653.97, "total_tokens": 969060560}
+{"current_steps": 22165, "total_steps": 29704, "loss": 0.31572632789611815, "lr": 1.5073500252801988e-06, "epoch": 1.4923915970913009, "percentage": 74.62, "elapsed_time": "1 day, 23:37:03", "remaining_time": "16:11:46", "throughput": 5654.36, "total_tokens": 969287272}
+{"current_steps": 22170, "total_steps": 29704, "loss": 0.2992145776748657, "lr": 1.5054584593636412e-06, "epoch": 1.492728252087261, "percentage": 74.64, "elapsed_time": "1 day, 23:37:39", "remaining_time": "16:11:06", "throughput": 5654.51, "total_tokens": 969517872}
+{"current_steps": 22175, "total_steps": 29704, "loss": 0.3096607208251953, "lr": 1.5035678706844814e-06, "epoch": 1.4930649070832211, "percentage": 74.65, "elapsed_time": "1 day, 23:38:16", "remaining_time": "16:10:27", "throughput": 5654.54, "total_tokens": 969734072}
+{"current_steps": 22180, "total_steps": 29704, "loss": 0.311397123336792, "lr": 1.501678259771418e-06, "epoch": 1.4934015620791812, "percentage": 74.67, "elapsed_time": "1 day, 23:39:01", "remaining_time": "16:09:51", "throughput": 5654.34, "total_tokens": 969956280}
+{"current_steps": 22185, "total_steps": 29704, "loss": 0.3107875108718872, "lr": 1.499789627152874e-06, "epoch": 1.4937382170751414, "percentage": 74.69, "elapsed_time": "1 day, 23:39:44", "remaining_time": "16:09:13", "throughput": 5654.26, "total_tokens": 970180640}
+{"current_steps": 22190, "total_steps": 29704, "loss": 0.26401643753051757, "lr": 1.497901973356999e-06, "epoch": 1.4940748720711015, "percentage": 74.7, "elapsed_time": "1 day, 23:40:21", "remaining_time": "16:08:34", "throughput": 5654.33, "total_tokens": 970405336}
+{"current_steps": 22195, "total_steps": 29704, "loss": 0.29029550552368166, "lr": 1.4960152989116694e-06, "epoch": 1.4944115270670617, "percentage": 74.72, "elapsed_time": "1 day, 23:40:59", "remaining_time": "16:07:55", "throughput": 5654.37, "total_tokens": 970628528}
+{"current_steps": 22200, "total_steps": 29704, "loss": 0.31101787090301514, "lr": 1.4941296043444869e-06, "epoch": 1.4947481820630217, "percentage": 74.74, "elapsed_time": "1 day, 23:41:40", "remaining_time": "16:07:17", "throughput": 5654.23, "total_tokens": 970833256}
+{"current_steps": 22205, "total_steps": 29704, "loss": 0.2868825912475586, "lr": 1.4922448901827847e-06, "epoch": 1.495084837058982, "percentage": 74.75, "elapsed_time": "1 day, 23:42:26", "remaining_time": "16:06:41", "throughput": 5654.0, "total_tokens": 971056192}
+{"current_steps": 22210, "total_steps": 29704, "loss": 0.27285065650939944, "lr": 1.4903611569536103e-06, "epoch": 1.495421492054942, "percentage": 74.77, "elapsed_time": "1 day, 23:43:04", "remaining_time": "16:06:02", "throughput": 5653.9, "total_tokens": 971251664}
+{"current_steps": 22215, "total_steps": 29704, "loss": 0.3076265096664429, "lr": 1.488478405183748e-06, "epoch": 1.4957581470509023, "percentage": 74.79, "elapsed_time": "1 day, 23:43:39", "remaining_time": "16:05:22", "throughput": 5653.96, "total_tokens": 971461824}
+{"current_steps": 22220, "total_steps": 29704, "loss": 0.2991771221160889, "lr": 1.4865966353997024e-06, "epoch": 1.4960948020468623, "percentage": 74.8, "elapsed_time": "1 day, 23:44:13", "remaining_time": "16:04:42", "throughput": 5654.17, "total_tokens": 971691984}
+{"current_steps": 22225, "total_steps": 29704, "loss": 0.27810006141662597, "lr": 1.4847158481277047e-06, "epoch": 1.4964314570428225, "percentage": 74.82, "elapsed_time": "1 day, 23:44:52", "remaining_time": "16:04:04", "throughput": 5654.17, "total_tokens": 971910752}
+{"current_steps": 22230, "total_steps": 29704, "loss": 0.274126935005188, "lr": 1.48283604389371e-06, "epoch": 1.4967681120387826, "percentage": 74.84, "elapsed_time": "1 day, 23:45:29", "remaining_time": "16:03:24", "throughput": 5654.17, "total_tokens": 972116168}
+{"current_steps": 22235, "total_steps": 29704, "loss": 0.28500995635986326, "lr": 1.480957223223401e-06, "epoch": 1.4971047670347428, "percentage": 74.86, "elapsed_time": "1 day, 23:46:06", "remaining_time": "16:02:45", "throughput": 5654.21, "total_tokens": 972336272}
+{"current_steps": 22240, "total_steps": 29704, "loss": 0.32386600971221924, "lr": 1.4790793866421826e-06, "epoch": 1.497441422030703, "percentage": 74.87, "elapsed_time": "1 day, 23:46:43", "remaining_time": "16:02:06", "throughput": 5654.25, "total_tokens": 972548536}
+{"current_steps": 22245, "total_steps": 29704, "loss": 0.30510640144348145, "lr": 1.4772025346751862e-06, "epoch": 1.497778077026663, "percentage": 74.89, "elapsed_time": "1 day, 23:47:15", "remaining_time": "16:01:25", "throughput": 5654.46, "total_tokens": 972767336}
+{"current_steps": 22250, "total_steps": 29704, "loss": 0.3076331615447998, "lr": 1.475326667847266e-06, "epoch": 1.4981147320226231, "percentage": 74.91, "elapsed_time": "1 day, 23:47:53", "remaining_time": "16:00:46", "throughput": 5654.49, "total_tokens": 972987240}
+{"current_steps": 22255, "total_steps": 29704, "loss": 0.30481748580932616, "lr": 1.4734517866830073e-06, "epoch": 1.4984513870185834, "percentage": 74.92, "elapsed_time": "1 day, 23:48:27", "remaining_time": "16:00:06", "throughput": 5654.71, "total_tokens": 973215336}
+{"current_steps": 22260, "total_steps": 29704, "loss": 0.303895378112793, "lr": 1.4715778917067081e-06, "epoch": 1.4987880420145436, "percentage": 74.94, "elapsed_time": "1 day, 23:49:09", "remaining_time": "15:59:28", "throughput": 5654.54, "total_tokens": 973425600}
+{"current_steps": 22265, "total_steps": 29704, "loss": 0.29153730869293215, "lr": 1.4697049834424016e-06, "epoch": 1.4991246970105037, "percentage": 74.96, "elapsed_time": "1 day, 23:49:48", "remaining_time": "15:58:50", "throughput": 5654.51, "total_tokens": 973643560}
+{"current_steps": 22270, "total_steps": 29704, "loss": 0.285983943939209, "lr": 1.4678330624138387e-06, "epoch": 1.4994613520064637, "percentage": 74.97, "elapsed_time": "1 day, 23:50:21", "remaining_time": "15:58:09", "throughput": 5654.65, "total_tokens": 973854824}
+{"current_steps": 22275, "total_steps": 29704, "loss": 0.2954267978668213, "lr": 1.465962129144497e-06, "epoch": 1.499798007002424, "percentage": 74.99, "elapsed_time": "1 day, 23:50:56", "remaining_time": "15:57:29", "throughput": 5654.8, "total_tokens": 974073608}
+{"current_steps": 22280, "total_steps": 29704, "loss": 0.32407686710357664, "lr": 1.4640921841575766e-06, "epoch": 1.5001346619983842, "percentage": 75.01, "elapsed_time": "1 day, 23:51:32", "remaining_time": "15:56:50", "throughput": 5654.96, "total_tokens": 974307744}
+{"current_steps": 22285, "total_steps": 29704, "loss": 0.29003033638000486, "lr": 1.4622232279760018e-06, "epoch": 1.5004713169943442, "percentage": 75.02, "elapsed_time": "1 day, 23:52:06", "remaining_time": "15:56:10", "throughput": 5655.12, "total_tokens": 974527448}
+{"current_steps": 22290, "total_steps": 29704, "loss": 0.3260036468505859, "lr": 1.4603552611224202e-06, "epoch": 1.5008079719903042, "percentage": 75.04, "elapsed_time": "1 day, 23:52:44", "remaining_time": "15:55:31", "throughput": 5655.05, "total_tokens": 974732192}
+{"current_steps": 22295, "total_steps": 29704, "loss": 0.3202426195144653, "lr": 1.4584882841192027e-06, "epoch": 1.5011446269862645, "percentage": 75.06, "elapsed_time": "1 day, 23:53:22", "remaining_time": "15:54:52", "throughput": 5655.08, "total_tokens": 974950184}
+{"current_steps": 22300, "total_steps": 29704, "loss": 0.2875128030776978, "lr": 1.456622297488442e-06, "epoch": 1.5014812819822247, "percentage": 75.07, "elapsed_time": "1 day, 23:53:56", "remaining_time": "15:54:12", "throughput": 5655.17, "total_tokens": 975160280}
+{"current_steps": 22305, "total_steps": 29704, "loss": 0.29420223236083987, "lr": 1.4547573017519595e-06, "epoch": 1.5018179369781848, "percentage": 75.09, "elapsed_time": "1 day, 23:54:36", "remaining_time": "15:53:33", "throughput": 5655.07, "total_tokens": 975368840}
+{"current_steps": 22310, "total_steps": 29704, "loss": 0.28504467010498047, "lr": 1.4528932974312897e-06, "epoch": 1.5021545919741448, "percentage": 75.11, "elapsed_time": "1 day, 23:55:19", "remaining_time": "15:52:56", "throughput": 5655.03, "total_tokens": 975600768}
+{"current_steps": 22315, "total_steps": 29704, "loss": 0.29707143306732176, "lr": 1.4510302850477005e-06, "epoch": 1.502491246970105, "percentage": 75.12, "elapsed_time": "1 day, 23:55:57", "remaining_time": "15:52:17", "throughput": 5655.04, "total_tokens": 975818888}
+{"current_steps": 22320, "total_steps": 29704, "loss": 0.289170503616333, "lr": 1.4491682651221746e-06, "epoch": 1.5028279019660653, "percentage": 75.14, "elapsed_time": "1 day, 23:56:33", "remaining_time": "15:51:38", "throughput": 5655.04, "total_tokens": 976024992}
+{"current_steps": 22325, "total_steps": 29704, "loss": 0.298907732963562, "lr": 1.447307238175421e-06, "epoch": 1.5031645569620253, "percentage": 75.16, "elapsed_time": "1 day, 23:57:14", "remaining_time": "15:51:00", "throughput": 5654.89, "total_tokens": 976231792}
+{"current_steps": 22330, "total_steps": 29704, "loss": 0.29713735580444334, "lr": 1.4454472047278688e-06, "epoch": 1.5035012119579854, "percentage": 75.18, "elapsed_time": "1 day, 23:57:56", "remaining_time": "15:50:22", "throughput": 5654.76, "total_tokens": 976442768}
+{"current_steps": 22335, "total_steps": 29704, "loss": 0.2721333265304565, "lr": 1.443588165299672e-06, "epoch": 1.5038378669539456, "percentage": 75.19, "elapsed_time": "1 day, 23:58:41", "remaining_time": "15:49:46", "throughput": 5654.59, "total_tokens": 976670000}
+{"current_steps": 22340, "total_steps": 29704, "loss": 0.31305713653564454, "lr": 1.4417301204107037e-06, "epoch": 1.5041745219499059, "percentage": 75.21, "elapsed_time": "1 day, 23:59:17", "remaining_time": "15:49:06", "throughput": 5654.7, "total_tokens": 976892192}
+{"current_steps": 22345, "total_steps": 29704, "loss": 0.3005058288574219, "lr": 1.4398730705805607e-06, "epoch": 1.504511176945866, "percentage": 75.23, "elapsed_time": "1 day, 23:59:55", "remaining_time": "15:48:27", "throughput": 5654.76, "total_tokens": 977116432}
+{"current_steps": 22350, "total_steps": 29704, "loss": 0.3000321388244629, "lr": 1.43801701632856e-06, "epoch": 1.504847831941826, "percentage": 75.24, "elapsed_time": "2 days, 0:00:31", "remaining_time": "15:47:48", "throughput": 5654.89, "total_tokens": 977343248}
+{"current_steps": 22355, "total_steps": 29704, "loss": 0.32420189380645753, "lr": 1.436161958173745e-06, "epoch": 1.5051844869377862, "percentage": 75.26, "elapsed_time": "2 days, 0:01:08", "remaining_time": "15:47:08", "throughput": 5654.98, "total_tokens": 977567152}
+{"current_steps": 22360, "total_steps": 29704, "loss": 0.3067188262939453, "lr": 1.4343078966348718e-06, "epoch": 1.5055211419337464, "percentage": 75.28, "elapsed_time": "2 days, 0:01:50", "remaining_time": "15:46:31", "throughput": 5654.83, "total_tokens": 977777560}
+{"current_steps": 22365, "total_steps": 29704, "loss": 0.28587291240692136, "lr": 1.432454832230426e-06, "epoch": 1.5058577969297064, "percentage": 75.29, "elapsed_time": "2 days, 0:02:33", "remaining_time": "15:45:54", "throughput": 5654.75, "total_tokens": 978009320}
+{"current_steps": 22370, "total_steps": 29704, "loss": 0.28476083278656006, "lr": 1.4306027654786103e-06, "epoch": 1.5061944519256665, "percentage": 75.31, "elapsed_time": "2 days, 0:03:11", "remaining_time": "15:45:15", "throughput": 5654.83, "total_tokens": 978236352}
+{"current_steps": 22375, "total_steps": 29704, "loss": 0.3085874080657959, "lr": 1.4287516968973491e-06, "epoch": 1.5065311069216267, "percentage": 75.33, "elapsed_time": "2 days, 0:03:48", "remaining_time": "15:44:36", "throughput": 5654.87, "total_tokens": 978453896}
+{"current_steps": 22380, "total_steps": 29704, "loss": 0.3164313077926636, "lr": 1.4269016270042884e-06, "epoch": 1.506867761917587, "percentage": 75.34, "elapsed_time": "2 days, 0:04:29", "remaining_time": "15:43:58", "throughput": 5654.81, "total_tokens": 978673624}
+{"current_steps": 22385, "total_steps": 29704, "loss": 0.2872413873672485, "lr": 1.425052556316794e-06, "epoch": 1.507204416913547, "percentage": 75.36, "elapsed_time": "2 days, 0:05:10", "remaining_time": "15:43:20", "throughput": 5654.74, "total_tokens": 978891952}
+{"current_steps": 22390, "total_steps": 29704, "loss": 0.3223658323287964, "lr": 1.4232044853519522e-06, "epoch": 1.507541071909507, "percentage": 75.38, "elapsed_time": "2 days, 0:05:49", "remaining_time": "15:42:41", "throughput": 5654.65, "total_tokens": 979097208}
+{"current_steps": 22395, "total_steps": 29704, "loss": 0.28975491523742675, "lr": 1.4213574146265708e-06, "epoch": 1.5078777269054673, "percentage": 75.39, "elapsed_time": "2 days, 0:06:24", "remaining_time": "15:42:01", "throughput": 5654.7, "total_tokens": 979304416}
+{"current_steps": 22400, "total_steps": 29704, "loss": 0.32438082695007325, "lr": 1.4195113446571752e-06, "epoch": 1.5082143819014275, "percentage": 75.41, "elapsed_time": "2 days, 0:06:59", "remaining_time": "15:41:22", "throughput": 5654.83, "total_tokens": 979527856}
+{"current_steps": 22405, "total_steps": 29704, "loss": 0.2848972797393799, "lr": 1.417666275960018e-06, "epoch": 1.5085510368973876, "percentage": 75.43, "elapsed_time": "2 days, 0:07:30", "remaining_time": "15:40:40", "throughput": 5655.17, "total_tokens": 979761880}
+{"current_steps": 22410, "total_steps": 29704, "loss": 0.27116708755493163, "lr": 1.4158222090510605e-06, "epoch": 1.5088876918933476, "percentage": 75.44, "elapsed_time": "2 days, 0:08:09", "remaining_time": "15:40:02", "throughput": 5655.22, "total_tokens": 979990616}
+{"current_steps": 22415, "total_steps": 29704, "loss": 0.3101805686950684, "lr": 1.4139791444459945e-06, "epoch": 1.5092243468893078, "percentage": 75.46, "elapsed_time": "2 days, 0:08:43", "remaining_time": "15:39:21", "throughput": 5655.39, "total_tokens": 980209928}
+{"current_steps": 22420, "total_steps": 29704, "loss": 0.3069303512573242, "lr": 1.4121370826602254e-06, "epoch": 1.509561001885268, "percentage": 75.48, "elapsed_time": "2 days, 0:09:23", "remaining_time": "15:38:43", "throughput": 5655.29, "total_tokens": 980417304}
+{"current_steps": 22425, "total_steps": 29704, "loss": 0.3054825782775879, "lr": 1.4102960242088803e-06, "epoch": 1.5098976568812281, "percentage": 75.49, "elapsed_time": "2 days, 0:10:07", "remaining_time": "15:38:06", "throughput": 5655.09, "total_tokens": 980633968}
+{"current_steps": 22430, "total_steps": 29704, "loss": 0.28913288116455077, "lr": 1.4084559696068057e-06, "epoch": 1.5102343118771882, "percentage": 75.51, "elapsed_time": "2 days, 0:10:45", "remaining_time": "15:37:27", "throughput": 5655.09, "total_tokens": 980850136}
+{"current_steps": 22435, "total_steps": 29704, "loss": 0.2985984325408936, "lr": 1.406616919368567e-06, "epoch": 1.5105709668731484, "percentage": 75.53, "elapsed_time": "2 days, 0:11:28", "remaining_time": "15:36:50", "throughput": 5654.95, "total_tokens": 981066808}
+{"current_steps": 22440, "total_steps": 29704, "loss": 0.29711313247680665, "lr": 1.4047788740084474e-06, "epoch": 1.5109076218691087, "percentage": 75.55, "elapsed_time": "2 days, 0:12:10", "remaining_time": "15:36:12", "throughput": 5654.88, "total_tokens": 981290632}
+{"current_steps": 22445, "total_steps": 29704, "loss": 0.2857701539993286, "lr": 1.402941834040452e-06, "epoch": 1.5112442768650687, "percentage": 75.56, "elapsed_time": "2 days, 0:12:51", "remaining_time": "15:35:35", "throughput": 5654.88, "total_tokens": 981524520}
+{"current_steps": 22450, "total_steps": 29704, "loss": 0.28029732704162597, "lr": 1.4011057999783006e-06, "epoch": 1.5115809318610287, "percentage": 75.58, "elapsed_time": "2 days, 0:13:33", "remaining_time": "15:34:57", "throughput": 5654.72, "total_tokens": 981733352}
+{"current_steps": 22455, "total_steps": 29704, "loss": 0.3050365924835205, "lr": 1.3992707723354398e-06, "epoch": 1.511917586856989, "percentage": 75.6, "elapsed_time": "2 days, 0:14:11", "remaining_time": "15:34:18", "throughput": 5654.74, "total_tokens": 981955216}
+{"current_steps": 22460, "total_steps": 29704, "loss": 0.2846760988235474, "lr": 1.397436751625022e-06, "epoch": 1.5122542418529492, "percentage": 75.61, "elapsed_time": "2 days, 0:14:52", "remaining_time": "15:33:40", "throughput": 5654.64, "total_tokens": 982166184}
+{"current_steps": 22465, "total_steps": 29704, "loss": 0.2913294792175293, "lr": 1.3956037383599308e-06, "epoch": 1.5125908968489092, "percentage": 75.63, "elapsed_time": "2 days, 0:15:31", "remaining_time": "15:33:02", "throughput": 5654.63, "total_tokens": 982385952}
+{"current_steps": 22470, "total_steps": 29704, "loss": 0.27708570957183837, "lr": 1.39377173305276e-06, "epoch": 1.5129275518448693, "percentage": 75.65, "elapsed_time": "2 days, 0:16:09", "remaining_time": "15:32:23", "throughput": 5654.59, "total_tokens": 982592624}
+{"current_steps": 22475, "total_steps": 29704, "loss": 0.3209893465042114, "lr": 1.3919407362158254e-06, "epoch": 1.5132642068408295, "percentage": 75.66, "elapsed_time": "2 days, 0:16:46", "remaining_time": "15:31:44", "throughput": 5654.72, "total_tokens": 982825112}
+{"current_steps": 22480, "total_steps": 29704, "loss": 0.2871732711791992, "lr": 1.3901107483611582e-06, "epoch": 1.5136008618367898, "percentage": 75.68, "elapsed_time": "2 days, 0:17:24", "remaining_time": "15:31:05", "throughput": 5654.66, "total_tokens": 983031648}
+{"current_steps": 22485, "total_steps": 29704, "loss": 0.2984759330749512, "lr": 1.3882817700005091e-06, "epoch": 1.5139375168327498, "percentage": 75.7, "elapsed_time": "2 days, 0:18:12", "remaining_time": "15:30:29", "throughput": 5654.34, "total_tokens": 983247128}
+{"current_steps": 22490, "total_steps": 29704, "loss": 0.3003609418869019, "lr": 1.3864538016453466e-06, "epoch": 1.5142741718287098, "percentage": 75.71, "elapsed_time": "2 days, 0:18:49", "remaining_time": "15:29:50", "throughput": 5654.29, "total_tokens": 983445264}
+{"current_steps": 22495, "total_steps": 29704, "loss": 0.304855489730835, "lr": 1.384626843806855e-06, "epoch": 1.51461082682467, "percentage": 75.73, "elapsed_time": "2 days, 0:19:29", "remaining_time": "15:29:12", "throughput": 5654.14, "total_tokens": 983646472}
+{"current_steps": 22500, "total_steps": 29704, "loss": 0.31099758148193357, "lr": 1.3828008969959367e-06, "epoch": 1.5149474818206303, "percentage": 75.75, "elapsed_time": "2 days, 0:20:09", "remaining_time": "15:28:33", "throughput": 5654.16, "total_tokens": 983875960}
+{"current_steps": 22505, "total_steps": 29704, "loss": 0.2885129928588867, "lr": 1.3809759617232166e-06, "epoch": 1.5152841368165904, "percentage": 75.76, "elapsed_time": "2 days, 0:20:50", "remaining_time": "15:27:56", "throughput": 5654.05, "total_tokens": 984089952}
+{"current_steps": 22510, "total_steps": 29704, "loss": 0.29050498008728026, "lr": 1.3791520384990247e-06, "epoch": 1.5156207918125504, "percentage": 75.78, "elapsed_time": "2 days, 0:21:25", "remaining_time": "15:27:16", "throughput": 5654.14, "total_tokens": 984303984}
+{"current_steps": 22515, "total_steps": 29704, "loss": 0.2980940818786621, "lr": 1.3773291278334216e-06, "epoch": 1.5159574468085106, "percentage": 75.8, "elapsed_time": "2 days, 0:22:02", "remaining_time": "15:26:37", "throughput": 5654.17, "total_tokens": 984518832}
+{"current_steps": 22520, "total_steps": 29704, "loss": 0.3086568355560303, "lr": 1.3755072302361754e-06, "epoch": 1.5162941018044709, "percentage": 75.81, "elapsed_time": "2 days, 0:22:43", "remaining_time": "15:25:59", "throughput": 5654.18, "total_tokens": 984751312}
+{"current_steps": 22525, "total_steps": 29704, "loss": 0.28640379905700686, "lr": 1.3736863462167749e-06, "epoch": 1.516630756800431, "percentage": 75.83, "elapsed_time": "2 days, 0:23:23", "remaining_time": "15:25:20", "throughput": 5654.03, "total_tokens": 984953136}
+{"current_steps": 22530, "total_steps": 29704, "loss": 0.28820300102233887, "lr": 1.3718664762844242e-06, "epoch": 1.516967411796391, "percentage": 75.85, "elapsed_time": "2 days, 0:23:58", "remaining_time": "15:24:40", "throughput": 5654.12, "total_tokens": 985162064}
+{"current_steps": 22535, "total_steps": 29704, "loss": 0.2942743062973022, "lr": 1.370047620948044e-06, "epoch": 1.5173040667923512, "percentage": 75.87, "elapsed_time": "2 days, 0:24:39", "remaining_time": "15:24:02", "throughput": 5654.05, "total_tokens": 985382480}
+{"current_steps": 22540, "total_steps": 29704, "loss": 0.3100618839263916, "lr": 1.368229780716272e-06, "epoch": 1.5176407217883114, "percentage": 75.88, "elapsed_time": "2 days, 0:25:10", "remaining_time": "15:23:21", "throughput": 5654.29, "total_tokens": 985602688}
+{"current_steps": 22545, "total_steps": 29704, "loss": 0.2899731397628784, "lr": 1.3664129560974604e-06, "epoch": 1.5179773767842715, "percentage": 75.9, "elapsed_time": "2 days, 0:25:46", "remaining_time": "15:22:42", "throughput": 5654.36, "total_tokens": 985820032}
+{"current_steps": 22550, "total_steps": 29704, "loss": 0.29310288429260256, "lr": 1.3645971475996777e-06, "epoch": 1.5183140317802315, "percentage": 75.92, "elapsed_time": "2 days, 0:26:22", "remaining_time": "15:22:02", "throughput": 5654.38, "total_tokens": 986025552}
+{"current_steps": 22555, "total_steps": 29704, "loss": 0.29608545303344724, "lr": 1.3627823557307135e-06, "epoch": 1.5186506867761917, "percentage": 75.93, "elapsed_time": "2 days, 0:27:04", "remaining_time": "15:21:25", "throughput": 5654.36, "total_tokens": 986257600}
+{"current_steps": 22560, "total_steps": 29704, "loss": 0.3058366298675537, "lr": 1.3609685809980616e-06, "epoch": 1.518987341772152, "percentage": 75.95, "elapsed_time": "2 days, 0:27:46", "remaining_time": "15:20:47", "throughput": 5654.3, "total_tokens": 986486440}
+{"current_steps": 22565, "total_steps": 29704, "loss": 0.2807319164276123, "lr": 1.3591558239089443e-06, "epoch": 1.519323996768112, "percentage": 75.97, "elapsed_time": "2 days, 0:28:32", "remaining_time": "15:20:11", "throughput": 5654.08, "total_tokens": 986707576}
+{"current_steps": 22570, "total_steps": 29704, "loss": 0.317004919052124, "lr": 1.3573440849702902e-06, "epoch": 1.519660651764072, "percentage": 75.98, "elapsed_time": "2 days, 0:29:14", "remaining_time": "15:19:33", "throughput": 5654.05, "total_tokens": 986937912}
+{"current_steps": 22575, "total_steps": 29704, "loss": 0.3014618635177612, "lr": 1.3555333646887476e-06, "epoch": 1.5199973067600323, "percentage": 76.0, "elapsed_time": "2 days, 0:29:49", "remaining_time": "15:18:54", "throughput": 5654.19, "total_tokens": 987163824}
+{"current_steps": 22580, "total_steps": 29704, "loss": 0.29477739334106445, "lr": 1.3537236635706779e-06, "epoch": 1.5203339617559926, "percentage": 76.02, "elapsed_time": "2 days, 0:30:23", "remaining_time": "15:18:13", "throughput": 5654.44, "total_tokens": 987396616}
+{"current_steps": 22585, "total_steps": 29704, "loss": 0.27982220649719236, "lr": 1.351914982122159e-06, "epoch": 1.5206706167519526, "percentage": 76.03, "elapsed_time": "2 days, 0:30:59", "remaining_time": "15:17:34", "throughput": 5654.56, "total_tokens": 987623048}
+{"current_steps": 22590, "total_steps": 29704, "loss": 0.3012231349945068, "lr": 1.3501073208489823e-06, "epoch": 1.5210072717479126, "percentage": 76.05, "elapsed_time": "2 days, 0:31:36", "remaining_time": "15:16:55", "throughput": 5654.61, "total_tokens": 987842064}
+{"current_steps": 22595, "total_steps": 29704, "loss": 0.29522716999053955, "lr": 1.3483006802566546e-06, "epoch": 1.5213439267438729, "percentage": 76.07, "elapsed_time": "2 days, 0:32:13", "remaining_time": "15:16:15", "throughput": 5654.71, "total_tokens": 988065680}
+{"current_steps": 22600, "total_steps": 29704, "loss": 0.3093279838562012, "lr": 1.3464950608503957e-06, "epoch": 1.5216805817398331, "percentage": 76.08, "elapsed_time": "2 days, 0:32:52", "remaining_time": "15:15:37", "throughput": 5654.68, "total_tokens": 988279680}
+{"current_steps": 22605, "total_steps": 29704, "loss": 0.30634384155273436, "lr": 1.344690463135146e-06, "epoch": 1.5220172367357931, "percentage": 76.1, "elapsed_time": "2 days, 0:33:37", "remaining_time": "15:15:00", "throughput": 5654.45, "total_tokens": 988495592}
+{"current_steps": 22610, "total_steps": 29704, "loss": 0.3291673183441162, "lr": 1.3428868876155493e-06, "epoch": 1.5223538917317532, "percentage": 76.12, "elapsed_time": "2 days, 0:34:13", "remaining_time": "15:14:21", "throughput": 5654.43, "total_tokens": 988698000}
+{"current_steps": 22615, "total_steps": 29704, "loss": 0.2833380699157715, "lr": 1.3410843347959745e-06, "epoch": 1.5226905467277134, "percentage": 76.13, "elapsed_time": "2 days, 0:34:51", "remaining_time": "15:13:42", "throughput": 5654.44, "total_tokens": 988914088}
+{"current_steps": 22620, "total_steps": 29704, "loss": 0.3043062686920166, "lr": 1.3392828051804979e-06, "epoch": 1.5230272017236737, "percentage": 76.15, "elapsed_time": "2 days, 0:35:34", "remaining_time": "15:13:04", "throughput": 5654.35, "total_tokens": 989141848}
+{"current_steps": 22625, "total_steps": 29704, "loss": 0.3176832437515259, "lr": 1.3374822992729114e-06, "epoch": 1.5233638567196337, "percentage": 76.17, "elapsed_time": "2 days, 0:36:17", "remaining_time": "15:12:27", "throughput": 5654.22, "total_tokens": 989359560}
+{"current_steps": 22630, "total_steps": 29704, "loss": 0.30327634811401366, "lr": 1.3356828175767212e-06, "epoch": 1.5237005117155937, "percentage": 76.19, "elapsed_time": "2 days, 0:36:56", "remaining_time": "15:11:48", "throughput": 5654.15, "total_tokens": 989566784}
+{"current_steps": 22635, "total_steps": 29704, "loss": 0.27550592422485354, "lr": 1.3338843605951462e-06, "epoch": 1.524037166711554, "percentage": 76.2, "elapsed_time": "2 days, 0:37:38", "remaining_time": "15:11:11", "throughput": 5654.09, "total_tokens": 989793688}
+{"current_steps": 22640, "total_steps": 29704, "loss": 0.2927609205245972, "lr": 1.33208692883112e-06, "epoch": 1.5243738217075142, "percentage": 76.22, "elapsed_time": "2 days, 0:38:19", "remaining_time": "15:10:33", "throughput": 5654.11, "total_tokens": 990030416}
+{"current_steps": 22645, "total_steps": 29704, "loss": 0.2735952377319336, "lr": 1.3302905227872876e-06, "epoch": 1.5247104767034743, "percentage": 76.24, "elapsed_time": "2 days, 0:38:58", "remaining_time": "15:09:54", "throughput": 5654.21, "total_tokens": 990266880}
+{"current_steps": 22650, "total_steps": 29704, "loss": 0.3053655862808228, "lr": 1.328495142966007e-06, "epoch": 1.5250471316994343, "percentage": 76.25, "elapsed_time": "2 days, 0:39:36", "remaining_time": "15:09:16", "throughput": 5654.18, "total_tokens": 990479584}
+{"current_steps": 22655, "total_steps": 29704, "loss": 0.29980125427246096, "lr": 1.3267007898693552e-06, "epoch": 1.5253837866953945, "percentage": 76.27, "elapsed_time": "2 days, 0:40:23", "remaining_time": "15:08:40", "throughput": 5653.99, "total_tokens": 990712176}
+{"current_steps": 22660, "total_steps": 29704, "loss": 0.3055927276611328, "lr": 1.324907463999111e-06, "epoch": 1.5257204416913548, "percentage": 76.29, "elapsed_time": "2 days, 0:41:01", "remaining_time": "15:08:00", "throughput": 5654.07, "total_tokens": 990938448}
+{"current_steps": 22665, "total_steps": 29704, "loss": 0.3090615510940552, "lr": 1.3231151658567776e-06, "epoch": 1.5260570966873148, "percentage": 76.3, "elapsed_time": "2 days, 0:41:34", "remaining_time": "15:07:20", "throughput": 5654.2, "total_tokens": 991152664}
+{"current_steps": 22670, "total_steps": 29704, "loss": 0.3019981861114502, "lr": 1.321323895943563e-06, "epoch": 1.5263937516832748, "percentage": 76.32, "elapsed_time": "2 days, 0:42:15", "remaining_time": "15:06:42", "throughput": 5654.14, "total_tokens": 991370424}
+{"current_steps": 22675, "total_steps": 29704, "loss": 0.31381664276123045, "lr": 1.3195336547603905e-06, "epoch": 1.526730406679235, "percentage": 76.34, "elapsed_time": "2 days, 0:42:58", "remaining_time": "15:06:05", "throughput": 5654.0, "total_tokens": 991588192}
+{"current_steps": 22680, "total_steps": 29704, "loss": 0.28606882095336916, "lr": 1.3177444428078951e-06, "epoch": 1.5270670616751953, "percentage": 76.35, "elapsed_time": "2 days, 0:43:33", "remaining_time": "15:05:25", "throughput": 5654.05, "total_tokens": 991798856}
+{"current_steps": 22685, "total_steps": 29704, "loss": 0.30185737609863283, "lr": 1.3159562605864245e-06, "epoch": 1.5274037166711554, "percentage": 76.37, "elapsed_time": "2 days, 0:44:17", "remaining_time": "15:04:48", "throughput": 5653.94, "total_tokens": 992026624}
+{"current_steps": 22690, "total_steps": 29704, "loss": 0.29338903427124025, "lr": 1.3141691085960385e-06, "epoch": 1.5277403716671154, "percentage": 76.39, "elapsed_time": "2 days, 0:44:56", "remaining_time": "15:04:10", "throughput": 5654.04, "total_tokens": 992265864}
+{"current_steps": 22695, "total_steps": 29704, "loss": 0.2825155258178711, "lr": 1.3123829873365073e-06, "epoch": 1.5280770266630757, "percentage": 76.4, "elapsed_time": "2 days, 0:45:35", "remaining_time": "15:03:31", "throughput": 5654.01, "total_tokens": 992478536}
+{"current_steps": 22700, "total_steps": 29704, "loss": 0.2889784097671509, "lr": 1.3105978973073136e-06, "epoch": 1.528413681659036, "percentage": 76.42, "elapsed_time": "2 days, 0:46:17", "remaining_time": "15:02:53", "throughput": 5653.87, "total_tokens": 992692016}
+{"current_steps": 22705, "total_steps": 29704, "loss": 0.3144156694412231, "lr": 1.3088138390076565e-06, "epoch": 1.528750336654996, "percentage": 76.44, "elapsed_time": "2 days, 0:46:50", "remaining_time": "15:02:13", "throughput": 5654.07, "total_tokens": 992915000}
+{"current_steps": 22710, "total_steps": 29704, "loss": 0.28853483200073243, "lr": 1.3070308129364357e-06, "epoch": 1.529086991650956, "percentage": 76.45, "elapsed_time": "2 days, 0:47:33", "remaining_time": "15:01:35", "throughput": 5653.93, "total_tokens": 993131448}
+{"current_steps": 22715, "total_steps": 29704, "loss": 0.28805036544799806, "lr": 1.3052488195922736e-06, "epoch": 1.5294236466469162, "percentage": 76.47, "elapsed_time": "2 days, 0:48:13", "remaining_time": "15:00:57", "throughput": 5653.81, "total_tokens": 993337608}
+{"current_steps": 22720, "total_steps": 29704, "loss": 0.2852388620376587, "lr": 1.3034678594734972e-06, "epoch": 1.5297603016428765, "percentage": 76.49, "elapsed_time": "2 days, 0:48:51", "remaining_time": "15:00:18", "throughput": 5653.85, "total_tokens": 993560896}
+{"current_steps": 22725, "total_steps": 29704, "loss": 0.30547394752502444, "lr": 1.3016879330781468e-06, "epoch": 1.5300969566388365, "percentage": 76.5, "elapsed_time": "2 days, 0:49:28", "remaining_time": "14:59:39", "throughput": 5653.83, "total_tokens": 993766904}
+{"current_steps": 22730, "total_steps": 29704, "loss": 0.29537832736968994, "lr": 1.2999090409039728e-06, "epoch": 1.5304336116347965, "percentage": 76.52, "elapsed_time": "2 days, 0:50:06", "remaining_time": "14:59:00", "throughput": 5653.89, "total_tokens": 993988440}
+{"current_steps": 22735, "total_steps": 29704, "loss": 0.30858845710754396, "lr": 1.2981311834484367e-06, "epoch": 1.5307702666307568, "percentage": 76.54, "elapsed_time": "2 days, 0:50:51", "remaining_time": "14:58:24", "throughput": 5653.66, "total_tokens": 994207440}
+{"current_steps": 22740, "total_steps": 29704, "loss": 0.28605356216430666, "lr": 1.296354361208711e-06, "epoch": 1.531106921626717, "percentage": 76.56, "elapsed_time": "2 days, 0:51:31", "remaining_time": "14:57:45", "throughput": 5653.61, "total_tokens": 994421664}
+{"current_steps": 22745, "total_steps": 29704, "loss": 0.3216970920562744, "lr": 1.294578574681678e-06, "epoch": 1.531443576622677, "percentage": 76.57, "elapsed_time": "2 days, 0:52:08", "remaining_time": "14:57:06", "throughput": 5653.59, "total_tokens": 994629248}
+{"current_steps": 22750, "total_steps": 29704, "loss": 0.28469209671020507, "lr": 1.292803824363929e-06, "epoch": 1.531780231618637, "percentage": 76.59, "elapsed_time": "2 days, 0:52:47", "remaining_time": "14:56:28", "throughput": 5653.61, "total_tokens": 994850920}
+{"current_steps": 22755, "total_steps": 29704, "loss": 0.3232132434844971, "lr": 1.2910301107517726e-06, "epoch": 1.5321168866145973, "percentage": 76.61, "elapsed_time": "2 days, 0:53:28", "remaining_time": "14:55:50", "throughput": 5653.56, "total_tokens": 995074552}
+{"current_steps": 22760, "total_steps": 29704, "loss": 0.319903826713562, "lr": 1.2892574343412156e-06, "epoch": 1.5324535416105576, "percentage": 76.62, "elapsed_time": "2 days, 0:54:03", "remaining_time": "14:55:10", "throughput": 5653.7, "total_tokens": 995294736}
+{"current_steps": 22765, "total_steps": 29704, "loss": 0.3038845777511597, "lr": 1.2874857956279858e-06, "epoch": 1.5327901966065176, "percentage": 76.64, "elapsed_time": "2 days, 0:54:41", "remaining_time": "14:54:31", "throughput": 5653.67, "total_tokens": 995505008}
+{"current_steps": 22770, "total_steps": 29704, "loss": 0.3022770404815674, "lr": 1.2857151951075147e-06, "epoch": 1.5331268516024776, "percentage": 76.66, "elapsed_time": "2 days, 0:55:20", "remaining_time": "14:53:52", "throughput": 5653.68, "total_tokens": 995729520}
+{"current_steps": 22775, "total_steps": 29704, "loss": 0.2998102903366089, "lr": 1.2839456332749456e-06, "epoch": 1.5334635065984379, "percentage": 76.67, "elapsed_time": "2 days, 0:55:58", "remaining_time": "14:53:13", "throughput": 5653.69, "total_tokens": 995943808}
+{"current_steps": 22780, "total_steps": 29704, "loss": 0.2987299919128418, "lr": 1.2821771106251308e-06, "epoch": 1.5338001615943981, "percentage": 76.69, "elapsed_time": "2 days, 0:56:34", "remaining_time": "14:52:34", "throughput": 5653.75, "total_tokens": 996160472}
+{"current_steps": 22785, "total_steps": 29704, "loss": 0.27775073051452637, "lr": 1.2804096276526312e-06, "epoch": 1.5341368165903582, "percentage": 76.71, "elapsed_time": "2 days, 0:57:11", "remaining_time": "14:51:55", "throughput": 5653.82, "total_tokens": 996378168}
+{"current_steps": 22790, "total_steps": 29704, "loss": 0.2855050086975098, "lr": 1.2786431848517184e-06, "epoch": 1.5344734715863182, "percentage": 76.72, "elapsed_time": "2 days, 0:57:47", "remaining_time": "14:51:15", "throughput": 5653.83, "total_tokens": 996584112}
+{"current_steps": 22795, "total_steps": 29704, "loss": 0.29274682998657225, "lr": 1.2768777827163725e-06, "epoch": 1.5348101265822784, "percentage": 76.74, "elapsed_time": "2 days, 0:58:25", "remaining_time": "14:50:36", "throughput": 5653.82, "total_tokens": 996797912}
+{"current_steps": 22800, "total_steps": 29704, "loss": 0.29671099185943606, "lr": 1.2751134217402811e-06, "epoch": 1.5351467815782387, "percentage": 76.76, "elapsed_time": "2 days, 0:59:07", "remaining_time": "14:49:59", "throughput": 5653.64, "total_tokens": 997003488}
+{"current_steps": 22805, "total_steps": 29704, "loss": 0.2848902940750122, "lr": 1.2733501024168465e-06, "epoch": 1.5354834365741987, "percentage": 76.77, "elapsed_time": "2 days, 0:59:41", "remaining_time": "14:49:19", "throughput": 5653.8, "total_tokens": 997225312}
+{"current_steps": 22810, "total_steps": 29704, "loss": 0.30264275074005126, "lr": 1.2715878252391695e-06, "epoch": 1.5358200915701588, "percentage": 76.79, "elapsed_time": "2 days, 1:00:21", "remaining_time": "14:48:40", "throughput": 5653.68, "total_tokens": 997430736}
+{"current_steps": 22815, "total_steps": 29704, "loss": 0.3026357412338257, "lr": 1.269826590700069e-06, "epoch": 1.536156746566119, "percentage": 76.81, "elapsed_time": "2 days, 1:00:54", "remaining_time": "14:48:00", "throughput": 5653.88, "total_tokens": 997650400}
+{"current_steps": 22820, "total_steps": 29704, "loss": 0.2799959659576416, "lr": 1.2680663992920684e-06, "epoch": 1.5364934015620793, "percentage": 76.82, "elapsed_time": "2 days, 1:01:35", "remaining_time": "14:47:22", "throughput": 5653.8, "total_tokens": 997872240}
+{"current_steps": 22825, "total_steps": 29704, "loss": 0.30673794746398925, "lr": 1.2663072515073988e-06, "epoch": 1.5368300565580393, "percentage": 76.84, "elapsed_time": "2 days, 1:02:15", "remaining_time": "14:46:44", "throughput": 5653.75, "total_tokens": 998084992}
+{"current_steps": 22830, "total_steps": 29704, "loss": 0.29927687644958495, "lr": 1.2645491478380002e-06, "epoch": 1.5371667115539993, "percentage": 76.86, "elapsed_time": "2 days, 1:02:57", "remaining_time": "14:46:06", "throughput": 5653.59, "total_tokens": 998297904}
+{"current_steps": 22835, "total_steps": 29704, "loss": 0.3054951190948486, "lr": 1.2627920887755212e-06, "epoch": 1.5375033665499596, "percentage": 76.88, "elapsed_time": "2 days, 1:03:33", "remaining_time": "14:45:27", "throughput": 5653.64, "total_tokens": 998508592}
+{"current_steps": 22840, "total_steps": 29704, "loss": 0.3088068962097168, "lr": 1.261036074811317e-06, "epoch": 1.5378400215459198, "percentage": 76.89, "elapsed_time": "2 days, 1:04:05", "remaining_time": "14:44:46", "throughput": 5653.89, "total_tokens": 998733664}
+{"current_steps": 22845, "total_steps": 29704, "loss": 0.2938983917236328, "lr": 1.2592811064364524e-06, "epoch": 1.5381766765418798, "percentage": 76.91, "elapsed_time": "2 days, 1:04:49", "remaining_time": "14:44:09", "throughput": 5653.74, "total_tokens": 998954872}
+{"current_steps": 22850, "total_steps": 29704, "loss": 0.3196579456329346, "lr": 1.2575271841416963e-06, "epoch": 1.5385133315378399, "percentage": 76.93, "elapsed_time": "2 days, 1:05:27", "remaining_time": "14:43:30", "throughput": 5653.6, "total_tokens": 999145416}
+{"current_steps": 22855, "total_steps": 29704, "loss": 0.27573988437652586, "lr": 1.2557743084175327e-06, "epoch": 1.5388499865338001, "percentage": 76.94, "elapsed_time": "2 days, 1:06:02", "remaining_time": "14:42:50", "throughput": 5653.69, "total_tokens": 999360320}
+{"current_steps": 22860, "total_steps": 29704, "loss": 0.2842655420303345, "lr": 1.254022479754141e-06, "epoch": 1.5391866415297604, "percentage": 76.96, "elapsed_time": "2 days, 1:06:37", "remaining_time": "14:42:10", "throughput": 5653.87, "total_tokens": 999588440}
+{"current_steps": 22865, "total_steps": 29704, "loss": 0.28894410133361814, "lr": 1.252271698641419e-06, "epoch": 1.5395232965257204, "percentage": 76.98, "elapsed_time": "2 days, 1:07:15", "remaining_time": "14:41:32", "throughput": 5653.85, "total_tokens": 999804216}
+{"current_steps": 22870, "total_steps": 29704, "loss": 0.310650110244751, "lr": 1.2505219655689659e-06, "epoch": 1.5398599515216804, "percentage": 76.99, "elapsed_time": "2 days, 1:07:48", "remaining_time": "14:40:51", "throughput": 5654.12, "total_tokens": 1000033680}
+{"current_steps": 22875, "total_steps": 29704, "loss": 0.3046865701675415, "lr": 1.2487732810260877e-06, "epoch": 1.5401966065176407, "percentage": 77.01, "elapsed_time": "2 days, 1:08:27", "remaining_time": "14:40:13", "throughput": 5654.09, "total_tokens": 1000253168}
+{"current_steps": 22880, "total_steps": 29704, "loss": 0.2814844846725464, "lr": 1.2470256455018027e-06, "epoch": 1.540533261513601, "percentage": 77.03, "elapsed_time": "2 days, 1:09:08", "remaining_time": "14:39:35", "throughput": 5654.0, "total_tokens": 1000469584}
+{"current_steps": 22885, "total_steps": 29704, "loss": 0.28022444248199463, "lr": 1.2452790594848257e-06, "epoch": 1.540869916509561, "percentage": 77.04, "elapsed_time": "2 days, 1:09:54", "remaining_time": "14:38:58", "throughput": 5653.74, "total_tokens": 1000680152}
+{"current_steps": 22890, "total_steps": 29704, "loss": 0.3237956762313843, "lr": 1.2435335234635892e-06, "epoch": 1.541206571505521, "percentage": 77.06, "elapsed_time": "2 days, 1:10:31", "remaining_time": "14:38:19", "throughput": 5653.77, "total_tokens": 1000898768}
+{"current_steps": 22895, "total_steps": 29704, "loss": 0.29699585437774656, "lr": 1.2417890379262221e-06, "epoch": 1.5415432265014812, "percentage": 77.08, "elapsed_time": "2 days, 1:11:17", "remaining_time": "14:37:42", "throughput": 5653.57, "total_tokens": 1001118440}
+{"current_steps": 22900, "total_steps": 29704, "loss": 0.26516332626342776, "lr": 1.2400456033605672e-06, "epoch": 1.5418798814974415, "percentage": 77.09, "elapsed_time": "2 days, 1:11:58", "remaining_time": "14:37:05", "throughput": 5653.45, "total_tokens": 1001331680}
+{"current_steps": 22905, "total_steps": 29704, "loss": 0.3010143756866455, "lr": 1.2383032202541706e-06, "epoch": 1.5422165364934015, "percentage": 77.11, "elapsed_time": "2 days, 1:12:36", "remaining_time": "14:36:26", "throughput": 5653.47, "total_tokens": 1001546592}
+{"current_steps": 22910, "total_steps": 29704, "loss": 0.32308430671691896, "lr": 1.2365618890942827e-06, "epoch": 1.5425531914893615, "percentage": 77.13, "elapsed_time": "2 days, 1:13:11", "remaining_time": "14:35:46", "throughput": 5653.58, "total_tokens": 1001766568}
+{"current_steps": 22915, "total_steps": 29704, "loss": 0.3221110820770264, "lr": 1.2348216103678618e-06, "epoch": 1.5428898464853218, "percentage": 77.14, "elapsed_time": "2 days, 1:13:42", "remaining_time": "14:35:05", "throughput": 5653.82, "total_tokens": 1001986120}
+{"current_steps": 22920, "total_steps": 29704, "loss": 0.2979999780654907, "lr": 1.2330823845615713e-06, "epoch": 1.543226501481282, "percentage": 77.16, "elapsed_time": "2 days, 1:14:20", "remaining_time": "14:34:26", "throughput": 5653.73, "total_tokens": 1002184920}
+{"current_steps": 22925, "total_steps": 29704, "loss": 0.29711506366729734, "lr": 1.2313442121617785e-06, "epoch": 1.543563156477242, "percentage": 77.18, "elapsed_time": "2 days, 1:15:01", "remaining_time": "14:33:48", "throughput": 5653.75, "total_tokens": 1002416768}
+{"current_steps": 22930, "total_steps": 29704, "loss": 0.2836009502410889, "lr": 1.229607093654563e-06, "epoch": 1.543899811473202, "percentage": 77.19, "elapsed_time": "2 days, 1:15:33", "remaining_time": "14:33:08", "throughput": 5654.0, "total_tokens": 1002644952}
+{"current_steps": 22935, "total_steps": 29704, "loss": 0.30063090324401853, "lr": 1.2278710295256974e-06, "epoch": 1.5442364664691623, "percentage": 77.21, "elapsed_time": "2 days, 1:16:08", "remaining_time": "14:32:28", "throughput": 5654.2, "total_tokens": 1002875328}
+{"current_steps": 22940, "total_steps": 29704, "loss": 0.2941617965698242, "lr": 1.2261360202606725e-06, "epoch": 1.5445731214651226, "percentage": 77.23, "elapsed_time": "2 days, 1:16:46", "remaining_time": "14:31:49", "throughput": 5654.3, "total_tokens": 1003108632}
+{"current_steps": 22945, "total_steps": 29704, "loss": 0.3176748752593994, "lr": 1.2244020663446731e-06, "epoch": 1.5449097764610826, "percentage": 77.25, "elapsed_time": "2 days, 1:17:29", "remaining_time": "14:31:11", "throughput": 5654.13, "total_tokens": 1003320896}
+{"current_steps": 22950, "total_steps": 29704, "loss": 0.29861812591552733, "lr": 1.222669168262598e-06, "epoch": 1.5452464314570427, "percentage": 77.26, "elapsed_time": "2 days, 1:18:06", "remaining_time": "14:30:32", "throughput": 5654.19, "total_tokens": 1003542056}
+{"current_steps": 22955, "total_steps": 29704, "loss": 0.29040777683258057, "lr": 1.2209373264990448e-06, "epoch": 1.545583086453003, "percentage": 77.28, "elapsed_time": "2 days, 1:18:39", "remaining_time": "14:29:52", "throughput": 5654.37, "total_tokens": 1003762224}
+{"current_steps": 22960, "total_steps": 29704, "loss": 0.29322617053985595, "lr": 1.2192065415383176e-06, "epoch": 1.5459197414489632, "percentage": 77.3, "elapsed_time": "2 days, 1:19:19", "remaining_time": "14:29:14", "throughput": 5654.31, "total_tokens": 1003976720}
+{"current_steps": 22965, "total_steps": 29704, "loss": 0.3234560966491699, "lr": 1.217476813864425e-06, "epoch": 1.5462563964449232, "percentage": 77.31, "elapsed_time": "2 days, 1:20:04", "remaining_time": "14:28:37", "throughput": 5654.18, "total_tokens": 1004209864}
+{"current_steps": 22970, "total_steps": 29704, "loss": 0.27009384632110595, "lr": 1.2157481439610801e-06, "epoch": 1.5465930514408834, "percentage": 77.33, "elapsed_time": "2 days, 1:20:47", "remaining_time": "14:28:00", "throughput": 5654.05, "total_tokens": 1004428792}
+{"current_steps": 22975, "total_steps": 29704, "loss": 0.2998237609863281, "lr": 1.2140205323116976e-06, "epoch": 1.5469297064368437, "percentage": 77.35, "elapsed_time": "2 days, 1:21:22", "remaining_time": "14:27:20", "throughput": 5654.07, "total_tokens": 1004626616}
+{"current_steps": 22980, "total_steps": 29704, "loss": 0.2867993116378784, "lr": 1.2122939793994043e-06, "epoch": 1.5472663614328037, "percentage": 77.36, "elapsed_time": "2 days, 1:21:56", "remaining_time": "14:26:40", "throughput": 5654.18, "total_tokens": 1004841312}
+{"current_steps": 22985, "total_steps": 29704, "loss": 0.292344331741333, "lr": 1.2105684857070182e-06, "epoch": 1.5476030164287637, "percentage": 77.38, "elapsed_time": "2 days, 1:22:34", "remaining_time": "14:26:01", "throughput": 5654.23, "total_tokens": 1005062880}
+{"current_steps": 22990, "total_steps": 29704, "loss": 0.30412988662719725, "lr": 1.2088440517170729e-06, "epoch": 1.547939671424724, "percentage": 77.4, "elapsed_time": "2 days, 1:23:10", "remaining_time": "14:25:21", "throughput": 5654.35, "total_tokens": 1005287360}
+{"current_steps": 22995, "total_steps": 29704, "loss": 0.3056650161743164, "lr": 1.2071206779117994e-06, "epoch": 1.5482763264206842, "percentage": 77.41, "elapsed_time": "2 days, 1:23:47", "remaining_time": "14:24:42", "throughput": 5654.41, "total_tokens": 1005509112}
+{"current_steps": 23000, "total_steps": 29704, "loss": 0.29947643280029296, "lr": 1.2053983647731337e-06, "epoch": 1.5486129814166443, "percentage": 77.43, "elapsed_time": "2 days, 1:24:22", "remaining_time": "14:24:03", "throughput": 5654.61, "total_tokens": 1005742432}
+{"current_steps": 23005, "total_steps": 29704, "loss": 0.3036951541900635, "lr": 1.2036771127827152e-06, "epoch": 1.5489496364126043, "percentage": 77.45, "elapsed_time": "2 days, 1:24:56", "remaining_time": "14:23:23", "throughput": 5654.67, "total_tokens": 1005948408}
+{"current_steps": 23010, "total_steps": 29704, "loss": 0.3014035224914551, "lr": 1.2019569224218869e-06, "epoch": 1.5492862914085646, "percentage": 77.46, "elapsed_time": "2 days, 1:25:40", "remaining_time": "14:22:45", "throughput": 5654.62, "total_tokens": 1006184672}
+{"current_steps": 23015, "total_steps": 29704, "loss": 0.27741825580596924, "lr": 1.2002377941716936e-06, "epoch": 1.5496229464045248, "percentage": 77.48, "elapsed_time": "2 days, 1:26:18", "remaining_time": "14:22:06", "throughput": 5654.62, "total_tokens": 1006397816}
+{"current_steps": 23020, "total_steps": 29704, "loss": 0.30047073364257815, "lr": 1.1985197285128853e-06, "epoch": 1.5499596014004848, "percentage": 77.5, "elapsed_time": "2 days, 1:27:04", "remaining_time": "14:21:30", "throughput": 5654.32, "total_tokens": 1006606872}
+{"current_steps": 23025, "total_steps": 29704, "loss": 0.29756712913513184, "lr": 1.1968027259259107e-06, "epoch": 1.5502962563964449, "percentage": 77.51, "elapsed_time": "2 days, 1:27:41", "remaining_time": "14:20:51", "throughput": 5654.28, "total_tokens": 1006810960}
+{"current_steps": 23030, "total_steps": 29704, "loss": 0.27481400966644287, "lr": 1.1950867868909293e-06, "epoch": 1.5506329113924051, "percentage": 77.53, "elapsed_time": "2 days, 1:28:24", "remaining_time": "14:20:13", "throughput": 5654.11, "total_tokens": 1007021520}
+{"current_steps": 23035, "total_steps": 29704, "loss": 0.30696120262146, "lr": 1.1933719118877923e-06, "epoch": 1.5509695663883654, "percentage": 77.55, "elapsed_time": "2 days, 1:28:54", "remaining_time": "14:19:32", "throughput": 5654.51, "total_tokens": 1007265144}
+{"current_steps": 23040, "total_steps": 29704, "loss": 0.2742566347122192, "lr": 1.1916581013960632e-06, "epoch": 1.5513062213843254, "percentage": 77.57, "elapsed_time": "2 days, 1:29:36", "remaining_time": "14:18:55", "throughput": 5654.53, "total_tokens": 1007502816}
+{"current_steps": 23045, "total_steps": 29704, "loss": 0.31151793003082273, "lr": 1.189945355895002e-06, "epoch": 1.5516428763802854, "percentage": 77.58, "elapsed_time": "2 days, 1:30:17", "remaining_time": "14:18:16", "throughput": 5654.44, "total_tokens": 1007718880}
+{"current_steps": 23050, "total_steps": 29704, "loss": 0.27026982307434083, "lr": 1.1882336758635727e-06, "epoch": 1.5519795313762457, "percentage": 77.6, "elapsed_time": "2 days, 1:30:52", "remaining_time": "14:17:37", "throughput": 5654.62, "total_tokens": 1007951160}
+{"current_steps": 23055, "total_steps": 29704, "loss": 0.32370221614837646, "lr": 1.1865230617804412e-06, "epoch": 1.552316186372206, "percentage": 77.62, "elapsed_time": "2 days, 1:31:28", "remaining_time": "14:16:57", "throughput": 5654.81, "total_tokens": 1008186432}
+{"current_steps": 23060, "total_steps": 29704, "loss": 0.2949348211288452, "lr": 1.1848135141239753e-06, "epoch": 1.552652841368166, "percentage": 77.63, "elapsed_time": "2 days, 1:32:09", "remaining_time": "14:16:19", "throughput": 5654.72, "total_tokens": 1008403400}
+{"current_steps": 23065, "total_steps": 29704, "loss": 0.3110039710998535, "lr": 1.1831050333722438e-06, "epoch": 1.552989496364126, "percentage": 77.65, "elapsed_time": "2 days, 1:32:44", "remaining_time": "14:15:40", "throughput": 5654.82, "total_tokens": 1008619128}
+{"current_steps": 23070, "total_steps": 29704, "loss": 0.29149832725524905, "lr": 1.1813976200030191e-06, "epoch": 1.5533261513600862, "percentage": 77.67, "elapsed_time": "2 days, 1:33:20", "remaining_time": "14:15:00", "throughput": 5654.89, "total_tokens": 1008834768}
+{"current_steps": 23075, "total_steps": 29704, "loss": 0.31109683513641356, "lr": 1.1796912744937717e-06, "epoch": 1.5536628063560465, "percentage": 77.68, "elapsed_time": "2 days, 1:33:57", "remaining_time": "14:14:21", "throughput": 5654.98, "total_tokens": 1009058816}
+{"current_steps": 23080, "total_steps": 29704, "loss": 0.2948103904724121, "lr": 1.1779859973216802e-06, "epoch": 1.5539994613520065, "percentage": 77.7, "elapsed_time": "2 days, 1:34:39", "remaining_time": "14:13:44", "throughput": 5654.82, "total_tokens": 1009270824}
+{"current_steps": 23085, "total_steps": 29704, "loss": 0.29932336807250975, "lr": 1.1762817889636142e-06, "epoch": 1.5543361163479665, "percentage": 77.72, "elapsed_time": "2 days, 1:35:18", "remaining_time": "14:13:05", "throughput": 5654.71, "total_tokens": 1009470192}
+{"current_steps": 23090, "total_steps": 29704, "loss": 0.31789186000823977, "lr": 1.1745786498961542e-06, "epoch": 1.5546727713439268, "percentage": 77.73, "elapsed_time": "2 days, 1:35:57", "remaining_time": "14:12:26", "throughput": 5654.8, "total_tokens": 1009708792}
+{"current_steps": 23095, "total_steps": 29704, "loss": 0.2769877195358276, "lr": 1.1728765805955766e-06, "epoch": 1.555009426339887, "percentage": 77.75, "elapsed_time": "2 days, 1:36:32", "remaining_time": "14:11:46", "throughput": 5654.91, "total_tokens": 1009922120}
+{"current_steps": 23100, "total_steps": 29704, "loss": 0.32232234477996824, "lr": 1.1711755815378595e-06, "epoch": 1.555346081335847, "percentage": 77.77, "elapsed_time": "2 days, 1:37:06", "remaining_time": "14:11:06", "throughput": 5655.0, "total_tokens": 1010130920}
+{"current_steps": 23105, "total_steps": 29704, "loss": 0.293909478187561, "lr": 1.1694756531986818e-06, "epoch": 1.555682736331807, "percentage": 77.78, "elapsed_time": "2 days, 1:37:50", "remaining_time": "14:10:29", "throughput": 5654.88, "total_tokens": 1010358016}
+{"current_steps": 23110, "total_steps": 29704, "loss": 0.3079185962677002, "lr": 1.1677767960534232e-06, "epoch": 1.5560193913277673, "percentage": 77.8, "elapsed_time": "2 days, 1:38:34", "remaining_time": "14:09:52", "throughput": 5654.7, "total_tokens": 1010578448}
+{"current_steps": 23115, "total_steps": 29704, "loss": 0.31560115814208983, "lr": 1.166079010577163e-06, "epoch": 1.5563560463237276, "percentage": 77.82, "elapsed_time": "2 days, 1:39:12", "remaining_time": "14:09:13", "throughput": 5654.68, "total_tokens": 1010787032}
+{"current_steps": 23120, "total_steps": 29704, "loss": 0.3215439796447754, "lr": 1.1643822972446823e-06, "epoch": 1.5566927013196876, "percentage": 77.83, "elapsed_time": "2 days, 1:39:46", "remaining_time": "14:08:33", "throughput": 5654.86, "total_tokens": 1011011240}
+{"current_steps": 23125, "total_steps": 29704, "loss": 0.2977125883102417, "lr": 1.1626866565304594e-06, "epoch": 1.5570293563156477, "percentage": 77.85, "elapsed_time": "2 days, 1:40:28", "remaining_time": "14:07:56", "throughput": 5654.82, "total_tokens": 1011244280}
+{"current_steps": 23130, "total_steps": 29704, "loss": 0.2924909830093384, "lr": 1.16099208890868e-06, "epoch": 1.557366011311608, "percentage": 77.87, "elapsed_time": "2 days, 1:41:07", "remaining_time": "14:07:17", "throughput": 5654.81, "total_tokens": 1011459816}
+{"current_steps": 23135, "total_steps": 29704, "loss": 0.29182374477386475, "lr": 1.1592985948532187e-06, "epoch": 1.5577026663075682, "percentage": 77.89, "elapsed_time": "2 days, 1:41:41", "remaining_time": "14:06:37", "throughput": 5654.9, "total_tokens": 1011668760}
+{"current_steps": 23140, "total_steps": 29704, "loss": 0.28834197521209715, "lr": 1.1576061748376594e-06, "epoch": 1.5580393213035282, "percentage": 77.9, "elapsed_time": "2 days, 1:42:21", "remaining_time": "14:05:59", "throughput": 5654.85, "total_tokens": 1011885552}
+{"current_steps": 23145, "total_steps": 29704, "loss": 0.293502950668335, "lr": 1.1559148293352805e-06, "epoch": 1.5583759762994882, "percentage": 77.92, "elapsed_time": "2 days, 1:43:01", "remaining_time": "14:05:21", "throughput": 5654.83, "total_tokens": 1012108496}
+{"current_steps": 23150, "total_steps": 29704, "loss": 0.2947830677032471, "lr": 1.154224558819062e-06, "epoch": 1.5587126312954485, "percentage": 77.94, "elapsed_time": "2 days, 1:43:41", "remaining_time": "14:04:42", "throughput": 5654.79, "total_tokens": 1012331792}
+{"current_steps": 23155, "total_steps": 29704, "loss": 0.29758148193359374, "lr": 1.1525353637616821e-06, "epoch": 1.5590492862914087, "percentage": 77.95, "elapsed_time": "2 days, 1:44:18", "remaining_time": "14:04:03", "throughput": 5654.87, "total_tokens": 1012552976}
+{"current_steps": 23160, "total_steps": 29704, "loss": 0.3053610324859619, "lr": 1.1508472446355196e-06, "epoch": 1.5593859412873687, "percentage": 77.97, "elapsed_time": "2 days, 1:44:59", "remaining_time": "14:03:25", "throughput": 5654.88, "total_tokens": 1012785584}
+{"current_steps": 23165, "total_steps": 29704, "loss": 0.2998149871826172, "lr": 1.1491602019126501e-06, "epoch": 1.5597225962833288, "percentage": 77.99, "elapsed_time": "2 days, 1:45:29", "remaining_time": "14:02:44", "throughput": 5655.23, "total_tokens": 1013018136}
+{"current_steps": 23170, "total_steps": 29704, "loss": 0.31585729122161865, "lr": 1.1474742360648515e-06, "epoch": 1.560059251279289, "percentage": 78.0, "elapsed_time": "2 days, 1:46:14", "remaining_time": "14:02:07", "throughput": 5655.07, "total_tokens": 1013244520}
+{"current_steps": 23175, "total_steps": 29704, "loss": 0.3203889846801758, "lr": 1.1457893475635968e-06, "epoch": 1.5603959062752493, "percentage": 78.02, "elapsed_time": "2 days, 1:46:59", "remaining_time": "14:01:30", "throughput": 5654.84, "total_tokens": 1013457472}
+{"current_steps": 23180, "total_steps": 29704, "loss": 0.2820396900177002, "lr": 1.144105536880063e-06, "epoch": 1.5607325612712093, "percentage": 78.04, "elapsed_time": "2 days, 1:47:32", "remaining_time": "14:00:50", "throughput": 5655.0, "total_tokens": 1013672336}
+{"current_steps": 23185, "total_steps": 29704, "loss": 0.29037587642669677, "lr": 1.1424228044851176e-06, "epoch": 1.5610692162671693, "percentage": 78.05, "elapsed_time": "2 days, 1:48:08", "remaining_time": "14:00:10", "throughput": 5654.92, "total_tokens": 1013859016}
+{"current_steps": 23190, "total_steps": 29704, "loss": 0.3020007133483887, "lr": 1.1407411508493355e-06, "epoch": 1.5614058712631296, "percentage": 78.07, "elapsed_time": "2 days, 1:48:45", "remaining_time": "13:59:32", "throughput": 5654.89, "total_tokens": 1014067352}
+{"current_steps": 23195, "total_steps": 29704, "loss": 0.2727111101150513, "lr": 1.1390605764429846e-06, "epoch": 1.5617425262590898, "percentage": 78.09, "elapsed_time": "2 days, 1:49:25", "remaining_time": "13:58:53", "throughput": 5654.82, "total_tokens": 1014277952}
+{"current_steps": 23200, "total_steps": 29704, "loss": 0.29233598709106445, "lr": 1.1373810817360314e-06, "epoch": 1.5620791812550499, "percentage": 78.1, "elapsed_time": "2 days, 1:50:03", "remaining_time": "13:58:14", "throughput": 5654.91, "total_tokens": 1014509144}
+{"current_steps": 23205, "total_steps": 29704, "loss": 0.3186174392700195, "lr": 1.135702667198142e-06, "epoch": 1.5624158362510099, "percentage": 78.12, "elapsed_time": "2 days, 1:50:36", "remaining_time": "13:57:34", "throughput": 5655.13, "total_tokens": 1014733760}
+{"current_steps": 23210, "total_steps": 29704, "loss": 0.2658507823944092, "lr": 1.134025333298679e-06, "epoch": 1.5627524912469701, "percentage": 78.14, "elapsed_time": "2 days, 1:51:13", "remaining_time": "13:56:55", "throughput": 5655.05, "total_tokens": 1014933448}
+{"current_steps": 23215, "total_steps": 29704, "loss": 0.27293944358825684, "lr": 1.1323490805067045e-06, "epoch": 1.5630891462429304, "percentage": 78.15, "elapsed_time": "2 days, 1:51:54", "remaining_time": "13:56:17", "throughput": 5654.99, "total_tokens": 1015151416}
+{"current_steps": 23220, "total_steps": 29704, "loss": 0.27212421894073485, "lr": 1.1306739092909757e-06, "epoch": 1.5634258012388904, "percentage": 78.17, "elapsed_time": "2 days, 1:52:38", "remaining_time": "13:55:40", "throughput": 5654.76, "total_tokens": 1015357104}
+{"current_steps": 23225, "total_steps": 29704, "loss": 0.28475046157836914, "lr": 1.1289998201199493e-06, "epoch": 1.5637624562348504, "percentage": 78.19, "elapsed_time": "2 days, 1:53:18", "remaining_time": "13:55:01", "throughput": 5654.72, "total_tokens": 1015578216}
+{"current_steps": 23230, "total_steps": 29704, "loss": 0.3049982786178589, "lr": 1.1273268134617816e-06, "epoch": 1.5640991112308107, "percentage": 78.2, "elapsed_time": "2 days, 1:53:57", "remaining_time": "13:54:23", "throughput": 5654.75, "total_tokens": 1015803864}
+{"current_steps": 23235, "total_steps": 29704, "loss": 0.2998204708099365, "lr": 1.1256548897843189e-06, "epoch": 1.564435766226771, "percentage": 78.22, "elapsed_time": "2 days, 1:54:36", "remaining_time": "13:53:44", "throughput": 5654.8, "total_tokens": 1016035144}
+{"current_steps": 23240, "total_steps": 29704, "loss": 0.2948010921478271, "lr": 1.1239840495551136e-06, "epoch": 1.564772421222731, "percentage": 78.24, "elapsed_time": "2 days, 1:55:16", "remaining_time": "13:53:06", "throughput": 5654.78, "total_tokens": 1016256464}
+{"current_steps": 23245, "total_steps": 29704, "loss": 0.3241511583328247, "lr": 1.1223142932414087e-06, "epoch": 1.565109076218691, "percentage": 78.26, "elapsed_time": "2 days, 1:55:58", "remaining_time": "13:52:28", "throughput": 5654.72, "total_tokens": 1016483368}
+{"current_steps": 23250, "total_steps": 29704, "loss": 0.2887975454330444, "lr": 1.120645621310147e-06, "epoch": 1.5654457312146512, "percentage": 78.27, "elapsed_time": "2 days, 1:56:40", "remaining_time": "13:51:51", "throughput": 5654.59, "total_tokens": 1016700976}
+{"current_steps": 23255, "total_steps": 29704, "loss": 0.2927762508392334, "lr": 1.1189780342279666e-06, "epoch": 1.5657823862106115, "percentage": 78.29, "elapsed_time": "2 days, 1:57:15", "remaining_time": "13:51:11", "throughput": 5654.65, "total_tokens": 1016907680}
+{"current_steps": 23260, "total_steps": 29704, "loss": 0.30506865978240966, "lr": 1.1173115324612033e-06, "epoch": 1.5661190412065715, "percentage": 78.31, "elapsed_time": "2 days, 1:57:56", "remaining_time": "13:50:33", "throughput": 5654.49, "total_tokens": 1017110400}
+{"current_steps": 23265, "total_steps": 29704, "loss": 0.3330164909362793, "lr": 1.1156461164758891e-06, "epoch": 1.5664556962025316, "percentage": 78.32, "elapsed_time": "2 days, 1:58:33", "remaining_time": "13:49:54", "throughput": 5654.59, "total_tokens": 1017339120}
+{"current_steps": 23270, "total_steps": 29704, "loss": 0.2630832433700562, "lr": 1.1139817867377522e-06, "epoch": 1.5667923511984918, "percentage": 78.34, "elapsed_time": "2 days, 1:59:11", "remaining_time": "13:49:15", "throughput": 5654.67, "total_tokens": 1017568520}
+{"current_steps": 23275, "total_steps": 29704, "loss": 0.29378499984741213, "lr": 1.1123185437122153e-06, "epoch": 1.567129006194452, "percentage": 78.36, "elapsed_time": "2 days, 1:59:55", "remaining_time": "13:48:38", "throughput": 5654.58, "total_tokens": 1017797256}
+{"current_steps": 23280, "total_steps": 29704, "loss": 0.30549144744873047, "lr": 1.1106563878644038e-06, "epoch": 1.567465661190412, "percentage": 78.37, "elapsed_time": "2 days, 2:00:35", "remaining_time": "13:48:00", "throughput": 5654.56, "total_tokens": 1018024880}
+{"current_steps": 23285, "total_steps": 29704, "loss": 0.30190200805664064, "lr": 1.1089953196591286e-06, "epoch": 1.5678023161863721, "percentage": 78.39, "elapsed_time": "2 days, 2:01:17", "remaining_time": "13:47:22", "throughput": 5654.45, "total_tokens": 1018241712}
+{"current_steps": 23290, "total_steps": 29704, "loss": 0.3117091417312622, "lr": 1.107335339560906e-06, "epoch": 1.5681389711823324, "percentage": 78.41, "elapsed_time": "2 days, 2:01:55", "remaining_time": "13:46:43", "throughput": 5654.42, "total_tokens": 1018448032}
+{"current_steps": 23295, "total_steps": 29704, "loss": 0.3014981269836426, "lr": 1.1056764480339427e-06, "epoch": 1.5684756261782926, "percentage": 78.42, "elapsed_time": "2 days, 2:02:38", "remaining_time": "13:46:05", "throughput": 5654.32, "total_tokens": 1018674080}
+{"current_steps": 23300, "total_steps": 29704, "loss": 0.33310580253601074, "lr": 1.1040186455421425e-06, "epoch": 1.5688122811742526, "percentage": 78.44, "elapsed_time": "2 days, 2:03:14", "remaining_time": "13:45:26", "throughput": 5654.48, "total_tokens": 1018908720}
+{"current_steps": 23305, "total_steps": 29704, "loss": 0.310105037689209, "lr": 1.1023619325491052e-06, "epoch": 1.5691489361702127, "percentage": 78.46, "elapsed_time": "2 days, 2:03:54", "remaining_time": "13:44:48", "throughput": 5654.44, "total_tokens": 1019122152}
+{"current_steps": 23310, "total_steps": 29704, "loss": 0.27553586959838866, "lr": 1.1007063095181248e-06, "epoch": 1.569485591166173, "percentage": 78.47, "elapsed_time": "2 days, 2:04:38", "remaining_time": "13:44:11", "throughput": 5654.26, "total_tokens": 1019342832}
+{"current_steps": 23315, "total_steps": 29704, "loss": 0.29708888530731203, "lr": 1.0990517769121905e-06, "epoch": 1.5698222461621332, "percentage": 78.49, "elapsed_time": "2 days, 2:05:14", "remaining_time": "13:43:31", "throughput": 5654.27, "total_tokens": 1019546240}
+{"current_steps": 23320, "total_steps": 29704, "loss": 0.3073767900466919, "lr": 1.0973983351939876e-06, "epoch": 1.5701589011580932, "percentage": 78.51, "elapsed_time": "2 days, 2:05:52", "remaining_time": "13:42:52", "throughput": 5654.31, "total_tokens": 1019767384}
+{"current_steps": 23325, "total_steps": 29704, "loss": 0.2895050048828125, "lr": 1.095745984825895e-06, "epoch": 1.5704955561540532, "percentage": 78.52, "elapsed_time": "2 days, 2:06:26", "remaining_time": "13:42:12", "throughput": 5654.52, "total_tokens": 1019996880}
+{"current_steps": 23330, "total_steps": 29704, "loss": 0.30101957321166994, "lr": 1.0940947262699902e-06, "epoch": 1.5708322111500135, "percentage": 78.54, "elapsed_time": "2 days, 2:07:08", "remaining_time": "13:41:35", "throughput": 5654.31, "total_tokens": 1020200320}
+{"current_steps": 23335, "total_steps": 29704, "loss": 0.28864340782165526, "lr": 1.092444559988038e-06, "epoch": 1.5711688661459737, "percentage": 78.56, "elapsed_time": "2 days, 2:07:46", "remaining_time": "13:40:56", "throughput": 5654.34, "total_tokens": 1020417768}
+{"current_steps": 23340, "total_steps": 29704, "loss": 0.29140138626098633, "lr": 1.0907954864415048e-06, "epoch": 1.5715055211419338, "percentage": 78.58, "elapsed_time": "2 days, 2:08:27", "remaining_time": "13:40:18", "throughput": 5654.3, "total_tokens": 1020645240}
+{"current_steps": 23345, "total_steps": 29704, "loss": 0.28663978576660154, "lr": 1.089147506091549e-06, "epoch": 1.5718421761378938, "percentage": 78.59, "elapsed_time": "2 days, 2:09:02", "remaining_time": "13:39:38", "throughput": 5654.42, "total_tokens": 1020864992}
+{"current_steps": 23350, "total_steps": 29704, "loss": 0.3154577732086182, "lr": 1.0875006193990217e-06, "epoch": 1.572178831133854, "percentage": 78.61, "elapsed_time": "2 days, 2:09:44", "remaining_time": "13:39:00", "throughput": 5654.3, "total_tokens": 1021076760}
+{"current_steps": 23355, "total_steps": 29704, "loss": 0.2915508270263672, "lr": 1.0858548268244706e-06, "epoch": 1.5725154861298143, "percentage": 78.63, "elapsed_time": "2 days, 2:10:23", "remaining_time": "13:38:21", "throughput": 5654.29, "total_tokens": 1021295960}
+{"current_steps": 23360, "total_steps": 29704, "loss": 0.293584418296814, "lr": 1.0842101288281359e-06, "epoch": 1.5728521411257743, "percentage": 78.64, "elapsed_time": "2 days, 2:11:02", "remaining_time": "13:37:43", "throughput": 5654.27, "total_tokens": 1021516488}
+{"current_steps": 23365, "total_steps": 29704, "loss": 0.28592419624328613, "lr": 1.082566525869952e-06, "epoch": 1.5731887961217343, "percentage": 78.66, "elapsed_time": "2 days, 2:11:42", "remaining_time": "13:37:05", "throughput": 5654.29, "total_tokens": 1021745864}
+{"current_steps": 23370, "total_steps": 29704, "loss": 0.3062607288360596, "lr": 1.0809240184095476e-06, "epoch": 1.5735254511176946, "percentage": 78.68, "elapsed_time": "2 days, 2:12:17", "remaining_time": "13:36:25", "throughput": 5654.4, "total_tokens": 1021959216}
+{"current_steps": 23375, "total_steps": 29704, "loss": 0.26804351806640625, "lr": 1.0792826069062429e-06, "epoch": 1.5738621061136548, "percentage": 78.69, "elapsed_time": "2 days, 2:12:55", "remaining_time": "13:35:46", "throughput": 5654.42, "total_tokens": 1022178816}
+{"current_steps": 23380, "total_steps": 29704, "loss": 0.31443793773651124, "lr": 1.0776422918190576e-06, "epoch": 1.5741987611096149, "percentage": 78.71, "elapsed_time": "2 days, 2:13:26", "remaining_time": "13:35:05", "throughput": 5654.69, "total_tokens": 1022402224}
+{"current_steps": 23385, "total_steps": 29704, "loss": 0.28395633697509765, "lr": 1.0760030736066952e-06, "epoch": 1.574535416105575, "percentage": 78.73, "elapsed_time": "2 days, 2:14:06", "remaining_time": "13:34:27", "throughput": 5654.56, "total_tokens": 1022607104}
+{"current_steps": 23390, "total_steps": 29704, "loss": 0.3159285068511963, "lr": 1.0743649527275619e-06, "epoch": 1.5748720711015352, "percentage": 78.74, "elapsed_time": "2 days, 2:14:48", "remaining_time": "13:33:49", "throughput": 5654.51, "total_tokens": 1022834416}
+{"current_steps": 23395, "total_steps": 29704, "loss": 0.3103215456008911, "lr": 1.0727279296397514e-06, "epoch": 1.5752087260974954, "percentage": 78.76, "elapsed_time": "2 days, 2:15:32", "remaining_time": "13:33:12", "throughput": 5654.4, "total_tokens": 1023064384}
+{"current_steps": 23400, "total_steps": 29704, "loss": 0.3091814279556274, "lr": 1.0710920048010526e-06, "epoch": 1.5755453810934554, "percentage": 78.78, "elapsed_time": "2 days, 2:16:13", "remaining_time": "13:32:34", "throughput": 5654.33, "total_tokens": 1023286312}
+{"current_steps": 23405, "total_steps": 29704, "loss": 0.29776315689086913, "lr": 1.0694571786689462e-06, "epoch": 1.5758820360894155, "percentage": 78.79, "elapsed_time": "2 days, 2:16:52", "remaining_time": "13:31:56", "throughput": 5654.45, "total_tokens": 1023528528}
+{"current_steps": 23410, "total_steps": 29704, "loss": 0.29261269569396975, "lr": 1.067823451700606e-06, "epoch": 1.5762186910853757, "percentage": 78.81, "elapsed_time": "2 days, 2:17:27", "remaining_time": "13:31:16", "throughput": 5654.58, "total_tokens": 1023748656}
+{"current_steps": 23415, "total_steps": 29704, "loss": 0.29994802474975585, "lr": 1.066190824352899e-06, "epoch": 1.576555346081336, "percentage": 78.83, "elapsed_time": "2 days, 2:18:07", "remaining_time": "13:30:38", "throughput": 5654.58, "total_tokens": 1023974480}
+{"current_steps": 23420, "total_steps": 29704, "loss": 0.27131869792938235, "lr": 1.0645592970823837e-06, "epoch": 1.576892001077296, "percentage": 78.84, "elapsed_time": "2 days, 2:18:43", "remaining_time": "13:29:58", "throughput": 5654.58, "total_tokens": 1024177832}
+{"current_steps": 23425, "total_steps": 29704, "loss": 0.3091407299041748, "lr": 1.0629288703453105e-06, "epoch": 1.577228656073256, "percentage": 78.86, "elapsed_time": "2 days, 2:19:23", "remaining_time": "13:29:20", "throughput": 5654.57, "total_tokens": 1024403608}
+{"current_steps": 23430, "total_steps": 29704, "loss": 0.29852406978607177, "lr": 1.061299544597627e-06, "epoch": 1.5775653110692163, "percentage": 78.88, "elapsed_time": "2 days, 2:20:01", "remaining_time": "13:28:41", "throughput": 5654.51, "total_tokens": 1024608360}
+{"current_steps": 23435, "total_steps": 29704, "loss": 0.3082388162612915, "lr": 1.0596713202949637e-06, "epoch": 1.5779019660651765, "percentage": 78.9, "elapsed_time": "2 days, 2:20:44", "remaining_time": "13:28:03", "throughput": 5654.45, "total_tokens": 1024836472}
+{"current_steps": 23440, "total_steps": 29704, "loss": 0.3000252485275269, "lr": 1.0580441978926527e-06, "epoch": 1.5782386210611365, "percentage": 78.91, "elapsed_time": "2 days, 2:21:24", "remaining_time": "13:27:25", "throughput": 5654.37, "total_tokens": 1025047408}
+{"current_steps": 23445, "total_steps": 29704, "loss": 0.2908348083496094, "lr": 1.056418177845711e-06, "epoch": 1.5785752760570966, "percentage": 78.93, "elapsed_time": "2 days, 2:22:06", "remaining_time": "13:26:47", "throughput": 5654.18, "total_tokens": 1025252528}
+{"current_steps": 23450, "total_steps": 29704, "loss": 0.3012411594390869, "lr": 1.0547932606088506e-06, "epoch": 1.5789119310530568, "percentage": 78.95, "elapsed_time": "2 days, 2:22:45", "remaining_time": "13:26:09", "throughput": 5654.12, "total_tokens": 1025462120}
+{"current_steps": 23455, "total_steps": 29704, "loss": 0.29930336475372316, "lr": 1.053169446636475e-06, "epoch": 1.579248586049017, "percentage": 78.96, "elapsed_time": "2 days, 2:23:22", "remaining_time": "13:25:30", "throughput": 5654.14, "total_tokens": 1025676472}
+{"current_steps": 23460, "total_steps": 29704, "loss": 0.3104248046875, "lr": 1.0515467363826782e-06, "epoch": 1.579585241044977, "percentage": 78.98, "elapsed_time": "2 days, 2:23:56", "remaining_time": "13:24:50", "throughput": 5654.34, "total_tokens": 1025904840}
+{"current_steps": 23465, "total_steps": 29704, "loss": 0.2840702533721924, "lr": 1.0499251303012454e-06, "epoch": 1.5799218960409371, "percentage": 79.0, "elapsed_time": "2 days, 2:24:32", "remaining_time": "13:24:10", "throughput": 5654.45, "total_tokens": 1026129016}
+{"current_steps": 23470, "total_steps": 29704, "loss": 0.30331013202667234, "lr": 1.0483046288456544e-06, "epoch": 1.5802585510368974, "percentage": 79.01, "elapsed_time": "2 days, 2:25:08", "remaining_time": "13:23:31", "throughput": 5654.47, "total_tokens": 1026336600}
+{"current_steps": 23475, "total_steps": 29704, "loss": 0.30037317276000974, "lr": 1.0466852324690719e-06, "epoch": 1.5805952060328576, "percentage": 79.03, "elapsed_time": "2 days, 2:25:47", "remaining_time": "13:22:52", "throughput": 5654.44, "total_tokens": 1026549792}
+{"current_steps": 23480, "total_steps": 29704, "loss": 0.2942227363586426, "lr": 1.0450669416243596e-06, "epoch": 1.5809318610288177, "percentage": 79.05, "elapsed_time": "2 days, 2:26:25", "remaining_time": "13:22:14", "throughput": 5654.49, "total_tokens": 1026774584}
+{"current_steps": 23485, "total_steps": 29704, "loss": 0.3083514213562012, "lr": 1.0434497567640639e-06, "epoch": 1.5812685160247777, "percentage": 79.06, "elapsed_time": "2 days, 2:27:02", "remaining_time": "13:21:34", "throughput": 5654.57, "total_tokens": 1026996640}
+{"current_steps": 23490, "total_steps": 29704, "loss": 0.3067812204360962, "lr": 1.0418336783404282e-06, "epoch": 1.581605171020738, "percentage": 79.08, "elapsed_time": "2 days, 2:27:40", "remaining_time": "13:20:56", "throughput": 5654.59, "total_tokens": 1027217736}
+{"current_steps": 23495, "total_steps": 29704, "loss": 0.3050783395767212, "lr": 1.0402187068053825e-06, "epoch": 1.5819418260166982, "percentage": 79.1, "elapsed_time": "2 days, 2:28:14", "remaining_time": "13:20:16", "throughput": 5654.72, "total_tokens": 1027433032}
+{"current_steps": 23500, "total_steps": 29704, "loss": 0.28964962959289553, "lr": 1.0386048426105495e-06, "epoch": 1.5822784810126582, "percentage": 79.11, "elapsed_time": "2 days, 2:28:52", "remaining_time": "13:19:37", "throughput": 5654.74, "total_tokens": 1027649992}
+{"current_steps": 23505, "total_steps": 29704, "loss": 0.28251526355743406, "lr": 1.0369920862072396e-06, "epoch": 1.5826151360086183, "percentage": 79.13, "elapsed_time": "2 days, 2:29:36", "remaining_time": "13:19:00", "throughput": 5654.52, "total_tokens": 1027858824}
+{"current_steps": 23510, "total_steps": 29704, "loss": 0.28255496025085447, "lr": 1.0353804380464556e-06, "epoch": 1.5829517910045785, "percentage": 79.15, "elapsed_time": "2 days, 2:30:18", "remaining_time": "13:18:22", "throughput": 5654.42, "total_tokens": 1028079384}
+{"current_steps": 23515, "total_steps": 29704, "loss": 0.2897408246994019, "lr": 1.0337698985788903e-06, "epoch": 1.5832884460005388, "percentage": 79.16, "elapsed_time": "2 days, 2:30:59", "remaining_time": "13:17:44", "throughput": 5654.39, "total_tokens": 1028306808}
+{"current_steps": 23520, "total_steps": 29704, "loss": 0.27561049461364745, "lr": 1.0321604682549247e-06, "epoch": 1.5836251009964988, "percentage": 79.18, "elapsed_time": "2 days, 2:31:43", "remaining_time": "13:17:07", "throughput": 5654.31, "total_tokens": 1028540048}
+{"current_steps": 23525, "total_steps": 29704, "loss": 0.3138261318206787, "lr": 1.0305521475246311e-06, "epoch": 1.5839617559924588, "percentage": 79.2, "elapsed_time": "2 days, 2:32:20", "remaining_time": "13:16:28", "throughput": 5654.31, "total_tokens": 1028749872}
+{"current_steps": 23530, "total_steps": 29704, "loss": 0.31621651649475097, "lr": 1.028944936837774e-06, "epoch": 1.584298410988419, "percentage": 79.21, "elapsed_time": "2 days, 2:32:54", "remaining_time": "13:15:48", "throughput": 5654.36, "total_tokens": 1028951280}
+{"current_steps": 23535, "total_steps": 29704, "loss": 0.30298333168029784, "lr": 1.0273388366438003e-06, "epoch": 1.5846350659843793, "percentage": 79.23, "elapsed_time": "2 days, 2:33:34", "remaining_time": "13:15:09", "throughput": 5654.35, "total_tokens": 1029175832}
+{"current_steps": 23540, "total_steps": 29704, "loss": 0.2973572015762329, "lr": 1.0257338473918538e-06, "epoch": 1.5849717209803393, "percentage": 79.25, "elapsed_time": "2 days, 2:34:12", "remaining_time": "13:14:30", "throughput": 5654.41, "total_tokens": 1029399728}
+{"current_steps": 23545, "total_steps": 29704, "loss": 0.3015413761138916, "lr": 1.0241299695307644e-06, "epoch": 1.5853083759762994, "percentage": 79.27, "elapsed_time": "2 days, 2:34:53", "remaining_time": "13:13:52", "throughput": 5654.35, "total_tokens": 1029616800}
+{"current_steps": 23550, "total_steps": 29704, "loss": 0.282063889503479, "lr": 1.0225272035090506e-06, "epoch": 1.5856450309722596, "percentage": 79.28, "elapsed_time": "2 days, 2:35:28", "remaining_time": "13:13:13", "throughput": 5654.53, "total_tokens": 1029848952}
+{"current_steps": 23555, "total_steps": 29704, "loss": 0.30583910942077636, "lr": 1.0209255497749209e-06, "epoch": 1.5859816859682199, "percentage": 79.3, "elapsed_time": "2 days, 2:36:10", "remaining_time": "13:12:35", "throughput": 5654.43, "total_tokens": 1030073088}
+{"current_steps": 23560, "total_steps": 29704, "loss": 0.28136019706726073, "lr": 1.0193250087762735e-06, "epoch": 1.58631834096418, "percentage": 79.32, "elapsed_time": "2 days, 2:36:52", "remaining_time": "13:11:57", "throughput": 5654.35, "total_tokens": 1030294120}
+{"current_steps": 23565, "total_steps": 29704, "loss": 0.29512906074523926, "lr": 1.0177255809606934e-06, "epoch": 1.58665499596014, "percentage": 79.33, "elapsed_time": "2 days, 2:37:25", "remaining_time": "13:11:17", "throughput": 5654.56, "total_tokens": 1030518512}
+{"current_steps": 23570, "total_steps": 29704, "loss": 0.2986447334289551, "lr": 1.0161272667754562e-06, "epoch": 1.5869916509561002, "percentage": 79.35, "elapsed_time": "2 days, 2:38:11", "remaining_time": "13:10:40", "throughput": 5654.43, "total_tokens": 1030753488}
+{"current_steps": 23575, "total_steps": 29704, "loss": 0.2920704364776611, "lr": 1.014530066667524e-06, "epoch": 1.5873283059520604, "percentage": 79.37, "elapsed_time": "2 days, 2:38:58", "remaining_time": "13:10:04", "throughput": 5654.22, "total_tokens": 1030980864}
+{"current_steps": 23580, "total_steps": 29704, "loss": 0.26558356285095214, "lr": 1.0129339810835526e-06, "epoch": 1.5876649609480205, "percentage": 79.38, "elapsed_time": "2 days, 2:39:40", "remaining_time": "13:09:26", "throughput": 5654.04, "total_tokens": 1031184200}
+{"current_steps": 23585, "total_steps": 29704, "loss": 0.2880537986755371, "lr": 1.0113390104698767e-06, "epoch": 1.5880016159439805, "percentage": 79.4, "elapsed_time": "2 days, 2:40:15", "remaining_time": "13:08:46", "throughput": 5654.2, "total_tokens": 1031415416}
+{"current_steps": 23590, "total_steps": 29704, "loss": 0.32445645332336426, "lr": 1.009745155272529e-06, "epoch": 1.5883382709399407, "percentage": 79.42, "elapsed_time": "2 days, 2:40:52", "remaining_time": "13:08:07", "throughput": 5654.29, "total_tokens": 1031642064}
+{"current_steps": 23595, "total_steps": 29704, "loss": 0.30813934803009035, "lr": 1.0081524159372246e-06, "epoch": 1.588674925935901, "percentage": 79.43, "elapsed_time": "2 days, 2:41:29", "remaining_time": "13:07:28", "throughput": 5654.32, "total_tokens": 1031852072}
+{"current_steps": 23600, "total_steps": 29704, "loss": 0.2949239253997803, "lr": 1.0065607929093685e-06, "epoch": 1.589011580931861, "percentage": 79.45, "elapsed_time": "2 days, 2:42:10", "remaining_time": "13:06:50", "throughput": 5654.22, "total_tokens": 1032066320}
+{"current_steps": 23605, "total_steps": 29704, "loss": 0.2600874423980713, "lr": 1.0049702866340521e-06, "epoch": 1.589348235927821, "percentage": 79.47, "elapsed_time": "2 days, 2:42:51", "remaining_time": "13:06:12", "throughput": 5654.16, "total_tokens": 1032291736}
+{"current_steps": 23610, "total_steps": 29704, "loss": 0.28646035194396974, "lr": 1.0033808975560556e-06, "epoch": 1.5896848909237813, "percentage": 79.48, "elapsed_time": "2 days, 2:43:29", "remaining_time": "13:05:33", "throughput": 5654.17, "total_tokens": 1032506240}
+{"current_steps": 23615, "total_steps": 29704, "loss": 0.29119462966918946, "lr": 1.0017926261198473e-06, "epoch": 1.5900215459197415, "percentage": 79.5, "elapsed_time": "2 days, 2:44:09", "remaining_time": "13:04:55", "throughput": 5654.12, "total_tokens": 1032720936}
+{"current_steps": 23620, "total_steps": 29704, "loss": 0.30643343925476074, "lr": 1.0002054727695814e-06, "epoch": 1.5903582009157016, "percentage": 79.52, "elapsed_time": "2 days, 2:44:50", "remaining_time": "13:04:17", "throughput": 5653.99, "total_tokens": 1032929224}
+{"current_steps": 23625, "total_steps": 29704, "loss": 0.29436190128326417, "lr": 9.986194379490993e-07, "epoch": 1.5906948559116616, "percentage": 79.53, "elapsed_time": "2 days, 2:45:26", "remaining_time": "13:03:37", "throughput": 5654.02, "total_tokens": 1033140248}
+{"current_steps": 23630, "total_steps": 29704, "loss": 0.30518579483032227, "lr": 9.970345221019345e-07, "epoch": 1.5910315109076218, "percentage": 79.55, "elapsed_time": "2 days, 2:46:08", "remaining_time": "13:02:59", "throughput": 5653.9, "total_tokens": 1033354080}
+{"current_steps": 23635, "total_steps": 29704, "loss": 0.30224905014038084, "lr": 9.954507256712987e-07, "epoch": 1.591368165903582, "percentage": 79.57, "elapsed_time": "2 days, 2:46:44", "remaining_time": "13:02:20", "throughput": 5653.93, "total_tokens": 1033566992}
+{"current_steps": 23640, "total_steps": 29704, "loss": 0.3042686700820923, "lr": 9.938680491000991e-07, "epoch": 1.5917048208995421, "percentage": 79.59, "elapsed_time": "2 days, 2:47:27", "remaining_time": "13:01:43", "throughput": 5653.76, "total_tokens": 1033777624}
+{"current_steps": 23645, "total_steps": 29704, "loss": 0.2958357810974121, "lr": 9.92286492830925e-07, "epoch": 1.5920414758955022, "percentage": 79.6, "elapsed_time": "2 days, 2:48:07", "remaining_time": "13:01:04", "throughput": 5653.78, "total_tokens": 1034004184}
+{"current_steps": 23650, "total_steps": 29704, "loss": 0.31322293281555175, "lr": 9.907060573060535e-07, "epoch": 1.5923781308914624, "percentage": 79.62, "elapsed_time": "2 days, 2:48:41", "remaining_time": "13:00:24", "throughput": 5653.95, "total_tokens": 1034231736}
+{"current_steps": 23655, "total_steps": 29704, "loss": 0.3153421401977539, "lr": 9.891267429674484e-07, "epoch": 1.5927147858874227, "percentage": 79.64, "elapsed_time": "2 days, 2:49:20", "remaining_time": "12:59:46", "throughput": 5653.91, "total_tokens": 1034440472}
+{"current_steps": 23660, "total_steps": 29704, "loss": 0.28733248710632325, "lr": 9.8754855025676e-07, "epoch": 1.5930514408833827, "percentage": 79.65, "elapsed_time": "2 days, 2:49:56", "remaining_time": "12:59:06", "throughput": 5653.97, "total_tokens": 1034659216}
+{"current_steps": 23665, "total_steps": 29704, "loss": 0.29557249546051023, "lr": 9.85971479615324e-07, "epoch": 1.5933880958793427, "percentage": 79.67, "elapsed_time": "2 days, 2:50:37", "remaining_time": "12:58:28", "throughput": 5653.97, "total_tokens": 1034888536}
+{"current_steps": 23670, "total_steps": 29704, "loss": 0.28273298740386965, "lr": 9.843955314841647e-07, "epoch": 1.593724750875303, "percentage": 79.69, "elapsed_time": "2 days, 2:51:10", "remaining_time": "12:57:48", "throughput": 5654.17, "total_tokens": 1035112904}
+{"current_steps": 23675, "total_steps": 29704, "loss": 0.29895691871643065, "lr": 9.82820706303989e-07, "epoch": 1.5940614058712632, "percentage": 79.7, "elapsed_time": "2 days, 2:51:54", "remaining_time": "12:57:11", "throughput": 5654.06, "total_tokens": 1035338280}
+{"current_steps": 23680, "total_steps": 29704, "loss": 0.3054283618927002, "lr": 9.812470045151952e-07, "epoch": 1.5943980608672232, "percentage": 79.72, "elapsed_time": "2 days, 2:52:31", "remaining_time": "12:56:32", "throughput": 5654.01, "total_tokens": 1035542816}
+{"current_steps": 23685, "total_steps": 29704, "loss": 0.29663498401641847, "lr": 9.79674426557859e-07, "epoch": 1.5947347158631833, "percentage": 79.74, "elapsed_time": "2 days, 2:53:13", "remaining_time": "12:55:54", "throughput": 5653.86, "total_tokens": 1035752536}
+{"current_steps": 23690, "total_steps": 29704, "loss": 0.2824697732925415, "lr": 9.781029728717513e-07, "epoch": 1.5950713708591435, "percentage": 79.75, "elapsed_time": "2 days, 2:53:51", "remaining_time": "12:55:15", "throughput": 5653.87, "total_tokens": 1035967072}
+{"current_steps": 23695, "total_steps": 29704, "loss": 0.31726827621459963, "lr": 9.765326438963218e-07, "epoch": 1.5954080258551038, "percentage": 79.77, "elapsed_time": "2 days, 2:54:28", "remaining_time": "12:54:36", "throughput": 5653.9, "total_tokens": 1036183056}
+{"current_steps": 23700, "total_steps": 29704, "loss": 0.29088969230651857, "lr": 9.749634400707087e-07, "epoch": 1.5957446808510638, "percentage": 79.79, "elapsed_time": "2 days, 2:55:04", "remaining_time": "12:53:57", "throughput": 5653.92, "total_tokens": 1036386224}
+{"current_steps": 23705, "total_steps": 29704, "loss": 0.3045504570007324, "lr": 9.733953618337344e-07, "epoch": 1.5960813358470238, "percentage": 79.8, "elapsed_time": "2 days, 2:55:45", "remaining_time": "12:53:19", "throughput": 5653.75, "total_tokens": 1036590880}
+{"current_steps": 23710, "total_steps": 29704, "loss": 0.315799617767334, "lr": 9.718284096239077e-07, "epoch": 1.596417990842984, "percentage": 79.82, "elapsed_time": "2 days, 2:56:24", "remaining_time": "12:52:40", "throughput": 5653.86, "total_tokens": 1036831896}
+{"current_steps": 23715, "total_steps": 29704, "loss": 0.28366994857788086, "lr": 9.702625838794215e-07, "epoch": 1.5967546458389443, "percentage": 79.84, "elapsed_time": "2 days, 2:57:04", "remaining_time": "12:52:02", "throughput": 5653.84, "total_tokens": 1037051528}
+{"current_steps": 23720, "total_steps": 29704, "loss": 0.28309922218322753, "lr": 9.686978850381535e-07, "epoch": 1.5970913008349044, "percentage": 79.85, "elapsed_time": "2 days, 2:57:43", "remaining_time": "12:51:23", "throughput": 5653.86, "total_tokens": 1037274432}
+{"current_steps": 23725, "total_steps": 29704, "loss": 0.28775806427001954, "lr": 9.67134313537666e-07, "epoch": 1.5974279558308644, "percentage": 79.87, "elapsed_time": "2 days, 2:58:18", "remaining_time": "12:50:44", "throughput": 5653.92, "total_tokens": 1037487448}
+{"current_steps": 23730, "total_steps": 29704, "loss": 0.2938704013824463, "lr": 9.6557186981521e-07, "epoch": 1.5977646108268246, "percentage": 79.89, "elapsed_time": "2 days, 2:58:52", "remaining_time": "12:50:04", "throughput": 5654.1, "total_tokens": 1037711184}
+{"current_steps": 23735, "total_steps": 29704, "loss": 0.29409012794494627, "lr": 9.640105543077133e-07, "epoch": 1.5981012658227849, "percentage": 79.91, "elapsed_time": "2 days, 2:59:29", "remaining_time": "12:49:25", "throughput": 5654.14, "total_tokens": 1037930696}
+{"current_steps": 23740, "total_steps": 29704, "loss": 0.31141953468322753, "lr": 9.624503674517972e-07, "epoch": 1.598437920818745, "percentage": 79.92, "elapsed_time": "2 days, 3:00:16", "remaining_time": "12:48:48", "throughput": 5653.93, "total_tokens": 1038151728}
+{"current_steps": 23745, "total_steps": 29704, "loss": 0.2785954475402832, "lr": 9.608913096837603e-07, "epoch": 1.598774575814705, "percentage": 79.94, "elapsed_time": "2 days, 3:00:58", "remaining_time": "12:48:10", "throughput": 5653.84, "total_tokens": 1038375320}
+{"current_steps": 23750, "total_steps": 29704, "loss": 0.30640616416931155, "lr": 9.593333814395889e-07, "epoch": 1.5991112308106652, "percentage": 79.96, "elapsed_time": "2 days, 3:01:30", "remaining_time": "12:47:30", "throughput": 5653.97, "total_tokens": 1038579600}
+{"current_steps": 23755, "total_steps": 29704, "loss": 0.29427356719970704, "lr": 9.577765831549529e-07, "epoch": 1.5994478858066254, "percentage": 79.97, "elapsed_time": "2 days, 3:02:03", "remaining_time": "12:46:50", "throughput": 5654.16, "total_tokens": 1038803840}
+{"current_steps": 23760, "total_steps": 29704, "loss": 0.30373120307922363, "lr": 9.562209152652058e-07, "epoch": 1.5997845408025855, "percentage": 79.99, "elapsed_time": "2 days, 3:02:46", "remaining_time": "12:46:12", "throughput": 5653.99, "total_tokens": 1039013760}
+{"current_steps": 23765, "total_steps": 29704, "loss": 0.3227171182632446, "lr": 9.54666378205385e-07, "epoch": 1.6001211957985455, "percentage": 80.01, "elapsed_time": "2 days, 3:03:33", "remaining_time": "12:45:36", "throughput": 5653.78, "total_tokens": 1039242488}
+{"current_steps": 23770, "total_steps": 29704, "loss": 0.2936609983444214, "lr": 9.531129724102117e-07, "epoch": 1.6004578507945058, "percentage": 80.02, "elapsed_time": "2 days, 3:04:12", "remaining_time": "12:44:57", "throughput": 5653.75, "total_tokens": 1039454848}
+{"current_steps": 23775, "total_steps": 29704, "loss": 0.2982936859130859, "lr": 9.515606983140896e-07, "epoch": 1.600794505790466, "percentage": 80.04, "elapsed_time": "2 days, 3:04:46", "remaining_time": "12:44:17", "throughput": 5653.9, "total_tokens": 1039677264}
+{"current_steps": 23780, "total_steps": 29704, "loss": 0.3003803253173828, "lr": 9.500095563511119e-07, "epoch": 1.601131160786426, "percentage": 80.06, "elapsed_time": "2 days, 3:05:30", "remaining_time": "12:43:40", "throughput": 5653.79, "total_tokens": 1039906360}
+{"current_steps": 23785, "total_steps": 29704, "loss": 0.28896503448486327, "lr": 9.484595469550445e-07, "epoch": 1.601467815782386, "percentage": 80.07, "elapsed_time": "2 days, 3:06:10", "remaining_time": "12:43:01", "throughput": 5653.79, "total_tokens": 1040131192}
+{"current_steps": 23790, "total_steps": 29704, "loss": 0.27666046619415285, "lr": 9.469106705593462e-07, "epoch": 1.6018044707783463, "percentage": 80.09, "elapsed_time": "2 days, 3:06:49", "remaining_time": "12:42:23", "throughput": 5653.7, "total_tokens": 1040333032}
+{"current_steps": 23795, "total_steps": 29704, "loss": 0.29598097801208495, "lr": 9.453629275971549e-07, "epoch": 1.6021411257743066, "percentage": 80.11, "elapsed_time": "2 days, 3:07:27", "remaining_time": "12:41:44", "throughput": 5653.72, "total_tokens": 1040552408}
+{"current_steps": 23800, "total_steps": 29704, "loss": 0.28930160999298093, "lr": 9.438163185012916e-07, "epoch": 1.6024777807702666, "percentage": 80.12, "elapsed_time": "2 days, 3:08:11", "remaining_time": "12:41:07", "throughput": 5653.51, "total_tokens": 1040764544}
+{"current_steps": 23805, "total_steps": 29704, "loss": 0.2819626808166504, "lr": 9.422708437042604e-07, "epoch": 1.6028144357662266, "percentage": 80.14, "elapsed_time": "2 days, 3:08:55", "remaining_time": "12:40:29", "throughput": 5653.41, "total_tokens": 1040990752}
+{"current_steps": 23810, "total_steps": 29704, "loss": 0.31176731586456297, "lr": 9.40726503638249e-07, "epoch": 1.6031510907621869, "percentage": 80.16, "elapsed_time": "2 days, 3:09:27", "remaining_time": "12:39:49", "throughput": 5653.59, "total_tokens": 1041206528}
+{"current_steps": 23815, "total_steps": 29704, "loss": 0.29704809188842773, "lr": 9.391832987351268e-07, "epoch": 1.6034877457581471, "percentage": 80.17, "elapsed_time": "2 days, 3:10:02", "remaining_time": "12:39:09", "throughput": 5653.68, "total_tokens": 1041421728}
+{"current_steps": 23820, "total_steps": 29704, "loss": 0.29626688957214353, "lr": 9.376412294264458e-07, "epoch": 1.6038244007541071, "percentage": 80.19, "elapsed_time": "2 days, 3:10:45", "remaining_time": "12:38:32", "throughput": 5653.57, "total_tokens": 1041643056}
+{"current_steps": 23825, "total_steps": 29704, "loss": 0.29646663665771483, "lr": 9.361002961434401e-07, "epoch": 1.6041610557500672, "percentage": 80.21, "elapsed_time": "2 days, 3:11:26", "remaining_time": "12:37:54", "throughput": 5653.47, "total_tokens": 1041859688}
+{"current_steps": 23830, "total_steps": 29704, "loss": 0.31340832710266114, "lr": 9.345604993170299e-07, "epoch": 1.6044977107460274, "percentage": 80.22, "elapsed_time": "2 days, 3:12:04", "remaining_time": "12:37:15", "throughput": 5653.56, "total_tokens": 1042088840}
+{"current_steps": 23835, "total_steps": 29704, "loss": 0.29099485874176023, "lr": 9.330218393778101e-07, "epoch": 1.6048343657419877, "percentage": 80.24, "elapsed_time": "2 days, 3:12:39", "remaining_time": "12:36:35", "throughput": 5653.61, "total_tokens": 1042295560}
+{"current_steps": 23840, "total_steps": 29704, "loss": 0.2913378715515137, "lr": 9.314843167560656e-07, "epoch": 1.6051710207379477, "percentage": 80.26, "elapsed_time": "2 days, 3:13:13", "remaining_time": "12:35:55", "throughput": 5653.74, "total_tokens": 1042510744}
+{"current_steps": 23845, "total_steps": 29704, "loss": 0.29028098583221434, "lr": 9.299479318817578e-07, "epoch": 1.6055076757339077, "percentage": 80.28, "elapsed_time": "2 days, 3:13:53", "remaining_time": "12:35:17", "throughput": 5653.66, "total_tokens": 1042725176}
+{"current_steps": 23850, "total_steps": 29704, "loss": 0.2781973361968994, "lr": 9.284126851845316e-07, "epoch": 1.605844330729868, "percentage": 80.29, "elapsed_time": "2 days, 3:14:33", "remaining_time": "12:34:39", "throughput": 5653.73, "total_tokens": 1042960480}
+{"current_steps": 23855, "total_steps": 29704, "loss": 0.3008106231689453, "lr": 9.268785770937172e-07, "epoch": 1.6061809857258282, "percentage": 80.31, "elapsed_time": "2 days, 3:15:13", "remaining_time": "12:34:00", "throughput": 5653.65, "total_tokens": 1043174632}
+{"current_steps": 23860, "total_steps": 29704, "loss": 0.3095564842224121, "lr": 9.253456080383178e-07, "epoch": 1.6065176407217883, "percentage": 80.33, "elapsed_time": "2 days, 3:15:50", "remaining_time": "12:33:21", "throughput": 5653.72, "total_tokens": 1043395416}
+{"current_steps": 23865, "total_steps": 29704, "loss": 0.32625675201416016, "lr": 9.238137784470286e-07, "epoch": 1.6068542957177483, "percentage": 80.34, "elapsed_time": "2 days, 3:16:22", "remaining_time": "12:32:41", "throughput": 5653.86, "total_tokens": 1043602960}
+{"current_steps": 23870, "total_steps": 29704, "loss": 0.2661672353744507, "lr": 9.222830887482153e-07, "epoch": 1.6071909507137085, "percentage": 80.36, "elapsed_time": "2 days, 3:16:59", "remaining_time": "12:32:02", "throughput": 5653.92, "total_tokens": 1043826208}
+{"current_steps": 23875, "total_steps": 29704, "loss": 0.3128221035003662, "lr": 9.207535393699351e-07, "epoch": 1.6075276057096688, "percentage": 80.38, "elapsed_time": "2 days, 3:17:38", "remaining_time": "12:31:23", "throughput": 5653.87, "total_tokens": 1044034312}
+{"current_steps": 23880, "total_steps": 29704, "loss": 0.29769034385681153, "lr": 9.192251307399197e-07, "epoch": 1.6078642607056288, "percentage": 80.39, "elapsed_time": "2 days, 3:18:14", "remaining_time": "12:30:44", "throughput": 5653.95, "total_tokens": 1044252784}
+{"current_steps": 23885, "total_steps": 29704, "loss": 0.2840586185455322, "lr": 9.176978632855842e-07, "epoch": 1.6082009157015889, "percentage": 80.41, "elapsed_time": "2 days, 3:18:52", "remaining_time": "12:30:05", "throughput": 5653.98, "total_tokens": 1044471704}
+{"current_steps": 23890, "total_steps": 29704, "loss": 0.2887522459030151, "lr": 9.161717374340235e-07, "epoch": 1.608537570697549, "percentage": 80.43, "elapsed_time": "2 days, 3:19:30", "remaining_time": "12:29:26", "throughput": 5654.1, "total_tokens": 1044709256}
+{"current_steps": 23895, "total_steps": 29704, "loss": 0.3057317495346069, "lr": 9.14646753612014e-07, "epoch": 1.6088742256935094, "percentage": 80.44, "elapsed_time": "2 days, 3:20:02", "remaining_time": "12:28:46", "throughput": 5654.31, "total_tokens": 1044932856}
+{"current_steps": 23900, "total_steps": 29704, "loss": 0.29195594787597656, "lr": 9.131229122460112e-07, "epoch": 1.6092108806894694, "percentage": 80.46, "elapsed_time": "2 days, 3:20:47", "remaining_time": "12:28:09", "throughput": 5654.22, "total_tokens": 1045169176}
+{"current_steps": 23905, "total_steps": 29704, "loss": 0.30215911865234374, "lr": 9.116002137621566e-07, "epoch": 1.6095475356854294, "percentage": 80.48, "elapsed_time": "2 days, 3:21:28", "remaining_time": "12:27:31", "throughput": 5654.15, "total_tokens": 1045388344}
+{"current_steps": 23910, "total_steps": 29704, "loss": 0.33268914222717283, "lr": 9.100786585862626e-07, "epoch": 1.6098841906813897, "percentage": 80.49, "elapsed_time": "2 days, 3:22:12", "remaining_time": "12:26:53", "throughput": 5653.92, "total_tokens": 1045590472}
+{"current_steps": 23915, "total_steps": 29704, "loss": 0.3007976531982422, "lr": 9.085582471438326e-07, "epoch": 1.61022084567735, "percentage": 80.51, "elapsed_time": "2 days, 3:22:48", "remaining_time": "12:26:14", "throughput": 5653.98, "total_tokens": 1045807448}
+{"current_steps": 23920, "total_steps": 29704, "loss": 0.3141302585601807, "lr": 9.070389798600394e-07, "epoch": 1.61055750067331, "percentage": 80.53, "elapsed_time": "2 days, 3:23:28", "remaining_time": "12:25:36", "throughput": 5653.92, "total_tokens": 1046022008}
+{"current_steps": 23925, "total_steps": 29704, "loss": 0.3039546489715576, "lr": 9.055208571597451e-07, "epoch": 1.61089415566927, "percentage": 80.54, "elapsed_time": "2 days, 3:24:06", "remaining_time": "12:24:57", "throughput": 5653.94, "total_tokens": 1046240136}
+{"current_steps": 23930, "total_steps": 29704, "loss": 0.30896403789520266, "lr": 9.040038794674872e-07, "epoch": 1.6112308106652302, "percentage": 80.56, "elapsed_time": "2 days, 3:24:47", "remaining_time": "12:24:19", "throughput": 5653.85, "total_tokens": 1046460096}
+{"current_steps": 23935, "total_steps": 29704, "loss": 0.28701066970825195, "lr": 9.024880472074831e-07, "epoch": 1.6115674656611905, "percentage": 80.58, "elapsed_time": "2 days, 3:25:32", "remaining_time": "12:23:42", "throughput": 5653.59, "total_tokens": 1046662408}
+{"current_steps": 23940, "total_steps": 29704, "loss": 0.30707645416259766, "lr": 9.009733608036308e-07, "epoch": 1.6119041206571505, "percentage": 80.6, "elapsed_time": "2 days, 3:26:09", "remaining_time": "12:23:03", "throughput": 5653.67, "total_tokens": 1046889064}
+{"current_steps": 23945, "total_steps": 29704, "loss": 0.29622645378112794, "lr": 8.994598206795068e-07, "epoch": 1.6122407756531107, "percentage": 80.61, "elapsed_time": "2 days, 3:26:45", "remaining_time": "12:22:23", "throughput": 5653.75, "total_tokens": 1047104032}
+{"current_steps": 23950, "total_steps": 29704, "loss": 0.28798418045043944, "lr": 8.979474272583672e-07, "epoch": 1.612577430649071, "percentage": 80.63, "elapsed_time": "2 days, 3:27:27", "remaining_time": "12:21:45", "throughput": 5653.59, "total_tokens": 1047312440}
+{"current_steps": 23955, "total_steps": 29704, "loss": 0.3181246280670166, "lr": 8.964361809631517e-07, "epoch": 1.612914085645031, "percentage": 80.65, "elapsed_time": "2 days, 3:28:09", "remaining_time": "12:21:08", "throughput": 5653.52, "total_tokens": 1047539480}
+{"current_steps": 23960, "total_steps": 29704, "loss": 0.31255478858947755, "lr": 8.949260822164707e-07, "epoch": 1.613250740640991, "percentage": 80.66, "elapsed_time": "2 days, 3:28:46", "remaining_time": "12:20:28", "throughput": 5653.6, "total_tokens": 1047761560}
+{"current_steps": 23965, "total_steps": 29704, "loss": 0.2988784074783325, "lr": 8.934171314406226e-07, "epoch": 1.6135873956369513, "percentage": 80.68, "elapsed_time": "2 days, 3:29:21", "remaining_time": "12:19:49", "throughput": 5653.74, "total_tokens": 1047987312}
+{"current_steps": 23970, "total_steps": 29704, "loss": 0.30140008926391604, "lr": 8.919093290575765e-07, "epoch": 1.6139240506329116, "percentage": 80.7, "elapsed_time": "2 days, 3:30:05", "remaining_time": "12:19:11", "throughput": 5653.54, "total_tokens": 1048194320}
+{"current_steps": 23975, "total_steps": 29704, "loss": 0.30414538383483886, "lr": 8.904026754889877e-07, "epoch": 1.6142607056288716, "percentage": 80.71, "elapsed_time": "2 days, 3:30:44", "remaining_time": "12:18:33", "throughput": 5653.46, "total_tokens": 1048404056}
+{"current_steps": 23980, "total_steps": 29704, "loss": 0.32020394802093505, "lr": 8.888971711561867e-07, "epoch": 1.6145973606248316, "percentage": 80.73, "elapsed_time": "2 days, 3:31:26", "remaining_time": "12:17:55", "throughput": 5653.31, "total_tokens": 1048613480}
+{"current_steps": 23985, "total_steps": 29704, "loss": 0.32930703163146974, "lr": 8.873928164801821e-07, "epoch": 1.6149340156207919, "percentage": 80.75, "elapsed_time": "2 days, 3:32:05", "remaining_time": "12:17:16", "throughput": 5653.33, "total_tokens": 1048836768}
+{"current_steps": 23990, "total_steps": 29704, "loss": 0.2910914421081543, "lr": 8.858896118816624e-07, "epoch": 1.6152706706167521, "percentage": 80.76, "elapsed_time": "2 days, 3:32:48", "remaining_time": "12:16:39", "throughput": 5653.05, "total_tokens": 1049026440}
+{"current_steps": 23995, "total_steps": 29704, "loss": 0.2792020797729492, "lr": 8.84387557780994e-07, "epoch": 1.6156073256127121, "percentage": 80.78, "elapsed_time": "2 days, 3:33:31", "remaining_time": "12:16:01", "throughput": 5652.86, "total_tokens": 1049235720}
+{"current_steps": 24000, "total_steps": 29704, "loss": 0.2844292163848877, "lr": 8.828866545982196e-07, "epoch": 1.6159439806086722, "percentage": 80.8, "elapsed_time": "2 days, 3:34:12", "remaining_time": "12:15:23", "throughput": 5652.79, "total_tokens": 1049455984}
+{"current_steps": 24005, "total_steps": 29704, "loss": 0.28838918209075926, "lr": 8.813869027530664e-07, "epoch": 1.6162806356046324, "percentage": 80.81, "elapsed_time": "2 days, 3:34:58", "remaining_time": "12:14:46", "throughput": 5652.59, "total_tokens": 1049680856}
+{"current_steps": 24010, "total_steps": 29704, "loss": 0.31560914516448973, "lr": 8.798883026649302e-07, "epoch": 1.6166172906005927, "percentage": 80.83, "elapsed_time": "2 days, 3:35:35", "remaining_time": "12:14:07", "throughput": 5652.74, "total_tokens": 1049914184}
+{"current_steps": 24015, "total_steps": 29704, "loss": 0.2625638246536255, "lr": 8.783908547528947e-07, "epoch": 1.6169539455965527, "percentage": 80.85, "elapsed_time": "2 days, 3:36:14", "remaining_time": "12:13:28", "throughput": 5652.7, "total_tokens": 1050126648}
+{"current_steps": 24020, "total_steps": 29704, "loss": 0.29282054901123045, "lr": 8.768945594357109e-07, "epoch": 1.6172906005925127, "percentage": 80.86, "elapsed_time": "2 days, 3:36:53", "remaining_time": "12:12:50", "throughput": 5652.65, "total_tokens": 1050339456}
+{"current_steps": 24025, "total_steps": 29704, "loss": 0.3160220146179199, "lr": 8.753994171318175e-07, "epoch": 1.617627255588473, "percentage": 80.88, "elapsed_time": "2 days, 3:37:31", "remaining_time": "12:12:11", "throughput": 5652.63, "total_tokens": 1050547328}
+{"current_steps": 24030, "total_steps": 29704, "loss": 0.3164478063583374, "lr": 8.739054282593245e-07, "epoch": 1.6179639105844332, "percentage": 80.9, "elapsed_time": "2 days, 3:38:10", "remaining_time": "12:11:32", "throughput": 5652.64, "total_tokens": 1050771952}
+{"current_steps": 24035, "total_steps": 29704, "loss": 0.3083709716796875, "lr": 8.724125932360217e-07, "epoch": 1.6183005655803933, "percentage": 80.92, "elapsed_time": "2 days, 3:38:49", "remaining_time": "12:10:54", "throughput": 5652.6, "total_tokens": 1050984192}
+{"current_steps": 24040, "total_steps": 29704, "loss": 0.3174473762512207, "lr": 8.70920912479375e-07, "epoch": 1.6186372205763533, "percentage": 80.93, "elapsed_time": "2 days, 3:39:24", "remaining_time": "12:10:14", "throughput": 5652.68, "total_tokens": 1051196368}
+{"current_steps": 24045, "total_steps": 29704, "loss": 0.29400835037231443, "lr": 8.694303864065285e-07, "epoch": 1.6189738755723135, "percentage": 80.95, "elapsed_time": "2 days, 3:40:05", "remaining_time": "12:09:36", "throughput": 5652.67, "total_tokens": 1051425808}
+{"current_steps": 24050, "total_steps": 29704, "loss": 0.29439725875854494, "lr": 8.679410154343015e-07, "epoch": 1.6193105305682738, "percentage": 80.97, "elapsed_time": "2 days, 3:40:42", "remaining_time": "12:08:57", "throughput": 5652.71, "total_tokens": 1051646216}
+{"current_steps": 24055, "total_steps": 29704, "loss": 0.3235488414764404, "lr": 8.664527999791961e-07, "epoch": 1.6196471855642338, "percentage": 80.98, "elapsed_time": "2 days, 3:41:22", "remaining_time": "12:08:19", "throughput": 5652.68, "total_tokens": 1051864880}
+{"current_steps": 24060, "total_steps": 29704, "loss": 0.29723272323608396, "lr": 8.649657404573819e-07, "epoch": 1.6199838405601938, "percentage": 81.0, "elapsed_time": "2 days, 3:42:06", "remaining_time": "12:07:41", "throughput": 5652.57, "total_tokens": 1052093008}
+{"current_steps": 24065, "total_steps": 29704, "loss": 0.317378568649292, "lr": 8.634798372847148e-07, "epoch": 1.620320495556154, "percentage": 81.02, "elapsed_time": "2 days, 3:42:42", "remaining_time": "12:07:02", "throughput": 5652.69, "total_tokens": 1052316944}
+{"current_steps": 24070, "total_steps": 29704, "loss": 0.2666431665420532, "lr": 8.619950908767189e-07, "epoch": 1.6206571505521143, "percentage": 81.03, "elapsed_time": "2 days, 3:43:25", "remaining_time": "12:06:24", "throughput": 5652.45, "total_tokens": 1052516136}
+{"current_steps": 24075, "total_steps": 29704, "loss": 0.2880539894104004, "lr": 8.605115016486016e-07, "epoch": 1.6209938055480744, "percentage": 81.05, "elapsed_time": "2 days, 3:44:03", "remaining_time": "12:05:45", "throughput": 5652.55, "total_tokens": 1052748400}
+{"current_steps": 24080, "total_steps": 29704, "loss": 0.2833811044692993, "lr": 8.590290700152426e-07, "epoch": 1.6213304605440344, "percentage": 81.07, "elapsed_time": "2 days, 3:44:49", "remaining_time": "12:05:08", "throughput": 5652.31, "total_tokens": 1052963936}
+{"current_steps": 24085, "total_steps": 29704, "loss": 0.29882135391235354, "lr": 8.575477963912005e-07, "epoch": 1.6216671155399947, "percentage": 81.08, "elapsed_time": "2 days, 3:45:29", "remaining_time": "12:04:30", "throughput": 5652.28, "total_tokens": 1053188096}
+{"current_steps": 24090, "total_steps": 29704, "loss": 0.3093993663787842, "lr": 8.560676811907071e-07, "epoch": 1.622003770535955, "percentage": 81.1, "elapsed_time": "2 days, 3:46:10", "remaining_time": "12:03:52", "throughput": 5652.23, "total_tokens": 1053409680}
+{"current_steps": 24095, "total_steps": 29704, "loss": 0.29143998622894285, "lr": 8.54588724827673e-07, "epoch": 1.622340425531915, "percentage": 81.12, "elapsed_time": "2 days, 3:46:52", "remaining_time": "12:03:14", "throughput": 5652.01, "total_tokens": 1053604304}
+{"current_steps": 24100, "total_steps": 29704, "loss": 0.2952460050582886, "lr": 8.531109277156818e-07, "epoch": 1.622677080527875, "percentage": 81.13, "elapsed_time": "2 days, 3:47:25", "remaining_time": "12:02:34", "throughput": 5652.25, "total_tokens": 1053836472}
+{"current_steps": 24105, "total_steps": 29704, "loss": 0.26687233448028563, "lr": 8.516342902679986e-07, "epoch": 1.6230137355238352, "percentage": 81.15, "elapsed_time": "2 days, 3:48:07", "remaining_time": "12:01:56", "throughput": 5652.1, "total_tokens": 1054042928}
+{"current_steps": 24110, "total_steps": 29704, "loss": 0.274417781829834, "lr": 8.501588128975557e-07, "epoch": 1.6233503905197955, "percentage": 81.17, "elapsed_time": "2 days, 3:48:45", "remaining_time": "12:01:17", "throughput": 5652.13, "total_tokens": 1054266256}
+{"current_steps": 24115, "total_steps": 29704, "loss": 0.28249123096466067, "lr": 8.486844960169704e-07, "epoch": 1.6236870455157555, "percentage": 81.18, "elapsed_time": "2 days, 3:49:25", "remaining_time": "12:00:39", "throughput": 5652.07, "total_tokens": 1054484008}
+{"current_steps": 24120, "total_steps": 29704, "loss": 0.2758941650390625, "lr": 8.472113400385257e-07, "epoch": 1.6240237005117155, "percentage": 81.2, "elapsed_time": "2 days, 3:50:05", "remaining_time": "12:00:00", "throughput": 5652.08, "total_tokens": 1054711848}
+{"current_steps": 24125, "total_steps": 29704, "loss": 0.3019557476043701, "lr": 8.457393453741886e-07, "epoch": 1.6243603555076758, "percentage": 81.22, "elapsed_time": "2 days, 3:50:48", "remaining_time": "11:59:23", "throughput": 5651.94, "total_tokens": 1054923912}
+{"current_steps": 24130, "total_steps": 29704, "loss": 0.28956966400146483, "lr": 8.442685124355965e-07, "epoch": 1.624697010503636, "percentage": 81.23, "elapsed_time": "2 days, 3:51:29", "remaining_time": "11:58:45", "throughput": 5651.83, "total_tokens": 1055138184}
+{"current_steps": 24135, "total_steps": 29704, "loss": 0.31919891834259034, "lr": 8.427988416340633e-07, "epoch": 1.625033665499596, "percentage": 81.25, "elapsed_time": "2 days, 3:52:16", "remaining_time": "11:58:08", "throughput": 5651.5, "total_tokens": 1055341616}
+{"current_steps": 24140, "total_steps": 29704, "loss": 0.2868987560272217, "lr": 8.413303333805777e-07, "epoch": 1.625370320495556, "percentage": 81.27, "elapsed_time": "2 days, 3:52:58", "remaining_time": "11:57:30", "throughput": 5651.36, "total_tokens": 1055555328}
+{"current_steps": 24145, "total_steps": 29704, "loss": 0.3065673589706421, "lr": 8.398629880858034e-07, "epoch": 1.6257069754915163, "percentage": 81.29, "elapsed_time": "2 days, 3:53:34", "remaining_time": "11:56:51", "throughput": 5651.52, "total_tokens": 1055788776}
+{"current_steps": 24150, "total_steps": 29704, "loss": 0.30824875831604004, "lr": 8.383968061600778e-07, "epoch": 1.6260436304874766, "percentage": 81.3, "elapsed_time": "2 days, 3:54:11", "remaining_time": "11:56:11", "throughput": 5651.55, "total_tokens": 1055998904}
+{"current_steps": 24155, "total_steps": 29704, "loss": 0.2875575304031372, "lr": 8.36931788013417e-07, "epoch": 1.6263802854834366, "percentage": 81.32, "elapsed_time": "2 days, 3:54:45", "remaining_time": "11:55:32", "throughput": 5651.7, "total_tokens": 1056224128}
+{"current_steps": 24160, "total_steps": 29704, "loss": 0.30007798671722413, "lr": 8.354679340555044e-07, "epoch": 1.6267169404793966, "percentage": 81.34, "elapsed_time": "2 days, 3:55:18", "remaining_time": "11:54:52", "throughput": 5651.84, "total_tokens": 1056436296}
+{"current_steps": 24165, "total_steps": 29704, "loss": 0.28886971473693845, "lr": 8.34005244695707e-07, "epoch": 1.6270535954753569, "percentage": 81.35, "elapsed_time": "2 days, 3:55:53", "remaining_time": "11:54:12", "throughput": 5651.95, "total_tokens": 1056648824}
+{"current_steps": 24170, "total_steps": 29704, "loss": 0.29457979202270507, "lr": 8.325437203430558e-07, "epoch": 1.6273902504713171, "percentage": 81.37, "elapsed_time": "2 days, 3:56:35", "remaining_time": "11:53:34", "throughput": 5651.85, "total_tokens": 1056867768}
+{"current_steps": 24175, "total_steps": 29704, "loss": 0.295581316947937, "lr": 8.310833614062652e-07, "epoch": 1.6277269054672772, "percentage": 81.39, "elapsed_time": "2 days, 3:57:12", "remaining_time": "11:52:55", "throughput": 5651.87, "total_tokens": 1057083272}
+{"current_steps": 24180, "total_steps": 29704, "loss": 0.2983792543411255, "lr": 8.296241682937189e-07, "epoch": 1.6280635604632372, "percentage": 81.4, "elapsed_time": "2 days, 3:57:54", "remaining_time": "11:52:17", "throughput": 5651.8, "total_tokens": 1057309512}
+{"current_steps": 24185, "total_steps": 29704, "loss": 0.2972233295440674, "lr": 8.281661414134761e-07, "epoch": 1.6284002154591974, "percentage": 81.42, "elapsed_time": "2 days, 3:58:28", "remaining_time": "11:51:38", "throughput": 5651.91, "total_tokens": 1057522792}
+{"current_steps": 24190, "total_steps": 29704, "loss": 0.31084840297698973, "lr": 8.267092811732686e-07, "epoch": 1.6287368704551577, "percentage": 81.44, "elapsed_time": "2 days, 3:59:08", "remaining_time": "11:50:59", "throughput": 5651.9, "total_tokens": 1057745072}
+{"current_steps": 24195, "total_steps": 29704, "loss": 0.2759894847869873, "lr": 8.252535879805029e-07, "epoch": 1.6290735254511177, "percentage": 81.45, "elapsed_time": "2 days, 3:59:47", "remaining_time": "11:50:21", "throughput": 5651.94, "total_tokens": 1057972008}
+{"current_steps": 24200, "total_steps": 29704, "loss": 0.289616322517395, "lr": 8.237990622422576e-07, "epoch": 1.6294101804470777, "percentage": 81.47, "elapsed_time": "2 days, 4:00:23", "remaining_time": "11:49:41", "throughput": 5652.01, "total_tokens": 1058191104}
+{"current_steps": 24205, "total_steps": 29704, "loss": 0.3101085662841797, "lr": 8.223457043652905e-07, "epoch": 1.629746835443038, "percentage": 81.49, "elapsed_time": "2 days, 4:00:56", "remaining_time": "11:49:01", "throughput": 5652.22, "total_tokens": 1058412784}
+{"current_steps": 24210, "total_steps": 29704, "loss": 0.2965271472930908, "lr": 8.208935147560227e-07, "epoch": 1.6300834904389983, "percentage": 81.5, "elapsed_time": "2 days, 4:01:28", "remaining_time": "11:48:21", "throughput": 5652.31, "total_tokens": 1058609720}
+{"current_steps": 24215, "total_steps": 29704, "loss": 0.3259326696395874, "lr": 8.194424938205597e-07, "epoch": 1.6304201454349583, "percentage": 81.52, "elapsed_time": "2 days, 4:02:08", "remaining_time": "11:47:43", "throughput": 5652.36, "total_tokens": 1058845344}
+{"current_steps": 24220, "total_steps": 29704, "loss": 0.28400454521179197, "lr": 8.179926419646705e-07, "epoch": 1.6307568004309183, "percentage": 81.54, "elapsed_time": "2 days, 4:02:51", "remaining_time": "11:47:05", "throughput": 5652.2, "total_tokens": 1059062096}
+{"current_steps": 24225, "total_steps": 29704, "loss": 0.32569429874420164, "lr": 8.165439595938047e-07, "epoch": 1.6310934554268786, "percentage": 81.55, "elapsed_time": "2 days, 4:03:34", "remaining_time": "11:46:27", "throughput": 5651.96, "total_tokens": 1059262648}
+{"current_steps": 24230, "total_steps": 29704, "loss": 0.2735768795013428, "lr": 8.150964471130801e-07, "epoch": 1.6314301104228388, "percentage": 81.57, "elapsed_time": "2 days, 4:04:08", "remaining_time": "11:45:48", "throughput": 5652.17, "total_tokens": 1059491256}
+{"current_steps": 24235, "total_steps": 29704, "loss": 0.30480170249938965, "lr": 8.136501049272899e-07, "epoch": 1.6317667654187988, "percentage": 81.59, "elapsed_time": "2 days, 4:04:45", "remaining_time": "11:45:08", "throughput": 5652.21, "total_tokens": 1059706760}
+{"current_steps": 24240, "total_steps": 29704, "loss": 0.29867599010467527, "lr": 8.122049334408983e-07, "epoch": 1.6321034204147589, "percentage": 81.61, "elapsed_time": "2 days, 4:05:22", "remaining_time": "11:44:29", "throughput": 5652.23, "total_tokens": 1059920888}
+{"current_steps": 24245, "total_steps": 29704, "loss": 0.29615049362182616, "lr": 8.107609330580429e-07, "epoch": 1.6324400754107191, "percentage": 81.62, "elapsed_time": "2 days, 4:06:03", "remaining_time": "11:43:51", "throughput": 5652.17, "total_tokens": 1060137784}
+{"current_steps": 24250, "total_steps": 29704, "loss": 0.3028998613357544, "lr": 8.093181041825332e-07, "epoch": 1.6327767304066794, "percentage": 81.64, "elapsed_time": "2 days, 4:06:43", "remaining_time": "11:43:13", "throughput": 5652.13, "total_tokens": 1060360840}
+{"current_steps": 24255, "total_steps": 29704, "loss": 0.29851455688476564, "lr": 8.078764472178552e-07, "epoch": 1.6331133854026394, "percentage": 81.66, "elapsed_time": "2 days, 4:07:23", "remaining_time": "11:42:35", "throughput": 5652.06, "total_tokens": 1060573408}
+{"current_steps": 24260, "total_steps": 29704, "loss": 0.32530434131622316, "lr": 8.06435962567158e-07, "epoch": 1.6334500403985994, "percentage": 81.67, "elapsed_time": "2 days, 4:08:05", "remaining_time": "11:41:57", "throughput": 5651.83, "total_tokens": 1060766968}
+{"current_steps": 24265, "total_steps": 29704, "loss": 0.26018400192260743, "lr": 8.049966506332746e-07, "epoch": 1.6337866953945597, "percentage": 81.69, "elapsed_time": "2 days, 4:08:38", "remaining_time": "11:41:17", "throughput": 5652.0, "total_tokens": 1060985712}
+{"current_steps": 24270, "total_steps": 29704, "loss": 0.2798137664794922, "lr": 8.035585118186984e-07, "epoch": 1.63412335039052, "percentage": 81.71, "elapsed_time": "2 days, 4:09:11", "remaining_time": "11:40:37", "throughput": 5652.14, "total_tokens": 1061195896}
+{"current_steps": 24275, "total_steps": 29704, "loss": 0.2788818836212158, "lr": 8.021215465256038e-07, "epoch": 1.63446000538648, "percentage": 81.72, "elapsed_time": "2 days, 4:09:55", "remaining_time": "11:39:59", "throughput": 5652.0, "total_tokens": 1061420544}
+{"current_steps": 24280, "total_steps": 29704, "loss": 0.30008883476257325, "lr": 8.006857551558328e-07, "epoch": 1.63479666038244, "percentage": 81.74, "elapsed_time": "2 days, 4:10:42", "remaining_time": "11:39:22", "throughput": 5651.74, "total_tokens": 1061636496}
+{"current_steps": 24285, "total_steps": 29704, "loss": 0.3101008415222168, "lr": 7.992511381108997e-07, "epoch": 1.6351333153784002, "percentage": 81.76, "elapsed_time": "2 days, 4:11:15", "remaining_time": "11:38:42", "throughput": 5652.04, "total_tokens": 1061878752}
+{"current_steps": 24290, "total_steps": 29704, "loss": 0.3050835132598877, "lr": 7.978176957919909e-07, "epoch": 1.6354699703743605, "percentage": 81.77, "elapsed_time": "2 days, 4:12:01", "remaining_time": "11:38:05", "throughput": 5651.77, "total_tokens": 1062088984}
+{"current_steps": 24295, "total_steps": 29704, "loss": 0.27643623352050783, "lr": 7.963854285999634e-07, "epoch": 1.6358066253703205, "percentage": 81.79, "elapsed_time": "2 days, 4:12:46", "remaining_time": "11:37:28", "throughput": 5651.58, "total_tokens": 1062308496}
+{"current_steps": 24300, "total_steps": 29704, "loss": 0.30392279624938967, "lr": 7.949543369353452e-07, "epoch": 1.6361432803662805, "percentage": 81.81, "elapsed_time": "2 days, 4:13:28", "remaining_time": "11:36:50", "throughput": 5651.38, "total_tokens": 1062508472}
+{"current_steps": 24305, "total_steps": 29704, "loss": 0.3222590446472168, "lr": 7.9352442119834e-07, "epoch": 1.6364799353622408, "percentage": 81.82, "elapsed_time": "2 days, 4:14:09", "remaining_time": "11:36:12", "throughput": 5651.28, "total_tokens": 1062721224}
+{"current_steps": 24310, "total_steps": 29704, "loss": 0.3079147577285767, "lr": 7.920956817888148e-07, "epoch": 1.636816590358201, "percentage": 81.84, "elapsed_time": "2 days, 4:14:45", "remaining_time": "11:35:33", "throughput": 5651.35, "total_tokens": 1062938784}
+{"current_steps": 24315, "total_steps": 29704, "loss": 0.2923039436340332, "lr": 7.906681191063165e-07, "epoch": 1.637153245354161, "percentage": 81.86, "elapsed_time": "2 days, 4:15:22", "remaining_time": "11:34:54", "throughput": 5651.42, "total_tokens": 1063160648}
+{"current_steps": 24320, "total_steps": 29704, "loss": 0.3094671964645386, "lr": 7.892417335500541e-07, "epoch": 1.637489900350121, "percentage": 81.87, "elapsed_time": "2 days, 4:15:58", "remaining_time": "11:34:14", "throughput": 5651.45, "total_tokens": 1063371504}
+{"current_steps": 24325, "total_steps": 29704, "loss": 0.32140753269195554, "lr": 7.878165255189146e-07, "epoch": 1.6378265553460813, "percentage": 81.89, "elapsed_time": "2 days, 4:16:36", "remaining_time": "11:33:35", "throughput": 5651.42, "total_tokens": 1063574440}
+{"current_steps": 24330, "total_steps": 29704, "loss": 0.28619048595428465, "lr": 7.863924954114522e-07, "epoch": 1.6381632103420416, "percentage": 81.91, "elapsed_time": "2 days, 4:17:12", "remaining_time": "11:32:56", "throughput": 5651.43, "total_tokens": 1063782584}
+{"current_steps": 24335, "total_steps": 29704, "loss": 0.2759582042694092, "lr": 7.849696436258919e-07, "epoch": 1.6384998653380016, "percentage": 81.92, "elapsed_time": "2 days, 4:17:52", "remaining_time": "11:32:18", "throughput": 5651.37, "total_tokens": 1063995392}
+{"current_steps": 24340, "total_steps": 29704, "loss": 0.30858869552612306, "lr": 7.835479705601307e-07, "epoch": 1.6388365203339617, "percentage": 81.94, "elapsed_time": "2 days, 4:18:30", "remaining_time": "11:31:39", "throughput": 5651.31, "total_tokens": 1064201992}
+{"current_steps": 24345, "total_steps": 29704, "loss": 0.3018312692642212, "lr": 7.821274766117337e-07, "epoch": 1.639173175329922, "percentage": 81.96, "elapsed_time": "2 days, 4:19:14", "remaining_time": "11:31:01", "throughput": 5651.15, "total_tokens": 1064418504}
+{"current_steps": 24350, "total_steps": 29704, "loss": 0.3180276393890381, "lr": 7.807081621779367e-07, "epoch": 1.6395098303258822, "percentage": 81.98, "elapsed_time": "2 days, 4:19:56", "remaining_time": "11:30:24", "throughput": 5651.06, "total_tokens": 1064640776}
+{"current_steps": 24355, "total_steps": 29704, "loss": 0.30577759742736815, "lr": 7.792900276556509e-07, "epoch": 1.6398464853218422, "percentage": 81.99, "elapsed_time": "2 days, 4:20:37", "remaining_time": "11:29:45", "throughput": 5651.03, "total_tokens": 1064866144}
+{"current_steps": 24360, "total_steps": 29704, "loss": 0.3030556201934814, "lr": 7.778730734414469e-07, "epoch": 1.6401831403178022, "percentage": 82.01, "elapsed_time": "2 days, 4:21:18", "remaining_time": "11:29:07", "throughput": 5650.94, "total_tokens": 1065082080}
+{"current_steps": 24365, "total_steps": 29704, "loss": 0.2959816217422485, "lr": 7.764572999315772e-07, "epoch": 1.6405197953137625, "percentage": 82.03, "elapsed_time": "2 days, 4:21:54", "remaining_time": "11:28:28", "throughput": 5651.08, "total_tokens": 1065308184}
+{"current_steps": 24370, "total_steps": 29704, "loss": 0.3010838031768799, "lr": 7.750427075219536e-07, "epoch": 1.6408564503097227, "percentage": 82.04, "elapsed_time": "2 days, 4:22:34", "remaining_time": "11:27:50", "throughput": 5650.97, "total_tokens": 1065519384}
+{"current_steps": 24375, "total_steps": 29704, "loss": 0.28882603645324706, "lr": 7.736292966081655e-07, "epoch": 1.6411931053056827, "percentage": 82.06, "elapsed_time": "2 days, 4:23:12", "remaining_time": "11:27:11", "throughput": 5650.87, "total_tokens": 1065712464}
+{"current_steps": 24380, "total_steps": 29704, "loss": 0.26800405979156494, "lr": 7.722170675854673e-07, "epoch": 1.6415297603016428, "percentage": 82.08, "elapsed_time": "2 days, 4:23:49", "remaining_time": "11:26:32", "throughput": 5650.97, "total_tokens": 1065939064}
+{"current_steps": 24385, "total_steps": 29704, "loss": 0.2844860076904297, "lr": 7.70806020848785e-07, "epoch": 1.641866415297603, "percentage": 82.09, "elapsed_time": "2 days, 4:24:20", "remaining_time": "11:25:51", "throughput": 5651.19, "total_tokens": 1066159000}
+{"current_steps": 24390, "total_steps": 29704, "loss": 0.2823920726776123, "lr": 7.693961567927133e-07, "epoch": 1.6422030702935633, "percentage": 82.11, "elapsed_time": "2 days, 4:25:00", "remaining_time": "11:25:13", "throughput": 5651.2, "total_tokens": 1066381848}
+{"current_steps": 24395, "total_steps": 29704, "loss": 0.2843328475952148, "lr": 7.679874758115153e-07, "epoch": 1.6425397252895233, "percentage": 82.13, "elapsed_time": "2 days, 4:25:38", "remaining_time": "11:24:34", "throughput": 5651.28, "total_tokens": 1066612256}
+{"current_steps": 24400, "total_steps": 29704, "loss": 0.30366389751434325, "lr": 7.66579978299124e-07, "epoch": 1.6428763802854833, "percentage": 82.14, "elapsed_time": "2 days, 4:26:15", "remaining_time": "11:23:55", "throughput": 5651.37, "total_tokens": 1066840032}
+{"current_steps": 24405, "total_steps": 29704, "loss": 0.32686214447021483, "lr": 7.651736646491447e-07, "epoch": 1.6432130352814436, "percentage": 82.16, "elapsed_time": "2 days, 4:26:52", "remaining_time": "11:23:16", "throughput": 5651.32, "total_tokens": 1067042480}
+{"current_steps": 24410, "total_steps": 29704, "loss": 0.27104051113128663, "lr": 7.63768535254844e-07, "epoch": 1.6435496902774038, "percentage": 82.18, "elapsed_time": "2 days, 4:27:29", "remaining_time": "11:22:37", "throughput": 5651.39, "total_tokens": 1067263472}
+{"current_steps": 24415, "total_steps": 29704, "loss": 0.3036444902420044, "lr": 7.623645905091664e-07, "epoch": 1.6438863452733639, "percentage": 82.19, "elapsed_time": "2 days, 4:28:12", "remaining_time": "11:21:59", "throughput": 5651.26, "total_tokens": 1067483856}
+{"current_steps": 24420, "total_steps": 29704, "loss": 0.3053473472595215, "lr": 7.609618308047156e-07, "epoch": 1.6442230002693239, "percentage": 82.21, "elapsed_time": "2 days, 4:28:45", "remaining_time": "11:21:19", "throughput": 5651.44, "total_tokens": 1067700112}
+{"current_steps": 24425, "total_steps": 29704, "loss": 0.29787912368774416, "lr": 7.595602565337729e-07, "epoch": 1.6445596552652841, "percentage": 82.23, "elapsed_time": "2 days, 4:29:33", "remaining_time": "11:20:42", "throughput": 5651.22, "total_tokens": 1067928768}
+{"current_steps": 24430, "total_steps": 29704, "loss": 0.31963634490966797, "lr": 7.581598680882829e-07, "epoch": 1.6448963102612444, "percentage": 82.24, "elapsed_time": "2 days, 4:30:12", "remaining_time": "11:20:04", "throughput": 5651.23, "total_tokens": 1068150624}
+{"current_steps": 24435, "total_steps": 29704, "loss": 0.29708251953125, "lr": 7.567606658598592e-07, "epoch": 1.6452329652572044, "percentage": 82.26, "elapsed_time": "2 days, 4:30:50", "remaining_time": "11:19:25", "throughput": 5651.24, "total_tokens": 1068367600}
+{"current_steps": 24440, "total_steps": 29704, "loss": 0.3157409906387329, "lr": 7.553626502397848e-07, "epoch": 1.6455696202531644, "percentage": 82.28, "elapsed_time": "2 days, 4:31:34", "remaining_time": "11:18:48", "throughput": 5651.09, "total_tokens": 1068589480}
+{"current_steps": 24445, "total_steps": 29704, "loss": 0.2788280248641968, "lr": 7.539658216190099e-07, "epoch": 1.6459062752491247, "percentage": 82.3, "elapsed_time": "2 days, 4:32:13", "remaining_time": "11:18:09", "throughput": 5651.08, "total_tokens": 1068810256}
+{"current_steps": 24450, "total_steps": 29704, "loss": 0.305745792388916, "lr": 7.525701803881525e-07, "epoch": 1.646242930245085, "percentage": 82.31, "elapsed_time": "2 days, 4:32:54", "remaining_time": "11:17:31", "throughput": 5650.99, "total_tokens": 1069020600}
+{"current_steps": 24455, "total_steps": 29704, "loss": 0.2875369548797607, "lr": 7.511757269375019e-07, "epoch": 1.646579585241045, "percentage": 82.33, "elapsed_time": "2 days, 4:33:40", "remaining_time": "11:16:54", "throughput": 5650.7, "total_tokens": 1069228640}
+{"current_steps": 24460, "total_steps": 29704, "loss": 0.31629951000213624, "lr": 7.497824616570087e-07, "epoch": 1.646916240237005, "percentage": 82.35, "elapsed_time": "2 days, 4:34:14", "remaining_time": "11:16:14", "throughput": 5650.75, "total_tokens": 1069428344}
+{"current_steps": 24465, "total_steps": 29704, "loss": 0.3073263168334961, "lr": 7.483903849362995e-07, "epoch": 1.6472528952329653, "percentage": 82.36, "elapsed_time": "2 days, 4:34:53", "remaining_time": "11:15:35", "throughput": 5650.76, "total_tokens": 1069649936}
+{"current_steps": 24470, "total_steps": 29704, "loss": 0.3005094289779663, "lr": 7.469994971646594e-07, "epoch": 1.6475895502289255, "percentage": 82.38, "elapsed_time": "2 days, 4:35:37", "remaining_time": "11:14:58", "throughput": 5650.67, "total_tokens": 1069880992}
+{"current_steps": 24475, "total_steps": 29704, "loss": 0.3060446262359619, "lr": 7.456097987310495e-07, "epoch": 1.6479262052248855, "percentage": 82.4, "elapsed_time": "2 days, 4:36:15", "remaining_time": "11:14:19", "throughput": 5650.64, "total_tokens": 1070091864}
+{"current_steps": 24480, "total_steps": 29704, "loss": 0.3100743770599365, "lr": 7.442212900240931e-07, "epoch": 1.6482628602208456, "percentage": 82.41, "elapsed_time": "2 days, 4:36:55", "remaining_time": "11:13:40", "throughput": 5650.57, "total_tokens": 1070303776}
+{"current_steps": 24485, "total_steps": 29704, "loss": 0.2700855255126953, "lr": 7.42833971432082e-07, "epoch": 1.6485995152168058, "percentage": 82.43, "elapsed_time": "2 days, 4:37:27", "remaining_time": "11:13:00", "throughput": 5650.73, "total_tokens": 1070514192}
+{"current_steps": 24490, "total_steps": 29704, "loss": 0.3064833641052246, "lr": 7.414478433429756e-07, "epoch": 1.648936170212766, "percentage": 82.45, "elapsed_time": "2 days, 4:38:04", "remaining_time": "11:12:21", "throughput": 5650.77, "total_tokens": 1070735968}
+{"current_steps": 24495, "total_steps": 29704, "loss": 0.29147264957427976, "lr": 7.400629061444003e-07, "epoch": 1.649272825208726, "percentage": 82.46, "elapsed_time": "2 days, 4:38:48", "remaining_time": "11:11:44", "throughput": 5650.6, "total_tokens": 1070953008}
+{"current_steps": 24500, "total_steps": 29704, "loss": 0.32079496383666994, "lr": 7.386791602236482e-07, "epoch": 1.6496094802046861, "percentage": 82.48, "elapsed_time": "2 days, 4:39:23", "remaining_time": "11:11:04", "throughput": 5650.78, "total_tokens": 1071182864}
+{"current_steps": 24505, "total_steps": 29704, "loss": 0.34608256816864014, "lr": 7.372966059676834e-07, "epoch": 1.6499461352006464, "percentage": 82.5, "elapsed_time": "2 days, 4:40:04", "remaining_time": "11:10:26", "throughput": 5650.72, "total_tokens": 1071404416}
+{"current_steps": 24510, "total_steps": 29704, "loss": 0.31566901206970216, "lr": 7.359152437631273e-07, "epoch": 1.6502827901966066, "percentage": 82.51, "elapsed_time": "2 days, 4:40:42", "remaining_time": "11:09:47", "throughput": 5650.8, "total_tokens": 1071628384}
+{"current_steps": 24515, "total_steps": 29704, "loss": 0.29537224769592285, "lr": 7.345350739962781e-07, "epoch": 1.6506194451925666, "percentage": 82.53, "elapsed_time": "2 days, 4:41:22", "remaining_time": "11:09:09", "throughput": 5650.75, "total_tokens": 1071849024}
+{"current_steps": 24520, "total_steps": 29704, "loss": 0.29905951023101807, "lr": 7.331560970530921e-07, "epoch": 1.6509561001885267, "percentage": 82.55, "elapsed_time": "2 days, 4:42:04", "remaining_time": "11:08:31", "throughput": 5650.69, "total_tokens": 1072071032}
+{"current_steps": 24525, "total_steps": 29704, "loss": 0.30941028594970704, "lr": 7.317783133191986e-07, "epoch": 1.651292755184487, "percentage": 82.56, "elapsed_time": "2 days, 4:42:34", "remaining_time": "11:07:50", "throughput": 5650.97, "total_tokens": 1072293992}
+{"current_steps": 24530, "total_steps": 29704, "loss": 0.2879075050354004, "lr": 7.304017231798893e-07, "epoch": 1.6516294101804472, "percentage": 82.58, "elapsed_time": "2 days, 4:43:13", "remaining_time": "11:07:12", "throughput": 5650.99, "total_tokens": 1072519464}
+{"current_steps": 24535, "total_steps": 29704, "loss": 0.29528236389160156, "lr": 7.290263270201231e-07, "epoch": 1.6519660651764072, "percentage": 82.6, "elapsed_time": "2 days, 4:43:50", "remaining_time": "11:06:33", "throughput": 5651.02, "total_tokens": 1072738280}
+{"current_steps": 24540, "total_steps": 29704, "loss": 0.28215551376342773, "lr": 7.276521252245261e-07, "epoch": 1.6523027201723672, "percentage": 82.62, "elapsed_time": "2 days, 4:44:26", "remaining_time": "11:05:54", "throughput": 5651.06, "total_tokens": 1072948632}
+{"current_steps": 24545, "total_steps": 29704, "loss": 0.29338951110839845, "lr": 7.262791181773882e-07, "epoch": 1.6526393751683275, "percentage": 82.63, "elapsed_time": "2 days, 4:45:11", "remaining_time": "11:05:16", "throughput": 5650.93, "total_tokens": 1073174384}
+{"current_steps": 24550, "total_steps": 29704, "loss": 0.30844817161560056, "lr": 7.24907306262666e-07, "epoch": 1.6529760301642877, "percentage": 82.65, "elapsed_time": "2 days, 4:45:55", "remaining_time": "11:04:38", "throughput": 5650.73, "total_tokens": 1073386448}
+{"current_steps": 24555, "total_steps": 29704, "loss": 0.3130614995956421, "lr": 7.235366898639856e-07, "epoch": 1.6533126851602478, "percentage": 82.67, "elapsed_time": "2 days, 4:46:36", "remaining_time": "11:04:00", "throughput": 5650.67, "total_tokens": 1073610192}
+{"current_steps": 24560, "total_steps": 29704, "loss": 0.3168208122253418, "lr": 7.221672693646309e-07, "epoch": 1.6536493401562078, "percentage": 82.68, "elapsed_time": "2 days, 4:47:16", "remaining_time": "11:03:22", "throughput": 5650.68, "total_tokens": 1073834968}
+{"current_steps": 24565, "total_steps": 29704, "loss": 0.30944645404815674, "lr": 7.207990451475594e-07, "epoch": 1.653985995152168, "percentage": 82.7, "elapsed_time": "2 days, 4:47:49", "remaining_time": "11:02:42", "throughput": 5650.93, "total_tokens": 1074068032}
+{"current_steps": 24570, "total_steps": 29704, "loss": 0.3064124584197998, "lr": 7.194320175953901e-07, "epoch": 1.6543226501481283, "percentage": 82.72, "elapsed_time": "2 days, 4:48:34", "remaining_time": "11:02:05", "throughput": 5650.71, "total_tokens": 1074280488}
+{"current_steps": 24575, "total_steps": 29704, "loss": 0.285313081741333, "lr": 7.180661870904066e-07, "epoch": 1.6546593051440883, "percentage": 82.73, "elapsed_time": "2 days, 4:49:15", "remaining_time": "11:01:26", "throughput": 5650.66, "total_tokens": 1074502024}
+{"current_steps": 24580, "total_steps": 29704, "loss": 0.2826228618621826, "lr": 7.167015540145599e-07, "epoch": 1.6549959601400483, "percentage": 82.75, "elapsed_time": "2 days, 4:49:54", "remaining_time": "11:00:48", "throughput": 5650.6, "total_tokens": 1074710072}
+{"current_steps": 24585, "total_steps": 29704, "loss": 0.3077234745025635, "lr": 7.153381187494657e-07, "epoch": 1.6553326151360086, "percentage": 82.77, "elapsed_time": "2 days, 4:50:28", "remaining_time": "11:00:08", "throughput": 5650.7, "total_tokens": 1074926392}
+{"current_steps": 24590, "total_steps": 29704, "loss": 0.28848824501037595, "lr": 7.139758816764036e-07, "epoch": 1.6556692701319689, "percentage": 82.78, "elapsed_time": "2 days, 4:51:06", "remaining_time": "10:59:29", "throughput": 5650.67, "total_tokens": 1075131776}
+{"current_steps": 24595, "total_steps": 29704, "loss": 0.28489990234375, "lr": 7.126148431763191e-07, "epoch": 1.6560059251279289, "percentage": 82.8, "elapsed_time": "2 days, 4:51:41", "remaining_time": "10:58:50", "throughput": 5650.83, "total_tokens": 1075362760}
+{"current_steps": 24600, "total_steps": 29704, "loss": 0.30712342262268066, "lr": 7.11255003629821e-07, "epoch": 1.656342580123889, "percentage": 82.82, "elapsed_time": "2 days, 4:52:19", "remaining_time": "10:58:11", "throughput": 5650.91, "total_tokens": 1075591552}
+{"current_steps": 24605, "total_steps": 29704, "loss": 0.3017916679382324, "lr": 7.098963634171874e-07, "epoch": 1.6566792351198492, "percentage": 82.83, "elapsed_time": "2 days, 4:52:56", "remaining_time": "10:57:32", "throughput": 5651.07, "total_tokens": 1075828544}
+{"current_steps": 24610, "total_steps": 29704, "loss": 0.3059419631958008, "lr": 7.085389229183537e-07, "epoch": 1.6570158901158094, "percentage": 82.85, "elapsed_time": "2 days, 4:53:38", "remaining_time": "10:56:54", "throughput": 5650.83, "total_tokens": 1076024056}
+{"current_steps": 24615, "total_steps": 29704, "loss": 0.28927106857299806, "lr": 7.071826825129264e-07, "epoch": 1.6573525451117694, "percentage": 82.87, "elapsed_time": "2 days, 4:54:21", "remaining_time": "10:56:16", "throughput": 5650.68, "total_tokens": 1076235696}
+{"current_steps": 24620, "total_steps": 29704, "loss": 0.30096821784973143, "lr": 7.058276425801735e-07, "epoch": 1.6576892001077295, "percentage": 82.88, "elapsed_time": "2 days, 4:55:02", "remaining_time": "10:55:38", "throughput": 5650.63, "total_tokens": 1076456872}
+{"current_steps": 24625, "total_steps": 29704, "loss": 0.29817709922790525, "lr": 7.044738034990273e-07, "epoch": 1.6580258551036897, "percentage": 82.9, "elapsed_time": "2 days, 4:55:33", "remaining_time": "10:54:58", "throughput": 5650.77, "total_tokens": 1076662016}
+{"current_steps": 24630, "total_steps": 29704, "loss": 0.3162134885787964, "lr": 7.031211656480841e-07, "epoch": 1.65836251009965, "percentage": 82.92, "elapsed_time": "2 days, 4:56:13", "remaining_time": "10:54:19", "throughput": 5650.71, "total_tokens": 1076875184}
+{"current_steps": 24635, "total_steps": 29704, "loss": 0.27853074073791506, "lr": 7.017697294056058e-07, "epoch": 1.65869916509561, "percentage": 82.93, "elapsed_time": "2 days, 4:56:55", "remaining_time": "10:53:41", "throughput": 5650.6, "total_tokens": 1077089776}
+{"current_steps": 24640, "total_steps": 29704, "loss": 0.3217106819152832, "lr": 7.004194951495163e-07, "epoch": 1.65903582009157, "percentage": 82.95, "elapsed_time": "2 days, 4:57:29", "remaining_time": "10:53:02", "throughput": 5650.77, "total_tokens": 1077318536}
+{"current_steps": 24645, "total_steps": 29704, "loss": 0.29686970710754396, "lr": 6.990704632574046e-07, "epoch": 1.6593724750875303, "percentage": 82.97, "elapsed_time": "2 days, 4:58:05", "remaining_time": "10:52:23", "throughput": 5650.91, "total_tokens": 1077547440}
+{"current_steps": 24650, "total_steps": 29704, "loss": 0.299147891998291, "lr": 6.977226341065219e-07, "epoch": 1.6597091300834905, "percentage": 82.99, "elapsed_time": "2 days, 4:58:43", "remaining_time": "10:51:44", "throughput": 5650.93, "total_tokens": 1077764672}
+{"current_steps": 24655, "total_steps": 29704, "loss": 0.3067599296569824, "lr": 6.96376008073788e-07, "epoch": 1.6600457850794506, "percentage": 83.0, "elapsed_time": "2 days, 4:59:18", "remaining_time": "10:51:04", "throughput": 5650.97, "total_tokens": 1077969096}
+{"current_steps": 24660, "total_steps": 29704, "loss": 0.27429049015045165, "lr": 6.950305855357781e-07, "epoch": 1.6603824400754106, "percentage": 83.02, "elapsed_time": "2 days, 4:59:52", "remaining_time": "10:50:25", "throughput": 5651.06, "total_tokens": 1078181392}
+{"current_steps": 24665, "total_steps": 29704, "loss": 0.30942230224609374, "lr": 6.936863668687383e-07, "epoch": 1.6607190950713708, "percentage": 83.04, "elapsed_time": "2 days, 5:00:30", "remaining_time": "10:49:46", "throughput": 5651.08, "total_tokens": 1078397824}
+{"current_steps": 24670, "total_steps": 29704, "loss": 0.323834753036499, "lr": 6.923433524485745e-07, "epoch": 1.661055750067331, "percentage": 83.05, "elapsed_time": "2 days, 5:01:12", "remaining_time": "10:49:08", "throughput": 5651.0, "total_tokens": 1078622016}
+{"current_steps": 24675, "total_steps": 29704, "loss": 0.296146297454834, "lr": 6.910015426508554e-07, "epoch": 1.6613924050632911, "percentage": 83.07, "elapsed_time": "2 days, 5:01:52", "remaining_time": "10:48:29", "throughput": 5651.01, "total_tokens": 1078847744}
+{"current_steps": 24680, "total_steps": 29704, "loss": 0.309325647354126, "lr": 6.896609378508152e-07, "epoch": 1.6617290600592511, "percentage": 83.09, "elapsed_time": "2 days, 5:02:30", "remaining_time": "10:47:50", "throughput": 5650.95, "total_tokens": 1079050512}
+{"current_steps": 24685, "total_steps": 29704, "loss": 0.2978062629699707, "lr": 6.883215384233488e-07, "epoch": 1.6620657150552114, "percentage": 83.1, "elapsed_time": "2 days, 5:03:11", "remaining_time": "10:47:12", "throughput": 5650.94, "total_tokens": 1079282736}
+{"current_steps": 24690, "total_steps": 29704, "loss": 0.28819808959960935, "lr": 6.869833447430152e-07, "epoch": 1.6624023700511716, "percentage": 83.12, "elapsed_time": "2 days, 5:03:57", "remaining_time": "10:46:35", "throughput": 5650.65, "total_tokens": 1079488008}
+{"current_steps": 24695, "total_steps": 29704, "loss": 0.2954545021057129, "lr": 6.856463571840361e-07, "epoch": 1.6627390250471317, "percentage": 83.14, "elapsed_time": "2 days, 5:04:34", "remaining_time": "10:45:56", "throughput": 5650.63, "total_tokens": 1079693528}
+{"current_steps": 24700, "total_steps": 29704, "loss": 0.3160905599594116, "lr": 6.843105761202945e-07, "epoch": 1.6630756800430917, "percentage": 83.15, "elapsed_time": "2 days, 5:05:16", "remaining_time": "10:45:18", "throughput": 5650.6, "total_tokens": 1079922112}
+{"current_steps": 24705, "total_steps": 29704, "loss": 0.3114809513092041, "lr": 6.829760019253412e-07, "epoch": 1.663412335039052, "percentage": 83.17, "elapsed_time": "2 days, 5:05:53", "remaining_time": "10:44:39", "throughput": 5650.63, "total_tokens": 1080138072}
+{"current_steps": 24710, "total_steps": 29704, "loss": 0.30275850296020507, "lr": 6.816426349723804e-07, "epoch": 1.6637489900350122, "percentage": 83.19, "elapsed_time": "2 days, 5:06:31", "remaining_time": "10:44:00", "throughput": 5650.62, "total_tokens": 1080349800}
+{"current_steps": 24715, "total_steps": 29704, "loss": 0.2700210094451904, "lr": 6.803104756342877e-07, "epoch": 1.6640856450309722, "percentage": 83.2, "elapsed_time": "2 days, 5:07:07", "remaining_time": "10:43:21", "throughput": 5650.72, "total_tokens": 1080576104}
+{"current_steps": 24720, "total_steps": 29704, "loss": 0.26948862075805663, "lr": 6.789795242835962e-07, "epoch": 1.6644223000269323, "percentage": 83.22, "elapsed_time": "2 days, 5:07:49", "remaining_time": "10:42:43", "throughput": 5650.64, "total_tokens": 1080793368}
+{"current_steps": 24725, "total_steps": 29704, "loss": 0.3047426700592041, "lr": 6.776497812925015e-07, "epoch": 1.6647589550228925, "percentage": 83.24, "elapsed_time": "2 days, 5:08:27", "remaining_time": "10:42:04", "throughput": 5650.6, "total_tokens": 1081001160}
+{"current_steps": 24730, "total_steps": 29704, "loss": 0.3334406614303589, "lr": 6.763212470328628e-07, "epoch": 1.6650956100188528, "percentage": 83.25, "elapsed_time": "2 days, 5:09:08", "remaining_time": "10:41:26", "throughput": 5650.53, "total_tokens": 1081221944}
+{"current_steps": 24735, "total_steps": 29704, "loss": 0.2895099878311157, "lr": 6.749939218762003e-07, "epoch": 1.6654322650148128, "percentage": 83.27, "elapsed_time": "2 days, 5:09:50", "remaining_time": "10:40:48", "throughput": 5650.44, "total_tokens": 1081439048}
+{"current_steps": 24740, "total_steps": 29704, "loss": 0.3205715656280518, "lr": 6.736678061936958e-07, "epoch": 1.6657689200107728, "percentage": 83.29, "elapsed_time": "2 days, 5:10:32", "remaining_time": "10:40:10", "throughput": 5650.42, "total_tokens": 1081670648}
+{"current_steps": 24745, "total_steps": 29704, "loss": 0.325761604309082, "lr": 6.723429003561937e-07, "epoch": 1.666105575006733, "percentage": 83.31, "elapsed_time": "2 days, 5:11:06", "remaining_time": "10:39:30", "throughput": 5650.65, "total_tokens": 1081907840}
+{"current_steps": 24750, "total_steps": 29704, "loss": 0.2971907615661621, "lr": 6.710192047341984e-07, "epoch": 1.6664422300026933, "percentage": 83.32, "elapsed_time": "2 days, 5:11:45", "remaining_time": "10:38:52", "throughput": 5650.65, "total_tokens": 1082129744}
+{"current_steps": 24755, "total_steps": 29704, "loss": 0.29376702308654784, "lr": 6.69696719697881e-07, "epoch": 1.6667788849986533, "percentage": 83.34, "elapsed_time": "2 days, 5:12:24", "remaining_time": "10:38:13", "throughput": 5650.67, "total_tokens": 1082352688}
+{"current_steps": 24760, "total_steps": 29704, "loss": 0.27995355129241944, "lr": 6.683754456170655e-07, "epoch": 1.6671155399946134, "percentage": 83.36, "elapsed_time": "2 days, 5:12:56", "remaining_time": "10:37:33", "throughput": 5650.83, "total_tokens": 1082568000}
+{"current_steps": 24765, "total_steps": 29704, "loss": 0.28558523654937745, "lr": 6.670553828612448e-07, "epoch": 1.6674521949905736, "percentage": 83.37, "elapsed_time": "2 days, 5:13:35", "remaining_time": "10:36:54", "throughput": 5650.69, "total_tokens": 1082762664}
+{"current_steps": 24770, "total_steps": 29704, "loss": 0.2899418592453003, "lr": 6.657365317995696e-07, "epoch": 1.6677888499865339, "percentage": 83.39, "elapsed_time": "2 days, 5:14:16", "remaining_time": "10:36:16", "throughput": 5650.64, "total_tokens": 1082980808}
+{"current_steps": 24775, "total_steps": 29704, "loss": 0.31790828704833984, "lr": 6.644188928008521e-07, "epoch": 1.668125504982494, "percentage": 83.41, "elapsed_time": "2 days, 5:14:56", "remaining_time": "10:35:38", "throughput": 5650.55, "total_tokens": 1083191992}
+{"current_steps": 24780, "total_steps": 29704, "loss": 0.31216111183166506, "lr": 6.631024662335667e-07, "epoch": 1.668462159978454, "percentage": 83.42, "elapsed_time": "2 days, 5:15:45", "remaining_time": "10:35:01", "throughput": 5650.23, "total_tokens": 1083409256}
+{"current_steps": 24785, "total_steps": 29704, "loss": 0.29103267192840576, "lr": 6.617872524658469e-07, "epoch": 1.6687988149744142, "percentage": 83.44, "elapsed_time": "2 days, 5:16:21", "remaining_time": "10:34:22", "throughput": 5650.39, "total_tokens": 1083638560}
+{"current_steps": 24790, "total_steps": 29704, "loss": 0.3304905414581299, "lr": 6.604732518654882e-07, "epoch": 1.6691354699703744, "percentage": 83.46, "elapsed_time": "2 days, 5:17:03", "remaining_time": "10:33:44", "throughput": 5650.26, "total_tokens": 1083853344}
+{"current_steps": 24795, "total_steps": 29704, "loss": 0.32892765998840334, "lr": 6.591604647999473e-07, "epoch": 1.6694721249663345, "percentage": 83.47, "elapsed_time": "2 days, 5:17:40", "remaining_time": "10:33:05", "throughput": 5650.37, "total_tokens": 1084083632}
+{"current_steps": 24800, "total_steps": 29704, "loss": 0.2843611478805542, "lr": 6.57848891636339e-07, "epoch": 1.6698087799622945, "percentage": 83.49, "elapsed_time": "2 days, 5:18:23", "remaining_time": "10:32:27", "throughput": 5650.28, "total_tokens": 1084309072}
+{"current_steps": 24805, "total_steps": 29704, "loss": 0.30639631748199464, "lr": 6.565385327414441e-07, "epoch": 1.6701454349582547, "percentage": 83.51, "elapsed_time": "2 days, 5:18:57", "remaining_time": "10:31:47", "throughput": 5650.46, "total_tokens": 1084535760}
+{"current_steps": 24810, "total_steps": 29704, "loss": 0.2900113582611084, "lr": 6.552293884816968e-07, "epoch": 1.670482089954215, "percentage": 83.52, "elapsed_time": "2 days, 5:19:37", "remaining_time": "10:31:09", "throughput": 5650.46, "total_tokens": 1084761472}
+{"current_steps": 24815, "total_steps": 29704, "loss": 0.29627652168273927, "lr": 6.539214592231968e-07, "epoch": 1.670818744950175, "percentage": 83.54, "elapsed_time": "2 days, 5:20:18", "remaining_time": "10:30:31", "throughput": 5650.43, "total_tokens": 1084987552}
+{"current_steps": 24820, "total_steps": 29704, "loss": 0.26366333961486815, "lr": 6.526147453317028e-07, "epoch": 1.671155399946135, "percentage": 83.56, "elapsed_time": "2 days, 5:20:58", "remaining_time": "10:29:52", "throughput": 5650.38, "total_tokens": 1085201944}
+{"current_steps": 24825, "total_steps": 29704, "loss": 0.299773645401001, "lr": 6.513092471726307e-07, "epoch": 1.6714920549420953, "percentage": 83.57, "elapsed_time": "2 days, 5:21:39", "remaining_time": "10:29:14", "throughput": 5650.39, "total_tokens": 1085434936}
+{"current_steps": 24830, "total_steps": 29704, "loss": 0.2928581714630127, "lr": 6.500049651110635e-07, "epoch": 1.6718287099380555, "percentage": 83.59, "elapsed_time": "2 days, 5:22:19", "remaining_time": "10:28:36", "throughput": 5650.35, "total_tokens": 1085656760}
+{"current_steps": 24835, "total_steps": 29704, "loss": 0.2924539089202881, "lr": 6.487018995117339e-07, "epoch": 1.6721653649340156, "percentage": 83.61, "elapsed_time": "2 days, 5:22:54", "remaining_time": "10:27:56", "throughput": 5650.56, "total_tokens": 1085894200}
+{"current_steps": 24840, "total_steps": 29704, "loss": 0.2911822319030762, "lr": 6.474000507390455e-07, "epoch": 1.6725020199299756, "percentage": 83.63, "elapsed_time": "2 days, 5:23:28", "remaining_time": "10:27:17", "throughput": 5650.69, "total_tokens": 1086112280}
+{"current_steps": 24845, "total_steps": 29704, "loss": 0.3034616470336914, "lr": 6.460994191570513e-07, "epoch": 1.6728386749259359, "percentage": 83.64, "elapsed_time": "2 days, 5:24:06", "remaining_time": "10:26:38", "throughput": 5650.75, "total_tokens": 1086334184}
+{"current_steps": 24850, "total_steps": 29704, "loss": 0.2885445594787598, "lr": 6.448000051294717e-07, "epoch": 1.673175329921896, "percentage": 83.66, "elapsed_time": "2 days, 5:24:44", "remaining_time": "10:25:59", "throughput": 5650.75, "total_tokens": 1086553800}
+{"current_steps": 24855, "total_steps": 29704, "loss": 0.30068202018737794, "lr": 6.435018090196832e-07, "epoch": 1.6735119849178561, "percentage": 83.68, "elapsed_time": "2 days, 5:25:21", "remaining_time": "10:25:20", "throughput": 5650.86, "total_tokens": 1086783696}
+{"current_steps": 24860, "total_steps": 29704, "loss": 0.2921877861022949, "lr": 6.42204831190722e-07, "epoch": 1.6738486399138162, "percentage": 83.69, "elapsed_time": "2 days, 5:25:56", "remaining_time": "10:24:40", "throughput": 5651.01, "total_tokens": 1087006792}
+{"current_steps": 24865, "total_steps": 29704, "loss": 0.30269834995269773, "lr": 6.409090720052835e-07, "epoch": 1.6741852949097764, "percentage": 83.71, "elapsed_time": "2 days, 5:26:36", "remaining_time": "10:24:02", "throughput": 5650.95, "total_tokens": 1087221944}
+{"current_steps": 24870, "total_steps": 29704, "loss": 0.27388882637023926, "lr": 6.396145318257229e-07, "epoch": 1.6745219499057367, "percentage": 83.73, "elapsed_time": "2 days, 5:27:13", "remaining_time": "10:23:23", "throughput": 5651.01, "total_tokens": 1087443824}
+{"current_steps": 24875, "total_steps": 29704, "loss": 0.31777362823486327, "lr": 6.383212110140535e-07, "epoch": 1.6748586049016967, "percentage": 83.74, "elapsed_time": "2 days, 5:27:49", "remaining_time": "10:22:44", "throughput": 5651.08, "total_tokens": 1087659568}
+{"current_steps": 24880, "total_steps": 29704, "loss": 0.3115004301071167, "lr": 6.370291099319509e-07, "epoch": 1.6751952598976567, "percentage": 83.76, "elapsed_time": "2 days, 5:28:25", "remaining_time": "10:22:04", "throughput": 5651.02, "total_tokens": 1087852096}
+{"current_steps": 24885, "total_steps": 29704, "loss": 0.29903695583343504, "lr": 6.357382289407433e-07, "epoch": 1.675531914893617, "percentage": 83.78, "elapsed_time": "2 days, 5:29:00", "remaining_time": "10:21:25", "throughput": 5651.19, "total_tokens": 1088082304}
+{"current_steps": 24890, "total_steps": 29704, "loss": 0.3122910499572754, "lr": 6.344485684014251e-07, "epoch": 1.6758685698895772, "percentage": 83.79, "elapsed_time": "2 days, 5:29:38", "remaining_time": "10:20:46", "throughput": 5651.17, "total_tokens": 1088296296}
+{"current_steps": 24895, "total_steps": 29704, "loss": 0.31540684700012206, "lr": 6.33160128674642e-07, "epoch": 1.6762052248855372, "percentage": 83.81, "elapsed_time": "2 days, 5:30:15", "remaining_time": "10:20:07", "throughput": 5651.16, "total_tokens": 1088501864}
+{"current_steps": 24900, "total_steps": 29704, "loss": 0.2925147533416748, "lr": 6.318729101207055e-07, "epoch": 1.6765418798814973, "percentage": 83.83, "elapsed_time": "2 days, 5:30:57", "remaining_time": "10:19:29", "throughput": 5651.1, "total_tokens": 1088727288}
+{"current_steps": 24905, "total_steps": 29704, "loss": 0.30594768524169924, "lr": 6.305869130995807e-07, "epoch": 1.6768785348774575, "percentage": 83.84, "elapsed_time": "2 days, 5:31:32", "remaining_time": "10:18:50", "throughput": 5651.23, "total_tokens": 1088947848}
+{"current_steps": 24910, "total_steps": 29704, "loss": 0.30313916206359864, "lr": 6.293021379708924e-07, "epoch": 1.6772151898734178, "percentage": 83.86, "elapsed_time": "2 days, 5:32:12", "remaining_time": "10:18:11", "throughput": 5651.21, "total_tokens": 1089171240}
+{"current_steps": 24915, "total_steps": 29704, "loss": 0.31810295581817627, "lr": 6.280185850939241e-07, "epoch": 1.6775518448693778, "percentage": 83.88, "elapsed_time": "2 days, 5:32:54", "remaining_time": "10:17:33", "throughput": 5651.11, "total_tokens": 1089387216}
+{"current_steps": 24920, "total_steps": 29704, "loss": 0.30693349838256834, "lr": 6.267362548276173e-07, "epoch": 1.677888499865338, "percentage": 83.89, "elapsed_time": "2 days, 5:33:31", "remaining_time": "10:16:54", "throughput": 5651.1, "total_tokens": 1089597552}
+{"current_steps": 24925, "total_steps": 29704, "loss": 0.30268676280975343, "lr": 6.254551475305703e-07, "epoch": 1.6782251548612983, "percentage": 83.91, "elapsed_time": "2 days, 5:34:09", "remaining_time": "10:16:16", "throughput": 5650.98, "total_tokens": 1089790616}
+{"current_steps": 24930, "total_steps": 29704, "loss": 0.31978979110717776, "lr": 6.241752635610443e-07, "epoch": 1.6785618098572583, "percentage": 83.93, "elapsed_time": "2 days, 5:34:46", "remaining_time": "10:15:36", "throughput": 5651.04, "total_tokens": 1090007856}
+{"current_steps": 24935, "total_steps": 29704, "loss": 0.31137659549713137, "lr": 6.228966032769496e-07, "epoch": 1.6788984648532184, "percentage": 83.94, "elapsed_time": "2 days, 5:35:22", "remaining_time": "10:14:57", "throughput": 5651.09, "total_tokens": 1090222152}
+{"current_steps": 24940, "total_steps": 29704, "loss": 0.29123806953430176, "lr": 6.216191670358651e-07, "epoch": 1.6792351198491786, "percentage": 83.96, "elapsed_time": "2 days, 5:36:05", "remaining_time": "10:14:19", "throughput": 5650.95, "total_tokens": 1090437952}
+{"current_steps": 24945, "total_steps": 29704, "loss": 0.32757062911987306, "lr": 6.203429551950163e-07, "epoch": 1.6795717748451389, "percentage": 83.98, "elapsed_time": "2 days, 5:36:51", "remaining_time": "10:13:42", "throughput": 5650.77, "total_tokens": 1090662752}
+{"current_steps": 24950, "total_steps": 29704, "loss": 0.28361237049102783, "lr": 6.19067968111296e-07, "epoch": 1.679908429841099, "percentage": 84.0, "elapsed_time": "2 days, 5:37:28", "remaining_time": "10:13:03", "throughput": 5650.75, "total_tokens": 1090869792}
+{"current_steps": 24955, "total_steps": 29704, "loss": 0.2768688678741455, "lr": 6.177942061412478e-07, "epoch": 1.680245084837059, "percentage": 84.01, "elapsed_time": "2 days, 5:38:08", "remaining_time": "10:12:25", "throughput": 5650.7, "total_tokens": 1091083944}
+{"current_steps": 24960, "total_steps": 29704, "loss": 0.2864533424377441, "lr": 6.16521669641077e-07, "epoch": 1.6805817398330192, "percentage": 84.03, "elapsed_time": "2 days, 5:38:40", "remaining_time": "10:11:45", "throughput": 5650.94, "total_tokens": 1091313248}
+{"current_steps": 24965, "total_steps": 29704, "loss": 0.30624988079071047, "lr": 6.152503589666426e-07, "epoch": 1.6809183948289794, "percentage": 84.05, "elapsed_time": "2 days, 5:39:17", "remaining_time": "10:11:06", "throughput": 5651.05, "total_tokens": 1091544272}
+{"current_steps": 24970, "total_steps": 29704, "loss": 0.26999435424804685, "lr": 6.139802744734636e-07, "epoch": 1.6812550498249395, "percentage": 84.06, "elapsed_time": "2 days, 5:39:53", "remaining_time": "10:10:27", "throughput": 5651.13, "total_tokens": 1091761648}
+{"current_steps": 24975, "total_steps": 29704, "loss": 0.26949467658996584, "lr": 6.127114165167142e-07, "epoch": 1.6815917048208995, "percentage": 84.08, "elapsed_time": "2 days, 5:40:28", "remaining_time": "10:09:47", "throughput": 5651.17, "total_tokens": 1091969864}
+{"current_steps": 24980, "total_steps": 29704, "loss": 0.30141243934631345, "lr": 6.11443785451229e-07, "epoch": 1.6819283598168597, "percentage": 84.1, "elapsed_time": "2 days, 5:41:09", "remaining_time": "10:09:09", "throughput": 5651.13, "total_tokens": 1092190952}
+{"current_steps": 24985, "total_steps": 29704, "loss": 0.29298968315124513, "lr": 6.101773816314926e-07, "epoch": 1.68226501481282, "percentage": 84.11, "elapsed_time": "2 days, 5:41:46", "remaining_time": "10:08:30", "throughput": 5651.21, "total_tokens": 1092417104}
+{"current_steps": 24990, "total_steps": 29704, "loss": 0.3086401462554932, "lr": 6.089122054116558e-07, "epoch": 1.68260166980878, "percentage": 84.13, "elapsed_time": "2 days, 5:42:22", "remaining_time": "10:07:51", "throughput": 5651.36, "total_tokens": 1092645416}
+{"current_steps": 24995, "total_steps": 29704, "loss": 0.2916858196258545, "lr": 6.076482571455161e-07, "epoch": 1.68293832480474, "percentage": 84.15, "elapsed_time": "2 days, 5:43:03", "remaining_time": "10:07:12", "throughput": 5651.31, "total_tokens": 1092868304}
+{"current_steps": 25000, "total_steps": 29704, "loss": 0.30973148345947266, "lr": 6.063855371865369e-07, "epoch": 1.6832749798007003, "percentage": 84.16, "elapsed_time": "2 days, 5:43:40", "remaining_time": "10:06:33", "throughput": 5651.29, "total_tokens": 1093073112}
+{"current_steps": 25005, "total_steps": 29704, "loss": 0.3318582057952881, "lr": 6.051240458878316e-07, "epoch": 1.6836116347966605, "percentage": 84.18, "elapsed_time": "2 days, 5:44:20", "remaining_time": "10:05:55", "throughput": 5651.25, "total_tokens": 1093292280}
+{"current_steps": 25010, "total_steps": 29704, "loss": 0.30891571044921873, "lr": 6.038637836021721e-07, "epoch": 1.6839482897926206, "percentage": 84.2, "elapsed_time": "2 days, 5:45:00", "remaining_time": "10:05:17", "throughput": 5651.23, "total_tokens": 1093512488}
+{"current_steps": 25015, "total_steps": 29704, "loss": 0.29696998596191404, "lr": 6.026047506819882e-07, "epoch": 1.6842849447885806, "percentage": 84.21, "elapsed_time": "2 days, 5:45:35", "remaining_time": "10:04:37", "throughput": 5651.27, "total_tokens": 1093724280}
+{"current_steps": 25020, "total_steps": 29704, "loss": 0.2949728012084961, "lr": 6.013469474793632e-07, "epoch": 1.6846215997845408, "percentage": 84.23, "elapsed_time": "2 days, 5:46:15", "remaining_time": "10:03:59", "throughput": 5651.26, "total_tokens": 1093945480}
+{"current_steps": 25025, "total_steps": 29704, "loss": 0.2946459770202637, "lr": 6.000903743460368e-07, "epoch": 1.684958254780501, "percentage": 84.25, "elapsed_time": "2 days, 5:46:59", "remaining_time": "10:03:21", "throughput": 5651.12, "total_tokens": 1094164496}
+{"current_steps": 25030, "total_steps": 29704, "loss": 0.296980881690979, "lr": 5.98835031633409e-07, "epoch": 1.6852949097764611, "percentage": 84.26, "elapsed_time": "2 days, 5:47:38", "remaining_time": "10:02:42", "throughput": 5651.14, "total_tokens": 1094387696}
+{"current_steps": 25035, "total_steps": 29704, "loss": 0.30903480052947996, "lr": 5.97580919692528e-07, "epoch": 1.6856315647724212, "percentage": 84.28, "elapsed_time": "2 days, 5:48:10", "remaining_time": "10:02:03", "throughput": 5651.4, "total_tokens": 1094621216}
+{"current_steps": 25040, "total_steps": 29704, "loss": 0.2778623104095459, "lr": 5.963280388741072e-07, "epoch": 1.6859682197683814, "percentage": 84.3, "elapsed_time": "2 days, 5:48:46", "remaining_time": "10:01:23", "throughput": 5651.5, "total_tokens": 1094843312}
+{"current_steps": 25045, "total_steps": 29704, "loss": 0.3040936946868896, "lr": 5.950763895285055e-07, "epoch": 1.6863048747643417, "percentage": 84.32, "elapsed_time": "2 days, 5:49:26", "remaining_time": "10:00:45", "throughput": 5651.37, "total_tokens": 1095048760}
+{"current_steps": 25050, "total_steps": 29704, "loss": 0.29845240116119387, "lr": 5.938259720057465e-07, "epoch": 1.6866415297603017, "percentage": 84.33, "elapsed_time": "2 days, 5:50:03", "remaining_time": "10:00:06", "throughput": 5651.47, "total_tokens": 1095275144}
+{"current_steps": 25055, "total_steps": 29704, "loss": 0.28672802448272705, "lr": 5.925767866555043e-07, "epoch": 1.6869781847562617, "percentage": 84.35, "elapsed_time": "2 days, 5:50:46", "remaining_time": "9:59:28", "throughput": 5651.33, "total_tokens": 1095488208}
+{"current_steps": 25060, "total_steps": 29704, "loss": 0.31831324100494385, "lr": 5.913288338271095e-07, "epoch": 1.687314839752222, "percentage": 84.37, "elapsed_time": "2 days, 5:51:24", "remaining_time": "9:58:49", "throughput": 5651.35, "total_tokens": 1095711504}
+{"current_steps": 25065, "total_steps": 29704, "loss": 0.29593048095703123, "lr": 5.900821138695478e-07, "epoch": 1.6876514947481822, "percentage": 84.38, "elapsed_time": "2 days, 5:52:04", "remaining_time": "9:58:11", "throughput": 5651.33, "total_tokens": 1095931728}
+{"current_steps": 25070, "total_steps": 29704, "loss": 0.2997978448867798, "lr": 5.888366271314605e-07, "epoch": 1.6879881497441422, "percentage": 84.4, "elapsed_time": "2 days, 5:52:50", "remaining_time": "9:57:33", "throughput": 5651.16, "total_tokens": 1096158920}
+{"current_steps": 25075, "total_steps": 29704, "loss": 0.3102496385574341, "lr": 5.875923739611439e-07, "epoch": 1.6883248047401023, "percentage": 84.42, "elapsed_time": "2 days, 5:53:25", "remaining_time": "9:56:54", "throughput": 5651.31, "total_tokens": 1096386400}
+{"current_steps": 25080, "total_steps": 29704, "loss": 0.288178825378418, "lr": 5.863493547065513e-07, "epoch": 1.6886614597360625, "percentage": 84.43, "elapsed_time": "2 days, 5:54:06", "remaining_time": "9:56:16", "throughput": 5651.21, "total_tokens": 1096598208}
+{"current_steps": 25085, "total_steps": 29704, "loss": 0.2890732765197754, "lr": 5.851075697152858e-07, "epoch": 1.6889981147320228, "percentage": 84.45, "elapsed_time": "2 days, 5:54:49", "remaining_time": "9:55:38", "throughput": 5651.17, "total_tokens": 1096830840}
+{"current_steps": 25090, "total_steps": 29704, "loss": 0.3075998067855835, "lr": 5.838670193346118e-07, "epoch": 1.6893347697279828, "percentage": 84.47, "elapsed_time": "2 days, 5:55:31", "remaining_time": "9:55:00", "throughput": 5651.04, "total_tokens": 1097042768}
+{"current_steps": 25095, "total_steps": 29704, "loss": 0.2823533773422241, "lr": 5.826277039114426e-07, "epoch": 1.6896714247239428, "percentage": 84.48, "elapsed_time": "2 days, 5:56:13", "remaining_time": "9:54:22", "throughput": 5650.92, "total_tokens": 1097258936}
+{"current_steps": 25100, "total_steps": 29704, "loss": 0.33135075569152833, "lr": 5.813896237923505e-07, "epoch": 1.690008079719903, "percentage": 84.5, "elapsed_time": "2 days, 5:56:58", "remaining_time": "9:53:44", "throughput": 5650.75, "total_tokens": 1097479600}
+{"current_steps": 25105, "total_steps": 29704, "loss": 0.2957238912582397, "lr": 5.801527793235606e-07, "epoch": 1.6903447347158633, "percentage": 84.52, "elapsed_time": "2 days, 5:57:38", "remaining_time": "9:53:06", "throughput": 5650.77, "total_tokens": 1097708984}
+{"current_steps": 25110, "total_steps": 29704, "loss": 0.28771932125091554, "lr": 5.789171708509516e-07, "epoch": 1.6906813897118234, "percentage": 84.53, "elapsed_time": "2 days, 5:58:19", "remaining_time": "9:52:28", "throughput": 5650.59, "total_tokens": 1097906256}
+{"current_steps": 25115, "total_steps": 29704, "loss": 0.30194315910339353, "lr": 5.776827987200584e-07, "epoch": 1.6910180447077834, "percentage": 84.55, "elapsed_time": "2 days, 5:58:56", "remaining_time": "9:51:49", "throughput": 5650.65, "total_tokens": 1098128136}
+{"current_steps": 25120, "total_steps": 29704, "loss": 0.2923448085784912, "lr": 5.764496632760691e-07, "epoch": 1.6913546997037436, "percentage": 84.57, "elapsed_time": "2 days, 5:59:33", "remaining_time": "9:51:10", "throughput": 5650.64, "total_tokens": 1098334664}
+{"current_steps": 25125, "total_steps": 29704, "loss": 0.27889845371246336, "lr": 5.752177648638241e-07, "epoch": 1.6916913546997039, "percentage": 84.58, "elapsed_time": "2 days, 6:00:08", "remaining_time": "9:50:30", "throughput": 5650.78, "total_tokens": 1098559728}
+{"current_steps": 25130, "total_steps": 29704, "loss": 0.3101338863372803, "lr": 5.739871038278239e-07, "epoch": 1.692028009695664, "percentage": 84.6, "elapsed_time": "2 days, 6:00:40", "remaining_time": "9:49:50", "throughput": 5651.0, "total_tokens": 1098782712}
+{"current_steps": 25135, "total_steps": 29704, "loss": 0.28267052173614504, "lr": 5.727576805122142e-07, "epoch": 1.692364664691624, "percentage": 84.62, "elapsed_time": "2 days, 6:01:20", "remaining_time": "9:49:12", "throughput": 5650.91, "total_tokens": 1098991872}
+{"current_steps": 25140, "total_steps": 29704, "loss": 0.30869436264038086, "lr": 5.715294952608041e-07, "epoch": 1.6927013196875842, "percentage": 84.64, "elapsed_time": "2 days, 6:01:52", "remaining_time": "9:48:32", "throughput": 5651.19, "total_tokens": 1099224528}
+{"current_steps": 25145, "total_steps": 29704, "loss": 0.3042963981628418, "lr": 5.703025484170471e-07, "epoch": 1.6930379746835444, "percentage": 84.65, "elapsed_time": "2 days, 6:02:28", "remaining_time": "9:47:53", "throughput": 5651.32, "total_tokens": 1099455640}
+{"current_steps": 25150, "total_steps": 29704, "loss": 0.3153714656829834, "lr": 5.690768403240587e-07, "epoch": 1.6933746296795045, "percentage": 84.67, "elapsed_time": "2 days, 6:03:05", "remaining_time": "9:47:14", "throughput": 5651.38, "total_tokens": 1099676720}
+{"current_steps": 25155, "total_steps": 29704, "loss": 0.30834920406341554, "lr": 5.67852371324602e-07, "epoch": 1.6937112846754645, "percentage": 84.69, "elapsed_time": "2 days, 6:03:40", "remaining_time": "9:46:34", "throughput": 5651.55, "total_tokens": 1099907704}
+{"current_steps": 25160, "total_steps": 29704, "loss": 0.2930551528930664, "lr": 5.66629141761097e-07, "epoch": 1.6940479396714248, "percentage": 84.7, "elapsed_time": "2 days, 6:04:17", "remaining_time": "9:45:55", "throughput": 5651.69, "total_tokens": 1100142256}
+{"current_steps": 25165, "total_steps": 29704, "loss": 0.30519957542419435, "lr": 5.654071519756154e-07, "epoch": 1.694384594667385, "percentage": 84.72, "elapsed_time": "2 days, 6:05:04", "remaining_time": "9:45:18", "throughput": 5651.48, "total_tokens": 1100366192}
+{"current_steps": 25170, "total_steps": 29704, "loss": 0.33337717056274413, "lr": 5.641864023098831e-07, "epoch": 1.694721249663345, "percentage": 84.74, "elapsed_time": "2 days, 6:05:45", "remaining_time": "9:44:40", "throughput": 5651.39, "total_tokens": 1100584800}
+{"current_steps": 25175, "total_steps": 29704, "loss": 0.29319195747375487, "lr": 5.629668931052773e-07, "epoch": 1.695057904659305, "percentage": 84.75, "elapsed_time": "2 days, 6:06:25", "remaining_time": "9:44:01", "throughput": 5651.45, "total_tokens": 1100817816}
+{"current_steps": 25180, "total_steps": 29704, "loss": 0.31841015815734863, "lr": 5.61748624702833e-07, "epoch": 1.6953945596552653, "percentage": 84.77, "elapsed_time": "2 days, 6:07:00", "remaining_time": "9:43:22", "throughput": 5651.58, "total_tokens": 1101043616}
+{"current_steps": 25185, "total_steps": 29704, "loss": 0.29940404891967776, "lr": 5.605315974432313e-07, "epoch": 1.6957312146512256, "percentage": 84.79, "elapsed_time": "2 days, 6:07:43", "remaining_time": "9:42:44", "throughput": 5651.46, "total_tokens": 1101262632}
+{"current_steps": 25190, "total_steps": 29704, "loss": 0.2952594757080078, "lr": 5.593158116668146e-07, "epoch": 1.6960678696471856, "percentage": 84.8, "elapsed_time": "2 days, 6:08:24", "remaining_time": "9:42:06", "throughput": 5651.5, "total_tokens": 1101502480}
+{"current_steps": 25195, "total_steps": 29704, "loss": 0.2999080181121826, "lr": 5.581012677135683e-07, "epoch": 1.6964045246431456, "percentage": 84.82, "elapsed_time": "2 days, 6:08:59", "remaining_time": "9:41:27", "throughput": 5651.54, "total_tokens": 1101706720}
+{"current_steps": 25200, "total_steps": 29704, "loss": 0.30147387981414797, "lr": 5.56887965923139e-07, "epoch": 1.6967411796391059, "percentage": 84.84, "elapsed_time": "2 days, 6:09:40", "remaining_time": "9:40:48", "throughput": 5651.51, "total_tokens": 1101936320}
+{"current_steps": 25205, "total_steps": 29704, "loss": 0.29673826694488525, "lr": 5.556759066348227e-07, "epoch": 1.6970778346350661, "percentage": 84.85, "elapsed_time": "2 days, 6:10:18", "remaining_time": "9:40:10", "throughput": 5651.55, "total_tokens": 1102155368}
+{"current_steps": 25210, "total_steps": 29704, "loss": 0.3003216743469238, "lr": 5.54465090187567e-07, "epoch": 1.6974144896310261, "percentage": 84.87, "elapsed_time": "2 days, 6:10:53", "remaining_time": "9:39:30", "throughput": 5651.62, "total_tokens": 1102367352}
+{"current_steps": 25215, "total_steps": 29704, "loss": 0.29103655815124513, "lr": 5.532555169199727e-07, "epoch": 1.6977511446269862, "percentage": 84.89, "elapsed_time": "2 days, 6:11:27", "remaining_time": "9:38:51", "throughput": 5651.74, "total_tokens": 1102583040}
+{"current_steps": 25220, "total_steps": 29704, "loss": 0.28641371726989745, "lr": 5.520471871702937e-07, "epoch": 1.6980877996229464, "percentage": 84.9, "elapsed_time": "2 days, 6:12:07", "remaining_time": "9:38:12", "throughput": 5651.74, "total_tokens": 1102810576}
+{"current_steps": 25225, "total_steps": 29704, "loss": 0.2999880313873291, "lr": 5.508401012764341e-07, "epoch": 1.6984244546189067, "percentage": 84.92, "elapsed_time": "2 days, 6:12:37", "remaining_time": "9:37:32", "throughput": 5651.98, "total_tokens": 1103028192}
+{"current_steps": 25230, "total_steps": 29704, "loss": 0.28559165000915526, "lr": 5.496342595759552e-07, "epoch": 1.6987611096148667, "percentage": 84.94, "elapsed_time": "2 days, 6:13:24", "remaining_time": "9:36:55", "throughput": 5651.81, "total_tokens": 1103256944}
+{"current_steps": 25235, "total_steps": 29704, "loss": 0.2883117198944092, "lr": 5.484296624060626e-07, "epoch": 1.6990977646108267, "percentage": 84.95, "elapsed_time": "2 days, 6:14:09", "remaining_time": "9:36:17", "throughput": 5651.65, "total_tokens": 1103482616}
+{"current_steps": 25240, "total_steps": 29704, "loss": 0.29885265827178953, "lr": 5.472263101036212e-07, "epoch": 1.699434419606787, "percentage": 84.97, "elapsed_time": "2 days, 6:14:45", "remaining_time": "9:35:38", "throughput": 5651.76, "total_tokens": 1103706952}
+{"current_steps": 25245, "total_steps": 29704, "loss": 0.3131263732910156, "lr": 5.460242030051421e-07, "epoch": 1.6997710746027472, "percentage": 84.99, "elapsed_time": "2 days, 6:15:22", "remaining_time": "9:34:59", "throughput": 5651.76, "total_tokens": 1103918168}
+{"current_steps": 25250, "total_steps": 29704, "loss": 0.3075102806091309, "lr": 5.448233414467924e-07, "epoch": 1.7001077295987073, "percentage": 85.01, "elapsed_time": "2 days, 6:15:58", "remaining_time": "9:34:20", "throughput": 5651.92, "total_tokens": 1104153064}
+{"current_steps": 25255, "total_steps": 29704, "loss": 0.3154341697692871, "lr": 5.43623725764389e-07, "epoch": 1.7004443845946673, "percentage": 85.02, "elapsed_time": "2 days, 6:16:36", "remaining_time": "9:33:41", "throughput": 5651.94, "total_tokens": 1104367200}
+{"current_steps": 25260, "total_steps": 29704, "loss": 0.29883978366851804, "lr": 5.424253562934001e-07, "epoch": 1.7007810395906275, "percentage": 85.04, "elapsed_time": "2 days, 6:17:21", "remaining_time": "9:33:04", "throughput": 5651.81, "total_tokens": 1104597088}
+{"current_steps": 25265, "total_steps": 29704, "loss": 0.29061620235443114, "lr": 5.412282333689461e-07, "epoch": 1.7011176945865878, "percentage": 85.06, "elapsed_time": "2 days, 6:18:00", "remaining_time": "9:32:25", "throughput": 5651.84, "total_tokens": 1104824056}
+{"current_steps": 25270, "total_steps": 29704, "loss": 0.29516527652740476, "lr": 5.400323573257987e-07, "epoch": 1.7014543495825478, "percentage": 85.07, "elapsed_time": "2 days, 6:18:37", "remaining_time": "9:31:46", "throughput": 5651.94, "total_tokens": 1105049880}
+{"current_steps": 25275, "total_steps": 29704, "loss": 0.287148642539978, "lr": 5.388377284983803e-07, "epoch": 1.7017910045785078, "percentage": 85.09, "elapsed_time": "2 days, 6:19:13", "remaining_time": "9:31:07", "throughput": 5652.05, "total_tokens": 1105277608}
+{"current_steps": 25280, "total_steps": 29704, "loss": 0.2909886360168457, "lr": 5.37644347220767e-07, "epoch": 1.702127659574468, "percentage": 85.11, "elapsed_time": "2 days, 6:19:53", "remaining_time": "9:30:28", "throughput": 5651.98, "total_tokens": 1105492920}
+{"current_steps": 25285, "total_steps": 29704, "loss": 0.30830817222595214, "lr": 5.364522138266809e-07, "epoch": 1.7024643145704283, "percentage": 85.12, "elapsed_time": "2 days, 6:20:29", "remaining_time": "9:29:49", "throughput": 5652.13, "total_tokens": 1105725216}
+{"current_steps": 25290, "total_steps": 29704, "loss": 0.2986649990081787, "lr": 5.352613286495018e-07, "epoch": 1.7028009695663884, "percentage": 85.14, "elapsed_time": "2 days, 6:21:12", "remaining_time": "9:29:11", "throughput": 5651.96, "total_tokens": 1105936272}
+{"current_steps": 25295, "total_steps": 29704, "loss": 0.28066387176513674, "lr": 5.340716920222527e-07, "epoch": 1.7031376245623484, "percentage": 85.16, "elapsed_time": "2 days, 6:21:51", "remaining_time": "9:28:33", "throughput": 5651.98, "total_tokens": 1106155736}
+{"current_steps": 25300, "total_steps": 29704, "loss": 0.3220468521118164, "lr": 5.32883304277616e-07, "epoch": 1.7034742795583087, "percentage": 85.17, "elapsed_time": "2 days, 6:22:34", "remaining_time": "9:27:55", "throughput": 5651.82, "total_tokens": 1106367880}
+{"current_steps": 25305, "total_steps": 29704, "loss": 0.2808229446411133, "lr": 5.31696165747918e-07, "epoch": 1.703810934554269, "percentage": 85.19, "elapsed_time": "2 days, 6:23:13", "remaining_time": "9:27:16", "throughput": 5651.82, "total_tokens": 1106591160}
+{"current_steps": 25310, "total_steps": 29704, "loss": 0.28664450645446776, "lr": 5.305102767651393e-07, "epoch": 1.704147589550229, "percentage": 85.21, "elapsed_time": "2 days, 6:24:00", "remaining_time": "9:26:39", "throughput": 5651.63, "total_tokens": 1106820392}
+{"current_steps": 25315, "total_steps": 29704, "loss": 0.30523157119750977, "lr": 5.293256376609091e-07, "epoch": 1.704484244546189, "percentage": 85.22, "elapsed_time": "2 days, 6:24:42", "remaining_time": "9:26:01", "throughput": 5651.55, "total_tokens": 1107043152}
+{"current_steps": 25320, "total_steps": 29704, "loss": 0.27099974155426027, "lr": 5.281422487665089e-07, "epoch": 1.7048208995421492, "percentage": 85.24, "elapsed_time": "2 days, 6:25:22", "remaining_time": "9:25:22", "throughput": 5651.6, "total_tokens": 1107273600}
+{"current_steps": 25325, "total_steps": 29704, "loss": 0.2887356042861938, "lr": 5.269601104128674e-07, "epoch": 1.7051575545381095, "percentage": 85.26, "elapsed_time": "2 days, 6:26:08", "remaining_time": "9:24:45", "throughput": 5651.42, "total_tokens": 1107500568}
+{"current_steps": 25330, "total_steps": 29704, "loss": 0.3096314430236816, "lr": 5.257792229305702e-07, "epoch": 1.7054942095340695, "percentage": 85.27, "elapsed_time": "2 days, 6:26:49", "remaining_time": "9:24:06", "throughput": 5651.38, "total_tokens": 1107721408}
+{"current_steps": 25335, "total_steps": 29704, "loss": 0.27949187755584715, "lr": 5.245995866498438e-07, "epoch": 1.7058308645300295, "percentage": 85.29, "elapsed_time": "2 days, 6:27:30", "remaining_time": "9:23:28", "throughput": 5651.25, "total_tokens": 1107927904}
+{"current_steps": 25340, "total_steps": 29704, "loss": 0.3069715738296509, "lr": 5.234212019005741e-07, "epoch": 1.7061675195259898, "percentage": 85.31, "elapsed_time": "2 days, 6:28:12", "remaining_time": "9:22:50", "throughput": 5651.09, "total_tokens": 1108137040}
+{"current_steps": 25345, "total_steps": 29704, "loss": 0.3008026838302612, "lr": 5.222440690122882e-07, "epoch": 1.70650417452195, "percentage": 85.33, "elapsed_time": "2 days, 6:28:53", "remaining_time": "9:22:12", "throughput": 5651.02, "total_tokens": 1108352000}
+{"current_steps": 25350, "total_steps": 29704, "loss": 0.2749685764312744, "lr": 5.210681883141716e-07, "epoch": 1.70684082951791, "percentage": 85.34, "elapsed_time": "2 days, 6:29:29", "remaining_time": "9:21:33", "throughput": 5651.09, "total_tokens": 1108571744}
+{"current_steps": 25355, "total_steps": 29704, "loss": 0.3137876749038696, "lr": 5.19893560135053e-07, "epoch": 1.70717748451387, "percentage": 85.36, "elapsed_time": "2 days, 6:30:03", "remaining_time": "9:20:53", "throughput": 5651.28, "total_tokens": 1108803560}
+{"current_steps": 25360, "total_steps": 29704, "loss": 0.3042750835418701, "lr": 5.187201848034146e-07, "epoch": 1.7075141395098303, "percentage": 85.38, "elapsed_time": "2 days, 6:30:45", "remaining_time": "9:20:15", "throughput": 5651.21, "total_tokens": 1109023488}
+{"current_steps": 25365, "total_steps": 29704, "loss": 0.290797758102417, "lr": 5.175480626473873e-07, "epoch": 1.7078507945057906, "percentage": 85.39, "elapsed_time": "2 days, 6:31:24", "remaining_time": "9:19:36", "throughput": 5651.16, "total_tokens": 1109234112}
+{"current_steps": 25370, "total_steps": 29704, "loss": 0.30904016494750974, "lr": 5.163771939947504e-07, "epoch": 1.7081874495017506, "percentage": 85.41, "elapsed_time": "2 days, 6:32:04", "remaining_time": "9:18:58", "throughput": 5651.13, "total_tokens": 1109454048}
+{"current_steps": 25375, "total_steps": 29704, "loss": 0.3244988203048706, "lr": 5.152075791729327e-07, "epoch": 1.7085241044977106, "percentage": 85.43, "elapsed_time": "2 days, 6:32:44", "remaining_time": "9:18:19", "throughput": 5651.18, "total_tokens": 1109687648}
+{"current_steps": 25380, "total_steps": 29704, "loss": 0.2983409881591797, "lr": 5.140392185090165e-07, "epoch": 1.7088607594936709, "percentage": 85.44, "elapsed_time": "2 days, 6:33:25", "remaining_time": "9:17:41", "throughput": 5651.02, "total_tokens": 1109893064}
+{"current_steps": 25385, "total_steps": 29704, "loss": 0.3025853157043457, "lr": 5.128721123297265e-07, "epoch": 1.7091974144896311, "percentage": 85.46, "elapsed_time": "2 days, 6:34:08", "remaining_time": "9:17:03", "throughput": 5650.99, "total_tokens": 1110127392}
+{"current_steps": 25390, "total_steps": 29704, "loss": 0.2883389949798584, "lr": 5.117062609614438e-07, "epoch": 1.7095340694855912, "percentage": 85.48, "elapsed_time": "2 days, 6:34:44", "remaining_time": "9:16:24", "throughput": 5651.08, "total_tokens": 1110349016}
+{"current_steps": 25395, "total_steps": 29704, "loss": 0.2935046195983887, "lr": 5.105416647301909e-07, "epoch": 1.7098707244815512, "percentage": 85.49, "elapsed_time": "2 days, 6:35:27", "remaining_time": "9:15:46", "throughput": 5650.97, "total_tokens": 1110572696}
+{"current_steps": 25400, "total_steps": 29704, "loss": 0.26650362014770507, "lr": 5.093783239616468e-07, "epoch": 1.7102073794775114, "percentage": 85.51, "elapsed_time": "2 days, 6:36:05", "remaining_time": "9:15:07", "throughput": 5650.94, "total_tokens": 1110779632}
+{"current_steps": 25405, "total_steps": 29704, "loss": 0.28749451637268064, "lr": 5.082162389811352e-07, "epoch": 1.7105440344734717, "percentage": 85.53, "elapsed_time": "2 days, 6:36:40", "remaining_time": "9:14:28", "throughput": 5651.04, "total_tokens": 1110997128}
+{"current_steps": 25410, "total_steps": 29704, "loss": 0.2878044605255127, "lr": 5.070554101136288e-07, "epoch": 1.7108806894694317, "percentage": 85.54, "elapsed_time": "2 days, 6:37:17", "remaining_time": "9:13:49", "throughput": 5651.02, "total_tokens": 1111200552}
+{"current_steps": 25415, "total_steps": 29704, "loss": 0.2986098289489746, "lr": 5.058958376837497e-07, "epoch": 1.7112173444653918, "percentage": 85.56, "elapsed_time": "2 days, 6:37:58", "remaining_time": "9:13:11", "throughput": 5651.0, "total_tokens": 1111427560}
+{"current_steps": 25420, "total_steps": 29704, "loss": 0.3050894021987915, "lr": 5.047375220157691e-07, "epoch": 1.711553999461352, "percentage": 85.58, "elapsed_time": "2 days, 6:38:38", "remaining_time": "9:12:32", "throughput": 5650.98, "total_tokens": 1111652600}
+{"current_steps": 25425, "total_steps": 29704, "loss": 0.3212940216064453, "lr": 5.035804634336056e-07, "epoch": 1.7118906544573123, "percentage": 85.59, "elapsed_time": "2 days, 6:39:20", "remaining_time": "9:11:54", "throughput": 5650.89, "total_tokens": 1111871248}
+{"current_steps": 25430, "total_steps": 29704, "loss": 0.293403172492981, "lr": 5.024246622608292e-07, "epoch": 1.7122273094532723, "percentage": 85.61, "elapsed_time": "2 days, 6:39:59", "remaining_time": "9:11:15", "throughput": 5650.97, "total_tokens": 1112108600}
+{"current_steps": 25435, "total_steps": 29704, "loss": 0.3186427354812622, "lr": 5.01270118820652e-07, "epoch": 1.7125639644492323, "percentage": 85.63, "elapsed_time": "2 days, 6:40:34", "remaining_time": "9:10:36", "throughput": 5651.13, "total_tokens": 1112335328}
+{"current_steps": 25440, "total_steps": 29704, "loss": 0.3163825511932373, "lr": 5.001168334359429e-07, "epoch": 1.7129006194451926, "percentage": 85.65, "elapsed_time": "2 days, 6:41:07", "remaining_time": "9:09:56", "throughput": 5651.36, "total_tokens": 1112566808}
+{"current_steps": 25445, "total_steps": 29704, "loss": 0.3128488540649414, "lr": 4.989648064292102e-07, "epoch": 1.7132372744411528, "percentage": 85.66, "elapsed_time": "2 days, 6:41:45", "remaining_time": "9:09:18", "throughput": 5651.42, "total_tokens": 1112795792}
+{"current_steps": 25450, "total_steps": 29704, "loss": 0.31338427066802976, "lr": 4.978140381226177e-07, "epoch": 1.7135739294371128, "percentage": 85.68, "elapsed_time": "2 days, 6:42:30", "remaining_time": "9:08:40", "throughput": 5651.25, "total_tokens": 1113015816}
+{"current_steps": 25455, "total_steps": 29704, "loss": 0.2886411428451538, "lr": 4.966645288379729e-07, "epoch": 1.7139105844330729, "percentage": 85.7, "elapsed_time": "2 days, 6:43:06", "remaining_time": "9:08:01", "throughput": 5651.35, "total_tokens": 1113242208}
+{"current_steps": 25460, "total_steps": 29704, "loss": 0.3108631134033203, "lr": 4.955162788967327e-07, "epoch": 1.7142472394290331, "percentage": 85.71, "elapsed_time": "2 days, 6:43:48", "remaining_time": "9:07:23", "throughput": 5651.12, "total_tokens": 1113433336}
+{"current_steps": 25465, "total_steps": 29704, "loss": 0.30652599334716796, "lr": 4.943692886200013e-07, "epoch": 1.7145838944249934, "percentage": 85.73, "elapsed_time": "2 days, 6:44:25", "remaining_time": "9:06:44", "throughput": 5651.14, "total_tokens": 1113646408}
+{"current_steps": 25470, "total_steps": 29704, "loss": 0.31150734424591064, "lr": 4.932235583285305e-07, "epoch": 1.7149205494209534, "percentage": 85.75, "elapsed_time": "2 days, 6:45:04", "remaining_time": "9:06:05", "throughput": 5651.23, "total_tokens": 1113881008}
+{"current_steps": 25475, "total_steps": 29704, "loss": 0.2943288326263428, "lr": 4.920790883427201e-07, "epoch": 1.7152572044169134, "percentage": 85.76, "elapsed_time": "2 days, 6:45:48", "remaining_time": "9:05:27", "throughput": 5650.98, "total_tokens": 1114079888}
+{"current_steps": 25480, "total_steps": 29704, "loss": 0.3120744228363037, "lr": 4.909358789826196e-07, "epoch": 1.7155938594128737, "percentage": 85.78, "elapsed_time": "2 days, 6:46:26", "remaining_time": "9:04:48", "throughput": 5650.85, "total_tokens": 1114268728}
+{"current_steps": 25485, "total_steps": 29704, "loss": 0.30831499099731446, "lr": 4.897939305679206e-07, "epoch": 1.715930514408834, "percentage": 85.8, "elapsed_time": "2 days, 6:46:57", "remaining_time": "9:04:09", "throughput": 5651.11, "total_tokens": 1114498760}
+{"current_steps": 25490, "total_steps": 29704, "loss": 0.2864993572235107, "lr": 4.886532434179686e-07, "epoch": 1.716267169404794, "percentage": 85.81, "elapsed_time": "2 days, 6:47:41", "remaining_time": "9:03:31", "throughput": 5650.95, "total_tokens": 1114714920}
+{"current_steps": 25495, "total_steps": 29704, "loss": 0.2663728713989258, "lr": 4.875138178517497e-07, "epoch": 1.716603824400754, "percentage": 85.83, "elapsed_time": "2 days, 6:48:20", "remaining_time": "9:02:52", "throughput": 5650.94, "total_tokens": 1114931832}
+{"current_steps": 25500, "total_steps": 29704, "loss": 0.31972756385803225, "lr": 4.863756541879028e-07, "epoch": 1.7169404793967142, "percentage": 85.85, "elapsed_time": "2 days, 6:48:58", "remaining_time": "9:02:13", "throughput": 5650.97, "total_tokens": 1115156536}
+{"current_steps": 25505, "total_steps": 29704, "loss": 0.2761407375335693, "lr": 4.852387527447117e-07, "epoch": 1.7172771343926745, "percentage": 85.86, "elapsed_time": "2 days, 6:49:35", "remaining_time": "9:01:34", "throughput": 5651.12, "total_tokens": 1115391624}
+{"current_steps": 25510, "total_steps": 29704, "loss": 0.2971747398376465, "lr": 4.84103113840107e-07, "epoch": 1.7176137893886345, "percentage": 85.88, "elapsed_time": "2 days, 6:50:10", "remaining_time": "9:00:55", "throughput": 5651.28, "total_tokens": 1115621104}
+{"current_steps": 25515, "total_steps": 29704, "loss": 0.2947702407836914, "lr": 4.829687377916664e-07, "epoch": 1.7179504443845945, "percentage": 85.9, "elapsed_time": "2 days, 6:50:41", "remaining_time": "9:00:15", "throughput": 5651.46, "total_tokens": 1115833104}
+{"current_steps": 25520, "total_steps": 29704, "loss": 0.2821854591369629, "lr": 4.81835624916614e-07, "epoch": 1.7182870993805548, "percentage": 85.91, "elapsed_time": "2 days, 6:51:19", "remaining_time": "8:59:36", "throughput": 5651.48, "total_tokens": 1116048240}
+{"current_steps": 25525, "total_steps": 29704, "loss": 0.28911077976226807, "lr": 4.807037755318211e-07, "epoch": 1.718623754376515, "percentage": 85.93, "elapsed_time": "2 days, 6:51:52", "remaining_time": "8:58:57", "throughput": 5651.68, "total_tokens": 1116280624}
+{"current_steps": 25530, "total_steps": 29704, "loss": 0.3094839096069336, "lr": 4.795731899538086e-07, "epoch": 1.718960409372475, "percentage": 85.95, "elapsed_time": "2 days, 6:52:40", "remaining_time": "8:58:19", "throughput": 5651.39, "total_tokens": 1116492320}
+{"current_steps": 25535, "total_steps": 29704, "loss": 0.2828222274780273, "lr": 4.78443868498737e-07, "epoch": 1.719297064368435, "percentage": 85.96, "elapsed_time": "2 days, 6:53:16", "remaining_time": "8:57:40", "throughput": 5651.47, "total_tokens": 1116713608}
+{"current_steps": 25540, "total_steps": 29704, "loss": 0.28137338161468506, "lr": 4.773158114824206e-07, "epoch": 1.7196337193643954, "percentage": 85.98, "elapsed_time": "2 days, 6:53:52", "remaining_time": "8:57:01", "throughput": 5651.6, "total_tokens": 1116938536}
+{"current_steps": 25545, "total_steps": 29704, "loss": 0.2812469959259033, "lr": 4.7618901922031445e-07, "epoch": 1.7199703743603556, "percentage": 86.0, "elapsed_time": "2 days, 6:54:30", "remaining_time": "8:56:22", "throughput": 5651.62, "total_tokens": 1117161488}
+{"current_steps": 25550, "total_steps": 29704, "loss": 0.3022423267364502, "lr": 4.750634920275243e-07, "epoch": 1.7203070293563156, "percentage": 86.02, "elapsed_time": "2 days, 6:55:12", "remaining_time": "8:55:44", "throughput": 5651.52, "total_tokens": 1117374592}
+{"current_steps": 25555, "total_steps": 29704, "loss": 0.319273042678833, "lr": 4.739392302188001e-07, "epoch": 1.7206436843522757, "percentage": 86.03, "elapsed_time": "2 days, 6:55:50", "remaining_time": "8:55:05", "throughput": 5651.57, "total_tokens": 1117603384}
+{"current_steps": 25560, "total_steps": 29704, "loss": 0.31023268699645995, "lr": 4.728162341085368e-07, "epoch": 1.720980339348236, "percentage": 86.05, "elapsed_time": "2 days, 6:56:33", "remaining_time": "8:54:27", "throughput": 5651.46, "total_tokens": 1117820784}
+{"current_steps": 25565, "total_steps": 29704, "loss": 0.312652063369751, "lr": 4.7169450401077843e-07, "epoch": 1.7213169943441962, "percentage": 86.07, "elapsed_time": "2 days, 6:57:16", "remaining_time": "8:53:49", "throughput": 5651.36, "total_tokens": 1118048160}
+{"current_steps": 25570, "total_steps": 29704, "loss": 0.3073885440826416, "lr": 4.7057404023921174e-07, "epoch": 1.7216536493401562, "percentage": 86.08, "elapsed_time": "2 days, 6:57:52", "remaining_time": "8:53:10", "throughput": 5651.49, "total_tokens": 1118274672}
+{"current_steps": 25575, "total_steps": 29704, "loss": 0.31366183757781985, "lr": 4.694548431071705e-07, "epoch": 1.7219903043361162, "percentage": 86.1, "elapsed_time": "2 days, 6:58:39", "remaining_time": "8:52:33", "throughput": 5651.25, "total_tokens": 1118492208}
+{"current_steps": 25580, "total_steps": 29704, "loss": 0.3049338340759277, "lr": 4.683369129276377e-07, "epoch": 1.7223269593320765, "percentage": 86.12, "elapsed_time": "2 days, 6:59:13", "remaining_time": "8:51:53", "throughput": 5651.42, "total_tokens": 1118716936}
+{"current_steps": 25585, "total_steps": 29704, "loss": 0.2772920846939087, "lr": 4.6722025001323444e-07, "epoch": 1.7226636143280367, "percentage": 86.13, "elapsed_time": "2 days, 6:59:58", "remaining_time": "8:51:16", "throughput": 5651.14, "total_tokens": 1118917560}
+{"current_steps": 25590, "total_steps": 29704, "loss": 0.3002648830413818, "lr": 4.6610485467623643e-07, "epoch": 1.7230002693239967, "percentage": 86.15, "elapsed_time": "2 days, 7:00:37", "remaining_time": "8:50:37", "throughput": 5651.06, "total_tokens": 1119122208}
+{"current_steps": 25595, "total_steps": 29704, "loss": 0.29230647087097167, "lr": 4.649907272285564e-07, "epoch": 1.7233369243199568, "percentage": 86.17, "elapsed_time": "2 days, 7:01:17", "remaining_time": "8:49:59", "throughput": 5651.02, "total_tokens": 1119341064}
+{"current_steps": 25600, "total_steps": 29704, "loss": 0.28714590072631835, "lr": 4.638778679817596e-07, "epoch": 1.723673579315917, "percentage": 86.18, "elapsed_time": "2 days, 7:02:00", "remaining_time": "8:49:21", "throughput": 5650.96, "total_tokens": 1119570192}
+{"current_steps": 25605, "total_steps": 29704, "loss": 0.3071947813034058, "lr": 4.6276627724705204e-07, "epoch": 1.7240102343118773, "percentage": 86.2, "elapsed_time": "2 days, 7:02:41", "remaining_time": "8:48:42", "throughput": 5650.88, "total_tokens": 1119786040}
+{"current_steps": 25610, "total_steps": 29704, "loss": 0.2772822380065918, "lr": 4.616559553352873e-07, "epoch": 1.7243468893078373, "percentage": 86.22, "elapsed_time": "2 days, 7:03:17", "remaining_time": "8:48:03", "throughput": 5650.94, "total_tokens": 1120003688}
+{"current_steps": 25615, "total_steps": 29704, "loss": 0.3121823310852051, "lr": 4.6054690255696256e-07, "epoch": 1.7246835443037973, "percentage": 86.23, "elapsed_time": "2 days, 7:03:52", "remaining_time": "8:47:24", "throughput": 5651.07, "total_tokens": 1120225000}
+{"current_steps": 25620, "total_steps": 29704, "loss": 0.299558162689209, "lr": 4.594391192222214e-07, "epoch": 1.7250201992997576, "percentage": 86.25, "elapsed_time": "2 days, 7:04:35", "remaining_time": "8:46:46", "throughput": 5650.91, "total_tokens": 1120436736}
+{"current_steps": 25625, "total_steps": 29704, "loss": 0.2828629732131958, "lr": 4.583326056408505e-07, "epoch": 1.7253568542957178, "percentage": 86.27, "elapsed_time": "2 days, 7:05:18", "remaining_time": "8:46:08", "throughput": 5650.8, "total_tokens": 1120656400}
+{"current_steps": 25630, "total_steps": 29704, "loss": 0.3062956094741821, "lr": 4.5722736212228613e-07, "epoch": 1.7256935092916779, "percentage": 86.28, "elapsed_time": "2 days, 7:06:00", "remaining_time": "8:45:30", "throughput": 5650.58, "total_tokens": 1120853112}
+{"current_steps": 25635, "total_steps": 29704, "loss": 0.3126507759094238, "lr": 4.5612338897560227e-07, "epoch": 1.726030164287638, "percentage": 86.3, "elapsed_time": "2 days, 7:06:43", "remaining_time": "8:44:52", "throughput": 5650.52, "total_tokens": 1121085288}
+{"current_steps": 25640, "total_steps": 29704, "loss": 0.31250529289245604, "lr": 4.5502068650952527e-07, "epoch": 1.7263668192835981, "percentage": 86.32, "elapsed_time": "2 days, 7:07:23", "remaining_time": "8:44:13", "throughput": 5650.48, "total_tokens": 1121299992}
+{"current_steps": 25645, "total_steps": 29704, "loss": 0.2995558023452759, "lr": 4.5391925503241894e-07, "epoch": 1.7267034742795584, "percentage": 86.34, "elapsed_time": "2 days, 7:08:02", "remaining_time": "8:43:35", "throughput": 5650.5, "total_tokens": 1121526960}
+{"current_steps": 25650, "total_steps": 29704, "loss": 0.29502243995666505, "lr": 4.5281909485229704e-07, "epoch": 1.7270401292755184, "percentage": 86.35, "elapsed_time": "2 days, 7:08:43", "remaining_time": "8:42:56", "throughput": 5650.41, "total_tokens": 1121740648}
+{"current_steps": 25655, "total_steps": 29704, "loss": 0.27261307239532473, "lr": 4.51720206276815e-07, "epoch": 1.7273767842714784, "percentage": 86.37, "elapsed_time": "2 days, 7:09:19", "remaining_time": "8:42:17", "throughput": 5650.46, "total_tokens": 1121954360}
+{"current_steps": 25660, "total_steps": 29704, "loss": 0.2796020984649658, "lr": 4.506225896132743e-07, "epoch": 1.7277134392674387, "percentage": 86.39, "elapsed_time": "2 days, 7:09:57", "remaining_time": "8:41:38", "throughput": 5650.48, "total_tokens": 1122168368}
+{"current_steps": 25665, "total_steps": 29704, "loss": 0.2953200101852417, "lr": 4.4952624516861944e-07, "epoch": 1.728050094263399, "percentage": 86.4, "elapsed_time": "2 days, 7:10:32", "remaining_time": "8:40:59", "throughput": 5650.51, "total_tokens": 1122372424}
+{"current_steps": 25670, "total_steps": 29704, "loss": 0.2733217716217041, "lr": 4.4843117324944005e-07, "epoch": 1.728386749259359, "percentage": 86.42, "elapsed_time": "2 days, 7:11:11", "remaining_time": "8:40:20", "throughput": 5650.58, "total_tokens": 1122608584}
+{"current_steps": 25675, "total_steps": 29704, "loss": 0.2921212911605835, "lr": 4.4733737416196745e-07, "epoch": 1.728723404255319, "percentage": 86.44, "elapsed_time": "2 days, 7:11:45", "remaining_time": "8:39:41", "throughput": 5650.49, "total_tokens": 1122786752}
+{"current_steps": 25680, "total_steps": 29704, "loss": 0.29300737380981445, "lr": 4.4624484821208346e-07, "epoch": 1.7290600592512793, "percentage": 86.45, "elapsed_time": "2 days, 7:12:23", "remaining_time": "8:39:02", "throughput": 5650.49, "total_tokens": 1123000352}
+{"current_steps": 25685, "total_steps": 29704, "loss": 0.26680479049682615, "lr": 4.451535957053044e-07, "epoch": 1.7293967142472395, "percentage": 86.47, "elapsed_time": "2 days, 7:13:06", "remaining_time": "8:38:24", "throughput": 5650.36, "total_tokens": 1123218520}
+{"current_steps": 25690, "total_steps": 29704, "loss": 0.28789448738098145, "lr": 4.440636169467999e-07, "epoch": 1.7297333692431995, "percentage": 86.49, "elapsed_time": "2 days, 7:13:48", "remaining_time": "8:37:46", "throughput": 5650.25, "total_tokens": 1123429840}
+{"current_steps": 25695, "total_steps": 29704, "loss": 0.2902831077575684, "lr": 4.429749122413757e-07, "epoch": 1.7300700242391596, "percentage": 86.5, "elapsed_time": "2 days, 7:14:31", "remaining_time": "8:37:08", "throughput": 5650.11, "total_tokens": 1123644168}
+{"current_steps": 25700, "total_steps": 29704, "loss": 0.29162254333496096, "lr": 4.418874818934865e-07, "epoch": 1.7304066792351198, "percentage": 86.52, "elapsed_time": "2 days, 7:15:07", "remaining_time": "8:36:29", "throughput": 5650.2, "total_tokens": 1123867240}
+{"current_steps": 25705, "total_steps": 29704, "loss": 0.30682687759399413, "lr": 4.4080132620722806e-07, "epoch": 1.73074333423108, "percentage": 86.54, "elapsed_time": "2 days, 7:15:44", "remaining_time": "8:35:50", "throughput": 5650.13, "total_tokens": 1124064016}
+{"current_steps": 25710, "total_steps": 29704, "loss": 0.30285561084747314, "lr": 4.3971644548634116e-07, "epoch": 1.73107998922704, "percentage": 86.55, "elapsed_time": "2 days, 7:16:23", "remaining_time": "8:35:11", "throughput": 5650.1, "total_tokens": 1124276384}
+{"current_steps": 25715, "total_steps": 29704, "loss": 0.3177463531494141, "lr": 4.3863284003420826e-07, "epoch": 1.7314166442230001, "percentage": 86.57, "elapsed_time": "2 days, 7:17:03", "remaining_time": "8:34:33", "throughput": 5650.05, "total_tokens": 1124494424}
+{"current_steps": 25720, "total_steps": 29704, "loss": 0.2902905702590942, "lr": 4.375505101538563e-07, "epoch": 1.7317532992189604, "percentage": 86.59, "elapsed_time": "2 days, 7:17:41", "remaining_time": "8:33:54", "throughput": 5650.08, "total_tokens": 1124712920}
+{"current_steps": 25725, "total_steps": 29704, "loss": 0.3163978099822998, "lr": 4.3646945614795487e-07, "epoch": 1.7320899542149206, "percentage": 86.6, "elapsed_time": "2 days, 7:18:15", "remaining_time": "8:33:14", "throughput": 5650.24, "total_tokens": 1124936520}
+{"current_steps": 25730, "total_steps": 29704, "loss": 0.31358091831207274, "lr": 4.3538967831881995e-07, "epoch": 1.7324266092108807, "percentage": 86.62, "elapsed_time": "2 days, 7:18:52", "remaining_time": "8:32:36", "throughput": 5650.34, "total_tokens": 1125167472}
+{"current_steps": 25735, "total_steps": 29704, "loss": 0.29058811664581297, "lr": 4.343111769684039e-07, "epoch": 1.7327632642068407, "percentage": 86.64, "elapsed_time": "2 days, 7:19:32", "remaining_time": "8:31:57", "throughput": 5650.25, "total_tokens": 1125373784}
+{"current_steps": 25740, "total_steps": 29704, "loss": 0.27836108207702637, "lr": 4.332339523983103e-07, "epoch": 1.733099919202801, "percentage": 86.65, "elapsed_time": "2 days, 7:20:12", "remaining_time": "8:31:19", "throughput": 5650.24, "total_tokens": 1125601280}
+{"current_steps": 25745, "total_steps": 29704, "loss": 0.28482909202575685, "lr": 4.321580049097773e-07, "epoch": 1.7334365741987612, "percentage": 86.67, "elapsed_time": "2 days, 7:20:46", "remaining_time": "8:30:39", "throughput": 5650.4, "total_tokens": 1125825176}
+{"current_steps": 25750, "total_steps": 29704, "loss": 0.27461097240447996, "lr": 4.310833348036941e-07, "epoch": 1.7337732291947212, "percentage": 86.69, "elapsed_time": "2 days, 7:21:30", "remaining_time": "8:30:01", "throughput": 5650.24, "total_tokens": 1126041128}
+{"current_steps": 25755, "total_steps": 29704, "loss": 0.2958423614501953, "lr": 4.300099423805865e-07, "epoch": 1.7341098841906812, "percentage": 86.71, "elapsed_time": "2 days, 7:22:13", "remaining_time": "8:29:23", "throughput": 5650.05, "total_tokens": 1126244560}
+{"current_steps": 25760, "total_steps": 29704, "loss": 0.2995717525482178, "lr": 4.289378279406264e-07, "epoch": 1.7344465391866415, "percentage": 86.72, "elapsed_time": "2 days, 7:22:46", "remaining_time": "8:28:44", "throughput": 5650.23, "total_tokens": 1126468312}
+{"current_steps": 25765, "total_steps": 29704, "loss": 0.30245399475097656, "lr": 4.2786699178362647e-07, "epoch": 1.7347831941826017, "percentage": 86.74, "elapsed_time": "2 days, 7:23:27", "remaining_time": "8:28:05", "throughput": 5650.19, "total_tokens": 1126689568}
+{"current_steps": 25770, "total_steps": 29704, "loss": 0.3159316539764404, "lr": 4.2679743420904254e-07, "epoch": 1.7351198491785618, "percentage": 86.76, "elapsed_time": "2 days, 7:24:01", "remaining_time": "8:27:26", "throughput": 5650.37, "total_tokens": 1126917656}
+{"current_steps": 25775, "total_steps": 29704, "loss": 0.26204819679260255, "lr": 4.257291555159726e-07, "epoch": 1.7354565041745218, "percentage": 86.77, "elapsed_time": "2 days, 7:24:34", "remaining_time": "8:26:46", "throughput": 5650.5, "total_tokens": 1127132592}
+{"current_steps": 25780, "total_steps": 29704, "loss": 0.2924259424209595, "lr": 4.246621560031594e-07, "epoch": 1.735793159170482, "percentage": 86.79, "elapsed_time": "2 days, 7:25:09", "remaining_time": "8:26:07", "throughput": 5650.56, "total_tokens": 1127339712}
+{"current_steps": 25785, "total_steps": 29704, "loss": 0.27674455642700196, "lr": 4.235964359689837e-07, "epoch": 1.7361298141664423, "percentage": 86.81, "elapsed_time": "2 days, 7:25:48", "remaining_time": "8:25:28", "throughput": 5650.58, "total_tokens": 1127566704}
+{"current_steps": 25790, "total_steps": 29704, "loss": 0.3013903617858887, "lr": 4.225319957114726e-07, "epoch": 1.7364664691624023, "percentage": 86.82, "elapsed_time": "2 days, 7:26:28", "remaining_time": "8:24:50", "throughput": 5650.48, "total_tokens": 1127772128}
+{"current_steps": 25795, "total_steps": 29704, "loss": 0.28219919204711913, "lr": 4.214688355282909e-07, "epoch": 1.7368031241583624, "percentage": 86.84, "elapsed_time": "2 days, 7:27:11", "remaining_time": "8:24:12", "throughput": 5650.34, "total_tokens": 1127988664}
+{"current_steps": 25800, "total_steps": 29704, "loss": 0.2839889764785767, "lr": 4.204069557167506e-07, "epoch": 1.7371397791543226, "percentage": 86.86, "elapsed_time": "2 days, 7:27:47", "remaining_time": "8:23:33", "throughput": 5650.4, "total_tokens": 1128202648}
+{"current_steps": 25805, "total_steps": 29704, "loss": 0.304730224609375, "lr": 4.1934635657380153e-07, "epoch": 1.7374764341502829, "percentage": 86.87, "elapsed_time": "2 days, 7:28:22", "remaining_time": "8:22:54", "throughput": 5650.52, "total_tokens": 1128423984}
+{"current_steps": 25810, "total_steps": 29704, "loss": 0.2796186447143555, "lr": 4.1828703839603637e-07, "epoch": 1.7378130891462429, "percentage": 86.89, "elapsed_time": "2 days, 7:29:00", "remaining_time": "8:22:15", "throughput": 5650.58, "total_tokens": 1128651136}
+{"current_steps": 25815, "total_steps": 29704, "loss": 0.32165372371673584, "lr": 4.172290014796926e-07, "epoch": 1.738149744142203, "percentage": 86.91, "elapsed_time": "2 days, 7:29:35", "remaining_time": "8:21:35", "throughput": 5650.69, "total_tokens": 1128867360}
+{"current_steps": 25820, "total_steps": 29704, "loss": 0.29702818393707275, "lr": 4.1617224612064353e-07, "epoch": 1.7384863991381632, "percentage": 86.92, "elapsed_time": "2 days, 7:30:15", "remaining_time": "8:20:57", "throughput": 5650.59, "total_tokens": 1129077456}
+{"current_steps": 25825, "total_steps": 29704, "loss": 0.2909691333770752, "lr": 4.151167726144101e-07, "epoch": 1.7388230541341234, "percentage": 86.94, "elapsed_time": "2 days, 7:30:50", "remaining_time": "8:20:18", "throughput": 5650.8, "total_tokens": 1129312768}
+{"current_steps": 25830, "total_steps": 29704, "loss": 0.3070362567901611, "lr": 4.1406258125615106e-07, "epoch": 1.7391597091300834, "percentage": 86.96, "elapsed_time": "2 days, 7:31:20", "remaining_time": "8:19:38", "throughput": 5651.06, "total_tokens": 1129537904}
+{"current_steps": 25835, "total_steps": 29704, "loss": 0.29037325382232665, "lr": 4.130096723406674e-07, "epoch": 1.7394963641260435, "percentage": 86.97, "elapsed_time": "2 days, 7:31:55", "remaining_time": "8:18:58", "throughput": 5651.17, "total_tokens": 1129757208}
+{"current_steps": 25840, "total_steps": 29704, "loss": 0.3026682615280151, "lr": 4.119580461624023e-07, "epoch": 1.7398330191220037, "percentage": 86.99, "elapsed_time": "2 days, 7:32:31", "remaining_time": "8:18:19", "throughput": 5651.19, "total_tokens": 1129966384}
+{"current_steps": 25845, "total_steps": 29704, "loss": 0.30320086479187014, "lr": 4.109077030154396e-07, "epoch": 1.740169674117964, "percentage": 87.01, "elapsed_time": "2 days, 7:33:06", "remaining_time": "8:17:40", "throughput": 5651.32, "total_tokens": 1130186152}
+{"current_steps": 25850, "total_steps": 29704, "loss": 0.3050692081451416, "lr": 4.0985864319350366e-07, "epoch": 1.740506329113924, "percentage": 87.03, "elapsed_time": "2 days, 7:33:50", "remaining_time": "8:17:02", "throughput": 5651.18, "total_tokens": 1130406368}
+{"current_steps": 25855, "total_steps": 29704, "loss": 0.2959964513778687, "lr": 4.0881086698996154e-07, "epoch": 1.740842984109884, "percentage": 87.04, "elapsed_time": "2 days, 7:34:32", "remaining_time": "8:16:24", "throughput": 5651.11, "total_tokens": 1130629712}
+{"current_steps": 25860, "total_steps": 29704, "loss": 0.2970879316329956, "lr": 4.077643746978194e-07, "epoch": 1.7411796391058443, "percentage": 87.06, "elapsed_time": "2 days, 7:35:07", "remaining_time": "8:15:45", "throughput": 5651.23, "total_tokens": 1130854672}
+{"current_steps": 25865, "total_steps": 29704, "loss": 0.29431335926055907, "lr": 4.067191666097281e-07, "epoch": 1.7415162941018045, "percentage": 87.08, "elapsed_time": "2 days, 7:35:43", "remaining_time": "8:15:06", "throughput": 5651.3, "total_tokens": 1131071712}
+{"current_steps": 25870, "total_steps": 29704, "loss": 0.28345699310302735, "lr": 4.056752430179728e-07, "epoch": 1.7418529490977646, "percentage": 87.09, "elapsed_time": "2 days, 7:36:27", "remaining_time": "8:14:28", "throughput": 5651.16, "total_tokens": 1131294176}
+{"current_steps": 25875, "total_steps": 29704, "loss": 0.2875986576080322, "lr": 4.046326042144866e-07, "epoch": 1.7421896040937246, "percentage": 87.11, "elapsed_time": "2 days, 7:37:08", "remaining_time": "8:13:49", "throughput": 5651.14, "total_tokens": 1131517072}
+{"current_steps": 25880, "total_steps": 29704, "loss": 0.2836933135986328, "lr": 4.0359125049083956e-07, "epoch": 1.7425262590896848, "percentage": 87.13, "elapsed_time": "2 days, 7:37:40", "remaining_time": "8:13:10", "throughput": 5651.31, "total_tokens": 1131734008}
+{"current_steps": 25885, "total_steps": 29704, "loss": 0.30028643608093264, "lr": 4.025511821382422e-07, "epoch": 1.742862914085645, "percentage": 87.14, "elapsed_time": "2 days, 7:38:26", "remaining_time": "8:12:32", "throughput": 5651.07, "total_tokens": 1131948944}
+{"current_steps": 25890, "total_steps": 29704, "loss": 0.28558173179626467, "lr": 4.015123994475462e-07, "epoch": 1.7431995690816051, "percentage": 87.16, "elapsed_time": "2 days, 7:39:06", "remaining_time": "8:11:54", "throughput": 5650.99, "total_tokens": 1132155112}
+{"current_steps": 25895, "total_steps": 29704, "loss": 0.29801325798034667, "lr": 4.0047490270924417e-07, "epoch": 1.7435362240775654, "percentage": 87.18, "elapsed_time": "2 days, 7:39:37", "remaining_time": "8:11:14", "throughput": 5651.29, "total_tokens": 1132391904}
+{"current_steps": 25900, "total_steps": 29704, "loss": 0.2968616247177124, "lr": 3.994386922134691e-07, "epoch": 1.7438728790735256, "percentage": 87.19, "elapsed_time": "2 days, 7:40:10", "remaining_time": "8:10:34", "throughput": 5651.47, "total_tokens": 1132613664}
+{"current_steps": 25905, "total_steps": 29704, "loss": 0.2899972438812256, "lr": 3.984037682499936e-07, "epoch": 1.7442095340694856, "percentage": 87.21, "elapsed_time": "2 days, 7:40:48", "remaining_time": "8:09:55", "throughput": 5651.52, "total_tokens": 1132836560}
+{"current_steps": 25910, "total_steps": 29704, "loss": 0.31725754737854006, "lr": 3.973701311082301e-07, "epoch": 1.7445461890654457, "percentage": 87.23, "elapsed_time": "2 days, 7:41:29", "remaining_time": "8:09:17", "throughput": 5651.43, "total_tokens": 1133050712}
+{"current_steps": 25915, "total_steps": 29704, "loss": 0.2900426387786865, "lr": 3.9633778107723454e-07, "epoch": 1.744882844061406, "percentage": 87.24, "elapsed_time": "2 days, 7:42:06", "remaining_time": "8:08:38", "throughput": 5651.4, "total_tokens": 1133254224}
+{"current_steps": 25920, "total_steps": 29704, "loss": 0.3090181350708008, "lr": 3.953067184456966e-07, "epoch": 1.7452194990573662, "percentage": 87.26, "elapsed_time": "2 days, 7:42:44", "remaining_time": "8:07:59", "throughput": 5651.39, "total_tokens": 1133470824}
+{"current_steps": 25925, "total_steps": 29704, "loss": 0.31003541946411134, "lr": 3.9427694350195224e-07, "epoch": 1.7455561540533262, "percentage": 87.28, "elapsed_time": "2 days, 7:43:21", "remaining_time": "8:07:21", "throughput": 5651.5, "total_tokens": 1133700592}
+{"current_steps": 25930, "total_steps": 29704, "loss": 0.3075569152832031, "lr": 3.9324845653397446e-07, "epoch": 1.7458928090492862, "percentage": 87.29, "elapsed_time": "2 days, 7:43:59", "remaining_time": "8:06:42", "throughput": 5651.55, "total_tokens": 1133925704}
+{"current_steps": 25935, "total_steps": 29704, "loss": 0.2702242374420166, "lr": 3.922212578293755e-07, "epoch": 1.7462294640452465, "percentage": 87.31, "elapsed_time": "2 days, 7:44:42", "remaining_time": "8:06:04", "throughput": 5651.41, "total_tokens": 1134140200}
+{"current_steps": 25940, "total_steps": 29704, "loss": 0.2941112518310547, "lr": 3.911953476754088e-07, "epoch": 1.7465661190412067, "percentage": 87.33, "elapsed_time": "2 days, 7:45:15", "remaining_time": "8:05:24", "throughput": 5651.6, "total_tokens": 1134363752}
+{"current_steps": 25945, "total_steps": 29704, "loss": 0.31668860912323, "lr": 3.9017072635896716e-07, "epoch": 1.7469027740371668, "percentage": 87.35, "elapsed_time": "2 days, 7:45:56", "remaining_time": "8:04:46", "throughput": 5651.56, "total_tokens": 1134590888}
+{"current_steps": 25950, "total_steps": 29704, "loss": 0.3079123020172119, "lr": 3.891473941665819e-07, "epoch": 1.7472394290331268, "percentage": 87.36, "elapsed_time": "2 days, 7:46:35", "remaining_time": "8:04:07", "throughput": 5651.62, "total_tokens": 1134821944}
+{"current_steps": 25955, "total_steps": 29704, "loss": 0.30869007110595703, "lr": 3.881253513844246e-07, "epoch": 1.747576084029087, "percentage": 87.38, "elapsed_time": "2 days, 7:47:16", "remaining_time": "8:03:29", "throughput": 5651.49, "total_tokens": 1135026112}
+{"current_steps": 25960, "total_steps": 29704, "loss": 0.29468765258789065, "lr": 3.871045982983063e-07, "epoch": 1.7479127390250473, "percentage": 87.4, "elapsed_time": "2 days, 7:47:57", "remaining_time": "8:02:50", "throughput": 5651.46, "total_tokens": 1135250784}
+{"current_steps": 25965, "total_steps": 29704, "loss": 0.2700496673583984, "lr": 3.8608513519367897e-07, "epoch": 1.7482493940210073, "percentage": 87.41, "elapsed_time": "2 days, 7:48:42", "remaining_time": "8:02:13", "throughput": 5651.34, "total_tokens": 1135481416}
+{"current_steps": 25970, "total_steps": 29704, "loss": 0.2968393087387085, "lr": 3.8506696235562915e-07, "epoch": 1.7485860490169673, "percentage": 87.43, "elapsed_time": "2 days, 7:49:20", "remaining_time": "8:01:34", "throughput": 5651.33, "total_tokens": 1135693800}
+{"current_steps": 25975, "total_steps": 29704, "loss": 0.2980025768280029, "lr": 3.840500800688884e-07, "epoch": 1.7489227040129276, "percentage": 87.45, "elapsed_time": "2 days, 7:50:06", "remaining_time": "8:00:56", "throughput": 5651.18, "total_tokens": 1135925744}
+{"current_steps": 25980, "total_steps": 29704, "loss": 0.2952871322631836, "lr": 3.8303448861782333e-07, "epoch": 1.7492593590088878, "percentage": 87.46, "elapsed_time": "2 days, 7:50:40", "remaining_time": "8:00:17", "throughput": 5651.24, "total_tokens": 1136130424}
+{"current_steps": 25985, "total_steps": 29704, "loss": 0.27895636558532716, "lr": 3.8202018828644105e-07, "epoch": 1.7495960140048479, "percentage": 87.48, "elapsed_time": "2 days, 7:51:19", "remaining_time": "7:59:38", "throughput": 5651.22, "total_tokens": 1136347560}
+{"current_steps": 25990, "total_steps": 29704, "loss": 0.32080807685852053, "lr": 3.8100717935838804e-07, "epoch": 1.749932669000808, "percentage": 87.5, "elapsed_time": "2 days, 7:52:01", "remaining_time": "7:59:00", "throughput": 5651.13, "total_tokens": 1136562648}
+{"current_steps": 25995, "total_steps": 29704, "loss": 0.27611637115478516, "lr": 3.799954621169477e-07, "epoch": 1.7502693239967682, "percentage": 87.51, "elapsed_time": "2 days, 7:52:40", "remaining_time": "7:58:21", "throughput": 5651.12, "total_tokens": 1136781352}
+{"current_steps": 26000, "total_steps": 29704, "loss": 0.29705018997192384, "lr": 3.7898503684504483e-07, "epoch": 1.7506059789927284, "percentage": 87.53, "elapsed_time": "2 days, 7:53:20", "remaining_time": "7:57:43", "throughput": 5651.06, "total_tokens": 1136994720}
+{"current_steps": 26005, "total_steps": 29704, "loss": 0.28871927261352537, "lr": 3.7797590382524066e-07, "epoch": 1.7509426339886884, "percentage": 87.55, "elapsed_time": "2 days, 7:54:02", "remaining_time": "7:57:05", "throughput": 5650.88, "total_tokens": 1137197696}
+{"current_steps": 26010, "total_steps": 29704, "loss": 0.3024580955505371, "lr": 3.76968063339736e-07, "epoch": 1.7512792889846485, "percentage": 87.56, "elapsed_time": "2 days, 7:54:34", "remaining_time": "7:56:25", "throughput": 5651.14, "total_tokens": 1137430488}
+{"current_steps": 26015, "total_steps": 29704, "loss": 0.28331038951873777, "lr": 3.759615156703722e-07, "epoch": 1.7516159439806087, "percentage": 87.58, "elapsed_time": "2 days, 7:55:12", "remaining_time": "7:55:46", "throughput": 5651.14, "total_tokens": 1137644120}
+{"current_steps": 26020, "total_steps": 29704, "loss": 0.3102069854736328, "lr": 3.749562610986235e-07, "epoch": 1.751952598976569, "percentage": 87.6, "elapsed_time": "2 days, 7:55:48", "remaining_time": "7:55:07", "throughput": 5651.19, "total_tokens": 1137861320}
+{"current_steps": 26025, "total_steps": 29704, "loss": 0.28222415447235105, "lr": 3.739522999056094e-07, "epoch": 1.752289253972529, "percentage": 87.61, "elapsed_time": "2 days, 7:56:25", "remaining_time": "7:54:28", "throughput": 5651.18, "total_tokens": 1138066152}
+{"current_steps": 26030, "total_steps": 29704, "loss": 0.3015716552734375, "lr": 3.7294963237208326e-07, "epoch": 1.752625908968489, "percentage": 87.63, "elapsed_time": "2 days, 7:57:10", "remaining_time": "7:53:50", "throughput": 5651.02, "total_tokens": 1138288896}
+{"current_steps": 26035, "total_steps": 29704, "loss": 0.2880394458770752, "lr": 3.7194825877843786e-07, "epoch": 1.7529625639644493, "percentage": 87.65, "elapsed_time": "2 days, 7:57:59", "remaining_time": "7:53:13", "throughput": 5650.76, "total_tokens": 1138513200}
+{"current_steps": 26040, "total_steps": 29704, "loss": 0.30432782173156736, "lr": 3.7094817940470375e-07, "epoch": 1.7532992189604095, "percentage": 87.66, "elapsed_time": "2 days, 7:58:40", "remaining_time": "7:52:35", "throughput": 5650.66, "total_tokens": 1138721632}
+{"current_steps": 26045, "total_steps": 29704, "loss": 0.2994675159454346, "lr": 3.6994939453055044e-07, "epoch": 1.7536358739563696, "percentage": 87.68, "elapsed_time": "2 days, 7:59:21", "remaining_time": "7:51:56", "throughput": 5650.63, "total_tokens": 1138951136}
+{"current_steps": 26050, "total_steps": 29704, "loss": 0.2943981170654297, "lr": 3.6895190443528515e-07, "epoch": 1.7539725289523296, "percentage": 87.7, "elapsed_time": "2 days, 8:00:01", "remaining_time": "7:51:18", "throughput": 5650.62, "total_tokens": 1139175408}
+{"current_steps": 26055, "total_steps": 29704, "loss": 0.312309193611145, "lr": 3.679557093978525e-07, "epoch": 1.7543091839482898, "percentage": 87.72, "elapsed_time": "2 days, 8:00:38", "remaining_time": "7:50:39", "throughput": 5650.69, "total_tokens": 1139396600}
+{"current_steps": 26060, "total_steps": 29704, "loss": 0.29059772491455077, "lr": 3.669608096968341e-07, "epoch": 1.75464583894425, "percentage": 87.73, "elapsed_time": "2 days, 8:01:21", "remaining_time": "7:50:01", "throughput": 5650.51, "total_tokens": 1139603752}
+{"current_steps": 26065, "total_steps": 29704, "loss": 0.2955317974090576, "lr": 3.6596720561045295e-07, "epoch": 1.75498249394021, "percentage": 87.75, "elapsed_time": "2 days, 8:01:58", "remaining_time": "7:49:22", "throughput": 5650.58, "total_tokens": 1139829360}
+{"current_steps": 26070, "total_steps": 29704, "loss": 0.2989501476287842, "lr": 3.6497489741656455e-07, "epoch": 1.7553191489361701, "percentage": 87.77, "elapsed_time": "2 days, 8:02:34", "remaining_time": "7:48:43", "throughput": 5650.69, "total_tokens": 1140053168}
+{"current_steps": 26075, "total_steps": 29704, "loss": 0.29015383720397947, "lr": 3.639838853926669e-07, "epoch": 1.7556558039321304, "percentage": 87.78, "elapsed_time": "2 days, 8:03:13", "remaining_time": "7:48:04", "throughput": 5650.74, "total_tokens": 1140284016}
+{"current_steps": 26080, "total_steps": 29704, "loss": 0.3135251998901367, "lr": 3.629941698158923e-07, "epoch": 1.7559924589280906, "percentage": 87.8, "elapsed_time": "2 days, 8:03:57", "remaining_time": "7:47:26", "throughput": 5650.61, "total_tokens": 1140503592}
+{"current_steps": 26085, "total_steps": 29704, "loss": 0.2774399757385254, "lr": 3.6200575096301207e-07, "epoch": 1.7563291139240507, "percentage": 87.82, "elapsed_time": "2 days, 8:04:33", "remaining_time": "7:46:47", "throughput": 5650.68, "total_tokens": 1140721216}
+{"current_steps": 26090, "total_steps": 29704, "loss": 0.32251529693603515, "lr": 3.61018629110434e-07, "epoch": 1.7566657689200107, "percentage": 87.83, "elapsed_time": "2 days, 8:05:07", "remaining_time": "7:46:08", "throughput": 5650.83, "total_tokens": 1140946680}
+{"current_steps": 26095, "total_steps": 29704, "loss": 0.2859219551086426, "lr": 3.6003280453420386e-07, "epoch": 1.757002423915971, "percentage": 87.85, "elapsed_time": "2 days, 8:05:40", "remaining_time": "7:45:28", "throughput": 5651.0, "total_tokens": 1141168176}
+{"current_steps": 26100, "total_steps": 29704, "loss": 0.28550260066986083, "lr": 3.59048277510004e-07, "epoch": 1.7573390789119312, "percentage": 87.87, "elapsed_time": "2 days, 8:06:21", "remaining_time": "7:44:50", "throughput": 5650.97, "total_tokens": 1141389904}
+{"current_steps": 26105, "total_steps": 29704, "loss": 0.28779993057250974, "lr": 3.5806504831315423e-07, "epoch": 1.7576757339078912, "percentage": 87.88, "elapsed_time": "2 days, 8:07:00", "remaining_time": "7:44:11", "throughput": 5650.94, "total_tokens": 1141608056}
+{"current_steps": 26110, "total_steps": 29704, "loss": 0.3048838138580322, "lr": 3.570831172186112e-07, "epoch": 1.7580123889038513, "percentage": 87.9, "elapsed_time": "2 days, 8:07:39", "remaining_time": "7:43:33", "throughput": 5650.96, "total_tokens": 1141832104}
+{"current_steps": 26115, "total_steps": 29704, "loss": 0.30502030849456785, "lr": 3.5610248450097084e-07, "epoch": 1.7583490438998115, "percentage": 87.92, "elapsed_time": "2 days, 8:08:20", "remaining_time": "7:42:54", "throughput": 5650.91, "total_tokens": 1142049928}
+{"current_steps": 26120, "total_steps": 29704, "loss": 0.2915775775909424, "lr": 3.551231504344604e-07, "epoch": 1.7586856988957718, "percentage": 87.93, "elapsed_time": "2 days, 8:08:58", "remaining_time": "7:42:16", "throughput": 5650.93, "total_tokens": 1142270008}
+{"current_steps": 26125, "total_steps": 29704, "loss": 0.30109832286834715, "lr": 3.5414511529295036e-07, "epoch": 1.7590223538917318, "percentage": 87.95, "elapsed_time": "2 days, 8:09:31", "remaining_time": "7:41:36", "throughput": 5651.08, "total_tokens": 1142484296}
+{"current_steps": 26130, "total_steps": 29704, "loss": 0.3126358509063721, "lr": 3.5316837934994465e-07, "epoch": 1.7593590088876918, "percentage": 87.97, "elapsed_time": "2 days, 8:10:11", "remaining_time": "7:40:57", "throughput": 5651.04, "total_tokens": 1142703184}
+{"current_steps": 26135, "total_steps": 29704, "loss": 0.3012665271759033, "lr": 3.521929428785836e-07, "epoch": 1.759695663883652, "percentage": 87.98, "elapsed_time": "2 days, 8:10:49", "remaining_time": "7:40:19", "throughput": 5651.1, "total_tokens": 1142932000}
+{"current_steps": 26140, "total_steps": 29704, "loss": 0.3177218198776245, "lr": 3.5121880615164584e-07, "epoch": 1.7600323188796123, "percentage": 88.0, "elapsed_time": "2 days, 8:11:28", "remaining_time": "7:39:40", "throughput": 5651.06, "total_tokens": 1143147936}
+{"current_steps": 26145, "total_steps": 29704, "loss": 0.30140066146850586, "lr": 3.5024596944154445e-07, "epoch": 1.7603689738755723, "percentage": 88.02, "elapsed_time": "2 days, 8:12:07", "remaining_time": "7:39:01", "throughput": 5651.05, "total_tokens": 1143365592}
+{"current_steps": 26150, "total_steps": 29704, "loss": 0.2727382659912109, "lr": 3.4927443302033127e-07, "epoch": 1.7607056288715324, "percentage": 88.04, "elapsed_time": "2 days, 8:12:55", "remaining_time": "7:38:24", "throughput": 5650.84, "total_tokens": 1143591488}
+{"current_steps": 26155, "total_steps": 29704, "loss": 0.2696976900100708, "lr": 3.4830419715969233e-07, "epoch": 1.7610422838674926, "percentage": 88.05, "elapsed_time": "2 days, 8:13:30", "remaining_time": "7:37:45", "throughput": 5650.9, "total_tokens": 1143804104}
+{"current_steps": 26160, "total_steps": 29704, "loss": 0.30320181846618655, "lr": 3.4733526213095114e-07, "epoch": 1.7613789388634529, "percentage": 88.07, "elapsed_time": "2 days, 8:14:13", "remaining_time": "7:37:07", "throughput": 5650.81, "total_tokens": 1144024536}
+{"current_steps": 26165, "total_steps": 29704, "loss": 0.3035079002380371, "lr": 3.4636762820506876e-07, "epoch": 1.761715593859413, "percentage": 88.09, "elapsed_time": "2 days, 8:14:52", "remaining_time": "7:36:28", "throughput": 5650.8, "total_tokens": 1144244512}
+{"current_steps": 26170, "total_steps": 29704, "loss": 0.30885040760040283, "lr": 3.4540129565263867e-07, "epoch": 1.762052248855373, "percentage": 88.1, "elapsed_time": "2 days, 8:15:28", "remaining_time": "7:35:49", "throughput": 5650.84, "total_tokens": 1144457424}
+{"current_steps": 26175, "total_steps": 29704, "loss": 0.30580987930297854, "lr": 3.444362647438942e-07, "epoch": 1.7623889038513332, "percentage": 88.12, "elapsed_time": "2 days, 8:16:06", "remaining_time": "7:35:10", "throughput": 5650.92, "total_tokens": 1144685136}
+{"current_steps": 26180, "total_steps": 29704, "loss": 0.29648699760437014, "lr": 3.434725357487029e-07, "epoch": 1.7627255588472934, "percentage": 88.14, "elapsed_time": "2 days, 8:16:48", "remaining_time": "7:34:32", "throughput": 5650.85, "total_tokens": 1144912488}
+{"current_steps": 26185, "total_steps": 29704, "loss": 0.3241448402404785, "lr": 3.4251010893656844e-07, "epoch": 1.7630622138432535, "percentage": 88.15, "elapsed_time": "2 days, 8:17:26", "remaining_time": "7:33:53", "throughput": 5650.87, "total_tokens": 1145132240}
+{"current_steps": 26190, "total_steps": 29704, "loss": 0.2964663505554199, "lr": 3.4154898457663065e-07, "epoch": 1.7633988688392135, "percentage": 88.17, "elapsed_time": "2 days, 8:18:04", "remaining_time": "7:33:14", "throughput": 5650.83, "total_tokens": 1145339072}
+{"current_steps": 26195, "total_steps": 29704, "loss": 0.2934283256530762, "lr": 3.40589162937664e-07, "epoch": 1.7637355238351737, "percentage": 88.19, "elapsed_time": "2 days, 8:18:43", "remaining_time": "7:32:36", "throughput": 5650.75, "total_tokens": 1145537280}
+{"current_steps": 26200, "total_steps": 29704, "loss": 0.2671190738677979, "lr": 3.3963064428808034e-07, "epoch": 1.764072178831134, "percentage": 88.2, "elapsed_time": "2 days, 8:19:25", "remaining_time": "7:31:57", "throughput": 5650.58, "total_tokens": 1145744432}
+{"current_steps": 26205, "total_steps": 29704, "loss": 0.29965527057647706, "lr": 3.386734288959265e-07, "epoch": 1.764408833827094, "percentage": 88.22, "elapsed_time": "2 days, 8:20:07", "remaining_time": "7:31:19", "throughput": 5650.38, "total_tokens": 1145941000}
+{"current_steps": 26210, "total_steps": 29704, "loss": 0.3009481430053711, "lr": 3.377175170288838e-07, "epoch": 1.764745488823054, "percentage": 88.24, "elapsed_time": "2 days, 8:20:41", "remaining_time": "7:30:40", "throughput": 5650.55, "total_tokens": 1146164576}
+{"current_steps": 26215, "total_steps": 29704, "loss": 0.2912009239196777, "lr": 3.367629089542712e-07, "epoch": 1.7650821438190143, "percentage": 88.25, "elapsed_time": "2 days, 8:21:29", "remaining_time": "7:30:02", "throughput": 5650.36, "total_tokens": 1146397144}
+{"current_steps": 26220, "total_steps": 29704, "loss": 0.3059426546096802, "lr": 3.3580960493904024e-07, "epoch": 1.7654187988149745, "percentage": 88.27, "elapsed_time": "2 days, 8:22:10", "remaining_time": "7:29:24", "throughput": 5650.3, "total_tokens": 1146616808}
+{"current_steps": 26225, "total_steps": 29704, "loss": 0.31472842693328856, "lr": 3.3485760524978093e-07, "epoch": 1.7657554538109346, "percentage": 88.29, "elapsed_time": "2 days, 8:22:42", "remaining_time": "7:28:44", "throughput": 5650.42, "total_tokens": 1146822992}
+{"current_steps": 26230, "total_steps": 29704, "loss": 0.29501938819885254, "lr": 3.3390691015271646e-07, "epoch": 1.7660921088068946, "percentage": 88.3, "elapsed_time": "2 days, 8:23:18", "remaining_time": "7:28:05", "throughput": 5650.49, "total_tokens": 1147041408}
+{"current_steps": 26235, "total_steps": 29704, "loss": 0.2826894521713257, "lr": 3.329575199137053e-07, "epoch": 1.7664287638028549, "percentage": 88.32, "elapsed_time": "2 days, 8:23:57", "remaining_time": "7:27:27", "throughput": 5650.51, "total_tokens": 1147263888}
+{"current_steps": 26240, "total_steps": 29704, "loss": 0.327055025100708, "lr": 3.3200943479824113e-07, "epoch": 1.766765418798815, "percentage": 88.34, "elapsed_time": "2 days, 8:24:39", "remaining_time": "7:26:48", "throughput": 5650.4, "total_tokens": 1147480536}
+{"current_steps": 26245, "total_steps": 29704, "loss": 0.2585421562194824, "lr": 3.3106265507145406e-07, "epoch": 1.7671020737947751, "percentage": 88.36, "elapsed_time": "2 days, 8:25:15", "remaining_time": "7:26:09", "throughput": 5650.44, "total_tokens": 1147695160}
+{"current_steps": 26250, "total_steps": 29704, "loss": 0.28537108898162844, "lr": 3.3011718099810687e-07, "epoch": 1.7674387287907352, "percentage": 88.37, "elapsed_time": "2 days, 8:25:55", "remaining_time": "7:25:31", "throughput": 5650.39, "total_tokens": 1147904984}
+{"current_steps": 26255, "total_steps": 29704, "loss": 0.3015407085418701, "lr": 3.2917301284259904e-07, "epoch": 1.7677753837866954, "percentage": 88.39, "elapsed_time": "2 days, 8:26:35", "remaining_time": "7:24:52", "throughput": 5650.38, "total_tokens": 1148133136}
+{"current_steps": 26260, "total_steps": 29704, "loss": 0.31117079257965086, "lr": 3.2823015086896335e-07, "epoch": 1.7681120387826557, "percentage": 88.41, "elapsed_time": "2 days, 8:27:11", "remaining_time": "7:24:13", "throughput": 5650.46, "total_tokens": 1148351808}
+{"current_steps": 26265, "total_steps": 29704, "loss": 0.28980340957641604, "lr": 3.2728859534087e-07, "epoch": 1.7684486937786157, "percentage": 88.42, "elapsed_time": "2 days, 8:27:50", "remaining_time": "7:23:35", "throughput": 5650.5, "total_tokens": 1148578048}
+{"current_steps": 26270, "total_steps": 29704, "loss": 0.30661208629608155, "lr": 3.263483465216194e-07, "epoch": 1.7687853487745757, "percentage": 88.44, "elapsed_time": "2 days, 8:28:24", "remaining_time": "7:22:55", "throughput": 5650.67, "total_tokens": 1148808400}
+{"current_steps": 26275, "total_steps": 29704, "loss": 0.29364948272705077, "lr": 3.2540940467415126e-07, "epoch": 1.769122003770536, "percentage": 88.46, "elapsed_time": "2 days, 8:29:04", "remaining_time": "7:22:17", "throughput": 5650.65, "total_tokens": 1149026320}
+{"current_steps": 26280, "total_steps": 29704, "loss": 0.29972095489501954, "lr": 3.2447177006103726e-07, "epoch": 1.7694586587664962, "percentage": 88.47, "elapsed_time": "2 days, 8:29:45", "remaining_time": "7:21:38", "throughput": 5650.56, "total_tokens": 1149241504}
+{"current_steps": 26285, "total_steps": 29704, "loss": 0.29863905906677246, "lr": 3.235354429444831e-07, "epoch": 1.7697953137624562, "percentage": 88.49, "elapsed_time": "2 days, 8:30:20", "remaining_time": "7:20:59", "throughput": 5650.58, "total_tokens": 1149442960}
+{"current_steps": 26290, "total_steps": 29704, "loss": 0.3004556655883789, "lr": 3.226004235863306e-07, "epoch": 1.7701319687584163, "percentage": 88.51, "elapsed_time": "2 days, 8:31:04", "remaining_time": "7:20:21", "throughput": 5650.45, "total_tokens": 1149664768}
+{"current_steps": 26295, "total_steps": 29704, "loss": 0.2912909984588623, "lr": 3.2166671224805437e-07, "epoch": 1.7704686237543765, "percentage": 88.52, "elapsed_time": "2 days, 8:31:44", "remaining_time": "7:19:43", "throughput": 5650.45, "total_tokens": 1149894072}
+{"current_steps": 26300, "total_steps": 29704, "loss": 0.28990769386291504, "lr": 3.207343091907633e-07, "epoch": 1.7708052787503368, "percentage": 88.54, "elapsed_time": "2 days, 8:32:27", "remaining_time": "7:19:05", "throughput": 5650.27, "total_tokens": 1150100408}
+{"current_steps": 26305, "total_steps": 29704, "loss": 0.30245256423950195, "lr": 3.1980321467520226e-07, "epoch": 1.7711419337462968, "percentage": 88.56, "elapsed_time": "2 days, 8:33:07", "remaining_time": "7:18:26", "throughput": 5650.16, "total_tokens": 1150300712}
+{"current_steps": 26310, "total_steps": 29704, "loss": 0.31411175727844237, "lr": 3.1887342896174735e-07, "epoch": 1.7714785887422568, "percentage": 88.57, "elapsed_time": "2 days, 8:33:50", "remaining_time": "7:17:48", "throughput": 5649.95, "total_tokens": 1150501592}
+{"current_steps": 26315, "total_steps": 29704, "loss": 0.28406760692596433, "lr": 3.1794495231041276e-07, "epoch": 1.771815243738217, "percentage": 88.59, "elapsed_time": "2 days, 8:34:31", "remaining_time": "7:17:10", "throughput": 5649.92, "total_tokens": 1150727232}
+{"current_steps": 26320, "total_steps": 29704, "loss": 0.3084291458129883, "lr": 3.1701778498084035e-07, "epoch": 1.7721518987341773, "percentage": 88.61, "elapsed_time": "2 days, 8:35:12", "remaining_time": "7:16:31", "throughput": 5649.78, "total_tokens": 1150932320}
+{"current_steps": 26325, "total_steps": 29704, "loss": 0.3075510263442993, "lr": 3.160919272323121e-07, "epoch": 1.7724885537301374, "percentage": 88.62, "elapsed_time": "2 days, 8:35:50", "remaining_time": "7:15:52", "throughput": 5649.87, "total_tokens": 1151162984}
+{"current_steps": 26330, "total_steps": 29704, "loss": 0.30257492065429686, "lr": 3.1516737932374087e-07, "epoch": 1.7728252087260974, "percentage": 88.64, "elapsed_time": "2 days, 8:36:24", "remaining_time": "7:15:13", "throughput": 5649.94, "total_tokens": 1151371272}
+{"current_steps": 26335, "total_steps": 29704, "loss": 0.29022831916809083, "lr": 3.142441415136727e-07, "epoch": 1.7731618637220576, "percentage": 88.66, "elapsed_time": "2 days, 8:36:57", "remaining_time": "7:14:34", "throughput": 5650.15, "total_tokens": 1151601408}
+{"current_steps": 26340, "total_steps": 29704, "loss": 0.2794740438461304, "lr": 3.133222140602893e-07, "epoch": 1.773498518718018, "percentage": 88.67, "elapsed_time": "2 days, 8:37:31", "remaining_time": "7:13:54", "throughput": 5650.22, "total_tokens": 1151807024}
+{"current_steps": 26345, "total_steps": 29704, "loss": 0.30714993476867675, "lr": 3.124015972214034e-07, "epoch": 1.773835173713978, "percentage": 88.69, "elapsed_time": "2 days, 8:38:00", "remaining_time": "7:13:14", "throughput": 5650.51, "total_tokens": 1152030904}
+{"current_steps": 26350, "total_steps": 29704, "loss": 0.31725311279296875, "lr": 3.114822912544635e-07, "epoch": 1.774171828709938, "percentage": 88.71, "elapsed_time": "2 days, 8:38:41", "remaining_time": "7:12:36", "throughput": 5650.41, "total_tokens": 1152240824}
+{"current_steps": 26355, "total_steps": 29704, "loss": 0.2949615716934204, "lr": 3.1056429641655007e-07, "epoch": 1.7745084837058982, "percentage": 88.73, "elapsed_time": "2 days, 8:39:23", "remaining_time": "7:11:58", "throughput": 5650.25, "total_tokens": 1152447328}
+{"current_steps": 26360, "total_steps": 29704, "loss": 0.2921284198760986, "lr": 3.0964761296437665e-07, "epoch": 1.7748451387018584, "percentage": 88.74, "elapsed_time": "2 days, 8:40:05", "remaining_time": "7:11:19", "throughput": 5650.21, "total_tokens": 1152671024}
+{"current_steps": 26365, "total_steps": 29704, "loss": 0.3069296836853027, "lr": 3.087322411542937e-07, "epoch": 1.7751817936978185, "percentage": 88.76, "elapsed_time": "2 days, 8:40:43", "remaining_time": "7:10:41", "throughput": 5650.25, "total_tokens": 1152898216}
+{"current_steps": 26370, "total_steps": 29704, "loss": 0.27917664051055907, "lr": 3.0781818124227867e-07, "epoch": 1.7755184486937785, "percentage": 88.78, "elapsed_time": "2 days, 8:41:23", "remaining_time": "7:10:02", "throughput": 5650.19, "total_tokens": 1153109064}
+{"current_steps": 26375, "total_steps": 29704, "loss": 0.2973165988922119, "lr": 3.0690543348394776e-07, "epoch": 1.7758551036897388, "percentage": 88.79, "elapsed_time": "2 days, 8:42:04", "remaining_time": "7:09:24", "throughput": 5650.14, "total_tokens": 1153332104}
+{"current_steps": 26380, "total_steps": 29704, "loss": 0.2835052490234375, "lr": 3.059939981345467e-07, "epoch": 1.776191758685699, "percentage": 88.81, "elapsed_time": "2 days, 8:42:33", "remaining_time": "7:08:44", "throughput": 5650.39, "total_tokens": 1153547680}
+{"current_steps": 26385, "total_steps": 29704, "loss": 0.28968124389648436, "lr": 3.050838754489566e-07, "epoch": 1.776528413681659, "percentage": 88.83, "elapsed_time": "2 days, 8:43:17", "remaining_time": "7:08:06", "throughput": 5650.23, "total_tokens": 1153764144}
+{"current_steps": 26390, "total_steps": 29704, "loss": 0.29620354175567626, "lr": 3.041750656816894e-07, "epoch": 1.776865068677619, "percentage": 88.84, "elapsed_time": "2 days, 8:43:55", "remaining_time": "7:07:27", "throughput": 5650.25, "total_tokens": 1153982600}
+{"current_steps": 26395, "total_steps": 29704, "loss": 0.3176565647125244, "lr": 3.0326756908689135e-07, "epoch": 1.7772017236735793, "percentage": 88.86, "elapsed_time": "2 days, 8:44:34", "remaining_time": "7:06:48", "throughput": 5650.27, "total_tokens": 1154204320}
+{"current_steps": 26400, "total_steps": 29704, "loss": 0.2815841197967529, "lr": 3.0236138591833995e-07, "epoch": 1.7775383786695396, "percentage": 88.88, "elapsed_time": "2 days, 8:45:12", "remaining_time": "7:06:10", "throughput": 5650.29, "total_tokens": 1154426800}
+{"current_steps": 26405, "total_steps": 29704, "loss": 0.31723904609680176, "lr": 3.0145651642944863e-07, "epoch": 1.7778750336654996, "percentage": 88.89, "elapsed_time": "2 days, 8:45:57", "remaining_time": "7:05:32", "throughput": 5650.19, "total_tokens": 1154656600}
+{"current_steps": 26410, "total_steps": 29704, "loss": 0.26880898475646975, "lr": 3.005529608732588e-07, "epoch": 1.7782116886614596, "percentage": 88.91, "elapsed_time": "2 days, 8:46:34", "remaining_time": "7:04:53", "throughput": 5650.2, "total_tokens": 1154869552}
+{"current_steps": 26415, "total_steps": 29704, "loss": 0.322553014755249, "lr": 2.996507195024495e-07, "epoch": 1.7785483436574199, "percentage": 88.93, "elapsed_time": "2 days, 8:47:15", "remaining_time": "7:04:14", "throughput": 5650.21, "total_tokens": 1155101520}
+{"current_steps": 26420, "total_steps": 29704, "loss": 0.2975711107254028, "lr": 2.9874979256932614e-07, "epoch": 1.7788849986533801, "percentage": 88.94, "elapsed_time": "2 days, 8:47:53", "remaining_time": "7:03:35", "throughput": 5650.24, "total_tokens": 1155322552}
+{"current_steps": 26425, "total_steps": 29704, "loss": 0.3099265813827515, "lr": 2.9785018032583325e-07, "epoch": 1.7792216536493402, "percentage": 88.96, "elapsed_time": "2 days, 8:48:36", "remaining_time": "7:02:57", "throughput": 5650.14, "total_tokens": 1155544168}
+{"current_steps": 26430, "total_steps": 29704, "loss": 0.298630952835083, "lr": 2.969518830235435e-07, "epoch": 1.7795583086453002, "percentage": 88.98, "elapsed_time": "2 days, 8:49:09", "remaining_time": "7:02:18", "throughput": 5650.29, "total_tokens": 1155762856}
+{"current_steps": 26435, "total_steps": 29704, "loss": 0.30402772426605223, "lr": 2.960549009136626e-07, "epoch": 1.7798949636412604, "percentage": 88.99, "elapsed_time": "2 days, 8:49:50", "remaining_time": "7:01:40", "throughput": 5650.23, "total_tokens": 1155983360}
+{"current_steps": 26440, "total_steps": 29704, "loss": 0.33509268760681155, "lr": 2.951592342470289e-07, "epoch": 1.7802316186372207, "percentage": 89.01, "elapsed_time": "2 days, 8:50:25", "remaining_time": "7:01:00", "throughput": 5650.39, "total_tokens": 1156211056}
+{"current_steps": 26445, "total_steps": 29704, "loss": 0.28286046981811525, "lr": 2.942648832741124e-07, "epoch": 1.7805682736331807, "percentage": 89.03, "elapsed_time": "2 days, 8:51:03", "remaining_time": "7:00:22", "throughput": 5650.34, "total_tokens": 1156416184}
+{"current_steps": 26450, "total_steps": 29704, "loss": 0.2894281387329102, "lr": 2.933718482450154e-07, "epoch": 1.7809049286291407, "percentage": 89.05, "elapsed_time": "2 days, 8:51:38", "remaining_time": "6:59:42", "throughput": 5650.44, "total_tokens": 1156635048}
+{"current_steps": 26455, "total_steps": 29704, "loss": 0.3163520097732544, "lr": 2.9248012940947423e-07, "epoch": 1.781241583625101, "percentage": 89.06, "elapsed_time": "2 days, 8:52:09", "remaining_time": "6:59:03", "throughput": 5650.7, "total_tokens": 1156864984}
+{"current_steps": 26460, "total_steps": 29704, "loss": 0.2976935625076294, "lr": 2.915897270168522e-07, "epoch": 1.7815782386210612, "percentage": 89.08, "elapsed_time": "2 days, 8:52:47", "remaining_time": "6:58:24", "throughput": 5650.66, "total_tokens": 1157071688}
+{"current_steps": 26465, "total_steps": 29704, "loss": 0.30093798637390134, "lr": 2.9070064131615004e-07, "epoch": 1.7819148936170213, "percentage": 89.1, "elapsed_time": "2 days, 8:53:29", "remaining_time": "6:57:46", "throughput": 5650.51, "total_tokens": 1157281584}
+{"current_steps": 26470, "total_steps": 29704, "loss": 0.291579532623291, "lr": 2.8981287255599567e-07, "epoch": 1.7822515486129813, "percentage": 89.11, "elapsed_time": "2 days, 8:54:03", "remaining_time": "6:57:07", "throughput": 5650.52, "total_tokens": 1157474720}
+{"current_steps": 26475, "total_steps": 29704, "loss": 0.2964937210083008, "lr": 2.8892642098465205e-07, "epoch": 1.7825882036089415, "percentage": 89.13, "elapsed_time": "2 days, 8:54:47", "remaining_time": "6:56:28", "throughput": 5650.4, "total_tokens": 1157694912}
+{"current_steps": 26480, "total_steps": 29704, "loss": 0.29356560707092283, "lr": 2.8804128685001263e-07, "epoch": 1.7829248586049018, "percentage": 89.15, "elapsed_time": "2 days, 8:55:30", "remaining_time": "6:55:50", "throughput": 5650.32, "total_tokens": 1157923424}
+{"current_steps": 26485, "total_steps": 29704, "loss": 0.2721066474914551, "lr": 2.8715747039960216e-07, "epoch": 1.7832615136008618, "percentage": 89.16, "elapsed_time": "2 days, 8:56:02", "remaining_time": "6:55:11", "throughput": 5650.46, "total_tokens": 1158134336}
+{"current_steps": 26490, "total_steps": 29704, "loss": 0.30223267078399657, "lr": 2.8627497188057684e-07, "epoch": 1.7835981685968219, "percentage": 89.18, "elapsed_time": "2 days, 8:56:42", "remaining_time": "6:54:32", "throughput": 5650.45, "total_tokens": 1158358632}
+{"current_steps": 26495, "total_steps": 29704, "loss": 0.32183716297149656, "lr": 2.853937915397248e-07, "epoch": 1.783934823592782, "percentage": 89.2, "elapsed_time": "2 days, 8:57:21", "remaining_time": "6:53:54", "throughput": 5650.45, "total_tokens": 1158575032}
+{"current_steps": 26500, "total_steps": 29704, "loss": 0.28379087448120116, "lr": 2.845139296234639e-07, "epoch": 1.7842714785887424, "percentage": 89.21, "elapsed_time": "2 days, 8:57:57", "remaining_time": "6:53:15", "throughput": 5650.56, "total_tokens": 1158803496}
+{"current_steps": 26505, "total_steps": 29704, "loss": 0.29067995548248293, "lr": 2.836353863778479e-07, "epoch": 1.7846081335847024, "percentage": 89.23, "elapsed_time": "2 days, 8:58:34", "remaining_time": "6:52:36", "throughput": 5650.6, "total_tokens": 1159023072}
+{"current_steps": 26510, "total_steps": 29704, "loss": 0.3040937900543213, "lr": 2.8275816204855534e-07, "epoch": 1.7849447885806624, "percentage": 89.25, "elapsed_time": "2 days, 8:59:23", "remaining_time": "6:51:58", "throughput": 5650.31, "total_tokens": 1159239184}
+{"current_steps": 26515, "total_steps": 29704, "loss": 0.31920318603515624, "lr": 2.818822568809015e-07, "epoch": 1.7852814435766227, "percentage": 89.26, "elapsed_time": "2 days, 8:59:59", "remaining_time": "6:51:19", "throughput": 5650.32, "total_tokens": 1159445280}
+{"current_steps": 26520, "total_steps": 29704, "loss": 0.2857333183288574, "lr": 2.810076711198284e-07, "epoch": 1.785618098572583, "percentage": 89.28, "elapsed_time": "2 days, 9:00:38", "remaining_time": "6:50:41", "throughput": 5650.3, "total_tokens": 1159658816}
+{"current_steps": 26525, "total_steps": 29704, "loss": 0.3130965709686279, "lr": 2.801344050099131e-07, "epoch": 1.785954753568543, "percentage": 89.3, "elapsed_time": "2 days, 9:01:15", "remaining_time": "6:50:02", "throughput": 5650.45, "total_tokens": 1159896520}
+{"current_steps": 26530, "total_steps": 29704, "loss": 0.3018080711364746, "lr": 2.7926245879536076e-07, "epoch": 1.786291408564503, "percentage": 89.31, "elapsed_time": "2 days, 9:01:58", "remaining_time": "6:49:23", "throughput": 5650.33, "total_tokens": 1160116400}
+{"current_steps": 26535, "total_steps": 29704, "loss": 0.28679165840148924, "lr": 2.7839183272000857e-07, "epoch": 1.7866280635604632, "percentage": 89.33, "elapsed_time": "2 days, 9:02:33", "remaining_time": "6:48:44", "throughput": 5650.49, "total_tokens": 1160350120}
+{"current_steps": 26540, "total_steps": 29704, "loss": 0.30443391799926756, "lr": 2.775225270273241e-07, "epoch": 1.7869647185564235, "percentage": 89.35, "elapsed_time": "2 days, 9:03:08", "remaining_time": "6:48:05", "throughput": 5650.64, "total_tokens": 1160577328}
+{"current_steps": 26545, "total_steps": 29704, "loss": 0.2736891508102417, "lr": 2.7665454196040665e-07, "epoch": 1.7873013735523835, "percentage": 89.37, "elapsed_time": "2 days, 9:03:46", "remaining_time": "6:47:26", "throughput": 5650.7, "total_tokens": 1160803320}
+{"current_steps": 26550, "total_steps": 29704, "loss": 0.31282496452331543, "lr": 2.7578787776198325e-07, "epoch": 1.7876380285483435, "percentage": 89.38, "elapsed_time": "2 days, 9:04:24", "remaining_time": "6:46:48", "throughput": 5650.68, "total_tokens": 1161013768}
+{"current_steps": 26555, "total_steps": 29704, "loss": 0.2743107318878174, "lr": 2.7492253467441775e-07, "epoch": 1.7879746835443038, "percentage": 89.4, "elapsed_time": "2 days, 9:05:00", "remaining_time": "6:46:09", "throughput": 5650.75, "total_tokens": 1161234120}
+{"current_steps": 26560, "total_steps": 29704, "loss": 0.30170412063598634, "lr": 2.740585129396961e-07, "epoch": 1.788311338540264, "percentage": 89.42, "elapsed_time": "2 days, 9:05:45", "remaining_time": "6:45:31", "throughput": 5650.63, "total_tokens": 1161458560}
+{"current_steps": 26565, "total_steps": 29704, "loss": 0.2769881248474121, "lr": 2.731958127994433e-07, "epoch": 1.788647993536224, "percentage": 89.43, "elapsed_time": "2 days, 9:06:26", "remaining_time": "6:44:52", "throughput": 5650.53, "total_tokens": 1161674768}
+{"current_steps": 26570, "total_steps": 29704, "loss": 0.307672905921936, "lr": 2.7233443449490695e-07, "epoch": 1.788984648532184, "percentage": 89.45, "elapsed_time": "2 days, 9:07:10", "remaining_time": "6:44:14", "throughput": 5650.39, "total_tokens": 1161890360}
+{"current_steps": 26575, "total_steps": 29704, "loss": 0.29656333923339845, "lr": 2.7147437826697153e-07, "epoch": 1.7893213035281443, "percentage": 89.47, "elapsed_time": "2 days, 9:07:51", "remaining_time": "6:43:36", "throughput": 5650.3, "total_tokens": 1162107656}
+{"current_steps": 26580, "total_steps": 29704, "loss": 0.3016365528106689, "lr": 2.706156443561481e-07, "epoch": 1.7896579585241046, "percentage": 89.48, "elapsed_time": "2 days, 9:08:29", "remaining_time": "6:42:57", "throughput": 5650.4, "total_tokens": 1162338136}
+{"current_steps": 26585, "total_steps": 29704, "loss": 0.29532361030578613, "lr": 2.6975823300257884e-07, "epoch": 1.7899946135200646, "percentage": 89.5, "elapsed_time": "2 days, 9:09:02", "remaining_time": "6:42:18", "throughput": 5650.51, "total_tokens": 1162551688}
+{"current_steps": 26590, "total_steps": 29704, "loss": 0.29342079162597656, "lr": 2.6890214444603604e-07, "epoch": 1.7903312685160246, "percentage": 89.52, "elapsed_time": "2 days, 9:09:35", "remaining_time": "6:41:38", "throughput": 5650.69, "total_tokens": 1162773064}
+{"current_steps": 26595, "total_steps": 29704, "loss": 0.30418787002563474, "lr": 2.6804737892592193e-07, "epoch": 1.790667923511985, "percentage": 89.53, "elapsed_time": "2 days, 9:10:24", "remaining_time": "6:41:01", "throughput": 5650.38, "total_tokens": 1162987848}
+{"current_steps": 26600, "total_steps": 29704, "loss": 0.30772976875305175, "lr": 2.671939366812687e-07, "epoch": 1.7910045785079451, "percentage": 89.55, "elapsed_time": "2 days, 9:11:02", "remaining_time": "6:40:22", "throughput": 5650.38, "total_tokens": 1163202312}
+{"current_steps": 26605, "total_steps": 29704, "loss": 0.29974985122680664, "lr": 2.66341817950741e-07, "epoch": 1.7913412335039052, "percentage": 89.57, "elapsed_time": "2 days, 9:11:38", "remaining_time": "6:39:43", "throughput": 5650.55, "total_tokens": 1163438080}
+{"current_steps": 26610, "total_steps": 29704, "loss": 0.29571521282196045, "lr": 2.6549102297262764e-07, "epoch": 1.7916778884998652, "percentage": 89.58, "elapsed_time": "2 days, 9:12:19", "remaining_time": "6:39:05", "throughput": 5650.49, "total_tokens": 1163660200}
+{"current_steps": 26615, "total_steps": 29704, "loss": 0.30173745155334475, "lr": 2.6464155198485387e-07, "epoch": 1.7920145434958255, "percentage": 89.6, "elapsed_time": "2 days, 9:12:59", "remaining_time": "6:38:26", "throughput": 5650.51, "total_tokens": 1163887248}
+{"current_steps": 26620, "total_steps": 29704, "loss": 0.29706802368164065, "lr": 2.6379340522496853e-07, "epoch": 1.7923511984917857, "percentage": 89.62, "elapsed_time": "2 days, 9:13:33", "remaining_time": "6:37:47", "throughput": 5650.65, "total_tokens": 1164110704}
+{"current_steps": 26625, "total_steps": 29704, "loss": 0.3205864429473877, "lr": 2.6294658293015584e-07, "epoch": 1.7926878534877457, "percentage": 89.63, "elapsed_time": "2 days, 9:14:10", "remaining_time": "6:37:08", "throughput": 5650.73, "total_tokens": 1164337200}
+{"current_steps": 26630, "total_steps": 29704, "loss": 0.31758651733398435, "lr": 2.621010853372258e-07, "epoch": 1.7930245084837058, "percentage": 89.65, "elapsed_time": "2 days, 9:14:51", "remaining_time": "6:36:29", "throughput": 5650.61, "total_tokens": 1164544968}
+{"current_steps": 26635, "total_steps": 29704, "loss": 0.3288443088531494, "lr": 2.612569126826187e-07, "epoch": 1.793361163479666, "percentage": 89.67, "elapsed_time": "2 days, 9:15:29", "remaining_time": "6:35:51", "throughput": 5650.64, "total_tokens": 1164763392}
+{"current_steps": 26640, "total_steps": 29704, "loss": 0.3208475589752197, "lr": 2.6041406520240563e-07, "epoch": 1.7936978184756263, "percentage": 89.68, "elapsed_time": "2 days, 9:16:02", "remaining_time": "6:35:11", "throughput": 5650.86, "total_tokens": 1164996760}
+{"current_steps": 26645, "total_steps": 29704, "loss": 0.3083490371704102, "lr": 2.5957254313228595e-07, "epoch": 1.7940344734715863, "percentage": 89.7, "elapsed_time": "2 days, 9:16:43", "remaining_time": "6:34:33", "throughput": 5650.78, "total_tokens": 1165209184}
+{"current_steps": 26650, "total_steps": 29704, "loss": 0.2793558597564697, "lr": 2.5873234670758753e-07, "epoch": 1.7943711284675463, "percentage": 89.72, "elapsed_time": "2 days, 9:17:25", "remaining_time": "6:33:55", "throughput": 5650.74, "total_tokens": 1165437296}
+{"current_steps": 26655, "total_steps": 29704, "loss": 0.29536702632904055, "lr": 2.578934761632712e-07, "epoch": 1.7947077834635066, "percentage": 89.74, "elapsed_time": "2 days, 9:17:56", "remaining_time": "6:33:15", "throughput": 5651.0, "total_tokens": 1165668480}
+{"current_steps": 26660, "total_steps": 29704, "loss": 0.30143041610717775, "lr": 2.570559317339211e-07, "epoch": 1.7950444384594668, "percentage": 89.75, "elapsed_time": "2 days, 9:18:38", "remaining_time": "6:32:37", "throughput": 5650.94, "total_tokens": 1165893536}
+{"current_steps": 26665, "total_steps": 29704, "loss": 0.2995075941085815, "lr": 2.56219713653757e-07, "epoch": 1.7953810934554268, "percentage": 89.77, "elapsed_time": "2 days, 9:19:12", "remaining_time": "6:31:57", "throughput": 5651.11, "total_tokens": 1166122240}
+{"current_steps": 26670, "total_steps": 29704, "loss": 0.3236057758331299, "lr": 2.5538482215662187e-07, "epoch": 1.7957177484513869, "percentage": 89.79, "elapsed_time": "2 days, 9:19:46", "remaining_time": "6:31:18", "throughput": 5651.23, "total_tokens": 1166338248}
+{"current_steps": 26675, "total_steps": 29704, "loss": 0.2701665163040161, "lr": 2.545512574759923e-07, "epoch": 1.7960544034473471, "percentage": 89.8, "elapsed_time": "2 days, 9:20:33", "remaining_time": "6:30:40", "throughput": 5650.99, "total_tokens": 1166555592}
+{"current_steps": 26680, "total_steps": 29704, "loss": 0.31007416248321534, "lr": 2.537190198449713e-07, "epoch": 1.7963910584433074, "percentage": 89.82, "elapsed_time": "2 days, 9:21:07", "remaining_time": "6:30:01", "throughput": 5651.12, "total_tokens": 1166772216}
+{"current_steps": 26685, "total_steps": 29704, "loss": 0.3006875991821289, "lr": 2.5288810949629207e-07, "epoch": 1.7967277134392674, "percentage": 89.84, "elapsed_time": "2 days, 9:21:49", "remaining_time": "6:29:23", "throughput": 5651.06, "total_tokens": 1166996856}
+{"current_steps": 26690, "total_steps": 29704, "loss": 0.2778863668441772, "lr": 2.5205852666231544e-07, "epoch": 1.7970643684352274, "percentage": 89.85, "elapsed_time": "2 days, 9:22:32", "remaining_time": "6:28:45", "throughput": 5650.91, "total_tokens": 1167208832}
+{"current_steps": 26695, "total_steps": 29704, "loss": 0.28261337280273435, "lr": 2.5123027157503144e-07, "epoch": 1.7974010234311877, "percentage": 89.87, "elapsed_time": "2 days, 9:23:10", "remaining_time": "6:28:06", "throughput": 5650.84, "total_tokens": 1167408464}
+{"current_steps": 26700, "total_steps": 29704, "loss": 0.3025941848754883, "lr": 2.504033444660592e-07, "epoch": 1.797737678427148, "percentage": 89.89, "elapsed_time": "2 days, 9:23:47", "remaining_time": "6:27:27", "throughput": 5650.97, "total_tokens": 1167643608}
+{"current_steps": 26705, "total_steps": 29704, "loss": 0.28271756172180174, "lr": 2.4957774556664706e-07, "epoch": 1.798074333423108, "percentage": 89.9, "elapsed_time": "2 days, 9:24:30", "remaining_time": "6:26:49", "throughput": 5650.87, "total_tokens": 1167868528}
+{"current_steps": 26710, "total_steps": 29704, "loss": 0.29103024005889894, "lr": 2.487534751076692e-07, "epoch": 1.798410988419068, "percentage": 89.92, "elapsed_time": "2 days, 9:25:15", "remaining_time": "6:26:11", "throughput": 5650.72, "total_tokens": 1168090128}
+{"current_steps": 26715, "total_steps": 29704, "loss": 0.2806957721710205, "lr": 2.47930533319633e-07, "epoch": 1.7987476434150282, "percentage": 89.94, "elapsed_time": "2 days, 9:25:57", "remaining_time": "6:25:33", "throughput": 5650.58, "total_tokens": 1168302824}
+{"current_steps": 26720, "total_steps": 29704, "loss": 0.3256166219711304, "lr": 2.4710892043266823e-07, "epoch": 1.7990842984109885, "percentage": 89.95, "elapsed_time": "2 days, 9:26:33", "remaining_time": "6:24:53", "throughput": 5650.7, "total_tokens": 1168526496}
+{"current_steps": 26725, "total_steps": 29704, "loss": 0.2893906354904175, "lr": 2.462886366765388e-07, "epoch": 1.7994209534069485, "percentage": 89.97, "elapsed_time": "2 days, 9:27:14", "remaining_time": "6:24:15", "throughput": 5650.63, "total_tokens": 1168747720}
+{"current_steps": 26730, "total_steps": 29704, "loss": 0.3119854211807251, "lr": 2.454696822806341e-07, "epoch": 1.7997576084029085, "percentage": 89.99, "elapsed_time": "2 days, 9:27:56", "remaining_time": "6:23:37", "throughput": 5650.49, "total_tokens": 1168955928}
+{"current_steps": 26735, "total_steps": 29704, "loss": 0.2833410739898682, "lr": 2.446520574739708e-07, "epoch": 1.8000942633988688, "percentage": 90.0, "elapsed_time": "2 days, 9:28:30", "remaining_time": "6:22:58", "throughput": 5650.67, "total_tokens": 1169181640}
+{"current_steps": 26740, "total_steps": 29704, "loss": 0.265745735168457, "lr": 2.438357624851967e-07, "epoch": 1.800430918394829, "percentage": 90.02, "elapsed_time": "2 days, 9:29:04", "remaining_time": "6:22:18", "throughput": 5650.78, "total_tokens": 1169399504}
+{"current_steps": 26745, "total_steps": 29704, "loss": 0.30634517669677735, "lr": 2.430207975425847e-07, "epoch": 1.800767573390789, "percentage": 90.04, "elapsed_time": "2 days, 9:29:46", "remaining_time": "6:21:40", "throughput": 5650.73, "total_tokens": 1169627072}
+{"current_steps": 26750, "total_steps": 29704, "loss": 0.306353759765625, "lr": 2.4220716287403744e-07, "epoch": 1.801104228386749, "percentage": 90.06, "elapsed_time": "2 days, 9:30:24", "remaining_time": "6:21:01", "throughput": 5650.83, "total_tokens": 1169859232}
+{"current_steps": 26755, "total_steps": 29704, "loss": 0.2988481044769287, "lr": 2.413948587070869e-07, "epoch": 1.8014408833827094, "percentage": 90.07, "elapsed_time": "2 days, 9:31:03", "remaining_time": "6:20:23", "throughput": 5650.63, "total_tokens": 1170041408}
+{"current_steps": 26760, "total_steps": 29704, "loss": 0.3097994327545166, "lr": 2.405838852688891e-07, "epoch": 1.8017775383786696, "percentage": 90.09, "elapsed_time": "2 days, 9:31:35", "remaining_time": "6:19:43", "throughput": 5650.87, "total_tokens": 1170267480}
+{"current_steps": 26765, "total_steps": 29704, "loss": 0.2666957378387451, "lr": 2.397742427862315e-07, "epoch": 1.8021141933746296, "percentage": 90.11, "elapsed_time": "2 days, 9:32:09", "remaining_time": "6:19:04", "throughput": 5651.04, "total_tokens": 1170495840}
+{"current_steps": 26770, "total_steps": 29704, "loss": 0.3030317544937134, "lr": 2.3896593148552695e-07, "epoch": 1.8024508483705897, "percentage": 90.12, "elapsed_time": "2 days, 9:32:44", "remaining_time": "6:18:25", "throughput": 5651.09, "total_tokens": 1170707968}
+{"current_steps": 26775, "total_steps": 29704, "loss": 0.28127307891845704, "lr": 2.3815895159281844e-07, "epoch": 1.80278750336655, "percentage": 90.14, "elapsed_time": "2 days, 9:33:21", "remaining_time": "6:17:46", "throughput": 5651.18, "total_tokens": 1170932136}
+{"current_steps": 26780, "total_steps": 29704, "loss": 0.2792118310928345, "lr": 2.3735330333377493e-07, "epoch": 1.8031241583625102, "percentage": 90.16, "elapsed_time": "2 days, 9:33:58", "remaining_time": "6:17:07", "throughput": 5651.29, "total_tokens": 1171162408}
+{"current_steps": 26785, "total_steps": 29704, "loss": 0.29738240242004393, "lr": 2.3654898693369232e-07, "epoch": 1.8034608133584702, "percentage": 90.17, "elapsed_time": "2 days, 9:34:41", "remaining_time": "6:16:29", "throughput": 5651.14, "total_tokens": 1171377408}
+{"current_steps": 26790, "total_steps": 29704, "loss": 0.30119688510894777, "lr": 2.3574600261749735e-07, "epoch": 1.8037974683544302, "percentage": 90.19, "elapsed_time": "2 days, 9:35:21", "remaining_time": "6:15:50", "throughput": 5651.12, "total_tokens": 1171597216}
+{"current_steps": 26795, "total_steps": 29704, "loss": 0.3168250322341919, "lr": 2.3494435060973875e-07, "epoch": 1.8041341233503905, "percentage": 90.21, "elapsed_time": "2 days, 9:35:55", "remaining_time": "6:15:11", "throughput": 5651.26, "total_tokens": 1171822008}
+{"current_steps": 26800, "total_steps": 29704, "loss": 0.30097527503967286, "lr": 2.3414403113459937e-07, "epoch": 1.8044707783463507, "percentage": 90.22, "elapsed_time": "2 days, 9:36:32", "remaining_time": "6:14:32", "throughput": 5651.32, "total_tokens": 1172040272}
+{"current_steps": 26805, "total_steps": 29704, "loss": 0.2781896352767944, "lr": 2.3334504441588413e-07, "epoch": 1.8048074333423108, "percentage": 90.24, "elapsed_time": "2 days, 9:37:12", "remaining_time": "6:13:54", "throughput": 5651.29, "total_tokens": 1172260704}
+{"current_steps": 26810, "total_steps": 29704, "loss": 0.2797673225402832, "lr": 2.3254739067702704e-07, "epoch": 1.8051440883382708, "percentage": 90.26, "elapsed_time": "2 days, 9:37:49", "remaining_time": "6:13:15", "throughput": 5651.35, "total_tokens": 1172485680}
+{"current_steps": 26815, "total_steps": 29704, "loss": 0.2838028907775879, "lr": 2.3175107014109077e-07, "epoch": 1.805480743334231, "percentage": 90.27, "elapsed_time": "2 days, 9:38:27", "remaining_time": "6:12:36", "throughput": 5651.4, "total_tokens": 1172707720}
+{"current_steps": 26820, "total_steps": 29704, "loss": 0.311909818649292, "lr": 2.3095608303076223e-07, "epoch": 1.8058173983301913, "percentage": 90.29, "elapsed_time": "2 days, 9:39:11", "remaining_time": "6:11:58", "throughput": 5651.26, "total_tokens": 1172928712}
+{"current_steps": 26825, "total_steps": 29704, "loss": 0.31490211486816405, "lr": 2.3016242956835798e-07, "epoch": 1.8061540533261513, "percentage": 90.31, "elapsed_time": "2 days, 9:39:54", "remaining_time": "6:11:20", "throughput": 5651.19, "total_tokens": 1173154336}
+{"current_steps": 26830, "total_steps": 29704, "loss": 0.284251070022583, "lr": 2.2937010997581999e-07, "epoch": 1.8064907083221113, "percentage": 90.32, "elapsed_time": "2 days, 9:40:34", "remaining_time": "6:10:41", "throughput": 5651.21, "total_tokens": 1173384008}
+{"current_steps": 26835, "total_steps": 29704, "loss": 0.2854511260986328, "lr": 2.2857912447471818e-07, "epoch": 1.8068273633180716, "percentage": 90.34, "elapsed_time": "2 days, 9:41:10", "remaining_time": "6:10:02", "throughput": 5651.21, "total_tokens": 1173591424}
+{"current_steps": 26840, "total_steps": 29704, "loss": 0.2829423427581787, "lr": 2.2778947328625012e-07, "epoch": 1.8071640183140318, "percentage": 90.36, "elapsed_time": "2 days, 9:41:47", "remaining_time": "6:09:23", "throughput": 5651.19, "total_tokens": 1173795056}
+{"current_steps": 26845, "total_steps": 29704, "loss": 0.3229234218597412, "lr": 2.2700115663123745e-07, "epoch": 1.8075006733099919, "percentage": 90.38, "elapsed_time": "2 days, 9:42:24", "remaining_time": "6:08:44", "throughput": 5651.22, "total_tokens": 1174010232}
+{"current_steps": 26850, "total_steps": 29704, "loss": 0.28131952285766604, "lr": 2.2621417473013164e-07, "epoch": 1.807837328305952, "percentage": 90.39, "elapsed_time": "2 days, 9:43:05", "remaining_time": "6:08:06", "throughput": 5651.17, "total_tokens": 1174230336}
+{"current_steps": 26855, "total_steps": 29704, "loss": 0.30530600547790526, "lr": 2.254285278030094e-07, "epoch": 1.8081739833019121, "percentage": 90.41, "elapsed_time": "2 days, 9:43:39", "remaining_time": "6:07:27", "throughput": 5651.35, "total_tokens": 1174459216}
+{"current_steps": 26860, "total_steps": 29704, "loss": 0.30002517700195314, "lr": 2.24644216069575e-07, "epoch": 1.8085106382978724, "percentage": 90.43, "elapsed_time": "2 days, 9:44:26", "remaining_time": "6:06:49", "throughput": 5651.07, "total_tokens": 1174667136}
+{"current_steps": 26865, "total_steps": 29704, "loss": 0.2729137420654297, "lr": 2.238612397491574e-07, "epoch": 1.8088472932938324, "percentage": 90.44, "elapsed_time": "2 days, 9:45:05", "remaining_time": "6:06:10", "throughput": 5651.09, "total_tokens": 1174893104}
+{"current_steps": 26870, "total_steps": 29704, "loss": 0.29819502830505373, "lr": 2.230795990607143e-07, "epoch": 1.8091839482897925, "percentage": 90.46, "elapsed_time": "2 days, 9:45:46", "remaining_time": "6:05:32", "throughput": 5651.07, "total_tokens": 1175120424}
+{"current_steps": 26875, "total_steps": 29704, "loss": 0.27527265548706054, "lr": 2.2229929422282915e-07, "epoch": 1.809520603285753, "percentage": 90.48, "elapsed_time": "2 days, 9:46:22", "remaining_time": "6:04:53", "throughput": 5651.13, "total_tokens": 1175334208}
+{"current_steps": 26880, "total_steps": 29704, "loss": 0.3183107614517212, "lr": 2.215203254537107e-07, "epoch": 1.809857258281713, "percentage": 90.49, "elapsed_time": "2 days, 9:47:02", "remaining_time": "6:04:14", "throughput": 5651.09, "total_tokens": 1175555296}
+{"current_steps": 26885, "total_steps": 29704, "loss": 0.3029907464981079, "lr": 2.2074269297119588e-07, "epoch": 1.810193913277673, "percentage": 90.51, "elapsed_time": "2 days, 9:47:36", "remaining_time": "6:03:35", "throughput": 5651.27, "total_tokens": 1175780944}
+{"current_steps": 26890, "total_steps": 29704, "loss": 0.29438438415527346, "lr": 2.199663969927479e-07, "epoch": 1.8105305682736332, "percentage": 90.53, "elapsed_time": "2 days, 9:48:16", "remaining_time": "6:02:56", "throughput": 5651.21, "total_tokens": 1175996056}
+{"current_steps": 26895, "total_steps": 29704, "loss": 0.29373531341552733, "lr": 2.1919143773545316e-07, "epoch": 1.8108672232695935, "percentage": 90.54, "elapsed_time": "2 days, 9:48:54", "remaining_time": "6:02:18", "throughput": 5651.25, "total_tokens": 1176219504}
+{"current_steps": 26900, "total_steps": 29704, "loss": 0.29395351409912107, "lr": 2.184178154160288e-07, "epoch": 1.8112038782655535, "percentage": 90.56, "elapsed_time": "2 days, 9:49:30", "remaining_time": "6:01:39", "throughput": 5651.31, "total_tokens": 1176436808}
+{"current_steps": 26905, "total_steps": 29704, "loss": 0.2931424379348755, "lr": 2.1764553025081514e-07, "epoch": 1.8115405332615135, "percentage": 90.58, "elapsed_time": "2 days, 9:50:07", "remaining_time": "6:01:00", "throughput": 5651.37, "total_tokens": 1176656768}
+{"current_steps": 26910, "total_steps": 29704, "loss": 0.28325557708740234, "lr": 2.168745824557794e-07, "epoch": 1.8118771882574738, "percentage": 90.59, "elapsed_time": "2 days, 9:50:50", "remaining_time": "6:00:22", "throughput": 5651.22, "total_tokens": 1176871952}
+{"current_steps": 26915, "total_steps": 29704, "loss": 0.28316502571105956, "lr": 2.161049722465136e-07, "epoch": 1.812213843253434, "percentage": 90.61, "elapsed_time": "2 days, 9:51:33", "remaining_time": "5:59:43", "throughput": 5651.09, "total_tokens": 1177088104}
+{"current_steps": 26920, "total_steps": 29704, "loss": 0.29533112049102783, "lr": 2.1533669983823835e-07, "epoch": 1.812550498249394, "percentage": 90.63, "elapsed_time": "2 days, 9:52:16", "remaining_time": "5:59:05", "throughput": 5651.04, "total_tokens": 1177320360}
+{"current_steps": 26925, "total_steps": 29704, "loss": 0.2751606464385986, "lr": 2.1456976544579732e-07, "epoch": 1.812887153245354, "percentage": 90.64, "elapsed_time": "2 days, 9:53:02", "remaining_time": "5:58:27", "throughput": 5650.96, "total_tokens": 1177560024}
+{"current_steps": 26930, "total_steps": 29704, "loss": 0.29282474517822266, "lr": 2.138041692836623e-07, "epoch": 1.8132238082413143, "percentage": 90.66, "elapsed_time": "2 days, 9:53:34", "remaining_time": "5:57:48", "throughput": 5651.13, "total_tokens": 1177776744}
+{"current_steps": 26935, "total_steps": 29704, "loss": 0.29790616035461426, "lr": 2.1303991156592818e-07, "epoch": 1.8135604632372746, "percentage": 90.68, "elapsed_time": "2 days, 9:54:12", "remaining_time": "5:57:09", "throughput": 5651.1, "total_tokens": 1177986848}
+{"current_steps": 26940, "total_steps": 29704, "loss": 0.2797182559967041, "lr": 2.122769925063195e-07, "epoch": 1.8138971182332346, "percentage": 90.69, "elapsed_time": "2 days, 9:54:53", "remaining_time": "5:56:31", "throughput": 5651.02, "total_tokens": 1178200832}
+{"current_steps": 26945, "total_steps": 29704, "loss": 0.2855216503143311, "lr": 2.1151541231818174e-07, "epoch": 1.8142337732291947, "percentage": 90.71, "elapsed_time": "2 days, 9:55:36", "remaining_time": "5:55:52", "throughput": 5650.9, "total_tokens": 1178415944}
+{"current_steps": 26950, "total_steps": 29704, "loss": 0.31327362060546876, "lr": 2.1075517121449063e-07, "epoch": 1.814570428225155, "percentage": 90.73, "elapsed_time": "2 days, 9:56:19", "remaining_time": "5:55:14", "throughput": 5650.79, "total_tokens": 1178637904}
+{"current_steps": 26955, "total_steps": 29704, "loss": 0.27339811325073243, "lr": 2.0999626940784334e-07, "epoch": 1.8149070832211152, "percentage": 90.75, "elapsed_time": "2 days, 9:56:55", "remaining_time": "5:54:35", "throughput": 5650.79, "total_tokens": 1178843008}
+{"current_steps": 26960, "total_steps": 29704, "loss": 0.2946287155151367, "lr": 2.0923870711046568e-07, "epoch": 1.8152437382170752, "percentage": 90.76, "elapsed_time": "2 days, 9:57:33", "remaining_time": "5:53:56", "throughput": 5650.84, "total_tokens": 1179067688}
+{"current_steps": 26965, "total_steps": 29704, "loss": 0.2847181797027588, "lr": 2.0848248453420704e-07, "epoch": 1.8155803932130352, "percentage": 90.78, "elapsed_time": "2 days, 9:58:10", "remaining_time": "5:53:17", "throughput": 5650.93, "total_tokens": 1179293168}
+{"current_steps": 26970, "total_steps": 29704, "loss": 0.2915675163269043, "lr": 2.077276018905422e-07, "epoch": 1.8159170482089955, "percentage": 90.8, "elapsed_time": "2 days, 9:58:49", "remaining_time": "5:52:39", "throughput": 5650.92, "total_tokens": 1179511568}
+{"current_steps": 26975, "total_steps": 29704, "loss": 0.3022061824798584, "lr": 2.0697405939057225e-07, "epoch": 1.8162537032049557, "percentage": 90.81, "elapsed_time": "2 days, 9:59:23", "remaining_time": "5:52:00", "throughput": 5650.94, "total_tokens": 1179708384}
+{"current_steps": 26980, "total_steps": 29704, "loss": 0.27937405109405516, "lr": 2.0622185724502253e-07, "epoch": 1.8165903582009157, "percentage": 90.83, "elapsed_time": "2 days, 9:59:57", "remaining_time": "5:51:21", "throughput": 5651.04, "total_tokens": 1179925616}
+{"current_steps": 26985, "total_steps": 29704, "loss": 0.2782083511352539, "lr": 2.0547099566424367e-07, "epoch": 1.8169270131968758, "percentage": 90.85, "elapsed_time": "2 days, 10:00:42", "remaining_time": "5:50:42", "throughput": 5650.88, "total_tokens": 1180142704}
+{"current_steps": 26990, "total_steps": 29704, "loss": 0.27927365303039553, "lr": 2.0472147485821325e-07, "epoch": 1.817263668192836, "percentage": 90.86, "elapsed_time": "2 days, 10:01:19", "remaining_time": "5:50:04", "throughput": 5650.85, "total_tokens": 1180348248}
+{"current_steps": 26995, "total_steps": 29704, "loss": 0.2773773670196533, "lr": 2.0397329503653084e-07, "epoch": 1.8176003231887963, "percentage": 90.88, "elapsed_time": "2 days, 10:01:56", "remaining_time": "5:49:25", "throughput": 5650.88, "total_tokens": 1180563544}
+{"current_steps": 27000, "total_steps": 29704, "loss": 0.3106630802154541, "lr": 2.03226456408423e-07, "epoch": 1.8179369781847563, "percentage": 90.9, "elapsed_time": "2 days, 10:02:40", "remaining_time": "5:48:46", "throughput": 5650.66, "total_tokens": 1180762848}
+{"current_steps": 27005, "total_steps": 29704, "loss": 0.30523037910461426, "lr": 2.024809591827409e-07, "epoch": 1.8182736331807163, "percentage": 90.91, "elapsed_time": "2 days, 10:03:22", "remaining_time": "5:48:08", "throughput": 5650.59, "total_tokens": 1180988840}
+{"current_steps": 27010, "total_steps": 29704, "loss": 0.26985788345336914, "lr": 2.0173680356796066e-07, "epoch": 1.8186102881766766, "percentage": 90.93, "elapsed_time": "2 days, 10:04:01", "remaining_time": "5:47:29", "throughput": 5650.51, "total_tokens": 1181189432}
+{"current_steps": 27015, "total_steps": 29704, "loss": 0.2943392515182495, "lr": 2.00993989772183e-07, "epoch": 1.8189469431726368, "percentage": 90.95, "elapsed_time": "2 days, 10:04:33", "remaining_time": "5:46:50", "throughput": 5650.68, "total_tokens": 1181407992}
+{"current_steps": 27020, "total_steps": 29704, "loss": 0.3112603187561035, "lr": 2.0025251800313284e-07, "epoch": 1.8192835981685969, "percentage": 90.96, "elapsed_time": "2 days, 10:05:20", "remaining_time": "5:46:12", "throughput": 5650.44, "total_tokens": 1181622192}
+{"current_steps": 27025, "total_steps": 29704, "loss": 0.3185489892959595, "lr": 1.9951238846816045e-07, "epoch": 1.8196202531645569, "percentage": 90.98, "elapsed_time": "2 days, 10:05:52", "remaining_time": "5:45:33", "throughput": 5650.69, "total_tokens": 1181854088}
+{"current_steps": 27030, "total_steps": 29704, "loss": 0.3113109111785889, "lr": 1.987736013742414e-07, "epoch": 1.8199569081605171, "percentage": 91.0, "elapsed_time": "2 days, 10:06:31", "remaining_time": "5:44:54", "throughput": 5650.65, "total_tokens": 1182065984}
+{"current_steps": 27035, "total_steps": 29704, "loss": 0.27271738052368166, "lr": 1.9803615692797375e-07, "epoch": 1.8202935631564774, "percentage": 91.01, "elapsed_time": "2 days, 10:07:08", "remaining_time": "5:44:15", "throughput": 5650.7, "total_tokens": 1182287536}
+{"current_steps": 27040, "total_steps": 29704, "loss": 0.29187893867492676, "lr": 1.9730005533558417e-07, "epoch": 1.8206302181524374, "percentage": 91.03, "elapsed_time": "2 days, 10:07:52", "remaining_time": "5:43:37", "throughput": 5650.61, "total_tokens": 1182520232}
+{"current_steps": 27045, "total_steps": 29704, "loss": 0.27612121105194093, "lr": 1.9656529680291802e-07, "epoch": 1.8209668731483974, "percentage": 91.05, "elapsed_time": "2 days, 10:08:32", "remaining_time": "5:42:59", "throughput": 5650.61, "total_tokens": 1182742904}
+{"current_steps": 27050, "total_steps": 29704, "loss": 0.28553037643432616, "lr": 1.9583188153544986e-07, "epoch": 1.8213035281443577, "percentage": 91.07, "elapsed_time": "2 days, 10:09:07", "remaining_time": "5:42:20", "throughput": 5650.62, "total_tokens": 1182945672}
+{"current_steps": 27055, "total_steps": 29704, "loss": 0.2979405641555786, "lr": 1.9509980973827724e-07, "epoch": 1.821640183140318, "percentage": 91.08, "elapsed_time": "2 days, 10:09:46", "remaining_time": "5:41:41", "throughput": 5650.62, "total_tokens": 1183165608}
+{"current_steps": 27060, "total_steps": 29704, "loss": 0.3077999114990234, "lr": 1.943690816161209e-07, "epoch": 1.821976838136278, "percentage": 91.1, "elapsed_time": "2 days, 10:10:24", "remaining_time": "5:41:02", "throughput": 5650.71, "total_tokens": 1183395432}
+{"current_steps": 27065, "total_steps": 29704, "loss": 0.2753284454345703, "lr": 1.9363969737332677e-07, "epoch": 1.822313493132238, "percentage": 91.12, "elapsed_time": "2 days, 10:11:00", "remaining_time": "5:40:23", "throughput": 5650.8, "total_tokens": 1183617608}
+{"current_steps": 27070, "total_steps": 29704, "loss": 0.27479655742645265, "lr": 1.9291165721386507e-07, "epoch": 1.8226501481281983, "percentage": 91.13, "elapsed_time": "2 days, 10:11:35", "remaining_time": "5:39:44", "throughput": 5650.85, "total_tokens": 1183826768}
+{"current_steps": 27075, "total_steps": 29704, "loss": 0.3049477100372314, "lr": 1.9218496134132957e-07, "epoch": 1.8229868031241585, "percentage": 91.15, "elapsed_time": "2 days, 10:12:11", "remaining_time": "5:39:05", "throughput": 5650.95, "total_tokens": 1184052800}
+{"current_steps": 27080, "total_steps": 29704, "loss": 0.282161283493042, "lr": 1.9145960995893831e-07, "epoch": 1.8233234581201185, "percentage": 91.17, "elapsed_time": "2 days, 10:12:47", "remaining_time": "5:38:26", "throughput": 5651.03, "total_tokens": 1184270392}
+{"current_steps": 27085, "total_steps": 29704, "loss": 0.3012186050415039, "lr": 1.9073560326953344e-07, "epoch": 1.8236601131160786, "percentage": 91.18, "elapsed_time": "2 days, 10:13:21", "remaining_time": "5:37:47", "throughput": 5651.18, "total_tokens": 1184494848}
+{"current_steps": 27090, "total_steps": 29704, "loss": 0.3109905242919922, "lr": 1.9001294147558303e-07, "epoch": 1.8239967681120388, "percentage": 91.2, "elapsed_time": "2 days, 10:14:01", "remaining_time": "5:37:09", "throughput": 5651.11, "total_tokens": 1184709808}
+{"current_steps": 27095, "total_steps": 29704, "loss": 0.3057056188583374, "lr": 1.892916247791743e-07, "epoch": 1.824333423107999, "percentage": 91.22, "elapsed_time": "2 days, 10:14:38", "remaining_time": "5:36:30", "throughput": 5651.13, "total_tokens": 1184920824}
+{"current_steps": 27100, "total_steps": 29704, "loss": 0.2988601207733154, "lr": 1.885716533820231e-07, "epoch": 1.824670078103959, "percentage": 91.23, "elapsed_time": "2 days, 10:15:21", "remaining_time": "5:35:51", "throughput": 5651.01, "total_tokens": 1185141448}
+{"current_steps": 27105, "total_steps": 29704, "loss": 0.2867612838745117, "lr": 1.878530274854662e-07, "epoch": 1.8250067330999191, "percentage": 91.25, "elapsed_time": "2 days, 10:15:57", "remaining_time": "5:35:12", "throughput": 5651.09, "total_tokens": 1185358376}
+{"current_steps": 27110, "total_steps": 29704, "loss": 0.3091785669326782, "lr": 1.8713574729046557e-07, "epoch": 1.8253433880958794, "percentage": 91.27, "elapsed_time": "2 days, 10:16:33", "remaining_time": "5:34:33", "throughput": 5651.21, "total_tokens": 1185587344}
+{"current_steps": 27115, "total_steps": 29704, "loss": 0.29340548515319825, "lr": 1.8641981299760582e-07, "epoch": 1.8256800430918396, "percentage": 91.28, "elapsed_time": "2 days, 10:17:06", "remaining_time": "5:33:54", "throughput": 5651.31, "total_tokens": 1185795904}
+{"current_steps": 27120, "total_steps": 29704, "loss": 0.30235800743103025, "lr": 1.857052248070962e-07, "epoch": 1.8260166980877997, "percentage": 91.3, "elapsed_time": "2 days, 10:17:48", "remaining_time": "5:33:16", "throughput": 5651.19, "total_tokens": 1186003992}
+{"current_steps": 27125, "total_steps": 29704, "loss": 0.2838723659515381, "lr": 1.8499198291876907e-07, "epoch": 1.8263533530837597, "percentage": 91.32, "elapsed_time": "2 days, 10:18:28", "remaining_time": "5:32:37", "throughput": 5651.06, "total_tokens": 1186206928}
+{"current_steps": 27130, "total_steps": 29704, "loss": 0.3044004440307617, "lr": 1.8428008753207937e-07, "epoch": 1.82669000807972, "percentage": 91.33, "elapsed_time": "2 days, 10:19:12", "remaining_time": "5:31:59", "throughput": 5650.95, "total_tokens": 1186430512}
+{"current_steps": 27135, "total_steps": 29704, "loss": 0.3086956262588501, "lr": 1.8356953884610673e-07, "epoch": 1.8270266630756802, "percentage": 91.35, "elapsed_time": "2 days, 10:19:56", "remaining_time": "5:31:21", "throughput": 5650.82, "total_tokens": 1186652392}
+{"current_steps": 27140, "total_steps": 29704, "loss": 0.2920703411102295, "lr": 1.8286033705955554e-07, "epoch": 1.8273633180716402, "percentage": 91.37, "elapsed_time": "2 days, 10:20:28", "remaining_time": "5:30:42", "throughput": 5651.03, "total_tokens": 1186878432}
+{"current_steps": 27145, "total_steps": 29704, "loss": 0.2958489656448364, "lr": 1.8215248237074879e-07, "epoch": 1.8276999730676002, "percentage": 91.38, "elapsed_time": "2 days, 10:21:07", "remaining_time": "5:30:03", "throughput": 5651.03, "total_tokens": 1187100664}
+{"current_steps": 27150, "total_steps": 29704, "loss": 0.2915064334869385, "lr": 1.8144597497763816e-07, "epoch": 1.8280366280635605, "percentage": 91.4, "elapsed_time": "2 days, 10:21:45", "remaining_time": "5:29:24", "throughput": 5651.04, "total_tokens": 1187314688}
+{"current_steps": 27155, "total_steps": 29704, "loss": 0.2950667142868042, "lr": 1.8074081507779506e-07, "epoch": 1.8283732830595207, "percentage": 91.42, "elapsed_time": "2 days, 10:22:25", "remaining_time": "5:28:46", "throughput": 5651.06, "total_tokens": 1187543272}
+{"current_steps": 27160, "total_steps": 29704, "loss": 0.3046703815460205, "lr": 1.8003700286841563e-07, "epoch": 1.8287099380554808, "percentage": 91.44, "elapsed_time": "2 days, 10:23:05", "remaining_time": "5:28:07", "throughput": 5651.03, "total_tokens": 1187764808}
+{"current_steps": 27165, "total_steps": 29704, "loss": 0.2969427824020386, "lr": 1.793345385463191e-07, "epoch": 1.8290465930514408, "percentage": 91.45, "elapsed_time": "2 days, 10:23:45", "remaining_time": "5:27:28", "throughput": 5651.01, "total_tokens": 1187985984}
+{"current_steps": 27170, "total_steps": 29704, "loss": 0.306946849822998, "lr": 1.7863342230794667e-07, "epoch": 1.829383248047401, "percentage": 91.47, "elapsed_time": "2 days, 10:24:16", "remaining_time": "5:26:49", "throughput": 5651.23, "total_tokens": 1188205016}
+{"current_steps": 27175, "total_steps": 29704, "loss": 0.2987263917922974, "lr": 1.7793365434936316e-07, "epoch": 1.8297199030433613, "percentage": 91.49, "elapsed_time": "2 days, 10:24:56", "remaining_time": "5:26:10", "throughput": 5651.25, "total_tokens": 1188436064}
+{"current_steps": 27180, "total_steps": 29704, "loss": 0.29759798049926756, "lr": 1.7723523486625704e-07, "epoch": 1.8300565580393213, "percentage": 91.5, "elapsed_time": "2 days, 10:25:37", "remaining_time": "5:25:32", "throughput": 5651.14, "total_tokens": 1188648440}
+{"current_steps": 27185, "total_steps": 29704, "loss": 0.3033345937728882, "lr": 1.7653816405393766e-07, "epoch": 1.8303932130352814, "percentage": 91.52, "elapsed_time": "2 days, 10:26:17", "remaining_time": "5:24:53", "throughput": 5651.11, "total_tokens": 1188864712}
+{"current_steps": 27190, "total_steps": 29704, "loss": 0.308400821685791, "lr": 1.758424421073418e-07, "epoch": 1.8307298680312416, "percentage": 91.54, "elapsed_time": "2 days, 10:26:54", "remaining_time": "5:24:15", "throughput": 5651.19, "total_tokens": 1189093376}
+{"current_steps": 27195, "total_steps": 29704, "loss": 0.2787260293960571, "lr": 1.7514806922102222e-07, "epoch": 1.8310665230272019, "percentage": 91.55, "elapsed_time": "2 days, 10:27:32", "remaining_time": "5:23:36", "throughput": 5651.12, "total_tokens": 1189290992}
+{"current_steps": 27200, "total_steps": 29704, "loss": 0.2868196964263916, "lr": 1.7445504558916027e-07, "epoch": 1.8314031780231619, "percentage": 91.57, "elapsed_time": "2 days, 10:28:02", "remaining_time": "5:22:56", "throughput": 5651.32, "total_tokens": 1189504064}
+{"current_steps": 27205, "total_steps": 29704, "loss": 0.283552360534668, "lr": 1.7376337140555754e-07, "epoch": 1.831739833019122, "percentage": 91.59, "elapsed_time": "2 days, 10:28:38", "remaining_time": "5:22:17", "throughput": 5651.33, "total_tokens": 1189712800}
+{"current_steps": 27210, "total_steps": 29704, "loss": 0.30210127830505373, "lr": 1.7307304686363825e-07, "epoch": 1.8320764880150822, "percentage": 91.6, "elapsed_time": "2 days, 10:29:17", "remaining_time": "5:21:39", "throughput": 5651.35, "total_tokens": 1189932032}
+{"current_steps": 27215, "total_steps": 29704, "loss": 0.2896436214447021, "lr": 1.723840721564496e-07, "epoch": 1.8324131430110424, "percentage": 91.62, "elapsed_time": "2 days, 10:29:56", "remaining_time": "5:21:00", "throughput": 5651.39, "total_tokens": 1190162824}
+{"current_steps": 27220, "total_steps": 29704, "loss": 0.29418466091156004, "lr": 1.7169644747666081e-07, "epoch": 1.8327497980070024, "percentage": 91.64, "elapsed_time": "2 days, 10:30:35", "remaining_time": "5:20:21", "throughput": 5651.42, "total_tokens": 1190392976}
+{"current_steps": 27225, "total_steps": 29704, "loss": 0.30214219093322753, "lr": 1.710101730165642e-07, "epoch": 1.8330864530029625, "percentage": 91.65, "elapsed_time": "2 days, 10:31:13", "remaining_time": "5:19:43", "throughput": 5651.48, "total_tokens": 1190618688}
+{"current_steps": 27230, "total_steps": 29704, "loss": 0.27968459129333495, "lr": 1.7032524896807402e-07, "epoch": 1.8334231079989227, "percentage": 91.67, "elapsed_time": "2 days, 10:31:52", "remaining_time": "5:19:04", "throughput": 5651.41, "total_tokens": 1190825912}
+{"current_steps": 27235, "total_steps": 29704, "loss": 0.3041102647781372, "lr": 1.6964167552272702e-07, "epoch": 1.833759762994883, "percentage": 91.69, "elapsed_time": "2 days, 10:32:30", "remaining_time": "5:18:25", "throughput": 5651.39, "total_tokens": 1191031648}
+{"current_steps": 27240, "total_steps": 29704, "loss": 0.30144453048706055, "lr": 1.6895945287168314e-07, "epoch": 1.834096417990843, "percentage": 91.7, "elapsed_time": "2 days, 10:33:07", "remaining_time": "5:17:46", "throughput": 5651.41, "total_tokens": 1191243504}
+{"current_steps": 27245, "total_steps": 29704, "loss": 0.31502351760864256, "lr": 1.682785812057225e-07, "epoch": 1.834433072986803, "percentage": 91.72, "elapsed_time": "2 days, 10:33:38", "remaining_time": "5:17:07", "throughput": 5651.53, "total_tokens": 1191444928}
+{"current_steps": 27250, "total_steps": 29704, "loss": 0.2920320987701416, "lr": 1.6759906071525e-07, "epoch": 1.8347697279827633, "percentage": 91.74, "elapsed_time": "2 days, 10:34:22", "remaining_time": "5:16:29", "throughput": 5651.39, "total_tokens": 1191662640}
+{"current_steps": 27255, "total_steps": 29704, "loss": 0.33261871337890625, "lr": 1.669208915902909e-07, "epoch": 1.8351063829787235, "percentage": 91.76, "elapsed_time": "2 days, 10:34:57", "remaining_time": "5:15:50", "throughput": 5651.55, "total_tokens": 1191900592}
+{"current_steps": 27260, "total_steps": 29704, "loss": 0.2747644901275635, "lr": 1.6624407402049237e-07, "epoch": 1.8354430379746836, "percentage": 91.77, "elapsed_time": "2 days, 10:35:31", "remaining_time": "5:15:11", "throughput": 5651.64, "total_tokens": 1192112584}
+{"current_steps": 27265, "total_steps": 29704, "loss": 0.28942217826843264, "lr": 1.655686081951252e-07, "epoch": 1.8357796929706436, "percentage": 91.79, "elapsed_time": "2 days, 10:36:02", "remaining_time": "5:14:31", "throughput": 5651.84, "total_tokens": 1192328048}
+{"current_steps": 27270, "total_steps": 29704, "loss": 0.32457714080810546, "lr": 1.6489449430308058e-07, "epoch": 1.8361163479666038, "percentage": 91.81, "elapsed_time": "2 days, 10:36:46", "remaining_time": "5:13:53", "throughput": 5651.75, "total_tokens": 1192555872}
+{"current_steps": 27275, "total_steps": 29704, "loss": 0.31779980659484863, "lr": 1.6422173253287323e-07, "epoch": 1.836453002962564, "percentage": 91.82, "elapsed_time": "2 days, 10:37:24", "remaining_time": "5:13:14", "throughput": 5651.82, "total_tokens": 1192784792}
+{"current_steps": 27280, "total_steps": 29704, "loss": 0.2909559726715088, "lr": 1.6355032307263764e-07, "epoch": 1.8367896579585241, "percentage": 91.84, "elapsed_time": "2 days, 10:38:07", "remaining_time": "5:12:36", "throughput": 5651.73, "total_tokens": 1193008024}
+{"current_steps": 27285, "total_steps": 29704, "loss": 0.27639720439910886, "lr": 1.628802661101314e-07, "epoch": 1.8371263129544841, "percentage": 91.86, "elapsed_time": "2 days, 10:38:46", "remaining_time": "5:11:57", "throughput": 5651.7, "total_tokens": 1193222040}
+{"current_steps": 27290, "total_steps": 29704, "loss": 0.2929367542266846, "lr": 1.6221156183273524e-07, "epoch": 1.8374629679504444, "percentage": 91.87, "elapsed_time": "2 days, 10:39:26", "remaining_time": "5:11:19", "throughput": 5651.69, "total_tokens": 1193445752}
+{"current_steps": 27295, "total_steps": 29704, "loss": 0.2990311622619629, "lr": 1.615442104274484e-07, "epoch": 1.8377996229464046, "percentage": 91.89, "elapsed_time": "2 days, 10:40:03", "remaining_time": "5:10:40", "throughput": 5651.79, "total_tokens": 1193676024}
+{"current_steps": 27300, "total_steps": 29704, "loss": 0.2992110252380371, "lr": 1.6087821208089492e-07, "epoch": 1.8381362779423647, "percentage": 91.91, "elapsed_time": "2 days, 10:40:39", "remaining_time": "5:10:01", "throughput": 5651.86, "total_tokens": 1193898896}
+{"current_steps": 27305, "total_steps": 29704, "loss": 0.2737618684768677, "lr": 1.6021356697931866e-07, "epoch": 1.8384729329383247, "percentage": 91.92, "elapsed_time": "2 days, 10:41:14", "remaining_time": "5:09:22", "throughput": 5652.0, "total_tokens": 1194124640}
+{"current_steps": 27310, "total_steps": 29704, "loss": 0.27455065250396726, "lr": 1.5955027530858537e-07, "epoch": 1.838809587934285, "percentage": 91.94, "elapsed_time": "2 days, 10:41:51", "remaining_time": "5:08:43", "throughput": 5652.07, "total_tokens": 1194345288}
+{"current_steps": 27315, "total_steps": 29704, "loss": 0.2740659236907959, "lr": 1.5888833725418275e-07, "epoch": 1.8391462429302452, "percentage": 91.96, "elapsed_time": "2 days, 10:42:27", "remaining_time": "5:08:04", "throughput": 5652.13, "total_tokens": 1194562672}
+{"current_steps": 27320, "total_steps": 29704, "loss": 0.26207771301269533, "lr": 1.582277530012194e-07, "epoch": 1.8394828979262052, "percentage": 91.97, "elapsed_time": "2 days, 10:43:00", "remaining_time": "5:07:25", "throughput": 5652.35, "total_tokens": 1194797416}
+{"current_steps": 27325, "total_steps": 29704, "loss": 0.29030356407165525, "lr": 1.5756852273442536e-07, "epoch": 1.8398195529221653, "percentage": 91.99, "elapsed_time": "2 days, 10:43:42", "remaining_time": "5:06:47", "throughput": 5652.22, "total_tokens": 1195005200}
+{"current_steps": 27330, "total_steps": 29704, "loss": 0.3182931900024414, "lr": 1.569106466381526e-07, "epoch": 1.8401562079181255, "percentage": 92.01, "elapsed_time": "2 days, 10:44:22", "remaining_time": "5:06:08", "throughput": 5652.19, "total_tokens": 1195227200}
+{"current_steps": 27335, "total_steps": 29704, "loss": 0.27002391815185545, "lr": 1.5625412489637337e-07, "epoch": 1.8404928629140858, "percentage": 92.02, "elapsed_time": "2 days, 10:45:07", "remaining_time": "5:05:30", "throughput": 5652.08, "total_tokens": 1195454536}
+{"current_steps": 27340, "total_steps": 29704, "loss": 0.2888377904891968, "lr": 1.555989576926842e-07, "epoch": 1.8408295179100458, "percentage": 92.04, "elapsed_time": "2 days, 10:45:49", "remaining_time": "5:04:51", "throughput": 5652.03, "total_tokens": 1195681944}
+{"current_steps": 27345, "total_steps": 29704, "loss": 0.29750452041625974, "lr": 1.5494514521029736e-07, "epoch": 1.8411661729060058, "percentage": 92.06, "elapsed_time": "2 days, 10:46:28", "remaining_time": "5:04:13", "throughput": 5652.03, "total_tokens": 1195905376}
+{"current_steps": 27350, "total_steps": 29704, "loss": 0.2998412370681763, "lr": 1.5429268763205108e-07, "epoch": 1.841502827901966, "percentage": 92.08, "elapsed_time": "2 days, 10:47:15", "remaining_time": "5:03:35", "throughput": 5651.82, "total_tokens": 1196127744}
+{"current_steps": 27355, "total_steps": 29704, "loss": 0.3214261531829834, "lr": 1.5364158514040328e-07, "epoch": 1.8418394828979263, "percentage": 92.09, "elapsed_time": "2 days, 10:47:56", "remaining_time": "5:02:56", "throughput": 5651.77, "total_tokens": 1196344424}
+{"current_steps": 27360, "total_steps": 29704, "loss": 0.30311391353607176, "lr": 1.5299183791743223e-07, "epoch": 1.8421761378938863, "percentage": 92.11, "elapsed_time": "2 days, 10:48:32", "remaining_time": "5:02:17", "throughput": 5651.78, "total_tokens": 1196551984}
+{"current_steps": 27365, "total_steps": 29704, "loss": 0.31781580448150637, "lr": 1.523434461448381e-07, "epoch": 1.8425127928898464, "percentage": 92.13, "elapsed_time": "2 days, 10:49:10", "remaining_time": "5:01:39", "throughput": 5651.86, "total_tokens": 1196784808}
+{"current_steps": 27370, "total_steps": 29704, "loss": 0.29927232265472414, "lr": 1.5169641000394086e-07, "epoch": 1.8428494478858066, "percentage": 92.14, "elapsed_time": "2 days, 10:49:49", "remaining_time": "5:01:00", "throughput": 5651.93, "total_tokens": 1197018080}
+{"current_steps": 27375, "total_steps": 29704, "loss": 0.2902745246887207, "lr": 1.5105072967568246e-07, "epoch": 1.8431861028817669, "percentage": 92.16, "elapsed_time": "2 days, 10:50:27", "remaining_time": "5:00:21", "throughput": 5651.91, "total_tokens": 1197229392}
+{"current_steps": 27380, "total_steps": 29704, "loss": 0.29667439460754397, "lr": 1.504064053406251e-07, "epoch": 1.843522757877727, "percentage": 92.18, "elapsed_time": "2 days, 10:51:09", "remaining_time": "4:59:43", "throughput": 5651.79, "total_tokens": 1197443992}
+{"current_steps": 27385, "total_steps": 29704, "loss": 0.3001372814178467, "lr": 1.4976343717895192e-07, "epoch": 1.843859412873687, "percentage": 92.19, "elapsed_time": "2 days, 10:51:42", "remaining_time": "4:59:04", "throughput": 5652.0, "total_tokens": 1197673640}
+{"current_steps": 27390, "total_steps": 29704, "loss": 0.30607056617736816, "lr": 1.491218253704685e-07, "epoch": 1.8441960678696472, "percentage": 92.21, "elapsed_time": "2 days, 10:52:23", "remaining_time": "4:58:25", "throughput": 5651.98, "total_tokens": 1197898760}
+{"current_steps": 27395, "total_steps": 29704, "loss": 0.30979294776916505, "lr": 1.484815700945974e-07, "epoch": 1.8445327228656074, "percentage": 92.23, "elapsed_time": "2 days, 10:53:02", "remaining_time": "4:57:47", "throughput": 5651.96, "total_tokens": 1198115856}
+{"current_steps": 27400, "total_steps": 29704, "loss": 0.2884451150894165, "lr": 1.478426715303849e-07, "epoch": 1.8448693778615675, "percentage": 92.24, "elapsed_time": "2 days, 10:53:41", "remaining_time": "4:57:08", "throughput": 5651.89, "total_tokens": 1198320608}
+{"current_steps": 27405, "total_steps": 29704, "loss": 0.29123849868774415, "lr": 1.4720512985649694e-07, "epoch": 1.8452060328575275, "percentage": 92.26, "elapsed_time": "2 days, 10:54:23", "remaining_time": "4:56:29", "throughput": 5651.76, "total_tokens": 1198532416}
+{"current_steps": 27410, "total_steps": 29704, "loss": 0.30306599140167234, "lr": 1.4656894525122034e-07, "epoch": 1.8455426878534877, "percentage": 92.28, "elapsed_time": "2 days, 10:55:05", "remaining_time": "4:55:51", "throughput": 5651.59, "total_tokens": 1198734552}
+{"current_steps": 27415, "total_steps": 29704, "loss": 0.30881667137145996, "lr": 1.459341178924617e-07, "epoch": 1.845879342849448, "percentage": 92.29, "elapsed_time": "2 days, 10:55:44", "remaining_time": "4:55:12", "throughput": 5651.59, "total_tokens": 1198955512}
+{"current_steps": 27420, "total_steps": 29704, "loss": 0.28842964172363283, "lr": 1.453006479577479e-07, "epoch": 1.846215997845408, "percentage": 92.31, "elapsed_time": "2 days, 10:56:20", "remaining_time": "4:54:34", "throughput": 5651.63, "total_tokens": 1199167896}
+{"current_steps": 27425, "total_steps": 29704, "loss": 0.3094759464263916, "lr": 1.4466853562422778e-07, "epoch": 1.846552652841368, "percentage": 92.33, "elapsed_time": "2 days, 10:56:54", "remaining_time": "4:53:54", "throughput": 5651.67, "total_tokens": 1199366136}
+{"current_steps": 27430, "total_steps": 29704, "loss": 0.294615650177002, "lr": 1.440377810686694e-07, "epoch": 1.8468893078373283, "percentage": 92.34, "elapsed_time": "2 days, 10:57:27", "remaining_time": "4:53:15", "throughput": 5651.83, "total_tokens": 1199589824}
+{"current_steps": 27435, "total_steps": 29704, "loss": 0.2963638067245483, "lr": 1.4340838446746052e-07, "epoch": 1.8472259628332885, "percentage": 92.36, "elapsed_time": "2 days, 10:58:04", "remaining_time": "4:52:36", "throughput": 5651.93, "total_tokens": 1199816200}
+{"current_steps": 27440, "total_steps": 29704, "loss": 0.2769758701324463, "lr": 1.4278034599661094e-07, "epoch": 1.8475626178292486, "percentage": 92.38, "elapsed_time": "2 days, 10:58:34", "remaining_time": "4:51:57", "throughput": 5652.14, "total_tokens": 1200031328}
+{"current_steps": 27445, "total_steps": 29704, "loss": 0.2689784288406372, "lr": 1.421536658317485e-07, "epoch": 1.8478992728252086, "percentage": 92.39, "elapsed_time": "2 days, 10:59:09", "remaining_time": "4:51:18", "throughput": 5652.13, "total_tokens": 1200227096}
+{"current_steps": 27450, "total_steps": 29704, "loss": 0.28493242263793944, "lr": 1.41528344148123e-07, "epoch": 1.8482359278211689, "percentage": 92.41, "elapsed_time": "2 days, 10:59:46", "remaining_time": "4:50:39", "throughput": 5652.25, "total_tokens": 1200462264}
+{"current_steps": 27455, "total_steps": 29704, "loss": 0.2554482936859131, "lr": 1.40904381120604e-07, "epoch": 1.848572582817129, "percentage": 92.43, "elapsed_time": "2 days, 11:00:17", "remaining_time": "4:50:00", "throughput": 5652.44, "total_tokens": 1200679968}
+{"current_steps": 27460, "total_steps": 29704, "loss": 0.2979857921600342, "lr": 1.4028177692367973e-07, "epoch": 1.8489092378130891, "percentage": 92.45, "elapsed_time": "2 days, 11:00:55", "remaining_time": "4:49:21", "throughput": 5652.43, "total_tokens": 1200888896}
+{"current_steps": 27465, "total_steps": 29704, "loss": 0.29952003955841067, "lr": 1.3966053173146032e-07, "epoch": 1.8492458928090492, "percentage": 92.46, "elapsed_time": "2 days, 11:01:29", "remaining_time": "4:48:42", "throughput": 5652.59, "total_tokens": 1201115544}
+{"current_steps": 27470, "total_steps": 29704, "loss": 0.29733192920684814, "lr": 1.3904064571767462e-07, "epoch": 1.8495825478050094, "percentage": 92.48, "elapsed_time": "2 days, 11:02:09", "remaining_time": "4:48:03", "throughput": 5652.59, "total_tokens": 1201343992}
+{"current_steps": 27475, "total_steps": 29704, "loss": 0.2762616157531738, "lr": 1.3842211905567227e-07, "epoch": 1.8499192028009697, "percentage": 92.5, "elapsed_time": "2 days, 11:02:41", "remaining_time": "4:47:24", "throughput": 5652.8, "total_tokens": 1201569224}
+{"current_steps": 27480, "total_steps": 29704, "loss": 0.2911623239517212, "lr": 1.378049519184216e-07, "epoch": 1.8502558577969297, "percentage": 92.51, "elapsed_time": "2 days, 11:03:19", "remaining_time": "4:46:46", "throughput": 5652.79, "total_tokens": 1201779840}
+{"current_steps": 27485, "total_steps": 29704, "loss": 0.2653970718383789, "lr": 1.3718914447851116e-07, "epoch": 1.8505925127928897, "percentage": 92.53, "elapsed_time": "2 days, 11:03:58", "remaining_time": "4:46:07", "throughput": 5652.81, "total_tokens": 1202007472}
+{"current_steps": 27490, "total_steps": 29704, "loss": 0.28401851654052734, "lr": 1.3657469690815106e-07, "epoch": 1.85092916778885, "percentage": 92.55, "elapsed_time": "2 days, 11:04:31", "remaining_time": "4:45:28", "throughput": 5653.07, "total_tokens": 1202246288}
+{"current_steps": 27495, "total_steps": 29704, "loss": 0.3170768737792969, "lr": 1.359616093791677e-07, "epoch": 1.8512658227848102, "percentage": 92.56, "elapsed_time": "2 days, 11:05:15", "remaining_time": "4:44:49", "throughput": 5652.99, "total_tokens": 1202478016}
+{"current_steps": 27500, "total_steps": 29704, "loss": 0.29583718776702883, "lr": 1.353498820630106e-07, "epoch": 1.8516024777807703, "percentage": 92.58, "elapsed_time": "2 days, 11:05:54", "remaining_time": "4:44:11", "throughput": 5653.0, "total_tokens": 1202698624}
+{"current_steps": 27505, "total_steps": 29704, "loss": 0.31031112670898436, "lr": 1.3473951513074625e-07, "epoch": 1.8519391327767303, "percentage": 92.6, "elapsed_time": "2 days, 11:06:35", "remaining_time": "4:43:32", "throughput": 5652.94, "total_tokens": 1202918504}
+{"current_steps": 27510, "total_steps": 29704, "loss": 0.29344775676727297, "lr": 1.341305087530631e-07, "epoch": 1.8522757877726905, "percentage": 92.61, "elapsed_time": "2 days, 11:07:06", "remaining_time": "4:42:53", "throughput": 5653.13, "total_tokens": 1203135384}
+{"current_steps": 27515, "total_steps": 29704, "loss": 0.3105172634124756, "lr": 1.3352286310026663e-07, "epoch": 1.8526124427686508, "percentage": 92.63, "elapsed_time": "2 days, 11:07:44", "remaining_time": "4:42:14", "throughput": 5653.12, "total_tokens": 1203348136}
+{"current_steps": 27520, "total_steps": 29704, "loss": 0.31962890625, "lr": 1.329165783422831e-07, "epoch": 1.8529490977646108, "percentage": 92.65, "elapsed_time": "2 days, 11:08:18", "remaining_time": "4:41:35", "throughput": 5653.2, "total_tokens": 1203560776}
+{"current_steps": 27525, "total_steps": 29704, "loss": 0.29946393966674806, "lr": 1.3231165464865914e-07, "epoch": 1.8532857527605708, "percentage": 92.66, "elapsed_time": "2 days, 11:08:49", "remaining_time": "4:40:56", "throughput": 5653.41, "total_tokens": 1203778736}
+{"current_steps": 27530, "total_steps": 29704, "loss": 0.29173548221588136, "lr": 1.3170809218855884e-07, "epoch": 1.853622407756531, "percentage": 92.68, "elapsed_time": "2 days, 11:09:26", "remaining_time": "4:40:17", "throughput": 5653.5, "total_tokens": 1204009952}
+{"current_steps": 27535, "total_steps": 29704, "loss": 0.2878772258758545, "lr": 1.3110589113076554e-07, "epoch": 1.8539590627524913, "percentage": 92.7, "elapsed_time": "2 days, 11:10:08", "remaining_time": "4:39:39", "throughput": 5653.44, "total_tokens": 1204230200}
+{"current_steps": 27540, "total_steps": 29704, "loss": 0.29323248863220214, "lr": 1.3050505164368566e-07, "epoch": 1.8542957177484514, "percentage": 92.71, "elapsed_time": "2 days, 11:10:39", "remaining_time": "4:38:59", "throughput": 5653.65, "total_tokens": 1204451192}
+{"current_steps": 27545, "total_steps": 29704, "loss": 0.2807599067687988, "lr": 1.299055738953392e-07, "epoch": 1.8546323727444114, "percentage": 92.73, "elapsed_time": "2 days, 11:11:20", "remaining_time": "4:38:21", "throughput": 5653.49, "total_tokens": 1204650768}
+{"current_steps": 27550, "total_steps": 29704, "loss": 0.28776376247406005, "lr": 1.293074580533704e-07, "epoch": 1.8549690277403716, "percentage": 92.75, "elapsed_time": "2 days, 11:12:07", "remaining_time": "4:37:43", "throughput": 5653.26, "total_tokens": 1204864152}
+{"current_steps": 27555, "total_steps": 29704, "loss": 0.3120842456817627, "lr": 1.2871070428503885e-07, "epoch": 1.855305682736332, "percentage": 92.77, "elapsed_time": "2 days, 11:12:46", "remaining_time": "4:37:04", "throughput": 5653.19, "total_tokens": 1205072184}
+{"current_steps": 27560, "total_steps": 29704, "loss": 0.3159989595413208, "lr": 1.2811531275722543e-07, "epoch": 1.855642337732292, "percentage": 92.78, "elapsed_time": "2 days, 11:13:22", "remaining_time": "4:36:25", "throughput": 5653.33, "total_tokens": 1205302840}
+{"current_steps": 27565, "total_steps": 29704, "loss": 0.30814902782440184, "lr": 1.275212836364298e-07, "epoch": 1.855978992728252, "percentage": 92.8, "elapsed_time": "2 days, 11:14:06", "remaining_time": "4:35:47", "throughput": 5653.19, "total_tokens": 1205521608}
+{"current_steps": 27570, "total_steps": 29704, "loss": 0.28333234786987305, "lr": 1.2692861708877014e-07, "epoch": 1.8563156477242122, "percentage": 92.82, "elapsed_time": "2 days, 11:14:49", "remaining_time": "4:35:09", "throughput": 5653.04, "total_tokens": 1205730248}
+{"current_steps": 27575, "total_steps": 29704, "loss": 0.2803944110870361, "lr": 1.2633731327998332e-07, "epoch": 1.8566523027201725, "percentage": 92.83, "elapsed_time": "2 days, 11:15:22", "remaining_time": "4:34:30", "throughput": 5653.17, "total_tokens": 1205950784}
+{"current_steps": 27580, "total_steps": 29704, "loss": 0.2885521411895752, "lr": 1.25747372375426e-07, "epoch": 1.8569889577161325, "percentage": 92.85, "elapsed_time": "2 days, 11:16:08", "remaining_time": "4:33:51", "throughput": 5652.95, "total_tokens": 1206159224}
+{"current_steps": 27585, "total_steps": 29704, "loss": 0.3000046253204346, "lr": 1.251587945400723e-07, "epoch": 1.8573256127120925, "percentage": 92.87, "elapsed_time": "2 days, 11:16:48", "remaining_time": "4:33:13", "throughput": 5652.9, "total_tokens": 1206379280}
+{"current_steps": 27590, "total_steps": 29704, "loss": 0.2905135154724121, "lr": 1.2457157993851832e-07, "epoch": 1.8576622677080528, "percentage": 92.88, "elapsed_time": "2 days, 11:17:26", "remaining_time": "4:32:34", "throughput": 5652.96, "total_tokens": 1206604064}
+{"current_steps": 27595, "total_steps": 29704, "loss": 0.2975008487701416, "lr": 1.2398572873497438e-07, "epoch": 1.857998922704013, "percentage": 92.9, "elapsed_time": "2 days, 11:18:07", "remaining_time": "4:31:56", "throughput": 5652.9, "total_tokens": 1206821040}
+{"current_steps": 27600, "total_steps": 29704, "loss": 0.2905877590179443, "lr": 1.2340124109327334e-07, "epoch": 1.858335577699973, "percentage": 92.92, "elapsed_time": "2 days, 11:18:46", "remaining_time": "4:31:17", "throughput": 5652.84, "total_tokens": 1207029960}
+{"current_steps": 27605, "total_steps": 29704, "loss": 0.3129683494567871, "lr": 1.22818117176865e-07, "epoch": 1.858672232695933, "percentage": 92.93, "elapsed_time": "2 days, 11:19:31", "remaining_time": "4:30:39", "throughput": 5652.68, "total_tokens": 1207251408}
+{"current_steps": 27610, "total_steps": 29704, "loss": 0.30173707008361816, "lr": 1.2223635714881777e-07, "epoch": 1.8590088876918933, "percentage": 92.95, "elapsed_time": "2 days, 11:20:06", "remaining_time": "4:30:00", "throughput": 5652.74, "total_tokens": 1207460752}
+{"current_steps": 27615, "total_steps": 29704, "loss": 0.28788735866546633, "lr": 1.2165596117181931e-07, "epoch": 1.8593455426878536, "percentage": 92.97, "elapsed_time": "2 days, 11:20:40", "remaining_time": "4:29:21", "throughput": 5652.89, "total_tokens": 1207688200}
+{"current_steps": 27620, "total_steps": 29704, "loss": 0.30665135383605957, "lr": 1.2107692940817594e-07, "epoch": 1.8596821976838136, "percentage": 92.98, "elapsed_time": "2 days, 11:21:22", "remaining_time": "4:28:42", "throughput": 5652.8, "total_tokens": 1207904320}
+{"current_steps": 27625, "total_steps": 29704, "loss": 0.2934282302856445, "lr": 1.2049926201981088e-07, "epoch": 1.8600188526797736, "percentage": 93.0, "elapsed_time": "2 days, 11:22:03", "remaining_time": "4:28:04", "throughput": 5652.75, "total_tokens": 1208124344}
+{"current_steps": 27630, "total_steps": 29704, "loss": 0.3092188835144043, "lr": 1.199229591682677e-07, "epoch": 1.8603555076757339, "percentage": 93.02, "elapsed_time": "2 days, 11:22:37", "remaining_time": "4:27:25", "throughput": 5652.89, "total_tokens": 1208348632}
+{"current_steps": 27635, "total_steps": 29704, "loss": 0.3101773262023926, "lr": 1.193480210147069e-07, "epoch": 1.8606921626716941, "percentage": 93.03, "elapsed_time": "2 days, 11:23:13", "remaining_time": "4:26:46", "throughput": 5653.02, "total_tokens": 1208576560}
+{"current_steps": 27640, "total_steps": 29704, "loss": 0.2830918788909912, "lr": 1.1877444771991042e-07, "epoch": 1.8610288176676542, "percentage": 93.05, "elapsed_time": "2 days, 11:23:50", "remaining_time": "4:26:07", "throughput": 5653.09, "total_tokens": 1208805448}
+{"current_steps": 27645, "total_steps": 29704, "loss": 0.28035907745361327, "lr": 1.1820223944427279e-07, "epoch": 1.8613654726636142, "percentage": 93.07, "elapsed_time": "2 days, 11:24:33", "remaining_time": "4:25:29", "throughput": 5652.95, "total_tokens": 1209015368}
+{"current_steps": 27650, "total_steps": 29704, "loss": 0.29651951789855957, "lr": 1.1763139634781262e-07, "epoch": 1.8617021276595744, "percentage": 93.09, "elapsed_time": "2 days, 11:25:11", "remaining_time": "4:24:50", "throughput": 5653.01, "total_tokens": 1209240952}
+{"current_steps": 27655, "total_steps": 29704, "loss": 0.3263824224472046, "lr": 1.1706191859016392e-07, "epoch": 1.8620387826555347, "percentage": 93.1, "elapsed_time": "2 days, 11:25:53", "remaining_time": "4:24:12", "throughput": 5652.9, "total_tokens": 1209461064}
+{"current_steps": 27660, "total_steps": 29704, "loss": 0.3037287950515747, "lr": 1.1649380633057927e-07, "epoch": 1.8623754376514947, "percentage": 93.12, "elapsed_time": "2 days, 11:26:37", "remaining_time": "4:23:33", "throughput": 5652.76, "total_tokens": 1209677248}
+{"current_steps": 27665, "total_steps": 29704, "loss": 0.27420315742492674, "lr": 1.1592705972792939e-07, "epoch": 1.8627120926474547, "percentage": 93.14, "elapsed_time": "2 days, 11:27:19", "remaining_time": "4:22:55", "throughput": 5652.66, "total_tokens": 1209892576}
+{"current_steps": 27670, "total_steps": 29704, "loss": 0.3063500165939331, "lr": 1.1536167894070249e-07, "epoch": 1.863048747643415, "percentage": 93.15, "elapsed_time": "2 days, 11:28:02", "remaining_time": "4:22:17", "throughput": 5652.58, "total_tokens": 1210118104}
+{"current_steps": 27675, "total_steps": 29704, "loss": 0.33033363819122313, "lr": 1.1479766412700655e-07, "epoch": 1.8633854026393752, "percentage": 93.17, "elapsed_time": "2 days, 11:28:41", "remaining_time": "4:21:38", "throughput": 5652.62, "total_tokens": 1210345472}
+{"current_steps": 27680, "total_steps": 29704, "loss": 0.31739757061004636, "lr": 1.1423501544456595e-07, "epoch": 1.8637220576353353, "percentage": 93.19, "elapsed_time": "2 days, 11:29:16", "remaining_time": "4:20:59", "throughput": 5652.69, "total_tokens": 1210558512}
+{"current_steps": 27685, "total_steps": 29704, "loss": 0.2853879451751709, "lr": 1.1367373305072315e-07, "epoch": 1.8640587126312953, "percentage": 93.2, "elapsed_time": "2 days, 11:29:52", "remaining_time": "4:20:20", "throughput": 5652.76, "total_tokens": 1210779944}
+{"current_steps": 27690, "total_steps": 29704, "loss": 0.2670130729675293, "lr": 1.1311381710244096e-07, "epoch": 1.8643953676272556, "percentage": 93.22, "elapsed_time": "2 days, 11:30:33", "remaining_time": "4:19:42", "throughput": 5652.73, "total_tokens": 1211003328}
+{"current_steps": 27695, "total_steps": 29704, "loss": 0.27195014953613283, "lr": 1.1255526775629521e-07, "epoch": 1.8647320226232158, "percentage": 93.24, "elapsed_time": "2 days, 11:31:18", "remaining_time": "4:19:03", "throughput": 5652.53, "total_tokens": 1211215040}
+{"current_steps": 27700, "total_steps": 29704, "loss": 0.30287680625915525, "lr": 1.1199808516848487e-07, "epoch": 1.8650686776191758, "percentage": 93.25, "elapsed_time": "2 days, 11:32:04", "remaining_time": "4:18:25", "throughput": 5652.33, "total_tokens": 1211429152}
+{"current_steps": 27705, "total_steps": 29704, "loss": 0.28324642181396487, "lr": 1.1144226949482306e-07, "epoch": 1.8654053326151359, "percentage": 93.27, "elapsed_time": "2 days, 11:32:40", "remaining_time": "4:17:46", "throughput": 5652.35, "total_tokens": 1211638376}
+{"current_steps": 27710, "total_steps": 29704, "loss": 0.27295451164245604, "lr": 1.1088782089074157e-07, "epoch": 1.865741987611096, "percentage": 93.29, "elapsed_time": "2 days, 11:33:20", "remaining_time": "4:17:08", "throughput": 5652.3, "total_tokens": 1211854440}
+{"current_steps": 27715, "total_steps": 29704, "loss": 0.3024445056915283, "lr": 1.1033473951129137e-07, "epoch": 1.8660786426070564, "percentage": 93.3, "elapsed_time": "2 days, 11:33:54", "remaining_time": "4:16:29", "throughput": 5652.43, "total_tokens": 1212077896}
+{"current_steps": 27720, "total_steps": 29704, "loss": 0.3134200096130371, "lr": 1.0978302551113929e-07, "epoch": 1.8664152976030164, "percentage": 93.32, "elapsed_time": "2 days, 11:34:36", "remaining_time": "4:15:50", "throughput": 5652.41, "total_tokens": 1212305392}
+{"current_steps": 27725, "total_steps": 29704, "loss": 0.30133209228515623, "lr": 1.0923267904457025e-07, "epoch": 1.8667519525989764, "percentage": 93.34, "elapsed_time": "2 days, 11:35:16", "remaining_time": "4:15:12", "throughput": 5652.39, "total_tokens": 1212533032}
+{"current_steps": 27730, "total_steps": 29704, "loss": 0.30214452743530273, "lr": 1.086837002654867e-07, "epoch": 1.8670886075949367, "percentage": 93.35, "elapsed_time": "2 days, 11:35:58", "remaining_time": "4:14:33", "throughput": 5652.27, "total_tokens": 1212743440}
+{"current_steps": 27735, "total_steps": 29704, "loss": 0.3034868001937866, "lr": 1.0813608932740915e-07, "epoch": 1.867425262590897, "percentage": 93.37, "elapsed_time": "2 days, 11:36:34", "remaining_time": "4:13:54", "throughput": 5652.41, "total_tokens": 1212977168}
+{"current_steps": 27740, "total_steps": 29704, "loss": 0.2956989765167236, "lr": 1.0758984638347569e-07, "epoch": 1.867761917586857, "percentage": 93.39, "elapsed_time": "2 days, 11:37:09", "remaining_time": "4:13:15", "throughput": 5652.57, "total_tokens": 1213207904}
+{"current_steps": 27745, "total_steps": 29704, "loss": 0.32259302139282225, "lr": 1.0704497158644022e-07, "epoch": 1.868098572582817, "percentage": 93.4, "elapsed_time": "2 days, 11:37:53", "remaining_time": "4:12:37", "throughput": 5652.41, "total_tokens": 1213425128}
+{"current_steps": 27750, "total_steps": 29704, "loss": 0.2915182590484619, "lr": 1.0650146508867642e-07, "epoch": 1.8684352275787772, "percentage": 93.42, "elapsed_time": "2 days, 11:38:32", "remaining_time": "4:11:58", "throughput": 5652.43, "total_tokens": 1213645624}
+{"current_steps": 27755, "total_steps": 29704, "loss": 0.2860418796539307, "lr": 1.0595932704217327e-07, "epoch": 1.8687718825747375, "percentage": 93.44, "elapsed_time": "2 days, 11:39:07", "remaining_time": "4:11:19", "throughput": 5652.54, "total_tokens": 1213868720}
+{"current_steps": 27760, "total_steps": 29704, "loss": 0.2798731327056885, "lr": 1.0541855759853837e-07, "epoch": 1.8691085375706975, "percentage": 93.46, "elapsed_time": "2 days, 11:39:36", "remaining_time": "4:10:40", "throughput": 5652.85, "total_tokens": 1214099608}
+{"current_steps": 27765, "total_steps": 29704, "loss": 0.29963858127593995, "lr": 1.0487915690899685e-07, "epoch": 1.8694451925666575, "percentage": 93.47, "elapsed_time": "2 days, 11:40:14", "remaining_time": "4:10:01", "throughput": 5652.88, "total_tokens": 1214319216}
+{"current_steps": 27770, "total_steps": 29704, "loss": 0.29176318645477295, "lr": 1.0434112512438921e-07, "epoch": 1.8697818475626178, "percentage": 93.49, "elapsed_time": "2 days, 11:40:52", "remaining_time": "4:09:23", "throughput": 5652.87, "total_tokens": 1214533368}
+{"current_steps": 27775, "total_steps": 29704, "loss": 0.2727836608886719, "lr": 1.0380446239517506e-07, "epoch": 1.870118502558578, "percentage": 93.51, "elapsed_time": "2 days, 11:41:33", "remaining_time": "4:08:44", "throughput": 5652.7, "total_tokens": 1214730928}
+{"current_steps": 27780, "total_steps": 29704, "loss": 0.2895357608795166, "lr": 1.032691688714299e-07, "epoch": 1.870455157554538, "percentage": 93.52, "elapsed_time": "2 days, 11:42:19", "remaining_time": "4:08:06", "throughput": 5652.38, "total_tokens": 1214918680}
+{"current_steps": 27785, "total_steps": 29704, "loss": 0.3172266483306885, "lr": 1.0273524470284734e-07, "epoch": 1.870791812550498, "percentage": 93.54, "elapsed_time": "2 days, 11:42:56", "remaining_time": "4:07:27", "throughput": 5652.47, "total_tokens": 1215146136}
+{"current_steps": 27790, "total_steps": 29704, "loss": 0.29100470542907714, "lr": 1.0220269003873739e-07, "epoch": 1.8711284675464583, "percentage": 93.56, "elapsed_time": "2 days, 11:43:31", "remaining_time": "4:06:48", "throughput": 5652.56, "total_tokens": 1215365024}
+{"current_steps": 27795, "total_steps": 29704, "loss": 0.2920061111450195, "lr": 1.0167150502802703e-07, "epoch": 1.8714651225424186, "percentage": 93.57, "elapsed_time": "2 days, 11:44:09", "remaining_time": "4:06:09", "throughput": 5652.59, "total_tokens": 1215587488}
+{"current_steps": 27800, "total_steps": 29704, "loss": 0.31276350021362304, "lr": 1.0114168981926075e-07, "epoch": 1.8718017775383786, "percentage": 93.59, "elapsed_time": "2 days, 11:44:51", "remaining_time": "4:05:31", "throughput": 5652.53, "total_tokens": 1215810000}
+{"current_steps": 27805, "total_steps": 29704, "loss": 0.3080112457275391, "lr": 1.006132445605995e-07, "epoch": 1.8721384325343386, "percentage": 93.61, "elapsed_time": "2 days, 11:45:24", "remaining_time": "4:04:52", "throughput": 5652.62, "total_tokens": 1216014256}
+{"current_steps": 27810, "total_steps": 29704, "loss": 0.28853521347045896, "lr": 1.0008616939982118e-07, "epoch": 1.872475087530299, "percentage": 93.62, "elapsed_time": "2 days, 11:46:02", "remaining_time": "4:04:13", "throughput": 5652.63, "total_tokens": 1216231488}
+{"current_steps": 27815, "total_steps": 29704, "loss": 0.283016300201416, "lr": 9.956046448432177e-08, "epoch": 1.8728117425262591, "percentage": 93.64, "elapsed_time": "2 days, 11:46:38", "remaining_time": "4:03:34", "throughput": 5652.68, "total_tokens": 1216451408}
+{"current_steps": 27820, "total_steps": 29704, "loss": 0.30707416534423826, "lr": 9.90361299611109e-08, "epoch": 1.8731483975222192, "percentage": 93.66, "elapsed_time": "2 days, 11:47:11", "remaining_time": "4:02:55", "throughput": 5652.89, "total_tokens": 1216680872}
+{"current_steps": 27825, "total_steps": 29704, "loss": 0.3057239055633545, "lr": 9.851316597681959e-08, "epoch": 1.8734850525181792, "percentage": 93.67, "elapsed_time": "2 days, 11:47:48", "remaining_time": "4:02:16", "throughput": 5652.89, "total_tokens": 1216892024}
+{"current_steps": 27830, "total_steps": 29704, "loss": 0.3025531768798828, "lr": 9.799157267769088e-08, "epoch": 1.8738217075141395, "percentage": 93.69, "elapsed_time": "2 days, 11:48:20", "remaining_time": "4:01:37", "throughput": 5653.1, "total_tokens": 1217114336}
+{"current_steps": 27835, "total_steps": 29704, "loss": 0.2927044153213501, "lr": 9.747135020958753e-08, "epoch": 1.8741583625100997, "percentage": 93.71, "elapsed_time": "2 days, 11:49:00", "remaining_time": "4:00:59", "throughput": 5653.06, "total_tokens": 1217335976}
+{"current_steps": 27840, "total_steps": 29704, "loss": 0.2803203582763672, "lr": 9.695249871798818e-08, "epoch": 1.8744950175060597, "percentage": 93.72, "elapsed_time": "2 days, 11:49:48", "remaining_time": "4:00:21", "throughput": 5652.81, "total_tokens": 1217552464}
+{"current_steps": 27845, "total_steps": 29704, "loss": 0.30305957794189453, "lr": 9.643501834798786e-08, "epoch": 1.8748316725020198, "percentage": 93.74, "elapsed_time": "2 days, 11:50:29", "remaining_time": "3:59:42", "throughput": 5652.74, "total_tokens": 1217765392}
+{"current_steps": 27850, "total_steps": 29704, "loss": 0.2971980094909668, "lr": 9.591890924429858e-08, "epoch": 1.8751683274979802, "percentage": 93.76, "elapsed_time": "2 days, 11:51:10", "remaining_time": "3:59:04", "throughput": 5652.76, "total_tokens": 1218000984}
+{"current_steps": 27855, "total_steps": 29704, "loss": 0.278977370262146, "lr": 9.540417155124826e-08, "epoch": 1.8755049824939403, "percentage": 93.78, "elapsed_time": "2 days, 11:51:49", "remaining_time": "3:58:25", "throughput": 5652.66, "total_tokens": 1218198848}
+{"current_steps": 27860, "total_steps": 29704, "loss": 0.2986736536026001, "lr": 9.489080541278117e-08, "epoch": 1.8758416374899003, "percentage": 93.79, "elapsed_time": "2 days, 11:52:34", "remaining_time": "3:57:47", "throughput": 5652.43, "total_tokens": 1218407032}
+{"current_steps": 27865, "total_steps": 29704, "loss": 0.3069639205932617, "lr": 9.437881097246026e-08, "epoch": 1.8761782924858605, "percentage": 93.81, "elapsed_time": "2 days, 11:53:13", "remaining_time": "3:57:08", "throughput": 5652.45, "total_tokens": 1218632248}
+{"current_steps": 27870, "total_steps": 29704, "loss": 0.2972809076309204, "lr": 9.3868188373461e-08, "epoch": 1.8765149474818208, "percentage": 93.83, "elapsed_time": "2 days, 11:53:48", "remaining_time": "3:56:29", "throughput": 5652.52, "total_tokens": 1218846656}
+{"current_steps": 27875, "total_steps": 29704, "loss": 0.3106103420257568, "lr": 9.335893775857918e-08, "epoch": 1.8768516024777808, "percentage": 93.84, "elapsed_time": "2 days, 11:54:23", "remaining_time": "3:55:50", "throughput": 5652.58, "total_tokens": 1219056688}
+{"current_steps": 27880, "total_steps": 29704, "loss": 0.28227670192718507, "lr": 9.285105927022309e-08, "epoch": 1.8771882574737409, "percentage": 93.86, "elapsed_time": "2 days, 11:55:02", "remaining_time": "3:55:11", "throughput": 5652.61, "total_tokens": 1219278728}
+{"current_steps": 27885, "total_steps": 29704, "loss": 0.2822525978088379, "lr": 9.234455305042133e-08, "epoch": 1.877524912469701, "percentage": 93.88, "elapsed_time": "2 days, 11:55:50", "remaining_time": "3:54:33", "throughput": 5652.31, "total_tokens": 1219485440}
+{"current_steps": 27890, "total_steps": 29704, "loss": 0.2904726266860962, "lr": 9.183941924081563e-08, "epoch": 1.8778615674656614, "percentage": 93.89, "elapsed_time": "2 days, 11:56:27", "remaining_time": "3:53:55", "throughput": 5652.33, "total_tokens": 1219702416}
+{"current_steps": 27895, "total_steps": 29704, "loss": 0.2883297443389893, "lr": 9.133565798266574e-08, "epoch": 1.8781982224616214, "percentage": 93.91, "elapsed_time": "2 days, 11:57:00", "remaining_time": "3:53:16", "throughput": 5652.49, "total_tokens": 1219921880}
+{"current_steps": 27900, "total_steps": 29704, "loss": 0.3204617500305176, "lr": 9.083326941684622e-08, "epoch": 1.8785348774575814, "percentage": 93.93, "elapsed_time": "2 days, 11:57:35", "remaining_time": "3:52:37", "throughput": 5652.61, "total_tokens": 1220148280}
+{"current_steps": 27905, "total_steps": 29704, "loss": 0.3104442358016968, "lr": 9.033225368384912e-08, "epoch": 1.8788715324535417, "percentage": 93.94, "elapsed_time": "2 days, 11:58:17", "remaining_time": "3:51:58", "throughput": 5652.5, "total_tokens": 1220358200}
+{"current_steps": 27910, "total_steps": 29704, "loss": 0.27222867012023927, "lr": 8.98326109237807e-08, "epoch": 1.879208187449502, "percentage": 93.96, "elapsed_time": "2 days, 11:58:59", "remaining_time": "3:51:20", "throughput": 5652.4, "total_tokens": 1220576488}
+{"current_steps": 27915, "total_steps": 29704, "loss": 0.31929314136505127, "lr": 8.933434127636586e-08, "epoch": 1.879544842445462, "percentage": 93.98, "elapsed_time": "2 days, 11:59:43", "remaining_time": "3:50:41", "throughput": 5652.3, "total_tokens": 1220801344}
+{"current_steps": 27920, "total_steps": 29704, "loss": 0.3021604061126709, "lr": 8.883744488094314e-08, "epoch": 1.879881497441422, "percentage": 93.99, "elapsed_time": "2 days, 12:00:20", "remaining_time": "3:50:03", "throughput": 5652.31, "total_tokens": 1221012952}
+{"current_steps": 27925, "total_steps": 29704, "loss": 0.30646262168884275, "lr": 8.834192187646917e-08, "epoch": 1.8802181524373822, "percentage": 94.01, "elapsed_time": "2 days, 12:00:54", "remaining_time": "3:49:24", "throughput": 5652.48, "total_tokens": 1221243752}
+{"current_steps": 27930, "total_steps": 29704, "loss": 0.30849490165710447, "lr": 8.784777240151365e-08, "epoch": 1.8805548074333425, "percentage": 94.03, "elapsed_time": "2 days, 12:01:25", "remaining_time": "3:48:44", "throughput": 5652.72, "total_tokens": 1221469512}
+{"current_steps": 27935, "total_steps": 29704, "loss": 0.3049221754074097, "lr": 8.73549965942655e-08, "epoch": 1.8808914624293025, "percentage": 94.04, "elapsed_time": "2 days, 12:02:03", "remaining_time": "3:48:06", "throughput": 5652.76, "total_tokens": 1221695328}
+{"current_steps": 27940, "total_steps": 29704, "loss": 0.29515538215637205, "lr": 8.686359459252725e-08, "epoch": 1.8812281174252625, "percentage": 94.06, "elapsed_time": "2 days, 12:02:42", "remaining_time": "3:47:27", "throughput": 5652.73, "total_tokens": 1221906744}
+{"current_steps": 27945, "total_steps": 29704, "loss": 0.28774356842041016, "lr": 8.637356653371787e-08, "epoch": 1.8815647724212228, "percentage": 94.08, "elapsed_time": "2 days, 12:03:17", "remaining_time": "3:46:48", "throughput": 5652.85, "total_tokens": 1222134512}
+{"current_steps": 27950, "total_steps": 29704, "loss": 0.2720442056655884, "lr": 8.588491255487163e-08, "epoch": 1.881901427417183, "percentage": 94.1, "elapsed_time": "2 days, 12:03:59", "remaining_time": "3:46:10", "throughput": 5652.74, "total_tokens": 1222344728}
+{"current_steps": 27955, "total_steps": 29704, "loss": 0.2836963176727295, "lr": 8.539763279264035e-08, "epoch": 1.882238082413143, "percentage": 94.11, "elapsed_time": "2 days, 12:04:44", "remaining_time": "3:45:31", "throughput": 5652.55, "total_tokens": 1222560856}
+{"current_steps": 27960, "total_steps": 29704, "loss": 0.3163050651550293, "lr": 8.49117273832889e-08, "epoch": 1.882574737409103, "percentage": 94.13, "elapsed_time": "2 days, 12:05:18", "remaining_time": "3:44:52", "throughput": 5652.74, "total_tokens": 1222789864}
+{"current_steps": 27965, "total_steps": 29704, "loss": 0.28265206813812255, "lr": 8.442719646270082e-08, "epoch": 1.8829113924050633, "percentage": 94.15, "elapsed_time": "2 days, 12:05:57", "remaining_time": "3:44:14", "throughput": 5652.7, "total_tokens": 1223005096}
+{"current_steps": 27970, "total_steps": 29704, "loss": 0.28957691192626955, "lr": 8.394404016637159e-08, "epoch": 1.8832480474010236, "percentage": 94.16, "elapsed_time": "2 days, 12:06:39", "remaining_time": "3:43:35", "throughput": 5652.63, "total_tokens": 1223223864}
+{"current_steps": 27975, "total_steps": 29704, "loss": 0.28935163021087645, "lr": 8.346225862941592e-08, "epoch": 1.8835847023969836, "percentage": 94.18, "elapsed_time": "2 days, 12:07:14", "remaining_time": "3:42:56", "throughput": 5652.67, "total_tokens": 1223430136}
+{"current_steps": 27980, "total_steps": 29704, "loss": 0.31040134429931643, "lr": 8.298185198656161e-08, "epoch": 1.8839213573929436, "percentage": 94.2, "elapsed_time": "2 days, 12:07:55", "remaining_time": "3:42:18", "throughput": 5652.65, "total_tokens": 1223661888}
+{"current_steps": 27985, "total_steps": 29704, "loss": 0.2839097023010254, "lr": 8.25028203721534e-08, "epoch": 1.884258012388904, "percentage": 94.21, "elapsed_time": "2 days, 12:08:36", "remaining_time": "3:41:39", "throughput": 5652.57, "total_tokens": 1223872392}
+{"current_steps": 27990, "total_steps": 29704, "loss": 0.30760276317596436, "lr": 8.202516392015081e-08, "epoch": 1.8845946673848641, "percentage": 94.23, "elapsed_time": "2 days, 12:09:13", "remaining_time": "3:41:00", "throughput": 5652.63, "total_tokens": 1224095496}
+{"current_steps": 27995, "total_steps": 29704, "loss": 0.31410627365112304, "lr": 8.154888276412865e-08, "epoch": 1.8849313223808242, "percentage": 94.25, "elapsed_time": "2 days, 12:09:47", "remaining_time": "3:40:21", "throughput": 5652.68, "total_tokens": 1224301616}
+{"current_steps": 28000, "total_steps": 29704, "loss": 0.2736332416534424, "lr": 8.10739770372776e-08, "epoch": 1.8852679773767842, "percentage": 94.26, "elapsed_time": "2 days, 12:10:24", "remaining_time": "3:39:43", "throughput": 5652.75, "total_tokens": 1224525064}
+{"current_steps": 28005, "total_steps": 29704, "loss": 0.2702982425689697, "lr": 8.060044687240365e-08, "epoch": 1.8856046323727444, "percentage": 94.28, "elapsed_time": "2 days, 12:10:55", "remaining_time": "3:39:03", "throughput": 5653.06, "total_tokens": 1224764712}
+{"current_steps": 28010, "total_steps": 29704, "loss": 0.31306607723236085, "lr": 8.012829240192754e-08, "epoch": 1.8859412873687047, "percentage": 94.3, "elapsed_time": "2 days, 12:11:35", "remaining_time": "3:38:25", "throughput": 5652.99, "total_tokens": 1224979848}
+{"current_steps": 28015, "total_steps": 29704, "loss": 0.2802762985229492, "lr": 7.9657513757887e-08, "epoch": 1.8862779423646647, "percentage": 94.31, "elapsed_time": "2 days, 12:12:15", "remaining_time": "3:37:46", "throughput": 5652.97, "total_tokens": 1225200744}
+{"current_steps": 28020, "total_steps": 29704, "loss": 0.28357934951782227, "lr": 7.918811107193225e-08, "epoch": 1.8866145973606248, "percentage": 94.33, "elapsed_time": "2 days, 12:12:51", "remaining_time": "3:37:07", "throughput": 5653.1, "total_tokens": 1225430328}
+{"current_steps": 28025, "total_steps": 29704, "loss": 0.3160097122192383, "lr": 7.872008447533109e-08, "epoch": 1.886951252356585, "percentage": 94.35, "elapsed_time": "2 days, 12:13:28", "remaining_time": "3:36:29", "throughput": 5653.14, "total_tokens": 1225648792}
+{"current_steps": 28030, "total_steps": 29704, "loss": 0.3094397306442261, "lr": 7.825343409896547e-08, "epoch": 1.8872879073525453, "percentage": 94.36, "elapsed_time": "2 days, 12:14:04", "remaining_time": "3:35:50", "throughput": 5653.19, "total_tokens": 1225864272}
+{"current_steps": 28035, "total_steps": 29704, "loss": 0.3081014156341553, "lr": 7.778816007333268e-08, "epoch": 1.8876245623485053, "percentage": 94.38, "elapsed_time": "2 days, 12:14:47", "remaining_time": "3:35:11", "throughput": 5653.11, "total_tokens": 1226087184}
+{"current_steps": 28040, "total_steps": 29704, "loss": 0.29062952995300295, "lr": 7.73242625285453e-08, "epoch": 1.8879612173444653, "percentage": 94.4, "elapsed_time": "2 days, 12:15:25", "remaining_time": "3:34:33", "throughput": 5653.11, "total_tokens": 1226306464}
+{"current_steps": 28045, "total_steps": 29704, "loss": 0.3064660787582397, "lr": 7.686174159433069e-08, "epoch": 1.8882978723404256, "percentage": 94.41, "elapsed_time": "2 days, 12:16:06", "remaining_time": "3:33:54", "throughput": 5653.03, "total_tokens": 1226520296}
+{"current_steps": 28050, "total_steps": 29704, "loss": 0.2722774982452393, "lr": 7.640059740003091e-08, "epoch": 1.8886345273363858, "percentage": 94.43, "elapsed_time": "2 days, 12:16:46", "remaining_time": "3:33:16", "throughput": 5653.05, "total_tokens": 1226746648}
+{"current_steps": 28055, "total_steps": 29704, "loss": 0.29419803619384766, "lr": 7.594083007460396e-08, "epoch": 1.8889711823323458, "percentage": 94.45, "elapsed_time": "2 days, 12:17:28", "remaining_time": "3:32:37", "throughput": 5652.91, "total_tokens": 1226953312}
+{"current_steps": 28060, "total_steps": 29704, "loss": 0.2896998167037964, "lr": 7.54824397466225e-08, "epoch": 1.8893078373283059, "percentage": 94.47, "elapsed_time": "2 days, 12:18:04", "remaining_time": "3:31:58", "throughput": 5652.99, "total_tokens": 1227175648}
+{"current_steps": 28065, "total_steps": 29704, "loss": 0.28265087604522704, "lr": 7.502542654427347e-08, "epoch": 1.8896444923242661, "percentage": 94.48, "elapsed_time": "2 days, 12:18:36", "remaining_time": "3:31:19", "throughput": 5653.12, "total_tokens": 1227383304}
+{"current_steps": 28070, "total_steps": 29704, "loss": 0.29991984367370605, "lr": 7.456979059535907e-08, "epoch": 1.8899811473202264, "percentage": 94.5, "elapsed_time": "2 days, 12:19:21", "remaining_time": "3:30:41", "throughput": 5652.95, "total_tokens": 1227603976}
+{"current_steps": 28075, "total_steps": 29704, "loss": 0.2972327470779419, "lr": 7.411553202729682e-08, "epoch": 1.8903178023161864, "percentage": 94.52, "elapsed_time": "2 days, 12:20:02", "remaining_time": "3:30:02", "throughput": 5652.77, "total_tokens": 1227795992}
+{"current_steps": 28080, "total_steps": 29704, "loss": 0.2765605926513672, "lr": 7.3662650967119e-08, "epoch": 1.8906544573121464, "percentage": 94.53, "elapsed_time": "2 days, 12:20:42", "remaining_time": "3:29:24", "throughput": 5652.72, "total_tokens": 1228009920}
+{"current_steps": 28085, "total_steps": 29704, "loss": 0.28570716381072997, "lr": 7.321114754147151e-08, "epoch": 1.8909911123081067, "percentage": 94.55, "elapsed_time": "2 days, 12:21:20", "remaining_time": "3:28:45", "throughput": 5652.74, "total_tokens": 1228232816}
+{"current_steps": 28090, "total_steps": 29704, "loss": 0.31825072765350343, "lr": 7.276102187661615e-08, "epoch": 1.891327767304067, "percentage": 94.57, "elapsed_time": "2 days, 12:21:59", "remaining_time": "3:28:06", "throughput": 5652.72, "total_tokens": 1228448704}
+{"current_steps": 28095, "total_steps": 29704, "loss": 0.30093338489532473, "lr": 7.231227409842944e-08, "epoch": 1.891664422300027, "percentage": 94.58, "elapsed_time": "2 days, 12:22:35", "remaining_time": "3:27:27", "throughput": 5652.83, "total_tokens": 1228675704}
+{"current_steps": 28100, "total_steps": 29704, "loss": 0.30986011028289795, "lr": 7.186490433240156e-08, "epoch": 1.892001077295987, "percentage": 94.6, "elapsed_time": "2 days, 12:23:10", "remaining_time": "3:26:49", "throughput": 5652.85, "total_tokens": 1228877808}
+{"current_steps": 28105, "total_steps": 29704, "loss": 0.2832916259765625, "lr": 7.14189127036391e-08, "epoch": 1.8923377322919472, "percentage": 94.62, "elapsed_time": "2 days, 12:23:45", "remaining_time": "3:26:10", "throughput": 5652.95, "total_tokens": 1229095280}
+{"current_steps": 28110, "total_steps": 29704, "loss": 0.3074744701385498, "lr": 7.097429933686118e-08, "epoch": 1.8926743872879075, "percentage": 94.63, "elapsed_time": "2 days, 12:24:26", "remaining_time": "3:25:31", "throughput": 5652.87, "total_tokens": 1229309216}
+{"current_steps": 28115, "total_steps": 29704, "loss": 0.31830539703369143, "lr": 7.053106435640333e-08, "epoch": 1.8930110422838675, "percentage": 94.65, "elapsed_time": "2 days, 12:25:04", "remaining_time": "3:24:52", "throughput": 5652.96, "total_tokens": 1229541912}
+{"current_steps": 28120, "total_steps": 29704, "loss": 0.299678635597229, "lr": 7.008920788621366e-08, "epoch": 1.8933476972798275, "percentage": 94.67, "elapsed_time": "2 days, 12:25:45", "remaining_time": "3:24:14", "throughput": 5652.88, "total_tokens": 1229758840}
+{"current_steps": 28125, "total_steps": 29704, "loss": 0.3041519641876221, "lr": 6.964873004985717e-08, "epoch": 1.8936843522757878, "percentage": 94.68, "elapsed_time": "2 days, 12:26:22", "remaining_time": "3:23:35", "throughput": 5652.93, "total_tokens": 1229975280}
+{"current_steps": 28130, "total_steps": 29704, "loss": 0.3081005096435547, "lr": 6.920963097051148e-08, "epoch": 1.894021007271748, "percentage": 94.7, "elapsed_time": "2 days, 12:27:02", "remaining_time": "3:22:56", "throughput": 5652.9, "total_tokens": 1230195248}
+{"current_steps": 28135, "total_steps": 29704, "loss": 0.28580417633056643, "lr": 6.877191077096945e-08, "epoch": 1.894357662267708, "percentage": 94.72, "elapsed_time": "2 days, 12:27:43", "remaining_time": "3:22:18", "throughput": 5652.78, "total_tokens": 1230403720}
+{"current_steps": 28140, "total_steps": 29704, "loss": 0.31174564361572266, "lr": 6.833556957363818e-08, "epoch": 1.894694317263668, "percentage": 94.73, "elapsed_time": "2 days, 12:28:21", "remaining_time": "3:21:39", "throughput": 5652.78, "total_tokens": 1230619400}
+{"current_steps": 28145, "total_steps": 29704, "loss": 0.2922555446624756, "lr": 6.790060750053895e-08, "epoch": 1.8950309722596284, "percentage": 94.75, "elapsed_time": "2 days, 12:28:58", "remaining_time": "3:21:00", "throughput": 5652.8, "total_tokens": 1230833368}
+{"current_steps": 28150, "total_steps": 29704, "loss": 0.2662560701370239, "lr": 6.746702467330723e-08, "epoch": 1.8953676272555886, "percentage": 94.77, "elapsed_time": "2 days, 12:29:38", "remaining_time": "3:20:22", "throughput": 5652.84, "total_tokens": 1231064032}
+{"current_steps": 28155, "total_steps": 29704, "loss": 0.2952676296234131, "lr": 6.703482121319383e-08, "epoch": 1.8957042822515486, "percentage": 94.79, "elapsed_time": "2 days, 12:30:15", "remaining_time": "3:19:43", "throughput": 5652.87, "total_tokens": 1231282944}
+{"current_steps": 28160, "total_steps": 29704, "loss": 0.31438963413238524, "lr": 6.660399724106204e-08, "epoch": 1.8960409372475087, "percentage": 94.8, "elapsed_time": "2 days, 12:30:55", "remaining_time": "3:19:04", "throughput": 5652.9, "total_tokens": 1231513400}
+{"current_steps": 28165, "total_steps": 29704, "loss": 0.28926453590393064, "lr": 6.617455287739161e-08, "epoch": 1.896377592243469, "percentage": 94.82, "elapsed_time": "2 days, 12:31:29", "remaining_time": "3:18:25", "throughput": 5653.04, "total_tokens": 1231738296}
+{"current_steps": 28170, "total_steps": 29704, "loss": 0.29925999641418455, "lr": 6.574648824227369e-08, "epoch": 1.8967142472394292, "percentage": 94.84, "elapsed_time": "2 days, 12:32:04", "remaining_time": "3:17:47", "throughput": 5653.15, "total_tokens": 1231961800}
+{"current_steps": 28175, "total_steps": 29704, "loss": 0.3044144153594971, "lr": 6.5319803455417e-08, "epoch": 1.8970509022353892, "percentage": 94.85, "elapsed_time": "2 days, 12:32:45", "remaining_time": "3:17:08", "throughput": 5653.12, "total_tokens": 1232184856}
+{"current_steps": 28180, "total_steps": 29704, "loss": 0.3226012706756592, "lr": 6.489449863614106e-08, "epoch": 1.8973875572313492, "percentage": 94.87, "elapsed_time": "2 days, 12:33:23", "remaining_time": "3:16:29", "throughput": 5653.14, "total_tokens": 1232404696}
+{"current_steps": 28185, "total_steps": 29704, "loss": 0.29819202423095703, "lr": 6.44705739033824e-08, "epoch": 1.8977242122273095, "percentage": 94.89, "elapsed_time": "2 days, 12:34:03", "remaining_time": "3:15:51", "throughput": 5653.11, "total_tokens": 1232620728}
+{"current_steps": 28190, "total_steps": 29704, "loss": 0.283476448059082, "lr": 6.404802937568899e-08, "epoch": 1.8980608672232697, "percentage": 94.9, "elapsed_time": "2 days, 12:34:36", "remaining_time": "3:15:12", "throughput": 5653.17, "total_tokens": 1232822576}
+{"current_steps": 28195, "total_steps": 29704, "loss": 0.276761531829834, "lr": 6.362686517122463e-08, "epoch": 1.8983975222192297, "percentage": 94.92, "elapsed_time": "2 days, 12:35:15", "remaining_time": "3:14:33", "throughput": 5653.15, "total_tokens": 1233038648}
+{"current_steps": 28200, "total_steps": 29704, "loss": 0.28359341621398926, "lr": 6.320708140776676e-08, "epoch": 1.8987341772151898, "percentage": 94.94, "elapsed_time": "2 days, 12:35:58", "remaining_time": "3:13:55", "throughput": 5653.03, "total_tokens": 1233258352}
+{"current_steps": 28205, "total_steps": 29704, "loss": 0.299153208732605, "lr": 6.278867820270596e-08, "epoch": 1.89907083221115, "percentage": 94.95, "elapsed_time": "2 days, 12:36:35", "remaining_time": "3:13:16", "throughput": 5653.09, "total_tokens": 1233480288}
+{"current_steps": 28210, "total_steps": 29704, "loss": 0.29818384647369384, "lr": 6.237165567304749e-08, "epoch": 1.8994074872071103, "percentage": 94.97, "elapsed_time": "2 days, 12:37:17", "remaining_time": "3:12:37", "throughput": 5652.97, "total_tokens": 1233691368}
+{"current_steps": 28215, "total_steps": 29704, "loss": 0.30460848808288576, "lr": 6.19560139354114e-08, "epoch": 1.8997441422030703, "percentage": 94.99, "elapsed_time": "2 days, 12:37:50", "remaining_time": "3:11:58", "throughput": 5653.22, "total_tokens": 1233930016}
+{"current_steps": 28220, "total_steps": 29704, "loss": 0.30403151512146, "lr": 6.154175310602916e-08, "epoch": 1.9000807971990303, "percentage": 95.0, "elapsed_time": "2 days, 12:38:22", "remaining_time": "3:11:19", "throughput": 5653.4, "total_tokens": 1234154432}
+{"current_steps": 28225, "total_steps": 29704, "loss": 0.2886775255203247, "lr": 6.112887330074812e-08, "epoch": 1.9004174521949906, "percentage": 95.02, "elapsed_time": "2 days, 12:39:00", "remaining_time": "3:10:41", "throughput": 5653.39, "total_tokens": 1234363888}
+{"current_steps": 28230, "total_steps": 29704, "loss": 0.3039447069168091, "lr": 6.07173746350287e-08, "epoch": 1.9007541071909508, "percentage": 95.04, "elapsed_time": "2 days, 12:39:37", "remaining_time": "3:10:02", "throughput": 5653.46, "total_tokens": 1234586704}
+{"current_steps": 28235, "total_steps": 29704, "loss": 0.30956282615661623, "lr": 6.030725722394548e-08, "epoch": 1.9010907621869109, "percentage": 95.05, "elapsed_time": "2 days, 12:40:19", "remaining_time": "3:09:23", "throughput": 5653.4, "total_tokens": 1234811984}
+{"current_steps": 28240, "total_steps": 29704, "loss": 0.31682801246643066, "lr": 5.98985211821862e-08, "epoch": 1.901427417182871, "percentage": 95.07, "elapsed_time": "2 days, 12:41:03", "remaining_time": "3:08:45", "throughput": 5653.25, "total_tokens": 1235026864}
+{"current_steps": 28245, "total_steps": 29704, "loss": 0.29059295654296874, "lr": 5.9491166624052185e-08, "epoch": 1.9017640721788311, "percentage": 95.09, "elapsed_time": "2 days, 12:41:42", "remaining_time": "3:08:06", "throughput": 5653.3, "total_tokens": 1235257424}
+{"current_steps": 28250, "total_steps": 29704, "loss": 0.2929649829864502, "lr": 5.908519366345955e-08, "epoch": 1.9021007271747914, "percentage": 95.11, "elapsed_time": "2 days, 12:42:11", "remaining_time": "3:07:27", "throughput": 5653.5, "total_tokens": 1235465968}
+{"current_steps": 28255, "total_steps": 29704, "loss": 0.32121825218200684, "lr": 5.868060241393747e-08, "epoch": 1.9024373821707514, "percentage": 95.12, "elapsed_time": "2 days, 12:42:54", "remaining_time": "3:06:49", "throughput": 5653.47, "total_tokens": 1235701752}
+{"current_steps": 28260, "total_steps": 29704, "loss": 0.30273237228393557, "lr": 5.8277392988627665e-08, "epoch": 1.9027740371667115, "percentage": 95.14, "elapsed_time": "2 days, 12:43:33", "remaining_time": "3:06:10", "throughput": 5653.45, "total_tokens": 1235918048}
+{"current_steps": 28265, "total_steps": 29704, "loss": 0.2830909967422485, "lr": 5.787556550028772e-08, "epoch": 1.9031106921626717, "percentage": 95.16, "elapsed_time": "2 days, 12:44:15", "remaining_time": "3:05:31", "throughput": 5653.33, "total_tokens": 1236133248}
+{"current_steps": 28270, "total_steps": 29704, "loss": 0.3094156265258789, "lr": 5.747512006128553e-08, "epoch": 1.903447347158632, "percentage": 95.17, "elapsed_time": "2 days, 12:44:50", "remaining_time": "3:04:53", "throughput": 5653.49, "total_tokens": 1236365056}
+{"current_steps": 28275, "total_steps": 29704, "loss": 0.30304322242736814, "lr": 5.707605678360595e-08, "epoch": 1.903784002154592, "percentage": 95.19, "elapsed_time": "2 days, 12:45:29", "remaining_time": "3:04:14", "throughput": 5653.49, "total_tokens": 1236586568}
+{"current_steps": 28280, "total_steps": 29704, "loss": 0.2991356611251831, "lr": 5.6678375778845276e-08, "epoch": 1.904120657150552, "percentage": 95.21, "elapsed_time": "2 days, 12:46:01", "remaining_time": "3:03:35", "throughput": 5653.71, "total_tokens": 1236816112}
+{"current_steps": 28285, "total_steps": 29704, "loss": 0.29000029563903806, "lr": 5.6282077158213435e-08, "epoch": 1.9044573121465123, "percentage": 95.22, "elapsed_time": "2 days, 12:46:39", "remaining_time": "3:02:56", "throughput": 5653.75, "total_tokens": 1237040088}
+{"current_steps": 28290, "total_steps": 29704, "loss": 0.29954440593719484, "lr": 5.588716103253511e-08, "epoch": 1.9047939671424725, "percentage": 95.24, "elapsed_time": "2 days, 12:47:25", "remaining_time": "3:02:18", "throughput": 5653.62, "total_tokens": 1237269408}
+{"current_steps": 28295, "total_steps": 29704, "loss": 0.31754138469696047, "lr": 5.549362751224585e-08, "epoch": 1.9051306221384325, "percentage": 95.26, "elapsed_time": "2 days, 12:48:03", "remaining_time": "3:01:39", "throughput": 5653.58, "total_tokens": 1237476144}
+{"current_steps": 28300, "total_steps": 29704, "loss": 0.2879055976867676, "lr": 5.5101476707397075e-08, "epoch": 1.9054672771343926, "percentage": 95.27, "elapsed_time": "2 days, 12:48:42", "remaining_time": "3:01:01", "throughput": 5653.59, "total_tokens": 1237699056}
+{"current_steps": 28305, "total_steps": 29704, "loss": 0.31143577098846437, "lr": 5.471070872765216e-08, "epoch": 1.9058039321303528, "percentage": 95.29, "elapsed_time": "2 days, 12:49:22", "remaining_time": "3:00:22", "throughput": 5653.59, "total_tokens": 1237922344}
+{"current_steps": 28310, "total_steps": 29704, "loss": 0.3075572490692139, "lr": 5.432132368228815e-08, "epoch": 1.906140587126313, "percentage": 95.31, "elapsed_time": "2 days, 12:49:56", "remaining_time": "2:59:43", "throughput": 5653.71, "total_tokens": 1238144360}
+{"current_steps": 28315, "total_steps": 29704, "loss": 0.29381508827209474, "lr": 5.393332168019572e-08, "epoch": 1.906477242122273, "percentage": 95.32, "elapsed_time": "2 days, 12:50:38", "remaining_time": "2:59:04", "throughput": 5653.7, "total_tokens": 1238376024}
+{"current_steps": 28320, "total_steps": 29704, "loss": 0.31864123344421386, "lr": 5.354670282987695e-08, "epoch": 1.9068138971182331, "percentage": 95.34, "elapsed_time": "2 days, 12:51:11", "remaining_time": "2:58:26", "throughput": 5653.93, "total_tokens": 1238615352}
+{"current_steps": 28325, "total_steps": 29704, "loss": 0.3147014379501343, "lr": 5.316146723945037e-08, "epoch": 1.9071505521141934, "percentage": 95.36, "elapsed_time": "2 days, 12:51:49", "remaining_time": "2:57:47", "throughput": 5653.97, "total_tokens": 1238839600}
+{"current_steps": 28330, "total_steps": 29704, "loss": 0.28926112651824953, "lr": 5.277761501664535e-08, "epoch": 1.9074872071101536, "percentage": 95.37, "elapsed_time": "2 days, 12:52:31", "remaining_time": "2:57:08", "throughput": 5653.88, "total_tokens": 1239055736}
+{"current_steps": 28335, "total_steps": 29704, "loss": 0.3053768157958984, "lr": 5.2395146268804934e-08, "epoch": 1.9078238621061137, "percentage": 95.39, "elapsed_time": "2 days, 12:53:03", "remaining_time": "2:56:29", "throughput": 5654.07, "total_tokens": 1239282360}
+{"current_steps": 28340, "total_steps": 29704, "loss": 0.2954925537109375, "lr": 5.201406110288465e-08, "epoch": 1.9081605171020737, "percentage": 95.41, "elapsed_time": "2 days, 12:53:42", "remaining_time": "2:55:51", "throughput": 5654.1, "total_tokens": 1239505344}
+{"current_steps": 28345, "total_steps": 29704, "loss": 0.26157379150390625, "lr": 5.1634359625454266e-08, "epoch": 1.908497172098034, "percentage": 95.42, "elapsed_time": "2 days, 12:54:21", "remaining_time": "2:55:12", "throughput": 5654.06, "total_tokens": 1239718552}
+{"current_steps": 28350, "total_steps": 29704, "loss": 0.32343695163726804, "lr": 5.125604194269606e-08, "epoch": 1.9088338270939942, "percentage": 95.44, "elapsed_time": "2 days, 12:55:03", "remaining_time": "2:54:33", "throughput": 5653.95, "total_tokens": 1239933464}
+{"current_steps": 28355, "total_steps": 29704, "loss": 0.32165529727935793, "lr": 5.087910816040542e-08, "epoch": 1.9091704820899542, "percentage": 95.46, "elapsed_time": "2 days, 12:55:48", "remaining_time": "2:53:55", "throughput": 5653.79, "total_tokens": 1240153424}
+{"current_steps": 28360, "total_steps": 29704, "loss": 0.29596068859100344, "lr": 5.0503558383990235e-08, "epoch": 1.9095071370859142, "percentage": 95.48, "elapsed_time": "2 days, 12:56:31", "remaining_time": "2:53:17", "throughput": 5653.75, "total_tokens": 1240382984}
+{"current_steps": 28365, "total_steps": 29704, "loss": 0.29495375156402587, "lr": 5.0129392718472614e-08, "epoch": 1.9098437920818745, "percentage": 95.49, "elapsed_time": "2 days, 12:57:08", "remaining_time": "2:52:38", "throughput": 5653.8, "total_tokens": 1240602456}
+{"current_steps": 28370, "total_steps": 29704, "loss": 0.3231898307800293, "lr": 4.9756611268486077e-08, "epoch": 1.9101804470778347, "percentage": 95.51, "elapsed_time": "2 days, 12:57:50", "remaining_time": "2:51:59", "throughput": 5653.77, "total_tokens": 1240834056}
+{"current_steps": 28375, "total_steps": 29704, "loss": 0.3085433006286621, "lr": 4.9385214138277795e-08, "epoch": 1.9105171020737948, "percentage": 95.53, "elapsed_time": "2 days, 12:58:35", "remaining_time": "2:51:21", "throughput": 5653.61, "total_tokens": 1241053560}
+{"current_steps": 28380, "total_steps": 29704, "loss": 0.31458406448364257, "lr": 4.901520143170857e-08, "epoch": 1.9108537570697548, "percentage": 95.54, "elapsed_time": "2 days, 12:59:11", "remaining_time": "2:50:42", "throughput": 5653.7, "total_tokens": 1241278752}
+{"current_steps": 28385, "total_steps": 29704, "loss": 0.2959612846374512, "lr": 4.864657325225064e-08, "epoch": 1.911190412065715, "percentage": 95.56, "elapsed_time": "2 days, 12:59:52", "remaining_time": "2:50:04", "throughput": 5653.67, "total_tokens": 1241501272}
+{"current_steps": 28390, "total_steps": 29704, "loss": 0.28577983379364014, "lr": 4.827932970298932e-08, "epoch": 1.9115270670616753, "percentage": 95.58, "elapsed_time": "2 days, 13:00:31", "remaining_time": "2:49:25", "throughput": 5653.59, "total_tokens": 1241706992}
+{"current_steps": 28395, "total_steps": 29704, "loss": 0.30133938789367676, "lr": 4.7913470886624125e-08, "epoch": 1.9118637220576353, "percentage": 95.59, "elapsed_time": "2 days, 13:01:07", "remaining_time": "2:48:46", "throughput": 5653.67, "total_tokens": 1241929008}
+{"current_steps": 28400, "total_steps": 29704, "loss": 0.2801696538925171, "lr": 4.754899690546544e-08, "epoch": 1.9122003770535954, "percentage": 95.61, "elapsed_time": "2 days, 13:01:41", "remaining_time": "2:48:07", "throughput": 5653.8, "total_tokens": 1242147848}
+{"current_steps": 28405, "total_steps": 29704, "loss": 0.3026761531829834, "lr": 4.718590786143729e-08, "epoch": 1.9125370320495556, "percentage": 95.63, "elapsed_time": "2 days, 13:02:14", "remaining_time": "2:47:28", "throughput": 5653.94, "total_tokens": 1242368696}
+{"current_steps": 28410, "total_steps": 29704, "loss": 0.3060077905654907, "lr": 4.6824203856076776e-08, "epoch": 1.9128736870455159, "percentage": 95.64, "elapsed_time": "2 days, 13:02:50", "remaining_time": "2:46:49", "throughput": 5654.11, "total_tokens": 1242605672}
+{"current_steps": 28415, "total_steps": 29704, "loss": 0.2775508642196655, "lr": 4.646388499053356e-08, "epoch": 1.9132103420414759, "percentage": 95.66, "elapsed_time": "2 days, 13:03:26", "remaining_time": "2:46:11", "throughput": 5654.15, "total_tokens": 1242816568}
+{"current_steps": 28420, "total_steps": 29704, "loss": 0.2732607126235962, "lr": 4.6104951365568696e-08, "epoch": 1.913546997037436, "percentage": 95.68, "elapsed_time": "2 days, 13:04:02", "remaining_time": "2:45:32", "throughput": 5654.21, "total_tokens": 1243037664}
+{"current_steps": 28425, "total_steps": 29704, "loss": 0.2992802858352661, "lr": 4.574740308155801e-08, "epoch": 1.9138836520333962, "percentage": 95.69, "elapsed_time": "2 days, 13:04:44", "remaining_time": "2:44:53", "throughput": 5654.09, "total_tokens": 1243248672}
+{"current_steps": 28430, "total_steps": 29704, "loss": 0.2957939624786377, "lr": 4.539124023848762e-08, "epoch": 1.9142203070293564, "percentage": 95.71, "elapsed_time": "2 days, 13:05:24", "remaining_time": "2:44:15", "throughput": 5654.13, "total_tokens": 1243481672}
+{"current_steps": 28435, "total_steps": 29704, "loss": 0.32006833553314207, "lr": 4.503646293595787e-08, "epoch": 1.9145569620253164, "percentage": 95.73, "elapsed_time": "2 days, 13:05:58", "remaining_time": "2:43:36", "throughput": 5654.23, "total_tokens": 1243698024}
+{"current_steps": 28440, "total_steps": 29704, "loss": 0.2883774280548096, "lr": 4.468307127318105e-08, "epoch": 1.9148936170212765, "percentage": 95.74, "elapsed_time": "2 days, 13:06:34", "remaining_time": "2:42:57", "throughput": 5654.35, "total_tokens": 1243923240}
+{"current_steps": 28445, "total_steps": 29704, "loss": 0.2830281496047974, "lr": 4.4331065348982014e-08, "epoch": 1.9152302720172367, "percentage": 95.76, "elapsed_time": "2 days, 13:07:05", "remaining_time": "2:42:18", "throughput": 5654.54, "total_tokens": 1244140864}
+{"current_steps": 28450, "total_steps": 29704, "loss": 0.3066460371017456, "lr": 4.3980445261798125e-08, "epoch": 1.915566927013197, "percentage": 95.78, "elapsed_time": "2 days, 13:07:51", "remaining_time": "2:41:40", "throughput": 5654.35, "total_tokens": 1244363064}
+{"current_steps": 28455, "total_steps": 29704, "loss": 0.2869099140167236, "lr": 4.363121110967927e-08, "epoch": 1.915903582009157, "percentage": 95.8, "elapsed_time": "2 days, 13:08:31", "remaining_time": "2:41:01", "throughput": 5654.25, "total_tokens": 1244567184}
+{"current_steps": 28460, "total_steps": 29704, "loss": 0.27414395809173586, "lr": 4.328336299028735e-08, "epoch": 1.916240237005117, "percentage": 95.81, "elapsed_time": "2 days, 13:09:16", "remaining_time": "2:40:23", "throughput": 5654.05, "total_tokens": 1244775824}
+{"current_steps": 28465, "total_steps": 29704, "loss": 0.30220975875854494, "lr": 4.293690100089731e-08, "epoch": 1.9165768920010773, "percentage": 95.83, "elapsed_time": "2 days, 13:09:55", "remaining_time": "2:39:44", "throughput": 5654.01, "total_tokens": 1244988464}
+{"current_steps": 28470, "total_steps": 29704, "loss": 0.29154458045959475, "lr": 4.259182523839556e-08, "epoch": 1.9169135469970375, "percentage": 95.85, "elapsed_time": "2 days, 13:10:34", "remaining_time": "2:39:05", "throughput": 5654.04, "total_tokens": 1245214008}
+{"current_steps": 28475, "total_steps": 29704, "loss": 0.3210744857788086, "lr": 4.224813579928211e-08, "epoch": 1.9172502019929976, "percentage": 95.86, "elapsed_time": "2 days, 13:11:18", "remaining_time": "2:38:27", "throughput": 5653.91, "total_tokens": 1245434864}
+{"current_steps": 28480, "total_steps": 29704, "loss": 0.29584832191467286, "lr": 4.190583277966842e-08, "epoch": 1.9175868569889576, "percentage": 95.88, "elapsed_time": "2 days, 13:11:59", "remaining_time": "2:37:48", "throughput": 5653.84, "total_tokens": 1245653840}
+{"current_steps": 28485, "total_steps": 29704, "loss": 0.2821512699127197, "lr": 4.1564916275278477e-08, "epoch": 1.9179235119849178, "percentage": 95.9, "elapsed_time": "2 days, 13:12:38", "remaining_time": "2:37:10", "throughput": 5653.8, "total_tokens": 1245862296}
+{"current_steps": 28490, "total_steps": 29704, "loss": 0.2951045989990234, "lr": 4.122538638144824e-08, "epoch": 1.918260166980878, "percentage": 95.91, "elapsed_time": "2 days, 13:13:15", "remaining_time": "2:36:31", "throughput": 5653.84, "total_tokens": 1246082824}
+{"current_steps": 28495, "total_steps": 29704, "loss": 0.2757328271865845, "lr": 4.08872431931262e-08, "epoch": 1.9185968219768381, "percentage": 95.93, "elapsed_time": "2 days, 13:13:55", "remaining_time": "2:35:52", "throughput": 5653.78, "total_tokens": 1246295392}
+{"current_steps": 28500, "total_steps": 29704, "loss": 0.29179792404174804, "lr": 4.055048680487339e-08, "epoch": 1.9189334769727981, "percentage": 95.95, "elapsed_time": "2 days, 13:14:33", "remaining_time": "2:35:14", "throughput": 5653.84, "total_tokens": 1246522464}
+{"current_steps": 28505, "total_steps": 29704, "loss": 0.3171137809753418, "lr": 4.0215117310862806e-08, "epoch": 1.9192701319687584, "percentage": 95.96, "elapsed_time": "2 days, 13:15:14", "remaining_time": "2:34:35", "throughput": 5653.68, "total_tokens": 1246721184}
+{"current_steps": 28510, "total_steps": 29704, "loss": 0.3017882347106934, "lr": 3.9881134804878316e-08, "epoch": 1.9196067869647186, "percentage": 95.98, "elapsed_time": "2 days, 13:15:55", "remaining_time": "2:33:56", "throughput": 5653.55, "total_tokens": 1246923736}
+{"current_steps": 28515, "total_steps": 29704, "loss": 0.3179955005645752, "lr": 3.9548539380318e-08, "epoch": 1.9199434419606787, "percentage": 96.0, "elapsed_time": "2 days, 13:16:28", "remaining_time": "2:33:17", "throughput": 5653.75, "total_tokens": 1247152480}
+{"current_steps": 28520, "total_steps": 29704, "loss": 0.3016833305358887, "lr": 3.921733113019077e-08, "epoch": 1.9202800969566387, "percentage": 96.01, "elapsed_time": "2 days, 13:17:08", "remaining_time": "2:32:39", "throughput": 5653.63, "total_tokens": 1247354448}
+{"current_steps": 28525, "total_steps": 29704, "loss": 0.3128508567810059, "lr": 3.888751014711867e-08, "epoch": 1.920616751952599, "percentage": 96.03, "elapsed_time": "2 days, 13:17:42", "remaining_time": "2:32:00", "throughput": 5653.7, "total_tokens": 1247557312}
+{"current_steps": 28530, "total_steps": 29704, "loss": 0.28232321739196775, "lr": 3.855907652333402e-08, "epoch": 1.9209534069485592, "percentage": 96.05, "elapsed_time": "2 days, 13:18:22", "remaining_time": "2:31:21", "throughput": 5653.67, "total_tokens": 1247776920}
+{"current_steps": 28535, "total_steps": 29704, "loss": 0.2884651184082031, "lr": 3.8232030350683344e-08, "epoch": 1.9212900619445192, "percentage": 96.06, "elapsed_time": "2 days, 13:19:03", "remaining_time": "2:30:43", "throughput": 5653.65, "total_tokens": 1248006808}
+{"current_steps": 28540, "total_steps": 29704, "loss": 0.2990164041519165, "lr": 3.7906371720622946e-08, "epoch": 1.9216267169404793, "percentage": 96.08, "elapsed_time": "2 days, 13:19:42", "remaining_time": "2:30:04", "throughput": 5653.69, "total_tokens": 1248234992}
+{"current_steps": 28545, "total_steps": 29704, "loss": 0.28374738693237306, "lr": 3.758210072422275e-08, "epoch": 1.9219633719364395, "percentage": 96.1, "elapsed_time": "2 days, 13:20:18", "remaining_time": "2:29:25", "throughput": 5653.75, "total_tokens": 1248453088}
+{"current_steps": 28550, "total_steps": 29704, "loss": 0.30987684726715087, "lr": 3.72592174521641e-08, "epoch": 1.9223000269323998, "percentage": 96.12, "elapsed_time": "2 days, 13:21:01", "remaining_time": "2:28:47", "throughput": 5653.66, "total_tokens": 1248674912}
+{"current_steps": 28555, "total_steps": 29704, "loss": 0.30152506828308107, "lr": 3.693772199474033e-08, "epoch": 1.9226366819283598, "percentage": 96.13, "elapsed_time": "2 days, 13:21:39", "remaining_time": "2:28:08", "throughput": 5653.69, "total_tokens": 1248896320}
+{"current_steps": 28560, "total_steps": 29704, "loss": 0.30233216285705566, "lr": 3.661761444185674e-08, "epoch": 1.9229733369243198, "percentage": 96.15, "elapsed_time": "2 days, 13:22:21", "remaining_time": "2:27:30", "throughput": 5653.59, "total_tokens": 1249110256}
+{"current_steps": 28565, "total_steps": 29704, "loss": 0.3038165807723999, "lr": 3.6298894883030066e-08, "epoch": 1.92330999192028, "percentage": 96.17, "elapsed_time": "2 days, 13:22:55", "remaining_time": "2:26:51", "throughput": 5653.75, "total_tokens": 1249337408}
+{"current_steps": 28570, "total_steps": 29704, "loss": 0.3213151454925537, "lr": 3.5981563407389007e-08, "epoch": 1.9236466469162403, "percentage": 96.18, "elapsed_time": "2 days, 13:23:38", "remaining_time": "2:26:12", "throughput": 5653.63, "total_tokens": 1249559176}
+{"current_steps": 28575, "total_steps": 29704, "loss": 0.2826304197311401, "lr": 3.566562010367536e-08, "epoch": 1.9239833019122003, "percentage": 96.2, "elapsed_time": "2 days, 13:24:11", "remaining_time": "2:25:33", "throughput": 5653.74, "total_tokens": 1249766312}
+{"current_steps": 28580, "total_steps": 29704, "loss": 0.2978407621383667, "lr": 3.535106506024011e-08, "epoch": 1.9243199569081604, "percentage": 96.22, "elapsed_time": "2 days, 13:24:48", "remaining_time": "2:24:55", "throughput": 5653.76, "total_tokens": 1249983696}
+{"current_steps": 28585, "total_steps": 29704, "loss": 0.2644341230392456, "lr": 3.5037898365049005e-08, "epoch": 1.9246566119041206, "percentage": 96.23, "elapsed_time": "2 days, 13:25:31", "remaining_time": "2:24:16", "throughput": 5653.67, "total_tokens": 1250204168}
+{"current_steps": 28590, "total_steps": 29704, "loss": 0.29390702247619627, "lr": 3.472612010567755e-08, "epoch": 1.9249932669000809, "percentage": 96.25, "elapsed_time": "2 days, 13:26:07", "remaining_time": "2:23:37", "throughput": 5653.68, "total_tokens": 1250412232}
+{"current_steps": 28595, "total_steps": 29704, "loss": 0.3093965291976929, "lr": 3.441573036931323e-08, "epoch": 1.925329921896041, "percentage": 96.27, "elapsed_time": "2 days, 13:26:45", "remaining_time": "2:22:59", "throughput": 5653.74, "total_tokens": 1250638880}
+{"current_steps": 28600, "total_steps": 29704, "loss": 0.273404598236084, "lr": 3.410672924275604e-08, "epoch": 1.925666576892001, "percentage": 96.28, "elapsed_time": "2 days, 13:27:29", "remaining_time": "2:22:20", "throughput": 5653.67, "total_tokens": 1250870096}
+{"current_steps": 28605, "total_steps": 29704, "loss": 0.3114511489868164, "lr": 3.3799116812416876e-08, "epoch": 1.9260032318879612, "percentage": 96.3, "elapsed_time": "2 days, 13:28:05", "remaining_time": "2:21:41", "throughput": 5653.7, "total_tokens": 1251082376}
+{"current_steps": 28610, "total_steps": 29704, "loss": 0.31067667007446287, "lr": 3.349289316431803e-08, "epoch": 1.9263398868839214, "percentage": 96.32, "elapsed_time": "2 days, 13:28:42", "remaining_time": "2:21:02", "throughput": 5653.79, "total_tokens": 1251307384}
+{"current_steps": 28615, "total_steps": 29704, "loss": 0.29280407428741456, "lr": 3.3188058384095446e-08, "epoch": 1.9266765418798815, "percentage": 96.33, "elapsed_time": "2 days, 13:29:20", "remaining_time": "2:20:24", "throughput": 5653.8, "total_tokens": 1251525400}
+{"current_steps": 28620, "total_steps": 29704, "loss": 0.32468104362487793, "lr": 3.2884612556993714e-08, "epoch": 1.9270131968758415, "percentage": 96.35, "elapsed_time": "2 days, 13:30:03", "remaining_time": "2:19:45", "throughput": 5653.67, "total_tokens": 1251744936}
+{"current_steps": 28625, "total_steps": 29704, "loss": 0.3124075174331665, "lr": 3.258255576787161e-08, "epoch": 1.9273498518718017, "percentage": 96.37, "elapsed_time": "2 days, 13:30:43", "remaining_time": "2:19:07", "throughput": 5653.62, "total_tokens": 1251958328}
+{"current_steps": 28630, "total_steps": 29704, "loss": 0.31823225021362306, "lr": 3.22818881011977e-08, "epoch": 1.927686506867762, "percentage": 96.38, "elapsed_time": "2 days, 13:31:21", "remaining_time": "2:18:28", "throughput": 5653.7, "total_tokens": 1252192752}
+{"current_steps": 28635, "total_steps": 29704, "loss": 0.27967541217803954, "lr": 3.198260964105249e-08, "epoch": 1.928023161863722, "percentage": 96.4, "elapsed_time": "2 days, 13:31:59", "remaining_time": "2:17:49", "throughput": 5653.72, "total_tokens": 1252407792}
+{"current_steps": 28640, "total_steps": 29704, "loss": 0.2996687412261963, "lr": 3.168472047112903e-08, "epoch": 1.928359816859682, "percentage": 96.42, "elapsed_time": "2 days, 13:32:46", "remaining_time": "2:17:11", "throughput": 5653.5, "total_tokens": 1252623280}
+{"current_steps": 28645, "total_steps": 29704, "loss": 0.28585450649261473, "lr": 3.138822067473068e-08, "epoch": 1.9286964718556423, "percentage": 96.43, "elapsed_time": "2 days, 13:33:24", "remaining_time": "2:16:32", "throughput": 5653.48, "total_tokens": 1252835408}
+{"current_steps": 28650, "total_steps": 29704, "loss": 0.30099103450775144, "lr": 3.10931103347728e-08, "epoch": 1.9290331268516026, "percentage": 96.45, "elapsed_time": "2 days, 13:34:06", "remaining_time": "2:15:54", "throughput": 5653.4, "total_tokens": 1253057512}
+{"current_steps": 28655, "total_steps": 29704, "loss": 0.3233338356018066, "lr": 3.0799389533781587e-08, "epoch": 1.9293697818475626, "percentage": 96.47, "elapsed_time": "2 days, 13:34:43", "remaining_time": "2:15:15", "throughput": 5653.48, "total_tokens": 1253285920}
+{"current_steps": 28660, "total_steps": 29704, "loss": 0.28039982318878176, "lr": 3.050705835389578e-08, "epoch": 1.9297064368435226, "percentage": 96.49, "elapsed_time": "2 days, 13:35:22", "remaining_time": "2:14:36", "throughput": 5653.48, "total_tokens": 1253502504}
+{"current_steps": 28665, "total_steps": 29704, "loss": 0.2949859380722046, "lr": 3.021611687686443e-08, "epoch": 1.9300430918394829, "percentage": 96.5, "elapsed_time": "2 days, 13:35:59", "remaining_time": "2:13:57", "throughput": 5653.48, "total_tokens": 1253714448}
+{"current_steps": 28670, "total_steps": 29704, "loss": 0.3135673999786377, "lr": 2.992656518404857e-08, "epoch": 1.9303797468354431, "percentage": 96.52, "elapsed_time": "2 days, 13:36:38", "remaining_time": "2:13:19", "throughput": 5653.42, "total_tokens": 1253921528}
+{"current_steps": 28675, "total_steps": 29704, "loss": 0.28868966102600097, "lr": 2.96384033564201e-08, "epoch": 1.9307164018314031, "percentage": 96.54, "elapsed_time": "2 days, 13:37:15", "remaining_time": "2:12:40", "throughput": 5653.45, "total_tokens": 1254137480}
+{"current_steps": 28680, "total_steps": 29704, "loss": 0.27773294448852537, "lr": 2.935163147456288e-08, "epoch": 1.9310530568273632, "percentage": 96.55, "elapsed_time": "2 days, 13:37:56", "remaining_time": "2:12:01", "throughput": 5653.42, "total_tokens": 1254360912}
+{"current_steps": 28685, "total_steps": 29704, "loss": 0.3049351215362549, "lr": 2.906624961867166e-08, "epoch": 1.9313897118233234, "percentage": 96.57, "elapsed_time": "2 days, 13:38:33", "remaining_time": "2:11:23", "throughput": 5653.48, "total_tokens": 1254581976}
+{"current_steps": 28690, "total_steps": 29704, "loss": 0.2789355754852295, "lr": 2.8782257868553154e-08, "epoch": 1.9317263668192837, "percentage": 96.59, "elapsed_time": "2 days, 13:39:10", "remaining_time": "2:10:44", "throughput": 5653.5, "total_tokens": 1254798040}
+{"current_steps": 28695, "total_steps": 29704, "loss": 0.2943142890930176, "lr": 2.849965630362328e-08, "epoch": 1.9320630218152437, "percentage": 96.6, "elapsed_time": "2 days, 13:39:51", "remaining_time": "2:10:05", "throughput": 5653.4, "total_tokens": 1255008552}
+{"current_steps": 28700, "total_steps": 29704, "loss": 0.2914332389831543, "lr": 2.8218445002912153e-08, "epoch": 1.9323996768112037, "percentage": 96.62, "elapsed_time": "2 days, 13:40:32", "remaining_time": "2:09:27", "throughput": 5653.38, "total_tokens": 1255234552}
+{"current_steps": 28705, "total_steps": 29704, "loss": 0.2954190731048584, "lr": 2.7938624045059094e-08, "epoch": 1.932736331807164, "percentage": 96.64, "elapsed_time": "2 days, 13:41:07", "remaining_time": "2:08:48", "throughput": 5653.51, "total_tokens": 1255459816}
+{"current_steps": 28710, "total_steps": 29704, "loss": 0.3093540668487549, "lr": 2.766019350831428e-08, "epoch": 1.9330729868031242, "percentage": 96.65, "elapsed_time": "2 days, 13:41:46", "remaining_time": "2:08:09", "throughput": 5653.59, "total_tokens": 1255699664}
+{"current_steps": 28715, "total_steps": 29704, "loss": 0.3093677282333374, "lr": 2.7383153470541546e-08, "epoch": 1.9334096417990843, "percentage": 96.67, "elapsed_time": "2 days, 13:42:19", "remaining_time": "2:07:30", "throughput": 5653.79, "total_tokens": 1255930096}
+{"current_steps": 28720, "total_steps": 29704, "loss": 0.2810804843902588, "lr": 2.710750400921336e-08, "epoch": 1.9337462967950443, "percentage": 96.69, "elapsed_time": "2 days, 13:42:59", "remaining_time": "2:06:52", "throughput": 5653.68, "total_tokens": 1256129960}
+{"current_steps": 28725, "total_steps": 29704, "loss": 0.29244356155395507, "lr": 2.683324520141417e-08, "epoch": 1.9340829517910045, "percentage": 96.7, "elapsed_time": "2 days, 13:43:40", "remaining_time": "2:06:13", "throughput": 5653.63, "total_tokens": 1256355072}
+{"current_steps": 28730, "total_steps": 29704, "loss": 0.3020967483520508, "lr": 2.65603771238393e-08, "epoch": 1.9344196067869648, "percentage": 96.72, "elapsed_time": "2 days, 13:44:26", "remaining_time": "2:05:35", "throughput": 5653.41, "total_tokens": 1256563176}
+{"current_steps": 28735, "total_steps": 29704, "loss": 0.2978071689605713, "lr": 2.6288899852796036e-08, "epoch": 1.9347562617829248, "percentage": 96.74, "elapsed_time": "2 days, 13:45:06", "remaining_time": "2:04:56", "throughput": 5653.45, "total_tokens": 1256797928}
+{"current_steps": 28740, "total_steps": 29704, "loss": 0.302975058555603, "lr": 2.6018813464202543e-08, "epoch": 1.9350929167788848, "percentage": 96.75, "elapsed_time": "2 days, 13:45:41", "remaining_time": "2:04:17", "throughput": 5653.51, "total_tokens": 1257009296}
+{"current_steps": 28745, "total_steps": 29704, "loss": 0.29305691719055177, "lr": 2.575011803358618e-08, "epoch": 1.935429571774845, "percentage": 96.77, "elapsed_time": "2 days, 13:46:20", "remaining_time": "2:03:39", "throughput": 5653.54, "total_tokens": 1257233712}
+{"current_steps": 28750, "total_steps": 29704, "loss": 0.27554895877838137, "lr": 2.5482813636087955e-08, "epoch": 1.9357662267708053, "percentage": 96.79, "elapsed_time": "2 days, 13:47:03", "remaining_time": "2:03:00", "throughput": 5653.37, "total_tokens": 1257442168}
+{"current_steps": 28755, "total_steps": 29704, "loss": 0.2871532678604126, "lr": 2.5216900346458073e-08, "epoch": 1.9361028817667654, "percentage": 96.81, "elapsed_time": "2 days, 13:47:46", "remaining_time": "2:02:22", "throughput": 5653.27, "total_tokens": 1257660464}
+{"current_steps": 28760, "total_steps": 29704, "loss": 0.29356064796447756, "lr": 2.4952378239058716e-08, "epoch": 1.9364395367627254, "percentage": 96.82, "elapsed_time": "2 days, 13:48:19", "remaining_time": "2:01:43", "throughput": 5653.45, "total_tokens": 1257889496}
+{"current_steps": 28765, "total_steps": 29704, "loss": 0.2963290214538574, "lr": 2.4689247387862934e-08, "epoch": 1.9367761917586857, "percentage": 96.84, "elapsed_time": "2 days, 13:48:52", "remaining_time": "2:01:04", "throughput": 5653.57, "total_tokens": 1258103416}
+{"current_steps": 28770, "total_steps": 29704, "loss": 0.28399271965026857, "lr": 2.4427507866453537e-08, "epoch": 1.937112846754646, "percentage": 96.86, "elapsed_time": "2 days, 13:49:33", "remaining_time": "2:00:25", "throughput": 5653.38, "total_tokens": 1258294944}
+{"current_steps": 28775, "total_steps": 29704, "loss": 0.2864212989807129, "lr": 2.416715974802586e-08, "epoch": 1.937449501750606, "percentage": 96.87, "elapsed_time": "2 days, 13:50:08", "remaining_time": "1:59:46", "throughput": 5653.45, "total_tokens": 1258505472}
+{"current_steps": 28780, "total_steps": 29704, "loss": 0.30834760665893557, "lr": 2.3908203105384997e-08, "epoch": 1.937786156746566, "percentage": 96.89, "elapsed_time": "2 days, 13:50:42", "remaining_time": "1:59:08", "throughput": 5653.58, "total_tokens": 1258727136}
+{"current_steps": 28785, "total_steps": 29704, "loss": 0.2865851402282715, "lr": 2.365063801094747e-08, "epoch": 1.9381228117425262, "percentage": 96.91, "elapsed_time": "2 days, 13:51:16", "remaining_time": "1:58:29", "throughput": 5653.68, "total_tokens": 1258942528}
+{"current_steps": 28790, "total_steps": 29704, "loss": 0.30954954624176023, "lr": 2.3394464536740658e-08, "epoch": 1.9384594667384865, "percentage": 96.92, "elapsed_time": "2 days, 13:51:58", "remaining_time": "1:57:50", "throughput": 5653.67, "total_tokens": 1259175632}
+{"current_steps": 28795, "total_steps": 29704, "loss": 0.3149877548217773, "lr": 2.3139682754402816e-08, "epoch": 1.9387961217344465, "percentage": 96.94, "elapsed_time": "2 days, 13:52:35", "remaining_time": "1:57:11", "throughput": 5653.75, "total_tokens": 1259405216}
+{"current_steps": 28800, "total_steps": 29704, "loss": 0.2907993316650391, "lr": 2.2886292735182502e-08, "epoch": 1.9391327767304065, "percentage": 96.96, "elapsed_time": "2 days, 13:53:12", "remaining_time": "1:56:33", "throughput": 5653.87, "total_tokens": 1259637856}
+{"current_steps": 28805, "total_steps": 29704, "loss": 0.2934614896774292, "lr": 2.2634294549939705e-08, "epoch": 1.9394694317263668, "percentage": 96.97, "elapsed_time": "2 days, 13:53:53", "remaining_time": "1:55:54", "throughput": 5653.85, "total_tokens": 1259866088}
+{"current_steps": 28810, "total_steps": 29704, "loss": 0.3583078145980835, "lr": 2.2383688269144723e-08, "epoch": 1.939806086722327, "percentage": 96.99, "elapsed_time": "2 days, 13:54:33", "remaining_time": "1:55:15", "throughput": 5653.83, "total_tokens": 1260088088}
+{"current_steps": 28815, "total_steps": 29704, "loss": 0.2929667949676514, "lr": 2.2134473962878712e-08, "epoch": 1.940142741718287, "percentage": 97.01, "elapsed_time": "2 days, 13:55:09", "remaining_time": "1:54:37", "throughput": 5653.87, "total_tokens": 1260302240}
+{"current_steps": 28820, "total_steps": 29704, "loss": 0.28723697662353515, "lr": 2.1886651700833705e-08, "epoch": 1.940479396714247, "percentage": 97.02, "elapsed_time": "2 days, 13:55:44", "remaining_time": "1:53:58", "throughput": 5653.82, "total_tokens": 1260489696}
+{"current_steps": 28825, "total_steps": 29704, "loss": 0.3041691780090332, "lr": 2.1640221552312603e-08, "epoch": 1.9408160517102075, "percentage": 97.04, "elapsed_time": "2 days, 13:56:23", "remaining_time": "1:53:19", "throughput": 5653.81, "total_tokens": 1260707328}
+{"current_steps": 28830, "total_steps": 29704, "loss": 0.30638680458068845, "lr": 2.1395183586229163e-08, "epoch": 1.9411527067061676, "percentage": 97.06, "elapsed_time": "2 days, 13:57:02", "remaining_time": "1:52:41", "throughput": 5653.88, "total_tokens": 1260942368}
+{"current_steps": 28835, "total_steps": 29704, "loss": 0.28965296745300295, "lr": 2.1151537871106353e-08, "epoch": 1.9414893617021276, "percentage": 97.07, "elapsed_time": "2 days, 13:57:38", "remaining_time": "1:52:02", "throughput": 5653.9, "total_tokens": 1261152728}
+{"current_steps": 28840, "total_steps": 29704, "loss": 0.285279655456543, "lr": 2.0909284475080226e-08, "epoch": 1.9418260166980879, "percentage": 97.09, "elapsed_time": "2 days, 13:58:27", "remaining_time": "1:51:23", "throughput": 5653.66, "total_tokens": 1261373808}
+{"current_steps": 28845, "total_steps": 29704, "loss": 0.2754975318908691, "lr": 2.0668423465894928e-08, "epoch": 1.942162671694048, "percentage": 97.11, "elapsed_time": "2 days, 13:59:08", "remaining_time": "1:50:45", "throughput": 5653.59, "total_tokens": 1261592168}
+{"current_steps": 28850, "total_steps": 29704, "loss": 0.28652522563934324, "lr": 2.0428954910907684e-08, "epoch": 1.9424993266900081, "percentage": 97.12, "elapsed_time": "2 days, 13:59:46", "remaining_time": "1:50:06", "throughput": 5653.5, "total_tokens": 1261787104}
+{"current_steps": 28855, "total_steps": 29704, "loss": 0.3006443977355957, "lr": 2.019087887708382e-08, "epoch": 1.9428359816859682, "percentage": 97.14, "elapsed_time": "2 days, 14:00:27", "remaining_time": "1:49:28", "throughput": 5653.45, "total_tokens": 1262007728}
+{"current_steps": 28860, "total_steps": 29704, "loss": 0.29778175354003905, "lr": 1.9954195431001744e-08, "epoch": 1.9431726366819284, "percentage": 97.16, "elapsed_time": "2 days, 14:01:02", "remaining_time": "1:48:49", "throughput": 5653.59, "total_tokens": 1262233872}
+{"current_steps": 28865, "total_steps": 29704, "loss": 0.2813864231109619, "lr": 1.9718904638848513e-08, "epoch": 1.9435092916778887, "percentage": 97.18, "elapsed_time": "2 days, 14:01:45", "remaining_time": "1:48:10", "throughput": 5653.46, "total_tokens": 1262445256}
+{"current_steps": 28870, "total_steps": 29704, "loss": 0.2955453395843506, "lr": 1.9485006566422602e-08, "epoch": 1.9438459466738487, "percentage": 97.19, "elapsed_time": "2 days, 14:02:27", "remaining_time": "1:47:32", "throughput": 5653.44, "total_tokens": 1262680216}
+{"current_steps": 28875, "total_steps": 29704, "loss": 0.3031654357910156, "lr": 1.9252501279132806e-08, "epoch": 1.9441826016698087, "percentage": 97.21, "elapsed_time": "2 days, 14:03:11", "remaining_time": "1:46:53", "throughput": 5653.29, "total_tokens": 1262895920}
+{"current_steps": 28880, "total_steps": 29704, "loss": 0.2994793176651001, "lr": 1.902138884199878e-08, "epoch": 1.944519256665769, "percentage": 97.23, "elapsed_time": "2 days, 14:03:47", "remaining_time": "1:46:14", "throughput": 5653.4, "total_tokens": 1263122360}
+{"current_steps": 28885, "total_steps": 29704, "loss": 0.30690946578979494, "lr": 1.879166931964993e-08, "epoch": 1.9448559116617292, "percentage": 97.24, "elapsed_time": "2 days, 14:04:27", "remaining_time": "1:45:36", "throughput": 5653.34, "total_tokens": 1263337688}
+{"current_steps": 28890, "total_steps": 29704, "loss": 0.29476094245910645, "lr": 1.856334277632765e-08, "epoch": 1.9451925666576892, "percentage": 97.26, "elapsed_time": "2 days, 14:05:06", "remaining_time": "1:44:57", "throughput": 5653.41, "total_tokens": 1263574904}
+{"current_steps": 28895, "total_steps": 29704, "loss": 0.3002762794494629, "lr": 1.8336409275880872e-08, "epoch": 1.9455292216536493, "percentage": 97.28, "elapsed_time": "2 days, 14:05:52", "remaining_time": "1:44:19", "throughput": 5653.29, "total_tokens": 1263807760}
+{"current_steps": 28900, "total_steps": 29704, "loss": 0.2916655302047729, "lr": 1.8110868881772713e-08, "epoch": 1.9458658766496095, "percentage": 97.29, "elapsed_time": "2 days, 14:06:32", "remaining_time": "1:43:40", "throughput": 5653.21, "total_tokens": 1264016568}
+{"current_steps": 28905, "total_steps": 29704, "loss": 0.2689183712005615, "lr": 1.788672165707328e-08, "epoch": 1.9462025316455698, "percentage": 97.31, "elapsed_time": "2 days, 14:07:17", "remaining_time": "1:43:01", "throughput": 5653.06, "total_tokens": 1264236144}
+{"current_steps": 28910, "total_steps": 29704, "loss": 0.2718139886856079, "lr": 1.7663967664465763e-08, "epoch": 1.9465391866415298, "percentage": 97.33, "elapsed_time": "2 days, 14:07:57", "remaining_time": "1:42:23", "throughput": 5652.98, "total_tokens": 1264443160}
+{"current_steps": 28915, "total_steps": 29704, "loss": 0.3149592399597168, "lr": 1.7442606966242005e-08, "epoch": 1.9468758416374898, "percentage": 97.34, "elapsed_time": "2 days, 14:08:40", "remaining_time": "1:41:44", "throughput": 5652.95, "total_tokens": 1264681216}
+{"current_steps": 28920, "total_steps": 29704, "loss": 0.3035134315490723, "lr": 1.722263962430526e-08, "epoch": 1.94721249663345, "percentage": 97.36, "elapsed_time": "2 days, 14:09:15", "remaining_time": "1:41:05", "throughput": 5653.06, "total_tokens": 1264904128}
+{"current_steps": 28925, "total_steps": 29704, "loss": 0.28637185096740725, "lr": 1.7004065700168547e-08, "epoch": 1.9475491516294103, "percentage": 97.38, "elapsed_time": "2 days, 14:09:48", "remaining_time": "1:40:26", "throughput": 5653.24, "total_tokens": 1265126560}
+{"current_steps": 28930, "total_steps": 29704, "loss": 0.3050400733947754, "lr": 1.6786885254954644e-08, "epoch": 1.9478858066253704, "percentage": 97.39, "elapsed_time": "2 days, 14:10:24", "remaining_time": "1:39:48", "throughput": 5653.37, "total_tokens": 1265361040}
+{"current_steps": 28935, "total_steps": 29704, "loss": 0.2956183910369873, "lr": 1.6571098349398296e-08, "epoch": 1.9482224616213304, "percentage": 97.41, "elapsed_time": "2 days, 14:11:03", "remaining_time": "1:39:09", "throughput": 5653.29, "total_tokens": 1265565976}
+{"current_steps": 28940, "total_steps": 29704, "loss": 0.31615753173828126, "lr": 1.635670504384346e-08, "epoch": 1.9485591166172906, "percentage": 97.43, "elapsed_time": "2 days, 14:11:48", "remaining_time": "1:38:31", "throughput": 5653.04, "total_tokens": 1265765816}
+{"current_steps": 28945, "total_steps": 29704, "loss": 0.29006495475769045, "lr": 1.6143705398243837e-08, "epoch": 1.948895771613251, "percentage": 97.44, "elapsed_time": "2 days, 14:12:32", "remaining_time": "1:37:52", "throughput": 5652.86, "total_tokens": 1265969312}
+{"current_steps": 28950, "total_steps": 29704, "loss": 0.29108805656433107, "lr": 1.5932099472165674e-08, "epoch": 1.949232426609211, "percentage": 97.46, "elapsed_time": "2 days, 14:13:10", "remaining_time": "1:37:13", "throughput": 5652.82, "total_tokens": 1266179488}
+{"current_steps": 28955, "total_steps": 29704, "loss": 0.2916900157928467, "lr": 1.572188732478164e-08, "epoch": 1.949569081605171, "percentage": 97.48, "elapsed_time": "2 days, 14:13:50", "remaining_time": "1:36:35", "throughput": 5652.84, "total_tokens": 1266409776}
+{"current_steps": 28960, "total_steps": 29704, "loss": 0.3035112857818604, "lr": 1.551306901487859e-08, "epoch": 1.9499057366011312, "percentage": 97.5, "elapsed_time": "2 days, 14:14:34", "remaining_time": "1:35:56", "throughput": 5652.7, "total_tokens": 1266624136}
+{"current_steps": 28965, "total_steps": 29704, "loss": 0.2758214473724365, "lr": 1.5305644600852043e-08, "epoch": 1.9502423915970915, "percentage": 97.51, "elapsed_time": "2 days, 14:15:14", "remaining_time": "1:35:17", "throughput": 5652.71, "total_tokens": 1266855056}
+{"current_steps": 28970, "total_steps": 29704, "loss": 0.2854607105255127, "lr": 1.5099614140706154e-08, "epoch": 1.9505790465930515, "percentage": 97.53, "elapsed_time": "2 days, 14:15:58", "remaining_time": "1:34:39", "throughput": 5652.57, "total_tokens": 1267072344}
+{"current_steps": 28975, "total_steps": 29704, "loss": 0.31021819114685056, "lr": 1.489497769205761e-08, "epoch": 1.9509157015890115, "percentage": 97.55, "elapsed_time": "2 days, 14:16:39", "remaining_time": "1:34:00", "throughput": 5652.46, "total_tokens": 1267277496}
+{"current_steps": 28980, "total_steps": 29704, "loss": 0.2921774864196777, "lr": 1.4691735312132304e-08, "epoch": 1.9512523565849718, "percentage": 97.56, "elapsed_time": "2 days, 14:17:17", "remaining_time": "1:33:22", "throughput": 5652.51, "total_tokens": 1267506744}
+{"current_steps": 28985, "total_steps": 29704, "loss": 0.30820393562316895, "lr": 1.4489887057766439e-08, "epoch": 1.951589011580932, "percentage": 97.58, "elapsed_time": "2 days, 14:17:52", "remaining_time": "1:32:43", "throughput": 5652.6, "total_tokens": 1267724224}
+{"current_steps": 28990, "total_steps": 29704, "loss": 0.29095611572265623, "lr": 1.428943298540597e-08, "epoch": 1.951925666576892, "percentage": 97.6, "elapsed_time": "2 days, 14:18:29", "remaining_time": "1:32:04", "throughput": 5652.68, "total_tokens": 1267950120}
+{"current_steps": 28995, "total_steps": 29704, "loss": 0.3069890022277832, "lr": 1.4090373151107173e-08, "epoch": 1.952262321572852, "percentage": 97.61, "elapsed_time": "2 days, 14:19:02", "remaining_time": "1:31:25", "throughput": 5652.86, "total_tokens": 1268176376}
+{"current_steps": 29000, "total_steps": 29704, "loss": 0.2565622329711914, "lr": 1.3892707610536627e-08, "epoch": 1.9525989765688123, "percentage": 97.63, "elapsed_time": "2 days, 14:19:46", "remaining_time": "1:30:47", "throughput": 5652.68, "total_tokens": 1268384472}
+{"current_steps": 29005, "total_steps": 29704, "loss": 0.2830059051513672, "lr": 1.3696436418970672e-08, "epoch": 1.9529356315647726, "percentage": 97.65, "elapsed_time": "2 days, 14:20:28", "remaining_time": "1:30:08", "throughput": 5652.6, "total_tokens": 1268601536}
+{"current_steps": 29010, "total_steps": 29704, "loss": 0.31444830894470216, "lr": 1.3501559631296512e-08, "epoch": 1.9532722865607326, "percentage": 97.66, "elapsed_time": "2 days, 14:21:11", "remaining_time": "1:29:29", "throughput": 5652.42, "total_tokens": 1268805592}
+{"current_steps": 29015, "total_steps": 29704, "loss": 0.31547060012817385, "lr": 1.3308077302010557e-08, "epoch": 1.9536089415566926, "percentage": 97.68, "elapsed_time": "2 days, 14:21:49", "remaining_time": "1:28:51", "throughput": 5652.5, "total_tokens": 1269039096}
+{"current_steps": 29020, "total_steps": 29704, "loss": 0.2989495754241943, "lr": 1.3115989485218972e-08, "epoch": 1.9539455965526529, "percentage": 97.7, "elapsed_time": "2 days, 14:22:23", "remaining_time": "1:28:12", "throughput": 5652.56, "total_tokens": 1269246368}
+{"current_steps": 29025, "total_steps": 29704, "loss": 0.2807016849517822, "lr": 1.2925296234638784e-08, "epoch": 1.9542822515486131, "percentage": 97.71, "elapsed_time": "2 days, 14:23:04", "remaining_time": "1:27:33", "throughput": 5652.5, "total_tokens": 1269464104}
+{"current_steps": 29030, "total_steps": 29704, "loss": 0.29611780643463137, "lr": 1.2735997603597339e-08, "epoch": 1.9546189065445732, "percentage": 97.73, "elapsed_time": "2 days, 14:23:36", "remaining_time": "1:26:55", "throughput": 5652.74, "total_tokens": 1269699728}
+{"current_steps": 29035, "total_steps": 29704, "loss": 0.2928828001022339, "lr": 1.2548093645030623e-08, "epoch": 1.9549555615405332, "percentage": 97.75, "elapsed_time": "2 days, 14:24:16", "remaining_time": "1:26:16", "throughput": 5652.76, "total_tokens": 1269928480}
+{"current_steps": 29040, "total_steps": 29704, "loss": 0.28606395721435546, "lr": 1.236158441148605e-08, "epoch": 1.9552922165364934, "percentage": 97.76, "elapsed_time": "2 days, 14:24:52", "remaining_time": "1:25:37", "throughput": 5652.85, "total_tokens": 1270155816}
+{"current_steps": 29045, "total_steps": 29704, "loss": 0.32508511543273927, "lr": 1.2176469955119119e-08, "epoch": 1.9556288715324537, "percentage": 97.78, "elapsed_time": "2 days, 14:25:33", "remaining_time": "1:24:58", "throughput": 5652.87, "total_tokens": 1270385632}
+{"current_steps": 29050, "total_steps": 29704, "loss": 0.3148582696914673, "lr": 1.199275032769842e-08, "epoch": 1.9559655265284137, "percentage": 97.8, "elapsed_time": "2 days, 14:26:16", "remaining_time": "1:24:20", "throughput": 5652.77, "total_tokens": 1270610312}
+{"current_steps": 29055, "total_steps": 29704, "loss": 0.32245888710021975, "lr": 1.1810425580598971e-08, "epoch": 1.9563021815243737, "percentage": 97.82, "elapsed_time": "2 days, 14:26:57", "remaining_time": "1:23:41", "throughput": 5652.79, "total_tokens": 1270844152}
+{"current_steps": 29060, "total_steps": 29704, "loss": 0.29970650672912597, "lr": 1.1629495764807763e-08, "epoch": 1.956638836520334, "percentage": 97.83, "elapsed_time": "2 days, 14:27:41", "remaining_time": "1:23:03", "throughput": 5652.73, "total_tokens": 1271079472}
+{"current_steps": 29065, "total_steps": 29704, "loss": 0.29777934551239016, "lr": 1.1449960930921544e-08, "epoch": 1.9569754915162942, "percentage": 97.85, "elapsed_time": "2 days, 14:28:18", "remaining_time": "1:22:24", "throughput": 5652.74, "total_tokens": 1271292624}
+{"current_steps": 29070, "total_steps": 29704, "loss": 0.28283419609069826, "lr": 1.1271821129146266e-08, "epoch": 1.9573121465122543, "percentage": 97.87, "elapsed_time": "2 days, 14:28:52", "remaining_time": "1:21:45", "throughput": 5652.87, "total_tokens": 1271514960}
+{"current_steps": 29075, "total_steps": 29704, "loss": 0.28227386474609373, "lr": 1.1095076409298189e-08, "epoch": 1.9576488015082143, "percentage": 97.88, "elapsed_time": "2 days, 14:29:30", "remaining_time": "1:21:06", "throughput": 5652.86, "total_tokens": 1271727712}
+{"current_steps": 29080, "total_steps": 29704, "loss": 0.2864542484283447, "lr": 1.0919726820803889e-08, "epoch": 1.9579854565041745, "percentage": 97.9, "elapsed_time": "2 days, 14:30:07", "remaining_time": "1:20:28", "throughput": 5652.86, "total_tokens": 1271937480}
+{"current_steps": 29085, "total_steps": 29704, "loss": 0.2920374870300293, "lr": 1.0745772412698585e-08, "epoch": 1.9583221115001348, "percentage": 97.92, "elapsed_time": "2 days, 14:30:47", "remaining_time": "1:19:49", "throughput": 5652.82, "total_tokens": 1272150888}
+{"current_steps": 29090, "total_steps": 29704, "loss": 0.28984389305114744, "lr": 1.057321323362892e-08, "epoch": 1.9586587664960948, "percentage": 97.93, "elapsed_time": "2 days, 14:31:28", "remaining_time": "1:19:10", "throughput": 5652.82, "total_tokens": 1272385000}
+{"current_steps": 29095, "total_steps": 29704, "loss": 0.2945473432540894, "lr": 1.0402049331849629e-08, "epoch": 1.9589954214920549, "percentage": 97.95, "elapsed_time": "2 days, 14:32:00", "remaining_time": "1:18:32", "throughput": 5653.02, "total_tokens": 1272608520}
+{"current_steps": 29100, "total_steps": 29704, "loss": 0.30609171390533446, "lr": 1.0232280755226865e-08, "epoch": 1.959332076488015, "percentage": 97.97, "elapsed_time": "2 days, 14:32:44", "remaining_time": "1:17:53", "throughput": 5652.87, "total_tokens": 1272824152}
+{"current_steps": 29105, "total_steps": 29704, "loss": 0.2889437675476074, "lr": 1.006390755123543e-08, "epoch": 1.9596687314839754, "percentage": 97.98, "elapsed_time": "2 days, 14:33:24", "remaining_time": "1:17:14", "throughput": 5652.81, "total_tokens": 1273039232}
+{"current_steps": 29110, "total_steps": 29704, "loss": 0.3118439674377441, "lr": 9.89692976696044e-09, "epoch": 1.9600053864799354, "percentage": 98.0, "elapsed_time": "2 days, 14:34:10", "remaining_time": "1:16:36", "throughput": 5652.62, "total_tokens": 1273253968}
+{"current_steps": 29115, "total_steps": 29704, "loss": 0.31283159255981446, "lr": 9.731347449097316e-09, "epoch": 1.9603420414758954, "percentage": 98.02, "elapsed_time": "2 days, 14:34:52", "remaining_time": "1:15:57", "throughput": 5652.53, "total_tokens": 1273474224}
+{"current_steps": 29120, "total_steps": 29704, "loss": 0.30971055030822753, "lr": 9.56716064394958e-09, "epoch": 1.9606786964718557, "percentage": 98.03, "elapsed_time": "2 days, 14:35:24", "remaining_time": "1:15:18", "throughput": 5652.74, "total_tokens": 1273701776}
+{"current_steps": 29125, "total_steps": 29704, "loss": 0.287065315246582, "lr": 9.404369397432166e-09, "epoch": 1.961015351467816, "percentage": 98.05, "elapsed_time": "2 days, 14:36:00", "remaining_time": "1:14:40", "throughput": 5652.8, "total_tokens": 1273920400}
+{"current_steps": 29130, "total_steps": 29704, "loss": 0.2945993900299072, "lr": 9.242973755068663e-09, "epoch": 1.961352006463776, "percentage": 98.07, "elapsed_time": "2 days, 14:36:36", "remaining_time": "1:14:01", "throughput": 5652.91, "total_tokens": 1274143544}
+{"current_steps": 29135, "total_steps": 29704, "loss": 0.2969123363494873, "lr": 9.082973761993518e-09, "epoch": 1.961688661459736, "percentage": 98.08, "elapsed_time": "2 days, 14:37:17", "remaining_time": "1:13:22", "throughput": 5652.8, "total_tokens": 1274356376}
+{"current_steps": 29140, "total_steps": 29704, "loss": 0.3293659448623657, "lr": 8.924369462949834e-09, "epoch": 1.9620253164556962, "percentage": 98.1, "elapsed_time": "2 days, 14:37:58", "remaining_time": "1:12:44", "throughput": 5652.69, "total_tokens": 1274559280}
+{"current_steps": 29145, "total_steps": 29704, "loss": 0.2915841817855835, "lr": 8.767160902291016e-09, "epoch": 1.9623619714516565, "percentage": 98.12, "elapsed_time": "2 days, 14:38:34", "remaining_time": "1:12:05", "throughput": 5652.74, "total_tokens": 1274776792}
+{"current_steps": 29150, "total_steps": 29704, "loss": 0.30088181495666505, "lr": 8.61134812397968e-09, "epoch": 1.9626986264476165, "percentage": 98.13, "elapsed_time": "2 days, 14:39:11", "remaining_time": "1:11:26", "throughput": 5652.8, "total_tokens": 1274996960}
+{"current_steps": 29155, "total_steps": 29704, "loss": 0.2771022796630859, "lr": 8.45693117158819e-09, "epoch": 1.9630352814435765, "percentage": 98.15, "elapsed_time": "2 days, 14:39:47", "remaining_time": "1:10:47", "throughput": 5652.81, "total_tokens": 1275203528}
+{"current_steps": 29160, "total_steps": 29704, "loss": 0.2890202045440674, "lr": 8.303910088299228e-09, "epoch": 1.9633719364395368, "percentage": 98.17, "elapsed_time": "2 days, 14:40:27", "remaining_time": "1:10:09", "throughput": 5652.76, "total_tokens": 1275420104}
+{"current_steps": 29165, "total_steps": 29704, "loss": 0.297850513458252, "lr": 8.152284916904674e-09, "epoch": 1.963708591435497, "percentage": 98.19, "elapsed_time": "2 days, 14:41:05", "remaining_time": "1:09:30", "throughput": 5652.83, "total_tokens": 1275648864}
+{"current_steps": 29170, "total_steps": 29704, "loss": 0.2845508098602295, "lr": 8.002055699805612e-09, "epoch": 1.964045246431457, "percentage": 98.2, "elapsed_time": "2 days, 14:41:51", "remaining_time": "1:08:51", "throughput": 5652.61, "total_tokens": 1275859328}
+{"current_steps": 29175, "total_steps": 29704, "loss": 0.28490164279937746, "lr": 7.853222479013434e-09, "epoch": 1.964381901427417, "percentage": 98.22, "elapsed_time": "2 days, 14:42:27", "remaining_time": "1:08:13", "throughput": 5652.77, "total_tokens": 1276098304}
+{"current_steps": 29180, "total_steps": 29704, "loss": 0.29223477840423584, "lr": 7.705785296148737e-09, "epoch": 1.9647185564233773, "percentage": 98.24, "elapsed_time": "2 days, 14:43:02", "remaining_time": "1:07:34", "throughput": 5652.83, "total_tokens": 1276313944}
+{"current_steps": 29185, "total_steps": 29704, "loss": 0.3048643350601196, "lr": 7.55974419244132e-09, "epoch": 1.9650552114193376, "percentage": 98.25, "elapsed_time": "2 days, 14:43:35", "remaining_time": "1:06:55", "throughput": 5653.01, "total_tokens": 1276534976}
+{"current_steps": 29190, "total_steps": 29704, "loss": 0.2795270919799805, "lr": 7.415099208732402e-09, "epoch": 1.9653918664152976, "percentage": 98.27, "elapsed_time": "2 days, 14:44:10", "remaining_time": "1:06:16", "throughput": 5653.09, "total_tokens": 1276756464}
+{"current_steps": 29195, "total_steps": 29704, "loss": 0.3065217971801758, "lr": 7.2718503854707355e-09, "epoch": 1.9657285214112576, "percentage": 98.29, "elapsed_time": "2 days, 14:44:51", "remaining_time": "1:05:38", "throughput": 5653.07, "total_tokens": 1276983008}
+{"current_steps": 29200, "total_steps": 29704, "loss": 0.26297545433044434, "lr": 7.129997762715391e-09, "epoch": 1.966065176407218, "percentage": 98.3, "elapsed_time": "2 days, 14:45:29", "remaining_time": "1:04:59", "throughput": 5653.11, "total_tokens": 1277205144}
+{"current_steps": 29205, "total_steps": 29704, "loss": 0.2864915132522583, "lr": 6.98954138013519e-09, "epoch": 1.9664018314031781, "percentage": 98.32, "elapsed_time": "2 days, 14:46:07", "remaining_time": "1:04:20", "throughput": 5653.09, "total_tokens": 1277414784}
+{"current_steps": 29210, "total_steps": 29704, "loss": 0.3049713373184204, "lr": 6.850481277008159e-09, "epoch": 1.9667384863991382, "percentage": 98.34, "elapsed_time": "2 days, 14:46:54", "remaining_time": "1:03:42", "throughput": 5652.93, "total_tokens": 1277643008}
+{"current_steps": 29215, "total_steps": 29704, "loss": 0.32462234497070314, "lr": 6.712817492222079e-09, "epoch": 1.9670751413950982, "percentage": 98.35, "elapsed_time": "2 days, 14:47:34", "remaining_time": "1:03:03", "throughput": 5652.91, "total_tokens": 1277862488}
+{"current_steps": 29220, "total_steps": 29704, "loss": 0.29960381984710693, "lr": 6.576550064273934e-09, "epoch": 1.9674117963910585, "percentage": 98.37, "elapsed_time": "2 days, 14:48:07", "remaining_time": "1:02:24", "throughput": 5653.05, "total_tokens": 1278083144}
+{"current_steps": 29225, "total_steps": 29704, "loss": 0.30608859062194826, "lr": 6.44167903127102e-09, "epoch": 1.9677484513870187, "percentage": 98.39, "elapsed_time": "2 days, 14:48:48", "remaining_time": "1:01:46", "throughput": 5653.01, "total_tokens": 1278306920}
+{"current_steps": 29230, "total_steps": 29704, "loss": 0.2798271656036377, "lr": 6.3082044309287215e-09, "epoch": 1.9680851063829787, "percentage": 98.4, "elapsed_time": "2 days, 14:49:23", "remaining_time": "1:01:07", "throughput": 5653.13, "total_tokens": 1278530936}
+{"current_steps": 29235, "total_steps": 29704, "loss": 0.29613399505615234, "lr": 6.176126300573848e-09, "epoch": 1.9684217613789388, "percentage": 98.42, "elapsed_time": "2 days, 14:50:04", "remaining_time": "1:00:28", "throughput": 5653.07, "total_tokens": 1278747232}
+{"current_steps": 29240, "total_steps": 29704, "loss": 0.28275036811828613, "lr": 6.045444677140744e-09, "epoch": 1.968758416374899, "percentage": 98.44, "elapsed_time": "2 days, 14:50:41", "remaining_time": "0:59:50", "throughput": 5653.14, "total_tokens": 1278973944}
+{"current_steps": 29245, "total_steps": 29704, "loss": 0.2926478385925293, "lr": 5.91615959717462e-09, "epoch": 1.9690950713708593, "percentage": 98.45, "elapsed_time": "2 days, 14:51:20", "remaining_time": "0:59:11", "throughput": 5653.15, "total_tokens": 1279195664}
+{"current_steps": 29250, "total_steps": 29704, "loss": 0.27014522552490233, "lr": 5.78827109682989e-09, "epoch": 1.9694317263668193, "percentage": 98.47, "elapsed_time": "2 days, 14:51:54", "remaining_time": "0:58:32", "throughput": 5653.27, "total_tokens": 1279416864}
+{"current_steps": 29255, "total_steps": 29704, "loss": 0.2848729848861694, "lr": 5.661779211869056e-09, "epoch": 1.9697683813627793, "percentage": 98.49, "elapsed_time": "2 days, 14:52:32", "remaining_time": "0:57:54", "throughput": 5653.26, "total_tokens": 1279631848}
+{"current_steps": 29260, "total_steps": 29704, "loss": 0.3105973243713379, "lr": 5.536683977666601e-09, "epoch": 1.9701050363587396, "percentage": 98.51, "elapsed_time": "2 days, 14:53:05", "remaining_time": "0:57:15", "throughput": 5653.47, "total_tokens": 1279865120}
+{"current_steps": 29265, "total_steps": 29704, "loss": 0.28957743644714357, "lr": 5.412985429203988e-09, "epoch": 1.9704416913546998, "percentage": 98.52, "elapsed_time": "2 days, 14:53:43", "remaining_time": "0:56:36", "throughput": 5653.47, "total_tokens": 1280078376}
+{"current_steps": 29270, "total_steps": 29704, "loss": 0.2784770965576172, "lr": 5.290683601073543e-09, "epoch": 1.9707783463506598, "percentage": 98.54, "elapsed_time": "2 days, 14:54:19", "remaining_time": "0:55:57", "throughput": 5653.56, "total_tokens": 1280304192}
+{"current_steps": 29275, "total_steps": 29704, "loss": 0.299863862991333, "lr": 5.169778527476243e-09, "epoch": 1.9711150013466199, "percentage": 98.56, "elapsed_time": "2 days, 14:54:58", "remaining_time": "0:55:19", "throughput": 5653.59, "total_tokens": 1280532272}
+{"current_steps": 29280, "total_steps": 29704, "loss": 0.31504406929016116, "lr": 5.050270242222821e-09, "epoch": 1.9714516563425801, "percentage": 98.57, "elapsed_time": "2 days, 14:55:36", "remaining_time": "0:54:40", "throughput": 5653.52, "total_tokens": 1280731456}
+{"current_steps": 29285, "total_steps": 29704, "loss": 0.2948625087738037, "lr": 4.932158778733765e-09, "epoch": 1.9717883113385404, "percentage": 98.59, "elapsed_time": "2 days, 14:56:20", "remaining_time": "0:54:01", "throughput": 5653.42, "total_tokens": 1280954184}
+{"current_steps": 29290, "total_steps": 29704, "loss": 0.3150008201599121, "lr": 4.8154441700387636e-09, "epoch": 1.9721249663345004, "percentage": 98.61, "elapsed_time": "2 days, 14:56:53", "remaining_time": "0:53:23", "throughput": 5653.48, "total_tokens": 1281153920}
+{"current_steps": 29295, "total_steps": 29704, "loss": 0.31816887855529785, "lr": 4.7001264487761545e-09, "epoch": 1.9724616213304604, "percentage": 98.62, "elapsed_time": "2 days, 14:57:33", "remaining_time": "0:52:44", "throughput": 5653.48, "total_tokens": 1281377960}
+{"current_steps": 29300, "total_steps": 29704, "loss": 0.3077888250350952, "lr": 4.586205647194031e-09, "epoch": 1.9727982763264207, "percentage": 98.64, "elapsed_time": "2 days, 14:58:11", "remaining_time": "0:52:05", "throughput": 5653.49, "total_tokens": 1281598760}
+{"current_steps": 29305, "total_steps": 29704, "loss": 0.3019340991973877, "lr": 4.4736817971507974e-09, "epoch": 1.973134931322381, "percentage": 98.66, "elapsed_time": "2 days, 14:58:57", "remaining_time": "0:51:27", "throughput": 5653.29, "total_tokens": 1281813592}
+{"current_steps": 29310, "total_steps": 29704, "loss": 0.30186340808868406, "lr": 4.362554930112395e-09, "epoch": 1.973471586318341, "percentage": 98.67, "elapsed_time": "2 days, 14:59:35", "remaining_time": "0:50:48", "throughput": 5653.29, "total_tokens": 1282029480}
+{"current_steps": 29315, "total_steps": 29704, "loss": 0.2891343593597412, "lr": 4.252825077155631e-09, "epoch": 1.973808241314301, "percentage": 98.69, "elapsed_time": "2 days, 15:00:14", "remaining_time": "0:50:09", "throughput": 5653.29, "total_tokens": 1282250600}
+{"current_steps": 29320, "total_steps": 29704, "loss": 0.29642043113708494, "lr": 4.1444922689665156e-09, "epoch": 1.9741448963102612, "percentage": 98.71, "elapsed_time": "2 days, 15:00:55", "remaining_time": "0:49:31", "throughput": 5653.22, "total_tokens": 1282463216}
+{"current_steps": 29325, "total_steps": 29704, "loss": 0.2692773103713989, "lr": 4.037556535839149e-09, "epoch": 1.9744815513062215, "percentage": 98.72, "elapsed_time": "2 days, 15:01:27", "remaining_time": "0:48:52", "throughput": 5653.39, "total_tokens": 1282685880}
+{"current_steps": 29330, "total_steps": 29704, "loss": 0.28195557594299314, "lr": 3.932017907677943e-09, "epoch": 1.9748182063021815, "percentage": 98.74, "elapsed_time": "2 days, 15:02:11", "remaining_time": "0:48:13", "throughput": 5653.32, "total_tokens": 1282915496}
+{"current_steps": 29335, "total_steps": 29704, "loss": 0.310672664642334, "lr": 3.827876413997067e-09, "epoch": 1.9751548612981416, "percentage": 98.76, "elapsed_time": "2 days, 15:02:42", "remaining_time": "0:47:34", "throughput": 5653.41, "total_tokens": 1283112440}
+{"current_steps": 29340, "total_steps": 29704, "loss": 0.2937079668045044, "lr": 3.725132083918781e-09, "epoch": 1.9754915162941018, "percentage": 98.77, "elapsed_time": "2 days, 15:03:24", "remaining_time": "0:46:56", "throughput": 5653.29, "total_tokens": 1283319208}
+{"current_steps": 29345, "total_steps": 29704, "loss": 0.3103309631347656, "lr": 3.623784946175102e-09, "epoch": 1.975828171290062, "percentage": 98.79, "elapsed_time": "2 days, 15:04:04", "remaining_time": "0:46:17", "throughput": 5653.17, "total_tokens": 1283519312}
+{"current_steps": 29350, "total_steps": 29704, "loss": 0.28435301780700684, "lr": 3.5238350291083577e-09, "epoch": 1.976164826286022, "percentage": 98.81, "elapsed_time": "2 days, 15:04:45", "remaining_time": "0:45:38", "throughput": 5653.06, "total_tokens": 1283726744}
+{"current_steps": 29355, "total_steps": 29704, "loss": 0.31540963649749754, "lr": 3.4252823606678588e-09, "epoch": 1.976501481281982, "percentage": 98.83, "elapsed_time": "2 days, 15:05:18", "remaining_time": "0:45:00", "throughput": 5653.23, "total_tokens": 1283952928}
+{"current_steps": 29360, "total_steps": 29704, "loss": 0.3130497455596924, "lr": 3.328126968414336e-09, "epoch": 1.9768381362779424, "percentage": 98.84, "elapsed_time": "2 days, 15:05:54", "remaining_time": "0:44:21", "throughput": 5653.19, "total_tokens": 1284147936}
+{"current_steps": 29365, "total_steps": 29704, "loss": 0.31641981601715086, "lr": 3.232368879517167e-09, "epoch": 1.9771747912739026, "percentage": 98.86, "elapsed_time": "2 days, 15:06:40", "remaining_time": "0:43:42", "throughput": 5653.02, "total_tokens": 1284366488}
+{"current_steps": 29370, "total_steps": 29704, "loss": 0.27493464946746826, "lr": 3.1380081207543766e-09, "epoch": 1.9775114462698626, "percentage": 98.88, "elapsed_time": "2 days, 15:07:16", "remaining_time": "0:43:04", "throughput": 5653.05, "total_tokens": 1284580288}
+{"current_steps": 29375, "total_steps": 29704, "loss": 0.300215482711792, "lr": 3.0450447185137454e-09, "epoch": 1.9778481012658227, "percentage": 98.89, "elapsed_time": "2 days, 15:07:56", "remaining_time": "0:42:25", "throughput": 5652.98, "total_tokens": 1284790176}
+{"current_steps": 29380, "total_steps": 29704, "loss": 0.29309883117675783, "lr": 2.953478698792256e-09, "epoch": 1.978184756261783, "percentage": 98.91, "elapsed_time": "2 days, 15:08:31", "remaining_time": "0:41:46", "throughput": 5653.11, "total_tokens": 1285019536}
+{"current_steps": 29385, "total_steps": 29704, "loss": 0.28349852561950684, "lr": 2.8633100871960917e-09, "epoch": 1.9785214112577432, "percentage": 98.93, "elapsed_time": "2 days, 15:09:14", "remaining_time": "0:41:08", "throughput": 5652.97, "total_tokens": 1285225968}
+{"current_steps": 29390, "total_steps": 29704, "loss": 0.3020082950592041, "lr": 2.774538908940638e-09, "epoch": 1.9788580662537032, "percentage": 98.94, "elapsed_time": "2 days, 15:09:51", "remaining_time": "0:40:29", "throughput": 5652.94, "total_tokens": 1285433856}
+{"current_steps": 29395, "total_steps": 29704, "loss": 0.2933856248855591, "lr": 2.687165188849927e-09, "epoch": 1.9791947212496632, "percentage": 98.96, "elapsed_time": "2 days, 15:10:23", "remaining_time": "0:39:50", "throughput": 5653.16, "total_tokens": 1285662136}
+{"current_steps": 29400, "total_steps": 29704, "loss": 0.29178383350372317, "lr": 2.6011889513588574e-09, "epoch": 1.9795313762456235, "percentage": 98.98, "elapsed_time": "2 days, 15:11:01", "remaining_time": "0:39:11", "throughput": 5653.18, "total_tokens": 1285879472}
+{"current_steps": 29405, "total_steps": 29704, "loss": 0.31009225845336913, "lr": 2.5166102205093077e-09, "epoch": 1.9798680312415837, "percentage": 98.99, "elapsed_time": "2 days, 15:11:35", "remaining_time": "0:38:33", "throughput": 5653.31, "total_tokens": 1286104184}
+{"current_steps": 29410, "total_steps": 29704, "loss": 0.2966364860534668, "lr": 2.4334290199540254e-09, "epoch": 1.9802046862375438, "percentage": 99.01, "elapsed_time": "2 days, 15:12:13", "remaining_time": "0:37:54", "throughput": 5653.34, "total_tokens": 1286326480}
+{"current_steps": 29415, "total_steps": 29704, "loss": 0.27869043350219724, "lr": 2.351645372953848e-09, "epoch": 1.9805413412335038, "percentage": 99.03, "elapsed_time": "2 days, 15:12:54", "remaining_time": "0:37:15", "throughput": 5653.26, "total_tokens": 1286539544}
+{"current_steps": 29420, "total_steps": 29704, "loss": 0.292054557800293, "lr": 2.2712593023804795e-09, "epoch": 1.980877996229464, "percentage": 99.04, "elapsed_time": "2 days, 15:13:23", "remaining_time": "0:36:37", "throughput": 5653.48, "total_tokens": 1286749296}
+{"current_steps": 29425, "total_steps": 29704, "loss": 0.3023597240447998, "lr": 2.1922708307120512e-09, "epoch": 1.9812146512254243, "percentage": 99.06, "elapsed_time": "2 days, 15:14:06", "remaining_time": "0:35:58", "throughput": 5653.36, "total_tokens": 1286966040}
+{"current_steps": 29430, "total_steps": 29704, "loss": 0.3239266872406006, "lr": 2.1146799800386698e-09, "epoch": 1.9815513062213843, "percentage": 99.08, "elapsed_time": "2 days, 15:14:38", "remaining_time": "0:35:19", "throughput": 5653.54, "total_tokens": 1287186672}
+{"current_steps": 29435, "total_steps": 29704, "loss": 0.278280234336853, "lr": 2.0384867720579794e-09, "epoch": 1.9818879612173443, "percentage": 99.09, "elapsed_time": "2 days, 15:15:17", "remaining_time": "0:34:41", "throughput": 5653.47, "total_tokens": 1287391672}
+{"current_steps": 29440, "total_steps": 29704, "loss": 0.30411410331726074, "lr": 1.96369122807738e-09, "epoch": 1.9822246162133046, "percentage": 99.11, "elapsed_time": "2 days, 15:15:52", "remaining_time": "0:34:02", "throughput": 5653.59, "total_tokens": 1287620424}
+{"current_steps": 29445, "total_steps": 29704, "loss": 0.2814857721328735, "lr": 1.8902933690123637e-09, "epoch": 1.9825612712092648, "percentage": 99.13, "elapsed_time": "2 days, 15:16:23", "remaining_time": "0:33:23", "throughput": 5653.79, "total_tokens": 1287839296}
+{"current_steps": 29450, "total_steps": 29704, "loss": 0.3011634826660156, "lr": 1.8182932153892884e-09, "epoch": 1.9828979262052249, "percentage": 99.14, "elapsed_time": "2 days, 15:17:01", "remaining_time": "0:32:44", "throughput": 5653.81, "total_tokens": 1288057400}
+{"current_steps": 29455, "total_steps": 29704, "loss": 0.29094853401184084, "lr": 1.7476907873426041e-09, "epoch": 1.983234581201185, "percentage": 99.16, "elapsed_time": "2 days, 15:17:38", "remaining_time": "0:32:06", "throughput": 5653.88, "total_tokens": 1288283336}
+{"current_steps": 29460, "total_steps": 29704, "loss": 0.29619736671447755, "lr": 1.6784861046159617e-09, "epoch": 1.9835712361971451, "percentage": 99.18, "elapsed_time": "2 days, 15:18:24", "remaining_time": "0:31:27", "throughput": 5653.7, "total_tokens": 1288507000}
+{"current_steps": 29465, "total_steps": 29704, "loss": 0.3063995599746704, "lr": 1.610679186561659e-09, "epoch": 1.9839078911931054, "percentage": 99.2, "elapsed_time": "2 days, 15:19:04", "remaining_time": "0:30:48", "throughput": 5653.68, "total_tokens": 1288727264}
+{"current_steps": 29470, "total_steps": 29704, "loss": 0.2689430952072144, "lr": 1.5442700521428599e-09, "epoch": 1.9842445461890654, "percentage": 99.21, "elapsed_time": "2 days, 15:19:43", "remaining_time": "0:30:10", "throughput": 5653.64, "total_tokens": 1288937368}
+{"current_steps": 29475, "total_steps": 29704, "loss": 0.29396581649780273, "lr": 1.4792587199297103e-09, "epoch": 1.9845812011850255, "percentage": 99.23, "elapsed_time": "2 days, 15:20:25", "remaining_time": "0:29:31", "throughput": 5653.58, "total_tokens": 1289161136}
+{"current_steps": 29480, "total_steps": 29704, "loss": 0.310637903213501, "lr": 1.4156452081026672e-09, "epoch": 1.9849178561809857, "percentage": 99.25, "elapsed_time": "2 days, 15:21:06", "remaining_time": "0:28:52", "throughput": 5653.57, "total_tokens": 1289388008}
+{"current_steps": 29485, "total_steps": 29704, "loss": 0.29576449394226073, "lr": 1.3534295344513893e-09, "epoch": 1.985254511176946, "percentage": 99.26, "elapsed_time": "2 days, 15:21:50", "remaining_time": "0:28:14", "throughput": 5653.47, "total_tokens": 1289614784}
+{"current_steps": 29490, "total_steps": 29704, "loss": 0.2955544710159302, "lr": 1.2926117163741813e-09, "epoch": 1.985591166172906, "percentage": 99.28, "elapsed_time": "2 days, 15:22:29", "remaining_time": "0:27:35", "throughput": 5653.47, "total_tokens": 1289834936}
+{"current_steps": 29495, "total_steps": 29704, "loss": 0.3009219169616699, "lr": 1.233191770877995e-09, "epoch": 1.985927821168866, "percentage": 99.3, "elapsed_time": "2 days, 15:23:08", "remaining_time": "0:26:56", "throughput": 5653.47, "total_tokens": 1290054672}
+{"current_steps": 29500, "total_steps": 29704, "loss": 0.2817996978759766, "lr": 1.1751697145800934e-09, "epoch": 1.9862644761648263, "percentage": 99.31, "elapsed_time": "2 days, 15:23:44", "remaining_time": "0:26:18", "throughput": 5653.57, "total_tokens": 1290281064}
+{"current_steps": 29505, "total_steps": 29704, "loss": 0.3367082357406616, "lr": 1.1185455637052756e-09, "epoch": 1.9866011311607865, "percentage": 99.33, "elapsed_time": "2 days, 15:24:25", "remaining_time": "0:25:39", "throughput": 5653.54, "total_tokens": 1290509624}
+{"current_steps": 29510, "total_steps": 29704, "loss": 0.3116638660430908, "lr": 1.063319334089763e-09, "epoch": 1.9869377861567465, "percentage": 99.35, "elapsed_time": "2 days, 15:25:08", "remaining_time": "0:25:00", "throughput": 5653.42, "total_tokens": 1290724632}
+{"current_steps": 29515, "total_steps": 29704, "loss": 0.2789448261260986, "lr": 1.0094910411762027e-09, "epoch": 1.9872744411527066, "percentage": 99.36, "elapsed_time": "2 days, 15:25:53", "remaining_time": "0:24:22", "throughput": 5653.29, "total_tokens": 1290947232}
+{"current_steps": 29520, "total_steps": 29704, "loss": 0.31690254211425783, "lr": 9.570607000175536e-10, "epoch": 1.9876110961486668, "percentage": 99.38, "elapsed_time": "2 days, 15:26:36", "remaining_time": "0:23:43", "throughput": 5653.18, "total_tokens": 1291165048}
+{"current_steps": 29525, "total_steps": 29704, "loss": 0.29549479484558105, "lr": 9.060283252765311e-10, "epoch": 1.987947751144627, "percentage": 99.4, "elapsed_time": "2 days, 15:27:17", "remaining_time": "0:23:04", "throughput": 5653.1, "total_tokens": 1291376632}
+{"current_steps": 29530, "total_steps": 29704, "loss": 0.2834463357925415, "lr": 8.563939312233871e-10, "epoch": 1.988284406140587, "percentage": 99.41, "elapsed_time": "2 days, 15:27:55", "remaining_time": "0:22:26", "throughput": 5653.08, "total_tokens": 1291592160}
+{"current_steps": 29535, "total_steps": 29704, "loss": 0.29933602809906007, "lr": 8.081575317386847e-10, "epoch": 1.9886210611365471, "percentage": 99.43, "elapsed_time": "2 days, 15:28:32", "remaining_time": "0:21:47", "throughput": 5653.13, "total_tokens": 1291812360}
+{"current_steps": 29540, "total_steps": 29704, "loss": 0.30489234924316405, "lr": 7.613191403116338e-10, "epoch": 1.9889577161325074, "percentage": 99.45, "elapsed_time": "2 days, 15:29:05", "remaining_time": "0:21:08", "throughput": 5653.31, "total_tokens": 1292039232}
+{"current_steps": 29545, "total_steps": 29704, "loss": 0.3050061225891113, "lr": 7.158787700400905e-10, "epoch": 1.9892943711284676, "percentage": 99.46, "elapsed_time": "2 days, 15:29:48", "remaining_time": "0:20:30", "throughput": 5653.25, "total_tokens": 1292268968}
+{"current_steps": 29550, "total_steps": 29704, "loss": 0.2916685104370117, "lr": 6.718364336316674e-10, "epoch": 1.9896310261244277, "percentage": 99.48, "elapsed_time": "2 days, 15:30:27", "remaining_time": "0:19:51", "throughput": 5653.27, "total_tokens": 1292491832}
+{"current_steps": 29555, "total_steps": 29704, "loss": 0.29037137031555177, "lr": 6.291921434026238e-10, "epoch": 1.9899676811203877, "percentage": 99.5, "elapsed_time": "2 days, 15:31:04", "remaining_time": "0:19:12", "throughput": 5653.32, "total_tokens": 1292713384}
+{"current_steps": 29560, "total_steps": 29704, "loss": 0.26992483139038087, "lr": 5.879459112778651e-10, "epoch": 1.990304336116348, "percentage": 99.52, "elapsed_time": "2 days, 15:31:47", "remaining_time": "0:18:34", "throughput": 5653.15, "total_tokens": 1292916816}
+{"current_steps": 29565, "total_steps": 29704, "loss": 0.3113460302352905, "lr": 5.480977487920536e-10, "epoch": 1.9906409911123082, "percentage": 99.53, "elapsed_time": "2 days, 15:32:24", "remaining_time": "0:17:55", "throughput": 5653.26, "total_tokens": 1293152760}
+{"current_steps": 29570, "total_steps": 29704, "loss": 0.29065916538238523, "lr": 5.096476670890526e-10, "epoch": 1.9909776461082682, "percentage": 99.55, "elapsed_time": "2 days, 15:33:06", "remaining_time": "0:17:16", "throughput": 5653.19, "total_tokens": 1293370744}
+{"current_steps": 29575, "total_steps": 29704, "loss": 0.3043833255767822, "lr": 4.725956769202622e-10, "epoch": 1.9913143011042282, "percentage": 99.57, "elapsed_time": "2 days, 15:33:41", "remaining_time": "0:16:38", "throughput": 5653.22, "total_tokens": 1293580568}
+{"current_steps": 29580, "total_steps": 29704, "loss": 0.3031582832336426, "lr": 4.369417886485039e-10, "epoch": 1.9916509561001885, "percentage": 99.58, "elapsed_time": "2 days, 15:34:19", "remaining_time": "0:15:59", "throughput": 5653.21, "total_tokens": 1293788160}
+{"current_steps": 29585, "total_steps": 29704, "loss": 0.293507981300354, "lr": 4.026860122430254e-10, "epoch": 1.9919876110961487, "percentage": 99.6, "elapsed_time": "2 days, 15:34:51", "remaining_time": "0:15:20", "throughput": 5653.38, "total_tokens": 1294011512}
+{"current_steps": 29590, "total_steps": 29704, "loss": 0.29284181594848635, "lr": 3.698283572839412e-10, "epoch": 1.9923242660921088, "percentage": 99.62, "elapsed_time": "2 days, 15:35:30", "remaining_time": "0:14:41", "throughput": 5653.4, "total_tokens": 1294233288}
+{"current_steps": 29595, "total_steps": 29704, "loss": 0.2951065540313721, "lr": 3.38368832959457e-10, "epoch": 1.9926609210880688, "percentage": 99.63, "elapsed_time": "2 days, 15:36:06", "remaining_time": "0:14:03", "throughput": 5653.4, "total_tokens": 1294439056}
+{"current_steps": 29600, "total_steps": 29704, "loss": 0.30010027885437013, "lr": 3.0830744806753523e-10, "epoch": 1.992997576084029, "percentage": 99.65, "elapsed_time": "2 days, 15:36:42", "remaining_time": "0:13:24", "throughput": 5653.46, "total_tokens": 1294656608}
+{"current_steps": 29605, "total_steps": 29704, "loss": 0.29233975410461427, "lr": 2.7964421101478457e-10, "epoch": 1.9933342310799893, "percentage": 99.67, "elapsed_time": "2 days, 15:37:13", "remaining_time": "0:12:45", "throughput": 5653.61, "total_tokens": 1294866112}
+{"current_steps": 29610, "total_steps": 29704, "loss": 0.3044607162475586, "lr": 2.5237912981646016e-10, "epoch": 1.9936708860759493, "percentage": 99.68, "elapsed_time": "2 days, 15:37:48", "remaining_time": "0:12:07", "throughput": 5653.7, "total_tokens": 1295082752}
+{"current_steps": 29615, "total_steps": 29704, "loss": 0.2928770542144775, "lr": 2.265122120975738e-10, "epoch": 1.9940075410719094, "percentage": 99.7, "elapsed_time": "2 days, 15:38:28", "remaining_time": "0:11:28", "throughput": 5653.63, "total_tokens": 1295294464}
+{"current_steps": 29620, "total_steps": 29704, "loss": 0.2951038360595703, "lr": 2.020434650906733e-10, "epoch": 1.9943441960678696, "percentage": 99.72, "elapsed_time": "2 days, 15:39:04", "remaining_time": "0:10:49", "throughput": 5653.71, "total_tokens": 1295518440}
+{"current_steps": 29625, "total_steps": 29704, "loss": 0.28854942321777344, "lr": 1.7897289563972854e-10, "epoch": 1.9946808510638299, "percentage": 99.73, "elapsed_time": "2 days, 15:39:41", "remaining_time": "0:10:11", "throughput": 5653.81, "total_tokens": 1295752200}
+{"current_steps": 29630, "total_steps": 29704, "loss": 0.28860130310058596, "lr": 1.5730051019569037e-10, "epoch": 1.99501750605979, "percentage": 99.75, "elapsed_time": "2 days, 15:40:17", "remaining_time": "0:09:32", "throughput": 5653.87, "total_tokens": 1295966672}
+{"current_steps": 29635, "total_steps": 29704, "loss": 0.29309055805206297, "lr": 1.3702631481926633e-10, "epoch": 1.99535416105575, "percentage": 99.77, "elapsed_time": "2 days, 15:40:53", "remaining_time": "0:08:53", "throughput": 5653.95, "total_tokens": 1296186448}
+{"current_steps": 29640, "total_steps": 29704, "loss": 0.28788046836853026, "lr": 1.1815031517981023e-10, "epoch": 1.9956908160517102, "percentage": 99.78, "elapsed_time": "2 days, 15:41:24", "remaining_time": "0:08:15", "throughput": 5654.05, "total_tokens": 1296387216}
+{"current_steps": 29645, "total_steps": 29704, "loss": 0.28290114402770994, "lr": 1.0067251655643262e-10, "epoch": 1.9960274710476704, "percentage": 99.8, "elapsed_time": "2 days, 15:42:00", "remaining_time": "0:07:36", "throughput": 5654.16, "total_tokens": 1296614056}
+{"current_steps": 29650, "total_steps": 29704, "loss": 0.2950653076171875, "lr": 8.459292383633522e-11, "epoch": 1.9963641260436304, "percentage": 99.82, "elapsed_time": "2 days, 15:42:43", "remaining_time": "0:06:57", "throughput": 5653.97, "total_tokens": 1296815776}
+{"current_steps": 29655, "total_steps": 29704, "loss": 0.2833098888397217, "lr": 6.991154151592128e-11, "epoch": 1.9967007810395905, "percentage": 99.84, "elapsed_time": "2 days, 15:43:23", "remaining_time": "0:06:19", "throughput": 5653.87, "total_tokens": 1297021184}
+{"current_steps": 29660, "total_steps": 29704, "loss": 0.3091259956359863, "lr": 5.662837370135066e-11, "epoch": 1.9970374360355507, "percentage": 99.85, "elapsed_time": "2 days, 15:44:01", "remaining_time": "0:05:40", "throughput": 5653.94, "total_tokens": 1297247832}
+{"current_steps": 29665, "total_steps": 29704, "loss": 0.3012515068054199, "lr": 4.47434241074296e-11, "epoch": 1.997374091031511, "percentage": 99.87, "elapsed_time": "2 days, 15:44:38", "remaining_time": "0:05:01", "throughput": 5653.95, "total_tokens": 1297458968}
+{"current_steps": 29670, "total_steps": 29704, "loss": 0.3033904552459717, "lr": 3.4256696057055615e-11, "epoch": 1.997710746027471, "percentage": 99.89, "elapsed_time": "2 days, 15:45:22", "remaining_time": "0:04:23", "throughput": 5653.8, "total_tokens": 1297673200}
+{"current_steps": 29675, "total_steps": 29704, "loss": 0.30843775272369384, "lr": 2.5168192482882824e-11, "epoch": 1.998047401023431, "percentage": 99.9, "elapsed_time": "2 days, 15:45:56", "remaining_time": "0:03:44", "throughput": 5653.92, "total_tokens": 1297896616}
+{"current_steps": 29680, "total_steps": 29704, "loss": 0.27484722137451173, "lr": 1.747791592676684e-11, "epoch": 1.9983840560193913, "percentage": 99.92, "elapsed_time": "2 days, 15:46:29", "remaining_time": "0:03:05", "throughput": 5654.16, "total_tokens": 1298137256}
+{"current_steps": 29685, "total_steps": 29704, "loss": 0.3029062032699585, "lr": 1.1185868539209665e-11, "epoch": 1.9987207110153515, "percentage": 99.94, "elapsed_time": "2 days, 15:47:03", "remaining_time": "0:02:26", "throughput": 5654.25, "total_tokens": 1298347688}
+{"current_steps": 29690, "total_steps": 29704, "loss": 0.30252552032470703, "lr": 6.29205207991479e-12, "epoch": 1.9990573660113116, "percentage": 99.95, "elapsed_time": "2 days, 15:47:39", "remaining_time": "0:01:48", "throughput": 5654.24, "total_tokens": 1298551880}
+{"current_steps": 29695, "total_steps": 29704, "loss": 0.2877742528915405, "lr": 2.7964679166769815e-12, "epoch": 1.9993940210072716, "percentage": 99.97, "elapsed_time": "2 days, 15:48:19", "remaining_time": "0:01:09", "throughput": 5654.22, "total_tokens": 1298772440}
+{"current_steps": 29700, "total_steps": 29704, "loss": 0.26116390228271485, "lr": 6.991170281578364e-13, "epoch": 1.9997306760032318, "percentage": 99.99, "elapsed_time": "2 days, 15:48:57", "remaining_time": "0:00:30", "throughput": 5654.21, "total_tokens": 1298981184}
+{"current_steps": 29704, "total_steps": 29704, "epoch": 2.0, "percentage": 100.0, "elapsed_time": "2 days, 15:50:18", "remaining_time": "0:00:00", "throughput": 5652.95, "total_tokens": 1299153424}
diff --git a/trainer_state.json b/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..d303104e9e80d0d791589877258dc714fb4b300e
--- /dev/null
+++ b/trainer_state.json
@@ -0,0 +1,59444 @@
+{
+ "best_global_step": null,
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 2.0,
+ "eval_steps": 500,
+ "global_step": 29704,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.00033665499596014,
+ "grad_norm": 1.7158506568336664,
+ "learning_rate": 9.999999552565099e-06,
+ "loss": 0.8321187973022461,
+ "num_input_tokens_seen": 213976,
+ "step": 5,
+ "train_runtime": 48.5386,
+ "train_tokens_per_second": 4408.369
+ },
+ {
+ "epoch": 0.00067330999192028,
+ "grad_norm": 1.4989508496023307,
+ "learning_rate": 9.999997734860948e-06,
+ "loss": 0.6714894294738769,
+ "num_input_tokens_seen": 418192,
+ "step": 10,
+ "train_runtime": 79.1759,
+ "train_tokens_per_second": 5281.811
+ },
+ {
+ "epoch": 0.0010099649878804202,
+ "grad_norm": 1.3987943897146944,
+ "learning_rate": 9.999994518923374e-06,
+ "loss": 0.6414656162261962,
+ "num_input_tokens_seen": 631760,
+ "step": 15,
+ "train_runtime": 116.8887,
+ "train_tokens_per_second": 5404.799
+ },
+ {
+ "epoch": 0.00134661998384056,
+ "grad_norm": 1.3594234171992137,
+ "learning_rate": 9.999989904753278e-06,
+ "loss": 0.6119038105010987,
+ "num_input_tokens_seen": 847240,
+ "step": 20,
+ "train_runtime": 160.5101,
+ "train_tokens_per_second": 5278.422
+ },
+ {
+ "epoch": 0.0016832749798007002,
+ "grad_norm": 1.1837151081821848,
+ "learning_rate": 9.999983892351947e-06,
+ "loss": 0.5918544292449951,
+ "num_input_tokens_seen": 1067560,
+ "step": 25,
+ "train_runtime": 197.4905,
+ "train_tokens_per_second": 5405.628
+ },
+ {
+ "epoch": 0.0020199299757608404,
+ "grad_norm": 1.1504621450496095,
+ "learning_rate": 9.999976481721066e-06,
+ "loss": 0.5148830890655518,
+ "num_input_tokens_seen": 1288760,
+ "step": 30,
+ "train_runtime": 231.3357,
+ "train_tokens_per_second": 5570.951
+ },
+ {
+ "epoch": 0.0023565849717209803,
+ "grad_norm": 1.0663060499576293,
+ "learning_rate": 9.999967672862704e-06,
+ "loss": 0.554470157623291,
+ "num_input_tokens_seen": 1501896,
+ "step": 35,
+ "train_runtime": 271.9258,
+ "train_tokens_per_second": 5523.183
+ },
+ {
+ "epoch": 0.00269323996768112,
+ "grad_norm": 1.1179858029742544,
+ "learning_rate": 9.999957465779326e-06,
+ "loss": 0.5239121437072753,
+ "num_input_tokens_seen": 1716792,
+ "step": 40,
+ "train_runtime": 319.9491,
+ "train_tokens_per_second": 5365.829
+ },
+ {
+ "epoch": 0.0030298949636412605,
+ "grad_norm": 1.1555971967505698,
+ "learning_rate": 9.999945860473788e-06,
+ "loss": 0.5400435924530029,
+ "num_input_tokens_seen": 1920760,
+ "step": 45,
+ "train_runtime": 364.1591,
+ "train_tokens_per_second": 5274.507
+ },
+ {
+ "epoch": 0.0033665499596014004,
+ "grad_norm": 0.9321994843580264,
+ "learning_rate": 9.999932856949333e-06,
+ "loss": 0.5238595008850098,
+ "num_input_tokens_seen": 2140344,
+ "step": 50,
+ "train_runtime": 402.577,
+ "train_tokens_per_second": 5316.608
+ },
+ {
+ "epoch": 0.0037032049555615404,
+ "grad_norm": 1.1572766450172172,
+ "learning_rate": 9.999918455209599e-06,
+ "loss": 0.5059286117553711,
+ "num_input_tokens_seen": 2368296,
+ "step": 55,
+ "train_runtime": 438.0431,
+ "train_tokens_per_second": 5406.536
+ },
+ {
+ "epoch": 0.004039859951521681,
+ "grad_norm": 1.1277411399750283,
+ "learning_rate": 9.99990265525861e-06,
+ "loss": 0.5235630989074707,
+ "num_input_tokens_seen": 2589544,
+ "step": 60,
+ "train_runtime": 475.1915,
+ "train_tokens_per_second": 5449.474
+ },
+ {
+ "epoch": 0.004376514947481821,
+ "grad_norm": 1.233348290191874,
+ "learning_rate": 9.999885457100789e-06,
+ "loss": 0.5595571517944335,
+ "num_input_tokens_seen": 2809432,
+ "step": 65,
+ "train_runtime": 503.8866,
+ "train_tokens_per_second": 5575.524
+ },
+ {
+ "epoch": 0.0047131699434419605,
+ "grad_norm": 1.2298135105742523,
+ "learning_rate": 9.999866860740943e-06,
+ "loss": 0.5092310428619384,
+ "num_input_tokens_seen": 3020248,
+ "step": 70,
+ "train_runtime": 543.5456,
+ "train_tokens_per_second": 5556.568
+ },
+ {
+ "epoch": 0.005049824939402101,
+ "grad_norm": 0.9679470345624956,
+ "learning_rate": 9.999846866184274e-06,
+ "loss": 0.5114652633666992,
+ "num_input_tokens_seen": 3249600,
+ "step": 75,
+ "train_runtime": 585.283,
+ "train_tokens_per_second": 5552.186
+ },
+ {
+ "epoch": 0.00538647993536224,
+ "grad_norm": 0.979469640417179,
+ "learning_rate": 9.99982547343637e-06,
+ "loss": 0.5040130615234375,
+ "num_input_tokens_seen": 3443864,
+ "step": 80,
+ "train_runtime": 626.1684,
+ "train_tokens_per_second": 5499.901
+ },
+ {
+ "epoch": 0.005723134931322381,
+ "grad_norm": 1.078786163056917,
+ "learning_rate": 9.999802682503217e-06,
+ "loss": 0.4742586612701416,
+ "num_input_tokens_seen": 3661032,
+ "step": 85,
+ "train_runtime": 663.7311,
+ "train_tokens_per_second": 5515.836
+ },
+ {
+ "epoch": 0.006059789927282521,
+ "grad_norm": 0.9113383787124433,
+ "learning_rate": 9.999778493391186e-06,
+ "loss": 0.5044189453125,
+ "num_input_tokens_seen": 3888680,
+ "step": 90,
+ "train_runtime": 707.589,
+ "train_tokens_per_second": 5495.676
+ },
+ {
+ "epoch": 0.0063964449232426605,
+ "grad_norm": 1.0192977781689194,
+ "learning_rate": 9.999752906107043e-06,
+ "loss": 0.5348274230957031,
+ "num_input_tokens_seen": 4114248,
+ "step": 95,
+ "train_runtime": 744.4001,
+ "train_tokens_per_second": 5526.931
+ },
+ {
+ "epoch": 0.006733099919202801,
+ "grad_norm": 1.046539367979976,
+ "learning_rate": 9.999725920657944e-06,
+ "loss": 0.5229468822479248,
+ "num_input_tokens_seen": 4329256,
+ "step": 100,
+ "train_runtime": 790.5278,
+ "train_tokens_per_second": 5476.412
+ },
+ {
+ "epoch": 0.007069754915162941,
+ "grad_norm": 1.0823101637799084,
+ "learning_rate": 9.999697537051432e-06,
+ "loss": 0.4990852355957031,
+ "num_input_tokens_seen": 4533832,
+ "step": 105,
+ "train_runtime": 828.8033,
+ "train_tokens_per_second": 5470.335
+ },
+ {
+ "epoch": 0.007406409911123081,
+ "grad_norm": 1.0579674272002448,
+ "learning_rate": 9.999667755295449e-06,
+ "loss": 0.48107209205627444,
+ "num_input_tokens_seen": 4749088,
+ "step": 110,
+ "train_runtime": 864.874,
+ "train_tokens_per_second": 5491.075
+ },
+ {
+ "epoch": 0.007743064907083221,
+ "grad_norm": 1.2773224732968154,
+ "learning_rate": 9.999636575398319e-06,
+ "loss": 0.4919854164123535,
+ "num_input_tokens_seen": 4960752,
+ "step": 115,
+ "train_runtime": 904.8595,
+ "train_tokens_per_second": 5482.345
+ },
+ {
+ "epoch": 0.008079719903043361,
+ "grad_norm": 1.0488778079229073,
+ "learning_rate": 9.999603997368766e-06,
+ "loss": 0.4375304222106934,
+ "num_input_tokens_seen": 5167864,
+ "step": 120,
+ "train_runtime": 939.3678,
+ "train_tokens_per_second": 5501.428
+ },
+ {
+ "epoch": 0.008416374899003501,
+ "grad_norm": 1.1229102076855217,
+ "learning_rate": 9.999570021215895e-06,
+ "loss": 0.4794501304626465,
+ "num_input_tokens_seen": 5384648,
+ "step": 125,
+ "train_runtime": 973.7305,
+ "train_tokens_per_second": 5529.916
+ },
+ {
+ "epoch": 0.008753029894963642,
+ "grad_norm": 1.0387093548265027,
+ "learning_rate": 9.99953464694921e-06,
+ "loss": 0.5004319667816162,
+ "num_input_tokens_seen": 5590392,
+ "step": 130,
+ "train_runtime": 1013.6767,
+ "train_tokens_per_second": 5514.965
+ },
+ {
+ "epoch": 0.009089684890923782,
+ "grad_norm": 1.0508132347808834,
+ "learning_rate": 9.999497874578606e-06,
+ "loss": 0.495189905166626,
+ "num_input_tokens_seen": 5819224,
+ "step": 135,
+ "train_runtime": 1052.6891,
+ "train_tokens_per_second": 5527.961
+ },
+ {
+ "epoch": 0.009426339886883921,
+ "grad_norm": 1.0896002910221911,
+ "learning_rate": 9.999459704114361e-06,
+ "loss": 0.5080117225646973,
+ "num_input_tokens_seen": 6037808,
+ "step": 140,
+ "train_runtime": 1094.5671,
+ "train_tokens_per_second": 5516.161
+ },
+ {
+ "epoch": 0.00976299488284406,
+ "grad_norm": 1.034914657869065,
+ "learning_rate": 9.999420135567153e-06,
+ "loss": 0.526894474029541,
+ "num_input_tokens_seen": 6252752,
+ "step": 145,
+ "train_runtime": 1135.4311,
+ "train_tokens_per_second": 5506.941
+ },
+ {
+ "epoch": 0.010099649878804202,
+ "grad_norm": 1.164626155534281,
+ "learning_rate": 9.999379168948047e-06,
+ "loss": 0.5105450630187989,
+ "num_input_tokens_seen": 6476720,
+ "step": 150,
+ "train_runtime": 1172.7989,
+ "train_tokens_per_second": 5522.447
+ },
+ {
+ "epoch": 0.010436304874764341,
+ "grad_norm": 1.2400950788317526,
+ "learning_rate": 9.999336804268496e-06,
+ "loss": 0.47139749526977537,
+ "num_input_tokens_seen": 6698024,
+ "step": 155,
+ "train_runtime": 1217.419,
+ "train_tokens_per_second": 5501.823
+ },
+ {
+ "epoch": 0.01077295987072448,
+ "grad_norm": 1.0587023212526414,
+ "learning_rate": 9.999293041540351e-06,
+ "loss": 0.5010284900665283,
+ "num_input_tokens_seen": 6910784,
+ "step": 160,
+ "train_runtime": 1258.4852,
+ "train_tokens_per_second": 5491.351
+ },
+ {
+ "epoch": 0.011109614866684622,
+ "grad_norm": 1.0126894215000632,
+ "learning_rate": 9.999247880775849e-06,
+ "loss": 0.49982900619506837,
+ "num_input_tokens_seen": 7115568,
+ "step": 165,
+ "train_runtime": 1295.6174,
+ "train_tokens_per_second": 5492.029
+ },
+ {
+ "epoch": 0.011446269862644761,
+ "grad_norm": 1.198929638798561,
+ "learning_rate": 9.999201321987617e-06,
+ "loss": 0.46059427261352537,
+ "num_input_tokens_seen": 7329488,
+ "step": 170,
+ "train_runtime": 1337.1094,
+ "train_tokens_per_second": 5481.592
+ },
+ {
+ "epoch": 0.011782924858604901,
+ "grad_norm": 1.0114658976479987,
+ "learning_rate": 9.999153365188676e-06,
+ "loss": 0.5039016723632812,
+ "num_input_tokens_seen": 7547880,
+ "step": 175,
+ "train_runtime": 1383.6734,
+ "train_tokens_per_second": 5454.958
+ },
+ {
+ "epoch": 0.012119579854565042,
+ "grad_norm": 1.0579154349493107,
+ "learning_rate": 9.99910401039244e-06,
+ "loss": 0.4546662330627441,
+ "num_input_tokens_seen": 7757728,
+ "step": 180,
+ "train_runtime": 1424.6608,
+ "train_tokens_per_second": 5445.316
+ },
+ {
+ "epoch": 0.012456234850525182,
+ "grad_norm": 1.1327328565153654,
+ "learning_rate": 9.999053257612708e-06,
+ "loss": 0.464589786529541,
+ "num_input_tokens_seen": 7984416,
+ "step": 185,
+ "train_runtime": 1468.0361,
+ "train_tokens_per_second": 5438.842
+ },
+ {
+ "epoch": 0.012792889846485321,
+ "grad_norm": 1.0695885705791155,
+ "learning_rate": 9.999001106863672e-06,
+ "loss": 0.5309792518615722,
+ "num_input_tokens_seen": 8204976,
+ "step": 190,
+ "train_runtime": 1506.2271,
+ "train_tokens_per_second": 5447.37
+ },
+ {
+ "epoch": 0.013129544842445462,
+ "grad_norm": 1.129268767285916,
+ "learning_rate": 9.998947558159916e-06,
+ "loss": 0.4964266300201416,
+ "num_input_tokens_seen": 8428800,
+ "step": 195,
+ "train_runtime": 1546.2799,
+ "train_tokens_per_second": 5451.018
+ },
+ {
+ "epoch": 0.013466199838405602,
+ "grad_norm": 1.0645176443218387,
+ "learning_rate": 9.99889261151642e-06,
+ "loss": 0.44077634811401367,
+ "num_input_tokens_seen": 8642192,
+ "step": 200,
+ "train_runtime": 1586.0871,
+ "train_tokens_per_second": 5448.75
+ },
+ {
+ "epoch": 0.013802854834365741,
+ "grad_norm": 1.0273158071004191,
+ "learning_rate": 9.998836266948542e-06,
+ "loss": 0.47046546936035155,
+ "num_input_tokens_seen": 8852840,
+ "step": 205,
+ "train_runtime": 1618.9934,
+ "train_tokens_per_second": 5468.114
+ },
+ {
+ "epoch": 0.014139509830325882,
+ "grad_norm": 1.2260591863245542,
+ "learning_rate": 9.998778524472043e-06,
+ "loss": 0.4429631233215332,
+ "num_input_tokens_seen": 9059368,
+ "step": 210,
+ "train_runtime": 1658.2941,
+ "train_tokens_per_second": 5463.065
+ },
+ {
+ "epoch": 0.014476164826286022,
+ "grad_norm": 1.1035215533216065,
+ "learning_rate": 9.99871938410307e-06,
+ "loss": 0.46187715530395507,
+ "num_input_tokens_seen": 9273936,
+ "step": 215,
+ "train_runtime": 1690.8398,
+ "train_tokens_per_second": 5484.811
+ },
+ {
+ "epoch": 0.014812819822246161,
+ "grad_norm": 1.500582938758766,
+ "learning_rate": 9.99865884585816e-06,
+ "loss": 0.5123334407806397,
+ "num_input_tokens_seen": 9489816,
+ "step": 220,
+ "train_runtime": 1724.7509,
+ "train_tokens_per_second": 5502.137
+ },
+ {
+ "epoch": 0.015149474818206303,
+ "grad_norm": 1.1104716846300207,
+ "learning_rate": 9.998596909754245e-06,
+ "loss": 0.4839674949645996,
+ "num_input_tokens_seen": 9710720,
+ "step": 225,
+ "train_runtime": 1769.0601,
+ "train_tokens_per_second": 5489.197
+ },
+ {
+ "epoch": 0.015486129814166442,
+ "grad_norm": 1.1763018457719354,
+ "learning_rate": 9.998533575808642e-06,
+ "loss": 0.455305290222168,
+ "num_input_tokens_seen": 9922160,
+ "step": 230,
+ "train_runtime": 1804.4173,
+ "train_tokens_per_second": 5498.817
+ },
+ {
+ "epoch": 0.015822784810126583,
+ "grad_norm": 0.9508268973242634,
+ "learning_rate": 9.998468844039064e-06,
+ "loss": 0.4588268280029297,
+ "num_input_tokens_seen": 10132480,
+ "step": 235,
+ "train_runtime": 1843.1618,
+ "train_tokens_per_second": 5497.336
+ },
+ {
+ "epoch": 0.016159439806086723,
+ "grad_norm": 0.982580320370019,
+ "learning_rate": 9.998402714463613e-06,
+ "loss": 0.49381322860717775,
+ "num_input_tokens_seen": 10356032,
+ "step": 240,
+ "train_runtime": 1886.9781,
+ "train_tokens_per_second": 5488.157
+ },
+ {
+ "epoch": 0.016496094802046862,
+ "grad_norm": 0.9967583322638394,
+ "learning_rate": 9.998335187100782e-06,
+ "loss": 0.4929532051086426,
+ "num_input_tokens_seen": 10571688,
+ "step": 245,
+ "train_runtime": 1931.0773,
+ "train_tokens_per_second": 5474.503
+ },
+ {
+ "epoch": 0.016832749798007002,
+ "grad_norm": 1.0005940352447382,
+ "learning_rate": 9.998266261969453e-06,
+ "loss": 0.488722038269043,
+ "num_input_tokens_seen": 10787736,
+ "step": 250,
+ "train_runtime": 1967.3657,
+ "train_tokens_per_second": 5483.341
+ },
+ {
+ "epoch": 0.01716940479396714,
+ "grad_norm": 0.9740927827374989,
+ "learning_rate": 9.998195939088905e-06,
+ "loss": 0.46301860809326173,
+ "num_input_tokens_seen": 11014240,
+ "step": 255,
+ "train_runtime": 1995.3214,
+ "train_tokens_per_second": 5520.033
+ },
+ {
+ "epoch": 0.017506059789927284,
+ "grad_norm": 1.0005631289772583,
+ "learning_rate": 9.998124218478797e-06,
+ "loss": 0.4852250099182129,
+ "num_input_tokens_seen": 11227928,
+ "step": 260,
+ "train_runtime": 2040.8305,
+ "train_tokens_per_second": 5501.646
+ },
+ {
+ "epoch": 0.017842714785887424,
+ "grad_norm": 3.2898314540824733,
+ "learning_rate": 9.998051100159192e-06,
+ "loss": 0.4806126594543457,
+ "num_input_tokens_seen": 11457336,
+ "step": 265,
+ "train_runtime": 2073.3223,
+ "train_tokens_per_second": 5526.076
+ },
+ {
+ "epoch": 0.018179369781847563,
+ "grad_norm": 1.0137684438609407,
+ "learning_rate": 9.997976584150533e-06,
+ "loss": 0.4908778667449951,
+ "num_input_tokens_seen": 11669408,
+ "step": 270,
+ "train_runtime": 2105.6176,
+ "train_tokens_per_second": 5542.036
+ },
+ {
+ "epoch": 0.018516024777807703,
+ "grad_norm": 0.882254713488144,
+ "learning_rate": 9.997900670473662e-06,
+ "loss": 0.4400303840637207,
+ "num_input_tokens_seen": 11900376,
+ "step": 275,
+ "train_runtime": 2145.8648,
+ "train_tokens_per_second": 5545.725
+ },
+ {
+ "epoch": 0.018852679773767842,
+ "grad_norm": 1.0103112069849245,
+ "learning_rate": 9.997823359149804e-06,
+ "loss": 0.47128896713256835,
+ "num_input_tokens_seen": 12115296,
+ "step": 280,
+ "train_runtime": 2180.0754,
+ "train_tokens_per_second": 5557.283
+ },
+ {
+ "epoch": 0.01918933476972798,
+ "grad_norm": 2.383562470887594,
+ "learning_rate": 9.997744650200579e-06,
+ "loss": 0.5039300441741943,
+ "num_input_tokens_seen": 12335296,
+ "step": 285,
+ "train_runtime": 2224.7672,
+ "train_tokens_per_second": 5544.533
+ },
+ {
+ "epoch": 0.01952598976568812,
+ "grad_norm": 1.012511096525962,
+ "learning_rate": 9.997664543648e-06,
+ "loss": 0.47290778160095215,
+ "num_input_tokens_seen": 12550016,
+ "step": 290,
+ "train_runtime": 2269.2912,
+ "train_tokens_per_second": 5530.368
+ },
+ {
+ "epoch": 0.019862644761648264,
+ "grad_norm": 0.9200150844694661,
+ "learning_rate": 9.997583039514469e-06,
+ "loss": 0.47549943923950194,
+ "num_input_tokens_seen": 12769944,
+ "step": 295,
+ "train_runtime": 2310.0213,
+ "train_tokens_per_second": 5528.063
+ },
+ {
+ "epoch": 0.020199299757608404,
+ "grad_norm": 1.0603122997620704,
+ "learning_rate": 9.997500137822776e-06,
+ "loss": 0.48865842819213867,
+ "num_input_tokens_seen": 12994152,
+ "step": 300,
+ "train_runtime": 2347.3612,
+ "train_tokens_per_second": 5535.642
+ },
+ {
+ "epoch": 0.020535954753568543,
+ "grad_norm": 1.1560224095031963,
+ "learning_rate": 9.997415838596106e-06,
+ "loss": 0.46552267074584963,
+ "num_input_tokens_seen": 13206104,
+ "step": 305,
+ "train_runtime": 2390.8704,
+ "train_tokens_per_second": 5523.555
+ },
+ {
+ "epoch": 0.020872609749528682,
+ "grad_norm": 0.9496962272740035,
+ "learning_rate": 9.997330141858032e-06,
+ "loss": 0.4945693492889404,
+ "num_input_tokens_seen": 13423544,
+ "step": 310,
+ "train_runtime": 2429.9223,
+ "train_tokens_per_second": 5524.269
+ },
+ {
+ "epoch": 0.021209264745488822,
+ "grad_norm": 1.1164280869632712,
+ "learning_rate": 9.997243047632519e-06,
+ "loss": 0.47294368743896487,
+ "num_input_tokens_seen": 13640584,
+ "step": 315,
+ "train_runtime": 2463.856,
+ "train_tokens_per_second": 5536.275
+ },
+ {
+ "epoch": 0.02154591974144896,
+ "grad_norm": 0.9211260067991657,
+ "learning_rate": 9.997154555943923e-06,
+ "loss": 0.43665332794189454,
+ "num_input_tokens_seen": 13858888,
+ "step": 320,
+ "train_runtime": 2506.1,
+ "train_tokens_per_second": 5530.062
+ },
+ {
+ "epoch": 0.021882574737409104,
+ "grad_norm": 1.0719268322873283,
+ "learning_rate": 9.997064666816992e-06,
+ "loss": 0.4625236034393311,
+ "num_input_tokens_seen": 14076648,
+ "step": 325,
+ "train_runtime": 2544.8417,
+ "train_tokens_per_second": 5531.443
+ },
+ {
+ "epoch": 0.022219229733369244,
+ "grad_norm": 1.146001730574971,
+ "learning_rate": 9.996973380276858e-06,
+ "loss": 0.5440494060516358,
+ "num_input_tokens_seen": 14299864,
+ "step": 330,
+ "train_runtime": 2575.9234,
+ "train_tokens_per_second": 5551.355
+ },
+ {
+ "epoch": 0.022555884729329383,
+ "grad_norm": 1.1489261273369993,
+ "learning_rate": 9.996880696349053e-06,
+ "loss": 0.45033721923828124,
+ "num_input_tokens_seen": 14522560,
+ "step": 335,
+ "train_runtime": 2614.1655,
+ "train_tokens_per_second": 5555.333
+ },
+ {
+ "epoch": 0.022892539725289523,
+ "grad_norm": 1.0606322220744733,
+ "learning_rate": 9.996786615059497e-06,
+ "loss": 0.5329171180725097,
+ "num_input_tokens_seen": 14746648,
+ "step": 340,
+ "train_runtime": 2655.6018,
+ "train_tokens_per_second": 5553.034
+ },
+ {
+ "epoch": 0.023229194721249662,
+ "grad_norm": 0.9692502297767218,
+ "learning_rate": 9.996691136434498e-06,
+ "loss": 0.4342483997344971,
+ "num_input_tokens_seen": 14969080,
+ "step": 345,
+ "train_runtime": 2690.8338,
+ "train_tokens_per_second": 5562.989
+ },
+ {
+ "epoch": 0.023565849717209802,
+ "grad_norm": 0.8994912857959015,
+ "learning_rate": 9.996594260500753e-06,
+ "loss": 0.507387924194336,
+ "num_input_tokens_seen": 15193912,
+ "step": 350,
+ "train_runtime": 2729.4748,
+ "train_tokens_per_second": 5566.606
+ },
+ {
+ "epoch": 0.023902504713169945,
+ "grad_norm": 0.9930018613669586,
+ "learning_rate": 9.996495987285358e-06,
+ "loss": 0.4612261772155762,
+ "num_input_tokens_seen": 15411408,
+ "step": 355,
+ "train_runtime": 2773.0354,
+ "train_tokens_per_second": 5557.595
+ },
+ {
+ "epoch": 0.024239159709130084,
+ "grad_norm": 1.0882060582433957,
+ "learning_rate": 9.996396316815793e-06,
+ "loss": 0.4673755645751953,
+ "num_input_tokens_seen": 15630480,
+ "step": 360,
+ "train_runtime": 2815.6019,
+ "train_tokens_per_second": 5551.381
+ },
+ {
+ "epoch": 0.024575814705090224,
+ "grad_norm": 1.1223066651934839,
+ "learning_rate": 9.99629524911993e-06,
+ "loss": 0.4566653251647949,
+ "num_input_tokens_seen": 15847552,
+ "step": 365,
+ "train_runtime": 2852.9783,
+ "train_tokens_per_second": 5554.74
+ },
+ {
+ "epoch": 0.024912469701050363,
+ "grad_norm": 1.0964642711464192,
+ "learning_rate": 9.996192784226033e-06,
+ "loss": 0.440608549118042,
+ "num_input_tokens_seen": 16049488,
+ "step": 370,
+ "train_runtime": 2887.3093,
+ "train_tokens_per_second": 5558.631
+ },
+ {
+ "epoch": 0.025249124697010503,
+ "grad_norm": 0.9458333109972462,
+ "learning_rate": 9.996088922162755e-06,
+ "loss": 0.4601402759552002,
+ "num_input_tokens_seen": 16283544,
+ "step": 375,
+ "train_runtime": 2920.1965,
+ "train_tokens_per_second": 5576.181
+ },
+ {
+ "epoch": 0.025585779692970642,
+ "grad_norm": 0.9639187009881571,
+ "learning_rate": 9.995983662959141e-06,
+ "loss": 0.4678368091583252,
+ "num_input_tokens_seen": 16507176,
+ "step": 380,
+ "train_runtime": 2959.819,
+ "train_tokens_per_second": 5577.09
+ },
+ {
+ "epoch": 0.025922434688930785,
+ "grad_norm": 0.9786802993443171,
+ "learning_rate": 9.995877006644627e-06,
+ "loss": 0.4893208026885986,
+ "num_input_tokens_seen": 16739664,
+ "step": 385,
+ "train_runtime": 3003.5655,
+ "train_tokens_per_second": 5573.264
+ },
+ {
+ "epoch": 0.026259089684890925,
+ "grad_norm": 0.9983372688388842,
+ "learning_rate": 9.995768953249038e-06,
+ "loss": 0.4782050132751465,
+ "num_input_tokens_seen": 16966112,
+ "step": 390,
+ "train_runtime": 3042.5667,
+ "train_tokens_per_second": 5576.25
+ },
+ {
+ "epoch": 0.026595744680851064,
+ "grad_norm": 0.9763900936703368,
+ "learning_rate": 9.995659502802594e-06,
+ "loss": 0.4591067314147949,
+ "num_input_tokens_seen": 17184144,
+ "step": 395,
+ "train_runtime": 3081.608,
+ "train_tokens_per_second": 5576.356
+ },
+ {
+ "epoch": 0.026932399676811204,
+ "grad_norm": 1.1889179128899698,
+ "learning_rate": 9.995548655335899e-06,
+ "loss": 0.4847611427307129,
+ "num_input_tokens_seen": 17393504,
+ "step": 400,
+ "train_runtime": 3119.919,
+ "train_tokens_per_second": 5574.986
+ },
+ {
+ "epoch": 0.027269054672771343,
+ "grad_norm": 1.0385231933952215,
+ "learning_rate": 9.995436410879951e-06,
+ "loss": 0.5011202812194824,
+ "num_input_tokens_seen": 17618352,
+ "step": 405,
+ "train_runtime": 3158.5186,
+ "train_tokens_per_second": 5578.043
+ },
+ {
+ "epoch": 0.027605709668731482,
+ "grad_norm": 1.0315411320040826,
+ "learning_rate": 9.99532276946614e-06,
+ "loss": 0.4390664577484131,
+ "num_input_tokens_seen": 17833504,
+ "step": 410,
+ "train_runtime": 3197.1139,
+ "train_tokens_per_second": 5578.001
+ },
+ {
+ "epoch": 0.027942364664691625,
+ "grad_norm": 0.9798271051589648,
+ "learning_rate": 9.995207731126246e-06,
+ "loss": 0.48116722106933596,
+ "num_input_tokens_seen": 18060040,
+ "step": 415,
+ "train_runtime": 3233.4074,
+ "train_tokens_per_second": 5585.451
+ },
+ {
+ "epoch": 0.028279019660651765,
+ "grad_norm": 0.9349852018887401,
+ "learning_rate": 9.99509129589244e-06,
+ "loss": 0.4614968776702881,
+ "num_input_tokens_seen": 18273536,
+ "step": 420,
+ "train_runtime": 3276.2797,
+ "train_tokens_per_second": 5577.526
+ },
+ {
+ "epoch": 0.028615674656611904,
+ "grad_norm": 1.1608131463568248,
+ "learning_rate": 9.994973463797279e-06,
+ "loss": 0.4571842193603516,
+ "num_input_tokens_seen": 18501288,
+ "step": 425,
+ "train_runtime": 3312.0947,
+ "train_tokens_per_second": 5585.978
+ },
+ {
+ "epoch": 0.028952329652572044,
+ "grad_norm": 0.9117559518264567,
+ "learning_rate": 9.99485423487372e-06,
+ "loss": 0.42873506546020507,
+ "num_input_tokens_seen": 18713400,
+ "step": 430,
+ "train_runtime": 3354.8456,
+ "train_tokens_per_second": 5578.021
+ },
+ {
+ "epoch": 0.029288984648532183,
+ "grad_norm": 1.0351008634262573,
+ "learning_rate": 9.994733609155097e-06,
+ "loss": 0.46686439514160155,
+ "num_input_tokens_seen": 18927424,
+ "step": 435,
+ "train_runtime": 3390.8404,
+ "train_tokens_per_second": 5581.927
+ },
+ {
+ "epoch": 0.029625639644492323,
+ "grad_norm": 0.9515109413699648,
+ "learning_rate": 9.994611586675149e-06,
+ "loss": 0.4384657382965088,
+ "num_input_tokens_seen": 19153848,
+ "step": 440,
+ "train_runtime": 3427.1383,
+ "train_tokens_per_second": 5588.875
+ },
+ {
+ "epoch": 0.029962294640452466,
+ "grad_norm": 2.170654371130382,
+ "learning_rate": 9.994488167468e-06,
+ "loss": 0.457914924621582,
+ "num_input_tokens_seen": 19367752,
+ "step": 445,
+ "train_runtime": 3463.4205,
+ "train_tokens_per_second": 5592.088
+ },
+ {
+ "epoch": 0.030298949636412605,
+ "grad_norm": 0.9213402764770521,
+ "learning_rate": 9.994363351568158e-06,
+ "loss": 0.49051642417907715,
+ "num_input_tokens_seen": 19581376,
+ "step": 450,
+ "train_runtime": 3507.0672,
+ "train_tokens_per_second": 5583.405
+ },
+ {
+ "epoch": 0.030635604632372745,
+ "grad_norm": 1.0129019967889514,
+ "learning_rate": 9.994237139010532e-06,
+ "loss": 0.4277812480926514,
+ "num_input_tokens_seen": 19787728,
+ "step": 455,
+ "train_runtime": 3546.6942,
+ "train_tokens_per_second": 5579.203
+ },
+ {
+ "epoch": 0.030972259628332884,
+ "grad_norm": 0.9251976717430972,
+ "learning_rate": 9.994109529830414e-06,
+ "loss": 0.4721639633178711,
+ "num_input_tokens_seen": 20024672,
+ "step": 460,
+ "train_runtime": 3585.3338,
+ "train_tokens_per_second": 5585.163
+ },
+ {
+ "epoch": 0.03130891462429303,
+ "grad_norm": 0.8419132637197803,
+ "learning_rate": 9.993980524063494e-06,
+ "loss": 0.44678316116333006,
+ "num_input_tokens_seen": 20247816,
+ "step": 465,
+ "train_runtime": 3622.8878,
+ "train_tokens_per_second": 5588.861
+ },
+ {
+ "epoch": 0.03164556962025317,
+ "grad_norm": 0.8727373261991417,
+ "learning_rate": 9.993850121745842e-06,
+ "loss": 0.448548698425293,
+ "num_input_tokens_seen": 20469008,
+ "step": 470,
+ "train_runtime": 3663.8929,
+ "train_tokens_per_second": 5586.683
+ },
+ {
+ "epoch": 0.031982224616213306,
+ "grad_norm": 0.9661829335825901,
+ "learning_rate": 9.99371832291393e-06,
+ "loss": 0.44652185440063474,
+ "num_input_tokens_seen": 20684368,
+ "step": 475,
+ "train_runtime": 3702.3594,
+ "train_tokens_per_second": 5586.807
+ },
+ {
+ "epoch": 0.032318879612173446,
+ "grad_norm": 1.052144245514416,
+ "learning_rate": 9.993585127604612e-06,
+ "loss": 0.4458156108856201,
+ "num_input_tokens_seen": 20888824,
+ "step": 480,
+ "train_runtime": 3737.8839,
+ "train_tokens_per_second": 5588.409
+ },
+ {
+ "epoch": 0.032655534608133585,
+ "grad_norm": 0.8727464499212149,
+ "learning_rate": 9.993450535855136e-06,
+ "loss": 0.4528946876525879,
+ "num_input_tokens_seen": 21093776,
+ "step": 485,
+ "train_runtime": 3771.0576,
+ "train_tokens_per_second": 5593.597
+ },
+ {
+ "epoch": 0.032992189604093725,
+ "grad_norm": 1.4490470715769006,
+ "learning_rate": 9.993314547703142e-06,
+ "loss": 0.43402805328369143,
+ "num_input_tokens_seen": 21314528,
+ "step": 490,
+ "train_runtime": 3810.8049,
+ "train_tokens_per_second": 5593.183
+ },
+ {
+ "epoch": 0.033328844600053864,
+ "grad_norm": 0.913157610231942,
+ "learning_rate": 9.993177163186655e-06,
+ "loss": 0.44025378227233886,
+ "num_input_tokens_seen": 21529648,
+ "step": 495,
+ "train_runtime": 3846.3087,
+ "train_tokens_per_second": 5597.483
+ },
+ {
+ "epoch": 0.033665499596014004,
+ "grad_norm": 1.0732207036056458,
+ "learning_rate": 9.993038382344097e-06,
+ "loss": 0.43718814849853516,
+ "num_input_tokens_seen": 21744008,
+ "step": 500,
+ "train_runtime": 3886.9027,
+ "train_tokens_per_second": 5594.174
+ },
+ {
+ "epoch": 0.03400215459197414,
+ "grad_norm": 0.9385843989678325,
+ "learning_rate": 9.992898205214279e-06,
+ "loss": 0.44336519241333006,
+ "num_input_tokens_seen": 21963168,
+ "step": 505,
+ "train_runtime": 3919.9586,
+ "train_tokens_per_second": 5602.908
+ },
+ {
+ "epoch": 0.03433880958793428,
+ "grad_norm": 0.9966215411858433,
+ "learning_rate": 9.992756631836397e-06,
+ "loss": 0.4398432731628418,
+ "num_input_tokens_seen": 22181088,
+ "step": 510,
+ "train_runtime": 3967.3085,
+ "train_tokens_per_second": 5590.966
+ },
+ {
+ "epoch": 0.03467546458389442,
+ "grad_norm": 0.8920648260369749,
+ "learning_rate": 9.992613662250044e-06,
+ "loss": 0.4598443031311035,
+ "num_input_tokens_seen": 22402688,
+ "step": 515,
+ "train_runtime": 4007.4675,
+ "train_tokens_per_second": 5590.236
+ },
+ {
+ "epoch": 0.03501211957985457,
+ "grad_norm": 0.9290820640509749,
+ "learning_rate": 9.992469296495201e-06,
+ "loss": 0.4575651168823242,
+ "num_input_tokens_seen": 22628840,
+ "step": 520,
+ "train_runtime": 4042.7584,
+ "train_tokens_per_second": 5597.376
+ },
+ {
+ "epoch": 0.03534877457581471,
+ "grad_norm": 0.9109745499097042,
+ "learning_rate": 9.992323534612238e-06,
+ "loss": 0.414502477645874,
+ "num_input_tokens_seen": 22848504,
+ "step": 525,
+ "train_runtime": 4082.2649,
+ "train_tokens_per_second": 5597.017
+ },
+ {
+ "epoch": 0.03568542957177485,
+ "grad_norm": 1.1344769076017303,
+ "learning_rate": 9.99217637664192e-06,
+ "loss": 0.45860743522644043,
+ "num_input_tokens_seen": 23072192,
+ "step": 530,
+ "train_runtime": 4119.8107,
+ "train_tokens_per_second": 5600.304
+ },
+ {
+ "epoch": 0.03602208456773499,
+ "grad_norm": 0.940449394205852,
+ "learning_rate": 9.992027822625397e-06,
+ "loss": 0.45654983520507814,
+ "num_input_tokens_seen": 23281952,
+ "step": 535,
+ "train_runtime": 4161.0794,
+ "train_tokens_per_second": 5595.171
+ },
+ {
+ "epoch": 0.036358739563695126,
+ "grad_norm": 1.0036660879417987,
+ "learning_rate": 9.99187787260421e-06,
+ "loss": 0.46324939727783204,
+ "num_input_tokens_seen": 23492592,
+ "step": 540,
+ "train_runtime": 4199.7934,
+ "train_tokens_per_second": 5593.749
+ },
+ {
+ "epoch": 0.036695394559655266,
+ "grad_norm": 0.9529512682682689,
+ "learning_rate": 9.991726526620294e-06,
+ "loss": 0.4310856819152832,
+ "num_input_tokens_seen": 23724784,
+ "step": 545,
+ "train_runtime": 4235.651,
+ "train_tokens_per_second": 5601.213
+ },
+ {
+ "epoch": 0.037032049555615405,
+ "grad_norm": 1.0092720255970657,
+ "learning_rate": 9.991573784715973e-06,
+ "loss": 0.4721970081329346,
+ "num_input_tokens_seen": 23941816,
+ "step": 550,
+ "train_runtime": 4278.956,
+ "train_tokens_per_second": 5595.247
+ },
+ {
+ "epoch": 0.037368704551575545,
+ "grad_norm": 1.0108753586712804,
+ "learning_rate": 9.99141964693396e-06,
+ "loss": 0.48006014823913573,
+ "num_input_tokens_seen": 24165232,
+ "step": 555,
+ "train_runtime": 4315.3047,
+ "train_tokens_per_second": 5599.89
+ },
+ {
+ "epoch": 0.037705359547535684,
+ "grad_norm": 1.0343909095504493,
+ "learning_rate": 9.99126411331736e-06,
+ "loss": 0.4417973518371582,
+ "num_input_tokens_seen": 24389896,
+ "step": 560,
+ "train_runtime": 4359.1735,
+ "train_tokens_per_second": 5595.074
+ },
+ {
+ "epoch": 0.038042014543495824,
+ "grad_norm": 1.0113397276422895,
+ "learning_rate": 9.991107183909665e-06,
+ "loss": 0.4680752754211426,
+ "num_input_tokens_seen": 24614992,
+ "step": 565,
+ "train_runtime": 4391.4505,
+ "train_tokens_per_second": 5605.208
+ },
+ {
+ "epoch": 0.03837866953945596,
+ "grad_norm": 0.9954842488566088,
+ "learning_rate": 9.990948858754763e-06,
+ "loss": 0.4640792369842529,
+ "num_input_tokens_seen": 24823080,
+ "step": 570,
+ "train_runtime": 4434.0745,
+ "train_tokens_per_second": 5598.255
+ },
+ {
+ "epoch": 0.0387153245354161,
+ "grad_norm": 0.9153320849806976,
+ "learning_rate": 9.990789137896927e-06,
+ "loss": 0.4609560966491699,
+ "num_input_tokens_seen": 25044784,
+ "step": 575,
+ "train_runtime": 4471.0247,
+ "train_tokens_per_second": 5601.576
+ },
+ {
+ "epoch": 0.03905197953137624,
+ "grad_norm": 0.9718791869837801,
+ "learning_rate": 9.990628021380823e-06,
+ "loss": 0.47663059234619143,
+ "num_input_tokens_seen": 25262200,
+ "step": 580,
+ "train_runtime": 4509.6496,
+ "train_tokens_per_second": 5601.81
+ },
+ {
+ "epoch": 0.03938863452733639,
+ "grad_norm": 1.218415508866825,
+ "learning_rate": 9.990465509251508e-06,
+ "loss": 0.46056556701660156,
+ "num_input_tokens_seen": 25473136,
+ "step": 585,
+ "train_runtime": 4541.7073,
+ "train_tokens_per_second": 5608.714
+ },
+ {
+ "epoch": 0.03972528952329653,
+ "grad_norm": 1.0200294193913204,
+ "learning_rate": 9.990301601554425e-06,
+ "loss": 0.48001799583435056,
+ "num_input_tokens_seen": 25671088,
+ "step": 590,
+ "train_runtime": 4581.4528,
+ "train_tokens_per_second": 5603.264
+ },
+ {
+ "epoch": 0.04006194451925667,
+ "grad_norm": 1.0450169479009095,
+ "learning_rate": 9.990136298335414e-06,
+ "loss": 0.40839109420776365,
+ "num_input_tokens_seen": 25887472,
+ "step": 595,
+ "train_runtime": 4618.2668,
+ "train_tokens_per_second": 5605.452
+ },
+ {
+ "epoch": 0.04039859951521681,
+ "grad_norm": 0.9533941163271269,
+ "learning_rate": 9.989969599640698e-06,
+ "loss": 0.4223615646362305,
+ "num_input_tokens_seen": 26117456,
+ "step": 600,
+ "train_runtime": 4650.9664,
+ "train_tokens_per_second": 5615.49
+ },
+ {
+ "epoch": 0.040735254511176947,
+ "grad_norm": 0.9258754840147914,
+ "learning_rate": 9.989801505516895e-06,
+ "loss": 0.505678129196167,
+ "num_input_tokens_seen": 26348056,
+ "step": 605,
+ "train_runtime": 4690.0532,
+ "train_tokens_per_second": 5617.859
+ },
+ {
+ "epoch": 0.041071909507137086,
+ "grad_norm": 0.9866589002080994,
+ "learning_rate": 9.989632016011012e-06,
+ "loss": 0.47707581520080566,
+ "num_input_tokens_seen": 26572040,
+ "step": 610,
+ "train_runtime": 4721.6631,
+ "train_tokens_per_second": 5627.687
+ },
+ {
+ "epoch": 0.041408564503097225,
+ "grad_norm": 1.0374837134071888,
+ "learning_rate": 9.989461131170448e-06,
+ "loss": 0.4833499908447266,
+ "num_input_tokens_seen": 26769320,
+ "step": 615,
+ "train_runtime": 4762.2337,
+ "train_tokens_per_second": 5621.169
+ },
+ {
+ "epoch": 0.041745219499057365,
+ "grad_norm": 0.8823212064118572,
+ "learning_rate": 9.98928885104299e-06,
+ "loss": 0.4514059543609619,
+ "num_input_tokens_seen": 26988024,
+ "step": 620,
+ "train_runtime": 4795.3541,
+ "train_tokens_per_second": 5627.952
+ },
+ {
+ "epoch": 0.042081874495017504,
+ "grad_norm": 0.8780469782436839,
+ "learning_rate": 9.989115175676812e-06,
+ "loss": 0.4226998329162598,
+ "num_input_tokens_seen": 27212624,
+ "step": 625,
+ "train_runtime": 4834.4544,
+ "train_tokens_per_second": 5628.892
+ },
+ {
+ "epoch": 0.042418529490977644,
+ "grad_norm": 0.9554506078310793,
+ "learning_rate": 9.988940105120483e-06,
+ "loss": 0.47803831100463867,
+ "num_input_tokens_seen": 27422208,
+ "step": 630,
+ "train_runtime": 4877.5946,
+ "train_tokens_per_second": 5622.076
+ },
+ {
+ "epoch": 0.04275518448693778,
+ "grad_norm": 0.93348166885423,
+ "learning_rate": 9.988763639422965e-06,
+ "loss": 0.4802694797515869,
+ "num_input_tokens_seen": 27650384,
+ "step": 635,
+ "train_runtime": 4915.966,
+ "train_tokens_per_second": 5624.608
+ },
+ {
+ "epoch": 0.04309183948289792,
+ "grad_norm": 0.9769454728471132,
+ "learning_rate": 9.988585778633604e-06,
+ "loss": 0.4367190361022949,
+ "num_input_tokens_seen": 27851424,
+ "step": 640,
+ "train_runtime": 4948.8643,
+ "train_tokens_per_second": 5627.841
+ },
+ {
+ "epoch": 0.04342849447885807,
+ "grad_norm": 0.9848655689806975,
+ "learning_rate": 9.988406522802134e-06,
+ "loss": 0.45001683235168455,
+ "num_input_tokens_seen": 28060088,
+ "step": 645,
+ "train_runtime": 4991.2246,
+ "train_tokens_per_second": 5621.884
+ },
+ {
+ "epoch": 0.04376514947481821,
+ "grad_norm": 0.9626370029590127,
+ "learning_rate": 9.98822587197869e-06,
+ "loss": 0.41652770042419435,
+ "num_input_tokens_seen": 28274912,
+ "step": 650,
+ "train_runtime": 5025.7373,
+ "train_tokens_per_second": 5626.023
+ },
+ {
+ "epoch": 0.04410180447077835,
+ "grad_norm": 0.9983674144400133,
+ "learning_rate": 9.988043826213785e-06,
+ "loss": 0.4282545566558838,
+ "num_input_tokens_seen": 28481968,
+ "step": 655,
+ "train_runtime": 5062.1603,
+ "train_tokens_per_second": 5626.445
+ },
+ {
+ "epoch": 0.04443845946673849,
+ "grad_norm": 2.4559661675719884,
+ "learning_rate": 9.98786038555833e-06,
+ "loss": 0.4085527420043945,
+ "num_input_tokens_seen": 28702472,
+ "step": 660,
+ "train_runtime": 5099.8424,
+ "train_tokens_per_second": 5628.11
+ },
+ {
+ "epoch": 0.04477511446269863,
+ "grad_norm": 0.8332845163143295,
+ "learning_rate": 9.987675550063625e-06,
+ "loss": 0.44841432571411133,
+ "num_input_tokens_seen": 28932512,
+ "step": 665,
+ "train_runtime": 5135.2428,
+ "train_tokens_per_second": 5634.108
+ },
+ {
+ "epoch": 0.04511176945865877,
+ "grad_norm": 1.0017265357539382,
+ "learning_rate": 9.987489319781355e-06,
+ "loss": 0.4431906223297119,
+ "num_input_tokens_seen": 29153640,
+ "step": 670,
+ "train_runtime": 5165.908,
+ "train_tokens_per_second": 5643.469
+ },
+ {
+ "epoch": 0.045448424454618906,
+ "grad_norm": 0.8381101119904498,
+ "learning_rate": 9.987301694763602e-06,
+ "loss": 0.4798222541809082,
+ "num_input_tokens_seen": 29368992,
+ "step": 675,
+ "train_runtime": 5201.341,
+ "train_tokens_per_second": 5646.427
+ },
+ {
+ "epoch": 0.045785079450579046,
+ "grad_norm": 0.9739030010713555,
+ "learning_rate": 9.987112675062834e-06,
+ "loss": 0.4200302600860596,
+ "num_input_tokens_seen": 29582872,
+ "step": 680,
+ "train_runtime": 5240.8223,
+ "train_tokens_per_second": 5644.7
+ },
+ {
+ "epoch": 0.046121734446539185,
+ "grad_norm": 0.9923641076223235,
+ "learning_rate": 9.986922260731908e-06,
+ "loss": 0.42931652069091797,
+ "num_input_tokens_seen": 29791056,
+ "step": 685,
+ "train_runtime": 5280.43,
+ "train_tokens_per_second": 5641.786
+ },
+ {
+ "epoch": 0.046458389442499325,
+ "grad_norm": 0.8786561915980274,
+ "learning_rate": 9.986730451824073e-06,
+ "loss": 0.4656373977661133,
+ "num_input_tokens_seen": 30001648,
+ "step": 690,
+ "train_runtime": 5325.5768,
+ "train_tokens_per_second": 5633.502
+ },
+ {
+ "epoch": 0.046795044438459464,
+ "grad_norm": 0.9155252116619295,
+ "learning_rate": 9.986537248392969e-06,
+ "loss": 0.4756736755371094,
+ "num_input_tokens_seen": 30223896,
+ "step": 695,
+ "train_runtime": 5365.8859,
+ "train_tokens_per_second": 5632.601
+ },
+ {
+ "epoch": 0.047131699434419604,
+ "grad_norm": 1.0345486643678798,
+ "learning_rate": 9.986342650492626e-06,
+ "loss": 0.4506352424621582,
+ "num_input_tokens_seen": 30445352,
+ "step": 700,
+ "train_runtime": 5399.4485,
+ "train_tokens_per_second": 5638.604
+ },
+ {
+ "epoch": 0.04746835443037975,
+ "grad_norm": 0.919553088389443,
+ "learning_rate": 9.98614665817746e-06,
+ "loss": 0.42203187942504883,
+ "num_input_tokens_seen": 30642128,
+ "step": 705,
+ "train_runtime": 5439.3583,
+ "train_tokens_per_second": 5633.409
+ },
+ {
+ "epoch": 0.04780500942633989,
+ "grad_norm": 0.9271703985798777,
+ "learning_rate": 9.98594927150228e-06,
+ "loss": 0.427701473236084,
+ "num_input_tokens_seen": 30873224,
+ "step": 710,
+ "train_runtime": 5478.7451,
+ "train_tokens_per_second": 5635.09
+ },
+ {
+ "epoch": 0.04814166442230003,
+ "grad_norm": 0.9508538400970585,
+ "learning_rate": 9.985750490522286e-06,
+ "loss": 0.39564857482910154,
+ "num_input_tokens_seen": 31087512,
+ "step": 715,
+ "train_runtime": 5521.2724,
+ "train_tokens_per_second": 5630.498
+ },
+ {
+ "epoch": 0.04847831941826017,
+ "grad_norm": 0.9372110505017411,
+ "learning_rate": 9.985550315293066e-06,
+ "loss": 0.4636754035949707,
+ "num_input_tokens_seen": 31318912,
+ "step": 720,
+ "train_runtime": 5557.4337,
+ "train_tokens_per_second": 5635.499
+ },
+ {
+ "epoch": 0.04881497441422031,
+ "grad_norm": 0.9970520932843633,
+ "learning_rate": 9.985348745870598e-06,
+ "loss": 0.4597306728363037,
+ "num_input_tokens_seen": 31546096,
+ "step": 725,
+ "train_runtime": 5592.8911,
+ "train_tokens_per_second": 5640.392
+ },
+ {
+ "epoch": 0.04915162941018045,
+ "grad_norm": 0.8271247348731543,
+ "learning_rate": 9.98514578231125e-06,
+ "loss": 0.4194610595703125,
+ "num_input_tokens_seen": 31774968,
+ "step": 730,
+ "train_runtime": 5630.8722,
+ "train_tokens_per_second": 5642.992
+ },
+ {
+ "epoch": 0.04948828440614059,
+ "grad_norm": 0.9452005671897916,
+ "learning_rate": 9.984941424671781e-06,
+ "loss": 0.4705995559692383,
+ "num_input_tokens_seen": 31991384,
+ "step": 735,
+ "train_runtime": 5672.8388,
+ "train_tokens_per_second": 5639.396
+ },
+ {
+ "epoch": 0.049824939402100726,
+ "grad_norm": 0.9711026791829106,
+ "learning_rate": 9.984735673009337e-06,
+ "loss": 0.43364248275756834,
+ "num_input_tokens_seen": 32211232,
+ "step": 740,
+ "train_runtime": 5717.5318,
+ "train_tokens_per_second": 5633.765
+ },
+ {
+ "epoch": 0.050161594398060866,
+ "grad_norm": 0.8583208108971163,
+ "learning_rate": 9.984528527381458e-06,
+ "loss": 0.4496613502502441,
+ "num_input_tokens_seen": 32432504,
+ "step": 745,
+ "train_runtime": 5758.1136,
+ "train_tokens_per_second": 5632.488
+ },
+ {
+ "epoch": 0.050498249394021005,
+ "grad_norm": 1.0983008551303333,
+ "learning_rate": 9.984319987846072e-06,
+ "loss": 0.4476910591125488,
+ "num_input_tokens_seen": 32644184,
+ "step": 750,
+ "train_runtime": 5795.8989,
+ "train_tokens_per_second": 5632.29
+ },
+ {
+ "epoch": 0.050834904389981145,
+ "grad_norm": 0.9341698177556362,
+ "learning_rate": 9.984110054461495e-06,
+ "loss": 0.4494589328765869,
+ "num_input_tokens_seen": 32883080,
+ "step": 755,
+ "train_runtime": 5831.2135,
+ "train_tokens_per_second": 5639.149
+ },
+ {
+ "epoch": 0.051171559385941284,
+ "grad_norm": 2.9705690058280205,
+ "learning_rate": 9.983898727286435e-06,
+ "loss": 0.44231281280517576,
+ "num_input_tokens_seen": 33102688,
+ "step": 760,
+ "train_runtime": 5864.3725,
+ "train_tokens_per_second": 5644.711
+ },
+ {
+ "epoch": 0.05150821438190143,
+ "grad_norm": 0.9183883314753698,
+ "learning_rate": 9.983686006379985e-06,
+ "loss": 0.45505475997924805,
+ "num_input_tokens_seen": 33319336,
+ "step": 765,
+ "train_runtime": 5901.491,
+ "train_tokens_per_second": 5645.918
+ },
+ {
+ "epoch": 0.05184486937786157,
+ "grad_norm": 0.9313957301525887,
+ "learning_rate": 9.983471891801638e-06,
+ "loss": 0.437741756439209,
+ "num_input_tokens_seen": 33545192,
+ "step": 770,
+ "train_runtime": 5938.7864,
+ "train_tokens_per_second": 5648.493
+ },
+ {
+ "epoch": 0.05218152437382171,
+ "grad_norm": 0.8973015582826919,
+ "learning_rate": 9.983256383611267e-06,
+ "loss": 0.441961669921875,
+ "num_input_tokens_seen": 33767136,
+ "step": 775,
+ "train_runtime": 5974.9184,
+ "train_tokens_per_second": 5651.481
+ },
+ {
+ "epoch": 0.05251817936978185,
+ "grad_norm": 1.380873298582274,
+ "learning_rate": 9.983039481869138e-06,
+ "loss": 0.42367143630981446,
+ "num_input_tokens_seen": 33985712,
+ "step": 780,
+ "train_runtime": 6009.8556,
+ "train_tokens_per_second": 5654.996
+ },
+ {
+ "epoch": 0.05285483436574199,
+ "grad_norm": 0.9605495456782687,
+ "learning_rate": 9.98282118663591e-06,
+ "loss": 0.4347693920135498,
+ "num_input_tokens_seen": 34208000,
+ "step": 785,
+ "train_runtime": 6046.7667,
+ "train_tokens_per_second": 5657.238
+ },
+ {
+ "epoch": 0.05319148936170213,
+ "grad_norm": 0.9570184798319451,
+ "learning_rate": 9.982601497972624e-06,
+ "loss": 0.4524986267089844,
+ "num_input_tokens_seen": 34425040,
+ "step": 790,
+ "train_runtime": 6085.7762,
+ "train_tokens_per_second": 5656.639
+ },
+ {
+ "epoch": 0.05352814435766227,
+ "grad_norm": 1.293632659414026,
+ "learning_rate": 9.982380415940717e-06,
+ "loss": 0.4259688377380371,
+ "num_input_tokens_seen": 34634712,
+ "step": 795,
+ "train_runtime": 6120.1944,
+ "train_tokens_per_second": 5659.087
+ },
+ {
+ "epoch": 0.05386479935362241,
+ "grad_norm": 0.9262120217395047,
+ "learning_rate": 9.982157940602015e-06,
+ "loss": 0.46526622772216797,
+ "num_input_tokens_seen": 34865376,
+ "step": 800,
+ "train_runtime": 6156.1798,
+ "train_tokens_per_second": 5663.476
+ },
+ {
+ "epoch": 0.05420145434958255,
+ "grad_norm": 0.892992639712076,
+ "learning_rate": 9.981934072018732e-06,
+ "loss": 0.44102907180786133,
+ "num_input_tokens_seen": 35098448,
+ "step": 805,
+ "train_runtime": 6189.6138,
+ "train_tokens_per_second": 5670.539
+ },
+ {
+ "epoch": 0.054538109345542686,
+ "grad_norm": 0.8797013169732244,
+ "learning_rate": 9.981708810253472e-06,
+ "loss": 0.4654733657836914,
+ "num_input_tokens_seen": 35317184,
+ "step": 810,
+ "train_runtime": 6223.9712,
+ "train_tokens_per_second": 5674.381
+ },
+ {
+ "epoch": 0.054874764341502826,
+ "grad_norm": 1.1385550766773516,
+ "learning_rate": 9.981482155369229e-06,
+ "loss": 0.4581299781799316,
+ "num_input_tokens_seen": 35543024,
+ "step": 815,
+ "train_runtime": 6264.1777,
+ "train_tokens_per_second": 5674.013
+ },
+ {
+ "epoch": 0.055211419337462965,
+ "grad_norm": 1.0108160361368144,
+ "learning_rate": 9.981254107429385e-06,
+ "loss": 0.4073037147521973,
+ "num_input_tokens_seen": 35752976,
+ "step": 820,
+ "train_runtime": 6301.9605,
+ "train_tokens_per_second": 5673.31
+ },
+ {
+ "epoch": 0.05554807433342311,
+ "grad_norm": 0.8271492373593813,
+ "learning_rate": 9.981024666497716e-06,
+ "loss": 0.4087496757507324,
+ "num_input_tokens_seen": 35970808,
+ "step": 825,
+ "train_runtime": 6336.9136,
+ "train_tokens_per_second": 5676.392
+ },
+ {
+ "epoch": 0.05588472932938325,
+ "grad_norm": 1.2448406201928153,
+ "learning_rate": 9.98079383263838e-06,
+ "loss": 0.451265811920166,
+ "num_input_tokens_seen": 36175264,
+ "step": 830,
+ "train_runtime": 6372.1661,
+ "train_tokens_per_second": 5677.075
+ },
+ {
+ "epoch": 0.05622138432534339,
+ "grad_norm": 1.0960090803895477,
+ "learning_rate": 9.980561605915932e-06,
+ "loss": 0.43454971313476565,
+ "num_input_tokens_seen": 36399968,
+ "step": 835,
+ "train_runtime": 6409.522,
+ "train_tokens_per_second": 5679.046
+ },
+ {
+ "epoch": 0.05655803932130353,
+ "grad_norm": 0.865350004624438,
+ "learning_rate": 9.980327986395316e-06,
+ "loss": 0.44866161346435546,
+ "num_input_tokens_seen": 36614352,
+ "step": 840,
+ "train_runtime": 6450.8502,
+ "train_tokens_per_second": 5675.896
+ },
+ {
+ "epoch": 0.05689469431726367,
+ "grad_norm": 0.9132401043618537,
+ "learning_rate": 9.980092974141855e-06,
+ "loss": 0.43543548583984376,
+ "num_input_tokens_seen": 36842440,
+ "step": 845,
+ "train_runtime": 6489.8829,
+ "train_tokens_per_second": 5676.904
+ },
+ {
+ "epoch": 0.05723134931322381,
+ "grad_norm": 0.9515826536482608,
+ "learning_rate": 9.979856569221276e-06,
+ "loss": 0.434114933013916,
+ "num_input_tokens_seen": 37064928,
+ "step": 850,
+ "train_runtime": 6535.3585,
+ "train_tokens_per_second": 5671.445
+ },
+ {
+ "epoch": 0.05756800430918395,
+ "grad_norm": 0.905243476379667,
+ "learning_rate": 9.979618771699688e-06,
+ "loss": 0.47480268478393556,
+ "num_input_tokens_seen": 37286696,
+ "step": 855,
+ "train_runtime": 6571.3163,
+ "train_tokens_per_second": 5674.159
+ },
+ {
+ "epoch": 0.05790465930514409,
+ "grad_norm": 0.9973188371129253,
+ "learning_rate": 9.979379581643588e-06,
+ "loss": 0.4271403312683105,
+ "num_input_tokens_seen": 37505104,
+ "step": 860,
+ "train_runtime": 6612.0758,
+ "train_tokens_per_second": 5672.213
+ },
+ {
+ "epoch": 0.05824131430110423,
+ "grad_norm": 0.891663822548588,
+ "learning_rate": 9.979138999119866e-06,
+ "loss": 0.46260747909545896,
+ "num_input_tokens_seen": 37721104,
+ "step": 865,
+ "train_runtime": 6657.2054,
+ "train_tokens_per_second": 5666.207
+ },
+ {
+ "epoch": 0.05857796929706437,
+ "grad_norm": 0.8193884764682856,
+ "learning_rate": 9.978897024195801e-06,
+ "loss": 0.43666720390319824,
+ "num_input_tokens_seen": 37959848,
+ "step": 870,
+ "train_runtime": 6692.9435,
+ "train_tokens_per_second": 5671.622
+ },
+ {
+ "epoch": 0.058914624293024506,
+ "grad_norm": 1.1669037394876847,
+ "learning_rate": 9.978653656939058e-06,
+ "loss": 0.4272760391235352,
+ "num_input_tokens_seen": 38179560,
+ "step": 875,
+ "train_runtime": 6736.655,
+ "train_tokens_per_second": 5667.436
+ },
+ {
+ "epoch": 0.059251279288984646,
+ "grad_norm": 1.0468766117182693,
+ "learning_rate": 9.978408897417699e-06,
+ "loss": 0.4613694190979004,
+ "num_input_tokens_seen": 38405040,
+ "step": 880,
+ "train_runtime": 6776.3735,
+ "train_tokens_per_second": 5667.492
+ },
+ {
+ "epoch": 0.059587934284944785,
+ "grad_norm": 0.8352176899590911,
+ "learning_rate": 9.978162745700165e-06,
+ "loss": 0.4417067527770996,
+ "num_input_tokens_seen": 38631232,
+ "step": 885,
+ "train_runtime": 6814.8051,
+ "train_tokens_per_second": 5668.721
+ },
+ {
+ "epoch": 0.05992458928090493,
+ "grad_norm": 1.0696188249598721,
+ "learning_rate": 9.977915201855293e-06,
+ "loss": 0.42413697242736814,
+ "num_input_tokens_seen": 38849712,
+ "step": 890,
+ "train_runtime": 6850.9387,
+ "train_tokens_per_second": 5670.714
+ },
+ {
+ "epoch": 0.06026124427686507,
+ "grad_norm": 0.9281445460422576,
+ "learning_rate": 9.977666265952309e-06,
+ "loss": 0.4329718589782715,
+ "num_input_tokens_seen": 39067416,
+ "step": 895,
+ "train_runtime": 6890.6802,
+ "train_tokens_per_second": 5669.602
+ },
+ {
+ "epoch": 0.06059789927282521,
+ "grad_norm": 0.8685965904886482,
+ "learning_rate": 9.977415938060823e-06,
+ "loss": 0.4434166431427002,
+ "num_input_tokens_seen": 39288624,
+ "step": 900,
+ "train_runtime": 6926.1271,
+ "train_tokens_per_second": 5672.524
+ },
+ {
+ "epoch": 0.06093455426878535,
+ "grad_norm": 0.9417641978342199,
+ "learning_rate": 9.977164218250846e-06,
+ "loss": 0.444110107421875,
+ "num_input_tokens_seen": 39505328,
+ "step": 905,
+ "train_runtime": 6968.8615,
+ "train_tokens_per_second": 5668.835
+ },
+ {
+ "epoch": 0.06127120926474549,
+ "grad_norm": 0.8557175345597763,
+ "learning_rate": 9.976911106592765e-06,
+ "loss": 0.4071072578430176,
+ "num_input_tokens_seen": 39717144,
+ "step": 910,
+ "train_runtime": 7009.1269,
+ "train_tokens_per_second": 5666.49
+ },
+ {
+ "epoch": 0.06160786426070563,
+ "grad_norm": 1.1399112064858985,
+ "learning_rate": 9.976656603157363e-06,
+ "loss": 0.4384052276611328,
+ "num_input_tokens_seen": 39933000,
+ "step": 915,
+ "train_runtime": 7045.7591,
+ "train_tokens_per_second": 5667.665
+ },
+ {
+ "epoch": 0.06194451925666577,
+ "grad_norm": 0.9115769460735773,
+ "learning_rate": 9.976400708015811e-06,
+ "loss": 0.42393884658813474,
+ "num_input_tokens_seen": 40145168,
+ "step": 920,
+ "train_runtime": 7079.4173,
+ "train_tokens_per_second": 5670.688
+ },
+ {
+ "epoch": 0.06228117425262591,
+ "grad_norm": 0.8414795692849011,
+ "learning_rate": 9.97614342123967e-06,
+ "loss": 0.4269254684448242,
+ "num_input_tokens_seen": 40379264,
+ "step": 925,
+ "train_runtime": 7111.1373,
+ "train_tokens_per_second": 5678.313
+ },
+ {
+ "epoch": 0.06261782924858605,
+ "grad_norm": 0.9352365482641929,
+ "learning_rate": 9.97588474290089e-06,
+ "loss": 0.4331661224365234,
+ "num_input_tokens_seen": 40594944,
+ "step": 930,
+ "train_runtime": 7152.76,
+ "train_tokens_per_second": 5675.424
+ },
+ {
+ "epoch": 0.0629544842445462,
+ "grad_norm": 1.0614310774414917,
+ "learning_rate": 9.975624673071806e-06,
+ "loss": 0.4371979236602783,
+ "num_input_tokens_seen": 40808856,
+ "step": 935,
+ "train_runtime": 7197.718,
+ "train_tokens_per_second": 5669.694
+ },
+ {
+ "epoch": 0.06329113924050633,
+ "grad_norm": 1.0079121416060646,
+ "learning_rate": 9.97536321182515e-06,
+ "loss": 0.4044677734375,
+ "num_input_tokens_seen": 41018224,
+ "step": 940,
+ "train_runtime": 7227.9115,
+ "train_tokens_per_second": 5674.976
+ },
+ {
+ "epoch": 0.06362779423646647,
+ "grad_norm": 0.9987688379119654,
+ "learning_rate": 9.975100359234037e-06,
+ "loss": 0.4564094066619873,
+ "num_input_tokens_seen": 41239392,
+ "step": 945,
+ "train_runtime": 7265.9862,
+ "train_tokens_per_second": 5675.677
+ },
+ {
+ "epoch": 0.06396444923242661,
+ "grad_norm": 0.9165255721069369,
+ "learning_rate": 9.974836115371972e-06,
+ "loss": 0.46511135101318357,
+ "num_input_tokens_seen": 41454352,
+ "step": 950,
+ "train_runtime": 7304.5238,
+ "train_tokens_per_second": 5675.161
+ },
+ {
+ "epoch": 0.06430110422838675,
+ "grad_norm": 0.9087923038014859,
+ "learning_rate": 9.974570480312852e-06,
+ "loss": 0.40421581268310547,
+ "num_input_tokens_seen": 41680576,
+ "step": 955,
+ "train_runtime": 7346.7849,
+ "train_tokens_per_second": 5673.308
+ },
+ {
+ "epoch": 0.06463775922434689,
+ "grad_norm": 1.0231903244202105,
+ "learning_rate": 9.974303454130962e-06,
+ "loss": 0.4464391231536865,
+ "num_input_tokens_seen": 41882360,
+ "step": 960,
+ "train_runtime": 7380.8428,
+ "train_tokens_per_second": 5674.469
+ },
+ {
+ "epoch": 0.06497441422030703,
+ "grad_norm": 0.8595399990273458,
+ "learning_rate": 9.97403503690097e-06,
+ "loss": 0.42637953758239744,
+ "num_input_tokens_seen": 42103128,
+ "step": 965,
+ "train_runtime": 7415.439,
+ "train_tokens_per_second": 5677.766
+ },
+ {
+ "epoch": 0.06531106921626717,
+ "grad_norm": 0.8422231085957664,
+ "learning_rate": 9.973765228697942e-06,
+ "loss": 0.39764556884765623,
+ "num_input_tokens_seen": 42328192,
+ "step": 970,
+ "train_runtime": 7453.2347,
+ "train_tokens_per_second": 5679.171
+ },
+ {
+ "epoch": 0.06564772421222731,
+ "grad_norm": 0.9247166172844082,
+ "learning_rate": 9.973494029597327e-06,
+ "loss": 0.43375320434570314,
+ "num_input_tokens_seen": 42531664,
+ "step": 975,
+ "train_runtime": 7487.6836,
+ "train_tokens_per_second": 5680.217
+ },
+ {
+ "epoch": 0.06598437920818745,
+ "grad_norm": 1.0151970555059149,
+ "learning_rate": 9.973221439674968e-06,
+ "loss": 0.4309089660644531,
+ "num_input_tokens_seen": 42740856,
+ "step": 980,
+ "train_runtime": 7527.1188,
+ "train_tokens_per_second": 5678.249
+ },
+ {
+ "epoch": 0.06632103420414759,
+ "grad_norm": 0.882649641197766,
+ "learning_rate": 9.972947459007089e-06,
+ "loss": 0.45548033714294434,
+ "num_input_tokens_seen": 42956288,
+ "step": 985,
+ "train_runtime": 7562.7131,
+ "train_tokens_per_second": 5680.01
+ },
+ {
+ "epoch": 0.06665768920010773,
+ "grad_norm": 0.9104741890803707,
+ "learning_rate": 9.972672087670313e-06,
+ "loss": 0.4385095119476318,
+ "num_input_tokens_seen": 43177272,
+ "step": 990,
+ "train_runtime": 7599.4941,
+ "train_tokens_per_second": 5681.598
+ },
+ {
+ "epoch": 0.06699434419606787,
+ "grad_norm": 0.8108451361340878,
+ "learning_rate": 9.972395325741644e-06,
+ "loss": 0.413985538482666,
+ "num_input_tokens_seen": 43400152,
+ "step": 995,
+ "train_runtime": 7640.1047,
+ "train_tokens_per_second": 5680.57
+ },
+ {
+ "epoch": 0.06733099919202801,
+ "grad_norm": 0.9025973522517067,
+ "learning_rate": 9.972117173298476e-06,
+ "loss": 0.38621523380279543,
+ "num_input_tokens_seen": 43614968,
+ "step": 1000,
+ "train_runtime": 7674.2806,
+ "train_tokens_per_second": 5683.265
+ },
+ {
+ "epoch": 0.06766765418798815,
+ "grad_norm": 0.971243318554936,
+ "learning_rate": 9.971837630418597e-06,
+ "loss": 0.42676420211791993,
+ "num_input_tokens_seen": 43835904,
+ "step": 1005,
+ "train_runtime": 7717.1168,
+ "train_tokens_per_second": 5680.347
+ },
+ {
+ "epoch": 0.06800430918394829,
+ "grad_norm": 0.9278937330533324,
+ "learning_rate": 9.971556697180179e-06,
+ "loss": 0.44001970291137693,
+ "num_input_tokens_seen": 44048512,
+ "step": 1010,
+ "train_runtime": 7756.4813,
+ "train_tokens_per_second": 5678.93
+ },
+ {
+ "epoch": 0.06834096417990843,
+ "grad_norm": 0.9429692040194982,
+ "learning_rate": 9.971274373661784e-06,
+ "loss": 0.4194046974182129,
+ "num_input_tokens_seen": 44271000,
+ "step": 1015,
+ "train_runtime": 7787.4236,
+ "train_tokens_per_second": 5684.935
+ },
+ {
+ "epoch": 0.06867761917586857,
+ "grad_norm": 1.111228553028809,
+ "learning_rate": 9.97099065994236e-06,
+ "loss": 0.39097912311553956,
+ "num_input_tokens_seen": 44499576,
+ "step": 1020,
+ "train_runtime": 7822.0558,
+ "train_tokens_per_second": 5688.987
+ },
+ {
+ "epoch": 0.0690142741718287,
+ "grad_norm": 1.0049309947436895,
+ "learning_rate": 9.970705556101253e-06,
+ "loss": 0.43691391944885255,
+ "num_input_tokens_seen": 44702000,
+ "step": 1025,
+ "train_runtime": 7857.7887,
+ "train_tokens_per_second": 5688.878
+ },
+ {
+ "epoch": 0.06935092916778884,
+ "grad_norm": 0.8982051488791603,
+ "learning_rate": 9.970419062218186e-06,
+ "loss": 0.45786590576171876,
+ "num_input_tokens_seen": 44936832,
+ "step": 1030,
+ "train_runtime": 7899.821,
+ "train_tokens_per_second": 5688.336
+ },
+ {
+ "epoch": 0.06968758416374898,
+ "grad_norm": 0.9779591960140872,
+ "learning_rate": 9.970131178373276e-06,
+ "loss": 0.4594112396240234,
+ "num_input_tokens_seen": 45166720,
+ "step": 1035,
+ "train_runtime": 7937.6344,
+ "train_tokens_per_second": 5690.199
+ },
+ {
+ "epoch": 0.07002423915970914,
+ "grad_norm": 0.9403137936019832,
+ "learning_rate": 9.969841904647033e-06,
+ "loss": 0.48036918640136717,
+ "num_input_tokens_seen": 45386112,
+ "step": 1040,
+ "train_runtime": 7975.3873,
+ "train_tokens_per_second": 5690.772
+ },
+ {
+ "epoch": 0.07036089415566928,
+ "grad_norm": 0.9489869208797291,
+ "learning_rate": 9.969551241120349e-06,
+ "loss": 0.4358417510986328,
+ "num_input_tokens_seen": 45615968,
+ "step": 1045,
+ "train_runtime": 8008.3857,
+ "train_tokens_per_second": 5696.025
+ },
+ {
+ "epoch": 0.07069754915162942,
+ "grad_norm": 0.8565455428368292,
+ "learning_rate": 9.969259187874507e-06,
+ "loss": 0.40404400825500486,
+ "num_input_tokens_seen": 45846704,
+ "step": 1050,
+ "train_runtime": 8044.3345,
+ "train_tokens_per_second": 5699.254
+ },
+ {
+ "epoch": 0.07103420414758956,
+ "grad_norm": 0.9631616045121916,
+ "learning_rate": 9.968965744991178e-06,
+ "loss": 0.43367910385131836,
+ "num_input_tokens_seen": 46069384,
+ "step": 1055,
+ "train_runtime": 8075.3735,
+ "train_tokens_per_second": 5704.923
+ },
+ {
+ "epoch": 0.0713708591435497,
+ "grad_norm": 0.923891258768515,
+ "learning_rate": 9.968670912552422e-06,
+ "loss": 0.41605100631713865,
+ "num_input_tokens_seen": 46289280,
+ "step": 1060,
+ "train_runtime": 8110.2287,
+ "train_tokens_per_second": 5707.518
+ },
+ {
+ "epoch": 0.07170751413950983,
+ "grad_norm": 0.9199449654930428,
+ "learning_rate": 9.968374690640692e-06,
+ "loss": 0.4259674072265625,
+ "num_input_tokens_seen": 46513832,
+ "step": 1065,
+ "train_runtime": 8149.1273,
+ "train_tokens_per_second": 5707.83
+ },
+ {
+ "epoch": 0.07204416913546997,
+ "grad_norm": 0.9349037827416551,
+ "learning_rate": 9.968077079338822e-06,
+ "loss": 0.44624719619750974,
+ "num_input_tokens_seen": 46732528,
+ "step": 1070,
+ "train_runtime": 8183.763,
+ "train_tokens_per_second": 5710.396
+ },
+ {
+ "epoch": 0.07238082413143011,
+ "grad_norm": 0.9411351171686876,
+ "learning_rate": 9.967778078730038e-06,
+ "loss": 0.4426886558532715,
+ "num_input_tokens_seen": 46948464,
+ "step": 1075,
+ "train_runtime": 8220.2593,
+ "train_tokens_per_second": 5711.312
+ },
+ {
+ "epoch": 0.07271747912739025,
+ "grad_norm": 0.9433124357981822,
+ "learning_rate": 9.967477688897957e-06,
+ "loss": 0.44336910247802735,
+ "num_input_tokens_seen": 47171032,
+ "step": 1080,
+ "train_runtime": 8259.7389,
+ "train_tokens_per_second": 5710.959
+ },
+ {
+ "epoch": 0.07305413412335039,
+ "grad_norm": 0.8570101786695654,
+ "learning_rate": 9.967175909926578e-06,
+ "loss": 0.46260600090026854,
+ "num_input_tokens_seen": 47402968,
+ "step": 1085,
+ "train_runtime": 8293.6887,
+ "train_tokens_per_second": 5715.547
+ },
+ {
+ "epoch": 0.07339078911931053,
+ "grad_norm": 0.967581695257143,
+ "learning_rate": 9.966872741900296e-06,
+ "loss": 0.4760030746459961,
+ "num_input_tokens_seen": 47634040,
+ "step": 1090,
+ "train_runtime": 8342.26,
+ "train_tokens_per_second": 5709.968
+ },
+ {
+ "epoch": 0.07372744411527067,
+ "grad_norm": 1.3602823694752813,
+ "learning_rate": 9.96656818490389e-06,
+ "loss": 0.4523357391357422,
+ "num_input_tokens_seen": 47849424,
+ "step": 1095,
+ "train_runtime": 8378.4116,
+ "train_tokens_per_second": 5711.038
+ },
+ {
+ "epoch": 0.07406409911123081,
+ "grad_norm": 0.8798163617081339,
+ "learning_rate": 9.966262239022529e-06,
+ "loss": 0.48172898292541505,
+ "num_input_tokens_seen": 48074544,
+ "step": 1100,
+ "train_runtime": 8414.2774,
+ "train_tokens_per_second": 5713.449
+ },
+ {
+ "epoch": 0.07440075410719095,
+ "grad_norm": 0.8689729621566785,
+ "learning_rate": 9.965954904341768e-06,
+ "loss": 0.4594748497009277,
+ "num_input_tokens_seen": 48294808,
+ "step": 1105,
+ "train_runtime": 8453.2631,
+ "train_tokens_per_second": 5713.156
+ },
+ {
+ "epoch": 0.07473740910315109,
+ "grad_norm": 0.9298500312875317,
+ "learning_rate": 9.965646180947555e-06,
+ "loss": 0.44074625968933107,
+ "num_input_tokens_seen": 48509312,
+ "step": 1110,
+ "train_runtime": 8496.7093,
+ "train_tokens_per_second": 5709.188
+ },
+ {
+ "epoch": 0.07507406409911123,
+ "grad_norm": 0.9049700587066084,
+ "learning_rate": 9.96533606892622e-06,
+ "loss": 0.42275848388671877,
+ "num_input_tokens_seen": 48715504,
+ "step": 1115,
+ "train_runtime": 8533.5132,
+ "train_tokens_per_second": 5708.728
+ },
+ {
+ "epoch": 0.07541071909507137,
+ "grad_norm": 0.882919522521347,
+ "learning_rate": 9.965024568364487e-06,
+ "loss": 0.4448400497436523,
+ "num_input_tokens_seen": 48944392,
+ "step": 1120,
+ "train_runtime": 8571.3237,
+ "train_tokens_per_second": 5710.249
+ },
+ {
+ "epoch": 0.07574737409103151,
+ "grad_norm": 0.9374688066893563,
+ "learning_rate": 9.964711679349467e-06,
+ "loss": 0.41461782455444335,
+ "num_input_tokens_seen": 49155824,
+ "step": 1125,
+ "train_runtime": 8611.0892,
+ "train_tokens_per_second": 5708.433
+ },
+ {
+ "epoch": 0.07608402908699165,
+ "grad_norm": 0.7995128542211772,
+ "learning_rate": 9.964397401968656e-06,
+ "loss": 0.46135854721069336,
+ "num_input_tokens_seen": 49380912,
+ "step": 1130,
+ "train_runtime": 8650.8346,
+ "train_tokens_per_second": 5708.225
+ },
+ {
+ "epoch": 0.07642068408295179,
+ "grad_norm": 0.8607214899503258,
+ "learning_rate": 9.964081736309942e-06,
+ "loss": 0.41400885581970215,
+ "num_input_tokens_seen": 49617456,
+ "step": 1135,
+ "train_runtime": 8688.4749,
+ "train_tokens_per_second": 5710.721
+ },
+ {
+ "epoch": 0.07675733907891193,
+ "grad_norm": 0.8715957366665356,
+ "learning_rate": 9.963764682461599e-06,
+ "loss": 0.4346445083618164,
+ "num_input_tokens_seen": 49848584,
+ "step": 1140,
+ "train_runtime": 8722.1234,
+ "train_tokens_per_second": 5715.189
+ },
+ {
+ "epoch": 0.07709399407487207,
+ "grad_norm": 0.8561235152822287,
+ "learning_rate": 9.96344624051229e-06,
+ "loss": 0.4340538501739502,
+ "num_input_tokens_seen": 50065696,
+ "step": 1145,
+ "train_runtime": 8762.0281,
+ "train_tokens_per_second": 5713.939
+ },
+ {
+ "epoch": 0.0774306490708322,
+ "grad_norm": 0.929804620150381,
+ "learning_rate": 9.96312641055107e-06,
+ "loss": 0.4169576644897461,
+ "num_input_tokens_seen": 50283968,
+ "step": 1150,
+ "train_runtime": 8801.5092,
+ "train_tokens_per_second": 5713.107
+ },
+ {
+ "epoch": 0.07776730406679234,
+ "grad_norm": 0.752419655730966,
+ "learning_rate": 9.962805192667375e-06,
+ "loss": 0.41555070877075195,
+ "num_input_tokens_seen": 50500904,
+ "step": 1155,
+ "train_runtime": 8848.2982,
+ "train_tokens_per_second": 5707.414
+ },
+ {
+ "epoch": 0.07810395906275248,
+ "grad_norm": 1.0612801529163063,
+ "learning_rate": 9.962482586951033e-06,
+ "loss": 0.4202773094177246,
+ "num_input_tokens_seen": 50733408,
+ "step": 1160,
+ "train_runtime": 8889.0849,
+ "train_tokens_per_second": 5707.383
+ },
+ {
+ "epoch": 0.07844061405871264,
+ "grad_norm": 0.9990837210097588,
+ "learning_rate": 9.962158593492258e-06,
+ "loss": 0.442533016204834,
+ "num_input_tokens_seen": 50939416,
+ "step": 1165,
+ "train_runtime": 8930.0327,
+ "train_tokens_per_second": 5704.281
+ },
+ {
+ "epoch": 0.07877726905467278,
+ "grad_norm": 0.8491329656110541,
+ "learning_rate": 9.961833212381657e-06,
+ "loss": 0.4386699676513672,
+ "num_input_tokens_seen": 51160880,
+ "step": 1170,
+ "train_runtime": 8967.7862,
+ "train_tokens_per_second": 5704.962
+ },
+ {
+ "epoch": 0.07911392405063292,
+ "grad_norm": 0.8233317233024117,
+ "learning_rate": 9.961506443710221e-06,
+ "loss": 0.4202608108520508,
+ "num_input_tokens_seen": 51382192,
+ "step": 1175,
+ "train_runtime": 9009.4779,
+ "train_tokens_per_second": 5703.126
+ },
+ {
+ "epoch": 0.07945057904659306,
+ "grad_norm": 1.1740908151711091,
+ "learning_rate": 9.961178287569326e-06,
+ "loss": 0.4113266944885254,
+ "num_input_tokens_seen": 51610400,
+ "step": 1180,
+ "train_runtime": 9043.5038,
+ "train_tokens_per_second": 5706.903
+ },
+ {
+ "epoch": 0.0797872340425532,
+ "grad_norm": 0.9683671789990694,
+ "learning_rate": 9.960848744050747e-06,
+ "loss": 0.44871106147766116,
+ "num_input_tokens_seen": 51828672,
+ "step": 1185,
+ "train_runtime": 9085.7121,
+ "train_tokens_per_second": 5704.415
+ },
+ {
+ "epoch": 0.08012388903851334,
+ "grad_norm": 0.7384402440175095,
+ "learning_rate": 9.960517813246633e-06,
+ "loss": 0.44126176834106445,
+ "num_input_tokens_seen": 52038752,
+ "step": 1190,
+ "train_runtime": 9126.3974,
+ "train_tokens_per_second": 5702.004
+ },
+ {
+ "epoch": 0.08046054403447347,
+ "grad_norm": 0.9518446142368773,
+ "learning_rate": 9.96018549524953e-06,
+ "loss": 0.4168846607208252,
+ "num_input_tokens_seen": 52257272,
+ "step": 1195,
+ "train_runtime": 9165.705,
+ "train_tokens_per_second": 5701.391
+ },
+ {
+ "epoch": 0.08079719903043361,
+ "grad_norm": 0.9426346640072512,
+ "learning_rate": 9.959851790152371e-06,
+ "loss": 0.41556696891784667,
+ "num_input_tokens_seen": 52478032,
+ "step": 1200,
+ "train_runtime": 9204.8312,
+ "train_tokens_per_second": 5701.14
+ },
+ {
+ "epoch": 0.08113385402639375,
+ "grad_norm": 0.8462737285766736,
+ "learning_rate": 9.959516698048475e-06,
+ "loss": 0.3835282325744629,
+ "num_input_tokens_seen": 52684592,
+ "step": 1205,
+ "train_runtime": 9245.4602,
+ "train_tokens_per_second": 5698.428
+ },
+ {
+ "epoch": 0.08147050902235389,
+ "grad_norm": 0.8765077742279136,
+ "learning_rate": 9.959180219031551e-06,
+ "loss": 0.4106024742126465,
+ "num_input_tokens_seen": 52895048,
+ "step": 1210,
+ "train_runtime": 9284.6342,
+ "train_tokens_per_second": 5697.052
+ },
+ {
+ "epoch": 0.08180716401831403,
+ "grad_norm": 0.8799539327811132,
+ "learning_rate": 9.958842353195688e-06,
+ "loss": 0.4241660118103027,
+ "num_input_tokens_seen": 53122496,
+ "step": 1215,
+ "train_runtime": 9323.4499,
+ "train_tokens_per_second": 5697.73
+ },
+ {
+ "epoch": 0.08214381901427417,
+ "grad_norm": 0.8851017529748658,
+ "learning_rate": 9.958503100635377e-06,
+ "loss": 0.43586392402648927,
+ "num_input_tokens_seen": 53331344,
+ "step": 1220,
+ "train_runtime": 9359.4211,
+ "train_tokens_per_second": 5698.146
+ },
+ {
+ "epoch": 0.08248047401023431,
+ "grad_norm": 0.9611985454606802,
+ "learning_rate": 9.958162461445484e-06,
+ "loss": 0.4607269287109375,
+ "num_input_tokens_seen": 53560160,
+ "step": 1225,
+ "train_runtime": 9398.2675,
+ "train_tokens_per_second": 5698.94
+ },
+ {
+ "epoch": 0.08281712900619445,
+ "grad_norm": 0.9067497311826802,
+ "learning_rate": 9.957820435721271e-06,
+ "loss": 0.4349203586578369,
+ "num_input_tokens_seen": 53775560,
+ "step": 1230,
+ "train_runtime": 9435.4003,
+ "train_tokens_per_second": 5699.341
+ },
+ {
+ "epoch": 0.08315378400215459,
+ "grad_norm": 0.963108721688269,
+ "learning_rate": 9.95747702355838e-06,
+ "loss": 0.45670351982116697,
+ "num_input_tokens_seen": 54000808,
+ "step": 1235,
+ "train_runtime": 9470.0321,
+ "train_tokens_per_second": 5702.284
+ },
+ {
+ "epoch": 0.08349043899811473,
+ "grad_norm": 0.8513390607173822,
+ "learning_rate": 9.957132225052848e-06,
+ "loss": 0.4656820297241211,
+ "num_input_tokens_seen": 54218336,
+ "step": 1240,
+ "train_runtime": 9509.4298,
+ "train_tokens_per_second": 5701.534
+ },
+ {
+ "epoch": 0.08382709399407487,
+ "grad_norm": 0.8033639657062732,
+ "learning_rate": 9.956786040301098e-06,
+ "loss": 0.407243537902832,
+ "num_input_tokens_seen": 54435576,
+ "step": 1245,
+ "train_runtime": 9549.0218,
+ "train_tokens_per_second": 5700.644
+ },
+ {
+ "epoch": 0.08416374899003501,
+ "grad_norm": 0.9112338258739832,
+ "learning_rate": 9.956438469399936e-06,
+ "loss": 0.42780628204345705,
+ "num_input_tokens_seen": 54646448,
+ "step": 1250,
+ "train_runtime": 9584.5449,
+ "train_tokens_per_second": 5701.517
+ },
+ {
+ "epoch": 0.08450040398599515,
+ "grad_norm": 0.898524764750954,
+ "learning_rate": 9.956089512446562e-06,
+ "loss": 0.4281005859375,
+ "num_input_tokens_seen": 54849384,
+ "step": 1255,
+ "train_runtime": 9624.8869,
+ "train_tokens_per_second": 5698.704
+ },
+ {
+ "epoch": 0.08483705898195529,
+ "grad_norm": 0.9393978257221618,
+ "learning_rate": 9.955739169538559e-06,
+ "loss": 0.43431816101074217,
+ "num_input_tokens_seen": 55062944,
+ "step": 1260,
+ "train_runtime": 9664.3501,
+ "train_tokens_per_second": 5697.532
+ },
+ {
+ "epoch": 0.08517371397791543,
+ "grad_norm": 0.9086736133443781,
+ "learning_rate": 9.955387440773902e-06,
+ "loss": 0.4328322887420654,
+ "num_input_tokens_seen": 55267032,
+ "step": 1265,
+ "train_runtime": 9704.5196,
+ "train_tokens_per_second": 5694.979
+ },
+ {
+ "epoch": 0.08551036897387557,
+ "grad_norm": 1.1824810596928264,
+ "learning_rate": 9.955034326250947e-06,
+ "loss": 0.42726478576660154,
+ "num_input_tokens_seen": 55485176,
+ "step": 1270,
+ "train_runtime": 9740.2982,
+ "train_tokens_per_second": 5696.456
+ },
+ {
+ "epoch": 0.0858470239698357,
+ "grad_norm": 0.981019269531651,
+ "learning_rate": 9.954679826068444e-06,
+ "loss": 0.431905460357666,
+ "num_input_tokens_seen": 55697472,
+ "step": 1275,
+ "train_runtime": 9778.2861,
+ "train_tokens_per_second": 5696.036
+ },
+ {
+ "epoch": 0.08618367896579585,
+ "grad_norm": 0.9167947068742166,
+ "learning_rate": 9.954323940325526e-06,
+ "loss": 0.4367355823516846,
+ "num_input_tokens_seen": 55909208,
+ "step": 1280,
+ "train_runtime": 9815.8422,
+ "train_tokens_per_second": 5695.814
+ },
+ {
+ "epoch": 0.086520333961756,
+ "grad_norm": 0.8284962122347568,
+ "learning_rate": 9.953966669121715e-06,
+ "loss": 0.4562966346740723,
+ "num_input_tokens_seen": 56144336,
+ "step": 1285,
+ "train_runtime": 9855.4694,
+ "train_tokens_per_second": 5696.769
+ },
+ {
+ "epoch": 0.08685698895771614,
+ "grad_norm": 0.9908373153090002,
+ "learning_rate": 9.953608012556924e-06,
+ "loss": 0.4428553581237793,
+ "num_input_tokens_seen": 56370632,
+ "step": 1290,
+ "train_runtime": 9900.5066,
+ "train_tokens_per_second": 5693.712
+ },
+ {
+ "epoch": 0.08719364395367628,
+ "grad_norm": 0.8603920936193332,
+ "learning_rate": 9.953247970731449e-06,
+ "loss": 0.4518705368041992,
+ "num_input_tokens_seen": 56581824,
+ "step": 1295,
+ "train_runtime": 9941.9268,
+ "train_tokens_per_second": 5691.233
+ },
+ {
+ "epoch": 0.08753029894963642,
+ "grad_norm": 1.0065353834353872,
+ "learning_rate": 9.952886543745973e-06,
+ "loss": 0.45711684226989746,
+ "num_input_tokens_seen": 56797912,
+ "step": 1300,
+ "train_runtime": 9983.8473,
+ "train_tokens_per_second": 5688.98
+ },
+ {
+ "epoch": 0.08786695394559656,
+ "grad_norm": 0.9704989049669684,
+ "learning_rate": 9.952523731701566e-06,
+ "loss": 0.43997964859008787,
+ "num_input_tokens_seen": 57028400,
+ "step": 1305,
+ "train_runtime": 10023.0235,
+ "train_tokens_per_second": 5689.74
+ },
+ {
+ "epoch": 0.0882036089415567,
+ "grad_norm": 0.7746848672115061,
+ "learning_rate": 9.952159534699693e-06,
+ "loss": 0.39839365482330324,
+ "num_input_tokens_seen": 57239112,
+ "step": 1310,
+ "train_runtime": 10060.0427,
+ "train_tokens_per_second": 5689.748
+ },
+ {
+ "epoch": 0.08854026393751684,
+ "grad_norm": 0.8703667437003134,
+ "learning_rate": 9.951793952842197e-06,
+ "loss": 0.46162776947021483,
+ "num_input_tokens_seen": 57451840,
+ "step": 1315,
+ "train_runtime": 10090.9799,
+ "train_tokens_per_second": 5693.386
+ },
+ {
+ "epoch": 0.08887691893347698,
+ "grad_norm": 0.9970523155610451,
+ "learning_rate": 9.95142698623131e-06,
+ "loss": 0.4125502109527588,
+ "num_input_tokens_seen": 57649616,
+ "step": 1320,
+ "train_runtime": 10134.3943,
+ "train_tokens_per_second": 5688.511
+ },
+ {
+ "epoch": 0.08921357392943711,
+ "grad_norm": 0.8982754367798736,
+ "learning_rate": 9.951058634969657e-06,
+ "loss": 0.4594625473022461,
+ "num_input_tokens_seen": 57859504,
+ "step": 1325,
+ "train_runtime": 10173.3422,
+ "train_tokens_per_second": 5687.364
+ },
+ {
+ "epoch": 0.08955022892539725,
+ "grad_norm": 0.8441984007212928,
+ "learning_rate": 9.950688899160244e-06,
+ "loss": 0.42917823791503906,
+ "num_input_tokens_seen": 58092104,
+ "step": 1330,
+ "train_runtime": 10208.5539,
+ "train_tokens_per_second": 5690.532
+ },
+ {
+ "epoch": 0.0898868839213574,
+ "grad_norm": 0.8326739767212783,
+ "learning_rate": 9.950317778906469e-06,
+ "loss": 0.4345374584197998,
+ "num_input_tokens_seen": 58319256,
+ "step": 1335,
+ "train_runtime": 10249.3358,
+ "train_tokens_per_second": 5690.052
+ },
+ {
+ "epoch": 0.09022353891731753,
+ "grad_norm": 0.8294890352349942,
+ "learning_rate": 9.949945274312109e-06,
+ "loss": 0.40816512107849123,
+ "num_input_tokens_seen": 58540424,
+ "step": 1340,
+ "train_runtime": 10287.6498,
+ "train_tokens_per_second": 5690.359
+ },
+ {
+ "epoch": 0.09056019391327767,
+ "grad_norm": 0.968521031703479,
+ "learning_rate": 9.94957138548134e-06,
+ "loss": 0.4445356369018555,
+ "num_input_tokens_seen": 58736800,
+ "step": 1345,
+ "train_runtime": 10326.024,
+ "train_tokens_per_second": 5688.23
+ },
+ {
+ "epoch": 0.09089684890923781,
+ "grad_norm": 1.007040150938243,
+ "learning_rate": 9.949196112518715e-06,
+ "loss": 0.41266183853149413,
+ "num_input_tokens_seen": 58963840,
+ "step": 1350,
+ "train_runtime": 10361.3193,
+ "train_tokens_per_second": 5690.766
+ },
+ {
+ "epoch": 0.09123350390519795,
+ "grad_norm": 0.8518178617284553,
+ "learning_rate": 9.94881945552918e-06,
+ "loss": 0.43543500900268556,
+ "num_input_tokens_seen": 59186072,
+ "step": 1355,
+ "train_runtime": 10406.6964,
+ "train_tokens_per_second": 5687.307
+ },
+ {
+ "epoch": 0.09157015890115809,
+ "grad_norm": 0.9351761092832128,
+ "learning_rate": 9.948441414618064e-06,
+ "loss": 0.45530238151550295,
+ "num_input_tokens_seen": 59396096,
+ "step": 1360,
+ "train_runtime": 10448.5625,
+ "train_tokens_per_second": 5684.619
+ },
+ {
+ "epoch": 0.09190681389711823,
+ "grad_norm": 0.976556409327906,
+ "learning_rate": 9.948061989891085e-06,
+ "loss": 0.41725835800170896,
+ "num_input_tokens_seen": 59621960,
+ "step": 1365,
+ "train_runtime": 10484.4296,
+ "train_tokens_per_second": 5686.715
+ },
+ {
+ "epoch": 0.09224346889307837,
+ "grad_norm": 0.9413072947091584,
+ "learning_rate": 9.94768118145435e-06,
+ "loss": 0.4238771438598633,
+ "num_input_tokens_seen": 59834688,
+ "step": 1370,
+ "train_runtime": 10521.7898,
+ "train_tokens_per_second": 5686.74
+ },
+ {
+ "epoch": 0.09258012388903851,
+ "grad_norm": 0.8072753824687519,
+ "learning_rate": 9.947298989414349e-06,
+ "loss": 0.4197520732879639,
+ "num_input_tokens_seen": 60061512,
+ "step": 1375,
+ "train_runtime": 10562.6863,
+ "train_tokens_per_second": 5686.197
+ },
+ {
+ "epoch": 0.09291677888499865,
+ "grad_norm": 0.8243851964827824,
+ "learning_rate": 9.946915413877963e-06,
+ "loss": 0.4294243812561035,
+ "num_input_tokens_seen": 60289152,
+ "step": 1380,
+ "train_runtime": 10608.7449,
+ "train_tokens_per_second": 5682.967
+ },
+ {
+ "epoch": 0.09325343388095879,
+ "grad_norm": 0.8639433441786879,
+ "learning_rate": 9.946530454952454e-06,
+ "loss": 0.44629898071289065,
+ "num_input_tokens_seen": 60516576,
+ "step": 1385,
+ "train_runtime": 10649.4064,
+ "train_tokens_per_second": 5682.624
+ },
+ {
+ "epoch": 0.09359008887691893,
+ "grad_norm": 0.9381874141159465,
+ "learning_rate": 9.946144112745477e-06,
+ "loss": 0.4549723625183105,
+ "num_input_tokens_seen": 60744096,
+ "step": 1390,
+ "train_runtime": 10690.6621,
+ "train_tokens_per_second": 5681.977
+ },
+ {
+ "epoch": 0.09392674387287907,
+ "grad_norm": 0.9598970905259178,
+ "learning_rate": 9.94575638736507e-06,
+ "loss": 0.4265571594238281,
+ "num_input_tokens_seen": 60970464,
+ "step": 1395,
+ "train_runtime": 10730.6636,
+ "train_tokens_per_second": 5681.891
+ },
+ {
+ "epoch": 0.09426339886883921,
+ "grad_norm": 0.8882564184072146,
+ "learning_rate": 9.945367278919662e-06,
+ "loss": 0.469175910949707,
+ "num_input_tokens_seen": 61195536,
+ "step": 1400,
+ "train_runtime": 10771.6796,
+ "train_tokens_per_second": 5681.151
+ },
+ {
+ "epoch": 0.09460005386479936,
+ "grad_norm": 0.8345645071811238,
+ "learning_rate": 9.944976787518064e-06,
+ "loss": 0.43855986595153806,
+ "num_input_tokens_seen": 61415400,
+ "step": 1405,
+ "train_runtime": 10804.9607,
+ "train_tokens_per_second": 5684.0
+ },
+ {
+ "epoch": 0.0949367088607595,
+ "grad_norm": 1.0139184695279284,
+ "learning_rate": 9.944584913269474e-06,
+ "loss": 0.43834495544433594,
+ "num_input_tokens_seen": 61645440,
+ "step": 1410,
+ "train_runtime": 10846.1692,
+ "train_tokens_per_second": 5683.614
+ },
+ {
+ "epoch": 0.09527336385671964,
+ "grad_norm": 0.8171057323619708,
+ "learning_rate": 9.944191656283481e-06,
+ "loss": 0.41381168365478516,
+ "num_input_tokens_seen": 61868952,
+ "step": 1415,
+ "train_runtime": 10885.6156,
+ "train_tokens_per_second": 5683.551
+ },
+ {
+ "epoch": 0.09561001885267978,
+ "grad_norm": 0.8671727475302144,
+ "learning_rate": 9.943797016670059e-06,
+ "loss": 0.41778225898742677,
+ "num_input_tokens_seen": 62081744,
+ "step": 1420,
+ "train_runtime": 10924.1913,
+ "train_tokens_per_second": 5682.96
+ },
+ {
+ "epoch": 0.09594667384863992,
+ "grad_norm": 0.9139643587899275,
+ "learning_rate": 9.943400994539565e-06,
+ "loss": 0.41177783012390134,
+ "num_input_tokens_seen": 62299552,
+ "step": 1425,
+ "train_runtime": 10963.2034,
+ "train_tokens_per_second": 5682.605
+ },
+ {
+ "epoch": 0.09628332884460006,
+ "grad_norm": 0.8942790678032806,
+ "learning_rate": 9.943003590002744e-06,
+ "loss": 0.4324612617492676,
+ "num_input_tokens_seen": 62524424,
+ "step": 1430,
+ "train_runtime": 11003.5976,
+ "train_tokens_per_second": 5682.18
+ },
+ {
+ "epoch": 0.0966199838405602,
+ "grad_norm": 0.9134672447344226,
+ "learning_rate": 9.942604803170733e-06,
+ "loss": 0.4207320690155029,
+ "num_input_tokens_seen": 62736736,
+ "step": 1435,
+ "train_runtime": 11041.0071,
+ "train_tokens_per_second": 5682.157
+ },
+ {
+ "epoch": 0.09695663883652034,
+ "grad_norm": 1.078284335277287,
+ "learning_rate": 9.94220463415505e-06,
+ "loss": 0.4587702751159668,
+ "num_input_tokens_seen": 62950552,
+ "step": 1440,
+ "train_runtime": 11082.5134,
+ "train_tokens_per_second": 5680.169
+ },
+ {
+ "epoch": 0.09729329383248048,
+ "grad_norm": 0.8211105532018887,
+ "learning_rate": 9.9418030830676e-06,
+ "loss": 0.4369199752807617,
+ "num_input_tokens_seen": 63173304,
+ "step": 1445,
+ "train_runtime": 11119.7396,
+ "train_tokens_per_second": 5681.186
+ },
+ {
+ "epoch": 0.09762994882844062,
+ "grad_norm": 0.9975227482376734,
+ "learning_rate": 9.941400150020676e-06,
+ "loss": 0.45250940322875977,
+ "num_input_tokens_seen": 63376456,
+ "step": 1450,
+ "train_runtime": 11153.3434,
+ "train_tokens_per_second": 5682.283
+ },
+ {
+ "epoch": 0.09796660382440076,
+ "grad_norm": 0.875023853319168,
+ "learning_rate": 9.940995835126957e-06,
+ "loss": 0.4510367393493652,
+ "num_input_tokens_seen": 63598704,
+ "step": 1455,
+ "train_runtime": 11190.4484,
+ "train_tokens_per_second": 5683.303
+ },
+ {
+ "epoch": 0.0983032588203609,
+ "grad_norm": 0.7795257917882069,
+ "learning_rate": 9.940590138499508e-06,
+ "loss": 0.4070897102355957,
+ "num_input_tokens_seen": 63824984,
+ "step": 1460,
+ "train_runtime": 11228.0323,
+ "train_tokens_per_second": 5684.432
+ },
+ {
+ "epoch": 0.09863991381632103,
+ "grad_norm": 0.9802634486624268,
+ "learning_rate": 9.940183060251783e-06,
+ "loss": 0.424533748626709,
+ "num_input_tokens_seen": 64020440,
+ "step": 1465,
+ "train_runtime": 11269.0538,
+ "train_tokens_per_second": 5681.084
+ },
+ {
+ "epoch": 0.09897656881228117,
+ "grad_norm": 0.8129225031494536,
+ "learning_rate": 9.939774600497616e-06,
+ "loss": 0.3772727489471436,
+ "num_input_tokens_seen": 64237912,
+ "step": 1470,
+ "train_runtime": 11305.1846,
+ "train_tokens_per_second": 5682.164
+ },
+ {
+ "epoch": 0.09931322380824131,
+ "grad_norm": 0.876095085095231,
+ "learning_rate": 9.939364759351236e-06,
+ "loss": 0.43614888191223145,
+ "num_input_tokens_seen": 64452808,
+ "step": 1475,
+ "train_runtime": 11342.3271,
+ "train_tokens_per_second": 5682.503
+ },
+ {
+ "epoch": 0.09964987880420145,
+ "grad_norm": 0.8249502349163025,
+ "learning_rate": 9.938953536927249e-06,
+ "loss": 0.4052250862121582,
+ "num_input_tokens_seen": 64679536,
+ "step": 1480,
+ "train_runtime": 11380.8832,
+ "train_tokens_per_second": 5683.174
+ },
+ {
+ "epoch": 0.09998653380016159,
+ "grad_norm": 0.8004037055954487,
+ "learning_rate": 9.938540933340657e-06,
+ "loss": 0.43981304168701174,
+ "num_input_tokens_seen": 64907672,
+ "step": 1485,
+ "train_runtime": 11414.3496,
+ "train_tokens_per_second": 5686.498
+ },
+ {
+ "epoch": 0.10032318879612173,
+ "grad_norm": 0.7556539983829699,
+ "learning_rate": 9.938126948706839e-06,
+ "loss": 0.4235081195831299,
+ "num_input_tokens_seen": 65134656,
+ "step": 1490,
+ "train_runtime": 11451.1493,
+ "train_tokens_per_second": 5688.045
+ },
+ {
+ "epoch": 0.10065984379208187,
+ "grad_norm": 0.7974202081471553,
+ "learning_rate": 9.937711583141567e-06,
+ "loss": 0.41340045928955077,
+ "num_input_tokens_seen": 65353344,
+ "step": 1495,
+ "train_runtime": 11488.8753,
+ "train_tokens_per_second": 5688.402
+ },
+ {
+ "epoch": 0.10099649878804201,
+ "grad_norm": 0.9113055103223531,
+ "learning_rate": 9.937294836760995e-06,
+ "loss": 0.43211612701416013,
+ "num_input_tokens_seen": 65570904,
+ "step": 1500,
+ "train_runtime": 11525.754,
+ "train_tokens_per_second": 5689.077
+ },
+ {
+ "epoch": 0.10133315378400215,
+ "grad_norm": 0.9362482868425025,
+ "learning_rate": 9.936876709681668e-06,
+ "loss": 0.4765140533447266,
+ "num_input_tokens_seen": 65784200,
+ "step": 1505,
+ "train_runtime": 11567.0824,
+ "train_tokens_per_second": 5687.19
+ },
+ {
+ "epoch": 0.10166980877996229,
+ "grad_norm": 0.7345279591609306,
+ "learning_rate": 9.93645720202051e-06,
+ "loss": 0.3693789005279541,
+ "num_input_tokens_seen": 66002240,
+ "step": 1510,
+ "train_runtime": 11603.3588,
+ "train_tokens_per_second": 5688.201
+ },
+ {
+ "epoch": 0.10200646377592243,
+ "grad_norm": 0.8440448421494396,
+ "learning_rate": 9.936036313894836e-06,
+ "loss": 0.46529722213745117,
+ "num_input_tokens_seen": 66217160,
+ "step": 1515,
+ "train_runtime": 11637.8846,
+ "train_tokens_per_second": 5689.793
+ },
+ {
+ "epoch": 0.10234311877188257,
+ "grad_norm": 0.9288537021329782,
+ "learning_rate": 9.935614045422349e-06,
+ "loss": 0.4474299907684326,
+ "num_input_tokens_seen": 66437216,
+ "step": 1520,
+ "train_runtime": 11674.3858,
+ "train_tokens_per_second": 5690.853
+ },
+ {
+ "epoch": 0.10267977376784271,
+ "grad_norm": 0.7816250239377797,
+ "learning_rate": 9.935190396721132e-06,
+ "loss": 0.4093039035797119,
+ "num_input_tokens_seen": 66659632,
+ "step": 1525,
+ "train_runtime": 11714.7321,
+ "train_tokens_per_second": 5690.24
+ },
+ {
+ "epoch": 0.10301642876380286,
+ "grad_norm": 0.868350723790852,
+ "learning_rate": 9.934765367909658e-06,
+ "loss": 0.442030143737793,
+ "num_input_tokens_seen": 66874200,
+ "step": 1530,
+ "train_runtime": 11751.0775,
+ "train_tokens_per_second": 5690.899
+ },
+ {
+ "epoch": 0.103353083759763,
+ "grad_norm": 0.9654553160320544,
+ "learning_rate": 9.934338959106783e-06,
+ "loss": 0.4125958442687988,
+ "num_input_tokens_seen": 67096584,
+ "step": 1535,
+ "train_runtime": 11787.6988,
+ "train_tokens_per_second": 5692.085
+ },
+ {
+ "epoch": 0.10368973875572314,
+ "grad_norm": 1.0001250180683936,
+ "learning_rate": 9.933911170431753e-06,
+ "loss": 0.45245895385742185,
+ "num_input_tokens_seen": 67300344,
+ "step": 1540,
+ "train_runtime": 11825.444,
+ "train_tokens_per_second": 5691.147
+ },
+ {
+ "epoch": 0.10402639375168328,
+ "grad_norm": 0.9480950052674862,
+ "learning_rate": 9.933482002004198e-06,
+ "loss": 0.4369757652282715,
+ "num_input_tokens_seen": 67528208,
+ "step": 1545,
+ "train_runtime": 11868.3124,
+ "train_tokens_per_second": 5689.79
+ },
+ {
+ "epoch": 0.10436304874764342,
+ "grad_norm": 1.049191818835591,
+ "learning_rate": 9.933051453944134e-06,
+ "loss": 0.47658519744873046,
+ "num_input_tokens_seen": 67752368,
+ "step": 1550,
+ "train_runtime": 11899.3433,
+ "train_tokens_per_second": 5693.791
+ },
+ {
+ "epoch": 0.10469970374360356,
+ "grad_norm": 0.8519840991796084,
+ "learning_rate": 9.932619526371958e-06,
+ "loss": 0.4100327014923096,
+ "num_input_tokens_seen": 67972272,
+ "step": 1555,
+ "train_runtime": 11933.0537,
+ "train_tokens_per_second": 5696.134
+ },
+ {
+ "epoch": 0.1050363587395637,
+ "grad_norm": 0.8587262694316211,
+ "learning_rate": 9.932186219408463e-06,
+ "loss": 0.40094866752624514,
+ "num_input_tokens_seen": 68182312,
+ "step": 1560,
+ "train_runtime": 11973.5123,
+ "train_tokens_per_second": 5694.429
+ },
+ {
+ "epoch": 0.10537301373552384,
+ "grad_norm": 0.9397299864907294,
+ "learning_rate": 9.931751533174819e-06,
+ "loss": 0.44454145431518555,
+ "num_input_tokens_seen": 68389480,
+ "step": 1565,
+ "train_runtime": 12014.0801,
+ "train_tokens_per_second": 5692.444
+ },
+ {
+ "epoch": 0.10570966873148398,
+ "grad_norm": 0.9745376383539546,
+ "learning_rate": 9.931315467792584e-06,
+ "loss": 0.45273237228393554,
+ "num_input_tokens_seen": 68625472,
+ "step": 1570,
+ "train_runtime": 12045.9213,
+ "train_tokens_per_second": 5696.988
+ },
+ {
+ "epoch": 0.10604632372744412,
+ "grad_norm": 0.796989168878711,
+ "learning_rate": 9.930878023383705e-06,
+ "loss": 0.4475082874298096,
+ "num_input_tokens_seen": 68842256,
+ "step": 1575,
+ "train_runtime": 12081.8379,
+ "train_tokens_per_second": 5697.995
+ },
+ {
+ "epoch": 0.10638297872340426,
+ "grad_norm": 0.7777899346165759,
+ "learning_rate": 9.93043920007051e-06,
+ "loss": 0.39358346462249755,
+ "num_input_tokens_seen": 69063336,
+ "step": 1580,
+ "train_runtime": 12120.9279,
+ "train_tokens_per_second": 5697.859
+ },
+ {
+ "epoch": 0.1067196337193644,
+ "grad_norm": 0.856296656865995,
+ "learning_rate": 9.929998997975712e-06,
+ "loss": 0.40912957191467286,
+ "num_input_tokens_seen": 69297392,
+ "step": 1585,
+ "train_runtime": 12157.5274,
+ "train_tokens_per_second": 5699.958
+ },
+ {
+ "epoch": 0.10705628871532454,
+ "grad_norm": 0.8406798760366443,
+ "learning_rate": 9.929557417222417e-06,
+ "loss": 0.4154947280883789,
+ "num_input_tokens_seen": 69502272,
+ "step": 1590,
+ "train_runtime": 12195.0467,
+ "train_tokens_per_second": 5699.221
+ },
+ {
+ "epoch": 0.10739294371128467,
+ "grad_norm": 0.8506547801105618,
+ "learning_rate": 9.929114457934109e-06,
+ "loss": 0.433414888381958,
+ "num_input_tokens_seen": 69718712,
+ "step": 1595,
+ "train_runtime": 12235.8524,
+ "train_tokens_per_second": 5697.904
+ },
+ {
+ "epoch": 0.10772959870724481,
+ "grad_norm": 0.8306444786085215,
+ "learning_rate": 9.92867012023466e-06,
+ "loss": 0.429976749420166,
+ "num_input_tokens_seen": 69942920,
+ "step": 1600,
+ "train_runtime": 12275.1053,
+ "train_tokens_per_second": 5697.949
+ },
+ {
+ "epoch": 0.10806625370320495,
+ "grad_norm": 0.8083632874129185,
+ "learning_rate": 9.92822440424833e-06,
+ "loss": 0.465771484375,
+ "num_input_tokens_seen": 70167408,
+ "step": 1605,
+ "train_runtime": 12311.856,
+ "train_tokens_per_second": 5699.174
+ },
+ {
+ "epoch": 0.1084029086991651,
+ "grad_norm": 0.8623025104703945,
+ "learning_rate": 9.927777310099759e-06,
+ "loss": 0.4795114040374756,
+ "num_input_tokens_seen": 70398096,
+ "step": 1610,
+ "train_runtime": 12355.4012,
+ "train_tokens_per_second": 5697.759
+ },
+ {
+ "epoch": 0.10873956369512523,
+ "grad_norm": 0.8734280942811361,
+ "learning_rate": 9.927328837913976e-06,
+ "loss": 0.42298336029052735,
+ "num_input_tokens_seen": 70621368,
+ "step": 1615,
+ "train_runtime": 12392.0717,
+ "train_tokens_per_second": 5698.915
+ },
+ {
+ "epoch": 0.10907621869108537,
+ "grad_norm": 0.8755533500883431,
+ "learning_rate": 9.926878987816397e-06,
+ "loss": 0.4128897666931152,
+ "num_input_tokens_seen": 70850496,
+ "step": 1620,
+ "train_runtime": 12432.0003,
+ "train_tokens_per_second": 5699.042
+ },
+ {
+ "epoch": 0.10941287368704551,
+ "grad_norm": 0.7849562344468639,
+ "learning_rate": 9.92642775993282e-06,
+ "loss": 0.40384368896484374,
+ "num_input_tokens_seen": 71078912,
+ "step": 1625,
+ "train_runtime": 12472.6839,
+ "train_tokens_per_second": 5698.766
+ },
+ {
+ "epoch": 0.10974952868300565,
+ "grad_norm": 0.8314524460612794,
+ "learning_rate": 9.925975154389428e-06,
+ "loss": 0.39646482467651367,
+ "num_input_tokens_seen": 71291944,
+ "step": 1630,
+ "train_runtime": 12512.7857,
+ "train_tokens_per_second": 5697.528
+ },
+ {
+ "epoch": 0.11008618367896579,
+ "grad_norm": 0.8397438217714046,
+ "learning_rate": 9.925521171312792e-06,
+ "loss": 0.41344375610351564,
+ "num_input_tokens_seen": 71513056,
+ "step": 1635,
+ "train_runtime": 12552.1031,
+ "train_tokens_per_second": 5697.297
+ },
+ {
+ "epoch": 0.11042283867492593,
+ "grad_norm": 0.8717786503893655,
+ "learning_rate": 9.925065810829868e-06,
+ "loss": 0.43964757919311526,
+ "num_input_tokens_seen": 71712776,
+ "step": 1640,
+ "train_runtime": 12592.8232,
+ "train_tokens_per_second": 5694.734
+ },
+ {
+ "epoch": 0.11075949367088607,
+ "grad_norm": 0.8466265387114118,
+ "learning_rate": 9.924609073067994e-06,
+ "loss": 0.45009913444519045,
+ "num_input_tokens_seen": 71918888,
+ "step": 1645,
+ "train_runtime": 12637.095,
+ "train_tokens_per_second": 5691.093
+ },
+ {
+ "epoch": 0.11109614866684622,
+ "grad_norm": 0.8202088060176734,
+ "learning_rate": 9.924150958154897e-06,
+ "loss": 0.43289871215820314,
+ "num_input_tokens_seen": 72144880,
+ "step": 1650,
+ "train_runtime": 12679.4137,
+ "train_tokens_per_second": 5689.922
+ },
+ {
+ "epoch": 0.11143280366280636,
+ "grad_norm": 0.8241473767552461,
+ "learning_rate": 9.923691466218686e-06,
+ "loss": 0.4033914566040039,
+ "num_input_tokens_seen": 72368424,
+ "step": 1655,
+ "train_runtime": 12722.761,
+ "train_tokens_per_second": 5688.107
+ },
+ {
+ "epoch": 0.1117694586587665,
+ "grad_norm": 0.9179601239629808,
+ "learning_rate": 9.923230597387856e-06,
+ "loss": 0.4340330123901367,
+ "num_input_tokens_seen": 72604440,
+ "step": 1660,
+ "train_runtime": 12760.3543,
+ "train_tokens_per_second": 5689.845
+ },
+ {
+ "epoch": 0.11210611365472664,
+ "grad_norm": 0.9667829858136964,
+ "learning_rate": 9.922768351791289e-06,
+ "loss": 0.45328373908996583,
+ "num_input_tokens_seen": 72822440,
+ "step": 1665,
+ "train_runtime": 12804.6293,
+ "train_tokens_per_second": 5687.196
+ },
+ {
+ "epoch": 0.11244276865068678,
+ "grad_norm": 0.9239911563296863,
+ "learning_rate": 9.92230472955825e-06,
+ "loss": 0.41826558113098145,
+ "num_input_tokens_seen": 73042800,
+ "step": 1670,
+ "train_runtime": 12838.5402,
+ "train_tokens_per_second": 5689.338
+ },
+ {
+ "epoch": 0.11277942364664692,
+ "grad_norm": 0.8636254640310012,
+ "learning_rate": 9.92183973081839e-06,
+ "loss": 0.3946269750595093,
+ "num_input_tokens_seen": 73268056,
+ "step": 1675,
+ "train_runtime": 12882.6482,
+ "train_tokens_per_second": 5687.344
+ },
+ {
+ "epoch": 0.11311607864260706,
+ "grad_norm": 0.7861261803530913,
+ "learning_rate": 9.921373355701743e-06,
+ "loss": 0.46369400024414065,
+ "num_input_tokens_seen": 73499832,
+ "step": 1680,
+ "train_runtime": 12924.32,
+ "train_tokens_per_second": 5686.94
+ },
+ {
+ "epoch": 0.1134527336385672,
+ "grad_norm": 0.8384955151965564,
+ "learning_rate": 9.920905604338732e-06,
+ "loss": 0.45543785095214845,
+ "num_input_tokens_seen": 73720264,
+ "step": 1685,
+ "train_runtime": 12963.3961,
+ "train_tokens_per_second": 5686.802
+ },
+ {
+ "epoch": 0.11378938863452734,
+ "grad_norm": 1.0833650625320947,
+ "learning_rate": 9.920436476860158e-06,
+ "loss": 0.38962454795837403,
+ "num_input_tokens_seen": 73931824,
+ "step": 1690,
+ "train_runtime": 13004.1451,
+ "train_tokens_per_second": 5685.251
+ },
+ {
+ "epoch": 0.11412604363048748,
+ "grad_norm": 0.8250025055357804,
+ "learning_rate": 9.919965973397214e-06,
+ "loss": 0.40491409301757814,
+ "num_input_tokens_seen": 74150736,
+ "step": 1695,
+ "train_runtime": 13037.0352,
+ "train_tokens_per_second": 5687.699
+ },
+ {
+ "epoch": 0.11446269862644762,
+ "grad_norm": 0.841239081826391,
+ "learning_rate": 9.919494094081475e-06,
+ "loss": 0.45804147720336913,
+ "num_input_tokens_seen": 74352136,
+ "step": 1700,
+ "train_runtime": 13073.6778,
+ "train_tokens_per_second": 5687.163
+ },
+ {
+ "epoch": 0.11479935362240776,
+ "grad_norm": 0.9629476764947537,
+ "learning_rate": 9.919020839044899e-06,
+ "loss": 0.4247718334197998,
+ "num_input_tokens_seen": 74588328,
+ "step": 1705,
+ "train_runtime": 13108.3031,
+ "train_tokens_per_second": 5690.159
+ },
+ {
+ "epoch": 0.1151360086183679,
+ "grad_norm": 0.8166202703347624,
+ "learning_rate": 9.918546208419832e-06,
+ "loss": 0.4040073394775391,
+ "num_input_tokens_seen": 74810232,
+ "step": 1710,
+ "train_runtime": 13152.6429,
+ "train_tokens_per_second": 5687.848
+ },
+ {
+ "epoch": 0.11547266361432804,
+ "grad_norm": 0.9143911353222378,
+ "learning_rate": 9.918070202339e-06,
+ "loss": 0.41643967628479006,
+ "num_input_tokens_seen": 75038248,
+ "step": 1715,
+ "train_runtime": 13188.009,
+ "train_tokens_per_second": 5689.885
+ },
+ {
+ "epoch": 0.11580931861028818,
+ "grad_norm": 0.854163675009985,
+ "learning_rate": 9.91759282093552e-06,
+ "loss": 0.4160469532012939,
+ "num_input_tokens_seen": 75259984,
+ "step": 1720,
+ "train_runtime": 13221.5917,
+ "train_tokens_per_second": 5692.203
+ },
+ {
+ "epoch": 0.11614597360624832,
+ "grad_norm": 0.9111231961657891,
+ "learning_rate": 9.917114064342888e-06,
+ "loss": 0.38855648040771484,
+ "num_input_tokens_seen": 75468520,
+ "step": 1725,
+ "train_runtime": 13264.901,
+ "train_tokens_per_second": 5689.339
+ },
+ {
+ "epoch": 0.11648262860220845,
+ "grad_norm": 0.8667773958295081,
+ "learning_rate": 9.916633932694988e-06,
+ "loss": 0.3962873458862305,
+ "num_input_tokens_seen": 75690392,
+ "step": 1730,
+ "train_runtime": 13301.6754,
+ "train_tokens_per_second": 5690.29
+ },
+ {
+ "epoch": 0.1168192835981686,
+ "grad_norm": 0.8572351568987933,
+ "learning_rate": 9.916152426126086e-06,
+ "loss": 0.3940844774246216,
+ "num_input_tokens_seen": 75908632,
+ "step": 1735,
+ "train_runtime": 13340.3356,
+ "train_tokens_per_second": 5690.159
+ },
+ {
+ "epoch": 0.11715593859412873,
+ "grad_norm": 0.8920380227031728,
+ "learning_rate": 9.915669544770837e-06,
+ "loss": 0.41963586807250974,
+ "num_input_tokens_seen": 76143760,
+ "step": 1740,
+ "train_runtime": 13373.7492,
+ "train_tokens_per_second": 5693.524
+ },
+ {
+ "epoch": 0.11749259359008887,
+ "grad_norm": 0.8151573693579499,
+ "learning_rate": 9.915185288764272e-06,
+ "loss": 0.4006998062133789,
+ "num_input_tokens_seen": 76381464,
+ "step": 1745,
+ "train_runtime": 13411.6709,
+ "train_tokens_per_second": 5695.149
+ },
+ {
+ "epoch": 0.11782924858604901,
+ "grad_norm": 0.7823689731999574,
+ "learning_rate": 9.914699658241815e-06,
+ "loss": 0.39071149826049806,
+ "num_input_tokens_seen": 76607736,
+ "step": 1750,
+ "train_runtime": 13455.8214,
+ "train_tokens_per_second": 5693.278
+ },
+ {
+ "epoch": 0.11816590358200915,
+ "grad_norm": 0.9650032542802038,
+ "learning_rate": 9.91421265333927e-06,
+ "loss": 0.44945535659790037,
+ "num_input_tokens_seen": 76836152,
+ "step": 1755,
+ "train_runtime": 13495.7655,
+ "train_tokens_per_second": 5693.353
+ },
+ {
+ "epoch": 0.11850255857796929,
+ "grad_norm": 0.8293434376681832,
+ "learning_rate": 9.913724274192829e-06,
+ "loss": 0.4147852897644043,
+ "num_input_tokens_seen": 77069784,
+ "step": 1760,
+ "train_runtime": 13537.8959,
+ "train_tokens_per_second": 5692.892
+ },
+ {
+ "epoch": 0.11883921357392943,
+ "grad_norm": 0.9258862128909312,
+ "learning_rate": 9.91323452093906e-06,
+ "loss": 0.41230435371398927,
+ "num_input_tokens_seen": 77288624,
+ "step": 1765,
+ "train_runtime": 13574.917,
+ "train_tokens_per_second": 5693.488
+ },
+ {
+ "epoch": 0.11917586856988957,
+ "grad_norm": 0.848930776896082,
+ "learning_rate": 9.912743393714927e-06,
+ "loss": 0.3945260286331177,
+ "num_input_tokens_seen": 77519384,
+ "step": 1770,
+ "train_runtime": 13605.814,
+ "train_tokens_per_second": 5697.519
+ },
+ {
+ "epoch": 0.11951252356584972,
+ "grad_norm": 0.8278888004348279,
+ "learning_rate": 9.912250892657769e-06,
+ "loss": 0.40961732864379885,
+ "num_input_tokens_seen": 77735336,
+ "step": 1775,
+ "train_runtime": 13644.1096,
+ "train_tokens_per_second": 5697.355
+ },
+ {
+ "epoch": 0.11984917856180986,
+ "grad_norm": 0.9501953857186206,
+ "learning_rate": 9.911757017905312e-06,
+ "loss": 0.41979494094848635,
+ "num_input_tokens_seen": 77943944,
+ "step": 1780,
+ "train_runtime": 13675.7254,
+ "train_tokens_per_second": 5699.438
+ },
+ {
+ "epoch": 0.12018583355777,
+ "grad_norm": 0.8175400797161728,
+ "learning_rate": 9.911261769595668e-06,
+ "loss": 0.4165688991546631,
+ "num_input_tokens_seen": 78176688,
+ "step": 1785,
+ "train_runtime": 13720.1102,
+ "train_tokens_per_second": 5697.964
+ },
+ {
+ "epoch": 0.12052248855373014,
+ "grad_norm": 0.9695481492778111,
+ "learning_rate": 9.910765147867332e-06,
+ "loss": 0.4288275718688965,
+ "num_input_tokens_seen": 78398920,
+ "step": 1790,
+ "train_runtime": 13754.606,
+ "train_tokens_per_second": 5699.83
+ },
+ {
+ "epoch": 0.12085914354969028,
+ "grad_norm": 0.9043853969674001,
+ "learning_rate": 9.91026715285918e-06,
+ "loss": 0.42678542137145997,
+ "num_input_tokens_seen": 78617400,
+ "step": 1795,
+ "train_runtime": 13795.7712,
+ "train_tokens_per_second": 5698.659
+ },
+ {
+ "epoch": 0.12119579854565042,
+ "grad_norm": 0.9643745725188065,
+ "learning_rate": 9.909767784710476e-06,
+ "loss": 0.4547886371612549,
+ "num_input_tokens_seen": 78858720,
+ "step": 1800,
+ "train_runtime": 13832.1459,
+ "train_tokens_per_second": 5701.12
+ },
+ {
+ "epoch": 0.12153245354161056,
+ "grad_norm": 0.8832616111060988,
+ "learning_rate": 9.909267043560868e-06,
+ "loss": 0.4295345783233643,
+ "num_input_tokens_seen": 79081264,
+ "step": 1805,
+ "train_runtime": 13867.5931,
+ "train_tokens_per_second": 5702.595
+ },
+ {
+ "epoch": 0.1218691085375707,
+ "grad_norm": 0.8001666467072275,
+ "learning_rate": 9.908764929550388e-06,
+ "loss": 0.4215728759765625,
+ "num_input_tokens_seen": 79301792,
+ "step": 1810,
+ "train_runtime": 13907.1584,
+ "train_tokens_per_second": 5702.228
+ },
+ {
+ "epoch": 0.12220576353353084,
+ "grad_norm": 0.9182690866647383,
+ "learning_rate": 9.908261442819444e-06,
+ "loss": 0.4215451717376709,
+ "num_input_tokens_seen": 79503840,
+ "step": 1815,
+ "train_runtime": 13955.1411,
+ "train_tokens_per_second": 5697.1
+ },
+ {
+ "epoch": 0.12254241852949098,
+ "grad_norm": 0.7856797264096972,
+ "learning_rate": 9.907756583508842e-06,
+ "loss": 0.3922400951385498,
+ "num_input_tokens_seen": 79740056,
+ "step": 1820,
+ "train_runtime": 13995.629,
+ "train_tokens_per_second": 5697.497
+ },
+ {
+ "epoch": 0.12287907352545112,
+ "grad_norm": 0.791802883681199,
+ "learning_rate": 9.907250351759761e-06,
+ "loss": 0.40973472595214844,
+ "num_input_tokens_seen": 79961848,
+ "step": 1825,
+ "train_runtime": 14035.6538,
+ "train_tokens_per_second": 5697.052
+ },
+ {
+ "epoch": 0.12321572852141126,
+ "grad_norm": 1.0344260725436623,
+ "learning_rate": 9.906742747713766e-06,
+ "loss": 0.4293796539306641,
+ "num_input_tokens_seen": 80180448,
+ "step": 1830,
+ "train_runtime": 14071.4195,
+ "train_tokens_per_second": 5698.107
+ },
+ {
+ "epoch": 0.1235523835173714,
+ "grad_norm": 1.0115683734845813,
+ "learning_rate": 9.906233771512808e-06,
+ "loss": 0.4379068374633789,
+ "num_input_tokens_seen": 80391288,
+ "step": 1835,
+ "train_runtime": 14108.8045,
+ "train_tokens_per_second": 5697.952
+ },
+ {
+ "epoch": 0.12388903851333154,
+ "grad_norm": 0.9113985281993394,
+ "learning_rate": 9.905723423299222e-06,
+ "loss": 0.44842100143432617,
+ "num_input_tokens_seen": 80619704,
+ "step": 1840,
+ "train_runtime": 14148.1364,
+ "train_tokens_per_second": 5698.256
+ },
+ {
+ "epoch": 0.12422569350929168,
+ "grad_norm": 0.8655689498703674,
+ "learning_rate": 9.905211703215723e-06,
+ "loss": 0.402877140045166,
+ "num_input_tokens_seen": 80832568,
+ "step": 1845,
+ "train_runtime": 14183.917,
+ "train_tokens_per_second": 5698.889
+ },
+ {
+ "epoch": 0.12456234850525182,
+ "grad_norm": 0.8753972943680609,
+ "learning_rate": 9.904698611405412e-06,
+ "loss": 0.4264510631561279,
+ "num_input_tokens_seen": 81058296,
+ "step": 1850,
+ "train_runtime": 14223.3173,
+ "train_tokens_per_second": 5698.973
+ },
+ {
+ "epoch": 0.12489900350121196,
+ "grad_norm": 0.7688694545812488,
+ "learning_rate": 9.904184148011774e-06,
+ "loss": 0.438901424407959,
+ "num_input_tokens_seen": 81274840,
+ "step": 1855,
+ "train_runtime": 14263.055,
+ "train_tokens_per_second": 5698.277
+ },
+ {
+ "epoch": 0.1252356584971721,
+ "grad_norm": 0.9233391292188109,
+ "learning_rate": 9.903668313178681e-06,
+ "loss": 0.44609222412109373,
+ "num_input_tokens_seen": 81500576,
+ "step": 1860,
+ "train_runtime": 14299.671,
+ "train_tokens_per_second": 5699.472
+ },
+ {
+ "epoch": 0.12557231349313225,
+ "grad_norm": 1.0943906900012805,
+ "learning_rate": 9.903151107050377e-06,
+ "loss": 0.396915864944458,
+ "num_input_tokens_seen": 81707648,
+ "step": 1865,
+ "train_runtime": 14337.8178,
+ "train_tokens_per_second": 5698.751
+ },
+ {
+ "epoch": 0.1259089684890924,
+ "grad_norm": 0.7168785500619042,
+ "learning_rate": 9.902632529771502e-06,
+ "loss": 0.42331228256225584,
+ "num_input_tokens_seen": 81926600,
+ "step": 1870,
+ "train_runtime": 14376.4633,
+ "train_tokens_per_second": 5698.662
+ },
+ {
+ "epoch": 0.12624562348505253,
+ "grad_norm": 1.0835704108533986,
+ "learning_rate": 9.902112581487074e-06,
+ "loss": 0.42675061225891114,
+ "num_input_tokens_seen": 82148216,
+ "step": 1875,
+ "train_runtime": 14416.4946,
+ "train_tokens_per_second": 5698.21
+ },
+ {
+ "epoch": 0.12658227848101267,
+ "grad_norm": 0.9867817197311566,
+ "learning_rate": 9.901591262342494e-06,
+ "loss": 0.4384635925292969,
+ "num_input_tokens_seen": 82358848,
+ "step": 1880,
+ "train_runtime": 14459.8702,
+ "train_tokens_per_second": 5695.684
+ },
+ {
+ "epoch": 0.1269189334769728,
+ "grad_norm": 0.9106268020914177,
+ "learning_rate": 9.901068572483548e-06,
+ "loss": 0.4060489654541016,
+ "num_input_tokens_seen": 82576440,
+ "step": 1885,
+ "train_runtime": 14497.7161,
+ "train_tokens_per_second": 5695.824
+ },
+ {
+ "epoch": 0.12725558847293295,
+ "grad_norm": 0.9009980190191491,
+ "learning_rate": 9.900544512056402e-06,
+ "loss": 0.47153472900390625,
+ "num_input_tokens_seen": 82807688,
+ "step": 1890,
+ "train_runtime": 14534.4393,
+ "train_tokens_per_second": 5697.343
+ },
+ {
+ "epoch": 0.12759224346889309,
+ "grad_norm": 0.8549638938689581,
+ "learning_rate": 9.900019081207612e-06,
+ "loss": 0.4209029197692871,
+ "num_input_tokens_seen": 83038168,
+ "step": 1895,
+ "train_runtime": 14569.2093,
+ "train_tokens_per_second": 5699.566
+ },
+ {
+ "epoch": 0.12792889846485322,
+ "grad_norm": 0.8968781974119465,
+ "learning_rate": 9.89949228008411e-06,
+ "loss": 0.42281227111816405,
+ "num_input_tokens_seen": 83267720,
+ "step": 1900,
+ "train_runtime": 14606.2305,
+ "train_tokens_per_second": 5700.836
+ },
+ {
+ "epoch": 0.12826555346081336,
+ "grad_norm": 0.8595709799025085,
+ "learning_rate": 9.898964108833216e-06,
+ "loss": 0.42650370597839354,
+ "num_input_tokens_seen": 83492616,
+ "step": 1905,
+ "train_runtime": 14646.7617,
+ "train_tokens_per_second": 5700.415
+ },
+ {
+ "epoch": 0.1286022084567735,
+ "grad_norm": 0.8815455063429253,
+ "learning_rate": 9.898434567602629e-06,
+ "loss": 0.4016720771789551,
+ "num_input_tokens_seen": 83715840,
+ "step": 1910,
+ "train_runtime": 14686.2575,
+ "train_tokens_per_second": 5700.284
+ },
+ {
+ "epoch": 0.12893886345273364,
+ "grad_norm": 0.9694766961124608,
+ "learning_rate": 9.897903656540437e-06,
+ "loss": 0.4289149284362793,
+ "num_input_tokens_seen": 83934184,
+ "step": 1915,
+ "train_runtime": 14724.9891,
+ "train_tokens_per_second": 5700.119
+ },
+ {
+ "epoch": 0.12927551844869378,
+ "grad_norm": 0.9823894967580632,
+ "learning_rate": 9.897371375795103e-06,
+ "loss": 0.43784418106079104,
+ "num_input_tokens_seen": 84150312,
+ "step": 1920,
+ "train_runtime": 14759.5087,
+ "train_tokens_per_second": 5701.43
+ },
+ {
+ "epoch": 0.12961217344465392,
+ "grad_norm": 0.8194539046535433,
+ "learning_rate": 9.896837725515484e-06,
+ "loss": 0.44328908920288085,
+ "num_input_tokens_seen": 84361600,
+ "step": 1925,
+ "train_runtime": 14805.8899,
+ "train_tokens_per_second": 5697.841
+ },
+ {
+ "epoch": 0.12994882844061406,
+ "grad_norm": 0.7081955649991436,
+ "learning_rate": 9.89630270585081e-06,
+ "loss": 0.39231128692626954,
+ "num_input_tokens_seen": 84577168,
+ "step": 1930,
+ "train_runtime": 14842.2012,
+ "train_tokens_per_second": 5698.425
+ },
+ {
+ "epoch": 0.1302854834365742,
+ "grad_norm": 0.8873554761311145,
+ "learning_rate": 9.895766316950694e-06,
+ "loss": 0.4394209384918213,
+ "num_input_tokens_seen": 84796464,
+ "step": 1935,
+ "train_runtime": 14876.8277,
+ "train_tokens_per_second": 5699.902
+ },
+ {
+ "epoch": 0.13062213843253434,
+ "grad_norm": 0.8262969710012915,
+ "learning_rate": 9.89522855896514e-06,
+ "loss": 0.41849603652954104,
+ "num_input_tokens_seen": 85025304,
+ "step": 1940,
+ "train_runtime": 14919.7496,
+ "train_tokens_per_second": 5698.843
+ },
+ {
+ "epoch": 0.13095879342849448,
+ "grad_norm": 0.9120986383703955,
+ "learning_rate": 9.89468943204453e-06,
+ "loss": 0.4164949893951416,
+ "num_input_tokens_seen": 85239728,
+ "step": 1945,
+ "train_runtime": 14964.5159,
+ "train_tokens_per_second": 5696.123
+ },
+ {
+ "epoch": 0.13129544842445462,
+ "grad_norm": 1.0478669869240167,
+ "learning_rate": 9.89414893633963e-06,
+ "loss": 0.41168971061706544,
+ "num_input_tokens_seen": 85451488,
+ "step": 1950,
+ "train_runtime": 14998.4779,
+ "train_tokens_per_second": 5697.344
+ },
+ {
+ "epoch": 0.13163210342041476,
+ "grad_norm": 0.8072074904847002,
+ "learning_rate": 9.893607072001584e-06,
+ "loss": 0.42524333000183107,
+ "num_input_tokens_seen": 85681416,
+ "step": 1955,
+ "train_runtime": 15045.5899,
+ "train_tokens_per_second": 5694.786
+ },
+ {
+ "epoch": 0.1319687584163749,
+ "grad_norm": 0.7884630260384243,
+ "learning_rate": 9.893063839181925e-06,
+ "loss": 0.41834263801574706,
+ "num_input_tokens_seen": 85897864,
+ "step": 1960,
+ "train_runtime": 15086.3562,
+ "train_tokens_per_second": 5693.745
+ },
+ {
+ "epoch": 0.13230541341233504,
+ "grad_norm": 0.8242479851716092,
+ "learning_rate": 9.892519238032566e-06,
+ "loss": 0.4272817611694336,
+ "num_input_tokens_seen": 86126184,
+ "step": 1965,
+ "train_runtime": 15130.1067,
+ "train_tokens_per_second": 5692.371
+ },
+ {
+ "epoch": 0.13264206840829518,
+ "grad_norm": 0.8883896038795247,
+ "learning_rate": 9.891973268705804e-06,
+ "loss": 0.45201525688171384,
+ "num_input_tokens_seen": 86340160,
+ "step": 1970,
+ "train_runtime": 15166.855,
+ "train_tokens_per_second": 5692.687
+ },
+ {
+ "epoch": 0.13297872340425532,
+ "grad_norm": 0.7899276150283895,
+ "learning_rate": 9.891425931354316e-06,
+ "loss": 0.4250628471374512,
+ "num_input_tokens_seen": 86565168,
+ "step": 1975,
+ "train_runtime": 15203.4839,
+ "train_tokens_per_second": 5693.772
+ },
+ {
+ "epoch": 0.13331537840021546,
+ "grad_norm": 0.8879719340880936,
+ "learning_rate": 9.890877226131165e-06,
+ "loss": 0.41506524085998536,
+ "num_input_tokens_seen": 86796552,
+ "step": 1980,
+ "train_runtime": 15247.3097,
+ "train_tokens_per_second": 5692.581
+ },
+ {
+ "epoch": 0.1336520333961756,
+ "grad_norm": 0.916769222493085,
+ "learning_rate": 9.890327153189793e-06,
+ "loss": 0.4300036907196045,
+ "num_input_tokens_seen": 87016968,
+ "step": 1985,
+ "train_runtime": 15292.8026,
+ "train_tokens_per_second": 5690.06
+ },
+ {
+ "epoch": 0.13398868839213574,
+ "grad_norm": 0.9050174324296295,
+ "learning_rate": 9.889775712684026e-06,
+ "loss": 0.4278897285461426,
+ "num_input_tokens_seen": 87240744,
+ "step": 1990,
+ "train_runtime": 15335.4097,
+ "train_tokens_per_second": 5688.843
+ },
+ {
+ "epoch": 0.13432534338809587,
+ "grad_norm": 0.8780717724520911,
+ "learning_rate": 9.889222904768074e-06,
+ "loss": 0.42691898345947266,
+ "num_input_tokens_seen": 87458888,
+ "step": 1995,
+ "train_runtime": 15375.168,
+ "train_tokens_per_second": 5688.321
+ },
+ {
+ "epoch": 0.13466199838405601,
+ "grad_norm": 0.8749943321506966,
+ "learning_rate": 9.888668729596528e-06,
+ "loss": 0.41673784255981444,
+ "num_input_tokens_seen": 87661224,
+ "step": 2000,
+ "train_runtime": 15417.6729,
+ "train_tokens_per_second": 5685.762
+ },
+ {
+ "epoch": 0.13499865338001615,
+ "grad_norm": 0.9429761954235121,
+ "learning_rate": 9.88811318732436e-06,
+ "loss": 0.4196599006652832,
+ "num_input_tokens_seen": 87871224,
+ "step": 2005,
+ "train_runtime": 15454.7159,
+ "train_tokens_per_second": 5685.722
+ },
+ {
+ "epoch": 0.1353353083759763,
+ "grad_norm": 0.8303836661224822,
+ "learning_rate": 9.887556278106927e-06,
+ "loss": 0.43319454193115237,
+ "num_input_tokens_seen": 88101424,
+ "step": 2010,
+ "train_runtime": 15492.7733,
+ "train_tokens_per_second": 5686.614
+ },
+ {
+ "epoch": 0.13567196337193643,
+ "grad_norm": 1.2988017888943026,
+ "learning_rate": 9.886998002099964e-06,
+ "loss": 0.42733011245727537,
+ "num_input_tokens_seen": 88331320,
+ "step": 2015,
+ "train_runtime": 15533.3637,
+ "train_tokens_per_second": 5686.555
+ },
+ {
+ "epoch": 0.13600861836789657,
+ "grad_norm": 0.926333904941965,
+ "learning_rate": 9.886438359459595e-06,
+ "loss": 0.4501063346862793,
+ "num_input_tokens_seen": 88546592,
+ "step": 2020,
+ "train_runtime": 15572.5905,
+ "train_tokens_per_second": 5686.054
+ },
+ {
+ "epoch": 0.1363452733638567,
+ "grad_norm": 0.8812830943715737,
+ "learning_rate": 9.885877350342319e-06,
+ "loss": 0.3861879587173462,
+ "num_input_tokens_seen": 88774480,
+ "step": 2025,
+ "train_runtime": 15611.197,
+ "train_tokens_per_second": 5686.59
+ },
+ {
+ "epoch": 0.13668192835981685,
+ "grad_norm": 0.8007094863514034,
+ "learning_rate": 9.885314974905024e-06,
+ "loss": 0.3875585556030273,
+ "num_input_tokens_seen": 89011576,
+ "step": 2030,
+ "train_runtime": 15652.3238,
+ "train_tokens_per_second": 5686.796
+ },
+ {
+ "epoch": 0.137018583355777,
+ "grad_norm": 0.9125014304117678,
+ "learning_rate": 9.884751233304973e-06,
+ "loss": 0.39730267524719237,
+ "num_input_tokens_seen": 89217520,
+ "step": 2035,
+ "train_runtime": 15686.075,
+ "train_tokens_per_second": 5687.689
+ },
+ {
+ "epoch": 0.13735523835173713,
+ "grad_norm": 0.8474671445937998,
+ "learning_rate": 9.884186125699816e-06,
+ "loss": 0.42992243766784666,
+ "num_input_tokens_seen": 89431816,
+ "step": 2040,
+ "train_runtime": 15726.6605,
+ "train_tokens_per_second": 5686.637
+ },
+ {
+ "epoch": 0.13769189334769727,
+ "grad_norm": 0.9102001150203423,
+ "learning_rate": 9.883619652247584e-06,
+ "loss": 0.4123133659362793,
+ "num_input_tokens_seen": 89651968,
+ "step": 2045,
+ "train_runtime": 15768.4222,
+ "train_tokens_per_second": 5685.538
+ },
+ {
+ "epoch": 0.1380285483436574,
+ "grad_norm": 0.6804193522827889,
+ "learning_rate": 9.883051813106687e-06,
+ "loss": 0.3782381296157837,
+ "num_input_tokens_seen": 89874624,
+ "step": 2050,
+ "train_runtime": 15813.3958,
+ "train_tokens_per_second": 5683.449
+ },
+ {
+ "epoch": 0.13836520333961755,
+ "grad_norm": 0.8989541162037309,
+ "learning_rate": 9.882482608435924e-06,
+ "loss": 0.43795013427734375,
+ "num_input_tokens_seen": 90097216,
+ "step": 2055,
+ "train_runtime": 15851.8211,
+ "train_tokens_per_second": 5683.714
+ },
+ {
+ "epoch": 0.1387018583355777,
+ "grad_norm": 0.8434117541493056,
+ "learning_rate": 9.881912038394468e-06,
+ "loss": 0.42223529815673827,
+ "num_input_tokens_seen": 90317784,
+ "step": 2060,
+ "train_runtime": 15897.2071,
+ "train_tokens_per_second": 5681.362
+ },
+ {
+ "epoch": 0.13903851333153783,
+ "grad_norm": 0.7960282493279416,
+ "learning_rate": 9.881340103141878e-06,
+ "loss": 0.4242293357849121,
+ "num_input_tokens_seen": 90546968,
+ "step": 2065,
+ "train_runtime": 15940.8703,
+ "train_tokens_per_second": 5680.177
+ },
+ {
+ "epoch": 0.13937516832749797,
+ "grad_norm": 0.8413026941754382,
+ "learning_rate": 9.880766802838092e-06,
+ "loss": 0.4057013511657715,
+ "num_input_tokens_seen": 90762816,
+ "step": 2070,
+ "train_runtime": 15983.7133,
+ "train_tokens_per_second": 5678.456
+ },
+ {
+ "epoch": 0.1397118233234581,
+ "grad_norm": 0.7350898878720474,
+ "learning_rate": 9.880192137643434e-06,
+ "loss": 0.4075471878051758,
+ "num_input_tokens_seen": 90979408,
+ "step": 2075,
+ "train_runtime": 16017.6972,
+ "train_tokens_per_second": 5679.931
+ },
+ {
+ "epoch": 0.14004847831941827,
+ "grad_norm": 0.9141607370838677,
+ "learning_rate": 9.879616107718608e-06,
+ "loss": 0.4198787689208984,
+ "num_input_tokens_seen": 91197560,
+ "step": 2080,
+ "train_runtime": 16051.0779,
+ "train_tokens_per_second": 5681.709
+ },
+ {
+ "epoch": 0.1403851333153784,
+ "grad_norm": 0.7897438400002121,
+ "learning_rate": 9.879038713224695e-06,
+ "loss": 0.4208370685577393,
+ "num_input_tokens_seen": 91421872,
+ "step": 2085,
+ "train_runtime": 16093.9488,
+ "train_tokens_per_second": 5680.512
+ },
+ {
+ "epoch": 0.14072178831133855,
+ "grad_norm": 0.8205504842451986,
+ "learning_rate": 9.878459954323165e-06,
+ "loss": 0.4030628204345703,
+ "num_input_tokens_seen": 91647112,
+ "step": 2090,
+ "train_runtime": 16126.4411,
+ "train_tokens_per_second": 5683.034
+ },
+ {
+ "epoch": 0.1410584433072987,
+ "grad_norm": 0.8325253458879384,
+ "learning_rate": 9.877879831175865e-06,
+ "loss": 0.39506003856658933,
+ "num_input_tokens_seen": 91871896,
+ "step": 2095,
+ "train_runtime": 16160.6707,
+ "train_tokens_per_second": 5684.906
+ },
+ {
+ "epoch": 0.14139509830325883,
+ "grad_norm": 0.821317865316409,
+ "learning_rate": 9.877298343945025e-06,
+ "loss": 0.41777982711791994,
+ "num_input_tokens_seen": 92094456,
+ "step": 2100,
+ "train_runtime": 16203.3038,
+ "train_tokens_per_second": 5683.684
+ },
+ {
+ "epoch": 0.14173175329921897,
+ "grad_norm": 0.7872012950426368,
+ "learning_rate": 9.876715492793256e-06,
+ "loss": 0.41480426788330077,
+ "num_input_tokens_seen": 92315440,
+ "step": 2105,
+ "train_runtime": 16242.0831,
+ "train_tokens_per_second": 5683.719
+ },
+ {
+ "epoch": 0.1420684082951791,
+ "grad_norm": 0.8092561101034788,
+ "learning_rate": 9.87613127788355e-06,
+ "loss": 0.389608097076416,
+ "num_input_tokens_seen": 92527368,
+ "step": 2110,
+ "train_runtime": 16278.4874,
+ "train_tokens_per_second": 5684.027
+ },
+ {
+ "epoch": 0.14240506329113925,
+ "grad_norm": 0.8850860821410723,
+ "learning_rate": 9.87554569937928e-06,
+ "loss": 0.4558421611785889,
+ "num_input_tokens_seen": 92759672,
+ "step": 2115,
+ "train_runtime": 16314.9617,
+ "train_tokens_per_second": 5685.559
+ },
+ {
+ "epoch": 0.1427417182870994,
+ "grad_norm": 0.8891822985539864,
+ "learning_rate": 9.874958757444203e-06,
+ "loss": 0.45552263259887693,
+ "num_input_tokens_seen": 92995096,
+ "step": 2120,
+ "train_runtime": 16350.6171,
+ "train_tokens_per_second": 5687.559
+ },
+ {
+ "epoch": 0.14307837328305953,
+ "grad_norm": 0.8780645983704002,
+ "learning_rate": 9.874370452242454e-06,
+ "loss": 0.43328495025634767,
+ "num_input_tokens_seen": 93229944,
+ "step": 2125,
+ "train_runtime": 16384.0995,
+ "train_tokens_per_second": 5690.27
+ },
+ {
+ "epoch": 0.14341502827901967,
+ "grad_norm": 0.9402959761478824,
+ "learning_rate": 9.873780783938553e-06,
+ "loss": 0.4462221622467041,
+ "num_input_tokens_seen": 93450480,
+ "step": 2130,
+ "train_runtime": 16417.1629,
+ "train_tokens_per_second": 5692.243
+ },
+ {
+ "epoch": 0.1437516832749798,
+ "grad_norm": 0.8574770635812221,
+ "learning_rate": 9.873189752697396e-06,
+ "loss": 0.3947939395904541,
+ "num_input_tokens_seen": 93670472,
+ "step": 2135,
+ "train_runtime": 16459.3253,
+ "train_tokens_per_second": 5691.027
+ },
+ {
+ "epoch": 0.14408833827093995,
+ "grad_norm": 0.8947319210921507,
+ "learning_rate": 9.872597358684265e-06,
+ "loss": 0.458558464050293,
+ "num_input_tokens_seen": 93883544,
+ "step": 2140,
+ "train_runtime": 16489.5195,
+ "train_tokens_per_second": 5693.528
+ },
+ {
+ "epoch": 0.1444249932669001,
+ "grad_norm": 0.8538177526814695,
+ "learning_rate": 9.872003602064816e-06,
+ "loss": 0.4299613475799561,
+ "num_input_tokens_seen": 94116824,
+ "step": 2145,
+ "train_runtime": 16523.1274,
+ "train_tokens_per_second": 5696.066
+ },
+ {
+ "epoch": 0.14476164826286023,
+ "grad_norm": 0.7471306795366713,
+ "learning_rate": 9.8714084830051e-06,
+ "loss": 0.4079564571380615,
+ "num_input_tokens_seen": 94349480,
+ "step": 2150,
+ "train_runtime": 16564.7452,
+ "train_tokens_per_second": 5695.8
+ },
+ {
+ "epoch": 0.14509830325882037,
+ "grad_norm": 0.8218664635099376,
+ "learning_rate": 9.870812001671533e-06,
+ "loss": 0.40776524543762205,
+ "num_input_tokens_seen": 94577624,
+ "step": 2155,
+ "train_runtime": 16600.9249,
+ "train_tokens_per_second": 5697.13
+ },
+ {
+ "epoch": 0.1454349582547805,
+ "grad_norm": 1.0737295934493194,
+ "learning_rate": 9.870214158230922e-06,
+ "loss": 0.4177376747131348,
+ "num_input_tokens_seen": 94804128,
+ "step": 2160,
+ "train_runtime": 16635.1905,
+ "train_tokens_per_second": 5699.011
+ },
+ {
+ "epoch": 0.14577161325074064,
+ "grad_norm": 0.786781014793927,
+ "learning_rate": 9.869614952850454e-06,
+ "loss": 0.40900435447692873,
+ "num_input_tokens_seen": 95026192,
+ "step": 2165,
+ "train_runtime": 16669.2594,
+ "train_tokens_per_second": 5700.685
+ },
+ {
+ "epoch": 0.14610826824670078,
+ "grad_norm": 0.8360055927785602,
+ "learning_rate": 9.86901438569769e-06,
+ "loss": 0.42821083068847654,
+ "num_input_tokens_seen": 95254672,
+ "step": 2170,
+ "train_runtime": 16706.7948,
+ "train_tokens_per_second": 5701.553
+ },
+ {
+ "epoch": 0.14644492324266092,
+ "grad_norm": 0.8933164074109287,
+ "learning_rate": 9.86841245694058e-06,
+ "loss": 0.425125789642334,
+ "num_input_tokens_seen": 95450280,
+ "step": 2175,
+ "train_runtime": 16746.2086,
+ "train_tokens_per_second": 5699.814
+ },
+ {
+ "epoch": 0.14678157823862106,
+ "grad_norm": 0.8906147805897214,
+ "learning_rate": 9.86780916674745e-06,
+ "loss": 0.45731253623962403,
+ "num_input_tokens_seen": 95651304,
+ "step": 2180,
+ "train_runtime": 16783.8632,
+ "train_tokens_per_second": 5699.004
+ },
+ {
+ "epoch": 0.1471182332345812,
+ "grad_norm": 1.0525327714019657,
+ "learning_rate": 9.86720451528701e-06,
+ "loss": 0.40560040473937986,
+ "num_input_tokens_seen": 95859952,
+ "step": 2185,
+ "train_runtime": 16822.6242,
+ "train_tokens_per_second": 5698.276
+ },
+ {
+ "epoch": 0.14745488823054134,
+ "grad_norm": 0.816498968853468,
+ "learning_rate": 9.866598502728348e-06,
+ "loss": 0.43126611709594725,
+ "num_input_tokens_seen": 96069168,
+ "step": 2190,
+ "train_runtime": 16862.6901,
+ "train_tokens_per_second": 5697.144
+ },
+ {
+ "epoch": 0.14779154322650148,
+ "grad_norm": 0.841739118951704,
+ "learning_rate": 9.865991129240931e-06,
+ "loss": 0.41145782470703124,
+ "num_input_tokens_seen": 96283672,
+ "step": 2195,
+ "train_runtime": 16896.2709,
+ "train_tokens_per_second": 5698.516
+ },
+ {
+ "epoch": 0.14812819822246162,
+ "grad_norm": 0.878868960575425,
+ "learning_rate": 9.865382394994614e-06,
+ "loss": 0.40999040603637693,
+ "num_input_tokens_seen": 96502528,
+ "step": 2200,
+ "train_runtime": 16939.3291,
+ "train_tokens_per_second": 5696.951
+ },
+ {
+ "epoch": 0.14846485321842176,
+ "grad_norm": 0.6967776248464405,
+ "learning_rate": 9.864772300159622e-06,
+ "loss": 0.3990957021713257,
+ "num_input_tokens_seen": 96720064,
+ "step": 2205,
+ "train_runtime": 16973.4551,
+ "train_tokens_per_second": 5698.313
+ },
+ {
+ "epoch": 0.1488015082143819,
+ "grad_norm": 0.9032740614218077,
+ "learning_rate": 9.864160844906571e-06,
+ "loss": 0.43683781623840334,
+ "num_input_tokens_seen": 96941160,
+ "step": 2210,
+ "train_runtime": 17009.3974,
+ "train_tokens_per_second": 5699.271
+ },
+ {
+ "epoch": 0.14913816321034204,
+ "grad_norm": 0.8444434220131386,
+ "learning_rate": 9.863548029406449e-06,
+ "loss": 0.419160795211792,
+ "num_input_tokens_seen": 97161584,
+ "step": 2215,
+ "train_runtime": 17050.6002,
+ "train_tokens_per_second": 5698.426
+ },
+ {
+ "epoch": 0.14947481820630218,
+ "grad_norm": 0.7872642768120427,
+ "learning_rate": 9.86293385383063e-06,
+ "loss": 0.3916934490203857,
+ "num_input_tokens_seen": 97372312,
+ "step": 2220,
+ "train_runtime": 17095.7206,
+ "train_tokens_per_second": 5695.713
+ },
+ {
+ "epoch": 0.14981147320226232,
+ "grad_norm": 0.9092720440628393,
+ "learning_rate": 9.862318318350865e-06,
+ "loss": 0.4371610641479492,
+ "num_input_tokens_seen": 97594144,
+ "step": 2225,
+ "train_runtime": 17132.445,
+ "train_tokens_per_second": 5696.452
+ },
+ {
+ "epoch": 0.15014812819822246,
+ "grad_norm": 0.7914448222714119,
+ "learning_rate": 9.861701423139287e-06,
+ "loss": 0.39911181926727296,
+ "num_input_tokens_seen": 97820888,
+ "step": 2230,
+ "train_runtime": 17165.9269,
+ "train_tokens_per_second": 5698.55
+ },
+ {
+ "epoch": 0.1504847831941826,
+ "grad_norm": 0.8316284130963478,
+ "learning_rate": 9.861083168368411e-06,
+ "loss": 0.41806850433349607,
+ "num_input_tokens_seen": 98045776,
+ "step": 2235,
+ "train_runtime": 17204.7719,
+ "train_tokens_per_second": 5698.755
+ },
+ {
+ "epoch": 0.15082143819014274,
+ "grad_norm": 0.9088287061584689,
+ "learning_rate": 9.860463554211125e-06,
+ "loss": 0.3891399383544922,
+ "num_input_tokens_seen": 98268344,
+ "step": 2240,
+ "train_runtime": 17246.0796,
+ "train_tokens_per_second": 5698.011
+ },
+ {
+ "epoch": 0.15115809318610288,
+ "grad_norm": 0.9381951784290052,
+ "learning_rate": 9.859842580840706e-06,
+ "loss": 0.4640520095825195,
+ "num_input_tokens_seen": 98493000,
+ "step": 2245,
+ "train_runtime": 17289.4391,
+ "train_tokens_per_second": 5696.715
+ },
+ {
+ "epoch": 0.15149474818206302,
+ "grad_norm": 0.7432535531903158,
+ "learning_rate": 9.859220248430806e-06,
+ "loss": 0.41865386962890627,
+ "num_input_tokens_seen": 98731240,
+ "step": 2250,
+ "train_runtime": 17329.2789,
+ "train_tokens_per_second": 5697.366
+ },
+ {
+ "epoch": 0.15183140317802316,
+ "grad_norm": 0.7752690728179633,
+ "learning_rate": 9.85859655715546e-06,
+ "loss": 0.39298291206359864,
+ "num_input_tokens_seen": 98953224,
+ "step": 2255,
+ "train_runtime": 17375.0272,
+ "train_tokens_per_second": 5695.141
+ },
+ {
+ "epoch": 0.1521680581739833,
+ "grad_norm": 0.8610212435397341,
+ "learning_rate": 9.857971507189077e-06,
+ "loss": 0.41016359329223634,
+ "num_input_tokens_seen": 99171512,
+ "step": 2260,
+ "train_runtime": 17414.0463,
+ "train_tokens_per_second": 5694.915
+ },
+ {
+ "epoch": 0.15250471316994343,
+ "grad_norm": 0.8902354817859172,
+ "learning_rate": 9.857345098706455e-06,
+ "loss": 0.4255502700805664,
+ "num_input_tokens_seen": 99402080,
+ "step": 2265,
+ "train_runtime": 17451.574,
+ "train_tokens_per_second": 5695.88
+ },
+ {
+ "epoch": 0.15284136816590357,
+ "grad_norm": 0.9086587862733272,
+ "learning_rate": 9.856717331882765e-06,
+ "loss": 0.4265317440032959,
+ "num_input_tokens_seen": 99607944,
+ "step": 2270,
+ "train_runtime": 17486.7082,
+ "train_tokens_per_second": 5696.209
+ },
+ {
+ "epoch": 0.1531780231618637,
+ "grad_norm": 0.9432280136780563,
+ "learning_rate": 9.85608820689356e-06,
+ "loss": 0.39440102577209474,
+ "num_input_tokens_seen": 99830656,
+ "step": 2275,
+ "train_runtime": 17528.0918,
+ "train_tokens_per_second": 5695.466
+ },
+ {
+ "epoch": 0.15351467815782385,
+ "grad_norm": 0.84487279707662,
+ "learning_rate": 9.855457723914772e-06,
+ "loss": 0.4084943771362305,
+ "num_input_tokens_seen": 100050880,
+ "step": 2280,
+ "train_runtime": 17562.2543,
+ "train_tokens_per_second": 5696.927
+ },
+ {
+ "epoch": 0.153851333153784,
+ "grad_norm": 0.9062843980819959,
+ "learning_rate": 9.854825883122713e-06,
+ "loss": 0.4020846366882324,
+ "num_input_tokens_seen": 100260608,
+ "step": 2285,
+ "train_runtime": 17602.1432,
+ "train_tokens_per_second": 5695.932
+ },
+ {
+ "epoch": 0.15418798814974413,
+ "grad_norm": 0.765345818981944,
+ "learning_rate": 9.85419268469408e-06,
+ "loss": 0.39870567321777345,
+ "num_input_tokens_seen": 100466752,
+ "step": 2290,
+ "train_runtime": 17638.1952,
+ "train_tokens_per_second": 5695.977
+ },
+ {
+ "epoch": 0.15452464314570427,
+ "grad_norm": 0.8051620604928534,
+ "learning_rate": 9.85355812880594e-06,
+ "loss": 0.3938899517059326,
+ "num_input_tokens_seen": 100686544,
+ "step": 2295,
+ "train_runtime": 17677.5734,
+ "train_tokens_per_second": 5695.722
+ },
+ {
+ "epoch": 0.1548612981416644,
+ "grad_norm": 0.7457579703786071,
+ "learning_rate": 9.852922215635747e-06,
+ "loss": 0.37209017276763917,
+ "num_input_tokens_seen": 100908792,
+ "step": 2300,
+ "train_runtime": 17719.4452,
+ "train_tokens_per_second": 5694.805
+ },
+ {
+ "epoch": 0.15519795313762455,
+ "grad_norm": 0.8179184151627187,
+ "learning_rate": 9.85228494536133e-06,
+ "loss": 0.4363281726837158,
+ "num_input_tokens_seen": 101132472,
+ "step": 2305,
+ "train_runtime": 17753.0155,
+ "train_tokens_per_second": 5696.636
+ },
+ {
+ "epoch": 0.1555346081335847,
+ "grad_norm": 0.8389619253246731,
+ "learning_rate": 9.851646318160901e-06,
+ "loss": 0.41831111907958984,
+ "num_input_tokens_seen": 101349024,
+ "step": 2310,
+ "train_runtime": 17793.8083,
+ "train_tokens_per_second": 5695.747
+ },
+ {
+ "epoch": 0.15587126312954483,
+ "grad_norm": 0.7507564820666337,
+ "learning_rate": 9.85100633421305e-06,
+ "loss": 0.39124910831451415,
+ "num_input_tokens_seen": 101558936,
+ "step": 2315,
+ "train_runtime": 17835.905,
+ "train_tokens_per_second": 5694.072
+ },
+ {
+ "epoch": 0.15620791812550497,
+ "grad_norm": 0.8895097039801884,
+ "learning_rate": 9.850364993696749e-06,
+ "loss": 0.42600722312927247,
+ "num_input_tokens_seen": 101772184,
+ "step": 2320,
+ "train_runtime": 17875.446,
+ "train_tokens_per_second": 5693.407
+ },
+ {
+ "epoch": 0.15654457312146514,
+ "grad_norm": 0.9520687878700897,
+ "learning_rate": 9.849722296791339e-06,
+ "loss": 0.4083821773529053,
+ "num_input_tokens_seen": 101982520,
+ "step": 2325,
+ "train_runtime": 17909.3087,
+ "train_tokens_per_second": 5694.386
+ },
+ {
+ "epoch": 0.15688122811742528,
+ "grad_norm": 0.9282345518369051,
+ "learning_rate": 9.849078243676558e-06,
+ "loss": 0.4323273658752441,
+ "num_input_tokens_seen": 102196704,
+ "step": 2330,
+ "train_runtime": 17950.3047,
+ "train_tokens_per_second": 5693.313
+ },
+ {
+ "epoch": 0.15721788311338541,
+ "grad_norm": 0.7571706215220008,
+ "learning_rate": 9.848432834532506e-06,
+ "loss": 0.3904409885406494,
+ "num_input_tokens_seen": 102413872,
+ "step": 2335,
+ "train_runtime": 17990.1951,
+ "train_tokens_per_second": 5692.76
+ },
+ {
+ "epoch": 0.15755453810934555,
+ "grad_norm": 0.7812435917007698,
+ "learning_rate": 9.847786069539673e-06,
+ "loss": 0.42157392501831054,
+ "num_input_tokens_seen": 102636848,
+ "step": 2340,
+ "train_runtime": 18027.5652,
+ "train_tokens_per_second": 5693.328
+ },
+ {
+ "epoch": 0.1578911931053057,
+ "grad_norm": 0.8268032160233915,
+ "learning_rate": 9.847137948878926e-06,
+ "loss": 0.4385071277618408,
+ "num_input_tokens_seen": 102861544,
+ "step": 2345,
+ "train_runtime": 18065.377,
+ "train_tokens_per_second": 5693.85
+ },
+ {
+ "epoch": 0.15822784810126583,
+ "grad_norm": 0.9872902684937591,
+ "learning_rate": 9.846488472731505e-06,
+ "loss": 0.46036734580993655,
+ "num_input_tokens_seen": 103092600,
+ "step": 2350,
+ "train_runtime": 18100.6003,
+ "train_tokens_per_second": 5695.535
+ },
+ {
+ "epoch": 0.15856450309722597,
+ "grad_norm": 0.8069383596278824,
+ "learning_rate": 9.845837641279038e-06,
+ "loss": 0.4151791572570801,
+ "num_input_tokens_seen": 103323896,
+ "step": 2355,
+ "train_runtime": 18131.0861,
+ "train_tokens_per_second": 5698.715
+ },
+ {
+ "epoch": 0.1589011580931861,
+ "grad_norm": 0.918784982161899,
+ "learning_rate": 9.845185454703528e-06,
+ "loss": 0.42508544921875,
+ "num_input_tokens_seen": 103541848,
+ "step": 2360,
+ "train_runtime": 18171.0294,
+ "train_tokens_per_second": 5698.183
+ },
+ {
+ "epoch": 0.15923781308914625,
+ "grad_norm": 0.817441107070197,
+ "learning_rate": 9.844531913187356e-06,
+ "loss": 0.3977517604827881,
+ "num_input_tokens_seen": 103759048,
+ "step": 2365,
+ "train_runtime": 18209.2359,
+ "train_tokens_per_second": 5698.155
+ },
+ {
+ "epoch": 0.1595744680851064,
+ "grad_norm": 0.8849599843025514,
+ "learning_rate": 9.84387701691328e-06,
+ "loss": 0.41399388313293456,
+ "num_input_tokens_seen": 103958648,
+ "step": 2370,
+ "train_runtime": 18252.5289,
+ "train_tokens_per_second": 5695.575
+ },
+ {
+ "epoch": 0.15991112308106653,
+ "grad_norm": 0.9775050084644975,
+ "learning_rate": 9.843220766064443e-06,
+ "loss": 0.4396827697753906,
+ "num_input_tokens_seen": 104164280,
+ "step": 2375,
+ "train_runtime": 18287.1992,
+ "train_tokens_per_second": 5696.022
+ },
+ {
+ "epoch": 0.16024777807702667,
+ "grad_norm": 1.026818097492493,
+ "learning_rate": 9.842563160824364e-06,
+ "loss": 0.4241004943847656,
+ "num_input_tokens_seen": 104376504,
+ "step": 2380,
+ "train_runtime": 18322.4835,
+ "train_tokens_per_second": 5696.635
+ },
+ {
+ "epoch": 0.1605844330729868,
+ "grad_norm": 0.8310388237023464,
+ "learning_rate": 9.841904201376939e-06,
+ "loss": 0.44525814056396484,
+ "num_input_tokens_seen": 104607616,
+ "step": 2385,
+ "train_runtime": 18361.4521,
+ "train_tokens_per_second": 5697.132
+ },
+ {
+ "epoch": 0.16092108806894695,
+ "grad_norm": 0.8107352807912601,
+ "learning_rate": 9.841243887906443e-06,
+ "loss": 0.3924037218093872,
+ "num_input_tokens_seen": 104827648,
+ "step": 2390,
+ "train_runtime": 18397.6205,
+ "train_tokens_per_second": 5697.892
+ },
+ {
+ "epoch": 0.1612577430649071,
+ "grad_norm": 0.8255737690801053,
+ "learning_rate": 9.840582220597532e-06,
+ "loss": 0.38776795864105223,
+ "num_input_tokens_seen": 105048616,
+ "step": 2395,
+ "train_runtime": 18437.9099,
+ "train_tokens_per_second": 5697.425
+ },
+ {
+ "epoch": 0.16159439806086723,
+ "grad_norm": 1.0393091287356393,
+ "learning_rate": 9.839919199635238e-06,
+ "loss": 0.44150123596191404,
+ "num_input_tokens_seen": 105271480,
+ "step": 2400,
+ "train_runtime": 18481.0333,
+ "train_tokens_per_second": 5696.19
+ },
+ {
+ "epoch": 0.16193105305682737,
+ "grad_norm": 1.778936138437422,
+ "learning_rate": 9.839254825204973e-06,
+ "loss": 0.4216805934906006,
+ "num_input_tokens_seen": 105497528,
+ "step": 2405,
+ "train_runtime": 18522.7039,
+ "train_tokens_per_second": 5695.579
+ },
+ {
+ "epoch": 0.1622677080527875,
+ "grad_norm": 0.8730955239731051,
+ "learning_rate": 9.838589097492527e-06,
+ "loss": 0.4195371627807617,
+ "num_input_tokens_seen": 105713952,
+ "step": 2410,
+ "train_runtime": 18562.2885,
+ "train_tokens_per_second": 5695.093
+ },
+ {
+ "epoch": 0.16260436304874765,
+ "grad_norm": 1.003318291974337,
+ "learning_rate": 9.83792201668407e-06,
+ "loss": 0.4166804313659668,
+ "num_input_tokens_seen": 105940976,
+ "step": 2415,
+ "train_runtime": 18598.4307,
+ "train_tokens_per_second": 5696.232
+ },
+ {
+ "epoch": 0.16294101804470779,
+ "grad_norm": 0.9882562378903975,
+ "learning_rate": 9.837253582966146e-06,
+ "loss": 0.4339381217956543,
+ "num_input_tokens_seen": 106165352,
+ "step": 2420,
+ "train_runtime": 18633.1302,
+ "train_tokens_per_second": 5697.666
+ },
+ {
+ "epoch": 0.16327767304066793,
+ "grad_norm": 0.8141828153275021,
+ "learning_rate": 9.836583796525686e-06,
+ "loss": 0.3936171531677246,
+ "num_input_tokens_seen": 106364536,
+ "step": 2425,
+ "train_runtime": 18675.1796,
+ "train_tokens_per_second": 5695.503
+ },
+ {
+ "epoch": 0.16361432803662807,
+ "grad_norm": 0.861811860528164,
+ "learning_rate": 9.835912657549986e-06,
+ "loss": 0.4514013290405273,
+ "num_input_tokens_seen": 106587560,
+ "step": 2430,
+ "train_runtime": 18717.8406,
+ "train_tokens_per_second": 5694.437
+ },
+ {
+ "epoch": 0.1639509830325882,
+ "grad_norm": 0.8811857541516364,
+ "learning_rate": 9.835240166226734e-06,
+ "loss": 0.39427995681762695,
+ "num_input_tokens_seen": 106799528,
+ "step": 2435,
+ "train_runtime": 18761.1279,
+ "train_tokens_per_second": 5692.596
+ },
+ {
+ "epoch": 0.16428763802854834,
+ "grad_norm": 0.8015505121470864,
+ "learning_rate": 9.834566322743987e-06,
+ "loss": 0.43557186126708985,
+ "num_input_tokens_seen": 107012480,
+ "step": 2440,
+ "train_runtime": 18795.981,
+ "train_tokens_per_second": 5693.37
+ },
+ {
+ "epoch": 0.16462429302450848,
+ "grad_norm": 0.8088311122015132,
+ "learning_rate": 9.833891127290186e-06,
+ "loss": 0.4289062976837158,
+ "num_input_tokens_seen": 107250496,
+ "step": 2445,
+ "train_runtime": 18827.4248,
+ "train_tokens_per_second": 5696.504
+ },
+ {
+ "epoch": 0.16496094802046862,
+ "grad_norm": 0.7483438381257065,
+ "learning_rate": 9.833214580054145e-06,
+ "loss": 0.3959768295288086,
+ "num_input_tokens_seen": 107469136,
+ "step": 2450,
+ "train_runtime": 18863.1469,
+ "train_tokens_per_second": 5697.307
+ },
+ {
+ "epoch": 0.16529760301642876,
+ "grad_norm": 0.8277673951140269,
+ "learning_rate": 9.832536681225058e-06,
+ "loss": 0.44450016021728517,
+ "num_input_tokens_seen": 107691864,
+ "step": 2455,
+ "train_runtime": 18898.3354,
+ "train_tokens_per_second": 5698.484
+ },
+ {
+ "epoch": 0.1656342580123889,
+ "grad_norm": 0.8054080794016197,
+ "learning_rate": 9.831857430992497e-06,
+ "loss": 0.3906950235366821,
+ "num_input_tokens_seen": 107905928,
+ "step": 2460,
+ "train_runtime": 18933.5537,
+ "train_tokens_per_second": 5699.19
+ },
+ {
+ "epoch": 0.16597091300834904,
+ "grad_norm": 0.8380957144167158,
+ "learning_rate": 9.831176829546416e-06,
+ "loss": 0.41658635139465333,
+ "num_input_tokens_seen": 108126176,
+ "step": 2465,
+ "train_runtime": 18970.3816,
+ "train_tokens_per_second": 5699.736
+ },
+ {
+ "epoch": 0.16630756800430918,
+ "grad_norm": 0.9452365824285418,
+ "learning_rate": 9.830494877077137e-06,
+ "loss": 0.46334171295166016,
+ "num_input_tokens_seen": 108358872,
+ "step": 2470,
+ "train_runtime": 19004.9571,
+ "train_tokens_per_second": 5701.611
+ },
+ {
+ "epoch": 0.16664422300026932,
+ "grad_norm": 1.1889597050506375,
+ "learning_rate": 9.82981157377537e-06,
+ "loss": 0.4094019889831543,
+ "num_input_tokens_seen": 108572432,
+ "step": 2475,
+ "train_runtime": 19047.273,
+ "train_tokens_per_second": 5700.156
+ },
+ {
+ "epoch": 0.16698087799622946,
+ "grad_norm": 0.8308872657286545,
+ "learning_rate": 9.829126919832198e-06,
+ "loss": 0.3822948455810547,
+ "num_input_tokens_seen": 108789616,
+ "step": 2480,
+ "train_runtime": 19086.2476,
+ "train_tokens_per_second": 5699.895
+ },
+ {
+ "epoch": 0.1673175329921896,
+ "grad_norm": 0.7290133116509023,
+ "learning_rate": 9.82844091543908e-06,
+ "loss": 0.4369968414306641,
+ "num_input_tokens_seen": 109023616,
+ "step": 2485,
+ "train_runtime": 19125.718,
+ "train_tokens_per_second": 5700.367
+ },
+ {
+ "epoch": 0.16765418798814974,
+ "grad_norm": 0.7990719470039959,
+ "learning_rate": 9.82775356078786e-06,
+ "loss": 0.4084890842437744,
+ "num_input_tokens_seen": 109238536,
+ "step": 2490,
+ "train_runtime": 19167.2805,
+ "train_tokens_per_second": 5699.219
+ },
+ {
+ "epoch": 0.16799084298410988,
+ "grad_norm": 0.7393931696314984,
+ "learning_rate": 9.82706485607075e-06,
+ "loss": 0.3479023456573486,
+ "num_input_tokens_seen": 109448064,
+ "step": 2495,
+ "train_runtime": 19202.9034,
+ "train_tokens_per_second": 5699.558
+ },
+ {
+ "epoch": 0.16832749798007002,
+ "grad_norm": 0.8223012662270877,
+ "learning_rate": 9.826374801480346e-06,
+ "loss": 0.4044170379638672,
+ "num_input_tokens_seen": 109654600,
+ "step": 2500,
+ "train_runtime": 19240.4274,
+ "train_tokens_per_second": 5699.177
+ },
+ {
+ "epoch": 0.16866415297603016,
+ "grad_norm": 0.7578035444329189,
+ "learning_rate": 9.825683397209617e-06,
+ "loss": 0.4128159999847412,
+ "num_input_tokens_seen": 109844880,
+ "step": 2505,
+ "train_runtime": 19277.0081,
+ "train_tokens_per_second": 5698.233
+ },
+ {
+ "epoch": 0.1690008079719903,
+ "grad_norm": 0.7754496384426975,
+ "learning_rate": 9.824990643451915e-06,
+ "loss": 0.40271682739257814,
+ "num_input_tokens_seen": 110072368,
+ "step": 2510,
+ "train_runtime": 19317.0275,
+ "train_tokens_per_second": 5698.204
+ },
+ {
+ "epoch": 0.16933746296795044,
+ "grad_norm": 0.7630225519369642,
+ "learning_rate": 9.824296540400965e-06,
+ "loss": 0.4032273769378662,
+ "num_input_tokens_seen": 110306168,
+ "step": 2515,
+ "train_runtime": 19359.4719,
+ "train_tokens_per_second": 5697.788
+ },
+ {
+ "epoch": 0.16967411796391058,
+ "grad_norm": 0.9445770110284926,
+ "learning_rate": 9.823601088250872e-06,
+ "loss": 0.43148112297058105,
+ "num_input_tokens_seen": 110517392,
+ "step": 2520,
+ "train_runtime": 19399.6091,
+ "train_tokens_per_second": 5696.888
+ },
+ {
+ "epoch": 0.17001077295987072,
+ "grad_norm": 0.8318843666554462,
+ "learning_rate": 9.822904287196114e-06,
+ "loss": 0.4280102729797363,
+ "num_input_tokens_seen": 110729000,
+ "step": 2525,
+ "train_runtime": 19437.1917,
+ "train_tokens_per_second": 5696.759
+ },
+ {
+ "epoch": 0.17034742795583085,
+ "grad_norm": 0.8164720266895754,
+ "learning_rate": 9.822206137431553e-06,
+ "loss": 0.4008312702178955,
+ "num_input_tokens_seen": 110951576,
+ "step": 2530,
+ "train_runtime": 19473.1465,
+ "train_tokens_per_second": 5697.671
+ },
+ {
+ "epoch": 0.170684082951791,
+ "grad_norm": 0.7906437056018215,
+ "learning_rate": 9.821506639152421e-06,
+ "loss": 0.42267236709594724,
+ "num_input_tokens_seen": 111165912,
+ "step": 2535,
+ "train_runtime": 19511.9498,
+ "train_tokens_per_second": 5697.325
+ },
+ {
+ "epoch": 0.17102073794775113,
+ "grad_norm": 0.8027232200064034,
+ "learning_rate": 9.820805792554334e-06,
+ "loss": 0.3978217124938965,
+ "num_input_tokens_seen": 111390648,
+ "step": 2540,
+ "train_runtime": 19548.9277,
+ "train_tokens_per_second": 5698.044
+ },
+ {
+ "epoch": 0.17135739294371127,
+ "grad_norm": 0.9732933783691021,
+ "learning_rate": 9.820103597833278e-06,
+ "loss": 0.40553808212280273,
+ "num_input_tokens_seen": 111613680,
+ "step": 2545,
+ "train_runtime": 19588.8813,
+ "train_tokens_per_second": 5697.808
+ },
+ {
+ "epoch": 0.1716940479396714,
+ "grad_norm": 0.82729851843596,
+ "learning_rate": 9.819400055185622e-06,
+ "loss": 0.3968106031417847,
+ "num_input_tokens_seen": 111830672,
+ "step": 2550,
+ "train_runtime": 19624.1494,
+ "train_tokens_per_second": 5698.625
+ },
+ {
+ "epoch": 0.17203070293563155,
+ "grad_norm": 0.8468905353541687,
+ "learning_rate": 9.818695164808108e-06,
+ "loss": 0.40885286331176757,
+ "num_input_tokens_seen": 112047832,
+ "step": 2555,
+ "train_runtime": 19668.7374,
+ "train_tokens_per_second": 5696.748
+ },
+ {
+ "epoch": 0.1723673579315917,
+ "grad_norm": 0.7939112227025558,
+ "learning_rate": 9.817988926897856e-06,
+ "loss": 0.39391307830810546,
+ "num_input_tokens_seen": 112267160,
+ "step": 2560,
+ "train_runtime": 19710.9584,
+ "train_tokens_per_second": 5695.672
+ },
+ {
+ "epoch": 0.17270401292755186,
+ "grad_norm": 0.8639441951747879,
+ "learning_rate": 9.817281341652367e-06,
+ "loss": 0.4082347869873047,
+ "num_input_tokens_seen": 112491800,
+ "step": 2565,
+ "train_runtime": 19749.2522,
+ "train_tokens_per_second": 5696.003
+ },
+ {
+ "epoch": 0.173040667923512,
+ "grad_norm": 0.716613288966692,
+ "learning_rate": 9.816572409269508e-06,
+ "loss": 0.3824393033981323,
+ "num_input_tokens_seen": 112719560,
+ "step": 2570,
+ "train_runtime": 19784.3744,
+ "train_tokens_per_second": 5697.403
+ },
+ {
+ "epoch": 0.17337732291947214,
+ "grad_norm": 0.7563163756690298,
+ "learning_rate": 9.815862129947537e-06,
+ "loss": 0.4002216339111328,
+ "num_input_tokens_seen": 112940512,
+ "step": 2575,
+ "train_runtime": 19821.7603,
+ "train_tokens_per_second": 5697.804
+ },
+ {
+ "epoch": 0.17371397791543228,
+ "grad_norm": 0.8087466973345561,
+ "learning_rate": 9.815150503885078e-06,
+ "loss": 0.37178959846496584,
+ "num_input_tokens_seen": 113165624,
+ "step": 2580,
+ "train_runtime": 19858.1999,
+ "train_tokens_per_second": 5698.685
+ },
+ {
+ "epoch": 0.17405063291139242,
+ "grad_norm": 0.8994912714777122,
+ "learning_rate": 9.814437531281133e-06,
+ "loss": 0.40114603042602537,
+ "num_input_tokens_seen": 113386744,
+ "step": 2585,
+ "train_runtime": 19903.4827,
+ "train_tokens_per_second": 5696.829
+ },
+ {
+ "epoch": 0.17438728790735256,
+ "grad_norm": 0.9269225867551159,
+ "learning_rate": 9.813723212335085e-06,
+ "loss": 0.43936843872070314,
+ "num_input_tokens_seen": 113593688,
+ "step": 2590,
+ "train_runtime": 19940.3903,
+ "train_tokens_per_second": 5696.663
+ },
+ {
+ "epoch": 0.1747239429033127,
+ "grad_norm": 0.8457918670255721,
+ "learning_rate": 9.813007547246691e-06,
+ "loss": 0.39149506092071534,
+ "num_input_tokens_seen": 113810536,
+ "step": 2595,
+ "train_runtime": 19976.29,
+ "train_tokens_per_second": 5697.281
+ },
+ {
+ "epoch": 0.17506059789927284,
+ "grad_norm": 0.8422559964504273,
+ "learning_rate": 9.812290536216083e-06,
+ "loss": 0.4519649982452393,
+ "num_input_tokens_seen": 114012584,
+ "step": 2600,
+ "train_runtime": 20014.4208,
+ "train_tokens_per_second": 5696.522
+ },
+ {
+ "epoch": 0.17539725289523297,
+ "grad_norm": 0.7614963966994434,
+ "learning_rate": 9.811572179443773e-06,
+ "loss": 0.3978102445602417,
+ "num_input_tokens_seen": 114240040,
+ "step": 2605,
+ "train_runtime": 20049.8879,
+ "train_tokens_per_second": 5697.789
+ },
+ {
+ "epoch": 0.17573390789119311,
+ "grad_norm": 0.7655048641611386,
+ "learning_rate": 9.810852477130645e-06,
+ "loss": 0.40557031631469725,
+ "num_input_tokens_seen": 114449696,
+ "step": 2610,
+ "train_runtime": 20089.5257,
+ "train_tokens_per_second": 5696.983
+ },
+ {
+ "epoch": 0.17607056288715325,
+ "grad_norm": 0.8093163549967093,
+ "learning_rate": 9.810131429477964e-06,
+ "loss": 0.38374016284942625,
+ "num_input_tokens_seen": 114677776,
+ "step": 2615,
+ "train_runtime": 20121.5227,
+ "train_tokens_per_second": 5699.259
+ },
+ {
+ "epoch": 0.1764072178831134,
+ "grad_norm": 0.8258310946925135,
+ "learning_rate": 9.809409036687364e-06,
+ "loss": 0.396789026260376,
+ "num_input_tokens_seen": 114896336,
+ "step": 2620,
+ "train_runtime": 20158.9581,
+ "train_tokens_per_second": 5699.518
+ },
+ {
+ "epoch": 0.17674387287907353,
+ "grad_norm": 0.7382001307664883,
+ "learning_rate": 9.808685298960867e-06,
+ "loss": 0.4137075424194336,
+ "num_input_tokens_seen": 115125264,
+ "step": 2625,
+ "train_runtime": 20201.5972,
+ "train_tokens_per_second": 5698.82
+ },
+ {
+ "epoch": 0.17708052787503367,
+ "grad_norm": 0.7663556283529158,
+ "learning_rate": 9.807960216500858e-06,
+ "loss": 0.4165202617645264,
+ "num_input_tokens_seen": 115343304,
+ "step": 2630,
+ "train_runtime": 20236.3656,
+ "train_tokens_per_second": 5699.803
+ },
+ {
+ "epoch": 0.1774171828709938,
+ "grad_norm": 0.7884797087616878,
+ "learning_rate": 9.807233789510106e-06,
+ "loss": 0.38329482078552246,
+ "num_input_tokens_seen": 115555208,
+ "step": 2635,
+ "train_runtime": 20279.8657,
+ "train_tokens_per_second": 5698.026
+ },
+ {
+ "epoch": 0.17775383786695395,
+ "grad_norm": 0.7323486746827084,
+ "learning_rate": 9.806506018191754e-06,
+ "loss": 0.39183108806610106,
+ "num_input_tokens_seen": 115777312,
+ "step": 2640,
+ "train_runtime": 20317.1207,
+ "train_tokens_per_second": 5698.51
+ },
+ {
+ "epoch": 0.1780904928629141,
+ "grad_norm": 0.8712184035601627,
+ "learning_rate": 9.80577690274932e-06,
+ "loss": 0.39883904457092284,
+ "num_input_tokens_seen": 116000320,
+ "step": 2645,
+ "train_runtime": 20353.9248,
+ "train_tokens_per_second": 5699.162
+ },
+ {
+ "epoch": 0.17842714785887423,
+ "grad_norm": 0.6517931345598202,
+ "learning_rate": 9.8050464433867e-06,
+ "loss": 0.3788264751434326,
+ "num_input_tokens_seen": 116228264,
+ "step": 2650,
+ "train_runtime": 20388.2032,
+ "train_tokens_per_second": 5700.761
+ },
+ {
+ "epoch": 0.17876380285483437,
+ "grad_norm": 0.8947080972086597,
+ "learning_rate": 9.804314640308165e-06,
+ "loss": 0.43610692024230957,
+ "num_input_tokens_seen": 116459784,
+ "step": 2655,
+ "train_runtime": 20433.6777,
+ "train_tokens_per_second": 5699.404
+ },
+ {
+ "epoch": 0.1791004578507945,
+ "grad_norm": 0.8316768119557644,
+ "learning_rate": 9.803581493718361e-06,
+ "loss": 0.4152191162109375,
+ "num_input_tokens_seen": 116677976,
+ "step": 2660,
+ "train_runtime": 20471.066,
+ "train_tokens_per_second": 5699.653
+ },
+ {
+ "epoch": 0.17943711284675465,
+ "grad_norm": 0.7694745863442054,
+ "learning_rate": 9.80284700382231e-06,
+ "loss": 0.4158869743347168,
+ "num_input_tokens_seen": 116913096,
+ "step": 2665,
+ "train_runtime": 20513.9974,
+ "train_tokens_per_second": 5699.186
+ },
+ {
+ "epoch": 0.1797737678427148,
+ "grad_norm": 0.7759018142960451,
+ "learning_rate": 9.802111170825408e-06,
+ "loss": 0.40905094146728516,
+ "num_input_tokens_seen": 117139040,
+ "step": 2670,
+ "train_runtime": 20555.3803,
+ "train_tokens_per_second": 5698.705
+ },
+ {
+ "epoch": 0.18011042283867493,
+ "grad_norm": 0.7209579007108823,
+ "learning_rate": 9.80137399493343e-06,
+ "loss": 0.4070711135864258,
+ "num_input_tokens_seen": 117354016,
+ "step": 2675,
+ "train_runtime": 20594.1226,
+ "train_tokens_per_second": 5698.423
+ },
+ {
+ "epoch": 0.18044707783463507,
+ "grad_norm": 0.7855863516525835,
+ "learning_rate": 9.800635476352525e-06,
+ "loss": 0.39305026531219484,
+ "num_input_tokens_seen": 117552920,
+ "step": 2680,
+ "train_runtime": 20636.4809,
+ "train_tokens_per_second": 5696.365
+ },
+ {
+ "epoch": 0.1807837328305952,
+ "grad_norm": 0.8011882919088659,
+ "learning_rate": 9.799895615289218e-06,
+ "loss": 0.38626153469085694,
+ "num_input_tokens_seen": 117762232,
+ "step": 2685,
+ "train_runtime": 20677.4885,
+ "train_tokens_per_second": 5695.19
+ },
+ {
+ "epoch": 0.18112038782655535,
+ "grad_norm": 0.7959442838200258,
+ "learning_rate": 9.799154411950407e-06,
+ "loss": 0.44461402893066404,
+ "num_input_tokens_seen": 117998240,
+ "step": 2690,
+ "train_runtime": 20716.7034,
+ "train_tokens_per_second": 5695.802
+ },
+ {
+ "epoch": 0.18145704282251549,
+ "grad_norm": 0.7750011575792003,
+ "learning_rate": 9.79841186654337e-06,
+ "loss": 0.4056252479553223,
+ "num_input_tokens_seen": 118211832,
+ "step": 2695,
+ "train_runtime": 20752.9555,
+ "train_tokens_per_second": 5696.144
+ },
+ {
+ "epoch": 0.18179369781847562,
+ "grad_norm": 0.8899930605283304,
+ "learning_rate": 9.797667979275752e-06,
+ "loss": 0.43903465270996095,
+ "num_input_tokens_seen": 118421416,
+ "step": 2700,
+ "train_runtime": 20785.8453,
+ "train_tokens_per_second": 5697.214
+ },
+ {
+ "epoch": 0.18213035281443576,
+ "grad_norm": 0.7964676861353408,
+ "learning_rate": 9.796922750355584e-06,
+ "loss": 0.3816461801528931,
+ "num_input_tokens_seen": 118635408,
+ "step": 2705,
+ "train_runtime": 20823.8564,
+ "train_tokens_per_second": 5697.091
+ },
+ {
+ "epoch": 0.1824670078103959,
+ "grad_norm": 0.8627424628290897,
+ "learning_rate": 9.796176179991266e-06,
+ "loss": 0.4243772029876709,
+ "num_input_tokens_seen": 118853520,
+ "step": 2710,
+ "train_runtime": 20861.0922,
+ "train_tokens_per_second": 5697.378
+ },
+ {
+ "epoch": 0.18280366280635604,
+ "grad_norm": 0.8313804598527397,
+ "learning_rate": 9.795428268391572e-06,
+ "loss": 0.39929049015045165,
+ "num_input_tokens_seen": 119057464,
+ "step": 2715,
+ "train_runtime": 20899.2219,
+ "train_tokens_per_second": 5696.741
+ },
+ {
+ "epoch": 0.18314031780231618,
+ "grad_norm": 0.9224656362131587,
+ "learning_rate": 9.794679015765655e-06,
+ "loss": 0.4025382995605469,
+ "num_input_tokens_seen": 119270504,
+ "step": 2720,
+ "train_runtime": 20936.4359,
+ "train_tokens_per_second": 5696.791
+ },
+ {
+ "epoch": 0.18347697279827632,
+ "grad_norm": 0.8830106782970203,
+ "learning_rate": 9.793928422323041e-06,
+ "loss": 0.40293426513671876,
+ "num_input_tokens_seen": 119494864,
+ "step": 2725,
+ "train_runtime": 20982.2238,
+ "train_tokens_per_second": 5695.052
+ },
+ {
+ "epoch": 0.18381362779423646,
+ "grad_norm": 0.9585116874337107,
+ "learning_rate": 9.793176488273629e-06,
+ "loss": 0.40497751235961915,
+ "num_input_tokens_seen": 119700000,
+ "step": 2730,
+ "train_runtime": 21018.3132,
+ "train_tokens_per_second": 5695.034
+ },
+ {
+ "epoch": 0.1841502827901966,
+ "grad_norm": 0.9047740005031505,
+ "learning_rate": 9.792423213827697e-06,
+ "loss": 0.3891148567199707,
+ "num_input_tokens_seen": 119916696,
+ "step": 2735,
+ "train_runtime": 21057.8627,
+ "train_tokens_per_second": 5694.628
+ },
+ {
+ "epoch": 0.18448693778615674,
+ "grad_norm": 0.8310766014320071,
+ "learning_rate": 9.791668599195897e-06,
+ "loss": 0.3831428289413452,
+ "num_input_tokens_seen": 120142080,
+ "step": 2740,
+ "train_runtime": 21101.6665,
+ "train_tokens_per_second": 5693.488
+ },
+ {
+ "epoch": 0.18482359278211688,
+ "grad_norm": 0.7855666850049595,
+ "learning_rate": 9.790912644589248e-06,
+ "loss": 0.42138309478759767,
+ "num_input_tokens_seen": 120348496,
+ "step": 2745,
+ "train_runtime": 21146.0201,
+ "train_tokens_per_second": 5691.307
+ },
+ {
+ "epoch": 0.18516024777807702,
+ "grad_norm": 0.7620864321495552,
+ "learning_rate": 9.790155350219159e-06,
+ "loss": 0.3965245246887207,
+ "num_input_tokens_seen": 120570160,
+ "step": 2750,
+ "train_runtime": 21186.6351,
+ "train_tokens_per_second": 5690.859
+ },
+ {
+ "epoch": 0.18549690277403716,
+ "grad_norm": 0.7543478062618101,
+ "learning_rate": 9.789396716297399e-06,
+ "loss": 0.4091955661773682,
+ "num_input_tokens_seen": 120800688,
+ "step": 2755,
+ "train_runtime": 21220.0669,
+ "train_tokens_per_second": 5692.757
+ },
+ {
+ "epoch": 0.1858335577699973,
+ "grad_norm": 0.8640772187026432,
+ "learning_rate": 9.788636743036122e-06,
+ "loss": 0.38339416980743407,
+ "num_input_tokens_seen": 121012328,
+ "step": 2760,
+ "train_runtime": 21257.8451,
+ "train_tokens_per_second": 5692.596
+ },
+ {
+ "epoch": 0.18617021276595744,
+ "grad_norm": 0.7868554451548372,
+ "learning_rate": 9.787875430647847e-06,
+ "loss": 0.3950460195541382,
+ "num_input_tokens_seen": 121232032,
+ "step": 2765,
+ "train_runtime": 21294.5091,
+ "train_tokens_per_second": 5693.112
+ },
+ {
+ "epoch": 0.18650686776191758,
+ "grad_norm": 0.8537247075271003,
+ "learning_rate": 9.787112779345477e-06,
+ "loss": 0.39361560344696045,
+ "num_input_tokens_seen": 121436200,
+ "step": 2770,
+ "train_runtime": 21331.5416,
+ "train_tokens_per_second": 5692.8
+ },
+ {
+ "epoch": 0.18684352275787772,
+ "grad_norm": 0.8358704371476329,
+ "learning_rate": 9.786348789342283e-06,
+ "loss": 0.4345618724822998,
+ "num_input_tokens_seen": 121656544,
+ "step": 2775,
+ "train_runtime": 21370.3261,
+ "train_tokens_per_second": 5692.779
+ },
+ {
+ "epoch": 0.18718017775383786,
+ "grad_norm": 0.7754752799232651,
+ "learning_rate": 9.785583460851912e-06,
+ "loss": 0.39746851921081544,
+ "num_input_tokens_seen": 121881368,
+ "step": 2780,
+ "train_runtime": 21412.528,
+ "train_tokens_per_second": 5692.059
+ },
+ {
+ "epoch": 0.187516832749798,
+ "grad_norm": 0.7537619511207324,
+ "learning_rate": 9.784816794088386e-06,
+ "loss": 0.4182771682739258,
+ "num_input_tokens_seen": 122116272,
+ "step": 2785,
+ "train_runtime": 21450.3054,
+ "train_tokens_per_second": 5692.985
+ },
+ {
+ "epoch": 0.18785348774575814,
+ "grad_norm": 0.86793765719075,
+ "learning_rate": 9.784048789266101e-06,
+ "loss": 0.4000124931335449,
+ "num_input_tokens_seen": 122324024,
+ "step": 2790,
+ "train_runtime": 21493.8907,
+ "train_tokens_per_second": 5691.107
+ },
+ {
+ "epoch": 0.18819014274171827,
+ "grad_norm": 0.7757014341183018,
+ "learning_rate": 9.78327944659983e-06,
+ "loss": 0.411564302444458,
+ "num_input_tokens_seen": 122539808,
+ "step": 2795,
+ "train_runtime": 21541.5867,
+ "train_tokens_per_second": 5688.523
+ },
+ {
+ "epoch": 0.18852679773767841,
+ "grad_norm": 0.8345683649256179,
+ "learning_rate": 9.78250876630471e-06,
+ "loss": 0.40423192977905276,
+ "num_input_tokens_seen": 122747320,
+ "step": 2800,
+ "train_runtime": 21579.8161,
+ "train_tokens_per_second": 5688.061
+ },
+ {
+ "epoch": 0.18886345273363855,
+ "grad_norm": 0.8512506107699013,
+ "learning_rate": 9.781736748596267e-06,
+ "loss": 0.39595718383789064,
+ "num_input_tokens_seen": 122957984,
+ "step": 2805,
+ "train_runtime": 21616.5661,
+ "train_tokens_per_second": 5688.137
+ },
+ {
+ "epoch": 0.18920010772959872,
+ "grad_norm": 0.8124112022835439,
+ "learning_rate": 9.78096339369039e-06,
+ "loss": 0.3889493942260742,
+ "num_input_tokens_seen": 123167712,
+ "step": 2810,
+ "train_runtime": 21651.6405,
+ "train_tokens_per_second": 5688.609
+ },
+ {
+ "epoch": 0.18953676272555886,
+ "grad_norm": 0.7367215294516509,
+ "learning_rate": 9.780188701803345e-06,
+ "loss": 0.42559328079223635,
+ "num_input_tokens_seen": 123386632,
+ "step": 2815,
+ "train_runtime": 21694.1793,
+ "train_tokens_per_second": 5687.546
+ },
+ {
+ "epoch": 0.189873417721519,
+ "grad_norm": 0.7954578075141111,
+ "learning_rate": 9.779412673151772e-06,
+ "loss": 0.40504965782165525,
+ "num_input_tokens_seen": 123614712,
+ "step": 2820,
+ "train_runtime": 21735.5555,
+ "train_tokens_per_second": 5687.212
+ },
+ {
+ "epoch": 0.19021007271747914,
+ "grad_norm": 0.7248412230311774,
+ "learning_rate": 9.778635307952686e-06,
+ "loss": 0.4043741703033447,
+ "num_input_tokens_seen": 123839728,
+ "step": 2825,
+ "train_runtime": 21777.9441,
+ "train_tokens_per_second": 5686.475
+ },
+ {
+ "epoch": 0.19054672771343928,
+ "grad_norm": 0.9358958553909015,
+ "learning_rate": 9.777856606423472e-06,
+ "loss": 0.403910493850708,
+ "num_input_tokens_seen": 124065288,
+ "step": 2830,
+ "train_runtime": 21817.8593,
+ "train_tokens_per_second": 5686.41
+ },
+ {
+ "epoch": 0.19088338270939942,
+ "grad_norm": 0.8463489849936312,
+ "learning_rate": 9.777076568781897e-06,
+ "loss": 0.43606057167053225,
+ "num_input_tokens_seen": 124264360,
+ "step": 2835,
+ "train_runtime": 21855.3285,
+ "train_tokens_per_second": 5685.769
+ },
+ {
+ "epoch": 0.19122003770535956,
+ "grad_norm": 0.9165712627873297,
+ "learning_rate": 9.77629519524609e-06,
+ "loss": 0.3660429954528809,
+ "num_input_tokens_seen": 124479288,
+ "step": 2840,
+ "train_runtime": 21894.9089,
+ "train_tokens_per_second": 5685.307
+ },
+ {
+ "epoch": 0.1915566927013197,
+ "grad_norm": 0.7656937544219163,
+ "learning_rate": 9.775512486034564e-06,
+ "loss": 0.4081143379211426,
+ "num_input_tokens_seen": 124692848,
+ "step": 2845,
+ "train_runtime": 21938.4076,
+ "train_tokens_per_second": 5683.769
+ },
+ {
+ "epoch": 0.19189334769727984,
+ "grad_norm": 0.867416455867072,
+ "learning_rate": 9.774728441366197e-06,
+ "loss": 0.4261101245880127,
+ "num_input_tokens_seen": 124899520,
+ "step": 2850,
+ "train_runtime": 21975.3012,
+ "train_tokens_per_second": 5683.632
+ },
+ {
+ "epoch": 0.19223000269323998,
+ "grad_norm": 0.760427761575941,
+ "learning_rate": 9.773943061460249e-06,
+ "loss": 0.39615979194641116,
+ "num_input_tokens_seen": 125109776,
+ "step": 2855,
+ "train_runtime": 22015.122,
+ "train_tokens_per_second": 5682.902
+ },
+ {
+ "epoch": 0.19256665768920012,
+ "grad_norm": 0.8768768754679016,
+ "learning_rate": 9.773156346536345e-06,
+ "loss": 0.42858476638793946,
+ "num_input_tokens_seen": 125326624,
+ "step": 2860,
+ "train_runtime": 22054.6435,
+ "train_tokens_per_second": 5682.55
+ },
+ {
+ "epoch": 0.19290331268516026,
+ "grad_norm": 0.7656261692386461,
+ "learning_rate": 9.772368296814489e-06,
+ "loss": 0.3920739650726318,
+ "num_input_tokens_seen": 125545392,
+ "step": 2865,
+ "train_runtime": 22088.3487,
+ "train_tokens_per_second": 5683.784
+ },
+ {
+ "epoch": 0.1932399676811204,
+ "grad_norm": 0.7687451750642236,
+ "learning_rate": 9.771578912515058e-06,
+ "loss": 0.439731502532959,
+ "num_input_tokens_seen": 125763512,
+ "step": 2870,
+ "train_runtime": 22123.8551,
+ "train_tokens_per_second": 5684.521
+ },
+ {
+ "epoch": 0.19357662267708053,
+ "grad_norm": 3.7881124274468654,
+ "learning_rate": 9.7707881938588e-06,
+ "loss": 0.3899986743927002,
+ "num_input_tokens_seen": 125983312,
+ "step": 2875,
+ "train_runtime": 22166.3825,
+ "train_tokens_per_second": 5683.531
+ },
+ {
+ "epoch": 0.19391327767304067,
+ "grad_norm": 0.7637738370885714,
+ "learning_rate": 9.769996141066836e-06,
+ "loss": 0.4218841552734375,
+ "num_input_tokens_seen": 126221208,
+ "step": 2880,
+ "train_runtime": 22196.5764,
+ "train_tokens_per_second": 5686.517
+ },
+ {
+ "epoch": 0.1942499326690008,
+ "grad_norm": 0.7196684362616566,
+ "learning_rate": 9.76920275436066e-06,
+ "loss": 0.41857566833496096,
+ "num_input_tokens_seen": 126460024,
+ "step": 2885,
+ "train_runtime": 22232.6411,
+ "train_tokens_per_second": 5688.034
+ },
+ {
+ "epoch": 0.19458658766496095,
+ "grad_norm": 0.7681501572608938,
+ "learning_rate": 9.768408033962145e-06,
+ "loss": 0.4293510437011719,
+ "num_input_tokens_seen": 126682488,
+ "step": 2890,
+ "train_runtime": 22274.9037,
+ "train_tokens_per_second": 5687.229
+ },
+ {
+ "epoch": 0.1949232426609211,
+ "grad_norm": 0.7776932982886595,
+ "learning_rate": 9.767611980093525e-06,
+ "loss": 0.40836372375488283,
+ "num_input_tokens_seen": 126903552,
+ "step": 2895,
+ "train_runtime": 22315.0672,
+ "train_tokens_per_second": 5686.9
+ },
+ {
+ "epoch": 0.19525989765688123,
+ "grad_norm": 0.7725164291944147,
+ "learning_rate": 9.76681459297742e-06,
+ "loss": 0.41483469009399415,
+ "num_input_tokens_seen": 127124488,
+ "step": 2900,
+ "train_runtime": 22361.1218,
+ "train_tokens_per_second": 5685.068
+ },
+ {
+ "epoch": 0.19559655265284137,
+ "grad_norm": 1.0523171028632952,
+ "learning_rate": 9.766015872836814e-06,
+ "loss": 0.4586020469665527,
+ "num_input_tokens_seen": 127351272,
+ "step": 2905,
+ "train_runtime": 22403.7587,
+ "train_tokens_per_second": 5684.371
+ },
+ {
+ "epoch": 0.1959332076488015,
+ "grad_norm": 0.981785930598382,
+ "learning_rate": 9.765215819895066e-06,
+ "loss": 0.4265925407409668,
+ "num_input_tokens_seen": 127584848,
+ "step": 2910,
+ "train_runtime": 22435.5709,
+ "train_tokens_per_second": 5686.722
+ },
+ {
+ "epoch": 0.19626986264476165,
+ "grad_norm": 0.7598810830140263,
+ "learning_rate": 9.76441443437591e-06,
+ "loss": 0.3775201320648193,
+ "num_input_tokens_seen": 127815392,
+ "step": 2915,
+ "train_runtime": 22468.3308,
+ "train_tokens_per_second": 5688.691
+ },
+ {
+ "epoch": 0.1966065176407218,
+ "grad_norm": 0.8036869153351719,
+ "learning_rate": 9.76361171650345e-06,
+ "loss": 0.4249995708465576,
+ "num_input_tokens_seen": 128025072,
+ "step": 2920,
+ "train_runtime": 22508.1392,
+ "train_tokens_per_second": 5687.946
+ },
+ {
+ "epoch": 0.19694317263668193,
+ "grad_norm": 0.8854228803465163,
+ "learning_rate": 9.762807666502164e-06,
+ "loss": 0.41265301704406737,
+ "num_input_tokens_seen": 128236720,
+ "step": 2925,
+ "train_runtime": 22548.1044,
+ "train_tokens_per_second": 5687.251
+ },
+ {
+ "epoch": 0.19727982763264207,
+ "grad_norm": 1.302088487080232,
+ "learning_rate": 9.762002284596898e-06,
+ "loss": 0.4183157444000244,
+ "num_input_tokens_seen": 128448432,
+ "step": 2930,
+ "train_runtime": 22591.5125,
+ "train_tokens_per_second": 5685.694
+ },
+ {
+ "epoch": 0.1976164826286022,
+ "grad_norm": 0.8232254317537885,
+ "learning_rate": 9.761195571012882e-06,
+ "loss": 0.4257832050323486,
+ "num_input_tokens_seen": 128657896,
+ "step": 2935,
+ "train_runtime": 22629.0733,
+ "train_tokens_per_second": 5685.513
+ },
+ {
+ "epoch": 0.19795313762456235,
+ "grad_norm": 0.8172096711574582,
+ "learning_rate": 9.760387525975705e-06,
+ "loss": 0.395475172996521,
+ "num_input_tokens_seen": 128883360,
+ "step": 2940,
+ "train_runtime": 22662.7742,
+ "train_tokens_per_second": 5687.007
+ },
+ {
+ "epoch": 0.1982897926205225,
+ "grad_norm": 0.7517384205877706,
+ "learning_rate": 9.759578149711337e-06,
+ "loss": 0.35533039569854735,
+ "num_input_tokens_seen": 129112664,
+ "step": 2945,
+ "train_runtime": 22703.1966,
+ "train_tokens_per_second": 5686.982
+ },
+ {
+ "epoch": 0.19862644761648263,
+ "grad_norm": 0.7164265548546924,
+ "learning_rate": 9.758767442446117e-06,
+ "loss": 0.38588006496429444,
+ "num_input_tokens_seen": 129342864,
+ "step": 2950,
+ "train_runtime": 22748.3032,
+ "train_tokens_per_second": 5685.825
+ },
+ {
+ "epoch": 0.19896310261244277,
+ "grad_norm": 0.733779366224309,
+ "learning_rate": 9.757955404406755e-06,
+ "loss": 0.38382346630096437,
+ "num_input_tokens_seen": 129566032,
+ "step": 2955,
+ "train_runtime": 22786.3627,
+ "train_tokens_per_second": 5686.122
+ },
+ {
+ "epoch": 0.1992997576084029,
+ "grad_norm": 0.9448244412990339,
+ "learning_rate": 9.757142035820337e-06,
+ "loss": 0.3917106628417969,
+ "num_input_tokens_seen": 129766056,
+ "step": 2960,
+ "train_runtime": 22824.9574,
+ "train_tokens_per_second": 5685.27
+ },
+ {
+ "epoch": 0.19963641260436304,
+ "grad_norm": 0.9552752355322248,
+ "learning_rate": 9.756327336914318e-06,
+ "loss": 0.4134823322296143,
+ "num_input_tokens_seen": 129977048,
+ "step": 2965,
+ "train_runtime": 22868.2844,
+ "train_tokens_per_second": 5683.725
+ },
+ {
+ "epoch": 0.19997306760032318,
+ "grad_norm": 1.0910385569978054,
+ "learning_rate": 9.755511307916526e-06,
+ "loss": 0.4055294990539551,
+ "num_input_tokens_seen": 130194224,
+ "step": 2970,
+ "train_runtime": 22901.3701,
+ "train_tokens_per_second": 5684.997
+ },
+ {
+ "epoch": 0.20030972259628332,
+ "grad_norm": 0.8363361731098393,
+ "learning_rate": 9.754693949055161e-06,
+ "loss": 0.4119466781616211,
+ "num_input_tokens_seen": 130423144,
+ "step": 2975,
+ "train_runtime": 22941.2287,
+ "train_tokens_per_second": 5685.098
+ },
+ {
+ "epoch": 0.20064637759224346,
+ "grad_norm": 0.8509098752743555,
+ "learning_rate": 9.753875260558795e-06,
+ "loss": 0.43303780555725097,
+ "num_input_tokens_seen": 130623168,
+ "step": 2980,
+ "train_runtime": 22977.1333,
+ "train_tokens_per_second": 5684.92
+ },
+ {
+ "epoch": 0.2009830325882036,
+ "grad_norm": 0.7851629802721982,
+ "learning_rate": 9.75305524265637e-06,
+ "loss": 0.39323015213012696,
+ "num_input_tokens_seen": 130853536,
+ "step": 2985,
+ "train_runtime": 23010.865,
+ "train_tokens_per_second": 5686.598
+ },
+ {
+ "epoch": 0.20131968758416374,
+ "grad_norm": 0.8423251979034995,
+ "learning_rate": 9.752233895577205e-06,
+ "loss": 0.4042644023895264,
+ "num_input_tokens_seen": 131059592,
+ "step": 2990,
+ "train_runtime": 23047.2444,
+ "train_tokens_per_second": 5686.562
+ },
+ {
+ "epoch": 0.20165634258012388,
+ "grad_norm": 0.7311264203424918,
+ "learning_rate": 9.751411219550984e-06,
+ "loss": 0.40361328125,
+ "num_input_tokens_seen": 131276760,
+ "step": 2995,
+ "train_runtime": 23082.7777,
+ "train_tokens_per_second": 5687.217
+ },
+ {
+ "epoch": 0.20199299757608402,
+ "grad_norm": 0.9307523548267392,
+ "learning_rate": 9.750587214807766e-06,
+ "loss": 0.39671123027801514,
+ "num_input_tokens_seen": 131486248,
+ "step": 3000,
+ "train_runtime": 23124.9978,
+ "train_tokens_per_second": 5685.892
+ },
+ {
+ "epoch": 0.20232965257204416,
+ "grad_norm": 0.8339744148987415,
+ "learning_rate": 9.749761881577981e-06,
+ "loss": 0.42069306373596194,
+ "num_input_tokens_seen": 131716280,
+ "step": 3005,
+ "train_runtime": 23166.3136,
+ "train_tokens_per_second": 5685.681
+ },
+ {
+ "epoch": 0.2026663075680043,
+ "grad_norm": 0.7849226835237555,
+ "learning_rate": 9.748935220092434e-06,
+ "loss": 0.39777138233184817,
+ "num_input_tokens_seen": 131935832,
+ "step": 3010,
+ "train_runtime": 23202.4925,
+ "train_tokens_per_second": 5686.278
+ },
+ {
+ "epoch": 0.20300296256396444,
+ "grad_norm": 0.8473889903914309,
+ "learning_rate": 9.748107230582295e-06,
+ "loss": 0.39002056121826173,
+ "num_input_tokens_seen": 132140024,
+ "step": 3015,
+ "train_runtime": 23239.8205,
+ "train_tokens_per_second": 5685.931
+ },
+ {
+ "epoch": 0.20333961755992458,
+ "grad_norm": 0.7565501165020906,
+ "learning_rate": 9.747277913279109e-06,
+ "loss": 0.4608659267425537,
+ "num_input_tokens_seen": 132358504,
+ "step": 3020,
+ "train_runtime": 23285.5396,
+ "train_tokens_per_second": 5684.15
+ },
+ {
+ "epoch": 0.20367627255588472,
+ "grad_norm": 0.71921389289753,
+ "learning_rate": 9.746447268414794e-06,
+ "loss": 0.44081912040710447,
+ "num_input_tokens_seen": 132589576,
+ "step": 3025,
+ "train_runtime": 23324.3416,
+ "train_tokens_per_second": 5684.601
+ },
+ {
+ "epoch": 0.20401292755184486,
+ "grad_norm": 0.8592739422898477,
+ "learning_rate": 9.745615296221635e-06,
+ "loss": 0.40747461318969724,
+ "num_input_tokens_seen": 132819496,
+ "step": 3030,
+ "train_runtime": 23358.8471,
+ "train_tokens_per_second": 5686.047
+ },
+ {
+ "epoch": 0.204349582547805,
+ "grad_norm": 0.7375243033673482,
+ "learning_rate": 9.74478199693229e-06,
+ "loss": 0.42429242134094236,
+ "num_input_tokens_seen": 133045880,
+ "step": 3035,
+ "train_runtime": 23401.9716,
+ "train_tokens_per_second": 5685.242
+ },
+ {
+ "epoch": 0.20468623754376514,
+ "grad_norm": 0.7984195828599883,
+ "learning_rate": 9.743947370779791e-06,
+ "loss": 0.4262086391448975,
+ "num_input_tokens_seen": 133273848,
+ "step": 3040,
+ "train_runtime": 23442.6366,
+ "train_tokens_per_second": 5685.105
+ },
+ {
+ "epoch": 0.20502289253972528,
+ "grad_norm": 0.7817233028800991,
+ "learning_rate": 9.743111417997536e-06,
+ "loss": 0.41955132484436036,
+ "num_input_tokens_seen": 133500640,
+ "step": 3045,
+ "train_runtime": 23481.7864,
+ "train_tokens_per_second": 5685.285
+ },
+ {
+ "epoch": 0.20535954753568542,
+ "grad_norm": 0.8413661304916547,
+ "learning_rate": 9.742274138819298e-06,
+ "loss": 0.416639518737793,
+ "num_input_tokens_seen": 133719848,
+ "step": 3050,
+ "train_runtime": 23518.5528,
+ "train_tokens_per_second": 5685.718
+ },
+ {
+ "epoch": 0.20569620253164558,
+ "grad_norm": 0.8835193289624108,
+ "learning_rate": 9.74143553347922e-06,
+ "loss": 0.45115389823913576,
+ "num_input_tokens_seen": 133944240,
+ "step": 3055,
+ "train_runtime": 23558.0231,
+ "train_tokens_per_second": 5685.716
+ },
+ {
+ "epoch": 0.20603285752760572,
+ "grad_norm": 0.7584811920041505,
+ "learning_rate": 9.740595602211814e-06,
+ "loss": 0.395220947265625,
+ "num_input_tokens_seen": 134166880,
+ "step": 3060,
+ "train_runtime": 23594.0834,
+ "train_tokens_per_second": 5686.463
+ },
+ {
+ "epoch": 0.20636951252356586,
+ "grad_norm": 0.779533118470965,
+ "learning_rate": 9.739754345251963e-06,
+ "loss": 0.3915586471557617,
+ "num_input_tokens_seen": 134373184,
+ "step": 3065,
+ "train_runtime": 23633.4149,
+ "train_tokens_per_second": 5685.729
+ },
+ {
+ "epoch": 0.206706167519526,
+ "grad_norm": 0.752085146799199,
+ "learning_rate": 9.738911762834924e-06,
+ "loss": 0.36250555515289307,
+ "num_input_tokens_seen": 134594840,
+ "step": 3070,
+ "train_runtime": 23669.854,
+ "train_tokens_per_second": 5686.34
+ },
+ {
+ "epoch": 0.20704282251548614,
+ "grad_norm": 0.9101615257113173,
+ "learning_rate": 9.738067855196323e-06,
+ "loss": 0.40388903617858884,
+ "num_input_tokens_seen": 134814144,
+ "step": 3075,
+ "train_runtime": 23702.4871,
+ "train_tokens_per_second": 5687.764
+ },
+ {
+ "epoch": 0.20737947751144628,
+ "grad_norm": 0.937886524311701,
+ "learning_rate": 9.737222622572153e-06,
+ "loss": 0.4489736557006836,
+ "num_input_tokens_seen": 135025088,
+ "step": 3080,
+ "train_runtime": 23747.5708,
+ "train_tokens_per_second": 5685.848
+ },
+ {
+ "epoch": 0.20771613250740642,
+ "grad_norm": 0.939422935382809,
+ "learning_rate": 9.736376065198784e-06,
+ "loss": 0.4101541519165039,
+ "num_input_tokens_seen": 135253064,
+ "step": 3085,
+ "train_runtime": 23786.6187,
+ "train_tokens_per_second": 5686.099
+ },
+ {
+ "epoch": 0.20805278750336656,
+ "grad_norm": 0.8449440163538374,
+ "learning_rate": 9.735528183312951e-06,
+ "loss": 0.39399616718292235,
+ "num_input_tokens_seen": 135486752,
+ "step": 3090,
+ "train_runtime": 23827.8882,
+ "train_tokens_per_second": 5686.058
+ },
+ {
+ "epoch": 0.2083894424993267,
+ "grad_norm": 0.7846090848755455,
+ "learning_rate": 9.734678977151762e-06,
+ "loss": 0.3820789098739624,
+ "num_input_tokens_seen": 135702632,
+ "step": 3095,
+ "train_runtime": 23868.8149,
+ "train_tokens_per_second": 5685.353
+ },
+ {
+ "epoch": 0.20872609749528684,
+ "grad_norm": 0.7998350921308711,
+ "learning_rate": 9.733828446952696e-06,
+ "loss": 0.40059328079223633,
+ "num_input_tokens_seen": 135925976,
+ "step": 3100,
+ "train_runtime": 23904.8211,
+ "train_tokens_per_second": 5686.132
+ },
+ {
+ "epoch": 0.20906275249124698,
+ "grad_norm": 0.7794995917456119,
+ "learning_rate": 9.732976592953598e-06,
+ "loss": 0.3992907524108887,
+ "num_input_tokens_seen": 136155256,
+ "step": 3105,
+ "train_runtime": 23947.0473,
+ "train_tokens_per_second": 5685.68
+ },
+ {
+ "epoch": 0.20939940748720712,
+ "grad_norm": 0.8028611006413868,
+ "learning_rate": 9.73212341539269e-06,
+ "loss": 0.40617761611938474,
+ "num_input_tokens_seen": 136377968,
+ "step": 3110,
+ "train_runtime": 23995.829,
+ "train_tokens_per_second": 5683.403
+ },
+ {
+ "epoch": 0.20973606248316726,
+ "grad_norm": 0.7679301554806387,
+ "learning_rate": 9.731268914508556e-06,
+ "loss": 0.39726805686950684,
+ "num_input_tokens_seen": 136595144,
+ "step": 3115,
+ "train_runtime": 24037.8732,
+ "train_tokens_per_second": 5682.497
+ },
+ {
+ "epoch": 0.2100727174791274,
+ "grad_norm": 0.7757940100269232,
+ "learning_rate": 9.730413090540158e-06,
+ "loss": 0.3819112777709961,
+ "num_input_tokens_seen": 136818256,
+ "step": 3120,
+ "train_runtime": 24075.9596,
+ "train_tokens_per_second": 5682.775
+ },
+ {
+ "epoch": 0.21040937247508754,
+ "grad_norm": 0.7507007143030089,
+ "learning_rate": 9.729555943726825e-06,
+ "loss": 0.40000243186950685,
+ "num_input_tokens_seen": 137031872,
+ "step": 3125,
+ "train_runtime": 24120.3223,
+ "train_tokens_per_second": 5681.179
+ },
+ {
+ "epoch": 0.21074602747104768,
+ "grad_norm": 0.8053852599963457,
+ "learning_rate": 9.728697474308253e-06,
+ "loss": 0.40857748985290526,
+ "num_input_tokens_seen": 137246200,
+ "step": 3130,
+ "train_runtime": 24159.6984,
+ "train_tokens_per_second": 5680.791
+ },
+ {
+ "epoch": 0.21108268246700782,
+ "grad_norm": 0.8178855876896101,
+ "learning_rate": 9.72783768252451e-06,
+ "loss": 0.38510382175445557,
+ "num_input_tokens_seen": 137454056,
+ "step": 3135,
+ "train_runtime": 24197.6049,
+ "train_tokens_per_second": 5680.482
+ },
+ {
+ "epoch": 0.21141933746296795,
+ "grad_norm": 0.7840878331434453,
+ "learning_rate": 9.726976568616036e-06,
+ "loss": 0.40147695541381834,
+ "num_input_tokens_seen": 137663648,
+ "step": 3140,
+ "train_runtime": 24241.393,
+ "train_tokens_per_second": 5678.867
+ },
+ {
+ "epoch": 0.2117559924589281,
+ "grad_norm": 0.9228616942450966,
+ "learning_rate": 9.726114132823638e-06,
+ "loss": 0.389466381072998,
+ "num_input_tokens_seen": 137882432,
+ "step": 3145,
+ "train_runtime": 24280.2706,
+ "train_tokens_per_second": 5678.785
+ },
+ {
+ "epoch": 0.21209264745488823,
+ "grad_norm": 0.6934203083016008,
+ "learning_rate": 9.725250375388493e-06,
+ "loss": 0.4213107109069824,
+ "num_input_tokens_seen": 138093344,
+ "step": 3150,
+ "train_runtime": 24320.3928,
+ "train_tokens_per_second": 5678.089
+ },
+ {
+ "epoch": 0.21242930245084837,
+ "grad_norm": 0.8669084820101542,
+ "learning_rate": 9.724385296552147e-06,
+ "loss": 0.43596372604370115,
+ "num_input_tokens_seen": 138316336,
+ "step": 3155,
+ "train_runtime": 24362.0648,
+ "train_tokens_per_second": 5677.529
+ },
+ {
+ "epoch": 0.2127659574468085,
+ "grad_norm": 0.9337337457982863,
+ "learning_rate": 9.72351889655652e-06,
+ "loss": 0.4109373092651367,
+ "num_input_tokens_seen": 138541888,
+ "step": 3160,
+ "train_runtime": 24397.2747,
+ "train_tokens_per_second": 5678.58
+ },
+ {
+ "epoch": 0.21310261244276865,
+ "grad_norm": 0.8342554979170712,
+ "learning_rate": 9.722651175643895e-06,
+ "loss": 0.40399460792541503,
+ "num_input_tokens_seen": 138760928,
+ "step": 3165,
+ "train_runtime": 24436.674,
+ "train_tokens_per_second": 5678.388
+ },
+ {
+ "epoch": 0.2134392674387288,
+ "grad_norm": 0.7886653586812639,
+ "learning_rate": 9.721782134056928e-06,
+ "loss": 0.36252574920654296,
+ "num_input_tokens_seen": 138982088,
+ "step": 3170,
+ "train_runtime": 24480.3769,
+ "train_tokens_per_second": 5677.285
+ },
+ {
+ "epoch": 0.21377592243468893,
+ "grad_norm": 0.7172325919553721,
+ "learning_rate": 9.720911772038645e-06,
+ "loss": 0.4204305648803711,
+ "num_input_tokens_seen": 139208968,
+ "step": 3175,
+ "train_runtime": 24514.7644,
+ "train_tokens_per_second": 5678.577
+ },
+ {
+ "epoch": 0.21411257743064907,
+ "grad_norm": 0.8184330706813875,
+ "learning_rate": 9.720040089832436e-06,
+ "loss": 0.4118339538574219,
+ "num_input_tokens_seen": 139404296,
+ "step": 3180,
+ "train_runtime": 24555.2547,
+ "train_tokens_per_second": 5677.168
+ },
+ {
+ "epoch": 0.2144492324266092,
+ "grad_norm": 0.7810238480356294,
+ "learning_rate": 9.719167087682069e-06,
+ "loss": 0.38251235485076907,
+ "num_input_tokens_seen": 139626528,
+ "step": 3185,
+ "train_runtime": 24596.095,
+ "train_tokens_per_second": 5676.776
+ },
+ {
+ "epoch": 0.21478588742256935,
+ "grad_norm": 0.8237949114328076,
+ "learning_rate": 9.718292765831673e-06,
+ "loss": 0.39778890609741213,
+ "num_input_tokens_seen": 139852040,
+ "step": 3190,
+ "train_runtime": 24636.5885,
+ "train_tokens_per_second": 5676.599
+ },
+ {
+ "epoch": 0.2151225424185295,
+ "grad_norm": 0.823181949661376,
+ "learning_rate": 9.71741712452575e-06,
+ "loss": 0.4149151802062988,
+ "num_input_tokens_seen": 140074344,
+ "step": 3195,
+ "train_runtime": 24669.1536,
+ "train_tokens_per_second": 5678.117
+ },
+ {
+ "epoch": 0.21545919741448963,
+ "grad_norm": 0.7987011374780209,
+ "learning_rate": 9.716540164009174e-06,
+ "loss": 0.3865767478942871,
+ "num_input_tokens_seen": 140277680,
+ "step": 3200,
+ "train_runtime": 24707.2868,
+ "train_tokens_per_second": 5677.583
+ },
+ {
+ "epoch": 0.21579585241044977,
+ "grad_norm": 0.9131858261391301,
+ "learning_rate": 9.715661884527178e-06,
+ "loss": 0.4173252582550049,
+ "num_input_tokens_seen": 140503936,
+ "step": 3205,
+ "train_runtime": 24750.0835,
+ "train_tokens_per_second": 5676.908
+ },
+ {
+ "epoch": 0.2161325074064099,
+ "grad_norm": 0.9283199650974979,
+ "learning_rate": 9.714782286325374e-06,
+ "loss": 0.4102613925933838,
+ "num_input_tokens_seen": 140723000,
+ "step": 3210,
+ "train_runtime": 24793.1175,
+ "train_tokens_per_second": 5675.89
+ },
+ {
+ "epoch": 0.21646916240237005,
+ "grad_norm": 0.9449783149453059,
+ "learning_rate": 9.713901369649738e-06,
+ "loss": 0.3924571990966797,
+ "num_input_tokens_seen": 140943200,
+ "step": 3215,
+ "train_runtime": 24831.5013,
+ "train_tokens_per_second": 5675.984
+ },
+ {
+ "epoch": 0.2168058173983302,
+ "grad_norm": 0.7571957619819966,
+ "learning_rate": 9.713019134746618e-06,
+ "loss": 0.38917341232299807,
+ "num_input_tokens_seen": 141178808,
+ "step": 3220,
+ "train_runtime": 24867.52,
+ "train_tokens_per_second": 5677.237
+ },
+ {
+ "epoch": 0.21714247239429033,
+ "grad_norm": 0.8569813128796843,
+ "learning_rate": 9.712135581862722e-06,
+ "loss": 0.39778847694396974,
+ "num_input_tokens_seen": 141410600,
+ "step": 3225,
+ "train_runtime": 24909.1487,
+ "train_tokens_per_second": 5677.055
+ },
+ {
+ "epoch": 0.21747912739025047,
+ "grad_norm": 0.7957771696921552,
+ "learning_rate": 9.711250711245139e-06,
+ "loss": 0.41150903701782227,
+ "num_input_tokens_seen": 141622744,
+ "step": 3230,
+ "train_runtime": 24947.3479,
+ "train_tokens_per_second": 5676.866
+ },
+ {
+ "epoch": 0.2178157823862106,
+ "grad_norm": 0.7674615378265653,
+ "learning_rate": 9.710364523141317e-06,
+ "loss": 0.4173532009124756,
+ "num_input_tokens_seen": 141845232,
+ "step": 3235,
+ "train_runtime": 24990.9712,
+ "train_tokens_per_second": 5675.859
+ },
+ {
+ "epoch": 0.21815243738217074,
+ "grad_norm": 0.881520768069259,
+ "learning_rate": 9.709477017799076e-06,
+ "loss": 0.4302227973937988,
+ "num_input_tokens_seen": 142068920,
+ "step": 3240,
+ "train_runtime": 25022.7411,
+ "train_tokens_per_second": 5677.592
+ },
+ {
+ "epoch": 0.21848909237813088,
+ "grad_norm": 0.9412217646326109,
+ "learning_rate": 9.708588195466604e-06,
+ "loss": 0.4126433849334717,
+ "num_input_tokens_seen": 142289600,
+ "step": 3245,
+ "train_runtime": 25054.9859,
+ "train_tokens_per_second": 5679.093
+ },
+ {
+ "epoch": 0.21882574737409102,
+ "grad_norm": 0.7430440131191046,
+ "learning_rate": 9.707698056392458e-06,
+ "loss": 0.4167231559753418,
+ "num_input_tokens_seen": 142506920,
+ "step": 3250,
+ "train_runtime": 25090.3796,
+ "train_tokens_per_second": 5679.743
+ },
+ {
+ "epoch": 0.21916240237005116,
+ "grad_norm": 0.8971811831202605,
+ "learning_rate": 9.706806600825562e-06,
+ "loss": 0.4131289005279541,
+ "num_input_tokens_seen": 142719992,
+ "step": 3255,
+ "train_runtime": 25133.8043,
+ "train_tokens_per_second": 5678.408
+ },
+ {
+ "epoch": 0.2194990573660113,
+ "grad_norm": 0.9964244321532095,
+ "learning_rate": 9.70591382901521e-06,
+ "loss": 0.42434329986572267,
+ "num_input_tokens_seen": 142935376,
+ "step": 3260,
+ "train_runtime": 25170.1091,
+ "train_tokens_per_second": 5678.775
+ },
+ {
+ "epoch": 0.21983571236197144,
+ "grad_norm": 0.7540863225432197,
+ "learning_rate": 9.705019741211061e-06,
+ "loss": 0.39461374282836914,
+ "num_input_tokens_seen": 143155712,
+ "step": 3265,
+ "train_runtime": 25208.8229,
+ "train_tokens_per_second": 5678.794
+ },
+ {
+ "epoch": 0.22017236735793158,
+ "grad_norm": 0.8596938023198728,
+ "learning_rate": 9.704124337663145e-06,
+ "loss": 0.38963048458099364,
+ "num_input_tokens_seen": 143381832,
+ "step": 3270,
+ "train_runtime": 25243.421,
+ "train_tokens_per_second": 5679.968
+ },
+ {
+ "epoch": 0.22050902235389172,
+ "grad_norm": 0.8540309395774371,
+ "learning_rate": 9.703227618621859e-06,
+ "loss": 0.4044978618621826,
+ "num_input_tokens_seen": 143607328,
+ "step": 3275,
+ "train_runtime": 25284.0238,
+ "train_tokens_per_second": 5679.766
+ },
+ {
+ "epoch": 0.22084567734985186,
+ "grad_norm": 0.8072539756195105,
+ "learning_rate": 9.702329584337965e-06,
+ "loss": 0.41426868438720704,
+ "num_input_tokens_seen": 143809000,
+ "step": 3280,
+ "train_runtime": 25326.637,
+ "train_tokens_per_second": 5678.172
+ },
+ {
+ "epoch": 0.221182332345812,
+ "grad_norm": 0.9295650810498607,
+ "learning_rate": 9.701430235062599e-06,
+ "loss": 0.43207788467407227,
+ "num_input_tokens_seen": 144023696,
+ "step": 3285,
+ "train_runtime": 25366.0897,
+ "train_tokens_per_second": 5677.804
+ },
+ {
+ "epoch": 0.22151898734177214,
+ "grad_norm": 0.9183012776415354,
+ "learning_rate": 9.700529571047259e-06,
+ "loss": 0.4370114803314209,
+ "num_input_tokens_seen": 144236616,
+ "step": 3290,
+ "train_runtime": 25403.6973,
+ "train_tokens_per_second": 5677.78
+ },
+ {
+ "epoch": 0.22185564233773228,
+ "grad_norm": 0.8676275172270271,
+ "learning_rate": 9.699627592543814e-06,
+ "loss": 0.3935758829116821,
+ "num_input_tokens_seen": 144451272,
+ "step": 3295,
+ "train_runtime": 25449.4742,
+ "train_tokens_per_second": 5676.002
+ },
+ {
+ "epoch": 0.22219229733369245,
+ "grad_norm": 0.7791903636055438,
+ "learning_rate": 9.698724299804498e-06,
+ "loss": 0.41451497077941896,
+ "num_input_tokens_seen": 144675240,
+ "step": 3300,
+ "train_runtime": 25489.8713,
+ "train_tokens_per_second": 5675.793
+ },
+ {
+ "epoch": 0.22252895232965259,
+ "grad_norm": 0.8026534486815711,
+ "learning_rate": 9.697819693081915e-06,
+ "loss": 0.40555124282836913,
+ "num_input_tokens_seen": 144894992,
+ "step": 3305,
+ "train_runtime": 25526.7574,
+ "train_tokens_per_second": 5676.2
+ },
+ {
+ "epoch": 0.22286560732561272,
+ "grad_norm": 0.7189128013140877,
+ "learning_rate": 9.696913772629034e-06,
+ "loss": 0.38943185806274416,
+ "num_input_tokens_seen": 145119152,
+ "step": 3310,
+ "train_runtime": 25563.6911,
+ "train_tokens_per_second": 5676.768
+ },
+ {
+ "epoch": 0.22320226232157286,
+ "grad_norm": 0.788476880680751,
+ "learning_rate": 9.696006538699195e-06,
+ "loss": 0.42933268547058107,
+ "num_input_tokens_seen": 145340576,
+ "step": 3315,
+ "train_runtime": 25595.5361,
+ "train_tokens_per_second": 5678.356
+ },
+ {
+ "epoch": 0.223538917317533,
+ "grad_norm": 0.9173858981265539,
+ "learning_rate": 9.695097991546102e-06,
+ "loss": 0.41726884841918943,
+ "num_input_tokens_seen": 145536392,
+ "step": 3320,
+ "train_runtime": 25635.2268,
+ "train_tokens_per_second": 5677.203
+ },
+ {
+ "epoch": 0.22387557231349314,
+ "grad_norm": 0.85644319812075,
+ "learning_rate": 9.694188131423828e-06,
+ "loss": 0.38661761283874513,
+ "num_input_tokens_seen": 145752592,
+ "step": 3325,
+ "train_runtime": 25675.7087,
+ "train_tokens_per_second": 5676.673
+ },
+ {
+ "epoch": 0.22421222730945328,
+ "grad_norm": 0.8283061394820891,
+ "learning_rate": 9.693276958586809e-06,
+ "loss": 0.4395936965942383,
+ "num_input_tokens_seen": 145975632,
+ "step": 3330,
+ "train_runtime": 25717.6903,
+ "train_tokens_per_second": 5676.079
+ },
+ {
+ "epoch": 0.22454888230541342,
+ "grad_norm": 1.107175434088717,
+ "learning_rate": 9.692364473289856e-06,
+ "loss": 0.37718255519866944,
+ "num_input_tokens_seen": 146190016,
+ "step": 3335,
+ "train_runtime": 25758.607,
+ "train_tokens_per_second": 5675.385
+ },
+ {
+ "epoch": 0.22488553730137356,
+ "grad_norm": 0.8144408251561523,
+ "learning_rate": 9.69145067578814e-06,
+ "loss": 0.38504073619842527,
+ "num_input_tokens_seen": 146415880,
+ "step": 3340,
+ "train_runtime": 25799.3872,
+ "train_tokens_per_second": 5675.169
+ },
+ {
+ "epoch": 0.2252221922973337,
+ "grad_norm": 0.7888334595920984,
+ "learning_rate": 9.690535566337205e-06,
+ "loss": 0.44390153884887695,
+ "num_input_tokens_seen": 146637056,
+ "step": 3345,
+ "train_runtime": 25841.4558,
+ "train_tokens_per_second": 5674.489
+ },
+ {
+ "epoch": 0.22555884729329384,
+ "grad_norm": 0.8121918469969017,
+ "learning_rate": 9.689619145192955e-06,
+ "loss": 0.4186522960662842,
+ "num_input_tokens_seen": 146858960,
+ "step": 3350,
+ "train_runtime": 25884.465,
+ "train_tokens_per_second": 5673.633
+ },
+ {
+ "epoch": 0.22589550228925398,
+ "grad_norm": 0.7357528337250359,
+ "learning_rate": 9.688701412611663e-06,
+ "loss": 0.39981067180633545,
+ "num_input_tokens_seen": 147088760,
+ "step": 3355,
+ "train_runtime": 25924.1782,
+ "train_tokens_per_second": 5673.806
+ },
+ {
+ "epoch": 0.22623215728521412,
+ "grad_norm": 0.8039258104050244,
+ "learning_rate": 9.687782368849973e-06,
+ "loss": 0.40915794372558595,
+ "num_input_tokens_seen": 147306920,
+ "step": 3360,
+ "train_runtime": 25967.0787,
+ "train_tokens_per_second": 5672.834
+ },
+ {
+ "epoch": 0.22656881228117426,
+ "grad_norm": 0.8366828009872761,
+ "learning_rate": 9.686862014164893e-06,
+ "loss": 0.3879998207092285,
+ "num_input_tokens_seen": 147534144,
+ "step": 3365,
+ "train_runtime": 26004.3358,
+ "train_tokens_per_second": 5673.444
+ },
+ {
+ "epoch": 0.2269054672771344,
+ "grad_norm": 0.8583728877321222,
+ "learning_rate": 9.685940348813795e-06,
+ "loss": 0.3802268266677856,
+ "num_input_tokens_seen": 147751544,
+ "step": 3370,
+ "train_runtime": 26042.4838,
+ "train_tokens_per_second": 5673.481
+ },
+ {
+ "epoch": 0.22724212227309454,
+ "grad_norm": 0.8211994293131717,
+ "learning_rate": 9.685017373054421e-06,
+ "loss": 0.39569883346557616,
+ "num_input_tokens_seen": 147971208,
+ "step": 3375,
+ "train_runtime": 26081.1089,
+ "train_tokens_per_second": 5673.501
+ },
+ {
+ "epoch": 0.22757877726905468,
+ "grad_norm": 0.7908688808511072,
+ "learning_rate": 9.68409308714488e-06,
+ "loss": 0.3963681697845459,
+ "num_input_tokens_seen": 148186648,
+ "step": 3380,
+ "train_runtime": 26119.5816,
+ "train_tokens_per_second": 5673.393
+ },
+ {
+ "epoch": 0.22791543226501482,
+ "grad_norm": 0.7696519906828798,
+ "learning_rate": 9.68316749134364e-06,
+ "loss": 0.4038090705871582,
+ "num_input_tokens_seen": 148417752,
+ "step": 3385,
+ "train_runtime": 26162.0325,
+ "train_tokens_per_second": 5673.021
+ },
+ {
+ "epoch": 0.22825208726097496,
+ "grad_norm": 0.7784468571247383,
+ "learning_rate": 9.682240585909545e-06,
+ "loss": 0.38401403427124026,
+ "num_input_tokens_seen": 148628520,
+ "step": 3390,
+ "train_runtime": 26201.9074,
+ "train_tokens_per_second": 5672.431
+ },
+ {
+ "epoch": 0.2285887422569351,
+ "grad_norm": 0.7682945918583601,
+ "learning_rate": 9.681312371101803e-06,
+ "loss": 0.36235072612762453,
+ "num_input_tokens_seen": 148860120,
+ "step": 3395,
+ "train_runtime": 26235.4949,
+ "train_tokens_per_second": 5673.997
+ },
+ {
+ "epoch": 0.22892539725289524,
+ "grad_norm": 0.7562252932302131,
+ "learning_rate": 9.680382847179984e-06,
+ "loss": 0.4153308868408203,
+ "num_input_tokens_seen": 149075360,
+ "step": 3400,
+ "train_runtime": 26279.0336,
+ "train_tokens_per_second": 5672.787
+ },
+ {
+ "epoch": 0.22926205224885537,
+ "grad_norm": 0.8026900227545827,
+ "learning_rate": 9.679452014404025e-06,
+ "loss": 0.39875361919403074,
+ "num_input_tokens_seen": 149283904,
+ "step": 3405,
+ "train_runtime": 26318.8542,
+ "train_tokens_per_second": 5672.128
+ },
+ {
+ "epoch": 0.22959870724481551,
+ "grad_norm": 0.8571181299293755,
+ "learning_rate": 9.678519873034234e-06,
+ "loss": 0.435106086730957,
+ "num_input_tokens_seen": 149511344,
+ "step": 3410,
+ "train_runtime": 26357.0843,
+ "train_tokens_per_second": 5672.53
+ },
+ {
+ "epoch": 0.22993536224077565,
+ "grad_norm": 0.9132428672371381,
+ "learning_rate": 9.677586423331277e-06,
+ "loss": 0.42625818252563474,
+ "num_input_tokens_seen": 149734672,
+ "step": 3415,
+ "train_runtime": 26396.1327,
+ "train_tokens_per_second": 5672.599
+ },
+ {
+ "epoch": 0.2302720172367358,
+ "grad_norm": 0.7227848772197648,
+ "learning_rate": 9.676651665556194e-06,
+ "loss": 0.3853285312652588,
+ "num_input_tokens_seen": 149966984,
+ "step": 3420,
+ "train_runtime": 26437.1674,
+ "train_tokens_per_second": 5672.581
+ },
+ {
+ "epoch": 0.23060867223269593,
+ "grad_norm": 0.9758464763691787,
+ "learning_rate": 9.675715599970383e-06,
+ "loss": 0.39388408660888674,
+ "num_input_tokens_seen": 150173312,
+ "step": 3425,
+ "train_runtime": 26476.8347,
+ "train_tokens_per_second": 5671.876
+ },
+ {
+ "epoch": 0.23094532722865607,
+ "grad_norm": 0.8789640499008409,
+ "learning_rate": 9.674778226835617e-06,
+ "loss": 0.43226184844970705,
+ "num_input_tokens_seen": 150378728,
+ "step": 3430,
+ "train_runtime": 26514.5982,
+ "train_tokens_per_second": 5671.545
+ },
+ {
+ "epoch": 0.2312819822246162,
+ "grad_norm": 0.8223781330799458,
+ "learning_rate": 9.673839546414026e-06,
+ "loss": 0.4151242733001709,
+ "num_input_tokens_seen": 150589928,
+ "step": 3435,
+ "train_runtime": 26555.5388,
+ "train_tokens_per_second": 5670.754
+ },
+ {
+ "epoch": 0.23161863722057635,
+ "grad_norm": 0.6819707841325117,
+ "learning_rate": 9.672899558968107e-06,
+ "loss": 0.39006340503692627,
+ "num_input_tokens_seen": 150802664,
+ "step": 3440,
+ "train_runtime": 26599.7743,
+ "train_tokens_per_second": 5669.321
+ },
+ {
+ "epoch": 0.2319552922165365,
+ "grad_norm": 0.8192017234463131,
+ "learning_rate": 9.67195826476073e-06,
+ "loss": 0.44136791229248046,
+ "num_input_tokens_seen": 151027880,
+ "step": 3445,
+ "train_runtime": 26637.6563,
+ "train_tokens_per_second": 5669.714
+ },
+ {
+ "epoch": 0.23229194721249663,
+ "grad_norm": 0.8137299353827566,
+ "learning_rate": 9.671015664055122e-06,
+ "loss": 0.3967195749282837,
+ "num_input_tokens_seen": 151239928,
+ "step": 3450,
+ "train_runtime": 26671.0692,
+ "train_tokens_per_second": 5670.561
+ },
+ {
+ "epoch": 0.23262860220845677,
+ "grad_norm": 0.819274273484707,
+ "learning_rate": 9.670071757114877e-06,
+ "loss": 0.4030034065246582,
+ "num_input_tokens_seen": 151459832,
+ "step": 3455,
+ "train_runtime": 26710.598,
+ "train_tokens_per_second": 5670.402
+ },
+ {
+ "epoch": 0.2329652572044169,
+ "grad_norm": 0.8052967688256787,
+ "learning_rate": 9.669126544203957e-06,
+ "loss": 0.38636322021484376,
+ "num_input_tokens_seen": 151667896,
+ "step": 3460,
+ "train_runtime": 26749.1234,
+ "train_tokens_per_second": 5670.014
+ },
+ {
+ "epoch": 0.23330191220037705,
+ "grad_norm": 0.8630240510385624,
+ "learning_rate": 9.668180025586689e-06,
+ "loss": 0.4077507972717285,
+ "num_input_tokens_seen": 151902352,
+ "step": 3465,
+ "train_runtime": 26791.3212,
+ "train_tokens_per_second": 5669.834
+ },
+ {
+ "epoch": 0.2336385671963372,
+ "grad_norm": 0.7786921906276983,
+ "learning_rate": 9.667232201527761e-06,
+ "loss": 0.3916687250137329,
+ "num_input_tokens_seen": 152102680,
+ "step": 3470,
+ "train_runtime": 26827.5433,
+ "train_tokens_per_second": 5669.646
+ },
+ {
+ "epoch": 0.23397522219229733,
+ "grad_norm": 0.8593939179148706,
+ "learning_rate": 9.666283072292234e-06,
+ "loss": 0.41549954414367674,
+ "num_input_tokens_seen": 152303592,
+ "step": 3475,
+ "train_runtime": 26864.7416,
+ "train_tokens_per_second": 5669.274
+ },
+ {
+ "epoch": 0.23431187718825747,
+ "grad_norm": 0.7173015967840941,
+ "learning_rate": 9.665332638145522e-06,
+ "loss": 0.40798091888427734,
+ "num_input_tokens_seen": 152526320,
+ "step": 3480,
+ "train_runtime": 26904.6608,
+ "train_tokens_per_second": 5669.141
+ },
+ {
+ "epoch": 0.2346485321842176,
+ "grad_norm": 0.7787780767474121,
+ "learning_rate": 9.664380899353414e-06,
+ "loss": 0.44217610359191895,
+ "num_input_tokens_seen": 152747800,
+ "step": 3485,
+ "train_runtime": 26938.7633,
+ "train_tokens_per_second": 5670.186
+ },
+ {
+ "epoch": 0.23498518718017775,
+ "grad_norm": 0.875181117161752,
+ "learning_rate": 9.663427856182062e-06,
+ "loss": 0.418949556350708,
+ "num_input_tokens_seen": 152978904,
+ "step": 3490,
+ "train_runtime": 26976.7683,
+ "train_tokens_per_second": 5670.765
+ },
+ {
+ "epoch": 0.23532184217613789,
+ "grad_norm": 0.8234151432189463,
+ "learning_rate": 9.662473508897981e-06,
+ "loss": 0.41005191802978513,
+ "num_input_tokens_seen": 153200600,
+ "step": 3495,
+ "train_runtime": 27021.1553,
+ "train_tokens_per_second": 5669.654
+ },
+ {
+ "epoch": 0.23565849717209802,
+ "grad_norm": 0.9381121440419455,
+ "learning_rate": 9.661517857768052e-06,
+ "loss": 0.43746633529663087,
+ "num_input_tokens_seen": 153413320,
+ "step": 3500,
+ "train_runtime": 27058.2447,
+ "train_tokens_per_second": 5669.744
+ },
+ {
+ "epoch": 0.23599515216805816,
+ "grad_norm": 0.9159040524887876,
+ "learning_rate": 9.660560903059517e-06,
+ "loss": 0.4074262619018555,
+ "num_input_tokens_seen": 153619928,
+ "step": 3505,
+ "train_runtime": 27099.4693,
+ "train_tokens_per_second": 5668.743
+ },
+ {
+ "epoch": 0.2363318071640183,
+ "grad_norm": 0.7088057637221193,
+ "learning_rate": 9.659602645039988e-06,
+ "loss": 0.4169961929321289,
+ "num_input_tokens_seen": 153844808,
+ "step": 3510,
+ "train_runtime": 27137.4232,
+ "train_tokens_per_second": 5669.102
+ },
+ {
+ "epoch": 0.23666846215997844,
+ "grad_norm": 0.7425326417567343,
+ "learning_rate": 9.658643083977435e-06,
+ "loss": 0.37222681045532224,
+ "num_input_tokens_seen": 154057008,
+ "step": 3515,
+ "train_runtime": 27172.217,
+ "train_tokens_per_second": 5669.652
+ },
+ {
+ "epoch": 0.23700511715593858,
+ "grad_norm": 0.7835915209475461,
+ "learning_rate": 9.657682220140202e-06,
+ "loss": 0.3870618104934692,
+ "num_input_tokens_seen": 154257952,
+ "step": 3520,
+ "train_runtime": 27216.0159,
+ "train_tokens_per_second": 5667.911
+ },
+ {
+ "epoch": 0.23734177215189872,
+ "grad_norm": 0.944898207115128,
+ "learning_rate": 9.656720053796987e-06,
+ "loss": 0.4269230365753174,
+ "num_input_tokens_seen": 154468968,
+ "step": 3525,
+ "train_runtime": 27253.774,
+ "train_tokens_per_second": 5667.801
+ },
+ {
+ "epoch": 0.23767842714785886,
+ "grad_norm": 0.964924059220392,
+ "learning_rate": 9.655756585216859e-06,
+ "loss": 0.39426770210266116,
+ "num_input_tokens_seen": 154679920,
+ "step": 3530,
+ "train_runtime": 27297.8899,
+ "train_tokens_per_second": 5666.369
+ },
+ {
+ "epoch": 0.238015082143819,
+ "grad_norm": 0.8389859916982031,
+ "learning_rate": 9.654791814669246e-06,
+ "loss": 0.41727919578552247,
+ "num_input_tokens_seen": 154894088,
+ "step": 3535,
+ "train_runtime": 27338.9009,
+ "train_tokens_per_second": 5665.703
+ },
+ {
+ "epoch": 0.23835173713977914,
+ "grad_norm": 0.7799448759874524,
+ "learning_rate": 9.653825742423949e-06,
+ "loss": 0.36776251792907716,
+ "num_input_tokens_seen": 155118416,
+ "step": 3540,
+ "train_runtime": 27380.1523,
+ "train_tokens_per_second": 5665.36
+ },
+ {
+ "epoch": 0.2386883921357393,
+ "grad_norm": 0.7611172988563449,
+ "learning_rate": 9.652858368751118e-06,
+ "loss": 0.38267481327056885,
+ "num_input_tokens_seen": 155325944,
+ "step": 3545,
+ "train_runtime": 27419.2328,
+ "train_tokens_per_second": 5664.854
+ },
+ {
+ "epoch": 0.23902504713169945,
+ "grad_norm": 0.849616732171608,
+ "learning_rate": 9.651889693921284e-06,
+ "loss": 0.39499473571777344,
+ "num_input_tokens_seen": 155553936,
+ "step": 3550,
+ "train_runtime": 27453.0014,
+ "train_tokens_per_second": 5666.191
+ },
+ {
+ "epoch": 0.2393617021276596,
+ "grad_norm": 0.8290842696777342,
+ "learning_rate": 9.650919718205331e-06,
+ "loss": 0.46364564895629884,
+ "num_input_tokens_seen": 155767440,
+ "step": 3555,
+ "train_runtime": 27489.7425,
+ "train_tokens_per_second": 5666.384
+ },
+ {
+ "epoch": 0.23969835712361973,
+ "grad_norm": 0.6972330732061462,
+ "learning_rate": 9.649948441874509e-06,
+ "loss": 0.41176652908325195,
+ "num_input_tokens_seen": 155991432,
+ "step": 3560,
+ "train_runtime": 27533.7281,
+ "train_tokens_per_second": 5665.467
+ },
+ {
+ "epoch": 0.24003501211957987,
+ "grad_norm": 0.8232987599322095,
+ "learning_rate": 9.648975865200434e-06,
+ "loss": 0.3719303607940674,
+ "num_input_tokens_seen": 156200840,
+ "step": 3565,
+ "train_runtime": 27572.2717,
+ "train_tokens_per_second": 5665.142
+ },
+ {
+ "epoch": 0.24037166711554,
+ "grad_norm": 0.7784058710951964,
+ "learning_rate": 9.648001988455082e-06,
+ "loss": 0.42696056365966795,
+ "num_input_tokens_seen": 156423232,
+ "step": 3570,
+ "train_runtime": 27611.0095,
+ "train_tokens_per_second": 5665.249
+ },
+ {
+ "epoch": 0.24070832211150014,
+ "grad_norm": 0.8101799377558996,
+ "learning_rate": 9.647026811910794e-06,
+ "loss": 0.42109375,
+ "num_input_tokens_seen": 156638144,
+ "step": 3575,
+ "train_runtime": 27652.042,
+ "train_tokens_per_second": 5664.614
+ },
+ {
+ "epoch": 0.24104497710746028,
+ "grad_norm": 0.7776435744711748,
+ "learning_rate": 9.646050335840279e-06,
+ "loss": 0.440277624130249,
+ "num_input_tokens_seen": 156867440,
+ "step": 3580,
+ "train_runtime": 27693.6123,
+ "train_tokens_per_second": 5664.391
+ },
+ {
+ "epoch": 0.24138163210342042,
+ "grad_norm": 0.8251722449916711,
+ "learning_rate": 9.6450725605166e-06,
+ "loss": 0.4013186454772949,
+ "num_input_tokens_seen": 157079728,
+ "step": 3585,
+ "train_runtime": 27726.0789,
+ "train_tokens_per_second": 5665.414
+ },
+ {
+ "epoch": 0.24171828709938056,
+ "grad_norm": 0.8433224749813183,
+ "learning_rate": 9.644093486213192e-06,
+ "loss": 0.4329217910766602,
+ "num_input_tokens_seen": 157312856,
+ "step": 3590,
+ "train_runtime": 27767.0627,
+ "train_tokens_per_second": 5665.448
+ },
+ {
+ "epoch": 0.2420549420953407,
+ "grad_norm": 0.8662917363975451,
+ "learning_rate": 9.64311311320385e-06,
+ "loss": 0.4125063896179199,
+ "num_input_tokens_seen": 157536824,
+ "step": 3595,
+ "train_runtime": 27801.6779,
+ "train_tokens_per_second": 5666.45
+ },
+ {
+ "epoch": 0.24239159709130084,
+ "grad_norm": 0.8971134461583352,
+ "learning_rate": 9.642131441762732e-06,
+ "loss": 0.39645483493804934,
+ "num_input_tokens_seen": 157744672,
+ "step": 3600,
+ "train_runtime": 27838.8811,
+ "train_tokens_per_second": 5666.344
+ },
+ {
+ "epoch": 0.24272825208726098,
+ "grad_norm": 0.8431994851395936,
+ "learning_rate": 9.641148472164359e-06,
+ "loss": 0.3727452516555786,
+ "num_input_tokens_seen": 157957176,
+ "step": 3605,
+ "train_runtime": 27875.6937,
+ "train_tokens_per_second": 5666.484
+ },
+ {
+ "epoch": 0.24306490708322112,
+ "grad_norm": 0.8088195809393041,
+ "learning_rate": 9.640164204683616e-06,
+ "loss": 0.39656991958618165,
+ "num_input_tokens_seen": 158185920,
+ "step": 3610,
+ "train_runtime": 27916.062,
+ "train_tokens_per_second": 5666.484
+ },
+ {
+ "epoch": 0.24340156207918126,
+ "grad_norm": 0.8326446982063453,
+ "learning_rate": 9.639178639595748e-06,
+ "loss": 0.43434457778930663,
+ "num_input_tokens_seen": 158410048,
+ "step": 3615,
+ "train_runtime": 27944.4625,
+ "train_tokens_per_second": 5668.746
+ },
+ {
+ "epoch": 0.2437382170751414,
+ "grad_norm": 0.7829852555366933,
+ "learning_rate": 9.638191777176367e-06,
+ "loss": 0.40036301612854003,
+ "num_input_tokens_seen": 158632248,
+ "step": 3620,
+ "train_runtime": 27978.7949,
+ "train_tokens_per_second": 5669.731
+ },
+ {
+ "epoch": 0.24407487207110154,
+ "grad_norm": 0.7554602216351339,
+ "learning_rate": 9.637203617701446e-06,
+ "loss": 0.41867575645446775,
+ "num_input_tokens_seen": 158861832,
+ "step": 3625,
+ "train_runtime": 28021.0376,
+ "train_tokens_per_second": 5669.377
+ },
+ {
+ "epoch": 0.24441152706706168,
+ "grad_norm": 0.7700724836949714,
+ "learning_rate": 9.63621416144732e-06,
+ "loss": 0.3974162578582764,
+ "num_input_tokens_seen": 159076104,
+ "step": 3630,
+ "train_runtime": 28057.1542,
+ "train_tokens_per_second": 5669.716
+ },
+ {
+ "epoch": 0.24474818206302182,
+ "grad_norm": 0.819901947145382,
+ "learning_rate": 9.635223408690688e-06,
+ "loss": 0.39289240837097167,
+ "num_input_tokens_seen": 159275688,
+ "step": 3635,
+ "train_runtime": 28092.5481,
+ "train_tokens_per_second": 5669.678
+ },
+ {
+ "epoch": 0.24508483705898196,
+ "grad_norm": 0.74792422760683,
+ "learning_rate": 9.63423135970861e-06,
+ "loss": 0.4178638935089111,
+ "num_input_tokens_seen": 159506232,
+ "step": 3640,
+ "train_runtime": 28131.9808,
+ "train_tokens_per_second": 5669.925
+ },
+ {
+ "epoch": 0.2454214920549421,
+ "grad_norm": 0.8690926777352274,
+ "learning_rate": 9.633238014778511e-06,
+ "loss": 0.4228395462036133,
+ "num_input_tokens_seen": 159735712,
+ "step": 3645,
+ "train_runtime": 28164.6263,
+ "train_tokens_per_second": 5671.501
+ },
+ {
+ "epoch": 0.24575814705090224,
+ "grad_norm": 0.8408876506004788,
+ "learning_rate": 9.632243374178173e-06,
+ "loss": 0.3977930784225464,
+ "num_input_tokens_seen": 159959000,
+ "step": 3650,
+ "train_runtime": 28206.3804,
+ "train_tokens_per_second": 5671.022
+ },
+ {
+ "epoch": 0.24609480204686238,
+ "grad_norm": 0.8178577594154023,
+ "learning_rate": 9.63124743818575e-06,
+ "loss": 0.42597079277038574,
+ "num_input_tokens_seen": 160183304,
+ "step": 3655,
+ "train_runtime": 28242.8523,
+ "train_tokens_per_second": 5671.64
+ },
+ {
+ "epoch": 0.24643145704282252,
+ "grad_norm": 0.7174819842350753,
+ "learning_rate": 9.630250207079745e-06,
+ "loss": 0.385437536239624,
+ "num_input_tokens_seen": 160396208,
+ "step": 3660,
+ "train_runtime": 28279.1927,
+ "train_tokens_per_second": 5671.881
+ },
+ {
+ "epoch": 0.24676811203878266,
+ "grad_norm": 0.9558291181876568,
+ "learning_rate": 9.629251681139038e-06,
+ "loss": 0.41692662239074707,
+ "num_input_tokens_seen": 160619648,
+ "step": 3665,
+ "train_runtime": 28315.8408,
+ "train_tokens_per_second": 5672.431
+ },
+ {
+ "epoch": 0.2471047670347428,
+ "grad_norm": 0.7393493036785912,
+ "learning_rate": 9.628251860642857e-06,
+ "loss": 0.40501389503479,
+ "num_input_tokens_seen": 160832936,
+ "step": 3670,
+ "train_runtime": 28358.3536,
+ "train_tokens_per_second": 5671.448
+ },
+ {
+ "epoch": 0.24744142203070293,
+ "grad_norm": 1.3913304565406266,
+ "learning_rate": 9.627250745870805e-06,
+ "loss": 0.4110020637512207,
+ "num_input_tokens_seen": 161045560,
+ "step": 3675,
+ "train_runtime": 28395.0175,
+ "train_tokens_per_second": 5671.613
+ },
+ {
+ "epoch": 0.24777807702666307,
+ "grad_norm": 0.7989443375204172,
+ "learning_rate": 9.626248337102834e-06,
+ "loss": 0.4464590072631836,
+ "num_input_tokens_seen": 161267264,
+ "step": 3680,
+ "train_runtime": 28433.9863,
+ "train_tokens_per_second": 5671.638
+ },
+ {
+ "epoch": 0.2481147320226232,
+ "grad_norm": 0.8100743623465356,
+ "learning_rate": 9.625244634619269e-06,
+ "loss": 0.42216014862060547,
+ "num_input_tokens_seen": 161484928,
+ "step": 3685,
+ "train_runtime": 28470.6875,
+ "train_tokens_per_second": 5671.971
+ },
+ {
+ "epoch": 0.24845138701858335,
+ "grad_norm": 0.8140092686493631,
+ "learning_rate": 9.62423963870079e-06,
+ "loss": 0.4016411781311035,
+ "num_input_tokens_seen": 161695760,
+ "step": 3690,
+ "train_runtime": 28513.9311,
+ "train_tokens_per_second": 5670.764
+ },
+ {
+ "epoch": 0.2487880420145435,
+ "grad_norm": 0.7547948701222302,
+ "learning_rate": 9.623233349628441e-06,
+ "loss": 0.36588501930236816,
+ "num_input_tokens_seen": 161906344,
+ "step": 3695,
+ "train_runtime": 28554.4813,
+ "train_tokens_per_second": 5670.085
+ },
+ {
+ "epoch": 0.24912469701050363,
+ "grad_norm": 0.7855444375692547,
+ "learning_rate": 9.62222576768363e-06,
+ "loss": 0.37874155044555663,
+ "num_input_tokens_seen": 162121952,
+ "step": 3700,
+ "train_runtime": 28593.7113,
+ "train_tokens_per_second": 5669.846
+ },
+ {
+ "epoch": 0.24946135200646377,
+ "grad_norm": 0.8013508326751752,
+ "learning_rate": 9.62121689314812e-06,
+ "loss": 0.4508801460266113,
+ "num_input_tokens_seen": 162339592,
+ "step": 3705,
+ "train_runtime": 28637.9282,
+ "train_tokens_per_second": 5668.692
+ },
+ {
+ "epoch": 0.2497980070024239,
+ "grad_norm": 0.8341179639748463,
+ "learning_rate": 9.620206726304045e-06,
+ "loss": 0.4121876239776611,
+ "num_input_tokens_seen": 162551488,
+ "step": 3710,
+ "train_runtime": 28677.3311,
+ "train_tokens_per_second": 5668.292
+ },
+ {
+ "epoch": 0.2501346619983841,
+ "grad_norm": 0.7606955484077755,
+ "learning_rate": 9.61919526743389e-06,
+ "loss": 0.4007285118103027,
+ "num_input_tokens_seen": 162775296,
+ "step": 3715,
+ "train_runtime": 28717.1122,
+ "train_tokens_per_second": 5668.233
+ },
+ {
+ "epoch": 0.2504713169943442,
+ "grad_norm": 0.7586767868766064,
+ "learning_rate": 9.618182516820507e-06,
+ "loss": 0.4169614315032959,
+ "num_input_tokens_seen": 163004008,
+ "step": 3720,
+ "train_runtime": 28752.7905,
+ "train_tokens_per_second": 5669.154
+ },
+ {
+ "epoch": 0.25080797199030436,
+ "grad_norm": 0.7958372407148936,
+ "learning_rate": 9.61716847474711e-06,
+ "loss": 0.38545975685119627,
+ "num_input_tokens_seen": 163216408,
+ "step": 3725,
+ "train_runtime": 28788.1056,
+ "train_tokens_per_second": 5669.578
+ },
+ {
+ "epoch": 0.2511446269862645,
+ "grad_norm": 0.7821165797023588,
+ "learning_rate": 9.616153141497273e-06,
+ "loss": 0.41805257797241213,
+ "num_input_tokens_seen": 163430088,
+ "step": 3730,
+ "train_runtime": 28827.4553,
+ "train_tokens_per_second": 5669.251
+ },
+ {
+ "epoch": 0.25148128198222464,
+ "grad_norm": 0.8340007896342425,
+ "learning_rate": 9.61513651735493e-06,
+ "loss": 0.37319297790527345,
+ "num_input_tokens_seen": 163642328,
+ "step": 3735,
+ "train_runtime": 28865.8432,
+ "train_tokens_per_second": 5669.065
+ },
+ {
+ "epoch": 0.2518179369781848,
+ "grad_norm": 0.8830424913719485,
+ "learning_rate": 9.614118602604376e-06,
+ "loss": 0.3945745944976807,
+ "num_input_tokens_seen": 163838104,
+ "step": 3740,
+ "train_runtime": 28903.991,
+ "train_tokens_per_second": 5668.356
+ },
+ {
+ "epoch": 0.2521545919741449,
+ "grad_norm": 0.8052297510787393,
+ "learning_rate": 9.613099397530267e-06,
+ "loss": 0.3722787380218506,
+ "num_input_tokens_seen": 164064200,
+ "step": 3745,
+ "train_runtime": 28947.9133,
+ "train_tokens_per_second": 5667.566
+ },
+ {
+ "epoch": 0.25249124697010505,
+ "grad_norm": 0.7862742185087508,
+ "learning_rate": 9.612078902417623e-06,
+ "loss": 0.4092113018035889,
+ "num_input_tokens_seen": 164293832,
+ "step": 3750,
+ "train_runtime": 28984.6022,
+ "train_tokens_per_second": 5668.314
+ },
+ {
+ "epoch": 0.2528279019660652,
+ "grad_norm": 0.917777159281807,
+ "learning_rate": 9.611057117551821e-06,
+ "loss": 0.4049520492553711,
+ "num_input_tokens_seen": 164494120,
+ "step": 3755,
+ "train_runtime": 29022.09,
+ "train_tokens_per_second": 5667.894
+ },
+ {
+ "epoch": 0.25316455696202533,
+ "grad_norm": 0.7487164986896401,
+ "learning_rate": 9.610034043218598e-06,
+ "loss": 0.41929073333740235,
+ "num_input_tokens_seen": 164721320,
+ "step": 3760,
+ "train_runtime": 29064.6791,
+ "train_tokens_per_second": 5667.405
+ },
+ {
+ "epoch": 0.2535012119579855,
+ "grad_norm": 0.8821854609765899,
+ "learning_rate": 9.609009679704057e-06,
+ "loss": 0.4089076995849609,
+ "num_input_tokens_seen": 164941328,
+ "step": 3765,
+ "train_runtime": 29100.0265,
+ "train_tokens_per_second": 5668.082
+ },
+ {
+ "epoch": 0.2538378669539456,
+ "grad_norm": 0.8451569510904378,
+ "learning_rate": 9.607984027294654e-06,
+ "loss": 0.4017925262451172,
+ "num_input_tokens_seen": 165148144,
+ "step": 3770,
+ "train_runtime": 29142.825,
+ "train_tokens_per_second": 5666.854
+ },
+ {
+ "epoch": 0.25417452194990575,
+ "grad_norm": 0.7139702054990582,
+ "learning_rate": 9.606957086277213e-06,
+ "loss": 0.4037941455841064,
+ "num_input_tokens_seen": 165372064,
+ "step": 3775,
+ "train_runtime": 29183.2104,
+ "train_tokens_per_second": 5666.685
+ },
+ {
+ "epoch": 0.2545111769458659,
+ "grad_norm": 0.8399965732594856,
+ "learning_rate": 9.605928856938913e-06,
+ "loss": 0.4064283847808838,
+ "num_input_tokens_seen": 165589000,
+ "step": 3780,
+ "train_runtime": 29221.4526,
+ "train_tokens_per_second": 5666.693
+ },
+ {
+ "epoch": 0.25484783194182603,
+ "grad_norm": 0.7840999876703054,
+ "learning_rate": 9.604899339567298e-06,
+ "loss": 0.3694040060043335,
+ "num_input_tokens_seen": 165806496,
+ "step": 3785,
+ "train_runtime": 29262.2531,
+ "train_tokens_per_second": 5666.225
+ },
+ {
+ "epoch": 0.25518448693778617,
+ "grad_norm": 0.8254996088572744,
+ "learning_rate": 9.603868534450263e-06,
+ "loss": 0.370687198638916,
+ "num_input_tokens_seen": 166040152,
+ "step": 3790,
+ "train_runtime": 29297.3995,
+ "train_tokens_per_second": 5667.402
+ },
+ {
+ "epoch": 0.2555211419337463,
+ "grad_norm": 0.8205302355074221,
+ "learning_rate": 9.602836441876074e-06,
+ "loss": 0.3572968006134033,
+ "num_input_tokens_seen": 166264712,
+ "step": 3795,
+ "train_runtime": 29337.7818,
+ "train_tokens_per_second": 5667.256
+ },
+ {
+ "epoch": 0.25585779692970645,
+ "grad_norm": 0.9811797993807015,
+ "learning_rate": 9.601803062133353e-06,
+ "loss": 0.3995164394378662,
+ "num_input_tokens_seen": 166486544,
+ "step": 3800,
+ "train_runtime": 29381.5758,
+ "train_tokens_per_second": 5666.359
+ },
+ {
+ "epoch": 0.2561944519256666,
+ "grad_norm": 0.7702336235429141,
+ "learning_rate": 9.600768395511078e-06,
+ "loss": 0.3709728717803955,
+ "num_input_tokens_seen": 166705432,
+ "step": 3805,
+ "train_runtime": 29417.1853,
+ "train_tokens_per_second": 5666.94
+ },
+ {
+ "epoch": 0.25653110692162673,
+ "grad_norm": 0.8067121262918427,
+ "learning_rate": 9.599732442298592e-06,
+ "loss": 0.39358603954315186,
+ "num_input_tokens_seen": 166928072,
+ "step": 3810,
+ "train_runtime": 29453.7851,
+ "train_tokens_per_second": 5667.457
+ },
+ {
+ "epoch": 0.25686776191758687,
+ "grad_norm": 0.8388091751069785,
+ "learning_rate": 9.598695202785596e-06,
+ "loss": 0.3980567216873169,
+ "num_input_tokens_seen": 167135336,
+ "step": 3815,
+ "train_runtime": 29489.9025,
+ "train_tokens_per_second": 5667.545
+ },
+ {
+ "epoch": 0.257204416913547,
+ "grad_norm": 0.7655332462464155,
+ "learning_rate": 9.59765667726215e-06,
+ "loss": 0.3841102600097656,
+ "num_input_tokens_seen": 167356664,
+ "step": 3820,
+ "train_runtime": 29526.6684,
+ "train_tokens_per_second": 5667.983
+ },
+ {
+ "epoch": 0.25754107190950715,
+ "grad_norm": 0.8418059528589851,
+ "learning_rate": 9.596616866018677e-06,
+ "loss": 0.46039524078369143,
+ "num_input_tokens_seen": 167568600,
+ "step": 3825,
+ "train_runtime": 29568.9722,
+ "train_tokens_per_second": 5667.042
+ },
+ {
+ "epoch": 0.2578777269054673,
+ "grad_norm": 0.87917393714771,
+ "learning_rate": 9.595575769345955e-06,
+ "loss": 0.3880587100982666,
+ "num_input_tokens_seen": 167766736,
+ "step": 3830,
+ "train_runtime": 29608.4677,
+ "train_tokens_per_second": 5666.174
+ },
+ {
+ "epoch": 0.2582143819014274,
+ "grad_norm": 0.7766137760871757,
+ "learning_rate": 9.594533387535122e-06,
+ "loss": 0.41129217147827146,
+ "num_input_tokens_seen": 167995600,
+ "step": 3835,
+ "train_runtime": 29647.1244,
+ "train_tokens_per_second": 5666.506
+ },
+ {
+ "epoch": 0.25855103689738757,
+ "grad_norm": 0.7405296564338093,
+ "learning_rate": 9.593489720877678e-06,
+ "loss": 0.4008028984069824,
+ "num_input_tokens_seen": 168214488,
+ "step": 3840,
+ "train_runtime": 29683.6625,
+ "train_tokens_per_second": 5666.905
+ },
+ {
+ "epoch": 0.2588876918933477,
+ "grad_norm": 0.8634330150776178,
+ "learning_rate": 9.592444769665482e-06,
+ "loss": 0.37677698135375975,
+ "num_input_tokens_seen": 168446672,
+ "step": 3845,
+ "train_runtime": 29726.006,
+ "train_tokens_per_second": 5666.643
+ },
+ {
+ "epoch": 0.25922434688930784,
+ "grad_norm": 0.967218365398896,
+ "learning_rate": 9.59139853419075e-06,
+ "loss": 0.3776779413223267,
+ "num_input_tokens_seen": 168649368,
+ "step": 3850,
+ "train_runtime": 29767.8639,
+ "train_tokens_per_second": 5665.484
+ },
+ {
+ "epoch": 0.259561001885268,
+ "grad_norm": 0.7776571735726057,
+ "learning_rate": 9.590351014746059e-06,
+ "loss": 0.40406084060668945,
+ "num_input_tokens_seen": 168875024,
+ "step": 3855,
+ "train_runtime": 29804.1223,
+ "train_tokens_per_second": 5666.163
+ },
+ {
+ "epoch": 0.2598976568812281,
+ "grad_norm": 1.1756300159580646,
+ "learning_rate": 9.589302211624343e-06,
+ "loss": 0.39687304496765136,
+ "num_input_tokens_seen": 169090984,
+ "step": 3860,
+ "train_runtime": 29840.2711,
+ "train_tokens_per_second": 5666.536
+ },
+ {
+ "epoch": 0.26023431187718826,
+ "grad_norm": 0.819456740850995,
+ "learning_rate": 9.5882521251189e-06,
+ "loss": 0.3968550682067871,
+ "num_input_tokens_seen": 169310120,
+ "step": 3865,
+ "train_runtime": 29877.9357,
+ "train_tokens_per_second": 5666.728
+ },
+ {
+ "epoch": 0.2605709668731484,
+ "grad_norm": 0.8515658291218985,
+ "learning_rate": 9.58720075552338e-06,
+ "loss": 0.42056074142456057,
+ "num_input_tokens_seen": 169524504,
+ "step": 3870,
+ "train_runtime": 29909.434,
+ "train_tokens_per_second": 5667.928
+ },
+ {
+ "epoch": 0.26090762186910854,
+ "grad_norm": 0.8783050270072679,
+ "learning_rate": 9.586148103131797e-06,
+ "loss": 0.40169625282287597,
+ "num_input_tokens_seen": 169733248,
+ "step": 3875,
+ "train_runtime": 29944.6725,
+ "train_tokens_per_second": 5668.229
+ },
+ {
+ "epoch": 0.2612442768650687,
+ "grad_norm": 0.857473518999431,
+ "learning_rate": 9.585094168238518e-06,
+ "loss": 0.4272635459899902,
+ "num_input_tokens_seen": 169943480,
+ "step": 3880,
+ "train_runtime": 29983.9133,
+ "train_tokens_per_second": 5667.822
+ },
+ {
+ "epoch": 0.2615809318610288,
+ "grad_norm": 0.8162084475991911,
+ "learning_rate": 9.58403895113828e-06,
+ "loss": 0.39615912437438966,
+ "num_input_tokens_seen": 170164936,
+ "step": 3885,
+ "train_runtime": 30023.2923,
+ "train_tokens_per_second": 5667.764
+ },
+ {
+ "epoch": 0.26191758685698896,
+ "grad_norm": 0.773095299777449,
+ "learning_rate": 9.582982452126164e-06,
+ "loss": 0.380422306060791,
+ "num_input_tokens_seen": 170380728,
+ "step": 3890,
+ "train_runtime": 30061.2728,
+ "train_tokens_per_second": 5667.782
+ },
+ {
+ "epoch": 0.2622542418529491,
+ "grad_norm": 0.7250445142657057,
+ "learning_rate": 9.58192467149762e-06,
+ "loss": 0.3720268726348877,
+ "num_input_tokens_seen": 170591568,
+ "step": 3895,
+ "train_runtime": 30102.8751,
+ "train_tokens_per_second": 5666.953
+ },
+ {
+ "epoch": 0.26259089684890924,
+ "grad_norm": 0.8455822434092976,
+ "learning_rate": 9.580865609548452e-06,
+ "loss": 0.4109041213989258,
+ "num_input_tokens_seen": 170793416,
+ "step": 3900,
+ "train_runtime": 30139.1143,
+ "train_tokens_per_second": 5666.836
+ },
+ {
+ "epoch": 0.2629275518448694,
+ "grad_norm": 0.669931714570705,
+ "learning_rate": 9.579805266574824e-06,
+ "loss": 0.3842399835586548,
+ "num_input_tokens_seen": 171006328,
+ "step": 3905,
+ "train_runtime": 30178.5838,
+ "train_tokens_per_second": 5666.48
+ },
+ {
+ "epoch": 0.2632642068408295,
+ "grad_norm": 0.6870277390309487,
+ "learning_rate": 9.578743642873258e-06,
+ "loss": 0.40047216415405273,
+ "num_input_tokens_seen": 171209880,
+ "step": 3910,
+ "train_runtime": 30217.8046,
+ "train_tokens_per_second": 5665.861
+ },
+ {
+ "epoch": 0.26360086183678966,
+ "grad_norm": 0.7479204866431858,
+ "learning_rate": 9.57768073874063e-06,
+ "loss": 0.40533103942871096,
+ "num_input_tokens_seen": 171434584,
+ "step": 3915,
+ "train_runtime": 30250.4257,
+ "train_tokens_per_second": 5667.179
+ },
+ {
+ "epoch": 0.2639375168327498,
+ "grad_norm": 0.822857586143697,
+ "learning_rate": 9.576616554474182e-06,
+ "loss": 0.41611337661743164,
+ "num_input_tokens_seen": 171637768,
+ "step": 3920,
+ "train_runtime": 30287.808,
+ "train_tokens_per_second": 5666.893
+ },
+ {
+ "epoch": 0.26427417182870994,
+ "grad_norm": 0.7335208763069258,
+ "learning_rate": 9.575551090371512e-06,
+ "loss": 0.39514660835266113,
+ "num_input_tokens_seen": 171845072,
+ "step": 3925,
+ "train_runtime": 30328.6771,
+ "train_tokens_per_second": 5666.092
+ },
+ {
+ "epoch": 0.2646108268246701,
+ "grad_norm": 0.7927281370222846,
+ "learning_rate": 9.574484346730564e-06,
+ "loss": 0.4031528472900391,
+ "num_input_tokens_seen": 172066528,
+ "step": 3930,
+ "train_runtime": 30370.1573,
+ "train_tokens_per_second": 5665.645
+ },
+ {
+ "epoch": 0.2649474818206302,
+ "grad_norm": 0.7830404478638062,
+ "learning_rate": 9.573416323849658e-06,
+ "loss": 0.4126886367797852,
+ "num_input_tokens_seen": 172285504,
+ "step": 3935,
+ "train_runtime": 30411.2446,
+ "train_tokens_per_second": 5665.191
+ },
+ {
+ "epoch": 0.26528413681659035,
+ "grad_norm": 0.7639494210600385,
+ "learning_rate": 9.572347022027461e-06,
+ "loss": 0.3947685718536377,
+ "num_input_tokens_seen": 172512448,
+ "step": 3940,
+ "train_runtime": 30452.9824,
+ "train_tokens_per_second": 5664.879
+ },
+ {
+ "epoch": 0.2656207918125505,
+ "grad_norm": 0.7465504294757535,
+ "learning_rate": 9.571276441563e-06,
+ "loss": 0.40018043518066404,
+ "num_input_tokens_seen": 172712640,
+ "step": 3945,
+ "train_runtime": 30484.642,
+ "train_tokens_per_second": 5665.562
+ },
+ {
+ "epoch": 0.26595744680851063,
+ "grad_norm": 0.79021748664721,
+ "learning_rate": 9.570204582755659e-06,
+ "loss": 0.38366117477416994,
+ "num_input_tokens_seen": 172934512,
+ "step": 3950,
+ "train_runtime": 30524.2614,
+ "train_tokens_per_second": 5665.477
+ },
+ {
+ "epoch": 0.2662941018044708,
+ "grad_norm": 0.9006277156621565,
+ "learning_rate": 9.569131445905178e-06,
+ "loss": 0.4079860210418701,
+ "num_input_tokens_seen": 173148536,
+ "step": 3955,
+ "train_runtime": 30557.4244,
+ "train_tokens_per_second": 5666.333
+ },
+ {
+ "epoch": 0.2666307568004309,
+ "grad_norm": 0.8299161339057768,
+ "learning_rate": 9.568057031311659e-06,
+ "loss": 0.41454520225524905,
+ "num_input_tokens_seen": 173357456,
+ "step": 3960,
+ "train_runtime": 30592.1621,
+ "train_tokens_per_second": 5666.728
+ },
+ {
+ "epoch": 0.26696741179639105,
+ "grad_norm": 0.9401299888719756,
+ "learning_rate": 9.566981339275557e-06,
+ "loss": 0.43401298522949217,
+ "num_input_tokens_seen": 173565960,
+ "step": 3965,
+ "train_runtime": 30636.8629,
+ "train_tokens_per_second": 5665.265
+ },
+ {
+ "epoch": 0.2673040667923512,
+ "grad_norm": 0.7744089765372347,
+ "learning_rate": 9.565904370097687e-06,
+ "loss": 0.4027921199798584,
+ "num_input_tokens_seen": 173765184,
+ "step": 3970,
+ "train_runtime": 30675.2756,
+ "train_tokens_per_second": 5664.666
+ },
+ {
+ "epoch": 0.26764072178831133,
+ "grad_norm": 1.1387875256699147,
+ "learning_rate": 9.564826124079219e-06,
+ "loss": 0.3860454082489014,
+ "num_input_tokens_seen": 173982664,
+ "step": 3975,
+ "train_runtime": 30722.6571,
+ "train_tokens_per_second": 5663.008
+ },
+ {
+ "epoch": 0.26797737678427147,
+ "grad_norm": 0.7717245266267511,
+ "learning_rate": 9.563746601521681e-06,
+ "loss": 0.37142066955566405,
+ "num_input_tokens_seen": 174207720,
+ "step": 3980,
+ "train_runtime": 30764.4205,
+ "train_tokens_per_second": 5662.636
+ },
+ {
+ "epoch": 0.2683140317802316,
+ "grad_norm": 0.8633313009939835,
+ "learning_rate": 9.562665802726961e-06,
+ "loss": 0.39812779426574707,
+ "num_input_tokens_seen": 174437200,
+ "step": 3985,
+ "train_runtime": 30809.6747,
+ "train_tokens_per_second": 5661.767
+ },
+ {
+ "epoch": 0.26865068677619175,
+ "grad_norm": 0.79235769535237,
+ "learning_rate": 9.561583727997297e-06,
+ "loss": 0.3973848819732666,
+ "num_input_tokens_seen": 174649000,
+ "step": 3990,
+ "train_runtime": 30852.8543,
+ "train_tokens_per_second": 5660.708
+ },
+ {
+ "epoch": 0.2689873417721519,
+ "grad_norm": 0.866887420036326,
+ "learning_rate": 9.560500377635288e-06,
+ "loss": 0.4132582664489746,
+ "num_input_tokens_seen": 174869968,
+ "step": 3995,
+ "train_runtime": 30894.8991,
+ "train_tokens_per_second": 5660.157
+ },
+ {
+ "epoch": 0.26932399676811203,
+ "grad_norm": 0.8490880534158571,
+ "learning_rate": 9.559415751943893e-06,
+ "loss": 0.4121055603027344,
+ "num_input_tokens_seen": 175087088,
+ "step": 4000,
+ "train_runtime": 30928.2656,
+ "train_tokens_per_second": 5661.07
+ },
+ {
+ "epoch": 0.26966065176407217,
+ "grad_norm": 0.7247750937341004,
+ "learning_rate": 9.558329851226422e-06,
+ "loss": 0.3906848430633545,
+ "num_input_tokens_seen": 175314216,
+ "step": 4005,
+ "train_runtime": 30967.3219,
+ "train_tokens_per_second": 5661.265
+ },
+ {
+ "epoch": 0.2699973067600323,
+ "grad_norm": 0.8575764814098764,
+ "learning_rate": 9.557242675786542e-06,
+ "loss": 0.4193889141082764,
+ "num_input_tokens_seen": 175541064,
+ "step": 4010,
+ "train_runtime": 31013.366,
+ "train_tokens_per_second": 5660.175
+ },
+ {
+ "epoch": 0.27033396175599245,
+ "grad_norm": 0.7331252032970815,
+ "learning_rate": 9.55615422592828e-06,
+ "loss": 0.4430998802185059,
+ "num_input_tokens_seen": 175770592,
+ "step": 4015,
+ "train_runtime": 31053.1239,
+ "train_tokens_per_second": 5660.319
+ },
+ {
+ "epoch": 0.2706706167519526,
+ "grad_norm": 0.8233208505066181,
+ "learning_rate": 9.555064501956018e-06,
+ "loss": 0.4185193061828613,
+ "num_input_tokens_seen": 175988096,
+ "step": 4020,
+ "train_runtime": 31090.813,
+ "train_tokens_per_second": 5660.453
+ },
+ {
+ "epoch": 0.2710072717479127,
+ "grad_norm": 0.7929784850811431,
+ "learning_rate": 9.553973504174493e-06,
+ "loss": 0.3700920581817627,
+ "num_input_tokens_seen": 176213488,
+ "step": 4025,
+ "train_runtime": 31131.4846,
+ "train_tokens_per_second": 5660.298
+ },
+ {
+ "epoch": 0.27134392674387287,
+ "grad_norm": 0.8802925347262501,
+ "learning_rate": 9.5528812328888e-06,
+ "loss": 0.41269850730895996,
+ "num_input_tokens_seen": 176420144,
+ "step": 4030,
+ "train_runtime": 31173.5245,
+ "train_tokens_per_second": 5659.294
+ },
+ {
+ "epoch": 0.271680581739833,
+ "grad_norm": 0.7894741861959437,
+ "learning_rate": 9.551787688404386e-06,
+ "loss": 0.3877572536468506,
+ "num_input_tokens_seen": 176635200,
+ "step": 4035,
+ "train_runtime": 31218.475,
+ "train_tokens_per_second": 5658.034
+ },
+ {
+ "epoch": 0.27201723673579314,
+ "grad_norm": 0.7609981119661086,
+ "learning_rate": 9.550692871027061e-06,
+ "loss": 0.3883471727371216,
+ "num_input_tokens_seen": 176856552,
+ "step": 4040,
+ "train_runtime": 31261.3151,
+ "train_tokens_per_second": 5657.361
+ },
+ {
+ "epoch": 0.2723538917317533,
+ "grad_norm": 0.7842118778601046,
+ "learning_rate": 9.549596781062984e-06,
+ "loss": 0.37834415435791013,
+ "num_input_tokens_seen": 177085864,
+ "step": 4045,
+ "train_runtime": 31299.8686,
+ "train_tokens_per_second": 5657.719
+ },
+ {
+ "epoch": 0.2726905467277134,
+ "grad_norm": 0.8324923888304009,
+ "learning_rate": 9.548499418818677e-06,
+ "loss": 0.3912312984466553,
+ "num_input_tokens_seen": 177292800,
+ "step": 4050,
+ "train_runtime": 31343.6672,
+ "train_tokens_per_second": 5656.415
+ },
+ {
+ "epoch": 0.27302720172367356,
+ "grad_norm": 0.886343399880937,
+ "learning_rate": 9.547400784601011e-06,
+ "loss": 0.448984956741333,
+ "num_input_tokens_seen": 177509120,
+ "step": 4055,
+ "train_runtime": 31388.1483,
+ "train_tokens_per_second": 5655.291
+ },
+ {
+ "epoch": 0.2733638567196337,
+ "grad_norm": 0.7611731029617864,
+ "learning_rate": 9.546300878717216e-06,
+ "loss": 0.41724767684936526,
+ "num_input_tokens_seen": 177737192,
+ "step": 4060,
+ "train_runtime": 31419.4811,
+ "train_tokens_per_second": 5656.91
+ },
+ {
+ "epoch": 0.27370051171559384,
+ "grad_norm": 0.7391294054271582,
+ "learning_rate": 9.545199701474877e-06,
+ "loss": 0.3876931667327881,
+ "num_input_tokens_seen": 177955528,
+ "step": 4065,
+ "train_runtime": 31460.3252,
+ "train_tokens_per_second": 5656.506
+ },
+ {
+ "epoch": 0.274037166711554,
+ "grad_norm": 0.8122534322473342,
+ "learning_rate": 9.544097253181935e-06,
+ "loss": 0.37918825149536134,
+ "num_input_tokens_seen": 178168688,
+ "step": 4070,
+ "train_runtime": 31501.2931,
+ "train_tokens_per_second": 5655.917
+ },
+ {
+ "epoch": 0.2743738217075141,
+ "grad_norm": 0.7594531409775707,
+ "learning_rate": 9.542993534146687e-06,
+ "loss": 0.41271028518676756,
+ "num_input_tokens_seen": 178382216,
+ "step": 4075,
+ "train_runtime": 31548.6623,
+ "train_tokens_per_second": 5654.193
+ },
+ {
+ "epoch": 0.27471047670347426,
+ "grad_norm": 0.813877169259893,
+ "learning_rate": 9.541888544677784e-06,
+ "loss": 0.4018410682678223,
+ "num_input_tokens_seen": 178601400,
+ "step": 4080,
+ "train_runtime": 31583.5769,
+ "train_tokens_per_second": 5654.882
+ },
+ {
+ "epoch": 0.2750471316994344,
+ "grad_norm": 0.8060293108187504,
+ "learning_rate": 9.540782285084232e-06,
+ "loss": 0.36272602081298827,
+ "num_input_tokens_seen": 178815704,
+ "step": 4085,
+ "train_runtime": 31624.6421,
+ "train_tokens_per_second": 5654.316
+ },
+ {
+ "epoch": 0.27538378669539454,
+ "grad_norm": 0.7676915231843366,
+ "learning_rate": 9.539674755675392e-06,
+ "loss": 0.39566807746887206,
+ "num_input_tokens_seen": 179033752,
+ "step": 4090,
+ "train_runtime": 31667.4816,
+ "train_tokens_per_second": 5653.552
+ },
+ {
+ "epoch": 0.2757204416913547,
+ "grad_norm": 0.6913331471393089,
+ "learning_rate": 9.538565956760983e-06,
+ "loss": 0.35824832916259763,
+ "num_input_tokens_seen": 179269312,
+ "step": 4095,
+ "train_runtime": 31701.6169,
+ "train_tokens_per_second": 5654.895
+ },
+ {
+ "epoch": 0.2760570966873148,
+ "grad_norm": 0.7630923484514573,
+ "learning_rate": 9.537455888651078e-06,
+ "loss": 0.3908623933792114,
+ "num_input_tokens_seen": 179500456,
+ "step": 4100,
+ "train_runtime": 31737.6882,
+ "train_tokens_per_second": 5655.751
+ },
+ {
+ "epoch": 0.27639375168327496,
+ "grad_norm": 0.8057751017404143,
+ "learning_rate": 9.536344551656103e-06,
+ "loss": 0.4113442420959473,
+ "num_input_tokens_seen": 179727568,
+ "step": 4105,
+ "train_runtime": 31778.3965,
+ "train_tokens_per_second": 5655.653
+ },
+ {
+ "epoch": 0.2767304066792351,
+ "grad_norm": 0.7452483379248684,
+ "learning_rate": 9.535231946086838e-06,
+ "loss": 0.392077112197876,
+ "num_input_tokens_seen": 179947920,
+ "step": 4110,
+ "train_runtime": 31810.5777,
+ "train_tokens_per_second": 5656.858
+ },
+ {
+ "epoch": 0.27706706167519524,
+ "grad_norm": 0.7935004896667632,
+ "learning_rate": 9.534118072254421e-06,
+ "loss": 0.41229848861694335,
+ "num_input_tokens_seen": 180172368,
+ "step": 4115,
+ "train_runtime": 31849.5051,
+ "train_tokens_per_second": 5656.991
+ },
+ {
+ "epoch": 0.2774037166711554,
+ "grad_norm": 0.782426266760514,
+ "learning_rate": 9.533002930470345e-06,
+ "loss": 0.38343167304992676,
+ "num_input_tokens_seen": 180392712,
+ "step": 4120,
+ "train_runtime": 31893.6268,
+ "train_tokens_per_second": 5656.074
+ },
+ {
+ "epoch": 0.2777403716671155,
+ "grad_norm": 0.6585746266135624,
+ "learning_rate": 9.531886521046452e-06,
+ "loss": 0.38598432540893557,
+ "num_input_tokens_seen": 180594944,
+ "step": 4125,
+ "train_runtime": 31934.9141,
+ "train_tokens_per_second": 5655.094
+ },
+ {
+ "epoch": 0.27807702666307565,
+ "grad_norm": 0.8150349479481921,
+ "learning_rate": 9.530768844294947e-06,
+ "loss": 0.41448087692260743,
+ "num_input_tokens_seen": 180816024,
+ "step": 4130,
+ "train_runtime": 31975.4857,
+ "train_tokens_per_second": 5654.833
+ },
+ {
+ "epoch": 0.2784136816590358,
+ "grad_norm": 0.8985901076929287,
+ "learning_rate": 9.52964990052838e-06,
+ "loss": 0.41909160614013674,
+ "num_input_tokens_seen": 181035024,
+ "step": 4135,
+ "train_runtime": 32014.2767,
+ "train_tokens_per_second": 5654.822
+ },
+ {
+ "epoch": 0.27875033665499593,
+ "grad_norm": 0.7912879897563827,
+ "learning_rate": 9.528529690059663e-06,
+ "loss": 0.3741894245147705,
+ "num_input_tokens_seen": 181258672,
+ "step": 4140,
+ "train_runtime": 32055.2042,
+ "train_tokens_per_second": 5654.579
+ },
+ {
+ "epoch": 0.2790869916509561,
+ "grad_norm": 0.9153777980141117,
+ "learning_rate": 9.52740821320206e-06,
+ "loss": 0.4044219970703125,
+ "num_input_tokens_seen": 181475056,
+ "step": 4145,
+ "train_runtime": 32086.7668,
+ "train_tokens_per_second": 5655.76
+ },
+ {
+ "epoch": 0.2794236466469162,
+ "grad_norm": 0.7181310809160721,
+ "learning_rate": 9.526285470269185e-06,
+ "loss": 0.4395021915435791,
+ "num_input_tokens_seen": 181701504,
+ "step": 4150,
+ "train_runtime": 32121.2174,
+ "train_tokens_per_second": 5656.744
+ },
+ {
+ "epoch": 0.2797603016428764,
+ "grad_norm": 0.833451101611962,
+ "learning_rate": 9.525161461575012e-06,
+ "loss": 0.4143718719482422,
+ "num_input_tokens_seen": 181905448,
+ "step": 4155,
+ "train_runtime": 32162.0174,
+ "train_tokens_per_second": 5655.909
+ },
+ {
+ "epoch": 0.28009695663883655,
+ "grad_norm": 0.9347376730252263,
+ "learning_rate": 9.524036187433867e-06,
+ "loss": 0.38724517822265625,
+ "num_input_tokens_seen": 182111832,
+ "step": 4160,
+ "train_runtime": 32199.5756,
+ "train_tokens_per_second": 5655.722
+ },
+ {
+ "epoch": 0.2804336116347967,
+ "grad_norm": 0.7395436682357247,
+ "learning_rate": 9.522909648160427e-06,
+ "loss": 0.38469388484954836,
+ "num_input_tokens_seen": 182331456,
+ "step": 4165,
+ "train_runtime": 32237.7321,
+ "train_tokens_per_second": 5655.84
+ },
+ {
+ "epoch": 0.2807702666307568,
+ "grad_norm": 0.9082810498252065,
+ "learning_rate": 9.521781844069727e-06,
+ "loss": 0.4097011566162109,
+ "num_input_tokens_seen": 182555536,
+ "step": 4170,
+ "train_runtime": 32282.4632,
+ "train_tokens_per_second": 5654.944
+ },
+ {
+ "epoch": 0.28110692162671697,
+ "grad_norm": 0.8211424225834406,
+ "learning_rate": 9.520652775477154e-06,
+ "loss": 0.38606297969818115,
+ "num_input_tokens_seen": 182776104,
+ "step": 4175,
+ "train_runtime": 32324.1374,
+ "train_tokens_per_second": 5654.477
+ },
+ {
+ "epoch": 0.2814435766226771,
+ "grad_norm": 0.6834380753616917,
+ "learning_rate": 9.519522442698447e-06,
+ "loss": 0.36697864532470703,
+ "num_input_tokens_seen": 182986640,
+ "step": 4180,
+ "train_runtime": 32354.2522,
+ "train_tokens_per_second": 5655.722
+ },
+ {
+ "epoch": 0.28178023161863724,
+ "grad_norm": 0.8351889375982321,
+ "learning_rate": 9.518390846049699e-06,
+ "loss": 0.4045860290527344,
+ "num_input_tokens_seen": 183200160,
+ "step": 4185,
+ "train_runtime": 32398.1959,
+ "train_tokens_per_second": 5654.641
+ },
+ {
+ "epoch": 0.2821168866145974,
+ "grad_norm": 0.7607561247725104,
+ "learning_rate": 9.517257985847361e-06,
+ "loss": 0.42084197998046874,
+ "num_input_tokens_seen": 183411672,
+ "step": 4190,
+ "train_runtime": 32441.1387,
+ "train_tokens_per_second": 5653.676
+ },
+ {
+ "epoch": 0.2824535416105575,
+ "grad_norm": 0.7318149860850908,
+ "learning_rate": 9.516123862408232e-06,
+ "loss": 0.4293764591217041,
+ "num_input_tokens_seen": 183636784,
+ "step": 4195,
+ "train_runtime": 32485.1985,
+ "train_tokens_per_second": 5652.937
+ },
+ {
+ "epoch": 0.28279019660651766,
+ "grad_norm": 1.0902481858366875,
+ "learning_rate": 9.514988476049466e-06,
+ "loss": 0.3744691848754883,
+ "num_input_tokens_seen": 183850024,
+ "step": 4200,
+ "train_runtime": 32521.0388,
+ "train_tokens_per_second": 5653.264
+ },
+ {
+ "epoch": 0.2831268516024778,
+ "grad_norm": 0.7196863893637454,
+ "learning_rate": 9.513851827088567e-06,
+ "loss": 0.39441556930541993,
+ "num_input_tokens_seen": 184055312,
+ "step": 4205,
+ "train_runtime": 32558.0875,
+ "train_tokens_per_second": 5653.136
+ },
+ {
+ "epoch": 0.28346350659843794,
+ "grad_norm": 0.901667167722795,
+ "learning_rate": 9.512713915843402e-06,
+ "loss": 0.399395227432251,
+ "num_input_tokens_seen": 184280840,
+ "step": 4210,
+ "train_runtime": 32591.1094,
+ "train_tokens_per_second": 5654.329
+ },
+ {
+ "epoch": 0.2838001615943981,
+ "grad_norm": 0.7871649191207917,
+ "learning_rate": 9.511574742632177e-06,
+ "loss": 0.38000011444091797,
+ "num_input_tokens_seen": 184499800,
+ "step": 4215,
+ "train_runtime": 32635.0953,
+ "train_tokens_per_second": 5653.417
+ },
+ {
+ "epoch": 0.2841368165903582,
+ "grad_norm": 0.7805206391000944,
+ "learning_rate": 9.510434307773464e-06,
+ "loss": 0.45735921859741213,
+ "num_input_tokens_seen": 184725080,
+ "step": 4220,
+ "train_runtime": 32673.8214,
+ "train_tokens_per_second": 5653.611
+ },
+ {
+ "epoch": 0.28447347158631836,
+ "grad_norm": 0.6332223820184592,
+ "learning_rate": 9.509292611586179e-06,
+ "loss": 0.35365097522735595,
+ "num_input_tokens_seen": 184912768,
+ "step": 4225,
+ "train_runtime": 32714.0896,
+ "train_tokens_per_second": 5652.389
+ },
+ {
+ "epoch": 0.2848101265822785,
+ "grad_norm": 0.8305072132980625,
+ "learning_rate": 9.508149654389592e-06,
+ "loss": 0.40491819381713867,
+ "num_input_tokens_seen": 185136408,
+ "step": 4230,
+ "train_runtime": 32755.4509,
+ "train_tokens_per_second": 5652.079
+ },
+ {
+ "epoch": 0.28514678157823864,
+ "grad_norm": 0.6650135309711892,
+ "learning_rate": 9.507005436503331e-06,
+ "loss": 0.3746334552764893,
+ "num_input_tokens_seen": 185367152,
+ "step": 4235,
+ "train_runtime": 32798.7805,
+ "train_tokens_per_second": 5651.648
+ },
+ {
+ "epoch": 0.2854834365741988,
+ "grad_norm": 0.8834317004719436,
+ "learning_rate": 9.505859958247373e-06,
+ "loss": 0.4024499893188477,
+ "num_input_tokens_seen": 185580528,
+ "step": 4240,
+ "train_runtime": 32837.3509,
+ "train_tokens_per_second": 5651.507
+ },
+ {
+ "epoch": 0.2858200915701589,
+ "grad_norm": 0.703931424320413,
+ "learning_rate": 9.504713219942045e-06,
+ "loss": 0.3693699359893799,
+ "num_input_tokens_seen": 185800152,
+ "step": 4245,
+ "train_runtime": 32873.2494,
+ "train_tokens_per_second": 5652.017
+ },
+ {
+ "epoch": 0.28615674656611906,
+ "grad_norm": 0.7673763574951983,
+ "learning_rate": 9.503565221908029e-06,
+ "loss": 0.3989689350128174,
+ "num_input_tokens_seen": 186023072,
+ "step": 4250,
+ "train_runtime": 32908.5929,
+ "train_tokens_per_second": 5652.72
+ },
+ {
+ "epoch": 0.2864934015620792,
+ "grad_norm": 0.7942180767418593,
+ "learning_rate": 9.50241596446636e-06,
+ "loss": 0.41822104454040526,
+ "num_input_tokens_seen": 186251344,
+ "step": 4255,
+ "train_runtime": 32948.6784,
+ "train_tokens_per_second": 5652.771
+ },
+ {
+ "epoch": 0.28683005655803934,
+ "grad_norm": 0.8790228905730333,
+ "learning_rate": 9.501265447938423e-06,
+ "loss": 0.40343165397644043,
+ "num_input_tokens_seen": 186459392,
+ "step": 4260,
+ "train_runtime": 32985.9245,
+ "train_tokens_per_second": 5652.696
+ },
+ {
+ "epoch": 0.2871667115539995,
+ "grad_norm": 0.8360538773951802,
+ "learning_rate": 9.500113672645958e-06,
+ "loss": 0.41559224128723143,
+ "num_input_tokens_seen": 186674536,
+ "step": 4265,
+ "train_runtime": 33020.78,
+ "train_tokens_per_second": 5653.244
+ },
+ {
+ "epoch": 0.2875033665499596,
+ "grad_norm": 0.8611168073988301,
+ "learning_rate": 9.498960638911054e-06,
+ "loss": 0.3698106288909912,
+ "num_input_tokens_seen": 186883568,
+ "step": 4270,
+ "train_runtime": 33066.6343,
+ "train_tokens_per_second": 5651.726
+ },
+ {
+ "epoch": 0.28784002154591976,
+ "grad_norm": 0.7947411629350094,
+ "learning_rate": 9.497806347056153e-06,
+ "loss": 0.394811749458313,
+ "num_input_tokens_seen": 187101448,
+ "step": 4275,
+ "train_runtime": 33108.0971,
+ "train_tokens_per_second": 5651.229
+ },
+ {
+ "epoch": 0.2881766765418799,
+ "grad_norm": 0.8271288744866215,
+ "learning_rate": 9.49665079740405e-06,
+ "loss": 0.41505231857299807,
+ "num_input_tokens_seen": 187323568,
+ "step": 4280,
+ "train_runtime": 33144.1133,
+ "train_tokens_per_second": 5651.79
+ },
+ {
+ "epoch": 0.28851333153784003,
+ "grad_norm": 0.8870087883370243,
+ "learning_rate": 9.495493990277889e-06,
+ "loss": 0.41215314865112307,
+ "num_input_tokens_seen": 187536768,
+ "step": 4285,
+ "train_runtime": 33177.0201,
+ "train_tokens_per_second": 5652.61
+ },
+ {
+ "epoch": 0.2888499865338002,
+ "grad_norm": 0.7416815526876334,
+ "learning_rate": 9.49433592600117e-06,
+ "loss": 0.39543790817260743,
+ "num_input_tokens_seen": 187761936,
+ "step": 4290,
+ "train_runtime": 33218.3652,
+ "train_tokens_per_second": 5652.353
+ },
+ {
+ "epoch": 0.2891866415297603,
+ "grad_norm": 0.8456820374140082,
+ "learning_rate": 9.49317660489774e-06,
+ "loss": 0.38694300651550295,
+ "num_input_tokens_seen": 187989768,
+ "step": 4295,
+ "train_runtime": 33254.7613,
+ "train_tokens_per_second": 5653.018
+ },
+ {
+ "epoch": 0.28952329652572045,
+ "grad_norm": 0.7340597270803788,
+ "learning_rate": 9.4920160272918e-06,
+ "loss": 0.39570274353027346,
+ "num_input_tokens_seen": 188221328,
+ "step": 4300,
+ "train_runtime": 33292.1232,
+ "train_tokens_per_second": 5653.629
+ },
+ {
+ "epoch": 0.2898599515216806,
+ "grad_norm": 0.9417045729052645,
+ "learning_rate": 9.490854193507904e-06,
+ "loss": 0.40100975036621095,
+ "num_input_tokens_seen": 188436984,
+ "step": 4305,
+ "train_runtime": 33333.4329,
+ "train_tokens_per_second": 5653.093
+ },
+ {
+ "epoch": 0.29019660651764073,
+ "grad_norm": 0.8307091778797395,
+ "learning_rate": 9.489691103870951e-06,
+ "loss": 0.4165069580078125,
+ "num_input_tokens_seen": 188666800,
+ "step": 4310,
+ "train_runtime": 33370.0506,
+ "train_tokens_per_second": 5653.776
+ },
+ {
+ "epoch": 0.29053326151360087,
+ "grad_norm": 0.6898420962550031,
+ "learning_rate": 9.488526758706198e-06,
+ "loss": 0.3968646764755249,
+ "num_input_tokens_seen": 188890864,
+ "step": 4315,
+ "train_runtime": 33408.9132,
+ "train_tokens_per_second": 5653.906
+ },
+ {
+ "epoch": 0.290869916509561,
+ "grad_norm": 0.9744166501910104,
+ "learning_rate": 9.48736115833925e-06,
+ "loss": 0.42871532440185545,
+ "num_input_tokens_seen": 189097152,
+ "step": 4320,
+ "train_runtime": 33444.6308,
+ "train_tokens_per_second": 5654.036
+ },
+ {
+ "epoch": 0.29120657150552115,
+ "grad_norm": 0.7709225593851357,
+ "learning_rate": 9.486194303096062e-06,
+ "loss": 0.388767147064209,
+ "num_input_tokens_seen": 189319032,
+ "step": 4325,
+ "train_runtime": 33480.301,
+ "train_tokens_per_second": 5654.639
+ },
+ {
+ "epoch": 0.2915432265014813,
+ "grad_norm": 0.6810949642678763,
+ "learning_rate": 9.485026193302945e-06,
+ "loss": 0.37795631885528563,
+ "num_input_tokens_seen": 189540056,
+ "step": 4330,
+ "train_runtime": 33515.0635,
+ "train_tokens_per_second": 5655.369
+ },
+ {
+ "epoch": 0.29187988149744143,
+ "grad_norm": 0.6374851744198567,
+ "learning_rate": 9.483856829286552e-06,
+ "loss": 0.43641042709350586,
+ "num_input_tokens_seen": 189765816,
+ "step": 4335,
+ "train_runtime": 33556.7326,
+ "train_tokens_per_second": 5655.074
+ },
+ {
+ "epoch": 0.29221653649340157,
+ "grad_norm": 0.8345515619506504,
+ "learning_rate": 9.482686211373896e-06,
+ "loss": 0.4038511276245117,
+ "num_input_tokens_seen": 189992032,
+ "step": 4340,
+ "train_runtime": 33597.0529,
+ "train_tokens_per_second": 5655.021
+ },
+ {
+ "epoch": 0.2925531914893617,
+ "grad_norm": 0.8775359668232541,
+ "learning_rate": 9.481514339892335e-06,
+ "loss": 0.4172847747802734,
+ "num_input_tokens_seen": 190224704,
+ "step": 4345,
+ "train_runtime": 33633.3907,
+ "train_tokens_per_second": 5655.829
+ },
+ {
+ "epoch": 0.29288984648532185,
+ "grad_norm": 0.7728830127459773,
+ "learning_rate": 9.48034121516958e-06,
+ "loss": 0.41391868591308595,
+ "num_input_tokens_seen": 190437840,
+ "step": 4350,
+ "train_runtime": 33671.0816,
+ "train_tokens_per_second": 5655.828
+ },
+ {
+ "epoch": 0.293226501481282,
+ "grad_norm": 0.7347575506312092,
+ "learning_rate": 9.47916683753369e-06,
+ "loss": 0.38980252742767335,
+ "num_input_tokens_seen": 190656112,
+ "step": 4355,
+ "train_runtime": 33713.543,
+ "train_tokens_per_second": 5655.179
+ },
+ {
+ "epoch": 0.2935631564772421,
+ "grad_norm": 0.7225462237696316,
+ "learning_rate": 9.477991207313077e-06,
+ "loss": 0.4024055480957031,
+ "num_input_tokens_seen": 190883776,
+ "step": 4360,
+ "train_runtime": 33754.4757,
+ "train_tokens_per_second": 5655.066
+ },
+ {
+ "epoch": 0.29389981147320227,
+ "grad_norm": 0.7312109916350326,
+ "learning_rate": 9.476814324836504e-06,
+ "loss": 0.38675341606140134,
+ "num_input_tokens_seen": 191111960,
+ "step": 4365,
+ "train_runtime": 33789.2531,
+ "train_tokens_per_second": 5655.998
+ },
+ {
+ "epoch": 0.2942364664691624,
+ "grad_norm": 0.7752639421468126,
+ "learning_rate": 9.475636190433078e-06,
+ "loss": 0.44519591331481934,
+ "num_input_tokens_seen": 191343016,
+ "step": 4370,
+ "train_runtime": 33827.4824,
+ "train_tokens_per_second": 5656.437
+ },
+ {
+ "epoch": 0.29457312146512254,
+ "grad_norm": 0.7480674780381356,
+ "learning_rate": 9.474456804432264e-06,
+ "loss": 0.3989989995956421,
+ "num_input_tokens_seen": 191572600,
+ "step": 4375,
+ "train_runtime": 33864.2914,
+ "train_tokens_per_second": 5657.068
+ },
+ {
+ "epoch": 0.2949097764610827,
+ "grad_norm": 0.7492609939786371,
+ "learning_rate": 9.473276167163872e-06,
+ "loss": 0.41669330596923826,
+ "num_input_tokens_seen": 191802056,
+ "step": 4380,
+ "train_runtime": 33903.6008,
+ "train_tokens_per_second": 5657.277
+ },
+ {
+ "epoch": 0.2952464314570428,
+ "grad_norm": 0.7254851965917017,
+ "learning_rate": 9.472094278958066e-06,
+ "loss": 0.41675920486450196,
+ "num_input_tokens_seen": 192033856,
+ "step": 4385,
+ "train_runtime": 33947.504,
+ "train_tokens_per_second": 5656.789
+ },
+ {
+ "epoch": 0.29558308645300296,
+ "grad_norm": 0.7301224106455536,
+ "learning_rate": 9.470911140145353e-06,
+ "loss": 0.3885148763656616,
+ "num_input_tokens_seen": 192244376,
+ "step": 4390,
+ "train_runtime": 33990.9485,
+ "train_tokens_per_second": 5655.752
+ },
+ {
+ "epoch": 0.2959197414489631,
+ "grad_norm": 0.8599934571941987,
+ "learning_rate": 9.469726751056599e-06,
+ "loss": 0.3994889736175537,
+ "num_input_tokens_seen": 192466432,
+ "step": 4395,
+ "train_runtime": 34031.0021,
+ "train_tokens_per_second": 5655.62
+ },
+ {
+ "epoch": 0.29625639644492324,
+ "grad_norm": 0.8672385958944913,
+ "learning_rate": 9.468541112023009e-06,
+ "loss": 0.42194247245788574,
+ "num_input_tokens_seen": 192676720,
+ "step": 4400,
+ "train_runtime": 34068.9438,
+ "train_tokens_per_second": 5655.494
+ },
+ {
+ "epoch": 0.2965930514408834,
+ "grad_norm": 0.8654785396115573,
+ "learning_rate": 9.467354223376149e-06,
+ "loss": 0.4138471603393555,
+ "num_input_tokens_seen": 192867824,
+ "step": 4405,
+ "train_runtime": 34101.5758,
+ "train_tokens_per_second": 5655.687
+ },
+ {
+ "epoch": 0.2969297064368435,
+ "grad_norm": 0.7586880855501626,
+ "learning_rate": 9.466166085447923e-06,
+ "loss": 0.4040337562561035,
+ "num_input_tokens_seen": 193079552,
+ "step": 4410,
+ "train_runtime": 34144.2523,
+ "train_tokens_per_second": 5654.819
+ },
+ {
+ "epoch": 0.29726636143280366,
+ "grad_norm": 0.8413204922562729,
+ "learning_rate": 9.464976698570595e-06,
+ "loss": 0.3852673053741455,
+ "num_input_tokens_seen": 193274704,
+ "step": 4415,
+ "train_runtime": 34179.0678,
+ "train_tokens_per_second": 5654.768
+ },
+ {
+ "epoch": 0.2976030164287638,
+ "grad_norm": 0.8227530497102407,
+ "learning_rate": 9.463786063076769e-06,
+ "loss": 0.42119736671447755,
+ "num_input_tokens_seen": 193494016,
+ "step": 4420,
+ "train_runtime": 34217.4917,
+ "train_tokens_per_second": 5654.828
+ },
+ {
+ "epoch": 0.29793967142472394,
+ "grad_norm": 0.8498857055875949,
+ "learning_rate": 9.462594179299408e-06,
+ "loss": 0.34813380241394043,
+ "num_input_tokens_seen": 193708944,
+ "step": 4425,
+ "train_runtime": 34255.6112,
+ "train_tokens_per_second": 5654.809
+ },
+ {
+ "epoch": 0.2982763264206841,
+ "grad_norm": 0.7591384035910306,
+ "learning_rate": 9.461401047571811e-06,
+ "loss": 0.39094252586364747,
+ "num_input_tokens_seen": 193940712,
+ "step": 4430,
+ "train_runtime": 34291.0594,
+ "train_tokens_per_second": 5655.722
+ },
+ {
+ "epoch": 0.2986129814166442,
+ "grad_norm": 0.8057409790041082,
+ "learning_rate": 9.460206668227639e-06,
+ "loss": 0.4155745983123779,
+ "num_input_tokens_seen": 194168936,
+ "step": 4435,
+ "train_runtime": 34330.974,
+ "train_tokens_per_second": 5655.795
+ },
+ {
+ "epoch": 0.29894963641260436,
+ "grad_norm": 0.7522439279097726,
+ "learning_rate": 9.459011041600895e-06,
+ "loss": 0.39428043365478516,
+ "num_input_tokens_seen": 194397872,
+ "step": 4440,
+ "train_runtime": 34371.2051,
+ "train_tokens_per_second": 5655.835
+ },
+ {
+ "epoch": 0.2992862914085645,
+ "grad_norm": 0.8885125640048409,
+ "learning_rate": 9.457814168025932e-06,
+ "loss": 0.3819078207015991,
+ "num_input_tokens_seen": 194618856,
+ "step": 4445,
+ "train_runtime": 34409.2545,
+ "train_tokens_per_second": 5656.003
+ },
+ {
+ "epoch": 0.29962294640452464,
+ "grad_norm": 0.8798594554422206,
+ "learning_rate": 9.456616047837452e-06,
+ "loss": 0.4244089603424072,
+ "num_input_tokens_seen": 194815632,
+ "step": 4450,
+ "train_runtime": 34451.3067,
+ "train_tokens_per_second": 5654.811
+ },
+ {
+ "epoch": 0.2999596014004848,
+ "grad_norm": 0.7140830042785486,
+ "learning_rate": 9.455416681370506e-06,
+ "loss": 0.43105688095092776,
+ "num_input_tokens_seen": 195043320,
+ "step": 4455,
+ "train_runtime": 34491.908,
+ "train_tokens_per_second": 5654.756
+ },
+ {
+ "epoch": 0.3002962563964449,
+ "grad_norm": 0.7933555904953391,
+ "learning_rate": 9.45421606896049e-06,
+ "loss": 0.4107870101928711,
+ "num_input_tokens_seen": 195273800,
+ "step": 4460,
+ "train_runtime": 34532.6899,
+ "train_tokens_per_second": 5654.752
+ },
+ {
+ "epoch": 0.30063291139240506,
+ "grad_norm": 0.7656195298280324,
+ "learning_rate": 9.453014210943155e-06,
+ "loss": 0.39834353923797605,
+ "num_input_tokens_seen": 195492912,
+ "step": 4465,
+ "train_runtime": 34574.8621,
+ "train_tokens_per_second": 5654.192
+ },
+ {
+ "epoch": 0.3009695663883652,
+ "grad_norm": 0.7817448952040295,
+ "learning_rate": 9.451811107654596e-06,
+ "loss": 0.39757986068725587,
+ "num_input_tokens_seen": 195716896,
+ "step": 4470,
+ "train_runtime": 34621.7737,
+ "train_tokens_per_second": 5653.0
+ },
+ {
+ "epoch": 0.30130622138432533,
+ "grad_norm": 0.7841726368249392,
+ "learning_rate": 9.450606759431255e-06,
+ "loss": 0.40091800689697266,
+ "num_input_tokens_seen": 195945064,
+ "step": 4475,
+ "train_runtime": 34656.1692,
+ "train_tokens_per_second": 5653.974
+ },
+ {
+ "epoch": 0.3016428763802855,
+ "grad_norm": 0.8817669510853672,
+ "learning_rate": 9.449401166609928e-06,
+ "loss": 0.42542014122009275,
+ "num_input_tokens_seen": 196155296,
+ "step": 4480,
+ "train_runtime": 34691.0919,
+ "train_tokens_per_second": 5654.342
+ },
+ {
+ "epoch": 0.3019795313762456,
+ "grad_norm": 0.958339871121062,
+ "learning_rate": 9.448194329527752e-06,
+ "loss": 0.40885372161865235,
+ "num_input_tokens_seen": 196367168,
+ "step": 4485,
+ "train_runtime": 34725.1749,
+ "train_tokens_per_second": 5654.894
+ },
+ {
+ "epoch": 0.30231618637220575,
+ "grad_norm": 0.7219539133285781,
+ "learning_rate": 9.446986248522216e-06,
+ "loss": 0.38100175857543944,
+ "num_input_tokens_seen": 196585760,
+ "step": 4490,
+ "train_runtime": 34763.5979,
+ "train_tokens_per_second": 5654.931
+ },
+ {
+ "epoch": 0.3026528413681659,
+ "grad_norm": 0.7623511745984738,
+ "learning_rate": 9.445776923931157e-06,
+ "loss": 0.3696560859680176,
+ "num_input_tokens_seen": 196799296,
+ "step": 4495,
+ "train_runtime": 34801.3851,
+ "train_tokens_per_second": 5654.927
+ },
+ {
+ "epoch": 0.30298949636412603,
+ "grad_norm": 0.7900772624336759,
+ "learning_rate": 9.444566356092754e-06,
+ "loss": 0.3985780715942383,
+ "num_input_tokens_seen": 197027464,
+ "step": 4500,
+ "train_runtime": 34835.2173,
+ "train_tokens_per_second": 5655.985
+ },
+ {
+ "epoch": 0.30332615136008617,
+ "grad_norm": 0.8352145795658186,
+ "learning_rate": 9.443354545345545e-06,
+ "loss": 0.36351020336151124,
+ "num_input_tokens_seen": 197246488,
+ "step": 4505,
+ "train_runtime": 34872.1679,
+ "train_tokens_per_second": 5656.273
+ },
+ {
+ "epoch": 0.3036628063560463,
+ "grad_norm": 0.7235873259829829,
+ "learning_rate": 9.442141492028406e-06,
+ "loss": 0.3931256294250488,
+ "num_input_tokens_seen": 197464240,
+ "step": 4510,
+ "train_runtime": 34916.8698,
+ "train_tokens_per_second": 5655.268
+ },
+ {
+ "epoch": 0.30399946135200645,
+ "grad_norm": 0.7287455155410504,
+ "learning_rate": 9.440927196480563e-06,
+ "loss": 0.3859192609786987,
+ "num_input_tokens_seen": 197693416,
+ "step": 4515,
+ "train_runtime": 34952.6701,
+ "train_tokens_per_second": 5656.032
+ },
+ {
+ "epoch": 0.3043361163479666,
+ "grad_norm": 0.7124742821058738,
+ "learning_rate": 9.439711659041593e-06,
+ "loss": 0.3971871376037598,
+ "num_input_tokens_seen": 197908880,
+ "step": 4520,
+ "train_runtime": 34995.5331,
+ "train_tokens_per_second": 5655.261
+ },
+ {
+ "epoch": 0.30467277134392673,
+ "grad_norm": 0.8049967305897996,
+ "learning_rate": 9.438494880051413e-06,
+ "loss": 0.4050449371337891,
+ "num_input_tokens_seen": 198130240,
+ "step": 4525,
+ "train_runtime": 35031.404,
+ "train_tokens_per_second": 5655.789
+ },
+ {
+ "epoch": 0.30500942633988687,
+ "grad_norm": 0.8167232313014808,
+ "learning_rate": 9.437276859850293e-06,
+ "loss": 0.39906351566314696,
+ "num_input_tokens_seen": 198354624,
+ "step": 4530,
+ "train_runtime": 35070.2405,
+ "train_tokens_per_second": 5655.924
+ },
+ {
+ "epoch": 0.305346081335847,
+ "grad_norm": 0.8219257479433354,
+ "learning_rate": 9.436057598778848e-06,
+ "loss": 0.41982173919677734,
+ "num_input_tokens_seen": 198569592,
+ "step": 4535,
+ "train_runtime": 35103.488,
+ "train_tokens_per_second": 5656.691
+ },
+ {
+ "epoch": 0.30568273633180715,
+ "grad_norm": 0.8373417043715061,
+ "learning_rate": 9.434837097178043e-06,
+ "loss": 0.44964447021484377,
+ "num_input_tokens_seen": 198790848,
+ "step": 4540,
+ "train_runtime": 35136.0232,
+ "train_tokens_per_second": 5657.75
+ },
+ {
+ "epoch": 0.3060193913277673,
+ "grad_norm": 0.8551654564560078,
+ "learning_rate": 9.433615355389183e-06,
+ "loss": 0.4145000457763672,
+ "num_input_tokens_seen": 198995864,
+ "step": 4545,
+ "train_runtime": 35171.0896,
+ "train_tokens_per_second": 5657.939
+ },
+ {
+ "epoch": 0.3063560463237274,
+ "grad_norm": 0.8539775390580207,
+ "learning_rate": 9.432392373753926e-06,
+ "loss": 0.38867299556732177,
+ "num_input_tokens_seen": 199228112,
+ "step": 4550,
+ "train_runtime": 35204.7853,
+ "train_tokens_per_second": 5659.12
+ },
+ {
+ "epoch": 0.30669270131968757,
+ "grad_norm": 0.7417346548045687,
+ "learning_rate": 9.431168152614278e-06,
+ "loss": 0.398314380645752,
+ "num_input_tokens_seen": 199444448,
+ "step": 4555,
+ "train_runtime": 35246.8496,
+ "train_tokens_per_second": 5658.504
+ },
+ {
+ "epoch": 0.3070293563156477,
+ "grad_norm": 0.7867984496679544,
+ "learning_rate": 9.429942692312585e-06,
+ "loss": 0.36063923835754397,
+ "num_input_tokens_seen": 199665776,
+ "step": 4560,
+ "train_runtime": 35277.6142,
+ "train_tokens_per_second": 5659.844
+ },
+ {
+ "epoch": 0.30736601131160785,
+ "grad_norm": 0.6064626251289161,
+ "learning_rate": 9.428715993191546e-06,
+ "loss": 0.38531041145324707,
+ "num_input_tokens_seen": 199888952,
+ "step": 4565,
+ "train_runtime": 35318.7352,
+ "train_tokens_per_second": 5659.573
+ },
+ {
+ "epoch": 0.307702666307568,
+ "grad_norm": 0.7823320659003977,
+ "learning_rate": 9.427488055594199e-06,
+ "loss": 0.4015589714050293,
+ "num_input_tokens_seen": 200094448,
+ "step": 4570,
+ "train_runtime": 35356.6905,
+ "train_tokens_per_second": 5659.309
+ },
+ {
+ "epoch": 0.3080393213035281,
+ "grad_norm": 0.7797326812244397,
+ "learning_rate": 9.426258879863937e-06,
+ "loss": 0.40995278358459475,
+ "num_input_tokens_seen": 200330936,
+ "step": 4575,
+ "train_runtime": 35388.9695,
+ "train_tokens_per_second": 5660.83
+ },
+ {
+ "epoch": 0.30837597629948826,
+ "grad_norm": 0.6558180029153803,
+ "learning_rate": 9.425028466344494e-06,
+ "loss": 0.35400965213775637,
+ "num_input_tokens_seen": 200550336,
+ "step": 4580,
+ "train_runtime": 35423.4598,
+ "train_tokens_per_second": 5661.512
+ },
+ {
+ "epoch": 0.3087126312954484,
+ "grad_norm": 0.9357449426999332,
+ "learning_rate": 9.42379681537995e-06,
+ "loss": 0.40514564514160156,
+ "num_input_tokens_seen": 200767584,
+ "step": 4585,
+ "train_runtime": 35466.8621,
+ "train_tokens_per_second": 5660.709
+ },
+ {
+ "epoch": 0.30904928629140854,
+ "grad_norm": 1.1029634658189478,
+ "learning_rate": 9.422563927314732e-06,
+ "loss": 0.39676432609558104,
+ "num_input_tokens_seen": 200990368,
+ "step": 4590,
+ "train_runtime": 35503.267,
+ "train_tokens_per_second": 5661.18
+ },
+ {
+ "epoch": 0.3093859412873687,
+ "grad_norm": 0.7960058470606449,
+ "learning_rate": 9.421329802493615e-06,
+ "loss": 0.3798659801483154,
+ "num_input_tokens_seen": 201213136,
+ "step": 4595,
+ "train_runtime": 35545.6953,
+ "train_tokens_per_second": 5660.689
+ },
+ {
+ "epoch": 0.3097225962833288,
+ "grad_norm": 0.7764466461774601,
+ "learning_rate": 9.420094441261717e-06,
+ "loss": 0.4128886222839355,
+ "num_input_tokens_seen": 201436008,
+ "step": 4600,
+ "train_runtime": 35591.0272,
+ "train_tokens_per_second": 5659.741
+ },
+ {
+ "epoch": 0.31005925127928896,
+ "grad_norm": 0.8514809314658448,
+ "learning_rate": 9.418857843964506e-06,
+ "loss": 0.3874027252197266,
+ "num_input_tokens_seen": 201651024,
+ "step": 4605,
+ "train_runtime": 35633.1312,
+ "train_tokens_per_second": 5659.088
+ },
+ {
+ "epoch": 0.3103959062752491,
+ "grad_norm": 0.791242525351627,
+ "learning_rate": 9.417620010947786e-06,
+ "loss": 0.41231350898742675,
+ "num_input_tokens_seen": 201862568,
+ "step": 4610,
+ "train_runtime": 35672.8097,
+ "train_tokens_per_second": 5658.724
+ },
+ {
+ "epoch": 0.31073256127120924,
+ "grad_norm": 0.8196831043483923,
+ "learning_rate": 9.416380942557719e-06,
+ "loss": 0.3759624719619751,
+ "num_input_tokens_seen": 202081848,
+ "step": 4615,
+ "train_runtime": 35711.7461,
+ "train_tokens_per_second": 5658.694
+ },
+ {
+ "epoch": 0.3110692162671694,
+ "grad_norm": 0.814113401127112,
+ "learning_rate": 9.415140639140803e-06,
+ "loss": 0.40959873199462893,
+ "num_input_tokens_seen": 202288416,
+ "step": 4620,
+ "train_runtime": 35748.2065,
+ "train_tokens_per_second": 5658.701
+ },
+ {
+ "epoch": 0.3114058712631295,
+ "grad_norm": 0.6867352095609621,
+ "learning_rate": 9.413899101043887e-06,
+ "loss": 0.4299426555633545,
+ "num_input_tokens_seen": 202507280,
+ "step": 4625,
+ "train_runtime": 35787.2021,
+ "train_tokens_per_second": 5658.651
+ },
+ {
+ "epoch": 0.31174252625908966,
+ "grad_norm": 0.9127540749426457,
+ "learning_rate": 9.412656328614162e-06,
+ "loss": 0.4051331043243408,
+ "num_input_tokens_seen": 202708520,
+ "step": 4630,
+ "train_runtime": 35826.4689,
+ "train_tokens_per_second": 5658.066
+ },
+ {
+ "epoch": 0.3120791812550498,
+ "grad_norm": 0.7769819809945303,
+ "learning_rate": 9.411412322199165e-06,
+ "loss": 0.40700321197509765,
+ "num_input_tokens_seen": 202938288,
+ "step": 4635,
+ "train_runtime": 35860.2769,
+ "train_tokens_per_second": 5659.139
+ },
+ {
+ "epoch": 0.31241583625100994,
+ "grad_norm": 0.7763840440333488,
+ "learning_rate": 9.41016708214678e-06,
+ "loss": 0.397929310798645,
+ "num_input_tokens_seen": 203140768,
+ "step": 4640,
+ "train_runtime": 35900.8969,
+ "train_tokens_per_second": 5658.376
+ },
+ {
+ "epoch": 0.31275249124697013,
+ "grad_norm": 0.838793433476314,
+ "learning_rate": 9.408920608805233e-06,
+ "loss": 0.3729093074798584,
+ "num_input_tokens_seen": 203353200,
+ "step": 4645,
+ "train_runtime": 35943.6665,
+ "train_tokens_per_second": 5657.553
+ },
+ {
+ "epoch": 0.31308914624293027,
+ "grad_norm": 0.759403553618091,
+ "learning_rate": 9.407672902523097e-06,
+ "loss": 0.42061967849731446,
+ "num_input_tokens_seen": 203576960,
+ "step": 4650,
+ "train_runtime": 35989.9311,
+ "train_tokens_per_second": 5656.498
+ },
+ {
+ "epoch": 0.3134258012388904,
+ "grad_norm": 0.8516208752301061,
+ "learning_rate": 9.406423963649287e-06,
+ "loss": 0.39535982608795167,
+ "num_input_tokens_seen": 203780664,
+ "step": 4655,
+ "train_runtime": 36033.4005,
+ "train_tokens_per_second": 5655.327
+ },
+ {
+ "epoch": 0.31376245623485055,
+ "grad_norm": 0.6661646334234551,
+ "learning_rate": 9.405173792533069e-06,
+ "loss": 0.3896958827972412,
+ "num_input_tokens_seen": 204002504,
+ "step": 4660,
+ "train_runtime": 36071.9272,
+ "train_tokens_per_second": 5655.437
+ },
+ {
+ "epoch": 0.3140991112308107,
+ "grad_norm": 0.802128758388854,
+ "learning_rate": 9.403922389524045e-06,
+ "loss": 0.383258056640625,
+ "num_input_tokens_seen": 204219216,
+ "step": 4665,
+ "train_runtime": 36107.3083,
+ "train_tokens_per_second": 5655.897
+ },
+ {
+ "epoch": 0.31443576622677083,
+ "grad_norm": 0.8311848290104593,
+ "learning_rate": 9.40266975497217e-06,
+ "loss": 0.41370835304260256,
+ "num_input_tokens_seen": 204435080,
+ "step": 4670,
+ "train_runtime": 36146.1312,
+ "train_tokens_per_second": 5655.794
+ },
+ {
+ "epoch": 0.31477242122273097,
+ "grad_norm": 0.7524391894573029,
+ "learning_rate": 9.401415889227736e-06,
+ "loss": 0.4065183162689209,
+ "num_input_tokens_seen": 204665984,
+ "step": 4675,
+ "train_runtime": 36189.9551,
+ "train_tokens_per_second": 5655.326
+ },
+ {
+ "epoch": 0.3151090762186911,
+ "grad_norm": 0.6516968499884368,
+ "learning_rate": 9.400160792641385e-06,
+ "loss": 0.40553603172302244,
+ "num_input_tokens_seen": 204900584,
+ "step": 4680,
+ "train_runtime": 36230.3148,
+ "train_tokens_per_second": 5655.501
+ },
+ {
+ "epoch": 0.31544573121465125,
+ "grad_norm": 0.713149562707507,
+ "learning_rate": 9.398904465564097e-06,
+ "loss": 0.3963156700134277,
+ "num_input_tokens_seen": 205121896,
+ "step": 4685,
+ "train_runtime": 36261.5038,
+ "train_tokens_per_second": 5656.74
+ },
+ {
+ "epoch": 0.3157823862106114,
+ "grad_norm": 0.7445101762458055,
+ "learning_rate": 9.397646908347203e-06,
+ "loss": 0.3701439142227173,
+ "num_input_tokens_seen": 205337424,
+ "step": 4690,
+ "train_runtime": 36299.7979,
+ "train_tokens_per_second": 5656.71
+ },
+ {
+ "epoch": 0.3161190412065715,
+ "grad_norm": 0.7876242336850148,
+ "learning_rate": 9.396388121342375e-06,
+ "loss": 0.37079999446868894,
+ "num_input_tokens_seen": 205566880,
+ "step": 4695,
+ "train_runtime": 36337.2628,
+ "train_tokens_per_second": 5657.192
+ },
+ {
+ "epoch": 0.31645569620253167,
+ "grad_norm": 0.8752489859691854,
+ "learning_rate": 9.395128104901628e-06,
+ "loss": 0.382697606086731,
+ "num_input_tokens_seen": 205780336,
+ "step": 4700,
+ "train_runtime": 36373.0101,
+ "train_tokens_per_second": 5657.501
+ },
+ {
+ "epoch": 0.3167923511984918,
+ "grad_norm": 0.6981702689866264,
+ "learning_rate": 9.393866859377321e-06,
+ "loss": 0.4276883125305176,
+ "num_input_tokens_seen": 206004992,
+ "step": 4705,
+ "train_runtime": 36409.7995,
+ "train_tokens_per_second": 5657.955
+ },
+ {
+ "epoch": 0.31712900619445195,
+ "grad_norm": 0.7289408984257015,
+ "learning_rate": 9.392604385122157e-06,
+ "loss": 0.39430766105651854,
+ "num_input_tokens_seen": 206237344,
+ "step": 4710,
+ "train_runtime": 36441.1815,
+ "train_tokens_per_second": 5659.458
+ },
+ {
+ "epoch": 0.3174656611904121,
+ "grad_norm": 0.7707325345782555,
+ "learning_rate": 9.391340682489185e-06,
+ "loss": 0.3808866024017334,
+ "num_input_tokens_seen": 206449496,
+ "step": 4715,
+ "train_runtime": 36480.1405,
+ "train_tokens_per_second": 5659.23
+ },
+ {
+ "epoch": 0.3178023161863722,
+ "grad_norm": 0.8379775589672267,
+ "learning_rate": 9.390075751831794e-06,
+ "loss": 0.39603471755981445,
+ "num_input_tokens_seen": 206677304,
+ "step": 4720,
+ "train_runtime": 36513.5133,
+ "train_tokens_per_second": 5660.296
+ },
+ {
+ "epoch": 0.31813897118233236,
+ "grad_norm": 1.1929541297616924,
+ "learning_rate": 9.388809593503718e-06,
+ "loss": 0.41028714179992676,
+ "num_input_tokens_seen": 206901904,
+ "step": 4725,
+ "train_runtime": 36554.4289,
+ "train_tokens_per_second": 5660.105
+ },
+ {
+ "epoch": 0.3184756261782925,
+ "grad_norm": 0.8413335424516502,
+ "learning_rate": 9.387542207859034e-06,
+ "loss": 0.428537654876709,
+ "num_input_tokens_seen": 207122320,
+ "step": 4730,
+ "train_runtime": 36584.6009,
+ "train_tokens_per_second": 5661.462
+ },
+ {
+ "epoch": 0.31881228117425264,
+ "grad_norm": 0.7898247112897376,
+ "learning_rate": 9.386273595252161e-06,
+ "loss": 0.41155662536621096,
+ "num_input_tokens_seen": 207352296,
+ "step": 4735,
+ "train_runtime": 36620.4486,
+ "train_tokens_per_second": 5662.2
+ },
+ {
+ "epoch": 0.3191489361702128,
+ "grad_norm": 0.8389868484209729,
+ "learning_rate": 9.385003756037865e-06,
+ "loss": 0.4223154067993164,
+ "num_input_tokens_seen": 207575072,
+ "step": 4740,
+ "train_runtime": 36662.9004,
+ "train_tokens_per_second": 5661.72
+ },
+ {
+ "epoch": 0.3194855911661729,
+ "grad_norm": 0.827214827358242,
+ "learning_rate": 9.383732690571253e-06,
+ "loss": 0.40798358917236327,
+ "num_input_tokens_seen": 207817360,
+ "step": 4745,
+ "train_runtime": 36697.2613,
+ "train_tokens_per_second": 5663.021
+ },
+ {
+ "epoch": 0.31982224616213306,
+ "grad_norm": 0.8033752581626638,
+ "learning_rate": 9.382460399207769e-06,
+ "loss": 0.40032100677490234,
+ "num_input_tokens_seen": 208045464,
+ "step": 4750,
+ "train_runtime": 36739.6396,
+ "train_tokens_per_second": 5662.697
+ },
+ {
+ "epoch": 0.3201589011580932,
+ "grad_norm": 0.7745249893682031,
+ "learning_rate": 9.381186882303212e-06,
+ "loss": 0.3821758985519409,
+ "num_input_tokens_seen": 208245512,
+ "step": 4755,
+ "train_runtime": 36775.3408,
+ "train_tokens_per_second": 5662.64
+ },
+ {
+ "epoch": 0.32049555615405334,
+ "grad_norm": 0.7835752828676862,
+ "learning_rate": 9.379912140213713e-06,
+ "loss": 0.38662168979644773,
+ "num_input_tokens_seen": 208458376,
+ "step": 4760,
+ "train_runtime": 36810.474,
+ "train_tokens_per_second": 5663.018
+ },
+ {
+ "epoch": 0.3208322111500135,
+ "grad_norm": 0.7427789671372642,
+ "learning_rate": 9.37863617329575e-06,
+ "loss": 0.37845721244812014,
+ "num_input_tokens_seen": 208662464,
+ "step": 4765,
+ "train_runtime": 36848.709,
+ "train_tokens_per_second": 5662.68
+ },
+ {
+ "epoch": 0.3211688661459736,
+ "grad_norm": 0.7906549867893844,
+ "learning_rate": 9.377358981906145e-06,
+ "loss": 0.3855098485946655,
+ "num_input_tokens_seen": 208883752,
+ "step": 4770,
+ "train_runtime": 36885.8469,
+ "train_tokens_per_second": 5662.978
+ },
+ {
+ "epoch": 0.32150552114193376,
+ "grad_norm": 0.7000304206526682,
+ "learning_rate": 9.376080566402059e-06,
+ "loss": 0.41453986167907714,
+ "num_input_tokens_seen": 209114544,
+ "step": 4775,
+ "train_runtime": 36923.9104,
+ "train_tokens_per_second": 5663.391
+ },
+ {
+ "epoch": 0.3218421761378939,
+ "grad_norm": 0.7569304220647727,
+ "learning_rate": 9.374800927140994e-06,
+ "loss": 0.33689126968383787,
+ "num_input_tokens_seen": 209342880,
+ "step": 4780,
+ "train_runtime": 36959.0708,
+ "train_tokens_per_second": 5664.181
+ },
+ {
+ "epoch": 0.32217883113385404,
+ "grad_norm": 0.7693618085921944,
+ "learning_rate": 9.373520064480804e-06,
+ "loss": 0.3944025278091431,
+ "num_input_tokens_seen": 209567016,
+ "step": 4785,
+ "train_runtime": 36999.5953,
+ "train_tokens_per_second": 5664.035
+ },
+ {
+ "epoch": 0.3225154861298142,
+ "grad_norm": 0.7545108220793344,
+ "learning_rate": 9.372237978779672e-06,
+ "loss": 0.3935252666473389,
+ "num_input_tokens_seen": 209798224,
+ "step": 4790,
+ "train_runtime": 37035.8347,
+ "train_tokens_per_second": 5664.736
+ },
+ {
+ "epoch": 0.3228521411257743,
+ "grad_norm": 0.7487128271214056,
+ "learning_rate": 9.37095467039613e-06,
+ "loss": 0.3892039775848389,
+ "num_input_tokens_seen": 209998416,
+ "step": 4795,
+ "train_runtime": 37074.0037,
+ "train_tokens_per_second": 5664.304
+ },
+ {
+ "epoch": 0.32318879612173446,
+ "grad_norm": 0.7480661049890698,
+ "learning_rate": 9.369670139689056e-06,
+ "loss": 0.3860762119293213,
+ "num_input_tokens_seen": 210212136,
+ "step": 4800,
+ "train_runtime": 37113.4829,
+ "train_tokens_per_second": 5664.037
+ },
+ {
+ "epoch": 0.3235254511176946,
+ "grad_norm": 0.7846044101372698,
+ "learning_rate": 9.368384387017659e-06,
+ "loss": 0.377323055267334,
+ "num_input_tokens_seen": 210412912,
+ "step": 4805,
+ "train_runtime": 37150.0136,
+ "train_tokens_per_second": 5663.872
+ },
+ {
+ "epoch": 0.32386210611365474,
+ "grad_norm": 0.77183510086721,
+ "learning_rate": 9.367097412741497e-06,
+ "loss": 0.3832876443862915,
+ "num_input_tokens_seen": 210626632,
+ "step": 4810,
+ "train_runtime": 37185.5052,
+ "train_tokens_per_second": 5664.213
+ },
+ {
+ "epoch": 0.3241987611096149,
+ "grad_norm": 0.9148498149260199,
+ "learning_rate": 9.36580921722047e-06,
+ "loss": 0.41832733154296875,
+ "num_input_tokens_seen": 210847304,
+ "step": 4815,
+ "train_runtime": 37226.3372,
+ "train_tokens_per_second": 5663.928
+ },
+ {
+ "epoch": 0.324535416105575,
+ "grad_norm": 0.6839164180369833,
+ "learning_rate": 9.364519800814818e-06,
+ "loss": 0.3626012563705444,
+ "num_input_tokens_seen": 211071840,
+ "step": 4820,
+ "train_runtime": 37269.7466,
+ "train_tokens_per_second": 5663.356
+ },
+ {
+ "epoch": 0.32487207110153515,
+ "grad_norm": 0.6658908418064057,
+ "learning_rate": 9.36322916388512e-06,
+ "loss": 0.3882687330245972,
+ "num_input_tokens_seen": 211297288,
+ "step": 4825,
+ "train_runtime": 37311.4159,
+ "train_tokens_per_second": 5663.073
+ },
+ {
+ "epoch": 0.3252087260974953,
+ "grad_norm": 0.7420928499905488,
+ "learning_rate": 9.3619373067923e-06,
+ "loss": 0.3612262487411499,
+ "num_input_tokens_seen": 211529896,
+ "step": 4830,
+ "train_runtime": 37348.4533,
+ "train_tokens_per_second": 5663.686
+ },
+ {
+ "epoch": 0.32554538109345543,
+ "grad_norm": 0.7491528413015758,
+ "learning_rate": 9.360644229897622e-06,
+ "loss": 0.39133119583129883,
+ "num_input_tokens_seen": 211744104,
+ "step": 4835,
+ "train_runtime": 37390.6137,
+ "train_tokens_per_second": 5663.028
+ },
+ {
+ "epoch": 0.32588203608941557,
+ "grad_norm": 0.8362580597115173,
+ "learning_rate": 9.35934993356269e-06,
+ "loss": 0.4070117950439453,
+ "num_input_tokens_seen": 211960664,
+ "step": 4840,
+ "train_runtime": 37434.3008,
+ "train_tokens_per_second": 5662.204
+ },
+ {
+ "epoch": 0.3262186910853757,
+ "grad_norm": 0.8872020316370295,
+ "learning_rate": 9.35805441814945e-06,
+ "loss": 0.37361793518066405,
+ "num_input_tokens_seen": 212184888,
+ "step": 4845,
+ "train_runtime": 37481.2897,
+ "train_tokens_per_second": 5661.088
+ },
+ {
+ "epoch": 0.32655534608133585,
+ "grad_norm": 0.7655564350170675,
+ "learning_rate": 9.356757684020188e-06,
+ "loss": 0.40865292549133303,
+ "num_input_tokens_seen": 212387000,
+ "step": 4850,
+ "train_runtime": 37521.1842,
+ "train_tokens_per_second": 5660.456
+ },
+ {
+ "epoch": 0.326892001077296,
+ "grad_norm": 0.9840119081755966,
+ "learning_rate": 9.355459731537533e-06,
+ "loss": 0.37503221035003664,
+ "num_input_tokens_seen": 212602240,
+ "step": 4855,
+ "train_runtime": 37563.6776,
+ "train_tokens_per_second": 5659.782
+ },
+ {
+ "epoch": 0.32722865607325613,
+ "grad_norm": 0.8416963207567933,
+ "learning_rate": 9.354160561064451e-06,
+ "loss": 0.4150728225708008,
+ "num_input_tokens_seen": 212829424,
+ "step": 4860,
+ "train_runtime": 37600.8083,
+ "train_tokens_per_second": 5660.235
+ },
+ {
+ "epoch": 0.32756531106921627,
+ "grad_norm": 0.6549002669311522,
+ "learning_rate": 9.352860172964254e-06,
+ "loss": 0.41104774475097655,
+ "num_input_tokens_seen": 213059616,
+ "step": 4865,
+ "train_runtime": 37639.635,
+ "train_tokens_per_second": 5660.512
+ },
+ {
+ "epoch": 0.3279019660651764,
+ "grad_norm": 0.7662647212420475,
+ "learning_rate": 9.35155856760059e-06,
+ "loss": 0.37230610847473145,
+ "num_input_tokens_seen": 213291344,
+ "step": 4870,
+ "train_runtime": 37683.6107,
+ "train_tokens_per_second": 5660.056
+ },
+ {
+ "epoch": 0.32823862106113655,
+ "grad_norm": 0.8032964428930659,
+ "learning_rate": 9.35025574533745e-06,
+ "loss": 0.4028165817260742,
+ "num_input_tokens_seen": 213518768,
+ "step": 4875,
+ "train_runtime": 37716.795,
+ "train_tokens_per_second": 5661.106
+ },
+ {
+ "epoch": 0.3285752760570967,
+ "grad_norm": 0.709809886638744,
+ "learning_rate": 9.34895170653916e-06,
+ "loss": 0.37210307121276853,
+ "num_input_tokens_seen": 213742008,
+ "step": 4880,
+ "train_runtime": 37761.13,
+ "train_tokens_per_second": 5660.371
+ },
+ {
+ "epoch": 0.3289119310530568,
+ "grad_norm": 1.0245077977220034,
+ "learning_rate": 9.347646451570394e-06,
+ "loss": 0.3780977725982666,
+ "num_input_tokens_seen": 213967120,
+ "step": 4885,
+ "train_runtime": 37795.7217,
+ "train_tokens_per_second": 5661.147
+ },
+ {
+ "epoch": 0.32924858604901697,
+ "grad_norm": 0.7783625701111361,
+ "learning_rate": 9.346339980796162e-06,
+ "loss": 0.44457039833068845,
+ "num_input_tokens_seen": 214204152,
+ "step": 4890,
+ "train_runtime": 37830.001,
+ "train_tokens_per_second": 5662.282
+ },
+ {
+ "epoch": 0.3295852410449771,
+ "grad_norm": 0.7619788379384177,
+ "learning_rate": 9.345032294581813e-06,
+ "loss": 0.3991098403930664,
+ "num_input_tokens_seen": 214426328,
+ "step": 4895,
+ "train_runtime": 37861.0229,
+ "train_tokens_per_second": 5663.511
+ },
+ {
+ "epoch": 0.32992189604093725,
+ "grad_norm": 0.7063552896906414,
+ "learning_rate": 9.343723393293038e-06,
+ "loss": 0.37960023880004884,
+ "num_input_tokens_seen": 214641400,
+ "step": 4900,
+ "train_runtime": 37900.5478,
+ "train_tokens_per_second": 5663.28
+ },
+ {
+ "epoch": 0.3302585510368974,
+ "grad_norm": 0.9300418965561537,
+ "learning_rate": 9.342413277295869e-06,
+ "loss": 0.3947901725769043,
+ "num_input_tokens_seen": 214865368,
+ "step": 4905,
+ "train_runtime": 37941.8194,
+ "train_tokens_per_second": 5663.022
+ },
+ {
+ "epoch": 0.3305952060328575,
+ "grad_norm": 0.660978808262332,
+ "learning_rate": 9.341101946956672e-06,
+ "loss": 0.39415407180786133,
+ "num_input_tokens_seen": 215077016,
+ "step": 4910,
+ "train_runtime": 37976.1785,
+ "train_tokens_per_second": 5663.472
+ },
+ {
+ "epoch": 0.33093186102881766,
+ "grad_norm": 0.7480448807625937,
+ "learning_rate": 9.33978940264216e-06,
+ "loss": 0.4069984436035156,
+ "num_input_tokens_seen": 215281168,
+ "step": 4915,
+ "train_runtime": 38020.9215,
+ "train_tokens_per_second": 5662.176
+ },
+ {
+ "epoch": 0.3312685160247778,
+ "grad_norm": 0.9363661586418124,
+ "learning_rate": 9.338475644719377e-06,
+ "loss": 0.396682071685791,
+ "num_input_tokens_seen": 215508928,
+ "step": 4920,
+ "train_runtime": 38060.6534,
+ "train_tokens_per_second": 5662.25
+ },
+ {
+ "epoch": 0.33160517102073794,
+ "grad_norm": 0.7522573238030015,
+ "learning_rate": 9.337160673555719e-06,
+ "loss": 0.3950789451599121,
+ "num_input_tokens_seen": 215730176,
+ "step": 4925,
+ "train_runtime": 38096.6721,
+ "train_tokens_per_second": 5662.704
+ },
+ {
+ "epoch": 0.3319418260166981,
+ "grad_norm": 0.7866963830157505,
+ "learning_rate": 9.335844489518905e-06,
+ "loss": 0.3960768222808838,
+ "num_input_tokens_seen": 215941544,
+ "step": 4930,
+ "train_runtime": 38135.6021,
+ "train_tokens_per_second": 5662.466
+ },
+ {
+ "epoch": 0.3322784810126582,
+ "grad_norm": 0.7771252275599534,
+ "learning_rate": 9.334527092977008e-06,
+ "loss": 0.3927053928375244,
+ "num_input_tokens_seen": 216139784,
+ "step": 4935,
+ "train_runtime": 38172.7334,
+ "train_tokens_per_second": 5662.151
+ },
+ {
+ "epoch": 0.33261513600861836,
+ "grad_norm": 0.7031312284914388,
+ "learning_rate": 9.33320848429843e-06,
+ "loss": 0.3905383825302124,
+ "num_input_tokens_seen": 216376872,
+ "step": 4940,
+ "train_runtime": 38213.8052,
+ "train_tokens_per_second": 5662.27
+ },
+ {
+ "epoch": 0.3329517910045785,
+ "grad_norm": 0.7992397342780773,
+ "learning_rate": 9.331888663851917e-06,
+ "loss": 0.41385631561279296,
+ "num_input_tokens_seen": 216602432,
+ "step": 4945,
+ "train_runtime": 38247.7905,
+ "train_tokens_per_second": 5663.136
+ },
+ {
+ "epoch": 0.33328844600053864,
+ "grad_norm": 0.8325843513269255,
+ "learning_rate": 9.330567632006552e-06,
+ "loss": 0.402661657333374,
+ "num_input_tokens_seen": 216820040,
+ "step": 4950,
+ "train_runtime": 38282.5886,
+ "train_tokens_per_second": 5663.672
+ },
+ {
+ "epoch": 0.3336251009964988,
+ "grad_norm": 0.8157610005939018,
+ "learning_rate": 9.329245389131759e-06,
+ "loss": 0.4041293144226074,
+ "num_input_tokens_seen": 217040648,
+ "step": 4955,
+ "train_runtime": 38321.0859,
+ "train_tokens_per_second": 5663.74
+ },
+ {
+ "epoch": 0.3339617559924589,
+ "grad_norm": 0.7514981142846812,
+ "learning_rate": 9.327921935597298e-06,
+ "loss": 0.373291015625,
+ "num_input_tokens_seen": 217253544,
+ "step": 4960,
+ "train_runtime": 38362.9141,
+ "train_tokens_per_second": 5663.114
+ },
+ {
+ "epoch": 0.33429841098841906,
+ "grad_norm": 0.9584412932125773,
+ "learning_rate": 9.326597271773267e-06,
+ "loss": 0.39936108589172364,
+ "num_input_tokens_seen": 217452896,
+ "step": 4965,
+ "train_runtime": 38404.9607,
+ "train_tokens_per_second": 5662.104
+ },
+ {
+ "epoch": 0.3346350659843792,
+ "grad_norm": 0.7914463925442367,
+ "learning_rate": 9.325271398030107e-06,
+ "loss": 0.3928532600402832,
+ "num_input_tokens_seen": 217666152,
+ "step": 4970,
+ "train_runtime": 38441.97,
+ "train_tokens_per_second": 5662.201
+ },
+ {
+ "epoch": 0.33497172098033934,
+ "grad_norm": 0.8843279935185989,
+ "learning_rate": 9.323944314738591e-06,
+ "loss": 0.40926666259765626,
+ "num_input_tokens_seen": 217886336,
+ "step": 4975,
+ "train_runtime": 38486.6001,
+ "train_tokens_per_second": 5661.356
+ },
+ {
+ "epoch": 0.3353083759762995,
+ "grad_norm": 0.7413319586045699,
+ "learning_rate": 9.322616022269838e-06,
+ "loss": 0.3675239562988281,
+ "num_input_tokens_seen": 218092776,
+ "step": 4980,
+ "train_runtime": 38523.0427,
+ "train_tokens_per_second": 5661.359
+ },
+ {
+ "epoch": 0.3356450309722596,
+ "grad_norm": 0.8050536335877444,
+ "learning_rate": 9.321286520995297e-06,
+ "loss": 0.40448784828186035,
+ "num_input_tokens_seen": 218303136,
+ "step": 4985,
+ "train_runtime": 38563.0671,
+ "train_tokens_per_second": 5660.938
+ },
+ {
+ "epoch": 0.33598168596821976,
+ "grad_norm": 0.8220117119004713,
+ "learning_rate": 9.319955811286758e-06,
+ "loss": 0.38973236083984375,
+ "num_input_tokens_seen": 218521848,
+ "step": 4990,
+ "train_runtime": 38606.842,
+ "train_tokens_per_second": 5660.184
+ },
+ {
+ "epoch": 0.3363183409641799,
+ "grad_norm": 0.724656690749584,
+ "learning_rate": 9.318623893516354e-06,
+ "loss": 0.38318960666656493,
+ "num_input_tokens_seen": 218733712,
+ "step": 4995,
+ "train_runtime": 38643.6596,
+ "train_tokens_per_second": 5660.274
+ },
+ {
+ "epoch": 0.33665499596014004,
+ "grad_norm": 0.777455286936359,
+ "learning_rate": 9.317290768056548e-06,
+ "loss": 0.3933337450027466,
+ "num_input_tokens_seen": 218942384,
+ "step": 5000,
+ "train_runtime": 38685.6918,
+ "train_tokens_per_second": 5659.518
+ },
+ {
+ "epoch": 0.3369916509561002,
+ "grad_norm": 0.7933540714498255,
+ "learning_rate": 9.315956435280147e-06,
+ "loss": 0.40426931381225584,
+ "num_input_tokens_seen": 219148584,
+ "step": 5005,
+ "train_runtime": 38726.8749,
+ "train_tokens_per_second": 5658.824
+ },
+ {
+ "epoch": 0.3373283059520603,
+ "grad_norm": 0.7310653600558,
+ "learning_rate": 9.314620895560292e-06,
+ "loss": 0.3823359251022339,
+ "num_input_tokens_seen": 219366640,
+ "step": 5010,
+ "train_runtime": 38759.9606,
+ "train_tokens_per_second": 5659.62
+ },
+ {
+ "epoch": 0.33766496094802045,
+ "grad_norm": 0.8607865029821252,
+ "learning_rate": 9.313284149270459e-06,
+ "loss": 0.4273562431335449,
+ "num_input_tokens_seen": 219580136,
+ "step": 5015,
+ "train_runtime": 38797.085,
+ "train_tokens_per_second": 5659.707
+ },
+ {
+ "epoch": 0.3380016159439806,
+ "grad_norm": 0.8291229922855889,
+ "learning_rate": 9.311946196784469e-06,
+ "loss": 0.396860933303833,
+ "num_input_tokens_seen": 219795256,
+ "step": 5020,
+ "train_runtime": 38838.6299,
+ "train_tokens_per_second": 5659.192
+ },
+ {
+ "epoch": 0.33833827093994073,
+ "grad_norm": 0.746756264356882,
+ "learning_rate": 9.310607038476476e-06,
+ "loss": 0.40080647468566893,
+ "num_input_tokens_seen": 220007056,
+ "step": 5025,
+ "train_runtime": 38882.3681,
+ "train_tokens_per_second": 5658.273
+ },
+ {
+ "epoch": 0.3386749259359009,
+ "grad_norm": 0.7715704037440936,
+ "learning_rate": 9.30926667472097e-06,
+ "loss": 0.3821887016296387,
+ "num_input_tokens_seen": 220230040,
+ "step": 5030,
+ "train_runtime": 38929.962,
+ "train_tokens_per_second": 5657.083
+ },
+ {
+ "epoch": 0.339011580931861,
+ "grad_norm": 0.695121253798547,
+ "learning_rate": 9.307925105892779e-06,
+ "loss": 0.4106111526489258,
+ "num_input_tokens_seen": 220463304,
+ "step": 5035,
+ "train_runtime": 38964.7555,
+ "train_tokens_per_second": 5658.018
+ },
+ {
+ "epoch": 0.33934823592782115,
+ "grad_norm": 0.7852274341067195,
+ "learning_rate": 9.30658233236707e-06,
+ "loss": 0.38834619522094727,
+ "num_input_tokens_seen": 220695456,
+ "step": 5040,
+ "train_runtime": 39007.1954,
+ "train_tokens_per_second": 5657.814
+ },
+ {
+ "epoch": 0.3396848909237813,
+ "grad_norm": 0.759484699190671,
+ "learning_rate": 9.30523835451934e-06,
+ "loss": 0.3923861265182495,
+ "num_input_tokens_seen": 220915344,
+ "step": 5045,
+ "train_runtime": 39053.4813,
+ "train_tokens_per_second": 5656.739
+ },
+ {
+ "epoch": 0.34002154591974143,
+ "grad_norm": 0.7964278768577083,
+ "learning_rate": 9.303893172725437e-06,
+ "loss": 0.4306797981262207,
+ "num_input_tokens_seen": 221146072,
+ "step": 5050,
+ "train_runtime": 39091.2343,
+ "train_tokens_per_second": 5657.178
+ },
+ {
+ "epoch": 0.34035820091570157,
+ "grad_norm": 0.7340454424901695,
+ "learning_rate": 9.302546787361533e-06,
+ "loss": 0.38288087844848634,
+ "num_input_tokens_seen": 221371344,
+ "step": 5055,
+ "train_runtime": 39133.7258,
+ "train_tokens_per_second": 5656.792
+ },
+ {
+ "epoch": 0.3406948559116617,
+ "grad_norm": 0.7383658204672374,
+ "learning_rate": 9.301199198804139e-06,
+ "loss": 0.37805628776550293,
+ "num_input_tokens_seen": 221583280,
+ "step": 5060,
+ "train_runtime": 39172.7195,
+ "train_tokens_per_second": 5656.571
+ },
+ {
+ "epoch": 0.34103151090762185,
+ "grad_norm": 0.8155446108957558,
+ "learning_rate": 9.299850407430102e-06,
+ "loss": 0.4092425346374512,
+ "num_input_tokens_seen": 221808808,
+ "step": 5065,
+ "train_runtime": 39215.053,
+ "train_tokens_per_second": 5656.216
+ },
+ {
+ "epoch": 0.341368165903582,
+ "grad_norm": 0.8253751293294798,
+ "learning_rate": 9.298500413616613e-06,
+ "loss": 0.3920405864715576,
+ "num_input_tokens_seen": 222029984,
+ "step": 5070,
+ "train_runtime": 39251.0729,
+ "train_tokens_per_second": 5656.66
+ },
+ {
+ "epoch": 0.3417048208995421,
+ "grad_norm": 0.7618503900471122,
+ "learning_rate": 9.297149217741188e-06,
+ "loss": 0.3707792043685913,
+ "num_input_tokens_seen": 222251080,
+ "step": 5075,
+ "train_runtime": 39288.8528,
+ "train_tokens_per_second": 5656.848
+ },
+ {
+ "epoch": 0.34204147589550227,
+ "grad_norm": 0.7834580864284257,
+ "learning_rate": 9.295796820181688e-06,
+ "loss": 0.4238715171813965,
+ "num_input_tokens_seen": 222475224,
+ "step": 5080,
+ "train_runtime": 39326.0211,
+ "train_tokens_per_second": 5657.201
+ },
+ {
+ "epoch": 0.3423781308914624,
+ "grad_norm": 0.7875688230457064,
+ "learning_rate": 9.294443221316305e-06,
+ "loss": 0.3979609727859497,
+ "num_input_tokens_seen": 222689712,
+ "step": 5085,
+ "train_runtime": 39362.9107,
+ "train_tokens_per_second": 5657.349
+ },
+ {
+ "epoch": 0.34271478588742255,
+ "grad_norm": 0.6945475678075447,
+ "learning_rate": 9.293088421523569e-06,
+ "loss": 0.41632776260375975,
+ "num_input_tokens_seen": 222903112,
+ "step": 5090,
+ "train_runtime": 39399.1126,
+ "train_tokens_per_second": 5657.567
+ },
+ {
+ "epoch": 0.3430514408833827,
+ "grad_norm": 0.8727362030270556,
+ "learning_rate": 9.291732421182346e-06,
+ "loss": 0.3804980754852295,
+ "num_input_tokens_seen": 223116536,
+ "step": 5095,
+ "train_runtime": 39444.3485,
+ "train_tokens_per_second": 5656.489
+ },
+ {
+ "epoch": 0.3433880958793428,
+ "grad_norm": 0.7868108132642002,
+ "learning_rate": 9.290375220671837e-06,
+ "loss": 0.3959723949432373,
+ "num_input_tokens_seen": 223331328,
+ "step": 5100,
+ "train_runtime": 39483.2714,
+ "train_tokens_per_second": 5656.353
+ },
+ {
+ "epoch": 0.34372475087530296,
+ "grad_norm": 0.8977037421010565,
+ "learning_rate": 9.28901682037158e-06,
+ "loss": 0.3575240135192871,
+ "num_input_tokens_seen": 223545784,
+ "step": 5105,
+ "train_runtime": 39520.4864,
+ "train_tokens_per_second": 5656.453
+ },
+ {
+ "epoch": 0.3440614058712631,
+ "grad_norm": 0.724529969116944,
+ "learning_rate": 9.287657220661442e-06,
+ "loss": 0.41797566413879395,
+ "num_input_tokens_seen": 223772168,
+ "step": 5110,
+ "train_runtime": 39556.4135,
+ "train_tokens_per_second": 5657.039
+ },
+ {
+ "epoch": 0.34439806086722324,
+ "grad_norm": 0.8314015078496961,
+ "learning_rate": 9.286296421921636e-06,
+ "loss": 0.3681803703308105,
+ "num_input_tokens_seen": 223987880,
+ "step": 5115,
+ "train_runtime": 39597.3882,
+ "train_tokens_per_second": 5656.633
+ },
+ {
+ "epoch": 0.3447347158631834,
+ "grad_norm": 0.8879862781401929,
+ "learning_rate": 9.284934424532706e-06,
+ "loss": 0.4501188278198242,
+ "num_input_tokens_seen": 224197912,
+ "step": 5120,
+ "train_runtime": 39637.597,
+ "train_tokens_per_second": 5656.193
+ },
+ {
+ "epoch": 0.3450713708591435,
+ "grad_norm": 0.7067262341475377,
+ "learning_rate": 9.283571228875525e-06,
+ "loss": 0.35322113037109376,
+ "num_input_tokens_seen": 224398888,
+ "step": 5125,
+ "train_runtime": 39676.3313,
+ "train_tokens_per_second": 5655.737
+ },
+ {
+ "epoch": 0.3454080258551037,
+ "grad_norm": 0.7651046532838097,
+ "learning_rate": 9.282206835331311e-06,
+ "loss": 0.4238561153411865,
+ "num_input_tokens_seen": 224624840,
+ "step": 5130,
+ "train_runtime": 39714.5021,
+ "train_tokens_per_second": 5655.99
+ },
+ {
+ "epoch": 0.34574468085106386,
+ "grad_norm": 1.0540025361763135,
+ "learning_rate": 9.28084124428161e-06,
+ "loss": 0.3581863880157471,
+ "num_input_tokens_seen": 224840616,
+ "step": 5135,
+ "train_runtime": 39760.1229,
+ "train_tokens_per_second": 5654.928
+ },
+ {
+ "epoch": 0.346081335847024,
+ "grad_norm": 0.712972521802771,
+ "learning_rate": 9.279474456108305e-06,
+ "loss": 0.37324137687683107,
+ "num_input_tokens_seen": 225076296,
+ "step": 5140,
+ "train_runtime": 39795.5671,
+ "train_tokens_per_second": 5655.813
+ },
+ {
+ "epoch": 0.34641799084298414,
+ "grad_norm": 0.7436296711193636,
+ "learning_rate": 9.278106471193615e-06,
+ "loss": 0.36896820068359376,
+ "num_input_tokens_seen": 225294528,
+ "step": 5145,
+ "train_runtime": 39833.6798,
+ "train_tokens_per_second": 5655.88
+ },
+ {
+ "epoch": 0.3467546458389443,
+ "grad_norm": 0.7808819136322714,
+ "learning_rate": 9.276737289920093e-06,
+ "loss": 0.4088866710662842,
+ "num_input_tokens_seen": 225512648,
+ "step": 5150,
+ "train_runtime": 39875.84,
+ "train_tokens_per_second": 5655.37
+ },
+ {
+ "epoch": 0.3470913008349044,
+ "grad_norm": 0.74354681758287,
+ "learning_rate": 9.275366912670624e-06,
+ "loss": 0.4070155143737793,
+ "num_input_tokens_seen": 225724000,
+ "step": 5155,
+ "train_runtime": 39910.7695,
+ "train_tokens_per_second": 5655.717
+ },
+ {
+ "epoch": 0.34742795583086455,
+ "grad_norm": 0.7443700356039168,
+ "learning_rate": 9.273995339828432e-06,
+ "loss": 0.3892085075378418,
+ "num_input_tokens_seen": 225939192,
+ "step": 5160,
+ "train_runtime": 39950.0434,
+ "train_tokens_per_second": 5655.543
+ },
+ {
+ "epoch": 0.3477646108268247,
+ "grad_norm": 0.8699331338781421,
+ "learning_rate": 9.272622571777072e-06,
+ "loss": 0.42869272232055666,
+ "num_input_tokens_seen": 226161472,
+ "step": 5165,
+ "train_runtime": 39986.0253,
+ "train_tokens_per_second": 5656.013
+ },
+ {
+ "epoch": 0.34810126582278483,
+ "grad_norm": 0.7946248476997148,
+ "learning_rate": 9.271248608900434e-06,
+ "loss": 0.3663766860961914,
+ "num_input_tokens_seen": 226388672,
+ "step": 5170,
+ "train_runtime": 40022.8625,
+ "train_tokens_per_second": 5656.484
+ },
+ {
+ "epoch": 0.348437920818745,
+ "grad_norm": 0.820181936155836,
+ "learning_rate": 9.269873451582741e-06,
+ "loss": 0.3990020275115967,
+ "num_input_tokens_seen": 226604448,
+ "step": 5175,
+ "train_runtime": 40065.8484,
+ "train_tokens_per_second": 5655.801
+ },
+ {
+ "epoch": 0.3487745758147051,
+ "grad_norm": 0.7865052705818104,
+ "learning_rate": 9.268497100208556e-06,
+ "loss": 0.40553035736083987,
+ "num_input_tokens_seen": 226814192,
+ "step": 5180,
+ "train_runtime": 40108.7169,
+ "train_tokens_per_second": 5654.985
+ },
+ {
+ "epoch": 0.34911123081066525,
+ "grad_norm": 0.8597790154802004,
+ "learning_rate": 9.267119555162765e-06,
+ "loss": 0.40463762283325194,
+ "num_input_tokens_seen": 227044824,
+ "step": 5185,
+ "train_runtime": 40148.6974,
+ "train_tokens_per_second": 5655.098
+ },
+ {
+ "epoch": 0.3494478858066254,
+ "grad_norm": 0.7383395894951612,
+ "learning_rate": 9.2657408168306e-06,
+ "loss": 0.3876547336578369,
+ "num_input_tokens_seen": 227264048,
+ "step": 5190,
+ "train_runtime": 40181.6708,
+ "train_tokens_per_second": 5655.913
+ },
+ {
+ "epoch": 0.34978454080258553,
+ "grad_norm": 0.8478101930950076,
+ "learning_rate": 9.264360885597617e-06,
+ "loss": 0.3948791027069092,
+ "num_input_tokens_seen": 227482568,
+ "step": 5195,
+ "train_runtime": 40224.9641,
+ "train_tokens_per_second": 5655.258
+ },
+ {
+ "epoch": 0.35012119579854567,
+ "grad_norm": 0.8312712862431509,
+ "learning_rate": 9.262979761849709e-06,
+ "loss": 0.3999017238616943,
+ "num_input_tokens_seen": 227701672,
+ "step": 5200,
+ "train_runtime": 40265.2495,
+ "train_tokens_per_second": 5655.042
+ },
+ {
+ "epoch": 0.3504578507945058,
+ "grad_norm": 0.8047649086754547,
+ "learning_rate": 9.261597445973106e-06,
+ "loss": 0.40303988456726075,
+ "num_input_tokens_seen": 227930736,
+ "step": 5205,
+ "train_runtime": 40300.5889,
+ "train_tokens_per_second": 5655.767
+ },
+ {
+ "epoch": 0.35079450579046595,
+ "grad_norm": 0.7279256779386029,
+ "learning_rate": 9.260213938354365e-06,
+ "loss": 0.38801677227020265,
+ "num_input_tokens_seen": 228142248,
+ "step": 5210,
+ "train_runtime": 40339.465,
+ "train_tokens_per_second": 5655.56
+ },
+ {
+ "epoch": 0.3511311607864261,
+ "grad_norm": 0.703365188592268,
+ "learning_rate": 9.25882923938038e-06,
+ "loss": 0.40598435401916505,
+ "num_input_tokens_seen": 228355000,
+ "step": 5215,
+ "train_runtime": 40382.8232,
+ "train_tokens_per_second": 5654.756
+ },
+ {
+ "epoch": 0.35146781578238623,
+ "grad_norm": 0.8641270145143162,
+ "learning_rate": 9.257443349438382e-06,
+ "loss": 0.42745413780212405,
+ "num_input_tokens_seen": 228579256,
+ "step": 5220,
+ "train_runtime": 40419.0219,
+ "train_tokens_per_second": 5655.24
+ },
+ {
+ "epoch": 0.35180447077834637,
+ "grad_norm": 0.8147972074351507,
+ "learning_rate": 9.256056268915924e-06,
+ "loss": 0.40657625198364256,
+ "num_input_tokens_seen": 228801840,
+ "step": 5225,
+ "train_runtime": 40461.4517,
+ "train_tokens_per_second": 5654.81
+ },
+ {
+ "epoch": 0.3521411257743065,
+ "grad_norm": 0.7053549338628146,
+ "learning_rate": 9.254667998200906e-06,
+ "loss": 0.4108288764953613,
+ "num_input_tokens_seen": 229023432,
+ "step": 5230,
+ "train_runtime": 40502.8889,
+ "train_tokens_per_second": 5654.496
+ },
+ {
+ "epoch": 0.35247778077026665,
+ "grad_norm": 0.8352493252466918,
+ "learning_rate": 9.253278537681548e-06,
+ "loss": 0.4314168930053711,
+ "num_input_tokens_seen": 229259792,
+ "step": 5235,
+ "train_runtime": 40538.046,
+ "train_tokens_per_second": 5655.423
+ },
+ {
+ "epoch": 0.3528144357662268,
+ "grad_norm": 0.7807713230559138,
+ "learning_rate": 9.251887887746407e-06,
+ "loss": 0.3931559085845947,
+ "num_input_tokens_seen": 229485656,
+ "step": 5240,
+ "train_runtime": 40577.5224,
+ "train_tokens_per_second": 5655.487
+ },
+ {
+ "epoch": 0.3531510907621869,
+ "grad_norm": 0.8636190729399873,
+ "learning_rate": 9.25049604878438e-06,
+ "loss": 0.38311750888824464,
+ "num_input_tokens_seen": 229704840,
+ "step": 5245,
+ "train_runtime": 40615.5331,
+ "train_tokens_per_second": 5655.591
+ },
+ {
+ "epoch": 0.35348774575814707,
+ "grad_norm": 0.8592849372799231,
+ "learning_rate": 9.249103021184684e-06,
+ "loss": 0.3960615634918213,
+ "num_input_tokens_seen": 229920256,
+ "step": 5250,
+ "train_runtime": 40653.9984,
+ "train_tokens_per_second": 5655.539
+ },
+ {
+ "epoch": 0.3538244007541072,
+ "grad_norm": 0.7502281989758054,
+ "learning_rate": 9.24770880533688e-06,
+ "loss": 0.4205930709838867,
+ "num_input_tokens_seen": 230150768,
+ "step": 5255,
+ "train_runtime": 40691.9738,
+ "train_tokens_per_second": 5655.925
+ },
+ {
+ "epoch": 0.35416105575006734,
+ "grad_norm": 0.9736374450794365,
+ "learning_rate": 9.24631340163085e-06,
+ "loss": 0.41542859077453614,
+ "num_input_tokens_seen": 230338192,
+ "step": 5260,
+ "train_runtime": 40727.6217,
+ "train_tokens_per_second": 5655.577
+ },
+ {
+ "epoch": 0.3544977107460275,
+ "grad_norm": 0.7629428550659803,
+ "learning_rate": 9.244916810456822e-06,
+ "loss": 0.3574668884277344,
+ "num_input_tokens_seen": 230568472,
+ "step": 5265,
+ "train_runtime": 40758.3906,
+ "train_tokens_per_second": 5656.957
+ },
+ {
+ "epoch": 0.3548343657419876,
+ "grad_norm": 0.7107248914856642,
+ "learning_rate": 9.24351903220534e-06,
+ "loss": 0.40234909057617185,
+ "num_input_tokens_seen": 230793504,
+ "step": 5270,
+ "train_runtime": 40798.996,
+ "train_tokens_per_second": 5656.843
+ },
+ {
+ "epoch": 0.35517102073794776,
+ "grad_norm": 0.8468122239159248,
+ "learning_rate": 9.242120067267295e-06,
+ "loss": 0.3589898347854614,
+ "num_input_tokens_seen": 231011256,
+ "step": 5275,
+ "train_runtime": 40837.4422,
+ "train_tokens_per_second": 5656.849
+ },
+ {
+ "epoch": 0.3555076757339079,
+ "grad_norm": 0.8384350564515061,
+ "learning_rate": 9.2407199160339e-06,
+ "loss": 0.3916470050811768,
+ "num_input_tokens_seen": 231224304,
+ "step": 5280,
+ "train_runtime": 40879.0573,
+ "train_tokens_per_second": 5656.302
+ },
+ {
+ "epoch": 0.35584433072986804,
+ "grad_norm": 0.7878128731634584,
+ "learning_rate": 9.239318578896701e-06,
+ "loss": 0.3869880199432373,
+ "num_input_tokens_seen": 231440576,
+ "step": 5285,
+ "train_runtime": 40917.8128,
+ "train_tokens_per_second": 5656.23
+ },
+ {
+ "epoch": 0.3561809857258282,
+ "grad_norm": 0.9404923069120379,
+ "learning_rate": 9.237916056247583e-06,
+ "loss": 0.40935673713684084,
+ "num_input_tokens_seen": 231653280,
+ "step": 5290,
+ "train_runtime": 40955.4085,
+ "train_tokens_per_second": 5656.232
+ },
+ {
+ "epoch": 0.3565176407217883,
+ "grad_norm": 0.790253055764951,
+ "learning_rate": 9.23651234847875e-06,
+ "loss": 0.4153427600860596,
+ "num_input_tokens_seen": 231867296,
+ "step": 5295,
+ "train_runtime": 40993.5671,
+ "train_tokens_per_second": 5656.187
+ },
+ {
+ "epoch": 0.35685429571774846,
+ "grad_norm": 0.8209369098367891,
+ "learning_rate": 9.23510745598275e-06,
+ "loss": 0.3877981185913086,
+ "num_input_tokens_seen": 232073536,
+ "step": 5300,
+ "train_runtime": 41032.3162,
+ "train_tokens_per_second": 5655.872
+ },
+ {
+ "epoch": 0.3571909507137086,
+ "grad_norm": 0.6660402351658175,
+ "learning_rate": 9.233701379152456e-06,
+ "loss": 0.3829216957092285,
+ "num_input_tokens_seen": 232295728,
+ "step": 5305,
+ "train_runtime": 41070.6587,
+ "train_tokens_per_second": 5656.002
+ },
+ {
+ "epoch": 0.35752760570966874,
+ "grad_norm": 0.7635535443595722,
+ "learning_rate": 9.232294118381069e-06,
+ "loss": 0.41249933242797854,
+ "num_input_tokens_seen": 232520456,
+ "step": 5310,
+ "train_runtime": 41107.4158,
+ "train_tokens_per_second": 5656.411
+ },
+ {
+ "epoch": 0.3578642607056289,
+ "grad_norm": 0.7665309192538949,
+ "learning_rate": 9.230885674062129e-06,
+ "loss": 0.3704015493392944,
+ "num_input_tokens_seen": 232740384,
+ "step": 5315,
+ "train_runtime": 41146.5227,
+ "train_tokens_per_second": 5656.38
+ },
+ {
+ "epoch": 0.358200915701589,
+ "grad_norm": 0.805325429526549,
+ "learning_rate": 9.2294760465895e-06,
+ "loss": 0.4049508094787598,
+ "num_input_tokens_seen": 232969144,
+ "step": 5320,
+ "train_runtime": 41180.0542,
+ "train_tokens_per_second": 5657.33
+ },
+ {
+ "epoch": 0.35853757069754916,
+ "grad_norm": 0.8309484529457033,
+ "learning_rate": 9.228065236357383e-06,
+ "loss": 0.39788720607757566,
+ "num_input_tokens_seen": 233196464,
+ "step": 5325,
+ "train_runtime": 41217.3194,
+ "train_tokens_per_second": 5657.73
+ },
+ {
+ "epoch": 0.3588742256935093,
+ "grad_norm": 0.7008865285408351,
+ "learning_rate": 9.226653243760304e-06,
+ "loss": 0.3763267993927002,
+ "num_input_tokens_seen": 233413024,
+ "step": 5330,
+ "train_runtime": 41251.2257,
+ "train_tokens_per_second": 5658.329
+ },
+ {
+ "epoch": 0.35921088068946944,
+ "grad_norm": 0.7229398353747315,
+ "learning_rate": 9.225240069193124e-06,
+ "loss": 0.4120461463928223,
+ "num_input_tokens_seen": 233629584,
+ "step": 5335,
+ "train_runtime": 41295.9805,
+ "train_tokens_per_second": 5657.441
+ },
+ {
+ "epoch": 0.3595475356854296,
+ "grad_norm": 0.854235630422482,
+ "learning_rate": 9.223825713051034e-06,
+ "loss": 0.4018904685974121,
+ "num_input_tokens_seen": 233844536,
+ "step": 5340,
+ "train_runtime": 41326.7978,
+ "train_tokens_per_second": 5658.424
+ },
+ {
+ "epoch": 0.3598841906813897,
+ "grad_norm": 0.8244749095839433,
+ "learning_rate": 9.222410175729549e-06,
+ "loss": 0.38492045402526853,
+ "num_input_tokens_seen": 234063096,
+ "step": 5345,
+ "train_runtime": 41367.4954,
+ "train_tokens_per_second": 5658.14
+ },
+ {
+ "epoch": 0.36022084567734985,
+ "grad_norm": 0.7424522527415252,
+ "learning_rate": 9.220993457624523e-06,
+ "loss": 0.3723012924194336,
+ "num_input_tokens_seen": 234279256,
+ "step": 5350,
+ "train_runtime": 41414.7783,
+ "train_tokens_per_second": 5656.9
+ },
+ {
+ "epoch": 0.36055750067331,
+ "grad_norm": 0.8296218729091308,
+ "learning_rate": 9.219575559132136e-06,
+ "loss": 0.4064267635345459,
+ "num_input_tokens_seen": 234496160,
+ "step": 5355,
+ "train_runtime": 41454.7573,
+ "train_tokens_per_second": 5656.677
+ },
+ {
+ "epoch": 0.36089415566927013,
+ "grad_norm": 0.8270954179905022,
+ "learning_rate": 9.2181564806489e-06,
+ "loss": 0.3711550235748291,
+ "num_input_tokens_seen": 234705872,
+ "step": 5360,
+ "train_runtime": 41488.3956,
+ "train_tokens_per_second": 5657.145
+ },
+ {
+ "epoch": 0.3612308106652303,
+ "grad_norm": 0.7010804881835582,
+ "learning_rate": 9.216736222571655e-06,
+ "loss": 0.36968867778778075,
+ "num_input_tokens_seen": 234928056,
+ "step": 5365,
+ "train_runtime": 41528.5306,
+ "train_tokens_per_second": 5657.028
+ },
+ {
+ "epoch": 0.3615674656611904,
+ "grad_norm": 0.7909884318309988,
+ "learning_rate": 9.215314785297572e-06,
+ "loss": 0.4384300708770752,
+ "num_input_tokens_seen": 235127224,
+ "step": 5370,
+ "train_runtime": 41566.5449,
+ "train_tokens_per_second": 5656.646
+ },
+ {
+ "epoch": 0.36190412065715055,
+ "grad_norm": 0.7730294782111428,
+ "learning_rate": 9.21389216922415e-06,
+ "loss": 0.3896491527557373,
+ "num_input_tokens_seen": 235350232,
+ "step": 5375,
+ "train_runtime": 41604.4719,
+ "train_tokens_per_second": 5656.849
+ },
+ {
+ "epoch": 0.3622407756531107,
+ "grad_norm": 1.5247558034207205,
+ "learning_rate": 9.212468374749222e-06,
+ "loss": 0.4335320949554443,
+ "num_input_tokens_seen": 235570704,
+ "step": 5380,
+ "train_runtime": 41642.8804,
+ "train_tokens_per_second": 5656.926
+ },
+ {
+ "epoch": 0.36257743064907083,
+ "grad_norm": 0.7761987245970543,
+ "learning_rate": 9.211043402270942e-06,
+ "loss": 0.39116473197937013,
+ "num_input_tokens_seen": 235785280,
+ "step": 5385,
+ "train_runtime": 41679.7538,
+ "train_tokens_per_second": 5657.07
+ },
+ {
+ "epoch": 0.36291408564503097,
+ "grad_norm": 0.7417904936648765,
+ "learning_rate": 9.209617252187806e-06,
+ "loss": 0.3792844533920288,
+ "num_input_tokens_seen": 235999696,
+ "step": 5390,
+ "train_runtime": 41717.5624,
+ "train_tokens_per_second": 5657.083
+ },
+ {
+ "epoch": 0.3632507406409911,
+ "grad_norm": 0.8168071916778651,
+ "learning_rate": 9.208189924898627e-06,
+ "loss": 0.3464975833892822,
+ "num_input_tokens_seen": 236210480,
+ "step": 5395,
+ "train_runtime": 41760.5338,
+ "train_tokens_per_second": 5656.309
+ },
+ {
+ "epoch": 0.36358739563695125,
+ "grad_norm": 0.7857550769816507,
+ "learning_rate": 9.206761420802554e-06,
+ "loss": 0.4210227966308594,
+ "num_input_tokens_seen": 236432304,
+ "step": 5400,
+ "train_runtime": 41802.7339,
+ "train_tokens_per_second": 5655.905
+ },
+ {
+ "epoch": 0.3639240506329114,
+ "grad_norm": 0.724781136378123,
+ "learning_rate": 9.205331740299065e-06,
+ "loss": 0.3803566932678223,
+ "num_input_tokens_seen": 236642552,
+ "step": 5405,
+ "train_runtime": 41844.665,
+ "train_tokens_per_second": 5655.262
+ },
+ {
+ "epoch": 0.36426070562887153,
+ "grad_norm": 0.7382233662482024,
+ "learning_rate": 9.203900883787967e-06,
+ "loss": 0.4222862243652344,
+ "num_input_tokens_seen": 236866080,
+ "step": 5410,
+ "train_runtime": 41880.1302,
+ "train_tokens_per_second": 5655.811
+ },
+ {
+ "epoch": 0.36459736062483167,
+ "grad_norm": 0.7707623946075276,
+ "learning_rate": 9.202468851669388e-06,
+ "loss": 0.42064590454101564,
+ "num_input_tokens_seen": 237084656,
+ "step": 5415,
+ "train_runtime": 41922.4674,
+ "train_tokens_per_second": 5655.313
+ },
+ {
+ "epoch": 0.3649340156207918,
+ "grad_norm": 0.7505014032426987,
+ "learning_rate": 9.201035644343797e-06,
+ "loss": 0.40491809844970705,
+ "num_input_tokens_seen": 237299088,
+ "step": 5420,
+ "train_runtime": 41963.0407,
+ "train_tokens_per_second": 5654.955
+ },
+ {
+ "epoch": 0.36527067061675195,
+ "grad_norm": 0.8419562349359976,
+ "learning_rate": 9.199601262211985e-06,
+ "loss": 0.3938578128814697,
+ "num_input_tokens_seen": 237505024,
+ "step": 5425,
+ "train_runtime": 41999.6624,
+ "train_tokens_per_second": 5654.927
+ },
+ {
+ "epoch": 0.3656073256127121,
+ "grad_norm": 0.7844116167185397,
+ "learning_rate": 9.19816570567507e-06,
+ "loss": 0.4014750480651855,
+ "num_input_tokens_seen": 237729312,
+ "step": 5430,
+ "train_runtime": 42038.7844,
+ "train_tokens_per_second": 5655.0
+ },
+ {
+ "epoch": 0.3659439806086722,
+ "grad_norm": 0.8709777369310681,
+ "learning_rate": 9.196728975134503e-06,
+ "loss": 0.41324663162231445,
+ "num_input_tokens_seen": 237937544,
+ "step": 5435,
+ "train_runtime": 42082.2656,
+ "train_tokens_per_second": 5654.105
+ },
+ {
+ "epoch": 0.36628063560463237,
+ "grad_norm": 0.943067660878324,
+ "learning_rate": 9.195291070992061e-06,
+ "loss": 0.39972093105316164,
+ "num_input_tokens_seen": 238144328,
+ "step": 5440,
+ "train_runtime": 42119.7987,
+ "train_tokens_per_second": 5653.976
+ },
+ {
+ "epoch": 0.3666172906005925,
+ "grad_norm": 0.7866408242613544,
+ "learning_rate": 9.193851993649849e-06,
+ "loss": 0.36598644256591795,
+ "num_input_tokens_seen": 238371752,
+ "step": 5445,
+ "train_runtime": 42157.0955,
+ "train_tokens_per_second": 5654.368
+ },
+ {
+ "epoch": 0.36695394559655264,
+ "grad_norm": 0.7572508667451172,
+ "learning_rate": 9.1924117435103e-06,
+ "loss": 0.3939794063568115,
+ "num_input_tokens_seen": 238586128,
+ "step": 5450,
+ "train_runtime": 42198.1981,
+ "train_tokens_per_second": 5653.941
+ },
+ {
+ "epoch": 0.3672906005925128,
+ "grad_norm": 0.885004223430614,
+ "learning_rate": 9.190970320976176e-06,
+ "loss": 0.3719717264175415,
+ "num_input_tokens_seen": 238811344,
+ "step": 5455,
+ "train_runtime": 42232.691,
+ "train_tokens_per_second": 5654.656
+ },
+ {
+ "epoch": 0.3676272555884729,
+ "grad_norm": 0.8940362227103387,
+ "learning_rate": 9.189527726450565e-06,
+ "loss": 0.4350133895874023,
+ "num_input_tokens_seen": 239003208,
+ "step": 5460,
+ "train_runtime": 42266.5178,
+ "train_tokens_per_second": 5654.67
+ },
+ {
+ "epoch": 0.36796391058443306,
+ "grad_norm": 0.6837019733666716,
+ "learning_rate": 9.188083960336885e-06,
+ "loss": 0.3788605213165283,
+ "num_input_tokens_seen": 239225040,
+ "step": 5465,
+ "train_runtime": 42305.7631,
+ "train_tokens_per_second": 5654.668
+ },
+ {
+ "epoch": 0.3683005655803932,
+ "grad_norm": 0.7278725929539929,
+ "learning_rate": 9.186639023038879e-06,
+ "loss": 0.4101285934448242,
+ "num_input_tokens_seen": 239445688,
+ "step": 5470,
+ "train_runtime": 42345.9353,
+ "train_tokens_per_second": 5654.514
+ },
+ {
+ "epoch": 0.36863722057635334,
+ "grad_norm": 0.8240745816268156,
+ "learning_rate": 9.185192914960622e-06,
+ "loss": 0.43002777099609374,
+ "num_input_tokens_seen": 239660856,
+ "step": 5475,
+ "train_runtime": 42383.8589,
+ "train_tokens_per_second": 5654.531
+ },
+ {
+ "epoch": 0.3689738755723135,
+ "grad_norm": 0.7561111927294512,
+ "learning_rate": 9.183745636506511e-06,
+ "loss": 0.400468635559082,
+ "num_input_tokens_seen": 239863160,
+ "step": 5480,
+ "train_runtime": 42415.6347,
+ "train_tokens_per_second": 5655.065
+ },
+ {
+ "epoch": 0.3693105305682736,
+ "grad_norm": 0.6657976994723397,
+ "learning_rate": 9.182297188081274e-06,
+ "loss": 0.39147305488586426,
+ "num_input_tokens_seen": 240088504,
+ "step": 5485,
+ "train_runtime": 42445.5983,
+ "train_tokens_per_second": 5656.382
+ },
+ {
+ "epoch": 0.36964718556423376,
+ "grad_norm": 0.7438674827381799,
+ "learning_rate": 9.180847570089964e-06,
+ "loss": 0.38025264739990233,
+ "num_input_tokens_seen": 240285080,
+ "step": 5490,
+ "train_runtime": 42487.4443,
+ "train_tokens_per_second": 5655.437
+ },
+ {
+ "epoch": 0.3699838405601939,
+ "grad_norm": 0.7687571314606845,
+ "learning_rate": 9.179396782937965e-06,
+ "loss": 0.4285304546356201,
+ "num_input_tokens_seen": 240518464,
+ "step": 5495,
+ "train_runtime": 42522.1167,
+ "train_tokens_per_second": 5656.314
+ },
+ {
+ "epoch": 0.37032049555615404,
+ "grad_norm": 0.8258932109796826,
+ "learning_rate": 9.17794482703098e-06,
+ "loss": 0.41852726936340334,
+ "num_input_tokens_seen": 240737112,
+ "step": 5500,
+ "train_runtime": 42561.0578,
+ "train_tokens_per_second": 5656.277
+ },
+ {
+ "epoch": 0.3706571505521142,
+ "grad_norm": 0.7712097331858169,
+ "learning_rate": 9.176491702775049e-06,
+ "loss": 0.4076284408569336,
+ "num_input_tokens_seen": 240940632,
+ "step": 5505,
+ "train_runtime": 42605.8265,
+ "train_tokens_per_second": 5655.11
+ },
+ {
+ "epoch": 0.3709938055480743,
+ "grad_norm": 0.7293507629162065,
+ "learning_rate": 9.17503741057653e-06,
+ "loss": 0.42336077690124513,
+ "num_input_tokens_seen": 241163512,
+ "step": 5510,
+ "train_runtime": 42641.1976,
+ "train_tokens_per_second": 5655.646
+ },
+ {
+ "epoch": 0.37133046054403446,
+ "grad_norm": 0.7873802897373652,
+ "learning_rate": 9.173581950842111e-06,
+ "loss": 0.3858926773071289,
+ "num_input_tokens_seen": 241380280,
+ "step": 5515,
+ "train_runtime": 42677.1161,
+ "train_tokens_per_second": 5655.965
+ },
+ {
+ "epoch": 0.3716671155399946,
+ "grad_norm": 0.833898462602849,
+ "learning_rate": 9.172125323978811e-06,
+ "loss": 0.39653089046478274,
+ "num_input_tokens_seen": 241601112,
+ "step": 5520,
+ "train_runtime": 42713.21,
+ "train_tokens_per_second": 5656.356
+ },
+ {
+ "epoch": 0.37200377053595474,
+ "grad_norm": 0.7933645694043111,
+ "learning_rate": 9.170667530393966e-06,
+ "loss": 0.4098936080932617,
+ "num_input_tokens_seen": 241828432,
+ "step": 5525,
+ "train_runtime": 42747.8553,
+ "train_tokens_per_second": 5657.089
+ },
+ {
+ "epoch": 0.3723404255319149,
+ "grad_norm": 0.724581396473804,
+ "learning_rate": 9.169208570495247e-06,
+ "loss": 0.3672953605651855,
+ "num_input_tokens_seen": 242051344,
+ "step": 5530,
+ "train_runtime": 42787.8177,
+ "train_tokens_per_second": 5657.015
+ },
+ {
+ "epoch": 0.372677080527875,
+ "grad_norm": 0.720329801926469,
+ "learning_rate": 9.167748444690646e-06,
+ "loss": 0.3897575855255127,
+ "num_input_tokens_seen": 242250584,
+ "step": 5535,
+ "train_runtime": 42826.4052,
+ "train_tokens_per_second": 5656.571
+ },
+ {
+ "epoch": 0.37301373552383515,
+ "grad_norm": 0.7934647008536321,
+ "learning_rate": 9.16628715338848e-06,
+ "loss": 0.3871308326721191,
+ "num_input_tokens_seen": 242459936,
+ "step": 5540,
+ "train_runtime": 42859.1014,
+ "train_tokens_per_second": 5657.14
+ },
+ {
+ "epoch": 0.3733503905197953,
+ "grad_norm": 0.7931279904448103,
+ "learning_rate": 9.164824696997398e-06,
+ "loss": 0.37985944747924805,
+ "num_input_tokens_seen": 242665952,
+ "step": 5545,
+ "train_runtime": 42896.1293,
+ "train_tokens_per_second": 5657.059
+ },
+ {
+ "epoch": 0.37368704551575543,
+ "grad_norm": 0.7548635507515563,
+ "learning_rate": 9.163361075926371e-06,
+ "loss": 0.4154071807861328,
+ "num_input_tokens_seen": 242885960,
+ "step": 5550,
+ "train_runtime": 42932.7348,
+ "train_tokens_per_second": 5657.361
+ },
+ {
+ "epoch": 0.3740237005117156,
+ "grad_norm": 0.824128032145257,
+ "learning_rate": 9.161896290584695e-06,
+ "loss": 0.37793383598327634,
+ "num_input_tokens_seen": 243108432,
+ "step": 5555,
+ "train_runtime": 42974.4088,
+ "train_tokens_per_second": 5657.051
+ },
+ {
+ "epoch": 0.3743603555076757,
+ "grad_norm": 0.755788883617862,
+ "learning_rate": 9.160430341381991e-06,
+ "loss": 0.3929650545120239,
+ "num_input_tokens_seen": 243321200,
+ "step": 5560,
+ "train_runtime": 43020.7871,
+ "train_tokens_per_second": 5655.898
+ },
+ {
+ "epoch": 0.37469701050363585,
+ "grad_norm": 0.8046945349779168,
+ "learning_rate": 9.15896322872821e-06,
+ "loss": 0.39426612854003906,
+ "num_input_tokens_seen": 243541160,
+ "step": 5565,
+ "train_runtime": 43065.6274,
+ "train_tokens_per_second": 5655.117
+ },
+ {
+ "epoch": 0.375033665499596,
+ "grad_norm": 0.8408220299669169,
+ "learning_rate": 9.157494953033625e-06,
+ "loss": 0.4226834297180176,
+ "num_input_tokens_seen": 243766296,
+ "step": 5570,
+ "train_runtime": 43100.4977,
+ "train_tokens_per_second": 5655.765
+ },
+ {
+ "epoch": 0.37537032049555613,
+ "grad_norm": 0.7701423462069609,
+ "learning_rate": 9.156025514708832e-06,
+ "loss": 0.40192432403564454,
+ "num_input_tokens_seen": 244002616,
+ "step": 5575,
+ "train_runtime": 43146.2968,
+ "train_tokens_per_second": 5655.239
+ },
+ {
+ "epoch": 0.37570697549151627,
+ "grad_norm": 0.7773289349591549,
+ "learning_rate": 9.154554914164757e-06,
+ "loss": 0.38442695140838623,
+ "num_input_tokens_seen": 244237240,
+ "step": 5580,
+ "train_runtime": 43188.3958,
+ "train_tokens_per_second": 5655.159
+ },
+ {
+ "epoch": 0.3760436304874764,
+ "grad_norm": 0.6847338403952183,
+ "learning_rate": 9.153083151812649e-06,
+ "loss": 0.377176308631897,
+ "num_input_tokens_seen": 244463576,
+ "step": 5585,
+ "train_runtime": 43225.3626,
+ "train_tokens_per_second": 5655.559
+ },
+ {
+ "epoch": 0.37638028548343655,
+ "grad_norm": 0.780657947894897,
+ "learning_rate": 9.151610228064078e-06,
+ "loss": 0.40382232666015627,
+ "num_input_tokens_seen": 244685816,
+ "step": 5590,
+ "train_runtime": 43264.9573,
+ "train_tokens_per_second": 5655.52
+ },
+ {
+ "epoch": 0.3767169404793967,
+ "grad_norm": 0.7617598851673253,
+ "learning_rate": 9.150136143330949e-06,
+ "loss": 0.42247867584228516,
+ "num_input_tokens_seen": 244901992,
+ "step": 5595,
+ "train_runtime": 43305.9839,
+ "train_tokens_per_second": 5655.154
+ },
+ {
+ "epoch": 0.37705359547535683,
+ "grad_norm": 0.6965545191686713,
+ "learning_rate": 9.14866089802548e-06,
+ "loss": 0.3760505676269531,
+ "num_input_tokens_seen": 245127560,
+ "step": 5600,
+ "train_runtime": 43348.427,
+ "train_tokens_per_second": 5654.82
+ },
+ {
+ "epoch": 0.37739025047131697,
+ "grad_norm": 0.8578077401564244,
+ "learning_rate": 9.147184492560219e-06,
+ "loss": 0.3698588848114014,
+ "num_input_tokens_seen": 245339952,
+ "step": 5605,
+ "train_runtime": 43388.8794,
+ "train_tokens_per_second": 5654.443
+ },
+ {
+ "epoch": 0.3777269054672771,
+ "grad_norm": 1.0507318076347205,
+ "learning_rate": 9.14570692734804e-06,
+ "loss": 0.3787434816360474,
+ "num_input_tokens_seen": 245565672,
+ "step": 5610,
+ "train_runtime": 43425.1962,
+ "train_tokens_per_second": 5654.912
+ },
+ {
+ "epoch": 0.37806356046323725,
+ "grad_norm": 0.8234417174719115,
+ "learning_rate": 9.144228202802137e-06,
+ "loss": 0.41345486640930174,
+ "num_input_tokens_seen": 245793416,
+ "step": 5615,
+ "train_runtime": 43457.0254,
+ "train_tokens_per_second": 5656.011
+ },
+ {
+ "epoch": 0.37840021545919744,
+ "grad_norm": 0.8147806646148233,
+ "learning_rate": 9.142748319336033e-06,
+ "loss": 0.39602549076080323,
+ "num_input_tokens_seen": 246014600,
+ "step": 5620,
+ "train_runtime": 43495.3144,
+ "train_tokens_per_second": 5656.117
+ },
+ {
+ "epoch": 0.3787368704551576,
+ "grad_norm": 0.8294007935657565,
+ "learning_rate": 9.141267277363573e-06,
+ "loss": 0.3865189552307129,
+ "num_input_tokens_seen": 246209040,
+ "step": 5625,
+ "train_runtime": 43530.2862,
+ "train_tokens_per_second": 5656.04
+ },
+ {
+ "epoch": 0.3790735254511177,
+ "grad_norm": 0.8057103649691175,
+ "learning_rate": 9.139785077298921e-06,
+ "loss": 0.3669655084609985,
+ "num_input_tokens_seen": 246435304,
+ "step": 5630,
+ "train_runtime": 43570.2665,
+ "train_tokens_per_second": 5656.043
+ },
+ {
+ "epoch": 0.37941018044707786,
+ "grad_norm": 0.7570731119383132,
+ "learning_rate": 9.138301719556577e-06,
+ "loss": 0.3826982736587524,
+ "num_input_tokens_seen": 246652712,
+ "step": 5635,
+ "train_runtime": 43609.2254,
+ "train_tokens_per_second": 5655.976
+ },
+ {
+ "epoch": 0.379746835443038,
+ "grad_norm": 0.7474769926348979,
+ "learning_rate": 9.136817204551352e-06,
+ "loss": 0.386492919921875,
+ "num_input_tokens_seen": 246882568,
+ "step": 5640,
+ "train_runtime": 43646.3585,
+ "train_tokens_per_second": 5656.43
+ },
+ {
+ "epoch": 0.38008349043899814,
+ "grad_norm": 0.7350249063601988,
+ "learning_rate": 9.135331532698385e-06,
+ "loss": 0.3880929470062256,
+ "num_input_tokens_seen": 247094472,
+ "step": 5645,
+ "train_runtime": 43685.4708,
+ "train_tokens_per_second": 5656.216
+ },
+ {
+ "epoch": 0.3804201454349583,
+ "grad_norm": 0.7440003142918237,
+ "learning_rate": 9.133844704413144e-06,
+ "loss": 0.37475876808166503,
+ "num_input_tokens_seen": 247305056,
+ "step": 5650,
+ "train_runtime": 43724.0377,
+ "train_tokens_per_second": 5656.043
+ },
+ {
+ "epoch": 0.3807568004309184,
+ "grad_norm": 0.829580642455565,
+ "learning_rate": 9.132356720111413e-06,
+ "loss": 0.40270652770996096,
+ "num_input_tokens_seen": 247521856,
+ "step": 5655,
+ "train_runtime": 43766.9087,
+ "train_tokens_per_second": 5655.457
+ },
+ {
+ "epoch": 0.38109345542687856,
+ "grad_norm": 0.7890155830011499,
+ "learning_rate": 9.1308675802093e-06,
+ "loss": 0.3874132394790649,
+ "num_input_tokens_seen": 247735120,
+ "step": 5660,
+ "train_runtime": 43801.888,
+ "train_tokens_per_second": 5655.809
+ },
+ {
+ "epoch": 0.3814301104228387,
+ "grad_norm": 0.7482568242628747,
+ "learning_rate": 9.129377285123241e-06,
+ "loss": 0.3984661102294922,
+ "num_input_tokens_seen": 247949328,
+ "step": 5665,
+ "train_runtime": 43845.4472,
+ "train_tokens_per_second": 5655.076
+ },
+ {
+ "epoch": 0.38176676541879884,
+ "grad_norm": 0.7947458109924397,
+ "learning_rate": 9.127885835269996e-06,
+ "loss": 0.4090867519378662,
+ "num_input_tokens_seen": 248174032,
+ "step": 5670,
+ "train_runtime": 43879.3187,
+ "train_tokens_per_second": 5655.831
+ },
+ {
+ "epoch": 0.382103420414759,
+ "grad_norm": 0.726446478217637,
+ "learning_rate": 9.126393231066636e-06,
+ "loss": 0.3848613977432251,
+ "num_input_tokens_seen": 248384368,
+ "step": 5675,
+ "train_runtime": 43917.9285,
+ "train_tokens_per_second": 5655.649
+ },
+ {
+ "epoch": 0.3824400754107191,
+ "grad_norm": 0.7750970581531168,
+ "learning_rate": 9.124899472930568e-06,
+ "loss": 0.4047283172607422,
+ "num_input_tokens_seen": 248607600,
+ "step": 5680,
+ "train_runtime": 43961.6528,
+ "train_tokens_per_second": 5655.101
+ },
+ {
+ "epoch": 0.38277673040667926,
+ "grad_norm": 1.9536806017675672,
+ "learning_rate": 9.123404561279517e-06,
+ "loss": 0.41053128242492676,
+ "num_input_tokens_seen": 248811928,
+ "step": 5685,
+ "train_runtime": 44006.8302,
+ "train_tokens_per_second": 5653.939
+ },
+ {
+ "epoch": 0.3831133854026394,
+ "grad_norm": 0.7123153967912388,
+ "learning_rate": 9.121908496531527e-06,
+ "loss": 0.3802008867263794,
+ "num_input_tokens_seen": 249035424,
+ "step": 5690,
+ "train_runtime": 44040.1743,
+ "train_tokens_per_second": 5654.733
+ },
+ {
+ "epoch": 0.38345004039859953,
+ "grad_norm": 0.7190654994858774,
+ "learning_rate": 9.120411279104971e-06,
+ "loss": 0.3904706954956055,
+ "num_input_tokens_seen": 249254456,
+ "step": 5695,
+ "train_runtime": 44083.466,
+ "train_tokens_per_second": 5654.148
+ },
+ {
+ "epoch": 0.3837866953945597,
+ "grad_norm": 0.9187897813904784,
+ "learning_rate": 9.118912909418538e-06,
+ "loss": 0.39504170417785645,
+ "num_input_tokens_seen": 249477272,
+ "step": 5700,
+ "train_runtime": 44127.086,
+ "train_tokens_per_second": 5653.609
+ },
+ {
+ "epoch": 0.3841233503905198,
+ "grad_norm": 0.7651587963325609,
+ "learning_rate": 9.117413387891246e-06,
+ "loss": 0.4095571041107178,
+ "num_input_tokens_seen": 249693520,
+ "step": 5705,
+ "train_runtime": 44170.6884,
+ "train_tokens_per_second": 5652.923
+ },
+ {
+ "epoch": 0.38446000538647995,
+ "grad_norm": 0.7879046146044965,
+ "learning_rate": 9.11591271494243e-06,
+ "loss": 0.39027183055877684,
+ "num_input_tokens_seen": 249909984,
+ "step": 5710,
+ "train_runtime": 44209.3015,
+ "train_tokens_per_second": 5652.882
+ },
+ {
+ "epoch": 0.3847966603824401,
+ "grad_norm": 0.7861690179999627,
+ "learning_rate": 9.114410890991748e-06,
+ "loss": 0.40212602615356446,
+ "num_input_tokens_seen": 250124664,
+ "step": 5715,
+ "train_runtime": 44250.9806,
+ "train_tokens_per_second": 5652.41
+ },
+ {
+ "epoch": 0.38513331537840023,
+ "grad_norm": 0.7798159280525913,
+ "learning_rate": 9.112907916459177e-06,
+ "loss": 0.37251665592193606,
+ "num_input_tokens_seen": 250317264,
+ "step": 5720,
+ "train_runtime": 44293.6615,
+ "train_tokens_per_second": 5651.311
+ },
+ {
+ "epoch": 0.38546997037436037,
+ "grad_norm": 0.7886435715633883,
+ "learning_rate": 9.111403791765025e-06,
+ "loss": 0.38798794746398924,
+ "num_input_tokens_seen": 250544576,
+ "step": 5725,
+ "train_runtime": 44329.9756,
+ "train_tokens_per_second": 5651.809
+ },
+ {
+ "epoch": 0.3858066253703205,
+ "grad_norm": 0.7210916045331931,
+ "learning_rate": 9.109898517329914e-06,
+ "loss": 0.3711621046066284,
+ "num_input_tokens_seen": 250774080,
+ "step": 5730,
+ "train_runtime": 44362.0555,
+ "train_tokens_per_second": 5652.896
+ },
+ {
+ "epoch": 0.38614328036628065,
+ "grad_norm": 0.757278528605975,
+ "learning_rate": 9.108392093574785e-06,
+ "loss": 0.3934319972991943,
+ "num_input_tokens_seen": 250992112,
+ "step": 5735,
+ "train_runtime": 44396.0233,
+ "train_tokens_per_second": 5653.482
+ },
+ {
+ "epoch": 0.3864799353622408,
+ "grad_norm": 0.7725316475525575,
+ "learning_rate": 9.10688452092091e-06,
+ "loss": 0.36261420249938964,
+ "num_input_tokens_seen": 251210912,
+ "step": 5740,
+ "train_runtime": 44426.4194,
+ "train_tokens_per_second": 5654.539
+ },
+ {
+ "epoch": 0.38681659035820093,
+ "grad_norm": 0.8668597404685281,
+ "learning_rate": 9.105375799789874e-06,
+ "loss": 0.39951817989349364,
+ "num_input_tokens_seen": 251426584,
+ "step": 5745,
+ "train_runtime": 44475.8422,
+ "train_tokens_per_second": 5653.105
+ },
+ {
+ "epoch": 0.38715324535416107,
+ "grad_norm": 0.8168129882131703,
+ "learning_rate": 9.103865930603586e-06,
+ "loss": 0.3850820064544678,
+ "num_input_tokens_seen": 251633720,
+ "step": 5750,
+ "train_runtime": 44516.691,
+ "train_tokens_per_second": 5652.57
+ },
+ {
+ "epoch": 0.3874899003501212,
+ "grad_norm": 0.7659015613854286,
+ "learning_rate": 9.102354913784279e-06,
+ "loss": 0.3853774070739746,
+ "num_input_tokens_seen": 251855296,
+ "step": 5755,
+ "train_runtime": 44559.5986,
+ "train_tokens_per_second": 5652.1
+ },
+ {
+ "epoch": 0.38782655534608135,
+ "grad_norm": 0.7966891158323898,
+ "learning_rate": 9.100842749754499e-06,
+ "loss": 0.3578346252441406,
+ "num_input_tokens_seen": 252080400,
+ "step": 5760,
+ "train_runtime": 44605.3404,
+ "train_tokens_per_second": 5651.35
+ },
+ {
+ "epoch": 0.3881632103420415,
+ "grad_norm": 1.3427031718401212,
+ "learning_rate": 9.099329438937122e-06,
+ "loss": 0.40854997634887696,
+ "num_input_tokens_seen": 252296312,
+ "step": 5765,
+ "train_runtime": 44649.458,
+ "train_tokens_per_second": 5650.602
+ },
+ {
+ "epoch": 0.3884998653380016,
+ "grad_norm": 0.6230050104908084,
+ "learning_rate": 9.09781498175534e-06,
+ "loss": 0.3587602138519287,
+ "num_input_tokens_seen": 252509656,
+ "step": 5770,
+ "train_runtime": 44685.4879,
+ "train_tokens_per_second": 5650.82
+ },
+ {
+ "epoch": 0.38883652033396177,
+ "grad_norm": 0.6750080511665485,
+ "learning_rate": 9.096299378632666e-06,
+ "loss": 0.37611684799194334,
+ "num_input_tokens_seen": 252740072,
+ "step": 5775,
+ "train_runtime": 44727.054,
+ "train_tokens_per_second": 5650.72
+ },
+ {
+ "epoch": 0.3891731753299219,
+ "grad_norm": 0.7734243231158888,
+ "learning_rate": 9.09478262999293e-06,
+ "loss": 0.4280585289001465,
+ "num_input_tokens_seen": 252973392,
+ "step": 5780,
+ "train_runtime": 44771.0866,
+ "train_tokens_per_second": 5650.374
+ },
+ {
+ "epoch": 0.38950983032588204,
+ "grad_norm": 0.8173587277273618,
+ "learning_rate": 9.09326473626029e-06,
+ "loss": 0.38141813278198244,
+ "num_input_tokens_seen": 253184320,
+ "step": 5785,
+ "train_runtime": 44809.6538,
+ "train_tokens_per_second": 5650.218
+ },
+ {
+ "epoch": 0.3898464853218422,
+ "grad_norm": 0.7487752402423106,
+ "learning_rate": 9.091745697859218e-06,
+ "loss": 0.41326370239257815,
+ "num_input_tokens_seen": 253411776,
+ "step": 5790,
+ "train_runtime": 44845.5045,
+ "train_tokens_per_second": 5650.773
+ },
+ {
+ "epoch": 0.3901831403178023,
+ "grad_norm": 0.8197406274404395,
+ "learning_rate": 9.090225515214512e-06,
+ "loss": 0.4117950439453125,
+ "num_input_tokens_seen": 253627512,
+ "step": 5795,
+ "train_runtime": 44885.8047,
+ "train_tokens_per_second": 5650.506
+ },
+ {
+ "epoch": 0.39051979531376246,
+ "grad_norm": 0.8857575262398083,
+ "learning_rate": 9.088704188751281e-06,
+ "loss": 0.4129201889038086,
+ "num_input_tokens_seen": 253840400,
+ "step": 5800,
+ "train_runtime": 44924.1283,
+ "train_tokens_per_second": 5650.425
+ },
+ {
+ "epoch": 0.3908564503097226,
+ "grad_norm": 0.85730454063038,
+ "learning_rate": 9.087181718894963e-06,
+ "loss": 0.39529869556427,
+ "num_input_tokens_seen": 254064096,
+ "step": 5805,
+ "train_runtime": 44968.4154,
+ "train_tokens_per_second": 5649.834
+ },
+ {
+ "epoch": 0.39119310530568274,
+ "grad_norm": 0.7118238052198951,
+ "learning_rate": 9.085658106071309e-06,
+ "loss": 0.39313058853149413,
+ "num_input_tokens_seen": 254281632,
+ "step": 5810,
+ "train_runtime": 45005.0615,
+ "train_tokens_per_second": 5650.067
+ },
+ {
+ "epoch": 0.3915297603016429,
+ "grad_norm": 0.7349746949660144,
+ "learning_rate": 9.084133350706394e-06,
+ "loss": 0.3712653398513794,
+ "num_input_tokens_seen": 254493944,
+ "step": 5815,
+ "train_runtime": 45042.1578,
+ "train_tokens_per_second": 5650.128
+ },
+ {
+ "epoch": 0.391866415297603,
+ "grad_norm": 0.7166943198737387,
+ "learning_rate": 9.082607453226613e-06,
+ "loss": 0.38478312492370603,
+ "num_input_tokens_seen": 254715712,
+ "step": 5820,
+ "train_runtime": 45077.3472,
+ "train_tokens_per_second": 5650.637
+ },
+ {
+ "epoch": 0.39220307029356316,
+ "grad_norm": 0.7543487043231831,
+ "learning_rate": 9.081080414058674e-06,
+ "loss": 0.3994553565979004,
+ "num_input_tokens_seen": 254927592,
+ "step": 5825,
+ "train_runtime": 45113.9999,
+ "train_tokens_per_second": 5650.742
+ },
+ {
+ "epoch": 0.3925397252895233,
+ "grad_norm": 0.7896679447300116,
+ "learning_rate": 9.079552233629612e-06,
+ "loss": 0.3710631847381592,
+ "num_input_tokens_seen": 255138464,
+ "step": 5830,
+ "train_runtime": 45152.0089,
+ "train_tokens_per_second": 5650.656
+ },
+ {
+ "epoch": 0.39287638028548344,
+ "grad_norm": 0.815361722369787,
+ "learning_rate": 9.078022912366776e-06,
+ "loss": 0.3914163827896118,
+ "num_input_tokens_seen": 255367048,
+ "step": 5835,
+ "train_runtime": 45192.1775,
+ "train_tokens_per_second": 5650.691
+ },
+ {
+ "epoch": 0.3932130352814436,
+ "grad_norm": 0.8005773109782048,
+ "learning_rate": 9.076492450697835e-06,
+ "loss": 0.40029544830322267,
+ "num_input_tokens_seen": 255598184,
+ "step": 5840,
+ "train_runtime": 45232.9765,
+ "train_tokens_per_second": 5650.704
+ },
+ {
+ "epoch": 0.3935496902774037,
+ "grad_norm": 0.849726633762188,
+ "learning_rate": 9.074960849050782e-06,
+ "loss": 0.3852508068084717,
+ "num_input_tokens_seen": 255819944,
+ "step": 5845,
+ "train_runtime": 45272.8634,
+ "train_tokens_per_second": 5650.624
+ },
+ {
+ "epoch": 0.39388634527336386,
+ "grad_norm": 0.7762826196998844,
+ "learning_rate": 9.07342810785392e-06,
+ "loss": 0.37957100868225097,
+ "num_input_tokens_seen": 256019616,
+ "step": 5850,
+ "train_runtime": 45311.4802,
+ "train_tokens_per_second": 5650.215
+ },
+ {
+ "epoch": 0.394223000269324,
+ "grad_norm": 0.7731972533551551,
+ "learning_rate": 9.071894227535878e-06,
+ "loss": 0.38977823257446287,
+ "num_input_tokens_seen": 256237256,
+ "step": 5855,
+ "train_runtime": 45352.5452,
+ "train_tokens_per_second": 5649.898
+ },
+ {
+ "epoch": 0.39455965526528414,
+ "grad_norm": 0.8423525268519388,
+ "learning_rate": 9.070359208525598e-06,
+ "loss": 0.406312084197998,
+ "num_input_tokens_seen": 256464368,
+ "step": 5860,
+ "train_runtime": 45388.423,
+ "train_tokens_per_second": 5650.436
+ },
+ {
+ "epoch": 0.3948963102612443,
+ "grad_norm": 0.8527879610056384,
+ "learning_rate": 9.068823051252346e-06,
+ "loss": 0.3981436491012573,
+ "num_input_tokens_seen": 256688032,
+ "step": 5865,
+ "train_runtime": 45429.2415,
+ "train_tokens_per_second": 5650.282
+ },
+ {
+ "epoch": 0.3952329652572044,
+ "grad_norm": 0.8385538303650153,
+ "learning_rate": 9.067285756145702e-06,
+ "loss": 0.3921481132507324,
+ "num_input_tokens_seen": 256917840,
+ "step": 5870,
+ "train_runtime": 45462.4848,
+ "train_tokens_per_second": 5651.205
+ },
+ {
+ "epoch": 0.39556962025316456,
+ "grad_norm": 0.8390456887457934,
+ "learning_rate": 9.065747323635565e-06,
+ "loss": 0.3327157020568848,
+ "num_input_tokens_seen": 257101736,
+ "step": 5875,
+ "train_runtime": 45501.4164,
+ "train_tokens_per_second": 5650.412
+ },
+ {
+ "epoch": 0.3959062752491247,
+ "grad_norm": 0.7851013810213687,
+ "learning_rate": 9.064207754152154e-06,
+ "loss": 0.3997276782989502,
+ "num_input_tokens_seen": 257317696,
+ "step": 5880,
+ "train_runtime": 45545.2236,
+ "train_tokens_per_second": 5649.719
+ },
+ {
+ "epoch": 0.39624293024508483,
+ "grad_norm": 0.7365128564544189,
+ "learning_rate": 9.062667048126005e-06,
+ "loss": 0.3864887714385986,
+ "num_input_tokens_seen": 257547392,
+ "step": 5885,
+ "train_runtime": 45589.9837,
+ "train_tokens_per_second": 5649.21
+ },
+ {
+ "epoch": 0.396579585241045,
+ "grad_norm": 0.8935927015959244,
+ "learning_rate": 9.061125205987971e-06,
+ "loss": 0.4157553672790527,
+ "num_input_tokens_seen": 257761408,
+ "step": 5890,
+ "train_runtime": 45630.8437,
+ "train_tokens_per_second": 5648.842
+ },
+ {
+ "epoch": 0.3969162402370051,
+ "grad_norm": 0.7876564044642139,
+ "learning_rate": 9.059582228169222e-06,
+ "loss": 0.37337889671325686,
+ "num_input_tokens_seen": 257963800,
+ "step": 5895,
+ "train_runtime": 45664.7,
+ "train_tokens_per_second": 5649.086
+ },
+ {
+ "epoch": 0.39725289523296525,
+ "grad_norm": 0.7288959690931758,
+ "learning_rate": 9.058038115101248e-06,
+ "loss": 0.4037358283996582,
+ "num_input_tokens_seen": 258195648,
+ "step": 5900,
+ "train_runtime": 45702.5616,
+ "train_tokens_per_second": 5649.479
+ },
+ {
+ "epoch": 0.3975895502289254,
+ "grad_norm": 0.7904683688779008,
+ "learning_rate": 9.056492867215857e-06,
+ "loss": 0.3476010084152222,
+ "num_input_tokens_seen": 258400784,
+ "step": 5905,
+ "train_runtime": 45748.63,
+ "train_tokens_per_second": 5648.274
+ },
+ {
+ "epoch": 0.39792620522488553,
+ "grad_norm": 0.6759275337501953,
+ "learning_rate": 9.05494648494517e-06,
+ "loss": 0.3933530330657959,
+ "num_input_tokens_seen": 258606968,
+ "step": 5910,
+ "train_runtime": 45786.7698,
+ "train_tokens_per_second": 5648.072
+ },
+ {
+ "epoch": 0.39826286022084567,
+ "grad_norm": 0.6825204404349337,
+ "learning_rate": 9.053398968721629e-06,
+ "loss": 0.37357821464538576,
+ "num_input_tokens_seen": 258832840,
+ "step": 5915,
+ "train_runtime": 45827.0425,
+ "train_tokens_per_second": 5648.037
+ },
+ {
+ "epoch": 0.3985995152168058,
+ "grad_norm": 0.7549266235464368,
+ "learning_rate": 9.05185031897799e-06,
+ "loss": 0.4264666080474854,
+ "num_input_tokens_seen": 259049920,
+ "step": 5920,
+ "train_runtime": 45860.4564,
+ "train_tokens_per_second": 5648.656
+ },
+ {
+ "epoch": 0.39893617021276595,
+ "grad_norm": 0.8046151442540065,
+ "learning_rate": 9.050300536147331e-06,
+ "loss": 0.42224712371826173,
+ "num_input_tokens_seen": 259285872,
+ "step": 5925,
+ "train_runtime": 45898.8286,
+ "train_tokens_per_second": 5649.074
+ },
+ {
+ "epoch": 0.3992728252087261,
+ "grad_norm": 0.8013279899857307,
+ "learning_rate": 9.048749620663044e-06,
+ "loss": 0.36397323608398435,
+ "num_input_tokens_seen": 259520536,
+ "step": 5930,
+ "train_runtime": 45935.2903,
+ "train_tokens_per_second": 5649.698
+ },
+ {
+ "epoch": 0.39960948020468623,
+ "grad_norm": 0.7188029194424288,
+ "learning_rate": 9.047197572958834e-06,
+ "loss": 0.37349615097045896,
+ "num_input_tokens_seen": 259742248,
+ "step": 5935,
+ "train_runtime": 45973.4138,
+ "train_tokens_per_second": 5649.836
+ },
+ {
+ "epoch": 0.39994613520064637,
+ "grad_norm": 0.8120677037958034,
+ "learning_rate": 9.04564439346873e-06,
+ "loss": 0.3917673110961914,
+ "num_input_tokens_seen": 259952080,
+ "step": 5940,
+ "train_runtime": 46008.964,
+ "train_tokens_per_second": 5650.031
+ },
+ {
+ "epoch": 0.4002827901966065,
+ "grad_norm": 0.8737926960149425,
+ "learning_rate": 9.044090082627071e-06,
+ "loss": 0.39895708560943605,
+ "num_input_tokens_seen": 260160320,
+ "step": 5945,
+ "train_runtime": 46045.9436,
+ "train_tokens_per_second": 5650.016
+ },
+ {
+ "epoch": 0.40061944519256665,
+ "grad_norm": 0.7302283729471113,
+ "learning_rate": 9.042534640868514e-06,
+ "loss": 0.4021873950958252,
+ "num_input_tokens_seen": 260368008,
+ "step": 5950,
+ "train_runtime": 46081.9071,
+ "train_tokens_per_second": 5650.114
+ },
+ {
+ "epoch": 0.4009561001885268,
+ "grad_norm": 0.749168823407156,
+ "learning_rate": 9.04097806862804e-06,
+ "loss": 0.40214033126831056,
+ "num_input_tokens_seen": 260592824,
+ "step": 5955,
+ "train_runtime": 46117.1368,
+ "train_tokens_per_second": 5650.672
+ },
+ {
+ "epoch": 0.4012927551844869,
+ "grad_norm": 0.723288812952783,
+ "learning_rate": 9.03942036634093e-06,
+ "loss": 0.4055626392364502,
+ "num_input_tokens_seen": 260810288,
+ "step": 5960,
+ "train_runtime": 46156.4921,
+ "train_tokens_per_second": 5650.566
+ },
+ {
+ "epoch": 0.40162941018044707,
+ "grad_norm": 0.7518699289987396,
+ "learning_rate": 9.037861534442797e-06,
+ "loss": 0.3734895706176758,
+ "num_input_tokens_seen": 261038312,
+ "step": 5965,
+ "train_runtime": 46198.2126,
+ "train_tokens_per_second": 5650.399
+ },
+ {
+ "epoch": 0.4019660651764072,
+ "grad_norm": 1.0026827049870985,
+ "learning_rate": 9.036301573369563e-06,
+ "loss": 0.40189208984375,
+ "num_input_tokens_seen": 261266152,
+ "step": 5970,
+ "train_runtime": 46239.4807,
+ "train_tokens_per_second": 5650.283
+ },
+ {
+ "epoch": 0.40230272017236735,
+ "grad_norm": 0.8605436747902855,
+ "learning_rate": 9.034740483557465e-06,
+ "loss": 0.3900024175643921,
+ "num_input_tokens_seen": 261483152,
+ "step": 5975,
+ "train_runtime": 46274.8434,
+ "train_tokens_per_second": 5650.655
+ },
+ {
+ "epoch": 0.4026393751683275,
+ "grad_norm": 0.7213971067668634,
+ "learning_rate": 9.033178265443055e-06,
+ "loss": 0.38559999465942385,
+ "num_input_tokens_seen": 261703416,
+ "step": 5980,
+ "train_runtime": 46306.9931,
+ "train_tokens_per_second": 5651.488
+ },
+ {
+ "epoch": 0.4029760301642876,
+ "grad_norm": 0.7467717729146578,
+ "learning_rate": 9.031614919463206e-06,
+ "loss": 0.3772167682647705,
+ "num_input_tokens_seen": 261926896,
+ "step": 5985,
+ "train_runtime": 46342.877,
+ "train_tokens_per_second": 5651.934
+ },
+ {
+ "epoch": 0.40331268516024776,
+ "grad_norm": 0.8052956550974576,
+ "learning_rate": 9.030050446055099e-06,
+ "loss": 0.3684276342391968,
+ "num_input_tokens_seen": 262143080,
+ "step": 5990,
+ "train_runtime": 46380.2479,
+ "train_tokens_per_second": 5652.041
+ },
+ {
+ "epoch": 0.4036493401562079,
+ "grad_norm": 0.6725874955047775,
+ "learning_rate": 9.028484845656235e-06,
+ "loss": 0.3965585231781006,
+ "num_input_tokens_seen": 262378952,
+ "step": 5995,
+ "train_runtime": 46413.7998,
+ "train_tokens_per_second": 5653.038
+ },
+ {
+ "epoch": 0.40398599515216804,
+ "grad_norm": 0.8065835959924061,
+ "learning_rate": 9.026918118704431e-06,
+ "loss": 0.42124166488647463,
+ "num_input_tokens_seen": 262599696,
+ "step": 6000,
+ "train_runtime": 46454.2053,
+ "train_tokens_per_second": 5652.872
+ },
+ {
+ "epoch": 0.4043226501481282,
+ "grad_norm": 0.6374762319564959,
+ "learning_rate": 9.025350265637816e-06,
+ "loss": 0.3700655221939087,
+ "num_input_tokens_seen": 262830992,
+ "step": 6005,
+ "train_runtime": 46490.7965,
+ "train_tokens_per_second": 5653.398
+ },
+ {
+ "epoch": 0.4046593051440883,
+ "grad_norm": 0.7573194160968504,
+ "learning_rate": 9.023781286894834e-06,
+ "loss": 0.38486456871032715,
+ "num_input_tokens_seen": 263044288,
+ "step": 6010,
+ "train_runtime": 46526.3581,
+ "train_tokens_per_second": 5653.662
+ },
+ {
+ "epoch": 0.40499596014004846,
+ "grad_norm": 0.8562717514351039,
+ "learning_rate": 9.022211182914247e-06,
+ "loss": 0.3843253135681152,
+ "num_input_tokens_seen": 263264800,
+ "step": 6015,
+ "train_runtime": 46564.9415,
+ "train_tokens_per_second": 5653.713
+ },
+ {
+ "epoch": 0.4053326151360086,
+ "grad_norm": 0.7322028797978756,
+ "learning_rate": 9.020639954135128e-06,
+ "loss": 0.3804231405258179,
+ "num_input_tokens_seen": 263476880,
+ "step": 6020,
+ "train_runtime": 46608.1754,
+ "train_tokens_per_second": 5653.019
+ },
+ {
+ "epoch": 0.40566927013196874,
+ "grad_norm": 0.7924756161102612,
+ "learning_rate": 9.019067600996867e-06,
+ "loss": 0.32609758377075193,
+ "num_input_tokens_seen": 263676440,
+ "step": 6025,
+ "train_runtime": 46647.9449,
+ "train_tokens_per_second": 5652.477
+ },
+ {
+ "epoch": 0.4060059251279289,
+ "grad_norm": 0.7406207884260587,
+ "learning_rate": 9.017494123939169e-06,
+ "loss": 0.3878431558609009,
+ "num_input_tokens_seen": 263891832,
+ "step": 6030,
+ "train_runtime": 46686.4579,
+ "train_tokens_per_second": 5652.428
+ },
+ {
+ "epoch": 0.406342580123889,
+ "grad_norm": 0.6859816815928503,
+ "learning_rate": 9.015919523402048e-06,
+ "loss": 0.38746066093444825,
+ "num_input_tokens_seen": 264108024,
+ "step": 6035,
+ "train_runtime": 46723.9959,
+ "train_tokens_per_second": 5652.514
+ },
+ {
+ "epoch": 0.40667923511984916,
+ "grad_norm": 1.0398516753802722,
+ "learning_rate": 9.014343799825838e-06,
+ "loss": 0.3903219699859619,
+ "num_input_tokens_seen": 264311560,
+ "step": 6040,
+ "train_runtime": 46758.4333,
+ "train_tokens_per_second": 5652.704
+ },
+ {
+ "epoch": 0.4070158901158093,
+ "grad_norm": 0.6954397754211817,
+ "learning_rate": 9.012766953651186e-06,
+ "loss": 0.42699203491210935,
+ "num_input_tokens_seen": 264536928,
+ "step": 6045,
+ "train_runtime": 46800.6106,
+ "train_tokens_per_second": 5652.425
+ },
+ {
+ "epoch": 0.40735254511176944,
+ "grad_norm": 0.6848650266078808,
+ "learning_rate": 9.01118898531905e-06,
+ "loss": 0.4040683746337891,
+ "num_input_tokens_seen": 264744560,
+ "step": 6050,
+ "train_runtime": 46841.7035,
+ "train_tokens_per_second": 5651.899
+ },
+ {
+ "epoch": 0.4076892001077296,
+ "grad_norm": 0.7157357093690987,
+ "learning_rate": 9.009609895270708e-06,
+ "loss": 0.3655755043029785,
+ "num_input_tokens_seen": 264973096,
+ "step": 6055,
+ "train_runtime": 46881.3659,
+ "train_tokens_per_second": 5651.992
+ },
+ {
+ "epoch": 0.4080258551036897,
+ "grad_norm": 0.8098322622781278,
+ "learning_rate": 9.008029683947743e-06,
+ "loss": 0.42575840950012206,
+ "num_input_tokens_seen": 265197544,
+ "step": 6060,
+ "train_runtime": 46917.7458,
+ "train_tokens_per_second": 5652.393
+ },
+ {
+ "epoch": 0.40836251009964986,
+ "grad_norm": 0.7104130729995994,
+ "learning_rate": 9.006448351792057e-06,
+ "loss": 0.3974879741668701,
+ "num_input_tokens_seen": 265418944,
+ "step": 6065,
+ "train_runtime": 46962.89,
+ "train_tokens_per_second": 5651.674
+ },
+ {
+ "epoch": 0.40869916509561,
+ "grad_norm": 0.6632525596263023,
+ "learning_rate": 9.004865899245864e-06,
+ "loss": 0.38395133018493655,
+ "num_input_tokens_seen": 265636776,
+ "step": 6070,
+ "train_runtime": 46999.967,
+ "train_tokens_per_second": 5651.85
+ },
+ {
+ "epoch": 0.40903582009157013,
+ "grad_norm": 0.7043912136666165,
+ "learning_rate": 9.003282326751694e-06,
+ "loss": 0.4008152961730957,
+ "num_input_tokens_seen": 265857800,
+ "step": 6075,
+ "train_runtime": 47039.3368,
+ "train_tokens_per_second": 5651.819
+ },
+ {
+ "epoch": 0.4093724750875303,
+ "grad_norm": 0.9655541200247993,
+ "learning_rate": 9.001697634752387e-06,
+ "loss": 0.40933728218078613,
+ "num_input_tokens_seen": 266070448,
+ "step": 6080,
+ "train_runtime": 47074.6096,
+ "train_tokens_per_second": 5652.101
+ },
+ {
+ "epoch": 0.4097091300834904,
+ "grad_norm": 0.683185307022826,
+ "learning_rate": 9.000111823691097e-06,
+ "loss": 0.37639498710632324,
+ "num_input_tokens_seen": 266291816,
+ "step": 6085,
+ "train_runtime": 47112.3669,
+ "train_tokens_per_second": 5652.27
+ },
+ {
+ "epoch": 0.41004578507945055,
+ "grad_norm": 0.8219233379550768,
+ "learning_rate": 8.99852489401129e-06,
+ "loss": 0.40433249473571775,
+ "num_input_tokens_seen": 266517904,
+ "step": 6090,
+ "train_runtime": 47148.2501,
+ "train_tokens_per_second": 5652.763
+ },
+ {
+ "epoch": 0.4103824400754107,
+ "grad_norm": 0.7423675573693415,
+ "learning_rate": 8.996936846156748e-06,
+ "loss": 0.3635870456695557,
+ "num_input_tokens_seen": 266736992,
+ "step": 6095,
+ "train_runtime": 47183.3361,
+ "train_tokens_per_second": 5653.203
+ },
+ {
+ "epoch": 0.41071909507137083,
+ "grad_norm": 0.8008363109798308,
+ "learning_rate": 8.995347680571563e-06,
+ "loss": 0.41033496856689455,
+ "num_input_tokens_seen": 266959272,
+ "step": 6100,
+ "train_runtime": 47217.625,
+ "train_tokens_per_second": 5653.806
+ },
+ {
+ "epoch": 0.41105575006733097,
+ "grad_norm": 0.6633014625775752,
+ "learning_rate": 8.993757397700137e-06,
+ "loss": 0.38593282699584963,
+ "num_input_tokens_seen": 267178696,
+ "step": 6105,
+ "train_runtime": 47254.1693,
+ "train_tokens_per_second": 5654.077
+ },
+ {
+ "epoch": 0.41139240506329117,
+ "grad_norm": 0.7915805562051228,
+ "learning_rate": 8.99216599798719e-06,
+ "loss": 0.42214293479919435,
+ "num_input_tokens_seen": 267415152,
+ "step": 6110,
+ "train_runtime": 47290.6934,
+ "train_tokens_per_second": 5654.71
+ },
+ {
+ "epoch": 0.4117290600592513,
+ "grad_norm": 0.7917765167657784,
+ "learning_rate": 8.990573481877753e-06,
+ "loss": 0.3639373779296875,
+ "num_input_tokens_seen": 267618552,
+ "step": 6115,
+ "train_runtime": 47325.6204,
+ "train_tokens_per_second": 5654.835
+ },
+ {
+ "epoch": 0.41206571505521145,
+ "grad_norm": 0.694124505400035,
+ "learning_rate": 8.988979849817167e-06,
+ "loss": 0.36659536361694334,
+ "num_input_tokens_seen": 267814904,
+ "step": 6120,
+ "train_runtime": 47366.4119,
+ "train_tokens_per_second": 5654.11
+ },
+ {
+ "epoch": 0.4124023700511716,
+ "grad_norm": 0.8642974554518021,
+ "learning_rate": 8.987385102251085e-06,
+ "loss": 0.3810924291610718,
+ "num_input_tokens_seen": 268015432,
+ "step": 6125,
+ "train_runtime": 47405.0635,
+ "train_tokens_per_second": 5653.73
+ },
+ {
+ "epoch": 0.4127390250471317,
+ "grad_norm": 0.7416982400357927,
+ "learning_rate": 8.985789239625475e-06,
+ "loss": 0.3897581100463867,
+ "num_input_tokens_seen": 268235696,
+ "step": 6130,
+ "train_runtime": 47449.5896,
+ "train_tokens_per_second": 5653.067
+ },
+ {
+ "epoch": 0.41307568004309186,
+ "grad_norm": 0.7586989342588731,
+ "learning_rate": 8.984192262386613e-06,
+ "loss": 0.36377105712890623,
+ "num_input_tokens_seen": 268454176,
+ "step": 6135,
+ "train_runtime": 47483.2101,
+ "train_tokens_per_second": 5653.665
+ },
+ {
+ "epoch": 0.413412335039052,
+ "grad_norm": 0.7553176125857815,
+ "learning_rate": 8.98259417098109e-06,
+ "loss": 0.3994285583496094,
+ "num_input_tokens_seen": 268685424,
+ "step": 6140,
+ "train_runtime": 47523.5954,
+ "train_tokens_per_second": 5653.727
+ },
+ {
+ "epoch": 0.41374899003501214,
+ "grad_norm": 0.7619380388493862,
+ "learning_rate": 8.980994965855808e-06,
+ "loss": 0.39436161518096924,
+ "num_input_tokens_seen": 268906160,
+ "step": 6145,
+ "train_runtime": 47557.2625,
+ "train_tokens_per_second": 5654.366
+ },
+ {
+ "epoch": 0.4140856450309723,
+ "grad_norm": 0.7194190068295008,
+ "learning_rate": 8.979394647457976e-06,
+ "loss": 0.3927819490432739,
+ "num_input_tokens_seen": 269128016,
+ "step": 6150,
+ "train_runtime": 47600.5297,
+ "train_tokens_per_second": 5653.887
+ },
+ {
+ "epoch": 0.4144223000269324,
+ "grad_norm": 0.741178107175373,
+ "learning_rate": 8.97779321623512e-06,
+ "loss": 0.35940072536468504,
+ "num_input_tokens_seen": 269333048,
+ "step": 6155,
+ "train_runtime": 47638.7497,
+ "train_tokens_per_second": 5653.655
+ },
+ {
+ "epoch": 0.41475895502289256,
+ "grad_norm": 0.7334172734354966,
+ "learning_rate": 8.976190672635077e-06,
+ "loss": 0.400145435333252,
+ "num_input_tokens_seen": 269557248,
+ "step": 6160,
+ "train_runtime": 47679.304,
+ "train_tokens_per_second": 5653.548
+ },
+ {
+ "epoch": 0.4150956100188527,
+ "grad_norm": 0.7612088304227586,
+ "learning_rate": 8.974587017105991e-06,
+ "loss": 0.41222319602966306,
+ "num_input_tokens_seen": 269768408,
+ "step": 6165,
+ "train_runtime": 47713.1048,
+ "train_tokens_per_second": 5653.969
+ },
+ {
+ "epoch": 0.41543226501481284,
+ "grad_norm": 0.7344591815245108,
+ "learning_rate": 8.97298225009632e-06,
+ "loss": 0.3931986570358276,
+ "num_input_tokens_seen": 269990368,
+ "step": 6170,
+ "train_runtime": 47754.3241,
+ "train_tokens_per_second": 5653.737
+ },
+ {
+ "epoch": 0.415768920010773,
+ "grad_norm": 0.7575071698285096,
+ "learning_rate": 8.971376372054829e-06,
+ "loss": 0.3943799495697021,
+ "num_input_tokens_seen": 270208056,
+ "step": 6175,
+ "train_runtime": 47794.7334,
+ "train_tokens_per_second": 5653.511
+ },
+ {
+ "epoch": 0.4161055750067331,
+ "grad_norm": 0.6823047182780082,
+ "learning_rate": 8.969769383430602e-06,
+ "loss": 0.3858570337295532,
+ "num_input_tokens_seen": 270432152,
+ "step": 6180,
+ "train_runtime": 47829.7321,
+ "train_tokens_per_second": 5654.06
+ },
+ {
+ "epoch": 0.41644223000269326,
+ "grad_norm": 0.7476934311068052,
+ "learning_rate": 8.968161284673027e-06,
+ "loss": 0.37152729034423826,
+ "num_input_tokens_seen": 270661216,
+ "step": 6185,
+ "train_runtime": 47868.2088,
+ "train_tokens_per_second": 5654.3
+ },
+ {
+ "epoch": 0.4167788849986534,
+ "grad_norm": 0.7528325409155248,
+ "learning_rate": 8.9665520762318e-06,
+ "loss": 0.37863969802856445,
+ "num_input_tokens_seen": 270875256,
+ "step": 6190,
+ "train_runtime": 47904.4792,
+ "train_tokens_per_second": 5654.487
+ },
+ {
+ "epoch": 0.41711553999461354,
+ "grad_norm": 0.7282239751820792,
+ "learning_rate": 8.964941758556934e-06,
+ "loss": 0.3976884126663208,
+ "num_input_tokens_seen": 271088288,
+ "step": 6195,
+ "train_runtime": 47948.7488,
+ "train_tokens_per_second": 5653.709
+ },
+ {
+ "epoch": 0.4174521949905737,
+ "grad_norm": 0.755222562109598,
+ "learning_rate": 8.963330332098747e-06,
+ "loss": 0.3814378261566162,
+ "num_input_tokens_seen": 271296528,
+ "step": 6200,
+ "train_runtime": 47985.6256,
+ "train_tokens_per_second": 5653.704
+ },
+ {
+ "epoch": 0.4177888499865338,
+ "grad_norm": 0.7830273045211564,
+ "learning_rate": 8.961717797307872e-06,
+ "loss": 0.37258970737457275,
+ "num_input_tokens_seen": 271515568,
+ "step": 6205,
+ "train_runtime": 48019.4934,
+ "train_tokens_per_second": 5654.278
+ },
+ {
+ "epoch": 0.41812550498249396,
+ "grad_norm": 0.7683395292659414,
+ "learning_rate": 8.960104154635248e-06,
+ "loss": 0.4039943695068359,
+ "num_input_tokens_seen": 271733704,
+ "step": 6210,
+ "train_runtime": 48057.6768,
+ "train_tokens_per_second": 5654.325
+ },
+ {
+ "epoch": 0.4184621599784541,
+ "grad_norm": 0.7964266950242961,
+ "learning_rate": 8.958489404532125e-06,
+ "loss": 0.4006314277648926,
+ "num_input_tokens_seen": 271941040,
+ "step": 6215,
+ "train_runtime": 48094.9785,
+ "train_tokens_per_second": 5654.25
+ },
+ {
+ "epoch": 0.41879881497441424,
+ "grad_norm": 0.7229961706218653,
+ "learning_rate": 8.956873547450062e-06,
+ "loss": 0.3996126651763916,
+ "num_input_tokens_seen": 272155560,
+ "step": 6220,
+ "train_runtime": 48135.1561,
+ "train_tokens_per_second": 5653.987
+ },
+ {
+ "epoch": 0.4191354699703744,
+ "grad_norm": 0.8692942989729212,
+ "learning_rate": 8.95525658384093e-06,
+ "loss": 0.4290035247802734,
+ "num_input_tokens_seen": 272369192,
+ "step": 6225,
+ "train_runtime": 48171.8576,
+ "train_tokens_per_second": 5654.114
+ },
+ {
+ "epoch": 0.4194721249663345,
+ "grad_norm": 0.7600990815023999,
+ "learning_rate": 8.953638514156908e-06,
+ "loss": 0.3810266494750977,
+ "num_input_tokens_seen": 272591248,
+ "step": 6230,
+ "train_runtime": 48208.9238,
+ "train_tokens_per_second": 5654.373
+ },
+ {
+ "epoch": 0.41980877996229465,
+ "grad_norm": 0.7355727181513495,
+ "learning_rate": 8.952019338850481e-06,
+ "loss": 0.3950758695602417,
+ "num_input_tokens_seen": 272810616,
+ "step": 6235,
+ "train_runtime": 48248.4869,
+ "train_tokens_per_second": 5654.283
+ },
+ {
+ "epoch": 0.4201454349582548,
+ "grad_norm": 0.7644200111226909,
+ "learning_rate": 8.950399058374447e-06,
+ "loss": 0.377358603477478,
+ "num_input_tokens_seen": 273036976,
+ "step": 6240,
+ "train_runtime": 48285.7901,
+ "train_tokens_per_second": 5654.603
+ },
+ {
+ "epoch": 0.42048208995421493,
+ "grad_norm": 0.7918087336685585,
+ "learning_rate": 8.948777673181913e-06,
+ "loss": 0.389499831199646,
+ "num_input_tokens_seen": 273260832,
+ "step": 6245,
+ "train_runtime": 48322.4015,
+ "train_tokens_per_second": 5654.951
+ },
+ {
+ "epoch": 0.42081874495017507,
+ "grad_norm": 0.7765030452628453,
+ "learning_rate": 8.947155183726296e-06,
+ "loss": 0.35937402248382566,
+ "num_input_tokens_seen": 273488360,
+ "step": 6250,
+ "train_runtime": 48361.4271,
+ "train_tokens_per_second": 5655.093
+ },
+ {
+ "epoch": 0.4211553999461352,
+ "grad_norm": 0.7014097499994082,
+ "learning_rate": 8.94553159046132e-06,
+ "loss": 0.3715068817138672,
+ "num_input_tokens_seen": 273707472,
+ "step": 6255,
+ "train_runtime": 48401.1614,
+ "train_tokens_per_second": 5654.977
+ },
+ {
+ "epoch": 0.42149205494209535,
+ "grad_norm": 0.6408398788945804,
+ "learning_rate": 8.943906893841015e-06,
+ "loss": 0.36748714447021485,
+ "num_input_tokens_seen": 273928544,
+ "step": 6260,
+ "train_runtime": 48433.3616,
+ "train_tokens_per_second": 5655.782
+ },
+ {
+ "epoch": 0.4218287099380555,
+ "grad_norm": 0.754509294899034,
+ "learning_rate": 8.942281094319721e-06,
+ "loss": 0.4157464027404785,
+ "num_input_tokens_seen": 274141592,
+ "step": 6265,
+ "train_runtime": 48468.7848,
+ "train_tokens_per_second": 5656.044
+ },
+ {
+ "epoch": 0.42216536493401563,
+ "grad_norm": 0.7840590763768849,
+ "learning_rate": 8.940654192352092e-06,
+ "loss": 0.40111403465270995,
+ "num_input_tokens_seen": 274349792,
+ "step": 6270,
+ "train_runtime": 48510.3168,
+ "train_tokens_per_second": 5655.494
+ },
+ {
+ "epoch": 0.42250201992997577,
+ "grad_norm": 0.8213882192132025,
+ "learning_rate": 8.939026188393083e-06,
+ "loss": 0.3853888034820557,
+ "num_input_tokens_seen": 274569104,
+ "step": 6275,
+ "train_runtime": 48556.531,
+ "train_tokens_per_second": 5654.628
+ },
+ {
+ "epoch": 0.4228386749259359,
+ "grad_norm": 0.8791839426260795,
+ "learning_rate": 8.937397082897963e-06,
+ "loss": 0.40306596755981444,
+ "num_input_tokens_seen": 274796616,
+ "step": 6280,
+ "train_runtime": 48596.7026,
+ "train_tokens_per_second": 5654.635
+ },
+ {
+ "epoch": 0.42317532992189605,
+ "grad_norm": 0.7843983011163462,
+ "learning_rate": 8.935766876322303e-06,
+ "loss": 0.4140591621398926,
+ "num_input_tokens_seen": 275015624,
+ "step": 6285,
+ "train_runtime": 48635.2618,
+ "train_tokens_per_second": 5654.655
+ },
+ {
+ "epoch": 0.4235119849178562,
+ "grad_norm": 0.7007974546940894,
+ "learning_rate": 8.934135569121985e-06,
+ "loss": 0.3959988832473755,
+ "num_input_tokens_seen": 275242840,
+ "step": 6290,
+ "train_runtime": 48673.0052,
+ "train_tokens_per_second": 5654.938
+ },
+ {
+ "epoch": 0.4238486399138163,
+ "grad_norm": 0.7078146309363205,
+ "learning_rate": 8.932503161753202e-06,
+ "loss": 0.38456072807312014,
+ "num_input_tokens_seen": 275462504,
+ "step": 6295,
+ "train_runtime": 48709.4026,
+ "train_tokens_per_second": 5655.222
+ },
+ {
+ "epoch": 0.42418529490977647,
+ "grad_norm": 1.0351275568049705,
+ "learning_rate": 8.930869654672449e-06,
+ "loss": 0.38416409492492676,
+ "num_input_tokens_seen": 275681808,
+ "step": 6300,
+ "train_runtime": 48747.6893,
+ "train_tokens_per_second": 5655.279
+ },
+ {
+ "epoch": 0.4245219499057366,
+ "grad_norm": 0.8236989821133165,
+ "learning_rate": 8.929235048336531e-06,
+ "loss": 0.37823896408081054,
+ "num_input_tokens_seen": 275903576,
+ "step": 6305,
+ "train_runtime": 48788.5888,
+ "train_tokens_per_second": 5655.084
+ },
+ {
+ "epoch": 0.42485860490169675,
+ "grad_norm": 0.8519274837856372,
+ "learning_rate": 8.927599343202563e-06,
+ "loss": 0.4197996139526367,
+ "num_input_tokens_seen": 276132792,
+ "step": 6310,
+ "train_runtime": 48821.7451,
+ "train_tokens_per_second": 5655.939
+ },
+ {
+ "epoch": 0.4251952598976569,
+ "grad_norm": 0.7601906981584227,
+ "learning_rate": 8.92596253972796e-06,
+ "loss": 0.39647321701049804,
+ "num_input_tokens_seen": 276373576,
+ "step": 6315,
+ "train_runtime": 48856.0141,
+ "train_tokens_per_second": 5656.9
+ },
+ {
+ "epoch": 0.425531914893617,
+ "grad_norm": 0.7436129276391935,
+ "learning_rate": 8.924324638370454e-06,
+ "loss": 0.4159084320068359,
+ "num_input_tokens_seen": 276601344,
+ "step": 6320,
+ "train_runtime": 48892.2285,
+ "train_tokens_per_second": 5657.368
+ },
+ {
+ "epoch": 0.42586856988957716,
+ "grad_norm": 0.619825666948129,
+ "learning_rate": 8.922685639588076e-06,
+ "loss": 0.3394161701202393,
+ "num_input_tokens_seen": 276801216,
+ "step": 6325,
+ "train_runtime": 48929.4586,
+ "train_tokens_per_second": 5657.149
+ },
+ {
+ "epoch": 0.4262052248855373,
+ "grad_norm": 0.7566990752584487,
+ "learning_rate": 8.921045543839167e-06,
+ "loss": 0.412354850769043,
+ "num_input_tokens_seen": 277025984,
+ "step": 6330,
+ "train_runtime": 48964.4429,
+ "train_tokens_per_second": 5657.697
+ },
+ {
+ "epoch": 0.42654187988149744,
+ "grad_norm": 0.7706529692481164,
+ "learning_rate": 8.919404351582376e-06,
+ "loss": 0.37722196578979494,
+ "num_input_tokens_seen": 277234256,
+ "step": 6335,
+ "train_runtime": 49000.4277,
+ "train_tokens_per_second": 5657.793
+ },
+ {
+ "epoch": 0.4268785348774576,
+ "grad_norm": 0.7056105235159689,
+ "learning_rate": 8.917762063276657e-06,
+ "loss": 0.38721909523010256,
+ "num_input_tokens_seen": 277448936,
+ "step": 6340,
+ "train_runtime": 49039.3589,
+ "train_tokens_per_second": 5657.679
+ },
+ {
+ "epoch": 0.4272151898734177,
+ "grad_norm": 0.7187047540510013,
+ "learning_rate": 8.916118679381268e-06,
+ "loss": 0.3728447198867798,
+ "num_input_tokens_seen": 277667920,
+ "step": 6345,
+ "train_runtime": 49076.9224,
+ "train_tokens_per_second": 5657.81
+ },
+ {
+ "epoch": 0.42755184486937786,
+ "grad_norm": 0.8958354766388671,
+ "learning_rate": 8.914474200355777e-06,
+ "loss": 0.39159350395202636,
+ "num_input_tokens_seen": 277885288,
+ "step": 6350,
+ "train_runtime": 49121.5781,
+ "train_tokens_per_second": 5657.092
+ },
+ {
+ "epoch": 0.427888499865338,
+ "grad_norm": 0.8411282788940809,
+ "learning_rate": 8.912828626660059e-06,
+ "loss": 0.4010343074798584,
+ "num_input_tokens_seen": 278112192,
+ "step": 6355,
+ "train_runtime": 49161.2354,
+ "train_tokens_per_second": 5657.144
+ },
+ {
+ "epoch": 0.42822515486129814,
+ "grad_norm": 0.7895834565221768,
+ "learning_rate": 8.911181958754294e-06,
+ "loss": 0.4207916259765625,
+ "num_input_tokens_seen": 278332088,
+ "step": 6360,
+ "train_runtime": 49198.4506,
+ "train_tokens_per_second": 5657.334
+ },
+ {
+ "epoch": 0.4285618098572583,
+ "grad_norm": 0.6974429090218124,
+ "learning_rate": 8.909534197098966e-06,
+ "loss": 0.4089661121368408,
+ "num_input_tokens_seen": 278560320,
+ "step": 6365,
+ "train_runtime": 49240.3447,
+ "train_tokens_per_second": 5657.156
+ },
+ {
+ "epoch": 0.4288984648532184,
+ "grad_norm": 0.7879065723514492,
+ "learning_rate": 8.907885342154865e-06,
+ "loss": 0.3871757507324219,
+ "num_input_tokens_seen": 278780408,
+ "step": 6370,
+ "train_runtime": 49273.5524,
+ "train_tokens_per_second": 5657.81
+ },
+ {
+ "epoch": 0.42923511984917856,
+ "grad_norm": 0.809098238394157,
+ "learning_rate": 8.906235394383089e-06,
+ "loss": 0.3723259925842285,
+ "num_input_tokens_seen": 278977576,
+ "step": 6375,
+ "train_runtime": 49314.5399,
+ "train_tokens_per_second": 5657.106
+ },
+ {
+ "epoch": 0.4295717748451387,
+ "grad_norm": 0.7788656469697295,
+ "learning_rate": 8.904584354245042e-06,
+ "loss": 0.38217949867248535,
+ "num_input_tokens_seen": 279198056,
+ "step": 6380,
+ "train_runtime": 49349.1158,
+ "train_tokens_per_second": 5657.61
+ },
+ {
+ "epoch": 0.42990842984109884,
+ "grad_norm": 0.7480156005511179,
+ "learning_rate": 8.902932222202433e-06,
+ "loss": 0.39810030460357665,
+ "num_input_tokens_seen": 279431128,
+ "step": 6385,
+ "train_runtime": 49390.7865,
+ "train_tokens_per_second": 5657.556
+ },
+ {
+ "epoch": 0.430245084837059,
+ "grad_norm": 0.7111926102410158,
+ "learning_rate": 8.901278998717272e-06,
+ "loss": 0.37800168991088867,
+ "num_input_tokens_seen": 279655568,
+ "step": 6390,
+ "train_runtime": 49420.9322,
+ "train_tokens_per_second": 5658.646
+ },
+ {
+ "epoch": 0.4305817398330191,
+ "grad_norm": 0.8127379256774018,
+ "learning_rate": 8.899624684251878e-06,
+ "loss": 0.38466432094573977,
+ "num_input_tokens_seen": 279872920,
+ "step": 6395,
+ "train_runtime": 49461.0295,
+ "train_tokens_per_second": 5658.453
+ },
+ {
+ "epoch": 0.43091839482897926,
+ "grad_norm": 0.7440397221517976,
+ "learning_rate": 8.897969279268877e-06,
+ "loss": 0.37961535453796386,
+ "num_input_tokens_seen": 280093688,
+ "step": 6400,
+ "train_runtime": 49494.6401,
+ "train_tokens_per_second": 5659.071
+ },
+ {
+ "epoch": 0.4312550498249394,
+ "grad_norm": 0.7095617451331785,
+ "learning_rate": 8.896312784231196e-06,
+ "loss": 0.3712193012237549,
+ "num_input_tokens_seen": 280305104,
+ "step": 6405,
+ "train_runtime": 49529.3102,
+ "train_tokens_per_second": 5659.378
+ },
+ {
+ "epoch": 0.43159170482089954,
+ "grad_norm": 0.8249055107614544,
+ "learning_rate": 8.89465519960207e-06,
+ "loss": 0.42307000160217284,
+ "num_input_tokens_seen": 280534896,
+ "step": 6410,
+ "train_runtime": 49562.7367,
+ "train_tokens_per_second": 5660.198
+ },
+ {
+ "epoch": 0.4319283598168597,
+ "grad_norm": 0.7044078667504021,
+ "learning_rate": 8.892996525845037e-06,
+ "loss": 0.39340283870697024,
+ "num_input_tokens_seen": 280760536,
+ "step": 6415,
+ "train_runtime": 49609.124,
+ "train_tokens_per_second": 5659.454
+ },
+ {
+ "epoch": 0.4322650148128198,
+ "grad_norm": 0.8758617954616612,
+ "learning_rate": 8.89133676342394e-06,
+ "loss": 0.4034512996673584,
+ "num_input_tokens_seen": 280959600,
+ "step": 6420,
+ "train_runtime": 49647.7511,
+ "train_tokens_per_second": 5659.06
+ },
+ {
+ "epoch": 0.43260166980877995,
+ "grad_norm": 0.7612824097484856,
+ "learning_rate": 8.889675912802923e-06,
+ "loss": 0.38754520416259763,
+ "num_input_tokens_seen": 281173760,
+ "step": 6425,
+ "train_runtime": 49684.4958,
+ "train_tokens_per_second": 5659.185
+ },
+ {
+ "epoch": 0.4329383248047401,
+ "grad_norm": 0.8243557630186539,
+ "learning_rate": 8.888013974446443e-06,
+ "loss": 0.4189424514770508,
+ "num_input_tokens_seen": 281395808,
+ "step": 6430,
+ "train_runtime": 49725.5827,
+ "train_tokens_per_second": 5658.975
+ },
+ {
+ "epoch": 0.43327497980070023,
+ "grad_norm": 0.8439524750777101,
+ "learning_rate": 8.886350948819253e-06,
+ "loss": 0.38129727840423583,
+ "num_input_tokens_seen": 281606072,
+ "step": 6435,
+ "train_runtime": 49758.4427,
+ "train_tokens_per_second": 5659.463
+ },
+ {
+ "epoch": 0.4336116347966604,
+ "grad_norm": 0.8007298009088516,
+ "learning_rate": 8.884686836386412e-06,
+ "loss": 0.38291594982147215,
+ "num_input_tokens_seen": 281822424,
+ "step": 6440,
+ "train_runtime": 49799.792,
+ "train_tokens_per_second": 5659.108
+ },
+ {
+ "epoch": 0.4339482897926205,
+ "grad_norm": 0.7881388654599728,
+ "learning_rate": 8.883021637613286e-06,
+ "loss": 0.4050453662872314,
+ "num_input_tokens_seen": 282049648,
+ "step": 6445,
+ "train_runtime": 49835.0256,
+ "train_tokens_per_second": 5659.667
+ },
+ {
+ "epoch": 0.43428494478858065,
+ "grad_norm": 0.6887341883009596,
+ "learning_rate": 8.881355352965537e-06,
+ "loss": 0.38397510051727296,
+ "num_input_tokens_seen": 282264512,
+ "step": 6450,
+ "train_runtime": 49867.6041,
+ "train_tokens_per_second": 5660.278
+ },
+ {
+ "epoch": 0.4346215997845408,
+ "grad_norm": 0.672101432469643,
+ "learning_rate": 8.879687982909145e-06,
+ "loss": 0.3574419736862183,
+ "num_input_tokens_seen": 282465912,
+ "step": 6455,
+ "train_runtime": 49905.6117,
+ "train_tokens_per_second": 5660.003
+ },
+ {
+ "epoch": 0.43495825478050093,
+ "grad_norm": 0.8916903644937464,
+ "learning_rate": 8.878019527910378e-06,
+ "loss": 0.41092991828918457,
+ "num_input_tokens_seen": 282670240,
+ "step": 6460,
+ "train_runtime": 49945.0146,
+ "train_tokens_per_second": 5659.629
+ },
+ {
+ "epoch": 0.43529490977646107,
+ "grad_norm": 0.8298316229675538,
+ "learning_rate": 8.876349988435815e-06,
+ "loss": 0.44412593841552733,
+ "num_input_tokens_seen": 282887856,
+ "step": 6465,
+ "train_runtime": 49983.0629,
+ "train_tokens_per_second": 5659.674
+ },
+ {
+ "epoch": 0.4356315647724212,
+ "grad_norm": 0.6431274180779993,
+ "learning_rate": 8.874679364952339e-06,
+ "loss": 0.38706722259521487,
+ "num_input_tokens_seen": 283127040,
+ "step": 6470,
+ "train_runtime": 50021.3951,
+ "train_tokens_per_second": 5660.119
+ },
+ {
+ "epoch": 0.43596821976838135,
+ "grad_norm": 0.7656387607672229,
+ "learning_rate": 8.873007657927135e-06,
+ "loss": 0.3893885135650635,
+ "num_input_tokens_seen": 283357232,
+ "step": 6475,
+ "train_runtime": 50062.5333,
+ "train_tokens_per_second": 5660.066
+ },
+ {
+ "epoch": 0.4363048747643415,
+ "grad_norm": 0.7096570879969786,
+ "learning_rate": 8.87133486782769e-06,
+ "loss": 0.35058715343475344,
+ "num_input_tokens_seen": 283558232,
+ "step": 6480,
+ "train_runtime": 50096.5148,
+ "train_tokens_per_second": 5660.239
+ },
+ {
+ "epoch": 0.4366415297603016,
+ "grad_norm": 0.8761505694419173,
+ "learning_rate": 8.869660995121792e-06,
+ "loss": 0.3999923229217529,
+ "num_input_tokens_seen": 283777800,
+ "step": 6485,
+ "train_runtime": 50131.8591,
+ "train_tokens_per_second": 5660.628
+ },
+ {
+ "epoch": 0.43697818475626177,
+ "grad_norm": 0.7555798070151492,
+ "learning_rate": 8.867986040277537e-06,
+ "loss": 0.4016421318054199,
+ "num_input_tokens_seen": 283985824,
+ "step": 6490,
+ "train_runtime": 50172.9,
+ "train_tokens_per_second": 5660.144
+ },
+ {
+ "epoch": 0.4373148397522219,
+ "grad_norm": 0.7001554230524034,
+ "learning_rate": 8.86631000376332e-06,
+ "loss": 0.4068891525268555,
+ "num_input_tokens_seen": 284193496,
+ "step": 6495,
+ "train_runtime": 50210.1755,
+ "train_tokens_per_second": 5660.078
+ },
+ {
+ "epoch": 0.43765149474818205,
+ "grad_norm": 0.809001796546686,
+ "learning_rate": 8.864632886047841e-06,
+ "loss": 0.3719168186187744,
+ "num_input_tokens_seen": 284406672,
+ "step": 6500,
+ "train_runtime": 50246.2209,
+ "train_tokens_per_second": 5660.26
+ },
+ {
+ "epoch": 0.4379881497441422,
+ "grad_norm": 0.8525357252115519,
+ "learning_rate": 8.862954687600097e-06,
+ "loss": 0.4161073684692383,
+ "num_input_tokens_seen": 284620064,
+ "step": 6505,
+ "train_runtime": 50283.9357,
+ "train_tokens_per_second": 5660.258
+ },
+ {
+ "epoch": 0.4383248047401023,
+ "grad_norm": 0.818348330515836,
+ "learning_rate": 8.861275408889393e-06,
+ "loss": 0.4044035911560059,
+ "num_input_tokens_seen": 284827680,
+ "step": 6510,
+ "train_runtime": 50319.756,
+ "train_tokens_per_second": 5660.355
+ },
+ {
+ "epoch": 0.43866145973606246,
+ "grad_norm": 0.8194841529764723,
+ "learning_rate": 8.859595050385332e-06,
+ "loss": 0.40093269348144533,
+ "num_input_tokens_seen": 285050288,
+ "step": 6515,
+ "train_runtime": 50356.7378,
+ "train_tokens_per_second": 5660.619
+ },
+ {
+ "epoch": 0.4389981147320226,
+ "grad_norm": 0.7603950111901223,
+ "learning_rate": 8.857913612557825e-06,
+ "loss": 0.39177913665771485,
+ "num_input_tokens_seen": 285275776,
+ "step": 6520,
+ "train_runtime": 50399.864,
+ "train_tokens_per_second": 5660.249
+ },
+ {
+ "epoch": 0.43933476972798274,
+ "grad_norm": 0.9171110192985658,
+ "learning_rate": 8.856231095877077e-06,
+ "loss": 0.37299246788024903,
+ "num_input_tokens_seen": 285483840,
+ "step": 6525,
+ "train_runtime": 50437.6759,
+ "train_tokens_per_second": 5660.131
+ },
+ {
+ "epoch": 0.4396714247239429,
+ "grad_norm": 0.6722265713323919,
+ "learning_rate": 8.8545475008136e-06,
+ "loss": 0.3728148937225342,
+ "num_input_tokens_seen": 285708616,
+ "step": 6530,
+ "train_runtime": 50478.218,
+ "train_tokens_per_second": 5660.038
+ },
+ {
+ "epoch": 0.440008079719903,
+ "grad_norm": 0.8225035496811262,
+ "learning_rate": 8.852862827838207e-06,
+ "loss": 0.37645254135131834,
+ "num_input_tokens_seen": 285914952,
+ "step": 6535,
+ "train_runtime": 50513.9654,
+ "train_tokens_per_second": 5660.117
+ },
+ {
+ "epoch": 0.44034473471586316,
+ "grad_norm": 0.672139181444301,
+ "learning_rate": 8.851177077422008e-06,
+ "loss": 0.4013407230377197,
+ "num_input_tokens_seen": 286146336,
+ "step": 6540,
+ "train_runtime": 50553.4611,
+ "train_tokens_per_second": 5660.272
+ },
+ {
+ "epoch": 0.4406813897118233,
+ "grad_norm": 0.7644327948595142,
+ "learning_rate": 8.849490250036421e-06,
+ "loss": 0.4249086380004883,
+ "num_input_tokens_seen": 286362088,
+ "step": 6545,
+ "train_runtime": 50590.1862,
+ "train_tokens_per_second": 5660.428
+ },
+ {
+ "epoch": 0.44101804470778344,
+ "grad_norm": 0.8087430911202192,
+ "learning_rate": 8.847802346153159e-06,
+ "loss": 0.3778714895248413,
+ "num_input_tokens_seen": 286576504,
+ "step": 6550,
+ "train_runtime": 50623.1732,
+ "train_tokens_per_second": 5660.975
+ },
+ {
+ "epoch": 0.4413546997037436,
+ "grad_norm": 0.822277412759685,
+ "learning_rate": 8.846113366244241e-06,
+ "loss": 0.40006532669067385,
+ "num_input_tokens_seen": 286802512,
+ "step": 6555,
+ "train_runtime": 50653.7568,
+ "train_tokens_per_second": 5662.019
+ },
+ {
+ "epoch": 0.4416913546997037,
+ "grad_norm": 0.7788883116407057,
+ "learning_rate": 8.844423310781986e-06,
+ "loss": 0.3659277677536011,
+ "num_input_tokens_seen": 287019968,
+ "step": 6560,
+ "train_runtime": 50693.4968,
+ "train_tokens_per_second": 5661.87
+ },
+ {
+ "epoch": 0.44202800969566386,
+ "grad_norm": 0.8189856612851919,
+ "learning_rate": 8.842732180239011e-06,
+ "loss": 0.417068338394165,
+ "num_input_tokens_seen": 287233880,
+ "step": 6565,
+ "train_runtime": 50732.9626,
+ "train_tokens_per_second": 5661.682
+ },
+ {
+ "epoch": 0.442364664691624,
+ "grad_norm": 0.7955988955008321,
+ "learning_rate": 8.841039975088234e-06,
+ "loss": 0.37861127853393556,
+ "num_input_tokens_seen": 287450920,
+ "step": 6570,
+ "train_runtime": 50770.9895,
+ "train_tokens_per_second": 5661.716
+ },
+ {
+ "epoch": 0.44270131968758414,
+ "grad_norm": 0.7673850184930923,
+ "learning_rate": 8.839346695802878e-06,
+ "loss": 0.44512476921081545,
+ "num_input_tokens_seen": 287680008,
+ "step": 6575,
+ "train_runtime": 50808.7713,
+ "train_tokens_per_second": 5662.015
+ },
+ {
+ "epoch": 0.4430379746835443,
+ "grad_norm": 0.707236454889657,
+ "learning_rate": 8.837652342856459e-06,
+ "loss": 0.4026740550994873,
+ "num_input_tokens_seen": 287904440,
+ "step": 6580,
+ "train_runtime": 50842.5258,
+ "train_tokens_per_second": 5662.67
+ },
+ {
+ "epoch": 0.4433746296795044,
+ "grad_norm": 0.7131491003211293,
+ "learning_rate": 8.835956916722803e-06,
+ "loss": 0.37171001434326173,
+ "num_input_tokens_seen": 288123072,
+ "step": 6585,
+ "train_runtime": 50884.3886,
+ "train_tokens_per_second": 5662.308
+ },
+ {
+ "epoch": 0.44371128467546456,
+ "grad_norm": 0.7279859643678576,
+ "learning_rate": 8.834260417876024e-06,
+ "loss": 0.39625725746154783,
+ "num_input_tokens_seen": 288337792,
+ "step": 6590,
+ "train_runtime": 50922.6708,
+ "train_tokens_per_second": 5662.268
+ },
+ {
+ "epoch": 0.4440479396714247,
+ "grad_norm": 0.6822187993121582,
+ "learning_rate": 8.832562846790549e-06,
+ "loss": 0.4032885074615479,
+ "num_input_tokens_seen": 288562504,
+ "step": 6595,
+ "train_runtime": 50958.16,
+ "train_tokens_per_second": 5662.734
+ },
+ {
+ "epoch": 0.4443845946673849,
+ "grad_norm": 0.7696796062355198,
+ "learning_rate": 8.830864203941092e-06,
+ "loss": 0.4127056121826172,
+ "num_input_tokens_seen": 288783304,
+ "step": 6600,
+ "train_runtime": 50994.9294,
+ "train_tokens_per_second": 5662.981
+ },
+ {
+ "epoch": 0.44472124966334503,
+ "grad_norm": 0.796046357490389,
+ "learning_rate": 8.829164489802677e-06,
+ "loss": 0.37214198112487795,
+ "num_input_tokens_seen": 288995032,
+ "step": 6605,
+ "train_runtime": 51030.6262,
+ "train_tokens_per_second": 5663.168
+ },
+ {
+ "epoch": 0.44505790465930517,
+ "grad_norm": 0.8194054840462127,
+ "learning_rate": 8.827463704850622e-06,
+ "loss": 0.3779730796813965,
+ "num_input_tokens_seen": 289211784,
+ "step": 6610,
+ "train_runtime": 51065.1791,
+ "train_tokens_per_second": 5663.581
+ },
+ {
+ "epoch": 0.4453945596552653,
+ "grad_norm": 0.853917923414557,
+ "learning_rate": 8.825761849560547e-06,
+ "loss": 0.3779446125030518,
+ "num_input_tokens_seen": 289426496,
+ "step": 6615,
+ "train_runtime": 51104.5336,
+ "train_tokens_per_second": 5663.421
+ },
+ {
+ "epoch": 0.44573121465122545,
+ "grad_norm": 0.8243805829529494,
+ "learning_rate": 8.824058924408371e-06,
+ "loss": 0.3913204431533813,
+ "num_input_tokens_seen": 289642104,
+ "step": 6620,
+ "train_runtime": 51148.3439,
+ "train_tokens_per_second": 5662.786
+ },
+ {
+ "epoch": 0.4460678696471856,
+ "grad_norm": 0.6977791985037581,
+ "learning_rate": 8.822354929870311e-06,
+ "loss": 0.3713568687438965,
+ "num_input_tokens_seen": 289855080,
+ "step": 6625,
+ "train_runtime": 51186.4829,
+ "train_tokens_per_second": 5662.727
+ },
+ {
+ "epoch": 0.44640452464314573,
+ "grad_norm": 0.6544439461875585,
+ "learning_rate": 8.820649866422884e-06,
+ "loss": 0.3891900539398193,
+ "num_input_tokens_seen": 290080272,
+ "step": 6630,
+ "train_runtime": 51221.0131,
+ "train_tokens_per_second": 5663.306
+ },
+ {
+ "epoch": 0.44674117963910587,
+ "grad_norm": 0.7628155627258327,
+ "learning_rate": 8.818943734542905e-06,
+ "loss": 0.38543388843536375,
+ "num_input_tokens_seen": 290310296,
+ "step": 6635,
+ "train_runtime": 51260.636,
+ "train_tokens_per_second": 5663.416
+ },
+ {
+ "epoch": 0.447077834635066,
+ "grad_norm": 0.7154748015782274,
+ "learning_rate": 8.817236534707486e-06,
+ "loss": 0.38082070350646974,
+ "num_input_tokens_seen": 290530408,
+ "step": 6640,
+ "train_runtime": 51298.209,
+ "train_tokens_per_second": 5663.559
+ },
+ {
+ "epoch": 0.44741448963102615,
+ "grad_norm": 0.719435209044012,
+ "learning_rate": 8.815528267394045e-06,
+ "loss": 0.38055622577667236,
+ "num_input_tokens_seen": 290752096,
+ "step": 6645,
+ "train_runtime": 51339.4277,
+ "train_tokens_per_second": 5663.33
+ },
+ {
+ "epoch": 0.4477511446269863,
+ "grad_norm": 0.8130810621777962,
+ "learning_rate": 8.81381893308029e-06,
+ "loss": 0.42238750457763674,
+ "num_input_tokens_seen": 290976696,
+ "step": 6650,
+ "train_runtime": 51376.6803,
+ "train_tokens_per_second": 5663.595
+ },
+ {
+ "epoch": 0.4480877996229464,
+ "grad_norm": 0.6802969800260558,
+ "learning_rate": 8.81210853224423e-06,
+ "loss": 0.397933292388916,
+ "num_input_tokens_seen": 291195128,
+ "step": 6655,
+ "train_runtime": 51416.0644,
+ "train_tokens_per_second": 5663.505
+ },
+ {
+ "epoch": 0.44842445461890656,
+ "grad_norm": 0.7683337052075077,
+ "learning_rate": 8.810397065364177e-06,
+ "loss": 0.3660165309906006,
+ "num_input_tokens_seen": 291397432,
+ "step": 6660,
+ "train_runtime": 51460.1895,
+ "train_tokens_per_second": 5662.58
+ },
+ {
+ "epoch": 0.4487611096148667,
+ "grad_norm": 0.7385595517387719,
+ "learning_rate": 8.808684532918735e-06,
+ "loss": 0.4035042762756348,
+ "num_input_tokens_seen": 291611984,
+ "step": 6665,
+ "train_runtime": 51497.2825,
+ "train_tokens_per_second": 5662.667
+ },
+ {
+ "epoch": 0.44909776461082684,
+ "grad_norm": 0.8670076691450054,
+ "learning_rate": 8.806970935386807e-06,
+ "loss": 0.37046058177948,
+ "num_input_tokens_seen": 291817952,
+ "step": 6670,
+ "train_runtime": 51539.7964,
+ "train_tokens_per_second": 5661.993
+ },
+ {
+ "epoch": 0.449434419606787,
+ "grad_norm": 0.845351499495225,
+ "learning_rate": 8.805256273247597e-06,
+ "loss": 0.3828955411911011,
+ "num_input_tokens_seen": 292030296,
+ "step": 6675,
+ "train_runtime": 51585.7162,
+ "train_tokens_per_second": 5661.069
+ },
+ {
+ "epoch": 0.4497710746027471,
+ "grad_norm": 0.7558486825710268,
+ "learning_rate": 8.803540546980605e-06,
+ "loss": 0.3911006450653076,
+ "num_input_tokens_seen": 292240392,
+ "step": 6680,
+ "train_runtime": 51625.5292,
+ "train_tokens_per_second": 5660.773
+ },
+ {
+ "epoch": 0.45010772959870726,
+ "grad_norm": 0.7143473072799023,
+ "learning_rate": 8.801823757065628e-06,
+ "loss": 0.39244418144226073,
+ "num_input_tokens_seen": 292456784,
+ "step": 6685,
+ "train_runtime": 51661.3439,
+ "train_tokens_per_second": 5661.037
+ },
+ {
+ "epoch": 0.4504443845946674,
+ "grad_norm": 0.7046306873714272,
+ "learning_rate": 8.800105903982758e-06,
+ "loss": 0.37194869518280027,
+ "num_input_tokens_seen": 292669672,
+ "step": 6690,
+ "train_runtime": 51695.7982,
+ "train_tokens_per_second": 5661.382
+ },
+ {
+ "epoch": 0.45078103959062754,
+ "grad_norm": 0.9038260012464076,
+ "learning_rate": 8.798386988212393e-06,
+ "loss": 0.36289196014404296,
+ "num_input_tokens_seen": 292866896,
+ "step": 6695,
+ "train_runtime": 51737.8021,
+ "train_tokens_per_second": 5660.598
+ },
+ {
+ "epoch": 0.4511176945865877,
+ "grad_norm": 0.7097725989673881,
+ "learning_rate": 8.796667010235216e-06,
+ "loss": 0.4039313793182373,
+ "num_input_tokens_seen": 293088104,
+ "step": 6700,
+ "train_runtime": 51775.0194,
+ "train_tokens_per_second": 5660.801
+ },
+ {
+ "epoch": 0.4514543495825478,
+ "grad_norm": 0.6983233354591538,
+ "learning_rate": 8.794945970532219e-06,
+ "loss": 0.38548543453216555,
+ "num_input_tokens_seen": 293312720,
+ "step": 6705,
+ "train_runtime": 51819.136,
+ "train_tokens_per_second": 5660.317
+ },
+ {
+ "epoch": 0.45179100457850796,
+ "grad_norm": 0.7715442099717464,
+ "learning_rate": 8.79322386958468e-06,
+ "loss": 0.42191410064697266,
+ "num_input_tokens_seen": 293540040,
+ "step": 6710,
+ "train_runtime": 51850.4773,
+ "train_tokens_per_second": 5661.279
+ },
+ {
+ "epoch": 0.4521276595744681,
+ "grad_norm": 0.8667322084375894,
+ "learning_rate": 8.791500707874183e-06,
+ "loss": 0.3590597152709961,
+ "num_input_tokens_seen": 293753504,
+ "step": 6715,
+ "train_runtime": 51887.9851,
+ "train_tokens_per_second": 5661.301
+ },
+ {
+ "epoch": 0.45246431457042824,
+ "grad_norm": 0.7001275719009605,
+ "learning_rate": 8.789776485882601e-06,
+ "loss": 0.40479679107666017,
+ "num_input_tokens_seen": 293980328,
+ "step": 6720,
+ "train_runtime": 51930.6229,
+ "train_tokens_per_second": 5661.021
+ },
+ {
+ "epoch": 0.4528009695663884,
+ "grad_norm": 0.7486619231659534,
+ "learning_rate": 8.788051204092112e-06,
+ "loss": 0.380597448348999,
+ "num_input_tokens_seen": 294201288,
+ "step": 6725,
+ "train_runtime": 51969.1487,
+ "train_tokens_per_second": 5661.076
+ },
+ {
+ "epoch": 0.4531376245623485,
+ "grad_norm": 0.837803412390275,
+ "learning_rate": 8.786324862985183e-06,
+ "loss": 0.39116497039794923,
+ "num_input_tokens_seen": 294414792,
+ "step": 6730,
+ "train_runtime": 52007.4528,
+ "train_tokens_per_second": 5661.012
+ },
+ {
+ "epoch": 0.45347427955830866,
+ "grad_norm": 0.7989569302827977,
+ "learning_rate": 8.78459746304458e-06,
+ "loss": 0.37044734954833985,
+ "num_input_tokens_seen": 294593136,
+ "step": 6735,
+ "train_runtime": 52048.2566,
+ "train_tokens_per_second": 5660.0
+ },
+ {
+ "epoch": 0.4538109345542688,
+ "grad_norm": 0.7910699073925987,
+ "learning_rate": 8.782869004753363e-06,
+ "loss": 0.40142025947570803,
+ "num_input_tokens_seen": 294808392,
+ "step": 6740,
+ "train_runtime": 52084.1003,
+ "train_tokens_per_second": 5660.238
+ },
+ {
+ "epoch": 0.45414758955022894,
+ "grad_norm": 0.7578062648097129,
+ "learning_rate": 8.781139488594892e-06,
+ "loss": 0.39638261795043944,
+ "num_input_tokens_seen": 295032712,
+ "step": 6745,
+ "train_runtime": 52116.6196,
+ "train_tokens_per_second": 5661.01
+ },
+ {
+ "epoch": 0.4544842445461891,
+ "grad_norm": 0.7479401557774803,
+ "learning_rate": 8.779408915052821e-06,
+ "loss": 0.3619338274002075,
+ "num_input_tokens_seen": 295250848,
+ "step": 6750,
+ "train_runtime": 52158.5344,
+ "train_tokens_per_second": 5660.643
+ },
+ {
+ "epoch": 0.4548208995421492,
+ "grad_norm": 0.7929634941210536,
+ "learning_rate": 8.777677284611096e-06,
+ "loss": 0.36865906715393065,
+ "num_input_tokens_seen": 295444808,
+ "step": 6755,
+ "train_runtime": 52194.881,
+ "train_tokens_per_second": 5660.417
+ },
+ {
+ "epoch": 0.45515755453810935,
+ "grad_norm": 0.7977612087147963,
+ "learning_rate": 8.775944597753969e-06,
+ "loss": 0.3817254304885864,
+ "num_input_tokens_seen": 295651640,
+ "step": 6760,
+ "train_runtime": 52235.5188,
+ "train_tokens_per_second": 5659.973
+ },
+ {
+ "epoch": 0.4554942095340695,
+ "grad_norm": 0.7276333392461809,
+ "learning_rate": 8.774210854965972e-06,
+ "loss": 0.4110917568206787,
+ "num_input_tokens_seen": 295872600,
+ "step": 6765,
+ "train_runtime": 52282.5363,
+ "train_tokens_per_second": 5659.11
+ },
+ {
+ "epoch": 0.45583086453002963,
+ "grad_norm": 0.7743347789370404,
+ "learning_rate": 8.772476056731946e-06,
+ "loss": 0.3540276050567627,
+ "num_input_tokens_seen": 296100760,
+ "step": 6770,
+ "train_runtime": 52321.5826,
+ "train_tokens_per_second": 5659.247
+ },
+ {
+ "epoch": 0.4561675195259898,
+ "grad_norm": 0.8417565391365261,
+ "learning_rate": 8.77074020353702e-06,
+ "loss": 0.37112271785736084,
+ "num_input_tokens_seen": 296310112,
+ "step": 6775,
+ "train_runtime": 52364.5089,
+ "train_tokens_per_second": 5658.606
+ },
+ {
+ "epoch": 0.4565041745219499,
+ "grad_norm": 0.8971045549426276,
+ "learning_rate": 8.76900329586662e-06,
+ "loss": 0.416597318649292,
+ "num_input_tokens_seen": 296529072,
+ "step": 6780,
+ "train_runtime": 52408.4219,
+ "train_tokens_per_second": 5658.042
+ },
+ {
+ "epoch": 0.45684082951791005,
+ "grad_norm": 0.7289171592528997,
+ "learning_rate": 8.767265334206467e-06,
+ "loss": 0.38317620754241943,
+ "num_input_tokens_seen": 296750896,
+ "step": 6785,
+ "train_runtime": 52453.3014,
+ "train_tokens_per_second": 5657.43
+ },
+ {
+ "epoch": 0.4571774845138702,
+ "grad_norm": 0.742990584892427,
+ "learning_rate": 8.765526319042577e-06,
+ "loss": 0.38066811561584474,
+ "num_input_tokens_seen": 296973576,
+ "step": 6790,
+ "train_runtime": 52484.833,
+ "train_tokens_per_second": 5658.274
+ },
+ {
+ "epoch": 0.45751413950983033,
+ "grad_norm": 0.7941387596625845,
+ "learning_rate": 8.763786250861258e-06,
+ "loss": 0.35711822509765623,
+ "num_input_tokens_seen": 297198488,
+ "step": 6795,
+ "train_runtime": 52525.9579,
+ "train_tokens_per_second": 5658.126
+ },
+ {
+ "epoch": 0.45785079450579047,
+ "grad_norm": 0.7637071617244607,
+ "learning_rate": 8.762045130149114e-06,
+ "loss": 0.37547504901885986,
+ "num_input_tokens_seen": 297424408,
+ "step": 6800,
+ "train_runtime": 52568.1627,
+ "train_tokens_per_second": 5657.881
+ },
+ {
+ "epoch": 0.4581874495017506,
+ "grad_norm": 0.7114371063573226,
+ "learning_rate": 8.76030295739305e-06,
+ "loss": 0.391666579246521,
+ "num_input_tokens_seen": 297652664,
+ "step": 6805,
+ "train_runtime": 52605.904,
+ "train_tokens_per_second": 5658.161
+ },
+ {
+ "epoch": 0.45852410449771075,
+ "grad_norm": 0.7263315072433297,
+ "learning_rate": 8.758559733080252e-06,
+ "loss": 0.4111741065979004,
+ "num_input_tokens_seen": 297874936,
+ "step": 6810,
+ "train_runtime": 52637.0942,
+ "train_tokens_per_second": 5659.031
+ },
+ {
+ "epoch": 0.4588607594936709,
+ "grad_norm": 0.7359316103796298,
+ "learning_rate": 8.75681545769821e-06,
+ "loss": 0.38690164089202883,
+ "num_input_tokens_seen": 298107584,
+ "step": 6815,
+ "train_runtime": 52669.678,
+ "train_tokens_per_second": 5659.947
+ },
+ {
+ "epoch": 0.45919741448963103,
+ "grad_norm": 0.8032263081480874,
+ "learning_rate": 8.755070131734706e-06,
+ "loss": 0.3867367744445801,
+ "num_input_tokens_seen": 298320208,
+ "step": 6820,
+ "train_runtime": 52707.8474,
+ "train_tokens_per_second": 5659.882
+ },
+ {
+ "epoch": 0.45953406948559117,
+ "grad_norm": 0.7833333891541101,
+ "learning_rate": 8.753323755677812e-06,
+ "loss": 0.4013802528381348,
+ "num_input_tokens_seen": 298533552,
+ "step": 6825,
+ "train_runtime": 52749.6034,
+ "train_tokens_per_second": 5659.446
+ },
+ {
+ "epoch": 0.4598707244815513,
+ "grad_norm": 0.7609318114113853,
+ "learning_rate": 8.7515763300159e-06,
+ "loss": 0.3463205575942993,
+ "num_input_tokens_seen": 298746696,
+ "step": 6830,
+ "train_runtime": 52789.3591,
+ "train_tokens_per_second": 5659.222
+ },
+ {
+ "epoch": 0.46020737947751145,
+ "grad_norm": 0.6930204184723102,
+ "learning_rate": 8.74982785523763e-06,
+ "loss": 0.40210418701171874,
+ "num_input_tokens_seen": 298972384,
+ "step": 6835,
+ "train_runtime": 52827.2223,
+ "train_tokens_per_second": 5659.438
+ },
+ {
+ "epoch": 0.4605440344734716,
+ "grad_norm": 0.7026299121313064,
+ "learning_rate": 8.748078331831957e-06,
+ "loss": 0.3934351682662964,
+ "num_input_tokens_seen": 299200112,
+ "step": 6840,
+ "train_runtime": 52862.9542,
+ "train_tokens_per_second": 5659.92
+ },
+ {
+ "epoch": 0.4608806894694317,
+ "grad_norm": 0.7019416941915204,
+ "learning_rate": 8.746327760288129e-06,
+ "loss": 0.3836989164352417,
+ "num_input_tokens_seen": 299437024,
+ "step": 6845,
+ "train_runtime": 52894.4779,
+ "train_tokens_per_second": 5661.026
+ },
+ {
+ "epoch": 0.46121734446539187,
+ "grad_norm": 0.8234006493355744,
+ "learning_rate": 8.744576141095691e-06,
+ "loss": 0.39553108215332033,
+ "num_input_tokens_seen": 299641752,
+ "step": 6850,
+ "train_runtime": 52932.8412,
+ "train_tokens_per_second": 5660.791
+ },
+ {
+ "epoch": 0.461553999461352,
+ "grad_norm": 0.846531534553998,
+ "learning_rate": 8.742823474744476e-06,
+ "loss": 0.39920735359191895,
+ "num_input_tokens_seen": 299852504,
+ "step": 6855,
+ "train_runtime": 52970.3403,
+ "train_tokens_per_second": 5660.762
+ },
+ {
+ "epoch": 0.46189065445731214,
+ "grad_norm": 0.7339872597188516,
+ "learning_rate": 8.74106976172461e-06,
+ "loss": 0.40232462882995607,
+ "num_input_tokens_seen": 300073192,
+ "step": 6860,
+ "train_runtime": 53013.6318,
+ "train_tokens_per_second": 5660.302
+ },
+ {
+ "epoch": 0.4622273094532723,
+ "grad_norm": 0.8137834490765111,
+ "learning_rate": 8.739315002526513e-06,
+ "loss": 0.40238103866577146,
+ "num_input_tokens_seen": 300287552,
+ "step": 6865,
+ "train_runtime": 53055.019,
+ "train_tokens_per_second": 5659.927
+ },
+ {
+ "epoch": 0.4625639644492324,
+ "grad_norm": 0.7042015061046794,
+ "learning_rate": 8.737559197640902e-06,
+ "loss": 0.3717548131942749,
+ "num_input_tokens_seen": 300496376,
+ "step": 6870,
+ "train_runtime": 53091.6564,
+ "train_tokens_per_second": 5659.955
+ },
+ {
+ "epoch": 0.46290061944519256,
+ "grad_norm": 0.7325334644495457,
+ "learning_rate": 8.735802347558778e-06,
+ "loss": 0.4027840614318848,
+ "num_input_tokens_seen": 300709416,
+ "step": 6875,
+ "train_runtime": 53135.1809,
+ "train_tokens_per_second": 5659.328
+ },
+ {
+ "epoch": 0.4632372744411527,
+ "grad_norm": 0.7655214239764673,
+ "learning_rate": 8.734044452771442e-06,
+ "loss": 0.38804852962493896,
+ "num_input_tokens_seen": 300918752,
+ "step": 6880,
+ "train_runtime": 53179.1079,
+ "train_tokens_per_second": 5658.59
+ },
+ {
+ "epoch": 0.46357392943711284,
+ "grad_norm": 0.8904799503928702,
+ "learning_rate": 8.73228551377048e-06,
+ "loss": 0.40845584869384766,
+ "num_input_tokens_seen": 301139720,
+ "step": 6885,
+ "train_runtime": 53208.9509,
+ "train_tokens_per_second": 5659.569
+ },
+ {
+ "epoch": 0.463910584433073,
+ "grad_norm": 0.6748289632604044,
+ "learning_rate": 8.730525531047776e-06,
+ "loss": 0.4137433052062988,
+ "num_input_tokens_seen": 301362752,
+ "step": 6890,
+ "train_runtime": 53254.1898,
+ "train_tokens_per_second": 5658.949
+ },
+ {
+ "epoch": 0.4642472394290331,
+ "grad_norm": 0.7035224245591345,
+ "learning_rate": 8.728764505095503e-06,
+ "loss": 0.4120682716369629,
+ "num_input_tokens_seen": 301587944,
+ "step": 6895,
+ "train_runtime": 53291.4984,
+ "train_tokens_per_second": 5659.213
+ },
+ {
+ "epoch": 0.46458389442499326,
+ "grad_norm": 0.6435345300249111,
+ "learning_rate": 8.727002436406126e-06,
+ "loss": 0.3724491834640503,
+ "num_input_tokens_seen": 301823392,
+ "step": 6900,
+ "train_runtime": 53331.8202,
+ "train_tokens_per_second": 5659.349
+ },
+ {
+ "epoch": 0.4649205494209534,
+ "grad_norm": 0.7100192481355228,
+ "learning_rate": 8.725239325472404e-06,
+ "loss": 0.4057363510131836,
+ "num_input_tokens_seen": 302046392,
+ "step": 6905,
+ "train_runtime": 53370.503,
+ "train_tokens_per_second": 5659.426
+ },
+ {
+ "epoch": 0.46525720441691354,
+ "grad_norm": 0.7998034880411997,
+ "learning_rate": 8.723475172787381e-06,
+ "loss": 0.3984755277633667,
+ "num_input_tokens_seen": 302270848,
+ "step": 6910,
+ "train_runtime": 53413.2206,
+ "train_tokens_per_second": 5659.102
+ },
+ {
+ "epoch": 0.4655938594128737,
+ "grad_norm": 0.9282661770223051,
+ "learning_rate": 8.7217099788444e-06,
+ "loss": 0.38761351108551023,
+ "num_input_tokens_seen": 302484424,
+ "step": 6915,
+ "train_runtime": 53452.6504,
+ "train_tokens_per_second": 5658.923
+ },
+ {
+ "epoch": 0.4659305144088338,
+ "grad_norm": 0.6829252086088465,
+ "learning_rate": 8.719943744137091e-06,
+ "loss": 0.3843079566955566,
+ "num_input_tokens_seen": 302706608,
+ "step": 6920,
+ "train_runtime": 53494.2752,
+ "train_tokens_per_second": 5658.673
+ },
+ {
+ "epoch": 0.46626716940479396,
+ "grad_norm": 0.7370197158639645,
+ "learning_rate": 8.718176469159378e-06,
+ "loss": 0.40040907859802244,
+ "num_input_tokens_seen": 302946080,
+ "step": 6925,
+ "train_runtime": 53527.8006,
+ "train_tokens_per_second": 5659.603
+ },
+ {
+ "epoch": 0.4666038244007541,
+ "grad_norm": 0.7744303086322805,
+ "learning_rate": 8.716408154405469e-06,
+ "loss": 0.370440411567688,
+ "num_input_tokens_seen": 303172472,
+ "step": 6930,
+ "train_runtime": 53558.5862,
+ "train_tokens_per_second": 5660.576
+ },
+ {
+ "epoch": 0.46694047939671424,
+ "grad_norm": 0.9149869901290688,
+ "learning_rate": 8.714638800369872e-06,
+ "loss": 0.4163699150085449,
+ "num_input_tokens_seen": 303374536,
+ "step": 6935,
+ "train_runtime": 53596.9127,
+ "train_tokens_per_second": 5660.299
+ },
+ {
+ "epoch": 0.4672771343926744,
+ "grad_norm": 0.6478644031670531,
+ "learning_rate": 8.712868407547378e-06,
+ "loss": 0.41262283325195315,
+ "num_input_tokens_seen": 303604704,
+ "step": 6940,
+ "train_runtime": 53634.4097,
+ "train_tokens_per_second": 5660.633
+ },
+ {
+ "epoch": 0.4676137893886345,
+ "grad_norm": 0.8540394960614904,
+ "learning_rate": 8.711096976433076e-06,
+ "loss": 0.40007991790771485,
+ "num_input_tokens_seen": 303830584,
+ "step": 6945,
+ "train_runtime": 53675.8336,
+ "train_tokens_per_second": 5660.473
+ },
+ {
+ "epoch": 0.46795044438459465,
+ "grad_norm": 0.6854573080771246,
+ "learning_rate": 8.709324507522337e-06,
+ "loss": 0.3609914779663086,
+ "num_input_tokens_seen": 304051832,
+ "step": 6950,
+ "train_runtime": 53710.0276,
+ "train_tokens_per_second": 5660.988
+ },
+ {
+ "epoch": 0.4682870993805548,
+ "grad_norm": 0.74529061138361,
+ "learning_rate": 8.707551001310828e-06,
+ "loss": 0.3842365264892578,
+ "num_input_tokens_seen": 304279352,
+ "step": 6955,
+ "train_runtime": 53749.8544,
+ "train_tokens_per_second": 5661.027
+ },
+ {
+ "epoch": 0.46862375437651493,
+ "grad_norm": 0.7026982064246363,
+ "learning_rate": 8.705776458294503e-06,
+ "loss": 0.387627911567688,
+ "num_input_tokens_seen": 304487200,
+ "step": 6960,
+ "train_runtime": 53785.1129,
+ "train_tokens_per_second": 5661.18
+ },
+ {
+ "epoch": 0.4689604093724751,
+ "grad_norm": 0.7214247021583069,
+ "learning_rate": 8.704000878969608e-06,
+ "loss": 0.37026023864746094,
+ "num_input_tokens_seen": 304693840,
+ "step": 6965,
+ "train_runtime": 53822.9634,
+ "train_tokens_per_second": 5661.038
+ },
+ {
+ "epoch": 0.4692970643684352,
+ "grad_norm": 0.6979359178232242,
+ "learning_rate": 8.702224263832679e-06,
+ "loss": 0.3628101825714111,
+ "num_input_tokens_seen": 304921392,
+ "step": 6970,
+ "train_runtime": 53861.4766,
+ "train_tokens_per_second": 5661.215
+ },
+ {
+ "epoch": 0.46963371936439535,
+ "grad_norm": 0.7813572135644429,
+ "learning_rate": 8.700446613380542e-06,
+ "loss": 0.4275547504425049,
+ "num_input_tokens_seen": 305149136,
+ "step": 6975,
+ "train_runtime": 53899.1275,
+ "train_tokens_per_second": 5661.486
+ },
+ {
+ "epoch": 0.4699703743603555,
+ "grad_norm": 0.800833875206026,
+ "learning_rate": 8.698667928110307e-06,
+ "loss": 0.373789119720459,
+ "num_input_tokens_seen": 305369336,
+ "step": 6980,
+ "train_runtime": 53936.8451,
+ "train_tokens_per_second": 5661.609
+ },
+ {
+ "epoch": 0.47030702935631563,
+ "grad_norm": 0.8980829202572613,
+ "learning_rate": 8.696888208519381e-06,
+ "loss": 0.3973226070404053,
+ "num_input_tokens_seen": 305581504,
+ "step": 6985,
+ "train_runtime": 53973.4435,
+ "train_tokens_per_second": 5661.701
+ },
+ {
+ "epoch": 0.47064368435227577,
+ "grad_norm": 0.7351189903914839,
+ "learning_rate": 8.695107455105454e-06,
+ "loss": 0.34683785438537595,
+ "num_input_tokens_seen": 305790632,
+ "step": 6990,
+ "train_runtime": 54006.1459,
+ "train_tokens_per_second": 5662.145
+ },
+ {
+ "epoch": 0.4709803393482359,
+ "grad_norm": 0.7928821233878864,
+ "learning_rate": 8.693325668366513e-06,
+ "loss": 0.39075660705566406,
+ "num_input_tokens_seen": 306004232,
+ "step": 6995,
+ "train_runtime": 54042.3536,
+ "train_tokens_per_second": 5662.304
+ },
+ {
+ "epoch": 0.47131699434419605,
+ "grad_norm": 0.8527776847132168,
+ "learning_rate": 8.691542848800825e-06,
+ "loss": 0.39600386619567873,
+ "num_input_tokens_seen": 306212208,
+ "step": 7000,
+ "train_runtime": 54084.5742,
+ "train_tokens_per_second": 5661.729
+ },
+ {
+ "epoch": 0.4716536493401562,
+ "grad_norm": 0.8808374134194763,
+ "learning_rate": 8.68975899690695e-06,
+ "loss": 0.38110928535461425,
+ "num_input_tokens_seen": 306437744,
+ "step": 7005,
+ "train_runtime": 54125.0428,
+ "train_tokens_per_second": 5661.663
+ },
+ {
+ "epoch": 0.47199030433611633,
+ "grad_norm": 0.7423709319863485,
+ "learning_rate": 8.687974113183738e-06,
+ "loss": 0.37776660919189453,
+ "num_input_tokens_seen": 306667400,
+ "step": 7010,
+ "train_runtime": 54160.5555,
+ "train_tokens_per_second": 5662.191
+ },
+ {
+ "epoch": 0.47232695933207647,
+ "grad_norm": 0.8031207862343954,
+ "learning_rate": 8.686188198130326e-06,
+ "loss": 0.3893289089202881,
+ "num_input_tokens_seen": 306891232,
+ "step": 7015,
+ "train_runtime": 54195.218,
+ "train_tokens_per_second": 5662.699
+ },
+ {
+ "epoch": 0.4726636143280366,
+ "grad_norm": 0.716418279114708,
+ "learning_rate": 8.684401252246138e-06,
+ "loss": 0.3991250038146973,
+ "num_input_tokens_seen": 307116336,
+ "step": 7020,
+ "train_runtime": 54233.7853,
+ "train_tokens_per_second": 5662.823
+ },
+ {
+ "epoch": 0.47300026932399675,
+ "grad_norm": 0.7731462314856246,
+ "learning_rate": 8.68261327603089e-06,
+ "loss": 0.42037010192871094,
+ "num_input_tokens_seen": 307328112,
+ "step": 7025,
+ "train_runtime": 54273.3793,
+ "train_tokens_per_second": 5662.594
+ },
+ {
+ "epoch": 0.4733369243199569,
+ "grad_norm": 0.7619151144650879,
+ "learning_rate": 8.68082426998458e-06,
+ "loss": 0.42117109298706057,
+ "num_input_tokens_seen": 307558384,
+ "step": 7030,
+ "train_runtime": 54315.1352,
+ "train_tokens_per_second": 5662.48
+ },
+ {
+ "epoch": 0.473673579315917,
+ "grad_norm": 0.8366328207039446,
+ "learning_rate": 8.6790342346075e-06,
+ "loss": 0.355437707901001,
+ "num_input_tokens_seen": 307757832,
+ "step": 7035,
+ "train_runtime": 54355.0453,
+ "train_tokens_per_second": 5661.992
+ },
+ {
+ "epoch": 0.47401023431187717,
+ "grad_norm": 0.694627660462694,
+ "learning_rate": 8.677243170400228e-06,
+ "loss": 0.3744319438934326,
+ "num_input_tokens_seen": 307964848,
+ "step": 7040,
+ "train_runtime": 54392.4804,
+ "train_tokens_per_second": 5661.901
+ },
+ {
+ "epoch": 0.4743468893078373,
+ "grad_norm": 0.7593548972261533,
+ "learning_rate": 8.675451077863632e-06,
+ "loss": 0.40117859840393066,
+ "num_input_tokens_seen": 308180176,
+ "step": 7045,
+ "train_runtime": 54429.867,
+ "train_tokens_per_second": 5661.968
+ },
+ {
+ "epoch": 0.47468354430379744,
+ "grad_norm": 0.7788878005901869,
+ "learning_rate": 8.67365795749886e-06,
+ "loss": 0.37619283199310305,
+ "num_input_tokens_seen": 308393256,
+ "step": 7050,
+ "train_runtime": 54475.6142,
+ "train_tokens_per_second": 5661.125
+ },
+ {
+ "epoch": 0.4750201992997576,
+ "grad_norm": 0.6935999303413757,
+ "learning_rate": 8.671863809807355e-06,
+ "loss": 0.36246466636657715,
+ "num_input_tokens_seen": 308598120,
+ "step": 7055,
+ "train_runtime": 54512.6696,
+ "train_tokens_per_second": 5661.035
+ },
+ {
+ "epoch": 0.4753568542957177,
+ "grad_norm": 0.7796482197914519,
+ "learning_rate": 8.670068635290844e-06,
+ "loss": 0.3815620899200439,
+ "num_input_tokens_seen": 308814312,
+ "step": 7060,
+ "train_runtime": 54553.6102,
+ "train_tokens_per_second": 5660.749
+ },
+ {
+ "epoch": 0.47569350929167786,
+ "grad_norm": 0.6394547128504924,
+ "learning_rate": 8.668272434451343e-06,
+ "loss": 0.3547971725463867,
+ "num_input_tokens_seen": 309036648,
+ "step": 7065,
+ "train_runtime": 54589.1907,
+ "train_tokens_per_second": 5661.133
+ },
+ {
+ "epoch": 0.476030164287638,
+ "grad_norm": 0.8424232555426785,
+ "learning_rate": 8.666475207791154e-06,
+ "loss": 0.3938549518585205,
+ "num_input_tokens_seen": 309237312,
+ "step": 7070,
+ "train_runtime": 54626.0249,
+ "train_tokens_per_second": 5660.989
+ },
+ {
+ "epoch": 0.47636681928359814,
+ "grad_norm": 0.7156676868042459,
+ "learning_rate": 8.664676955812863e-06,
+ "loss": 0.3805175065994263,
+ "num_input_tokens_seen": 309461072,
+ "step": 7075,
+ "train_runtime": 54665.5373,
+ "train_tokens_per_second": 5660.99
+ },
+ {
+ "epoch": 0.4767034742795583,
+ "grad_norm": 0.7803230987384645,
+ "learning_rate": 8.662877679019348e-06,
+ "loss": 0.38736543655395506,
+ "num_input_tokens_seen": 309676264,
+ "step": 7080,
+ "train_runtime": 54702.5852,
+ "train_tokens_per_second": 5661.09
+ },
+ {
+ "epoch": 0.4770401292755185,
+ "grad_norm": 0.811046251111573,
+ "learning_rate": 8.66107737791377e-06,
+ "loss": 0.40035252571105956,
+ "num_input_tokens_seen": 309908584,
+ "step": 7085,
+ "train_runtime": 54739.8422,
+ "train_tokens_per_second": 5661.481
+ },
+ {
+ "epoch": 0.4773767842714786,
+ "grad_norm": 0.8088100128547995,
+ "learning_rate": 8.659276052999574e-06,
+ "loss": 0.37158200740814207,
+ "num_input_tokens_seen": 310109536,
+ "step": 7090,
+ "train_runtime": 54784.3231,
+ "train_tokens_per_second": 5660.552
+ },
+ {
+ "epoch": 0.47771343926743876,
+ "grad_norm": 0.6607865351139958,
+ "learning_rate": 8.657473704780504e-06,
+ "loss": 0.3958035707473755,
+ "num_input_tokens_seen": 310316416,
+ "step": 7095,
+ "train_runtime": 54823.0111,
+ "train_tokens_per_second": 5660.331
+ },
+ {
+ "epoch": 0.4780500942633989,
+ "grad_norm": 0.7508776157927014,
+ "learning_rate": 8.65567033376057e-06,
+ "loss": 0.3861806392669678,
+ "num_input_tokens_seen": 310543280,
+ "step": 7100,
+ "train_runtime": 54858.1842,
+ "train_tokens_per_second": 5660.838
+ },
+ {
+ "epoch": 0.47838674925935903,
+ "grad_norm": 0.6796527894531441,
+ "learning_rate": 8.65386594044409e-06,
+ "loss": 0.38321571350097655,
+ "num_input_tokens_seen": 310767656,
+ "step": 7105,
+ "train_runtime": 54899.3833,
+ "train_tokens_per_second": 5660.677
+ },
+ {
+ "epoch": 0.4787234042553192,
+ "grad_norm": 0.7839875292722721,
+ "learning_rate": 8.652060525335647e-06,
+ "loss": 0.35958447456359866,
+ "num_input_tokens_seen": 310977648,
+ "step": 7110,
+ "train_runtime": 54935.0727,
+ "train_tokens_per_second": 5660.822
+ },
+ {
+ "epoch": 0.4790600592512793,
+ "grad_norm": 0.6805039530532168,
+ "learning_rate": 8.650254088940125e-06,
+ "loss": 0.37933621406555174,
+ "num_input_tokens_seen": 311184472,
+ "step": 7115,
+ "train_runtime": 54970.8187,
+ "train_tokens_per_second": 5660.903
+ },
+ {
+ "epoch": 0.47939671424723945,
+ "grad_norm": 0.8690524861477827,
+ "learning_rate": 8.648446631762686e-06,
+ "loss": 0.3870699882507324,
+ "num_input_tokens_seen": 311386608,
+ "step": 7120,
+ "train_runtime": 55007.3848,
+ "train_tokens_per_second": 5660.815
+ },
+ {
+ "epoch": 0.4797333692431996,
+ "grad_norm": 0.7834933561721031,
+ "learning_rate": 8.646638154308781e-06,
+ "loss": 0.3672234296798706,
+ "num_input_tokens_seen": 311600392,
+ "step": 7125,
+ "train_runtime": 55045.8437,
+ "train_tokens_per_second": 5660.743
+ },
+ {
+ "epoch": 0.48007002423915973,
+ "grad_norm": 0.6821280974528945,
+ "learning_rate": 8.644828657084144e-06,
+ "loss": 0.3536520004272461,
+ "num_input_tokens_seen": 311826176,
+ "step": 7130,
+ "train_runtime": 55086.3186,
+ "train_tokens_per_second": 5660.683
+ },
+ {
+ "epoch": 0.48040667923511987,
+ "grad_norm": 0.7090523007119766,
+ "learning_rate": 8.643018140594797e-06,
+ "loss": 0.35062260627746583,
+ "num_input_tokens_seen": 312040616,
+ "step": 7135,
+ "train_runtime": 55118.7948,
+ "train_tokens_per_second": 5661.238
+ },
+ {
+ "epoch": 0.48074333423108,
+ "grad_norm": 0.7694829828758608,
+ "learning_rate": 8.641206605347042e-06,
+ "loss": 0.3680727958679199,
+ "num_input_tokens_seen": 312250776,
+ "step": 7140,
+ "train_runtime": 55154.654,
+ "train_tokens_per_second": 5661.368
+ },
+ {
+ "epoch": 0.48107998922704015,
+ "grad_norm": 0.7834141310420747,
+ "learning_rate": 8.639394051847472e-06,
+ "loss": 0.3820206642150879,
+ "num_input_tokens_seen": 312469768,
+ "step": 7145,
+ "train_runtime": 55194.213,
+ "train_tokens_per_second": 5661.278
+ },
+ {
+ "epoch": 0.4814166442230003,
+ "grad_norm": 0.8086494042960466,
+ "learning_rate": 8.63758048060296e-06,
+ "loss": 0.3749295473098755,
+ "num_input_tokens_seen": 312686816,
+ "step": 7150,
+ "train_runtime": 55234.7643,
+ "train_tokens_per_second": 5661.051
+ },
+ {
+ "epoch": 0.48175329921896043,
+ "grad_norm": 1.146252827120329,
+ "learning_rate": 8.635765892120666e-06,
+ "loss": 0.35863730907440183,
+ "num_input_tokens_seen": 312928240,
+ "step": 7155,
+ "train_runtime": 55276.3406,
+ "train_tokens_per_second": 5661.161
+ },
+ {
+ "epoch": 0.48208995421492057,
+ "grad_norm": 0.6322723599846541,
+ "learning_rate": 8.633950286908032e-06,
+ "loss": 0.3957623243331909,
+ "num_input_tokens_seen": 313146400,
+ "step": 7160,
+ "train_runtime": 55318.9273,
+ "train_tokens_per_second": 5660.746
+ },
+ {
+ "epoch": 0.4824266092108807,
+ "grad_norm": 0.8242447392294052,
+ "learning_rate": 8.63213366547279e-06,
+ "loss": 0.41290960311889646,
+ "num_input_tokens_seen": 313353080,
+ "step": 7165,
+ "train_runtime": 55354.3316,
+ "train_tokens_per_second": 5660.859
+ },
+ {
+ "epoch": 0.48276326420684085,
+ "grad_norm": 0.8189218668862145,
+ "learning_rate": 8.630316028322952e-06,
+ "loss": 0.3599478960037231,
+ "num_input_tokens_seen": 313556520,
+ "step": 7170,
+ "train_runtime": 55393.1558,
+ "train_tokens_per_second": 5660.564
+ },
+ {
+ "epoch": 0.483099919202801,
+ "grad_norm": 0.8049613617878884,
+ "learning_rate": 8.628497375966811e-06,
+ "loss": 0.3688513278961182,
+ "num_input_tokens_seen": 313791304,
+ "step": 7175,
+ "train_runtime": 55433.9631,
+ "train_tokens_per_second": 5660.633
+ },
+ {
+ "epoch": 0.4834365741987611,
+ "grad_norm": 0.7403145334116261,
+ "learning_rate": 8.62667770891295e-06,
+ "loss": 0.4179924488067627,
+ "num_input_tokens_seen": 314025432,
+ "step": 7180,
+ "train_runtime": 55479.2927,
+ "train_tokens_per_second": 5660.228
+ },
+ {
+ "epoch": 0.48377322919472127,
+ "grad_norm": 0.6161204381365467,
+ "learning_rate": 8.624857027670232e-06,
+ "loss": 0.3718327522277832,
+ "num_input_tokens_seen": 314253744,
+ "step": 7185,
+ "train_runtime": 55518.1876,
+ "train_tokens_per_second": 5660.375
+ },
+ {
+ "epoch": 0.4841098841906814,
+ "grad_norm": 0.6673541170021607,
+ "learning_rate": 8.623035332747804e-06,
+ "loss": 0.4232837677001953,
+ "num_input_tokens_seen": 314457360,
+ "step": 7190,
+ "train_runtime": 55555.7684,
+ "train_tokens_per_second": 5660.211
+ },
+ {
+ "epoch": 0.48444653918664154,
+ "grad_norm": 0.8639989632411573,
+ "learning_rate": 8.6212126246551e-06,
+ "loss": 0.35295858383178713,
+ "num_input_tokens_seen": 314659296,
+ "step": 7195,
+ "train_runtime": 55590.9252,
+ "train_tokens_per_second": 5660.264
+ },
+ {
+ "epoch": 0.4847831941826017,
+ "grad_norm": 0.7194371682580046,
+ "learning_rate": 8.619388903901833e-06,
+ "loss": 0.397071385383606,
+ "num_input_tokens_seen": 314886400,
+ "step": 7200,
+ "train_runtime": 55632.7504,
+ "train_tokens_per_second": 5660.09
+ },
+ {
+ "epoch": 0.4851198491785618,
+ "grad_norm": 0.7895653948671645,
+ "learning_rate": 8.617564170997998e-06,
+ "loss": 0.393186354637146,
+ "num_input_tokens_seen": 315113560,
+ "step": 7205,
+ "train_runtime": 55665.6693,
+ "train_tokens_per_second": 5660.824
+ },
+ {
+ "epoch": 0.48545650417452196,
+ "grad_norm": 0.6916492568995722,
+ "learning_rate": 8.61573842645388e-06,
+ "loss": 0.38975393772125244,
+ "num_input_tokens_seen": 315339760,
+ "step": 7210,
+ "train_runtime": 55701.6461,
+ "train_tokens_per_second": 5661.229
+ },
+ {
+ "epoch": 0.4857931591704821,
+ "grad_norm": 0.7125114232933183,
+ "learning_rate": 8.61391167078004e-06,
+ "loss": 0.3773904085159302,
+ "num_input_tokens_seen": 315564560,
+ "step": 7215,
+ "train_runtime": 55739.932,
+ "train_tokens_per_second": 5661.373
+ },
+ {
+ "epoch": 0.48612981416644224,
+ "grad_norm": 0.7816578333674177,
+ "learning_rate": 8.612083904487324e-06,
+ "loss": 0.4223905563354492,
+ "num_input_tokens_seen": 315780664,
+ "step": 7220,
+ "train_runtime": 55782.1816,
+ "train_tokens_per_second": 5660.959
+ },
+ {
+ "epoch": 0.4864664691624024,
+ "grad_norm": 0.8225392117617899,
+ "learning_rate": 8.610255128086867e-06,
+ "loss": 0.39981255531311033,
+ "num_input_tokens_seen": 315985224,
+ "step": 7225,
+ "train_runtime": 55817.5809,
+ "train_tokens_per_second": 5661.034
+ },
+ {
+ "epoch": 0.4868031241583625,
+ "grad_norm": 0.7192041520493799,
+ "learning_rate": 8.60842534209007e-06,
+ "loss": 0.35159389972686766,
+ "num_input_tokens_seen": 316206120,
+ "step": 7230,
+ "train_runtime": 55856.1823,
+ "train_tokens_per_second": 5661.076
+ },
+ {
+ "epoch": 0.48713977915432266,
+ "grad_norm": 0.7538381499748801,
+ "learning_rate": 8.606594547008636e-06,
+ "loss": 0.33089404106140136,
+ "num_input_tokens_seen": 316428112,
+ "step": 7235,
+ "train_runtime": 55889.528,
+ "train_tokens_per_second": 5661.671
+ },
+ {
+ "epoch": 0.4874764341502828,
+ "grad_norm": 0.7734857420248509,
+ "learning_rate": 8.604762743354536e-06,
+ "loss": 0.39783239364624023,
+ "num_input_tokens_seen": 316666376,
+ "step": 7240,
+ "train_runtime": 55922.8766,
+ "train_tokens_per_second": 5662.555
+ },
+ {
+ "epoch": 0.48781308914624294,
+ "grad_norm": 0.7475280321551805,
+ "learning_rate": 8.602929931640029e-06,
+ "loss": 0.3995728254318237,
+ "num_input_tokens_seen": 316868832,
+ "step": 7245,
+ "train_runtime": 55959.5907,
+ "train_tokens_per_second": 5662.458
+ },
+ {
+ "epoch": 0.4881497441422031,
+ "grad_norm": 0.8349680082803242,
+ "learning_rate": 8.601096112377659e-06,
+ "loss": 0.4125559329986572,
+ "num_input_tokens_seen": 317082248,
+ "step": 7250,
+ "train_runtime": 56002.6917,
+ "train_tokens_per_second": 5661.911
+ },
+ {
+ "epoch": 0.4884863991381632,
+ "grad_norm": 0.6943776304621931,
+ "learning_rate": 8.59926128608024e-06,
+ "loss": 0.3738250255584717,
+ "num_input_tokens_seen": 317303120,
+ "step": 7255,
+ "train_runtime": 56039.6874,
+ "train_tokens_per_second": 5662.114
+ },
+ {
+ "epoch": 0.48882305413412336,
+ "grad_norm": 0.7789443419004642,
+ "learning_rate": 8.597425453260881e-06,
+ "loss": 0.3914709806442261,
+ "num_input_tokens_seen": 317515672,
+ "step": 7260,
+ "train_runtime": 56082.3268,
+ "train_tokens_per_second": 5661.599
+ },
+ {
+ "epoch": 0.4891597091300835,
+ "grad_norm": 0.8422191416131971,
+ "learning_rate": 8.595588614432967e-06,
+ "loss": 0.3669420719146729,
+ "num_input_tokens_seen": 317720352,
+ "step": 7265,
+ "train_runtime": 56128.1181,
+ "train_tokens_per_second": 5660.627
+ },
+ {
+ "epoch": 0.48949636412604364,
+ "grad_norm": 0.7163712412405564,
+ "learning_rate": 8.59375077011016e-06,
+ "loss": 0.36806163787841795,
+ "num_input_tokens_seen": 317948384,
+ "step": 7270,
+ "train_runtime": 56167.6356,
+ "train_tokens_per_second": 5660.704
+ },
+ {
+ "epoch": 0.4898330191220038,
+ "grad_norm": 0.8174970324207671,
+ "learning_rate": 8.591911920806412e-06,
+ "loss": 0.39014444351196287,
+ "num_input_tokens_seen": 318179072,
+ "step": 7275,
+ "train_runtime": 56213.9579,
+ "train_tokens_per_second": 5660.144
+ },
+ {
+ "epoch": 0.4901696741179639,
+ "grad_norm": 0.6974694369606549,
+ "learning_rate": 8.590072067035947e-06,
+ "loss": 0.38969347476959226,
+ "num_input_tokens_seen": 318410480,
+ "step": 7280,
+ "train_runtime": 56255.4338,
+ "train_tokens_per_second": 5660.084
+ },
+ {
+ "epoch": 0.49050632911392406,
+ "grad_norm": 0.6692328837890084,
+ "learning_rate": 8.588231209313278e-06,
+ "loss": 0.37391541004180906,
+ "num_input_tokens_seen": 318654632,
+ "step": 7285,
+ "train_runtime": 56291.6831,
+ "train_tokens_per_second": 5660.776
+ },
+ {
+ "epoch": 0.4908429841098842,
+ "grad_norm": 0.7608072412958533,
+ "learning_rate": 8.586389348153192e-06,
+ "loss": 0.3896606206893921,
+ "num_input_tokens_seen": 318874824,
+ "step": 7290,
+ "train_runtime": 56326.4539,
+ "train_tokens_per_second": 5661.191
+ },
+ {
+ "epoch": 0.49117963910584433,
+ "grad_norm": 0.7736109340706484,
+ "learning_rate": 8.584546484070762e-06,
+ "loss": 0.40265254974365233,
+ "num_input_tokens_seen": 319084200,
+ "step": 7295,
+ "train_runtime": 56365.3177,
+ "train_tokens_per_second": 5661.002
+ },
+ {
+ "epoch": 0.4915162941018045,
+ "grad_norm": 0.7875873984707813,
+ "learning_rate": 8.582702617581338e-06,
+ "loss": 0.37858309745788576,
+ "num_input_tokens_seen": 319307784,
+ "step": 7300,
+ "train_runtime": 56403.5288,
+ "train_tokens_per_second": 5661.131
+ },
+ {
+ "epoch": 0.4918529490977646,
+ "grad_norm": 0.7183058860492505,
+ "learning_rate": 8.58085774920055e-06,
+ "loss": 0.38918490409851075,
+ "num_input_tokens_seen": 319513304,
+ "step": 7305,
+ "train_runtime": 56438.5945,
+ "train_tokens_per_second": 5661.256
+ },
+ {
+ "epoch": 0.49218960409372475,
+ "grad_norm": 0.7356534041270457,
+ "learning_rate": 8.579011879444313e-06,
+ "loss": 0.414321231842041,
+ "num_input_tokens_seen": 319743560,
+ "step": 7310,
+ "train_runtime": 56483.4424,
+ "train_tokens_per_second": 5660.837
+ },
+ {
+ "epoch": 0.4925262590896849,
+ "grad_norm": 0.6584093897897996,
+ "learning_rate": 8.577165008828817e-06,
+ "loss": 0.34825422763824465,
+ "num_input_tokens_seen": 319947112,
+ "step": 7315,
+ "train_runtime": 56521.6053,
+ "train_tokens_per_second": 5660.616
+ },
+ {
+ "epoch": 0.49286291408564503,
+ "grad_norm": 0.7722746099858304,
+ "learning_rate": 8.57531713787053e-06,
+ "loss": 0.3858429193496704,
+ "num_input_tokens_seen": 320155920,
+ "step": 7320,
+ "train_runtime": 56558.1587,
+ "train_tokens_per_second": 5660.65
+ },
+ {
+ "epoch": 0.49319956908160517,
+ "grad_norm": 0.8389191394295502,
+ "learning_rate": 8.573468267086208e-06,
+ "loss": 0.40245862007141114,
+ "num_input_tokens_seen": 320372792,
+ "step": 7325,
+ "train_runtime": 56592.9849,
+ "train_tokens_per_second": 5660.998
+ },
+ {
+ "epoch": 0.4935362240775653,
+ "grad_norm": 0.7633923302039957,
+ "learning_rate": 8.57161839699288e-06,
+ "loss": 0.36266283988952636,
+ "num_input_tokens_seen": 320592160,
+ "step": 7330,
+ "train_runtime": 56631.6072,
+ "train_tokens_per_second": 5661.011
+ },
+ {
+ "epoch": 0.49387287907352545,
+ "grad_norm": 0.8780389979845686,
+ "learning_rate": 8.569767528107857e-06,
+ "loss": 0.3856183052062988,
+ "num_input_tokens_seen": 320807512,
+ "step": 7335,
+ "train_runtime": 56669.1815,
+ "train_tokens_per_second": 5661.058
+ },
+ {
+ "epoch": 0.4942095340694856,
+ "grad_norm": 0.6800471334043537,
+ "learning_rate": 8.567915660948727e-06,
+ "loss": 0.3747032880783081,
+ "num_input_tokens_seen": 321031440,
+ "step": 7340,
+ "train_runtime": 56709.5179,
+ "train_tokens_per_second": 5660.98
+ },
+ {
+ "epoch": 0.49454618906544573,
+ "grad_norm": 0.8334581237618652,
+ "learning_rate": 8.566062796033358e-06,
+ "loss": 0.4063011646270752,
+ "num_input_tokens_seen": 321244008,
+ "step": 7345,
+ "train_runtime": 56746.898,
+ "train_tokens_per_second": 5660.997
+ },
+ {
+ "epoch": 0.49488284406140587,
+ "grad_norm": 0.7667199760691221,
+ "learning_rate": 8.564208933879903e-06,
+ "loss": 0.37364883422851564,
+ "num_input_tokens_seen": 321456992,
+ "step": 7350,
+ "train_runtime": 56787.18,
+ "train_tokens_per_second": 5660.732
+ },
+ {
+ "epoch": 0.495219499057366,
+ "grad_norm": 0.7580149996088524,
+ "learning_rate": 8.562354075006782e-06,
+ "loss": 0.3439052104949951,
+ "num_input_tokens_seen": 321675512,
+ "step": 7355,
+ "train_runtime": 56825.1931,
+ "train_tokens_per_second": 5660.79
+ },
+ {
+ "epoch": 0.49555615405332615,
+ "grad_norm": 0.6724287776480686,
+ "learning_rate": 8.560498219932706e-06,
+ "loss": 0.4257340431213379,
+ "num_input_tokens_seen": 321898448,
+ "step": 7360,
+ "train_runtime": 56859.2761,
+ "train_tokens_per_second": 5661.318
+ },
+ {
+ "epoch": 0.4958928090492863,
+ "grad_norm": 0.735779732523222,
+ "learning_rate": 8.558641369176654e-06,
+ "loss": 0.33791351318359375,
+ "num_input_tokens_seen": 322099904,
+ "step": 7365,
+ "train_runtime": 56892.2085,
+ "train_tokens_per_second": 5661.582
+ },
+ {
+ "epoch": 0.4962294640452464,
+ "grad_norm": 0.7979574200557189,
+ "learning_rate": 8.556783523257892e-06,
+ "loss": 0.3972716569900513,
+ "num_input_tokens_seen": 322310456,
+ "step": 7370,
+ "train_runtime": 56935.1565,
+ "train_tokens_per_second": 5661.009
+ },
+ {
+ "epoch": 0.49656611904120657,
+ "grad_norm": 0.7637866611420476,
+ "learning_rate": 8.554924682695959e-06,
+ "loss": 0.3530308246612549,
+ "num_input_tokens_seen": 322518744,
+ "step": 7375,
+ "train_runtime": 56971.9645,
+ "train_tokens_per_second": 5661.008
+ },
+ {
+ "epoch": 0.4969027740371667,
+ "grad_norm": 0.7640709256586813,
+ "learning_rate": 8.553064848010677e-06,
+ "loss": 0.3955545902252197,
+ "num_input_tokens_seen": 322725520,
+ "step": 7380,
+ "train_runtime": 57007.6953,
+ "train_tokens_per_second": 5661.087
+ },
+ {
+ "epoch": 0.49723942903312685,
+ "grad_norm": 0.8082539381866035,
+ "learning_rate": 8.551204019722137e-06,
+ "loss": 0.37126519680023196,
+ "num_input_tokens_seen": 322938184,
+ "step": 7385,
+ "train_runtime": 57046.2643,
+ "train_tokens_per_second": 5660.987
+ },
+ {
+ "epoch": 0.497576084029087,
+ "grad_norm": 0.7610367285948725,
+ "learning_rate": 8.549342198350718e-06,
+ "loss": 0.40738329887390134,
+ "num_input_tokens_seen": 323160704,
+ "step": 7390,
+ "train_runtime": 57086.0056,
+ "train_tokens_per_second": 5660.944
+ },
+ {
+ "epoch": 0.4979127390250471,
+ "grad_norm": 0.7004854714973048,
+ "learning_rate": 8.54747938441707e-06,
+ "loss": 0.36683225631713867,
+ "num_input_tokens_seen": 323382024,
+ "step": 7395,
+ "train_runtime": 57128.4646,
+ "train_tokens_per_second": 5660.611
+ },
+ {
+ "epoch": 0.49824939402100726,
+ "grad_norm": 0.6981945654130406,
+ "learning_rate": 8.545615578442126e-06,
+ "loss": 0.35772018432617186,
+ "num_input_tokens_seen": 323601056,
+ "step": 7400,
+ "train_runtime": 57166.4275,
+ "train_tokens_per_second": 5660.684
+ },
+ {
+ "epoch": 0.4985860490169674,
+ "grad_norm": 0.8538725530295028,
+ "learning_rate": 8.54375078094709e-06,
+ "loss": 0.3746786117553711,
+ "num_input_tokens_seen": 323824464,
+ "step": 7405,
+ "train_runtime": 57216.1649,
+ "train_tokens_per_second": 5659.667
+ },
+ {
+ "epoch": 0.49892270401292754,
+ "grad_norm": 0.6772326544260556,
+ "learning_rate": 8.541884992453449e-06,
+ "loss": 0.3711742401123047,
+ "num_input_tokens_seen": 324047472,
+ "step": 7410,
+ "train_runtime": 57254.2655,
+ "train_tokens_per_second": 5659.796
+ },
+ {
+ "epoch": 0.4992593590088877,
+ "grad_norm": 0.6789887270206421,
+ "learning_rate": 8.540018213482966e-06,
+ "loss": 0.37660138607025145,
+ "num_input_tokens_seen": 324280688,
+ "step": 7415,
+ "train_runtime": 57294.4005,
+ "train_tokens_per_second": 5659.902
+ },
+ {
+ "epoch": 0.4995960140048478,
+ "grad_norm": 0.7605591952463279,
+ "learning_rate": 8.538150444557676e-06,
+ "loss": 0.3920082330703735,
+ "num_input_tokens_seen": 324488104,
+ "step": 7420,
+ "train_runtime": 57333.8655,
+ "train_tokens_per_second": 5659.624
+ },
+ {
+ "epoch": 0.49993266900080796,
+ "grad_norm": 0.826315107413246,
+ "learning_rate": 8.536281686199896e-06,
+ "loss": 0.385677433013916,
+ "num_input_tokens_seen": 324711216,
+ "step": 7425,
+ "train_runtime": 57366.6278,
+ "train_tokens_per_second": 5660.281
+ },
+ {
+ "epoch": 0.5002693239967682,
+ "grad_norm": 0.6877893318576644,
+ "learning_rate": 8.534411938932218e-06,
+ "loss": 0.3787369728088379,
+ "num_input_tokens_seen": 324939096,
+ "step": 7430,
+ "train_runtime": 57400.8374,
+ "train_tokens_per_second": 5660.877
+ },
+ {
+ "epoch": 0.5006059789927283,
+ "grad_norm": 0.7206196203304496,
+ "learning_rate": 8.532541203277515e-06,
+ "loss": 0.3697648525238037,
+ "num_input_tokens_seen": 325163640,
+ "step": 7435,
+ "train_runtime": 57437.2769,
+ "train_tokens_per_second": 5661.195
+ },
+ {
+ "epoch": 0.5009426339886884,
+ "grad_norm": 0.8219400190681228,
+ "learning_rate": 8.530669479758926e-06,
+ "loss": 0.37598912715911864,
+ "num_input_tokens_seen": 325383192,
+ "step": 7440,
+ "train_runtime": 57475.956,
+ "train_tokens_per_second": 5661.205
+ },
+ {
+ "epoch": 0.5012792889846486,
+ "grad_norm": 0.8185284831231338,
+ "learning_rate": 8.528796768899878e-06,
+ "loss": 0.39255619049072266,
+ "num_input_tokens_seen": 325585216,
+ "step": 7445,
+ "train_runtime": 57516.4354,
+ "train_tokens_per_second": 5660.734
+ },
+ {
+ "epoch": 0.5016159439806087,
+ "grad_norm": 0.748875770223375,
+ "learning_rate": 8.526923071224064e-06,
+ "loss": 0.39842636585235597,
+ "num_input_tokens_seen": 325810472,
+ "step": 7450,
+ "train_runtime": 57549.4748,
+ "train_tokens_per_second": 5661.398
+ },
+ {
+ "epoch": 0.5019525989765689,
+ "grad_norm": 0.7148185230181737,
+ "learning_rate": 8.525048387255461e-06,
+ "loss": 0.3887434482574463,
+ "num_input_tokens_seen": 326034504,
+ "step": 7455,
+ "train_runtime": 57590.0671,
+ "train_tokens_per_second": 5661.298
+ },
+ {
+ "epoch": 0.502289253972529,
+ "grad_norm": 0.7229968516896057,
+ "learning_rate": 8.523172717518315e-06,
+ "loss": 0.3696585178375244,
+ "num_input_tokens_seen": 326245696,
+ "step": 7460,
+ "train_runtime": 57627.1988,
+ "train_tokens_per_second": 5661.314
+ },
+ {
+ "epoch": 0.5026259089684891,
+ "grad_norm": 0.7305037275102111,
+ "learning_rate": 8.521296062537156e-06,
+ "loss": 0.37876601219177247,
+ "num_input_tokens_seen": 326460104,
+ "step": 7465,
+ "train_runtime": 57664.9318,
+ "train_tokens_per_second": 5661.328
+ },
+ {
+ "epoch": 0.5029625639644493,
+ "grad_norm": 0.7847466467811594,
+ "learning_rate": 8.51941842283678e-06,
+ "loss": 0.36045174598693847,
+ "num_input_tokens_seen": 326673104,
+ "step": 7470,
+ "train_runtime": 57704.8246,
+ "train_tokens_per_second": 5661.106
+ },
+ {
+ "epoch": 0.5032992189604094,
+ "grad_norm": 0.991284631688998,
+ "learning_rate": 8.517539798942265e-06,
+ "loss": 0.3735339641571045,
+ "num_input_tokens_seen": 326906960,
+ "step": 7475,
+ "train_runtime": 57737.7344,
+ "train_tokens_per_second": 5661.929
+ },
+ {
+ "epoch": 0.5036358739563696,
+ "grad_norm": 0.6872665714663292,
+ "learning_rate": 8.515660191378962e-06,
+ "loss": 0.39676947593688966,
+ "num_input_tokens_seen": 327128416,
+ "step": 7480,
+ "train_runtime": 57776.877,
+ "train_tokens_per_second": 5661.926
+ },
+ {
+ "epoch": 0.5039725289523297,
+ "grad_norm": 0.6751937767879514,
+ "learning_rate": 8.513779600672495e-06,
+ "loss": 0.3535287380218506,
+ "num_input_tokens_seen": 327364152,
+ "step": 7485,
+ "train_runtime": 57811.2991,
+ "train_tokens_per_second": 5662.633
+ },
+ {
+ "epoch": 0.5043091839482898,
+ "grad_norm": 0.7000671921171517,
+ "learning_rate": 8.511898027348771e-06,
+ "loss": 0.37773268222808837,
+ "num_input_tokens_seen": 327568048,
+ "step": 7490,
+ "train_runtime": 57853.1955,
+ "train_tokens_per_second": 5662.056
+ },
+ {
+ "epoch": 0.50464583894425,
+ "grad_norm": 0.6936107754040254,
+ "learning_rate": 8.51001547193396e-06,
+ "loss": 0.3704256296157837,
+ "num_input_tokens_seen": 327792152,
+ "step": 7495,
+ "train_runtime": 57893.3414,
+ "train_tokens_per_second": 5662.001
+ },
+ {
+ "epoch": 0.5049824939402101,
+ "grad_norm": 0.7889020414532445,
+ "learning_rate": 8.508131934954515e-06,
+ "loss": 0.36852757930755614,
+ "num_input_tokens_seen": 327994400,
+ "step": 7500,
+ "train_runtime": 57932.8569,
+ "train_tokens_per_second": 5661.63
+ },
+ {
+ "epoch": 0.5053191489361702,
+ "grad_norm": 0.715078307452266,
+ "learning_rate": 8.50624741693716e-06,
+ "loss": 0.3951726913452148,
+ "num_input_tokens_seen": 328200448,
+ "step": 7505,
+ "train_runtime": 57971.7531,
+ "train_tokens_per_second": 5661.386
+ },
+ {
+ "epoch": 0.5056558039321304,
+ "grad_norm": 1.068005291058121,
+ "learning_rate": 8.5043619184089e-06,
+ "loss": 0.4119250297546387,
+ "num_input_tokens_seen": 328421224,
+ "step": 7510,
+ "train_runtime": 58005.869,
+ "train_tokens_per_second": 5661.862
+ },
+ {
+ "epoch": 0.5059924589280905,
+ "grad_norm": 1.638455940254133,
+ "learning_rate": 8.502475439897e-06,
+ "loss": 0.4078369140625,
+ "num_input_tokens_seen": 328646192,
+ "step": 7515,
+ "train_runtime": 58043.9094,
+ "train_tokens_per_second": 5662.027
+ },
+ {
+ "epoch": 0.5063291139240507,
+ "grad_norm": 0.7358398144623932,
+ "learning_rate": 8.500587981929013e-06,
+ "loss": 0.3898489475250244,
+ "num_input_tokens_seen": 328867064,
+ "step": 7520,
+ "train_runtime": 58088.0612,
+ "train_tokens_per_second": 5661.526
+ },
+ {
+ "epoch": 0.5066657689200108,
+ "grad_norm": 0.7487957564819914,
+ "learning_rate": 8.498699545032758e-06,
+ "loss": 0.38454551696777345,
+ "num_input_tokens_seen": 329103168,
+ "step": 7525,
+ "train_runtime": 58124.1106,
+ "train_tokens_per_second": 5662.077
+ },
+ {
+ "epoch": 0.507002423915971,
+ "grad_norm": 0.7172633060502269,
+ "learning_rate": 8.496810129736334e-06,
+ "loss": 0.3850215435028076,
+ "num_input_tokens_seen": 329322920,
+ "step": 7530,
+ "train_runtime": 58166.9838,
+ "train_tokens_per_second": 5661.681
+ },
+ {
+ "epoch": 0.5073390789119311,
+ "grad_norm": 0.7747557646585903,
+ "learning_rate": 8.494919736568105e-06,
+ "loss": 0.3788848638534546,
+ "num_input_tokens_seen": 329533896,
+ "step": 7535,
+ "train_runtime": 58210.0232,
+ "train_tokens_per_second": 5661.119
+ },
+ {
+ "epoch": 0.5076757339078912,
+ "grad_norm": 0.7933496638959117,
+ "learning_rate": 8.49302836605672e-06,
+ "loss": 0.37194099426269533,
+ "num_input_tokens_seen": 329760624,
+ "step": 7540,
+ "train_runtime": 58246.3752,
+ "train_tokens_per_second": 5661.479
+ },
+ {
+ "epoch": 0.5080123889038514,
+ "grad_norm": 0.7365707403857205,
+ "learning_rate": 8.491136018731088e-06,
+ "loss": 0.3646608114242554,
+ "num_input_tokens_seen": 329984728,
+ "step": 7545,
+ "train_runtime": 58287.6488,
+ "train_tokens_per_second": 5661.315
+ },
+ {
+ "epoch": 0.5083490438998115,
+ "grad_norm": 0.8956737633683735,
+ "learning_rate": 8.4892426951204e-06,
+ "loss": 0.3987517118453979,
+ "num_input_tokens_seen": 330206096,
+ "step": 7550,
+ "train_runtime": 58316.2366,
+ "train_tokens_per_second": 5662.335
+ },
+ {
+ "epoch": 0.5086856988957716,
+ "grad_norm": 0.7170163461314129,
+ "learning_rate": 8.48734839575412e-06,
+ "loss": 0.3717711687088013,
+ "num_input_tokens_seen": 330437336,
+ "step": 7555,
+ "train_runtime": 58351.3409,
+ "train_tokens_per_second": 5662.892
+ },
+ {
+ "epoch": 0.5090223538917318,
+ "grad_norm": 0.6630558999546488,
+ "learning_rate": 8.485453121161983e-06,
+ "loss": 0.37380030155181887,
+ "num_input_tokens_seen": 330653616,
+ "step": 7560,
+ "train_runtime": 58392.093,
+ "train_tokens_per_second": 5662.644
+ },
+ {
+ "epoch": 0.5093590088876919,
+ "grad_norm": 0.7845896930720607,
+ "learning_rate": 8.483556871873993e-06,
+ "loss": 0.42113819122314455,
+ "num_input_tokens_seen": 330875672,
+ "step": 7565,
+ "train_runtime": 58433.2419,
+ "train_tokens_per_second": 5662.456
+ },
+ {
+ "epoch": 0.5096956638836521,
+ "grad_norm": 0.760051472590646,
+ "learning_rate": 8.481659648420433e-06,
+ "loss": 0.3786670207977295,
+ "num_input_tokens_seen": 331096488,
+ "step": 7570,
+ "train_runtime": 58475.0649,
+ "train_tokens_per_second": 5662.182
+ },
+ {
+ "epoch": 0.5100323188796122,
+ "grad_norm": 0.8835596263525625,
+ "learning_rate": 8.479761451331855e-06,
+ "loss": 0.3809337615966797,
+ "num_input_tokens_seen": 331285184,
+ "step": 7575,
+ "train_runtime": 58515.9798,
+ "train_tokens_per_second": 5661.448
+ },
+ {
+ "epoch": 0.5103689738755723,
+ "grad_norm": 0.9301429501501213,
+ "learning_rate": 8.477862281139082e-06,
+ "loss": 0.3540458917617798,
+ "num_input_tokens_seen": 331501760,
+ "step": 7580,
+ "train_runtime": 58558.4957,
+ "train_tokens_per_second": 5661.036
+ },
+ {
+ "epoch": 0.5107056288715325,
+ "grad_norm": 0.7609757852463125,
+ "learning_rate": 8.475962138373212e-06,
+ "loss": 0.3986123323440552,
+ "num_input_tokens_seen": 331702152,
+ "step": 7585,
+ "train_runtime": 58591.7614,
+ "train_tokens_per_second": 5661.242
+ },
+ {
+ "epoch": 0.5110422838674926,
+ "grad_norm": 0.6930333171792868,
+ "learning_rate": 8.474061023565614e-06,
+ "loss": 0.3940688371658325,
+ "num_input_tokens_seen": 331929992,
+ "step": 7590,
+ "train_runtime": 58631.4664,
+ "train_tokens_per_second": 5661.294
+ },
+ {
+ "epoch": 0.5113789388634528,
+ "grad_norm": 0.7801001592101691,
+ "learning_rate": 8.472158937247931e-06,
+ "loss": 0.40605816841125486,
+ "num_input_tokens_seen": 332148472,
+ "step": 7595,
+ "train_runtime": 58666.3698,
+ "train_tokens_per_second": 5661.65
+ },
+ {
+ "epoch": 0.5117155938594129,
+ "grad_norm": 0.7185541728168203,
+ "learning_rate": 8.47025587995207e-06,
+ "loss": 0.38570313453674315,
+ "num_input_tokens_seen": 332362728,
+ "step": 7600,
+ "train_runtime": 58704.4102,
+ "train_tokens_per_second": 5661.631
+ },
+ {
+ "epoch": 0.512052248855373,
+ "grad_norm": 0.7887783994842774,
+ "learning_rate": 8.46835185221022e-06,
+ "loss": 0.39829018115997317,
+ "num_input_tokens_seen": 332569544,
+ "step": 7605,
+ "train_runtime": 58743.9844,
+ "train_tokens_per_second": 5661.338
+ },
+ {
+ "epoch": 0.5123889038513332,
+ "grad_norm": 0.8602264236155862,
+ "learning_rate": 8.466446854554835e-06,
+ "loss": 0.3549684762954712,
+ "num_input_tokens_seen": 332778920,
+ "step": 7610,
+ "train_runtime": 58780.8818,
+ "train_tokens_per_second": 5661.346
+ },
+ {
+ "epoch": 0.5127255588472933,
+ "grad_norm": 0.7228150824538518,
+ "learning_rate": 8.464540887518638e-06,
+ "loss": 0.37906813621520996,
+ "num_input_tokens_seen": 332984080,
+ "step": 7615,
+ "train_runtime": 58823.2164,
+ "train_tokens_per_second": 5660.759
+ },
+ {
+ "epoch": 0.5130622138432535,
+ "grad_norm": 0.7996970703617401,
+ "learning_rate": 8.462633951634631e-06,
+ "loss": 0.3685762405395508,
+ "num_input_tokens_seen": 333202640,
+ "step": 7620,
+ "train_runtime": 58868.3048,
+ "train_tokens_per_second": 5660.136
+ },
+ {
+ "epoch": 0.5133988688392136,
+ "grad_norm": 0.6901634902422549,
+ "learning_rate": 8.46072604743608e-06,
+ "loss": 0.35547614097595215,
+ "num_input_tokens_seen": 333413904,
+ "step": 7625,
+ "train_runtime": 58905.0124,
+ "train_tokens_per_second": 5660.196
+ },
+ {
+ "epoch": 0.5137355238351737,
+ "grad_norm": 0.9033857052689097,
+ "learning_rate": 8.458817175456528e-06,
+ "loss": 0.3930670261383057,
+ "num_input_tokens_seen": 333630616,
+ "step": 7630,
+ "train_runtime": 58947.1829,
+ "train_tokens_per_second": 5659.823
+ },
+ {
+ "epoch": 0.5140721788311339,
+ "grad_norm": 0.7781771666334231,
+ "learning_rate": 8.45690733622978e-06,
+ "loss": 0.35927329063415525,
+ "num_input_tokens_seen": 333854600,
+ "step": 7635,
+ "train_runtime": 58986.6913,
+ "train_tokens_per_second": 5659.829
+ },
+ {
+ "epoch": 0.514408833827094,
+ "grad_norm": 0.8455368762157274,
+ "learning_rate": 8.45499653028992e-06,
+ "loss": 0.36787919998168944,
+ "num_input_tokens_seen": 334071192,
+ "step": 7640,
+ "train_runtime": 59022.2041,
+ "train_tokens_per_second": 5660.093
+ },
+ {
+ "epoch": 0.5147454888230542,
+ "grad_norm": 0.7586644939540588,
+ "learning_rate": 8.453084758171297e-06,
+ "loss": 0.37957746982574464,
+ "num_input_tokens_seen": 334291888,
+ "step": 7645,
+ "train_runtime": 59067.1738,
+ "train_tokens_per_second": 5659.521
+ },
+ {
+ "epoch": 0.5150821438190143,
+ "grad_norm": 0.7524816129003911,
+ "learning_rate": 8.451172020408532e-06,
+ "loss": 0.38945965766906737,
+ "num_input_tokens_seen": 334516424,
+ "step": 7650,
+ "train_runtime": 59101.459,
+ "train_tokens_per_second": 5660.037
+ },
+ {
+ "epoch": 0.5154187988149744,
+ "grad_norm": 0.7137520182282876,
+ "learning_rate": 8.449258317536517e-06,
+ "loss": 0.35926315784454343,
+ "num_input_tokens_seen": 334732808,
+ "step": 7655,
+ "train_runtime": 59134.7468,
+ "train_tokens_per_second": 5660.51
+ },
+ {
+ "epoch": 0.5157554538109346,
+ "grad_norm": 0.7129464795573726,
+ "learning_rate": 8.447343650090412e-06,
+ "loss": 0.35725347995758056,
+ "num_input_tokens_seen": 334954280,
+ "step": 7660,
+ "train_runtime": 59172.9555,
+ "train_tokens_per_second": 5660.597
+ },
+ {
+ "epoch": 0.5160921088068947,
+ "grad_norm": 0.7272448686794939,
+ "learning_rate": 8.445428018605647e-06,
+ "loss": 0.406490421295166,
+ "num_input_tokens_seen": 335178568,
+ "step": 7665,
+ "train_runtime": 59210.2513,
+ "train_tokens_per_second": 5660.82
+ },
+ {
+ "epoch": 0.5164287638028549,
+ "grad_norm": 0.6593329295977397,
+ "learning_rate": 8.443511423617924e-06,
+ "loss": 0.34410724639892576,
+ "num_input_tokens_seen": 335398608,
+ "step": 7670,
+ "train_runtime": 59246.6136,
+ "train_tokens_per_second": 5661.06
+ },
+ {
+ "epoch": 0.516765418798815,
+ "grad_norm": 0.7054753575271266,
+ "learning_rate": 8.441593865663213e-06,
+ "loss": 0.39874303340911865,
+ "num_input_tokens_seen": 335621696,
+ "step": 7675,
+ "train_runtime": 59285.0684,
+ "train_tokens_per_second": 5661.151
+ },
+ {
+ "epoch": 0.5171020737947751,
+ "grad_norm": 0.7429064696844442,
+ "learning_rate": 8.439675345277751e-06,
+ "loss": 0.40140485763549805,
+ "num_input_tokens_seen": 335839504,
+ "step": 7680,
+ "train_runtime": 59327.3379,
+ "train_tokens_per_second": 5660.788
+ },
+ {
+ "epoch": 0.5174387287907353,
+ "grad_norm": 0.6392657750321254,
+ "learning_rate": 8.437755862998047e-06,
+ "loss": 0.3619039297103882,
+ "num_input_tokens_seen": 336062080,
+ "step": 7685,
+ "train_runtime": 59362.2516,
+ "train_tokens_per_second": 5661.208
+ },
+ {
+ "epoch": 0.5177753837866954,
+ "grad_norm": 0.6333778064119364,
+ "learning_rate": 8.435835419360878e-06,
+ "loss": 0.3882591247558594,
+ "num_input_tokens_seen": 336299944,
+ "step": 7690,
+ "train_runtime": 59398.7224,
+ "train_tokens_per_second": 5661.737
+ },
+ {
+ "epoch": 0.5181120387826555,
+ "grad_norm": 0.7397126752113882,
+ "learning_rate": 8.433914014903291e-06,
+ "loss": 0.42388019561767576,
+ "num_input_tokens_seen": 336514968,
+ "step": 7695,
+ "train_runtime": 59437.773,
+ "train_tokens_per_second": 5661.635
+ },
+ {
+ "epoch": 0.5184486937786157,
+ "grad_norm": 0.6888095905405301,
+ "learning_rate": 8.431991650162599e-06,
+ "loss": 0.37606000900268555,
+ "num_input_tokens_seen": 336732600,
+ "step": 7700,
+ "train_runtime": 59483.4411,
+ "train_tokens_per_second": 5660.947
+ },
+ {
+ "epoch": 0.5187853487745758,
+ "grad_norm": 0.6980617402046243,
+ "learning_rate": 8.430068325676386e-06,
+ "loss": 0.39625885486602785,
+ "num_input_tokens_seen": 336948776,
+ "step": 7705,
+ "train_runtime": 59521.0447,
+ "train_tokens_per_second": 5661.002
+ },
+ {
+ "epoch": 0.519122003770536,
+ "grad_norm": 0.732968462897192,
+ "learning_rate": 8.428144041982502e-06,
+ "loss": 0.39479496479034426,
+ "num_input_tokens_seen": 337163520,
+ "step": 7710,
+ "train_runtime": 59556.4,
+ "train_tokens_per_second": 5661.247
+ },
+ {
+ "epoch": 0.5194586587664961,
+ "grad_norm": 0.6401198443952045,
+ "learning_rate": 8.42621879961907e-06,
+ "loss": 0.38160171508789065,
+ "num_input_tokens_seen": 337382992,
+ "step": 7715,
+ "train_runtime": 59597.2808,
+ "train_tokens_per_second": 5661.047
+ },
+ {
+ "epoch": 0.5197953137624562,
+ "grad_norm": 0.7011673418789361,
+ "learning_rate": 8.424292599124476e-06,
+ "loss": 0.3532627820968628,
+ "num_input_tokens_seen": 337609304,
+ "step": 7720,
+ "train_runtime": 59636.4965,
+ "train_tokens_per_second": 5661.119
+ },
+ {
+ "epoch": 0.5201319687584164,
+ "grad_norm": 0.7609717187854771,
+ "learning_rate": 8.422365441037374e-06,
+ "loss": 0.39478967189788816,
+ "num_input_tokens_seen": 337838072,
+ "step": 7725,
+ "train_runtime": 59669.1056,
+ "train_tokens_per_second": 5661.859
+ },
+ {
+ "epoch": 0.5204686237543765,
+ "grad_norm": 0.743645872276741,
+ "learning_rate": 8.420437325896692e-06,
+ "loss": 0.3906223297119141,
+ "num_input_tokens_seen": 338067512,
+ "step": 7730,
+ "train_runtime": 59705.1498,
+ "train_tokens_per_second": 5662.284
+ },
+ {
+ "epoch": 0.5208052787503367,
+ "grad_norm": 0.7586775310938808,
+ "learning_rate": 8.418508254241617e-06,
+ "loss": 0.3925325870513916,
+ "num_input_tokens_seen": 338284504,
+ "step": 7735,
+ "train_runtime": 59748.0729,
+ "train_tokens_per_second": 5661.848
+ },
+ {
+ "epoch": 0.5211419337462968,
+ "grad_norm": 0.8223238925961407,
+ "learning_rate": 8.416578226611612e-06,
+ "loss": 0.35642185211181643,
+ "num_input_tokens_seen": 338495440,
+ "step": 7740,
+ "train_runtime": 59783.2802,
+ "train_tokens_per_second": 5662.042
+ },
+ {
+ "epoch": 0.5214785887422569,
+ "grad_norm": 0.8532576047124709,
+ "learning_rate": 8.414647243546398e-06,
+ "loss": 0.38446238040924074,
+ "num_input_tokens_seen": 338717192,
+ "step": 7745,
+ "train_runtime": 59821.3869,
+ "train_tokens_per_second": 5662.142
+ },
+ {
+ "epoch": 0.5218152437382171,
+ "grad_norm": 0.6899241307861492,
+ "learning_rate": 8.412715305585973e-06,
+ "loss": 0.367059588432312,
+ "num_input_tokens_seen": 338946888,
+ "step": 7750,
+ "train_runtime": 59859.9792,
+ "train_tokens_per_second": 5662.329
+ },
+ {
+ "epoch": 0.5221518987341772,
+ "grad_norm": 0.7311441327867719,
+ "learning_rate": 8.410782413270594e-06,
+ "loss": 0.3799006462097168,
+ "num_input_tokens_seen": 339172224,
+ "step": 7755,
+ "train_runtime": 59900.1807,
+ "train_tokens_per_second": 5662.29
+ },
+ {
+ "epoch": 0.5224885537301374,
+ "grad_norm": 0.9420308727064326,
+ "learning_rate": 8.408848567140787e-06,
+ "loss": 0.37287948131561277,
+ "num_input_tokens_seen": 339364672,
+ "step": 7760,
+ "train_runtime": 59941.0763,
+ "train_tokens_per_second": 5661.638
+ },
+ {
+ "epoch": 0.5228252087260975,
+ "grad_norm": 0.7245874800766633,
+ "learning_rate": 8.406913767737353e-06,
+ "loss": 0.37642457485198977,
+ "num_input_tokens_seen": 339579064,
+ "step": 7765,
+ "train_runtime": 59982.2891,
+ "train_tokens_per_second": 5661.322
+ },
+ {
+ "epoch": 0.5231618637220576,
+ "grad_norm": 0.7199997285512201,
+ "learning_rate": 8.404978015601344e-06,
+ "loss": 0.3964075565338135,
+ "num_input_tokens_seen": 339801840,
+ "step": 7770,
+ "train_runtime": 60029.6628,
+ "train_tokens_per_second": 5660.566
+ },
+ {
+ "epoch": 0.5234985187180178,
+ "grad_norm": 0.697918705663631,
+ "learning_rate": 8.403041311274091e-06,
+ "loss": 0.39160215854644775,
+ "num_input_tokens_seen": 340025832,
+ "step": 7775,
+ "train_runtime": 60071.2747,
+ "train_tokens_per_second": 5660.373
+ },
+ {
+ "epoch": 0.5238351737139779,
+ "grad_norm": 0.8157700767931442,
+ "learning_rate": 8.401103655297188e-06,
+ "loss": 0.3895624876022339,
+ "num_input_tokens_seen": 340239616,
+ "step": 7780,
+ "train_runtime": 60105.9113,
+ "train_tokens_per_second": 5660.668
+ },
+ {
+ "epoch": 0.5241718287099381,
+ "grad_norm": 0.838923642880136,
+ "learning_rate": 8.399165048212494e-06,
+ "loss": 0.3589164257049561,
+ "num_input_tokens_seen": 340454984,
+ "step": 7785,
+ "train_runtime": 60145.3789,
+ "train_tokens_per_second": 5660.534
+ },
+ {
+ "epoch": 0.5245084837058982,
+ "grad_norm": 0.7246920315479125,
+ "learning_rate": 8.397225490562131e-06,
+ "loss": 0.4032485008239746,
+ "num_input_tokens_seen": 340684288,
+ "step": 7790,
+ "train_runtime": 60180.0663,
+ "train_tokens_per_second": 5661.082
+ },
+ {
+ "epoch": 0.5248451387018583,
+ "grad_norm": 0.7826576263190983,
+ "learning_rate": 8.395284982888494e-06,
+ "loss": 0.3588679313659668,
+ "num_input_tokens_seen": 340904512,
+ "step": 7795,
+ "train_runtime": 60216.628,
+ "train_tokens_per_second": 5661.302
+ },
+ {
+ "epoch": 0.5251817936978185,
+ "grad_norm": 0.7381573756482819,
+ "learning_rate": 8.393343525734239e-06,
+ "loss": 0.39098711013793946,
+ "num_input_tokens_seen": 341128688,
+ "step": 7800,
+ "train_runtime": 60254.8697,
+ "train_tokens_per_second": 5661.429
+ },
+ {
+ "epoch": 0.5255184486937786,
+ "grad_norm": 0.9510438503863269,
+ "learning_rate": 8.391401119642286e-06,
+ "loss": 0.39932818412780763,
+ "num_input_tokens_seen": 341326256,
+ "step": 7805,
+ "train_runtime": 60302.5843,
+ "train_tokens_per_second": 5660.226
+ },
+ {
+ "epoch": 0.5258551036897388,
+ "grad_norm": 0.7551888581389101,
+ "learning_rate": 8.389457765155824e-06,
+ "loss": 0.40273380279541016,
+ "num_input_tokens_seen": 341526840,
+ "step": 7810,
+ "train_runtime": 60347.108,
+ "train_tokens_per_second": 5659.374
+ },
+ {
+ "epoch": 0.5261917586856989,
+ "grad_norm": 0.7817537228902953,
+ "learning_rate": 8.387513462818309e-06,
+ "loss": 0.36842942237854004,
+ "num_input_tokens_seen": 341747720,
+ "step": 7815,
+ "train_runtime": 60388.0509,
+ "train_tokens_per_second": 5659.194
+ },
+ {
+ "epoch": 0.526528413681659,
+ "grad_norm": 0.76396158135225,
+ "learning_rate": 8.385568213173453e-06,
+ "loss": 0.3687218427658081,
+ "num_input_tokens_seen": 341961736,
+ "step": 7820,
+ "train_runtime": 60430.695,
+ "train_tokens_per_second": 5658.742
+ },
+ {
+ "epoch": 0.5268650686776192,
+ "grad_norm": 1.0538784566401151,
+ "learning_rate": 8.383622016765242e-06,
+ "loss": 0.3595340013504028,
+ "num_input_tokens_seen": 342180688,
+ "step": 7825,
+ "train_runtime": 60474.2774,
+ "train_tokens_per_second": 5658.285
+ },
+ {
+ "epoch": 0.5272017236735793,
+ "grad_norm": 0.7683842178996437,
+ "learning_rate": 8.381674874137926e-06,
+ "loss": 0.4014129638671875,
+ "num_input_tokens_seen": 342390504,
+ "step": 7830,
+ "train_runtime": 60507.7742,
+ "train_tokens_per_second": 5658.62
+ },
+ {
+ "epoch": 0.5275383786695395,
+ "grad_norm": 0.7767541075637882,
+ "learning_rate": 8.379726785836013e-06,
+ "loss": 0.3654167652130127,
+ "num_input_tokens_seen": 342602080,
+ "step": 7835,
+ "train_runtime": 60550.4658,
+ "train_tokens_per_second": 5658.125
+ },
+ {
+ "epoch": 0.5278750336654996,
+ "grad_norm": 0.6734458895156976,
+ "learning_rate": 8.377777752404283e-06,
+ "loss": 0.36977086067199705,
+ "num_input_tokens_seen": 342831800,
+ "step": 7840,
+ "train_runtime": 60591.1256,
+ "train_tokens_per_second": 5658.119
+ },
+ {
+ "epoch": 0.5282116886614597,
+ "grad_norm": 0.8168160522078569,
+ "learning_rate": 8.375827774387774e-06,
+ "loss": 0.34728164672851564,
+ "num_input_tokens_seen": 343029568,
+ "step": 7845,
+ "train_runtime": 60633.6042,
+ "train_tokens_per_second": 5657.417
+ },
+ {
+ "epoch": 0.5285483436574199,
+ "grad_norm": 0.6622532156086101,
+ "learning_rate": 8.373876852331793e-06,
+ "loss": 0.3438518762588501,
+ "num_input_tokens_seen": 343254472,
+ "step": 7850,
+ "train_runtime": 60671.9643,
+ "train_tokens_per_second": 5657.547
+ },
+ {
+ "epoch": 0.52888499865338,
+ "grad_norm": 0.7502437558347258,
+ "learning_rate": 8.371924986781909e-06,
+ "loss": 0.38301682472229004,
+ "num_input_tokens_seen": 343470008,
+ "step": 7855,
+ "train_runtime": 60709.3414,
+ "train_tokens_per_second": 5657.614
+ },
+ {
+ "epoch": 0.5292216536493402,
+ "grad_norm": 0.7587114327145728,
+ "learning_rate": 8.369972178283955e-06,
+ "loss": 0.36738762855529783,
+ "num_input_tokens_seen": 343698072,
+ "step": 7860,
+ "train_runtime": 60747.59,
+ "train_tokens_per_second": 5657.806
+ },
+ {
+ "epoch": 0.5295583086453003,
+ "grad_norm": 0.6881092460250557,
+ "learning_rate": 8.368018427384027e-06,
+ "loss": 0.38115482330322265,
+ "num_input_tokens_seen": 343919960,
+ "step": 7865,
+ "train_runtime": 60791.0097,
+ "train_tokens_per_second": 5657.415
+ },
+ {
+ "epoch": 0.5298949636412604,
+ "grad_norm": 0.8355237541845086,
+ "learning_rate": 8.366063734628485e-06,
+ "loss": 0.38156135082244874,
+ "num_input_tokens_seen": 344123568,
+ "step": 7870,
+ "train_runtime": 60827.6415,
+ "train_tokens_per_second": 5657.355
+ },
+ {
+ "epoch": 0.5302316186372206,
+ "grad_norm": 0.8303838221096392,
+ "learning_rate": 8.364108100563955e-06,
+ "loss": 0.4007874011993408,
+ "num_input_tokens_seen": 344341784,
+ "step": 7875,
+ "train_runtime": 60864.4296,
+ "train_tokens_per_second": 5657.521
+ },
+ {
+ "epoch": 0.5305682736331807,
+ "grad_norm": 0.7925795609225155,
+ "learning_rate": 8.362151525737321e-06,
+ "loss": 0.4212652206420898,
+ "num_input_tokens_seen": 344562552,
+ "step": 7880,
+ "train_runtime": 60901.6398,
+ "train_tokens_per_second": 5657.689
+ },
+ {
+ "epoch": 0.5309049286291408,
+ "grad_norm": 0.7468959848481416,
+ "learning_rate": 8.360194010695733e-06,
+ "loss": 0.3570136547088623,
+ "num_input_tokens_seen": 344782336,
+ "step": 7885,
+ "train_runtime": 60942.4135,
+ "train_tokens_per_second": 5657.51
+ },
+ {
+ "epoch": 0.531241583625101,
+ "grad_norm": 0.7105479533553207,
+ "learning_rate": 8.358235555986607e-06,
+ "loss": 0.39661598205566406,
+ "num_input_tokens_seen": 345008672,
+ "step": 7890,
+ "train_runtime": 60976.8653,
+ "train_tokens_per_second": 5658.026
+ },
+ {
+ "epoch": 0.5315782386210611,
+ "grad_norm": 0.768825896681359,
+ "learning_rate": 8.356276162157613e-06,
+ "loss": 0.3597188234329224,
+ "num_input_tokens_seen": 345216360,
+ "step": 7895,
+ "train_runtime": 61021.9355,
+ "train_tokens_per_second": 5657.25
+ },
+ {
+ "epoch": 0.5319148936170213,
+ "grad_norm": 0.711887791776799,
+ "learning_rate": 8.354315829756696e-06,
+ "loss": 0.37782256603240966,
+ "num_input_tokens_seen": 345442968,
+ "step": 7900,
+ "train_runtime": 61056.926,
+ "train_tokens_per_second": 5657.72
+ },
+ {
+ "epoch": 0.5322515486129814,
+ "grad_norm": 0.7271488313267995,
+ "learning_rate": 8.352354559332053e-06,
+ "loss": 0.36695542335510256,
+ "num_input_tokens_seen": 345663248,
+ "step": 7905,
+ "train_runtime": 61093.5998,
+ "train_tokens_per_second": 5657.929
+ },
+ {
+ "epoch": 0.5325882036089415,
+ "grad_norm": 0.7790624119649752,
+ "learning_rate": 8.350392351432146e-06,
+ "loss": 0.3877138376235962,
+ "num_input_tokens_seen": 345877640,
+ "step": 7910,
+ "train_runtime": 61125.7433,
+ "train_tokens_per_second": 5658.461
+ },
+ {
+ "epoch": 0.5329248586049017,
+ "grad_norm": 0.652480198172122,
+ "learning_rate": 8.348429206605702e-06,
+ "loss": 0.3623414754867554,
+ "num_input_tokens_seen": 346113080,
+ "step": 7915,
+ "train_runtime": 61170.1897,
+ "train_tokens_per_second": 5658.199
+ },
+ {
+ "epoch": 0.5332615136008618,
+ "grad_norm": 0.8166602948279773,
+ "learning_rate": 8.346465125401706e-06,
+ "loss": 0.41824941635131835,
+ "num_input_tokens_seen": 346331632,
+ "step": 7920,
+ "train_runtime": 61213.7632,
+ "train_tokens_per_second": 5657.741
+ },
+ {
+ "epoch": 0.533598168596822,
+ "grad_norm": 0.7338720699869766,
+ "learning_rate": 8.344500108369412e-06,
+ "loss": 0.3689335107803345,
+ "num_input_tokens_seen": 346540104,
+ "step": 7925,
+ "train_runtime": 61254.7218,
+ "train_tokens_per_second": 5657.361
+ },
+ {
+ "epoch": 0.5339348235927821,
+ "grad_norm": 0.7241659753104186,
+ "learning_rate": 8.342534156058326e-06,
+ "loss": 0.3848552942276001,
+ "num_input_tokens_seen": 346759048,
+ "step": 7930,
+ "train_runtime": 61293.0183,
+ "train_tokens_per_second": 5657.399
+ },
+ {
+ "epoch": 0.5342714785887422,
+ "grad_norm": 0.7284900345851684,
+ "learning_rate": 8.34056726901822e-06,
+ "loss": 0.39329664707183837,
+ "num_input_tokens_seen": 346971992,
+ "step": 7935,
+ "train_runtime": 61333.9961,
+ "train_tokens_per_second": 5657.091
+ },
+ {
+ "epoch": 0.5346081335847024,
+ "grad_norm": 0.9343840115025148,
+ "learning_rate": 8.338599447799132e-06,
+ "loss": 0.37961349487304685,
+ "num_input_tokens_seen": 347176784,
+ "step": 7940,
+ "train_runtime": 61367.1004,
+ "train_tokens_per_second": 5657.376
+ },
+ {
+ "epoch": 0.5349447885806625,
+ "grad_norm": 0.7650504516817562,
+ "learning_rate": 8.336630692951353e-06,
+ "loss": 0.3706023931503296,
+ "num_input_tokens_seen": 347396336,
+ "step": 7945,
+ "train_runtime": 61405.4086,
+ "train_tokens_per_second": 5657.422
+ },
+ {
+ "epoch": 0.5352814435766227,
+ "grad_norm": 0.7809182592661493,
+ "learning_rate": 8.334661005025441e-06,
+ "loss": 0.3996375560760498,
+ "num_input_tokens_seen": 347612392,
+ "step": 7950,
+ "train_runtime": 61443.353,
+ "train_tokens_per_second": 5657.445
+ },
+ {
+ "epoch": 0.5356180985725828,
+ "grad_norm": 1.0173782664960012,
+ "learning_rate": 8.332690384572213e-06,
+ "loss": 0.4039102077484131,
+ "num_input_tokens_seen": 347827352,
+ "step": 7955,
+ "train_runtime": 61481.7721,
+ "train_tokens_per_second": 5657.406
+ },
+ {
+ "epoch": 0.5359547535685429,
+ "grad_norm": 0.7319934278286511,
+ "learning_rate": 8.330718832142746e-06,
+ "loss": 0.3981315612792969,
+ "num_input_tokens_seen": 348060432,
+ "step": 7960,
+ "train_runtime": 61520.2514,
+ "train_tokens_per_second": 5657.656
+ },
+ {
+ "epoch": 0.5362914085645031,
+ "grad_norm": 0.6935185521555359,
+ "learning_rate": 8.328746348288379e-06,
+ "loss": 0.3804043769836426,
+ "num_input_tokens_seen": 348266616,
+ "step": 7965,
+ "train_runtime": 61562.5177,
+ "train_tokens_per_second": 5657.121
+ },
+ {
+ "epoch": 0.5366280635604632,
+ "grad_norm": 0.7663090177174399,
+ "learning_rate": 8.32677293356071e-06,
+ "loss": 0.42568235397338866,
+ "num_input_tokens_seen": 348498272,
+ "step": 7970,
+ "train_runtime": 61598.1572,
+ "train_tokens_per_second": 5657.609
+ },
+ {
+ "epoch": 0.5369647185564234,
+ "grad_norm": 0.8375018048980476,
+ "learning_rate": 8.324798588511598e-06,
+ "loss": 0.3702836275100708,
+ "num_input_tokens_seen": 348696416,
+ "step": 7975,
+ "train_runtime": 61635.4226,
+ "train_tokens_per_second": 5657.403
+ },
+ {
+ "epoch": 0.5373013735523835,
+ "grad_norm": 0.8187637529093583,
+ "learning_rate": 8.322823313693162e-06,
+ "loss": 0.38841800689697265,
+ "num_input_tokens_seen": 348910592,
+ "step": 7980,
+ "train_runtime": 61671.4891,
+ "train_tokens_per_second": 5657.567
+ },
+ {
+ "epoch": 0.5376380285483436,
+ "grad_norm": 0.7918621668094143,
+ "learning_rate": 8.320847109657782e-06,
+ "loss": 0.3783905506134033,
+ "num_input_tokens_seen": 349120256,
+ "step": 7985,
+ "train_runtime": 61712.9559,
+ "train_tokens_per_second": 5657.163
+ },
+ {
+ "epoch": 0.5379746835443038,
+ "grad_norm": 0.7362861473446518,
+ "learning_rate": 8.318869976958099e-06,
+ "loss": 0.3449576377868652,
+ "num_input_tokens_seen": 349340688,
+ "step": 7990,
+ "train_runtime": 61755.9567,
+ "train_tokens_per_second": 5656.793
+ },
+ {
+ "epoch": 0.5383113385402639,
+ "grad_norm": 0.7351144206496397,
+ "learning_rate": 8.316891916147007e-06,
+ "loss": 0.40518803596496583,
+ "num_input_tokens_seen": 349557408,
+ "step": 7995,
+ "train_runtime": 61795.1163,
+ "train_tokens_per_second": 5656.716
+ },
+ {
+ "epoch": 0.5386479935362241,
+ "grad_norm": 0.8155814621488839,
+ "learning_rate": 8.314912927777671e-06,
+ "loss": 0.3914914608001709,
+ "num_input_tokens_seen": 349776936,
+ "step": 8000,
+ "train_runtime": 61829.5854,
+ "train_tokens_per_second": 5657.113
+ },
+ {
+ "epoch": 0.5389846485321842,
+ "grad_norm": 0.7399640627157749,
+ "learning_rate": 8.312933012403503e-06,
+ "loss": 0.3830781698226929,
+ "num_input_tokens_seen": 349981944,
+ "step": 8005,
+ "train_runtime": 61870.1471,
+ "train_tokens_per_second": 5656.718
+ },
+ {
+ "epoch": 0.5393213035281443,
+ "grad_norm": 0.6837705266722391,
+ "learning_rate": 8.310952170578184e-06,
+ "loss": 0.3917949676513672,
+ "num_input_tokens_seen": 350196560,
+ "step": 8010,
+ "train_runtime": 61910.9608,
+ "train_tokens_per_second": 5656.455
+ },
+ {
+ "epoch": 0.5396579585241045,
+ "grad_norm": 0.8108035687999815,
+ "learning_rate": 8.308970402855647e-06,
+ "loss": 0.35423126220703127,
+ "num_input_tokens_seen": 350407320,
+ "step": 8015,
+ "train_runtime": 61942.9362,
+ "train_tokens_per_second": 5656.938
+ },
+ {
+ "epoch": 0.5399946135200646,
+ "grad_norm": 0.7391585249859578,
+ "learning_rate": 8.306987709790087e-06,
+ "loss": 0.36304192543029784,
+ "num_input_tokens_seen": 350621920,
+ "step": 8020,
+ "train_runtime": 61974.5184,
+ "train_tokens_per_second": 5657.517
+ },
+ {
+ "epoch": 0.5403312685160248,
+ "grad_norm": 0.8297400381885977,
+ "learning_rate": 8.305004091935962e-06,
+ "loss": 0.3827538013458252,
+ "num_input_tokens_seen": 350840936,
+ "step": 8025,
+ "train_runtime": 62010.6476,
+ "train_tokens_per_second": 5657.753
+ },
+ {
+ "epoch": 0.5406679235119849,
+ "grad_norm": 0.8790917583658322,
+ "learning_rate": 8.303019549847979e-06,
+ "loss": 0.40552120208740233,
+ "num_input_tokens_seen": 351057864,
+ "step": 8030,
+ "train_runtime": 62051.8481,
+ "train_tokens_per_second": 5657.492
+ },
+ {
+ "epoch": 0.541004578507945,
+ "grad_norm": 0.7794541942972262,
+ "learning_rate": 8.301034084081114e-06,
+ "loss": 0.38002500534057615,
+ "num_input_tokens_seen": 351285720,
+ "step": 8035,
+ "train_runtime": 62090.6336,
+ "train_tokens_per_second": 5657.628
+ },
+ {
+ "epoch": 0.5413412335039052,
+ "grad_norm": 0.8524170296510677,
+ "learning_rate": 8.299047695190592e-06,
+ "loss": 0.408983850479126,
+ "num_input_tokens_seen": 351491576,
+ "step": 8040,
+ "train_runtime": 62127.0664,
+ "train_tokens_per_second": 5657.624
+ },
+ {
+ "epoch": 0.5416778884998653,
+ "grad_norm": 0.8835316549178089,
+ "learning_rate": 8.297060383731903e-06,
+ "loss": 0.38953261375427245,
+ "num_input_tokens_seen": 351706376,
+ "step": 8045,
+ "train_runtime": 62162.9599,
+ "train_tokens_per_second": 5657.813
+ },
+ {
+ "epoch": 0.5420145434958255,
+ "grad_norm": 0.8130215721187897,
+ "learning_rate": 8.29507215026079e-06,
+ "loss": 0.3989473819732666,
+ "num_input_tokens_seen": 351925296,
+ "step": 8050,
+ "train_runtime": 62194.3013,
+ "train_tokens_per_second": 5658.481
+ },
+ {
+ "epoch": 0.5423511984917856,
+ "grad_norm": 0.7698865553206388,
+ "learning_rate": 8.293082995333257e-06,
+ "loss": 0.3805055618286133,
+ "num_input_tokens_seen": 352152728,
+ "step": 8055,
+ "train_runtime": 62235.2959,
+ "train_tokens_per_second": 5658.409
+ },
+ {
+ "epoch": 0.5426878534877457,
+ "grad_norm": 0.7925089143300325,
+ "learning_rate": 8.291092919505568e-06,
+ "loss": 0.359727144241333,
+ "num_input_tokens_seen": 352368040,
+ "step": 8060,
+ "train_runtime": 62272.5239,
+ "train_tokens_per_second": 5658.483
+ },
+ {
+ "epoch": 0.5430245084837059,
+ "grad_norm": 0.7473619474168591,
+ "learning_rate": 8.289101923334237e-06,
+ "loss": 0.3927903175354004,
+ "num_input_tokens_seen": 352592456,
+ "step": 8065,
+ "train_runtime": 62314.1115,
+ "train_tokens_per_second": 5658.308
+ },
+ {
+ "epoch": 0.543361163479666,
+ "grad_norm": 0.8359023777365873,
+ "learning_rate": 8.28711000737604e-06,
+ "loss": 0.37551441192626955,
+ "num_input_tokens_seen": 352801680,
+ "step": 8070,
+ "train_runtime": 62352.2218,
+ "train_tokens_per_second": 5658.205
+ },
+ {
+ "epoch": 0.5436978184756261,
+ "grad_norm": 0.8077808908921092,
+ "learning_rate": 8.285117172188013e-06,
+ "loss": 0.3938951253890991,
+ "num_input_tokens_seen": 353025760,
+ "step": 8075,
+ "train_runtime": 62386.0037,
+ "train_tokens_per_second": 5658.733
+ },
+ {
+ "epoch": 0.5440344734715863,
+ "grad_norm": 0.8228182913825491,
+ "learning_rate": 8.283123418327443e-06,
+ "loss": 0.4004195213317871,
+ "num_input_tokens_seen": 353244352,
+ "step": 8080,
+ "train_runtime": 62419.1441,
+ "train_tokens_per_second": 5659.231
+ },
+ {
+ "epoch": 0.5443711284675464,
+ "grad_norm": 0.693682105218099,
+ "learning_rate": 8.281128746351878e-06,
+ "loss": 0.3595161437988281,
+ "num_input_tokens_seen": 353475168,
+ "step": 8085,
+ "train_runtime": 62453.3189,
+ "train_tokens_per_second": 5659.83
+ },
+ {
+ "epoch": 0.5447077834635066,
+ "grad_norm": 0.7187659246186631,
+ "learning_rate": 8.279133156819123e-06,
+ "loss": 0.40300402641296384,
+ "num_input_tokens_seen": 353691088,
+ "step": 8090,
+ "train_runtime": 62496.7209,
+ "train_tokens_per_second": 5659.354
+ },
+ {
+ "epoch": 0.5450444384594667,
+ "grad_norm": 0.725512834705303,
+ "learning_rate": 8.277136650287237e-06,
+ "loss": 0.3694662570953369,
+ "num_input_tokens_seen": 353917984,
+ "step": 8095,
+ "train_runtime": 62536.3526,
+ "train_tokens_per_second": 5659.396
+ },
+ {
+ "epoch": 0.5453810934554268,
+ "grad_norm": 0.8301428582075541,
+ "learning_rate": 8.275139227314537e-06,
+ "loss": 0.3757363557815552,
+ "num_input_tokens_seen": 354125952,
+ "step": 8100,
+ "train_runtime": 62575.4142,
+ "train_tokens_per_second": 5659.187
+ },
+ {
+ "epoch": 0.545717748451387,
+ "grad_norm": 0.7999915594881919,
+ "learning_rate": 8.273140888459594e-06,
+ "loss": 0.4242837905883789,
+ "num_input_tokens_seen": 354358280,
+ "step": 8105,
+ "train_runtime": 62606.4679,
+ "train_tokens_per_second": 5660.091
+ },
+ {
+ "epoch": 0.5460544034473471,
+ "grad_norm": 0.9175431453565677,
+ "learning_rate": 8.27114163428124e-06,
+ "loss": 0.3934792518615723,
+ "num_input_tokens_seen": 354575664,
+ "step": 8110,
+ "train_runtime": 62642.1464,
+ "train_tokens_per_second": 5660.337
+ },
+ {
+ "epoch": 0.5463910584433073,
+ "grad_norm": 0.7408300214764466,
+ "learning_rate": 8.26914146533856e-06,
+ "loss": 0.38090806007385253,
+ "num_input_tokens_seen": 354808384,
+ "step": 8115,
+ "train_runtime": 62681.8186,
+ "train_tokens_per_second": 5660.467
+ },
+ {
+ "epoch": 0.5467277134392674,
+ "grad_norm": 0.7033405910036602,
+ "learning_rate": 8.267140382190892e-06,
+ "loss": 0.3828258037567139,
+ "num_input_tokens_seen": 355026976,
+ "step": 8120,
+ "train_runtime": 62720.5013,
+ "train_tokens_per_second": 5660.461
+ },
+ {
+ "epoch": 0.5470643684352275,
+ "grad_norm": 0.6635248689946966,
+ "learning_rate": 8.265138385397835e-06,
+ "loss": 0.3777566432952881,
+ "num_input_tokens_seen": 355244064,
+ "step": 8125,
+ "train_runtime": 62758.0924,
+ "train_tokens_per_second": 5660.53
+ },
+ {
+ "epoch": 0.5474010234311877,
+ "grad_norm": 0.6184317548675163,
+ "learning_rate": 8.263135475519241e-06,
+ "loss": 0.34886605739593507,
+ "num_input_tokens_seen": 355483272,
+ "step": 8130,
+ "train_runtime": 62794.3565,
+ "train_tokens_per_second": 5661.07
+ },
+ {
+ "epoch": 0.5477376784271478,
+ "grad_norm": 0.7287895081916465,
+ "learning_rate": 8.261131653115216e-06,
+ "loss": 0.42696261405944824,
+ "num_input_tokens_seen": 355711240,
+ "step": 8135,
+ "train_runtime": 62834.333,
+ "train_tokens_per_second": 5661.097
+ },
+ {
+ "epoch": 0.548074333423108,
+ "grad_norm": 0.7794435545394348,
+ "learning_rate": 8.259126918746122e-06,
+ "loss": 0.36356501579284667,
+ "num_input_tokens_seen": 355930576,
+ "step": 8140,
+ "train_runtime": 62868.698,
+ "train_tokens_per_second": 5661.491
+ },
+ {
+ "epoch": 0.5484109884190681,
+ "grad_norm": 0.6736849599162793,
+ "learning_rate": 8.25712127297258e-06,
+ "loss": 0.3778836250305176,
+ "num_input_tokens_seen": 356150816,
+ "step": 8145,
+ "train_runtime": 62904.5449,
+ "train_tokens_per_second": 5661.766
+ },
+ {
+ "epoch": 0.5487476434150282,
+ "grad_norm": 0.7082862793695406,
+ "learning_rate": 8.255114716355459e-06,
+ "loss": 0.4094264507293701,
+ "num_input_tokens_seen": 356382088,
+ "step": 8150,
+ "train_runtime": 62942.0397,
+ "train_tokens_per_second": 5662.068
+ },
+ {
+ "epoch": 0.5490842984109884,
+ "grad_norm": 0.6051005207909932,
+ "learning_rate": 8.253107249455887e-06,
+ "loss": 0.3917287826538086,
+ "num_input_tokens_seen": 356602352,
+ "step": 8155,
+ "train_runtime": 62975.3252,
+ "train_tokens_per_second": 5662.573
+ },
+ {
+ "epoch": 0.5494209534069485,
+ "grad_norm": 0.7713641151309204,
+ "learning_rate": 8.251098872835247e-06,
+ "loss": 0.39372997283935546,
+ "num_input_tokens_seen": 356825280,
+ "step": 8160,
+ "train_runtime": 63017.9451,
+ "train_tokens_per_second": 5662.28
+ },
+ {
+ "epoch": 0.5497576084029087,
+ "grad_norm": 0.6787537110984104,
+ "learning_rate": 8.249089587055175e-06,
+ "loss": 0.3526461601257324,
+ "num_input_tokens_seen": 357037664,
+ "step": 8165,
+ "train_runtime": 63059.5973,
+ "train_tokens_per_second": 5661.908
+ },
+ {
+ "epoch": 0.5500942633988688,
+ "grad_norm": 0.6748882087262432,
+ "learning_rate": 8.247079392677558e-06,
+ "loss": 0.3740200996398926,
+ "num_input_tokens_seen": 357257424,
+ "step": 8170,
+ "train_runtime": 63101.2168,
+ "train_tokens_per_second": 5661.657
+ },
+ {
+ "epoch": 0.5504309183948289,
+ "grad_norm": 0.7274044025423265,
+ "learning_rate": 8.245068290264544e-06,
+ "loss": 0.3697060108184814,
+ "num_input_tokens_seen": 357480144,
+ "step": 8175,
+ "train_runtime": 63140.8867,
+ "train_tokens_per_second": 5661.627
+ },
+ {
+ "epoch": 0.5507675733907891,
+ "grad_norm": 0.7084601443387462,
+ "learning_rate": 8.243056280378531e-06,
+ "loss": 0.37436251640319823,
+ "num_input_tokens_seen": 357699472,
+ "step": 8180,
+ "train_runtime": 63181.8065,
+ "train_tokens_per_second": 5661.432
+ },
+ {
+ "epoch": 0.5511042283867492,
+ "grad_norm": 0.7481357333059493,
+ "learning_rate": 8.241043363582171e-06,
+ "loss": 0.36437788009643557,
+ "num_input_tokens_seen": 357926808,
+ "step": 8185,
+ "train_runtime": 63218.2393,
+ "train_tokens_per_second": 5661.765
+ },
+ {
+ "epoch": 0.5514408833827094,
+ "grad_norm": 1.0989196799559033,
+ "learning_rate": 8.239029540438369e-06,
+ "loss": 0.3505790948867798,
+ "num_input_tokens_seen": 358145064,
+ "step": 8190,
+ "train_runtime": 63257.3639,
+ "train_tokens_per_second": 5661.713
+ },
+ {
+ "epoch": 0.5517775383786695,
+ "grad_norm": 0.7305407817056846,
+ "learning_rate": 8.237014811510284e-06,
+ "loss": 0.3656752109527588,
+ "num_input_tokens_seen": 358369824,
+ "step": 8195,
+ "train_runtime": 63304.7711,
+ "train_tokens_per_second": 5661.024
+ },
+ {
+ "epoch": 0.5521141933746296,
+ "grad_norm": 0.6472079948654392,
+ "learning_rate": 8.234999177361329e-06,
+ "loss": 0.3714550018310547,
+ "num_input_tokens_seen": 358596784,
+ "step": 8200,
+ "train_runtime": 63340.7687,
+ "train_tokens_per_second": 5661.39
+ },
+ {
+ "epoch": 0.5524508483705898,
+ "grad_norm": 0.7398046092203671,
+ "learning_rate": 8.232982638555169e-06,
+ "loss": 0.40174274444580077,
+ "num_input_tokens_seen": 358820312,
+ "step": 8205,
+ "train_runtime": 63372.9424,
+ "train_tokens_per_second": 5662.043
+ },
+ {
+ "epoch": 0.5527875033665499,
+ "grad_norm": 0.7062238114257713,
+ "learning_rate": 8.230965195655724e-06,
+ "loss": 0.39043726921081545,
+ "num_input_tokens_seen": 359055496,
+ "step": 8210,
+ "train_runtime": 63412.4919,
+ "train_tokens_per_second": 5662.22
+ },
+ {
+ "epoch": 0.55312415836251,
+ "grad_norm": 0.8164987537109517,
+ "learning_rate": 8.228946849227167e-06,
+ "loss": 0.3990142345428467,
+ "num_input_tokens_seen": 359287376,
+ "step": 8215,
+ "train_runtime": 63449.3085,
+ "train_tokens_per_second": 5662.589
+ },
+ {
+ "epoch": 0.5534608133584702,
+ "grad_norm": 0.6938755136640083,
+ "learning_rate": 8.226927599833916e-06,
+ "loss": 0.386507511138916,
+ "num_input_tokens_seen": 359514576,
+ "step": 8220,
+ "train_runtime": 63481.4052,
+ "train_tokens_per_second": 5663.305
+ },
+ {
+ "epoch": 0.5537974683544303,
+ "grad_norm": 0.6819170611796698,
+ "learning_rate": 8.224907448040655e-06,
+ "loss": 0.37993340492248534,
+ "num_input_tokens_seen": 359745160,
+ "step": 8225,
+ "train_runtime": 63513.0016,
+ "train_tokens_per_second": 5664.118
+ },
+ {
+ "epoch": 0.5541341233503905,
+ "grad_norm": 0.7464665475450891,
+ "learning_rate": 8.222886394412309e-06,
+ "loss": 0.4058523178100586,
+ "num_input_tokens_seen": 359965008,
+ "step": 8230,
+ "train_runtime": 63553.6566,
+ "train_tokens_per_second": 5663.954
+ },
+ {
+ "epoch": 0.5544707783463506,
+ "grad_norm": 0.6722302273522195,
+ "learning_rate": 8.220864439514057e-06,
+ "loss": 0.384306526184082,
+ "num_input_tokens_seen": 360194040,
+ "step": 8235,
+ "train_runtime": 63592.1988,
+ "train_tokens_per_second": 5664.123
+ },
+ {
+ "epoch": 0.5548074333423108,
+ "grad_norm": 0.6641141478245969,
+ "learning_rate": 8.218841583911338e-06,
+ "loss": 0.3680997848510742,
+ "num_input_tokens_seen": 360409248,
+ "step": 8240,
+ "train_runtime": 63634.667,
+ "train_tokens_per_second": 5663.725
+ },
+ {
+ "epoch": 0.5551440883382709,
+ "grad_norm": 0.7528778151335748,
+ "learning_rate": 8.216817828169834e-06,
+ "loss": 0.413088321685791,
+ "num_input_tokens_seen": 360624480,
+ "step": 8245,
+ "train_runtime": 63670.5285,
+ "train_tokens_per_second": 5663.915
+ },
+ {
+ "epoch": 0.555480743334231,
+ "grad_norm": 0.8364309542706772,
+ "learning_rate": 8.214793172855481e-06,
+ "loss": 0.36465370655059814,
+ "num_input_tokens_seen": 360837048,
+ "step": 8250,
+ "train_runtime": 63711.1159,
+ "train_tokens_per_second": 5663.644
+ },
+ {
+ "epoch": 0.5558173983301912,
+ "grad_norm": 0.6793083765205596,
+ "learning_rate": 8.21276761853447e-06,
+ "loss": 0.36249728202819825,
+ "num_input_tokens_seen": 361060272,
+ "step": 8255,
+ "train_runtime": 63754.6776,
+ "train_tokens_per_second": 5663.275
+ },
+ {
+ "epoch": 0.5561540533261513,
+ "grad_norm": 0.8280046723618791,
+ "learning_rate": 8.210741165773238e-06,
+ "loss": 0.39040985107421877,
+ "num_input_tokens_seen": 361294192,
+ "step": 8260,
+ "train_runtime": 63785.4799,
+ "train_tokens_per_second": 5664.207
+ },
+ {
+ "epoch": 0.5564907083221114,
+ "grad_norm": 0.7577410939409384,
+ "learning_rate": 8.208713815138476e-06,
+ "loss": 0.36567642688751223,
+ "num_input_tokens_seen": 361524240,
+ "step": 8265,
+ "train_runtime": 63826.3741,
+ "train_tokens_per_second": 5664.183
+ },
+ {
+ "epoch": 0.5568273633180716,
+ "grad_norm": 0.6700074247940968,
+ "learning_rate": 8.20668556719713e-06,
+ "loss": 0.35714569091796877,
+ "num_input_tokens_seen": 361732392,
+ "step": 8270,
+ "train_runtime": 63860.1654,
+ "train_tokens_per_second": 5664.445
+ },
+ {
+ "epoch": 0.5571640183140317,
+ "grad_norm": 0.7602580503776288,
+ "learning_rate": 8.204656422516388e-06,
+ "loss": 0.3372737646102905,
+ "num_input_tokens_seen": 361946008,
+ "step": 8275,
+ "train_runtime": 63901.2504,
+ "train_tokens_per_second": 5664.146
+ },
+ {
+ "epoch": 0.5575006733099919,
+ "grad_norm": 0.8510184000147942,
+ "learning_rate": 8.2026263816637e-06,
+ "loss": 0.38366384506225587,
+ "num_input_tokens_seen": 362152248,
+ "step": 8280,
+ "train_runtime": 63938.3717,
+ "train_tokens_per_second": 5664.083
+ },
+ {
+ "epoch": 0.557837328305952,
+ "grad_norm": 0.7613111925807169,
+ "learning_rate": 8.200595445206753e-06,
+ "loss": 0.37342875003814696,
+ "num_input_tokens_seen": 362371096,
+ "step": 8285,
+ "train_runtime": 63979.792,
+ "train_tokens_per_second": 5663.837
+ },
+ {
+ "epoch": 0.5581739833019121,
+ "grad_norm": 1.256734854809266,
+ "learning_rate": 8.198563613713497e-06,
+ "loss": 0.37401297092437746,
+ "num_input_tokens_seen": 362579576,
+ "step": 8290,
+ "train_runtime": 64020.2056,
+ "train_tokens_per_second": 5663.518
+ },
+ {
+ "epoch": 0.5585106382978723,
+ "grad_norm": 0.7385210193529426,
+ "learning_rate": 8.196530887752127e-06,
+ "loss": 0.3654287338256836,
+ "num_input_tokens_seen": 362803720,
+ "step": 8295,
+ "train_runtime": 64054.0322,
+ "train_tokens_per_second": 5664.026
+ },
+ {
+ "epoch": 0.5588472932938324,
+ "grad_norm": 0.8102882370091832,
+ "learning_rate": 8.194497267891086e-06,
+ "loss": 0.37125024795532224,
+ "num_input_tokens_seen": 363027432,
+ "step": 8300,
+ "train_runtime": 64092.3585,
+ "train_tokens_per_second": 5664.13
+ },
+ {
+ "epoch": 0.5591839482897926,
+ "grad_norm": 0.7432402409233543,
+ "learning_rate": 8.19246275469907e-06,
+ "loss": 0.3796513795852661,
+ "num_input_tokens_seen": 363254432,
+ "step": 8305,
+ "train_runtime": 64136.3098,
+ "train_tokens_per_second": 5663.788
+ },
+ {
+ "epoch": 0.5595206032857528,
+ "grad_norm": 0.6849455308836975,
+ "learning_rate": 8.190427348745027e-06,
+ "loss": 0.36992511749267576,
+ "num_input_tokens_seen": 363479976,
+ "step": 8310,
+ "train_runtime": 64171.4185,
+ "train_tokens_per_second": 5664.204
+ },
+ {
+ "epoch": 0.559857258281713,
+ "grad_norm": 0.6998379838638995,
+ "learning_rate": 8.188391050598146e-06,
+ "loss": 0.3632178544998169,
+ "num_input_tokens_seen": 363712504,
+ "step": 8315,
+ "train_runtime": 64216.5044,
+ "train_tokens_per_second": 5663.848
+ },
+ {
+ "epoch": 0.5601939132776731,
+ "grad_norm": 0.7737605990707936,
+ "learning_rate": 8.186353860827878e-06,
+ "loss": 0.37043955326080324,
+ "num_input_tokens_seen": 363919784,
+ "step": 8320,
+ "train_runtime": 64256.872,
+ "train_tokens_per_second": 5663.515
+ },
+ {
+ "epoch": 0.5605305682736332,
+ "grad_norm": 0.6957700417795362,
+ "learning_rate": 8.184315780003911e-06,
+ "loss": 0.3709959745407104,
+ "num_input_tokens_seen": 364140736,
+ "step": 8325,
+ "train_runtime": 64302.6454,
+ "train_tokens_per_second": 5662.92
+ },
+ {
+ "epoch": 0.5608672232695934,
+ "grad_norm": 0.64141155335251,
+ "learning_rate": 8.182276808696193e-06,
+ "loss": 0.40601282119750975,
+ "num_input_tokens_seen": 364348280,
+ "step": 8330,
+ "train_runtime": 64347.1042,
+ "train_tokens_per_second": 5662.233
+ },
+ {
+ "epoch": 0.5612038782655535,
+ "grad_norm": 0.6158998839242338,
+ "learning_rate": 8.18023694747491e-06,
+ "loss": 0.3827899694442749,
+ "num_input_tokens_seen": 364574968,
+ "step": 8335,
+ "train_runtime": 64381.872,
+ "train_tokens_per_second": 5662.696
+ },
+ {
+ "epoch": 0.5615405332615137,
+ "grad_norm": 0.7883261214385181,
+ "learning_rate": 8.178196196910508e-06,
+ "loss": 0.3656409740447998,
+ "num_input_tokens_seen": 364803736,
+ "step": 8340,
+ "train_runtime": 64421.7397,
+ "train_tokens_per_second": 5662.743
+ },
+ {
+ "epoch": 0.5618771882574738,
+ "grad_norm": 0.7366395620437122,
+ "learning_rate": 8.176154557573672e-06,
+ "loss": 0.4000543594360352,
+ "num_input_tokens_seen": 365037240,
+ "step": 8345,
+ "train_runtime": 64459.3416,
+ "train_tokens_per_second": 5663.062
+ },
+ {
+ "epoch": 0.5622138432534339,
+ "grad_norm": 0.6985009938577207,
+ "learning_rate": 8.174112030035344e-06,
+ "loss": 0.3629467010498047,
+ "num_input_tokens_seen": 365268200,
+ "step": 8350,
+ "train_runtime": 64493.078,
+ "train_tokens_per_second": 5663.681
+ },
+ {
+ "epoch": 0.5625504982493941,
+ "grad_norm": 0.7274783538197038,
+ "learning_rate": 8.172068614866707e-06,
+ "loss": 0.36340205669403075,
+ "num_input_tokens_seen": 365477184,
+ "step": 8355,
+ "train_runtime": 64531.8006,
+ "train_tokens_per_second": 5663.521
+ },
+ {
+ "epoch": 0.5628871532453542,
+ "grad_norm": 0.7368665647506253,
+ "learning_rate": 8.170024312639197e-06,
+ "loss": 0.39545679092407227,
+ "num_input_tokens_seen": 365707680,
+ "step": 8360,
+ "train_runtime": 64570.9604,
+ "train_tokens_per_second": 5663.656
+ },
+ {
+ "epoch": 0.5632238082413143,
+ "grad_norm": 0.7340645583889084,
+ "learning_rate": 8.167979123924496e-06,
+ "loss": 0.3859426498413086,
+ "num_input_tokens_seen": 365930048,
+ "step": 8365,
+ "train_runtime": 64604.6699,
+ "train_tokens_per_second": 5664.142
+ },
+ {
+ "epoch": 0.5635604632372745,
+ "grad_norm": 0.7462104536621238,
+ "learning_rate": 8.165933049294537e-06,
+ "loss": 0.4087996482849121,
+ "num_input_tokens_seen": 366147352,
+ "step": 8370,
+ "train_runtime": 64646.3443,
+ "train_tokens_per_second": 5663.852
+ },
+ {
+ "epoch": 0.5638971182332346,
+ "grad_norm": 0.7582565637830365,
+ "learning_rate": 8.163886089321493e-06,
+ "loss": 0.37345223426818847,
+ "num_input_tokens_seen": 366377160,
+ "step": 8375,
+ "train_runtime": 64688.6158,
+ "train_tokens_per_second": 5663.704
+ },
+ {
+ "epoch": 0.5642337732291948,
+ "grad_norm": 0.6581884199846296,
+ "learning_rate": 8.161838244577795e-06,
+ "loss": 0.3573953151702881,
+ "num_input_tokens_seen": 366596264,
+ "step": 8380,
+ "train_runtime": 64726.1749,
+ "train_tokens_per_second": 5663.802
+ },
+ {
+ "epoch": 0.5645704282251549,
+ "grad_norm": 0.7540331450420041,
+ "learning_rate": 8.159789515636114e-06,
+ "loss": 0.3621669769287109,
+ "num_input_tokens_seen": 366814672,
+ "step": 8385,
+ "train_runtime": 64768.6142,
+ "train_tokens_per_second": 5663.463
+ },
+ {
+ "epoch": 0.564907083221115,
+ "grad_norm": 0.7846236965599341,
+ "learning_rate": 8.15773990306937e-06,
+ "loss": 0.39387807846069334,
+ "num_input_tokens_seen": 367010840,
+ "step": 8390,
+ "train_runtime": 64803.6605,
+ "train_tokens_per_second": 5663.428
+ },
+ {
+ "epoch": 0.5652437382170752,
+ "grad_norm": 0.8105957793723372,
+ "learning_rate": 8.155689407450732e-06,
+ "loss": 0.37447366714477537,
+ "num_input_tokens_seen": 367235664,
+ "step": 8395,
+ "train_runtime": 64840.3814,
+ "train_tokens_per_second": 5663.688
+ },
+ {
+ "epoch": 0.5655803932130353,
+ "grad_norm": 0.7593251086886306,
+ "learning_rate": 8.153638029353615e-06,
+ "loss": 0.3721097707748413,
+ "num_input_tokens_seen": 367439472,
+ "step": 8400,
+ "train_runtime": 64880.3333,
+ "train_tokens_per_second": 5663.341
+ },
+ {
+ "epoch": 0.5659170482089955,
+ "grad_norm": 0.7080065499934168,
+ "learning_rate": 8.151585769351678e-06,
+ "loss": 0.416213321685791,
+ "num_input_tokens_seen": 367665000,
+ "step": 8405,
+ "train_runtime": 64915.8566,
+ "train_tokens_per_second": 5663.716
+ },
+ {
+ "epoch": 0.5662537032049556,
+ "grad_norm": 0.8290526825274906,
+ "learning_rate": 8.149532628018833e-06,
+ "loss": 0.3906120538711548,
+ "num_input_tokens_seen": 367873544,
+ "step": 8410,
+ "train_runtime": 64954.7801,
+ "train_tokens_per_second": 5663.533
+ },
+ {
+ "epoch": 0.5665903582009157,
+ "grad_norm": 0.7887226131302195,
+ "learning_rate": 8.147478605929228e-06,
+ "loss": 0.36734464168548586,
+ "num_input_tokens_seen": 368106424,
+ "step": 8415,
+ "train_runtime": 64988.7811,
+ "train_tokens_per_second": 5664.153
+ },
+ {
+ "epoch": 0.5669270131968759,
+ "grad_norm": 0.7105615187211392,
+ "learning_rate": 8.145423703657269e-06,
+ "loss": 0.3696259021759033,
+ "num_input_tokens_seen": 368329816,
+ "step": 8420,
+ "train_runtime": 65035.2437,
+ "train_tokens_per_second": 5663.542
+ },
+ {
+ "epoch": 0.567263668192836,
+ "grad_norm": 0.7106675383557719,
+ "learning_rate": 8.143367921777601e-06,
+ "loss": 0.3901944637298584,
+ "num_input_tokens_seen": 368556080,
+ "step": 8425,
+ "train_runtime": 65073.1284,
+ "train_tokens_per_second": 5663.722
+ },
+ {
+ "epoch": 0.5676003231887962,
+ "grad_norm": 0.7158037922920018,
+ "learning_rate": 8.141311260865117e-06,
+ "loss": 0.42763843536376955,
+ "num_input_tokens_seen": 368786216,
+ "step": 8430,
+ "train_runtime": 65107.0411,
+ "train_tokens_per_second": 5664.306
+ },
+ {
+ "epoch": 0.5679369781847563,
+ "grad_norm": 0.7147505710629092,
+ "learning_rate": 8.139253721494956e-06,
+ "loss": 0.4124480724334717,
+ "num_input_tokens_seen": 368998320,
+ "step": 8435,
+ "train_runtime": 65143.6774,
+ "train_tokens_per_second": 5664.377
+ },
+ {
+ "epoch": 0.5682736331807164,
+ "grad_norm": 0.690454346302624,
+ "learning_rate": 8.137195304242502e-06,
+ "loss": 0.3694611549377441,
+ "num_input_tokens_seen": 369221808,
+ "step": 8440,
+ "train_runtime": 65180.1258,
+ "train_tokens_per_second": 5664.638
+ },
+ {
+ "epoch": 0.5686102881766766,
+ "grad_norm": 0.6647768706585024,
+ "learning_rate": 8.135136009683387e-06,
+ "loss": 0.3659370899200439,
+ "num_input_tokens_seen": 369439952,
+ "step": 8445,
+ "train_runtime": 65214.1207,
+ "train_tokens_per_second": 5665.03
+ },
+ {
+ "epoch": 0.5689469431726367,
+ "grad_norm": 0.7123577020124194,
+ "learning_rate": 8.13307583839348e-06,
+ "loss": 0.37821600437164304,
+ "num_input_tokens_seen": 369665384,
+ "step": 8450,
+ "train_runtime": 65253.9634,
+ "train_tokens_per_second": 5665.026
+ },
+ {
+ "epoch": 0.5692835981685969,
+ "grad_norm": 0.7080847091486382,
+ "learning_rate": 8.131014790948908e-06,
+ "loss": 0.3932258367538452,
+ "num_input_tokens_seen": 369885352,
+ "step": 8455,
+ "train_runtime": 65292.5756,
+ "train_tokens_per_second": 5665.045
+ },
+ {
+ "epoch": 0.569620253164557,
+ "grad_norm": 0.8175401785926597,
+ "learning_rate": 8.128952867926035e-06,
+ "loss": 0.42728986740112307,
+ "num_input_tokens_seen": 370102168,
+ "step": 8460,
+ "train_runtime": 65326.5467,
+ "train_tokens_per_second": 5665.418
+ },
+ {
+ "epoch": 0.5699569081605171,
+ "grad_norm": 0.7636022480938197,
+ "learning_rate": 8.126890069901468e-06,
+ "loss": 0.3966690540313721,
+ "num_input_tokens_seen": 370323328,
+ "step": 8465,
+ "train_runtime": 65365.5452,
+ "train_tokens_per_second": 5665.421
+ },
+ {
+ "epoch": 0.5702935631564773,
+ "grad_norm": 0.8306888968108936,
+ "learning_rate": 8.124826397452064e-06,
+ "loss": 0.3729779005050659,
+ "num_input_tokens_seen": 370551792,
+ "step": 8470,
+ "train_runtime": 65403.3941,
+ "train_tokens_per_second": 5665.636
+ },
+ {
+ "epoch": 0.5706302181524374,
+ "grad_norm": 0.6975418197600278,
+ "learning_rate": 8.122761851154923e-06,
+ "loss": 0.37224202156066893,
+ "num_input_tokens_seen": 370769584,
+ "step": 8475,
+ "train_runtime": 65437.1957,
+ "train_tokens_per_second": 5666.037
+ },
+ {
+ "epoch": 0.5709668731483976,
+ "grad_norm": 0.7691847068239698,
+ "learning_rate": 8.120696431587387e-06,
+ "loss": 0.3793643474578857,
+ "num_input_tokens_seen": 370986856,
+ "step": 8480,
+ "train_runtime": 65477.4237,
+ "train_tokens_per_second": 5665.874
+ },
+ {
+ "epoch": 0.5713035281443577,
+ "grad_norm": 0.653136501814857,
+ "learning_rate": 8.118630139327046e-06,
+ "loss": 0.36144418716430665,
+ "num_input_tokens_seen": 371203736,
+ "step": 8485,
+ "train_runtime": 65520.6447,
+ "train_tokens_per_second": 5665.447
+ },
+ {
+ "epoch": 0.5716401831403178,
+ "grad_norm": 0.6688068417437787,
+ "learning_rate": 8.116562974951731e-06,
+ "loss": 0.3969851016998291,
+ "num_input_tokens_seen": 371433168,
+ "step": 8490,
+ "train_runtime": 65562.0835,
+ "train_tokens_per_second": 5665.366
+ },
+ {
+ "epoch": 0.571976838136278,
+ "grad_norm": 0.6693856135821081,
+ "learning_rate": 8.11449493903952e-06,
+ "loss": 0.37904407978057864,
+ "num_input_tokens_seen": 371661928,
+ "step": 8495,
+ "train_runtime": 65600.9804,
+ "train_tokens_per_second": 5665.494
+ },
+ {
+ "epoch": 0.5723134931322381,
+ "grad_norm": 0.6989787293355911,
+ "learning_rate": 8.112426032168727e-06,
+ "loss": 0.39436898231506345,
+ "num_input_tokens_seen": 371877904,
+ "step": 8500,
+ "train_runtime": 65639.9628,
+ "train_tokens_per_second": 5665.419
+ },
+ {
+ "epoch": 0.5726501481281983,
+ "grad_norm": 0.6921966447452655,
+ "learning_rate": 8.11035625491792e-06,
+ "loss": 0.35653374195098875,
+ "num_input_tokens_seen": 372083344,
+ "step": 8505,
+ "train_runtime": 65673.4132,
+ "train_tokens_per_second": 5665.662
+ },
+ {
+ "epoch": 0.5729868031241584,
+ "grad_norm": 0.7863535213377899,
+ "learning_rate": 8.108285607865907e-06,
+ "loss": 0.3748607158660889,
+ "num_input_tokens_seen": 372296008,
+ "step": 8510,
+ "train_runtime": 65709.8413,
+ "train_tokens_per_second": 5665.757
+ },
+ {
+ "epoch": 0.5733234581201185,
+ "grad_norm": 0.7103469377120155,
+ "learning_rate": 8.106214091591731e-06,
+ "loss": 0.36768136024475095,
+ "num_input_tokens_seen": 372517376,
+ "step": 8515,
+ "train_runtime": 65747.0301,
+ "train_tokens_per_second": 5665.919
+ },
+ {
+ "epoch": 0.5736601131160787,
+ "grad_norm": 0.8043097859632866,
+ "learning_rate": 8.104141706674693e-06,
+ "loss": 0.3505872249603271,
+ "num_input_tokens_seen": 372744072,
+ "step": 8520,
+ "train_runtime": 65786.4844,
+ "train_tokens_per_second": 5665.967
+ },
+ {
+ "epoch": 0.5739967681120388,
+ "grad_norm": 0.7899542715564848,
+ "learning_rate": 8.102068453694324e-06,
+ "loss": 0.39815869331359866,
+ "num_input_tokens_seen": 372965496,
+ "step": 8525,
+ "train_runtime": 65827.1156,
+ "train_tokens_per_second": 5665.834
+ },
+ {
+ "epoch": 0.574333423107999,
+ "grad_norm": 0.7007321773126729,
+ "learning_rate": 8.099994333230404e-06,
+ "loss": 0.39165263175964354,
+ "num_input_tokens_seen": 373176000,
+ "step": 8530,
+ "train_runtime": 65869.8446,
+ "train_tokens_per_second": 5665.354
+ },
+ {
+ "epoch": 0.5746700781039591,
+ "grad_norm": 0.8059453918533225,
+ "learning_rate": 8.097919345862952e-06,
+ "loss": 0.387595534324646,
+ "num_input_tokens_seen": 373375616,
+ "step": 8535,
+ "train_runtime": 65901.6062,
+ "train_tokens_per_second": 5665.653
+ },
+ {
+ "epoch": 0.5750067330999192,
+ "grad_norm": 0.7748964359890478,
+ "learning_rate": 8.095843492172235e-06,
+ "loss": 0.3876967906951904,
+ "num_input_tokens_seen": 373591328,
+ "step": 8540,
+ "train_runtime": 65941.7008,
+ "train_tokens_per_second": 5665.479
+ },
+ {
+ "epoch": 0.5753433880958794,
+ "grad_norm": 0.773534060871464,
+ "learning_rate": 8.093766772738758e-06,
+ "loss": 0.3301082611083984,
+ "num_input_tokens_seen": 373805712,
+ "step": 8545,
+ "train_runtime": 65984.3171,
+ "train_tokens_per_second": 5665.069
+ },
+ {
+ "epoch": 0.5756800430918395,
+ "grad_norm": 0.7746351168910522,
+ "learning_rate": 8.091689188143267e-06,
+ "loss": 0.3910914659500122,
+ "num_input_tokens_seen": 374017824,
+ "step": 8550,
+ "train_runtime": 66030.3105,
+ "train_tokens_per_second": 5664.335
+ },
+ {
+ "epoch": 0.5760166980877997,
+ "grad_norm": 0.7771670247365485,
+ "learning_rate": 8.089610738966754e-06,
+ "loss": 0.3839992046356201,
+ "num_input_tokens_seen": 374221088,
+ "step": 8555,
+ "train_runtime": 66067.2485,
+ "train_tokens_per_second": 5664.245
+ },
+ {
+ "epoch": 0.5763533530837598,
+ "grad_norm": 0.8314883209793272,
+ "learning_rate": 8.08753142579045e-06,
+ "loss": 0.41312179565429685,
+ "num_input_tokens_seen": 374438784,
+ "step": 8560,
+ "train_runtime": 66103.0863,
+ "train_tokens_per_second": 5664.468
+ },
+ {
+ "epoch": 0.5766900080797199,
+ "grad_norm": 0.7070329659969831,
+ "learning_rate": 8.085451249195828e-06,
+ "loss": 0.369646692276001,
+ "num_input_tokens_seen": 374674400,
+ "step": 8565,
+ "train_runtime": 66144.5198,
+ "train_tokens_per_second": 5664.481
+ },
+ {
+ "epoch": 0.5770266630756801,
+ "grad_norm": 0.728101829102978,
+ "learning_rate": 8.083370209764603e-06,
+ "loss": 0.38304858207702636,
+ "num_input_tokens_seen": 374895048,
+ "step": 8570,
+ "train_runtime": 66187.4516,
+ "train_tokens_per_second": 5664.141
+ },
+ {
+ "epoch": 0.5773633180716402,
+ "grad_norm": 0.7613212980235913,
+ "learning_rate": 8.081288308078731e-06,
+ "loss": 0.39395363330841066,
+ "num_input_tokens_seen": 375130432,
+ "step": 8575,
+ "train_runtime": 66220.5986,
+ "train_tokens_per_second": 5664.86
+ },
+ {
+ "epoch": 0.5776999730676003,
+ "grad_norm": 0.6846318822730251,
+ "learning_rate": 8.07920554472041e-06,
+ "loss": 0.35710582733154295,
+ "num_input_tokens_seen": 375356416,
+ "step": 8580,
+ "train_runtime": 66266.8733,
+ "train_tokens_per_second": 5664.315
+ },
+ {
+ "epoch": 0.5780366280635605,
+ "grad_norm": 0.791144672174043,
+ "learning_rate": 8.077121920272075e-06,
+ "loss": 0.3807799816131592,
+ "num_input_tokens_seen": 375586768,
+ "step": 8585,
+ "train_runtime": 66303.4383,
+ "train_tokens_per_second": 5664.665
+ },
+ {
+ "epoch": 0.5783732830595206,
+ "grad_norm": 0.6749521518269885,
+ "learning_rate": 8.07503743531641e-06,
+ "loss": 0.3493949890136719,
+ "num_input_tokens_seen": 375796816,
+ "step": 8590,
+ "train_runtime": 66347.6746,
+ "train_tokens_per_second": 5664.054
+ },
+ {
+ "epoch": 0.5787099380554808,
+ "grad_norm": 0.7233256642537537,
+ "learning_rate": 8.072952090436328e-06,
+ "loss": 0.34188032150268555,
+ "num_input_tokens_seen": 376025856,
+ "step": 8595,
+ "train_runtime": 66379.0077,
+ "train_tokens_per_second": 5664.831
+ },
+ {
+ "epoch": 0.5790465930514409,
+ "grad_norm": 0.8183957573318358,
+ "learning_rate": 8.070865886214996e-06,
+ "loss": 0.378603458404541,
+ "num_input_tokens_seen": 376248416,
+ "step": 8600,
+ "train_runtime": 66412.9945,
+ "train_tokens_per_second": 5665.283
+ },
+ {
+ "epoch": 0.579383248047401,
+ "grad_norm": 0.8303736324901154,
+ "learning_rate": 8.06877882323581e-06,
+ "loss": 0.3776103496551514,
+ "num_input_tokens_seen": 376468952,
+ "step": 8605,
+ "train_runtime": 66445.0637,
+ "train_tokens_per_second": 5665.868
+ },
+ {
+ "epoch": 0.5797199030433612,
+ "grad_norm": 0.6628257697173621,
+ "learning_rate": 8.066690902082413e-06,
+ "loss": 0.35461578369140623,
+ "num_input_tokens_seen": 376690720,
+ "step": 8610,
+ "train_runtime": 66482.3424,
+ "train_tokens_per_second": 5666.027
+ },
+ {
+ "epoch": 0.5800565580393213,
+ "grad_norm": 0.6914119662414026,
+ "learning_rate": 8.064602123338683e-06,
+ "loss": 0.4189821720123291,
+ "num_input_tokens_seen": 376926952,
+ "step": 8615,
+ "train_runtime": 66523.5883,
+ "train_tokens_per_second": 5666.065
+ },
+ {
+ "epoch": 0.5803932130352815,
+ "grad_norm": 0.7255952843828343,
+ "learning_rate": 8.06251248758874e-06,
+ "loss": 0.3706305980682373,
+ "num_input_tokens_seen": 377137424,
+ "step": 8620,
+ "train_runtime": 66560.9464,
+ "train_tokens_per_second": 5666.047
+ },
+ {
+ "epoch": 0.5807298680312416,
+ "grad_norm": 0.7888860114164491,
+ "learning_rate": 8.060421995416949e-06,
+ "loss": 0.4162487030029297,
+ "num_input_tokens_seen": 377341792,
+ "step": 8625,
+ "train_runtime": 66601.9228,
+ "train_tokens_per_second": 5665.629
+ },
+ {
+ "epoch": 0.5810665230272017,
+ "grad_norm": 0.8097744521376653,
+ "learning_rate": 8.058330647407904e-06,
+ "loss": 0.3742199420928955,
+ "num_input_tokens_seen": 377553480,
+ "step": 8630,
+ "train_runtime": 66640.7232,
+ "train_tokens_per_second": 5665.507
+ },
+ {
+ "epoch": 0.5814031780231619,
+ "grad_norm": 0.9311380768738048,
+ "learning_rate": 8.056238444146445e-06,
+ "loss": 0.4032620429992676,
+ "num_input_tokens_seen": 377768576,
+ "step": 8635,
+ "train_runtime": 66676.5105,
+ "train_tokens_per_second": 5665.692
+ },
+ {
+ "epoch": 0.581739833019122,
+ "grad_norm": 0.7908376307807878,
+ "learning_rate": 8.054145386217652e-06,
+ "loss": 0.35233349800109864,
+ "num_input_tokens_seen": 377990912,
+ "step": 8640,
+ "train_runtime": 66709.357,
+ "train_tokens_per_second": 5666.235
+ },
+ {
+ "epoch": 0.5820764880150822,
+ "grad_norm": 0.7544609324314604,
+ "learning_rate": 8.05205147420684e-06,
+ "loss": 0.36639404296875,
+ "num_input_tokens_seen": 378205832,
+ "step": 8645,
+ "train_runtime": 66742.4566,
+ "train_tokens_per_second": 5666.645
+ },
+ {
+ "epoch": 0.5824131430110423,
+ "grad_norm": 0.8852523330609511,
+ "learning_rate": 8.049956708699567e-06,
+ "loss": 0.41215500831604,
+ "num_input_tokens_seen": 378422632,
+ "step": 8650,
+ "train_runtime": 66779.3643,
+ "train_tokens_per_second": 5666.76
+ },
+ {
+ "epoch": 0.5827497980070024,
+ "grad_norm": 0.7748134898838014,
+ "learning_rate": 8.047861090281623e-06,
+ "loss": 0.4017993450164795,
+ "num_input_tokens_seen": 378636312,
+ "step": 8655,
+ "train_runtime": 66817.0207,
+ "train_tokens_per_second": 5666.764
+ },
+ {
+ "epoch": 0.5830864530029626,
+ "grad_norm": 0.7174936151093502,
+ "learning_rate": 8.045764619539047e-06,
+ "loss": 0.36281633377075195,
+ "num_input_tokens_seen": 378862480,
+ "step": 8660,
+ "train_runtime": 66859.5287,
+ "train_tokens_per_second": 5666.544
+ },
+ {
+ "epoch": 0.5834231079989227,
+ "grad_norm": 0.7177129776779663,
+ "learning_rate": 8.043667297058106e-06,
+ "loss": 0.3568310022354126,
+ "num_input_tokens_seen": 379073648,
+ "step": 8665,
+ "train_runtime": 66894.7683,
+ "train_tokens_per_second": 5666.716
+ },
+ {
+ "epoch": 0.5837597629948829,
+ "grad_norm": 0.671557492383721,
+ "learning_rate": 8.04156912342531e-06,
+ "loss": 0.3924489259719849,
+ "num_input_tokens_seen": 379296048,
+ "step": 8670,
+ "train_runtime": 66930.6354,
+ "train_tokens_per_second": 5667.002
+ },
+ {
+ "epoch": 0.584096417990843,
+ "grad_norm": 0.8338696138817213,
+ "learning_rate": 8.03947009922741e-06,
+ "loss": 0.3730125904083252,
+ "num_input_tokens_seen": 379514216,
+ "step": 8675,
+ "train_runtime": 66971.0506,
+ "train_tokens_per_second": 5666.84
+ },
+ {
+ "epoch": 0.5844330729868031,
+ "grad_norm": 0.8046977070640711,
+ "learning_rate": 8.037370225051387e-06,
+ "loss": 0.4212483406066895,
+ "num_input_tokens_seen": 379726496,
+ "step": 8680,
+ "train_runtime": 67011.5775,
+ "train_tokens_per_second": 5666.58
+ },
+ {
+ "epoch": 0.5847697279827633,
+ "grad_norm": 0.747137316814288,
+ "learning_rate": 8.035269501484467e-06,
+ "loss": 0.3820772886276245,
+ "num_input_tokens_seen": 379952664,
+ "step": 8685,
+ "train_runtime": 67045.976,
+ "train_tokens_per_second": 5667.047
+ },
+ {
+ "epoch": 0.5851063829787234,
+ "grad_norm": 0.7779439978184742,
+ "learning_rate": 8.03316792911411e-06,
+ "loss": 0.41135897636413576,
+ "num_input_tokens_seen": 380168720,
+ "step": 8690,
+ "train_runtime": 67086.3352,
+ "train_tokens_per_second": 5666.858
+ },
+ {
+ "epoch": 0.5854430379746836,
+ "grad_norm": 0.6815474179189787,
+ "learning_rate": 8.031065508528011e-06,
+ "loss": 0.35835254192352295,
+ "num_input_tokens_seen": 380396160,
+ "step": 8695,
+ "train_runtime": 67124.515,
+ "train_tokens_per_second": 5667.023
+ },
+ {
+ "epoch": 0.5857796929706437,
+ "grad_norm": 0.7133086557979279,
+ "learning_rate": 8.028962240314111e-06,
+ "loss": 0.3686984539031982,
+ "num_input_tokens_seen": 380620472,
+ "step": 8700,
+ "train_runtime": 67161.6095,
+ "train_tokens_per_second": 5667.233
+ },
+ {
+ "epoch": 0.5861163479666038,
+ "grad_norm": 0.7088226284029326,
+ "learning_rate": 8.026858125060577e-06,
+ "loss": 0.3879531383514404,
+ "num_input_tokens_seen": 380837592,
+ "step": 8705,
+ "train_runtime": 67201.0534,
+ "train_tokens_per_second": 5667.137
+ },
+ {
+ "epoch": 0.586453002962564,
+ "grad_norm": 0.7612804968151702,
+ "learning_rate": 8.024753163355821e-06,
+ "loss": 0.4247321128845215,
+ "num_input_tokens_seen": 381061216,
+ "step": 8710,
+ "train_runtime": 67242.9084,
+ "train_tokens_per_second": 5666.935
+ },
+ {
+ "epoch": 0.5867896579585241,
+ "grad_norm": 0.7005358220776335,
+ "learning_rate": 8.022647355788489e-06,
+ "loss": 0.3679481029510498,
+ "num_input_tokens_seen": 381280288,
+ "step": 8715,
+ "train_runtime": 67281.4067,
+ "train_tokens_per_second": 5666.949
+ },
+ {
+ "epoch": 0.5871263129544843,
+ "grad_norm": 0.728044784671465,
+ "learning_rate": 8.02054070294746e-06,
+ "loss": 0.40247626304626466,
+ "num_input_tokens_seen": 381496992,
+ "step": 8720,
+ "train_runtime": 67319.4589,
+ "train_tokens_per_second": 5666.965
+ },
+ {
+ "epoch": 0.5874629679504444,
+ "grad_norm": 0.7052792472604399,
+ "learning_rate": 8.018433205421859e-06,
+ "loss": 0.36869523525238035,
+ "num_input_tokens_seen": 381727008,
+ "step": 8725,
+ "train_runtime": 67359.8794,
+ "train_tokens_per_second": 5666.979
+ },
+ {
+ "epoch": 0.5877996229464045,
+ "grad_norm": 0.8148613184034657,
+ "learning_rate": 8.016324863801036e-06,
+ "loss": 0.3896385669708252,
+ "num_input_tokens_seen": 381950608,
+ "step": 8730,
+ "train_runtime": 67395.2795,
+ "train_tokens_per_second": 5667.32
+ },
+ {
+ "epoch": 0.5881362779423647,
+ "grad_norm": 0.7008873936428064,
+ "learning_rate": 8.014215678674581e-06,
+ "loss": 0.397304368019104,
+ "num_input_tokens_seen": 382173664,
+ "step": 8735,
+ "train_runtime": 67436.6612,
+ "train_tokens_per_second": 5667.15
+ },
+ {
+ "epoch": 0.5884729329383248,
+ "grad_norm": 0.6667026328000061,
+ "learning_rate": 8.012105650632325e-06,
+ "loss": 0.37210235595703123,
+ "num_input_tokens_seen": 382388568,
+ "step": 8740,
+ "train_runtime": 67476.723,
+ "train_tokens_per_second": 5666.97
+ },
+ {
+ "epoch": 0.588809587934285,
+ "grad_norm": 0.7893753618716253,
+ "learning_rate": 8.009994780264328e-06,
+ "loss": 0.38750429153442384,
+ "num_input_tokens_seen": 382584944,
+ "step": 8745,
+ "train_runtime": 67517.8366,
+ "train_tokens_per_second": 5666.428
+ },
+ {
+ "epoch": 0.5891462429302451,
+ "grad_norm": 0.7520095638811525,
+ "learning_rate": 8.00788306816089e-06,
+ "loss": 0.38128883838653566,
+ "num_input_tokens_seen": 382801808,
+ "step": 8750,
+ "train_runtime": 67554.4917,
+ "train_tokens_per_second": 5666.563
+ },
+ {
+ "epoch": 0.5894828979262052,
+ "grad_norm": 0.7841152997217926,
+ "learning_rate": 8.005770514912543e-06,
+ "loss": 0.39379191398620605,
+ "num_input_tokens_seen": 383025832,
+ "step": 8755,
+ "train_runtime": 67593.0084,
+ "train_tokens_per_second": 5666.649
+ },
+ {
+ "epoch": 0.5898195529221654,
+ "grad_norm": 0.728445499817985,
+ "learning_rate": 8.003657121110055e-06,
+ "loss": 0.3845071792602539,
+ "num_input_tokens_seen": 383260728,
+ "step": 8760,
+ "train_runtime": 67628.415,
+ "train_tokens_per_second": 5667.155
+ },
+ {
+ "epoch": 0.5901562079181255,
+ "grad_norm": 0.7109388567699519,
+ "learning_rate": 8.001542887344432e-06,
+ "loss": 0.36870722770690917,
+ "num_input_tokens_seen": 383481016,
+ "step": 8765,
+ "train_runtime": 67663.424,
+ "train_tokens_per_second": 5667.479
+ },
+ {
+ "epoch": 0.5904928629140856,
+ "grad_norm": 0.7379119456141738,
+ "learning_rate": 7.999427814206911e-06,
+ "loss": 0.38972065448760984,
+ "num_input_tokens_seen": 383699928,
+ "step": 8770,
+ "train_runtime": 67698.1314,
+ "train_tokens_per_second": 5667.807
+ },
+ {
+ "epoch": 0.5908295179100458,
+ "grad_norm": 0.7977299068456677,
+ "learning_rate": 7.997311902288968e-06,
+ "loss": 0.3479362964630127,
+ "num_input_tokens_seen": 383923208,
+ "step": 8775,
+ "train_runtime": 67735.622,
+ "train_tokens_per_second": 5667.966
+ },
+ {
+ "epoch": 0.5911661729060059,
+ "grad_norm": 0.6927052461533478,
+ "learning_rate": 7.995195152182307e-06,
+ "loss": 0.35241219997406004,
+ "num_input_tokens_seen": 384124912,
+ "step": 8780,
+ "train_runtime": 67775.1001,
+ "train_tokens_per_second": 5667.641
+ },
+ {
+ "epoch": 0.5915028279019661,
+ "grad_norm": 1.8520827992951523,
+ "learning_rate": 7.993077564478875e-06,
+ "loss": 0.367362380027771,
+ "num_input_tokens_seen": 384352936,
+ "step": 8785,
+ "train_runtime": 67813.9355,
+ "train_tokens_per_second": 5667.757
+ },
+ {
+ "epoch": 0.5918394828979262,
+ "grad_norm": 0.8136091755899332,
+ "learning_rate": 7.990959139770844e-06,
+ "loss": 0.3980868339538574,
+ "num_input_tokens_seen": 384586952,
+ "step": 8790,
+ "train_runtime": 67855.5637,
+ "train_tokens_per_second": 5667.729
+ },
+ {
+ "epoch": 0.5921761378938863,
+ "grad_norm": 0.7689488310801835,
+ "learning_rate": 7.988839878650628e-06,
+ "loss": 0.3829834461212158,
+ "num_input_tokens_seen": 384813912,
+ "step": 8795,
+ "train_runtime": 67899.407,
+ "train_tokens_per_second": 5667.412
+ },
+ {
+ "epoch": 0.5925127928898465,
+ "grad_norm": 0.7981067639524714,
+ "learning_rate": 7.986719781710872e-06,
+ "loss": 0.3936153411865234,
+ "num_input_tokens_seen": 385040600,
+ "step": 8800,
+ "train_runtime": 67941.2356,
+ "train_tokens_per_second": 5667.259
+ },
+ {
+ "epoch": 0.5928494478858066,
+ "grad_norm": 0.7888638773984543,
+ "learning_rate": 7.984598849544452e-06,
+ "loss": 0.37601919174194337,
+ "num_input_tokens_seen": 385256736,
+ "step": 8805,
+ "train_runtime": 67980.1555,
+ "train_tokens_per_second": 5667.194
+ },
+ {
+ "epoch": 0.5931861028817668,
+ "grad_norm": 0.8141275660860262,
+ "learning_rate": 7.98247708274448e-06,
+ "loss": 0.3897037744522095,
+ "num_input_tokens_seen": 385471192,
+ "step": 8810,
+ "train_runtime": 68015.2704,
+ "train_tokens_per_second": 5667.421
+ },
+ {
+ "epoch": 0.5935227578777269,
+ "grad_norm": 0.7312305067797448,
+ "learning_rate": 7.980354481904303e-06,
+ "loss": 0.4111155033111572,
+ "num_input_tokens_seen": 385695968,
+ "step": 8815,
+ "train_runtime": 68053.7353,
+ "train_tokens_per_second": 5667.521
+ },
+ {
+ "epoch": 0.593859412873687,
+ "grad_norm": 0.8597141063820571,
+ "learning_rate": 7.978231047617498e-06,
+ "loss": 0.41021156311035156,
+ "num_input_tokens_seen": 385910576,
+ "step": 8820,
+ "train_runtime": 68091.813,
+ "train_tokens_per_second": 5667.503
+ },
+ {
+ "epoch": 0.5941960678696472,
+ "grad_norm": 0.7977576371807428,
+ "learning_rate": 7.976106780477878e-06,
+ "loss": 0.4051943778991699,
+ "num_input_tokens_seen": 386135568,
+ "step": 8825,
+ "train_runtime": 68134.8361,
+ "train_tokens_per_second": 5667.227
+ },
+ {
+ "epoch": 0.5945327228656073,
+ "grad_norm": 0.7369358873224232,
+ "learning_rate": 7.973981681079487e-06,
+ "loss": 0.3718585729598999,
+ "num_input_tokens_seen": 386344944,
+ "step": 8830,
+ "train_runtime": 68175.9892,
+ "train_tokens_per_second": 5666.877
+ },
+ {
+ "epoch": 0.5948693778615675,
+ "grad_norm": 0.7361018680183479,
+ "learning_rate": 7.971855750016603e-06,
+ "loss": 0.37234504222869874,
+ "num_input_tokens_seen": 386561056,
+ "step": 8835,
+ "train_runtime": 68210.534,
+ "train_tokens_per_second": 5667.175
+ },
+ {
+ "epoch": 0.5952060328575276,
+ "grad_norm": 0.7646438523333068,
+ "learning_rate": 7.969728987883734e-06,
+ "loss": 0.386487340927124,
+ "num_input_tokens_seen": 386782376,
+ "step": 8840,
+ "train_runtime": 68250.0947,
+ "train_tokens_per_second": 5667.133
+ },
+ {
+ "epoch": 0.5955426878534877,
+ "grad_norm": 0.8191747882763178,
+ "learning_rate": 7.967601395275624e-06,
+ "loss": 0.38187594413757325,
+ "num_input_tokens_seen": 386999696,
+ "step": 8845,
+ "train_runtime": 68291.3193,
+ "train_tokens_per_second": 5666.894
+ },
+ {
+ "epoch": 0.5958793428494479,
+ "grad_norm": 0.6459146656871505,
+ "learning_rate": 7.965472972787247e-06,
+ "loss": 0.35875325202941893,
+ "num_input_tokens_seen": 387220408,
+ "step": 8850,
+ "train_runtime": 68328.0647,
+ "train_tokens_per_second": 5667.077
+ },
+ {
+ "epoch": 0.596215997845408,
+ "grad_norm": 0.7549464702278598,
+ "learning_rate": 7.963343721013808e-06,
+ "loss": 0.3718303680419922,
+ "num_input_tokens_seen": 387455840,
+ "step": 8855,
+ "train_runtime": 68366.913,
+ "train_tokens_per_second": 5667.3
+ },
+ {
+ "epoch": 0.5965526528413682,
+ "grad_norm": 0.7558527061932778,
+ "learning_rate": 7.961213640550748e-06,
+ "loss": 0.3902780771255493,
+ "num_input_tokens_seen": 387678720,
+ "step": 8860,
+ "train_runtime": 68410.3443,
+ "train_tokens_per_second": 5666.961
+ },
+ {
+ "epoch": 0.5968893078373283,
+ "grad_norm": 0.7082248157089,
+ "learning_rate": 7.959082731993736e-06,
+ "loss": 0.3638461589813232,
+ "num_input_tokens_seen": 387897464,
+ "step": 8865,
+ "train_runtime": 68443.0142,
+ "train_tokens_per_second": 5667.452
+ },
+ {
+ "epoch": 0.5972259628332884,
+ "grad_norm": 0.9079660049109616,
+ "learning_rate": 7.956950995938673e-06,
+ "loss": 0.3705005168914795,
+ "num_input_tokens_seen": 388107624,
+ "step": 8870,
+ "train_runtime": 68486.4151,
+ "train_tokens_per_second": 5666.929
+ },
+ {
+ "epoch": 0.5975626178292486,
+ "grad_norm": 0.7984526639513112,
+ "learning_rate": 7.954818432981694e-06,
+ "loss": 0.3910086631774902,
+ "num_input_tokens_seen": 388310680,
+ "step": 8875,
+ "train_runtime": 68525.481,
+ "train_tokens_per_second": 5666.661
+ },
+ {
+ "epoch": 0.5978992728252087,
+ "grad_norm": 0.7147181194308313,
+ "learning_rate": 7.952685043719162e-06,
+ "loss": 0.3656719923019409,
+ "num_input_tokens_seen": 388533832,
+ "step": 8880,
+ "train_runtime": 68568.5855,
+ "train_tokens_per_second": 5666.353
+ },
+ {
+ "epoch": 0.5982359278211689,
+ "grad_norm": 0.7782537183487828,
+ "learning_rate": 7.950550828747672e-06,
+ "loss": 0.37486002445220945,
+ "num_input_tokens_seen": 388754728,
+ "step": 8885,
+ "train_runtime": 68608.906,
+ "train_tokens_per_second": 5666.243
+ },
+ {
+ "epoch": 0.598572582817129,
+ "grad_norm": 0.6958894420319796,
+ "learning_rate": 7.948415788664052e-06,
+ "loss": 0.40519165992736816,
+ "num_input_tokens_seen": 388977056,
+ "step": 8890,
+ "train_runtime": 68641.9616,
+ "train_tokens_per_second": 5666.753
+ },
+ {
+ "epoch": 0.5989092378130891,
+ "grad_norm": 0.8244159660705994,
+ "learning_rate": 7.946279924065359e-06,
+ "loss": 0.3696116924285889,
+ "num_input_tokens_seen": 389195368,
+ "step": 8895,
+ "train_runtime": 68676.2104,
+ "train_tokens_per_second": 5667.106
+ },
+ {
+ "epoch": 0.5992458928090493,
+ "grad_norm": 0.7624772559058,
+ "learning_rate": 7.94414323554888e-06,
+ "loss": 0.3721637725830078,
+ "num_input_tokens_seen": 389424256,
+ "step": 8900,
+ "train_runtime": 68707.9383,
+ "train_tokens_per_second": 5667.82
+ },
+ {
+ "epoch": 0.5995825478050094,
+ "grad_norm": 0.7421871285525378,
+ "learning_rate": 7.942005723712135e-06,
+ "loss": 0.3689994812011719,
+ "num_input_tokens_seen": 389659456,
+ "step": 8905,
+ "train_runtime": 68742.7631,
+ "train_tokens_per_second": 5668.371
+ },
+ {
+ "epoch": 0.5999192028009696,
+ "grad_norm": 0.7598715011111157,
+ "learning_rate": 7.939867389152868e-06,
+ "loss": 0.36181044578552246,
+ "num_input_tokens_seen": 389878008,
+ "step": 8910,
+ "train_runtime": 68782.9286,
+ "train_tokens_per_second": 5668.238
+ },
+ {
+ "epoch": 0.6002558577969297,
+ "grad_norm": 0.7642969305957489,
+ "learning_rate": 7.937728232469062e-06,
+ "loss": 0.3904815435409546,
+ "num_input_tokens_seen": 390090680,
+ "step": 8915,
+ "train_runtime": 68823.5271,
+ "train_tokens_per_second": 5667.984
+ },
+ {
+ "epoch": 0.6005925127928898,
+ "grad_norm": 0.7743252489033806,
+ "learning_rate": 7.935588254258921e-06,
+ "loss": 0.407269811630249,
+ "num_input_tokens_seen": 390298200,
+ "step": 8920,
+ "train_runtime": 68859.6954,
+ "train_tokens_per_second": 5668.021
+ },
+ {
+ "epoch": 0.60092916778885,
+ "grad_norm": 0.7626396676106256,
+ "learning_rate": 7.933447455120889e-06,
+ "loss": 0.3707296848297119,
+ "num_input_tokens_seen": 390513112,
+ "step": 8925,
+ "train_runtime": 68906.8163,
+ "train_tokens_per_second": 5667.264
+ },
+ {
+ "epoch": 0.6012658227848101,
+ "grad_norm": 0.9270450940086636,
+ "learning_rate": 7.931305835653629e-06,
+ "loss": 0.34468765258789064,
+ "num_input_tokens_seen": 390741944,
+ "step": 8930,
+ "train_runtime": 68945.5046,
+ "train_tokens_per_second": 5667.403
+ },
+ {
+ "epoch": 0.6016024777807703,
+ "grad_norm": 0.7523711200271979,
+ "learning_rate": 7.929163396456038e-06,
+ "loss": 0.39664325714111326,
+ "num_input_tokens_seen": 390950400,
+ "step": 8935,
+ "train_runtime": 68982.333,
+ "train_tokens_per_second": 5667.399
+ },
+ {
+ "epoch": 0.6019391327767304,
+ "grad_norm": 0.7743378155770866,
+ "learning_rate": 7.927020138127245e-06,
+ "loss": 0.369353199005127,
+ "num_input_tokens_seen": 391172832,
+ "step": 8940,
+ "train_runtime": 69022.2251,
+ "train_tokens_per_second": 5667.346
+ },
+ {
+ "epoch": 0.6022757877726905,
+ "grad_norm": 0.6649737691389936,
+ "learning_rate": 7.924876061266603e-06,
+ "loss": 0.3477231502532959,
+ "num_input_tokens_seen": 391403392,
+ "step": 8945,
+ "train_runtime": 69060.7271,
+ "train_tokens_per_second": 5667.525
+ },
+ {
+ "epoch": 0.6026124427686507,
+ "grad_norm": 0.8428633382608517,
+ "learning_rate": 7.922731166473698e-06,
+ "loss": 0.40044050216674804,
+ "num_input_tokens_seen": 391619536,
+ "step": 8950,
+ "train_runtime": 69095.3564,
+ "train_tokens_per_second": 5667.813
+ },
+ {
+ "epoch": 0.6029490977646108,
+ "grad_norm": 0.7797139187172423,
+ "learning_rate": 7.920585454348341e-06,
+ "loss": 0.35779447555541993,
+ "num_input_tokens_seen": 391832424,
+ "step": 8955,
+ "train_runtime": 69130.8816,
+ "train_tokens_per_second": 5667.98
+ },
+ {
+ "epoch": 0.603285752760571,
+ "grad_norm": 0.6454426330600102,
+ "learning_rate": 7.918438925490578e-06,
+ "loss": 0.35099799633026124,
+ "num_input_tokens_seen": 392057736,
+ "step": 8960,
+ "train_runtime": 69168.5952,
+ "train_tokens_per_second": 5668.147
+ },
+ {
+ "epoch": 0.6036224077565311,
+ "grad_norm": 0.9240196513971288,
+ "learning_rate": 7.916291580500672e-06,
+ "loss": 0.37104010581970215,
+ "num_input_tokens_seen": 392274152,
+ "step": 8965,
+ "train_runtime": 69209.1039,
+ "train_tokens_per_second": 5667.956
+ },
+ {
+ "epoch": 0.6039590627524912,
+ "grad_norm": 0.753245828305525,
+ "learning_rate": 7.914143419979127e-06,
+ "loss": 0.36745285987854004,
+ "num_input_tokens_seen": 392499120,
+ "step": 8970,
+ "train_runtime": 69247.1372,
+ "train_tokens_per_second": 5668.092
+ },
+ {
+ "epoch": 0.6042957177484514,
+ "grad_norm": 0.6569111554819693,
+ "learning_rate": 7.911994444526667e-06,
+ "loss": 0.34250531196594236,
+ "num_input_tokens_seen": 392720016,
+ "step": 8975,
+ "train_runtime": 69281.6006,
+ "train_tokens_per_second": 5668.46
+ },
+ {
+ "epoch": 0.6046323727444115,
+ "grad_norm": 0.6710386941472357,
+ "learning_rate": 7.909844654744246e-06,
+ "loss": 0.3526078939437866,
+ "num_input_tokens_seen": 392945160,
+ "step": 8980,
+ "train_runtime": 69318.5011,
+ "train_tokens_per_second": 5668.691
+ },
+ {
+ "epoch": 0.6049690277403716,
+ "grad_norm": 0.7627659966441698,
+ "learning_rate": 7.907694051233044e-06,
+ "loss": 0.38707795143127444,
+ "num_input_tokens_seen": 393164952,
+ "step": 8985,
+ "train_runtime": 69355.9491,
+ "train_tokens_per_second": 5668.799
+ },
+ {
+ "epoch": 0.6053056827363318,
+ "grad_norm": 0.6996994069888867,
+ "learning_rate": 7.905542634594476e-06,
+ "loss": 0.3759605884552002,
+ "num_input_tokens_seen": 393392000,
+ "step": 8990,
+ "train_runtime": 69403.0227,
+ "train_tokens_per_second": 5668.226
+ },
+ {
+ "epoch": 0.6056423377322919,
+ "grad_norm": 0.7505686616004086,
+ "learning_rate": 7.903390405430174e-06,
+ "loss": 0.36977381706237794,
+ "num_input_tokens_seen": 393611632,
+ "step": 8995,
+ "train_runtime": 69443.7718,
+ "train_tokens_per_second": 5668.062
+ },
+ {
+ "epoch": 0.6059789927282521,
+ "grad_norm": 0.7461690774426525,
+ "learning_rate": 7.901237364342003e-06,
+ "loss": 0.3545074939727783,
+ "num_input_tokens_seen": 393813184,
+ "step": 9000,
+ "train_runtime": 69478.2997,
+ "train_tokens_per_second": 5668.147
+ },
+ {
+ "epoch": 0.6063156477242122,
+ "grad_norm": 0.692733389183626,
+ "learning_rate": 7.899083511932053e-06,
+ "loss": 0.3470455646514893,
+ "num_input_tokens_seen": 394027272,
+ "step": 9005,
+ "train_runtime": 69518.7288,
+ "train_tokens_per_second": 5667.93
+ },
+ {
+ "epoch": 0.6066523027201723,
+ "grad_norm": 0.7055838759339507,
+ "learning_rate": 7.896928848802645e-06,
+ "loss": 0.372318959236145,
+ "num_input_tokens_seen": 394252480,
+ "step": 9010,
+ "train_runtime": 69553.6455,
+ "train_tokens_per_second": 5668.322
+ },
+ {
+ "epoch": 0.6069889577161325,
+ "grad_norm": 0.7576301760274441,
+ "learning_rate": 7.894773375556322e-06,
+ "loss": 0.39665966033935546,
+ "num_input_tokens_seen": 394449024,
+ "step": 9015,
+ "train_runtime": 69593.8001,
+ "train_tokens_per_second": 5667.876
+ },
+ {
+ "epoch": 0.6073256127120926,
+ "grad_norm": 0.8095166727188953,
+ "learning_rate": 7.892617092795855e-06,
+ "loss": 0.38990225791931155,
+ "num_input_tokens_seen": 394672680,
+ "step": 9020,
+ "train_runtime": 69632.6302,
+ "train_tokens_per_second": 5667.927
+ },
+ {
+ "epoch": 0.6076622677080528,
+ "grad_norm": 0.6966710594837849,
+ "learning_rate": 7.890460001124242e-06,
+ "loss": 0.3693479061126709,
+ "num_input_tokens_seen": 394882720,
+ "step": 9025,
+ "train_runtime": 69668.2129,
+ "train_tokens_per_second": 5668.047
+ },
+ {
+ "epoch": 0.6079989227040129,
+ "grad_norm": 0.7467966463857733,
+ "learning_rate": 7.888302101144707e-06,
+ "loss": 0.3727553367614746,
+ "num_input_tokens_seen": 395096336,
+ "step": 9030,
+ "train_runtime": 69710.5053,
+ "train_tokens_per_second": 5667.673
+ },
+ {
+ "epoch": 0.608335577699973,
+ "grad_norm": 0.7033388740227385,
+ "learning_rate": 7.886143393460699e-06,
+ "loss": 0.38645596504211427,
+ "num_input_tokens_seen": 395327104,
+ "step": 9035,
+ "train_runtime": 69746.6807,
+ "train_tokens_per_second": 5668.042
+ },
+ {
+ "epoch": 0.6086722326959332,
+ "grad_norm": 0.717725417977876,
+ "learning_rate": 7.883983878675895e-06,
+ "loss": 0.3641960620880127,
+ "num_input_tokens_seen": 395561000,
+ "step": 9040,
+ "train_runtime": 69794.3801,
+ "train_tokens_per_second": 5667.519
+ },
+ {
+ "epoch": 0.6090088876918933,
+ "grad_norm": 0.737338173324883,
+ "learning_rate": 7.881823557394194e-06,
+ "loss": 0.3701021194458008,
+ "num_input_tokens_seen": 395790544,
+ "step": 9045,
+ "train_runtime": 69832.741,
+ "train_tokens_per_second": 5667.693
+ },
+ {
+ "epoch": 0.6093455426878535,
+ "grad_norm": 0.7527693128608011,
+ "learning_rate": 7.879662430219728e-06,
+ "loss": 0.35156404972076416,
+ "num_input_tokens_seen": 395987680,
+ "step": 9050,
+ "train_runtime": 69873.2033,
+ "train_tokens_per_second": 5667.232
+ },
+ {
+ "epoch": 0.6096821976838136,
+ "grad_norm": 0.8175940814840023,
+ "learning_rate": 7.877500497756843e-06,
+ "loss": 0.4006368637084961,
+ "num_input_tokens_seen": 396207224,
+ "step": 9055,
+ "train_runtime": 69914.8014,
+ "train_tokens_per_second": 5667.001
+ },
+ {
+ "epoch": 0.6100188526797737,
+ "grad_norm": 1.0466330679614098,
+ "learning_rate": 7.87533776061012e-06,
+ "loss": 0.3919075012207031,
+ "num_input_tokens_seen": 396431144,
+ "step": 9060,
+ "train_runtime": 69954.8331,
+ "train_tokens_per_second": 5666.959
+ },
+ {
+ "epoch": 0.6103555076757339,
+ "grad_norm": 0.7304596174173837,
+ "learning_rate": 7.873174219384362e-06,
+ "loss": 0.3918348789215088,
+ "num_input_tokens_seen": 396647760,
+ "step": 9065,
+ "train_runtime": 69996.3796,
+ "train_tokens_per_second": 5666.69
+ },
+ {
+ "epoch": 0.610692162671694,
+ "grad_norm": 0.7556787705072924,
+ "learning_rate": 7.871009874684595e-06,
+ "loss": 0.3941289186477661,
+ "num_input_tokens_seen": 396880640,
+ "step": 9070,
+ "train_runtime": 70028.8034,
+ "train_tokens_per_second": 5667.391
+ },
+ {
+ "epoch": 0.6110288176676542,
+ "grad_norm": 0.8397284206623034,
+ "learning_rate": 7.868844727116072e-06,
+ "loss": 0.3926485300064087,
+ "num_input_tokens_seen": 397103976,
+ "step": 9075,
+ "train_runtime": 70068.2945,
+ "train_tokens_per_second": 5667.385
+ },
+ {
+ "epoch": 0.6113654726636143,
+ "grad_norm": 0.6869244210168536,
+ "learning_rate": 7.866678777284267e-06,
+ "loss": 0.3840303897857666,
+ "num_input_tokens_seen": 397331160,
+ "step": 9080,
+ "train_runtime": 70102.9706,
+ "train_tokens_per_second": 5667.822
+ },
+ {
+ "epoch": 0.6117021276595744,
+ "grad_norm": 0.6366270970878026,
+ "learning_rate": 7.864512025794886e-06,
+ "loss": 0.3659891843795776,
+ "num_input_tokens_seen": 397555992,
+ "step": 9085,
+ "train_runtime": 70142.9573,
+ "train_tokens_per_second": 5667.796
+ },
+ {
+ "epoch": 0.6120387826555346,
+ "grad_norm": 0.7702173454778595,
+ "learning_rate": 7.86234447325385e-06,
+ "loss": 0.36196537017822267,
+ "num_input_tokens_seen": 397769720,
+ "step": 9090,
+ "train_runtime": 70185.3981,
+ "train_tokens_per_second": 5667.414
+ },
+ {
+ "epoch": 0.6123754376514947,
+ "grad_norm": 0.6638241416823153,
+ "learning_rate": 7.860176120267309e-06,
+ "loss": 0.4144433498382568,
+ "num_input_tokens_seen": 398006352,
+ "step": 9095,
+ "train_runtime": 70227.0746,
+ "train_tokens_per_second": 5667.42
+ },
+ {
+ "epoch": 0.6127120926474549,
+ "grad_norm": 0.7106604037911711,
+ "learning_rate": 7.858006967441637e-06,
+ "loss": 0.3602044105529785,
+ "num_input_tokens_seen": 398216992,
+ "step": 9100,
+ "train_runtime": 70268.2952,
+ "train_tokens_per_second": 5667.093
+ },
+ {
+ "epoch": 0.613048747643415,
+ "grad_norm": 0.8013079858059877,
+ "learning_rate": 7.85583701538343e-06,
+ "loss": 0.40687074661254885,
+ "num_input_tokens_seen": 398447240,
+ "step": 9105,
+ "train_runtime": 70313.64,
+ "train_tokens_per_second": 5666.713
+ },
+ {
+ "epoch": 0.6133854026393751,
+ "grad_norm": 0.8363243678158666,
+ "learning_rate": 7.853666264699506e-06,
+ "loss": 0.3638491153717041,
+ "num_input_tokens_seen": 398656928,
+ "step": 9110,
+ "train_runtime": 70355.8426,
+ "train_tokens_per_second": 5666.295
+ },
+ {
+ "epoch": 0.6137220576353353,
+ "grad_norm": 0.6855947504133412,
+ "learning_rate": 7.851494715996914e-06,
+ "loss": 0.41129069328308104,
+ "num_input_tokens_seen": 398870368,
+ "step": 9115,
+ "train_runtime": 70395.581,
+ "train_tokens_per_second": 5666.128
+ },
+ {
+ "epoch": 0.6140587126312954,
+ "grad_norm": 0.7613022857869545,
+ "learning_rate": 7.849322369882915e-06,
+ "loss": 0.3685478687286377,
+ "num_input_tokens_seen": 399088536,
+ "step": 9120,
+ "train_runtime": 70444.8636,
+ "train_tokens_per_second": 5665.261
+ },
+ {
+ "epoch": 0.6143953676272556,
+ "grad_norm": 0.6988300008519205,
+ "learning_rate": 7.847149226965003e-06,
+ "loss": 0.34872984886169434,
+ "num_input_tokens_seen": 399301232,
+ "step": 9125,
+ "train_runtime": 70485.0506,
+ "train_tokens_per_second": 5665.049
+ },
+ {
+ "epoch": 0.6147320226232157,
+ "grad_norm": 0.8477724491109673,
+ "learning_rate": 7.844975287850886e-06,
+ "loss": 0.35025968551635744,
+ "num_input_tokens_seen": 399523504,
+ "step": 9130,
+ "train_runtime": 70522.5255,
+ "train_tokens_per_second": 5665.19
+ },
+ {
+ "epoch": 0.6150686776191758,
+ "grad_norm": 0.7692634619323344,
+ "learning_rate": 7.842800553148506e-06,
+ "loss": 0.3998849630355835,
+ "num_input_tokens_seen": 399735256,
+ "step": 9135,
+ "train_runtime": 70558.7444,
+ "train_tokens_per_second": 5665.283
+ },
+ {
+ "epoch": 0.615405332615136,
+ "grad_norm": 0.7313920740237436,
+ "learning_rate": 7.840625023466014e-06,
+ "loss": 0.3812243938446045,
+ "num_input_tokens_seen": 399958824,
+ "step": 9140,
+ "train_runtime": 70598.2086,
+ "train_tokens_per_second": 5665.283
+ },
+ {
+ "epoch": 0.6157419876110961,
+ "grad_norm": 0.7274387978674163,
+ "learning_rate": 7.838448699411792e-06,
+ "loss": 0.3413673400878906,
+ "num_input_tokens_seen": 400181528,
+ "step": 9145,
+ "train_runtime": 70641.2437,
+ "train_tokens_per_second": 5664.984
+ },
+ {
+ "epoch": 0.6160786426070562,
+ "grad_norm": 0.7006449570449917,
+ "learning_rate": 7.836271581594442e-06,
+ "loss": 0.4210306167602539,
+ "num_input_tokens_seen": 400404728,
+ "step": 9150,
+ "train_runtime": 70677.6291,
+ "train_tokens_per_second": 5665.226
+ },
+ {
+ "epoch": 0.6164152976030164,
+ "grad_norm": 0.8552267430665972,
+ "learning_rate": 7.834093670622789e-06,
+ "loss": 0.3944708824157715,
+ "num_input_tokens_seen": 400620128,
+ "step": 9155,
+ "train_runtime": 70718.8549,
+ "train_tokens_per_second": 5664.969
+ },
+ {
+ "epoch": 0.6167519525989765,
+ "grad_norm": 0.7631479691553084,
+ "learning_rate": 7.831914967105878e-06,
+ "loss": 0.4071969985961914,
+ "num_input_tokens_seen": 400851832,
+ "step": 9160,
+ "train_runtime": 70759.3063,
+ "train_tokens_per_second": 5665.005
+ },
+ {
+ "epoch": 0.6170886075949367,
+ "grad_norm": 0.6980894707116475,
+ "learning_rate": 7.829735471652978e-06,
+ "loss": 0.3579000234603882,
+ "num_input_tokens_seen": 401070744,
+ "step": 9165,
+ "train_runtime": 70799.3413,
+ "train_tokens_per_second": 5664.894
+ },
+ {
+ "epoch": 0.6174252625908968,
+ "grad_norm": 0.8778577602385942,
+ "learning_rate": 7.827555184873575e-06,
+ "loss": 0.38968710899353026,
+ "num_input_tokens_seen": 401294816,
+ "step": 9170,
+ "train_runtime": 70831.4898,
+ "train_tokens_per_second": 5665.486
+ },
+ {
+ "epoch": 0.617761917586857,
+ "grad_norm": 0.6780563561113029,
+ "learning_rate": 7.825374107377381e-06,
+ "loss": 0.35830237865448,
+ "num_input_tokens_seen": 401509200,
+ "step": 9175,
+ "train_runtime": 70873.5987,
+ "train_tokens_per_second": 5665.145
+ },
+ {
+ "epoch": 0.6180985725828171,
+ "grad_norm": 0.7956649065370722,
+ "learning_rate": 7.823192239774327e-06,
+ "loss": 0.3784201145172119,
+ "num_input_tokens_seen": 401720584,
+ "step": 9180,
+ "train_runtime": 70913.365,
+ "train_tokens_per_second": 5664.949
+ },
+ {
+ "epoch": 0.6184352275787772,
+ "grad_norm": 0.689595538429294,
+ "learning_rate": 7.821009582674567e-06,
+ "loss": 0.3932128667831421,
+ "num_input_tokens_seen": 401937000,
+ "step": 9185,
+ "train_runtime": 70950.5951,
+ "train_tokens_per_second": 5665.026
+ },
+ {
+ "epoch": 0.6187718825747374,
+ "grad_norm": 0.7661957206566802,
+ "learning_rate": 7.818826136688473e-06,
+ "loss": 0.41967573165893557,
+ "num_input_tokens_seen": 402163016,
+ "step": 9190,
+ "train_runtime": 70988.3051,
+ "train_tokens_per_second": 5665.201
+ },
+ {
+ "epoch": 0.6191085375706975,
+ "grad_norm": 1.24632241828128,
+ "learning_rate": 7.816641902426638e-06,
+ "loss": 0.3871551513671875,
+ "num_input_tokens_seen": 402390608,
+ "step": 9195,
+ "train_runtime": 71026.2689,
+ "train_tokens_per_second": 5665.377
+ },
+ {
+ "epoch": 0.6194451925666576,
+ "grad_norm": 0.7754009003596685,
+ "learning_rate": 7.814456880499877e-06,
+ "loss": 0.39989705085754396,
+ "num_input_tokens_seen": 402608440,
+ "step": 9200,
+ "train_runtime": 71059.0686,
+ "train_tokens_per_second": 5665.828
+ },
+ {
+ "epoch": 0.6197818475626178,
+ "grad_norm": 0.8050073203615641,
+ "learning_rate": 7.812271071519222e-06,
+ "loss": 0.40456953048706057,
+ "num_input_tokens_seen": 402830072,
+ "step": 9205,
+ "train_runtime": 71092.5653,
+ "train_tokens_per_second": 5666.276
+ },
+ {
+ "epoch": 0.6201185025585779,
+ "grad_norm": 0.636871155540424,
+ "learning_rate": 7.810084476095932e-06,
+ "loss": 0.3537304401397705,
+ "num_input_tokens_seen": 403051544,
+ "step": 9210,
+ "train_runtime": 71132.4539,
+ "train_tokens_per_second": 5666.212
+ },
+ {
+ "epoch": 0.6204551575545381,
+ "grad_norm": 0.7771337802144807,
+ "learning_rate": 7.807897094841476e-06,
+ "loss": 0.3726407527923584,
+ "num_input_tokens_seen": 403260152,
+ "step": 9215,
+ "train_runtime": 71171.5092,
+ "train_tokens_per_second": 5666.033
+ },
+ {
+ "epoch": 0.6207918125504982,
+ "grad_norm": 0.8018367108579978,
+ "learning_rate": 7.805708928367555e-06,
+ "loss": 0.3734896183013916,
+ "num_input_tokens_seen": 403468824,
+ "step": 9220,
+ "train_runtime": 71211.2179,
+ "train_tokens_per_second": 5665.804
+ },
+ {
+ "epoch": 0.6211284675464583,
+ "grad_norm": 0.7409189256900955,
+ "learning_rate": 7.803519977286075e-06,
+ "loss": 0.4021895885467529,
+ "num_input_tokens_seen": 403677704,
+ "step": 9225,
+ "train_runtime": 71249.4495,
+ "train_tokens_per_second": 5665.696
+ },
+ {
+ "epoch": 0.6214651225424185,
+ "grad_norm": 0.7435590926911755,
+ "learning_rate": 7.801330242209176e-06,
+ "loss": 0.38091113567352297,
+ "num_input_tokens_seen": 403883200,
+ "step": 9230,
+ "train_runtime": 71290.1392,
+ "train_tokens_per_second": 5665.345
+ },
+ {
+ "epoch": 0.6218017775383786,
+ "grad_norm": 0.7623979983718003,
+ "learning_rate": 7.799139723749207e-06,
+ "loss": 0.3860699415206909,
+ "num_input_tokens_seen": 404093128,
+ "step": 9235,
+ "train_runtime": 71328.4839,
+ "train_tokens_per_second": 5665.242
+ },
+ {
+ "epoch": 0.6221384325343388,
+ "grad_norm": 0.752221691602039,
+ "learning_rate": 7.79694842251874e-06,
+ "loss": 0.3864139080047607,
+ "num_input_tokens_seen": 404319728,
+ "step": 9240,
+ "train_runtime": 71368.1022,
+ "train_tokens_per_second": 5665.272
+ },
+ {
+ "epoch": 0.6224750875302989,
+ "grad_norm": 0.669956786900923,
+ "learning_rate": 7.794756339130565e-06,
+ "loss": 0.3734030485153198,
+ "num_input_tokens_seen": 404541968,
+ "step": 9245,
+ "train_runtime": 71411.2121,
+ "train_tokens_per_second": 5664.964
+ },
+ {
+ "epoch": 0.622811742526259,
+ "grad_norm": 0.6983661416716165,
+ "learning_rate": 7.792563474197692e-06,
+ "loss": 0.3814718008041382,
+ "num_input_tokens_seen": 404754312,
+ "step": 9250,
+ "train_runtime": 71448.6745,
+ "train_tokens_per_second": 5664.966
+ },
+ {
+ "epoch": 0.6231483975222192,
+ "grad_norm": 0.7124131828610318,
+ "learning_rate": 7.79036982833335e-06,
+ "loss": 0.4030300617218018,
+ "num_input_tokens_seen": 404985704,
+ "step": 9255,
+ "train_runtime": 71479.1768,
+ "train_tokens_per_second": 5665.786
+ },
+ {
+ "epoch": 0.6234850525181793,
+ "grad_norm": 0.7886397068317147,
+ "learning_rate": 7.78817540215098e-06,
+ "loss": 0.3688653469085693,
+ "num_input_tokens_seen": 405208848,
+ "step": 9260,
+ "train_runtime": 71523.5445,
+ "train_tokens_per_second": 5665.391
+ },
+ {
+ "epoch": 0.6238217075141395,
+ "grad_norm": 0.7219385997208987,
+ "learning_rate": 7.78598019626425e-06,
+ "loss": 0.37583611011505125,
+ "num_input_tokens_seen": 405447552,
+ "step": 9265,
+ "train_runtime": 71555.7936,
+ "train_tokens_per_second": 5666.174
+ },
+ {
+ "epoch": 0.6241583625100996,
+ "grad_norm": 0.7695265603997767,
+ "learning_rate": 7.783784211287044e-06,
+ "loss": 0.4006516456604004,
+ "num_input_tokens_seen": 405651112,
+ "step": 9270,
+ "train_runtime": 71597.6416,
+ "train_tokens_per_second": 5665.705
+ },
+ {
+ "epoch": 0.6244950175060597,
+ "grad_norm": 0.9194045285642082,
+ "learning_rate": 7.78158744783346e-06,
+ "loss": 0.408467960357666,
+ "num_input_tokens_seen": 405873152,
+ "step": 9275,
+ "train_runtime": 71636.7797,
+ "train_tokens_per_second": 5665.709
+ },
+ {
+ "epoch": 0.6248316725020199,
+ "grad_norm": 0.7515447419413074,
+ "learning_rate": 7.779389906517813e-06,
+ "loss": 0.35638930797576907,
+ "num_input_tokens_seen": 406086632,
+ "step": 9280,
+ "train_runtime": 71670.2143,
+ "train_tokens_per_second": 5666.045
+ },
+ {
+ "epoch": 0.6251683274979801,
+ "grad_norm": 0.7184945703842209,
+ "learning_rate": 7.777191587954645e-06,
+ "loss": 0.3749531745910645,
+ "num_input_tokens_seen": 406298704,
+ "step": 9285,
+ "train_runtime": 71712.7407,
+ "train_tokens_per_second": 5665.642
+ },
+ {
+ "epoch": 0.6255049824939403,
+ "grad_norm": 0.8027326683853181,
+ "learning_rate": 7.774992492758704e-06,
+ "loss": 0.36041698455810545,
+ "num_input_tokens_seen": 406502360,
+ "step": 9290,
+ "train_runtime": 71750.3732,
+ "train_tokens_per_second": 5665.509
+ },
+ {
+ "epoch": 0.6258416374899004,
+ "grad_norm": 0.7186651340067036,
+ "learning_rate": 7.77279262154496e-06,
+ "loss": 0.38366947174072263,
+ "num_input_tokens_seen": 406727224,
+ "step": 9295,
+ "train_runtime": 71789.6701,
+ "train_tokens_per_second": 5665.54
+ },
+ {
+ "epoch": 0.6261782924858605,
+ "grad_norm": 0.8295991455328212,
+ "learning_rate": 7.770591974928604e-06,
+ "loss": 0.3997042655944824,
+ "num_input_tokens_seen": 406958480,
+ "step": 9300,
+ "train_runtime": 71824.8324,
+ "train_tokens_per_second": 5665.986
+ },
+ {
+ "epoch": 0.6265149474818207,
+ "grad_norm": 0.7658676057217327,
+ "learning_rate": 7.768390553525034e-06,
+ "loss": 0.3705620527267456,
+ "num_input_tokens_seen": 407186440,
+ "step": 9305,
+ "train_runtime": 71865.0099,
+ "train_tokens_per_second": 5665.99
+ },
+ {
+ "epoch": 0.6268516024777808,
+ "grad_norm": 0.7085619378845062,
+ "learning_rate": 7.766188357949875e-06,
+ "loss": 0.3714077711105347,
+ "num_input_tokens_seen": 407407560,
+ "step": 9310,
+ "train_runtime": 71909.2798,
+ "train_tokens_per_second": 5665.577
+ },
+ {
+ "epoch": 0.627188257473741,
+ "grad_norm": 0.8341480094252892,
+ "learning_rate": 7.763985388818963e-06,
+ "loss": 0.36980538368225097,
+ "num_input_tokens_seen": 407628232,
+ "step": 9315,
+ "train_runtime": 71949.141,
+ "train_tokens_per_second": 5665.505
+ },
+ {
+ "epoch": 0.6275249124697011,
+ "grad_norm": 0.8636262815852889,
+ "learning_rate": 7.76178164674835e-06,
+ "loss": 0.3548977613449097,
+ "num_input_tokens_seen": 407838184,
+ "step": 9320,
+ "train_runtime": 71979.0743,
+ "train_tokens_per_second": 5666.066
+ },
+ {
+ "epoch": 0.6278615674656612,
+ "grad_norm": 0.7192127713238392,
+ "learning_rate": 7.759577132354305e-06,
+ "loss": 0.3763836622238159,
+ "num_input_tokens_seen": 408065528,
+ "step": 9325,
+ "train_runtime": 72013.1228,
+ "train_tokens_per_second": 5666.544
+ },
+ {
+ "epoch": 0.6281982224616214,
+ "grad_norm": 0.7987348676964026,
+ "learning_rate": 7.757371846253315e-06,
+ "loss": 0.3501241683959961,
+ "num_input_tokens_seen": 408275816,
+ "step": 9330,
+ "train_runtime": 72052.1024,
+ "train_tokens_per_second": 5666.397
+ },
+ {
+ "epoch": 0.6285348774575815,
+ "grad_norm": 0.7470954183186648,
+ "learning_rate": 7.755165789062083e-06,
+ "loss": 0.4022618293762207,
+ "num_input_tokens_seen": 408498408,
+ "step": 9335,
+ "train_runtime": 72090.7302,
+ "train_tokens_per_second": 5666.448
+ },
+ {
+ "epoch": 0.6288715324535417,
+ "grad_norm": 0.7151603552727446,
+ "learning_rate": 7.752958961397522e-06,
+ "loss": 0.3877957105636597,
+ "num_input_tokens_seen": 408733352,
+ "step": 9340,
+ "train_runtime": 72120.5104,
+ "train_tokens_per_second": 5667.366
+ },
+ {
+ "epoch": 0.6292081874495018,
+ "grad_norm": 0.7351213604942984,
+ "learning_rate": 7.750751363876765e-06,
+ "loss": 0.39035401344299314,
+ "num_input_tokens_seen": 408941920,
+ "step": 9345,
+ "train_runtime": 72164.8249,
+ "train_tokens_per_second": 5666.776
+ },
+ {
+ "epoch": 0.6295448424454619,
+ "grad_norm": 0.8453874832584365,
+ "learning_rate": 7.748542997117162e-06,
+ "loss": 0.3737621307373047,
+ "num_input_tokens_seen": 409162768,
+ "step": 9350,
+ "train_runtime": 72196.2605,
+ "train_tokens_per_second": 5667.368
+ },
+ {
+ "epoch": 0.6298814974414221,
+ "grad_norm": 0.7429854652224369,
+ "learning_rate": 7.746333861736275e-06,
+ "loss": 0.37030327320098877,
+ "num_input_tokens_seen": 409372800,
+ "step": 9355,
+ "train_runtime": 72230.736,
+ "train_tokens_per_second": 5667.571
+ },
+ {
+ "epoch": 0.6302181524373822,
+ "grad_norm": 0.8081848245845143,
+ "learning_rate": 7.74412395835188e-06,
+ "loss": 0.3631394386291504,
+ "num_input_tokens_seen": 409583128,
+ "step": 9360,
+ "train_runtime": 72272.2265,
+ "train_tokens_per_second": 5667.227
+ },
+ {
+ "epoch": 0.6305548074333424,
+ "grad_norm": 0.6576701373107874,
+ "learning_rate": 7.74191328758197e-06,
+ "loss": 0.37232241630554197,
+ "num_input_tokens_seen": 409794192,
+ "step": 9365,
+ "train_runtime": 72313.6306,
+ "train_tokens_per_second": 5666.901
+ },
+ {
+ "epoch": 0.6308914624293025,
+ "grad_norm": 0.735743623178803,
+ "learning_rate": 7.739701850044752e-06,
+ "loss": 0.4075340270996094,
+ "num_input_tokens_seen": 410020536,
+ "step": 9370,
+ "train_runtime": 72350.0803,
+ "train_tokens_per_second": 5667.175
+ },
+ {
+ "epoch": 0.6312281174252626,
+ "grad_norm": 0.8241305453975977,
+ "learning_rate": 7.73748964635865e-06,
+ "loss": 0.3718372583389282,
+ "num_input_tokens_seen": 410241072,
+ "step": 9375,
+ "train_runtime": 72387.3064,
+ "train_tokens_per_second": 5667.307
+ },
+ {
+ "epoch": 0.6315647724212228,
+ "grad_norm": 0.6824612940355289,
+ "learning_rate": 7.735276677142297e-06,
+ "loss": 0.35228054523468016,
+ "num_input_tokens_seen": 410443760,
+ "step": 9380,
+ "train_runtime": 72423.3738,
+ "train_tokens_per_second": 5667.283
+ },
+ {
+ "epoch": 0.6319014274171829,
+ "grad_norm": 0.7420817513182338,
+ "learning_rate": 7.733062943014542e-06,
+ "loss": 0.3992206573486328,
+ "num_input_tokens_seen": 410670656,
+ "step": 9385,
+ "train_runtime": 72463.1943,
+ "train_tokens_per_second": 5667.3
+ },
+ {
+ "epoch": 0.632238082413143,
+ "grad_norm": 0.7090881615954795,
+ "learning_rate": 7.730848444594452e-06,
+ "loss": 0.3573267698287964,
+ "num_input_tokens_seen": 410903448,
+ "step": 9390,
+ "train_runtime": 72497.2497,
+ "train_tokens_per_second": 5667.849
+ },
+ {
+ "epoch": 0.6325747374091032,
+ "grad_norm": 0.7269251761731511,
+ "learning_rate": 7.7286331825013e-06,
+ "loss": 0.37146763801574706,
+ "num_input_tokens_seen": 411126128,
+ "step": 9395,
+ "train_runtime": 72531.5005,
+ "train_tokens_per_second": 5668.242
+ },
+ {
+ "epoch": 0.6329113924050633,
+ "grad_norm": 0.8641784035147556,
+ "learning_rate": 7.72641715735458e-06,
+ "loss": 0.3655902624130249,
+ "num_input_tokens_seen": 411336720,
+ "step": 9400,
+ "train_runtime": 72578.3797,
+ "train_tokens_per_second": 5667.483
+ },
+ {
+ "epoch": 0.6332480474010235,
+ "grad_norm": 0.7283249670371048,
+ "learning_rate": 7.724200369773998e-06,
+ "loss": 0.38719449043273924,
+ "num_input_tokens_seen": 411563872,
+ "step": 9405,
+ "train_runtime": 72618.0679,
+ "train_tokens_per_second": 5667.513
+ },
+ {
+ "epoch": 0.6335847023969836,
+ "grad_norm": 0.6863677356579423,
+ "learning_rate": 7.721982820379467e-06,
+ "loss": 0.37484476566314695,
+ "num_input_tokens_seen": 411787952,
+ "step": 9410,
+ "train_runtime": 72652.1342,
+ "train_tokens_per_second": 5667.94
+ },
+ {
+ "epoch": 0.6339213573929438,
+ "grad_norm": 0.7975234970060876,
+ "learning_rate": 7.719764509791123e-06,
+ "loss": 0.3741923332214355,
+ "num_input_tokens_seen": 411975928,
+ "step": 9415,
+ "train_runtime": 72685.292,
+ "train_tokens_per_second": 5667.941
+ },
+ {
+ "epoch": 0.6342580123889039,
+ "grad_norm": 0.735648228811728,
+ "learning_rate": 7.717545438629303e-06,
+ "loss": 0.43533172607421877,
+ "num_input_tokens_seen": 412214368,
+ "step": 9420,
+ "train_runtime": 72721.9474,
+ "train_tokens_per_second": 5668.363
+ },
+ {
+ "epoch": 0.634594667384864,
+ "grad_norm": 0.6852408582111305,
+ "learning_rate": 7.71532560751457e-06,
+ "loss": 0.3732184410095215,
+ "num_input_tokens_seen": 412434712,
+ "step": 9425,
+ "train_runtime": 72757.9756,
+ "train_tokens_per_second": 5668.584
+ },
+ {
+ "epoch": 0.6349313223808242,
+ "grad_norm": 0.7315681955128508,
+ "learning_rate": 7.713105017067686e-06,
+ "loss": 0.38740081787109376,
+ "num_input_tokens_seen": 412668752,
+ "step": 9430,
+ "train_runtime": 72791.6587,
+ "train_tokens_per_second": 5669.176
+ },
+ {
+ "epoch": 0.6352679773767843,
+ "grad_norm": 0.840365118006204,
+ "learning_rate": 7.710883667909639e-06,
+ "loss": 0.37178239822387693,
+ "num_input_tokens_seen": 412886888,
+ "step": 9435,
+ "train_runtime": 72827.5502,
+ "train_tokens_per_second": 5669.378
+ },
+ {
+ "epoch": 0.6356046323727444,
+ "grad_norm": 0.7096620735032497,
+ "learning_rate": 7.708661560661617e-06,
+ "loss": 0.3967950105667114,
+ "num_input_tokens_seen": 413111704,
+ "step": 9440,
+ "train_runtime": 72863.4631,
+ "train_tokens_per_second": 5669.669
+ },
+ {
+ "epoch": 0.6359412873687046,
+ "grad_norm": 0.7646005775520524,
+ "learning_rate": 7.706438695945026e-06,
+ "loss": 0.39545307159423826,
+ "num_input_tokens_seen": 413330656,
+ "step": 9445,
+ "train_runtime": 72901.529,
+ "train_tokens_per_second": 5669.712
+ },
+ {
+ "epoch": 0.6362779423646647,
+ "grad_norm": 0.7774443776922321,
+ "learning_rate": 7.704215074381484e-06,
+ "loss": 0.3966789960861206,
+ "num_input_tokens_seen": 413537792,
+ "step": 9450,
+ "train_runtime": 72937.8522,
+ "train_tokens_per_second": 5669.728
+ },
+ {
+ "epoch": 0.6366145973606249,
+ "grad_norm": 0.7509437079381308,
+ "learning_rate": 7.70199069659282e-06,
+ "loss": 0.38933987617492677,
+ "num_input_tokens_seen": 413751760,
+ "step": 9455,
+ "train_runtime": 72973.609,
+ "train_tokens_per_second": 5669.882
+ },
+ {
+ "epoch": 0.636951252356585,
+ "grad_norm": 0.7325824670692431,
+ "learning_rate": 7.699765563201075e-06,
+ "loss": 0.3942718982696533,
+ "num_input_tokens_seen": 413978016,
+ "step": 9460,
+ "train_runtime": 73003.9087,
+ "train_tokens_per_second": 5670.628
+ },
+ {
+ "epoch": 0.6372879073525451,
+ "grad_norm": 1.211650286474132,
+ "learning_rate": 7.697539674828498e-06,
+ "loss": 0.36391506195068357,
+ "num_input_tokens_seen": 414211120,
+ "step": 9465,
+ "train_runtime": 73043.0392,
+ "train_tokens_per_second": 5670.782
+ },
+ {
+ "epoch": 0.6376245623485053,
+ "grad_norm": 0.7046659464816453,
+ "learning_rate": 7.695313032097554e-06,
+ "loss": 0.35442147254943845,
+ "num_input_tokens_seen": 414430048,
+ "step": 9470,
+ "train_runtime": 73079.8991,
+ "train_tokens_per_second": 5670.917
+ },
+ {
+ "epoch": 0.6379612173444654,
+ "grad_norm": 0.6947182191976057,
+ "learning_rate": 7.693085635630913e-06,
+ "loss": 0.37603862285614015,
+ "num_input_tokens_seen": 414632320,
+ "step": 9475,
+ "train_runtime": 73126.3077,
+ "train_tokens_per_second": 5670.084
+ },
+ {
+ "epoch": 0.6382978723404256,
+ "grad_norm": 0.7520528416471214,
+ "learning_rate": 7.690857486051463e-06,
+ "loss": 0.3694300174713135,
+ "num_input_tokens_seen": 414859816,
+ "step": 9480,
+ "train_runtime": 73166.4344,
+ "train_tokens_per_second": 5670.084
+ },
+ {
+ "epoch": 0.6386345273363857,
+ "grad_norm": 0.7317534743877758,
+ "learning_rate": 7.688628583982298e-06,
+ "loss": 0.3734642267227173,
+ "num_input_tokens_seen": 415071856,
+ "step": 9485,
+ "train_runtime": 73208.2652,
+ "train_tokens_per_second": 5669.74
+ },
+ {
+ "epoch": 0.6389711823323458,
+ "grad_norm": 0.7042485073751668,
+ "learning_rate": 7.68639893004672e-06,
+ "loss": 0.3597568035125732,
+ "num_input_tokens_seen": 415294888,
+ "step": 9490,
+ "train_runtime": 73249.0074,
+ "train_tokens_per_second": 5669.632
+ },
+ {
+ "epoch": 0.639307837328306,
+ "grad_norm": 0.7092737777153273,
+ "learning_rate": 7.684168524868253e-06,
+ "loss": 0.3566474437713623,
+ "num_input_tokens_seen": 415526408,
+ "step": 9495,
+ "train_runtime": 73288.7331,
+ "train_tokens_per_second": 5669.717
+ },
+ {
+ "epoch": 0.6396444923242661,
+ "grad_norm": 0.788876930336864,
+ "learning_rate": 7.681937369070612e-06,
+ "loss": 0.37108328342437746,
+ "num_input_tokens_seen": 415741912,
+ "step": 9500,
+ "train_runtime": 73321.3834,
+ "train_tokens_per_second": 5670.132
+ },
+ {
+ "epoch": 0.6399811473202263,
+ "grad_norm": 0.7482496645801756,
+ "learning_rate": 7.67970546327774e-06,
+ "loss": 0.3800150156021118,
+ "num_input_tokens_seen": 415962440,
+ "step": 9505,
+ "train_runtime": 73359.2213,
+ "train_tokens_per_second": 5670.213
+ },
+ {
+ "epoch": 0.6403178023161864,
+ "grad_norm": 0.7003781550580643,
+ "learning_rate": 7.67747280811378e-06,
+ "loss": 0.3659864902496338,
+ "num_input_tokens_seen": 416179208,
+ "step": 9510,
+ "train_runtime": 73399.2881,
+ "train_tokens_per_second": 5670.071
+ },
+ {
+ "epoch": 0.6406544573121465,
+ "grad_norm": 0.6823118163683989,
+ "learning_rate": 7.675239404203088e-06,
+ "loss": 0.35662126541137695,
+ "num_input_tokens_seen": 416394640,
+ "step": 9515,
+ "train_runtime": 73439.5055,
+ "train_tokens_per_second": 5669.9
+ },
+ {
+ "epoch": 0.6409911123081067,
+ "grad_norm": 0.7167030632794295,
+ "learning_rate": 7.673005252170226e-06,
+ "loss": 0.3727959156036377,
+ "num_input_tokens_seen": 416620784,
+ "step": 9520,
+ "train_runtime": 73484.0265,
+ "train_tokens_per_second": 5669.542
+ },
+ {
+ "epoch": 0.6413277673040668,
+ "grad_norm": 0.7593758397546905,
+ "learning_rate": 7.670770352639968e-06,
+ "loss": 0.3673388957977295,
+ "num_input_tokens_seen": 416826168,
+ "step": 9525,
+ "train_runtime": 73525.2289,
+ "train_tokens_per_second": 5669.158
+ },
+ {
+ "epoch": 0.641664422300027,
+ "grad_norm": 0.7362340230809692,
+ "learning_rate": 7.668534706237299e-06,
+ "loss": 0.37580122947692873,
+ "num_input_tokens_seen": 417047480,
+ "step": 9530,
+ "train_runtime": 73556.9783,
+ "train_tokens_per_second": 5669.72
+ },
+ {
+ "epoch": 0.6420010772959871,
+ "grad_norm": 0.7543938330712561,
+ "learning_rate": 7.666298313587409e-06,
+ "loss": 0.4170052528381348,
+ "num_input_tokens_seen": 417268504,
+ "step": 9535,
+ "train_runtime": 73597.5716,
+ "train_tokens_per_second": 5669.596
+ },
+ {
+ "epoch": 0.6423377322919472,
+ "grad_norm": 0.8008472981307811,
+ "learning_rate": 7.664061175315696e-06,
+ "loss": 0.3715662956237793,
+ "num_input_tokens_seen": 417464000,
+ "step": 9540,
+ "train_runtime": 73631.8047,
+ "train_tokens_per_second": 5669.615
+ },
+ {
+ "epoch": 0.6426743872879074,
+ "grad_norm": 0.7188017157103265,
+ "learning_rate": 7.661823292047771e-06,
+ "loss": 0.3764479160308838,
+ "num_input_tokens_seen": 417683920,
+ "step": 9545,
+ "train_runtime": 73672.7623,
+ "train_tokens_per_second": 5669.448
+ },
+ {
+ "epoch": 0.6430110422838675,
+ "grad_norm": 0.8895228408009356,
+ "learning_rate": 7.659584664409452e-06,
+ "loss": 0.3806633949279785,
+ "num_input_tokens_seen": 417910080,
+ "step": 9550,
+ "train_runtime": 73714.3818,
+ "train_tokens_per_second": 5669.315
+ },
+ {
+ "epoch": 0.6433476972798277,
+ "grad_norm": 0.7813418710468062,
+ "learning_rate": 7.65734529302676e-06,
+ "loss": 0.33327679634094237,
+ "num_input_tokens_seen": 418114864,
+ "step": 9555,
+ "train_runtime": 73751.4559,
+ "train_tokens_per_second": 5669.242
+ },
+ {
+ "epoch": 0.6436843522757878,
+ "grad_norm": 0.7440032604844485,
+ "learning_rate": 7.655105178525932e-06,
+ "loss": 0.36725008487701416,
+ "num_input_tokens_seen": 418331608,
+ "step": 9560,
+ "train_runtime": 73789.4942,
+ "train_tokens_per_second": 5669.257
+ },
+ {
+ "epoch": 0.6440210072717479,
+ "grad_norm": 0.7428051543591697,
+ "learning_rate": 7.652864321533406e-06,
+ "loss": 0.37874999046325686,
+ "num_input_tokens_seen": 418542736,
+ "step": 9565,
+ "train_runtime": 73827.1886,
+ "train_tokens_per_second": 5669.222
+ },
+ {
+ "epoch": 0.6443576622677081,
+ "grad_norm": 0.6906490664289745,
+ "learning_rate": 7.650622722675833e-06,
+ "loss": 0.39217414855957033,
+ "num_input_tokens_seen": 418769104,
+ "step": 9570,
+ "train_runtime": 73869.6555,
+ "train_tokens_per_second": 5669.027
+ },
+ {
+ "epoch": 0.6446943172636682,
+ "grad_norm": 0.7836457733135137,
+ "learning_rate": 7.648380382580067e-06,
+ "loss": 0.3911130428314209,
+ "num_input_tokens_seen": 418983272,
+ "step": 9575,
+ "train_runtime": 73911.9877,
+ "train_tokens_per_second": 5668.678
+ },
+ {
+ "epoch": 0.6450309722596284,
+ "grad_norm": 0.7191438751807192,
+ "learning_rate": 7.646137301873173e-06,
+ "loss": 0.3877583026885986,
+ "num_input_tokens_seen": 419193152,
+ "step": 9580,
+ "train_runtime": 73952.718,
+ "train_tokens_per_second": 5668.394
+ },
+ {
+ "epoch": 0.6453676272555885,
+ "grad_norm": 0.8722963532886264,
+ "learning_rate": 7.643893481182421e-06,
+ "loss": 0.35512332916259765,
+ "num_input_tokens_seen": 419415144,
+ "step": 9585,
+ "train_runtime": 73990.0513,
+ "train_tokens_per_second": 5668.534
+ },
+ {
+ "epoch": 0.6457042822515486,
+ "grad_norm": 0.7893095386452317,
+ "learning_rate": 7.641648921135287e-06,
+ "loss": 0.4030734062194824,
+ "num_input_tokens_seen": 419642840,
+ "step": 9590,
+ "train_runtime": 74032.6519,
+ "train_tokens_per_second": 5668.348
+ },
+ {
+ "epoch": 0.6460409372475088,
+ "grad_norm": 0.8691991766270082,
+ "learning_rate": 7.639403622359455e-06,
+ "loss": 0.38401873111724855,
+ "num_input_tokens_seen": 419851960,
+ "step": 9595,
+ "train_runtime": 74068.9663,
+ "train_tokens_per_second": 5668.392
+ },
+ {
+ "epoch": 0.6463775922434689,
+ "grad_norm": 0.7669041161649485,
+ "learning_rate": 7.637157585482817e-06,
+ "loss": 0.3698650598526001,
+ "num_input_tokens_seen": 420069368,
+ "step": 9600,
+ "train_runtime": 74113.2241,
+ "train_tokens_per_second": 5667.941
+ },
+ {
+ "epoch": 0.646714247239429,
+ "grad_norm": 0.6764016385660067,
+ "learning_rate": 7.63491081113347e-06,
+ "loss": 0.3979948043823242,
+ "num_input_tokens_seen": 420301264,
+ "step": 9605,
+ "train_runtime": 74149.7053,
+ "train_tokens_per_second": 5668.28
+ },
+ {
+ "epoch": 0.6470509022353892,
+ "grad_norm": 0.7632828915951216,
+ "learning_rate": 7.632663299939718e-06,
+ "loss": 0.4009084701538086,
+ "num_input_tokens_seen": 420528480,
+ "step": 9610,
+ "train_runtime": 74191.7204,
+ "train_tokens_per_second": 5668.132
+ },
+ {
+ "epoch": 0.6473875572313493,
+ "grad_norm": 0.7759849429505897,
+ "learning_rate": 7.630415052530066e-06,
+ "loss": 0.37017059326171875,
+ "num_input_tokens_seen": 420744352,
+ "step": 9615,
+ "train_runtime": 74228.9204,
+ "train_tokens_per_second": 5668.2
+ },
+ {
+ "epoch": 0.6477242122273095,
+ "grad_norm": 0.7521586068543743,
+ "learning_rate": 7.628166069533235e-06,
+ "loss": 0.3800983905792236,
+ "num_input_tokens_seen": 420951552,
+ "step": 9620,
+ "train_runtime": 74264.9231,
+ "train_tokens_per_second": 5668.242
+ },
+ {
+ "epoch": 0.6480608672232696,
+ "grad_norm": 0.8746614520646498,
+ "learning_rate": 7.625916351578141e-06,
+ "loss": 0.4101721286773682,
+ "num_input_tokens_seen": 421168904,
+ "step": 9625,
+ "train_runtime": 74302.0251,
+ "train_tokens_per_second": 5668.337
+ },
+ {
+ "epoch": 0.6483975222192297,
+ "grad_norm": 0.645310594782184,
+ "learning_rate": 7.623665899293914e-06,
+ "loss": 0.35078365802764894,
+ "num_input_tokens_seen": 421386640,
+ "step": 9630,
+ "train_runtime": 74343.6656,
+ "train_tokens_per_second": 5668.091
+ },
+ {
+ "epoch": 0.6487341772151899,
+ "grad_norm": 0.709824008157462,
+ "learning_rate": 7.621414713309885e-06,
+ "loss": 0.3464890241622925,
+ "num_input_tokens_seen": 421604504,
+ "step": 9635,
+ "train_runtime": 74380.7993,
+ "train_tokens_per_second": 5668.19
+ },
+ {
+ "epoch": 0.64907083221115,
+ "grad_norm": 0.6987475287759364,
+ "learning_rate": 7.61916279425559e-06,
+ "loss": 0.38260612487792967,
+ "num_input_tokens_seen": 421833672,
+ "step": 9640,
+ "train_runtime": 74422.4312,
+ "train_tokens_per_second": 5668.099
+ },
+ {
+ "epoch": 0.6494074872071102,
+ "grad_norm": 0.727875668145333,
+ "learning_rate": 7.6169101427607716e-06,
+ "loss": 0.38509573936462405,
+ "num_input_tokens_seen": 422067496,
+ "step": 9645,
+ "train_runtime": 74456.6368,
+ "train_tokens_per_second": 5668.635
+ },
+ {
+ "epoch": 0.6497441422030703,
+ "grad_norm": 0.7549567194100304,
+ "learning_rate": 7.614656759455375e-06,
+ "loss": 0.38222179412841795,
+ "num_input_tokens_seen": 422289144,
+ "step": 9650,
+ "train_runtime": 74495.8893,
+ "train_tokens_per_second": 5668.623
+ },
+ {
+ "epoch": 0.6500807971990304,
+ "grad_norm": 0.8363753913347584,
+ "learning_rate": 7.612402644969555e-06,
+ "loss": 0.38720729351043703,
+ "num_input_tokens_seen": 422500848,
+ "step": 9655,
+ "train_runtime": 74537.09,
+ "train_tokens_per_second": 5668.33
+ },
+ {
+ "epoch": 0.6504174521949906,
+ "grad_norm": 0.718216642211442,
+ "learning_rate": 7.6101477999336646e-06,
+ "loss": 0.37707133293151857,
+ "num_input_tokens_seen": 422718672,
+ "step": 9660,
+ "train_runtime": 74584.7849,
+ "train_tokens_per_second": 5667.626
+ },
+ {
+ "epoch": 0.6507541071909507,
+ "grad_norm": 0.8872821374211818,
+ "learning_rate": 7.607892224978266e-06,
+ "loss": 0.3739673376083374,
+ "num_input_tokens_seen": 422929480,
+ "step": 9665,
+ "train_runtime": 74620.552,
+ "train_tokens_per_second": 5667.735
+ },
+ {
+ "epoch": 0.6510907621869109,
+ "grad_norm": 0.691277480453269,
+ "learning_rate": 7.605635920734122e-06,
+ "loss": 0.3418617010116577,
+ "num_input_tokens_seen": 423142128,
+ "step": 9670,
+ "train_runtime": 74655.9636,
+ "train_tokens_per_second": 5667.895
+ },
+ {
+ "epoch": 0.651427417182871,
+ "grad_norm": 0.9025153211455825,
+ "learning_rate": 7.603378887832202e-06,
+ "loss": 0.3855189561843872,
+ "num_input_tokens_seen": 423349056,
+ "step": 9675,
+ "train_runtime": 74695.4667,
+ "train_tokens_per_second": 5667.667
+ },
+ {
+ "epoch": 0.6517640721788311,
+ "grad_norm": 0.7073570329177757,
+ "learning_rate": 7.601121126903676e-06,
+ "loss": 0.3803261280059814,
+ "num_input_tokens_seen": 423573240,
+ "step": 9680,
+ "train_runtime": 74738.2061,
+ "train_tokens_per_second": 5667.426
+ },
+ {
+ "epoch": 0.6521007271747913,
+ "grad_norm": 0.7609618594052814,
+ "learning_rate": 7.598862638579922e-06,
+ "loss": 0.3738381385803223,
+ "num_input_tokens_seen": 423809440,
+ "step": 9685,
+ "train_runtime": 74781.8146,
+ "train_tokens_per_second": 5667.279
+ },
+ {
+ "epoch": 0.6524373821707514,
+ "grad_norm": 0.6698743309851123,
+ "learning_rate": 7.5966034234925205e-06,
+ "loss": 0.3761345624923706,
+ "num_input_tokens_seen": 424041528,
+ "step": 9690,
+ "train_runtime": 74821.3132,
+ "train_tokens_per_second": 5667.39
+ },
+ {
+ "epoch": 0.6527740371667116,
+ "grad_norm": 0.7414693193437045,
+ "learning_rate": 7.59434348227325e-06,
+ "loss": 0.43105735778808596,
+ "num_input_tokens_seen": 424268184,
+ "step": 9695,
+ "train_runtime": 74855.3245,
+ "train_tokens_per_second": 5667.842
+ },
+ {
+ "epoch": 0.6531106921626717,
+ "grad_norm": 0.8237582570303243,
+ "learning_rate": 7.592082815554098e-06,
+ "loss": 0.40675368309021,
+ "num_input_tokens_seen": 424492616,
+ "step": 9700,
+ "train_runtime": 74894.094,
+ "train_tokens_per_second": 5667.905
+ },
+ {
+ "epoch": 0.6534473471586318,
+ "grad_norm": 0.9212376319324829,
+ "learning_rate": 7.58982142396725e-06,
+ "loss": 0.419390869140625,
+ "num_input_tokens_seen": 424686848,
+ "step": 9705,
+ "train_runtime": 74939.3759,
+ "train_tokens_per_second": 5667.072
+ },
+ {
+ "epoch": 0.653784002154592,
+ "grad_norm": 0.7691414977900423,
+ "learning_rate": 7.5875593081451e-06,
+ "loss": 0.3695266008377075,
+ "num_input_tokens_seen": 424913632,
+ "step": 9710,
+ "train_runtime": 74985.8205,
+ "train_tokens_per_second": 5666.586
+ },
+ {
+ "epoch": 0.6541206571505521,
+ "grad_norm": 0.7263975698689714,
+ "learning_rate": 7.585296468720243e-06,
+ "loss": 0.3619475126266479,
+ "num_input_tokens_seen": 425139040,
+ "step": 9715,
+ "train_runtime": 75026.5575,
+ "train_tokens_per_second": 5666.514
+ },
+ {
+ "epoch": 0.6544573121465123,
+ "grad_norm": 0.712706963298683,
+ "learning_rate": 7.5830329063254705e-06,
+ "loss": 0.38277277946472166,
+ "num_input_tokens_seen": 425352824,
+ "step": 9720,
+ "train_runtime": 75065.2226,
+ "train_tokens_per_second": 5666.443
+ },
+ {
+ "epoch": 0.6547939671424724,
+ "grad_norm": 0.6764513937336157,
+ "learning_rate": 7.5807686215937835e-06,
+ "loss": 0.38765199184417726,
+ "num_input_tokens_seen": 425565784,
+ "step": 9725,
+ "train_runtime": 75105.3744,
+ "train_tokens_per_second": 5666.249
+ },
+ {
+ "epoch": 0.6551306221384325,
+ "grad_norm": 0.7196173883927505,
+ "learning_rate": 7.578503615158379e-06,
+ "loss": 0.3398409366607666,
+ "num_input_tokens_seen": 425778128,
+ "step": 9730,
+ "train_runtime": 75149.2065,
+ "train_tokens_per_second": 5665.77
+ },
+ {
+ "epoch": 0.6554672771343927,
+ "grad_norm": 0.646960134820036,
+ "learning_rate": 7.576237887652662e-06,
+ "loss": 0.38698830604553225,
+ "num_input_tokens_seen": 426007040,
+ "step": 9735,
+ "train_runtime": 75188.2698,
+ "train_tokens_per_second": 5665.871
+ },
+ {
+ "epoch": 0.6558039321303528,
+ "grad_norm": 0.8063318146524463,
+ "learning_rate": 7.573971439710235e-06,
+ "loss": 0.34668645858764646,
+ "num_input_tokens_seen": 426223488,
+ "step": 9740,
+ "train_runtime": 75225.1526,
+ "train_tokens_per_second": 5665.97
+ },
+ {
+ "epoch": 0.656140587126313,
+ "grad_norm": 0.8258091202754753,
+ "learning_rate": 7.571704271964902e-06,
+ "loss": 0.3675506591796875,
+ "num_input_tokens_seen": 426433528,
+ "step": 9745,
+ "train_runtime": 75260.6499,
+ "train_tokens_per_second": 5666.089
+ },
+ {
+ "epoch": 0.6564772421222731,
+ "grad_norm": 0.8027675726296949,
+ "learning_rate": 7.56943638505067e-06,
+ "loss": 0.3636334419250488,
+ "num_input_tokens_seen": 426658696,
+ "step": 9750,
+ "train_runtime": 75296.5611,
+ "train_tokens_per_second": 5666.377
+ },
+ {
+ "epoch": 0.6568138971182332,
+ "grad_norm": 0.7190477185688757,
+ "learning_rate": 7.567167779601746e-06,
+ "loss": 0.39123508930206297,
+ "num_input_tokens_seen": 426886792,
+ "step": 9755,
+ "train_runtime": 75338.0495,
+ "train_tokens_per_second": 5666.284
+ },
+ {
+ "epoch": 0.6571505521141934,
+ "grad_norm": 0.7906692823017727,
+ "learning_rate": 7.56489845625254e-06,
+ "loss": 0.413257360458374,
+ "num_input_tokens_seen": 427104192,
+ "step": 9760,
+ "train_runtime": 75386.2254,
+ "train_tokens_per_second": 5665.547
+ },
+ {
+ "epoch": 0.6574872071101535,
+ "grad_norm": 0.7301115269654299,
+ "learning_rate": 7.562628415637659e-06,
+ "loss": 0.369996976852417,
+ "num_input_tokens_seen": 427312504,
+ "step": 9765,
+ "train_runtime": 75426.5577,
+ "train_tokens_per_second": 5665.279
+ },
+ {
+ "epoch": 0.6578238621061137,
+ "grad_norm": 0.7224806256488935,
+ "learning_rate": 7.560357658391913e-06,
+ "loss": 0.4071043491363525,
+ "num_input_tokens_seen": 427533792,
+ "step": 9770,
+ "train_runtime": 75465.4869,
+ "train_tokens_per_second": 5665.289
+ },
+ {
+ "epoch": 0.6581605171020738,
+ "grad_norm": 0.7268098674561758,
+ "learning_rate": 7.558086185150311e-06,
+ "loss": 0.3804465770721436,
+ "num_input_tokens_seen": 427747608,
+ "step": 9775,
+ "train_runtime": 75507.0805,
+ "train_tokens_per_second": 5665.0
+ },
+ {
+ "epoch": 0.6584971720980339,
+ "grad_norm": 0.7850729912855515,
+ "learning_rate": 7.555813996548065e-06,
+ "loss": 0.40397090911865235,
+ "num_input_tokens_seen": 427972008,
+ "step": 9780,
+ "train_runtime": 75543.708,
+ "train_tokens_per_second": 5665.224
+ },
+ {
+ "epoch": 0.6588338270939941,
+ "grad_norm": 0.7662329566093665,
+ "learning_rate": 7.553541093220586e-06,
+ "loss": 0.3842537641525269,
+ "num_input_tokens_seen": 428186312,
+ "step": 9785,
+ "train_runtime": 75579.4178,
+ "train_tokens_per_second": 5665.383
+ },
+ {
+ "epoch": 0.6591704820899542,
+ "grad_norm": 0.70627363493871,
+ "learning_rate": 7.551267475803481e-06,
+ "loss": 0.37559702396392824,
+ "num_input_tokens_seen": 428413952,
+ "step": 9790,
+ "train_runtime": 75619.0124,
+ "train_tokens_per_second": 5665.426
+ },
+ {
+ "epoch": 0.6595071370859144,
+ "grad_norm": 0.7354101668176468,
+ "learning_rate": 7.5489931449325635e-06,
+ "loss": 0.3736457347869873,
+ "num_input_tokens_seen": 428630728,
+ "step": 9795,
+ "train_runtime": 75658.6215,
+ "train_tokens_per_second": 5665.326
+ },
+ {
+ "epoch": 0.6598437920818745,
+ "grad_norm": 0.7892531883838505,
+ "learning_rate": 7.546718101243842e-06,
+ "loss": 0.39138045310974123,
+ "num_input_tokens_seen": 428842760,
+ "step": 9800,
+ "train_runtime": 75693.8787,
+ "train_tokens_per_second": 5665.488
+ },
+ {
+ "epoch": 0.6601804470778346,
+ "grad_norm": 0.8775677410278766,
+ "learning_rate": 7.544442345373522e-06,
+ "loss": 0.3698113918304443,
+ "num_input_tokens_seen": 429051072,
+ "step": 9805,
+ "train_runtime": 75734.5916,
+ "train_tokens_per_second": 5665.193
+ },
+ {
+ "epoch": 0.6605171020737948,
+ "grad_norm": 0.6944135512445987,
+ "learning_rate": 7.542165877958014e-06,
+ "loss": 0.3717064380645752,
+ "num_input_tokens_seen": 429266232,
+ "step": 9810,
+ "train_runtime": 75771.7628,
+ "train_tokens_per_second": 5665.253
+ },
+ {
+ "epoch": 0.6608537570697549,
+ "grad_norm": 0.7361334690779209,
+ "learning_rate": 7.539888699633926e-06,
+ "loss": 0.38160529136657717,
+ "num_input_tokens_seen": 429495032,
+ "step": 9815,
+ "train_runtime": 75808.8176,
+ "train_tokens_per_second": 5665.502
+ },
+ {
+ "epoch": 0.661190412065715,
+ "grad_norm": 0.8126000202476551,
+ "learning_rate": 7.5376108110380606e-06,
+ "loss": 0.34899187088012695,
+ "num_input_tokens_seen": 429723384,
+ "step": 9820,
+ "train_runtime": 75847.4225,
+ "train_tokens_per_second": 5665.629
+ },
+ {
+ "epoch": 0.6615270670616752,
+ "grad_norm": 0.6957748548678448,
+ "learning_rate": 7.535332212807426e-06,
+ "loss": 0.3428369998931885,
+ "num_input_tokens_seen": 429924744,
+ "step": 9825,
+ "train_runtime": 75886.3897,
+ "train_tokens_per_second": 5665.374
+ },
+ {
+ "epoch": 0.6618637220576353,
+ "grad_norm": 0.7233083648231398,
+ "learning_rate": 7.53305290557922e-06,
+ "loss": 0.39725618362426757,
+ "num_input_tokens_seen": 430154928,
+ "step": 9830,
+ "train_runtime": 75925.3068,
+ "train_tokens_per_second": 5665.501
+ },
+ {
+ "epoch": 0.6622003770535955,
+ "grad_norm": 0.7049996354735961,
+ "learning_rate": 7.530772889990848e-06,
+ "loss": 0.3740149974822998,
+ "num_input_tokens_seen": 430386800,
+ "step": 9835,
+ "train_runtime": 75959.5503,
+ "train_tokens_per_second": 5666.0
+ },
+ {
+ "epoch": 0.6625370320495556,
+ "grad_norm": 0.6720383481048683,
+ "learning_rate": 7.528492166679907e-06,
+ "loss": 0.35007429122924805,
+ "num_input_tokens_seen": 430616592,
+ "step": 9840,
+ "train_runtime": 76001.567,
+ "train_tokens_per_second": 5665.891
+ },
+ {
+ "epoch": 0.6628736870455157,
+ "grad_norm": 0.6809307882652357,
+ "learning_rate": 7.526210736284194e-06,
+ "loss": 0.38547792434692385,
+ "num_input_tokens_seen": 430820368,
+ "step": 9845,
+ "train_runtime": 76036.0252,
+ "train_tokens_per_second": 5666.003
+ },
+ {
+ "epoch": 0.6632103420414759,
+ "grad_norm": 0.7241468747622457,
+ "learning_rate": 7.523928599441706e-06,
+ "loss": 0.3851293087005615,
+ "num_input_tokens_seen": 431045232,
+ "step": 9850,
+ "train_runtime": 76076.642,
+ "train_tokens_per_second": 5665.934
+ },
+ {
+ "epoch": 0.663546997037436,
+ "grad_norm": 0.8065697031009399,
+ "learning_rate": 7.5216457567906295e-06,
+ "loss": 0.3921393632888794,
+ "num_input_tokens_seen": 431272728,
+ "step": 9855,
+ "train_runtime": 76119.1303,
+ "train_tokens_per_second": 5665.76
+ },
+ {
+ "epoch": 0.6638836520333962,
+ "grad_norm": 0.7363581812525671,
+ "learning_rate": 7.519362208969362e-06,
+ "loss": 0.3589025974273682,
+ "num_input_tokens_seen": 431497576,
+ "step": 9860,
+ "train_runtime": 76164.9303,
+ "train_tokens_per_second": 5665.305
+ },
+ {
+ "epoch": 0.6642203070293563,
+ "grad_norm": 0.7028175355387544,
+ "learning_rate": 7.5170779566164855e-06,
+ "loss": 0.39649462699890137,
+ "num_input_tokens_seen": 431716768,
+ "step": 9865,
+ "train_runtime": 76207.7038,
+ "train_tokens_per_second": 5665.002
+ },
+ {
+ "epoch": 0.6645569620253164,
+ "grad_norm": 0.8083377822452007,
+ "learning_rate": 7.514793000370783e-06,
+ "loss": 0.39783761501312254,
+ "num_input_tokens_seen": 431927680,
+ "step": 9870,
+ "train_runtime": 76252.3393,
+ "train_tokens_per_second": 5664.452
+ },
+ {
+ "epoch": 0.6648936170212766,
+ "grad_norm": 0.7361462900334614,
+ "learning_rate": 7.512507340871239e-06,
+ "loss": 0.3664602756500244,
+ "num_input_tokens_seen": 432144784,
+ "step": 9875,
+ "train_runtime": 76293.2197,
+ "train_tokens_per_second": 5664.262
+ },
+ {
+ "epoch": 0.6652302720172367,
+ "grad_norm": 0.7771322596278541,
+ "learning_rate": 7.510220978757028e-06,
+ "loss": 0.39647390842437746,
+ "num_input_tokens_seen": 432361504,
+ "step": 9880,
+ "train_runtime": 76335.1474,
+ "train_tokens_per_second": 5663.99
+ },
+ {
+ "epoch": 0.6655669270131969,
+ "grad_norm": 0.9351257512839886,
+ "learning_rate": 7.507933914667524e-06,
+ "loss": 0.4201164722442627,
+ "num_input_tokens_seen": 432586992,
+ "step": 9885,
+ "train_runtime": 76375.8314,
+ "train_tokens_per_second": 5663.925
+ },
+ {
+ "epoch": 0.665903582009157,
+ "grad_norm": 0.7531976870702812,
+ "learning_rate": 7.505646149242299e-06,
+ "loss": 0.37460851669311523,
+ "num_input_tokens_seen": 432809080,
+ "step": 9890,
+ "train_runtime": 76415.9118,
+ "train_tokens_per_second": 5663.861
+ },
+ {
+ "epoch": 0.6662402370051171,
+ "grad_norm": 0.719734237163997,
+ "learning_rate": 7.503357683121119e-06,
+ "loss": 0.34871246814727785,
+ "num_input_tokens_seen": 433035120,
+ "step": 9895,
+ "train_runtime": 76462.7297,
+ "train_tokens_per_second": 5663.349
+ },
+ {
+ "epoch": 0.6665768920010773,
+ "grad_norm": 0.7130561120719486,
+ "learning_rate": 7.501068516943944e-06,
+ "loss": 0.3231820583343506,
+ "num_input_tokens_seen": 433248272,
+ "step": 9900,
+ "train_runtime": 76502.2615,
+ "train_tokens_per_second": 5663.209
+ },
+ {
+ "epoch": 0.6669135469970374,
+ "grad_norm": 0.6874890275442226,
+ "learning_rate": 7.498778651350934e-06,
+ "loss": 0.3865694284439087,
+ "num_input_tokens_seen": 433472920,
+ "step": 9905,
+ "train_runtime": 76539.0929,
+ "train_tokens_per_second": 5663.419
+ },
+ {
+ "epoch": 0.6672502019929976,
+ "grad_norm": 0.6837593508656135,
+ "learning_rate": 7.496488086982443e-06,
+ "loss": 0.3751338720321655,
+ "num_input_tokens_seen": 433694440,
+ "step": 9910,
+ "train_runtime": 76582.0454,
+ "train_tokens_per_second": 5663.135
+ },
+ {
+ "epoch": 0.6675868569889577,
+ "grad_norm": 0.834960347653905,
+ "learning_rate": 7.494196824479019e-06,
+ "loss": 0.37354316711425783,
+ "num_input_tokens_seen": 433909064,
+ "step": 9915,
+ "train_runtime": 76619.6521,
+ "train_tokens_per_second": 5663.156
+ },
+ {
+ "epoch": 0.6679235119849178,
+ "grad_norm": 0.6905943083082271,
+ "learning_rate": 7.491904864481406e-06,
+ "loss": 0.3818557024002075,
+ "num_input_tokens_seen": 434122768,
+ "step": 9920,
+ "train_runtime": 76661.7307,
+ "train_tokens_per_second": 5662.835
+ },
+ {
+ "epoch": 0.668260166980878,
+ "grad_norm": 0.6608540738794552,
+ "learning_rate": 7.489612207630543e-06,
+ "loss": 0.34175124168396,
+ "num_input_tokens_seen": 434354528,
+ "step": 9925,
+ "train_runtime": 76695.5171,
+ "train_tokens_per_second": 5663.363
+ },
+ {
+ "epoch": 0.6685968219768381,
+ "grad_norm": 0.747448764956391,
+ "learning_rate": 7.487318854567566e-06,
+ "loss": 0.37988905906677245,
+ "num_input_tokens_seen": 434564096,
+ "step": 9930,
+ "train_runtime": 76729.4177,
+ "train_tokens_per_second": 5663.592
+ },
+ {
+ "epoch": 0.6689334769727983,
+ "grad_norm": 0.7551042289550628,
+ "learning_rate": 7.485024805933803e-06,
+ "loss": 0.3599559307098389,
+ "num_input_tokens_seen": 434767584,
+ "step": 9935,
+ "train_runtime": 76765.8819,
+ "train_tokens_per_second": 5663.552
+ },
+ {
+ "epoch": 0.6692701319687584,
+ "grad_norm": 0.6631253047912142,
+ "learning_rate": 7.482730062370776e-06,
+ "loss": 0.3699958086013794,
+ "num_input_tokens_seen": 434981392,
+ "step": 9940,
+ "train_runtime": 76810.3453,
+ "train_tokens_per_second": 5663.057
+ },
+ {
+ "epoch": 0.6696067869647185,
+ "grad_norm": 0.7419282684235264,
+ "learning_rate": 7.480434624520205e-06,
+ "loss": 0.3593663215637207,
+ "num_input_tokens_seen": 435217648,
+ "step": 9945,
+ "train_runtime": 76843.0552,
+ "train_tokens_per_second": 5663.721
+ },
+ {
+ "epoch": 0.6699434419606787,
+ "grad_norm": 0.7057608866633803,
+ "learning_rate": 7.478138493024001e-06,
+ "loss": 0.3879536151885986,
+ "num_input_tokens_seen": 435444608,
+ "step": 9950,
+ "train_runtime": 76882.5546,
+ "train_tokens_per_second": 5663.764
+ },
+ {
+ "epoch": 0.6702800969566388,
+ "grad_norm": 0.740329239724491,
+ "learning_rate": 7.475841668524268e-06,
+ "loss": 0.34809188842773436,
+ "num_input_tokens_seen": 435656280,
+ "step": 9955,
+ "train_runtime": 76923.9627,
+ "train_tokens_per_second": 5663.466
+ },
+ {
+ "epoch": 0.670616751952599,
+ "grad_norm": 0.8121045332443452,
+ "learning_rate": 7.473544151663309e-06,
+ "loss": 0.4048673152923584,
+ "num_input_tokens_seen": 435875456,
+ "step": 9960,
+ "train_runtime": 76958.8069,
+ "train_tokens_per_second": 5663.75
+ },
+ {
+ "epoch": 0.6709534069485591,
+ "grad_norm": 0.8149107981141157,
+ "learning_rate": 7.471245943083615e-06,
+ "loss": 0.3677277326583862,
+ "num_input_tokens_seen": 436097568,
+ "step": 9965,
+ "train_runtime": 76998.2656,
+ "train_tokens_per_second": 5663.732
+ },
+ {
+ "epoch": 0.6712900619445192,
+ "grad_norm": 0.7104981842358721,
+ "learning_rate": 7.468947043427873e-06,
+ "loss": 0.3423744201660156,
+ "num_input_tokens_seen": 436314872,
+ "step": 9970,
+ "train_runtime": 77032.0267,
+ "train_tokens_per_second": 5664.071
+ },
+ {
+ "epoch": 0.6716267169404794,
+ "grad_norm": 0.8060621663622469,
+ "learning_rate": 7.466647453338965e-06,
+ "loss": 0.38165833950042727,
+ "num_input_tokens_seen": 436524088,
+ "step": 9975,
+ "train_runtime": 77070.1105,
+ "train_tokens_per_second": 5663.987
+ },
+ {
+ "epoch": 0.6719633719364395,
+ "grad_norm": 0.6817255880743095,
+ "learning_rate": 7.464347173459958e-06,
+ "loss": 0.3900636911392212,
+ "num_input_tokens_seen": 436763792,
+ "step": 9980,
+ "train_runtime": 77108.4939,
+ "train_tokens_per_second": 5664.276
+ },
+ {
+ "epoch": 0.6723000269323997,
+ "grad_norm": 0.5849966120883113,
+ "learning_rate": 7.462046204434126e-06,
+ "loss": 0.35174126625061036,
+ "num_input_tokens_seen": 437001592,
+ "step": 9985,
+ "train_runtime": 77143.495,
+ "train_tokens_per_second": 5664.789
+ },
+ {
+ "epoch": 0.6726366819283598,
+ "grad_norm": 0.819711833914903,
+ "learning_rate": 7.459744546904922e-06,
+ "loss": 0.37253947257995607,
+ "num_input_tokens_seen": 437217472,
+ "step": 9990,
+ "train_runtime": 77177.4484,
+ "train_tokens_per_second": 5665.094
+ },
+ {
+ "epoch": 0.6729733369243199,
+ "grad_norm": 0.7103670160362342,
+ "learning_rate": 7.457442201516001e-06,
+ "loss": 0.34635467529296876,
+ "num_input_tokens_seen": 437432256,
+ "step": 9995,
+ "train_runtime": 77222.7561,
+ "train_tokens_per_second": 5664.551
+ },
+ {
+ "epoch": 0.6733099919202801,
+ "grad_norm": 0.7772791959365979,
+ "learning_rate": 7.455139168911203e-06,
+ "loss": 0.3714813232421875,
+ "num_input_tokens_seen": 437632184,
+ "step": 10000,
+ "train_runtime": 77259.692,
+ "train_tokens_per_second": 5664.431
+ },
+ {
+ "epoch": 0.6736466469162402,
+ "grad_norm": 0.8663644820452563,
+ "learning_rate": 7.452835449734564e-06,
+ "loss": 0.37085671424865724,
+ "num_input_tokens_seen": 437852448,
+ "step": 10005,
+ "train_runtime": 77297.557,
+ "train_tokens_per_second": 5664.506
+ },
+ {
+ "epoch": 0.6739833019122003,
+ "grad_norm": 0.6628154115799231,
+ "learning_rate": 7.450531044630315e-06,
+ "loss": 0.35476658344268797,
+ "num_input_tokens_seen": 438072720,
+ "step": 10010,
+ "train_runtime": 77330.8077,
+ "train_tokens_per_second": 5664.918
+ },
+ {
+ "epoch": 0.6743199569081605,
+ "grad_norm": 0.7435766802454172,
+ "learning_rate": 7.448225954242874e-06,
+ "loss": 0.3909281253814697,
+ "num_input_tokens_seen": 438300752,
+ "step": 10015,
+ "train_runtime": 77372.9883,
+ "train_tokens_per_second": 5664.777
+ },
+ {
+ "epoch": 0.6746566119041206,
+ "grad_norm": 0.7076966807661021,
+ "learning_rate": 7.44592017921685e-06,
+ "loss": 0.37700977325439455,
+ "num_input_tokens_seen": 438521952,
+ "step": 10020,
+ "train_runtime": 77408.1985,
+ "train_tokens_per_second": 5665.058
+ },
+ {
+ "epoch": 0.6749932669000808,
+ "grad_norm": 0.7283120289124543,
+ "learning_rate": 7.443613720197048e-06,
+ "loss": 0.39572868347167967,
+ "num_input_tokens_seen": 438742144,
+ "step": 10025,
+ "train_runtime": 77451.6366,
+ "train_tokens_per_second": 5664.724
+ },
+ {
+ "epoch": 0.6753299218960409,
+ "grad_norm": 0.8184114176248545,
+ "learning_rate": 7.44130657782846e-06,
+ "loss": 0.35796802043914794,
+ "num_input_tokens_seen": 438946944,
+ "step": 10030,
+ "train_runtime": 77487.5196,
+ "train_tokens_per_second": 5664.744
+ },
+ {
+ "epoch": 0.675666576892001,
+ "grad_norm": 0.7168914852518904,
+ "learning_rate": 7.438998752756274e-06,
+ "loss": 0.3807486057281494,
+ "num_input_tokens_seen": 439164968,
+ "step": 10035,
+ "train_runtime": 77530.1548,
+ "train_tokens_per_second": 5664.441
+ },
+ {
+ "epoch": 0.6760032318879612,
+ "grad_norm": 0.7003642669524813,
+ "learning_rate": 7.436690245625864e-06,
+ "loss": 0.36953206062316896,
+ "num_input_tokens_seen": 439382032,
+ "step": 10040,
+ "train_runtime": 77568.57,
+ "train_tokens_per_second": 5664.434
+ },
+ {
+ "epoch": 0.6763398868839213,
+ "grad_norm": 0.8332693301417188,
+ "learning_rate": 7.434381057082797e-06,
+ "loss": 0.369028639793396,
+ "num_input_tokens_seen": 439590072,
+ "step": 10045,
+ "train_runtime": 77608.9784,
+ "train_tokens_per_second": 5664.165
+ },
+ {
+ "epoch": 0.6766765418798815,
+ "grad_norm": 0.7659413901414965,
+ "learning_rate": 7.4320711877728305e-06,
+ "loss": 0.3970766067504883,
+ "num_input_tokens_seen": 439814224,
+ "step": 10050,
+ "train_runtime": 77655.5893,
+ "train_tokens_per_second": 5663.652
+ },
+ {
+ "epoch": 0.6770131968758416,
+ "grad_norm": 0.7616766803709787,
+ "learning_rate": 7.429760638341909e-06,
+ "loss": 0.34726481437683104,
+ "num_input_tokens_seen": 440013360,
+ "step": 10055,
+ "train_runtime": 77694.3971,
+ "train_tokens_per_second": 5663.386
+ },
+ {
+ "epoch": 0.6773498518718017,
+ "grad_norm": 0.7214303745868783,
+ "learning_rate": 7.427449409436176e-06,
+ "loss": 0.42118000984191895,
+ "num_input_tokens_seen": 440232216,
+ "step": 10060,
+ "train_runtime": 77735.8693,
+ "train_tokens_per_second": 5663.18
+ },
+ {
+ "epoch": 0.6776865068677619,
+ "grad_norm": 0.6480309395843215,
+ "learning_rate": 7.425137501701955e-06,
+ "loss": 0.35333929061889646,
+ "num_input_tokens_seen": 440465120,
+ "step": 10065,
+ "train_runtime": 77778.1439,
+ "train_tokens_per_second": 5663.096
+ },
+ {
+ "epoch": 0.678023161863722,
+ "grad_norm": 0.8425975984479226,
+ "learning_rate": 7.422824915785765e-06,
+ "loss": 0.3732592582702637,
+ "num_input_tokens_seen": 440684520,
+ "step": 10070,
+ "train_runtime": 77813.6788,
+ "train_tokens_per_second": 5663.33
+ },
+ {
+ "epoch": 0.6783598168596822,
+ "grad_norm": 0.7751732244738783,
+ "learning_rate": 7.420511652334313e-06,
+ "loss": 0.3922509431838989,
+ "num_input_tokens_seen": 440888096,
+ "step": 10075,
+ "train_runtime": 77856.5574,
+ "train_tokens_per_second": 5662.825
+ },
+ {
+ "epoch": 0.6786964718556423,
+ "grad_norm": 0.746917656110605,
+ "learning_rate": 7.418197711994495e-06,
+ "loss": 0.3609792947769165,
+ "num_input_tokens_seen": 441099168,
+ "step": 10080,
+ "train_runtime": 77896.5389,
+ "train_tokens_per_second": 5662.629
+ },
+ {
+ "epoch": 0.6790331268516024,
+ "grad_norm": 0.7979148491415948,
+ "learning_rate": 7.4158830954133984e-06,
+ "loss": 0.37356295585632326,
+ "num_input_tokens_seen": 441329664,
+ "step": 10085,
+ "train_runtime": 77936.3536,
+ "train_tokens_per_second": 5662.693
+ },
+ {
+ "epoch": 0.6793697818475626,
+ "grad_norm": 0.7047169214626602,
+ "learning_rate": 7.413567803238298e-06,
+ "loss": 0.36446294784545896,
+ "num_input_tokens_seen": 441545160,
+ "step": 10090,
+ "train_runtime": 77973.8265,
+ "train_tokens_per_second": 5662.736
+ },
+ {
+ "epoch": 0.6797064368435227,
+ "grad_norm": 0.6385696916431561,
+ "learning_rate": 7.411251836116658e-06,
+ "loss": 0.3654582977294922,
+ "num_input_tokens_seen": 441770600,
+ "step": 10095,
+ "train_runtime": 78021.2974,
+ "train_tokens_per_second": 5662.18
+ },
+ {
+ "epoch": 0.6800430918394829,
+ "grad_norm": 0.675501708680357,
+ "learning_rate": 7.408935194696132e-06,
+ "loss": 0.38489809036254885,
+ "num_input_tokens_seen": 441989608,
+ "step": 10100,
+ "train_runtime": 78062.7748,
+ "train_tokens_per_second": 5661.977
+ },
+ {
+ "epoch": 0.680379746835443,
+ "grad_norm": 0.6976086448861777,
+ "learning_rate": 7.406617879624558e-06,
+ "loss": 0.39519965648651123,
+ "num_input_tokens_seen": 442219464,
+ "step": 10105,
+ "train_runtime": 78104.0438,
+ "train_tokens_per_second": 5661.928
+ },
+ {
+ "epoch": 0.6807164018314031,
+ "grad_norm": 0.6737767286617782,
+ "learning_rate": 7.40429989154997e-06,
+ "loss": 0.35456171035766604,
+ "num_input_tokens_seen": 442434504,
+ "step": 10110,
+ "train_runtime": 78146.9907,
+ "train_tokens_per_second": 5661.568
+ },
+ {
+ "epoch": 0.6810530568273633,
+ "grad_norm": 0.6824592254845565,
+ "learning_rate": 7.4019812311205826e-06,
+ "loss": 0.396740984916687,
+ "num_input_tokens_seen": 442663960,
+ "step": 10115,
+ "train_runtime": 78186.9172,
+ "train_tokens_per_second": 5661.612
+ },
+ {
+ "epoch": 0.6813897118233234,
+ "grad_norm": 0.6995329567504307,
+ "learning_rate": 7.399661898984805e-06,
+ "loss": 0.39820547103881837,
+ "num_input_tokens_seen": 442904544,
+ "step": 10120,
+ "train_runtime": 78228.9374,
+ "train_tokens_per_second": 5661.646
+ },
+ {
+ "epoch": 0.6817263668192836,
+ "grad_norm": 0.7146974067721376,
+ "learning_rate": 7.397341895791228e-06,
+ "loss": 0.3785873889923096,
+ "num_input_tokens_seen": 443119656,
+ "step": 10125,
+ "train_runtime": 78271.2656,
+ "train_tokens_per_second": 5661.332
+ },
+ {
+ "epoch": 0.6820630218152437,
+ "grad_norm": 0.7979315787216683,
+ "learning_rate": 7.395021222188634e-06,
+ "loss": 0.37588326930999755,
+ "num_input_tokens_seen": 443345048,
+ "step": 10130,
+ "train_runtime": 78307.7048,
+ "train_tokens_per_second": 5661.576
+ },
+ {
+ "epoch": 0.6823996768112038,
+ "grad_norm": 0.8569372783831029,
+ "learning_rate": 7.392699878825993e-06,
+ "loss": 0.4026289939880371,
+ "num_input_tokens_seen": 443550416,
+ "step": 10135,
+ "train_runtime": 78348.8182,
+ "train_tokens_per_second": 5661.227
+ },
+ {
+ "epoch": 0.682736331807164,
+ "grad_norm": 0.6799351048003278,
+ "learning_rate": 7.390377866352462e-06,
+ "loss": 0.3739679574966431,
+ "num_input_tokens_seen": 443760728,
+ "step": 10140,
+ "train_runtime": 78391.1203,
+ "train_tokens_per_second": 5660.855
+ },
+ {
+ "epoch": 0.6830729868031241,
+ "grad_norm": 0.7679162845866374,
+ "learning_rate": 7.388055185417382e-06,
+ "loss": 0.36473128795623777,
+ "num_input_tokens_seen": 443986040,
+ "step": 10145,
+ "train_runtime": 78429.781,
+ "train_tokens_per_second": 5660.937
+ },
+ {
+ "epoch": 0.6834096417990843,
+ "grad_norm": 0.7045812738235366,
+ "learning_rate": 7.3857318366702844e-06,
+ "loss": 0.35809712409973143,
+ "num_input_tokens_seen": 444206480,
+ "step": 10150,
+ "train_runtime": 78472.8507,
+ "train_tokens_per_second": 5660.639
+ },
+ {
+ "epoch": 0.6837462967950444,
+ "grad_norm": 0.7004993125342208,
+ "learning_rate": 7.383407820760885e-06,
+ "loss": 0.3713981151580811,
+ "num_input_tokens_seen": 444432096,
+ "step": 10155,
+ "train_runtime": 78515.6453,
+ "train_tokens_per_second": 5660.427
+ },
+ {
+ "epoch": 0.6840829517910045,
+ "grad_norm": 0.7774580479484972,
+ "learning_rate": 7.381083138339089e-06,
+ "loss": 0.38779325485229493,
+ "num_input_tokens_seen": 444643008,
+ "step": 10160,
+ "train_runtime": 78555.8979,
+ "train_tokens_per_second": 5660.212
+ },
+ {
+ "epoch": 0.6844196067869647,
+ "grad_norm": 0.7163976333149118,
+ "learning_rate": 7.378757790054988e-06,
+ "loss": 0.39658608436584475,
+ "num_input_tokens_seen": 444875528,
+ "step": 10165,
+ "train_runtime": 78592.4434,
+ "train_tokens_per_second": 5660.538
+ },
+ {
+ "epoch": 0.6847562617829248,
+ "grad_norm": 0.7470536381083663,
+ "learning_rate": 7.376431776558855e-06,
+ "loss": 0.3956130504608154,
+ "num_input_tokens_seen": 445097312,
+ "step": 10170,
+ "train_runtime": 78630.7483,
+ "train_tokens_per_second": 5660.601
+ },
+ {
+ "epoch": 0.685092916778885,
+ "grad_norm": 0.7408806601027974,
+ "learning_rate": 7.3741050985011546e-06,
+ "loss": 0.3585235118865967,
+ "num_input_tokens_seen": 445319152,
+ "step": 10175,
+ "train_runtime": 78670.4341,
+ "train_tokens_per_second": 5660.566
+ },
+ {
+ "epoch": 0.6854295717748451,
+ "grad_norm": 0.7086975222942481,
+ "learning_rate": 7.371777756532531e-06,
+ "loss": 0.34885153770446775,
+ "num_input_tokens_seen": 445544848,
+ "step": 10180,
+ "train_runtime": 78704.6359,
+ "train_tokens_per_second": 5660.973
+ },
+ {
+ "epoch": 0.6857662267708052,
+ "grad_norm": 0.7956956482990667,
+ "learning_rate": 7.369449751303823e-06,
+ "loss": 0.3933322191238403,
+ "num_input_tokens_seen": 445764768,
+ "step": 10185,
+ "train_runtime": 78741.4834,
+ "train_tokens_per_second": 5661.117
+ },
+ {
+ "epoch": 0.6861028817667654,
+ "grad_norm": 0.5998016401864463,
+ "learning_rate": 7.367121083466048e-06,
+ "loss": 0.3657109498977661,
+ "num_input_tokens_seen": 445988536,
+ "step": 10190,
+ "train_runtime": 78776.8049,
+ "train_tokens_per_second": 5661.419
+ },
+ {
+ "epoch": 0.6864395367627255,
+ "grad_norm": 0.8579022323586973,
+ "learning_rate": 7.364791753670409e-06,
+ "loss": 0.3847865104675293,
+ "num_input_tokens_seen": 446213600,
+ "step": 10195,
+ "train_runtime": 78810.0378,
+ "train_tokens_per_second": 5661.888
+ },
+ {
+ "epoch": 0.6867761917586857,
+ "grad_norm": 0.729478615131132,
+ "learning_rate": 7.362461762568298e-06,
+ "loss": 0.372448205947876,
+ "num_input_tokens_seen": 446432912,
+ "step": 10200,
+ "train_runtime": 78848.3501,
+ "train_tokens_per_second": 5661.918
+ },
+ {
+ "epoch": 0.6871128467546458,
+ "grad_norm": 0.67715047966365,
+ "learning_rate": 7.360131110811287e-06,
+ "loss": 0.3532183408737183,
+ "num_input_tokens_seen": 446656416,
+ "step": 10205,
+ "train_runtime": 78888.3323,
+ "train_tokens_per_second": 5661.882
+ },
+ {
+ "epoch": 0.6874495017506059,
+ "grad_norm": 0.7095724500319497,
+ "learning_rate": 7.357799799051138e-06,
+ "loss": 0.3632460355758667,
+ "num_input_tokens_seen": 446871712,
+ "step": 10210,
+ "train_runtime": 78927.4603,
+ "train_tokens_per_second": 5661.803
+ },
+ {
+ "epoch": 0.6877861567465661,
+ "grad_norm": 0.9476488222825441,
+ "learning_rate": 7.355467827939794e-06,
+ "loss": 0.38349783420562744,
+ "num_input_tokens_seen": 447082192,
+ "step": 10215,
+ "train_runtime": 78967.9316,
+ "train_tokens_per_second": 5661.566
+ },
+ {
+ "epoch": 0.6881228117425262,
+ "grad_norm": 0.7980347922355381,
+ "learning_rate": 7.353135198129382e-06,
+ "loss": 0.3912900686264038,
+ "num_input_tokens_seen": 447291168,
+ "step": 10220,
+ "train_runtime": 79002.7871,
+ "train_tokens_per_second": 5661.714
+ },
+ {
+ "epoch": 0.6884594667384863,
+ "grad_norm": 0.796413758238857,
+ "learning_rate": 7.350801910272216e-06,
+ "loss": 0.38165950775146484,
+ "num_input_tokens_seen": 447512392,
+ "step": 10225,
+ "train_runtime": 79046.6142,
+ "train_tokens_per_second": 5661.373
+ },
+ {
+ "epoch": 0.6887961217344465,
+ "grad_norm": 0.7746118196960639,
+ "learning_rate": 7.348467965020791e-06,
+ "loss": 0.34427103996276853,
+ "num_input_tokens_seen": 447731304,
+ "step": 10230,
+ "train_runtime": 79082.2762,
+ "train_tokens_per_second": 5661.588
+ },
+ {
+ "epoch": 0.6891327767304066,
+ "grad_norm": 0.8166556060856294,
+ "learning_rate": 7.3461333630277885e-06,
+ "loss": 0.3613192081451416,
+ "num_input_tokens_seen": 447922416,
+ "step": 10235,
+ "train_runtime": 79117.4476,
+ "train_tokens_per_second": 5661.487
+ },
+ {
+ "epoch": 0.6894694317263668,
+ "grad_norm": 0.6976640506234788,
+ "learning_rate": 7.343798104946074e-06,
+ "loss": 0.3512859344482422,
+ "num_input_tokens_seen": 448148520,
+ "step": 10240,
+ "train_runtime": 79156.1315,
+ "train_tokens_per_second": 5661.577
+ },
+ {
+ "epoch": 0.6898060867223269,
+ "grad_norm": 0.74195160383717,
+ "learning_rate": 7.341462191428692e-06,
+ "loss": 0.381636905670166,
+ "num_input_tokens_seen": 448364664,
+ "step": 10245,
+ "train_runtime": 79199.1829,
+ "train_tokens_per_second": 5661.228
+ },
+ {
+ "epoch": 0.690142741718287,
+ "grad_norm": 0.7527010867181296,
+ "learning_rate": 7.339125623128874e-06,
+ "loss": 0.3913971662521362,
+ "num_input_tokens_seen": 448587016,
+ "step": 10250,
+ "train_runtime": 79248.9634,
+ "train_tokens_per_second": 5660.478
+ },
+ {
+ "epoch": 0.6904793967142472,
+ "grad_norm": 0.7210165301469443,
+ "learning_rate": 7.336788400700035e-06,
+ "loss": 0.36610395908355714,
+ "num_input_tokens_seen": 448798728,
+ "step": 10255,
+ "train_runtime": 79292.6113,
+ "train_tokens_per_second": 5660.032
+ },
+ {
+ "epoch": 0.6908160517102074,
+ "grad_norm": 0.9096784607094398,
+ "learning_rate": 7.334450524795771e-06,
+ "loss": 0.408566951751709,
+ "num_input_tokens_seen": 449023728,
+ "step": 10260,
+ "train_runtime": 79333.8153,
+ "train_tokens_per_second": 5659.929
+ },
+ {
+ "epoch": 0.6911527067061676,
+ "grad_norm": 0.6784070074683265,
+ "learning_rate": 7.332111996069863e-06,
+ "loss": 0.3746421575546265,
+ "num_input_tokens_seen": 449245560,
+ "step": 10265,
+ "train_runtime": 79370.4334,
+ "train_tokens_per_second": 5660.112
+ },
+ {
+ "epoch": 0.6914893617021277,
+ "grad_norm": 0.8780985630278836,
+ "learning_rate": 7.32977281517627e-06,
+ "loss": 0.39997851848602295,
+ "num_input_tokens_seen": 449446632,
+ "step": 10270,
+ "train_runtime": 79406.7923,
+ "train_tokens_per_second": 5660.053
+ },
+ {
+ "epoch": 0.6918260166980879,
+ "grad_norm": 0.7238138438393178,
+ "learning_rate": 7.3274329827691406e-06,
+ "loss": 0.3859497308731079,
+ "num_input_tokens_seen": 449659944,
+ "step": 10275,
+ "train_runtime": 79449.7577,
+ "train_tokens_per_second": 5659.677
+ },
+ {
+ "epoch": 0.692162671694048,
+ "grad_norm": 0.6417831874350313,
+ "learning_rate": 7.325092499502796e-06,
+ "loss": 0.35598082542419435,
+ "num_input_tokens_seen": 449892656,
+ "step": 10280,
+ "train_runtime": 79489.514,
+ "train_tokens_per_second": 5659.774
+ },
+ {
+ "epoch": 0.6924993266900081,
+ "grad_norm": 0.6886975793380838,
+ "learning_rate": 7.322751366031751e-06,
+ "loss": 0.3739547967910767,
+ "num_input_tokens_seen": 450115160,
+ "step": 10285,
+ "train_runtime": 79528.7582,
+ "train_tokens_per_second": 5659.779
+ },
+ {
+ "epoch": 0.6928359816859683,
+ "grad_norm": 0.8067129855431404,
+ "learning_rate": 7.320409583010691e-06,
+ "loss": 0.3593981027603149,
+ "num_input_tokens_seen": 450339952,
+ "step": 10290,
+ "train_runtime": 79567.7249,
+ "train_tokens_per_second": 5659.832
+ },
+ {
+ "epoch": 0.6931726366819284,
+ "grad_norm": 0.720222949795357,
+ "learning_rate": 7.318067151094493e-06,
+ "loss": 0.4208518981933594,
+ "num_input_tokens_seen": 450562568,
+ "step": 10295,
+ "train_runtime": 79603.3438,
+ "train_tokens_per_second": 5660.096
+ },
+ {
+ "epoch": 0.6935092916778886,
+ "grad_norm": 0.6662609649687479,
+ "learning_rate": 7.3157240709382046e-06,
+ "loss": 0.3863611936569214,
+ "num_input_tokens_seen": 450794024,
+ "step": 10300,
+ "train_runtime": 79634.484,
+ "train_tokens_per_second": 5660.789
+ },
+ {
+ "epoch": 0.6938459466738487,
+ "grad_norm": 0.7343868155686462,
+ "learning_rate": 7.313380343197066e-06,
+ "loss": 0.3639280080795288,
+ "num_input_tokens_seen": 451027856,
+ "step": 10305,
+ "train_runtime": 79674.0903,
+ "train_tokens_per_second": 5660.91
+ },
+ {
+ "epoch": 0.6941826016698088,
+ "grad_norm": 0.7926474138155267,
+ "learning_rate": 7.311035968526492e-06,
+ "loss": 0.3794968605041504,
+ "num_input_tokens_seen": 451247432,
+ "step": 10310,
+ "train_runtime": 79715.3108,
+ "train_tokens_per_second": 5660.737
+ },
+ {
+ "epoch": 0.694519256665769,
+ "grad_norm": 0.8428213530948624,
+ "learning_rate": 7.308690947582078e-06,
+ "loss": 0.38078649044036866,
+ "num_input_tokens_seen": 451456696,
+ "step": 10315,
+ "train_runtime": 79750.434,
+ "train_tokens_per_second": 5660.868
+ },
+ {
+ "epoch": 0.6948559116617291,
+ "grad_norm": 0.7185545573840266,
+ "learning_rate": 7.306345281019602e-06,
+ "loss": 0.37413511276245115,
+ "num_input_tokens_seen": 451677704,
+ "step": 10320,
+ "train_runtime": 79790.1325,
+ "train_tokens_per_second": 5660.822
+ },
+ {
+ "epoch": 0.6951925666576892,
+ "grad_norm": 0.7678020136994413,
+ "learning_rate": 7.303998969495023e-06,
+ "loss": 0.40517454147338866,
+ "num_input_tokens_seen": 451892384,
+ "step": 10325,
+ "train_runtime": 79832.0377,
+ "train_tokens_per_second": 5660.539
+ },
+ {
+ "epoch": 0.6955292216536494,
+ "grad_norm": 0.7116247415151229,
+ "learning_rate": 7.301652013664479e-06,
+ "loss": 0.39863314628601076,
+ "num_input_tokens_seen": 452118752,
+ "step": 10330,
+ "train_runtime": 79868.7262,
+ "train_tokens_per_second": 5660.773
+ },
+ {
+ "epoch": 0.6958658766496095,
+ "grad_norm": 0.7656084452706404,
+ "learning_rate": 7.299304414184288e-06,
+ "loss": 0.40618147850036623,
+ "num_input_tokens_seen": 452338552,
+ "step": 10335,
+ "train_runtime": 79909.7465,
+ "train_tokens_per_second": 5660.618
+ },
+ {
+ "epoch": 0.6962025316455697,
+ "grad_norm": 0.6668364749984544,
+ "learning_rate": 7.296956171710951e-06,
+ "loss": 0.3879892587661743,
+ "num_input_tokens_seen": 452573592,
+ "step": 10340,
+ "train_runtime": 79946.2615,
+ "train_tokens_per_second": 5660.973
+ },
+ {
+ "epoch": 0.6965391866415298,
+ "grad_norm": 0.8038854323996706,
+ "learning_rate": 7.294607286901146e-06,
+ "loss": 0.3926121711730957,
+ "num_input_tokens_seen": 452795696,
+ "step": 10345,
+ "train_runtime": 79987.6707,
+ "train_tokens_per_second": 5660.819
+ },
+ {
+ "epoch": 0.69687584163749,
+ "grad_norm": 0.7769622893093939,
+ "learning_rate": 7.292257760411726e-06,
+ "loss": 0.3886493444442749,
+ "num_input_tokens_seen": 453009328,
+ "step": 10350,
+ "train_runtime": 80023.6895,
+ "train_tokens_per_second": 5660.94
+ },
+ {
+ "epoch": 0.6972124966334501,
+ "grad_norm": 0.7478012896643832,
+ "learning_rate": 7.289907592899737e-06,
+ "loss": 0.37845487594604493,
+ "num_input_tokens_seen": 453214464,
+ "step": 10355,
+ "train_runtime": 80062.9601,
+ "train_tokens_per_second": 5660.726
+ },
+ {
+ "epoch": 0.6975491516294102,
+ "grad_norm": 0.7047299342802962,
+ "learning_rate": 7.287556785022389e-06,
+ "loss": 0.3894767761230469,
+ "num_input_tokens_seen": 453437368,
+ "step": 10360,
+ "train_runtime": 80102.9256,
+ "train_tokens_per_second": 5660.684
+ },
+ {
+ "epoch": 0.6978858066253704,
+ "grad_norm": 0.7604979984461925,
+ "learning_rate": 7.285205337437082e-06,
+ "loss": 0.3892651557922363,
+ "num_input_tokens_seen": 453647680,
+ "step": 10365,
+ "train_runtime": 80141.1595,
+ "train_tokens_per_second": 5660.608
+ },
+ {
+ "epoch": 0.6982224616213305,
+ "grad_norm": 0.7970440490683166,
+ "learning_rate": 7.2828532508013895e-06,
+ "loss": 0.399712347984314,
+ "num_input_tokens_seen": 453867592,
+ "step": 10370,
+ "train_runtime": 80178.1608,
+ "train_tokens_per_second": 5660.738
+ },
+ {
+ "epoch": 0.6985591166172906,
+ "grad_norm": 0.7584286596844286,
+ "learning_rate": 7.2805005257730625e-06,
+ "loss": 0.3772284030914307,
+ "num_input_tokens_seen": 454087008,
+ "step": 10375,
+ "train_runtime": 80211.2544,
+ "train_tokens_per_second": 5661.138
+ },
+ {
+ "epoch": 0.6988957716132508,
+ "grad_norm": 0.6810825697097003,
+ "learning_rate": 7.278147163010037e-06,
+ "loss": 0.36341052055358886,
+ "num_input_tokens_seen": 454313224,
+ "step": 10380,
+ "train_runtime": 80250.7125,
+ "train_tokens_per_second": 5661.174
+ },
+ {
+ "epoch": 0.6992324266092109,
+ "grad_norm": 0.7849426670217893,
+ "learning_rate": 7.275793163170423e-06,
+ "loss": 0.36755235195159913,
+ "num_input_tokens_seen": 454539992,
+ "step": 10385,
+ "train_runtime": 80288.0816,
+ "train_tokens_per_second": 5661.363
+ },
+ {
+ "epoch": 0.6995690816051711,
+ "grad_norm": 0.7022327990344984,
+ "learning_rate": 7.273438526912506e-06,
+ "loss": 0.3956979990005493,
+ "num_input_tokens_seen": 454748824,
+ "step": 10390,
+ "train_runtime": 80325.0195,
+ "train_tokens_per_second": 5661.36
+ },
+ {
+ "epoch": 0.6999057366011312,
+ "grad_norm": 0.7244301627454164,
+ "learning_rate": 7.271083254894757e-06,
+ "loss": 0.36125593185424804,
+ "num_input_tokens_seen": 454962456,
+ "step": 10395,
+ "train_runtime": 80364.6433,
+ "train_tokens_per_second": 5661.227
+ },
+ {
+ "epoch": 0.7002423915970913,
+ "grad_norm": 0.7856991901608305,
+ "learning_rate": 7.268727347775815e-06,
+ "loss": 0.3766735076904297,
+ "num_input_tokens_seen": 455189544,
+ "step": 10400,
+ "train_runtime": 80398.3718,
+ "train_tokens_per_second": 5661.676
+ },
+ {
+ "epoch": 0.7005790465930515,
+ "grad_norm": 0.7076340782949039,
+ "learning_rate": 7.266370806214506e-06,
+ "loss": 0.32742795944213865,
+ "num_input_tokens_seen": 455409928,
+ "step": 10405,
+ "train_runtime": 80446.5155,
+ "train_tokens_per_second": 5661.027
+ },
+ {
+ "epoch": 0.7009157015890116,
+ "grad_norm": 0.6900523877236565,
+ "learning_rate": 7.264013630869827e-06,
+ "loss": 0.37685844898223875,
+ "num_input_tokens_seen": 455626104,
+ "step": 10410,
+ "train_runtime": 80491.2258,
+ "train_tokens_per_second": 5660.569
+ },
+ {
+ "epoch": 0.7012523565849718,
+ "grad_norm": 0.6710950318779684,
+ "learning_rate": 7.261655822400957e-06,
+ "loss": 0.39275636672973635,
+ "num_input_tokens_seen": 455857232,
+ "step": 10415,
+ "train_runtime": 80526.04,
+ "train_tokens_per_second": 5660.992
+ },
+ {
+ "epoch": 0.7015890115809319,
+ "grad_norm": 0.7198657361358307,
+ "learning_rate": 7.259297381467248e-06,
+ "loss": 0.38013458251953125,
+ "num_input_tokens_seen": 456077832,
+ "step": 10420,
+ "train_runtime": 80569.4365,
+ "train_tokens_per_second": 5660.68
+ },
+ {
+ "epoch": 0.701925666576892,
+ "grad_norm": 0.7005819646574629,
+ "learning_rate": 7.2569383087282285e-06,
+ "loss": 0.34854826927185056,
+ "num_input_tokens_seen": 456272960,
+ "step": 10425,
+ "train_runtime": 80611.1139,
+ "train_tokens_per_second": 5660.174
+ },
+ {
+ "epoch": 0.7022623215728522,
+ "grad_norm": 0.7072800232396435,
+ "learning_rate": 7.25457860484361e-06,
+ "loss": 0.39771556854248047,
+ "num_input_tokens_seen": 456493600,
+ "step": 10430,
+ "train_runtime": 80651.8402,
+ "train_tokens_per_second": 5660.052
+ },
+ {
+ "epoch": 0.7025989765688123,
+ "grad_norm": 0.7445696785976343,
+ "learning_rate": 7.252218270473274e-06,
+ "loss": 0.3760495662689209,
+ "num_input_tokens_seen": 456704912,
+ "step": 10435,
+ "train_runtime": 80692.0569,
+ "train_tokens_per_second": 5659.85
+ },
+ {
+ "epoch": 0.7029356315647725,
+ "grad_norm": 0.7457222056984448,
+ "learning_rate": 7.249857306277279e-06,
+ "loss": 0.369310188293457,
+ "num_input_tokens_seen": 456921864,
+ "step": 10440,
+ "train_runtime": 80726.6828,
+ "train_tokens_per_second": 5660.109
+ },
+ {
+ "epoch": 0.7032722865607326,
+ "grad_norm": 0.9221604408221112,
+ "learning_rate": 7.247495712915864e-06,
+ "loss": 0.35996944904327394,
+ "num_input_tokens_seen": 457143032,
+ "step": 10445,
+ "train_runtime": 80760.6102,
+ "train_tokens_per_second": 5660.47
+ },
+ {
+ "epoch": 0.7036089415566927,
+ "grad_norm": 0.6862985079879047,
+ "learning_rate": 7.245133491049439e-06,
+ "loss": 0.3736717939376831,
+ "num_input_tokens_seen": 457363400,
+ "step": 10450,
+ "train_runtime": 80799.8061,
+ "train_tokens_per_second": 5660.452
+ },
+ {
+ "epoch": 0.7039455965526529,
+ "grad_norm": 0.6935497313758425,
+ "learning_rate": 7.242770641338592e-06,
+ "loss": 0.4129630088806152,
+ "num_input_tokens_seen": 457584008,
+ "step": 10455,
+ "train_runtime": 80841.1453,
+ "train_tokens_per_second": 5660.286
+ },
+ {
+ "epoch": 0.704282251548613,
+ "grad_norm": 0.7202390162158359,
+ "learning_rate": 7.240407164444088e-06,
+ "loss": 0.36982312202453616,
+ "num_input_tokens_seen": 457798528,
+ "step": 10460,
+ "train_runtime": 80879.602,
+ "train_tokens_per_second": 5660.247
+ },
+ {
+ "epoch": 0.7046189065445732,
+ "grad_norm": 0.7863736874260931,
+ "learning_rate": 7.238043061026862e-06,
+ "loss": 0.34938392639160154,
+ "num_input_tokens_seen": 458006080,
+ "step": 10465,
+ "train_runtime": 80921.1996,
+ "train_tokens_per_second": 5659.902
+ },
+ {
+ "epoch": 0.7049555615405333,
+ "grad_norm": 0.7013473236004093,
+ "learning_rate": 7.235678331748034e-06,
+ "loss": 0.3546873092651367,
+ "num_input_tokens_seen": 458203208,
+ "step": 10470,
+ "train_runtime": 80956.6805,
+ "train_tokens_per_second": 5659.857
+ },
+ {
+ "epoch": 0.7052922165364934,
+ "grad_norm": 0.7707435582118967,
+ "learning_rate": 7.233312977268887e-06,
+ "loss": 0.40183773040771487,
+ "num_input_tokens_seen": 458427952,
+ "step": 10475,
+ "train_runtime": 80993.4895,
+ "train_tokens_per_second": 5660.059
+ },
+ {
+ "epoch": 0.7056288715324536,
+ "grad_norm": 0.7366337541289394,
+ "learning_rate": 7.230946998250889e-06,
+ "loss": 0.36978516578674314,
+ "num_input_tokens_seen": 458631504,
+ "step": 10480,
+ "train_runtime": 81028.6939,
+ "train_tokens_per_second": 5660.112
+ },
+ {
+ "epoch": 0.7059655265284137,
+ "grad_norm": 0.7121052684917011,
+ "learning_rate": 7.228580395355676e-06,
+ "loss": 0.34577772617340086,
+ "num_input_tokens_seen": 458845056,
+ "step": 10485,
+ "train_runtime": 81072.1146,
+ "train_tokens_per_second": 5659.715
+ },
+ {
+ "epoch": 0.7063021815243739,
+ "grad_norm": 0.7741334918584597,
+ "learning_rate": 7.226213169245062e-06,
+ "loss": 0.38037862777709963,
+ "num_input_tokens_seen": 459063872,
+ "step": 10490,
+ "train_runtime": 81113.0274,
+ "train_tokens_per_second": 5659.558
+ },
+ {
+ "epoch": 0.706638836520334,
+ "grad_norm": 0.7871663193272351,
+ "learning_rate": 7.223845320581035e-06,
+ "loss": 0.3649828672409058,
+ "num_input_tokens_seen": 459276768,
+ "step": 10495,
+ "train_runtime": 81158.0461,
+ "train_tokens_per_second": 5659.042
+ },
+ {
+ "epoch": 0.7069754915162941,
+ "grad_norm": 0.8550643751341503,
+ "learning_rate": 7.221476850025754e-06,
+ "loss": 0.3691579818725586,
+ "num_input_tokens_seen": 459504320,
+ "step": 10500,
+ "train_runtime": 81197.4298,
+ "train_tokens_per_second": 5659.099
+ },
+ {
+ "epoch": 0.7073121465122543,
+ "grad_norm": 0.6802216728098077,
+ "learning_rate": 7.219107758241557e-06,
+ "loss": 0.381423807144165,
+ "num_input_tokens_seen": 459726544,
+ "step": 10505,
+ "train_runtime": 81239.3131,
+ "train_tokens_per_second": 5658.917
+ },
+ {
+ "epoch": 0.7076488015082144,
+ "grad_norm": 0.7476015422415041,
+ "learning_rate": 7.2167380458909505e-06,
+ "loss": 0.3745572090148926,
+ "num_input_tokens_seen": 459949656,
+ "step": 10510,
+ "train_runtime": 81274.379,
+ "train_tokens_per_second": 5659.221
+ },
+ {
+ "epoch": 0.7079854565041745,
+ "grad_norm": 0.6742333237089346,
+ "learning_rate": 7.214367713636619e-06,
+ "loss": 0.375703763961792,
+ "num_input_tokens_seen": 460172128,
+ "step": 10515,
+ "train_runtime": 81312.8461,
+ "train_tokens_per_second": 5659.28
+ },
+ {
+ "epoch": 0.7083221115001347,
+ "grad_norm": 0.6954522791161296,
+ "learning_rate": 7.211996762141418e-06,
+ "loss": 0.37557039260864256,
+ "num_input_tokens_seen": 460390848,
+ "step": 10520,
+ "train_runtime": 81345.9866,
+ "train_tokens_per_second": 5659.663
+ },
+ {
+ "epoch": 0.7086587664960948,
+ "grad_norm": 0.7339967212433919,
+ "learning_rate": 7.209625192068374e-06,
+ "loss": 0.4028931617736816,
+ "num_input_tokens_seen": 460611832,
+ "step": 10525,
+ "train_runtime": 81379.8389,
+ "train_tokens_per_second": 5660.024
+ },
+ {
+ "epoch": 0.708995421492055,
+ "grad_norm": 0.701987120656059,
+ "learning_rate": 7.207253004080692e-06,
+ "loss": 0.37696285247802735,
+ "num_input_tokens_seen": 460844344,
+ "step": 10530,
+ "train_runtime": 81418.5196,
+ "train_tokens_per_second": 5660.191
+ },
+ {
+ "epoch": 0.7093320764880151,
+ "grad_norm": 0.7755594445584074,
+ "learning_rate": 7.204880198841746e-06,
+ "loss": 0.34694790840148926,
+ "num_input_tokens_seen": 461058504,
+ "step": 10535,
+ "train_runtime": 81453.7678,
+ "train_tokens_per_second": 5660.37
+ },
+ {
+ "epoch": 0.7096687314839752,
+ "grad_norm": 0.8426741622679441,
+ "learning_rate": 7.2025067770150835e-06,
+ "loss": 0.4015054225921631,
+ "num_input_tokens_seen": 461286472,
+ "step": 10540,
+ "train_runtime": 81493.9088,
+ "train_tokens_per_second": 5660.38
+ },
+ {
+ "epoch": 0.7100053864799354,
+ "grad_norm": 0.7557428415974053,
+ "learning_rate": 7.2001327392644235e-06,
+ "loss": 0.3497929573059082,
+ "num_input_tokens_seen": 461515496,
+ "step": 10545,
+ "train_runtime": 81528.2463,
+ "train_tokens_per_second": 5660.805
+ },
+ {
+ "epoch": 0.7103420414758955,
+ "grad_norm": 0.7845759323690452,
+ "learning_rate": 7.197758086253659e-06,
+ "loss": 0.39971365928649905,
+ "num_input_tokens_seen": 461738504,
+ "step": 10550,
+ "train_runtime": 81568.9369,
+ "train_tokens_per_second": 5660.715
+ },
+ {
+ "epoch": 0.7106786964718557,
+ "grad_norm": 0.7112358056324952,
+ "learning_rate": 7.195382818646854e-06,
+ "loss": 0.39182677268981936,
+ "num_input_tokens_seen": 461947016,
+ "step": 10555,
+ "train_runtime": 81608.3983,
+ "train_tokens_per_second": 5660.533
+ },
+ {
+ "epoch": 0.7110153514678158,
+ "grad_norm": 0.772603293296723,
+ "learning_rate": 7.1930069371082445e-06,
+ "loss": 0.36522927284240725,
+ "num_input_tokens_seen": 462152784,
+ "step": 10560,
+ "train_runtime": 81652.6367,
+ "train_tokens_per_second": 5659.986
+ },
+ {
+ "epoch": 0.7113520064637759,
+ "grad_norm": 0.7156871566744746,
+ "learning_rate": 7.190630442302238e-06,
+ "loss": 0.3833275079727173,
+ "num_input_tokens_seen": 462379128,
+ "step": 10565,
+ "train_runtime": 81694.2563,
+ "train_tokens_per_second": 5659.873
+ },
+ {
+ "epoch": 0.7116886614597361,
+ "grad_norm": 0.7033879162554953,
+ "learning_rate": 7.188253334893414e-06,
+ "loss": 0.3999313831329346,
+ "num_input_tokens_seen": 462598648,
+ "step": 10570,
+ "train_runtime": 81730.3714,
+ "train_tokens_per_second": 5660.058
+ },
+ {
+ "epoch": 0.7120253164556962,
+ "grad_norm": 0.8834420627247958,
+ "learning_rate": 7.18587561554652e-06,
+ "loss": 0.3910887956619263,
+ "num_input_tokens_seen": 462807208,
+ "step": 10575,
+ "train_runtime": 81771.4349,
+ "train_tokens_per_second": 5659.766
+ },
+ {
+ "epoch": 0.7123619714516564,
+ "grad_norm": 0.6918016219416666,
+ "learning_rate": 7.183497284926483e-06,
+ "loss": 0.36777448654174805,
+ "num_input_tokens_seen": 463042928,
+ "step": 10580,
+ "train_runtime": 81807.4141,
+ "train_tokens_per_second": 5660.158
+ },
+ {
+ "epoch": 0.7126986264476165,
+ "grad_norm": 0.7599277759479904,
+ "learning_rate": 7.181118343698392e-06,
+ "loss": 0.36260690689086916,
+ "num_input_tokens_seen": 463252424,
+ "step": 10585,
+ "train_runtime": 81851.2205,
+ "train_tokens_per_second": 5659.689
+ },
+ {
+ "epoch": 0.7130352814435766,
+ "grad_norm": 0.7224431307545425,
+ "learning_rate": 7.178738792527512e-06,
+ "loss": 0.39423255920410155,
+ "num_input_tokens_seen": 463469984,
+ "step": 10590,
+ "train_runtime": 81891.039,
+ "train_tokens_per_second": 5659.593
+ },
+ {
+ "epoch": 0.7133719364395368,
+ "grad_norm": 0.7536526351492208,
+ "learning_rate": 7.176358632079277e-06,
+ "loss": 0.37251434326171873,
+ "num_input_tokens_seen": 463682208,
+ "step": 10595,
+ "train_runtime": 81929.5391,
+ "train_tokens_per_second": 5659.524
+ },
+ {
+ "epoch": 0.7137085914354969,
+ "grad_norm": 0.6793493050362683,
+ "learning_rate": 7.173977863019288e-06,
+ "loss": 0.37181549072265624,
+ "num_input_tokens_seen": 463907464,
+ "step": 10600,
+ "train_runtime": 81970.1484,
+ "train_tokens_per_second": 5659.468
+ },
+ {
+ "epoch": 0.7140452464314571,
+ "grad_norm": 0.6822436534253042,
+ "learning_rate": 7.171596486013323e-06,
+ "loss": 0.34790878295898436,
+ "num_input_tokens_seen": 464138256,
+ "step": 10605,
+ "train_runtime": 82007.4481,
+ "train_tokens_per_second": 5659.709
+ },
+ {
+ "epoch": 0.7143819014274172,
+ "grad_norm": 0.8051213804159209,
+ "learning_rate": 7.169214501727326e-06,
+ "loss": 0.37795867919921877,
+ "num_input_tokens_seen": 464338936,
+ "step": 10610,
+ "train_runtime": 82043.2488,
+ "train_tokens_per_second": 5659.685
+ },
+ {
+ "epoch": 0.7147185564233773,
+ "grad_norm": 0.686749483511892,
+ "learning_rate": 7.16683191082741e-06,
+ "loss": 0.36943345069885253,
+ "num_input_tokens_seen": 464560144,
+ "step": 10615,
+ "train_runtime": 82079.0859,
+ "train_tokens_per_second": 5659.909
+ },
+ {
+ "epoch": 0.7150552114193375,
+ "grad_norm": 0.8746419009334871,
+ "learning_rate": 7.16444871397986e-06,
+ "loss": 0.368468713760376,
+ "num_input_tokens_seen": 464765496,
+ "step": 10620,
+ "train_runtime": 82117.3295,
+ "train_tokens_per_second": 5659.774
+ },
+ {
+ "epoch": 0.7153918664152976,
+ "grad_norm": 1.042013296164164,
+ "learning_rate": 7.162064911851129e-06,
+ "loss": 0.36743860244750975,
+ "num_input_tokens_seen": 464978408,
+ "step": 10625,
+ "train_runtime": 82160.6887,
+ "train_tokens_per_second": 5659.378
+ },
+ {
+ "epoch": 0.7157285214112578,
+ "grad_norm": 0.7321276232893942,
+ "learning_rate": 7.15968050510784e-06,
+ "loss": 0.40234060287475587,
+ "num_input_tokens_seen": 465192704,
+ "step": 10630,
+ "train_runtime": 82198.1606,
+ "train_tokens_per_second": 5659.405
+ },
+ {
+ "epoch": 0.7160651764072179,
+ "grad_norm": 0.8428920187826066,
+ "learning_rate": 7.1572954944167834e-06,
+ "loss": 0.3755232810974121,
+ "num_input_tokens_seen": 465408616,
+ "step": 10635,
+ "train_runtime": 82239.586,
+ "train_tokens_per_second": 5659.18
+ },
+ {
+ "epoch": 0.716401831403178,
+ "grad_norm": 0.679996578354662,
+ "learning_rate": 7.154909880444922e-06,
+ "loss": 0.3969583988189697,
+ "num_input_tokens_seen": 465633232,
+ "step": 10640,
+ "train_runtime": 82282.7769,
+ "train_tokens_per_second": 5658.939
+ },
+ {
+ "epoch": 0.7167384863991382,
+ "grad_norm": 1.5740350922344166,
+ "learning_rate": 7.152523663859384e-06,
+ "loss": 0.37885403633117676,
+ "num_input_tokens_seen": 465845256,
+ "step": 10645,
+ "train_runtime": 82324.9766,
+ "train_tokens_per_second": 5658.614
+ },
+ {
+ "epoch": 0.7170751413950983,
+ "grad_norm": 0.8328362306471117,
+ "learning_rate": 7.150136845327466e-06,
+ "loss": 0.3639181613922119,
+ "num_input_tokens_seen": 466056584,
+ "step": 10650,
+ "train_runtime": 82369.823,
+ "train_tokens_per_second": 5658.099
+ },
+ {
+ "epoch": 0.7174117963910585,
+ "grad_norm": 0.737238654335971,
+ "learning_rate": 7.147749425516637e-06,
+ "loss": 0.3720954418182373,
+ "num_input_tokens_seen": 466287992,
+ "step": 10655,
+ "train_runtime": 82407.102,
+ "train_tokens_per_second": 5658.347
+ },
+ {
+ "epoch": 0.7177484513870186,
+ "grad_norm": 0.7078523897720794,
+ "learning_rate": 7.145361405094529e-06,
+ "loss": 0.3578562021255493,
+ "num_input_tokens_seen": 466523176,
+ "step": 10660,
+ "train_runtime": 82439.2799,
+ "train_tokens_per_second": 5658.991
+ },
+ {
+ "epoch": 0.7180851063829787,
+ "grad_norm": 0.7362901383213949,
+ "learning_rate": 7.142972784728945e-06,
+ "loss": 0.3409886360168457,
+ "num_input_tokens_seen": 466734112,
+ "step": 10665,
+ "train_runtime": 82483.0478,
+ "train_tokens_per_second": 5658.546
+ },
+ {
+ "epoch": 0.7184217613789389,
+ "grad_norm": 0.8146924235270421,
+ "learning_rate": 7.140583565087856e-06,
+ "loss": 0.35494747161865237,
+ "num_input_tokens_seen": 466952312,
+ "step": 10670,
+ "train_runtime": 82519.1924,
+ "train_tokens_per_second": 5658.712
+ },
+ {
+ "epoch": 0.718758416374899,
+ "grad_norm": 0.7593849226158301,
+ "learning_rate": 7.138193746839399e-06,
+ "loss": 0.3933510065078735,
+ "num_input_tokens_seen": 467185248,
+ "step": 10675,
+ "train_runtime": 82565.1418,
+ "train_tokens_per_second": 5658.384
+ },
+ {
+ "epoch": 0.7190950713708592,
+ "grad_norm": 0.7169033125551184,
+ "learning_rate": 7.135803330651878e-06,
+ "loss": 0.37902536392211916,
+ "num_input_tokens_seen": 467421448,
+ "step": 10680,
+ "train_runtime": 82599.409,
+ "train_tokens_per_second": 5658.896
+ },
+ {
+ "epoch": 0.7194317263668193,
+ "grad_norm": 0.6927480355842544,
+ "learning_rate": 7.133412317193766e-06,
+ "loss": 0.3649748802185059,
+ "num_input_tokens_seen": 467648088,
+ "step": 10685,
+ "train_runtime": 82647.9029,
+ "train_tokens_per_second": 5658.318
+ },
+ {
+ "epoch": 0.7197683813627794,
+ "grad_norm": 0.8368737935217295,
+ "learning_rate": 7.131020707133703e-06,
+ "loss": 0.39012324810028076,
+ "num_input_tokens_seen": 467852784,
+ "step": 10690,
+ "train_runtime": 82685.2143,
+ "train_tokens_per_second": 5658.24
+ },
+ {
+ "epoch": 0.7201050363587396,
+ "grad_norm": 0.7246874489536738,
+ "learning_rate": 7.128628501140495e-06,
+ "loss": 0.3994007587432861,
+ "num_input_tokens_seen": 468062360,
+ "step": 10695,
+ "train_runtime": 82719.4298,
+ "train_tokens_per_second": 5658.433
+ },
+ {
+ "epoch": 0.7204416913546997,
+ "grad_norm": 0.7034507707765917,
+ "learning_rate": 7.126235699883115e-06,
+ "loss": 0.3543556213378906,
+ "num_input_tokens_seen": 468260752,
+ "step": 10700,
+ "train_runtime": 82762.0259,
+ "train_tokens_per_second": 5657.918
+ },
+ {
+ "epoch": 0.7207783463506598,
+ "grad_norm": 0.6713729985803661,
+ "learning_rate": 7.1238423040306995e-06,
+ "loss": 0.3659778594970703,
+ "num_input_tokens_seen": 468480296,
+ "step": 10705,
+ "train_runtime": 82807.3194,
+ "train_tokens_per_second": 5657.474
+ },
+ {
+ "epoch": 0.72111500134662,
+ "grad_norm": 0.6714547248755331,
+ "learning_rate": 7.121448314252557e-06,
+ "loss": 0.33471217155456545,
+ "num_input_tokens_seen": 468714640,
+ "step": 10710,
+ "train_runtime": 82849.0161,
+ "train_tokens_per_second": 5657.456
+ },
+ {
+ "epoch": 0.7214516563425801,
+ "grad_norm": 0.7542079623814812,
+ "learning_rate": 7.1190537312181586e-06,
+ "loss": 0.3590662479400635,
+ "num_input_tokens_seen": 468920200,
+ "step": 10715,
+ "train_runtime": 82883.6718,
+ "train_tokens_per_second": 5657.57
+ },
+ {
+ "epoch": 0.7217883113385403,
+ "grad_norm": 0.788689096156777,
+ "learning_rate": 7.116658555597141e-06,
+ "loss": 0.35546276569366453,
+ "num_input_tokens_seen": 469105088,
+ "step": 10720,
+ "train_runtime": 82920.6765,
+ "train_tokens_per_second": 5657.275
+ },
+ {
+ "epoch": 0.7221249663345004,
+ "grad_norm": 0.7720777482295464,
+ "learning_rate": 7.114262788059308e-06,
+ "loss": 0.36028232574462893,
+ "num_input_tokens_seen": 469316144,
+ "step": 10725,
+ "train_runtime": 82956.2558,
+ "train_tokens_per_second": 5657.393
+ },
+ {
+ "epoch": 0.7224616213304605,
+ "grad_norm": 0.7595426301739953,
+ "learning_rate": 7.111866429274627e-06,
+ "loss": 0.3734895706176758,
+ "num_input_tokens_seen": 469533344,
+ "step": 10730,
+ "train_runtime": 82995.9693,
+ "train_tokens_per_second": 5657.303
+ },
+ {
+ "epoch": 0.7227982763264207,
+ "grad_norm": 0.7895201856991956,
+ "learning_rate": 7.1094694799132356e-06,
+ "loss": 0.38394627571105955,
+ "num_input_tokens_seen": 469757776,
+ "step": 10735,
+ "train_runtime": 83039.5254,
+ "train_tokens_per_second": 5657.038
+ },
+ {
+ "epoch": 0.7231349313223808,
+ "grad_norm": 0.7170404350301777,
+ "learning_rate": 7.10707194064543e-06,
+ "loss": 0.355570125579834,
+ "num_input_tokens_seen": 469973048,
+ "step": 10740,
+ "train_runtime": 83082.4999,
+ "train_tokens_per_second": 5656.703
+ },
+ {
+ "epoch": 0.723471586318341,
+ "grad_norm": 0.751527765717788,
+ "learning_rate": 7.104673812141676e-06,
+ "loss": 0.4048286437988281,
+ "num_input_tokens_seen": 470200352,
+ "step": 10745,
+ "train_runtime": 83123.5825,
+ "train_tokens_per_second": 5656.642
+ },
+ {
+ "epoch": 0.7238082413143011,
+ "grad_norm": 0.7209513181054711,
+ "learning_rate": 7.102275095072601e-06,
+ "loss": 0.3921640157699585,
+ "num_input_tokens_seen": 470434416,
+ "step": 10750,
+ "train_runtime": 83170.1972,
+ "train_tokens_per_second": 5656.286
+ },
+ {
+ "epoch": 0.7241448963102612,
+ "grad_norm": 0.7141189337990387,
+ "learning_rate": 7.099875790108998e-06,
+ "loss": 0.36566712856292727,
+ "num_input_tokens_seen": 470662208,
+ "step": 10755,
+ "train_runtime": 83219.7805,
+ "train_tokens_per_second": 5655.653
+ },
+ {
+ "epoch": 0.7244815513062214,
+ "grad_norm": 0.6677452072067016,
+ "learning_rate": 7.097475897921827e-06,
+ "loss": 0.37137765884399415,
+ "num_input_tokens_seen": 470876192,
+ "step": 10760,
+ "train_runtime": 83259.577,
+ "train_tokens_per_second": 5655.52
+ },
+ {
+ "epoch": 0.7248182063021815,
+ "grad_norm": 0.8654836254283663,
+ "learning_rate": 7.09507541918221e-06,
+ "loss": 0.3979093790054321,
+ "num_input_tokens_seen": 471091920,
+ "step": 10765,
+ "train_runtime": 83298.0964,
+ "train_tokens_per_second": 5655.494
+ },
+ {
+ "epoch": 0.7251548612981417,
+ "grad_norm": 0.6072122160998616,
+ "learning_rate": 7.092674354561433e-06,
+ "loss": 0.3879514217376709,
+ "num_input_tokens_seen": 471316336,
+ "step": 10770,
+ "train_runtime": 83339.0663,
+ "train_tokens_per_second": 5655.407
+ },
+ {
+ "epoch": 0.7254915162941018,
+ "grad_norm": 0.753636374019264,
+ "learning_rate": 7.090272704730945e-06,
+ "loss": 0.3511503219604492,
+ "num_input_tokens_seen": 471518800,
+ "step": 10775,
+ "train_runtime": 83375.4431,
+ "train_tokens_per_second": 5655.368
+ },
+ {
+ "epoch": 0.7258281712900619,
+ "grad_norm": 0.7575044992088539,
+ "learning_rate": 7.08787047036236e-06,
+ "loss": 0.35472354888916013,
+ "num_input_tokens_seen": 471754616,
+ "step": 10780,
+ "train_runtime": 83418.0251,
+ "train_tokens_per_second": 5655.308
+ },
+ {
+ "epoch": 0.7261648262860221,
+ "grad_norm": 0.822516802699336,
+ "learning_rate": 7.085467652127457e-06,
+ "loss": 0.3795314788818359,
+ "num_input_tokens_seen": 471981872,
+ "step": 10785,
+ "train_runtime": 83450.133,
+ "train_tokens_per_second": 5655.855
+ },
+ {
+ "epoch": 0.7265014812819822,
+ "grad_norm": 0.6385132066085284,
+ "learning_rate": 7.083064250698175e-06,
+ "loss": 0.3529500961303711,
+ "num_input_tokens_seen": 472213144,
+ "step": 10790,
+ "train_runtime": 83491.9091,
+ "train_tokens_per_second": 5655.795
+ },
+ {
+ "epoch": 0.7268381362779424,
+ "grad_norm": 0.6595130665327932,
+ "learning_rate": 7.0806602667466175e-06,
+ "loss": 0.37914767265319826,
+ "num_input_tokens_seen": 472443584,
+ "step": 10795,
+ "train_runtime": 83534.7223,
+ "train_tokens_per_second": 5655.655
+ },
+ {
+ "epoch": 0.7271747912739025,
+ "grad_norm": 0.6988131699183094,
+ "learning_rate": 7.078255700945051e-06,
+ "loss": 0.38105590343475343,
+ "num_input_tokens_seen": 472655528,
+ "step": 10800,
+ "train_runtime": 83566.9737,
+ "train_tokens_per_second": 5656.009
+ },
+ {
+ "epoch": 0.7275114462698626,
+ "grad_norm": 0.7284700984007974,
+ "learning_rate": 7.075850553965904e-06,
+ "loss": 0.366654634475708,
+ "num_input_tokens_seen": 472863848,
+ "step": 10805,
+ "train_runtime": 83605.9871,
+ "train_tokens_per_second": 5655.861
+ },
+ {
+ "epoch": 0.7278481012658228,
+ "grad_norm": 0.7093309792405998,
+ "learning_rate": 7.07344482648177e-06,
+ "loss": 0.3818976879119873,
+ "num_input_tokens_seen": 473087584,
+ "step": 10810,
+ "train_runtime": 83637.9528,
+ "train_tokens_per_second": 5656.375
+ },
+ {
+ "epoch": 0.7281847562617829,
+ "grad_norm": 0.7655242160543194,
+ "learning_rate": 7.071038519165402e-06,
+ "loss": 0.3804640293121338,
+ "num_input_tokens_seen": 473302440,
+ "step": 10815,
+ "train_runtime": 83680.6746,
+ "train_tokens_per_second": 5656.054
+ },
+ {
+ "epoch": 0.7285214112577431,
+ "grad_norm": 0.7319764295438455,
+ "learning_rate": 7.068631632689716e-06,
+ "loss": 0.3819848299026489,
+ "num_input_tokens_seen": 473528888,
+ "step": 10820,
+ "train_runtime": 83719.4607,
+ "train_tokens_per_second": 5656.139
+ },
+ {
+ "epoch": 0.7288580662537032,
+ "grad_norm": 9.329714081971986,
+ "learning_rate": 7.066224167727791e-06,
+ "loss": 0.3868781566619873,
+ "num_input_tokens_seen": 473749232,
+ "step": 10825,
+ "train_runtime": 83753.7938,
+ "train_tokens_per_second": 5656.451
+ },
+ {
+ "epoch": 0.7291947212496633,
+ "grad_norm": 0.841160080727543,
+ "learning_rate": 7.063816124952865e-06,
+ "loss": 0.36863956451416013,
+ "num_input_tokens_seen": 473966688,
+ "step": 10830,
+ "train_runtime": 83791.8425,
+ "train_tokens_per_second": 5656.478
+ },
+ {
+ "epoch": 0.7295313762456235,
+ "grad_norm": 0.6549972032651611,
+ "learning_rate": 7.061407505038341e-06,
+ "loss": 0.3608518362045288,
+ "num_input_tokens_seen": 474183856,
+ "step": 10835,
+ "train_runtime": 83830.8355,
+ "train_tokens_per_second": 5656.437
+ },
+ {
+ "epoch": 0.7298680312415836,
+ "grad_norm": 0.6949899381534901,
+ "learning_rate": 7.058998308657782e-06,
+ "loss": 0.3866151809692383,
+ "num_input_tokens_seen": 474414672,
+ "step": 10840,
+ "train_runtime": 83863.1172,
+ "train_tokens_per_second": 5657.012
+ },
+ {
+ "epoch": 0.7302046862375438,
+ "grad_norm": 0.6794084004266899,
+ "learning_rate": 7.056588536484912e-06,
+ "loss": 0.35731284618377684,
+ "num_input_tokens_seen": 474622840,
+ "step": 10845,
+ "train_runtime": 83892.8777,
+ "train_tokens_per_second": 5657.487
+ },
+ {
+ "epoch": 0.7305413412335039,
+ "grad_norm": 1.3762092654034057,
+ "learning_rate": 7.0541781891936145e-06,
+ "loss": 0.35231194496154783,
+ "num_input_tokens_seen": 474842808,
+ "step": 10850,
+ "train_runtime": 83931.0911,
+ "train_tokens_per_second": 5657.532
+ },
+ {
+ "epoch": 0.730877996229464,
+ "grad_norm": 0.8182830044909798,
+ "learning_rate": 7.0517672674579364e-06,
+ "loss": 0.3739452838897705,
+ "num_input_tokens_seen": 475060040,
+ "step": 10855,
+ "train_runtime": 83971.3377,
+ "train_tokens_per_second": 5657.407
+ },
+ {
+ "epoch": 0.7312146512254242,
+ "grad_norm": 0.7405063263799381,
+ "learning_rate": 7.049355771952087e-06,
+ "loss": 0.367911171913147,
+ "num_input_tokens_seen": 475270960,
+ "step": 10860,
+ "train_runtime": 84012.5713,
+ "train_tokens_per_second": 5657.141
+ },
+ {
+ "epoch": 0.7315513062213843,
+ "grad_norm": 0.743519981806223,
+ "learning_rate": 7.046943703350429e-06,
+ "loss": 0.3706895112991333,
+ "num_input_tokens_seen": 475496232,
+ "step": 10865,
+ "train_runtime": 84056.0111,
+ "train_tokens_per_second": 5656.897
+ },
+ {
+ "epoch": 0.7318879612173445,
+ "grad_norm": 0.6596197156889615,
+ "learning_rate": 7.044531062327492e-06,
+ "loss": 0.356654691696167,
+ "num_input_tokens_seen": 475709832,
+ "step": 10870,
+ "train_runtime": 84089.5859,
+ "train_tokens_per_second": 5657.179
+ },
+ {
+ "epoch": 0.7322246162133046,
+ "grad_norm": 0.6526147436786898,
+ "learning_rate": 7.042117849557963e-06,
+ "loss": 0.3708888053894043,
+ "num_input_tokens_seen": 475932072,
+ "step": 10875,
+ "train_runtime": 84133.6504,
+ "train_tokens_per_second": 5656.858
+ },
+ {
+ "epoch": 0.7325612712092647,
+ "grad_norm": 0.744802883666199,
+ "learning_rate": 7.0397040657166885e-06,
+ "loss": 0.36887059211730955,
+ "num_input_tokens_seen": 476162776,
+ "step": 10880,
+ "train_runtime": 84169.2025,
+ "train_tokens_per_second": 5657.209
+ },
+ {
+ "epoch": 0.7328979262052249,
+ "grad_norm": 0.7068220775813847,
+ "learning_rate": 7.037289711478676e-06,
+ "loss": 0.34908170700073243,
+ "num_input_tokens_seen": 476376384,
+ "step": 10885,
+ "train_runtime": 84204.347,
+ "train_tokens_per_second": 5657.385
+ },
+ {
+ "epoch": 0.733234581201185,
+ "grad_norm": 0.7211408903358524,
+ "learning_rate": 7.034874787519095e-06,
+ "loss": 0.3759069204330444,
+ "num_input_tokens_seen": 476599400,
+ "step": 10890,
+ "train_runtime": 84242.2013,
+ "train_tokens_per_second": 5657.49
+ },
+ {
+ "epoch": 0.7335712361971451,
+ "grad_norm": 0.6965350386219624,
+ "learning_rate": 7.0324592945132655e-06,
+ "loss": 0.4069540023803711,
+ "num_input_tokens_seen": 476821304,
+ "step": 10895,
+ "train_runtime": 84276.594,
+ "train_tokens_per_second": 5657.814
+ },
+ {
+ "epoch": 0.7339078911931053,
+ "grad_norm": 0.7259339581846704,
+ "learning_rate": 7.0300432331366764e-06,
+ "loss": 0.3802814245223999,
+ "num_input_tokens_seen": 477036424,
+ "step": 10900,
+ "train_runtime": 84309.7162,
+ "train_tokens_per_second": 5658.143
+ },
+ {
+ "epoch": 0.7342445461890654,
+ "grad_norm": 0.726630192334144,
+ "learning_rate": 7.02762660406497e-06,
+ "loss": 0.3741856813430786,
+ "num_input_tokens_seen": 477257288,
+ "step": 10905,
+ "train_runtime": 84348.0236,
+ "train_tokens_per_second": 5658.192
+ },
+ {
+ "epoch": 0.7345812011850256,
+ "grad_norm": 0.6353064783947603,
+ "learning_rate": 7.02520940797395e-06,
+ "loss": 0.3721340179443359,
+ "num_input_tokens_seen": 477464672,
+ "step": 10910,
+ "train_runtime": 84385.7451,
+ "train_tokens_per_second": 5658.12
+ },
+ {
+ "epoch": 0.7349178561809857,
+ "grad_norm": 0.6280488461396918,
+ "learning_rate": 7.022791645539576e-06,
+ "loss": 0.37077898979187013,
+ "num_input_tokens_seen": 477670600,
+ "step": 10915,
+ "train_runtime": 84423.4019,
+ "train_tokens_per_second": 5658.035
+ },
+ {
+ "epoch": 0.7352545111769458,
+ "grad_norm": 0.7582582970874048,
+ "learning_rate": 7.02037331743797e-06,
+ "loss": 0.4068927764892578,
+ "num_input_tokens_seen": 477891112,
+ "step": 10920,
+ "train_runtime": 84461.3924,
+ "train_tokens_per_second": 5658.101
+ },
+ {
+ "epoch": 0.735591166172906,
+ "grad_norm": 0.6908837097564919,
+ "learning_rate": 7.0179544243454076e-06,
+ "loss": 0.3525757551193237,
+ "num_input_tokens_seen": 478120960,
+ "step": 10925,
+ "train_runtime": 84493.3059,
+ "train_tokens_per_second": 5658.684
+ },
+ {
+ "epoch": 0.7359278211688661,
+ "grad_norm": 0.8025906131360969,
+ "learning_rate": 7.0155349669383245e-06,
+ "loss": 0.36997184753417967,
+ "num_input_tokens_seen": 478344240,
+ "step": 10930,
+ "train_runtime": 84534.5442,
+ "train_tokens_per_second": 5658.565
+ },
+ {
+ "epoch": 0.7362644761648263,
+ "grad_norm": 0.691884724659927,
+ "learning_rate": 7.013114945893316e-06,
+ "loss": 0.36500933170318606,
+ "num_input_tokens_seen": 478567880,
+ "step": 10935,
+ "train_runtime": 84579.7591,
+ "train_tokens_per_second": 5658.184
+ },
+ {
+ "epoch": 0.7366011311607864,
+ "grad_norm": 0.7278702815389843,
+ "learning_rate": 7.010694361887131e-06,
+ "loss": 0.3656963109970093,
+ "num_input_tokens_seen": 478795792,
+ "step": 10940,
+ "train_runtime": 84624.1001,
+ "train_tokens_per_second": 5657.913
+ },
+ {
+ "epoch": 0.7369377861567465,
+ "grad_norm": 0.6970616747051929,
+ "learning_rate": 7.008273215596681e-06,
+ "loss": 0.3773782253265381,
+ "num_input_tokens_seen": 479014424,
+ "step": 10945,
+ "train_runtime": 84661.672,
+ "train_tokens_per_second": 5657.984
+ },
+ {
+ "epoch": 0.7372744411527067,
+ "grad_norm": 0.7390738693319422,
+ "learning_rate": 7.005851507699029e-06,
+ "loss": 0.3759654760360718,
+ "num_input_tokens_seen": 479243632,
+ "step": 10950,
+ "train_runtime": 84694.665,
+ "train_tokens_per_second": 5658.487
+ },
+ {
+ "epoch": 0.7376110961486668,
+ "grad_norm": 0.6808946119767743,
+ "learning_rate": 7.003429238871399e-06,
+ "loss": 0.3657572031021118,
+ "num_input_tokens_seen": 479473192,
+ "step": 10955,
+ "train_runtime": 84738.238,
+ "train_tokens_per_second": 5658.286
+ },
+ {
+ "epoch": 0.737947751144627,
+ "grad_norm": 0.7178504014412501,
+ "learning_rate": 7.001006409791171e-06,
+ "loss": 0.40654420852661133,
+ "num_input_tokens_seen": 479687920,
+ "step": 10960,
+ "train_runtime": 84776.0398,
+ "train_tokens_per_second": 5658.296
+ },
+ {
+ "epoch": 0.7382844061405871,
+ "grad_norm": 0.7628213758613887,
+ "learning_rate": 6.998583021135882e-06,
+ "loss": 0.3780509948730469,
+ "num_input_tokens_seen": 479906096,
+ "step": 10965,
+ "train_runtime": 84815.2532,
+ "train_tokens_per_second": 5658.252
+ },
+ {
+ "epoch": 0.7386210611365472,
+ "grad_norm": 0.8356715268382896,
+ "learning_rate": 6.996159073583224e-06,
+ "loss": 0.36737675666809083,
+ "num_input_tokens_seen": 480125696,
+ "step": 10970,
+ "train_runtime": 84852.9815,
+ "train_tokens_per_second": 5658.324
+ },
+ {
+ "epoch": 0.7389577161325074,
+ "grad_norm": 0.8055302742937802,
+ "learning_rate": 6.993734567811046e-06,
+ "loss": 0.34843399524688723,
+ "num_input_tokens_seen": 480343296,
+ "step": 10975,
+ "train_runtime": 84896.1427,
+ "train_tokens_per_second": 5658.011
+ },
+ {
+ "epoch": 0.7392943711284675,
+ "grad_norm": 0.6594571098967035,
+ "learning_rate": 6.991309504497352e-06,
+ "loss": 0.37664797306060793,
+ "num_input_tokens_seen": 480566544,
+ "step": 10980,
+ "train_runtime": 84934.0218,
+ "train_tokens_per_second": 5658.116
+ },
+ {
+ "epoch": 0.7396310261244277,
+ "grad_norm": 0.7694992058673898,
+ "learning_rate": 6.988883884320307e-06,
+ "loss": 0.3457742214202881,
+ "num_input_tokens_seen": 480774232,
+ "step": 10985,
+ "train_runtime": 84970.4518,
+ "train_tokens_per_second": 5658.134
+ },
+ {
+ "epoch": 0.7399676811203878,
+ "grad_norm": 0.7402878857378402,
+ "learning_rate": 6.9864577079582254e-06,
+ "loss": 0.39705471992492675,
+ "num_input_tokens_seen": 481000200,
+ "step": 10990,
+ "train_runtime": 85008.5973,
+ "train_tokens_per_second": 5658.254
+ },
+ {
+ "epoch": 0.7403043361163479,
+ "grad_norm": 0.8393816654846036,
+ "learning_rate": 6.984030976089581e-06,
+ "loss": 0.3594042301177979,
+ "num_input_tokens_seen": 481219544,
+ "step": 10995,
+ "train_runtime": 85043.9927,
+ "train_tokens_per_second": 5658.478
+ },
+ {
+ "epoch": 0.7406409911123081,
+ "grad_norm": 0.7035982175377999,
+ "learning_rate": 6.981603689392999e-06,
+ "loss": 0.3729727745056152,
+ "num_input_tokens_seen": 481433184,
+ "step": 11000,
+ "train_runtime": 85091.999,
+ "train_tokens_per_second": 5657.796
+ },
+ {
+ "epoch": 0.7409776461082682,
+ "grad_norm": 0.7063885505049592,
+ "learning_rate": 6.979175848547263e-06,
+ "loss": 0.3587386131286621,
+ "num_input_tokens_seen": 481648384,
+ "step": 11005,
+ "train_runtime": 85133.4023,
+ "train_tokens_per_second": 5657.572
+ },
+ {
+ "epoch": 0.7413143011042284,
+ "grad_norm": 0.818479059925005,
+ "learning_rate": 6.976747454231316e-06,
+ "loss": 0.36256041526794436,
+ "num_input_tokens_seen": 481853600,
+ "step": 11010,
+ "train_runtime": 85175.3796,
+ "train_tokens_per_second": 5657.193
+ },
+ {
+ "epoch": 0.7416509561001885,
+ "grad_norm": 0.7213537892974411,
+ "learning_rate": 6.974318507124245e-06,
+ "loss": 0.37796127796173096,
+ "num_input_tokens_seen": 482067848,
+ "step": 11015,
+ "train_runtime": 85211.3126,
+ "train_tokens_per_second": 5657.322
+ },
+ {
+ "epoch": 0.7419876110961486,
+ "grad_norm": 0.9553094338809663,
+ "learning_rate": 6.971889007905299e-06,
+ "loss": 0.38568458557128904,
+ "num_input_tokens_seen": 482272608,
+ "step": 11020,
+ "train_runtime": 85242.7243,
+ "train_tokens_per_second": 5657.64
+ },
+ {
+ "epoch": 0.7423242660921088,
+ "grad_norm": 0.6954588658200586,
+ "learning_rate": 6.9694589572538795e-06,
+ "loss": 0.4131626605987549,
+ "num_input_tokens_seen": 482500568,
+ "step": 11025,
+ "train_runtime": 85279.0869,
+ "train_tokens_per_second": 5657.9
+ },
+ {
+ "epoch": 0.7426609210880689,
+ "grad_norm": 0.6724190701417789,
+ "learning_rate": 6.9670283558495436e-06,
+ "loss": 0.3677205085754395,
+ "num_input_tokens_seen": 482727520,
+ "step": 11030,
+ "train_runtime": 85315.8408,
+ "train_tokens_per_second": 5658.123
+ },
+ {
+ "epoch": 0.742997576084029,
+ "grad_norm": 0.7736013776309341,
+ "learning_rate": 6.964597204371999e-06,
+ "loss": 0.3734050035476685,
+ "num_input_tokens_seen": 482943600,
+ "step": 11035,
+ "train_runtime": 85349.9838,
+ "train_tokens_per_second": 5658.391
+ },
+ {
+ "epoch": 0.7433342310799892,
+ "grad_norm": 0.7557593534274823,
+ "learning_rate": 6.962165503501111e-06,
+ "loss": 0.3419958591461182,
+ "num_input_tokens_seen": 483163040,
+ "step": 11040,
+ "train_runtime": 85392.7525,
+ "train_tokens_per_second": 5658.127
+ },
+ {
+ "epoch": 0.7436708860759493,
+ "grad_norm": 0.786614283633,
+ "learning_rate": 6.959733253916899e-06,
+ "loss": 0.39629015922546384,
+ "num_input_tokens_seen": 483387400,
+ "step": 11045,
+ "train_runtime": 85430.316,
+ "train_tokens_per_second": 5658.265
+ },
+ {
+ "epoch": 0.7440075410719095,
+ "grad_norm": 0.7118679551230915,
+ "learning_rate": 6.95730045629953e-06,
+ "loss": 0.3637987613677979,
+ "num_input_tokens_seen": 483586168,
+ "step": 11050,
+ "train_runtime": 85469.2598,
+ "train_tokens_per_second": 5658.013
+ },
+ {
+ "epoch": 0.7443441960678696,
+ "grad_norm": 0.775880823500199,
+ "learning_rate": 6.954867111329329e-06,
+ "loss": 0.3997496604919434,
+ "num_input_tokens_seen": 483812040,
+ "step": 11055,
+ "train_runtime": 85501.004,
+ "train_tokens_per_second": 5658.554
+ },
+ {
+ "epoch": 0.7446808510638298,
+ "grad_norm": 0.8104111912789507,
+ "learning_rate": 6.952433219686775e-06,
+ "loss": 0.36485040187835693,
+ "num_input_tokens_seen": 484033200,
+ "step": 11060,
+ "train_runtime": 85544.4427,
+ "train_tokens_per_second": 5658.266
+ },
+ {
+ "epoch": 0.7450175060597899,
+ "grad_norm": 0.745606714965393,
+ "learning_rate": 6.949998782052497e-06,
+ "loss": 0.38755149841308595,
+ "num_input_tokens_seen": 484251832,
+ "step": 11065,
+ "train_runtime": 85584.405,
+ "train_tokens_per_second": 5658.178
+ },
+ {
+ "epoch": 0.74535416105575,
+ "grad_norm": 0.7941311038883269,
+ "learning_rate": 6.947563799107277e-06,
+ "loss": 0.36263265609741213,
+ "num_input_tokens_seen": 484460952,
+ "step": 11070,
+ "train_runtime": 85617.994,
+ "train_tokens_per_second": 5658.401
+ },
+ {
+ "epoch": 0.7456908160517102,
+ "grad_norm": 0.795109174361431,
+ "learning_rate": 6.945128271532052e-06,
+ "loss": 0.37139124870300294,
+ "num_input_tokens_seen": 484696168,
+ "step": 11075,
+ "train_runtime": 85658.1904,
+ "train_tokens_per_second": 5658.492
+ },
+ {
+ "epoch": 0.7460274710476703,
+ "grad_norm": 0.7555629636652617,
+ "learning_rate": 6.942692200007908e-06,
+ "loss": 0.37421250343322754,
+ "num_input_tokens_seen": 484914304,
+ "step": 11080,
+ "train_runtime": 85693.7663,
+ "train_tokens_per_second": 5658.688
+ },
+ {
+ "epoch": 0.7463641260436304,
+ "grad_norm": 0.7710680161258006,
+ "learning_rate": 6.9402555852160845e-06,
+ "loss": 0.3868663787841797,
+ "num_input_tokens_seen": 485140952,
+ "step": 11085,
+ "train_runtime": 85733.9493,
+ "train_tokens_per_second": 5658.68
+ },
+ {
+ "epoch": 0.7467007810395906,
+ "grad_norm": 0.6618662744692935,
+ "learning_rate": 6.937818427837975e-06,
+ "loss": 0.35773367881774903,
+ "num_input_tokens_seen": 485364088,
+ "step": 11090,
+ "train_runtime": 85775.7043,
+ "train_tokens_per_second": 5658.526
+ },
+ {
+ "epoch": 0.7470374360355507,
+ "grad_norm": 0.8138674839150098,
+ "learning_rate": 6.93538072855512e-06,
+ "loss": 0.40668773651123047,
+ "num_input_tokens_seen": 485585456,
+ "step": 11095,
+ "train_runtime": 85811.2509,
+ "train_tokens_per_second": 5658.762
+ },
+ {
+ "epoch": 0.7473740910315109,
+ "grad_norm": 0.7237738810724516,
+ "learning_rate": 6.932942488049217e-06,
+ "loss": 0.3412586212158203,
+ "num_input_tokens_seen": 485799760,
+ "step": 11100,
+ "train_runtime": 85848.9089,
+ "train_tokens_per_second": 5658.776
+ },
+ {
+ "epoch": 0.747710746027471,
+ "grad_norm": 0.7192380582338419,
+ "learning_rate": 6.930503707002108e-06,
+ "loss": 0.39055523872375486,
+ "num_input_tokens_seen": 486017000,
+ "step": 11105,
+ "train_runtime": 85893.6142,
+ "train_tokens_per_second": 5658.36
+ },
+ {
+ "epoch": 0.7480474010234311,
+ "grad_norm": 0.7129551106844463,
+ "learning_rate": 6.928064386095796e-06,
+ "loss": 0.3696824789047241,
+ "num_input_tokens_seen": 486243416,
+ "step": 11110,
+ "train_runtime": 85934.5192,
+ "train_tokens_per_second": 5658.301
+ },
+ {
+ "epoch": 0.7483840560193913,
+ "grad_norm": 0.8296106328518675,
+ "learning_rate": 6.9256245260124264e-06,
+ "loss": 0.369446325302124,
+ "num_input_tokens_seen": 486469728,
+ "step": 11115,
+ "train_runtime": 85966.8967,
+ "train_tokens_per_second": 5658.803
+ },
+ {
+ "epoch": 0.7487207110153514,
+ "grad_norm": 0.8112267099332857,
+ "learning_rate": 6.923184127434298e-06,
+ "loss": 0.3864800453186035,
+ "num_input_tokens_seen": 486670960,
+ "step": 11120,
+ "train_runtime": 86001.8182,
+ "train_tokens_per_second": 5658.845
+ },
+ {
+ "epoch": 0.7490573660113116,
+ "grad_norm": 0.8933045922508607,
+ "learning_rate": 6.920743191043861e-06,
+ "loss": 0.35923840999603274,
+ "num_input_tokens_seen": 486881136,
+ "step": 11125,
+ "train_runtime": 86039.0507,
+ "train_tokens_per_second": 5658.839
+ },
+ {
+ "epoch": 0.7493940210072717,
+ "grad_norm": 0.6534193785965624,
+ "learning_rate": 6.918301717523714e-06,
+ "loss": 0.3482653617858887,
+ "num_input_tokens_seen": 487095072,
+ "step": 11130,
+ "train_runtime": 86077.5734,
+ "train_tokens_per_second": 5658.792
+ },
+ {
+ "epoch": 0.7497306760032318,
+ "grad_norm": 0.7160660990648459,
+ "learning_rate": 6.915859707556611e-06,
+ "loss": 0.3809635639190674,
+ "num_input_tokens_seen": 487312640,
+ "step": 11135,
+ "train_runtime": 86112.0956,
+ "train_tokens_per_second": 5659.05
+ },
+ {
+ "epoch": 0.750067330999192,
+ "grad_norm": 0.7880329367953478,
+ "learning_rate": 6.913417161825449e-06,
+ "loss": 0.40050134658813474,
+ "num_input_tokens_seen": 487525824,
+ "step": 11140,
+ "train_runtime": 86145.5605,
+ "train_tokens_per_second": 5659.326
+ },
+ {
+ "epoch": 0.7504039859951521,
+ "grad_norm": 0.7445156362898442,
+ "learning_rate": 6.91097408101328e-06,
+ "loss": 0.3789609432220459,
+ "num_input_tokens_seen": 487742328,
+ "step": 11145,
+ "train_runtime": 86186.4755,
+ "train_tokens_per_second": 5659.152
+ },
+ {
+ "epoch": 0.7507406409911123,
+ "grad_norm": 0.8940165351109152,
+ "learning_rate": 6.908530465803302e-06,
+ "loss": 0.370347785949707,
+ "num_input_tokens_seen": 487949632,
+ "step": 11150,
+ "train_runtime": 86217.4676,
+ "train_tokens_per_second": 5659.522
+ },
+ {
+ "epoch": 0.7510772959870724,
+ "grad_norm": 0.8568185419619945,
+ "learning_rate": 6.906086316878865e-06,
+ "loss": 0.3664633989334106,
+ "num_input_tokens_seen": 488152136,
+ "step": 11155,
+ "train_runtime": 86258.4672,
+ "train_tokens_per_second": 5659.179
+ },
+ {
+ "epoch": 0.7514139509830325,
+ "grad_norm": 0.6561053217546615,
+ "learning_rate": 6.903641634923468e-06,
+ "loss": 0.3477164268493652,
+ "num_input_tokens_seen": 488367048,
+ "step": 11160,
+ "train_runtime": 86297.7659,
+ "train_tokens_per_second": 5659.093
+ },
+ {
+ "epoch": 0.7517506059789927,
+ "grad_norm": 0.7394466907995108,
+ "learning_rate": 6.9011964206207585e-06,
+ "loss": 0.3994622230529785,
+ "num_input_tokens_seen": 488567560,
+ "step": 11165,
+ "train_runtime": 86330.4394,
+ "train_tokens_per_second": 5659.273
+ },
+ {
+ "epoch": 0.7520872609749528,
+ "grad_norm": 0.7890852129954726,
+ "learning_rate": 6.898750674654531e-06,
+ "loss": 0.36271369457244873,
+ "num_input_tokens_seen": 488788328,
+ "step": 11170,
+ "train_runtime": 86369.9487,
+ "train_tokens_per_second": 5659.241
+ },
+ {
+ "epoch": 0.752423915970913,
+ "grad_norm": 0.6327317994824692,
+ "learning_rate": 6.896304397708733e-06,
+ "loss": 0.35087008476257325,
+ "num_input_tokens_seen": 488995192,
+ "step": 11175,
+ "train_runtime": 86408.8053,
+ "train_tokens_per_second": 5659.09
+ },
+ {
+ "epoch": 0.7527605709668731,
+ "grad_norm": 0.7895910131617881,
+ "learning_rate": 6.893857590467456e-06,
+ "loss": 0.3524162292480469,
+ "num_input_tokens_seen": 489213016,
+ "step": 11180,
+ "train_runtime": 86445.1307,
+ "train_tokens_per_second": 5659.232
+ },
+ {
+ "epoch": 0.7530972259628332,
+ "grad_norm": 0.7380392223605537,
+ "learning_rate": 6.891410253614944e-06,
+ "loss": 0.34860227108001707,
+ "num_input_tokens_seen": 489418272,
+ "step": 11185,
+ "train_runtime": 86484.3079,
+ "train_tokens_per_second": 5659.041
+ },
+ {
+ "epoch": 0.7534338809587934,
+ "grad_norm": 0.702987221857376,
+ "learning_rate": 6.888962387835586e-06,
+ "loss": 0.384684419631958,
+ "num_input_tokens_seen": 489652568,
+ "step": 11190,
+ "train_runtime": 86517.2831,
+ "train_tokens_per_second": 5659.593
+ },
+ {
+ "epoch": 0.7537705359547535,
+ "grad_norm": 0.7485124246366612,
+ "learning_rate": 6.886513993813919e-06,
+ "loss": 0.38306422233581544,
+ "num_input_tokens_seen": 489864096,
+ "step": 11195,
+ "train_runtime": 86557.4583,
+ "train_tokens_per_second": 5659.409
+ },
+ {
+ "epoch": 0.7541071909507137,
+ "grad_norm": 0.7067279867145273,
+ "learning_rate": 6.88406507223463e-06,
+ "loss": 0.36569690704345703,
+ "num_input_tokens_seen": 490092880,
+ "step": 11200,
+ "train_runtime": 86594.5272,
+ "train_tokens_per_second": 5659.629
+ },
+ {
+ "epoch": 0.7544438459466738,
+ "grad_norm": 0.7595465228382173,
+ "learning_rate": 6.881615623782551e-06,
+ "loss": 0.35335478782653806,
+ "num_input_tokens_seen": 490301440,
+ "step": 11205,
+ "train_runtime": 86631.5997,
+ "train_tokens_per_second": 5659.614
+ },
+ {
+ "epoch": 0.7547805009426339,
+ "grad_norm": 0.7480898301661179,
+ "learning_rate": 6.879165649142663e-06,
+ "loss": 0.3659096956253052,
+ "num_input_tokens_seen": 490500792,
+ "step": 11210,
+ "train_runtime": 86669.1193,
+ "train_tokens_per_second": 5659.464
+ },
+ {
+ "epoch": 0.7551171559385941,
+ "grad_norm": 0.6467936248180215,
+ "learning_rate": 6.876715149000093e-06,
+ "loss": 0.36233794689178467,
+ "num_input_tokens_seen": 490705136,
+ "step": 11215,
+ "train_runtime": 86708.2047,
+ "train_tokens_per_second": 5659.27
+ },
+ {
+ "epoch": 0.7554538109345542,
+ "grad_norm": 0.7242780017786447,
+ "learning_rate": 6.874264124040117e-06,
+ "loss": 0.3791798114776611,
+ "num_input_tokens_seen": 490930904,
+ "step": 11220,
+ "train_runtime": 86744.2352,
+ "train_tokens_per_second": 5659.522
+ },
+ {
+ "epoch": 0.7557904659305144,
+ "grad_norm": 0.7534511874997005,
+ "learning_rate": 6.871812574948154e-06,
+ "loss": 0.394960880279541,
+ "num_input_tokens_seen": 491158544,
+ "step": 11225,
+ "train_runtime": 86777.1348,
+ "train_tokens_per_second": 5660.0
+ },
+ {
+ "epoch": 0.7561271209264745,
+ "grad_norm": 0.7151778561469025,
+ "learning_rate": 6.869360502409773e-06,
+ "loss": 0.3845766544342041,
+ "num_input_tokens_seen": 491385232,
+ "step": 11230,
+ "train_runtime": 86814.8579,
+ "train_tokens_per_second": 5660.151
+ },
+ {
+ "epoch": 0.7564637759224346,
+ "grad_norm": 0.7179238868789197,
+ "learning_rate": 6.866907907110689e-06,
+ "loss": 0.351468563079834,
+ "num_input_tokens_seen": 491593976,
+ "step": 11235,
+ "train_runtime": 86852.7406,
+ "train_tokens_per_second": 5660.086
+ },
+ {
+ "epoch": 0.7568004309183949,
+ "grad_norm": 0.6534240517661798,
+ "learning_rate": 6.864454789736763e-06,
+ "loss": 0.3472340822219849,
+ "num_input_tokens_seen": 491827504,
+ "step": 11240,
+ "train_runtime": 86890.0801,
+ "train_tokens_per_second": 5660.341
+ },
+ {
+ "epoch": 0.757137085914355,
+ "grad_norm": 0.6984388211704616,
+ "learning_rate": 6.8620011509740005e-06,
+ "loss": 0.3707188844680786,
+ "num_input_tokens_seen": 492042264,
+ "step": 11245,
+ "train_runtime": 86929.2774,
+ "train_tokens_per_second": 5660.259
+ },
+ {
+ "epoch": 0.7574737409103152,
+ "grad_norm": 0.8438964758084152,
+ "learning_rate": 6.859546991508552e-06,
+ "loss": 0.3541130542755127,
+ "num_input_tokens_seen": 492261592,
+ "step": 11250,
+ "train_runtime": 86971.6043,
+ "train_tokens_per_second": 5660.027
+ },
+ {
+ "epoch": 0.7578103959062753,
+ "grad_norm": 0.6972841594933251,
+ "learning_rate": 6.857092312026716e-06,
+ "loss": 0.3816833972930908,
+ "num_input_tokens_seen": 492495240,
+ "step": 11255,
+ "train_runtime": 87006.7607,
+ "train_tokens_per_second": 5660.425
+ },
+ {
+ "epoch": 0.7581470509022354,
+ "grad_norm": 0.7161736954331712,
+ "learning_rate": 6.8546371132149395e-06,
+ "loss": 0.39476380348205564,
+ "num_input_tokens_seen": 492712328,
+ "step": 11260,
+ "train_runtime": 87046.3192,
+ "train_tokens_per_second": 5660.346
+ },
+ {
+ "epoch": 0.7584837058981956,
+ "grad_norm": 0.7348844948766989,
+ "learning_rate": 6.852181395759807e-06,
+ "loss": 0.39330875873565674,
+ "num_input_tokens_seen": 492943248,
+ "step": 11265,
+ "train_runtime": 87087.1631,
+ "train_tokens_per_second": 5660.343
+ },
+ {
+ "epoch": 0.7588203608941557,
+ "grad_norm": 0.5683273529508269,
+ "learning_rate": 6.849725160348054e-06,
+ "loss": 0.3455686330795288,
+ "num_input_tokens_seen": 493152344,
+ "step": 11270,
+ "train_runtime": 87127.3625,
+ "train_tokens_per_second": 5660.132
+ },
+ {
+ "epoch": 0.7591570158901159,
+ "grad_norm": 0.7258002012386351,
+ "learning_rate": 6.847268407666557e-06,
+ "loss": 0.35138664245605467,
+ "num_input_tokens_seen": 493375584,
+ "step": 11275,
+ "train_runtime": 87171.3314,
+ "train_tokens_per_second": 5659.838
+ },
+ {
+ "epoch": 0.759493670886076,
+ "grad_norm": 0.7379175392341268,
+ "learning_rate": 6.844811138402341e-06,
+ "loss": 0.37623722553253175,
+ "num_input_tokens_seen": 493594016,
+ "step": 11280,
+ "train_runtime": 87212.4908,
+ "train_tokens_per_second": 5659.671
+ },
+ {
+ "epoch": 0.7598303258820361,
+ "grad_norm": 0.7131462906095328,
+ "learning_rate": 6.842353353242574e-06,
+ "loss": 0.36553401947021485,
+ "num_input_tokens_seen": 493803080,
+ "step": 11285,
+ "train_runtime": 87254.8113,
+ "train_tokens_per_second": 5659.322
+ },
+ {
+ "epoch": 0.7601669808779963,
+ "grad_norm": 0.7404989121841616,
+ "learning_rate": 6.839895052874565e-06,
+ "loss": 0.3759953737258911,
+ "num_input_tokens_seen": 494027672,
+ "step": 11290,
+ "train_runtime": 87290.4517,
+ "train_tokens_per_second": 5659.584
+ },
+ {
+ "epoch": 0.7605036358739564,
+ "grad_norm": 0.6838203757609232,
+ "learning_rate": 6.837436237985772e-06,
+ "loss": 0.3641194343566895,
+ "num_input_tokens_seen": 494238208,
+ "step": 11295,
+ "train_runtime": 87333.9419,
+ "train_tokens_per_second": 5659.177
+ },
+ {
+ "epoch": 0.7608402908699166,
+ "grad_norm": 0.850634796737577,
+ "learning_rate": 6.834976909263794e-06,
+ "loss": 0.37904021739959715,
+ "num_input_tokens_seen": 494450104,
+ "step": 11300,
+ "train_runtime": 87372.8846,
+ "train_tokens_per_second": 5659.08
+ },
+ {
+ "epoch": 0.7611769458658767,
+ "grad_norm": 0.7504214621662669,
+ "learning_rate": 6.832517067396374e-06,
+ "loss": 0.3586588859558105,
+ "num_input_tokens_seen": 494683264,
+ "step": 11305,
+ "train_runtime": 87416.6251,
+ "train_tokens_per_second": 5658.915
+ },
+ {
+ "epoch": 0.7615136008618368,
+ "grad_norm": 0.7181780717405477,
+ "learning_rate": 6.8300567130714e-06,
+ "loss": 0.39732561111450193,
+ "num_input_tokens_seen": 494912080,
+ "step": 11310,
+ "train_runtime": 87455.7279,
+ "train_tokens_per_second": 5659.001
+ },
+ {
+ "epoch": 0.761850255857797,
+ "grad_norm": 0.8326466499790485,
+ "learning_rate": 6.8275958469769035e-06,
+ "loss": 0.38528354167938234,
+ "num_input_tokens_seen": 495143768,
+ "step": 11315,
+ "train_runtime": 87496.2946,
+ "train_tokens_per_second": 5659.026
+ },
+ {
+ "epoch": 0.7621869108537571,
+ "grad_norm": 0.8043440720773757,
+ "learning_rate": 6.825134469801055e-06,
+ "loss": 0.3580173969268799,
+ "num_input_tokens_seen": 495373776,
+ "step": 11320,
+ "train_runtime": 87536.7056,
+ "train_tokens_per_second": 5659.041
+ },
+ {
+ "epoch": 0.7625235658497173,
+ "grad_norm": 0.7612366109464449,
+ "learning_rate": 6.8226725822321715e-06,
+ "loss": 0.4001301765441895,
+ "num_input_tokens_seen": 495581360,
+ "step": 11325,
+ "train_runtime": 87574.6303,
+ "train_tokens_per_second": 5658.96
+ },
+ {
+ "epoch": 0.7628602208456774,
+ "grad_norm": 0.7254035816528333,
+ "learning_rate": 6.820210184958712e-06,
+ "loss": 0.38148252964019774,
+ "num_input_tokens_seen": 495805664,
+ "step": 11330,
+ "train_runtime": 87611.2812,
+ "train_tokens_per_second": 5659.153
+ },
+ {
+ "epoch": 0.7631968758416375,
+ "grad_norm": 0.7816370531432046,
+ "learning_rate": 6.81774727866928e-06,
+ "loss": 0.38228158950805663,
+ "num_input_tokens_seen": 496015888,
+ "step": 11335,
+ "train_runtime": 87647.9698,
+ "train_tokens_per_second": 5659.183
+ },
+ {
+ "epoch": 0.7635335308375977,
+ "grad_norm": 0.9719681651052307,
+ "learning_rate": 6.815283864052617e-06,
+ "loss": 0.3840935468673706,
+ "num_input_tokens_seen": 496237176,
+ "step": 11340,
+ "train_runtime": 87687.7469,
+ "train_tokens_per_second": 5659.139
+ },
+ {
+ "epoch": 0.7638701858335578,
+ "grad_norm": 0.7821628809832356,
+ "learning_rate": 6.812819941797611e-06,
+ "loss": 0.38556504249572754,
+ "num_input_tokens_seen": 496442520,
+ "step": 11345,
+ "train_runtime": 87730.1254,
+ "train_tokens_per_second": 5658.746
+ },
+ {
+ "epoch": 0.764206840829518,
+ "grad_norm": 0.7793739338464833,
+ "learning_rate": 6.810355512593288e-06,
+ "loss": 0.3861824035644531,
+ "num_input_tokens_seen": 496669864,
+ "step": 11350,
+ "train_runtime": 87767.0891,
+ "train_tokens_per_second": 5658.953
+ },
+ {
+ "epoch": 0.7645434958254781,
+ "grad_norm": 0.7769049083652588,
+ "learning_rate": 6.807890577128817e-06,
+ "loss": 0.33897886276245115,
+ "num_input_tokens_seen": 496892112,
+ "step": 11355,
+ "train_runtime": 87813.3112,
+ "train_tokens_per_second": 5658.506
+ },
+ {
+ "epoch": 0.7648801508214382,
+ "grad_norm": 0.8235786764795517,
+ "learning_rate": 6.805425136093513e-06,
+ "loss": 0.3582730531692505,
+ "num_input_tokens_seen": 497111304,
+ "step": 11360,
+ "train_runtime": 87847.6951,
+ "train_tokens_per_second": 5658.786
+ },
+ {
+ "epoch": 0.7652168058173984,
+ "grad_norm": 0.825724303695909,
+ "learning_rate": 6.802959190176827e-06,
+ "loss": 0.38888885974884035,
+ "num_input_tokens_seen": 497318752,
+ "step": 11365,
+ "train_runtime": 87886.6126,
+ "train_tokens_per_second": 5658.641
+ },
+ {
+ "epoch": 0.7655534608133585,
+ "grad_norm": 0.7692237966789585,
+ "learning_rate": 6.800492740068352e-06,
+ "loss": 0.3655484676361084,
+ "num_input_tokens_seen": 497523256,
+ "step": 11370,
+ "train_runtime": 87923.25,
+ "train_tokens_per_second": 5658.609
+ },
+ {
+ "epoch": 0.7658901158093187,
+ "grad_norm": 0.8501602672533506,
+ "learning_rate": 6.798025786457825e-06,
+ "loss": 0.35955286026000977,
+ "num_input_tokens_seen": 497749392,
+ "step": 11375,
+ "train_runtime": 87963.9631,
+ "train_tokens_per_second": 5658.56
+ },
+ {
+ "epoch": 0.7662267708052788,
+ "grad_norm": 0.6654629734284263,
+ "learning_rate": 6.795558330035118e-06,
+ "loss": 0.3689566612243652,
+ "num_input_tokens_seen": 497978632,
+ "step": 11380,
+ "train_runtime": 88010.5357,
+ "train_tokens_per_second": 5658.171
+ },
+ {
+ "epoch": 0.7665634258012389,
+ "grad_norm": 0.571714653501898,
+ "learning_rate": 6.793090371490251e-06,
+ "loss": 0.37029447555541994,
+ "num_input_tokens_seen": 498211416,
+ "step": 11385,
+ "train_runtime": 88043.3102,
+ "train_tokens_per_second": 5658.708
+ },
+ {
+ "epoch": 0.7669000807971991,
+ "grad_norm": 0.6562894791866329,
+ "learning_rate": 6.790621911513378e-06,
+ "loss": 0.36082067489624026,
+ "num_input_tokens_seen": 498430376,
+ "step": 11390,
+ "train_runtime": 88076.9318,
+ "train_tokens_per_second": 5659.034
+ },
+ {
+ "epoch": 0.7672367357931592,
+ "grad_norm": 0.6830456569016362,
+ "learning_rate": 6.788152950794799e-06,
+ "loss": 0.3912408113479614,
+ "num_input_tokens_seen": 498646776,
+ "step": 11395,
+ "train_runtime": 88113.1004,
+ "train_tokens_per_second": 5659.167
+ },
+ {
+ "epoch": 0.7675733907891193,
+ "grad_norm": 0.8063375531181857,
+ "learning_rate": 6.7856834900249495e-06,
+ "loss": 0.3919992208480835,
+ "num_input_tokens_seen": 498865144,
+ "step": 11400,
+ "train_runtime": 88157.9923,
+ "train_tokens_per_second": 5658.763
+ },
+ {
+ "epoch": 0.7679100457850795,
+ "grad_norm": 0.7488462821998634,
+ "learning_rate": 6.783213529894405e-06,
+ "loss": 0.3473628044128418,
+ "num_input_tokens_seen": 499082328,
+ "step": 11405,
+ "train_runtime": 88202.6214,
+ "train_tokens_per_second": 5658.362
+ },
+ {
+ "epoch": 0.7682467007810396,
+ "grad_norm": 0.7845757383023176,
+ "learning_rate": 6.780743071093883e-06,
+ "loss": 0.35947589874267577,
+ "num_input_tokens_seen": 499306216,
+ "step": 11410,
+ "train_runtime": 88242.1656,
+ "train_tokens_per_second": 5658.363
+ },
+ {
+ "epoch": 0.7685833557769998,
+ "grad_norm": 1.4808480239457869,
+ "learning_rate": 6.77827211431424e-06,
+ "loss": 0.36936905384063723,
+ "num_input_tokens_seen": 499532072,
+ "step": 11415,
+ "train_runtime": 88280.0767,
+ "train_tokens_per_second": 5658.492
+ },
+ {
+ "epoch": 0.7689200107729599,
+ "grad_norm": 0.7193580497484903,
+ "learning_rate": 6.775800660246471e-06,
+ "loss": 0.36505095958709716,
+ "num_input_tokens_seen": 499732272,
+ "step": 11420,
+ "train_runtime": 88316.7765,
+ "train_tokens_per_second": 5658.407
+ },
+ {
+ "epoch": 0.76925666576892,
+ "grad_norm": 0.7225028793880158,
+ "learning_rate": 6.77332870958171e-06,
+ "loss": 0.3719770908355713,
+ "num_input_tokens_seen": 499963848,
+ "step": 11425,
+ "train_runtime": 88361.2969,
+ "train_tokens_per_second": 5658.177
+ },
+ {
+ "epoch": 0.7695933207648802,
+ "grad_norm": 0.7310175751597672,
+ "learning_rate": 6.770856263011229e-06,
+ "loss": 0.3673600196838379,
+ "num_input_tokens_seen": 500175216,
+ "step": 11430,
+ "train_runtime": 88394.2396,
+ "train_tokens_per_second": 5658.459
+ },
+ {
+ "epoch": 0.7699299757608403,
+ "grad_norm": 0.8011531340232532,
+ "learning_rate": 6.768383321226442e-06,
+ "loss": 0.38931114673614503,
+ "num_input_tokens_seen": 500380304,
+ "step": 11435,
+ "train_runtime": 88432.6121,
+ "train_tokens_per_second": 5658.323
+ },
+ {
+ "epoch": 0.7702666307568005,
+ "grad_norm": 1.540406850238615,
+ "learning_rate": 6.765909884918898e-06,
+ "loss": 0.3583209753036499,
+ "num_input_tokens_seen": 500590256,
+ "step": 11440,
+ "train_runtime": 88469.3611,
+ "train_tokens_per_second": 5658.346
+ },
+ {
+ "epoch": 0.7706032857527606,
+ "grad_norm": 0.6645503979893219,
+ "learning_rate": 6.763435954780286e-06,
+ "loss": 0.3920794725418091,
+ "num_input_tokens_seen": 500810896,
+ "step": 11445,
+ "train_runtime": 88504.9842,
+ "train_tokens_per_second": 5658.561
+ },
+ {
+ "epoch": 0.7709399407487207,
+ "grad_norm": 0.7669134143397409,
+ "learning_rate": 6.7609615315024325e-06,
+ "loss": 0.37000212669372556,
+ "num_input_tokens_seen": 501027440,
+ "step": 11450,
+ "train_runtime": 88543.2243,
+ "train_tokens_per_second": 5658.563
+ },
+ {
+ "epoch": 0.7712765957446809,
+ "grad_norm": 0.7415266590401455,
+ "learning_rate": 6.758486615777301e-06,
+ "loss": 0.37416849136352537,
+ "num_input_tokens_seen": 501241576,
+ "step": 11455,
+ "train_runtime": 88583.8287,
+ "train_tokens_per_second": 5658.387
+ },
+ {
+ "epoch": 0.771613250740641,
+ "grad_norm": 0.8407214839390016,
+ "learning_rate": 6.756011208296995e-06,
+ "loss": 0.3626854419708252,
+ "num_input_tokens_seen": 501454696,
+ "step": 11460,
+ "train_runtime": 88621.4278,
+ "train_tokens_per_second": 5658.391
+ },
+ {
+ "epoch": 0.7719499057366012,
+ "grad_norm": 0.7461388799032196,
+ "learning_rate": 6.753535309753755e-06,
+ "loss": 0.35544500350952146,
+ "num_input_tokens_seen": 501658528,
+ "step": 11465,
+ "train_runtime": 88662.546,
+ "train_tokens_per_second": 5658.066
+ },
+ {
+ "epoch": 0.7722865607325613,
+ "grad_norm": 0.6919457018289741,
+ "learning_rate": 6.7510589208399565e-06,
+ "loss": 0.3747244119644165,
+ "num_input_tokens_seen": 501877728,
+ "step": 11470,
+ "train_runtime": 88703.6872,
+ "train_tokens_per_second": 5657.913
+ },
+ {
+ "epoch": 0.7726232157285214,
+ "grad_norm": 0.7585317880232398,
+ "learning_rate": 6.748582042248115e-06,
+ "loss": 0.35962686538696287,
+ "num_input_tokens_seen": 502089032,
+ "step": 11475,
+ "train_runtime": 88740.6334,
+ "train_tokens_per_second": 5657.938
+ },
+ {
+ "epoch": 0.7729598707244816,
+ "grad_norm": 0.7198688949375671,
+ "learning_rate": 6.746104674670882e-06,
+ "loss": 0.38360817432403566,
+ "num_input_tokens_seen": 502313408,
+ "step": 11480,
+ "train_runtime": 88777.3916,
+ "train_tokens_per_second": 5658.123
+ },
+ {
+ "epoch": 0.7732965257204417,
+ "grad_norm": 0.6976205058667109,
+ "learning_rate": 6.743626818801042e-06,
+ "loss": 0.39516823291778563,
+ "num_input_tokens_seen": 502534360,
+ "step": 11485,
+ "train_runtime": 88818.4462,
+ "train_tokens_per_second": 5657.995
+ },
+ {
+ "epoch": 0.7736331807164019,
+ "grad_norm": 0.6984390114365983,
+ "learning_rate": 6.741148475331525e-06,
+ "loss": 0.37048795223236086,
+ "num_input_tokens_seen": 502757200,
+ "step": 11490,
+ "train_runtime": 88854.2343,
+ "train_tokens_per_second": 5658.224
+ },
+ {
+ "epoch": 0.773969835712362,
+ "grad_norm": 0.6953225919926989,
+ "learning_rate": 6.738669644955388e-06,
+ "loss": 0.3468164443969727,
+ "num_input_tokens_seen": 502970848,
+ "step": 11495,
+ "train_runtime": 88897.1339,
+ "train_tokens_per_second": 5657.897
+ },
+ {
+ "epoch": 0.7743064907083221,
+ "grad_norm": 0.7656686554212365,
+ "learning_rate": 6.736190328365828e-06,
+ "loss": 0.3672001838684082,
+ "num_input_tokens_seen": 503195432,
+ "step": 11500,
+ "train_runtime": 88938.4411,
+ "train_tokens_per_second": 5657.795
+ },
+ {
+ "epoch": 0.7746431457042823,
+ "grad_norm": 0.743345653178439,
+ "learning_rate": 6.73371052625618e-06,
+ "loss": 0.37738971710205077,
+ "num_input_tokens_seen": 503417816,
+ "step": 11505,
+ "train_runtime": 88979.7242,
+ "train_tokens_per_second": 5657.669
+ },
+ {
+ "epoch": 0.7749798007002424,
+ "grad_norm": 0.7491341805650183,
+ "learning_rate": 6.731230239319912e-06,
+ "loss": 0.37387096881866455,
+ "num_input_tokens_seen": 503642040,
+ "step": 11510,
+ "train_runtime": 89021.4956,
+ "train_tokens_per_second": 5657.533
+ },
+ {
+ "epoch": 0.7753164556962026,
+ "grad_norm": 0.6256746258505268,
+ "learning_rate": 6.728749468250627e-06,
+ "loss": 0.38176569938659666,
+ "num_input_tokens_seen": 503870440,
+ "step": 11515,
+ "train_runtime": 89058.3538,
+ "train_tokens_per_second": 5657.756
+ },
+ {
+ "epoch": 0.7756531106921627,
+ "grad_norm": 0.8527701221713213,
+ "learning_rate": 6.726268213742066e-06,
+ "loss": 0.375903582572937,
+ "num_input_tokens_seen": 504076280,
+ "step": 11520,
+ "train_runtime": 89100.0272,
+ "train_tokens_per_second": 5657.42
+ },
+ {
+ "epoch": 0.7759897656881228,
+ "grad_norm": 0.7746670396544606,
+ "learning_rate": 6.723786476488105e-06,
+ "loss": 0.38536639213562013,
+ "num_input_tokens_seen": 504303624,
+ "step": 11525,
+ "train_runtime": 89134.096,
+ "train_tokens_per_second": 5657.808
+ },
+ {
+ "epoch": 0.776326420684083,
+ "grad_norm": 0.6906699278173198,
+ "learning_rate": 6.72130425718275e-06,
+ "loss": 0.3504000186920166,
+ "num_input_tokens_seen": 504511120,
+ "step": 11530,
+ "train_runtime": 89170.4494,
+ "train_tokens_per_second": 5657.829
+ },
+ {
+ "epoch": 0.7766630756800431,
+ "grad_norm": 0.7640134079354167,
+ "learning_rate": 6.718821556520151e-06,
+ "loss": 0.3902331829071045,
+ "num_input_tokens_seen": 504720744,
+ "step": 11535,
+ "train_runtime": 89203.783,
+ "train_tokens_per_second": 5658.064
+ },
+ {
+ "epoch": 0.7769997306760033,
+ "grad_norm": 0.676852860064436,
+ "learning_rate": 6.716338375194583e-06,
+ "loss": 0.3558895826339722,
+ "num_input_tokens_seen": 504945672,
+ "step": 11540,
+ "train_runtime": 89239.0086,
+ "train_tokens_per_second": 5658.351
+ },
+ {
+ "epoch": 0.7773363856719634,
+ "grad_norm": 0.7081316231079949,
+ "learning_rate": 6.713854713900463e-06,
+ "loss": 0.39481263160705565,
+ "num_input_tokens_seen": 505168112,
+ "step": 11545,
+ "train_runtime": 89280.0653,
+ "train_tokens_per_second": 5658.241
+ },
+ {
+ "epoch": 0.7776730406679235,
+ "grad_norm": 0.7771142918437934,
+ "learning_rate": 6.7113705733323365e-06,
+ "loss": 0.3915176153182983,
+ "num_input_tokens_seen": 505373904,
+ "step": 11550,
+ "train_runtime": 89321.9305,
+ "train_tokens_per_second": 5657.893
+ },
+ {
+ "epoch": 0.7780096956638837,
+ "grad_norm": 0.7862100764061776,
+ "learning_rate": 6.708885954184885e-06,
+ "loss": 0.36997599601745607,
+ "num_input_tokens_seen": 505592960,
+ "step": 11555,
+ "train_runtime": 89362.1395,
+ "train_tokens_per_second": 5657.798
+ },
+ {
+ "epoch": 0.7783463506598438,
+ "grad_norm": 0.8044048640519208,
+ "learning_rate": 6.7064008571529275e-06,
+ "loss": 0.3574796199798584,
+ "num_input_tokens_seen": 505815992,
+ "step": 11560,
+ "train_runtime": 89407.5364,
+ "train_tokens_per_second": 5657.42
+ },
+ {
+ "epoch": 0.778683005655804,
+ "grad_norm": 0.63451019489799,
+ "learning_rate": 6.703915282931411e-06,
+ "loss": 0.37357444763183595,
+ "num_input_tokens_seen": 506041792,
+ "step": 11565,
+ "train_runtime": 89452.1107,
+ "train_tokens_per_second": 5657.125
+ },
+ {
+ "epoch": 0.7790196606517641,
+ "grad_norm": 0.7400296332193441,
+ "learning_rate": 6.701429232215418e-06,
+ "loss": 0.3822374105453491,
+ "num_input_tokens_seen": 506244224,
+ "step": 11570,
+ "train_runtime": 89491.3476,
+ "train_tokens_per_second": 5656.907
+ },
+ {
+ "epoch": 0.7793563156477242,
+ "grad_norm": 0.8414435559233742,
+ "learning_rate": 6.6989427057001665e-06,
+ "loss": 0.3804741144180298,
+ "num_input_tokens_seen": 506470096,
+ "step": 11575,
+ "train_runtime": 89532.7906,
+ "train_tokens_per_second": 5656.811
+ },
+ {
+ "epoch": 0.7796929706436844,
+ "grad_norm": 0.6787888680159854,
+ "learning_rate": 6.696455704081006e-06,
+ "loss": 0.3787699699401855,
+ "num_input_tokens_seen": 506688376,
+ "step": 11580,
+ "train_runtime": 89574.3736,
+ "train_tokens_per_second": 5656.622
+ },
+ {
+ "epoch": 0.7800296256396445,
+ "grad_norm": 0.711221275891421,
+ "learning_rate": 6.693968228053416e-06,
+ "loss": 0.38877334594726565,
+ "num_input_tokens_seen": 506908432,
+ "step": 11585,
+ "train_runtime": 89609.8238,
+ "train_tokens_per_second": 5656.84
+ },
+ {
+ "epoch": 0.7803662806356046,
+ "grad_norm": 0.7045284132631514,
+ "learning_rate": 6.691480278313013e-06,
+ "loss": 0.37210679054260254,
+ "num_input_tokens_seen": 507123272,
+ "step": 11590,
+ "train_runtime": 89650.9072,
+ "train_tokens_per_second": 5656.644
+ },
+ {
+ "epoch": 0.7807029356315648,
+ "grad_norm": 0.718080559096589,
+ "learning_rate": 6.688991855555545e-06,
+ "loss": 0.35350847244262695,
+ "num_input_tokens_seen": 507339456,
+ "step": 11595,
+ "train_runtime": 89694.1289,
+ "train_tokens_per_second": 5656.328
+ },
+ {
+ "epoch": 0.7810395906275249,
+ "grad_norm": 0.692068019249801,
+ "learning_rate": 6.686502960476888e-06,
+ "loss": 0.3558246612548828,
+ "num_input_tokens_seen": 507577136,
+ "step": 11600,
+ "train_runtime": 89731.2991,
+ "train_tokens_per_second": 5656.634
+ },
+ {
+ "epoch": 0.7813762456234851,
+ "grad_norm": 0.7053421542399363,
+ "learning_rate": 6.684013593773058e-06,
+ "loss": 0.3307861089706421,
+ "num_input_tokens_seen": 507811288,
+ "step": 11605,
+ "train_runtime": 89766.4366,
+ "train_tokens_per_second": 5657.028
+ },
+ {
+ "epoch": 0.7817129006194452,
+ "grad_norm": 0.7619528494233906,
+ "learning_rate": 6.6815237561401955e-06,
+ "loss": 0.33051750659942625,
+ "num_input_tokens_seen": 508006304,
+ "step": 11610,
+ "train_runtime": 89802.2646,
+ "train_tokens_per_second": 5656.943
+ },
+ {
+ "epoch": 0.7820495556154053,
+ "grad_norm": 0.8056101436691785,
+ "learning_rate": 6.679033448274577e-06,
+ "loss": 0.35503997802734377,
+ "num_input_tokens_seen": 508223832,
+ "step": 11615,
+ "train_runtime": 89850.1477,
+ "train_tokens_per_second": 5656.349
+ },
+ {
+ "epoch": 0.7823862106113655,
+ "grad_norm": 0.6611022339527542,
+ "learning_rate": 6.676542670872609e-06,
+ "loss": 0.3701477527618408,
+ "num_input_tokens_seen": 508440024,
+ "step": 11620,
+ "train_runtime": 89890.5249,
+ "train_tokens_per_second": 5656.214
+ },
+ {
+ "epoch": 0.7827228656073256,
+ "grad_norm": 0.6851834640607047,
+ "learning_rate": 6.674051424630829e-06,
+ "loss": 0.36185245513916015,
+ "num_input_tokens_seen": 508666440,
+ "step": 11625,
+ "train_runtime": 89924.4436,
+ "train_tokens_per_second": 5656.598
+ },
+ {
+ "epoch": 0.7830595206032858,
+ "grad_norm": 0.7081381172789725,
+ "learning_rate": 6.671559710245905e-06,
+ "loss": 0.36898825168609617,
+ "num_input_tokens_seen": 508892304,
+ "step": 11630,
+ "train_runtime": 89970.0326,
+ "train_tokens_per_second": 5656.242
+ },
+ {
+ "epoch": 0.7833961755992459,
+ "grad_norm": 0.7864088827874502,
+ "learning_rate": 6.669067528414639e-06,
+ "loss": 0.36539783477783205,
+ "num_input_tokens_seen": 509118040,
+ "step": 11635,
+ "train_runtime": 90012.286,
+ "train_tokens_per_second": 5656.095
+ },
+ {
+ "epoch": 0.783732830595206,
+ "grad_norm": 0.7072947789302337,
+ "learning_rate": 6.6665748798339606e-06,
+ "loss": 0.39262051582336427,
+ "num_input_tokens_seen": 509346248,
+ "step": 11640,
+ "train_runtime": 90050.1585,
+ "train_tokens_per_second": 5656.25
+ },
+ {
+ "epoch": 0.7840694855911662,
+ "grad_norm": 0.7696238239442139,
+ "learning_rate": 6.664081765200932e-06,
+ "loss": 0.38922972679138185,
+ "num_input_tokens_seen": 509547192,
+ "step": 11645,
+ "train_runtime": 90082.5361,
+ "train_tokens_per_second": 5656.448
+ },
+ {
+ "epoch": 0.7844061405871263,
+ "grad_norm": 0.9564255408577432,
+ "learning_rate": 6.661588185212745e-06,
+ "loss": 0.38661224842071534,
+ "num_input_tokens_seen": 509767456,
+ "step": 11650,
+ "train_runtime": 90124.0819,
+ "train_tokens_per_second": 5656.285
+ },
+ {
+ "epoch": 0.7847427955830865,
+ "grad_norm": 0.6920498506689123,
+ "learning_rate": 6.65909414056672e-06,
+ "loss": 0.37398757934570315,
+ "num_input_tokens_seen": 509978544,
+ "step": 11655,
+ "train_runtime": 90162.88,
+ "train_tokens_per_second": 5656.192
+ },
+ {
+ "epoch": 0.7850794505790466,
+ "grad_norm": 0.7983355414544516,
+ "learning_rate": 6.6565996319603075e-06,
+ "loss": 0.36649136543273925,
+ "num_input_tokens_seen": 510194488,
+ "step": 11660,
+ "train_runtime": 90205.1793,
+ "train_tokens_per_second": 5655.933
+ },
+ {
+ "epoch": 0.7854161055750067,
+ "grad_norm": 0.705582165882806,
+ "learning_rate": 6.6541046600910915e-06,
+ "loss": 0.35945792198181153,
+ "num_input_tokens_seen": 510408368,
+ "step": 11665,
+ "train_runtime": 90242.6397,
+ "train_tokens_per_second": 5655.956
+ },
+ {
+ "epoch": 0.7857527605709669,
+ "grad_norm": 0.7659229146292734,
+ "learning_rate": 6.6516092256567824e-06,
+ "loss": 0.36875619888305666,
+ "num_input_tokens_seen": 510635536,
+ "step": 11670,
+ "train_runtime": 90279.7373,
+ "train_tokens_per_second": 5656.148
+ },
+ {
+ "epoch": 0.786089415566927,
+ "grad_norm": 0.8578375495363305,
+ "learning_rate": 6.64911332935522e-06,
+ "loss": 0.39651241302490237,
+ "num_input_tokens_seen": 510860904,
+ "step": 11675,
+ "train_runtime": 90320.5501,
+ "train_tokens_per_second": 5656.087
+ },
+ {
+ "epoch": 0.7864260705628872,
+ "grad_norm": 0.7134409162325845,
+ "learning_rate": 6.646616971884373e-06,
+ "loss": 0.3647106885910034,
+ "num_input_tokens_seen": 511085760,
+ "step": 11680,
+ "train_runtime": 90359.7652,
+ "train_tokens_per_second": 5656.121
+ },
+ {
+ "epoch": 0.7867627255588473,
+ "grad_norm": 0.6706622363930574,
+ "learning_rate": 6.64412015394234e-06,
+ "loss": 0.35194456577301025,
+ "num_input_tokens_seen": 511285624,
+ "step": 11685,
+ "train_runtime": 90397.4925,
+ "train_tokens_per_second": 5655.971
+ },
+ {
+ "epoch": 0.7870993805548074,
+ "grad_norm": 0.8444783042140578,
+ "learning_rate": 6.641622876227349e-06,
+ "loss": 0.3812446594238281,
+ "num_input_tokens_seen": 511491128,
+ "step": 11690,
+ "train_runtime": 90439.6501,
+ "train_tokens_per_second": 5655.607
+ },
+ {
+ "epoch": 0.7874360355507676,
+ "grad_norm": 0.7494745665567035,
+ "learning_rate": 6.639125139437756e-06,
+ "loss": 0.3623605012893677,
+ "num_input_tokens_seen": 511707944,
+ "step": 11695,
+ "train_runtime": 90482.4407,
+ "train_tokens_per_second": 5655.329
+ },
+ {
+ "epoch": 0.7877726905467277,
+ "grad_norm": 0.7902023749862866,
+ "learning_rate": 6.6366269442720446e-06,
+ "loss": 0.36579647064208987,
+ "num_input_tokens_seen": 511918704,
+ "step": 11700,
+ "train_runtime": 90522.6671,
+ "train_tokens_per_second": 5655.144
+ },
+ {
+ "epoch": 0.7881093455426879,
+ "grad_norm": 0.7084908330150312,
+ "learning_rate": 6.634128291428827e-06,
+ "loss": 0.4204265594482422,
+ "num_input_tokens_seen": 512130360,
+ "step": 11705,
+ "train_runtime": 90561.121,
+ "train_tokens_per_second": 5655.08
+ },
+ {
+ "epoch": 0.788446000538648,
+ "grad_norm": 0.7463863432263244,
+ "learning_rate": 6.631629181606842e-06,
+ "loss": 0.37847514152526857,
+ "num_input_tokens_seen": 512353072,
+ "step": 11710,
+ "train_runtime": 90604.8421,
+ "train_tokens_per_second": 5654.809
+ },
+ {
+ "epoch": 0.7887826555346081,
+ "grad_norm": 0.7252447688219217,
+ "learning_rate": 6.629129615504961e-06,
+ "loss": 0.3778711795806885,
+ "num_input_tokens_seen": 512578040,
+ "step": 11715,
+ "train_runtime": 90644.5653,
+ "train_tokens_per_second": 5654.813
+ },
+ {
+ "epoch": 0.7891193105305683,
+ "grad_norm": 0.6877935716049668,
+ "learning_rate": 6.626629593822178e-06,
+ "loss": 0.349456787109375,
+ "num_input_tokens_seen": 512795568,
+ "step": 11720,
+ "train_runtime": 90683.8764,
+ "train_tokens_per_second": 5654.76
+ },
+ {
+ "epoch": 0.7894559655265284,
+ "grad_norm": 0.7159875503391986,
+ "learning_rate": 6.624129117257616e-06,
+ "loss": 0.3877539157867432,
+ "num_input_tokens_seen": 513020704,
+ "step": 11725,
+ "train_runtime": 90721.6257,
+ "train_tokens_per_second": 5654.889
+ },
+ {
+ "epoch": 0.7897926205224886,
+ "grad_norm": 0.8198647095115159,
+ "learning_rate": 6.621628186510524e-06,
+ "loss": 0.37906270027160643,
+ "num_input_tokens_seen": 513238728,
+ "step": 11730,
+ "train_runtime": 90763.6807,
+ "train_tokens_per_second": 5654.671
+ },
+ {
+ "epoch": 0.7901292755184487,
+ "grad_norm": 0.6710230144644209,
+ "learning_rate": 6.619126802280281e-06,
+ "loss": 0.35970292091369627,
+ "num_input_tokens_seen": 513454240,
+ "step": 11735,
+ "train_runtime": 90806.8266,
+ "train_tokens_per_second": 5654.357
+ },
+ {
+ "epoch": 0.7904659305144088,
+ "grad_norm": 0.6839883499284557,
+ "learning_rate": 6.616624965266392e-06,
+ "loss": 0.3722025632858276,
+ "num_input_tokens_seen": 513674120,
+ "step": 11740,
+ "train_runtime": 90845.0403,
+ "train_tokens_per_second": 5654.399
+ },
+ {
+ "epoch": 0.790802585510369,
+ "grad_norm": 0.7422766392719585,
+ "learning_rate": 6.614122676168486e-06,
+ "loss": 0.39596946239471437,
+ "num_input_tokens_seen": 513889520,
+ "step": 11745,
+ "train_runtime": 90889.19,
+ "train_tokens_per_second": 5654.022
+ },
+ {
+ "epoch": 0.7911392405063291,
+ "grad_norm": 0.8574462720949246,
+ "learning_rate": 6.611619935686321e-06,
+ "loss": 0.34925665855407717,
+ "num_input_tokens_seen": 514094568,
+ "step": 11750,
+ "train_runtime": 90923.7862,
+ "train_tokens_per_second": 5654.126
+ },
+ {
+ "epoch": 0.7914758955022893,
+ "grad_norm": 0.7396446955629874,
+ "learning_rate": 6.609116744519779e-06,
+ "loss": 0.3882757663726807,
+ "num_input_tokens_seen": 514319376,
+ "step": 11755,
+ "train_runtime": 90961.7449,
+ "train_tokens_per_second": 5654.238
+ },
+ {
+ "epoch": 0.7918125504982494,
+ "grad_norm": 0.8047711473168563,
+ "learning_rate": 6.60661310336887e-06,
+ "loss": 0.3384627103805542,
+ "num_input_tokens_seen": 514527584,
+ "step": 11760,
+ "train_runtime": 91008.8104,
+ "train_tokens_per_second": 5653.602
+ },
+ {
+ "epoch": 0.7921492054942095,
+ "grad_norm": 0.7981581690317775,
+ "learning_rate": 6.604109012933732e-06,
+ "loss": 0.37465848922729494,
+ "num_input_tokens_seen": 514729464,
+ "step": 11765,
+ "train_runtime": 91040.9107,
+ "train_tokens_per_second": 5653.826
+ },
+ {
+ "epoch": 0.7924858604901697,
+ "grad_norm": 0.7333032087882904,
+ "learning_rate": 6.601604473914622e-06,
+ "loss": 0.39129223823547366,
+ "num_input_tokens_seen": 514948712,
+ "step": 11770,
+ "train_runtime": 91080.7609,
+ "train_tokens_per_second": 5653.759
+ },
+ {
+ "epoch": 0.7928225154861298,
+ "grad_norm": 0.8203529810583842,
+ "learning_rate": 6.599099487011928e-06,
+ "loss": 0.3703568458557129,
+ "num_input_tokens_seen": 515174152,
+ "step": 11775,
+ "train_runtime": 91123.7113,
+ "train_tokens_per_second": 5653.569
+ },
+ {
+ "epoch": 0.79315917048209,
+ "grad_norm": 0.7692920757506985,
+ "learning_rate": 6.596594052926163e-06,
+ "loss": 0.37824716567993166,
+ "num_input_tokens_seen": 515398616,
+ "step": 11780,
+ "train_runtime": 91161.9049,
+ "train_tokens_per_second": 5653.662
+ },
+ {
+ "epoch": 0.7934958254780501,
+ "grad_norm": 0.6620821907912957,
+ "learning_rate": 6.5940881723579595e-06,
+ "loss": 0.36647372245788573,
+ "num_input_tokens_seen": 515618208,
+ "step": 11785,
+ "train_runtime": 91199.6136,
+ "train_tokens_per_second": 5653.732
+ },
+ {
+ "epoch": 0.7938324804740102,
+ "grad_norm": 0.7308558763801775,
+ "learning_rate": 6.591581846008085e-06,
+ "loss": 0.34111249446868896,
+ "num_input_tokens_seen": 515838000,
+ "step": 11790,
+ "train_runtime": 91242.8853,
+ "train_tokens_per_second": 5653.46
+ },
+ {
+ "epoch": 0.7941691354699704,
+ "grad_norm": 0.7510405382733509,
+ "learning_rate": 6.589075074577421e-06,
+ "loss": 0.3614863634109497,
+ "num_input_tokens_seen": 516058872,
+ "step": 11795,
+ "train_runtime": 91288.1701,
+ "train_tokens_per_second": 5653.075
+ },
+ {
+ "epoch": 0.7945057904659305,
+ "grad_norm": 0.682167345785408,
+ "learning_rate": 6.5865678587669805e-06,
+ "loss": 0.35341219902038573,
+ "num_input_tokens_seen": 516288480,
+ "step": 11800,
+ "train_runtime": 91324.6105,
+ "train_tokens_per_second": 5653.334
+ },
+ {
+ "epoch": 0.7948424454618906,
+ "grad_norm": 0.8727392399581251,
+ "learning_rate": 6.584060199277897e-06,
+ "loss": 0.3922135829925537,
+ "num_input_tokens_seen": 516500456,
+ "step": 11805,
+ "train_runtime": 91356.4687,
+ "train_tokens_per_second": 5653.682
+ },
+ {
+ "epoch": 0.7951791004578508,
+ "grad_norm": 0.7263526124399556,
+ "learning_rate": 6.581552096811428e-06,
+ "loss": 0.3736285209655762,
+ "num_input_tokens_seen": 516717184,
+ "step": 11810,
+ "train_runtime": 91391.2945,
+ "train_tokens_per_second": 5653.899
+ },
+ {
+ "epoch": 0.7955157554538109,
+ "grad_norm": 0.7483107004591556,
+ "learning_rate": 6.57904355206896e-06,
+ "loss": 0.3753064632415771,
+ "num_input_tokens_seen": 516937256,
+ "step": 11815,
+ "train_runtime": 91433.3951,
+ "train_tokens_per_second": 5653.703
+ },
+ {
+ "epoch": 0.7958524104497711,
+ "grad_norm": 0.741351883349472,
+ "learning_rate": 6.576534565751997e-06,
+ "loss": 0.35328757762908936,
+ "num_input_tokens_seen": 517155784,
+ "step": 11820,
+ "train_runtime": 91472.4934,
+ "train_tokens_per_second": 5653.675
+ },
+ {
+ "epoch": 0.7961890654457312,
+ "grad_norm": 0.7392430177231608,
+ "learning_rate": 6.574025138562172e-06,
+ "loss": 0.3843714475631714,
+ "num_input_tokens_seen": 517361632,
+ "step": 11825,
+ "train_runtime": 91507.0999,
+ "train_tokens_per_second": 5653.787
+ },
+ {
+ "epoch": 0.7965257204416913,
+ "grad_norm": 0.6186336035373811,
+ "learning_rate": 6.571515271201234e-06,
+ "loss": 0.3463775634765625,
+ "num_input_tokens_seen": 517585680,
+ "step": 11830,
+ "train_runtime": 91547.5139,
+ "train_tokens_per_second": 5653.738
+ },
+ {
+ "epoch": 0.7968623754376515,
+ "grad_norm": 0.6495706347541744,
+ "learning_rate": 6.569004964371062e-06,
+ "loss": 0.3861451864242554,
+ "num_input_tokens_seen": 517810808,
+ "step": 11835,
+ "train_runtime": 91580.6678,
+ "train_tokens_per_second": 5654.15
+ },
+ {
+ "epoch": 0.7971990304336116,
+ "grad_norm": 0.7143293886304013,
+ "learning_rate": 6.566494218773656e-06,
+ "loss": 0.3636291265487671,
+ "num_input_tokens_seen": 518036944,
+ "step": 11840,
+ "train_runtime": 91623.7035,
+ "train_tokens_per_second": 5653.962
+ },
+ {
+ "epoch": 0.7975356854295718,
+ "grad_norm": 0.7702610503844717,
+ "learning_rate": 6.563983035111136e-06,
+ "loss": 0.376003098487854,
+ "num_input_tokens_seen": 518266216,
+ "step": 11845,
+ "train_runtime": 91660.207,
+ "train_tokens_per_second": 5654.212
+ },
+ {
+ "epoch": 0.7978723404255319,
+ "grad_norm": 0.7600176188341217,
+ "learning_rate": 6.561471414085748e-06,
+ "loss": 0.36989893913269045,
+ "num_input_tokens_seen": 518477336,
+ "step": 11850,
+ "train_runtime": 91696.4011,
+ "train_tokens_per_second": 5654.282
+ },
+ {
+ "epoch": 0.798208995421492,
+ "grad_norm": 0.7119305326521156,
+ "learning_rate": 6.558959356399859e-06,
+ "loss": 0.3683455944061279,
+ "num_input_tokens_seen": 518691120,
+ "step": 11855,
+ "train_runtime": 91735.153,
+ "train_tokens_per_second": 5654.224
+ },
+ {
+ "epoch": 0.7985456504174522,
+ "grad_norm": 0.6724381741463443,
+ "learning_rate": 6.5564468627559544e-06,
+ "loss": 0.3579908609390259,
+ "num_input_tokens_seen": 518897696,
+ "step": 11860,
+ "train_runtime": 91775.8953,
+ "train_tokens_per_second": 5653.965
+ },
+ {
+ "epoch": 0.7988823054134123,
+ "grad_norm": 0.7387321689727897,
+ "learning_rate": 6.5539339338566495e-06,
+ "loss": 0.38613746166229246,
+ "num_input_tokens_seen": 519112712,
+ "step": 11865,
+ "train_runtime": 91816.066,
+ "train_tokens_per_second": 5653.833
+ },
+ {
+ "epoch": 0.7992189604093725,
+ "grad_norm": 0.8155036594396956,
+ "learning_rate": 6.5514205704046754e-06,
+ "loss": 0.34462246894836424,
+ "num_input_tokens_seen": 519326360,
+ "step": 11870,
+ "train_runtime": 91857.9272,
+ "train_tokens_per_second": 5653.582
+ },
+ {
+ "epoch": 0.7995556154053326,
+ "grad_norm": 0.7554873227873684,
+ "learning_rate": 6.548906773102886e-06,
+ "loss": 0.34732234477996826,
+ "num_input_tokens_seen": 519538040,
+ "step": 11875,
+ "train_runtime": 91892.7088,
+ "train_tokens_per_second": 5653.746
+ },
+ {
+ "epoch": 0.7998922704012927,
+ "grad_norm": 0.6865110717768457,
+ "learning_rate": 6.546392542654257e-06,
+ "loss": 0.3555797576904297,
+ "num_input_tokens_seen": 519754440,
+ "step": 11880,
+ "train_runtime": 91931.8443,
+ "train_tokens_per_second": 5653.693
+ },
+ {
+ "epoch": 0.8002289253972529,
+ "grad_norm": 0.7775168324740175,
+ "learning_rate": 6.543877879761882e-06,
+ "loss": 0.39305408000946046,
+ "num_input_tokens_seen": 519971256,
+ "step": 11885,
+ "train_runtime": 91972.3549,
+ "train_tokens_per_second": 5653.56
+ },
+ {
+ "epoch": 0.800565580393213,
+ "grad_norm": 0.6425158999557459,
+ "learning_rate": 6.5413627851289825e-06,
+ "loss": 0.3501169681549072,
+ "num_input_tokens_seen": 520197464,
+ "step": 11890,
+ "train_runtime": 92010.0574,
+ "train_tokens_per_second": 5653.702
+ },
+ {
+ "epoch": 0.8009022353891732,
+ "grad_norm": 0.7227536117889709,
+ "learning_rate": 6.538847259458896e-06,
+ "loss": 0.3551483631134033,
+ "num_input_tokens_seen": 520422616,
+ "step": 11895,
+ "train_runtime": 92042.9843,
+ "train_tokens_per_second": 5654.126
+ },
+ {
+ "epoch": 0.8012388903851333,
+ "grad_norm": 0.7042669271875746,
+ "learning_rate": 6.5363313034550805e-06,
+ "loss": 0.34134459495544434,
+ "num_input_tokens_seen": 520633392,
+ "step": 11900,
+ "train_runtime": 92087.6293,
+ "train_tokens_per_second": 5653.674
+ },
+ {
+ "epoch": 0.8015755453810934,
+ "grad_norm": 0.739311915732445,
+ "learning_rate": 6.533814917821114e-06,
+ "loss": 0.3754891395568848,
+ "num_input_tokens_seen": 520855784,
+ "step": 11905,
+ "train_runtime": 92127.8168,
+ "train_tokens_per_second": 5653.621
+ },
+ {
+ "epoch": 0.8019122003770536,
+ "grad_norm": 0.6302684231473084,
+ "learning_rate": 6.531298103260696e-06,
+ "loss": 0.3585964202880859,
+ "num_input_tokens_seen": 521066736,
+ "step": 11910,
+ "train_runtime": 92163.5443,
+ "train_tokens_per_second": 5653.719
+ },
+ {
+ "epoch": 0.8022488553730137,
+ "grad_norm": 0.7417671721607515,
+ "learning_rate": 6.528780860477648e-06,
+ "loss": 0.3797733783721924,
+ "num_input_tokens_seen": 521300416,
+ "step": 11915,
+ "train_runtime": 92200.3318,
+ "train_tokens_per_second": 5653.997
+ },
+ {
+ "epoch": 0.8025855103689739,
+ "grad_norm": 0.6890068438400968,
+ "learning_rate": 6.526263190175907e-06,
+ "loss": 0.3375946044921875,
+ "num_input_tokens_seen": 521532280,
+ "step": 11920,
+ "train_runtime": 92234.4047,
+ "train_tokens_per_second": 5654.422
+ },
+ {
+ "epoch": 0.802922165364934,
+ "grad_norm": 0.6823445253166819,
+ "learning_rate": 6.523745093059531e-06,
+ "loss": 0.3637096643447876,
+ "num_input_tokens_seen": 521751944,
+ "step": 11925,
+ "train_runtime": 92276.588,
+ "train_tokens_per_second": 5654.218
+ },
+ {
+ "epoch": 0.8032588203608941,
+ "grad_norm": 0.8198308621507191,
+ "learning_rate": 6.521226569832699e-06,
+ "loss": 0.3677360057830811,
+ "num_input_tokens_seen": 521974016,
+ "step": 11930,
+ "train_runtime": 92311.2607,
+ "train_tokens_per_second": 5654.5
+ },
+ {
+ "epoch": 0.8035954753568543,
+ "grad_norm": 0.7478623966170844,
+ "learning_rate": 6.518707621199707e-06,
+ "loss": 0.355226993560791,
+ "num_input_tokens_seen": 522190040,
+ "step": 11935,
+ "train_runtime": 92353.1434,
+ "train_tokens_per_second": 5654.275
+ },
+ {
+ "epoch": 0.8039321303528144,
+ "grad_norm": 0.7995961908090801,
+ "learning_rate": 6.516188247864972e-06,
+ "loss": 0.3752752304077148,
+ "num_input_tokens_seen": 522418072,
+ "step": 11940,
+ "train_runtime": 92392.9504,
+ "train_tokens_per_second": 5654.307
+ },
+ {
+ "epoch": 0.8042687853487746,
+ "grad_norm": 0.8084497458000778,
+ "learning_rate": 6.513668450533028e-06,
+ "loss": 0.3633857250213623,
+ "num_input_tokens_seen": 522645296,
+ "step": 11945,
+ "train_runtime": 92428.2209,
+ "train_tokens_per_second": 5654.607
+ },
+ {
+ "epoch": 0.8046054403447347,
+ "grad_norm": 0.7480700245413793,
+ "learning_rate": 6.511148229908531e-06,
+ "loss": 0.39172661304473877,
+ "num_input_tokens_seen": 522868752,
+ "step": 11950,
+ "train_runtime": 92461.7106,
+ "train_tokens_per_second": 5654.976
+ },
+ {
+ "epoch": 0.8049420953406948,
+ "grad_norm": 0.8708977640452219,
+ "learning_rate": 6.508627586696248e-06,
+ "loss": 0.3619298696517944,
+ "num_input_tokens_seen": 523082504,
+ "step": 11955,
+ "train_runtime": 92499.8318,
+ "train_tokens_per_second": 5654.956
+ },
+ {
+ "epoch": 0.805278750336655,
+ "grad_norm": 0.7261115677502934,
+ "learning_rate": 6.5061065216010695e-06,
+ "loss": 0.3667879581451416,
+ "num_input_tokens_seen": 523299872,
+ "step": 11960,
+ "train_runtime": 92542.5023,
+ "train_tokens_per_second": 5654.698
+ },
+ {
+ "epoch": 0.8056154053326151,
+ "grad_norm": 0.7422593678216647,
+ "learning_rate": 6.5035850353280075e-06,
+ "loss": 0.3360764503479004,
+ "num_input_tokens_seen": 523516992,
+ "step": 11965,
+ "train_runtime": 92581.276,
+ "train_tokens_per_second": 5654.675
+ },
+ {
+ "epoch": 0.8059520603285752,
+ "grad_norm": 0.7763633318074254,
+ "learning_rate": 6.501063128582183e-06,
+ "loss": 0.35983996391296386,
+ "num_input_tokens_seen": 523724536,
+ "step": 11970,
+ "train_runtime": 92615.6401,
+ "train_tokens_per_second": 5654.817
+ },
+ {
+ "epoch": 0.8062887153245354,
+ "grad_norm": 0.8599327424362851,
+ "learning_rate": 6.498540802068843e-06,
+ "loss": 0.36988580226898193,
+ "num_input_tokens_seen": 523941648,
+ "step": 11975,
+ "train_runtime": 92650.751,
+ "train_tokens_per_second": 5655.018
+ },
+ {
+ "epoch": 0.8066253703204955,
+ "grad_norm": 0.7313407988142705,
+ "learning_rate": 6.4960180564933454e-06,
+ "loss": 0.3595852851867676,
+ "num_input_tokens_seen": 524162560,
+ "step": 11980,
+ "train_runtime": 92691.1408,
+ "train_tokens_per_second": 5654.937
+ },
+ {
+ "epoch": 0.8069620253164557,
+ "grad_norm": 0.7521095206194225,
+ "learning_rate": 6.493494892561167e-06,
+ "loss": 0.3594093084335327,
+ "num_input_tokens_seen": 524374312,
+ "step": 11985,
+ "train_runtime": 92728.3173,
+ "train_tokens_per_second": 5654.953
+ },
+ {
+ "epoch": 0.8072986803124158,
+ "grad_norm": 0.8430159653893533,
+ "learning_rate": 6.490971310977907e-06,
+ "loss": 0.3831813812255859,
+ "num_input_tokens_seen": 524596088,
+ "step": 11990,
+ "train_runtime": 92765.8417,
+ "train_tokens_per_second": 5655.057
+ },
+ {
+ "epoch": 0.807635335308376,
+ "grad_norm": 0.7203196408716581,
+ "learning_rate": 6.488447312449273e-06,
+ "loss": 0.384428596496582,
+ "num_input_tokens_seen": 524821680,
+ "step": 11995,
+ "train_runtime": 92796.5099,
+ "train_tokens_per_second": 5655.619
+ },
+ {
+ "epoch": 0.8079719903043361,
+ "grad_norm": 0.7501646695656614,
+ "learning_rate": 6.485922897681095e-06,
+ "loss": 0.3725404739379883,
+ "num_input_tokens_seen": 525045696,
+ "step": 12000,
+ "train_runtime": 92832.3227,
+ "train_tokens_per_second": 5655.85
+ },
+ {
+ "epoch": 0.8083086453002962,
+ "grad_norm": 0.6490994037223768,
+ "learning_rate": 6.483398067379316e-06,
+ "loss": 0.4056462287902832,
+ "num_input_tokens_seen": 525279952,
+ "step": 12005,
+ "train_runtime": 92862.4479,
+ "train_tokens_per_second": 5656.538
+ },
+ {
+ "epoch": 0.8086453002962564,
+ "grad_norm": 0.719666275525647,
+ "learning_rate": 6.480872822249998e-06,
+ "loss": 0.3556082248687744,
+ "num_input_tokens_seen": 525506864,
+ "step": 12010,
+ "train_runtime": 92904.5673,
+ "train_tokens_per_second": 5656.416
+ },
+ {
+ "epoch": 0.8089819552922165,
+ "grad_norm": 0.764344635124093,
+ "learning_rate": 6.478347162999318e-06,
+ "loss": 0.3839968919754028,
+ "num_input_tokens_seen": 525719016,
+ "step": 12015,
+ "train_runtime": 92947.2954,
+ "train_tokens_per_second": 5656.098
+ },
+ {
+ "epoch": 0.8093186102881766,
+ "grad_norm": 0.6701860780093635,
+ "learning_rate": 6.475821090333568e-06,
+ "loss": 0.3664535999298096,
+ "num_input_tokens_seen": 525939808,
+ "step": 12020,
+ "train_runtime": 92994.8165,
+ "train_tokens_per_second": 5655.582
+ },
+ {
+ "epoch": 0.8096552652841368,
+ "grad_norm": 0.7604615691324128,
+ "learning_rate": 6.473294604959156e-06,
+ "loss": 0.38412063121795653,
+ "num_input_tokens_seen": 526150200,
+ "step": 12025,
+ "train_runtime": 93031.4012,
+ "train_tokens_per_second": 5655.619
+ },
+ {
+ "epoch": 0.8099919202800969,
+ "grad_norm": 0.7314264087625333,
+ "learning_rate": 6.470767707582605e-06,
+ "loss": 0.3502906322479248,
+ "num_input_tokens_seen": 526371776,
+ "step": 12030,
+ "train_runtime": 93073.7526,
+ "train_tokens_per_second": 5655.427
+ },
+ {
+ "epoch": 0.8103285752760571,
+ "grad_norm": 0.8031066646434316,
+ "learning_rate": 6.468240398910555e-06,
+ "loss": 0.36304702758789065,
+ "num_input_tokens_seen": 526566912,
+ "step": 12035,
+ "train_runtime": 93113.568,
+ "train_tokens_per_second": 5655.104
+ },
+ {
+ "epoch": 0.8106652302720172,
+ "grad_norm": 0.7056043206827206,
+ "learning_rate": 6.4657126796497595e-06,
+ "loss": 0.3545745849609375,
+ "num_input_tokens_seen": 526798856,
+ "step": 12040,
+ "train_runtime": 93156.2331,
+ "train_tokens_per_second": 5655.004
+ },
+ {
+ "epoch": 0.8110018852679773,
+ "grad_norm": 0.8107764954705456,
+ "learning_rate": 6.463184550507087e-06,
+ "loss": 0.37404961585998536,
+ "num_input_tokens_seen": 527012296,
+ "step": 12045,
+ "train_runtime": 93201.4479,
+ "train_tokens_per_second": 5654.551
+ },
+ {
+ "epoch": 0.8113385402639375,
+ "grad_norm": 0.6987203165103377,
+ "learning_rate": 6.46065601218952e-06,
+ "loss": 0.34106624126434326,
+ "num_input_tokens_seen": 527230672,
+ "step": 12050,
+ "train_runtime": 93238.6997,
+ "train_tokens_per_second": 5654.633
+ },
+ {
+ "epoch": 0.8116751952598976,
+ "grad_norm": 0.6797008656627719,
+ "learning_rate": 6.458127065404156e-06,
+ "loss": 0.3812570095062256,
+ "num_input_tokens_seen": 527453272,
+ "step": 12055,
+ "train_runtime": 93278.2031,
+ "train_tokens_per_second": 5654.625
+ },
+ {
+ "epoch": 0.8120118502558578,
+ "grad_norm": 0.6832021023137033,
+ "learning_rate": 6.45559771085821e-06,
+ "loss": 0.35934345722198485,
+ "num_input_tokens_seen": 527679752,
+ "step": 12060,
+ "train_runtime": 93316.0381,
+ "train_tokens_per_second": 5654.759
+ },
+ {
+ "epoch": 0.8123485052518179,
+ "grad_norm": 0.7280405860607188,
+ "learning_rate": 6.453067949259005e-06,
+ "loss": 0.35833373069763186,
+ "num_input_tokens_seen": 527905280,
+ "step": 12065,
+ "train_runtime": 93350.8689,
+ "train_tokens_per_second": 5655.066
+ },
+ {
+ "epoch": 0.812685160247778,
+ "grad_norm": 0.7042450928577643,
+ "learning_rate": 6.450537781313982e-06,
+ "loss": 0.33934841156005857,
+ "num_input_tokens_seen": 528118504,
+ "step": 12070,
+ "train_runtime": 93398.5502,
+ "train_tokens_per_second": 5654.461
+ },
+ {
+ "epoch": 0.8130218152437382,
+ "grad_norm": 0.8291985738023927,
+ "learning_rate": 6.448007207730693e-06,
+ "loss": 0.3756300926208496,
+ "num_input_tokens_seen": 528340504,
+ "step": 12075,
+ "train_runtime": 93436.5515,
+ "train_tokens_per_second": 5654.538
+ },
+ {
+ "epoch": 0.8133584702396983,
+ "grad_norm": 0.7204057314942438,
+ "learning_rate": 6.445476229216805e-06,
+ "loss": 0.35174586772918703,
+ "num_input_tokens_seen": 528575776,
+ "step": 12080,
+ "train_runtime": 93474.5152,
+ "train_tokens_per_second": 5654.758
+ },
+ {
+ "epoch": 0.8136951252356585,
+ "grad_norm": 0.7511444963498415,
+ "learning_rate": 6.442944846480099e-06,
+ "loss": 0.39254226684570315,
+ "num_input_tokens_seen": 528786904,
+ "step": 12085,
+ "train_runtime": 93515.4145,
+ "train_tokens_per_second": 5654.543
+ },
+ {
+ "epoch": 0.8140317802316186,
+ "grad_norm": 0.7816657648101096,
+ "learning_rate": 6.44041306022847e-06,
+ "loss": 0.39618897438049316,
+ "num_input_tokens_seen": 529015944,
+ "step": 12090,
+ "train_runtime": 93554.4561,
+ "train_tokens_per_second": 5654.631
+ },
+ {
+ "epoch": 0.8143684352275787,
+ "grad_norm": 0.8141588683065977,
+ "learning_rate": 6.43788087116992e-06,
+ "loss": 0.39859035015106203,
+ "num_input_tokens_seen": 529246224,
+ "step": 12095,
+ "train_runtime": 93595.1858,
+ "train_tokens_per_second": 5654.631
+ },
+ {
+ "epoch": 0.8147050902235389,
+ "grad_norm": 0.7932908932160116,
+ "learning_rate": 6.43534828001257e-06,
+ "loss": 0.3658564567565918,
+ "num_input_tokens_seen": 529476248,
+ "step": 12100,
+ "train_runtime": 93626.3021,
+ "train_tokens_per_second": 5655.208
+ },
+ {
+ "epoch": 0.815041745219499,
+ "grad_norm": 0.79019755327484,
+ "learning_rate": 6.43281528746465e-06,
+ "loss": 0.372475266456604,
+ "num_input_tokens_seen": 529683080,
+ "step": 12105,
+ "train_runtime": 93660.6895,
+ "train_tokens_per_second": 5655.34
+ },
+ {
+ "epoch": 0.8153784002154592,
+ "grad_norm": 0.6339454129210152,
+ "learning_rate": 6.430281894234504e-06,
+ "loss": 0.36925411224365234,
+ "num_input_tokens_seen": 529906632,
+ "step": 12110,
+ "train_runtime": 93709.4206,
+ "train_tokens_per_second": 5654.785
+ },
+ {
+ "epoch": 0.8157150552114193,
+ "grad_norm": 0.8085146313091093,
+ "learning_rate": 6.427748101030587e-06,
+ "loss": 0.3794065237045288,
+ "num_input_tokens_seen": 530122608,
+ "step": 12115,
+ "train_runtime": 93746.5407,
+ "train_tokens_per_second": 5654.85
+ },
+ {
+ "epoch": 0.8160517102073794,
+ "grad_norm": 0.7189030374735652,
+ "learning_rate": 6.425213908561468e-06,
+ "loss": 0.3644280433654785,
+ "num_input_tokens_seen": 530344760,
+ "step": 12120,
+ "train_runtime": 93782.6096,
+ "train_tokens_per_second": 5655.044
+ },
+ {
+ "epoch": 0.8163883652033396,
+ "grad_norm": 0.7379513095206941,
+ "learning_rate": 6.422679317535823e-06,
+ "loss": 0.3801795721054077,
+ "num_input_tokens_seen": 530554320,
+ "step": 12125,
+ "train_runtime": 93818.6237,
+ "train_tokens_per_second": 5655.107
+ },
+ {
+ "epoch": 0.8167250201992997,
+ "grad_norm": 0.7361384056056562,
+ "learning_rate": 6.420144328662441e-06,
+ "loss": 0.4110711097717285,
+ "num_input_tokens_seen": 530784216,
+ "step": 12130,
+ "train_runtime": 93858.9763,
+ "train_tokens_per_second": 5655.125
+ },
+ {
+ "epoch": 0.8170616751952599,
+ "grad_norm": 0.6984222353139848,
+ "learning_rate": 6.417608942650229e-06,
+ "loss": 0.36923315525054934,
+ "num_input_tokens_seen": 530985400,
+ "step": 12135,
+ "train_runtime": 93900.3001,
+ "train_tokens_per_second": 5654.779
+ },
+ {
+ "epoch": 0.81739833019122,
+ "grad_norm": 0.7696394382831785,
+ "learning_rate": 6.415073160208196e-06,
+ "loss": 0.368359637260437,
+ "num_input_tokens_seen": 531219992,
+ "step": 12140,
+ "train_runtime": 93938.5272,
+ "train_tokens_per_second": 5654.975
+ },
+ {
+ "epoch": 0.8177349851871801,
+ "grad_norm": 0.8778188370382649,
+ "learning_rate": 6.412536982045465e-06,
+ "loss": 0.38031327724456787,
+ "num_input_tokens_seen": 531416856,
+ "step": 12145,
+ "train_runtime": 93975.0195,
+ "train_tokens_per_second": 5654.874
+ },
+ {
+ "epoch": 0.8180716401831403,
+ "grad_norm": 0.7668958438378366,
+ "learning_rate": 6.410000408871273e-06,
+ "loss": 0.3821054458618164,
+ "num_input_tokens_seen": 531631960,
+ "step": 12150,
+ "train_runtime": 94010.159,
+ "train_tokens_per_second": 5655.048
+ },
+ {
+ "epoch": 0.8184082951791004,
+ "grad_norm": 0.7888864971284683,
+ "learning_rate": 6.407463441394961e-06,
+ "loss": 0.37601168155670167,
+ "num_input_tokens_seen": 531859288,
+ "step": 12155,
+ "train_runtime": 94046.1948,
+ "train_tokens_per_second": 5655.298
+ },
+ {
+ "epoch": 0.8187449501750605,
+ "grad_norm": 0.8191097049266588,
+ "learning_rate": 6.4049260803259874e-06,
+ "loss": 0.38929071426391604,
+ "num_input_tokens_seen": 532082152,
+ "step": 12160,
+ "train_runtime": 94088.2586,
+ "train_tokens_per_second": 5655.139
+ },
+ {
+ "epoch": 0.8190816051710207,
+ "grad_norm": 0.8221152177897472,
+ "learning_rate": 6.402388326373914e-06,
+ "loss": 0.3606018304824829,
+ "num_input_tokens_seen": 532304432,
+ "step": 12165,
+ "train_runtime": 94128.8133,
+ "train_tokens_per_second": 5655.064
+ },
+ {
+ "epoch": 0.8194182601669808,
+ "grad_norm": 0.6644138882185266,
+ "learning_rate": 6.399850180248418e-06,
+ "loss": 0.4054881572723389,
+ "num_input_tokens_seen": 532538000,
+ "step": 12170,
+ "train_runtime": 94167.2148,
+ "train_tokens_per_second": 5655.238
+ },
+ {
+ "epoch": 0.819754915162941,
+ "grad_norm": 0.7674282579500541,
+ "learning_rate": 6.397311642659283e-06,
+ "loss": 0.3806436538696289,
+ "num_input_tokens_seen": 532749856,
+ "step": 12175,
+ "train_runtime": 94203.2208,
+ "train_tokens_per_second": 5655.325
+ },
+ {
+ "epoch": 0.8200915701589011,
+ "grad_norm": 0.7532047074504019,
+ "learning_rate": 6.394772714316402e-06,
+ "loss": 0.3875831842422485,
+ "num_input_tokens_seen": 532981096,
+ "step": 12180,
+ "train_runtime": 94238.5397,
+ "train_tokens_per_second": 5655.66
+ },
+ {
+ "epoch": 0.8204282251548612,
+ "grad_norm": 0.691733853853237,
+ "learning_rate": 6.392233395929779e-06,
+ "loss": 0.34835638999938967,
+ "num_input_tokens_seen": 533193808,
+ "step": 12185,
+ "train_runtime": 94267.7706,
+ "train_tokens_per_second": 5656.162
+ },
+ {
+ "epoch": 0.8207648801508214,
+ "grad_norm": 0.6752837485492823,
+ "learning_rate": 6.389693688209526e-06,
+ "loss": 0.362264609336853,
+ "num_input_tokens_seen": 533413896,
+ "step": 12190,
+ "train_runtime": 94307.5026,
+ "train_tokens_per_second": 5656.113
+ },
+ {
+ "epoch": 0.8211015351467815,
+ "grad_norm": 0.6640861040569711,
+ "learning_rate": 6.387153591865867e-06,
+ "loss": 0.3703956604003906,
+ "num_input_tokens_seen": 533642648,
+ "step": 12195,
+ "train_runtime": 94351.3757,
+ "train_tokens_per_second": 5655.907
+ },
+ {
+ "epoch": 0.8214381901427417,
+ "grad_norm": 0.8639133957294087,
+ "learning_rate": 6.384613107609127e-06,
+ "loss": 0.38781538009643557,
+ "num_input_tokens_seen": 533855672,
+ "step": 12200,
+ "train_runtime": 94388.037,
+ "train_tokens_per_second": 5655.968
+ },
+ {
+ "epoch": 0.8217748451387018,
+ "grad_norm": 0.6461164084347707,
+ "learning_rate": 6.382072236149747e-06,
+ "loss": 0.3636336326599121,
+ "num_input_tokens_seen": 534086688,
+ "step": 12205,
+ "train_runtime": 94418.101,
+ "train_tokens_per_second": 5656.613
+ },
+ {
+ "epoch": 0.8221115001346619,
+ "grad_norm": 1.259849346653607,
+ "learning_rate": 6.379530978198273e-06,
+ "loss": 0.33962430953979494,
+ "num_input_tokens_seen": 534295064,
+ "step": 12210,
+ "train_runtime": 94452.6928,
+ "train_tokens_per_second": 5656.748
+ },
+ {
+ "epoch": 0.8224481551306222,
+ "grad_norm": 0.6840892751944132,
+ "learning_rate": 6.376989334465361e-06,
+ "loss": 0.34957318305969237,
+ "num_input_tokens_seen": 534518376,
+ "step": 12215,
+ "train_runtime": 94487.4762,
+ "train_tokens_per_second": 5657.029
+ },
+ {
+ "epoch": 0.8227848101265823,
+ "grad_norm": 0.7080671395957931,
+ "learning_rate": 6.374447305661772e-06,
+ "loss": 0.359330415725708,
+ "num_input_tokens_seen": 534753352,
+ "step": 12220,
+ "train_runtime": 94526.7452,
+ "train_tokens_per_second": 5657.165
+ },
+ {
+ "epoch": 0.8231214651225425,
+ "grad_norm": 0.7000813266501178,
+ "learning_rate": 6.3719048924983776e-06,
+ "loss": 0.3536032199859619,
+ "num_input_tokens_seen": 534959200,
+ "step": 12225,
+ "train_runtime": 94565.9434,
+ "train_tokens_per_second": 5656.996
+ },
+ {
+ "epoch": 0.8234581201185026,
+ "grad_norm": 0.7684419131244044,
+ "learning_rate": 6.369362095686152e-06,
+ "loss": 0.3803597688674927,
+ "num_input_tokens_seen": 535175808,
+ "step": 12230,
+ "train_runtime": 94601.305,
+ "train_tokens_per_second": 5657.172
+ },
+ {
+ "epoch": 0.8237947751144628,
+ "grad_norm": 0.6924107525351223,
+ "learning_rate": 6.366818915936185e-06,
+ "loss": 0.34799120426177976,
+ "num_input_tokens_seen": 535394008,
+ "step": 12235,
+ "train_runtime": 94637.8962,
+ "train_tokens_per_second": 5657.29
+ },
+ {
+ "epoch": 0.8241314301104229,
+ "grad_norm": 0.7585078799576096,
+ "learning_rate": 6.364275353959667e-06,
+ "loss": 0.3779747486114502,
+ "num_input_tokens_seen": 535603560,
+ "step": 12240,
+ "train_runtime": 94679.4987,
+ "train_tokens_per_second": 5657.017
+ },
+ {
+ "epoch": 0.824468085106383,
+ "grad_norm": 0.7685591544772424,
+ "learning_rate": 6.3617314104678966e-06,
+ "loss": 0.36225066184997556,
+ "num_input_tokens_seen": 535838080,
+ "step": 12245,
+ "train_runtime": 94720.1071,
+ "train_tokens_per_second": 5657.068
+ },
+ {
+ "epoch": 0.8248047401023432,
+ "grad_norm": 0.76161666147799,
+ "learning_rate": 6.359187086172278e-06,
+ "loss": 0.3977229118347168,
+ "num_input_tokens_seen": 536041408,
+ "step": 12250,
+ "train_runtime": 94755.4024,
+ "train_tokens_per_second": 5657.107
+ },
+ {
+ "epoch": 0.8251413950983033,
+ "grad_norm": 0.6732965305210872,
+ "learning_rate": 6.356642381784326e-06,
+ "loss": 0.3828271389007568,
+ "num_input_tokens_seen": 536255712,
+ "step": 12255,
+ "train_runtime": 94789.5374,
+ "train_tokens_per_second": 5657.33
+ },
+ {
+ "epoch": 0.8254780500942634,
+ "grad_norm": 0.9216909752320883,
+ "learning_rate": 6.354097298015658e-06,
+ "loss": 0.3526939392089844,
+ "num_input_tokens_seen": 536450456,
+ "step": 12260,
+ "train_runtime": 94828.1952,
+ "train_tokens_per_second": 5657.078
+ },
+ {
+ "epoch": 0.8258147050902236,
+ "grad_norm": 0.7392662084519794,
+ "learning_rate": 6.3515518355779985e-06,
+ "loss": 0.3932209014892578,
+ "num_input_tokens_seen": 536655032,
+ "step": 12265,
+ "train_runtime": 94865.9619,
+ "train_tokens_per_second": 5656.982
+ },
+ {
+ "epoch": 0.8261513600861837,
+ "grad_norm": 0.630920453324516,
+ "learning_rate": 6.349005995183177e-06,
+ "loss": 0.3485422134399414,
+ "num_input_tokens_seen": 536883576,
+ "step": 12270,
+ "train_runtime": 94899.668,
+ "train_tokens_per_second": 5657.381
+ },
+ {
+ "epoch": 0.8264880150821439,
+ "grad_norm": 0.6584126043910694,
+ "learning_rate": 6.346459777543131e-06,
+ "loss": 0.3577850341796875,
+ "num_input_tokens_seen": 537102112,
+ "step": 12275,
+ "train_runtime": 94933.4199,
+ "train_tokens_per_second": 5657.672
+ },
+ {
+ "epoch": 0.826824670078104,
+ "grad_norm": 0.69670427508635,
+ "learning_rate": 6.343913183369902e-06,
+ "loss": 0.3586315155029297,
+ "num_input_tokens_seen": 537306360,
+ "step": 12280,
+ "train_runtime": 94970.6969,
+ "train_tokens_per_second": 5657.602
+ },
+ {
+ "epoch": 0.8271613250740641,
+ "grad_norm": 0.6795435309811332,
+ "learning_rate": 6.341366213375637e-06,
+ "loss": 0.3537776231765747,
+ "num_input_tokens_seen": 537534592,
+ "step": 12285,
+ "train_runtime": 95003.3121,
+ "train_tokens_per_second": 5658.062
+ },
+ {
+ "epoch": 0.8274979800700243,
+ "grad_norm": 0.7334092343426839,
+ "learning_rate": 6.338818868272587e-06,
+ "loss": 0.3713988780975342,
+ "num_input_tokens_seen": 537747240,
+ "step": 12290,
+ "train_runtime": 95041.1213,
+ "train_tokens_per_second": 5658.048
+ },
+ {
+ "epoch": 0.8278346350659844,
+ "grad_norm": 0.7029457386112722,
+ "learning_rate": 6.336271148773111e-06,
+ "loss": 0.3642918109893799,
+ "num_input_tokens_seen": 537982080,
+ "step": 12295,
+ "train_runtime": 95085.5253,
+ "train_tokens_per_second": 5657.876
+ },
+ {
+ "epoch": 0.8281712900619446,
+ "grad_norm": 0.664085080042037,
+ "learning_rate": 6.333723055589669e-06,
+ "loss": 0.4124462127685547,
+ "num_input_tokens_seen": 538202248,
+ "step": 12300,
+ "train_runtime": 95123.8409,
+ "train_tokens_per_second": 5657.911
+ },
+ {
+ "epoch": 0.8285079450579047,
+ "grad_norm": 0.7689236487654127,
+ "learning_rate": 6.331174589434826e-06,
+ "loss": 0.3763313293457031,
+ "num_input_tokens_seen": 538420056,
+ "step": 12305,
+ "train_runtime": 95170.7251,
+ "train_tokens_per_second": 5657.413
+ },
+ {
+ "epoch": 0.8288446000538648,
+ "grad_norm": 0.7341729953622874,
+ "learning_rate": 6.328625751021254e-06,
+ "loss": 0.35441150665283205,
+ "num_input_tokens_seen": 538627360,
+ "step": 12310,
+ "train_runtime": 95214.1667,
+ "train_tokens_per_second": 5657.009
+ },
+ {
+ "epoch": 0.829181255049825,
+ "grad_norm": 0.7056989368636907,
+ "learning_rate": 6.326076541061729e-06,
+ "loss": 0.334460711479187,
+ "num_input_tokens_seen": 538844328,
+ "step": 12315,
+ "train_runtime": 95255.2861,
+ "train_tokens_per_second": 5656.844
+ },
+ {
+ "epoch": 0.8295179100457851,
+ "grad_norm": 0.7430926488416755,
+ "learning_rate": 6.323526960269127e-06,
+ "loss": 0.3949697971343994,
+ "num_input_tokens_seen": 539054992,
+ "step": 12320,
+ "train_runtime": 95291.3412,
+ "train_tokens_per_second": 5656.915
+ },
+ {
+ "epoch": 0.8298545650417453,
+ "grad_norm": 0.7296769222033309,
+ "learning_rate": 6.3209770093564315e-06,
+ "loss": 0.3561330556869507,
+ "num_input_tokens_seen": 539280216,
+ "step": 12325,
+ "train_runtime": 95332.3002,
+ "train_tokens_per_second": 5656.847
+ },
+ {
+ "epoch": 0.8301912200377054,
+ "grad_norm": 0.8506709358501154,
+ "learning_rate": 6.318426689036727e-06,
+ "loss": 0.3534719705581665,
+ "num_input_tokens_seen": 539494040,
+ "step": 12330,
+ "train_runtime": 95376.0518,
+ "train_tokens_per_second": 5656.494
+ },
+ {
+ "epoch": 0.8305278750336655,
+ "grad_norm": 0.7341499431044767,
+ "learning_rate": 6.315876000023202e-06,
+ "loss": 0.37946810722351076,
+ "num_input_tokens_seen": 539706008,
+ "step": 12335,
+ "train_runtime": 95416.5574,
+ "train_tokens_per_second": 5656.314
+ },
+ {
+ "epoch": 0.8308645300296257,
+ "grad_norm": 3.744199811122827,
+ "learning_rate": 6.313324943029151e-06,
+ "loss": 0.3778734922409058,
+ "num_input_tokens_seen": 539917272,
+ "step": 12340,
+ "train_runtime": 95460.2877,
+ "train_tokens_per_second": 5655.936
+ },
+ {
+ "epoch": 0.8312011850255858,
+ "grad_norm": 0.742412638594704,
+ "learning_rate": 6.310773518767967e-06,
+ "loss": 0.35590362548828125,
+ "num_input_tokens_seen": 540130136,
+ "step": 12345,
+ "train_runtime": 95493.567,
+ "train_tokens_per_second": 5656.194
+ },
+ {
+ "epoch": 0.831537840021546,
+ "grad_norm": 0.702835896250019,
+ "learning_rate": 6.30822172795315e-06,
+ "loss": 0.36078743934631347,
+ "num_input_tokens_seen": 540336200,
+ "step": 12350,
+ "train_runtime": 95528.2473,
+ "train_tokens_per_second": 5656.298
+ },
+ {
+ "epoch": 0.8318744950175061,
+ "grad_norm": 0.7843017900068701,
+ "learning_rate": 6.3056695712982965e-06,
+ "loss": 0.4154792785644531,
+ "num_input_tokens_seen": 540561200,
+ "step": 12355,
+ "train_runtime": 95559.8481,
+ "train_tokens_per_second": 5656.782
+ },
+ {
+ "epoch": 0.8322111500134662,
+ "grad_norm": 0.6757106331074613,
+ "learning_rate": 6.303117049517111e-06,
+ "loss": 0.3399226427078247,
+ "num_input_tokens_seen": 540776592,
+ "step": 12360,
+ "train_runtime": 95600.3642,
+ "train_tokens_per_second": 5656.637
+ },
+ {
+ "epoch": 0.8325478050094264,
+ "grad_norm": 0.7135184795154971,
+ "learning_rate": 6.300564163323398e-06,
+ "loss": 0.35541400909423826,
+ "num_input_tokens_seen": 540989568,
+ "step": 12365,
+ "train_runtime": 95640.6371,
+ "train_tokens_per_second": 5656.482
+ },
+ {
+ "epoch": 0.8328844600053865,
+ "grad_norm": 0.7511263678823955,
+ "learning_rate": 6.298010913431065e-06,
+ "loss": 0.36992754936218264,
+ "num_input_tokens_seen": 541205456,
+ "step": 12370,
+ "train_runtime": 95679.064,
+ "train_tokens_per_second": 5656.467
+ },
+ {
+ "epoch": 0.8332211150013467,
+ "grad_norm": 0.735359416996797,
+ "learning_rate": 6.295457300554119e-06,
+ "loss": 0.3586021184921265,
+ "num_input_tokens_seen": 541434640,
+ "step": 12375,
+ "train_runtime": 95717.4462,
+ "train_tokens_per_second": 5656.593
+ },
+ {
+ "epoch": 0.8335577699973068,
+ "grad_norm": 0.8567636961499998,
+ "learning_rate": 6.292903325406668e-06,
+ "loss": 0.38977057933807374,
+ "num_input_tokens_seen": 541635304,
+ "step": 12380,
+ "train_runtime": 95751.5012,
+ "train_tokens_per_second": 5656.677
+ },
+ {
+ "epoch": 0.8338944249932669,
+ "grad_norm": 0.7502505399469903,
+ "learning_rate": 6.2903489887029255e-06,
+ "loss": 0.34544246196746825,
+ "num_input_tokens_seen": 541858656,
+ "step": 12385,
+ "train_runtime": 95786.9128,
+ "train_tokens_per_second": 5656.917
+ },
+ {
+ "epoch": 0.8342310799892271,
+ "grad_norm": 0.6987916307726929,
+ "learning_rate": 6.287794291157204e-06,
+ "loss": 0.36299178600311277,
+ "num_input_tokens_seen": 542076592,
+ "step": 12390,
+ "train_runtime": 95820.7633,
+ "train_tokens_per_second": 5657.193
+ },
+ {
+ "epoch": 0.8345677349851872,
+ "grad_norm": 0.8150952905885787,
+ "learning_rate": 6.285239233483915e-06,
+ "loss": 0.36720116138458253,
+ "num_input_tokens_seen": 542309048,
+ "step": 12395,
+ "train_runtime": 95859.9652,
+ "train_tokens_per_second": 5657.305
+ },
+ {
+ "epoch": 0.8349043899811474,
+ "grad_norm": 0.683112485827057,
+ "learning_rate": 6.2826838163975724e-06,
+ "loss": 0.3783855438232422,
+ "num_input_tokens_seen": 542526280,
+ "step": 12400,
+ "train_runtime": 95898.4045,
+ "train_tokens_per_second": 5657.302
+ },
+ {
+ "epoch": 0.8352410449771075,
+ "grad_norm": 0.8138279320478745,
+ "learning_rate": 6.2801280406127905e-06,
+ "loss": 0.35349900722503663,
+ "num_input_tokens_seen": 542750064,
+ "step": 12405,
+ "train_runtime": 95934.8168,
+ "train_tokens_per_second": 5657.488
+ },
+ {
+ "epoch": 0.8355776999730676,
+ "grad_norm": 0.7615550252521133,
+ "learning_rate": 6.2775719068442845e-06,
+ "loss": 0.36704421043395996,
+ "num_input_tokens_seen": 542968240,
+ "step": 12410,
+ "train_runtime": 95972.339,
+ "train_tokens_per_second": 5657.549
+ },
+ {
+ "epoch": 0.8359143549690278,
+ "grad_norm": 0.7219105847794507,
+ "learning_rate": 6.275015415806869e-06,
+ "loss": 0.3532416343688965,
+ "num_input_tokens_seen": 543175408,
+ "step": 12415,
+ "train_runtime": 96016.4623,
+ "train_tokens_per_second": 5657.107
+ },
+ {
+ "epoch": 0.8362510099649879,
+ "grad_norm": 0.6790545597022978,
+ "learning_rate": 6.272458568215458e-06,
+ "loss": 0.3773714303970337,
+ "num_input_tokens_seen": 543389992,
+ "step": 12420,
+ "train_runtime": 96057.0753,
+ "train_tokens_per_second": 5656.949
+ },
+ {
+ "epoch": 0.836587664960948,
+ "grad_norm": 0.758978760003948,
+ "learning_rate": 6.269901364785066e-06,
+ "loss": 0.37525081634521484,
+ "num_input_tokens_seen": 543616944,
+ "step": 12425,
+ "train_runtime": 96095.6171,
+ "train_tokens_per_second": 5657.042
+ },
+ {
+ "epoch": 0.8369243199569082,
+ "grad_norm": 0.6823393482935673,
+ "learning_rate": 6.2673438062308055e-06,
+ "loss": 0.37595961093902586,
+ "num_input_tokens_seen": 543859984,
+ "step": 12430,
+ "train_runtime": 96131.3734,
+ "train_tokens_per_second": 5657.466
+ },
+ {
+ "epoch": 0.8372609749528683,
+ "grad_norm": 0.9689604562670696,
+ "learning_rate": 6.264785893267892e-06,
+ "loss": 0.3472005367279053,
+ "num_input_tokens_seen": 544063976,
+ "step": 12435,
+ "train_runtime": 96162.9294,
+ "train_tokens_per_second": 5657.731
+ },
+ {
+ "epoch": 0.8375976299488285,
+ "grad_norm": 0.7213840812199924,
+ "learning_rate": 6.2622276266116366e-06,
+ "loss": 0.36200590133666993,
+ "num_input_tokens_seen": 544288640,
+ "step": 12440,
+ "train_runtime": 96199.7795,
+ "train_tokens_per_second": 5657.899
+ },
+ {
+ "epoch": 0.8379342849447886,
+ "grad_norm": 0.6737267934047665,
+ "learning_rate": 6.25966900697745e-06,
+ "loss": 0.37799394130706787,
+ "num_input_tokens_seen": 544514152,
+ "step": 12445,
+ "train_runtime": 96237.3914,
+ "train_tokens_per_second": 5658.031
+ },
+ {
+ "epoch": 0.8382709399407487,
+ "grad_norm": 0.6929403643911132,
+ "learning_rate": 6.257110035080843e-06,
+ "loss": 0.3491935729980469,
+ "num_input_tokens_seen": 544741824,
+ "step": 12450,
+ "train_runtime": 96274.7948,
+ "train_tokens_per_second": 5658.198
+ },
+ {
+ "epoch": 0.8386075949367089,
+ "grad_norm": 0.6942505611698238,
+ "learning_rate": 6.254550711637422e-06,
+ "loss": 0.40814781188964844,
+ "num_input_tokens_seen": 544950448,
+ "step": 12455,
+ "train_runtime": 96314.5199,
+ "train_tokens_per_second": 5658.03
+ },
+ {
+ "epoch": 0.838944249932669,
+ "grad_norm": 0.7575909904692931,
+ "learning_rate": 6.251991037362897e-06,
+ "loss": 0.38295414447784426,
+ "num_input_tokens_seen": 545159072,
+ "step": 12460,
+ "train_runtime": 96349.3253,
+ "train_tokens_per_second": 5658.151
+ },
+ {
+ "epoch": 0.8392809049286292,
+ "grad_norm": 0.7581123803912536,
+ "learning_rate": 6.249431012973068e-06,
+ "loss": 0.37713871002197263,
+ "num_input_tokens_seen": 545367120,
+ "step": 12465,
+ "train_runtime": 96388.6672,
+ "train_tokens_per_second": 5658.0
+ },
+ {
+ "epoch": 0.8396175599245893,
+ "grad_norm": 0.7683081862438146,
+ "learning_rate": 6.246870639183843e-06,
+ "loss": 0.37398107051849366,
+ "num_input_tokens_seen": 545578664,
+ "step": 12470,
+ "train_runtime": 96426.5472,
+ "train_tokens_per_second": 5657.972
+ },
+ {
+ "epoch": 0.8399542149205494,
+ "grad_norm": 0.8223266832633437,
+ "learning_rate": 6.2443099167112185e-06,
+ "loss": 0.39701476097106936,
+ "num_input_tokens_seen": 545803912,
+ "step": 12475,
+ "train_runtime": 96462.5666,
+ "train_tokens_per_second": 5658.194
+ },
+ {
+ "epoch": 0.8402908699165096,
+ "grad_norm": 0.6969953488947742,
+ "learning_rate": 6.241748846271293e-06,
+ "loss": 0.3778507709503174,
+ "num_input_tokens_seen": 546030536,
+ "step": 12480,
+ "train_runtime": 96496.5077,
+ "train_tokens_per_second": 5658.552
+ },
+ {
+ "epoch": 0.8406275249124697,
+ "grad_norm": 0.7778677325039889,
+ "learning_rate": 6.239187428580263e-06,
+ "loss": 0.37636547088623046,
+ "num_input_tokens_seen": 546236768,
+ "step": 12485,
+ "train_runtime": 96538.6051,
+ "train_tokens_per_second": 5658.221
+ },
+ {
+ "epoch": 0.8409641799084299,
+ "grad_norm": 0.8059997970084556,
+ "learning_rate": 6.236625664354421e-06,
+ "loss": 0.3784951686859131,
+ "num_input_tokens_seen": 546462096,
+ "step": 12490,
+ "train_runtime": 96574.6617,
+ "train_tokens_per_second": 5658.442
+ },
+ {
+ "epoch": 0.84130083490439,
+ "grad_norm": 0.8520532583827878,
+ "learning_rate": 6.2340635543101545e-06,
+ "loss": 0.3676713466644287,
+ "num_input_tokens_seen": 546657168,
+ "step": 12495,
+ "train_runtime": 96614.4259,
+ "train_tokens_per_second": 5658.132
+ },
+ {
+ "epoch": 0.8416374899003501,
+ "grad_norm": 0.6539078891022843,
+ "learning_rate": 6.23150109916395e-06,
+ "loss": 0.33987085819244384,
+ "num_input_tokens_seen": 546881208,
+ "step": 12500,
+ "train_runtime": 96655.3189,
+ "train_tokens_per_second": 5658.056
+ },
+ {
+ "epoch": 0.8419741448963103,
+ "grad_norm": 0.6624040098906347,
+ "learning_rate": 6.22893829963239e-06,
+ "loss": 0.3859384059906006,
+ "num_input_tokens_seen": 547092816,
+ "step": 12505,
+ "train_runtime": 96691.3494,
+ "train_tokens_per_second": 5658.136
+ },
+ {
+ "epoch": 0.8423107998922704,
+ "grad_norm": 0.7298942090178581,
+ "learning_rate": 6.226375156432153e-06,
+ "loss": 0.357137393951416,
+ "num_input_tokens_seen": 547322656,
+ "step": 12510,
+ "train_runtime": 96731.546,
+ "train_tokens_per_second": 5658.161
+ },
+ {
+ "epoch": 0.8426474548882306,
+ "grad_norm": 0.7316486599091492,
+ "learning_rate": 6.2238116702800145e-06,
+ "loss": 0.3710299491882324,
+ "num_input_tokens_seen": 547537944,
+ "step": 12515,
+ "train_runtime": 96762.4833,
+ "train_tokens_per_second": 5658.577
+ },
+ {
+ "epoch": 0.8429841098841907,
+ "grad_norm": 0.7998109937320629,
+ "learning_rate": 6.221247841892846e-06,
+ "loss": 0.3879734992980957,
+ "num_input_tokens_seen": 547746624,
+ "step": 12520,
+ "train_runtime": 96805.6441,
+ "train_tokens_per_second": 5658.21
+ },
+ {
+ "epoch": 0.8433207648801508,
+ "grad_norm": 0.7307238083679223,
+ "learning_rate": 6.218683671987611e-06,
+ "loss": 0.3691586971282959,
+ "num_input_tokens_seen": 547983928,
+ "step": 12525,
+ "train_runtime": 96840.3547,
+ "train_tokens_per_second": 5658.632
+ },
+ {
+ "epoch": 0.843657419876111,
+ "grad_norm": 0.7271187490915773,
+ "learning_rate": 6.216119161281373e-06,
+ "loss": 0.37589943408966064,
+ "num_input_tokens_seen": 548209600,
+ "step": 12530,
+ "train_runtime": 96879.0798,
+ "train_tokens_per_second": 5658.699
+ },
+ {
+ "epoch": 0.8439940748720711,
+ "grad_norm": 0.7311632772691732,
+ "learning_rate": 6.213554310491291e-06,
+ "loss": 0.3672966003417969,
+ "num_input_tokens_seen": 548421688,
+ "step": 12535,
+ "train_runtime": 96928.6015,
+ "train_tokens_per_second": 5657.996
+ },
+ {
+ "epoch": 0.8443307298680313,
+ "grad_norm": 0.7450184498891292,
+ "learning_rate": 6.210989120334616e-06,
+ "loss": 0.362609338760376,
+ "num_input_tokens_seen": 548644072,
+ "step": 12540,
+ "train_runtime": 96971.5458,
+ "train_tokens_per_second": 5657.784
+ },
+ {
+ "epoch": 0.8446673848639914,
+ "grad_norm": 0.7758427572495085,
+ "learning_rate": 6.208423591528694e-06,
+ "loss": 0.38643317222595214,
+ "num_input_tokens_seen": 548854336,
+ "step": 12545,
+ "train_runtime": 97009.7887,
+ "train_tokens_per_second": 5657.721
+ },
+ {
+ "epoch": 0.8450040398599515,
+ "grad_norm": 0.688742133220875,
+ "learning_rate": 6.205857724790969e-06,
+ "loss": 0.3612468481063843,
+ "num_input_tokens_seen": 549072600,
+ "step": 12550,
+ "train_runtime": 97048.5889,
+ "train_tokens_per_second": 5657.708
+ },
+ {
+ "epoch": 0.8453406948559117,
+ "grad_norm": 0.7926232068697935,
+ "learning_rate": 6.203291520838974e-06,
+ "loss": 0.409790563583374,
+ "num_input_tokens_seen": 549303984,
+ "step": 12555,
+ "train_runtime": 97086.2605,
+ "train_tokens_per_second": 5657.896
+ },
+ {
+ "epoch": 0.8456773498518718,
+ "grad_norm": 0.7041401739211672,
+ "learning_rate": 6.200724980390344e-06,
+ "loss": 0.34305951595306394,
+ "num_input_tokens_seen": 549497696,
+ "step": 12560,
+ "train_runtime": 97125.3875,
+ "train_tokens_per_second": 5657.611
+ },
+ {
+ "epoch": 0.846014004847832,
+ "grad_norm": 0.6959581809215943,
+ "learning_rate": 6.198158104162804e-06,
+ "loss": 0.3178701400756836,
+ "num_input_tokens_seen": 549713600,
+ "step": 12565,
+ "train_runtime": 97158.9336,
+ "train_tokens_per_second": 5657.88
+ },
+ {
+ "epoch": 0.8463506598437921,
+ "grad_norm": 0.6766745725154476,
+ "learning_rate": 6.1955908928741685e-06,
+ "loss": 0.37602977752685546,
+ "num_input_tokens_seen": 549953184,
+ "step": 12570,
+ "train_runtime": 97195.8557,
+ "train_tokens_per_second": 5658.196
+ },
+ {
+ "epoch": 0.8466873148397522,
+ "grad_norm": 0.6881093744203187,
+ "learning_rate": 6.193023347242353e-06,
+ "loss": 0.38083457946777344,
+ "num_input_tokens_seen": 550184352,
+ "step": 12575,
+ "train_runtime": 97232.314,
+ "train_tokens_per_second": 5658.452
+ },
+ {
+ "epoch": 0.8470239698357124,
+ "grad_norm": 0.741931440728507,
+ "learning_rate": 6.190455467985361e-06,
+ "loss": 0.3633177042007446,
+ "num_input_tokens_seen": 550405552,
+ "step": 12580,
+ "train_runtime": 97272.8862,
+ "train_tokens_per_second": 5658.366
+ },
+ {
+ "epoch": 0.8473606248316725,
+ "grad_norm": 0.7792859259983645,
+ "learning_rate": 6.187887255821295e-06,
+ "loss": 0.3954836368560791,
+ "num_input_tokens_seen": 550625896,
+ "step": 12585,
+ "train_runtime": 97305.8601,
+ "train_tokens_per_second": 5658.713
+ },
+ {
+ "epoch": 0.8476972798276327,
+ "grad_norm": 0.7394004400669401,
+ "learning_rate": 6.185318711468345e-06,
+ "loss": 0.3501850128173828,
+ "num_input_tokens_seen": 550827088,
+ "step": 12590,
+ "train_runtime": 97346.1248,
+ "train_tokens_per_second": 5658.439
+ },
+ {
+ "epoch": 0.8480339348235928,
+ "grad_norm": 0.7906419360212633,
+ "learning_rate": 6.182749835644799e-06,
+ "loss": 0.3587897777557373,
+ "num_input_tokens_seen": 551036712,
+ "step": 12595,
+ "train_runtime": 97378.9075,
+ "train_tokens_per_second": 5658.687
+ },
+ {
+ "epoch": 0.8483705898195529,
+ "grad_norm": 0.6720676543784547,
+ "learning_rate": 6.180180629069031e-06,
+ "loss": 0.3534966468811035,
+ "num_input_tokens_seen": 551269008,
+ "step": 12600,
+ "train_runtime": 97416.9713,
+ "train_tokens_per_second": 5658.86
+ },
+ {
+ "epoch": 0.8487072448155131,
+ "grad_norm": 1.0844161686609408,
+ "learning_rate": 6.177611092459514e-06,
+ "loss": 0.3404214859008789,
+ "num_input_tokens_seen": 551486584,
+ "step": 12605,
+ "train_runtime": 97454.9229,
+ "train_tokens_per_second": 5658.889
+ },
+ {
+ "epoch": 0.8490438998114732,
+ "grad_norm": 0.8339846261840644,
+ "learning_rate": 6.175041226534809e-06,
+ "loss": 0.3584805965423584,
+ "num_input_tokens_seen": 551707952,
+ "step": 12610,
+ "train_runtime": 97487.377,
+ "train_tokens_per_second": 5659.276
+ },
+ {
+ "epoch": 0.8493805548074334,
+ "grad_norm": 0.688269893800755,
+ "learning_rate": 6.172471032013574e-06,
+ "loss": 0.327656888961792,
+ "num_input_tokens_seen": 551932560,
+ "step": 12615,
+ "train_runtime": 97526.9729,
+ "train_tokens_per_second": 5659.281
+ },
+ {
+ "epoch": 0.8497172098033935,
+ "grad_norm": 0.6595384550469138,
+ "learning_rate": 6.169900509614553e-06,
+ "loss": 0.3601629972457886,
+ "num_input_tokens_seen": 552163800,
+ "step": 12620,
+ "train_runtime": 97577.6237,
+ "train_tokens_per_second": 5658.713
+ },
+ {
+ "epoch": 0.8500538647993536,
+ "grad_norm": 0.8030731667400421,
+ "learning_rate": 6.167329660056585e-06,
+ "loss": 0.36720144748687744,
+ "num_input_tokens_seen": 552375472,
+ "step": 12625,
+ "train_runtime": 97617.7309,
+ "train_tokens_per_second": 5658.557
+ },
+ {
+ "epoch": 0.8503905197953138,
+ "grad_norm": 0.6580730799233167,
+ "learning_rate": 6.1647584840586e-06,
+ "loss": 0.36503446102142334,
+ "num_input_tokens_seen": 552599576,
+ "step": 12630,
+ "train_runtime": 97665.6991,
+ "train_tokens_per_second": 5658.072
+ },
+ {
+ "epoch": 0.8507271747912739,
+ "grad_norm": 0.7006154386468065,
+ "learning_rate": 6.1621869823396175e-06,
+ "loss": 0.37128496170043945,
+ "num_input_tokens_seen": 552832696,
+ "step": 12635,
+ "train_runtime": 97703.9353,
+ "train_tokens_per_second": 5658.244
+ },
+ {
+ "epoch": 0.851063829787234,
+ "grad_norm": 0.7225238597915238,
+ "learning_rate": 6.159615155618752e-06,
+ "loss": 0.37408173084259033,
+ "num_input_tokens_seen": 553059920,
+ "step": 12640,
+ "train_runtime": 97737.0325,
+ "train_tokens_per_second": 5658.653
+ },
+ {
+ "epoch": 0.8514004847831942,
+ "grad_norm": 0.8136467242343185,
+ "learning_rate": 6.157043004615207e-06,
+ "loss": 0.38596835136413576,
+ "num_input_tokens_seen": 553300640,
+ "step": 12645,
+ "train_runtime": 97780.3263,
+ "train_tokens_per_second": 5658.609
+ },
+ {
+ "epoch": 0.8517371397791543,
+ "grad_norm": 0.7939080541143891,
+ "learning_rate": 6.1544705300482755e-06,
+ "loss": 0.3616677761077881,
+ "num_input_tokens_seen": 553532032,
+ "step": 12650,
+ "train_runtime": 97820.9636,
+ "train_tokens_per_second": 5658.624
+ },
+ {
+ "epoch": 0.8520737947751145,
+ "grad_norm": 0.6610977560801387,
+ "learning_rate": 6.151897732637341e-06,
+ "loss": 0.3330955982208252,
+ "num_input_tokens_seen": 553739416,
+ "step": 12655,
+ "train_runtime": 97857.6515,
+ "train_tokens_per_second": 5658.622
+ },
+ {
+ "epoch": 0.8524104497710746,
+ "grad_norm": 0.8508497854348367,
+ "learning_rate": 6.149324613101877e-06,
+ "loss": 0.37527806758880616,
+ "num_input_tokens_seen": 553959976,
+ "step": 12660,
+ "train_runtime": 97897.0387,
+ "train_tokens_per_second": 5658.598
+ },
+ {
+ "epoch": 0.8527471047670347,
+ "grad_norm": 0.6193562313644957,
+ "learning_rate": 6.146751172161453e-06,
+ "loss": 0.33863201141357424,
+ "num_input_tokens_seen": 554189776,
+ "step": 12665,
+ "train_runtime": 97940.3461,
+ "train_tokens_per_second": 5658.442
+ },
+ {
+ "epoch": 0.8530837597629949,
+ "grad_norm": 0.7796518940843955,
+ "learning_rate": 6.144177410535719e-06,
+ "loss": 0.3652313470840454,
+ "num_input_tokens_seen": 554414232,
+ "step": 12670,
+ "train_runtime": 97979.2461,
+ "train_tokens_per_second": 5658.486
+ },
+ {
+ "epoch": 0.853420414758955,
+ "grad_norm": 0.7186651171193743,
+ "learning_rate": 6.14160332894442e-06,
+ "loss": 0.3651700973510742,
+ "num_input_tokens_seen": 554625192,
+ "step": 12675,
+ "train_runtime": 98015.4768,
+ "train_tokens_per_second": 5658.547
+ },
+ {
+ "epoch": 0.8537570697549152,
+ "grad_norm": 0.724709697937217,
+ "learning_rate": 6.139028928107391e-06,
+ "loss": 0.36736459732055665,
+ "num_input_tokens_seen": 554842240,
+ "step": 12680,
+ "train_runtime": 98056.2763,
+ "train_tokens_per_second": 5658.406
+ },
+ {
+ "epoch": 0.8540937247508753,
+ "grad_norm": 0.7018554477013618,
+ "learning_rate": 6.136454208744553e-06,
+ "loss": 0.3479247808456421,
+ "num_input_tokens_seen": 555060144,
+ "step": 12685,
+ "train_runtime": 98097.7581,
+ "train_tokens_per_second": 5658.235
+ },
+ {
+ "epoch": 0.8544303797468354,
+ "grad_norm": 0.7210589341441987,
+ "learning_rate": 6.133879171575921e-06,
+ "loss": 0.3790309906005859,
+ "num_input_tokens_seen": 555276648,
+ "step": 12690,
+ "train_runtime": 98135.4374,
+ "train_tokens_per_second": 5658.268
+ },
+ {
+ "epoch": 0.8547670347427956,
+ "grad_norm": 0.8012051731436295,
+ "learning_rate": 6.131303817321595e-06,
+ "loss": 0.3976943254470825,
+ "num_input_tokens_seen": 555506600,
+ "step": 12695,
+ "train_runtime": 98173.0099,
+ "train_tokens_per_second": 5658.445
+ },
+ {
+ "epoch": 0.8551036897387557,
+ "grad_norm": 0.6686702415410488,
+ "learning_rate": 6.128728146701763e-06,
+ "loss": 0.3907043933868408,
+ "num_input_tokens_seen": 555735640,
+ "step": 12700,
+ "train_runtime": 98208.5683,
+ "train_tokens_per_second": 5658.729
+ },
+ {
+ "epoch": 0.8554403447347159,
+ "grad_norm": 0.6745127609405563,
+ "learning_rate": 6.126152160436705e-06,
+ "loss": 0.3401200294494629,
+ "num_input_tokens_seen": 555954936,
+ "step": 12705,
+ "train_runtime": 98247.4191,
+ "train_tokens_per_second": 5658.723
+ },
+ {
+ "epoch": 0.855776999730676,
+ "grad_norm": 0.9952487891058109,
+ "learning_rate": 6.123575859246784e-06,
+ "loss": 0.3820429801940918,
+ "num_input_tokens_seen": 556162568,
+ "step": 12710,
+ "train_runtime": 98285.6421,
+ "train_tokens_per_second": 5658.635
+ },
+ {
+ "epoch": 0.8561136547266361,
+ "grad_norm": 0.7343846388254841,
+ "learning_rate": 6.120999243852459e-06,
+ "loss": 0.3488877773284912,
+ "num_input_tokens_seen": 556374088,
+ "step": 12715,
+ "train_runtime": 98325.3673,
+ "train_tokens_per_second": 5658.5
+ },
+ {
+ "epoch": 0.8564503097225963,
+ "grad_norm": 0.8403978577775663,
+ "learning_rate": 6.118422314974269e-06,
+ "loss": 0.3801596164703369,
+ "num_input_tokens_seen": 556596000,
+ "step": 12720,
+ "train_runtime": 98368.0103,
+ "train_tokens_per_second": 5658.303
+ },
+ {
+ "epoch": 0.8567869647185564,
+ "grad_norm": 0.7544062671681226,
+ "learning_rate": 6.115845073332845e-06,
+ "loss": 0.36686029434204104,
+ "num_input_tokens_seen": 556828456,
+ "step": 12725,
+ "train_runtime": 98413.3452,
+ "train_tokens_per_second": 5658.058
+ },
+ {
+ "epoch": 0.8571236197145166,
+ "grad_norm": 0.8566960890477198,
+ "learning_rate": 6.113267519648905e-06,
+ "loss": 0.39199421405792234,
+ "num_input_tokens_seen": 557067040,
+ "step": 12730,
+ "train_runtime": 98457.4483,
+ "train_tokens_per_second": 5657.947
+ },
+ {
+ "epoch": 0.8574602747104767,
+ "grad_norm": 0.6753240704632734,
+ "learning_rate": 6.110689654643253e-06,
+ "loss": 0.36382412910461426,
+ "num_input_tokens_seen": 557274264,
+ "step": 12735,
+ "train_runtime": 98497.8079,
+ "train_tokens_per_second": 5657.733
+ },
+ {
+ "epoch": 0.8577969297064368,
+ "grad_norm": 0.7623447924884228,
+ "learning_rate": 6.1081114790367805e-06,
+ "loss": 0.37306723594665525,
+ "num_input_tokens_seen": 557484064,
+ "step": 12740,
+ "train_runtime": 98531.9864,
+ "train_tokens_per_second": 5657.899
+ },
+ {
+ "epoch": 0.858133584702397,
+ "grad_norm": 0.7690727585148884,
+ "learning_rate": 6.105532993550467e-06,
+ "loss": 0.37079343795776365,
+ "num_input_tokens_seen": 557707168,
+ "step": 12745,
+ "train_runtime": 98570.6779,
+ "train_tokens_per_second": 5657.942
+ },
+ {
+ "epoch": 0.8584702396983571,
+ "grad_norm": 0.7227829734774792,
+ "learning_rate": 6.1029541989053765e-06,
+ "loss": 0.3675063610076904,
+ "num_input_tokens_seen": 557923328,
+ "step": 12750,
+ "train_runtime": 98611.8723,
+ "train_tokens_per_second": 5657.77
+ },
+ {
+ "epoch": 0.8588068946943173,
+ "grad_norm": 0.6216338336006675,
+ "learning_rate": 6.100375095822661e-06,
+ "loss": 0.3660135746002197,
+ "num_input_tokens_seen": 558141912,
+ "step": 12755,
+ "train_runtime": 98653.6004,
+ "train_tokens_per_second": 5657.593
+ },
+ {
+ "epoch": 0.8591435496902774,
+ "grad_norm": 0.7752180098632147,
+ "learning_rate": 6.097795685023558e-06,
+ "loss": 0.3654344081878662,
+ "num_input_tokens_seen": 558363664,
+ "step": 12760,
+ "train_runtime": 98689.3362,
+ "train_tokens_per_second": 5657.791
+ },
+ {
+ "epoch": 0.8594802046862375,
+ "grad_norm": 0.7139656897700937,
+ "learning_rate": 6.095215967229396e-06,
+ "loss": 0.3468952655792236,
+ "num_input_tokens_seen": 558569128,
+ "step": 12765,
+ "train_runtime": 98726.8939,
+ "train_tokens_per_second": 5657.72
+ },
+ {
+ "epoch": 0.8598168596821977,
+ "grad_norm": 0.6751611340302733,
+ "learning_rate": 6.092635943161578e-06,
+ "loss": 0.37009525299072266,
+ "num_input_tokens_seen": 558798144,
+ "step": 12770,
+ "train_runtime": 98770.45,
+ "train_tokens_per_second": 5657.544
+ },
+ {
+ "epoch": 0.8601535146781578,
+ "grad_norm": 0.7202350000001299,
+ "learning_rate": 6.090055613541603e-06,
+ "loss": 0.34846246242523193,
+ "num_input_tokens_seen": 559026000,
+ "step": 12775,
+ "train_runtime": 98813.4858,
+ "train_tokens_per_second": 5657.386
+ },
+ {
+ "epoch": 0.860490169674118,
+ "grad_norm": 0.7513541698981121,
+ "learning_rate": 6.087474979091052e-06,
+ "loss": 0.3586388111114502,
+ "num_input_tokens_seen": 559233040,
+ "step": 12780,
+ "train_runtime": 98860.409,
+ "train_tokens_per_second": 5656.795
+ },
+ {
+ "epoch": 0.8608268246700781,
+ "grad_norm": 0.7207910613458428,
+ "learning_rate": 6.084894040531591e-06,
+ "loss": 0.3598465919494629,
+ "num_input_tokens_seen": 559456240,
+ "step": 12785,
+ "train_runtime": 98903.5566,
+ "train_tokens_per_second": 5656.584
+ },
+ {
+ "epoch": 0.8611634796660382,
+ "grad_norm": 0.7450858357428686,
+ "learning_rate": 6.0823127985849705e-06,
+ "loss": 0.35327887535095215,
+ "num_input_tokens_seen": 559662696,
+ "step": 12790,
+ "train_runtime": 98942.3562,
+ "train_tokens_per_second": 5656.452
+ },
+ {
+ "epoch": 0.8615001346619984,
+ "grad_norm": 0.7413157970110851,
+ "learning_rate": 6.079731253973028e-06,
+ "loss": 0.3576143741607666,
+ "num_input_tokens_seen": 559886624,
+ "step": 12795,
+ "train_runtime": 98988.6398,
+ "train_tokens_per_second": 5656.069
+ },
+ {
+ "epoch": 0.8618367896579585,
+ "grad_norm": 0.7990871684605166,
+ "learning_rate": 6.077149407417683e-06,
+ "loss": 0.36043851375579833,
+ "num_input_tokens_seen": 560101432,
+ "step": 12800,
+ "train_runtime": 99024.1718,
+ "train_tokens_per_second": 5656.209
+ },
+ {
+ "epoch": 0.8621734446539187,
+ "grad_norm": 0.7364962437233615,
+ "learning_rate": 6.07456725964094e-06,
+ "loss": 0.3793959617614746,
+ "num_input_tokens_seen": 560309472,
+ "step": 12805,
+ "train_runtime": 99059.299,
+ "train_tokens_per_second": 5656.304
+ },
+ {
+ "epoch": 0.8625100996498788,
+ "grad_norm": 0.680598341183089,
+ "learning_rate": 6.07198481136489e-06,
+ "loss": 0.3436178684234619,
+ "num_input_tokens_seen": 560520408,
+ "step": 12810,
+ "train_runtime": 99092.681,
+ "train_tokens_per_second": 5656.527
+ },
+ {
+ "epoch": 0.8628467546458389,
+ "grad_norm": 0.765274818340009,
+ "learning_rate": 6.069402063311706e-06,
+ "loss": 0.3549325704574585,
+ "num_input_tokens_seen": 560731064,
+ "step": 12815,
+ "train_runtime": 99135.7209,
+ "train_tokens_per_second": 5656.196
+ },
+ {
+ "epoch": 0.8631834096417991,
+ "grad_norm": 0.8395744171280569,
+ "learning_rate": 6.066819016203646e-06,
+ "loss": 0.36306018829345704,
+ "num_input_tokens_seen": 560944360,
+ "step": 12820,
+ "train_runtime": 99173.0123,
+ "train_tokens_per_second": 5656.22
+ },
+ {
+ "epoch": 0.8635200646377592,
+ "grad_norm": 0.7436029507337527,
+ "learning_rate": 6.064235670763048e-06,
+ "loss": 0.36905884742736816,
+ "num_input_tokens_seen": 561164928,
+ "step": 12825,
+ "train_runtime": 99209.242,
+ "train_tokens_per_second": 5656.378
+ },
+ {
+ "epoch": 0.8638567196337193,
+ "grad_norm": 0.6763369772275482,
+ "learning_rate": 6.06165202771234e-06,
+ "loss": 0.3765472412109375,
+ "num_input_tokens_seen": 561394056,
+ "step": 12830,
+ "train_runtime": 99250.4089,
+ "train_tokens_per_second": 5656.34
+ },
+ {
+ "epoch": 0.8641933746296795,
+ "grad_norm": 0.6931570065850123,
+ "learning_rate": 6.059068087774026e-06,
+ "loss": 0.34474341869354247,
+ "num_input_tokens_seen": 561593424,
+ "step": 12835,
+ "train_runtime": 99291.1127,
+ "train_tokens_per_second": 5656.029
+ },
+ {
+ "epoch": 0.8645300296256396,
+ "grad_norm": 0.7306127243942253,
+ "learning_rate": 6.0564838516707005e-06,
+ "loss": 0.38959674835205077,
+ "num_input_tokens_seen": 561821784,
+ "step": 12840,
+ "train_runtime": 99328.0768,
+ "train_tokens_per_second": 5656.223
+ },
+ {
+ "epoch": 0.8648666846215998,
+ "grad_norm": 0.6692585369253214,
+ "learning_rate": 6.053899320125033e-06,
+ "loss": 0.34748311042785646,
+ "num_input_tokens_seen": 562042168,
+ "step": 12845,
+ "train_runtime": 99368.2816,
+ "train_tokens_per_second": 5656.153
+ },
+ {
+ "epoch": 0.8652033396175599,
+ "grad_norm": 0.731387810949299,
+ "learning_rate": 6.051314493859782e-06,
+ "loss": 0.37701342105865476,
+ "num_input_tokens_seen": 562268616,
+ "step": 12850,
+ "train_runtime": 99405.5774,
+ "train_tokens_per_second": 5656.309
+ },
+ {
+ "epoch": 0.86553999461352,
+ "grad_norm": 0.744983156104233,
+ "learning_rate": 6.048729373597786e-06,
+ "loss": 0.36400351524353025,
+ "num_input_tokens_seen": 562488728,
+ "step": 12855,
+ "train_runtime": 99441.1097,
+ "train_tokens_per_second": 5656.501
+ },
+ {
+ "epoch": 0.8658766496094802,
+ "grad_norm": 0.7052968569468162,
+ "learning_rate": 6.046143960061963e-06,
+ "loss": 0.38647964000701907,
+ "num_input_tokens_seen": 562715632,
+ "step": 12860,
+ "train_runtime": 99479.679,
+ "train_tokens_per_second": 5656.589
+ },
+ {
+ "epoch": 0.8662133046054403,
+ "grad_norm": 0.8490020320396742,
+ "learning_rate": 6.043558253975319e-06,
+ "loss": 0.3893282413482666,
+ "num_input_tokens_seen": 562949544,
+ "step": 12865,
+ "train_runtime": 99522.1085,
+ "train_tokens_per_second": 5656.528
+ },
+ {
+ "epoch": 0.8665499596014005,
+ "grad_norm": 0.7443300106386931,
+ "learning_rate": 6.040972256060936e-06,
+ "loss": 0.35498225688934326,
+ "num_input_tokens_seen": 563161168,
+ "step": 12870,
+ "train_runtime": 99567.3086,
+ "train_tokens_per_second": 5656.085
+ },
+ {
+ "epoch": 0.8668866145973606,
+ "grad_norm": 0.6849806622056446,
+ "learning_rate": 6.038385967041982e-06,
+ "loss": 0.3538069248199463,
+ "num_input_tokens_seen": 563369656,
+ "step": 12875,
+ "train_runtime": 99613.8289,
+ "train_tokens_per_second": 5655.537
+ },
+ {
+ "epoch": 0.8672232695933207,
+ "grad_norm": 0.6384169452977918,
+ "learning_rate": 6.035799387641703e-06,
+ "loss": 0.36878552436828616,
+ "num_input_tokens_seen": 563594552,
+ "step": 12880,
+ "train_runtime": 99655.4177,
+ "train_tokens_per_second": 5655.433
+ },
+ {
+ "epoch": 0.8675599245892809,
+ "grad_norm": 0.7636357717419208,
+ "learning_rate": 6.033212518583427e-06,
+ "loss": 0.3786178112030029,
+ "num_input_tokens_seen": 563803512,
+ "step": 12885,
+ "train_runtime": 99695.3116,
+ "train_tokens_per_second": 5655.266
+ },
+ {
+ "epoch": 0.867896579585241,
+ "grad_norm": 0.8323239828663773,
+ "learning_rate": 6.030625360590567e-06,
+ "loss": 0.35876643657684326,
+ "num_input_tokens_seen": 564030520,
+ "step": 12890,
+ "train_runtime": 99734.3991,
+ "train_tokens_per_second": 5655.326
+ },
+ {
+ "epoch": 0.8682332345812012,
+ "grad_norm": 0.7006071245579706,
+ "learning_rate": 6.028037914386609e-06,
+ "loss": 0.3620233297348022,
+ "num_input_tokens_seen": 564252688,
+ "step": 12895,
+ "train_runtime": 99773.9473,
+ "train_tokens_per_second": 5655.311
+ },
+ {
+ "epoch": 0.8685698895771613,
+ "grad_norm": 0.7617834642572114,
+ "learning_rate": 6.025450180695128e-06,
+ "loss": 0.38210210800170896,
+ "num_input_tokens_seen": 564471472,
+ "step": 12900,
+ "train_runtime": 99808.9271,
+ "train_tokens_per_second": 5655.521
+ },
+ {
+ "epoch": 0.8689065445731214,
+ "grad_norm": 0.8560111437943349,
+ "learning_rate": 6.022862160239774e-06,
+ "loss": 0.3630260467529297,
+ "num_input_tokens_seen": 564683208,
+ "step": 12905,
+ "train_runtime": 99850.9033,
+ "train_tokens_per_second": 5655.264
+ },
+ {
+ "epoch": 0.8692431995690816,
+ "grad_norm": 0.6306746195476955,
+ "learning_rate": 6.020273853744277e-06,
+ "loss": 0.3535295009613037,
+ "num_input_tokens_seen": 564903760,
+ "step": 12910,
+ "train_runtime": 99888.6442,
+ "train_tokens_per_second": 5655.335
+ },
+ {
+ "epoch": 0.8695798545650417,
+ "grad_norm": 0.6756821029513189,
+ "learning_rate": 6.017685261932452e-06,
+ "loss": 0.35518722534179686,
+ "num_input_tokens_seen": 565105760,
+ "step": 12915,
+ "train_runtime": 99923.6252,
+ "train_tokens_per_second": 5655.377
+ },
+ {
+ "epoch": 0.8699165095610019,
+ "grad_norm": 0.75004570307814,
+ "learning_rate": 6.015096385528189e-06,
+ "loss": 0.37440195083618166,
+ "num_input_tokens_seen": 565318768,
+ "step": 12920,
+ "train_runtime": 99960.4635,
+ "train_tokens_per_second": 5655.424
+ },
+ {
+ "epoch": 0.870253164556962,
+ "grad_norm": 0.725287349099034,
+ "learning_rate": 6.012507225255457e-06,
+ "loss": 0.3945824146270752,
+ "num_input_tokens_seen": 565515560,
+ "step": 12925,
+ "train_runtime": 99995.6077,
+ "train_tokens_per_second": 5655.404
+ },
+ {
+ "epoch": 0.8705898195529221,
+ "grad_norm": 0.6923174837526764,
+ "learning_rate": 6.009917781838309e-06,
+ "loss": 0.3706216812133789,
+ "num_input_tokens_seen": 565736648,
+ "step": 12930,
+ "train_runtime": 100027.8477,
+ "train_tokens_per_second": 5655.791
+ },
+ {
+ "epoch": 0.8709264745488823,
+ "grad_norm": 0.6934584410887805,
+ "learning_rate": 6.007328056000873e-06,
+ "loss": 0.37667386531829833,
+ "num_input_tokens_seen": 565960800,
+ "step": 12935,
+ "train_runtime": 100063.8699,
+ "train_tokens_per_second": 5655.996
+ },
+ {
+ "epoch": 0.8712631295448424,
+ "grad_norm": 0.7940274414990574,
+ "learning_rate": 6.004738048467359e-06,
+ "loss": 0.379984712600708,
+ "num_input_tokens_seen": 566180672,
+ "step": 12940,
+ "train_runtime": 100101.3217,
+ "train_tokens_per_second": 5656.076
+ },
+ {
+ "epoch": 0.8715997845408026,
+ "grad_norm": 0.7344390394871104,
+ "learning_rate": 6.002147759962054e-06,
+ "loss": 0.37138075828552247,
+ "num_input_tokens_seen": 566401496,
+ "step": 12945,
+ "train_runtime": 100139.2781,
+ "train_tokens_per_second": 5656.137
+ },
+ {
+ "epoch": 0.8719364395367627,
+ "grad_norm": 0.7421475803173249,
+ "learning_rate": 5.999557191209323e-06,
+ "loss": 0.39212737083435056,
+ "num_input_tokens_seen": 566630216,
+ "step": 12950,
+ "train_runtime": 100175.2546,
+ "train_tokens_per_second": 5656.389
+ },
+ {
+ "epoch": 0.8722730945327228,
+ "grad_norm": 0.8003491943698586,
+ "learning_rate": 5.996966342933611e-06,
+ "loss": 0.33372251987457274,
+ "num_input_tokens_seen": 566848176,
+ "step": 12955,
+ "train_runtime": 100212.8634,
+ "train_tokens_per_second": 5656.441
+ },
+ {
+ "epoch": 0.872609749528683,
+ "grad_norm": 0.7450010892993638,
+ "learning_rate": 5.99437521585944e-06,
+ "loss": 0.37457475662231443,
+ "num_input_tokens_seen": 567056408,
+ "step": 12960,
+ "train_runtime": 100250.1105,
+ "train_tokens_per_second": 5656.417
+ },
+ {
+ "epoch": 0.8729464045246431,
+ "grad_norm": 0.7217432409540151,
+ "learning_rate": 5.99178381071141e-06,
+ "loss": 0.3829956531524658,
+ "num_input_tokens_seen": 567275168,
+ "step": 12965,
+ "train_runtime": 100287.1326,
+ "train_tokens_per_second": 5656.51
+ },
+ {
+ "epoch": 0.8732830595206033,
+ "grad_norm": 0.8055483175054227,
+ "learning_rate": 5.989192128214201e-06,
+ "loss": 0.36388020515441893,
+ "num_input_tokens_seen": 567493544,
+ "step": 12970,
+ "train_runtime": 100327.3129,
+ "train_tokens_per_second": 5656.421
+ },
+ {
+ "epoch": 0.8736197145165634,
+ "grad_norm": 0.7033510804995021,
+ "learning_rate": 5.986600169092568e-06,
+ "loss": 0.350065803527832,
+ "num_input_tokens_seen": 567708176,
+ "step": 12975,
+ "train_runtime": 100364.1125,
+ "train_tokens_per_second": 5656.486
+ },
+ {
+ "epoch": 0.8739563695125235,
+ "grad_norm": 0.7167335785321889,
+ "learning_rate": 5.984007934071343e-06,
+ "loss": 0.3742049217224121,
+ "num_input_tokens_seen": 567941160,
+ "step": 12980,
+ "train_runtime": 100407.0573,
+ "train_tokens_per_second": 5656.387
+ },
+ {
+ "epoch": 0.8742930245084837,
+ "grad_norm": 0.7554361554619513,
+ "learning_rate": 5.981415423875436e-06,
+ "loss": 0.36548290252685545,
+ "num_input_tokens_seen": 568166040,
+ "step": 12985,
+ "train_runtime": 100442.582,
+ "train_tokens_per_second": 5656.625
+ },
+ {
+ "epoch": 0.8746296795044438,
+ "grad_norm": 0.9815756056409435,
+ "learning_rate": 5.978822639229836e-06,
+ "loss": 0.3484132528305054,
+ "num_input_tokens_seen": 568367984,
+ "step": 12990,
+ "train_runtime": 100479.4893,
+ "train_tokens_per_second": 5656.557
+ },
+ {
+ "epoch": 0.874966334500404,
+ "grad_norm": 0.6864898867694621,
+ "learning_rate": 5.976229580859607e-06,
+ "loss": 0.36139798164367676,
+ "num_input_tokens_seen": 568596000,
+ "step": 12995,
+ "train_runtime": 100516.4986,
+ "train_tokens_per_second": 5656.743
+ },
+ {
+ "epoch": 0.8753029894963641,
+ "grad_norm": 0.7186739597846133,
+ "learning_rate": 5.9736362494898895e-06,
+ "loss": 0.3539880275726318,
+ "num_input_tokens_seen": 568814440,
+ "step": 13000,
+ "train_runtime": 100555.0904,
+ "train_tokens_per_second": 5656.744
+ },
+ {
+ "epoch": 0.8756396444923242,
+ "grad_norm": 0.6506103454572264,
+ "learning_rate": 5.971042645845897e-06,
+ "loss": 0.3666597843170166,
+ "num_input_tokens_seen": 569029824,
+ "step": 13005,
+ "train_runtime": 100595.3102,
+ "train_tokens_per_second": 5656.624
+ },
+ {
+ "epoch": 0.8759762994882844,
+ "grad_norm": 0.7706124062857171,
+ "learning_rate": 5.968448770652926e-06,
+ "loss": 0.36510615348815917,
+ "num_input_tokens_seen": 569265096,
+ "step": 13010,
+ "train_runtime": 100627.6528,
+ "train_tokens_per_second": 5657.144
+ },
+ {
+ "epoch": 0.8763129544842445,
+ "grad_norm": 0.7419816582789281,
+ "learning_rate": 5.965854624636345e-06,
+ "loss": 0.3825402736663818,
+ "num_input_tokens_seen": 569479936,
+ "step": 13015,
+ "train_runtime": 100669.5038,
+ "train_tokens_per_second": 5656.926
+ },
+ {
+ "epoch": 0.8766496094802047,
+ "grad_norm": 0.7717112571968371,
+ "learning_rate": 5.963260208521598e-06,
+ "loss": 0.3698084354400635,
+ "num_input_tokens_seen": 569691888,
+ "step": 13020,
+ "train_runtime": 100708.4391,
+ "train_tokens_per_second": 5656.844
+ },
+ {
+ "epoch": 0.8769862644761648,
+ "grad_norm": 0.6761324100161831,
+ "learning_rate": 5.960665523034205e-06,
+ "loss": 0.38090927600860597,
+ "num_input_tokens_seen": 569901576,
+ "step": 13025,
+ "train_runtime": 100749.3523,
+ "train_tokens_per_second": 5656.628
+ },
+ {
+ "epoch": 0.8773229194721249,
+ "grad_norm": 0.7359973742678347,
+ "learning_rate": 5.958070568899762e-06,
+ "loss": 0.33312511444091797,
+ "num_input_tokens_seen": 570123424,
+ "step": 13030,
+ "train_runtime": 100790.5444,
+ "train_tokens_per_second": 5656.517
+ },
+ {
+ "epoch": 0.8776595744680851,
+ "grad_norm": 1.2136155199409513,
+ "learning_rate": 5.955475346843939e-06,
+ "loss": 0.3555187225341797,
+ "num_input_tokens_seen": 570337248,
+ "step": 13035,
+ "train_runtime": 100824.0572,
+ "train_tokens_per_second": 5656.758
+ },
+ {
+ "epoch": 0.8779962294640452,
+ "grad_norm": 0.6546063737811157,
+ "learning_rate": 5.952879857592482e-06,
+ "loss": 0.3440871000289917,
+ "num_input_tokens_seen": 570549824,
+ "step": 13040,
+ "train_runtime": 100867.9537,
+ "train_tokens_per_second": 5656.403
+ },
+ {
+ "epoch": 0.8783328844600053,
+ "grad_norm": 0.6844892975959102,
+ "learning_rate": 5.950284101871212e-06,
+ "loss": 0.38202319145202634,
+ "num_input_tokens_seen": 570757152,
+ "step": 13045,
+ "train_runtime": 100906.3146,
+ "train_tokens_per_second": 5656.308
+ },
+ {
+ "epoch": 0.8786695394559655,
+ "grad_norm": 0.747172974034045,
+ "learning_rate": 5.947688080406023e-06,
+ "loss": 0.3463796615600586,
+ "num_input_tokens_seen": 570960928,
+ "step": 13050,
+ "train_runtime": 100940.5074,
+ "train_tokens_per_second": 5656.41
+ },
+ {
+ "epoch": 0.8790061944519256,
+ "grad_norm": 0.7561590864609555,
+ "learning_rate": 5.945091793922885e-06,
+ "loss": 0.3679834842681885,
+ "num_input_tokens_seen": 571177360,
+ "step": 13055,
+ "train_runtime": 100974.1104,
+ "train_tokens_per_second": 5656.671
+ },
+ {
+ "epoch": 0.8793428494478858,
+ "grad_norm": 0.6782284385723677,
+ "learning_rate": 5.942495243147839e-06,
+ "loss": 0.3929899215698242,
+ "num_input_tokens_seen": 571399192,
+ "step": 13060,
+ "train_runtime": 101005.759,
+ "train_tokens_per_second": 5657.095
+ },
+ {
+ "epoch": 0.8796795044438459,
+ "grad_norm": 0.7143338491446609,
+ "learning_rate": 5.939898428807004e-06,
+ "loss": 0.3815101146697998,
+ "num_input_tokens_seen": 571612808,
+ "step": 13065,
+ "train_runtime": 101043.3454,
+ "train_tokens_per_second": 5657.105
+ },
+ {
+ "epoch": 0.880016159439806,
+ "grad_norm": 0.8449030830666151,
+ "learning_rate": 5.937301351626571e-06,
+ "loss": 0.379590368270874,
+ "num_input_tokens_seen": 571824200,
+ "step": 13070,
+ "train_runtime": 101081.8232,
+ "train_tokens_per_second": 5657.043
+ },
+ {
+ "epoch": 0.8803528144357662,
+ "grad_norm": 0.6830378970909649,
+ "learning_rate": 5.934704012332803e-06,
+ "loss": 0.35865001678466796,
+ "num_input_tokens_seen": 572052552,
+ "step": 13075,
+ "train_runtime": 101122.8991,
+ "train_tokens_per_second": 5657.003
+ },
+ {
+ "epoch": 0.8806894694317263,
+ "grad_norm": 0.8010624909706995,
+ "learning_rate": 5.93210641165204e-06,
+ "loss": 0.4059410095214844,
+ "num_input_tokens_seen": 572274280,
+ "step": 13080,
+ "train_runtime": 101165.6483,
+ "train_tokens_per_second": 5656.804
+ },
+ {
+ "epoch": 0.8810261244276865,
+ "grad_norm": 0.7989521029331256,
+ "learning_rate": 5.92950855031069e-06,
+ "loss": 0.36967930793762205,
+ "num_input_tokens_seen": 572493808,
+ "step": 13085,
+ "train_runtime": 101202.3764,
+ "train_tokens_per_second": 5656.921
+ },
+ {
+ "epoch": 0.8813627794236466,
+ "grad_norm": 0.8258054867788488,
+ "learning_rate": 5.926910429035239e-06,
+ "loss": 0.36150131225585935,
+ "num_input_tokens_seen": 572706072,
+ "step": 13090,
+ "train_runtime": 101239.3964,
+ "train_tokens_per_second": 5656.949
+ },
+ {
+ "epoch": 0.8816994344196067,
+ "grad_norm": 0.7636365729370889,
+ "learning_rate": 5.924312048552241e-06,
+ "loss": 0.3525059223175049,
+ "num_input_tokens_seen": 572932096,
+ "step": 13095,
+ "train_runtime": 101273.4837,
+ "train_tokens_per_second": 5657.276
+ },
+ {
+ "epoch": 0.8820360894155669,
+ "grad_norm": 0.7191821908315644,
+ "learning_rate": 5.9217134095883265e-06,
+ "loss": 0.3641670227050781,
+ "num_input_tokens_seen": 573150384,
+ "step": 13100,
+ "train_runtime": 101314.4917,
+ "train_tokens_per_second": 5657.141
+ },
+ {
+ "epoch": 0.882372744411527,
+ "grad_norm": 0.7363436743411855,
+ "learning_rate": 5.919114512870197e-06,
+ "loss": 0.3603955268859863,
+ "num_input_tokens_seen": 573381144,
+ "step": 13105,
+ "train_runtime": 101359.314,
+ "train_tokens_per_second": 5656.916
+ },
+ {
+ "epoch": 0.8827093994074872,
+ "grad_norm": 0.7568233135879606,
+ "learning_rate": 5.916515359124623e-06,
+ "loss": 0.3425842046737671,
+ "num_input_tokens_seen": 573581832,
+ "step": 13110,
+ "train_runtime": 101399.4218,
+ "train_tokens_per_second": 5656.658
+ },
+ {
+ "epoch": 0.8830460544034473,
+ "grad_norm": 0.7215795683330004,
+ "learning_rate": 5.913915949078453e-06,
+ "loss": 0.36979331970214846,
+ "num_input_tokens_seen": 573801512,
+ "step": 13115,
+ "train_runtime": 101438.6834,
+ "train_tokens_per_second": 5656.634
+ },
+ {
+ "epoch": 0.8833827093994074,
+ "grad_norm": 0.7764516396496984,
+ "learning_rate": 5.911316283458601e-06,
+ "loss": 0.3616600513458252,
+ "num_input_tokens_seen": 574007544,
+ "step": 13120,
+ "train_runtime": 101478.4371,
+ "train_tokens_per_second": 5656.448
+ },
+ {
+ "epoch": 0.8837193643953676,
+ "grad_norm": 0.7391044514688772,
+ "learning_rate": 5.908716362992057e-06,
+ "loss": 0.36803760528564455,
+ "num_input_tokens_seen": 574215752,
+ "step": 13125,
+ "train_runtime": 101514.9279,
+ "train_tokens_per_second": 5656.466
+ },
+ {
+ "epoch": 0.8840560193913277,
+ "grad_norm": 0.6863603049824145,
+ "learning_rate": 5.90611618840588e-06,
+ "loss": 0.379679536819458,
+ "num_input_tokens_seen": 574434272,
+ "step": 13130,
+ "train_runtime": 101554.971,
+ "train_tokens_per_second": 5656.388
+ },
+ {
+ "epoch": 0.8843926743872879,
+ "grad_norm": 0.9353489354148398,
+ "learning_rate": 5.903515760427198e-06,
+ "loss": 0.4000962257385254,
+ "num_input_tokens_seen": 574653864,
+ "step": 13135,
+ "train_runtime": 101588.2234,
+ "train_tokens_per_second": 5656.698
+ },
+ {
+ "epoch": 0.884729329383248,
+ "grad_norm": 0.732273202002582,
+ "learning_rate": 5.900915079783218e-06,
+ "loss": 0.3558906316757202,
+ "num_input_tokens_seen": 574887384,
+ "step": 13140,
+ "train_runtime": 101622.3485,
+ "train_tokens_per_second": 5657.096
+ },
+ {
+ "epoch": 0.8850659843792081,
+ "grad_norm": 0.7460917342917756,
+ "learning_rate": 5.898314147201207e-06,
+ "loss": 0.3416097640991211,
+ "num_input_tokens_seen": 575101544,
+ "step": 13145,
+ "train_runtime": 101664.6483,
+ "train_tokens_per_second": 5656.849
+ },
+ {
+ "epoch": 0.8854026393751683,
+ "grad_norm": 1.0430878064810838,
+ "learning_rate": 5.895712963408511e-06,
+ "loss": 0.3921904325485229,
+ "num_input_tokens_seen": 575309032,
+ "step": 13150,
+ "train_runtime": 101704.7114,
+ "train_tokens_per_second": 5656.661
+ },
+ {
+ "epoch": 0.8857392943711284,
+ "grad_norm": 0.7135808350871835,
+ "learning_rate": 5.893111529132539e-06,
+ "loss": 0.38542513847351073,
+ "num_input_tokens_seen": 575518184,
+ "step": 13155,
+ "train_runtime": 101742.2738,
+ "train_tokens_per_second": 5656.628
+ },
+ {
+ "epoch": 0.8860759493670886,
+ "grad_norm": 0.7661970012022706,
+ "learning_rate": 5.8905098451007755e-06,
+ "loss": 0.37914443016052246,
+ "num_input_tokens_seen": 575721848,
+ "step": 13160,
+ "train_runtime": 101781.9439,
+ "train_tokens_per_second": 5656.424
+ },
+ {
+ "epoch": 0.8864126043630487,
+ "grad_norm": 0.7203329040983677,
+ "learning_rate": 5.887907912040774e-06,
+ "loss": 0.3751526355743408,
+ "num_input_tokens_seen": 575946104,
+ "step": 13165,
+ "train_runtime": 101814.7009,
+ "train_tokens_per_second": 5656.807
+ },
+ {
+ "epoch": 0.8867492593590088,
+ "grad_norm": 0.7424424433197104,
+ "learning_rate": 5.885305730680158e-06,
+ "loss": 0.3788419723510742,
+ "num_input_tokens_seen": 576173784,
+ "step": 13170,
+ "train_runtime": 101847.3914,
+ "train_tokens_per_second": 5657.227
+ },
+ {
+ "epoch": 0.887085914354969,
+ "grad_norm": 0.7289665105864628,
+ "learning_rate": 5.882703301746616e-06,
+ "loss": 0.36290099620819094,
+ "num_input_tokens_seen": 576389120,
+ "step": 13175,
+ "train_runtime": 101889.8836,
+ "train_tokens_per_second": 5656.981
+ },
+ {
+ "epoch": 0.8874225693509291,
+ "grad_norm": 0.7262294984371385,
+ "learning_rate": 5.88010062596791e-06,
+ "loss": 0.35980806350708006,
+ "num_input_tokens_seen": 576598016,
+ "step": 13180,
+ "train_runtime": 101930.8807,
+ "train_tokens_per_second": 5656.755
+ },
+ {
+ "epoch": 0.8877592243468893,
+ "grad_norm": 0.601051210488872,
+ "learning_rate": 5.877497704071871e-06,
+ "loss": 0.3475220203399658,
+ "num_input_tokens_seen": 576831392,
+ "step": 13185,
+ "train_runtime": 101969.0556,
+ "train_tokens_per_second": 5656.926
+ },
+ {
+ "epoch": 0.8880958793428494,
+ "grad_norm": 0.7662456138801742,
+ "learning_rate": 5.874894536786396e-06,
+ "loss": 0.3340578079223633,
+ "num_input_tokens_seen": 577044568,
+ "step": 13190,
+ "train_runtime": 102009.2868,
+ "train_tokens_per_second": 5656.785
+ },
+ {
+ "epoch": 0.8884325343388096,
+ "grad_norm": 0.7991035565617537,
+ "learning_rate": 5.8722911248394555e-06,
+ "loss": 0.33777599334716796,
+ "num_input_tokens_seen": 577243936,
+ "step": 13195,
+ "train_runtime": 102050.722,
+ "train_tokens_per_second": 5656.441
+ },
+ {
+ "epoch": 0.8887691893347698,
+ "grad_norm": 0.7706395477682242,
+ "learning_rate": 5.8696874689590824e-06,
+ "loss": 0.3437623977661133,
+ "num_input_tokens_seen": 577453888,
+ "step": 13200,
+ "train_runtime": 102083.4149,
+ "train_tokens_per_second": 5656.687
+ },
+ {
+ "epoch": 0.8891058443307299,
+ "grad_norm": 0.8134775525528564,
+ "learning_rate": 5.867083569873383e-06,
+ "loss": 0.3634350776672363,
+ "num_input_tokens_seen": 577679688,
+ "step": 13205,
+ "train_runtime": 102117.5448,
+ "train_tokens_per_second": 5657.007
+ },
+ {
+ "epoch": 0.8894424993266901,
+ "grad_norm": 0.7673260905685215,
+ "learning_rate": 5.8644794283105255e-06,
+ "loss": 0.36774277687072754,
+ "num_input_tokens_seen": 577900960,
+ "step": 13210,
+ "train_runtime": 102155.9766,
+ "train_tokens_per_second": 5657.045
+ },
+ {
+ "epoch": 0.8897791543226502,
+ "grad_norm": 0.7882719202387701,
+ "learning_rate": 5.861875044998755e-06,
+ "loss": 0.40239391326904295,
+ "num_input_tokens_seen": 578125704,
+ "step": 13215,
+ "train_runtime": 102190.4444,
+ "train_tokens_per_second": 5657.336
+ },
+ {
+ "epoch": 0.8901158093186103,
+ "grad_norm": 0.7563037927913291,
+ "learning_rate": 5.859270420666377e-06,
+ "loss": 0.36860146522521975,
+ "num_input_tokens_seen": 578343816,
+ "step": 13220,
+ "train_runtime": 102228.533,
+ "train_tokens_per_second": 5657.362
+ },
+ {
+ "epoch": 0.8904524643145705,
+ "grad_norm": 0.7756762347173053,
+ "learning_rate": 5.856665556041764e-06,
+ "loss": 0.3481595993041992,
+ "num_input_tokens_seen": 578548144,
+ "step": 13225,
+ "train_runtime": 102265.8667,
+ "train_tokens_per_second": 5657.295
+ },
+ {
+ "epoch": 0.8907891193105306,
+ "grad_norm": 0.7183886665375777,
+ "learning_rate": 5.8540604518533605e-06,
+ "loss": 0.3859154224395752,
+ "num_input_tokens_seen": 578772776,
+ "step": 13230,
+ "train_runtime": 102299.6045,
+ "train_tokens_per_second": 5657.625
+ },
+ {
+ "epoch": 0.8911257743064908,
+ "grad_norm": 1.4271805656006054,
+ "learning_rate": 5.851455108829675e-06,
+ "loss": 0.34519567489624026,
+ "num_input_tokens_seen": 578981024,
+ "step": 13235,
+ "train_runtime": 102340.8563,
+ "train_tokens_per_second": 5657.379
+ },
+ {
+ "epoch": 0.8914624293024509,
+ "grad_norm": 0.6789149460339803,
+ "learning_rate": 5.848849527699283e-06,
+ "loss": 0.3698782205581665,
+ "num_input_tokens_seen": 579201768,
+ "step": 13240,
+ "train_runtime": 102379.4065,
+ "train_tokens_per_second": 5657.405
+ },
+ {
+ "epoch": 0.891799084298411,
+ "grad_norm": 0.7639170132777864,
+ "learning_rate": 5.846243709190827e-06,
+ "loss": 0.35757324695587156,
+ "num_input_tokens_seen": 579419880,
+ "step": 13245,
+ "train_runtime": 102417.7033,
+ "train_tokens_per_second": 5657.419
+ },
+ {
+ "epoch": 0.8921357392943712,
+ "grad_norm": 0.8837562529973417,
+ "learning_rate": 5.843637654033019e-06,
+ "loss": 0.37553791999816893,
+ "num_input_tokens_seen": 579620760,
+ "step": 13250,
+ "train_runtime": 102459.8571,
+ "train_tokens_per_second": 5657.052
+ },
+ {
+ "epoch": 0.8924723942903313,
+ "grad_norm": 0.676305723344451,
+ "learning_rate": 5.841031362954629e-06,
+ "loss": 0.39575533866882323,
+ "num_input_tokens_seen": 579849048,
+ "step": 13255,
+ "train_runtime": 102495.0631,
+ "train_tokens_per_second": 5657.336
+ },
+ {
+ "epoch": 0.8928090492862915,
+ "grad_norm": 0.6628658100615732,
+ "learning_rate": 5.8384248366845e-06,
+ "loss": 0.37047085762023924,
+ "num_input_tokens_seen": 580077352,
+ "step": 13260,
+ "train_runtime": 102540.1881,
+ "train_tokens_per_second": 5657.073
+ },
+ {
+ "epoch": 0.8931457042822516,
+ "grad_norm": 0.6825809402512039,
+ "learning_rate": 5.8358180759515396e-06,
+ "loss": 0.37846667766571046,
+ "num_input_tokens_seen": 580288200,
+ "step": 13265,
+ "train_runtime": 102577.1371,
+ "train_tokens_per_second": 5657.091
+ },
+ {
+ "epoch": 0.8934823592782117,
+ "grad_norm": 0.7471824938587698,
+ "learning_rate": 5.833211081484719e-06,
+ "loss": 0.37510037422180176,
+ "num_input_tokens_seen": 580506904,
+ "step": 13270,
+ "train_runtime": 102618.782,
+ "train_tokens_per_second": 5656.926
+ },
+ {
+ "epoch": 0.8938190142741719,
+ "grad_norm": 0.641717053888179,
+ "learning_rate": 5.830603854013077e-06,
+ "loss": 0.3673901081085205,
+ "num_input_tokens_seen": 580723408,
+ "step": 13275,
+ "train_runtime": 102660.3376,
+ "train_tokens_per_second": 5656.746
+ },
+ {
+ "epoch": 0.894155669270132,
+ "grad_norm": 0.8523312033525736,
+ "learning_rate": 5.827996394265716e-06,
+ "loss": 0.35067110061645507,
+ "num_input_tokens_seen": 580935112,
+ "step": 13280,
+ "train_runtime": 102701.6639,
+ "train_tokens_per_second": 5656.531
+ },
+ {
+ "epoch": 0.8944923242660922,
+ "grad_norm": 0.7794670905441262,
+ "learning_rate": 5.825388702971802e-06,
+ "loss": 0.35003342628479006,
+ "num_input_tokens_seen": 581151304,
+ "step": 13285,
+ "train_runtime": 102733.2999,
+ "train_tokens_per_second": 5656.893
+ },
+ {
+ "epoch": 0.8948289792620523,
+ "grad_norm": 0.7697508084044369,
+ "learning_rate": 5.822780780860571e-06,
+ "loss": 0.36042060852050783,
+ "num_input_tokens_seen": 581367952,
+ "step": 13290,
+ "train_runtime": 102770.8878,
+ "train_tokens_per_second": 5656.932
+ },
+ {
+ "epoch": 0.8951656342580124,
+ "grad_norm": 0.7094962387963322,
+ "learning_rate": 5.820172628661317e-06,
+ "loss": 0.3688685178756714,
+ "num_input_tokens_seen": 581603552,
+ "step": 13295,
+ "train_runtime": 102804.4905,
+ "train_tokens_per_second": 5657.375
+ },
+ {
+ "epoch": 0.8955022892539726,
+ "grad_norm": 0.7277970714037896,
+ "learning_rate": 5.8175642471034045e-06,
+ "loss": 0.34978251457214354,
+ "num_input_tokens_seen": 581823928,
+ "step": 13300,
+ "train_runtime": 102838.1084,
+ "train_tokens_per_second": 5657.669
+ },
+ {
+ "epoch": 0.8958389442499327,
+ "grad_norm": 0.7417598590201532,
+ "learning_rate": 5.8149556369162565e-06,
+ "loss": 0.3728696346282959,
+ "num_input_tokens_seen": 582031800,
+ "step": 13305,
+ "train_runtime": 102877.0099,
+ "train_tokens_per_second": 5657.55
+ },
+ {
+ "epoch": 0.8961755992458929,
+ "grad_norm": 0.7192634087406454,
+ "learning_rate": 5.812346798829361e-06,
+ "loss": 0.37227132320404055,
+ "num_input_tokens_seen": 582257072,
+ "step": 13310,
+ "train_runtime": 102918.7234,
+ "train_tokens_per_second": 5657.446
+ },
+ {
+ "epoch": 0.896512254241853,
+ "grad_norm": 0.7396821495476227,
+ "learning_rate": 5.809737733572276e-06,
+ "loss": 0.35695157051086424,
+ "num_input_tokens_seen": 582470280,
+ "step": 13315,
+ "train_runtime": 102957.1094,
+ "train_tokens_per_second": 5657.407
+ },
+ {
+ "epoch": 0.8968489092378131,
+ "grad_norm": 0.8286698107403867,
+ "learning_rate": 5.8071284418746166e-06,
+ "loss": 0.3608556747436523,
+ "num_input_tokens_seen": 582689968,
+ "step": 13320,
+ "train_runtime": 102997.8183,
+ "train_tokens_per_second": 5657.304
+ },
+ {
+ "epoch": 0.8971855642337733,
+ "grad_norm": 0.7429942677291922,
+ "learning_rate": 5.80451892446606e-06,
+ "loss": 0.38565864562988283,
+ "num_input_tokens_seen": 582917944,
+ "step": 13325,
+ "train_runtime": 103036.6727,
+ "train_tokens_per_second": 5657.383
+ },
+ {
+ "epoch": 0.8975222192297334,
+ "grad_norm": 0.6807484219365846,
+ "learning_rate": 5.801909182076354e-06,
+ "loss": 0.3657494068145752,
+ "num_input_tokens_seen": 583143528,
+ "step": 13330,
+ "train_runtime": 103071.4117,
+ "train_tokens_per_second": 5657.665
+ },
+ {
+ "epoch": 0.8978588742256935,
+ "grad_norm": 0.7911812244430602,
+ "learning_rate": 5.7992992154353e-06,
+ "loss": 0.3965157985687256,
+ "num_input_tokens_seen": 583360120,
+ "step": 13335,
+ "train_runtime": 103106.3833,
+ "train_tokens_per_second": 5657.847
+ },
+ {
+ "epoch": 0.8981955292216537,
+ "grad_norm": 0.7134344977635636,
+ "learning_rate": 5.79668902527277e-06,
+ "loss": 0.3728632926940918,
+ "num_input_tokens_seen": 583576392,
+ "step": 13340,
+ "train_runtime": 103143.3779,
+ "train_tokens_per_second": 5657.914
+ },
+ {
+ "epoch": 0.8985321842176138,
+ "grad_norm": 0.7068010881578768,
+ "learning_rate": 5.794078612318695e-06,
+ "loss": 0.36152513027191163,
+ "num_input_tokens_seen": 583793888,
+ "step": 13345,
+ "train_runtime": 103187.6757,
+ "train_tokens_per_second": 5657.593
+ },
+ {
+ "epoch": 0.898868839213574,
+ "grad_norm": 1.018985024468395,
+ "learning_rate": 5.791467977303068e-06,
+ "loss": 0.3539185285568237,
+ "num_input_tokens_seen": 584007880,
+ "step": 13350,
+ "train_runtime": 103221.6694,
+ "train_tokens_per_second": 5657.803
+ },
+ {
+ "epoch": 0.8992054942095341,
+ "grad_norm": 0.8787085802082749,
+ "learning_rate": 5.788857120955945e-06,
+ "loss": 0.3753284692764282,
+ "num_input_tokens_seen": 584222720,
+ "step": 13355,
+ "train_runtime": 103259.3517,
+ "train_tokens_per_second": 5657.819
+ },
+ {
+ "epoch": 0.8995421492054942,
+ "grad_norm": 0.7526792175486189,
+ "learning_rate": 5.786246044007443e-06,
+ "loss": 0.3408811092376709,
+ "num_input_tokens_seen": 584430032,
+ "step": 13360,
+ "train_runtime": 103303.2518,
+ "train_tokens_per_second": 5657.421
+ },
+ {
+ "epoch": 0.8998788042014544,
+ "grad_norm": 0.7042471293534047,
+ "learning_rate": 5.783634747187743e-06,
+ "loss": 0.38433132171630857,
+ "num_input_tokens_seen": 584657272,
+ "step": 13365,
+ "train_runtime": 103336.5354,
+ "train_tokens_per_second": 5657.798
+ },
+ {
+ "epoch": 0.9002154591974145,
+ "grad_norm": 0.7283278180613847,
+ "learning_rate": 5.781023231227084e-06,
+ "loss": 0.3663713693618774,
+ "num_input_tokens_seen": 584873064,
+ "step": 13370,
+ "train_runtime": 103383.5959,
+ "train_tokens_per_second": 5657.31
+ },
+ {
+ "epoch": 0.9005521141933747,
+ "grad_norm": 0.6987374788445736,
+ "learning_rate": 5.778411496855769e-06,
+ "loss": 0.3779282093048096,
+ "num_input_tokens_seen": 585095664,
+ "step": 13375,
+ "train_runtime": 103422.1497,
+ "train_tokens_per_second": 5657.354
+ },
+ {
+ "epoch": 0.9008887691893348,
+ "grad_norm": 0.7973255027303662,
+ "learning_rate": 5.775799544804161e-06,
+ "loss": 0.3911623477935791,
+ "num_input_tokens_seen": 585323224,
+ "step": 13380,
+ "train_runtime": 103458.2471,
+ "train_tokens_per_second": 5657.579
+ },
+ {
+ "epoch": 0.9012254241852949,
+ "grad_norm": 0.7376867678943736,
+ "learning_rate": 5.7731873758026845e-06,
+ "loss": 0.3567417860031128,
+ "num_input_tokens_seen": 585539800,
+ "step": 13385,
+ "train_runtime": 103499.6202,
+ "train_tokens_per_second": 5657.41
+ },
+ {
+ "epoch": 0.9015620791812551,
+ "grad_norm": 0.8164983344754712,
+ "learning_rate": 5.770574990581823e-06,
+ "loss": 0.3617391109466553,
+ "num_input_tokens_seen": 585756160,
+ "step": 13390,
+ "train_runtime": 103538.5541,
+ "train_tokens_per_second": 5657.372
+ },
+ {
+ "epoch": 0.9018987341772152,
+ "grad_norm": 0.7386008524112068,
+ "learning_rate": 5.767962389872124e-06,
+ "loss": 0.34462101459503175,
+ "num_input_tokens_seen": 585972312,
+ "step": 13395,
+ "train_runtime": 103578.7054,
+ "train_tokens_per_second": 5657.266
+ },
+ {
+ "epoch": 0.9022353891731754,
+ "grad_norm": 0.7749458765170661,
+ "learning_rate": 5.765349574404191e-06,
+ "loss": 0.3628375053405762,
+ "num_input_tokens_seen": 586175624,
+ "step": 13400,
+ "train_runtime": 103611.757,
+ "train_tokens_per_second": 5657.424
+ },
+ {
+ "epoch": 0.9025720441691355,
+ "grad_norm": 0.7292901746189114,
+ "learning_rate": 5.7627365449086905e-06,
+ "loss": 0.364549446105957,
+ "num_input_tokens_seen": 586370072,
+ "step": 13405,
+ "train_runtime": 103645.7307,
+ "train_tokens_per_second": 5657.445
+ },
+ {
+ "epoch": 0.9029086991650956,
+ "grad_norm": 0.7620757507465177,
+ "learning_rate": 5.760123302116345e-06,
+ "loss": 0.3901118040084839,
+ "num_input_tokens_seen": 586593560,
+ "step": 13410,
+ "train_runtime": 103677.957,
+ "train_tokens_per_second": 5657.843
+ },
+ {
+ "epoch": 0.9032453541610558,
+ "grad_norm": 0.7598225601483857,
+ "learning_rate": 5.757509846757945e-06,
+ "loss": 0.36999359130859377,
+ "num_input_tokens_seen": 586819408,
+ "step": 13415,
+ "train_runtime": 103716.3963,
+ "train_tokens_per_second": 5657.923
+ },
+ {
+ "epoch": 0.9035820091570159,
+ "grad_norm": 1.147309642657896,
+ "learning_rate": 5.75489617956433e-06,
+ "loss": 0.3971453428268433,
+ "num_input_tokens_seen": 587040712,
+ "step": 13420,
+ "train_runtime": 103762.395,
+ "train_tokens_per_second": 5657.548
+ },
+ {
+ "epoch": 0.9039186641529761,
+ "grad_norm": 0.6772540219988885,
+ "learning_rate": 5.752282301266406e-06,
+ "loss": 0.37274601459503176,
+ "num_input_tokens_seen": 587260024,
+ "step": 13425,
+ "train_runtime": 103799.3896,
+ "train_tokens_per_second": 5657.644
+ },
+ {
+ "epoch": 0.9042553191489362,
+ "grad_norm": 0.6865687048122345,
+ "learning_rate": 5.749668212595136e-06,
+ "loss": 0.36958420276641846,
+ "num_input_tokens_seen": 587475152,
+ "step": 13430,
+ "train_runtime": 103838.2234,
+ "train_tokens_per_second": 5657.6
+ },
+ {
+ "epoch": 0.9045919741448963,
+ "grad_norm": 0.7480660958831125,
+ "learning_rate": 5.747053914281539e-06,
+ "loss": 0.36184403896331785,
+ "num_input_tokens_seen": 587679656,
+ "step": 13435,
+ "train_runtime": 103881.14,
+ "train_tokens_per_second": 5657.231
+ },
+ {
+ "epoch": 0.9049286291408565,
+ "grad_norm": 0.7323661070573674,
+ "learning_rate": 5.744439407056699e-06,
+ "loss": 0.3511162757873535,
+ "num_input_tokens_seen": 587888648,
+ "step": 13440,
+ "train_runtime": 103918.9245,
+ "train_tokens_per_second": 5657.186
+ },
+ {
+ "epoch": 0.9052652841368166,
+ "grad_norm": 0.6204712137626076,
+ "learning_rate": 5.741824691651752e-06,
+ "loss": 0.3408759832382202,
+ "num_input_tokens_seen": 588114264,
+ "step": 13445,
+ "train_runtime": 103954.7773,
+ "train_tokens_per_second": 5657.405
+ },
+ {
+ "epoch": 0.9056019391327768,
+ "grad_norm": 0.8116090187953918,
+ "learning_rate": 5.739209768797896e-06,
+ "loss": 0.36322770118713377,
+ "num_input_tokens_seen": 588322904,
+ "step": 13450,
+ "train_runtime": 103991.6356,
+ "train_tokens_per_second": 5657.406
+ },
+ {
+ "epoch": 0.9059385941287369,
+ "grad_norm": 0.7146086921110422,
+ "learning_rate": 5.736594639226385e-06,
+ "loss": 0.32691378593444825,
+ "num_input_tokens_seen": 588542624,
+ "step": 13455,
+ "train_runtime": 104026.9412,
+ "train_tokens_per_second": 5657.598
+ },
+ {
+ "epoch": 0.906275249124697,
+ "grad_norm": 0.6758364668603787,
+ "learning_rate": 5.7339793036685306e-06,
+ "loss": 0.3772242546081543,
+ "num_input_tokens_seen": 588766448,
+ "step": 13460,
+ "train_runtime": 104068.0394,
+ "train_tokens_per_second": 5657.515
+ },
+ {
+ "epoch": 0.9066119041206572,
+ "grad_norm": 0.7332497573577879,
+ "learning_rate": 5.731363762855706e-06,
+ "loss": 0.3471216678619385,
+ "num_input_tokens_seen": 588981896,
+ "step": 13465,
+ "train_runtime": 104109.0854,
+ "train_tokens_per_second": 5657.353
+ },
+ {
+ "epoch": 0.9069485591166173,
+ "grad_norm": 0.8052750996260324,
+ "learning_rate": 5.728748017519337e-06,
+ "loss": 0.3998070001602173,
+ "num_input_tokens_seen": 589199016,
+ "step": 13470,
+ "train_runtime": 104146.4665,
+ "train_tokens_per_second": 5657.408
+ },
+ {
+ "epoch": 0.9072852141125775,
+ "grad_norm": 0.7072206601280142,
+ "learning_rate": 5.726132068390908e-06,
+ "loss": 0.3687425136566162,
+ "num_input_tokens_seen": 589402952,
+ "step": 13475,
+ "train_runtime": 104180.5112,
+ "train_tokens_per_second": 5657.516
+ },
+ {
+ "epoch": 0.9076218691085376,
+ "grad_norm": 0.7386177069645364,
+ "learning_rate": 5.723515916201962e-06,
+ "loss": 0.4100666046142578,
+ "num_input_tokens_seen": 589630440,
+ "step": 13480,
+ "train_runtime": 104218.7952,
+ "train_tokens_per_second": 5657.621
+ },
+ {
+ "epoch": 0.9079585241044977,
+ "grad_norm": 0.6255645042824444,
+ "learning_rate": 5.720899561684098e-06,
+ "loss": 0.3434640645980835,
+ "num_input_tokens_seen": 589849368,
+ "step": 13485,
+ "train_runtime": 104262.0828,
+ "train_tokens_per_second": 5657.372
+ },
+ {
+ "epoch": 0.9082951791004579,
+ "grad_norm": 0.7488072842696437,
+ "learning_rate": 5.71828300556897e-06,
+ "loss": 0.36190190315246584,
+ "num_input_tokens_seen": 590075552,
+ "step": 13490,
+ "train_runtime": 104304.1746,
+ "train_tokens_per_second": 5657.257
+ },
+ {
+ "epoch": 0.908631834096418,
+ "grad_norm": 0.595597879983287,
+ "learning_rate": 5.715666248588289e-06,
+ "loss": 0.3661900520324707,
+ "num_input_tokens_seen": 590300848,
+ "step": 13495,
+ "train_runtime": 104350.0273,
+ "train_tokens_per_second": 5656.93
+ },
+ {
+ "epoch": 0.9089684890923782,
+ "grad_norm": 0.7843852676552852,
+ "learning_rate": 5.713049291473825e-06,
+ "loss": 0.39053096771240237,
+ "num_input_tokens_seen": 590529208,
+ "step": 13500,
+ "train_runtime": 104390.3684,
+ "train_tokens_per_second": 5656.932
+ },
+ {
+ "epoch": 0.9093051440883383,
+ "grad_norm": 0.7754635800602863,
+ "learning_rate": 5.710432134957399e-06,
+ "loss": 0.3421627998352051,
+ "num_input_tokens_seen": 590729424,
+ "step": 13505,
+ "train_runtime": 104437.8595,
+ "train_tokens_per_second": 5656.277
+ },
+ {
+ "epoch": 0.9096417990842984,
+ "grad_norm": 0.7329858480957279,
+ "learning_rate": 5.707814779770892e-06,
+ "loss": 0.34261083602905273,
+ "num_input_tokens_seen": 590940928,
+ "step": 13510,
+ "train_runtime": 104484.5885,
+ "train_tokens_per_second": 5655.771
+ },
+ {
+ "epoch": 0.9099784540802586,
+ "grad_norm": 0.7345589839622834,
+ "learning_rate": 5.7051972266462395e-06,
+ "loss": 0.35787391662597656,
+ "num_input_tokens_seen": 591158192,
+ "step": 13515,
+ "train_runtime": 104521.8258,
+ "train_tokens_per_second": 5655.835
+ },
+ {
+ "epoch": 0.9103151090762187,
+ "grad_norm": 0.8008449716401149,
+ "learning_rate": 5.702579476315431e-06,
+ "loss": 0.3310979127883911,
+ "num_input_tokens_seen": 591376656,
+ "step": 13520,
+ "train_runtime": 104566.7223,
+ "train_tokens_per_second": 5655.496
+ },
+ {
+ "epoch": 0.9106517640721788,
+ "grad_norm": 0.7423340401003228,
+ "learning_rate": 5.699961529510513e-06,
+ "loss": 0.3578958988189697,
+ "num_input_tokens_seen": 591581336,
+ "step": 13525,
+ "train_runtime": 104607.1051,
+ "train_tokens_per_second": 5655.269
+ },
+ {
+ "epoch": 0.910988419068139,
+ "grad_norm": 0.7857402111034575,
+ "learning_rate": 5.697343386963586e-06,
+ "loss": 0.3766601800918579,
+ "num_input_tokens_seen": 591797696,
+ "step": 13530,
+ "train_runtime": 104651.9901,
+ "train_tokens_per_second": 5654.911
+ },
+ {
+ "epoch": 0.9113250740640991,
+ "grad_norm": 0.6657292352842593,
+ "learning_rate": 5.694725049406801e-06,
+ "loss": 0.3710474967956543,
+ "num_input_tokens_seen": 592014888,
+ "step": 13535,
+ "train_runtime": 104697.3063,
+ "train_tokens_per_second": 5654.538
+ },
+ {
+ "epoch": 0.9116617290600593,
+ "grad_norm": 0.6966081276914712,
+ "learning_rate": 5.692106517572373e-06,
+ "loss": 0.38451037406921384,
+ "num_input_tokens_seen": 592229672,
+ "step": 13540,
+ "train_runtime": 104731.9456,
+ "train_tokens_per_second": 5654.719
+ },
+ {
+ "epoch": 0.9119983840560194,
+ "grad_norm": 0.7724805146942423,
+ "learning_rate": 5.689487792192566e-06,
+ "loss": 0.3990262746810913,
+ "num_input_tokens_seen": 592451696,
+ "step": 13545,
+ "train_runtime": 104773.5682,
+ "train_tokens_per_second": 5654.591
+ },
+ {
+ "epoch": 0.9123350390519795,
+ "grad_norm": 0.7342026376072939,
+ "learning_rate": 5.686868873999696e-06,
+ "loss": 0.3797611236572266,
+ "num_input_tokens_seen": 592668960,
+ "step": 13550,
+ "train_runtime": 104816.4545,
+ "train_tokens_per_second": 5654.35
+ },
+ {
+ "epoch": 0.9126716940479397,
+ "grad_norm": 0.666648695750785,
+ "learning_rate": 5.684249763726135e-06,
+ "loss": 0.3764026641845703,
+ "num_input_tokens_seen": 592890144,
+ "step": 13555,
+ "train_runtime": 104853.017,
+ "train_tokens_per_second": 5654.488
+ },
+ {
+ "epoch": 0.9130083490438998,
+ "grad_norm": 0.796183302090131,
+ "learning_rate": 5.681630462104308e-06,
+ "loss": 0.3508880615234375,
+ "num_input_tokens_seen": 593092200,
+ "step": 13560,
+ "train_runtime": 104890.6314,
+ "train_tokens_per_second": 5654.387
+ },
+ {
+ "epoch": 0.91334500403986,
+ "grad_norm": 0.6941135923026022,
+ "learning_rate": 5.679010969866697e-06,
+ "loss": 0.3628666877746582,
+ "num_input_tokens_seen": 593304400,
+ "step": 13565,
+ "train_runtime": 104927.8794,
+ "train_tokens_per_second": 5654.402
+ },
+ {
+ "epoch": 0.9136816590358201,
+ "grad_norm": 0.7543532537354826,
+ "learning_rate": 5.676391287745835e-06,
+ "loss": 0.35071396827697754,
+ "num_input_tokens_seen": 593537816,
+ "step": 13570,
+ "train_runtime": 104959.059,
+ "train_tokens_per_second": 5654.946
+ },
+ {
+ "epoch": 0.9140183140317802,
+ "grad_norm": 0.6731321122510332,
+ "learning_rate": 5.673771416474305e-06,
+ "loss": 0.3439845085144043,
+ "num_input_tokens_seen": 593761520,
+ "step": 13575,
+ "train_runtime": 104997.0967,
+ "train_tokens_per_second": 5655.028
+ },
+ {
+ "epoch": 0.9143549690277404,
+ "grad_norm": 0.7587431284561748,
+ "learning_rate": 5.671151356784747e-06,
+ "loss": 0.3460673809051514,
+ "num_input_tokens_seen": 593970768,
+ "step": 13580,
+ "train_runtime": 105034.4592,
+ "train_tokens_per_second": 5655.009
+ },
+ {
+ "epoch": 0.9146916240237005,
+ "grad_norm": 0.6614026194133914,
+ "learning_rate": 5.668531109409851e-06,
+ "loss": 0.3715926170349121,
+ "num_input_tokens_seen": 594193064,
+ "step": 13585,
+ "train_runtime": 105071.7331,
+ "train_tokens_per_second": 5655.118
+ },
+ {
+ "epoch": 0.9150282790196607,
+ "grad_norm": 0.7852840297073502,
+ "learning_rate": 5.665910675082362e-06,
+ "loss": 0.36374545097351074,
+ "num_input_tokens_seen": 594409288,
+ "step": 13590,
+ "train_runtime": 105109.8348,
+ "train_tokens_per_second": 5655.125
+ },
+ {
+ "epoch": 0.9153649340156208,
+ "grad_norm": 0.7103415655589641,
+ "learning_rate": 5.663290054535078e-06,
+ "loss": 0.34845700263977053,
+ "num_input_tokens_seen": 594620336,
+ "step": 13595,
+ "train_runtime": 105150.7645,
+ "train_tokens_per_second": 5654.931
+ },
+ {
+ "epoch": 0.9157015890115809,
+ "grad_norm": 0.8295807713349123,
+ "learning_rate": 5.660669248500846e-06,
+ "loss": 0.39510331153869627,
+ "num_input_tokens_seen": 594838312,
+ "step": 13600,
+ "train_runtime": 105189.0392,
+ "train_tokens_per_second": 5654.946
+ },
+ {
+ "epoch": 0.9160382440075411,
+ "grad_norm": 0.7158029671235298,
+ "learning_rate": 5.658048257712563e-06,
+ "loss": 0.3466895818710327,
+ "num_input_tokens_seen": 595060816,
+ "step": 13605,
+ "train_runtime": 105234.3634,
+ "train_tokens_per_second": 5654.625
+ },
+ {
+ "epoch": 0.9163748990035012,
+ "grad_norm": 0.7282803359887597,
+ "learning_rate": 5.655427082903184e-06,
+ "loss": 0.3633090019226074,
+ "num_input_tokens_seen": 595279584,
+ "step": 13610,
+ "train_runtime": 105277.1838,
+ "train_tokens_per_second": 5654.403
+ },
+ {
+ "epoch": 0.9167115539994614,
+ "grad_norm": 0.6738750845313296,
+ "learning_rate": 5.652805724805711e-06,
+ "loss": 0.35727887153625487,
+ "num_input_tokens_seen": 595499192,
+ "step": 13615,
+ "train_runtime": 105314.5327,
+ "train_tokens_per_second": 5654.483
+ },
+ {
+ "epoch": 0.9170482089954215,
+ "grad_norm": 0.8264756955958018,
+ "learning_rate": 5.650184184153199e-06,
+ "loss": 0.3687946319580078,
+ "num_input_tokens_seen": 595723264,
+ "step": 13620,
+ "train_runtime": 105352.1128,
+ "train_tokens_per_second": 5654.592
+ },
+ {
+ "epoch": 0.9173848639913816,
+ "grad_norm": 0.7363635027786596,
+ "learning_rate": 5.647562461678754e-06,
+ "loss": 0.39588251113891604,
+ "num_input_tokens_seen": 595947688,
+ "step": 13625,
+ "train_runtime": 105388.882,
+ "train_tokens_per_second": 5654.749
+ },
+ {
+ "epoch": 0.9177215189873418,
+ "grad_norm": 0.7671112595712233,
+ "learning_rate": 5.64494055811553e-06,
+ "loss": 0.35439510345458985,
+ "num_input_tokens_seen": 596167280,
+ "step": 13630,
+ "train_runtime": 105432.2515,
+ "train_tokens_per_second": 5654.506
+ },
+ {
+ "epoch": 0.9180581739833019,
+ "grad_norm": 0.7391097804364258,
+ "learning_rate": 5.642318474196737e-06,
+ "loss": 0.3450098991394043,
+ "num_input_tokens_seen": 596389280,
+ "step": 13635,
+ "train_runtime": 105470.2709,
+ "train_tokens_per_second": 5654.572
+ },
+ {
+ "epoch": 0.9183948289792621,
+ "grad_norm": 0.7394314350713648,
+ "learning_rate": 5.63969621065563e-06,
+ "loss": 0.36347208023071287,
+ "num_input_tokens_seen": 596603992,
+ "step": 13640,
+ "train_runtime": 105510.1258,
+ "train_tokens_per_second": 5654.471
+ },
+ {
+ "epoch": 0.9187314839752222,
+ "grad_norm": 0.6523195737553367,
+ "learning_rate": 5.637073768225517e-06,
+ "loss": 0.3404127836227417,
+ "num_input_tokens_seen": 596814560,
+ "step": 13645,
+ "train_runtime": 105555.5698,
+ "train_tokens_per_second": 5654.032
+ },
+ {
+ "epoch": 0.9190681389711823,
+ "grad_norm": 0.7286950960994242,
+ "learning_rate": 5.634451147639758e-06,
+ "loss": 0.3368164300918579,
+ "num_input_tokens_seen": 597016352,
+ "step": 13650,
+ "train_runtime": 105596.0923,
+ "train_tokens_per_second": 5653.773
+ },
+ {
+ "epoch": 0.9194047939671425,
+ "grad_norm": 0.7381840389633337,
+ "learning_rate": 5.631828349631757e-06,
+ "loss": 0.34926304817199705,
+ "num_input_tokens_seen": 597241424,
+ "step": 13655,
+ "train_runtime": 105633.1844,
+ "train_tokens_per_second": 5653.919
+ },
+ {
+ "epoch": 0.9197414489631026,
+ "grad_norm": 0.7377168394438081,
+ "learning_rate": 5.629205374934974e-06,
+ "loss": 0.3418566703796387,
+ "num_input_tokens_seen": 597449976,
+ "step": 13660,
+ "train_runtime": 105670.7222,
+ "train_tokens_per_second": 5653.884
+ },
+ {
+ "epoch": 0.9200781039590628,
+ "grad_norm": 0.7062836493823704,
+ "learning_rate": 5.626582224282913e-06,
+ "loss": 0.36661295890808104,
+ "num_input_tokens_seen": 597675264,
+ "step": 13665,
+ "train_runtime": 105714.0467,
+ "train_tokens_per_second": 5653.698
+ },
+ {
+ "epoch": 0.9204147589550229,
+ "grad_norm": 0.7520494279886026,
+ "learning_rate": 5.623958898409132e-06,
+ "loss": 0.39702858924865725,
+ "num_input_tokens_seen": 597900752,
+ "step": 13670,
+ "train_runtime": 105743.9982,
+ "train_tokens_per_second": 5654.229
+ },
+ {
+ "epoch": 0.920751413950983,
+ "grad_norm": 0.6987426458318066,
+ "learning_rate": 5.621335398047234e-06,
+ "loss": 0.3848641157150269,
+ "num_input_tokens_seen": 598115680,
+ "step": 13675,
+ "train_runtime": 105787.0252,
+ "train_tokens_per_second": 5653.961
+ },
+ {
+ "epoch": 0.9210880689469432,
+ "grad_norm": 0.7307345712178913,
+ "learning_rate": 5.618711723930874e-06,
+ "loss": 0.36865854263305664,
+ "num_input_tokens_seen": 598333664,
+ "step": 13680,
+ "train_runtime": 105817.9321,
+ "train_tokens_per_second": 5654.369
+ },
+ {
+ "epoch": 0.9214247239429033,
+ "grad_norm": 0.7963356189127992,
+ "learning_rate": 5.616087876793753e-06,
+ "loss": 0.3818777561187744,
+ "num_input_tokens_seen": 598560984,
+ "step": 13685,
+ "train_runtime": 105853.2283,
+ "train_tokens_per_second": 5654.631
+ },
+ {
+ "epoch": 0.9217613789388635,
+ "grad_norm": 0.6658404353690085,
+ "learning_rate": 5.613463857369621e-06,
+ "loss": 0.3648639678955078,
+ "num_input_tokens_seen": 598784192,
+ "step": 13690,
+ "train_runtime": 105895.3239,
+ "train_tokens_per_second": 5654.491
+ },
+ {
+ "epoch": 0.9220980339348236,
+ "grad_norm": 0.7245035749434348,
+ "learning_rate": 5.610839666392278e-06,
+ "loss": 0.35527966022491453,
+ "num_input_tokens_seen": 599005456,
+ "step": 13695,
+ "train_runtime": 105941.0406,
+ "train_tokens_per_second": 5654.14
+ },
+ {
+ "epoch": 0.9224346889307837,
+ "grad_norm": 0.7564271755937382,
+ "learning_rate": 5.608215304595571e-06,
+ "loss": 0.3818879842758179,
+ "num_input_tokens_seen": 599222856,
+ "step": 13700,
+ "train_runtime": 105987.7905,
+ "train_tokens_per_second": 5653.697
+ },
+ {
+ "epoch": 0.9227713439267439,
+ "grad_norm": 0.71469916144582,
+ "learning_rate": 5.605590772713393e-06,
+ "loss": 0.35137481689453126,
+ "num_input_tokens_seen": 599421784,
+ "step": 13705,
+ "train_runtime": 106025.2333,
+ "train_tokens_per_second": 5653.577
+ },
+ {
+ "epoch": 0.923107998922704,
+ "grad_norm": 0.8680683000673615,
+ "learning_rate": 5.602966071479687e-06,
+ "loss": 0.3719889402389526,
+ "num_input_tokens_seen": 599626912,
+ "step": 13710,
+ "train_runtime": 106057.3995,
+ "train_tokens_per_second": 5653.796
+ },
+ {
+ "epoch": 0.9234446539186641,
+ "grad_norm": 0.7763025193380707,
+ "learning_rate": 5.600341201628439e-06,
+ "loss": 0.37358889579772947,
+ "num_input_tokens_seen": 599854192,
+ "step": 13715,
+ "train_runtime": 106090.67,
+ "train_tokens_per_second": 5654.165
+ },
+ {
+ "epoch": 0.9237813089146243,
+ "grad_norm": 0.6656080663010198,
+ "learning_rate": 5.597716163893692e-06,
+ "loss": 0.31069388389587405,
+ "num_input_tokens_seen": 600085080,
+ "step": 13720,
+ "train_runtime": 106134.8208,
+ "train_tokens_per_second": 5653.989
+ },
+ {
+ "epoch": 0.9241179639105844,
+ "grad_norm": 0.7327001774115326,
+ "learning_rate": 5.595090959009525e-06,
+ "loss": 0.34327847957611085,
+ "num_input_tokens_seen": 600310096,
+ "step": 13725,
+ "train_runtime": 106180.1722,
+ "train_tokens_per_second": 5653.693
+ },
+ {
+ "epoch": 0.9244546189065446,
+ "grad_norm": 0.6125120195863898,
+ "learning_rate": 5.592465587710068e-06,
+ "loss": 0.34142630100250243,
+ "num_input_tokens_seen": 600545032,
+ "step": 13730,
+ "train_runtime": 106218.5007,
+ "train_tokens_per_second": 5653.865
+ },
+ {
+ "epoch": 0.9247912739025047,
+ "grad_norm": 0.6167159170504363,
+ "learning_rate": 5.589840050729498e-06,
+ "loss": 0.3286783218383789,
+ "num_input_tokens_seen": 600768064,
+ "step": 13735,
+ "train_runtime": 106258.0895,
+ "train_tokens_per_second": 5653.857
+ },
+ {
+ "epoch": 0.9251279288984648,
+ "grad_norm": 0.7881839614694872,
+ "learning_rate": 5.587214348802039e-06,
+ "loss": 0.3909108638763428,
+ "num_input_tokens_seen": 600980632,
+ "step": 13740,
+ "train_runtime": 106298.4773,
+ "train_tokens_per_second": 5653.709
+ },
+ {
+ "epoch": 0.925464583894425,
+ "grad_norm": 0.7022898228386897,
+ "learning_rate": 5.5845884826619615e-06,
+ "loss": 0.36343495845794677,
+ "num_input_tokens_seen": 601207448,
+ "step": 13745,
+ "train_runtime": 106333.7277,
+ "train_tokens_per_second": 5653.968
+ },
+ {
+ "epoch": 0.9258012388903851,
+ "grad_norm": 0.6378832684361027,
+ "learning_rate": 5.5819624530435775e-06,
+ "loss": 0.3585641622543335,
+ "num_input_tokens_seen": 601419704,
+ "step": 13750,
+ "train_runtime": 106373.8251,
+ "train_tokens_per_second": 5653.832
+ },
+ {
+ "epoch": 0.9261378938863453,
+ "grad_norm": 0.7242428710010252,
+ "learning_rate": 5.57933626068125e-06,
+ "loss": 0.3933718681335449,
+ "num_input_tokens_seen": 601639824,
+ "step": 13755,
+ "train_runtime": 106414.6703,
+ "train_tokens_per_second": 5653.73
+ },
+ {
+ "epoch": 0.9264745488823054,
+ "grad_norm": 0.7473031695779977,
+ "learning_rate": 5.5767099063093835e-06,
+ "loss": 0.3709197998046875,
+ "num_input_tokens_seen": 601868536,
+ "step": 13760,
+ "train_runtime": 106452.2637,
+ "train_tokens_per_second": 5653.882
+ },
+ {
+ "epoch": 0.9268112038782655,
+ "grad_norm": 0.803164002917514,
+ "learning_rate": 5.574083390662431e-06,
+ "loss": 0.373383617401123,
+ "num_input_tokens_seen": 602098696,
+ "step": 13765,
+ "train_runtime": 106489.0364,
+ "train_tokens_per_second": 5654.091
+ },
+ {
+ "epoch": 0.9271478588742257,
+ "grad_norm": 0.7108278884261701,
+ "learning_rate": 5.57145671447489e-06,
+ "loss": 0.3655261993408203,
+ "num_input_tokens_seen": 602302480,
+ "step": 13770,
+ "train_runtime": 106525.8625,
+ "train_tokens_per_second": 5654.049
+ },
+ {
+ "epoch": 0.9274845138701858,
+ "grad_norm": 0.7613900230491544,
+ "learning_rate": 5.5688298784813e-06,
+ "loss": 0.3984712600708008,
+ "num_input_tokens_seen": 602522208,
+ "step": 13775,
+ "train_runtime": 106564.2964,
+ "train_tokens_per_second": 5654.072
+ },
+ {
+ "epoch": 0.927821168866146,
+ "grad_norm": 0.6708141182451115,
+ "learning_rate": 5.566202883416249e-06,
+ "loss": 0.3970597743988037,
+ "num_input_tokens_seen": 602750424,
+ "step": 13780,
+ "train_runtime": 106606.7053,
+ "train_tokens_per_second": 5653.964
+ },
+ {
+ "epoch": 0.9281578238621061,
+ "grad_norm": 0.7821647487746413,
+ "learning_rate": 5.563575730014366e-06,
+ "loss": 0.39171276092529295,
+ "num_input_tokens_seen": 602965320,
+ "step": 13785,
+ "train_runtime": 106643.4726,
+ "train_tokens_per_second": 5654.029
+ },
+ {
+ "epoch": 0.9284944788580662,
+ "grad_norm": 0.7438040498615117,
+ "learning_rate": 5.560948419010328e-06,
+ "loss": 0.3737471342086792,
+ "num_input_tokens_seen": 603186136,
+ "step": 13790,
+ "train_runtime": 106681.291,
+ "train_tokens_per_second": 5654.095
+ },
+ {
+ "epoch": 0.9288311338540264,
+ "grad_norm": 0.7446914306967766,
+ "learning_rate": 5.558320951138853e-06,
+ "loss": 0.34050416946411133,
+ "num_input_tokens_seen": 603397288,
+ "step": 13795,
+ "train_runtime": 106722.5651,
+ "train_tokens_per_second": 5653.887
+ },
+ {
+ "epoch": 0.9291677888499865,
+ "grad_norm": 0.7759821070900785,
+ "learning_rate": 5.5556933271347046e-06,
+ "loss": 0.3723226308822632,
+ "num_input_tokens_seen": 603606088,
+ "step": 13800,
+ "train_runtime": 106763.5139,
+ "train_tokens_per_second": 5653.674
+ },
+ {
+ "epoch": 0.9295044438459467,
+ "grad_norm": 0.7098216673132232,
+ "learning_rate": 5.55306554773269e-06,
+ "loss": 0.3821061611175537,
+ "num_input_tokens_seen": 603816144,
+ "step": 13805,
+ "train_runtime": 106805.761,
+ "train_tokens_per_second": 5653.404
+ },
+ {
+ "epoch": 0.9298410988419068,
+ "grad_norm": 0.9275669407149835,
+ "learning_rate": 5.550437613667658e-06,
+ "loss": 0.3593113660812378,
+ "num_input_tokens_seen": 604004552,
+ "step": 13810,
+ "train_runtime": 106836.865,
+ "train_tokens_per_second": 5653.522
+ },
+ {
+ "epoch": 0.9301777538378669,
+ "grad_norm": 0.6896177858346889,
+ "learning_rate": 5.547809525674501e-06,
+ "loss": 0.35454933643341063,
+ "num_input_tokens_seen": 604212440,
+ "step": 13815,
+ "train_runtime": 106875.1406,
+ "train_tokens_per_second": 5653.442
+ },
+ {
+ "epoch": 0.9305144088338271,
+ "grad_norm": 0.7927279590495904,
+ "learning_rate": 5.545181284488159e-06,
+ "loss": 0.38926100730895996,
+ "num_input_tokens_seen": 604439792,
+ "step": 13820,
+ "train_runtime": 106917.4476,
+ "train_tokens_per_second": 5653.332
+ },
+ {
+ "epoch": 0.9308510638297872,
+ "grad_norm": 0.8386569241979678,
+ "learning_rate": 5.5425528908436085e-06,
+ "loss": 0.35577256679534913,
+ "num_input_tokens_seen": 604643600,
+ "step": 13825,
+ "train_runtime": 106953.2903,
+ "train_tokens_per_second": 5653.343
+ },
+ {
+ "epoch": 0.9311877188257474,
+ "grad_norm": 0.7592336539171605,
+ "learning_rate": 5.539924345475873e-06,
+ "loss": 0.3907893180847168,
+ "num_input_tokens_seen": 604839736,
+ "step": 13830,
+ "train_runtime": 106988.9857,
+ "train_tokens_per_second": 5653.29
+ },
+ {
+ "epoch": 0.9315243738217075,
+ "grad_norm": 0.8126152719075438,
+ "learning_rate": 5.5372956491200145e-06,
+ "loss": 0.37813870906829833,
+ "num_input_tokens_seen": 605051728,
+ "step": 13835,
+ "train_runtime": 107025.6706,
+ "train_tokens_per_second": 5653.333
+ },
+ {
+ "epoch": 0.9318610288176676,
+ "grad_norm": 0.7879475237888696,
+ "learning_rate": 5.534666802511143e-06,
+ "loss": 0.36364994049072263,
+ "num_input_tokens_seen": 605251488,
+ "step": 13840,
+ "train_runtime": 107062.9642,
+ "train_tokens_per_second": 5653.229
+ },
+ {
+ "epoch": 0.9321976838136278,
+ "grad_norm": 0.6977984667014386,
+ "learning_rate": 5.532037806384404e-06,
+ "loss": 0.372878885269165,
+ "num_input_tokens_seen": 605478256,
+ "step": 13845,
+ "train_runtime": 107101.7476,
+ "train_tokens_per_second": 5653.299
+ },
+ {
+ "epoch": 0.9325343388095879,
+ "grad_norm": 0.7377782721135303,
+ "learning_rate": 5.529408661474989e-06,
+ "loss": 0.33782191276550294,
+ "num_input_tokens_seen": 605678152,
+ "step": 13850,
+ "train_runtime": 107140.4741,
+ "train_tokens_per_second": 5653.122
+ },
+ {
+ "epoch": 0.932870993805548,
+ "grad_norm": 0.7120793619791788,
+ "learning_rate": 5.526779368518129e-06,
+ "loss": 0.3900485038757324,
+ "num_input_tokens_seen": 605902976,
+ "step": 13855,
+ "train_runtime": 107174.2345,
+ "train_tokens_per_second": 5653.439
+ },
+ {
+ "epoch": 0.9332076488015082,
+ "grad_norm": 0.7800631196768846,
+ "learning_rate": 5.5241499282491e-06,
+ "loss": 0.35591723918914797,
+ "num_input_tokens_seen": 606103560,
+ "step": 13860,
+ "train_runtime": 107212.7655,
+ "train_tokens_per_second": 5653.278
+ },
+ {
+ "epoch": 0.9335443037974683,
+ "grad_norm": 0.6995200522098299,
+ "learning_rate": 5.5215203414032135e-06,
+ "loss": 0.3359647274017334,
+ "num_input_tokens_seen": 606332784,
+ "step": 13865,
+ "train_runtime": 107261.2791,
+ "train_tokens_per_second": 5652.858
+ },
+ {
+ "epoch": 0.9338809587934285,
+ "grad_norm": 0.7315710239059265,
+ "learning_rate": 5.518890608715828e-06,
+ "loss": 0.3636192798614502,
+ "num_input_tokens_seen": 606549080,
+ "step": 13870,
+ "train_runtime": 107306.9882,
+ "train_tokens_per_second": 5652.466
+ },
+ {
+ "epoch": 0.9342176137893886,
+ "grad_norm": 0.8443286259971129,
+ "learning_rate": 5.516260730922337e-06,
+ "loss": 0.3732717275619507,
+ "num_input_tokens_seen": 606754664,
+ "step": 13875,
+ "train_runtime": 107344.4184,
+ "train_tokens_per_second": 5652.41
+ },
+ {
+ "epoch": 0.9345542687853488,
+ "grad_norm": 0.7163150661251769,
+ "learning_rate": 5.51363070875818e-06,
+ "loss": 0.3665364503860474,
+ "num_input_tokens_seen": 606971336,
+ "step": 13880,
+ "train_runtime": 107380.8754,
+ "train_tokens_per_second": 5652.509
+ },
+ {
+ "epoch": 0.9348909237813089,
+ "grad_norm": 0.7826404929432437,
+ "learning_rate": 5.511000542958832e-06,
+ "loss": 0.3829598903656006,
+ "num_input_tokens_seen": 607182888,
+ "step": 13885,
+ "train_runtime": 107425.4489,
+ "train_tokens_per_second": 5652.133
+ },
+ {
+ "epoch": 0.935227578777269,
+ "grad_norm": 0.7974161880488172,
+ "learning_rate": 5.508370234259812e-06,
+ "loss": 0.3617416858673096,
+ "num_input_tokens_seen": 607395376,
+ "step": 13890,
+ "train_runtime": 107467.1083,
+ "train_tokens_per_second": 5651.919
+ },
+ {
+ "epoch": 0.9355642337732292,
+ "grad_norm": 0.7468198908899866,
+ "learning_rate": 5.505739783396678e-06,
+ "loss": 0.4197434425354004,
+ "num_input_tokens_seen": 607606560,
+ "step": 13895,
+ "train_runtime": 107501.3365,
+ "train_tokens_per_second": 5652.084
+ },
+ {
+ "epoch": 0.9359008887691893,
+ "grad_norm": 0.8018465922168921,
+ "learning_rate": 5.503109191105026e-06,
+ "loss": 0.368869423866272,
+ "num_input_tokens_seen": 607818184,
+ "step": 13900,
+ "train_runtime": 107542.256,
+ "train_tokens_per_second": 5651.901
+ },
+ {
+ "epoch": 0.9362375437651494,
+ "grad_norm": 0.6896918447153622,
+ "learning_rate": 5.500478458120493e-06,
+ "loss": 0.3878501892089844,
+ "num_input_tokens_seen": 608046144,
+ "step": 13905,
+ "train_runtime": 107578.9863,
+ "train_tokens_per_second": 5652.09
+ },
+ {
+ "epoch": 0.9365741987611096,
+ "grad_norm": 0.6502629376388114,
+ "learning_rate": 5.497847585178754e-06,
+ "loss": 0.37720532417297364,
+ "num_input_tokens_seen": 608274496,
+ "step": 13910,
+ "train_runtime": 107614.2489,
+ "train_tokens_per_second": 5652.36
+ },
+ {
+ "epoch": 0.9369108537570697,
+ "grad_norm": 0.671150617120138,
+ "learning_rate": 5.495216573015526e-06,
+ "loss": 0.34202401638031005,
+ "num_input_tokens_seen": 608487072,
+ "step": 13915,
+ "train_runtime": 107649.7272,
+ "train_tokens_per_second": 5652.472
+ },
+ {
+ "epoch": 0.9372475087530299,
+ "grad_norm": 0.7302905322159697,
+ "learning_rate": 5.492585422366562e-06,
+ "loss": 0.3395219802856445,
+ "num_input_tokens_seen": 608702136,
+ "step": 13920,
+ "train_runtime": 107690.4313,
+ "train_tokens_per_second": 5652.333
+ },
+ {
+ "epoch": 0.93758416374899,
+ "grad_norm": 0.7232012493162839,
+ "learning_rate": 5.489954133967657e-06,
+ "loss": 0.3353360652923584,
+ "num_input_tokens_seen": 608907648,
+ "step": 13925,
+ "train_runtime": 107730.6462,
+ "train_tokens_per_second": 5652.13
+ },
+ {
+ "epoch": 0.9379208187449501,
+ "grad_norm": 0.6700324546008681,
+ "learning_rate": 5.48732270855464e-06,
+ "loss": 0.3447211027145386,
+ "num_input_tokens_seen": 609137704,
+ "step": 13930,
+ "train_runtime": 107770.2254,
+ "train_tokens_per_second": 5652.189
+ },
+ {
+ "epoch": 0.9382574737409103,
+ "grad_norm": 0.744553222908747,
+ "learning_rate": 5.48469114686338e-06,
+ "loss": 0.36331336498260497,
+ "num_input_tokens_seen": 609342520,
+ "step": 13935,
+ "train_runtime": 107808.8795,
+ "train_tokens_per_second": 5652.062
+ },
+ {
+ "epoch": 0.9385941287368704,
+ "grad_norm": 0.7229277790841906,
+ "learning_rate": 5.482059449629788e-06,
+ "loss": 0.3539914131164551,
+ "num_input_tokens_seen": 609571440,
+ "step": 13940,
+ "train_runtime": 107847.7103,
+ "train_tokens_per_second": 5652.15
+ },
+ {
+ "epoch": 0.9389307837328306,
+ "grad_norm": 0.7003088104586751,
+ "learning_rate": 5.479427617589808e-06,
+ "loss": 0.3414954662322998,
+ "num_input_tokens_seen": 609787136,
+ "step": 13945,
+ "train_runtime": 107885.5574,
+ "train_tokens_per_second": 5652.167
+ },
+ {
+ "epoch": 0.9392674387287907,
+ "grad_norm": 0.6891150838884817,
+ "learning_rate": 5.4767956514794226e-06,
+ "loss": 0.3689284801483154,
+ "num_input_tokens_seen": 610015912,
+ "step": 13950,
+ "train_runtime": 107918.1469,
+ "train_tokens_per_second": 5652.58
+ },
+ {
+ "epoch": 0.9396040937247508,
+ "grad_norm": 0.8843526225142938,
+ "learning_rate": 5.474163552034655e-06,
+ "loss": 0.3568282127380371,
+ "num_input_tokens_seen": 610224872,
+ "step": 13955,
+ "train_runtime": 107958.8593,
+ "train_tokens_per_second": 5652.383
+ },
+ {
+ "epoch": 0.939940748720711,
+ "grad_norm": 0.7953669523701649,
+ "learning_rate": 5.4715313199915595e-06,
+ "loss": 0.33748788833618165,
+ "num_input_tokens_seen": 610433896,
+ "step": 13960,
+ "train_runtime": 107998.8796,
+ "train_tokens_per_second": 5652.224
+ },
+ {
+ "epoch": 0.9402774037166711,
+ "grad_norm": 0.8436524922865566,
+ "learning_rate": 5.468898956086236e-06,
+ "loss": 0.36934537887573243,
+ "num_input_tokens_seen": 610657928,
+ "step": 13965,
+ "train_runtime": 108037.0093,
+ "train_tokens_per_second": 5652.303
+ },
+ {
+ "epoch": 0.9406140587126313,
+ "grad_norm": 0.7062142551000551,
+ "learning_rate": 5.466266461054815e-06,
+ "loss": 0.3693251132965088,
+ "num_input_tokens_seen": 610879072,
+ "step": 13970,
+ "train_runtime": 108078.9259,
+ "train_tokens_per_second": 5652.157
+ },
+ {
+ "epoch": 0.9409507137085914,
+ "grad_norm": 0.7405368560790437,
+ "learning_rate": 5.463633835633464e-06,
+ "loss": 0.38323063850402833,
+ "num_input_tokens_seen": 611111488,
+ "step": 13975,
+ "train_runtime": 108116.6852,
+ "train_tokens_per_second": 5652.333
+ },
+ {
+ "epoch": 0.9412873687045515,
+ "grad_norm": 0.6906471137787829,
+ "learning_rate": 5.46100108055839e-06,
+ "loss": 0.37401399612426756,
+ "num_input_tokens_seen": 611338984,
+ "step": 13980,
+ "train_runtime": 108153.7009,
+ "train_tokens_per_second": 5652.502
+ },
+ {
+ "epoch": 0.9416240237005117,
+ "grad_norm": 0.699157303658586,
+ "learning_rate": 5.458368196565833e-06,
+ "loss": 0.34747042655944826,
+ "num_input_tokens_seen": 611557384,
+ "step": 13985,
+ "train_runtime": 108187.4002,
+ "train_tokens_per_second": 5652.76
+ },
+ {
+ "epoch": 0.9419606786964718,
+ "grad_norm": 0.6950434602213443,
+ "learning_rate": 5.455735184392073e-06,
+ "loss": 0.39068779945373533,
+ "num_input_tokens_seen": 611785288,
+ "step": 13990,
+ "train_runtime": 108222.1017,
+ "train_tokens_per_second": 5653.053
+ },
+ {
+ "epoch": 0.942297333692432,
+ "grad_norm": 0.8222775495837801,
+ "learning_rate": 5.4531020447734204e-06,
+ "loss": 0.3776577949523926,
+ "num_input_tokens_seen": 611999096,
+ "step": 13995,
+ "train_runtime": 108255.2451,
+ "train_tokens_per_second": 5653.297
+ },
+ {
+ "epoch": 0.9426339886883921,
+ "grad_norm": 0.6942977966354065,
+ "learning_rate": 5.450468778446226e-06,
+ "loss": 0.38014535903930663,
+ "num_input_tokens_seen": 612229616,
+ "step": 14000,
+ "train_runtime": 108302.2562,
+ "train_tokens_per_second": 5652.972
+ },
+ {
+ "epoch": 0.9429706436843522,
+ "grad_norm": 0.8181331571770365,
+ "learning_rate": 5.447835386146875e-06,
+ "loss": 0.3620244264602661,
+ "num_input_tokens_seen": 612457288,
+ "step": 14005,
+ "train_runtime": 108339.7787,
+ "train_tokens_per_second": 5653.116
+ },
+ {
+ "epoch": 0.9433072986803124,
+ "grad_norm": 0.7249133494502058,
+ "learning_rate": 5.445201868611784e-06,
+ "loss": 0.4034558296203613,
+ "num_input_tokens_seen": 612668352,
+ "step": 14010,
+ "train_runtime": 108376.897,
+ "train_tokens_per_second": 5653.127
+ },
+ {
+ "epoch": 0.9436439536762725,
+ "grad_norm": 0.7004943272407984,
+ "learning_rate": 5.442568226577413e-06,
+ "loss": 0.37315356731414795,
+ "num_input_tokens_seen": 612885272,
+ "step": 14015,
+ "train_runtime": 108415.5424,
+ "train_tokens_per_second": 5653.113
+ },
+ {
+ "epoch": 0.9439806086722327,
+ "grad_norm": 0.6490441295997635,
+ "learning_rate": 5.439934460780247e-06,
+ "loss": 0.3591346740722656,
+ "num_input_tokens_seen": 613107944,
+ "step": 14020,
+ "train_runtime": 108453.9122,
+ "train_tokens_per_second": 5653.166
+ },
+ {
+ "epoch": 0.9443172636681928,
+ "grad_norm": 0.6918392163792407,
+ "learning_rate": 5.437300571956813e-06,
+ "loss": 0.3745762586593628,
+ "num_input_tokens_seen": 613308168,
+ "step": 14025,
+ "train_runtime": 108489.5473,
+ "train_tokens_per_second": 5653.154
+ },
+ {
+ "epoch": 0.9446539186641529,
+ "grad_norm": 0.654742109915237,
+ "learning_rate": 5.434666560843667e-06,
+ "loss": 0.3551823616027832,
+ "num_input_tokens_seen": 613516376,
+ "step": 14030,
+ "train_runtime": 108525.2798,
+ "train_tokens_per_second": 5653.212
+ },
+ {
+ "epoch": 0.9449905736601131,
+ "grad_norm": 0.7025947454416406,
+ "learning_rate": 5.432032428177405e-06,
+ "loss": 0.3733973979949951,
+ "num_input_tokens_seen": 613743992,
+ "step": 14035,
+ "train_runtime": 108565.1399,
+ "train_tokens_per_second": 5653.233
+ },
+ {
+ "epoch": 0.9453272286560732,
+ "grad_norm": 0.7446809227599923,
+ "learning_rate": 5.429398174694651e-06,
+ "loss": 0.35695390701293944,
+ "num_input_tokens_seen": 613975032,
+ "step": 14040,
+ "train_runtime": 108600.6982,
+ "train_tokens_per_second": 5653.509
+ },
+ {
+ "epoch": 0.9456638836520334,
+ "grad_norm": 0.6898581042276205,
+ "learning_rate": 5.426763801132067e-06,
+ "loss": 0.3617269039154053,
+ "num_input_tokens_seen": 614162568,
+ "step": 14045,
+ "train_runtime": 108637.4509,
+ "train_tokens_per_second": 5653.323
+ },
+ {
+ "epoch": 0.9460005386479935,
+ "grad_norm": 0.7678862695446375,
+ "learning_rate": 5.424129308226347e-06,
+ "loss": 0.3843861103057861,
+ "num_input_tokens_seen": 614370656,
+ "step": 14050,
+ "train_runtime": 108675.5655,
+ "train_tokens_per_second": 5653.255
+ },
+ {
+ "epoch": 0.9463371936439536,
+ "grad_norm": 0.7654766109095031,
+ "learning_rate": 5.421494696714219e-06,
+ "loss": 0.3902265548706055,
+ "num_input_tokens_seen": 614608864,
+ "step": 14055,
+ "train_runtime": 108715.2383,
+ "train_tokens_per_second": 5653.383
+ },
+ {
+ "epoch": 0.9466738486399138,
+ "grad_norm": 0.7609882464523344,
+ "learning_rate": 5.418859967332442e-06,
+ "loss": 0.3707080841064453,
+ "num_input_tokens_seen": 614836608,
+ "step": 14060,
+ "train_runtime": 108755.1102,
+ "train_tokens_per_second": 5653.404
+ },
+ {
+ "epoch": 0.9470105036358739,
+ "grad_norm": 0.9394544427990754,
+ "learning_rate": 5.416225120817811e-06,
+ "loss": 0.3681238412857056,
+ "num_input_tokens_seen": 615027064,
+ "step": 14065,
+ "train_runtime": 108790.9955,
+ "train_tokens_per_second": 5653.29
+ },
+ {
+ "epoch": 0.947347158631834,
+ "grad_norm": 0.6539346876126566,
+ "learning_rate": 5.413590157907153e-06,
+ "loss": 0.3346075534820557,
+ "num_input_tokens_seen": 615270632,
+ "step": 14070,
+ "train_runtime": 108826.6066,
+ "train_tokens_per_second": 5653.678
+ },
+ {
+ "epoch": 0.9476838136277942,
+ "grad_norm": 0.7338889594818504,
+ "learning_rate": 5.4109550793373255e-06,
+ "loss": 0.355800724029541,
+ "num_input_tokens_seen": 615493608,
+ "step": 14075,
+ "train_runtime": 108868.2257,
+ "train_tokens_per_second": 5653.565
+ },
+ {
+ "epoch": 0.9480204686237543,
+ "grad_norm": 0.7464991259586956,
+ "learning_rate": 5.40831988584522e-06,
+ "loss": 0.38836262226104734,
+ "num_input_tokens_seen": 615719216,
+ "step": 14080,
+ "train_runtime": 108904.0303,
+ "train_tokens_per_second": 5653.778
+ },
+ {
+ "epoch": 0.9483571236197145,
+ "grad_norm": 0.7069248942260824,
+ "learning_rate": 5.40568457816776e-06,
+ "loss": 0.37587385177612304,
+ "num_input_tokens_seen": 615951848,
+ "step": 14085,
+ "train_runtime": 108947.057,
+ "train_tokens_per_second": 5653.68
+ },
+ {
+ "epoch": 0.9486937786156746,
+ "grad_norm": 0.6655009611716466,
+ "learning_rate": 5.403049157041903e-06,
+ "loss": 0.35560946464538573,
+ "num_input_tokens_seen": 616172928,
+ "step": 14090,
+ "train_runtime": 108984.6239,
+ "train_tokens_per_second": 5653.76
+ },
+ {
+ "epoch": 0.9490304336116347,
+ "grad_norm": 0.7730572202934075,
+ "learning_rate": 5.400413623204635e-06,
+ "loss": 0.35712380409240724,
+ "num_input_tokens_seen": 616390224,
+ "step": 14095,
+ "train_runtime": 109023.2559,
+ "train_tokens_per_second": 5653.75
+ },
+ {
+ "epoch": 0.9493670886075949,
+ "grad_norm": 0.7946160574063879,
+ "learning_rate": 5.397777977392973e-06,
+ "loss": 0.39092726707458497,
+ "num_input_tokens_seen": 616612064,
+ "step": 14100,
+ "train_runtime": 109071.1955,
+ "train_tokens_per_second": 5653.299
+ },
+ {
+ "epoch": 0.949703743603555,
+ "grad_norm": 0.7261331777714058,
+ "learning_rate": 5.395142220343967e-06,
+ "loss": 0.37581348419189453,
+ "num_input_tokens_seen": 616835560,
+ "step": 14105,
+ "train_runtime": 109114.6463,
+ "train_tokens_per_second": 5653.096
+ },
+ {
+ "epoch": 0.9500403985995152,
+ "grad_norm": 0.7685617426175168,
+ "learning_rate": 5.3925063527947e-06,
+ "loss": 0.34862990379333497,
+ "num_input_tokens_seen": 617051200,
+ "step": 14110,
+ "train_runtime": 109161.8492,
+ "train_tokens_per_second": 5652.627
+ },
+ {
+ "epoch": 0.9503770535954753,
+ "grad_norm": 0.771130071661086,
+ "learning_rate": 5.389870375482283e-06,
+ "loss": 0.37904350757598876,
+ "num_input_tokens_seen": 617248976,
+ "step": 14115,
+ "train_runtime": 109194.634,
+ "train_tokens_per_second": 5652.741
+ },
+ {
+ "epoch": 0.9507137085914354,
+ "grad_norm": 0.7205683359655156,
+ "learning_rate": 5.387234289143859e-06,
+ "loss": 0.331813383102417,
+ "num_input_tokens_seen": 617469104,
+ "step": 14120,
+ "train_runtime": 109226.9679,
+ "train_tokens_per_second": 5653.083
+ },
+ {
+ "epoch": 0.9510503635873956,
+ "grad_norm": 0.7693964177160706,
+ "learning_rate": 5.384598094516601e-06,
+ "loss": 0.3729023218154907,
+ "num_input_tokens_seen": 617692328,
+ "step": 14125,
+ "train_runtime": 109271.6235,
+ "train_tokens_per_second": 5652.816
+ },
+ {
+ "epoch": 0.9513870185833557,
+ "grad_norm": 0.6640672951882179,
+ "learning_rate": 5.381961792337712e-06,
+ "loss": 0.3759523153305054,
+ "num_input_tokens_seen": 617921872,
+ "step": 14130,
+ "train_runtime": 109307.5627,
+ "train_tokens_per_second": 5653.057
+ },
+ {
+ "epoch": 0.9517236735793159,
+ "grad_norm": 0.6678437973528463,
+ "learning_rate": 5.3793253833444245e-06,
+ "loss": 0.3401124715805054,
+ "num_input_tokens_seen": 618147624,
+ "step": 14135,
+ "train_runtime": 109342.9092,
+ "train_tokens_per_second": 5653.294
+ },
+ {
+ "epoch": 0.952060328575276,
+ "grad_norm": 0.6761734645912748,
+ "learning_rate": 5.3766888682740045e-06,
+ "loss": 0.3542716264724731,
+ "num_input_tokens_seen": 618375336,
+ "step": 14140,
+ "train_runtime": 109378.6235,
+ "train_tokens_per_second": 5653.53
+ },
+ {
+ "epoch": 0.9523969835712361,
+ "grad_norm": 0.7917338745636189,
+ "learning_rate": 5.374052247863745e-06,
+ "loss": 0.3719951152801514,
+ "num_input_tokens_seen": 618588568,
+ "step": 14145,
+ "train_runtime": 109412.8524,
+ "train_tokens_per_second": 5653.71
+ },
+ {
+ "epoch": 0.9527336385671963,
+ "grad_norm": 0.7643089958554735,
+ "learning_rate": 5.371415522850966e-06,
+ "loss": 0.34739990234375,
+ "num_input_tokens_seen": 618807136,
+ "step": 14150,
+ "train_runtime": 109455.6515,
+ "train_tokens_per_second": 5653.496
+ },
+ {
+ "epoch": 0.9530702935631564,
+ "grad_norm": 0.7334076755125779,
+ "learning_rate": 5.368778693973019e-06,
+ "loss": 0.35079226493835447,
+ "num_input_tokens_seen": 619020192,
+ "step": 14155,
+ "train_runtime": 109497.2351,
+ "train_tokens_per_second": 5653.295
+ },
+ {
+ "epoch": 0.9534069485591166,
+ "grad_norm": 0.6518007663729203,
+ "learning_rate": 5.366141761967287e-06,
+ "loss": 0.3731173753738403,
+ "num_input_tokens_seen": 619234312,
+ "step": 14160,
+ "train_runtime": 109535.8084,
+ "train_tokens_per_second": 5653.259
+ },
+ {
+ "epoch": 0.9537436035550767,
+ "grad_norm": 0.884684695372941,
+ "learning_rate": 5.363504727571179e-06,
+ "loss": 0.38480224609375,
+ "num_input_tokens_seen": 619455112,
+ "step": 14165,
+ "train_runtime": 109574.3358,
+ "train_tokens_per_second": 5653.286
+ },
+ {
+ "epoch": 0.954080258551037,
+ "grad_norm": 0.7444612393985186,
+ "learning_rate": 5.360867591522133e-06,
+ "loss": 0.4258725166320801,
+ "num_input_tokens_seen": 619686368,
+ "step": 14170,
+ "train_runtime": 109614.4844,
+ "train_tokens_per_second": 5653.326
+ },
+ {
+ "epoch": 0.9544169135469971,
+ "grad_norm": 0.766088989031395,
+ "learning_rate": 5.358230354557617e-06,
+ "loss": 0.34602572917938235,
+ "num_input_tokens_seen": 619912184,
+ "step": 14175,
+ "train_runtime": 109647.5189,
+ "train_tokens_per_second": 5653.682
+ },
+ {
+ "epoch": 0.9547535685429572,
+ "grad_norm": 0.8426997667284004,
+ "learning_rate": 5.355593017415123e-06,
+ "loss": 0.39844369888305664,
+ "num_input_tokens_seen": 620138400,
+ "step": 14180,
+ "train_runtime": 109681.0251,
+ "train_tokens_per_second": 5654.017
+ },
+ {
+ "epoch": 0.9550902235389174,
+ "grad_norm": 0.6855456694998082,
+ "learning_rate": 5.352955580832176e-06,
+ "loss": 0.36531069278717043,
+ "num_input_tokens_seen": 620363336,
+ "step": 14185,
+ "train_runtime": 109720.9773,
+ "train_tokens_per_second": 5654.008
+ },
+ {
+ "epoch": 0.9554268785348775,
+ "grad_norm": 0.6489568473620863,
+ "learning_rate": 5.350318045546326e-06,
+ "loss": 0.378409218788147,
+ "num_input_tokens_seen": 620586608,
+ "step": 14190,
+ "train_runtime": 109762.864,
+ "train_tokens_per_second": 5653.885
+ },
+ {
+ "epoch": 0.9557635335308377,
+ "grad_norm": 0.690203133783407,
+ "learning_rate": 5.347680412295152e-06,
+ "loss": 0.3777376413345337,
+ "num_input_tokens_seen": 620803016,
+ "step": 14195,
+ "train_runtime": 109803.0202,
+ "train_tokens_per_second": 5653.788
+ },
+ {
+ "epoch": 0.9561001885267978,
+ "grad_norm": 0.8763898868581147,
+ "learning_rate": 5.34504268181626e-06,
+ "loss": 0.37516131401062014,
+ "num_input_tokens_seen": 621009712,
+ "step": 14200,
+ "train_runtime": 109842.1757,
+ "train_tokens_per_second": 5653.655
+ },
+ {
+ "epoch": 0.9564368435227579,
+ "grad_norm": 0.8351648882525854,
+ "learning_rate": 5.342404854847282e-06,
+ "loss": 0.36751456260681153,
+ "num_input_tokens_seen": 621209912,
+ "step": 14205,
+ "train_runtime": 109878.6528,
+ "train_tokens_per_second": 5653.6
+ },
+ {
+ "epoch": 0.9567734985187181,
+ "grad_norm": 0.695513608466477,
+ "learning_rate": 5.3397669321258776e-06,
+ "loss": 0.3533796310424805,
+ "num_input_tokens_seen": 621439136,
+ "step": 14210,
+ "train_runtime": 109913.5419,
+ "train_tokens_per_second": 5653.891
+ },
+ {
+ "epoch": 0.9571101535146782,
+ "grad_norm": 0.7212650216714175,
+ "learning_rate": 5.337128914389734e-06,
+ "loss": 0.3940083980560303,
+ "num_input_tokens_seen": 621670256,
+ "step": 14215,
+ "train_runtime": 109948.6255,
+ "train_tokens_per_second": 5654.189
+ },
+ {
+ "epoch": 0.9574468085106383,
+ "grad_norm": 0.7045386935112185,
+ "learning_rate": 5.334490802376565e-06,
+ "loss": 0.3630239009857178,
+ "num_input_tokens_seen": 621896344,
+ "step": 14220,
+ "train_runtime": 109992.4071,
+ "train_tokens_per_second": 5653.993
+ },
+ {
+ "epoch": 0.9577834635065985,
+ "grad_norm": 1.0648426671565956,
+ "learning_rate": 5.33185259682411e-06,
+ "loss": 0.38904123306274413,
+ "num_input_tokens_seen": 622111488,
+ "step": 14225,
+ "train_runtime": 110028.3421,
+ "train_tokens_per_second": 5654.102
+ },
+ {
+ "epoch": 0.9581201185025586,
+ "grad_norm": 0.6854092676166333,
+ "learning_rate": 5.329214298470134e-06,
+ "loss": 0.3952740907669067,
+ "num_input_tokens_seen": 622335752,
+ "step": 14230,
+ "train_runtime": 110067.3201,
+ "train_tokens_per_second": 5654.137
+ },
+ {
+ "epoch": 0.9584567734985188,
+ "grad_norm": 0.6504683472772627,
+ "learning_rate": 5.326575908052428e-06,
+ "loss": 0.34938879013061525,
+ "num_input_tokens_seen": 622563512,
+ "step": 14235,
+ "train_runtime": 110103.0055,
+ "train_tokens_per_second": 5654.373
+ },
+ {
+ "epoch": 0.9587934284944789,
+ "grad_norm": 0.8743571550942602,
+ "learning_rate": 5.323937426308813e-06,
+ "loss": 0.4064329147338867,
+ "num_input_tokens_seen": 622789680,
+ "step": 14240,
+ "train_runtime": 110140.2833,
+ "train_tokens_per_second": 5654.513
+ },
+ {
+ "epoch": 0.959130083490439,
+ "grad_norm": 0.7364135099684984,
+ "learning_rate": 5.321298853977128e-06,
+ "loss": 0.3482068538665771,
+ "num_input_tokens_seen": 623030288,
+ "step": 14245,
+ "train_runtime": 110172.6755,
+ "train_tokens_per_second": 5655.035
+ },
+ {
+ "epoch": 0.9594667384863992,
+ "grad_norm": 1.04514521012973,
+ "learning_rate": 5.318660191795244e-06,
+ "loss": 0.35120339393615724,
+ "num_input_tokens_seen": 623250248,
+ "step": 14250,
+ "train_runtime": 110210.2613,
+ "train_tokens_per_second": 5655.102
+ },
+ {
+ "epoch": 0.9598033934823593,
+ "grad_norm": 0.6859815669971231,
+ "learning_rate": 5.316021440501053e-06,
+ "loss": 0.32753448486328124,
+ "num_input_tokens_seen": 623468032,
+ "step": 14255,
+ "train_runtime": 110254.6316,
+ "train_tokens_per_second": 5654.801
+ },
+ {
+ "epoch": 0.9601400484783195,
+ "grad_norm": 0.6696719492054304,
+ "learning_rate": 5.313382600832474e-06,
+ "loss": 0.32216849327087405,
+ "num_input_tokens_seen": 623687536,
+ "step": 14260,
+ "train_runtime": 110295.1281,
+ "train_tokens_per_second": 5654.715
+ },
+ {
+ "epoch": 0.9604767034742796,
+ "grad_norm": 0.6796334454419847,
+ "learning_rate": 5.310743673527451e-06,
+ "loss": 0.3820059776306152,
+ "num_input_tokens_seen": 623916888,
+ "step": 14265,
+ "train_runtime": 110337.4663,
+ "train_tokens_per_second": 5654.624
+ },
+ {
+ "epoch": 0.9608133584702397,
+ "grad_norm": 0.7257174247083755,
+ "learning_rate": 5.3081046593239496e-06,
+ "loss": 0.38223018646240237,
+ "num_input_tokens_seen": 624136416,
+ "step": 14270,
+ "train_runtime": 110378.6302,
+ "train_tokens_per_second": 5654.504
+ },
+ {
+ "epoch": 0.9611500134661999,
+ "grad_norm": 0.7285256715224561,
+ "learning_rate": 5.305465558959964e-06,
+ "loss": 0.366811728477478,
+ "num_input_tokens_seen": 624349152,
+ "step": 14275,
+ "train_runtime": 110421.354,
+ "train_tokens_per_second": 5654.243
+ },
+ {
+ "epoch": 0.96148666846216,
+ "grad_norm": 0.7064345566990189,
+ "learning_rate": 5.302826373173506e-06,
+ "loss": 0.3593759536743164,
+ "num_input_tokens_seen": 624577352,
+ "step": 14280,
+ "train_runtime": 110460.1396,
+ "train_tokens_per_second": 5654.323
+ },
+ {
+ "epoch": 0.9618233234581202,
+ "grad_norm": 0.7256414932456383,
+ "learning_rate": 5.3001871027026194e-06,
+ "loss": 0.3493657112121582,
+ "num_input_tokens_seen": 624788664,
+ "step": 14285,
+ "train_runtime": 110502.0095,
+ "train_tokens_per_second": 5654.093
+ },
+ {
+ "epoch": 0.9621599784540803,
+ "grad_norm": 0.7017567146140217,
+ "learning_rate": 5.2975477482853685e-06,
+ "loss": 0.3475620746612549,
+ "num_input_tokens_seen": 625013976,
+ "step": 14290,
+ "train_runtime": 110543.2223,
+ "train_tokens_per_second": 5654.023
+ },
+ {
+ "epoch": 0.9624966334500404,
+ "grad_norm": 0.8025787225363398,
+ "learning_rate": 5.294908310659838e-06,
+ "loss": 0.3636843204498291,
+ "num_input_tokens_seen": 625243416,
+ "step": 14295,
+ "train_runtime": 110579.9936,
+ "train_tokens_per_second": 5654.218
+ },
+ {
+ "epoch": 0.9628332884460006,
+ "grad_norm": 0.796752899535943,
+ "learning_rate": 5.292268790564138e-06,
+ "loss": 0.3458188772201538,
+ "num_input_tokens_seen": 625456472,
+ "step": 14300,
+ "train_runtime": 110621.9221,
+ "train_tokens_per_second": 5654.001
+ },
+ {
+ "epoch": 0.9631699434419607,
+ "grad_norm": 0.7622734681695542,
+ "learning_rate": 5.289629188736403e-06,
+ "loss": 0.39026923179626466,
+ "num_input_tokens_seen": 625676128,
+ "step": 14305,
+ "train_runtime": 110667.8056,
+ "train_tokens_per_second": 5653.642
+ },
+ {
+ "epoch": 0.9635065984379209,
+ "grad_norm": 0.973852297067902,
+ "learning_rate": 5.286989505914788e-06,
+ "loss": 0.38088412284851075,
+ "num_input_tokens_seen": 625904384,
+ "step": 14310,
+ "train_runtime": 110702.6608,
+ "train_tokens_per_second": 5653.924
+ },
+ {
+ "epoch": 0.963843253433881,
+ "grad_norm": 0.8147025206812348,
+ "learning_rate": 5.284349742837475e-06,
+ "loss": 0.3758919954299927,
+ "num_input_tokens_seen": 626134784,
+ "step": 14315,
+ "train_runtime": 110744.3819,
+ "train_tokens_per_second": 5653.874
+ },
+ {
+ "epoch": 0.9641799084298411,
+ "grad_norm": 0.8865610211887707,
+ "learning_rate": 5.281709900242662e-06,
+ "loss": 0.35522916316986086,
+ "num_input_tokens_seen": 626357296,
+ "step": 14320,
+ "train_runtime": 110789.1337,
+ "train_tokens_per_second": 5653.599
+ },
+ {
+ "epoch": 0.9645165634258013,
+ "grad_norm": 0.7685303095133553,
+ "learning_rate": 5.279069978868574e-06,
+ "loss": 0.3454322099685669,
+ "num_input_tokens_seen": 626576680,
+ "step": 14325,
+ "train_runtime": 110823.2787,
+ "train_tokens_per_second": 5653.836
+ },
+ {
+ "epoch": 0.9648532184217614,
+ "grad_norm": 0.7582279578264484,
+ "learning_rate": 5.276429979453455e-06,
+ "loss": 0.36722531318664553,
+ "num_input_tokens_seen": 626795800,
+ "step": 14330,
+ "train_runtime": 110855.2605,
+ "train_tokens_per_second": 5654.182
+ },
+ {
+ "epoch": 0.9651898734177216,
+ "grad_norm": 0.6609643186180265,
+ "learning_rate": 5.273789902735575e-06,
+ "loss": 0.36404809951782224,
+ "num_input_tokens_seen": 627011600,
+ "step": 14335,
+ "train_runtime": 110890.1684,
+ "train_tokens_per_second": 5654.348
+ },
+ {
+ "epoch": 0.9655265284136817,
+ "grad_norm": 0.7160660322102089,
+ "learning_rate": 5.27114974945322e-06,
+ "loss": 0.33545889854431155,
+ "num_input_tokens_seen": 627226888,
+ "step": 14340,
+ "train_runtime": 110937.4224,
+ "train_tokens_per_second": 5653.88
+ },
+ {
+ "epoch": 0.9658631834096418,
+ "grad_norm": 0.7371545286431122,
+ "learning_rate": 5.2685095203447046e-06,
+ "loss": 0.3596933841705322,
+ "num_input_tokens_seen": 627453768,
+ "step": 14345,
+ "train_runtime": 110976.5578,
+ "train_tokens_per_second": 5653.931
+ },
+ {
+ "epoch": 0.966199838405602,
+ "grad_norm": 0.6944067783154944,
+ "learning_rate": 5.265869216148357e-06,
+ "loss": 0.352811861038208,
+ "num_input_tokens_seen": 627673216,
+ "step": 14350,
+ "train_runtime": 111014.9066,
+ "train_tokens_per_second": 5653.954
+ },
+ {
+ "epoch": 0.9665364934015621,
+ "grad_norm": 0.7335669224993876,
+ "learning_rate": 5.263228837602532e-06,
+ "loss": 0.42299542427062986,
+ "num_input_tokens_seen": 627887552,
+ "step": 14355,
+ "train_runtime": 111050.8572,
+ "train_tokens_per_second": 5654.054
+ },
+ {
+ "epoch": 0.9668731483975223,
+ "grad_norm": 0.83794134450925,
+ "learning_rate": 5.2605883854456e-06,
+ "loss": 0.3737421989440918,
+ "num_input_tokens_seen": 628108328,
+ "step": 14360,
+ "train_runtime": 111089.5086,
+ "train_tokens_per_second": 5654.074
+ },
+ {
+ "epoch": 0.9672098033934824,
+ "grad_norm": 0.6342873275404923,
+ "learning_rate": 5.257947860415957e-06,
+ "loss": 0.3357997894287109,
+ "num_input_tokens_seen": 628335808,
+ "step": 14365,
+ "train_runtime": 111122.6788,
+ "train_tokens_per_second": 5654.434
+ },
+ {
+ "epoch": 0.9675464583894425,
+ "grad_norm": 0.8822334071658794,
+ "learning_rate": 5.255307263252019e-06,
+ "loss": 0.3541123867034912,
+ "num_input_tokens_seen": 628566792,
+ "step": 14370,
+ "train_runtime": 111156.9119,
+ "train_tokens_per_second": 5654.77
+ },
+ {
+ "epoch": 0.9678831133854027,
+ "grad_norm": 0.674709973682241,
+ "learning_rate": 5.2526665946922185e-06,
+ "loss": 0.3823538303375244,
+ "num_input_tokens_seen": 628784336,
+ "step": 14375,
+ "train_runtime": 111196.5312,
+ "train_tokens_per_second": 5654.712
+ },
+ {
+ "epoch": 0.9682197683813628,
+ "grad_norm": 0.658823173011451,
+ "learning_rate": 5.25002585547501e-06,
+ "loss": 0.35442066192626953,
+ "num_input_tokens_seen": 629008320,
+ "step": 14380,
+ "train_runtime": 111234.2162,
+ "train_tokens_per_second": 5654.81
+ },
+ {
+ "epoch": 0.968556423377323,
+ "grad_norm": 0.723111666266317,
+ "learning_rate": 5.2473850463388675e-06,
+ "loss": 0.3943564176559448,
+ "num_input_tokens_seen": 629232256,
+ "step": 14385,
+ "train_runtime": 111274.7039,
+ "train_tokens_per_second": 5654.765
+ },
+ {
+ "epoch": 0.9688930783732831,
+ "grad_norm": 0.6809256630686321,
+ "learning_rate": 5.244744168022288e-06,
+ "loss": 0.3620326995849609,
+ "num_input_tokens_seen": 629455456,
+ "step": 14390,
+ "train_runtime": 111314.9521,
+ "train_tokens_per_second": 5654.725
+ },
+ {
+ "epoch": 0.9692297333692432,
+ "grad_norm": 0.6956920321641313,
+ "learning_rate": 5.24210322126378e-06,
+ "loss": 0.36865873336791993,
+ "num_input_tokens_seen": 629691848,
+ "step": 14395,
+ "train_runtime": 111353.3824,
+ "train_tokens_per_second": 5654.896
+ },
+ {
+ "epoch": 0.9695663883652034,
+ "grad_norm": 0.7152647788363251,
+ "learning_rate": 5.23946220680188e-06,
+ "loss": 0.3680101871490479,
+ "num_input_tokens_seen": 629928368,
+ "step": 14400,
+ "train_runtime": 111391.3207,
+ "train_tokens_per_second": 5655.094
+ },
+ {
+ "epoch": 0.9699030433611635,
+ "grad_norm": 0.7371345942960084,
+ "learning_rate": 5.236821125375136e-06,
+ "loss": 0.3584733486175537,
+ "num_input_tokens_seen": 630150304,
+ "step": 14405,
+ "train_runtime": 111429.6066,
+ "train_tokens_per_second": 5655.142
+ },
+ {
+ "epoch": 0.9702396983571236,
+ "grad_norm": 0.6525643955944114,
+ "learning_rate": 5.234179977722119e-06,
+ "loss": 0.31973896026611326,
+ "num_input_tokens_seen": 630389640,
+ "step": 14410,
+ "train_runtime": 111465.3624,
+ "train_tokens_per_second": 5655.476
+ },
+ {
+ "epoch": 0.9705763533530838,
+ "grad_norm": 0.6728032683654267,
+ "learning_rate": 5.23153876458142e-06,
+ "loss": 0.36993358135223386,
+ "num_input_tokens_seen": 630599464,
+ "step": 14415,
+ "train_runtime": 111509.2822,
+ "train_tokens_per_second": 5655.13
+ },
+ {
+ "epoch": 0.9709130083490439,
+ "grad_norm": 0.7435706382156235,
+ "learning_rate": 5.2288974866916435e-06,
+ "loss": 0.36161553859710693,
+ "num_input_tokens_seen": 630794912,
+ "step": 14420,
+ "train_runtime": 111547.8849,
+ "train_tokens_per_second": 5654.925
+ },
+ {
+ "epoch": 0.9712496633450041,
+ "grad_norm": 0.702712902975055,
+ "learning_rate": 5.226256144791414e-06,
+ "loss": 0.36026706695556643,
+ "num_input_tokens_seen": 631000992,
+ "step": 14425,
+ "train_runtime": 111582.5203,
+ "train_tokens_per_second": 5655.016
+ },
+ {
+ "epoch": 0.9715863183409642,
+ "grad_norm": 0.7340375348827854,
+ "learning_rate": 5.223614739619375e-06,
+ "loss": 0.37513973712921145,
+ "num_input_tokens_seen": 631202192,
+ "step": 14430,
+ "train_runtime": 111620.6231,
+ "train_tokens_per_second": 5654.889
+ },
+ {
+ "epoch": 0.9719229733369243,
+ "grad_norm": 0.6505020134915022,
+ "learning_rate": 5.2209732719141855e-06,
+ "loss": 0.33454012870788574,
+ "num_input_tokens_seen": 631411712,
+ "step": 14435,
+ "train_runtime": 111660.3902,
+ "train_tokens_per_second": 5654.751
+ },
+ {
+ "epoch": 0.9722596283328845,
+ "grad_norm": 0.6834948719980648,
+ "learning_rate": 5.218331742414528e-06,
+ "loss": 0.3455739736557007,
+ "num_input_tokens_seen": 631641904,
+ "step": 14440,
+ "train_runtime": 111694.5696,
+ "train_tokens_per_second": 5655.082
+ },
+ {
+ "epoch": 0.9725962833288446,
+ "grad_norm": 0.7031614396342304,
+ "learning_rate": 5.215690151859094e-06,
+ "loss": 0.3380829572677612,
+ "num_input_tokens_seen": 631848056,
+ "step": 14445,
+ "train_runtime": 111733.8193,
+ "train_tokens_per_second": 5654.94
+ },
+ {
+ "epoch": 0.9729329383248048,
+ "grad_norm": 0.737272924329504,
+ "learning_rate": 5.213048500986597e-06,
+ "loss": 0.3667490243911743,
+ "num_input_tokens_seen": 632068664,
+ "step": 14450,
+ "train_runtime": 111782.1877,
+ "train_tokens_per_second": 5654.467
+ },
+ {
+ "epoch": 0.9732695933207649,
+ "grad_norm": 0.8384823634165054,
+ "learning_rate": 5.210406790535766e-06,
+ "loss": 0.3663896083831787,
+ "num_input_tokens_seen": 632279544,
+ "step": 14455,
+ "train_runtime": 111817.1549,
+ "train_tokens_per_second": 5654.584
+ },
+ {
+ "epoch": 0.973606248316725,
+ "grad_norm": 0.7339723372461647,
+ "learning_rate": 5.207765021245346e-06,
+ "loss": 0.3660223722457886,
+ "num_input_tokens_seen": 632492984,
+ "step": 14460,
+ "train_runtime": 111858.5212,
+ "train_tokens_per_second": 5654.401
+ },
+ {
+ "epoch": 0.9739429033126852,
+ "grad_norm": 0.7329956341586107,
+ "learning_rate": 5.2051231938541015e-06,
+ "loss": 0.3708993911743164,
+ "num_input_tokens_seen": 632694568,
+ "step": 14465,
+ "train_runtime": 111894.0881,
+ "train_tokens_per_second": 5654.406
+ },
+ {
+ "epoch": 0.9742795583086453,
+ "grad_norm": 0.7751460530884079,
+ "learning_rate": 5.202481309100808e-06,
+ "loss": 0.3687734127044678,
+ "num_input_tokens_seen": 632904360,
+ "step": 14470,
+ "train_runtime": 111935.2636,
+ "train_tokens_per_second": 5654.2
+ },
+ {
+ "epoch": 0.9746162133046055,
+ "grad_norm": 0.6898579692907033,
+ "learning_rate": 5.199839367724266e-06,
+ "loss": 0.3596921920776367,
+ "num_input_tokens_seen": 633132800,
+ "step": 14475,
+ "train_runtime": 111976.0175,
+ "train_tokens_per_second": 5654.182
+ },
+ {
+ "epoch": 0.9749528683005656,
+ "grad_norm": 0.7400137611963173,
+ "learning_rate": 5.197197370463278e-06,
+ "loss": 0.343227481842041,
+ "num_input_tokens_seen": 633370648,
+ "step": 14480,
+ "train_runtime": 112017.4444,
+ "train_tokens_per_second": 5654.214
+ },
+ {
+ "epoch": 0.9752895232965257,
+ "grad_norm": 0.6958590027819616,
+ "learning_rate": 5.194555318056675e-06,
+ "loss": 0.39154577255249023,
+ "num_input_tokens_seen": 633605816,
+ "step": 14485,
+ "train_runtime": 112050.1996,
+ "train_tokens_per_second": 5654.66
+ },
+ {
+ "epoch": 0.9756261782924859,
+ "grad_norm": 0.9461003642017359,
+ "learning_rate": 5.191913211243299e-06,
+ "loss": 0.37716875076293943,
+ "num_input_tokens_seen": 633832456,
+ "step": 14490,
+ "train_runtime": 112095.5877,
+ "train_tokens_per_second": 5654.393
+ },
+ {
+ "epoch": 0.975962833288446,
+ "grad_norm": 0.7296195050735405,
+ "learning_rate": 5.1892710507620055e-06,
+ "loss": 0.36735081672668457,
+ "num_input_tokens_seen": 634052208,
+ "step": 14495,
+ "train_runtime": 112132.123,
+ "train_tokens_per_second": 5654.51
+ },
+ {
+ "epoch": 0.9762994882844062,
+ "grad_norm": 0.713237266588106,
+ "learning_rate": 5.186628837351666e-06,
+ "loss": 0.3544626235961914,
+ "num_input_tokens_seen": 634282120,
+ "step": 14500,
+ "train_runtime": 112168.6707,
+ "train_tokens_per_second": 5654.717
+ },
+ {
+ "epoch": 0.9766361432803663,
+ "grad_norm": 0.7152784729674379,
+ "learning_rate": 5.183986571751167e-06,
+ "loss": 0.3681643962860107,
+ "num_input_tokens_seen": 634503040,
+ "step": 14505,
+ "train_runtime": 112208.6433,
+ "train_tokens_per_second": 5654.672
+ },
+ {
+ "epoch": 0.9769727982763264,
+ "grad_norm": 0.7143902821508025,
+ "learning_rate": 5.181344254699408e-06,
+ "loss": 0.384121036529541,
+ "num_input_tokens_seen": 634718224,
+ "step": 14510,
+ "train_runtime": 112246.9754,
+ "train_tokens_per_second": 5654.658
+ },
+ {
+ "epoch": 0.9773094532722866,
+ "grad_norm": 0.7563338962960594,
+ "learning_rate": 5.178701886935309e-06,
+ "loss": 0.3466785907745361,
+ "num_input_tokens_seen": 634926976,
+ "step": 14515,
+ "train_runtime": 112283.3188,
+ "train_tokens_per_second": 5654.687
+ },
+ {
+ "epoch": 0.9776461082682467,
+ "grad_norm": 0.6709897911941165,
+ "learning_rate": 5.176059469197796e-06,
+ "loss": 0.3563254356384277,
+ "num_input_tokens_seen": 635147104,
+ "step": 14520,
+ "train_runtime": 112318.4084,
+ "train_tokens_per_second": 5654.88
+ },
+ {
+ "epoch": 0.9779827632642069,
+ "grad_norm": 0.84998527536875,
+ "learning_rate": 5.173417002225815e-06,
+ "loss": 0.3569037914276123,
+ "num_input_tokens_seen": 635366792,
+ "step": 14525,
+ "train_runtime": 112357.718,
+ "train_tokens_per_second": 5654.857
+ },
+ {
+ "epoch": 0.978319418260167,
+ "grad_norm": 0.7892134632795219,
+ "learning_rate": 5.170774486758322e-06,
+ "loss": 0.36590094566345216,
+ "num_input_tokens_seen": 635560568,
+ "step": 14530,
+ "train_runtime": 112393.8137,
+ "train_tokens_per_second": 5654.765
+ },
+ {
+ "epoch": 0.9786560732561271,
+ "grad_norm": 0.6842756660337252,
+ "learning_rate": 5.168131923534287e-06,
+ "loss": 0.38362579345703124,
+ "num_input_tokens_seen": 635787144,
+ "step": 14535,
+ "train_runtime": 112428.7709,
+ "train_tokens_per_second": 5655.022
+ },
+ {
+ "epoch": 0.9789927282520873,
+ "grad_norm": 0.6691518942357926,
+ "learning_rate": 5.165489313292696e-06,
+ "loss": 0.35508084297180176,
+ "num_input_tokens_seen": 636000120,
+ "step": 14540,
+ "train_runtime": 112469.9454,
+ "train_tokens_per_second": 5654.845
+ },
+ {
+ "epoch": 0.9793293832480474,
+ "grad_norm": 0.7564357234909826,
+ "learning_rate": 5.162846656772548e-06,
+ "loss": 0.35072224140167235,
+ "num_input_tokens_seen": 636214280,
+ "step": 14545,
+ "train_runtime": 112512.921,
+ "train_tokens_per_second": 5654.589
+ },
+ {
+ "epoch": 0.9796660382440076,
+ "grad_norm": 0.7450101360653049,
+ "learning_rate": 5.16020395471285e-06,
+ "loss": 0.3438509464263916,
+ "num_input_tokens_seen": 636435816,
+ "step": 14550,
+ "train_runtime": 112554.2979,
+ "train_tokens_per_second": 5654.478
+ },
+ {
+ "epoch": 0.9800026932399677,
+ "grad_norm": 0.8137120525286944,
+ "learning_rate": 5.157561207852628e-06,
+ "loss": 0.3831623077392578,
+ "num_input_tokens_seen": 636640224,
+ "step": 14555,
+ "train_runtime": 112591.2086,
+ "train_tokens_per_second": 5654.44
+ },
+ {
+ "epoch": 0.9803393482359278,
+ "grad_norm": 0.7297592438951429,
+ "learning_rate": 5.154918416930915e-06,
+ "loss": 0.3411484479904175,
+ "num_input_tokens_seen": 636858520,
+ "step": 14560,
+ "train_runtime": 112629.598,
+ "train_tokens_per_second": 5654.451
+ },
+ {
+ "epoch": 0.980676003231888,
+ "grad_norm": 0.760019959304734,
+ "learning_rate": 5.152275582686761e-06,
+ "loss": 0.3839298486709595,
+ "num_input_tokens_seen": 637073856,
+ "step": 14565,
+ "train_runtime": 112666.3482,
+ "train_tokens_per_second": 5654.518
+ },
+ {
+ "epoch": 0.9810126582278481,
+ "grad_norm": 0.6959300895402221,
+ "learning_rate": 5.149632705859227e-06,
+ "loss": 0.36935174465179443,
+ "num_input_tokens_seen": 637299440,
+ "step": 14570,
+ "train_runtime": 112707.9899,
+ "train_tokens_per_second": 5654.43
+ },
+ {
+ "epoch": 0.9813493132238083,
+ "grad_norm": 0.7586977280516466,
+ "learning_rate": 5.146989787187383e-06,
+ "loss": 0.37570605278015134,
+ "num_input_tokens_seen": 637529440,
+ "step": 14575,
+ "train_runtime": 112743.4679,
+ "train_tokens_per_second": 5654.691
+ },
+ {
+ "epoch": 0.9816859682197684,
+ "grad_norm": 0.6825264977223254,
+ "learning_rate": 5.144346827410314e-06,
+ "loss": 0.3703890562057495,
+ "num_input_tokens_seen": 637754408,
+ "step": 14580,
+ "train_runtime": 112773.1649,
+ "train_tokens_per_second": 5655.197
+ },
+ {
+ "epoch": 0.9820226232157285,
+ "grad_norm": 0.826290290377258,
+ "learning_rate": 5.141703827267117e-06,
+ "loss": 0.3702056646347046,
+ "num_input_tokens_seen": 637982336,
+ "step": 14585,
+ "train_runtime": 112804.5446,
+ "train_tokens_per_second": 5655.644
+ },
+ {
+ "epoch": 0.9823592782116887,
+ "grad_norm": 0.7212214874544945,
+ "learning_rate": 5.139060787496893e-06,
+ "loss": 0.3754692077636719,
+ "num_input_tokens_seen": 638203872,
+ "step": 14590,
+ "train_runtime": 112841.7495,
+ "train_tokens_per_second": 5655.742
+ },
+ {
+ "epoch": 0.9826959332076488,
+ "grad_norm": 0.6942239873292527,
+ "learning_rate": 5.136417708838766e-06,
+ "loss": 0.3775701284408569,
+ "num_input_tokens_seen": 638407968,
+ "step": 14595,
+ "train_runtime": 112885.1355,
+ "train_tokens_per_second": 5655.377
+ },
+ {
+ "epoch": 0.983032588203609,
+ "grad_norm": 0.7575191844718635,
+ "learning_rate": 5.13377459203186e-06,
+ "loss": 0.3628222465515137,
+ "num_input_tokens_seen": 638616040,
+ "step": 14600,
+ "train_runtime": 112923.0341,
+ "train_tokens_per_second": 5655.321
+ },
+ {
+ "epoch": 0.9833692431995691,
+ "grad_norm": 0.7100885514863359,
+ "learning_rate": 5.131131437815318e-06,
+ "loss": 0.3371574401855469,
+ "num_input_tokens_seen": 638837176,
+ "step": 14605,
+ "train_runtime": 112963.1632,
+ "train_tokens_per_second": 5655.27
+ },
+ {
+ "epoch": 0.9837058981955292,
+ "grad_norm": 0.7208738472250812,
+ "learning_rate": 5.128488246928287e-06,
+ "loss": 0.36636881828308104,
+ "num_input_tokens_seen": 639065552,
+ "step": 14610,
+ "train_runtime": 113000.1458,
+ "train_tokens_per_second": 5655.44
+ },
+ {
+ "epoch": 0.9840425531914894,
+ "grad_norm": 0.7630145727500303,
+ "learning_rate": 5.125845020109925e-06,
+ "loss": 0.37039480209350584,
+ "num_input_tokens_seen": 639285272,
+ "step": 14615,
+ "train_runtime": 113047.3523,
+ "train_tokens_per_second": 5655.022
+ },
+ {
+ "epoch": 0.9843792081874495,
+ "grad_norm": 0.7720868800544706,
+ "learning_rate": 5.123201758099406e-06,
+ "loss": 0.36545164585113527,
+ "num_input_tokens_seen": 639499208,
+ "step": 14620,
+ "train_runtime": 113085.6686,
+ "train_tokens_per_second": 5654.998
+ },
+ {
+ "epoch": 0.9847158631834096,
+ "grad_norm": 0.7035125490360992,
+ "learning_rate": 5.12055846163591e-06,
+ "loss": 0.37010979652404785,
+ "num_input_tokens_seen": 639714312,
+ "step": 14625,
+ "train_runtime": 113123.7227,
+ "train_tokens_per_second": 5654.997
+ },
+ {
+ "epoch": 0.9850525181793698,
+ "grad_norm": 0.6519589735545973,
+ "learning_rate": 5.117915131458623e-06,
+ "loss": 0.3291551828384399,
+ "num_input_tokens_seen": 639938488,
+ "step": 14630,
+ "train_runtime": 113162.111,
+ "train_tokens_per_second": 5655.06
+ },
+ {
+ "epoch": 0.9853891731753299,
+ "grad_norm": 0.8145405370454951,
+ "learning_rate": 5.115271768306745e-06,
+ "loss": 0.36716470718383787,
+ "num_input_tokens_seen": 640152840,
+ "step": 14635,
+ "train_runtime": 113199.5845,
+ "train_tokens_per_second": 5655.081
+ },
+ {
+ "epoch": 0.9857258281712901,
+ "grad_norm": 0.790933823693933,
+ "learning_rate": 5.112628372919485e-06,
+ "loss": 0.3815732717514038,
+ "num_input_tokens_seen": 640369952,
+ "step": 14640,
+ "train_runtime": 113234.1823,
+ "train_tokens_per_second": 5655.271
+ },
+ {
+ "epoch": 0.9860624831672502,
+ "grad_norm": 0.9699782510514777,
+ "learning_rate": 5.1099849460360595e-06,
+ "loss": 0.34613332748413084,
+ "num_input_tokens_seen": 640567240,
+ "step": 14645,
+ "train_runtime": 113271.7361,
+ "train_tokens_per_second": 5655.137
+ },
+ {
+ "epoch": 0.9863991381632103,
+ "grad_norm": 0.6723948378020855,
+ "learning_rate": 5.107341488395695e-06,
+ "loss": 0.3379997730255127,
+ "num_input_tokens_seen": 640788384,
+ "step": 14650,
+ "train_runtime": 113310.4257,
+ "train_tokens_per_second": 5655.158
+ },
+ {
+ "epoch": 0.9867357931591705,
+ "grad_norm": 0.7000157996530737,
+ "learning_rate": 5.104698000737626e-06,
+ "loss": 0.37780165672302246,
+ "num_input_tokens_seen": 641011792,
+ "step": 14655,
+ "train_runtime": 113351.4983,
+ "train_tokens_per_second": 5655.08
+ },
+ {
+ "epoch": 0.9870724481551306,
+ "grad_norm": 0.7543705648632995,
+ "learning_rate": 5.102054483801094e-06,
+ "loss": 0.38533506393432615,
+ "num_input_tokens_seen": 641239984,
+ "step": 14660,
+ "train_runtime": 113390.3486,
+ "train_tokens_per_second": 5655.155
+ },
+ {
+ "epoch": 0.9874091031510908,
+ "grad_norm": 0.7193117605335543,
+ "learning_rate": 5.099410938325351e-06,
+ "loss": 0.3746023178100586,
+ "num_input_tokens_seen": 641468200,
+ "step": 14665,
+ "train_runtime": 113424.7218,
+ "train_tokens_per_second": 5655.453
+ },
+ {
+ "epoch": 0.9877457581470509,
+ "grad_norm": 0.8146622438199124,
+ "learning_rate": 5.096767365049657e-06,
+ "loss": 0.3502329111099243,
+ "num_input_tokens_seen": 641672792,
+ "step": 14670,
+ "train_runtime": 113464.964,
+ "train_tokens_per_second": 5655.25
+ },
+ {
+ "epoch": 0.988082413143011,
+ "grad_norm": 0.6875126263724606,
+ "learning_rate": 5.094123764713277e-06,
+ "loss": 0.3421436309814453,
+ "num_input_tokens_seen": 641904736,
+ "step": 14675,
+ "train_runtime": 113500.904,
+ "train_tokens_per_second": 5655.503
+ },
+ {
+ "epoch": 0.9884190681389712,
+ "grad_norm": 0.7258139194348168,
+ "learning_rate": 5.091480138055487e-06,
+ "loss": 0.36416258811950686,
+ "num_input_tokens_seen": 642136592,
+ "step": 14680,
+ "train_runtime": 113531.8425,
+ "train_tokens_per_second": 5656.004
+ },
+ {
+ "epoch": 0.9887557231349313,
+ "grad_norm": 0.7250476516557031,
+ "learning_rate": 5.088836485815568e-06,
+ "loss": 0.35428576469421386,
+ "num_input_tokens_seen": 642334648,
+ "step": 14685,
+ "train_runtime": 113566.8784,
+ "train_tokens_per_second": 5656.003
+ },
+ {
+ "epoch": 0.9890923781308915,
+ "grad_norm": 0.7385563837095339,
+ "learning_rate": 5.086192808732809e-06,
+ "loss": 0.35656602382659913,
+ "num_input_tokens_seen": 642538480,
+ "step": 14690,
+ "train_runtime": 113608.5759,
+ "train_tokens_per_second": 5655.722
+ },
+ {
+ "epoch": 0.9894290331268516,
+ "grad_norm": 0.7946980704838973,
+ "learning_rate": 5.083549107546505e-06,
+ "loss": 0.36766040325164795,
+ "num_input_tokens_seen": 642744792,
+ "step": 14695,
+ "train_runtime": 113647.6019,
+ "train_tokens_per_second": 5655.595
+ },
+ {
+ "epoch": 0.9897656881228117,
+ "grad_norm": 0.771038431226611,
+ "learning_rate": 5.080905382995961e-06,
+ "loss": 0.38895292282104493,
+ "num_input_tokens_seen": 642974608,
+ "step": 14700,
+ "train_runtime": 113683.9469,
+ "train_tokens_per_second": 5655.808
+ },
+ {
+ "epoch": 0.9901023431187719,
+ "grad_norm": 0.7343676499490034,
+ "learning_rate": 5.0782616358204835e-06,
+ "loss": 0.37378993034362795,
+ "num_input_tokens_seen": 643191928,
+ "step": 14705,
+ "train_runtime": 113723.5317,
+ "train_tokens_per_second": 5655.751
+ },
+ {
+ "epoch": 0.990438998114732,
+ "grad_norm": 0.8524692289540274,
+ "learning_rate": 5.075617866759389e-06,
+ "loss": 0.33492867946624755,
+ "num_input_tokens_seen": 643403376,
+ "step": 14710,
+ "train_runtime": 113766.092,
+ "train_tokens_per_second": 5655.493
+ },
+ {
+ "epoch": 0.9907756531106922,
+ "grad_norm": 0.7901150595595263,
+ "learning_rate": 5.0729740765519995e-06,
+ "loss": 0.4053798675537109,
+ "num_input_tokens_seen": 643617760,
+ "step": 14715,
+ "train_runtime": 113802.6369,
+ "train_tokens_per_second": 5655.561
+ },
+ {
+ "epoch": 0.9911123081066523,
+ "grad_norm": 0.7850703195564729,
+ "learning_rate": 5.070330265937643e-06,
+ "loss": 0.3686873435974121,
+ "num_input_tokens_seen": 643827376,
+ "step": 14720,
+ "train_runtime": 113838.4716,
+ "train_tokens_per_second": 5655.622
+ },
+ {
+ "epoch": 0.9914489631026124,
+ "grad_norm": 0.7339072075707702,
+ "learning_rate": 5.067686435655651e-06,
+ "loss": 0.3935819625854492,
+ "num_input_tokens_seen": 644051992,
+ "step": 14725,
+ "train_runtime": 113875.0085,
+ "train_tokens_per_second": 5655.78
+ },
+ {
+ "epoch": 0.9917856180985726,
+ "grad_norm": 0.7717913369926017,
+ "learning_rate": 5.065042586445363e-06,
+ "loss": 0.39838461875915526,
+ "num_input_tokens_seen": 644255512,
+ "step": 14730,
+ "train_runtime": 113907.4094,
+ "train_tokens_per_second": 5655.958
+ },
+ {
+ "epoch": 0.9921222730945327,
+ "grad_norm": 0.7852173126276834,
+ "learning_rate": 5.0623987190461245e-06,
+ "loss": 0.30464193820953367,
+ "num_input_tokens_seen": 644460824,
+ "step": 14735,
+ "train_runtime": 113944.9392,
+ "train_tokens_per_second": 5655.897
+ },
+ {
+ "epoch": 0.9924589280904929,
+ "grad_norm": 0.7913343599249734,
+ "learning_rate": 5.059754834197281e-06,
+ "loss": 0.3790989875793457,
+ "num_input_tokens_seen": 644675152,
+ "step": 14740,
+ "train_runtime": 113988.6481,
+ "train_tokens_per_second": 5655.608
+ },
+ {
+ "epoch": 0.992795583086453,
+ "grad_norm": 0.6313018834314664,
+ "learning_rate": 5.05711093263819e-06,
+ "loss": 0.32619643211364746,
+ "num_input_tokens_seen": 644890856,
+ "step": 14745,
+ "train_runtime": 114026.2898,
+ "train_tokens_per_second": 5655.633
+ },
+ {
+ "epoch": 0.9931322380824131,
+ "grad_norm": 0.7700737076224284,
+ "learning_rate": 5.054467015108208e-06,
+ "loss": 0.37743263244628905,
+ "num_input_tokens_seen": 645124328,
+ "step": 14750,
+ "train_runtime": 114061.9307,
+ "train_tokens_per_second": 5655.913
+ },
+ {
+ "epoch": 0.9934688930783733,
+ "grad_norm": 0.8548358470170475,
+ "learning_rate": 5.051823082346701e-06,
+ "loss": 0.37684605121612547,
+ "num_input_tokens_seen": 645342896,
+ "step": 14755,
+ "train_runtime": 114095.225,
+ "train_tokens_per_second": 5656.178
+ },
+ {
+ "epoch": 0.9938055480743334,
+ "grad_norm": 0.7487000463516748,
+ "learning_rate": 5.049179135093033e-06,
+ "loss": 0.3667085409164429,
+ "num_input_tokens_seen": 645572616,
+ "step": 14760,
+ "train_runtime": 114128.9739,
+ "train_tokens_per_second": 5656.518
+ },
+ {
+ "epoch": 0.9941422030702936,
+ "grad_norm": 0.7554586359729867,
+ "learning_rate": 5.046535174086575e-06,
+ "loss": 0.37410502433776854,
+ "num_input_tokens_seen": 645784112,
+ "step": 14765,
+ "train_runtime": 114171.4105,
+ "train_tokens_per_second": 5656.268
+ },
+ {
+ "epoch": 0.9944788580662537,
+ "grad_norm": 0.806201449115856,
+ "learning_rate": 5.043891200066706e-06,
+ "loss": 0.35855536460876464,
+ "num_input_tokens_seen": 646000480,
+ "step": 14770,
+ "train_runtime": 114211.5644,
+ "train_tokens_per_second": 5656.174
+ },
+ {
+ "epoch": 0.9948155130622138,
+ "grad_norm": 0.7970070550101245,
+ "learning_rate": 5.041247213772802e-06,
+ "loss": 0.3647798538208008,
+ "num_input_tokens_seen": 646228928,
+ "step": 14775,
+ "train_runtime": 114248.8453,
+ "train_tokens_per_second": 5656.328
+ },
+ {
+ "epoch": 0.995152168058174,
+ "grad_norm": 0.9274665922453337,
+ "learning_rate": 5.038603215944247e-06,
+ "loss": 0.3697342872619629,
+ "num_input_tokens_seen": 646427096,
+ "step": 14780,
+ "train_runtime": 114283.9377,
+ "train_tokens_per_second": 5656.325
+ },
+ {
+ "epoch": 0.9954888230541341,
+ "grad_norm": 0.7200893083392556,
+ "learning_rate": 5.0359592073204254e-06,
+ "loss": 0.34227933883666994,
+ "num_input_tokens_seen": 646643272,
+ "step": 14785,
+ "train_runtime": 114327.3802,
+ "train_tokens_per_second": 5656.067
+ },
+ {
+ "epoch": 0.9958254780500942,
+ "grad_norm": 1.0261926014877596,
+ "learning_rate": 5.033315188640725e-06,
+ "loss": 0.3627659797668457,
+ "num_input_tokens_seen": 646872200,
+ "step": 14790,
+ "train_runtime": 114366.7864,
+ "train_tokens_per_second": 5656.119
+ },
+ {
+ "epoch": 0.9961621330460544,
+ "grad_norm": 0.6681278213263397,
+ "learning_rate": 5.03067116064454e-06,
+ "loss": 0.3450484752655029,
+ "num_input_tokens_seen": 647094832,
+ "step": 14795,
+ "train_runtime": 114408.9697,
+ "train_tokens_per_second": 5655.98
+ },
+ {
+ "epoch": 0.9964987880420145,
+ "grad_norm": 0.6613774299958016,
+ "learning_rate": 5.028027124071262e-06,
+ "loss": 0.36185712814331056,
+ "num_input_tokens_seen": 647301568,
+ "step": 14800,
+ "train_runtime": 114449.8831,
+ "train_tokens_per_second": 5655.764
+ },
+ {
+ "epoch": 0.9968354430379747,
+ "grad_norm": 0.7005897120089515,
+ "learning_rate": 5.025383079660289e-06,
+ "loss": 0.3468491554260254,
+ "num_input_tokens_seen": 647531264,
+ "step": 14805,
+ "train_runtime": 114488.5353,
+ "train_tokens_per_second": 5655.861
+ },
+ {
+ "epoch": 0.9971720980339348,
+ "grad_norm": 0.7061938748348358,
+ "learning_rate": 5.0227390281510166e-06,
+ "loss": 0.3663314342498779,
+ "num_input_tokens_seen": 647753976,
+ "step": 14810,
+ "train_runtime": 114526.0357,
+ "train_tokens_per_second": 5655.954
+ },
+ {
+ "epoch": 0.997508753029895,
+ "grad_norm": 0.7212945140660744,
+ "learning_rate": 5.020094970282849e-06,
+ "loss": 0.3663675308227539,
+ "num_input_tokens_seen": 647955768,
+ "step": 14815,
+ "train_runtime": 114566.6744,
+ "train_tokens_per_second": 5655.709
+ },
+ {
+ "epoch": 0.9978454080258551,
+ "grad_norm": 0.7579147457391882,
+ "learning_rate": 5.017450906795187e-06,
+ "loss": 0.36784412860870364,
+ "num_input_tokens_seen": 648181384,
+ "step": 14820,
+ "train_runtime": 114601.4001,
+ "train_tokens_per_second": 5655.964
+ },
+ {
+ "epoch": 0.9981820630218152,
+ "grad_norm": 0.7496488759479608,
+ "learning_rate": 5.014806838427434e-06,
+ "loss": 0.36796035766601565,
+ "num_input_tokens_seen": 648382672,
+ "step": 14825,
+ "train_runtime": 114646.9011,
+ "train_tokens_per_second": 5655.475
+ },
+ {
+ "epoch": 0.9985187180177754,
+ "grad_norm": 0.6851905965880742,
+ "learning_rate": 5.0121627659189964e-06,
+ "loss": 0.3591431140899658,
+ "num_input_tokens_seen": 648612168,
+ "step": 14830,
+ "train_runtime": 114682.2822,
+ "train_tokens_per_second": 5655.731
+ },
+ {
+ "epoch": 0.9988553730137355,
+ "grad_norm": 0.6696970669236173,
+ "learning_rate": 5.009518690009281e-06,
+ "loss": 0.3566739082336426,
+ "num_input_tokens_seen": 648821208,
+ "step": 14835,
+ "train_runtime": 114724.3707,
+ "train_tokens_per_second": 5655.478
+ },
+ {
+ "epoch": 0.9991920280096956,
+ "grad_norm": 0.7614065048649256,
+ "learning_rate": 5.006874611437692e-06,
+ "loss": 0.3776280879974365,
+ "num_input_tokens_seen": 649046288,
+ "step": 14840,
+ "train_runtime": 114764.8122,
+ "train_tokens_per_second": 5655.447
+ },
+ {
+ "epoch": 0.9995286830056558,
+ "grad_norm": 0.6492499944472357,
+ "learning_rate": 5.004230530943641e-06,
+ "loss": 0.35755281448364257,
+ "num_input_tokens_seen": 649262232,
+ "step": 14845,
+ "train_runtime": 114804.5614,
+ "train_tokens_per_second": 5655.37
+ },
+ {
+ "epoch": 0.9998653380016159,
+ "grad_norm": 0.7184203567667401,
+ "learning_rate": 5.001586449266536e-06,
+ "loss": 0.3260646343231201,
+ "num_input_tokens_seen": 649487024,
+ "step": 14850,
+ "train_runtime": 114845.5757,
+ "train_tokens_per_second": 5655.307
+ },
+ {
+ "epoch": 1.000201992997576,
+ "grad_norm": 0.6358891287683051,
+ "learning_rate": 4.998942367145787e-06,
+ "loss": 0.35132179260253904,
+ "num_input_tokens_seen": 649706888,
+ "step": 14855,
+ "train_runtime": 114929.2136,
+ "train_tokens_per_second": 5653.105
+ },
+ {
+ "epoch": 1.0005386479935363,
+ "grad_norm": 0.7034269548859629,
+ "learning_rate": 4.996298285320798e-06,
+ "loss": 0.30704555511474607,
+ "num_input_tokens_seen": 649911248,
+ "step": 14860,
+ "train_runtime": 114965.2665,
+ "train_tokens_per_second": 5653.11
+ },
+ {
+ "epoch": 1.0008753029894963,
+ "grad_norm": 0.7771207057358878,
+ "learning_rate": 4.993654204530983e-06,
+ "loss": 0.2908186435699463,
+ "num_input_tokens_seen": 650136336,
+ "step": 14865,
+ "train_runtime": 115009.0891,
+ "train_tokens_per_second": 5652.913
+ },
+ {
+ "epoch": 1.0012119579854566,
+ "grad_norm": 0.7800505661917783,
+ "learning_rate": 4.99101012551575e-06,
+ "loss": 0.3061246633529663,
+ "num_input_tokens_seen": 650324944,
+ "step": 14870,
+ "train_runtime": 115047.7835,
+ "train_tokens_per_second": 5652.651
+ },
+ {
+ "epoch": 1.0015486129814166,
+ "grad_norm": 0.7451144290465576,
+ "learning_rate": 4.9883660490145075e-06,
+ "loss": 0.31706483364105226,
+ "num_input_tokens_seen": 650556192,
+ "step": 14875,
+ "train_runtime": 115083.1511,
+ "train_tokens_per_second": 5652.923
+ },
+ {
+ "epoch": 1.0018852679773769,
+ "grad_norm": 0.6924866106934799,
+ "learning_rate": 4.985721975766661e-06,
+ "loss": 0.32414817810058594,
+ "num_input_tokens_seen": 650775960,
+ "step": 14880,
+ "train_runtime": 115118.3264,
+ "train_tokens_per_second": 5653.105
+ },
+ {
+ "epoch": 1.002221922973337,
+ "grad_norm": 1.0051470668362659,
+ "learning_rate": 4.983077906511619e-06,
+ "loss": 0.32347331047058103,
+ "num_input_tokens_seen": 650983240,
+ "step": 14885,
+ "train_runtime": 115161.3444,
+ "train_tokens_per_second": 5652.793
+ },
+ {
+ "epoch": 1.0025585779692971,
+ "grad_norm": 0.7329517809959321,
+ "learning_rate": 4.980433841988786e-06,
+ "loss": 0.30762290954589844,
+ "num_input_tokens_seen": 651204568,
+ "step": 14890,
+ "train_runtime": 115197.7799,
+ "train_tokens_per_second": 5652.926
+ },
+ {
+ "epoch": 1.0028952329652572,
+ "grad_norm": 0.7126218012313592,
+ "learning_rate": 4.977789782937568e-06,
+ "loss": 0.3130316257476807,
+ "num_input_tokens_seen": 651431064,
+ "step": 14895,
+ "train_runtime": 115231.9768,
+ "train_tokens_per_second": 5653.214
+ },
+ {
+ "epoch": 1.0032318879612174,
+ "grad_norm": 0.7709912103285202,
+ "learning_rate": 4.975145730097364e-06,
+ "loss": 0.3236842155456543,
+ "num_input_tokens_seen": 651636984,
+ "step": 14900,
+ "train_runtime": 115269.0999,
+ "train_tokens_per_second": 5653.18
+ },
+ {
+ "epoch": 1.0035685429571775,
+ "grad_norm": 0.6705219562784165,
+ "learning_rate": 4.97250168420758e-06,
+ "loss": 0.3103510618209839,
+ "num_input_tokens_seen": 651849816,
+ "step": 14905,
+ "train_runtime": 115304.0536,
+ "train_tokens_per_second": 5653.312
+ },
+ {
+ "epoch": 1.0039051979531377,
+ "grad_norm": 0.737858641027806,
+ "learning_rate": 4.96985764600761e-06,
+ "loss": 0.3144733428955078,
+ "num_input_tokens_seen": 652078544,
+ "step": 14910,
+ "train_runtime": 115343.8311,
+ "train_tokens_per_second": 5653.346
+ },
+ {
+ "epoch": 1.0042418529490977,
+ "grad_norm": 0.7152864263979617,
+ "learning_rate": 4.967213616236855e-06,
+ "loss": 0.32990431785583496,
+ "num_input_tokens_seen": 652288816,
+ "step": 14915,
+ "train_runtime": 115379.8129,
+ "train_tokens_per_second": 5653.405
+ },
+ {
+ "epoch": 1.004578507945058,
+ "grad_norm": 0.750830922325084,
+ "learning_rate": 4.964569595634706e-06,
+ "loss": 0.3154884338378906,
+ "num_input_tokens_seen": 652509832,
+ "step": 14920,
+ "train_runtime": 115422.4071,
+ "train_tokens_per_second": 5653.234
+ },
+ {
+ "epoch": 1.004915162941018,
+ "grad_norm": 0.7045131043182794,
+ "learning_rate": 4.961925584940559e-06,
+ "loss": 0.3344886779785156,
+ "num_input_tokens_seen": 652737176,
+ "step": 14925,
+ "train_runtime": 115469.7957,
+ "train_tokens_per_second": 5652.882
+ },
+ {
+ "epoch": 1.0052518179369783,
+ "grad_norm": 0.7769156856429779,
+ "learning_rate": 4.959281584893799e-06,
+ "loss": 0.28301913738250734,
+ "num_input_tokens_seen": 652949264,
+ "step": 14930,
+ "train_runtime": 115507.9351,
+ "train_tokens_per_second": 5652.852
+ },
+ {
+ "epoch": 1.0055884729329383,
+ "grad_norm": 0.7698614993301071,
+ "learning_rate": 4.956637596233814e-06,
+ "loss": 0.3222982883453369,
+ "num_input_tokens_seen": 653149432,
+ "step": 14935,
+ "train_runtime": 115542.574,
+ "train_tokens_per_second": 5652.89
+ },
+ {
+ "epoch": 1.0059251279288985,
+ "grad_norm": 0.7927923790873723,
+ "learning_rate": 4.953993619699986e-06,
+ "loss": 0.29294121265411377,
+ "num_input_tokens_seen": 653371256,
+ "step": 14940,
+ "train_runtime": 115587.1554,
+ "train_tokens_per_second": 5652.629
+ },
+ {
+ "epoch": 1.0062617829248586,
+ "grad_norm": 0.7524347822729336,
+ "learning_rate": 4.951349656031698e-06,
+ "loss": 0.30971436500549315,
+ "num_input_tokens_seen": 653582728,
+ "step": 14945,
+ "train_runtime": 115628.4503,
+ "train_tokens_per_second": 5652.439
+ },
+ {
+ "epoch": 1.0065984379208188,
+ "grad_norm": 0.7572597064872421,
+ "learning_rate": 4.948705705968322e-06,
+ "loss": 0.3157053470611572,
+ "num_input_tokens_seen": 653808728,
+ "step": 14950,
+ "train_runtime": 115667.4033,
+ "train_tokens_per_second": 5652.489
+ },
+ {
+ "epoch": 1.0069350929167789,
+ "grad_norm": 0.7025308350419313,
+ "learning_rate": 4.946061770249233e-06,
+ "loss": 0.31903715133666993,
+ "num_input_tokens_seen": 654028536,
+ "step": 14955,
+ "train_runtime": 115709.3504,
+ "train_tokens_per_second": 5652.34
+ },
+ {
+ "epoch": 1.007271747912739,
+ "grad_norm": 0.746389149349819,
+ "learning_rate": 4.9434178496137955e-06,
+ "loss": 0.3309792518615723,
+ "num_input_tokens_seen": 654245296,
+ "step": 14960,
+ "train_runtime": 115746.6069,
+ "train_tokens_per_second": 5652.393
+ },
+ {
+ "epoch": 1.0076084029086991,
+ "grad_norm": 0.6143946299920816,
+ "learning_rate": 4.940773944801378e-06,
+ "loss": 0.31335494518280027,
+ "num_input_tokens_seen": 654464232,
+ "step": 14965,
+ "train_runtime": 115779.8261,
+ "train_tokens_per_second": 5652.662
+ },
+ {
+ "epoch": 1.0079450579046594,
+ "grad_norm": 0.8846423078456821,
+ "learning_rate": 4.938130056551337e-06,
+ "loss": 0.3319870948791504,
+ "num_input_tokens_seen": 654675824,
+ "step": 14970,
+ "train_runtime": 115819.5601,
+ "train_tokens_per_second": 5652.55
+ },
+ {
+ "epoch": 1.0082817129006194,
+ "grad_norm": 0.7105076447650682,
+ "learning_rate": 4.9354861856030305e-06,
+ "loss": 0.3344868183135986,
+ "num_input_tokens_seen": 654897216,
+ "step": 14975,
+ "train_runtime": 115859.9373,
+ "train_tokens_per_second": 5652.491
+ },
+ {
+ "epoch": 1.0086183678965797,
+ "grad_norm": 0.7528024963191167,
+ "learning_rate": 4.932842332695803e-06,
+ "loss": 0.28916707038879397,
+ "num_input_tokens_seen": 655099488,
+ "step": 14980,
+ "train_runtime": 115894.1166,
+ "train_tokens_per_second": 5652.569
+ },
+ {
+ "epoch": 1.0089550228925397,
+ "grad_norm": 0.6760065554104225,
+ "learning_rate": 4.9301984985690055e-06,
+ "loss": 0.31049747467041017,
+ "num_input_tokens_seen": 655334416,
+ "step": 14985,
+ "train_runtime": 115935.8108,
+ "train_tokens_per_second": 5652.562
+ },
+ {
+ "epoch": 1.0092916778885,
+ "grad_norm": 0.8434100907552518,
+ "learning_rate": 4.927554683961974e-06,
+ "loss": 0.3354443073272705,
+ "num_input_tokens_seen": 655543112,
+ "step": 14990,
+ "train_runtime": 115970.966,
+ "train_tokens_per_second": 5652.649
+ },
+ {
+ "epoch": 1.00962833288446,
+ "grad_norm": 0.7283088356363585,
+ "learning_rate": 4.924910889614046e-06,
+ "loss": 0.3087354898452759,
+ "num_input_tokens_seen": 655763808,
+ "step": 14995,
+ "train_runtime": 116009.5582,
+ "train_tokens_per_second": 5652.671
+ },
+ {
+ "epoch": 1.0099649878804202,
+ "grad_norm": 0.7618425432790579,
+ "learning_rate": 4.922267116264546e-06,
+ "loss": 0.33311076164245607,
+ "num_input_tokens_seen": 655973824,
+ "step": 15000,
+ "train_runtime": 116046.9467,
+ "train_tokens_per_second": 5652.659
+ },
+ {
+ "epoch": 1.0103016428763802,
+ "grad_norm": 0.8482961991460483,
+ "learning_rate": 4.919623364652801e-06,
+ "loss": 0.3059119462966919,
+ "num_input_tokens_seen": 656189544,
+ "step": 15005,
+ "train_runtime": 116087.885,
+ "train_tokens_per_second": 5652.524
+ },
+ {
+ "epoch": 1.0106382978723405,
+ "grad_norm": 0.7996774528917011,
+ "learning_rate": 4.916979635518124e-06,
+ "loss": 0.31961765289306643,
+ "num_input_tokens_seen": 656393600,
+ "step": 15010,
+ "train_runtime": 116127.8068,
+ "train_tokens_per_second": 5652.338
+ },
+ {
+ "epoch": 1.0109749528683005,
+ "grad_norm": 0.7485929128993108,
+ "learning_rate": 4.914335929599828e-06,
+ "loss": 0.3200997829437256,
+ "num_input_tokens_seen": 656618032,
+ "step": 15015,
+ "train_runtime": 116165.2315,
+ "train_tokens_per_second": 5652.449
+ },
+ {
+ "epoch": 1.0113116078642608,
+ "grad_norm": 0.8695129879078439,
+ "learning_rate": 4.9116922476372155e-06,
+ "loss": 0.32773504257202146,
+ "num_input_tokens_seen": 656838496,
+ "step": 15020,
+ "train_runtime": 116207.5873,
+ "train_tokens_per_second": 5652.286
+ },
+ {
+ "epoch": 1.0116482628602208,
+ "grad_norm": 0.8589619543857242,
+ "learning_rate": 4.909048590369586e-06,
+ "loss": 0.31574647426605223,
+ "num_input_tokens_seen": 657057672,
+ "step": 15025,
+ "train_runtime": 116240.7138,
+ "train_tokens_per_second": 5652.561
+ },
+ {
+ "epoch": 1.011984917856181,
+ "grad_norm": 0.7207841320471391,
+ "learning_rate": 4.906404958536227e-06,
+ "loss": 0.3148768901824951,
+ "num_input_tokens_seen": 657282040,
+ "step": 15030,
+ "train_runtime": 116286.0079,
+ "train_tokens_per_second": 5652.288
+ },
+ {
+ "epoch": 1.012321572852141,
+ "grad_norm": 0.7575770351222048,
+ "learning_rate": 4.903761352876422e-06,
+ "loss": 0.31647629737854005,
+ "num_input_tokens_seen": 657477448,
+ "step": 15035,
+ "train_runtime": 116329.5257,
+ "train_tokens_per_second": 5651.854
+ },
+ {
+ "epoch": 1.0126582278481013,
+ "grad_norm": 0.8492818928373452,
+ "learning_rate": 4.9011177741294485e-06,
+ "loss": 0.3354557275772095,
+ "num_input_tokens_seen": 657686648,
+ "step": 15040,
+ "train_runtime": 116368.5371,
+ "train_tokens_per_second": 5651.757
+ },
+ {
+ "epoch": 1.0129948828440614,
+ "grad_norm": 0.6899675763130493,
+ "learning_rate": 4.898474223034576e-06,
+ "loss": 0.3180086612701416,
+ "num_input_tokens_seen": 657917808,
+ "step": 15045,
+ "train_runtime": 116410.5268,
+ "train_tokens_per_second": 5651.704
+ },
+ {
+ "epoch": 1.0133315378400216,
+ "grad_norm": 0.7682260059994229,
+ "learning_rate": 4.895830700331061e-06,
+ "loss": 0.31407153606414795,
+ "num_input_tokens_seen": 658124800,
+ "step": 15050,
+ "train_runtime": 116451.6904,
+ "train_tokens_per_second": 5651.483
+ },
+ {
+ "epoch": 1.0136681928359816,
+ "grad_norm": 0.7119767891756066,
+ "learning_rate": 4.893187206758158e-06,
+ "loss": 0.3087860822677612,
+ "num_input_tokens_seen": 658352392,
+ "step": 15055,
+ "train_runtime": 116484.7713,
+ "train_tokens_per_second": 5651.832
+ },
+ {
+ "epoch": 1.014004847831942,
+ "grad_norm": 0.7727900894930202,
+ "learning_rate": 4.890543743055115e-06,
+ "loss": 0.3047621250152588,
+ "num_input_tokens_seen": 658570240,
+ "step": 15060,
+ "train_runtime": 116523.7373,
+ "train_tokens_per_second": 5651.812
+ },
+ {
+ "epoch": 1.014341502827902,
+ "grad_norm": 0.7153606277605876,
+ "learning_rate": 4.887900309961163e-06,
+ "loss": 0.310034441947937,
+ "num_input_tokens_seen": 658791472,
+ "step": 15065,
+ "train_runtime": 116562.1854,
+ "train_tokens_per_second": 5651.846
+ },
+ {
+ "epoch": 1.0146781578238622,
+ "grad_norm": 0.7071012991080629,
+ "learning_rate": 4.885256908215532e-06,
+ "loss": 0.32690412998199464,
+ "num_input_tokens_seen": 659014128,
+ "step": 15070,
+ "train_runtime": 116590.8598,
+ "train_tokens_per_second": 5652.365
+ },
+ {
+ "epoch": 1.0150148128198222,
+ "grad_norm": 0.7712968829018512,
+ "learning_rate": 4.88261353855744e-06,
+ "loss": 0.3006293296813965,
+ "num_input_tokens_seen": 659239712,
+ "step": 15075,
+ "train_runtime": 116632.4209,
+ "train_tokens_per_second": 5652.285
+ },
+ {
+ "epoch": 1.0153514678157824,
+ "grad_norm": 0.6949970194408363,
+ "learning_rate": 4.8799702017261e-06,
+ "loss": 0.28703746795654295,
+ "num_input_tokens_seen": 659455440,
+ "step": 15080,
+ "train_runtime": 116668.3446,
+ "train_tokens_per_second": 5652.394
+ },
+ {
+ "epoch": 1.0156881228117425,
+ "grad_norm": 0.7902536702292094,
+ "learning_rate": 4.877326898460709e-06,
+ "loss": 0.31946167945861814,
+ "num_input_tokens_seen": 659666248,
+ "step": 15085,
+ "train_runtime": 116710.1355,
+ "train_tokens_per_second": 5652.176
+ },
+ {
+ "epoch": 1.0160247778077027,
+ "grad_norm": 0.705619839622007,
+ "learning_rate": 4.87468362950046e-06,
+ "loss": 0.28933143615722656,
+ "num_input_tokens_seen": 659891120,
+ "step": 15090,
+ "train_runtime": 116756.1991,
+ "train_tokens_per_second": 5651.872
+ },
+ {
+ "epoch": 1.0163614328036628,
+ "grad_norm": 0.7513443644619009,
+ "learning_rate": 4.872040395584533e-06,
+ "loss": 0.3063974380493164,
+ "num_input_tokens_seen": 660099960,
+ "step": 15095,
+ "train_runtime": 116792.0331,
+ "train_tokens_per_second": 5651.926
+ },
+ {
+ "epoch": 1.016698087799623,
+ "grad_norm": 0.7643641805946824,
+ "learning_rate": 4.869397197452104e-06,
+ "loss": 0.2840407848358154,
+ "num_input_tokens_seen": 660319352,
+ "step": 15100,
+ "train_runtime": 116827.7471,
+ "train_tokens_per_second": 5652.076
+ },
+ {
+ "epoch": 1.017034742795583,
+ "grad_norm": 0.8009081378183989,
+ "learning_rate": 4.866754035842329e-06,
+ "loss": 0.3019397258758545,
+ "num_input_tokens_seen": 660540440,
+ "step": 15105,
+ "train_runtime": 116864.8835,
+ "train_tokens_per_second": 5652.172
+ },
+ {
+ "epoch": 1.0173713977915433,
+ "grad_norm": 0.7774094479500263,
+ "learning_rate": 4.8641109114943666e-06,
+ "loss": 0.30378923416137693,
+ "num_input_tokens_seen": 660755584,
+ "step": 15110,
+ "train_runtime": 116900.7001,
+ "train_tokens_per_second": 5652.281
+ },
+ {
+ "epoch": 1.0177080527875033,
+ "grad_norm": 0.6851052748232287,
+ "learning_rate": 4.861467825147352e-06,
+ "loss": 0.3153241157531738,
+ "num_input_tokens_seen": 660976736,
+ "step": 15115,
+ "train_runtime": 116933.8046,
+ "train_tokens_per_second": 5652.572
+ },
+ {
+ "epoch": 1.0180447077834636,
+ "grad_norm": 0.7308126890818304,
+ "learning_rate": 4.858824777540419e-06,
+ "loss": 0.3002683401107788,
+ "num_input_tokens_seen": 661194120,
+ "step": 15120,
+ "train_runtime": 116976.4465,
+ "train_tokens_per_second": 5652.37
+ },
+ {
+ "epoch": 1.0183813627794236,
+ "grad_norm": 0.7320711876294524,
+ "learning_rate": 4.856181769412686e-06,
+ "loss": 0.3039540767669678,
+ "num_input_tokens_seen": 661406256,
+ "step": 15125,
+ "train_runtime": 117021.2899,
+ "train_tokens_per_second": 5652.016
+ },
+ {
+ "epoch": 1.0187180177753838,
+ "grad_norm": 0.8338813052362888,
+ "learning_rate": 4.8535388015032644e-06,
+ "loss": 0.30543341636657717,
+ "num_input_tokens_seen": 661615064,
+ "step": 15130,
+ "train_runtime": 117059.1584,
+ "train_tokens_per_second": 5651.972
+ },
+ {
+ "epoch": 1.0190546727713439,
+ "grad_norm": 0.6951071164752909,
+ "learning_rate": 4.850895874551248e-06,
+ "loss": 0.32107832431793215,
+ "num_input_tokens_seen": 661843768,
+ "step": 15135,
+ "train_runtime": 117098.7017,
+ "train_tokens_per_second": 5652.016
+ },
+ {
+ "epoch": 1.0193913277673041,
+ "grad_norm": 0.6999606429191154,
+ "learning_rate": 4.848252989295726e-06,
+ "loss": 0.3267232894897461,
+ "num_input_tokens_seen": 662043496,
+ "step": 15140,
+ "train_runtime": 117135.0098,
+ "train_tokens_per_second": 5651.969
+ },
+ {
+ "epoch": 1.0197279827632642,
+ "grad_norm": 0.7298299606738377,
+ "learning_rate": 4.845610146475771e-06,
+ "loss": 0.2995804786682129,
+ "num_input_tokens_seen": 662264176,
+ "step": 15145,
+ "train_runtime": 117169.4311,
+ "train_tokens_per_second": 5652.192
+ },
+ {
+ "epoch": 1.0200646377592244,
+ "grad_norm": 0.7687173415714617,
+ "learning_rate": 4.842967346830447e-06,
+ "loss": 0.2979107141494751,
+ "num_input_tokens_seen": 662484464,
+ "step": 15150,
+ "train_runtime": 117205.462,
+ "train_tokens_per_second": 5652.334
+ },
+ {
+ "epoch": 1.0204012927551844,
+ "grad_norm": 0.8483763855881413,
+ "learning_rate": 4.840324591098803e-06,
+ "loss": 0.3103074073791504,
+ "num_input_tokens_seen": 662691688,
+ "step": 15155,
+ "train_runtime": 117245.4475,
+ "train_tokens_per_second": 5652.174
+ },
+ {
+ "epoch": 1.0207379477511447,
+ "grad_norm": 0.8319667788025759,
+ "learning_rate": 4.83768188001988e-06,
+ "loss": 0.3106090068817139,
+ "num_input_tokens_seen": 662895200,
+ "step": 15160,
+ "train_runtime": 117290.7287,
+ "train_tokens_per_second": 5651.727
+ },
+ {
+ "epoch": 1.0210746027471047,
+ "grad_norm": 0.7991516150174239,
+ "learning_rate": 4.835039214332699e-06,
+ "loss": 0.30320937633514405,
+ "num_input_tokens_seen": 663097200,
+ "step": 15165,
+ "train_runtime": 117331.5062,
+ "train_tokens_per_second": 5651.485
+ },
+ {
+ "epoch": 1.021411257743065,
+ "grad_norm": 0.7494866477414706,
+ "learning_rate": 4.832396594776277e-06,
+ "loss": 0.3088703155517578,
+ "num_input_tokens_seen": 663319712,
+ "step": 15170,
+ "train_runtime": 117375.1688,
+ "train_tokens_per_second": 5651.278
+ },
+ {
+ "epoch": 1.021747912739025,
+ "grad_norm": 0.7704595139249664,
+ "learning_rate": 4.82975402208961e-06,
+ "loss": 0.2921560764312744,
+ "num_input_tokens_seen": 663537384,
+ "step": 15175,
+ "train_runtime": 117413.8519,
+ "train_tokens_per_second": 5651.27
+ },
+ {
+ "epoch": 1.0220845677349852,
+ "grad_norm": 0.7282417186604674,
+ "learning_rate": 4.827111497011692e-06,
+ "loss": 0.284089994430542,
+ "num_input_tokens_seen": 663777040,
+ "step": 15180,
+ "train_runtime": 117452.9484,
+ "train_tokens_per_second": 5651.429
+ },
+ {
+ "epoch": 1.0224212227309453,
+ "grad_norm": 0.9566982948749372,
+ "learning_rate": 4.82446902028149e-06,
+ "loss": 0.3103042125701904,
+ "num_input_tokens_seen": 663998952,
+ "step": 15185,
+ "train_runtime": 117490.5025,
+ "train_tokens_per_second": 5651.512
+ },
+ {
+ "epoch": 1.0227578777269055,
+ "grad_norm": 0.8002891674482239,
+ "learning_rate": 4.821826592637968e-06,
+ "loss": 0.3256609201431274,
+ "num_input_tokens_seen": 664216976,
+ "step": 15190,
+ "train_runtime": 117525.3723,
+ "train_tokens_per_second": 5651.69
+ },
+ {
+ "epoch": 1.0230945327228655,
+ "grad_norm": 0.6734170041621031,
+ "learning_rate": 4.8191842148200694e-06,
+ "loss": 0.30333666801452636,
+ "num_input_tokens_seen": 664447704,
+ "step": 15195,
+ "train_runtime": 117563.4341,
+ "train_tokens_per_second": 5651.823
+ },
+ {
+ "epoch": 1.0234311877188258,
+ "grad_norm": 0.7333137152011099,
+ "learning_rate": 4.8165418875667315e-06,
+ "loss": 0.3009442090988159,
+ "num_input_tokens_seen": 664678616,
+ "step": 15200,
+ "train_runtime": 117609.1613,
+ "train_tokens_per_second": 5651.589
+ },
+ {
+ "epoch": 1.0237678427147858,
+ "grad_norm": 0.7015384084375479,
+ "learning_rate": 4.813899611616867e-06,
+ "loss": 0.3210268020629883,
+ "num_input_tokens_seen": 664897488,
+ "step": 15205,
+ "train_runtime": 117645.0557,
+ "train_tokens_per_second": 5651.725
+ },
+ {
+ "epoch": 1.024104497710746,
+ "grad_norm": 1.4380458993791456,
+ "learning_rate": 4.8112573877093855e-06,
+ "loss": 0.3092465400695801,
+ "num_input_tokens_seen": 665131288,
+ "step": 15210,
+ "train_runtime": 117686.9545,
+ "train_tokens_per_second": 5651.699
+ },
+ {
+ "epoch": 1.024441152706706,
+ "grad_norm": 0.7003471237764793,
+ "learning_rate": 4.808615216583171e-06,
+ "loss": 0.3192662000656128,
+ "num_input_tokens_seen": 665344872,
+ "step": 15215,
+ "train_runtime": 117720.8851,
+ "train_tokens_per_second": 5651.885
+ },
+ {
+ "epoch": 1.0247778077026664,
+ "grad_norm": 0.8436628825159369,
+ "learning_rate": 4.805973098977102e-06,
+ "loss": 0.32068495750427245,
+ "num_input_tokens_seen": 665556568,
+ "step": 15220,
+ "train_runtime": 117755.2974,
+ "train_tokens_per_second": 5652.031
+ },
+ {
+ "epoch": 1.0251144626986264,
+ "grad_norm": 0.7109267791884789,
+ "learning_rate": 4.803331035630037e-06,
+ "loss": 0.32355175018310545,
+ "num_input_tokens_seen": 665786432,
+ "step": 15225,
+ "train_runtime": 117788.2016,
+ "train_tokens_per_second": 5652.403
+ },
+ {
+ "epoch": 1.0254511176945866,
+ "grad_norm": 0.6877005988993926,
+ "learning_rate": 4.800689027280822e-06,
+ "loss": 0.29448685646057127,
+ "num_input_tokens_seen": 666001856,
+ "step": 15230,
+ "train_runtime": 117828.4813,
+ "train_tokens_per_second": 5652.299
+ },
+ {
+ "epoch": 1.0257877726905467,
+ "grad_norm": 0.7724290603372872,
+ "learning_rate": 4.798047074668283e-06,
+ "loss": 0.2994450330734253,
+ "num_input_tokens_seen": 666220128,
+ "step": 15235,
+ "train_runtime": 117866.0261,
+ "train_tokens_per_second": 5652.351
+ },
+ {
+ "epoch": 1.026124427686507,
+ "grad_norm": 0.6832618434512723,
+ "learning_rate": 4.795405178531236e-06,
+ "loss": 0.29120640754699706,
+ "num_input_tokens_seen": 666430824,
+ "step": 15240,
+ "train_runtime": 117899.98,
+ "train_tokens_per_second": 5652.51
+ },
+ {
+ "epoch": 1.026461082682467,
+ "grad_norm": 0.7900874797047402,
+ "learning_rate": 4.792763339608478e-06,
+ "loss": 0.31427597999572754,
+ "num_input_tokens_seen": 666637408,
+ "step": 15245,
+ "train_runtime": 117941.2447,
+ "train_tokens_per_second": 5652.284
+ },
+ {
+ "epoch": 1.0267977376784272,
+ "grad_norm": 0.830479588475876,
+ "learning_rate": 4.790121558638793e-06,
+ "loss": 0.2697123050689697,
+ "num_input_tokens_seen": 666856712,
+ "step": 15250,
+ "train_runtime": 117983.2012,
+ "train_tokens_per_second": 5652.133
+ },
+ {
+ "epoch": 1.0271343926743872,
+ "grad_norm": 0.702781552620645,
+ "learning_rate": 4.787479836360944e-06,
+ "loss": 0.33616783618927004,
+ "num_input_tokens_seen": 667072880,
+ "step": 15255,
+ "train_runtime": 118015.753,
+ "train_tokens_per_second": 5652.405
+ },
+ {
+ "epoch": 1.0274710476703475,
+ "grad_norm": 0.752937484002602,
+ "learning_rate": 4.784838173513682e-06,
+ "loss": 0.308072566986084,
+ "num_input_tokens_seen": 667290960,
+ "step": 15260,
+ "train_runtime": 118055.7427,
+ "train_tokens_per_second": 5652.338
+ },
+ {
+ "epoch": 1.0278077026663075,
+ "grad_norm": 0.8169678756972052,
+ "learning_rate": 4.782196570835737e-06,
+ "loss": 0.3036174297332764,
+ "num_input_tokens_seen": 667517912,
+ "step": 15265,
+ "train_runtime": 118092.0571,
+ "train_tokens_per_second": 5652.522
+ },
+ {
+ "epoch": 1.0281443576622677,
+ "grad_norm": 0.8398308050050739,
+ "learning_rate": 4.779555029065827e-06,
+ "loss": 0.29191629886627196,
+ "num_input_tokens_seen": 667717400,
+ "step": 15270,
+ "train_runtime": 118131.9033,
+ "train_tokens_per_second": 5652.304
+ },
+ {
+ "epoch": 1.0284810126582278,
+ "grad_norm": 0.7571501727231587,
+ "learning_rate": 4.776913548942649e-06,
+ "loss": 0.31882758140563966,
+ "num_input_tokens_seen": 667943704,
+ "step": 15275,
+ "train_runtime": 118171.6607,
+ "train_tokens_per_second": 5652.317
+ },
+ {
+ "epoch": 1.028817667654188,
+ "grad_norm": 0.7323168837010766,
+ "learning_rate": 4.774272131204889e-06,
+ "loss": 0.31909172534942626,
+ "num_input_tokens_seen": 668171240,
+ "step": 15280,
+ "train_runtime": 118214.9165,
+ "train_tokens_per_second": 5652.174
+ },
+ {
+ "epoch": 1.029154322650148,
+ "grad_norm": 0.8168602912557673,
+ "learning_rate": 4.7716307765912045e-06,
+ "loss": 0.3175906419754028,
+ "num_input_tokens_seen": 668389056,
+ "step": 15285,
+ "train_runtime": 118259.4761,
+ "train_tokens_per_second": 5651.886
+ },
+ {
+ "epoch": 1.0294909776461083,
+ "grad_norm": 0.686585468117989,
+ "learning_rate": 4.768989485840246e-06,
+ "loss": 0.3248094081878662,
+ "num_input_tokens_seen": 668619360,
+ "step": 15290,
+ "train_runtime": 118295.1221,
+ "train_tokens_per_second": 5652.13
+ },
+ {
+ "epoch": 1.0298276326420683,
+ "grad_norm": 0.7842217672393594,
+ "learning_rate": 4.766348259690641e-06,
+ "loss": 0.2917296886444092,
+ "num_input_tokens_seen": 668845224,
+ "step": 15295,
+ "train_runtime": 118327.5434,
+ "train_tokens_per_second": 5652.49
+ },
+ {
+ "epoch": 1.0301642876380286,
+ "grad_norm": 0.8697416761883611,
+ "learning_rate": 4.763707098881002e-06,
+ "loss": 0.3045883893966675,
+ "num_input_tokens_seen": 669059584,
+ "step": 15300,
+ "train_runtime": 118366.6141,
+ "train_tokens_per_second": 5652.435
+ },
+ {
+ "epoch": 1.0305009426339886,
+ "grad_norm": 0.8048541483957821,
+ "learning_rate": 4.761066004149917e-06,
+ "loss": 0.3286751270294189,
+ "num_input_tokens_seen": 669264992,
+ "step": 15305,
+ "train_runtime": 118405.4204,
+ "train_tokens_per_second": 5652.317
+ },
+ {
+ "epoch": 1.0308375976299489,
+ "grad_norm": 0.7590289473967594,
+ "learning_rate": 4.758424976235964e-06,
+ "loss": 0.3184011936187744,
+ "num_input_tokens_seen": 669485880,
+ "step": 15310,
+ "train_runtime": 118445.1028,
+ "train_tokens_per_second": 5652.288
+ },
+ {
+ "epoch": 1.031174252625909,
+ "grad_norm": 0.8012934218342038,
+ "learning_rate": 4.755784015877694e-06,
+ "loss": 0.33234100341796874,
+ "num_input_tokens_seen": 669718792,
+ "step": 15315,
+ "train_runtime": 118479.6014,
+ "train_tokens_per_second": 5652.608
+ },
+ {
+ "epoch": 1.0315109076218691,
+ "grad_norm": 0.7664598006252185,
+ "learning_rate": 4.753143123813646e-06,
+ "loss": 0.3180215358734131,
+ "num_input_tokens_seen": 669934720,
+ "step": 15320,
+ "train_runtime": 118516.7599,
+ "train_tokens_per_second": 5652.658
+ },
+ {
+ "epoch": 1.0318475626178292,
+ "grad_norm": 0.7487344131067605,
+ "learning_rate": 4.7505023007823365e-06,
+ "loss": 0.3165398597717285,
+ "num_input_tokens_seen": 670156656,
+ "step": 15325,
+ "train_runtime": 118558.1181,
+ "train_tokens_per_second": 5652.558
+ },
+ {
+ "epoch": 1.0321842176137894,
+ "grad_norm": 0.6711552412916334,
+ "learning_rate": 4.747861547522264e-06,
+ "loss": 0.28763623237609864,
+ "num_input_tokens_seen": 670377432,
+ "step": 15330,
+ "train_runtime": 118592.5418,
+ "train_tokens_per_second": 5652.779
+ },
+ {
+ "epoch": 1.0325208726097495,
+ "grad_norm": 0.7093961443291146,
+ "learning_rate": 4.745220864771904e-06,
+ "loss": 0.29185004234313966,
+ "num_input_tokens_seen": 670582320,
+ "step": 15335,
+ "train_runtime": 118635.1864,
+ "train_tokens_per_second": 5652.474
+ },
+ {
+ "epoch": 1.0328575276057097,
+ "grad_norm": 0.685756160543914,
+ "learning_rate": 4.742580253269718e-06,
+ "loss": 0.31283178329467776,
+ "num_input_tokens_seen": 670787368,
+ "step": 15340,
+ "train_runtime": 118672.7657,
+ "train_tokens_per_second": 5652.412
+ },
+ {
+ "epoch": 1.0331941826016697,
+ "grad_norm": 0.8426009554601754,
+ "learning_rate": 4.739939713754141e-06,
+ "loss": 0.28745934963226316,
+ "num_input_tokens_seen": 670998288,
+ "step": 15345,
+ "train_runtime": 118718.3964,
+ "train_tokens_per_second": 5652.016
+ },
+ {
+ "epoch": 1.03353083759763,
+ "grad_norm": 0.7932262204171826,
+ "learning_rate": 4.737299246963597e-06,
+ "loss": 0.3006887912750244,
+ "num_input_tokens_seen": 671207256,
+ "step": 15350,
+ "train_runtime": 118756.3685,
+ "train_tokens_per_second": 5651.969
+ },
+ {
+ "epoch": 1.03386749259359,
+ "grad_norm": 0.6568880971934117,
+ "learning_rate": 4.734658853636479e-06,
+ "loss": 0.3070893526077271,
+ "num_input_tokens_seen": 671429312,
+ "step": 15355,
+ "train_runtime": 118793.5087,
+ "train_tokens_per_second": 5652.071
+ },
+ {
+ "epoch": 1.0342041475895503,
+ "grad_norm": 0.8138534659581989,
+ "learning_rate": 4.732018534511167e-06,
+ "loss": 0.3325775146484375,
+ "num_input_tokens_seen": 671648736,
+ "step": 15360,
+ "train_runtime": 118828.7625,
+ "train_tokens_per_second": 5652.24
+ },
+ {
+ "epoch": 1.0345408025855103,
+ "grad_norm": 1.1941290710370875,
+ "learning_rate": 4.729378290326017e-06,
+ "loss": 0.2876759052276611,
+ "num_input_tokens_seen": 671854064,
+ "step": 15365,
+ "train_runtime": 118861.8808,
+ "train_tokens_per_second": 5652.393
+ },
+ {
+ "epoch": 1.0348774575814705,
+ "grad_norm": 0.6797186656709692,
+ "learning_rate": 4.726738121819365e-06,
+ "loss": 0.3200536727905273,
+ "num_input_tokens_seen": 672072384,
+ "step": 15370,
+ "train_runtime": 118895.929,
+ "train_tokens_per_second": 5652.611
+ },
+ {
+ "epoch": 1.0352141125774306,
+ "grad_norm": 0.8461920537238418,
+ "learning_rate": 4.7240980297295255e-06,
+ "loss": 0.2775277137756348,
+ "num_input_tokens_seen": 672283880,
+ "step": 15375,
+ "train_runtime": 118932.7251,
+ "train_tokens_per_second": 5652.64
+ },
+ {
+ "epoch": 1.0355507675733908,
+ "grad_norm": 0.6169331576620465,
+ "learning_rate": 4.721458014794795e-06,
+ "loss": 0.30312108993530273,
+ "num_input_tokens_seen": 672508584,
+ "step": 15380,
+ "train_runtime": 118965.3642,
+ "train_tokens_per_second": 5652.978
+ },
+ {
+ "epoch": 1.0358874225693508,
+ "grad_norm": 0.830035914264151,
+ "learning_rate": 4.718818077753439e-06,
+ "loss": 0.32550725936889646,
+ "num_input_tokens_seen": 672737840,
+ "step": 15385,
+ "train_runtime": 119001.6804,
+ "train_tokens_per_second": 5653.179
+ },
+ {
+ "epoch": 1.036224077565311,
+ "grad_norm": 0.7363851453554499,
+ "learning_rate": 4.716178219343712e-06,
+ "loss": 0.310680890083313,
+ "num_input_tokens_seen": 672951456,
+ "step": 15390,
+ "train_runtime": 119037.6094,
+ "train_tokens_per_second": 5653.268
+ },
+ {
+ "epoch": 1.0365607325612711,
+ "grad_norm": 0.9510291073057959,
+ "learning_rate": 4.713538440303841e-06,
+ "loss": 0.3095111846923828,
+ "num_input_tokens_seen": 673143664,
+ "step": 15395,
+ "train_runtime": 119069.9192,
+ "train_tokens_per_second": 5653.348
+ },
+ {
+ "epoch": 1.0368973875572314,
+ "grad_norm": 0.7829230030914317,
+ "learning_rate": 4.7108987413720335e-06,
+ "loss": 0.3282665491104126,
+ "num_input_tokens_seen": 673355776,
+ "step": 15400,
+ "train_runtime": 119108.1422,
+ "train_tokens_per_second": 5653.314
+ },
+ {
+ "epoch": 1.0372340425531914,
+ "grad_norm": 0.9090900544664394,
+ "learning_rate": 4.708259123286469e-06,
+ "loss": 0.31505331993103025,
+ "num_input_tokens_seen": 673568088,
+ "step": 15405,
+ "train_runtime": 119152.501,
+ "train_tokens_per_second": 5652.992
+ },
+ {
+ "epoch": 1.0375706975491517,
+ "grad_norm": 0.7723625134036335,
+ "learning_rate": 4.7056195867853105e-06,
+ "loss": 0.30636959075927733,
+ "num_input_tokens_seen": 673795064,
+ "step": 15410,
+ "train_runtime": 119195.1556,
+ "train_tokens_per_second": 5652.873
+ },
+ {
+ "epoch": 1.0379073525451117,
+ "grad_norm": 0.7552569039187913,
+ "learning_rate": 4.702980132606696e-06,
+ "loss": 0.2860846996307373,
+ "num_input_tokens_seen": 674021712,
+ "step": 15415,
+ "train_runtime": 119229.7446,
+ "train_tokens_per_second": 5653.134
+ },
+ {
+ "epoch": 1.038244007541072,
+ "grad_norm": 0.7757156949433505,
+ "learning_rate": 4.700340761488741e-06,
+ "loss": 0.297334623336792,
+ "num_input_tokens_seen": 674242240,
+ "step": 15420,
+ "train_runtime": 119259.447,
+ "train_tokens_per_second": 5653.575
+ },
+ {
+ "epoch": 1.038580662537032,
+ "grad_norm": 0.845930651529408,
+ "learning_rate": 4.6977014741695345e-06,
+ "loss": 0.3014695167541504,
+ "num_input_tokens_seen": 674464624,
+ "step": 15425,
+ "train_runtime": 119301.6217,
+ "train_tokens_per_second": 5653.441
+ },
+ {
+ "epoch": 1.0389173175329922,
+ "grad_norm": 0.6863209085436818,
+ "learning_rate": 4.69506227138715e-06,
+ "loss": 0.31081855297088623,
+ "num_input_tokens_seen": 674697456,
+ "step": 15430,
+ "train_runtime": 119344.807,
+ "train_tokens_per_second": 5653.346
+ },
+ {
+ "epoch": 1.0392539725289522,
+ "grad_norm": 0.8027138711324319,
+ "learning_rate": 4.6924231538796265e-06,
+ "loss": 0.3378600120544434,
+ "num_input_tokens_seen": 674921088,
+ "step": 15435,
+ "train_runtime": 119379.2034,
+ "train_tokens_per_second": 5653.59
+ },
+ {
+ "epoch": 1.0395906275249125,
+ "grad_norm": 0.7714672936432438,
+ "learning_rate": 4.6897841223849885e-06,
+ "loss": 0.3055275917053223,
+ "num_input_tokens_seen": 675136400,
+ "step": 15440,
+ "train_runtime": 119423.2981,
+ "train_tokens_per_second": 5653.306
+ },
+ {
+ "epoch": 1.0399272825208725,
+ "grad_norm": 0.7820475842478094,
+ "learning_rate": 4.68714517764123e-06,
+ "loss": 0.2911995887756348,
+ "num_input_tokens_seen": 675354608,
+ "step": 15445,
+ "train_runtime": 119463.4783,
+ "train_tokens_per_second": 5653.231
+ },
+ {
+ "epoch": 1.0402639375168328,
+ "grad_norm": 0.6960997029967657,
+ "learning_rate": 4.684506320386327e-06,
+ "loss": 0.2882180690765381,
+ "num_input_tokens_seen": 675560352,
+ "step": 15450,
+ "train_runtime": 119502.7651,
+ "train_tokens_per_second": 5653.094
+ },
+ {
+ "epoch": 1.0406005925127928,
+ "grad_norm": 0.7240646804471922,
+ "learning_rate": 4.6818675513582235e-06,
+ "loss": 0.2950881481170654,
+ "num_input_tokens_seen": 675780176,
+ "step": 15455,
+ "train_runtime": 119549.6104,
+ "train_tokens_per_second": 5652.718
+ },
+ {
+ "epoch": 1.040937247508753,
+ "grad_norm": 0.7871482559840214,
+ "learning_rate": 4.679228871294846e-06,
+ "loss": 0.2878568649291992,
+ "num_input_tokens_seen": 676001504,
+ "step": 15460,
+ "train_runtime": 119590.0384,
+ "train_tokens_per_second": 5652.657
+ },
+ {
+ "epoch": 1.041273902504713,
+ "grad_norm": 0.7023040971791071,
+ "learning_rate": 4.67659028093409e-06,
+ "loss": 0.31134207248687745,
+ "num_input_tokens_seen": 676224760,
+ "step": 15465,
+ "train_runtime": 119626.8329,
+ "train_tokens_per_second": 5652.785
+ },
+ {
+ "epoch": 1.0416105575006733,
+ "grad_norm": 0.7287009704666462,
+ "learning_rate": 4.673951781013834e-06,
+ "loss": 0.3257226228713989,
+ "num_input_tokens_seen": 676437448,
+ "step": 15470,
+ "train_runtime": 119663.8355,
+ "train_tokens_per_second": 5652.814
+ },
+ {
+ "epoch": 1.0419472124966334,
+ "grad_norm": 0.694252013991942,
+ "learning_rate": 4.67131337227192e-06,
+ "loss": 0.32440500259399413,
+ "num_input_tokens_seen": 676661152,
+ "step": 15475,
+ "train_runtime": 119696.4836,
+ "train_tokens_per_second": 5653.141
+ },
+ {
+ "epoch": 1.0422838674925936,
+ "grad_norm": 0.742357464257659,
+ "learning_rate": 4.668675055446175e-06,
+ "loss": 0.3229750394821167,
+ "num_input_tokens_seen": 676878032,
+ "step": 15480,
+ "train_runtime": 119731.4831,
+ "train_tokens_per_second": 5653.3
+ },
+ {
+ "epoch": 1.0426205224885536,
+ "grad_norm": 0.8706983479448045,
+ "learning_rate": 4.666036831274392e-06,
+ "loss": 0.27671065330505373,
+ "num_input_tokens_seen": 677092008,
+ "step": 15485,
+ "train_runtime": 119767.1182,
+ "train_tokens_per_second": 5653.405
+ },
+ {
+ "epoch": 1.0429571774845139,
+ "grad_norm": 0.7127829768191473,
+ "learning_rate": 4.663398700494346e-06,
+ "loss": 0.30718288421630857,
+ "num_input_tokens_seen": 677302040,
+ "step": 15490,
+ "train_runtime": 119808.142,
+ "train_tokens_per_second": 5653.222
+ },
+ {
+ "epoch": 1.043293832480474,
+ "grad_norm": 0.7433197799578122,
+ "learning_rate": 4.66076066384378e-06,
+ "loss": 0.3258528232574463,
+ "num_input_tokens_seen": 677525784,
+ "step": 15495,
+ "train_runtime": 119841.1792,
+ "train_tokens_per_second": 5653.531
+ },
+ {
+ "epoch": 1.0436304874764342,
+ "grad_norm": 0.6844094807815899,
+ "learning_rate": 4.658122722060413e-06,
+ "loss": 0.3136504888534546,
+ "num_input_tokens_seen": 677745832,
+ "step": 15500,
+ "train_runtime": 119873.5793,
+ "train_tokens_per_second": 5653.838
+ },
+ {
+ "epoch": 1.0439671424723942,
+ "grad_norm": 0.7516388636804747,
+ "learning_rate": 4.655484875881935e-06,
+ "loss": 0.3184205055236816,
+ "num_input_tokens_seen": 677956888,
+ "step": 15505,
+ "train_runtime": 119919.1449,
+ "train_tokens_per_second": 5653.45
+ },
+ {
+ "epoch": 1.0443037974683544,
+ "grad_norm": 0.7798917548462908,
+ "learning_rate": 4.652847126046014e-06,
+ "loss": 0.335153865814209,
+ "num_input_tokens_seen": 678150664,
+ "step": 15510,
+ "train_runtime": 119955.6333,
+ "train_tokens_per_second": 5653.346
+ },
+ {
+ "epoch": 1.0446404524643145,
+ "grad_norm": 0.7432288033280057,
+ "learning_rate": 4.650209473290286e-06,
+ "loss": 0.33000261783599855,
+ "num_input_tokens_seen": 678377032,
+ "step": 15515,
+ "train_runtime": 119990.1304,
+ "train_tokens_per_second": 5653.607
+ },
+ {
+ "epoch": 1.0449771074602747,
+ "grad_norm": 0.7736787887096265,
+ "learning_rate": 4.647571918352366e-06,
+ "loss": 0.31568212509155275,
+ "num_input_tokens_seen": 678595080,
+ "step": 15520,
+ "train_runtime": 120030.7386,
+ "train_tokens_per_second": 5653.511
+ },
+ {
+ "epoch": 1.0453137624562348,
+ "grad_norm": 0.8603539995040905,
+ "learning_rate": 4.644934461969833e-06,
+ "loss": 0.305867600440979,
+ "num_input_tokens_seen": 678818928,
+ "step": 15525,
+ "train_runtime": 120078.8018,
+ "train_tokens_per_second": 5653.112
+ },
+ {
+ "epoch": 1.045650417452195,
+ "grad_norm": 0.802437774953463,
+ "learning_rate": 4.642297104880247e-06,
+ "loss": 0.2917011260986328,
+ "num_input_tokens_seen": 679044840,
+ "step": 15530,
+ "train_runtime": 120117.9302,
+ "train_tokens_per_second": 5653.151
+ },
+ {
+ "epoch": 1.045987072448155,
+ "grad_norm": 0.8097716033620479,
+ "learning_rate": 4.639659847821133e-06,
+ "loss": 0.29921350479125974,
+ "num_input_tokens_seen": 679260504,
+ "step": 15535,
+ "train_runtime": 120155.5327,
+ "train_tokens_per_second": 5653.177
+ },
+ {
+ "epoch": 1.0463237274441153,
+ "grad_norm": 0.7081626022799157,
+ "learning_rate": 4.6370226915299945e-06,
+ "loss": 0.2979047060012817,
+ "num_input_tokens_seen": 679478208,
+ "step": 15540,
+ "train_runtime": 120188.8745,
+ "train_tokens_per_second": 5653.42
+ },
+ {
+ "epoch": 1.0466603824400753,
+ "grad_norm": 0.6768088643103284,
+ "learning_rate": 4.634385636744302e-06,
+ "loss": 0.3120341539382935,
+ "num_input_tokens_seen": 679708344,
+ "step": 15545,
+ "train_runtime": 120226.4885,
+ "train_tokens_per_second": 5653.566
+ },
+ {
+ "epoch": 1.0469970374360356,
+ "grad_norm": 1.008615167669033,
+ "learning_rate": 4.631748684201503e-06,
+ "loss": 0.3280055522918701,
+ "num_input_tokens_seen": 679937752,
+ "step": 15550,
+ "train_runtime": 120267.2677,
+ "train_tokens_per_second": 5653.556
+ },
+ {
+ "epoch": 1.0473336924319956,
+ "grad_norm": 0.7693743879402827,
+ "learning_rate": 4.629111834639008e-06,
+ "loss": 0.3046785593032837,
+ "num_input_tokens_seen": 680165064,
+ "step": 15555,
+ "train_runtime": 120305.2426,
+ "train_tokens_per_second": 5653.661
+ },
+ {
+ "epoch": 1.0476703474279558,
+ "grad_norm": 0.7036312918875381,
+ "learning_rate": 4.626475088794206e-06,
+ "loss": 0.3018321990966797,
+ "num_input_tokens_seen": 680389272,
+ "step": 15560,
+ "train_runtime": 120342.4244,
+ "train_tokens_per_second": 5653.777
+ },
+ {
+ "epoch": 1.0480070024239159,
+ "grad_norm": 0.7427333305833319,
+ "learning_rate": 4.623838447404454e-06,
+ "loss": 0.3129916191101074,
+ "num_input_tokens_seen": 680620144,
+ "step": 15565,
+ "train_runtime": 120385.4131,
+ "train_tokens_per_second": 5653.676
+ },
+ {
+ "epoch": 1.0483436574198761,
+ "grad_norm": 0.6803553069904376,
+ "learning_rate": 4.621201911207082e-06,
+ "loss": 0.29700002670288084,
+ "num_input_tokens_seen": 680832464,
+ "step": 15570,
+ "train_runtime": 120420.9855,
+ "train_tokens_per_second": 5653.769
+ },
+ {
+ "epoch": 1.0486803124158361,
+ "grad_norm": 0.6877375090372397,
+ "learning_rate": 4.618565480939387e-06,
+ "loss": 0.3175528526306152,
+ "num_input_tokens_seen": 681059856,
+ "step": 15575,
+ "train_runtime": 120454.7328,
+ "train_tokens_per_second": 5654.073
+ },
+ {
+ "epoch": 1.0490169674117964,
+ "grad_norm": 0.8334120115163696,
+ "learning_rate": 4.6159291573386395e-06,
+ "loss": 0.3175390958786011,
+ "num_input_tokens_seen": 681279128,
+ "step": 15580,
+ "train_runtime": 120491.4228,
+ "train_tokens_per_second": 5654.171
+ },
+ {
+ "epoch": 1.0493536224077564,
+ "grad_norm": 0.7481436859437062,
+ "learning_rate": 4.613292941142077e-06,
+ "loss": 0.30297222137451174,
+ "num_input_tokens_seen": 681498080,
+ "step": 15585,
+ "train_runtime": 120532.1304,
+ "train_tokens_per_second": 5654.078
+ },
+ {
+ "epoch": 1.0496902774037167,
+ "grad_norm": 0.7558163128951925,
+ "learning_rate": 4.610656833086912e-06,
+ "loss": 0.31702475547790526,
+ "num_input_tokens_seen": 681707736,
+ "step": 15590,
+ "train_runtime": 120571.1677,
+ "train_tokens_per_second": 5653.986
+ },
+ {
+ "epoch": 1.0500269323996767,
+ "grad_norm": 0.8628106650373581,
+ "learning_rate": 4.60802083391032e-06,
+ "loss": 0.3515254259109497,
+ "num_input_tokens_seen": 681929096,
+ "step": 15595,
+ "train_runtime": 120607.6322,
+ "train_tokens_per_second": 5654.112
+ },
+ {
+ "epoch": 1.050363587395637,
+ "grad_norm": 0.8187294344342202,
+ "learning_rate": 4.605384944349454e-06,
+ "loss": 0.29188697338104247,
+ "num_input_tokens_seen": 682138848,
+ "step": 15600,
+ "train_runtime": 120643.9645,
+ "train_tokens_per_second": 5654.148
+ },
+ {
+ "epoch": 1.050700242391597,
+ "grad_norm": 0.762024230166568,
+ "learning_rate": 4.602749165141429e-06,
+ "loss": 0.2967026948928833,
+ "num_input_tokens_seen": 682341504,
+ "step": 15605,
+ "train_runtime": 120680.7614,
+ "train_tokens_per_second": 5654.103
+ },
+ {
+ "epoch": 1.0510368973875572,
+ "grad_norm": 0.770669064131956,
+ "learning_rate": 4.600113497023333e-06,
+ "loss": 0.3200972080230713,
+ "num_input_tokens_seen": 682558744,
+ "step": 15610,
+ "train_runtime": 120718.1512,
+ "train_tokens_per_second": 5654.152
+ },
+ {
+ "epoch": 1.0513735523835175,
+ "grad_norm": 0.8428983487418248,
+ "learning_rate": 4.597477940732222e-06,
+ "loss": 0.2950571537017822,
+ "num_input_tokens_seen": 682762744,
+ "step": 15615,
+ "train_runtime": 120758.4342,
+ "train_tokens_per_second": 5653.955
+ },
+ {
+ "epoch": 1.0517102073794775,
+ "grad_norm": 0.8298281512031902,
+ "learning_rate": 4.594842497005124e-06,
+ "loss": 0.2987983226776123,
+ "num_input_tokens_seen": 682994336,
+ "step": 15620,
+ "train_runtime": 120791.1567,
+ "train_tokens_per_second": 5654.341
+ },
+ {
+ "epoch": 1.0520468623754375,
+ "grad_norm": 0.6720517085495755,
+ "learning_rate": 4.5922071665790264e-06,
+ "loss": 0.3123783111572266,
+ "num_input_tokens_seen": 683207456,
+ "step": 15625,
+ "train_runtime": 120830.8588,
+ "train_tokens_per_second": 5654.246
+ },
+ {
+ "epoch": 1.0523835173713978,
+ "grad_norm": 0.719767612563415,
+ "learning_rate": 4.589571950190897e-06,
+ "loss": 0.3103508949279785,
+ "num_input_tokens_seen": 683437384,
+ "step": 15630,
+ "train_runtime": 120873.1823,
+ "train_tokens_per_second": 5654.169
+ },
+ {
+ "epoch": 1.052720172367358,
+ "grad_norm": 0.7097307825735949,
+ "learning_rate": 4.586936848577661e-06,
+ "loss": 0.3118293285369873,
+ "num_input_tokens_seen": 683657552,
+ "step": 15635,
+ "train_runtime": 120907.5119,
+ "train_tokens_per_second": 5654.384
+ },
+ {
+ "epoch": 1.053056827363318,
+ "grad_norm": 0.7044277499214124,
+ "learning_rate": 4.584301862476219e-06,
+ "loss": 0.30995869636535645,
+ "num_input_tokens_seen": 683876008,
+ "step": 15640,
+ "train_runtime": 120944.8259,
+ "train_tokens_per_second": 5654.446
+ },
+ {
+ "epoch": 1.053393482359278,
+ "grad_norm": 0.7010017324067445,
+ "learning_rate": 4.581666992623435e-06,
+ "loss": 0.2860646963119507,
+ "num_input_tokens_seen": 684104824,
+ "step": 15645,
+ "train_runtime": 120984.5659,
+ "train_tokens_per_second": 5654.48
+ },
+ {
+ "epoch": 1.0537301373552383,
+ "grad_norm": 0.7902236702010951,
+ "learning_rate": 4.579032239756144e-06,
+ "loss": 0.3144806146621704,
+ "num_input_tokens_seen": 684326216,
+ "step": 15650,
+ "train_runtime": 121026.0735,
+ "train_tokens_per_second": 5654.37
+ },
+ {
+ "epoch": 1.0540667923511986,
+ "grad_norm": 0.7099398655983817,
+ "learning_rate": 4.576397604611143e-06,
+ "loss": 0.29763078689575195,
+ "num_input_tokens_seen": 684548424,
+ "step": 15655,
+ "train_runtime": 121060.9714,
+ "train_tokens_per_second": 5654.576
+ },
+ {
+ "epoch": 1.0544034473471586,
+ "grad_norm": 0.7151110990015658,
+ "learning_rate": 4.573763087925202e-06,
+ "loss": 0.30714902877807615,
+ "num_input_tokens_seen": 684772760,
+ "step": 15660,
+ "train_runtime": 121105.9773,
+ "train_tokens_per_second": 5654.327
+ },
+ {
+ "epoch": 1.0547401023431189,
+ "grad_norm": 0.7472326006657539,
+ "learning_rate": 4.571128690435053e-06,
+ "loss": 0.3167744159698486,
+ "num_input_tokens_seen": 684995472,
+ "step": 15665,
+ "train_runtime": 121149.0474,
+ "train_tokens_per_second": 5654.155
+ },
+ {
+ "epoch": 1.055076757339079,
+ "grad_norm": 0.8728278378658701,
+ "learning_rate": 4.568494412877401e-06,
+ "loss": 0.31170904636383057,
+ "num_input_tokens_seen": 685198200,
+ "step": 15670,
+ "train_runtime": 121178.4485,
+ "train_tokens_per_second": 5654.456
+ },
+ {
+ "epoch": 1.0554134123350392,
+ "grad_norm": 0.7744504434630247,
+ "learning_rate": 4.565860255988908e-06,
+ "loss": 0.31058855056762696,
+ "num_input_tokens_seen": 685420416,
+ "step": 15675,
+ "train_runtime": 121210.5142,
+ "train_tokens_per_second": 5654.793
+ },
+ {
+ "epoch": 1.0557500673309992,
+ "grad_norm": 0.7252974178645615,
+ "learning_rate": 4.563226220506212e-06,
+ "loss": 0.31565635204315184,
+ "num_input_tokens_seen": 685659776,
+ "step": 15680,
+ "train_runtime": 121247.6601,
+ "train_tokens_per_second": 5655.035
+ },
+ {
+ "epoch": 1.0560867223269594,
+ "grad_norm": 0.6919225790448248,
+ "learning_rate": 4.5605923071659085e-06,
+ "loss": 0.3112238645553589,
+ "num_input_tokens_seen": 685871672,
+ "step": 15685,
+ "train_runtime": 121283.9917,
+ "train_tokens_per_second": 5655.088
+ },
+ {
+ "epoch": 1.0564233773229195,
+ "grad_norm": 0.7233596710854652,
+ "learning_rate": 4.557958516704566e-06,
+ "loss": 0.2988229990005493,
+ "num_input_tokens_seen": 686092808,
+ "step": 15690,
+ "train_runtime": 121330.4161,
+ "train_tokens_per_second": 5654.747
+ },
+ {
+ "epoch": 1.0567600323188797,
+ "grad_norm": 0.8148391336751524,
+ "learning_rate": 4.5553248498587144e-06,
+ "loss": 0.3125143051147461,
+ "num_input_tokens_seen": 686309592,
+ "step": 15695,
+ "train_runtime": 121360.8019,
+ "train_tokens_per_second": 5655.117
+ },
+ {
+ "epoch": 1.0570966873148397,
+ "grad_norm": 0.8471401120479854,
+ "learning_rate": 4.552691307364853e-06,
+ "loss": 0.31246707439422605,
+ "num_input_tokens_seen": 686520592,
+ "step": 15700,
+ "train_runtime": 121401.4752,
+ "train_tokens_per_second": 5654.961
+ },
+ {
+ "epoch": 1.0574333423108,
+ "grad_norm": 0.8240701136904154,
+ "learning_rate": 4.5500578899594375e-06,
+ "loss": 0.33534066677093505,
+ "num_input_tokens_seen": 686743480,
+ "step": 15705,
+ "train_runtime": 121443.4876,
+ "train_tokens_per_second": 5654.84
+ },
+ {
+ "epoch": 1.05776999730676,
+ "grad_norm": 0.7338513258395319,
+ "learning_rate": 4.5474245983789e-06,
+ "loss": 0.3143591403961182,
+ "num_input_tokens_seen": 686964184,
+ "step": 15710,
+ "train_runtime": 121482.8751,
+ "train_tokens_per_second": 5654.823
+ },
+ {
+ "epoch": 1.0581066523027203,
+ "grad_norm": 0.7196154099105955,
+ "learning_rate": 4.54479143335963e-06,
+ "loss": 0.29776525497436523,
+ "num_input_tokens_seen": 687183192,
+ "step": 15715,
+ "train_runtime": 121528.6289,
+ "train_tokens_per_second": 5654.496
+ },
+ {
+ "epoch": 1.0584433072986803,
+ "grad_norm": 0.8284203342820305,
+ "learning_rate": 4.5421583956379855e-06,
+ "loss": 0.32465856075286864,
+ "num_input_tokens_seen": 687394088,
+ "step": 15720,
+ "train_runtime": 121563.9668,
+ "train_tokens_per_second": 5654.588
+ },
+ {
+ "epoch": 1.0587799622946406,
+ "grad_norm": 0.7128364600635472,
+ "learning_rate": 4.539525485950284e-06,
+ "loss": 0.29742043018341063,
+ "num_input_tokens_seen": 687621296,
+ "step": 15725,
+ "train_runtime": 121603.4696,
+ "train_tokens_per_second": 5654.619
+ },
+ {
+ "epoch": 1.0591166172906006,
+ "grad_norm": 0.8468437589127955,
+ "learning_rate": 4.536892705032815e-06,
+ "loss": 0.3188284397125244,
+ "num_input_tokens_seen": 687846312,
+ "step": 15730,
+ "train_runtime": 121637.3598,
+ "train_tokens_per_second": 5654.893
+ },
+ {
+ "epoch": 1.0594532722865608,
+ "grad_norm": 0.7689731083753425,
+ "learning_rate": 4.534260053621821e-06,
+ "loss": 0.29900522232055665,
+ "num_input_tokens_seen": 688062864,
+ "step": 15735,
+ "train_runtime": 121671.2264,
+ "train_tokens_per_second": 5655.099
+ },
+ {
+ "epoch": 1.0597899272825209,
+ "grad_norm": 0.7661666075190292,
+ "learning_rate": 4.531627532453519e-06,
+ "loss": 0.2740945816040039,
+ "num_input_tokens_seen": 688281760,
+ "step": 15740,
+ "train_runtime": 121719.5434,
+ "train_tokens_per_second": 5654.653
+ },
+ {
+ "epoch": 1.0601265822784811,
+ "grad_norm": 0.7197588574824276,
+ "learning_rate": 4.5289951422640826e-06,
+ "loss": 0.3090967655181885,
+ "num_input_tokens_seen": 688482288,
+ "step": 15745,
+ "train_runtime": 121757.3536,
+ "train_tokens_per_second": 5654.544
+ },
+ {
+ "epoch": 1.0604632372744411,
+ "grad_norm": 0.7581368758886011,
+ "learning_rate": 4.526362883789655e-06,
+ "loss": 0.29933021068572996,
+ "num_input_tokens_seen": 688706904,
+ "step": 15750,
+ "train_runtime": 121799.5328,
+ "train_tokens_per_second": 5654.43
+ },
+ {
+ "epoch": 1.0607998922704014,
+ "grad_norm": 0.7999491894487345,
+ "learning_rate": 4.5237307577663345e-06,
+ "loss": 0.32481019496917723,
+ "num_input_tokens_seen": 688934432,
+ "step": 15755,
+ "train_runtime": 121838.9872,
+ "train_tokens_per_second": 5654.466
+ },
+ {
+ "epoch": 1.0611365472663614,
+ "grad_norm": 0.668582256136596,
+ "learning_rate": 4.52109876493019e-06,
+ "loss": 0.3126739501953125,
+ "num_input_tokens_seen": 689154176,
+ "step": 15760,
+ "train_runtime": 121881.6703,
+ "train_tokens_per_second": 5654.289
+ },
+ {
+ "epoch": 1.0614732022623217,
+ "grad_norm": 0.7253320849615243,
+ "learning_rate": 4.518466906017249e-06,
+ "loss": 0.3287156343460083,
+ "num_input_tokens_seen": 689365536,
+ "step": 15765,
+ "train_runtime": 121920.7449,
+ "train_tokens_per_second": 5654.21
+ },
+ {
+ "epoch": 1.0618098572582817,
+ "grad_norm": 0.9312066051011625,
+ "learning_rate": 4.515835181763503e-06,
+ "loss": 0.29747960567474363,
+ "num_input_tokens_seen": 689584480,
+ "step": 15770,
+ "train_runtime": 121960.8859,
+ "train_tokens_per_second": 5654.145
+ },
+ {
+ "epoch": 1.062146512254242,
+ "grad_norm": 0.779107627410124,
+ "learning_rate": 4.513203592904903e-06,
+ "loss": 0.3129979848861694,
+ "num_input_tokens_seen": 689802248,
+ "step": 15775,
+ "train_runtime": 121996.8975,
+ "train_tokens_per_second": 5654.261
+ },
+ {
+ "epoch": 1.062483167250202,
+ "grad_norm": 0.7066715035398882,
+ "learning_rate": 4.510572140177368e-06,
+ "loss": 0.3110002040863037,
+ "num_input_tokens_seen": 690027296,
+ "step": 15780,
+ "train_runtime": 122041.4381,
+ "train_tokens_per_second": 5654.041
+ },
+ {
+ "epoch": 1.0628198222461622,
+ "grad_norm": 0.7657323857261208,
+ "learning_rate": 4.507940824316771e-06,
+ "loss": 0.31113996505737307,
+ "num_input_tokens_seen": 690244872,
+ "step": 15785,
+ "train_runtime": 122082.9515,
+ "train_tokens_per_second": 5653.901
+ },
+ {
+ "epoch": 1.0631564772421223,
+ "grad_norm": 0.778188387346321,
+ "learning_rate": 4.505309646058954e-06,
+ "loss": 0.306329345703125,
+ "num_input_tokens_seen": 690457904,
+ "step": 15790,
+ "train_runtime": 122123.6242,
+ "train_tokens_per_second": 5653.762
+ },
+ {
+ "epoch": 1.0634931322380825,
+ "grad_norm": 0.8721432879238993,
+ "learning_rate": 4.502678606139716e-06,
+ "loss": 0.290619421005249,
+ "num_input_tokens_seen": 690669592,
+ "step": 15795,
+ "train_runtime": 122161.8006,
+ "train_tokens_per_second": 5653.728
+ },
+ {
+ "epoch": 1.0638297872340425,
+ "grad_norm": 0.6919291628903627,
+ "learning_rate": 4.500047705294823e-06,
+ "loss": 0.28872241973876955,
+ "num_input_tokens_seen": 690892416,
+ "step": 15800,
+ "train_runtime": 122194.5033,
+ "train_tokens_per_second": 5654.038
+ },
+ {
+ "epoch": 1.0641664422300028,
+ "grad_norm": 0.7336900413158081,
+ "learning_rate": 4.497416944259993e-06,
+ "loss": 0.3109187364578247,
+ "num_input_tokens_seen": 691111312,
+ "step": 15805,
+ "train_runtime": 122227.6554,
+ "train_tokens_per_second": 5654.296
+ },
+ {
+ "epoch": 1.0645030972259628,
+ "grad_norm": 0.766581872834498,
+ "learning_rate": 4.494786323770912e-06,
+ "loss": 0.3253108263015747,
+ "num_input_tokens_seen": 691322056,
+ "step": 15810,
+ "train_runtime": 122271.3236,
+ "train_tokens_per_second": 5654.0
+ },
+ {
+ "epoch": 1.064839752221923,
+ "grad_norm": 0.6778315955670481,
+ "learning_rate": 4.492155844563224e-06,
+ "loss": 0.30454020500183104,
+ "num_input_tokens_seen": 691548176,
+ "step": 15815,
+ "train_runtime": 122307.168,
+ "train_tokens_per_second": 5654.192
+ },
+ {
+ "epoch": 1.065176407217883,
+ "grad_norm": 0.7604711413614016,
+ "learning_rate": 4.489525507372537e-06,
+ "loss": 0.3255009651184082,
+ "num_input_tokens_seen": 691767320,
+ "step": 15820,
+ "train_runtime": 122349.9913,
+ "train_tokens_per_second": 5654.004
+ },
+ {
+ "epoch": 1.0655130622138433,
+ "grad_norm": 0.7292388149896591,
+ "learning_rate": 4.486895312934411e-06,
+ "loss": 0.3073612689971924,
+ "num_input_tokens_seen": 691981536,
+ "step": 15825,
+ "train_runtime": 122391.6825,
+ "train_tokens_per_second": 5653.828
+ },
+ {
+ "epoch": 1.0658497172098034,
+ "grad_norm": 0.799787823954383,
+ "learning_rate": 4.484265261984378e-06,
+ "loss": 0.3512305974960327,
+ "num_input_tokens_seen": 692199872,
+ "step": 15830,
+ "train_runtime": 122434.3321,
+ "train_tokens_per_second": 5653.642
+ },
+ {
+ "epoch": 1.0661863722057636,
+ "grad_norm": 0.7027476090234593,
+ "learning_rate": 4.4816353552579165e-06,
+ "loss": 0.292151141166687,
+ "num_input_tokens_seen": 692412728,
+ "step": 15835,
+ "train_runtime": 122476.3937,
+ "train_tokens_per_second": 5653.438
+ },
+ {
+ "epoch": 1.0665230272017237,
+ "grad_norm": 0.8159747493895196,
+ "learning_rate": 4.479005593490477e-06,
+ "loss": 0.3161210536956787,
+ "num_input_tokens_seen": 692630960,
+ "step": 15840,
+ "train_runtime": 122513.6448,
+ "train_tokens_per_second": 5653.501
+ },
+ {
+ "epoch": 1.066859682197684,
+ "grad_norm": 0.7411839574837649,
+ "learning_rate": 4.476375977417461e-06,
+ "loss": 0.28519272804260254,
+ "num_input_tokens_seen": 692854048,
+ "step": 15845,
+ "train_runtime": 122555.3496,
+ "train_tokens_per_second": 5653.397
+ },
+ {
+ "epoch": 1.067196337193644,
+ "grad_norm": 0.7780714296454672,
+ "learning_rate": 4.473746507774235e-06,
+ "loss": 0.3064600944519043,
+ "num_input_tokens_seen": 693081408,
+ "step": 15850,
+ "train_runtime": 122591.5693,
+ "train_tokens_per_second": 5653.581
+ },
+ {
+ "epoch": 1.0675329921896042,
+ "grad_norm": 0.6762102135441501,
+ "learning_rate": 4.471117185296117e-06,
+ "loss": 0.28726582527160643,
+ "num_input_tokens_seen": 693299008,
+ "step": 15855,
+ "train_runtime": 122633.3438,
+ "train_tokens_per_second": 5653.43
+ },
+ {
+ "epoch": 1.0678696471855642,
+ "grad_norm": 0.6740157289152823,
+ "learning_rate": 4.468488010718394e-06,
+ "loss": 0.2859110593795776,
+ "num_input_tokens_seen": 693503552,
+ "step": 15860,
+ "train_runtime": 122669.1306,
+ "train_tokens_per_second": 5653.448
+ },
+ {
+ "epoch": 1.0682063021815245,
+ "grad_norm": 0.8206292602575027,
+ "learning_rate": 4.465858984776302e-06,
+ "loss": 0.31541666984558103,
+ "num_input_tokens_seen": 693719200,
+ "step": 15865,
+ "train_runtime": 122705.5825,
+ "train_tokens_per_second": 5653.526
+ },
+ {
+ "epoch": 1.0685429571774845,
+ "grad_norm": 0.8178983469469215,
+ "learning_rate": 4.463230108205044e-06,
+ "loss": 0.30719547271728515,
+ "num_input_tokens_seen": 693910832,
+ "step": 15870,
+ "train_runtime": 122745.905,
+ "train_tokens_per_second": 5653.23
+ },
+ {
+ "epoch": 1.0688796121734447,
+ "grad_norm": 0.7105445134284586,
+ "learning_rate": 4.460601381739772e-06,
+ "loss": 0.3158994674682617,
+ "num_input_tokens_seen": 694130912,
+ "step": 15875,
+ "train_runtime": 122784.3767,
+ "train_tokens_per_second": 5653.251
+ },
+ {
+ "epoch": 1.0692162671694048,
+ "grad_norm": 0.8182490674352466,
+ "learning_rate": 4.457972806115607e-06,
+ "loss": 0.33564248085021975,
+ "num_input_tokens_seen": 694345664,
+ "step": 15880,
+ "train_runtime": 122826.7003,
+ "train_tokens_per_second": 5653.052
+ },
+ {
+ "epoch": 1.069552922165365,
+ "grad_norm": 0.6520703883013584,
+ "learning_rate": 4.455344382067616e-06,
+ "loss": 0.31251869201660154,
+ "num_input_tokens_seen": 694563184,
+ "step": 15885,
+ "train_runtime": 122869.8883,
+ "train_tokens_per_second": 5652.835
+ },
+ {
+ "epoch": 1.069889577161325,
+ "grad_norm": 0.7437291287192324,
+ "learning_rate": 4.452716110330832e-06,
+ "loss": 0.3079527378082275,
+ "num_input_tokens_seen": 694784912,
+ "step": 15890,
+ "train_runtime": 122904.1033,
+ "train_tokens_per_second": 5653.065
+ },
+ {
+ "epoch": 1.0702262321572853,
+ "grad_norm": 0.8863362961200022,
+ "learning_rate": 4.450087991640242e-06,
+ "loss": 0.31774344444274905,
+ "num_input_tokens_seen": 694988912,
+ "step": 15895,
+ "train_runtime": 122939.5922,
+ "train_tokens_per_second": 5653.093
+ },
+ {
+ "epoch": 1.0705628871532453,
+ "grad_norm": 0.7816781374609185,
+ "learning_rate": 4.447460026730794e-06,
+ "loss": 0.3184868812561035,
+ "num_input_tokens_seen": 695201352,
+ "step": 15900,
+ "train_runtime": 122980.7158,
+ "train_tokens_per_second": 5652.93
+ },
+ {
+ "epoch": 1.0708995421492056,
+ "grad_norm": 0.7651374199427906,
+ "learning_rate": 4.444832216337384e-06,
+ "loss": 0.30800135135650636,
+ "num_input_tokens_seen": 695423576,
+ "step": 15905,
+ "train_runtime": 123019.1946,
+ "train_tokens_per_second": 5652.968
+ },
+ {
+ "epoch": 1.0712361971451656,
+ "grad_norm": 0.9298545556387856,
+ "learning_rate": 4.442204561194877e-06,
+ "loss": 0.29862637519836427,
+ "num_input_tokens_seen": 695642552,
+ "step": 15910,
+ "train_runtime": 123054.9881,
+ "train_tokens_per_second": 5653.103
+ },
+ {
+ "epoch": 1.0715728521411259,
+ "grad_norm": 0.679838258169619,
+ "learning_rate": 4.439577062038085e-06,
+ "loss": 0.29393310546875,
+ "num_input_tokens_seen": 695878256,
+ "step": 15915,
+ "train_runtime": 123095.8131,
+ "train_tokens_per_second": 5653.143
+ },
+ {
+ "epoch": 1.0719095071370859,
+ "grad_norm": 0.7450262188866336,
+ "learning_rate": 4.436949719601782e-06,
+ "loss": 0.3315741062164307,
+ "num_input_tokens_seen": 696096832,
+ "step": 15920,
+ "train_runtime": 123128.696,
+ "train_tokens_per_second": 5653.409
+ },
+ {
+ "epoch": 1.0722461621330461,
+ "grad_norm": 0.9095527446118012,
+ "learning_rate": 4.434322534620692e-06,
+ "loss": 0.3238795757293701,
+ "num_input_tokens_seen": 696300144,
+ "step": 15925,
+ "train_runtime": 123165.6915,
+ "train_tokens_per_second": 5653.361
+ },
+ {
+ "epoch": 1.0725828171290062,
+ "grad_norm": 0.6843805205148538,
+ "learning_rate": 4.431695507829504e-06,
+ "loss": 0.2765275716781616,
+ "num_input_tokens_seen": 696521264,
+ "step": 15930,
+ "train_runtime": 123205.1311,
+ "train_tokens_per_second": 5653.346
+ },
+ {
+ "epoch": 1.0729194721249664,
+ "grad_norm": 0.7473289414033233,
+ "learning_rate": 4.429068639962853e-06,
+ "loss": 0.3107874870300293,
+ "num_input_tokens_seen": 696737432,
+ "step": 15935,
+ "train_runtime": 123245.4597,
+ "train_tokens_per_second": 5653.25
+ },
+ {
+ "epoch": 1.0732561271209264,
+ "grad_norm": 0.7774027394469076,
+ "learning_rate": 4.426441931755337e-06,
+ "loss": 0.3024940490722656,
+ "num_input_tokens_seen": 696961616,
+ "step": 15940,
+ "train_runtime": 123286.7421,
+ "train_tokens_per_second": 5653.176
+ },
+ {
+ "epoch": 1.0735927821168867,
+ "grad_norm": 0.706587053052343,
+ "learning_rate": 4.4238153839415045e-06,
+ "loss": 0.32619800567626955,
+ "num_input_tokens_seen": 697193784,
+ "step": 15945,
+ "train_runtime": 123322.6979,
+ "train_tokens_per_second": 5653.41
+ },
+ {
+ "epoch": 1.0739294371128467,
+ "grad_norm": 0.8216197763150708,
+ "learning_rate": 4.421188997255865e-06,
+ "loss": 0.3372673749923706,
+ "num_input_tokens_seen": 697403080,
+ "step": 15950,
+ "train_runtime": 123358.9718,
+ "train_tokens_per_second": 5653.444
+ },
+ {
+ "epoch": 1.074266092108807,
+ "grad_norm": 0.7673913799601232,
+ "learning_rate": 4.418562772432874e-06,
+ "loss": 0.310252046585083,
+ "num_input_tokens_seen": 697621144,
+ "step": 15955,
+ "train_runtime": 123396.4433,
+ "train_tokens_per_second": 5653.495
+ },
+ {
+ "epoch": 1.074602747104767,
+ "grad_norm": 0.7842485455879602,
+ "learning_rate": 4.415936710206952e-06,
+ "loss": 0.31459197998046873,
+ "num_input_tokens_seen": 697855736,
+ "step": 15960,
+ "train_runtime": 123434.5672,
+ "train_tokens_per_second": 5653.649
+ },
+ {
+ "epoch": 1.0749394021007272,
+ "grad_norm": 0.8604702018619234,
+ "learning_rate": 4.413310811312465e-06,
+ "loss": 0.31325018405914307,
+ "num_input_tokens_seen": 698071128,
+ "step": 15965,
+ "train_runtime": 123468.5762,
+ "train_tokens_per_second": 5653.836
+ },
+ {
+ "epoch": 1.0752760570966873,
+ "grad_norm": 0.746270251899186,
+ "learning_rate": 4.41068507648374e-06,
+ "loss": 0.2819394826889038,
+ "num_input_tokens_seen": 698297384,
+ "step": 15970,
+ "train_runtime": 123503.2252,
+ "train_tokens_per_second": 5654.082
+ },
+ {
+ "epoch": 1.0756127120926475,
+ "grad_norm": 0.7078829630924763,
+ "learning_rate": 4.408059506455053e-06,
+ "loss": 0.29783015251159667,
+ "num_input_tokens_seen": 698536768,
+ "step": 15975,
+ "train_runtime": 123541.9583,
+ "train_tokens_per_second": 5654.247
+ },
+ {
+ "epoch": 1.0759493670886076,
+ "grad_norm": 0.6756334175310369,
+ "learning_rate": 4.40543410196064e-06,
+ "loss": 0.3107433795928955,
+ "num_input_tokens_seen": 698759848,
+ "step": 15980,
+ "train_runtime": 123579.0597,
+ "train_tokens_per_second": 5654.355
+ },
+ {
+ "epoch": 1.0762860220845678,
+ "grad_norm": 0.8743925394845493,
+ "learning_rate": 4.402808863734681e-06,
+ "loss": 0.32532343864440916,
+ "num_input_tokens_seen": 698978376,
+ "step": 15985,
+ "train_runtime": 123625.0383,
+ "train_tokens_per_second": 5654.019
+ },
+ {
+ "epoch": 1.0766226770805278,
+ "grad_norm": 0.7689380089547567,
+ "learning_rate": 4.40018379251132e-06,
+ "loss": 0.3170433044433594,
+ "num_input_tokens_seen": 699198368,
+ "step": 15990,
+ "train_runtime": 123668.6043,
+ "train_tokens_per_second": 5653.807
+ },
+ {
+ "epoch": 1.076959332076488,
+ "grad_norm": 0.8662735325666648,
+ "learning_rate": 4.397558889024649e-06,
+ "loss": 0.350922966003418,
+ "num_input_tokens_seen": 699398432,
+ "step": 15995,
+ "train_runtime": 123704.5755,
+ "train_tokens_per_second": 5653.78
+ },
+ {
+ "epoch": 1.0772959870724481,
+ "grad_norm": 0.6339471211052848,
+ "learning_rate": 4.394934154008715e-06,
+ "loss": 0.3178732395172119,
+ "num_input_tokens_seen": 699622136,
+ "step": 16000,
+ "train_runtime": 123751.9275,
+ "train_tokens_per_second": 5653.424
+ },
+ {
+ "epoch": 1.0776326420684084,
+ "grad_norm": 0.7488843367203586,
+ "learning_rate": 4.392309588197514e-06,
+ "loss": 0.3029392957687378,
+ "num_input_tokens_seen": 699844832,
+ "step": 16005,
+ "train_runtime": 123789.2847,
+ "train_tokens_per_second": 5653.517
+ },
+ {
+ "epoch": 1.0779692970643684,
+ "grad_norm": 0.7019722238512428,
+ "learning_rate": 4.389685192324999e-06,
+ "loss": 0.30281388759613037,
+ "num_input_tokens_seen": 700062776,
+ "step": 16010,
+ "train_runtime": 123831.9289,
+ "train_tokens_per_second": 5653.33
+ },
+ {
+ "epoch": 1.0783059520603286,
+ "grad_norm": 0.6922589184425343,
+ "learning_rate": 4.387060967125073e-06,
+ "loss": 0.3248000144958496,
+ "num_input_tokens_seen": 700270472,
+ "step": 16015,
+ "train_runtime": 123866.5766,
+ "train_tokens_per_second": 5653.426
+ },
+ {
+ "epoch": 1.0786426070562887,
+ "grad_norm": 0.8480144418378566,
+ "learning_rate": 4.384436913331596e-06,
+ "loss": 0.3387366533279419,
+ "num_input_tokens_seen": 700490600,
+ "step": 16020,
+ "train_runtime": 123908.611,
+ "train_tokens_per_second": 5653.284
+ },
+ {
+ "epoch": 1.078979262052249,
+ "grad_norm": 0.7822749047395506,
+ "learning_rate": 4.38181303167837e-06,
+ "loss": 0.3053700923919678,
+ "num_input_tokens_seen": 700708944,
+ "step": 16025,
+ "train_runtime": 123949.1805,
+ "train_tokens_per_second": 5653.195
+ },
+ {
+ "epoch": 1.079315917048209,
+ "grad_norm": 0.7151897163772738,
+ "learning_rate": 4.379189322899159e-06,
+ "loss": 0.29577248096466063,
+ "num_input_tokens_seen": 700917296,
+ "step": 16030,
+ "train_runtime": 123988.781,
+ "train_tokens_per_second": 5653.07
+ },
+ {
+ "epoch": 1.0796525720441692,
+ "grad_norm": 0.6994370560748215,
+ "learning_rate": 4.376565787727676e-06,
+ "loss": 0.30431132316589354,
+ "num_input_tokens_seen": 701140832,
+ "step": 16035,
+ "train_runtime": 124020.3139,
+ "train_tokens_per_second": 5653.435
+ },
+ {
+ "epoch": 1.0799892270401292,
+ "grad_norm": 0.7403775219598513,
+ "learning_rate": 4.373942426897581e-06,
+ "loss": 0.30581619739532473,
+ "num_input_tokens_seen": 701344984,
+ "step": 16040,
+ "train_runtime": 124056.6795,
+ "train_tokens_per_second": 5653.424
+ },
+ {
+ "epoch": 1.0803258820360895,
+ "grad_norm": 0.8197194371899179,
+ "learning_rate": 4.37131924114249e-06,
+ "loss": 0.32230820655822756,
+ "num_input_tokens_seen": 701550320,
+ "step": 16045,
+ "train_runtime": 124096.2964,
+ "train_tokens_per_second": 5653.274
+ },
+ {
+ "epoch": 1.0806625370320495,
+ "grad_norm": 0.7798066192591843,
+ "learning_rate": 4.368696231195968e-06,
+ "loss": 0.3164224624633789,
+ "num_input_tokens_seen": 701773408,
+ "step": 16050,
+ "train_runtime": 124133.9174,
+ "train_tokens_per_second": 5653.357
+ },
+ {
+ "epoch": 1.0809991920280098,
+ "grad_norm": 0.8232305871133864,
+ "learning_rate": 4.366073397791534e-06,
+ "loss": 0.29906353950500486,
+ "num_input_tokens_seen": 701989056,
+ "step": 16055,
+ "train_runtime": 124171.9449,
+ "train_tokens_per_second": 5653.363
+ },
+ {
+ "epoch": 1.0813358470239698,
+ "grad_norm": 0.6949559219075683,
+ "learning_rate": 4.363450741662651e-06,
+ "loss": 0.28748531341552735,
+ "num_input_tokens_seen": 702206288,
+ "step": 16060,
+ "train_runtime": 124208.7109,
+ "train_tokens_per_second": 5653.438
+ },
+ {
+ "epoch": 1.08167250201993,
+ "grad_norm": 0.7364548266367604,
+ "learning_rate": 4.360828263542739e-06,
+ "loss": 0.2912368059158325,
+ "num_input_tokens_seen": 702427720,
+ "step": 16065,
+ "train_runtime": 124248.5519,
+ "train_tokens_per_second": 5653.408
+ },
+ {
+ "epoch": 1.08200915701589,
+ "grad_norm": 0.7936024878904875,
+ "learning_rate": 4.358205964165164e-06,
+ "loss": 0.3195831298828125,
+ "num_input_tokens_seen": 702647704,
+ "step": 16070,
+ "train_runtime": 124288.6557,
+ "train_tokens_per_second": 5653.353
+ },
+ {
+ "epoch": 1.0823458120118503,
+ "grad_norm": 0.7229626629457014,
+ "learning_rate": 4.355583844263247e-06,
+ "loss": 0.32871217727661134,
+ "num_input_tokens_seen": 702872568,
+ "step": 16075,
+ "train_runtime": 124334.326,
+ "train_tokens_per_second": 5653.085
+ },
+ {
+ "epoch": 1.0826824670078103,
+ "grad_norm": 0.8615005299982201,
+ "learning_rate": 4.352961904570251e-06,
+ "loss": 0.3125025272369385,
+ "num_input_tokens_seen": 703089672,
+ "step": 16080,
+ "train_runtime": 124374.4941,
+ "train_tokens_per_second": 5653.005
+ },
+ {
+ "epoch": 1.0830191220037706,
+ "grad_norm": 0.7771590178180461,
+ "learning_rate": 4.350340145819397e-06,
+ "loss": 0.31962432861328127,
+ "num_input_tokens_seen": 703303976,
+ "step": 16085,
+ "train_runtime": 124407.3656,
+ "train_tokens_per_second": 5653.234
+ },
+ {
+ "epoch": 1.0833557769997306,
+ "grad_norm": 0.7052567726864358,
+ "learning_rate": 4.347718568743847e-06,
+ "loss": 0.31935825347900393,
+ "num_input_tokens_seen": 703525192,
+ "step": 16090,
+ "train_runtime": 124447.9997,
+ "train_tokens_per_second": 5653.166
+ },
+ {
+ "epoch": 1.0836924319956909,
+ "grad_norm": 0.6884996142175045,
+ "learning_rate": 4.34509717407672e-06,
+ "loss": 0.32219245433807375,
+ "num_input_tokens_seen": 703758184,
+ "step": 16095,
+ "train_runtime": 124485.0197,
+ "train_tokens_per_second": 5653.356
+ },
+ {
+ "epoch": 1.084029086991651,
+ "grad_norm": 0.7945771259917277,
+ "learning_rate": 4.342475962551081e-06,
+ "loss": 0.3197646141052246,
+ "num_input_tokens_seen": 703976600,
+ "step": 16100,
+ "train_runtime": 124520.4953,
+ "train_tokens_per_second": 5653.5
+ },
+ {
+ "epoch": 1.0843657419876112,
+ "grad_norm": 0.7193933737753864,
+ "learning_rate": 4.339854934899943e-06,
+ "loss": 0.2914067268371582,
+ "num_input_tokens_seen": 704189712,
+ "step": 16105,
+ "train_runtime": 124558.1197,
+ "train_tokens_per_second": 5653.503
+ },
+ {
+ "epoch": 1.0847023969835712,
+ "grad_norm": 0.6245767712435051,
+ "learning_rate": 4.337234091856265e-06,
+ "loss": 0.2758800983428955,
+ "num_input_tokens_seen": 704420104,
+ "step": 16110,
+ "train_runtime": 124598.8261,
+ "train_tokens_per_second": 5653.505
+ },
+ {
+ "epoch": 1.0850390519795314,
+ "grad_norm": 0.7731352613680776,
+ "learning_rate": 4.334613434152961e-06,
+ "loss": 0.3056293249130249,
+ "num_input_tokens_seen": 704628384,
+ "step": 16115,
+ "train_runtime": 124633.8287,
+ "train_tokens_per_second": 5653.589
+ },
+ {
+ "epoch": 1.0853757069754915,
+ "grad_norm": 0.8100290972348857,
+ "learning_rate": 4.331992962522888e-06,
+ "loss": 0.3311766147613525,
+ "num_input_tokens_seen": 704848816,
+ "step": 16120,
+ "train_runtime": 124667.6455,
+ "train_tokens_per_second": 5653.823
+ },
+ {
+ "epoch": 1.0857123619714517,
+ "grad_norm": 0.7302554603263116,
+ "learning_rate": 4.329372677698854e-06,
+ "loss": 0.3087399244308472,
+ "num_input_tokens_seen": 705076168,
+ "step": 16125,
+ "train_runtime": 124705.2351,
+ "train_tokens_per_second": 5653.942
+ },
+ {
+ "epoch": 1.0860490169674117,
+ "grad_norm": 0.7471321256250719,
+ "learning_rate": 4.3267525804136114e-06,
+ "loss": 0.3147927761077881,
+ "num_input_tokens_seen": 705296784,
+ "step": 16130,
+ "train_runtime": 124743.6802,
+ "train_tokens_per_second": 5653.968
+ },
+ {
+ "epoch": 1.086385671963372,
+ "grad_norm": 0.795048965234782,
+ "learning_rate": 4.324132671399864e-06,
+ "loss": 0.3137171268463135,
+ "num_input_tokens_seen": 705500048,
+ "step": 16135,
+ "train_runtime": 124785.7851,
+ "train_tokens_per_second": 5653.689
+ },
+ {
+ "epoch": 1.086722326959332,
+ "grad_norm": 0.7470988295180053,
+ "learning_rate": 4.321512951390258e-06,
+ "loss": 0.3282973289489746,
+ "num_input_tokens_seen": 705720120,
+ "step": 16140,
+ "train_runtime": 124823.8383,
+ "train_tokens_per_second": 5653.729
+ },
+ {
+ "epoch": 1.0870589819552923,
+ "grad_norm": 0.7567687111417263,
+ "learning_rate": 4.318893421117393e-06,
+ "loss": 0.3061366081237793,
+ "num_input_tokens_seen": 705941352,
+ "step": 16145,
+ "train_runtime": 124862.2769,
+ "train_tokens_per_second": 5653.76
+ },
+ {
+ "epoch": 1.0873956369512523,
+ "grad_norm": 0.7636628451001315,
+ "learning_rate": 4.31627408131381e-06,
+ "loss": 0.2968632698059082,
+ "num_input_tokens_seen": 706150776,
+ "step": 16150,
+ "train_runtime": 124908.0524,
+ "train_tokens_per_second": 5653.365
+ },
+ {
+ "epoch": 1.0877322919472125,
+ "grad_norm": 0.7225164738453893,
+ "learning_rate": 4.313654932712001e-06,
+ "loss": 0.2813676357269287,
+ "num_input_tokens_seen": 706367696,
+ "step": 16155,
+ "train_runtime": 124945.3002,
+ "train_tokens_per_second": 5653.415
+ },
+ {
+ "epoch": 1.0880689469431726,
+ "grad_norm": 0.7624100309618296,
+ "learning_rate": 4.3110359760444e-06,
+ "loss": 0.3084303617477417,
+ "num_input_tokens_seen": 706598168,
+ "step": 16160,
+ "train_runtime": 124977.2044,
+ "train_tokens_per_second": 5653.816
+ },
+ {
+ "epoch": 1.0884056019391328,
+ "grad_norm": 0.8194505401549996,
+ "learning_rate": 4.308417212043392e-06,
+ "loss": 0.2846923589706421,
+ "num_input_tokens_seen": 706810256,
+ "step": 16165,
+ "train_runtime": 125015.2796,
+ "train_tokens_per_second": 5653.791
+ },
+ {
+ "epoch": 1.0887422569350929,
+ "grad_norm": 1.0425941904759917,
+ "learning_rate": 4.305798641441304e-06,
+ "loss": 0.2945555210113525,
+ "num_input_tokens_seen": 707037952,
+ "step": 16170,
+ "train_runtime": 125057.1296,
+ "train_tokens_per_second": 5653.72
+ },
+ {
+ "epoch": 1.089078911931053,
+ "grad_norm": 0.7001009592549052,
+ "learning_rate": 4.303180264970416e-06,
+ "loss": 0.3075662851333618,
+ "num_input_tokens_seen": 707251272,
+ "step": 16175,
+ "train_runtime": 125103.0693,
+ "train_tokens_per_second": 5653.349
+ },
+ {
+ "epoch": 1.0894155669270131,
+ "grad_norm": 0.6728629250761784,
+ "learning_rate": 4.3005620833629415e-06,
+ "loss": 0.2843637466430664,
+ "num_input_tokens_seen": 707466896,
+ "step": 16180,
+ "train_runtime": 125144.1227,
+ "train_tokens_per_second": 5653.217
+ },
+ {
+ "epoch": 1.0897522219229734,
+ "grad_norm": 0.8207014522019808,
+ "learning_rate": 4.297944097351053e-06,
+ "loss": 0.3128535270690918,
+ "num_input_tokens_seen": 707669312,
+ "step": 16185,
+ "train_runtime": 125187.0812,
+ "train_tokens_per_second": 5652.894
+ },
+ {
+ "epoch": 1.0900888769189334,
+ "grad_norm": 0.7403515636473879,
+ "learning_rate": 4.295326307666856e-06,
+ "loss": 0.32403197288513186,
+ "num_input_tokens_seen": 707901856,
+ "step": 16190,
+ "train_runtime": 125223.3339,
+ "train_tokens_per_second": 5653.115
+ },
+ {
+ "epoch": 1.0904255319148937,
+ "grad_norm": 0.7331845978184556,
+ "learning_rate": 4.292708715042411e-06,
+ "loss": 0.3054851531982422,
+ "num_input_tokens_seen": 708121472,
+ "step": 16195,
+ "train_runtime": 125262.4718,
+ "train_tokens_per_second": 5653.102
+ },
+ {
+ "epoch": 1.0907621869108537,
+ "grad_norm": 0.8734471364431988,
+ "learning_rate": 4.290091320209718e-06,
+ "loss": 0.32215137481689454,
+ "num_input_tokens_seen": 708352952,
+ "step": 16200,
+ "train_runtime": 125297.9813,
+ "train_tokens_per_second": 5653.347
+ },
+ {
+ "epoch": 1.091098841906814,
+ "grad_norm": 0.814424549291456,
+ "learning_rate": 4.287474123900726e-06,
+ "loss": 0.3153803825378418,
+ "num_input_tokens_seen": 708559152,
+ "step": 16205,
+ "train_runtime": 125333.9985,
+ "train_tokens_per_second": 5653.367
+ },
+ {
+ "epoch": 1.091435496902774,
+ "grad_norm": 0.7017503991057531,
+ "learning_rate": 4.284857126847321e-06,
+ "loss": 0.3124639272689819,
+ "num_input_tokens_seen": 708773424,
+ "step": 16210,
+ "train_runtime": 125375.3591,
+ "train_tokens_per_second": 5653.212
+ },
+ {
+ "epoch": 1.0917721518987342,
+ "grad_norm": 0.7054843976620787,
+ "learning_rate": 4.28224032978134e-06,
+ "loss": 0.32616453170776366,
+ "num_input_tokens_seen": 709001032,
+ "step": 16215,
+ "train_runtime": 125415.5584,
+ "train_tokens_per_second": 5653.214
+ },
+ {
+ "epoch": 1.0921088068946943,
+ "grad_norm": 0.6285131408838311,
+ "learning_rate": 4.279623733434561e-06,
+ "loss": 0.29432287216186526,
+ "num_input_tokens_seen": 709220648,
+ "step": 16220,
+ "train_runtime": 125453.1068,
+ "train_tokens_per_second": 5653.273
+ },
+ {
+ "epoch": 1.0924454618906545,
+ "grad_norm": 0.7070132737008057,
+ "learning_rate": 4.277007338538711e-06,
+ "loss": 0.28021812438964844,
+ "num_input_tokens_seen": 709458880,
+ "step": 16225,
+ "train_runtime": 125489.4284,
+ "train_tokens_per_second": 5653.535
+ },
+ {
+ "epoch": 1.0927821168866145,
+ "grad_norm": 0.6568772287438737,
+ "learning_rate": 4.27439114582545e-06,
+ "loss": 0.278457498550415,
+ "num_input_tokens_seen": 709678888,
+ "step": 16230,
+ "train_runtime": 125528.1241,
+ "train_tokens_per_second": 5653.545
+ },
+ {
+ "epoch": 1.0931187718825748,
+ "grad_norm": 0.7614047551407584,
+ "learning_rate": 4.271775156026395e-06,
+ "loss": 0.3240532636642456,
+ "num_input_tokens_seen": 709881120,
+ "step": 16235,
+ "train_runtime": 125570.4013,
+ "train_tokens_per_second": 5653.252
+ },
+ {
+ "epoch": 1.0934554268785348,
+ "grad_norm": 0.6948637574677056,
+ "learning_rate": 4.269159369873093e-06,
+ "loss": 0.30025343894958495,
+ "num_input_tokens_seen": 710087584,
+ "step": 16240,
+ "train_runtime": 125608.5048,
+ "train_tokens_per_second": 5653.181
+ },
+ {
+ "epoch": 1.093792081874495,
+ "grad_norm": 0.6766287089415108,
+ "learning_rate": 4.266543788097043e-06,
+ "loss": 0.30265159606933595,
+ "num_input_tokens_seen": 710323016,
+ "step": 16245,
+ "train_runtime": 125642.5699,
+ "train_tokens_per_second": 5653.522
+ },
+ {
+ "epoch": 1.094128736870455,
+ "grad_norm": 0.775588372137403,
+ "learning_rate": 4.263928411429683e-06,
+ "loss": 0.34309751987457277,
+ "num_input_tokens_seen": 710535008,
+ "step": 16250,
+ "train_runtime": 125685.9241,
+ "train_tokens_per_second": 5653.258
+ },
+ {
+ "epoch": 1.0944653918664153,
+ "grad_norm": 0.7548771221924332,
+ "learning_rate": 4.261313240602398e-06,
+ "loss": 0.32427158355712893,
+ "num_input_tokens_seen": 710748160,
+ "step": 16255,
+ "train_runtime": 125721.4997,
+ "train_tokens_per_second": 5653.354
+ },
+ {
+ "epoch": 1.0948020468623754,
+ "grad_norm": 0.7250716525242203,
+ "learning_rate": 4.2586982763465064e-06,
+ "loss": 0.3176891326904297,
+ "num_input_tokens_seen": 710976808,
+ "step": 16260,
+ "train_runtime": 125755.3891,
+ "train_tokens_per_second": 5653.649
+ },
+ {
+ "epoch": 1.0951387018583356,
+ "grad_norm": 0.7160308232765491,
+ "learning_rate": 4.256083519393281e-06,
+ "loss": 0.30221972465515134,
+ "num_input_tokens_seen": 711195160,
+ "step": 16265,
+ "train_runtime": 125791.4899,
+ "train_tokens_per_second": 5653.762
+ },
+ {
+ "epoch": 1.0954753568542956,
+ "grad_norm": 0.7399122315680735,
+ "learning_rate": 4.253468970473925e-06,
+ "loss": 0.32169513702392577,
+ "num_input_tokens_seen": 711418152,
+ "step": 16270,
+ "train_runtime": 125823.4265,
+ "train_tokens_per_second": 5654.099
+ },
+ {
+ "epoch": 1.095812011850256,
+ "grad_norm": 0.7365264997369239,
+ "learning_rate": 4.250854630319593e-06,
+ "loss": 0.345171594619751,
+ "num_input_tokens_seen": 711640424,
+ "step": 16275,
+ "train_runtime": 125864.4895,
+ "train_tokens_per_second": 5654.021
+ },
+ {
+ "epoch": 1.096148666846216,
+ "grad_norm": 0.7088190228061766,
+ "learning_rate": 4.248240499661373e-06,
+ "loss": 0.3070409059524536,
+ "num_input_tokens_seen": 711848088,
+ "step": 16280,
+ "train_runtime": 125903.3761,
+ "train_tokens_per_second": 5653.924
+ },
+ {
+ "epoch": 1.0964853218421762,
+ "grad_norm": 0.7753388506945056,
+ "learning_rate": 4.245626579230302e-06,
+ "loss": 0.2884035587310791,
+ "num_input_tokens_seen": 712069664,
+ "step": 16285,
+ "train_runtime": 125935.0217,
+ "train_tokens_per_second": 5654.262
+ },
+ {
+ "epoch": 1.0968219768381362,
+ "grad_norm": 0.898320442866567,
+ "learning_rate": 4.243012869757351e-06,
+ "loss": 0.296952223777771,
+ "num_input_tokens_seen": 712306208,
+ "step": 16290,
+ "train_runtime": 125975.6337,
+ "train_tokens_per_second": 5654.317
+ },
+ {
+ "epoch": 1.0971586318340965,
+ "grad_norm": 0.6948335232020322,
+ "learning_rate": 4.240399371973438e-06,
+ "loss": 0.31108913421630857,
+ "num_input_tokens_seen": 712536912,
+ "step": 16295,
+ "train_runtime": 126015.4524,
+ "train_tokens_per_second": 5654.361
+ },
+ {
+ "epoch": 1.0974952868300565,
+ "grad_norm": 0.7536053030674238,
+ "learning_rate": 4.2377860866094165e-06,
+ "loss": 0.3221109390258789,
+ "num_input_tokens_seen": 712758800,
+ "step": 16300,
+ "train_runtime": 126050.8307,
+ "train_tokens_per_second": 5654.535
+ },
+ {
+ "epoch": 1.0978319418260167,
+ "grad_norm": 0.757140397293046,
+ "learning_rate": 4.235173014396088e-06,
+ "loss": 0.3079149007797241,
+ "num_input_tokens_seen": 712977904,
+ "step": 16305,
+ "train_runtime": 126086.8838,
+ "train_tokens_per_second": 5654.656
+ },
+ {
+ "epoch": 1.0981685968219768,
+ "grad_norm": 0.7077551119296236,
+ "learning_rate": 4.2325601560641854e-06,
+ "loss": 0.3280055046081543,
+ "num_input_tokens_seen": 713204864,
+ "step": 16310,
+ "train_runtime": 126127.6082,
+ "train_tokens_per_second": 5654.629
+ },
+ {
+ "epoch": 1.098505251817937,
+ "grad_norm": 0.742995716486435,
+ "learning_rate": 4.229947512344389e-06,
+ "loss": 0.3023355484008789,
+ "num_input_tokens_seen": 713422520,
+ "step": 16315,
+ "train_runtime": 126168.2023,
+ "train_tokens_per_second": 5654.535
+ },
+ {
+ "epoch": 1.098841906813897,
+ "grad_norm": 0.7574565860803252,
+ "learning_rate": 4.227335083967314e-06,
+ "loss": 0.3024130344390869,
+ "num_input_tokens_seen": 713640344,
+ "step": 16320,
+ "train_runtime": 126208.7161,
+ "train_tokens_per_second": 5654.446
+ },
+ {
+ "epoch": 1.0991785618098573,
+ "grad_norm": 0.6713642233636719,
+ "learning_rate": 4.224722871663522e-06,
+ "loss": 0.3010653018951416,
+ "num_input_tokens_seen": 713862120,
+ "step": 16325,
+ "train_runtime": 126241.4058,
+ "train_tokens_per_second": 5654.738
+ },
+ {
+ "epoch": 1.0995152168058173,
+ "grad_norm": 0.7224131004458003,
+ "learning_rate": 4.222110876163503e-06,
+ "loss": 0.3363258600234985,
+ "num_input_tokens_seen": 714091328,
+ "step": 16330,
+ "train_runtime": 126279.2124,
+ "train_tokens_per_second": 5654.86
+ },
+ {
+ "epoch": 1.0998518718017776,
+ "grad_norm": 0.788806961337755,
+ "learning_rate": 4.2194990981977e-06,
+ "loss": 0.2941534757614136,
+ "num_input_tokens_seen": 714318104,
+ "step": 16335,
+ "train_runtime": 126315.8939,
+ "train_tokens_per_second": 5655.014
+ },
+ {
+ "epoch": 1.1001885267977376,
+ "grad_norm": 0.7340574826834825,
+ "learning_rate": 4.216887538496482e-06,
+ "loss": 0.2995609760284424,
+ "num_input_tokens_seen": 714533544,
+ "step": 16340,
+ "train_runtime": 126356.39,
+ "train_tokens_per_second": 5654.906
+ },
+ {
+ "epoch": 1.1005251817936978,
+ "grad_norm": 0.7563051863525939,
+ "learning_rate": 4.2142761977901695e-06,
+ "loss": 0.3597081184387207,
+ "num_input_tokens_seen": 714751432,
+ "step": 16345,
+ "train_runtime": 126392.7269,
+ "train_tokens_per_second": 5655.004
+ },
+ {
+ "epoch": 1.1008618367896579,
+ "grad_norm": 0.7088542234512013,
+ "learning_rate": 4.21166507680901e-06,
+ "loss": 0.3166736364364624,
+ "num_input_tokens_seen": 714962360,
+ "step": 16350,
+ "train_runtime": 126422.3158,
+ "train_tokens_per_second": 5655.349
+ },
+ {
+ "epoch": 1.1011984917856181,
+ "grad_norm": 0.7396624505443076,
+ "learning_rate": 4.209054176283202e-06,
+ "loss": 0.31155028343200686,
+ "num_input_tokens_seen": 715172784,
+ "step": 16355,
+ "train_runtime": 126456.1758,
+ "train_tokens_per_second": 5655.499
+ },
+ {
+ "epoch": 1.1015351467815782,
+ "grad_norm": 0.762754435287207,
+ "learning_rate": 4.206443496942868e-06,
+ "loss": 0.3161408185958862,
+ "num_input_tokens_seen": 715390344,
+ "step": 16360,
+ "train_runtime": 126495.2215,
+ "train_tokens_per_second": 5655.473
+ },
+ {
+ "epoch": 1.1018718017775384,
+ "grad_norm": 0.976771706771798,
+ "learning_rate": 4.203833039518081e-06,
+ "loss": 0.2889822006225586,
+ "num_input_tokens_seen": 715601072,
+ "step": 16365,
+ "train_runtime": 126533.131,
+ "train_tokens_per_second": 5655.444
+ },
+ {
+ "epoch": 1.1022084567734984,
+ "grad_norm": 0.7390990491778792,
+ "learning_rate": 4.201222804738845e-06,
+ "loss": 0.29911801815032957,
+ "num_input_tokens_seen": 715832864,
+ "step": 16370,
+ "train_runtime": 126567.5723,
+ "train_tokens_per_second": 5655.737
+ },
+ {
+ "epoch": 1.1025451117694587,
+ "grad_norm": 0.7688021513330408,
+ "learning_rate": 4.198612793335107e-06,
+ "loss": 0.28797073364257814,
+ "num_input_tokens_seen": 716055080,
+ "step": 16375,
+ "train_runtime": 126611.3125,
+ "train_tokens_per_second": 5655.538
+ },
+ {
+ "epoch": 1.1028817667654187,
+ "grad_norm": 0.6787164905475817,
+ "learning_rate": 4.196003006036743e-06,
+ "loss": 0.3057976722717285,
+ "num_input_tokens_seen": 716266872,
+ "step": 16380,
+ "train_runtime": 126648.5655,
+ "train_tokens_per_second": 5655.547
+ },
+ {
+ "epoch": 1.103218421761379,
+ "grad_norm": 0.7945554316956748,
+ "learning_rate": 4.193393443573576e-06,
+ "loss": 0.3109589576721191,
+ "num_input_tokens_seen": 716490328,
+ "step": 16385,
+ "train_runtime": 126684.476,
+ "train_tokens_per_second": 5655.707
+ },
+ {
+ "epoch": 1.103555076757339,
+ "grad_norm": 0.7930645433989376,
+ "learning_rate": 4.19078410667536e-06,
+ "loss": 0.3338963985443115,
+ "num_input_tokens_seen": 716702856,
+ "step": 16390,
+ "train_runtime": 126724.8797,
+ "train_tokens_per_second": 5655.581
+ },
+ {
+ "epoch": 1.1038917317532992,
+ "grad_norm": 0.8073944896858126,
+ "learning_rate": 4.188174996071788e-06,
+ "loss": 0.32840120792388916,
+ "num_input_tokens_seen": 716893944,
+ "step": 16395,
+ "train_runtime": 126758.4514,
+ "train_tokens_per_second": 5655.591
+ },
+ {
+ "epoch": 1.1042283867492593,
+ "grad_norm": 0.7471631872859732,
+ "learning_rate": 4.18556611249249e-06,
+ "loss": 0.30612549781799314,
+ "num_input_tokens_seen": 717111816,
+ "step": 16400,
+ "train_runtime": 126792.6813,
+ "train_tokens_per_second": 5655.782
+ },
+ {
+ "epoch": 1.1045650417452195,
+ "grad_norm": 0.8308844608789487,
+ "learning_rate": 4.182957456667032e-06,
+ "loss": 0.3163015127182007,
+ "num_input_tokens_seen": 717335736,
+ "step": 16405,
+ "train_runtime": 126836.6466,
+ "train_tokens_per_second": 5655.587
+ },
+ {
+ "epoch": 1.1049016967411796,
+ "grad_norm": 0.7689707726304607,
+ "learning_rate": 4.1803490293249145e-06,
+ "loss": 0.3357135057449341,
+ "num_input_tokens_seen": 717557144,
+ "step": 16410,
+ "train_runtime": 126873.1786,
+ "train_tokens_per_second": 5655.704
+ },
+ {
+ "epoch": 1.1052383517371398,
+ "grad_norm": 0.866900091357806,
+ "learning_rate": 4.1777408311955785e-06,
+ "loss": 0.3057410717010498,
+ "num_input_tokens_seen": 717761472,
+ "step": 16415,
+ "train_runtime": 126909.3722,
+ "train_tokens_per_second": 5655.701
+ },
+ {
+ "epoch": 1.1055750067330998,
+ "grad_norm": 0.7300512821916881,
+ "learning_rate": 4.175132863008395e-06,
+ "loss": 0.33652937412261963,
+ "num_input_tokens_seen": 717986048,
+ "step": 16420,
+ "train_runtime": 126949.4267,
+ "train_tokens_per_second": 5655.686
+ },
+ {
+ "epoch": 1.10591166172906,
+ "grad_norm": 0.9194268863297489,
+ "learning_rate": 4.17252512549268e-06,
+ "loss": 0.3076220273971558,
+ "num_input_tokens_seen": 718179376,
+ "step": 16425,
+ "train_runtime": 126987.7182,
+ "train_tokens_per_second": 5655.503
+ },
+ {
+ "epoch": 1.10624831672502,
+ "grad_norm": 0.7503337161594836,
+ "learning_rate": 4.169917619377671e-06,
+ "loss": 0.3038554906845093,
+ "num_input_tokens_seen": 718392416,
+ "step": 16430,
+ "train_runtime": 127030.797,
+ "train_tokens_per_second": 5655.262
+ },
+ {
+ "epoch": 1.1065849717209804,
+ "grad_norm": 0.8669099466716282,
+ "learning_rate": 4.167310345392557e-06,
+ "loss": 0.34373929500579836,
+ "num_input_tokens_seen": 718606440,
+ "step": 16435,
+ "train_runtime": 127066.9236,
+ "train_tokens_per_second": 5655.338
+ },
+ {
+ "epoch": 1.1069216267169404,
+ "grad_norm": 0.7525899229778,
+ "learning_rate": 4.1647033042664455e-06,
+ "loss": 0.3425747871398926,
+ "num_input_tokens_seen": 718818792,
+ "step": 16440,
+ "train_runtime": 127105.1604,
+ "train_tokens_per_second": 5655.308
+ },
+ {
+ "epoch": 1.1072582817129006,
+ "grad_norm": 0.7395303188295309,
+ "learning_rate": 4.162096496728394e-06,
+ "loss": 0.3010833263397217,
+ "num_input_tokens_seen": 719039416,
+ "step": 16445,
+ "train_runtime": 127146.3098,
+ "train_tokens_per_second": 5655.213
+ },
+ {
+ "epoch": 1.1075949367088607,
+ "grad_norm": 0.7264360847991507,
+ "learning_rate": 4.1594899235073835e-06,
+ "loss": 0.2992356300354004,
+ "num_input_tokens_seen": 719268528,
+ "step": 16450,
+ "train_runtime": 127190.5278,
+ "train_tokens_per_second": 5655.048
+ },
+ {
+ "epoch": 1.107931591704821,
+ "grad_norm": 0.7474728546722258,
+ "learning_rate": 4.156883585332338e-06,
+ "loss": 0.33159723281860354,
+ "num_input_tokens_seen": 719497592,
+ "step": 16455,
+ "train_runtime": 127229.1668,
+ "train_tokens_per_second": 5655.131
+ },
+ {
+ "epoch": 1.108268246700781,
+ "grad_norm": 0.6867746878864566,
+ "learning_rate": 4.154277482932107e-06,
+ "loss": 0.31698899269104003,
+ "num_input_tokens_seen": 719703328,
+ "step": 16460,
+ "train_runtime": 127263.452,
+ "train_tokens_per_second": 5655.224
+ },
+ {
+ "epoch": 1.1086049016967412,
+ "grad_norm": 0.7188428455380581,
+ "learning_rate": 4.151671617035482e-06,
+ "loss": 0.30119948387145995,
+ "num_input_tokens_seen": 719915352,
+ "step": 16465,
+ "train_runtime": 127303.2856,
+ "train_tokens_per_second": 5655.12
+ },
+ {
+ "epoch": 1.1089415566927012,
+ "grad_norm": 0.7701204509330969,
+ "learning_rate": 4.149065988371184e-06,
+ "loss": 0.2912993669509888,
+ "num_input_tokens_seen": 720134040,
+ "step": 16470,
+ "train_runtime": 127336.2618,
+ "train_tokens_per_second": 5655.373
+ },
+ {
+ "epoch": 1.1092782116886615,
+ "grad_norm": 0.7513429280225311,
+ "learning_rate": 4.14646059766787e-06,
+ "loss": 0.30213978290557864,
+ "num_input_tokens_seen": 720337824,
+ "step": 16475,
+ "train_runtime": 127369.5743,
+ "train_tokens_per_second": 5655.494
+ },
+ {
+ "epoch": 1.1096148666846215,
+ "grad_norm": 0.6711604996544339,
+ "learning_rate": 4.143855445654127e-06,
+ "loss": 0.29933414459228513,
+ "num_input_tokens_seen": 720567936,
+ "step": 16480,
+ "train_runtime": 127409.6507,
+ "train_tokens_per_second": 5655.521
+ },
+ {
+ "epoch": 1.1099515216805818,
+ "grad_norm": 0.8475234440498322,
+ "learning_rate": 4.14125053305848e-06,
+ "loss": 0.2931232213973999,
+ "num_input_tokens_seen": 720770472,
+ "step": 16485,
+ "train_runtime": 127446.8624,
+ "train_tokens_per_second": 5655.459
+ },
+ {
+ "epoch": 1.1102881766765418,
+ "grad_norm": 0.7010224781863986,
+ "learning_rate": 4.1386458606093805e-06,
+ "loss": 0.30602593421936036,
+ "num_input_tokens_seen": 720990928,
+ "step": 16490,
+ "train_runtime": 127487.302,
+ "train_tokens_per_second": 5655.394
+ },
+ {
+ "epoch": 1.110624831672502,
+ "grad_norm": 0.7499446880256623,
+ "learning_rate": 4.1360414290352214e-06,
+ "loss": 0.3390794038772583,
+ "num_input_tokens_seen": 721222304,
+ "step": 16495,
+ "train_runtime": 127531.9109,
+ "train_tokens_per_second": 5655.23
+ },
+ {
+ "epoch": 1.110961486668462,
+ "grad_norm": 0.7785896374993136,
+ "learning_rate": 4.1334372390643205e-06,
+ "loss": 0.30769715309143064,
+ "num_input_tokens_seen": 721451296,
+ "step": 16500,
+ "train_runtime": 127572.4988,
+ "train_tokens_per_second": 5655.226
+ },
+ {
+ "epoch": 1.1112981416644223,
+ "grad_norm": 0.6703802322842832,
+ "learning_rate": 4.130833291424934e-06,
+ "loss": 0.32526416778564454,
+ "num_input_tokens_seen": 721672688,
+ "step": 16505,
+ "train_runtime": 127609.7092,
+ "train_tokens_per_second": 5655.312
+ },
+ {
+ "epoch": 1.1116347966603823,
+ "grad_norm": 0.8080822522673832,
+ "learning_rate": 4.128229586845244e-06,
+ "loss": 0.30047011375427246,
+ "num_input_tokens_seen": 721893296,
+ "step": 16510,
+ "train_runtime": 127652.4127,
+ "train_tokens_per_second": 5655.148
+ },
+ {
+ "epoch": 1.1119714516563426,
+ "grad_norm": 0.6398810192890917,
+ "learning_rate": 4.125626126053372e-06,
+ "loss": 0.2854645013809204,
+ "num_input_tokens_seen": 722117224,
+ "step": 16515,
+ "train_runtime": 127693.4791,
+ "train_tokens_per_second": 5655.083
+ },
+ {
+ "epoch": 1.1123081066523026,
+ "grad_norm": 0.7428717458853391,
+ "learning_rate": 4.123022909777364e-06,
+ "loss": 0.31157827377319336,
+ "num_input_tokens_seen": 722338424,
+ "step": 16520,
+ "train_runtime": 127724.0973,
+ "train_tokens_per_second": 5655.459
+ },
+ {
+ "epoch": 1.1126447616482629,
+ "grad_norm": 0.8547681854868342,
+ "learning_rate": 4.120419938745205e-06,
+ "loss": 0.31059443950653076,
+ "num_input_tokens_seen": 722572192,
+ "step": 16525,
+ "train_runtime": 127768.6413,
+ "train_tokens_per_second": 5655.317
+ },
+ {
+ "epoch": 1.112981416644223,
+ "grad_norm": 0.8535266815266339,
+ "learning_rate": 4.1178172136848035e-06,
+ "loss": 0.30442132949829104,
+ "num_input_tokens_seen": 722777960,
+ "step": 16530,
+ "train_runtime": 127805.8834,
+ "train_tokens_per_second": 5655.279
+ },
+ {
+ "epoch": 1.1133180716401831,
+ "grad_norm": 0.7715289136002668,
+ "learning_rate": 4.115214735324007e-06,
+ "loss": 0.29233555793762206,
+ "num_input_tokens_seen": 723011784,
+ "step": 16535,
+ "train_runtime": 127845.3251,
+ "train_tokens_per_second": 5655.363
+ },
+ {
+ "epoch": 1.1136547266361432,
+ "grad_norm": 0.6308478276358274,
+ "learning_rate": 4.112612504390586e-06,
+ "loss": 0.2731541872024536,
+ "num_input_tokens_seen": 723233976,
+ "step": 16540,
+ "train_runtime": 127887.594,
+ "train_tokens_per_second": 5655.232
+ },
+ {
+ "epoch": 1.1139913816321034,
+ "grad_norm": 0.7595556917317189,
+ "learning_rate": 4.11001052161225e-06,
+ "loss": 0.3273204803466797,
+ "num_input_tokens_seen": 723450040,
+ "step": 16545,
+ "train_runtime": 127924.3668,
+ "train_tokens_per_second": 5655.295
+ },
+ {
+ "epoch": 1.1143280366280637,
+ "grad_norm": 0.7539458359188388,
+ "learning_rate": 4.107408787716632e-06,
+ "loss": 0.3083867788314819,
+ "num_input_tokens_seen": 723674424,
+ "step": 16550,
+ "train_runtime": 127966.9008,
+ "train_tokens_per_second": 5655.169
+ },
+ {
+ "epoch": 1.1146646916240237,
+ "grad_norm": 0.6933389096869937,
+ "learning_rate": 4.104807303431303e-06,
+ "loss": 0.30321598052978516,
+ "num_input_tokens_seen": 723898176,
+ "step": 16555,
+ "train_runtime": 128013.6102,
+ "train_tokens_per_second": 5654.853
+ },
+ {
+ "epoch": 1.1150013466199837,
+ "grad_norm": 0.8314161994474075,
+ "learning_rate": 4.102206069483754e-06,
+ "loss": 0.32450571060180666,
+ "num_input_tokens_seen": 724110632,
+ "step": 16560,
+ "train_runtime": 128050.9045,
+ "train_tokens_per_second": 5654.865
+ },
+ {
+ "epoch": 1.115338001615944,
+ "grad_norm": 0.7582409248110241,
+ "learning_rate": 4.099605086601417e-06,
+ "loss": 0.28465938568115234,
+ "num_input_tokens_seen": 724312128,
+ "step": 16565,
+ "train_runtime": 128098.5725,
+ "train_tokens_per_second": 5654.334
+ },
+ {
+ "epoch": 1.1156746566119042,
+ "grad_norm": 0.7723226944103582,
+ "learning_rate": 4.097004355511645e-06,
+ "loss": 0.2872473239898682,
+ "num_input_tokens_seen": 724532224,
+ "step": 16570,
+ "train_runtime": 128135.0872,
+ "train_tokens_per_second": 5654.44
+ },
+ {
+ "epoch": 1.1160113116078643,
+ "grad_norm": 0.80551126446832,
+ "learning_rate": 4.094403876941728e-06,
+ "loss": 0.2786348819732666,
+ "num_input_tokens_seen": 724749712,
+ "step": 16575,
+ "train_runtime": 128172.2871,
+ "train_tokens_per_second": 5654.496
+ },
+ {
+ "epoch": 1.1163479666038243,
+ "grad_norm": 0.7081089427677039,
+ "learning_rate": 4.0918036516188755e-06,
+ "loss": 0.3147793531417847,
+ "num_input_tokens_seen": 724953832,
+ "step": 16580,
+ "train_runtime": 128215.8528,
+ "train_tokens_per_second": 5654.167
+ },
+ {
+ "epoch": 1.1166846215997845,
+ "grad_norm": 0.7234281848413817,
+ "learning_rate": 4.0892036802702395e-06,
+ "loss": 0.3022450923919678,
+ "num_input_tokens_seen": 725178424,
+ "step": 16585,
+ "train_runtime": 128258.0871,
+ "train_tokens_per_second": 5654.056
+ },
+ {
+ "epoch": 1.1170212765957448,
+ "grad_norm": 0.6801507964833571,
+ "learning_rate": 4.086603963622887e-06,
+ "loss": 0.2855158090591431,
+ "num_input_tokens_seen": 725402000,
+ "step": 16590,
+ "train_runtime": 128299.5805,
+ "train_tokens_per_second": 5653.97
+ },
+ {
+ "epoch": 1.1173579315917048,
+ "grad_norm": 0.712093664671178,
+ "learning_rate": 4.084004502403825e-06,
+ "loss": 0.32064228057861327,
+ "num_input_tokens_seen": 725622480,
+ "step": 16595,
+ "train_runtime": 128341.8001,
+ "train_tokens_per_second": 5653.828
+ },
+ {
+ "epoch": 1.1176945865876649,
+ "grad_norm": 0.7068967114722002,
+ "learning_rate": 4.081405297339982e-06,
+ "loss": 0.30630931854248045,
+ "num_input_tokens_seen": 725856392,
+ "step": 16600,
+ "train_runtime": 128375.2241,
+ "train_tokens_per_second": 5654.178
+ },
+ {
+ "epoch": 1.118031241583625,
+ "grad_norm": 0.7257036824000926,
+ "learning_rate": 4.0788063491582205e-06,
+ "loss": 0.3028933048248291,
+ "num_input_tokens_seen": 726063792,
+ "step": 16605,
+ "train_runtime": 128411.5161,
+ "train_tokens_per_second": 5654.195
+ },
+ {
+ "epoch": 1.1183678965795854,
+ "grad_norm": 0.8569267027392826,
+ "learning_rate": 4.076207658585323e-06,
+ "loss": 0.3076032638549805,
+ "num_input_tokens_seen": 726293720,
+ "step": 16610,
+ "train_runtime": 128453.0536,
+ "train_tokens_per_second": 5654.157
+ },
+ {
+ "epoch": 1.1187045515755454,
+ "grad_norm": 0.6859953010109835,
+ "learning_rate": 4.07360922634801e-06,
+ "loss": 0.31174705028533933,
+ "num_input_tokens_seen": 726522048,
+ "step": 16615,
+ "train_runtime": 128490.1199,
+ "train_tokens_per_second": 5654.303
+ },
+ {
+ "epoch": 1.1190412065715054,
+ "grad_norm": 0.8064627601454211,
+ "learning_rate": 4.071011053172922e-06,
+ "loss": 0.3052211284637451,
+ "num_input_tokens_seen": 726722760,
+ "step": 16620,
+ "train_runtime": 128528.4186,
+ "train_tokens_per_second": 5654.18
+ },
+ {
+ "epoch": 1.1193778615674657,
+ "grad_norm": 0.7719182024204673,
+ "learning_rate": 4.068413139786631e-06,
+ "loss": 0.29850754737854,
+ "num_input_tokens_seen": 726939496,
+ "step": 16625,
+ "train_runtime": 128567.5214,
+ "train_tokens_per_second": 5654.146
+ },
+ {
+ "epoch": 1.119714516563426,
+ "grad_norm": 0.7949931481273091,
+ "learning_rate": 4.065815486915635e-06,
+ "loss": 0.30380845069885254,
+ "num_input_tokens_seen": 727169648,
+ "step": 16630,
+ "train_runtime": 128605.7452,
+ "train_tokens_per_second": 5654.255
+ },
+ {
+ "epoch": 1.120051171559386,
+ "grad_norm": 0.7231190372167858,
+ "learning_rate": 4.063218095286359e-06,
+ "loss": 0.31917145252227785,
+ "num_input_tokens_seen": 727399600,
+ "step": 16635,
+ "train_runtime": 128642.9376,
+ "train_tokens_per_second": 5654.408
+ },
+ {
+ "epoch": 1.120387826555346,
+ "grad_norm": 0.7667748207910186,
+ "learning_rate": 4.0606209656251555e-06,
+ "loss": 0.30777583122253416,
+ "num_input_tokens_seen": 727623584,
+ "step": 16640,
+ "train_runtime": 128679.2099,
+ "train_tokens_per_second": 5654.554
+ },
+ {
+ "epoch": 1.1207244815513062,
+ "grad_norm": 0.7358231463616071,
+ "learning_rate": 4.058024098658304e-06,
+ "loss": 0.32628979682922366,
+ "num_input_tokens_seen": 727849768,
+ "step": 16645,
+ "train_runtime": 128721.1978,
+ "train_tokens_per_second": 5654.467
+ },
+ {
+ "epoch": 1.1210611365472665,
+ "grad_norm": 0.7289574965983563,
+ "learning_rate": 4.055427495112008e-06,
+ "loss": 0.33713221549987793,
+ "num_input_tokens_seen": 728072752,
+ "step": 16650,
+ "train_runtime": 128764.41,
+ "train_tokens_per_second": 5654.301
+ },
+ {
+ "epoch": 1.1213977915432265,
+ "grad_norm": 0.7302453800120214,
+ "learning_rate": 4.052831155712403e-06,
+ "loss": 0.28829092979431153,
+ "num_input_tokens_seen": 728297032,
+ "step": 16655,
+ "train_runtime": 128798.8316,
+ "train_tokens_per_second": 5654.531
+ },
+ {
+ "epoch": 1.1217344465391865,
+ "grad_norm": 0.6736314639659831,
+ "learning_rate": 4.050235081185544e-06,
+ "loss": 0.27930502891540526,
+ "num_input_tokens_seen": 728518648,
+ "step": 16660,
+ "train_runtime": 128837.9436,
+ "train_tokens_per_second": 5654.535
+ },
+ {
+ "epoch": 1.1220711015351468,
+ "grad_norm": 0.778226843935622,
+ "learning_rate": 4.047639272257415e-06,
+ "loss": 0.31568381786346433,
+ "num_input_tokens_seen": 728747776,
+ "step": 16665,
+ "train_runtime": 128881.478,
+ "train_tokens_per_second": 5654.403
+ },
+ {
+ "epoch": 1.122407756531107,
+ "grad_norm": 0.616922697066709,
+ "learning_rate": 4.045043729653927e-06,
+ "loss": 0.2794900894165039,
+ "num_input_tokens_seen": 728975240,
+ "step": 16670,
+ "train_runtime": 128920.0081,
+ "train_tokens_per_second": 5654.477
+ },
+ {
+ "epoch": 1.122744411527067,
+ "grad_norm": 0.7407200139322001,
+ "learning_rate": 4.0424484541009175e-06,
+ "loss": 0.3211031436920166,
+ "num_input_tokens_seen": 729207240,
+ "step": 16675,
+ "train_runtime": 128961.9626,
+ "train_tokens_per_second": 5654.437
+ },
+ {
+ "epoch": 1.1230810665230273,
+ "grad_norm": 0.8208779512348747,
+ "learning_rate": 4.039853446324141e-06,
+ "loss": 0.2974057197570801,
+ "num_input_tokens_seen": 729416088,
+ "step": 16680,
+ "train_runtime": 129004.1657,
+ "train_tokens_per_second": 5654.206
+ },
+ {
+ "epoch": 1.1234177215189873,
+ "grad_norm": 0.7898967563608909,
+ "learning_rate": 4.037258707049289e-06,
+ "loss": 0.2974997043609619,
+ "num_input_tokens_seen": 729633568,
+ "step": 16685,
+ "train_runtime": 129048.8248,
+ "train_tokens_per_second": 5653.934
+ },
+ {
+ "epoch": 1.1237543765149476,
+ "grad_norm": 0.6741454248192755,
+ "learning_rate": 4.034664237001969e-06,
+ "loss": 0.29579739570617675,
+ "num_input_tokens_seen": 729855424,
+ "step": 16690,
+ "train_runtime": 129087.6374,
+ "train_tokens_per_second": 5653.953
+ },
+ {
+ "epoch": 1.1240910315109076,
+ "grad_norm": 0.8311896058049114,
+ "learning_rate": 4.032070036907716e-06,
+ "loss": 0.3193542003631592,
+ "num_input_tokens_seen": 730062656,
+ "step": 16695,
+ "train_runtime": 129126.9171,
+ "train_tokens_per_second": 5653.838
+ },
+ {
+ "epoch": 1.1244276865068679,
+ "grad_norm": 0.7991665442055015,
+ "learning_rate": 4.029476107491991e-06,
+ "loss": 0.3203211784362793,
+ "num_input_tokens_seen": 730282736,
+ "step": 16700,
+ "train_runtime": 129171.6017,
+ "train_tokens_per_second": 5653.586
+ },
+ {
+ "epoch": 1.124764341502828,
+ "grad_norm": 0.7190813419301547,
+ "learning_rate": 4.0268824494801784e-06,
+ "loss": 0.30711615085601807,
+ "num_input_tokens_seen": 730503648,
+ "step": 16705,
+ "train_runtime": 129209.3295,
+ "train_tokens_per_second": 5653.645
+ },
+ {
+ "epoch": 1.1251009964987881,
+ "grad_norm": 0.7285352578803672,
+ "learning_rate": 4.024289063597585e-06,
+ "loss": 0.31215999126434324,
+ "num_input_tokens_seen": 730706704,
+ "step": 16710,
+ "train_runtime": 129251.1282,
+ "train_tokens_per_second": 5653.387
+ },
+ {
+ "epoch": 1.1254376514947482,
+ "grad_norm": 0.7346612645471816,
+ "learning_rate": 4.021695950569443e-06,
+ "loss": 0.29782519340515134,
+ "num_input_tokens_seen": 730929536,
+ "step": 16715,
+ "train_runtime": 129286.421,
+ "train_tokens_per_second": 5653.568
+ },
+ {
+ "epoch": 1.1257743064907082,
+ "grad_norm": 0.7837756415475194,
+ "learning_rate": 4.01910311112091e-06,
+ "loss": 0.3227576017379761,
+ "num_input_tokens_seen": 731146120,
+ "step": 16720,
+ "train_runtime": 129325.3701,
+ "train_tokens_per_second": 5653.54
+ },
+ {
+ "epoch": 1.1261109614866684,
+ "grad_norm": 0.742625616948461,
+ "learning_rate": 4.016510545977064e-06,
+ "loss": 0.30303683280944826,
+ "num_input_tokens_seen": 731374192,
+ "step": 16725,
+ "train_runtime": 129369.3334,
+ "train_tokens_per_second": 5653.381
+ },
+ {
+ "epoch": 1.1264476164826287,
+ "grad_norm": 0.8331891535993596,
+ "learning_rate": 4.013918255862908e-06,
+ "loss": 0.3152627944946289,
+ "num_input_tokens_seen": 731580072,
+ "step": 16730,
+ "train_runtime": 129413.4377,
+ "train_tokens_per_second": 5653.046
+ },
+ {
+ "epoch": 1.1267842714785887,
+ "grad_norm": 0.8423878577043675,
+ "learning_rate": 4.011326241503369e-06,
+ "loss": 0.2968592643737793,
+ "num_input_tokens_seen": 731789896,
+ "step": 16735,
+ "train_runtime": 129458.1727,
+ "train_tokens_per_second": 5652.713
+ },
+ {
+ "epoch": 1.127120926474549,
+ "grad_norm": 0.8161928635173408,
+ "learning_rate": 4.00873450362329e-06,
+ "loss": 0.32601251602172854,
+ "num_input_tokens_seen": 732012656,
+ "step": 16740,
+ "train_runtime": 129494.0456,
+ "train_tokens_per_second": 5652.867
+ },
+ {
+ "epoch": 1.127457581470509,
+ "grad_norm": 0.7114507295708142,
+ "learning_rate": 4.0061430429474476e-06,
+ "loss": 0.31506369113922117,
+ "num_input_tokens_seen": 732216280,
+ "step": 16745,
+ "train_runtime": 129530.2218,
+ "train_tokens_per_second": 5652.861
+ },
+ {
+ "epoch": 1.1277942364664693,
+ "grad_norm": 0.6959294825982753,
+ "learning_rate": 4.003551860200534e-06,
+ "loss": 0.3079033851623535,
+ "num_input_tokens_seen": 732421912,
+ "step": 16750,
+ "train_runtime": 129574.2624,
+ "train_tokens_per_second": 5652.526
+ },
+ {
+ "epoch": 1.1281308914624293,
+ "grad_norm": 0.7729992928736847,
+ "learning_rate": 4.000960956107167e-06,
+ "loss": 0.3259746551513672,
+ "num_input_tokens_seen": 732628200,
+ "step": 16755,
+ "train_runtime": 129616.0713,
+ "train_tokens_per_second": 5652.294
+ },
+ {
+ "epoch": 1.1284675464583895,
+ "grad_norm": 0.7253296171356767,
+ "learning_rate": 3.998370331391881e-06,
+ "loss": 0.3024310111999512,
+ "num_input_tokens_seen": 732828608,
+ "step": 16760,
+ "train_runtime": 129653.8465,
+ "train_tokens_per_second": 5652.193
+ },
+ {
+ "epoch": 1.1288042014543496,
+ "grad_norm": 0.8768820968683303,
+ "learning_rate": 3.995779986779139e-06,
+ "loss": 0.3032145738601685,
+ "num_input_tokens_seen": 733030240,
+ "step": 16765,
+ "train_runtime": 129696.2491,
+ "train_tokens_per_second": 5651.9
+ },
+ {
+ "epoch": 1.1291408564503098,
+ "grad_norm": 0.7571543524283587,
+ "learning_rate": 3.9931899229933204e-06,
+ "loss": 0.32830450534820554,
+ "num_input_tokens_seen": 733249296,
+ "step": 16770,
+ "train_runtime": 129729.1458,
+ "train_tokens_per_second": 5652.155
+ },
+ {
+ "epoch": 1.1294775114462698,
+ "grad_norm": 0.7464218260015815,
+ "learning_rate": 3.990600140758731e-06,
+ "loss": 0.3006105422973633,
+ "num_input_tokens_seen": 733474720,
+ "step": 16775,
+ "train_runtime": 129763.0451,
+ "train_tokens_per_second": 5652.416
+ },
+ {
+ "epoch": 1.12981416644223,
+ "grad_norm": 0.6221590957693296,
+ "learning_rate": 3.988010640799591e-06,
+ "loss": 0.2877049922943115,
+ "num_input_tokens_seen": 733694216,
+ "step": 16780,
+ "train_runtime": 129806.6164,
+ "train_tokens_per_second": 5652.21
+ },
+ {
+ "epoch": 1.1301508214381901,
+ "grad_norm": 0.8483974664089013,
+ "learning_rate": 3.985421423840051e-06,
+ "loss": 0.33561320304870607,
+ "num_input_tokens_seen": 733906352,
+ "step": 16785,
+ "train_runtime": 129844.4587,
+ "train_tokens_per_second": 5652.196
+ },
+ {
+ "epoch": 1.1304874764341504,
+ "grad_norm": 0.7448892261292801,
+ "learning_rate": 3.982832490604173e-06,
+ "loss": 0.28998355865478515,
+ "num_input_tokens_seen": 734133280,
+ "step": 16790,
+ "train_runtime": 129888.8092,
+ "train_tokens_per_second": 5652.013
+ },
+ {
+ "epoch": 1.1308241314301104,
+ "grad_norm": 0.8527357235171165,
+ "learning_rate": 3.9802438418159465e-06,
+ "loss": 0.3185850143432617,
+ "num_input_tokens_seen": 734371888,
+ "step": 16795,
+ "train_runtime": 129928.7672,
+ "train_tokens_per_second": 5652.112
+ },
+ {
+ "epoch": 1.1311607864260707,
+ "grad_norm": 0.7819044120460309,
+ "learning_rate": 3.977655478199275e-06,
+ "loss": 0.32431769371032715,
+ "num_input_tokens_seen": 734599608,
+ "step": 16800,
+ "train_runtime": 129969.4055,
+ "train_tokens_per_second": 5652.096
+ },
+ {
+ "epoch": 1.1314974414220307,
+ "grad_norm": 0.7998665202892938,
+ "learning_rate": 3.975067400477993e-06,
+ "loss": 0.3372927188873291,
+ "num_input_tokens_seen": 734795752,
+ "step": 16805,
+ "train_runtime": 130003.7743,
+ "train_tokens_per_second": 5652.111
+ },
+ {
+ "epoch": 1.131834096417991,
+ "grad_norm": 0.7814987669774647,
+ "learning_rate": 3.972479609375842e-06,
+ "loss": 0.31949615478515625,
+ "num_input_tokens_seen": 735021664,
+ "step": 16810,
+ "train_runtime": 130042.9188,
+ "train_tokens_per_second": 5652.147
+ },
+ {
+ "epoch": 1.132170751413951,
+ "grad_norm": 0.7598227623059015,
+ "learning_rate": 3.969892105616492e-06,
+ "loss": 0.31243975162506105,
+ "num_input_tokens_seen": 735246192,
+ "step": 16815,
+ "train_runtime": 130079.0722,
+ "train_tokens_per_second": 5652.302
+ },
+ {
+ "epoch": 1.1325074064099112,
+ "grad_norm": 1.1084494733823549,
+ "learning_rate": 3.967304889923529e-06,
+ "loss": 0.30039591789245607,
+ "num_input_tokens_seen": 735462560,
+ "step": 16820,
+ "train_runtime": 130118.6173,
+ "train_tokens_per_second": 5652.247
+ },
+ {
+ "epoch": 1.1328440614058712,
+ "grad_norm": 0.8395955074548532,
+ "learning_rate": 3.964717963020463e-06,
+ "loss": 0.3070765256881714,
+ "num_input_tokens_seen": 735678824,
+ "step": 16825,
+ "train_runtime": 130155.0309,
+ "train_tokens_per_second": 5652.327
+ },
+ {
+ "epoch": 1.1331807164018315,
+ "grad_norm": 0.7687120040297668,
+ "learning_rate": 3.962131325630715e-06,
+ "loss": 0.29083225727081297,
+ "num_input_tokens_seen": 735911416,
+ "step": 16830,
+ "train_runtime": 130203.9338,
+ "train_tokens_per_second": 5651.991
+ },
+ {
+ "epoch": 1.1335173713977915,
+ "grad_norm": 0.8425980087790549,
+ "learning_rate": 3.959544978477634e-06,
+ "loss": 0.33116788864135743,
+ "num_input_tokens_seen": 736118352,
+ "step": 16835,
+ "train_runtime": 130242.9795,
+ "train_tokens_per_second": 5651.885
+ },
+ {
+ "epoch": 1.1338540263937518,
+ "grad_norm": 0.6639463454020782,
+ "learning_rate": 3.9569589222844805e-06,
+ "loss": 0.31282644271850585,
+ "num_input_tokens_seen": 736333424,
+ "step": 16840,
+ "train_runtime": 130287.4137,
+ "train_tokens_per_second": 5651.608
+ },
+ {
+ "epoch": 1.1341906813897118,
+ "grad_norm": 0.7757792448332916,
+ "learning_rate": 3.954373157774439e-06,
+ "loss": 0.3130516052246094,
+ "num_input_tokens_seen": 736558888,
+ "step": 16845,
+ "train_runtime": 130324.2351,
+ "train_tokens_per_second": 5651.742
+ },
+ {
+ "epoch": 1.134527336385672,
+ "grad_norm": 0.8187765783249485,
+ "learning_rate": 3.951787685670609e-06,
+ "loss": 0.3079373359680176,
+ "num_input_tokens_seen": 736782672,
+ "step": 16850,
+ "train_runtime": 130363.4678,
+ "train_tokens_per_second": 5651.757
+ },
+ {
+ "epoch": 1.134863991381632,
+ "grad_norm": 0.7794193428094371,
+ "learning_rate": 3.949202506696012e-06,
+ "loss": 0.315033483505249,
+ "num_input_tokens_seen": 736994032,
+ "step": 16855,
+ "train_runtime": 130400.0871,
+ "train_tokens_per_second": 5651.791
+ },
+ {
+ "epoch": 1.1352006463775923,
+ "grad_norm": 0.7253716724703653,
+ "learning_rate": 3.946617621573581e-06,
+ "loss": 0.3076453685760498,
+ "num_input_tokens_seen": 737227128,
+ "step": 16860,
+ "train_runtime": 130441.2278,
+ "train_tokens_per_second": 5651.795
+ },
+ {
+ "epoch": 1.1355373013735524,
+ "grad_norm": 0.8240959096061262,
+ "learning_rate": 3.944033031026175e-06,
+ "loss": 0.2819002628326416,
+ "num_input_tokens_seen": 737425512,
+ "step": 16865,
+ "train_runtime": 130480.7005,
+ "train_tokens_per_second": 5651.606
+ },
+ {
+ "epoch": 1.1358739563695126,
+ "grad_norm": 0.7359564971034284,
+ "learning_rate": 3.941448735776563e-06,
+ "loss": 0.3037100791931152,
+ "num_input_tokens_seen": 737645856,
+ "step": 16870,
+ "train_runtime": 130515.6474,
+ "train_tokens_per_second": 5651.781
+ },
+ {
+ "epoch": 1.1362106113654726,
+ "grad_norm": 0.7004505644549724,
+ "learning_rate": 3.938864736547439e-06,
+ "loss": 0.31839513778686523,
+ "num_input_tokens_seen": 737874128,
+ "step": 16875,
+ "train_runtime": 130552.8691,
+ "train_tokens_per_second": 5651.918
+ },
+ {
+ "epoch": 1.1365472663614329,
+ "grad_norm": 1.9350046761541768,
+ "learning_rate": 3.936281034061405e-06,
+ "loss": 0.2992375135421753,
+ "num_input_tokens_seen": 738099488,
+ "step": 16880,
+ "train_runtime": 130593.8162,
+ "train_tokens_per_second": 5651.872
+ },
+ {
+ "epoch": 1.136883921357393,
+ "grad_norm": 0.7114720599403541,
+ "learning_rate": 3.93369762904099e-06,
+ "loss": 0.3111462354660034,
+ "num_input_tokens_seen": 738305920,
+ "step": 16885,
+ "train_runtime": 130639.1413,
+ "train_tokens_per_second": 5651.491
+ },
+ {
+ "epoch": 1.1372205763533532,
+ "grad_norm": 0.7657629856292947,
+ "learning_rate": 3.931114522208631e-06,
+ "loss": 0.3166337251663208,
+ "num_input_tokens_seen": 738508624,
+ "step": 16890,
+ "train_runtime": 130682.7388,
+ "train_tokens_per_second": 5651.157
+ },
+ {
+ "epoch": 1.1375572313493132,
+ "grad_norm": 0.7692834919287869,
+ "learning_rate": 3.928531714286689e-06,
+ "loss": 0.30945444107055664,
+ "num_input_tokens_seen": 738732992,
+ "step": 16895,
+ "train_runtime": 130719.4165,
+ "train_tokens_per_second": 5651.287
+ },
+ {
+ "epoch": 1.1378938863452734,
+ "grad_norm": 0.8278288727680423,
+ "learning_rate": 3.925949205997434e-06,
+ "loss": 0.32941608428955077,
+ "num_input_tokens_seen": 738942840,
+ "step": 16900,
+ "train_runtime": 130761.9137,
+ "train_tokens_per_second": 5651.056
+ },
+ {
+ "epoch": 1.1382305413412335,
+ "grad_norm": 0.7724135944921471,
+ "learning_rate": 3.923366998063062e-06,
+ "loss": 0.2874737739562988,
+ "num_input_tokens_seen": 739155240,
+ "step": 16905,
+ "train_runtime": 130801.6692,
+ "train_tokens_per_second": 5650.962
+ },
+ {
+ "epoch": 1.1385671963371937,
+ "grad_norm": 0.6482256296875065,
+ "learning_rate": 3.920785091205674e-06,
+ "loss": 0.27861971855163575,
+ "num_input_tokens_seen": 739385256,
+ "step": 16910,
+ "train_runtime": 130835.4441,
+ "train_tokens_per_second": 5651.261
+ },
+ {
+ "epoch": 1.1389038513331537,
+ "grad_norm": 0.8189363341827849,
+ "learning_rate": 3.918203486147294e-06,
+ "loss": 0.34842791557312014,
+ "num_input_tokens_seen": 739594936,
+ "step": 16915,
+ "train_runtime": 130876.0263,
+ "train_tokens_per_second": 5651.111
+ },
+ {
+ "epoch": 1.139240506329114,
+ "grad_norm": 0.8291380348485267,
+ "learning_rate": 3.9156221836098594e-06,
+ "loss": 0.3238081455230713,
+ "num_input_tokens_seen": 739819304,
+ "step": 16920,
+ "train_runtime": 130912.2289,
+ "train_tokens_per_second": 5651.262
+ },
+ {
+ "epoch": 1.139577161325074,
+ "grad_norm": 0.7047778409887597,
+ "learning_rate": 3.913041184315224e-06,
+ "loss": 0.3074013710021973,
+ "num_input_tokens_seen": 740033032,
+ "step": 16925,
+ "train_runtime": 130954.6922,
+ "train_tokens_per_second": 5651.062
+ },
+ {
+ "epoch": 1.1399138163210343,
+ "grad_norm": 0.8050080201528949,
+ "learning_rate": 3.910460488985154e-06,
+ "loss": 0.3118597030639648,
+ "num_input_tokens_seen": 740249656,
+ "step": 16930,
+ "train_runtime": 131003.4754,
+ "train_tokens_per_second": 5650.611
+ },
+ {
+ "epoch": 1.1402504713169943,
+ "grad_norm": 0.727734150500772,
+ "learning_rate": 3.9078800983413355e-06,
+ "loss": 0.3097679138183594,
+ "num_input_tokens_seen": 740474224,
+ "step": 16935,
+ "train_runtime": 131039.7311,
+ "train_tokens_per_second": 5650.761
+ },
+ {
+ "epoch": 1.1405871263129546,
+ "grad_norm": 0.7396507242655731,
+ "learning_rate": 3.905300013105365e-06,
+ "loss": 0.2923974752426147,
+ "num_input_tokens_seen": 740697112,
+ "step": 16940,
+ "train_runtime": 131079.5559,
+ "train_tokens_per_second": 5650.745
+ },
+ {
+ "epoch": 1.1409237813089146,
+ "grad_norm": 0.8383241709134363,
+ "learning_rate": 3.902720233998754e-06,
+ "loss": 0.30002388954162595,
+ "num_input_tokens_seen": 740910376,
+ "step": 16945,
+ "train_runtime": 131116.3686,
+ "train_tokens_per_second": 5650.785
+ },
+ {
+ "epoch": 1.1412604363048748,
+ "grad_norm": 0.7028544654776611,
+ "learning_rate": 3.90014076174293e-06,
+ "loss": 0.31175391674041747,
+ "num_input_tokens_seen": 741133216,
+ "step": 16950,
+ "train_runtime": 131156.2994,
+ "train_tokens_per_second": 5650.763
+ },
+ {
+ "epoch": 1.1415970913008349,
+ "grad_norm": 0.6785576189006745,
+ "learning_rate": 3.897561597059237e-06,
+ "loss": 0.2938354253768921,
+ "num_input_tokens_seen": 741357712,
+ "step": 16955,
+ "train_runtime": 131193.385,
+ "train_tokens_per_second": 5650.877
+ },
+ {
+ "epoch": 1.1419337462967951,
+ "grad_norm": 0.720906429135837,
+ "learning_rate": 3.894982740668927e-06,
+ "loss": 0.29590883255004885,
+ "num_input_tokens_seen": 741562560,
+ "step": 16960,
+ "train_runtime": 131227.1172,
+ "train_tokens_per_second": 5650.986
+ },
+ {
+ "epoch": 1.1422704012927551,
+ "grad_norm": 0.7501425891472119,
+ "learning_rate": 3.89240419329317e-06,
+ "loss": 0.3243727207183838,
+ "num_input_tokens_seen": 741804536,
+ "step": 16965,
+ "train_runtime": 131263.585,
+ "train_tokens_per_second": 5651.259
+ },
+ {
+ "epoch": 1.1426070562887154,
+ "grad_norm": 0.7110651173816877,
+ "learning_rate": 3.889825955653046e-06,
+ "loss": 0.31747913360595703,
+ "num_input_tokens_seen": 742023480,
+ "step": 16970,
+ "train_runtime": 131299.3451,
+ "train_tokens_per_second": 5651.388
+ },
+ {
+ "epoch": 1.1429437112846754,
+ "grad_norm": 0.7195903483502709,
+ "learning_rate": 3.887248028469558e-06,
+ "loss": 0.2832490921020508,
+ "num_input_tokens_seen": 742258408,
+ "step": 16975,
+ "train_runtime": 131333.0196,
+ "train_tokens_per_second": 5651.727
+ },
+ {
+ "epoch": 1.1432803662806357,
+ "grad_norm": 1.0763044494541907,
+ "learning_rate": 3.884670412463607e-06,
+ "loss": 0.2930821657180786,
+ "num_input_tokens_seen": 742474896,
+ "step": 16980,
+ "train_runtime": 131371.6188,
+ "train_tokens_per_second": 5651.715
+ },
+ {
+ "epoch": 1.1436170212765957,
+ "grad_norm": 0.7486899797692392,
+ "learning_rate": 3.8820931083560194e-06,
+ "loss": 0.31524713039398194,
+ "num_input_tokens_seen": 742680248,
+ "step": 16985,
+ "train_runtime": 131410.9057,
+ "train_tokens_per_second": 5651.588
+ },
+ {
+ "epoch": 1.143953676272556,
+ "grad_norm": 0.7165743405717724,
+ "learning_rate": 3.879516116867531e-06,
+ "loss": 0.29150216579437255,
+ "num_input_tokens_seen": 742894640,
+ "step": 16990,
+ "train_runtime": 131445.9029,
+ "train_tokens_per_second": 5651.714
+ },
+ {
+ "epoch": 1.144290331268516,
+ "grad_norm": 0.8121082083066077,
+ "learning_rate": 3.876939438718786e-06,
+ "loss": 0.30789361000061033,
+ "num_input_tokens_seen": 743112224,
+ "step": 16995,
+ "train_runtime": 131487.3253,
+ "train_tokens_per_second": 5651.588
+ },
+ {
+ "epoch": 1.1446269862644762,
+ "grad_norm": 0.7621739149221717,
+ "learning_rate": 3.874363074630345e-06,
+ "loss": 0.28911523818969725,
+ "num_input_tokens_seen": 743319176,
+ "step": 17000,
+ "train_runtime": 131524.9073,
+ "train_tokens_per_second": 5651.547
+ },
+ {
+ "epoch": 1.1449636412604363,
+ "grad_norm": 0.7737775948010384,
+ "learning_rate": 3.871787025322681e-06,
+ "loss": 0.3129148006439209,
+ "num_input_tokens_seen": 743545736,
+ "step": 17005,
+ "train_runtime": 131561.9387,
+ "train_tokens_per_second": 5651.678
+ },
+ {
+ "epoch": 1.1453002962563965,
+ "grad_norm": 0.7555308668544749,
+ "learning_rate": 3.86921129151618e-06,
+ "loss": 0.3087345361709595,
+ "num_input_tokens_seen": 743756824,
+ "step": 17010,
+ "train_runtime": 131596.877,
+ "train_tokens_per_second": 5651.782
+ },
+ {
+ "epoch": 1.1456369512523565,
+ "grad_norm": 0.7822771296052244,
+ "learning_rate": 3.866635873931133e-06,
+ "loss": 0.3323315143585205,
+ "num_input_tokens_seen": 743968848,
+ "step": 17015,
+ "train_runtime": 131635.333,
+ "train_tokens_per_second": 5651.741
+ },
+ {
+ "epoch": 1.1459736062483168,
+ "grad_norm": 0.750133051806931,
+ "learning_rate": 3.864060773287751e-06,
+ "loss": 0.3157732725143433,
+ "num_input_tokens_seen": 744191936,
+ "step": 17020,
+ "train_runtime": 131677.9874,
+ "train_tokens_per_second": 5651.605
+ },
+ {
+ "epoch": 1.1463102612442768,
+ "grad_norm": 0.750658874065227,
+ "learning_rate": 3.861485990306151e-06,
+ "loss": 0.3161588191986084,
+ "num_input_tokens_seen": 744416472,
+ "step": 17025,
+ "train_runtime": 131722.5155,
+ "train_tokens_per_second": 5651.399
+ },
+ {
+ "epoch": 1.146646916240237,
+ "grad_norm": 0.7657095158045504,
+ "learning_rate": 3.858911525706365e-06,
+ "loss": 0.3021253108978271,
+ "num_input_tokens_seen": 744645048,
+ "step": 17030,
+ "train_runtime": 131759.7192,
+ "train_tokens_per_second": 5651.538
+ },
+ {
+ "epoch": 1.146983571236197,
+ "grad_norm": 0.8025594627650351,
+ "learning_rate": 3.85633738020833e-06,
+ "loss": 0.34988272190093994,
+ "num_input_tokens_seen": 744878048,
+ "step": 17035,
+ "train_runtime": 131794.0086,
+ "train_tokens_per_second": 5651.835
+ },
+ {
+ "epoch": 1.1473202262321573,
+ "grad_norm": 0.7709002150890326,
+ "learning_rate": 3.853763554531902e-06,
+ "loss": 0.3174929618835449,
+ "num_input_tokens_seen": 745082920,
+ "step": 17040,
+ "train_runtime": 131829.24,
+ "train_tokens_per_second": 5651.879
+ },
+ {
+ "epoch": 1.1476568812281174,
+ "grad_norm": 0.715665166170693,
+ "learning_rate": 3.851190049396839e-06,
+ "loss": 0.2886343479156494,
+ "num_input_tokens_seen": 745302104,
+ "step": 17045,
+ "train_runtime": 131865.7716,
+ "train_tokens_per_second": 5651.975
+ },
+ {
+ "epoch": 1.1479935362240776,
+ "grad_norm": 0.7364852648051009,
+ "learning_rate": 3.848616865522815e-06,
+ "loss": 0.31344852447509763,
+ "num_input_tokens_seen": 745522704,
+ "step": 17050,
+ "train_runtime": 131907.8212,
+ "train_tokens_per_second": 5651.846
+ },
+ {
+ "epoch": 1.1483301912200377,
+ "grad_norm": 0.7409016478894407,
+ "learning_rate": 3.846044003629414e-06,
+ "loss": 0.30931525230407714,
+ "num_input_tokens_seen": 745742992,
+ "step": 17055,
+ "train_runtime": 131945.4567,
+ "train_tokens_per_second": 5651.904
+ },
+ {
+ "epoch": 1.148666846215998,
+ "grad_norm": 0.7178354231276903,
+ "learning_rate": 3.843471464436128e-06,
+ "loss": 0.2803271055221558,
+ "num_input_tokens_seen": 745942912,
+ "step": 17060,
+ "train_runtime": 131989.4344,
+ "train_tokens_per_second": 5651.535
+ },
+ {
+ "epoch": 1.149003501211958,
+ "grad_norm": 0.787297382407357,
+ "learning_rate": 3.840899248662358e-06,
+ "loss": 0.33543803691864016,
+ "num_input_tokens_seen": 746154376,
+ "step": 17065,
+ "train_runtime": 132025.9108,
+ "train_tokens_per_second": 5651.575
+ },
+ {
+ "epoch": 1.1493401562079182,
+ "grad_norm": 0.7696262614580517,
+ "learning_rate": 3.838327357027419e-06,
+ "loss": 0.32558441162109375,
+ "num_input_tokens_seen": 746367240,
+ "step": 17070,
+ "train_runtime": 132064.2315,
+ "train_tokens_per_second": 5651.547
+ },
+ {
+ "epoch": 1.1496768112038782,
+ "grad_norm": 0.726552691405123,
+ "learning_rate": 3.835755790250528e-06,
+ "loss": 0.3117891550064087,
+ "num_input_tokens_seen": 746592480,
+ "step": 17075,
+ "train_runtime": 132101.1976,
+ "train_tokens_per_second": 5651.671
+ },
+ {
+ "epoch": 1.1500134661998385,
+ "grad_norm": 0.650793107544695,
+ "learning_rate": 3.83318454905082e-06,
+ "loss": 0.2934911251068115,
+ "num_input_tokens_seen": 746818640,
+ "step": 17080,
+ "train_runtime": 132143.7046,
+ "train_tokens_per_second": 5651.564
+ },
+ {
+ "epoch": 1.1503501211957985,
+ "grad_norm": 0.7704916837868131,
+ "learning_rate": 3.830613634147331e-06,
+ "loss": 0.32155680656433105,
+ "num_input_tokens_seen": 747039496,
+ "step": 17085,
+ "train_runtime": 132184.8265,
+ "train_tokens_per_second": 5651.477
+ },
+ {
+ "epoch": 1.1506867761917587,
+ "grad_norm": 0.720317163934302,
+ "learning_rate": 3.82804304625901e-06,
+ "loss": 0.3073138236999512,
+ "num_input_tokens_seen": 747272792,
+ "step": 17090,
+ "train_runtime": 132220.525,
+ "train_tokens_per_second": 5651.716
+ },
+ {
+ "epoch": 1.1510234311877188,
+ "grad_norm": 0.7743031429870126,
+ "learning_rate": 3.825472786104715e-06,
+ "loss": 0.2988899707794189,
+ "num_input_tokens_seen": 747507048,
+ "step": 17095,
+ "train_runtime": 132256.7682,
+ "train_tokens_per_second": 5651.938
+ },
+ {
+ "epoch": 1.151360086183679,
+ "grad_norm": 0.8974847495064373,
+ "learning_rate": 3.8229028544032095e-06,
+ "loss": 0.3167643308639526,
+ "num_input_tokens_seen": 747707608,
+ "step": 17100,
+ "train_runtime": 132298.3666,
+ "train_tokens_per_second": 5651.677
+ },
+ {
+ "epoch": 1.151696741179639,
+ "grad_norm": 0.8824423710352778,
+ "learning_rate": 3.8203332518731666e-06,
+ "loss": 0.3144956111907959,
+ "num_input_tokens_seen": 747928560,
+ "step": 17105,
+ "train_runtime": 132333.3063,
+ "train_tokens_per_second": 5651.854
+ },
+ {
+ "epoch": 1.1520333961755993,
+ "grad_norm": 0.6533415062560854,
+ "learning_rate": 3.8177639792331695e-06,
+ "loss": 0.3034801483154297,
+ "num_input_tokens_seen": 748157856,
+ "step": 17110,
+ "train_runtime": 132372.8474,
+ "train_tokens_per_second": 5651.898
+ },
+ {
+ "epoch": 1.1523700511715593,
+ "grad_norm": 0.9418120724183658,
+ "learning_rate": 3.815195037201704e-06,
+ "loss": 0.34812791347503663,
+ "num_input_tokens_seen": 748364016,
+ "step": 17115,
+ "train_runtime": 132412.7976,
+ "train_tokens_per_second": 5651.75
+ },
+ {
+ "epoch": 1.1527067061675196,
+ "grad_norm": 0.7569217027983804,
+ "learning_rate": 3.8126264264971694e-06,
+ "loss": 0.321405029296875,
+ "num_input_tokens_seen": 748586232,
+ "step": 17120,
+ "train_runtime": 132454.6523,
+ "train_tokens_per_second": 5651.642
+ },
+ {
+ "epoch": 1.1530433611634796,
+ "grad_norm": 0.7524862294811008,
+ "learning_rate": 3.8100581478378673e-06,
+ "loss": 0.3137218952178955,
+ "num_input_tokens_seen": 748805128,
+ "step": 17125,
+ "train_runtime": 132492.9601,
+ "train_tokens_per_second": 5651.66
+ },
+ {
+ "epoch": 1.1533800161594399,
+ "grad_norm": 0.7796331102294818,
+ "learning_rate": 3.807490201942012e-06,
+ "loss": 0.3160563468933105,
+ "num_input_tokens_seen": 749027264,
+ "step": 17130,
+ "train_runtime": 132527.0569,
+ "train_tokens_per_second": 5651.882
+ },
+ {
+ "epoch": 1.1537166711553999,
+ "grad_norm": 0.774378899106255,
+ "learning_rate": 3.804922589527717e-06,
+ "loss": 0.29862165451049805,
+ "num_input_tokens_seen": 749243080,
+ "step": 17135,
+ "train_runtime": 132567.1494,
+ "train_tokens_per_second": 5651.8
+ },
+ {
+ "epoch": 1.1540533261513601,
+ "grad_norm": 0.7096741310187692,
+ "learning_rate": 3.8023553113130096e-06,
+ "loss": 0.3242902040481567,
+ "num_input_tokens_seen": 749478816,
+ "step": 17140,
+ "train_runtime": 132602.8293,
+ "train_tokens_per_second": 5652.058
+ },
+ {
+ "epoch": 1.1543899811473202,
+ "grad_norm": 0.7242850071624176,
+ "learning_rate": 3.799788368015821e-06,
+ "loss": 0.28427910804748535,
+ "num_input_tokens_seen": 749699336,
+ "step": 17145,
+ "train_runtime": 132642.0928,
+ "train_tokens_per_second": 5652.047
+ },
+ {
+ "epoch": 1.1547266361432804,
+ "grad_norm": 0.8411805719100419,
+ "learning_rate": 3.7972217603539873e-06,
+ "loss": 0.30901274681091306,
+ "num_input_tokens_seen": 749905168,
+ "step": 17150,
+ "train_runtime": 132681.4222,
+ "train_tokens_per_second": 5651.923
+ },
+ {
+ "epoch": 1.1550632911392404,
+ "grad_norm": 0.7527979126590726,
+ "learning_rate": 3.7946554890452524e-06,
+ "loss": 0.3213947772979736,
+ "num_input_tokens_seen": 750135464,
+ "step": 17155,
+ "train_runtime": 132719.6511,
+ "train_tokens_per_second": 5652.03
+ },
+ {
+ "epoch": 1.1553999461352007,
+ "grad_norm": 0.7550688889586102,
+ "learning_rate": 3.792089554807269e-06,
+ "loss": 0.3054837226867676,
+ "num_input_tokens_seen": 750362056,
+ "step": 17160,
+ "train_runtime": 132761.8356,
+ "train_tokens_per_second": 5651.941
+ },
+ {
+ "epoch": 1.1557366011311607,
+ "grad_norm": 0.8245885062079921,
+ "learning_rate": 3.789523958357587e-06,
+ "loss": 0.3161867618560791,
+ "num_input_tokens_seen": 750572160,
+ "step": 17165,
+ "train_runtime": 132802.089,
+ "train_tokens_per_second": 5651.81
+ },
+ {
+ "epoch": 1.156073256127121,
+ "grad_norm": 0.700436201047616,
+ "learning_rate": 3.7869587004136714e-06,
+ "loss": 0.2760681390762329,
+ "num_input_tokens_seen": 750802552,
+ "step": 17170,
+ "train_runtime": 132839.9318,
+ "train_tokens_per_second": 5651.934
+ },
+ {
+ "epoch": 1.156409911123081,
+ "grad_norm": 0.7117777773350229,
+ "learning_rate": 3.7843937816928864e-06,
+ "loss": 0.3197911262512207,
+ "num_input_tokens_seen": 751031672,
+ "step": 17175,
+ "train_runtime": 132884.5398,
+ "train_tokens_per_second": 5651.761
+ },
+ {
+ "epoch": 1.1567465661190413,
+ "grad_norm": 0.7963520839657313,
+ "learning_rate": 3.781829202912506e-06,
+ "loss": 0.3260814666748047,
+ "num_input_tokens_seen": 751243800,
+ "step": 17180,
+ "train_runtime": 132921.2705,
+ "train_tokens_per_second": 5651.795
+ },
+ {
+ "epoch": 1.1570832211150013,
+ "grad_norm": 0.735123307982365,
+ "learning_rate": 3.7792649647897023e-06,
+ "loss": 0.3070995807647705,
+ "num_input_tokens_seen": 751468240,
+ "step": 17185,
+ "train_runtime": 132963.3105,
+ "train_tokens_per_second": 5651.696
+ },
+ {
+ "epoch": 1.1574198761109615,
+ "grad_norm": 0.7383532177231918,
+ "learning_rate": 3.7767010680415595e-06,
+ "loss": 0.30820441246032715,
+ "num_input_tokens_seen": 751680248,
+ "step": 17190,
+ "train_runtime": 133002.5463,
+ "train_tokens_per_second": 5651.623
+ },
+ {
+ "epoch": 1.1577565311069216,
+ "grad_norm": 0.8213108428390675,
+ "learning_rate": 3.774137513385061e-06,
+ "loss": 0.3007169008255005,
+ "num_input_tokens_seen": 751897104,
+ "step": 17195,
+ "train_runtime": 133050.5633,
+ "train_tokens_per_second": 5651.213
+ },
+ {
+ "epoch": 1.1580931861028818,
+ "grad_norm": 0.6991961330303099,
+ "learning_rate": 3.7715743015371005e-06,
+ "loss": 0.316827392578125,
+ "num_input_tokens_seen": 752126600,
+ "step": 17200,
+ "train_runtime": 133089.0112,
+ "train_tokens_per_second": 5651.305
+ },
+ {
+ "epoch": 1.1584298410988418,
+ "grad_norm": 0.6752546760035666,
+ "learning_rate": 3.769011433214467e-06,
+ "loss": 0.28647899627685547,
+ "num_input_tokens_seen": 752350528,
+ "step": 17205,
+ "train_runtime": 133130.5351,
+ "train_tokens_per_second": 5651.224
+ },
+ {
+ "epoch": 1.158766496094802,
+ "grad_norm": 0.7193811538704217,
+ "learning_rate": 3.766448909133863e-06,
+ "loss": 0.32767200469970703,
+ "num_input_tokens_seen": 752583080,
+ "step": 17210,
+ "train_runtime": 133161.9958,
+ "train_tokens_per_second": 5651.636
+ },
+ {
+ "epoch": 1.1591031510907621,
+ "grad_norm": 0.774174630991004,
+ "learning_rate": 3.7638867300118864e-06,
+ "loss": 0.29516313076019285,
+ "num_input_tokens_seen": 752794792,
+ "step": 17215,
+ "train_runtime": 133198.9108,
+ "train_tokens_per_second": 5651.659
+ },
+ {
+ "epoch": 1.1594398060867224,
+ "grad_norm": 0.810354987252184,
+ "learning_rate": 3.7613248965650447e-06,
+ "loss": 0.2927243709564209,
+ "num_input_tokens_seen": 753018152,
+ "step": 17220,
+ "train_runtime": 133236.819,
+ "train_tokens_per_second": 5651.727
+ },
+ {
+ "epoch": 1.1597764610826824,
+ "grad_norm": 0.7109008764685462,
+ "learning_rate": 3.758763409509745e-06,
+ "loss": 0.3014249324798584,
+ "num_input_tokens_seen": 753245648,
+ "step": 17225,
+ "train_runtime": 133276.7031,
+ "train_tokens_per_second": 5651.743
+ },
+ {
+ "epoch": 1.1601131160786426,
+ "grad_norm": 0.8029128503628568,
+ "learning_rate": 3.7562022695623013e-06,
+ "loss": 0.3143294334411621,
+ "num_input_tokens_seen": 753463144,
+ "step": 17230,
+ "train_runtime": 133315.2382,
+ "train_tokens_per_second": 5651.741
+ },
+ {
+ "epoch": 1.1604497710746027,
+ "grad_norm": 0.6493145264817488,
+ "learning_rate": 3.7536414774389252e-06,
+ "loss": 0.3114176273345947,
+ "num_input_tokens_seen": 753687504,
+ "step": 17235,
+ "train_runtime": 133354.768,
+ "train_tokens_per_second": 5651.748
+ },
+ {
+ "epoch": 1.160786426070563,
+ "grad_norm": 0.7972007226513281,
+ "learning_rate": 3.7510810338557363e-06,
+ "loss": 0.2993874788284302,
+ "num_input_tokens_seen": 753893032,
+ "step": 17240,
+ "train_runtime": 133399.0853,
+ "train_tokens_per_second": 5651.411
+ },
+ {
+ "epoch": 1.161123081066523,
+ "grad_norm": 0.7446879758207031,
+ "learning_rate": 3.748520939528753e-06,
+ "loss": 0.30895147323608396,
+ "num_input_tokens_seen": 754112432,
+ "step": 17245,
+ "train_runtime": 133440.2549,
+ "train_tokens_per_second": 5651.311
+ },
+ {
+ "epoch": 1.1614597360624832,
+ "grad_norm": 0.7281872851418484,
+ "learning_rate": 3.7459611951739007e-06,
+ "loss": 0.3002096891403198,
+ "num_input_tokens_seen": 754338736,
+ "step": 17250,
+ "train_runtime": 133481.8765,
+ "train_tokens_per_second": 5651.245
+ },
+ {
+ "epoch": 1.1617963910584432,
+ "grad_norm": 0.7431634131632837,
+ "learning_rate": 3.7434018015069984e-06,
+ "loss": 0.3243889093399048,
+ "num_input_tokens_seen": 754559056,
+ "step": 17255,
+ "train_runtime": 133525.754,
+ "train_tokens_per_second": 5651.038
+ },
+ {
+ "epoch": 1.1621330460544035,
+ "grad_norm": 0.8324688332558176,
+ "learning_rate": 3.740842759243778e-06,
+ "loss": 0.3150045871734619,
+ "num_input_tokens_seen": 754778800,
+ "step": 17260,
+ "train_runtime": 133565.7133,
+ "train_tokens_per_second": 5650.992
+ },
+ {
+ "epoch": 1.1624697010503635,
+ "grad_norm": 0.6921071488758873,
+ "learning_rate": 3.738284069099862e-06,
+ "loss": 0.28803367614746095,
+ "num_input_tokens_seen": 755011760,
+ "step": 17265,
+ "train_runtime": 133605.0409,
+ "train_tokens_per_second": 5651.072
+ },
+ {
+ "epoch": 1.1628063560463238,
+ "grad_norm": 0.7280978950943039,
+ "learning_rate": 3.735725731790785e-06,
+ "loss": 0.31182174682617186,
+ "num_input_tokens_seen": 755236808,
+ "step": 17270,
+ "train_runtime": 133638.2275,
+ "train_tokens_per_second": 5651.353
+ },
+ {
+ "epoch": 1.1631430110422838,
+ "grad_norm": 0.7658534531818101,
+ "learning_rate": 3.733167748031974e-06,
+ "loss": 0.3179009914398193,
+ "num_input_tokens_seen": 755464248,
+ "step": 17275,
+ "train_runtime": 133676.324,
+ "train_tokens_per_second": 5651.444
+ },
+ {
+ "epoch": 1.163479666038244,
+ "grad_norm": 0.7211234703667541,
+ "learning_rate": 3.7306101185387646e-06,
+ "loss": 0.3087700366973877,
+ "num_input_tokens_seen": 755687464,
+ "step": 17280,
+ "train_runtime": 133717.1549,
+ "train_tokens_per_second": 5651.388
+ },
+ {
+ "epoch": 1.163816321034204,
+ "grad_norm": 0.7824239357354816,
+ "learning_rate": 3.728052844026386e-06,
+ "loss": 0.3141305923461914,
+ "num_input_tokens_seen": 755907672,
+ "step": 17285,
+ "train_runtime": 133760.0292,
+ "train_tokens_per_second": 5651.222
+ },
+ {
+ "epoch": 1.1641529760301643,
+ "grad_norm": 0.7951750585384382,
+ "learning_rate": 3.7254959252099745e-06,
+ "loss": 0.3007349014282227,
+ "num_input_tokens_seen": 756114320,
+ "step": 17290,
+ "train_runtime": 133800.4505,
+ "train_tokens_per_second": 5651.06
+ },
+ {
+ "epoch": 1.1644896310261243,
+ "grad_norm": 0.7638584288959342,
+ "learning_rate": 3.7229393628045617e-06,
+ "loss": 0.2961156368255615,
+ "num_input_tokens_seen": 756338136,
+ "step": 17295,
+ "train_runtime": 133836.5594,
+ "train_tokens_per_second": 5651.207
+ },
+ {
+ "epoch": 1.1648262860220846,
+ "grad_norm": 0.7194065626433747,
+ "learning_rate": 3.720383157525087e-06,
+ "loss": 0.2917820453643799,
+ "num_input_tokens_seen": 756556624,
+ "step": 17300,
+ "train_runtime": 133876.3071,
+ "train_tokens_per_second": 5651.161
+ },
+ {
+ "epoch": 1.1651629410180446,
+ "grad_norm": 0.8023605053763685,
+ "learning_rate": 3.717827310086379e-06,
+ "loss": 0.3004648685455322,
+ "num_input_tokens_seen": 756770512,
+ "step": 17305,
+ "train_runtime": 133914.1223,
+ "train_tokens_per_second": 5651.163
+ },
+ {
+ "epoch": 1.1654995960140049,
+ "grad_norm": 0.778357180423363,
+ "learning_rate": 3.7152718212031778e-06,
+ "loss": 0.31314473152160643,
+ "num_input_tokens_seen": 756987488,
+ "step": 17310,
+ "train_runtime": 133957.8798,
+ "train_tokens_per_second": 5650.937
+ },
+ {
+ "epoch": 1.165836251009965,
+ "grad_norm": 0.8275289683679823,
+ "learning_rate": 3.7127166915901123e-06,
+ "loss": 0.33815641403198243,
+ "num_input_tokens_seen": 757190888,
+ "step": 17315,
+ "train_runtime": 133995.0668,
+ "train_tokens_per_second": 5650.886
+ },
+ {
+ "epoch": 1.1661729060059252,
+ "grad_norm": 0.8112349556897875,
+ "learning_rate": 3.710161921961721e-06,
+ "loss": 0.3213828086853027,
+ "num_input_tokens_seen": 757406736,
+ "step": 17320,
+ "train_runtime": 134033.0013,
+ "train_tokens_per_second": 5650.897
+ },
+ {
+ "epoch": 1.1665095610018852,
+ "grad_norm": 0.7412368352707359,
+ "learning_rate": 3.707607513032434e-06,
+ "loss": 0.29859418869018556,
+ "num_input_tokens_seen": 757621448,
+ "step": 17325,
+ "train_runtime": 134070.8112,
+ "train_tokens_per_second": 5650.905
+ },
+ {
+ "epoch": 1.1668462159978454,
+ "grad_norm": 0.8530421631271973,
+ "learning_rate": 3.705053465516587e-06,
+ "loss": 0.2971299171447754,
+ "num_input_tokens_seen": 757842136,
+ "step": 17330,
+ "train_runtime": 134106.2945,
+ "train_tokens_per_second": 5651.056
+ },
+ {
+ "epoch": 1.1671828709938055,
+ "grad_norm": 0.7217367075989789,
+ "learning_rate": 3.702499780128407e-06,
+ "loss": 0.2950276851654053,
+ "num_input_tokens_seen": 758048560,
+ "step": 17335,
+ "train_runtime": 134146.644,
+ "train_tokens_per_second": 5650.895
+ },
+ {
+ "epoch": 1.1675195259897657,
+ "grad_norm": 1.8887784795675984,
+ "learning_rate": 3.6999464575820266e-06,
+ "loss": 0.3086616277694702,
+ "num_input_tokens_seen": 758266576,
+ "step": 17340,
+ "train_runtime": 134185.9605,
+ "train_tokens_per_second": 5650.864
+ },
+ {
+ "epoch": 1.1678561809857257,
+ "grad_norm": 0.6843592201114646,
+ "learning_rate": 3.697393498591473e-06,
+ "loss": 0.3014980792999268,
+ "num_input_tokens_seen": 758487384,
+ "step": 17345,
+ "train_runtime": 134230.8121,
+ "train_tokens_per_second": 5650.621
+ },
+ {
+ "epoch": 1.168192835981686,
+ "grad_norm": 0.8394323507587605,
+ "learning_rate": 3.694840903870676e-06,
+ "loss": 0.32812952995300293,
+ "num_input_tokens_seen": 758704816,
+ "step": 17350,
+ "train_runtime": 134266.4562,
+ "train_tokens_per_second": 5650.74
+ },
+ {
+ "epoch": 1.168529490977646,
+ "grad_norm": 0.7479174116708699,
+ "learning_rate": 3.6922886741334564e-06,
+ "loss": 0.3267998695373535,
+ "num_input_tokens_seen": 758924344,
+ "step": 17355,
+ "train_runtime": 134305.3103,
+ "train_tokens_per_second": 5650.74
+ },
+ {
+ "epoch": 1.1688661459736063,
+ "grad_norm": 0.7136771552457577,
+ "learning_rate": 3.689736810093541e-06,
+ "loss": 0.2991864442825317,
+ "num_input_tokens_seen": 759152032,
+ "step": 17360,
+ "train_runtime": 134342.9756,
+ "train_tokens_per_second": 5650.85
+ },
+ {
+ "epoch": 1.1692028009695663,
+ "grad_norm": 0.850190780847832,
+ "learning_rate": 3.6871853124645464e-06,
+ "loss": 0.31711387634277344,
+ "num_input_tokens_seen": 759378848,
+ "step": 17365,
+ "train_runtime": 134382.3965,
+ "train_tokens_per_second": 5650.88
+ },
+ {
+ "epoch": 1.1695394559655266,
+ "grad_norm": 0.6756862427689906,
+ "learning_rate": 3.684634181959994e-06,
+ "loss": 0.30228276252746583,
+ "num_input_tokens_seen": 759614152,
+ "step": 17370,
+ "train_runtime": 134422.1735,
+ "train_tokens_per_second": 5650.959
+ },
+ {
+ "epoch": 1.1698761109614866,
+ "grad_norm": 0.9293848953693671,
+ "learning_rate": 3.6820834192932974e-06,
+ "loss": 0.3172401666641235,
+ "num_input_tokens_seen": 759830040,
+ "step": 17375,
+ "train_runtime": 134460.803,
+ "train_tokens_per_second": 5650.941
+ },
+ {
+ "epoch": 1.1702127659574468,
+ "grad_norm": 0.7492817955305227,
+ "learning_rate": 3.679533025177772e-06,
+ "loss": 0.3354950904846191,
+ "num_input_tokens_seen": 760073280,
+ "step": 17380,
+ "train_runtime": 134496.7214,
+ "train_tokens_per_second": 5651.24
+ },
+ {
+ "epoch": 1.1705494209534069,
+ "grad_norm": 0.7424725386974647,
+ "learning_rate": 3.6769830003266234e-06,
+ "loss": 0.31007990837097166,
+ "num_input_tokens_seen": 760294432,
+ "step": 17385,
+ "train_runtime": 134528.7303,
+ "train_tokens_per_second": 5651.539
+ },
+ {
+ "epoch": 1.1708860759493671,
+ "grad_norm": 0.7790858025441731,
+ "learning_rate": 3.6744333454529607e-06,
+ "loss": 0.3161858081817627,
+ "num_input_tokens_seen": 760522824,
+ "step": 17390,
+ "train_runtime": 134571.7613,
+ "train_tokens_per_second": 5651.43
+ },
+ {
+ "epoch": 1.1712227309453271,
+ "grad_norm": 0.6903404727708745,
+ "learning_rate": 3.6718840612697847e-06,
+ "loss": 0.3075212001800537,
+ "num_input_tokens_seen": 760757728,
+ "step": 17395,
+ "train_runtime": 134611.0442,
+ "train_tokens_per_second": 5651.525
+ },
+ {
+ "epoch": 1.1715593859412874,
+ "grad_norm": 0.7365357926688636,
+ "learning_rate": 3.6693351484899986e-06,
+ "loss": 0.3011133670806885,
+ "num_input_tokens_seen": 760969040,
+ "step": 17400,
+ "train_runtime": 134643.3087,
+ "train_tokens_per_second": 5651.74
+ },
+ {
+ "epoch": 1.1718960409372474,
+ "grad_norm": 1.0260127216999277,
+ "learning_rate": 3.6667866078263937e-06,
+ "loss": 0.30281815528869627,
+ "num_input_tokens_seen": 761178944,
+ "step": 17405,
+ "train_runtime": 134680.0119,
+ "train_tokens_per_second": 5651.759
+ },
+ {
+ "epoch": 1.1722326959332077,
+ "grad_norm": 0.8315108294093327,
+ "learning_rate": 3.664238439991663e-06,
+ "loss": 0.34009361267089844,
+ "num_input_tokens_seen": 761396128,
+ "step": 17410,
+ "train_runtime": 134717.5677,
+ "train_tokens_per_second": 5651.795
+ },
+ {
+ "epoch": 1.1725693509291677,
+ "grad_norm": 0.7676177184346961,
+ "learning_rate": 3.6616906456983925e-06,
+ "loss": 0.3160113334655762,
+ "num_input_tokens_seen": 761617984,
+ "step": 17415,
+ "train_runtime": 134752.1162,
+ "train_tokens_per_second": 5651.993
+ },
+ {
+ "epoch": 1.172906005925128,
+ "grad_norm": 0.7720738410000789,
+ "learning_rate": 3.659143225659065e-06,
+ "loss": 0.32563726902008056,
+ "num_input_tokens_seen": 761835752,
+ "step": 17420,
+ "train_runtime": 134789.1882,
+ "train_tokens_per_second": 5652.054
+ },
+ {
+ "epoch": 1.173242660921088,
+ "grad_norm": 0.6926475609016922,
+ "learning_rate": 3.6565961805860595e-06,
+ "loss": 0.30316720008850095,
+ "num_input_tokens_seen": 762067720,
+ "step": 17425,
+ "train_runtime": 134833.4769,
+ "train_tokens_per_second": 5651.918
+ },
+ {
+ "epoch": 1.1735793159170482,
+ "grad_norm": 0.6768651518731568,
+ "learning_rate": 3.654049511191649e-06,
+ "loss": 0.3012946605682373,
+ "num_input_tokens_seen": 762275792,
+ "step": 17430,
+ "train_runtime": 134874.1565,
+ "train_tokens_per_second": 5651.756
+ },
+ {
+ "epoch": 1.1739159709130083,
+ "grad_norm": 0.7655463622291874,
+ "learning_rate": 3.6515032181879995e-06,
+ "loss": 0.3098176956176758,
+ "num_input_tokens_seen": 762491920,
+ "step": 17435,
+ "train_runtime": 134914.91,
+ "train_tokens_per_second": 5651.651
+ },
+ {
+ "epoch": 1.1742526259089685,
+ "grad_norm": 0.7242442794388552,
+ "learning_rate": 3.6489573022871755e-06,
+ "loss": 0.28149569034576416,
+ "num_input_tokens_seen": 762712840,
+ "step": 17440,
+ "train_runtime": 134952.0365,
+ "train_tokens_per_second": 5651.733
+ },
+ {
+ "epoch": 1.1745892809049285,
+ "grad_norm": 0.6756916904062972,
+ "learning_rate": 3.6464117642011333e-06,
+ "loss": 0.31446964740753175,
+ "num_input_tokens_seen": 762919536,
+ "step": 17445,
+ "train_runtime": 134991.2971,
+ "train_tokens_per_second": 5651.62
+ },
+ {
+ "epoch": 1.1749259359008888,
+ "grad_norm": 0.7799367018501008,
+ "learning_rate": 3.643866604641726e-06,
+ "loss": 0.31410171985626223,
+ "num_input_tokens_seen": 763152304,
+ "step": 17450,
+ "train_runtime": 135025.582,
+ "train_tokens_per_second": 5651.909
+ },
+ {
+ "epoch": 1.1752625908968488,
+ "grad_norm": 0.7436018486594765,
+ "learning_rate": 3.641321824320697e-06,
+ "loss": 0.29862222671508787,
+ "num_input_tokens_seen": 763371656,
+ "step": 17455,
+ "train_runtime": 135071.3304,
+ "train_tokens_per_second": 5651.619
+ },
+ {
+ "epoch": 1.175599245892809,
+ "grad_norm": 0.6881929025831984,
+ "learning_rate": 3.6387774239496893e-06,
+ "loss": 0.2896681785583496,
+ "num_input_tokens_seen": 763607368,
+ "step": 17460,
+ "train_runtime": 135110.4495,
+ "train_tokens_per_second": 5651.727
+ },
+ {
+ "epoch": 1.175935900888769,
+ "grad_norm": 0.7211281131146224,
+ "learning_rate": 3.6362334042402326e-06,
+ "loss": 0.3099272012710571,
+ "num_input_tokens_seen": 763825752,
+ "step": 17465,
+ "train_runtime": 135156.7264,
+ "train_tokens_per_second": 5651.408
+ },
+ {
+ "epoch": 1.1762725558847293,
+ "grad_norm": 0.7082063726811787,
+ "learning_rate": 3.6336897659037563e-06,
+ "loss": 0.2802994966506958,
+ "num_input_tokens_seen": 764035624,
+ "step": 17470,
+ "train_runtime": 135188.9582,
+ "train_tokens_per_second": 5651.613
+ },
+ {
+ "epoch": 1.1766092108806894,
+ "grad_norm": 0.721477429769724,
+ "learning_rate": 3.6311465096515787e-06,
+ "loss": 0.29577035903930665,
+ "num_input_tokens_seen": 764261488,
+ "step": 17475,
+ "train_runtime": 135224.5379,
+ "train_tokens_per_second": 5651.796
+ },
+ {
+ "epoch": 1.1769458658766496,
+ "grad_norm": 0.7174604232091135,
+ "learning_rate": 3.6286036361949175e-06,
+ "loss": 0.2771427154541016,
+ "num_input_tokens_seen": 764486984,
+ "step": 17480,
+ "train_runtime": 135262.7158,
+ "train_tokens_per_second": 5651.868
+ },
+ {
+ "epoch": 1.1772825208726099,
+ "grad_norm": 0.6898907133835245,
+ "learning_rate": 3.6260611462448736e-06,
+ "loss": 0.2587808847427368,
+ "num_input_tokens_seen": 764711328,
+ "step": 17485,
+ "train_runtime": 135302.057,
+ "train_tokens_per_second": 5651.882
+ },
+ {
+ "epoch": 1.17761917586857,
+ "grad_norm": 0.9411699136352466,
+ "learning_rate": 3.623519040512451e-06,
+ "loss": 0.3395826816558838,
+ "num_input_tokens_seen": 764916768,
+ "step": 17490,
+ "train_runtime": 135338.873,
+ "train_tokens_per_second": 5651.863
+ },
+ {
+ "epoch": 1.17795583086453,
+ "grad_norm": 0.6299891222606505,
+ "learning_rate": 3.6209773197085373e-06,
+ "loss": 0.2713226556777954,
+ "num_input_tokens_seen": 765139336,
+ "step": 17495,
+ "train_runtime": 135380.6388,
+ "train_tokens_per_second": 5651.763
+ },
+ {
+ "epoch": 1.1782924858604902,
+ "grad_norm": 0.7488235191732048,
+ "learning_rate": 3.6184359845439214e-06,
+ "loss": 0.31320526599884035,
+ "num_input_tokens_seen": 765373584,
+ "step": 17500,
+ "train_runtime": 135413.0388,
+ "train_tokens_per_second": 5652.141
+ },
+ {
+ "epoch": 1.1786291408564504,
+ "grad_norm": 0.7715560770139029,
+ "learning_rate": 3.615895035729275e-06,
+ "loss": 0.32944774627685547,
+ "num_input_tokens_seen": 765601768,
+ "step": 17505,
+ "train_runtime": 135459.3007,
+ "train_tokens_per_second": 5651.895
+ },
+ {
+ "epoch": 1.1789657958524105,
+ "grad_norm": 0.7026976712835434,
+ "learning_rate": 3.6133544739751703e-06,
+ "loss": 0.2919255256652832,
+ "num_input_tokens_seen": 765813632,
+ "step": 17510,
+ "train_runtime": 135499.1227,
+ "train_tokens_per_second": 5651.798
+ },
+ {
+ "epoch": 1.1793024508483705,
+ "grad_norm": 0.7763204505182639,
+ "learning_rate": 3.6108142999920633e-06,
+ "loss": 0.3174884796142578,
+ "num_input_tokens_seen": 766030160,
+ "step": 17515,
+ "train_runtime": 135540.5033,
+ "train_tokens_per_second": 5651.67
+ },
+ {
+ "epoch": 1.1796391058443307,
+ "grad_norm": 0.7315888831681889,
+ "learning_rate": 3.608274514490308e-06,
+ "loss": 0.30716781616210936,
+ "num_input_tokens_seen": 766254576,
+ "step": 17520,
+ "train_runtime": 135580.8511,
+ "train_tokens_per_second": 5651.643
+ },
+ {
+ "epoch": 1.179975760840291,
+ "grad_norm": 0.8351647015979019,
+ "learning_rate": 3.605735118180146e-06,
+ "loss": 0.2879929542541504,
+ "num_input_tokens_seen": 766474424,
+ "step": 17525,
+ "train_runtime": 135626.1303,
+ "train_tokens_per_second": 5651.377
+ },
+ {
+ "epoch": 1.180312415836251,
+ "grad_norm": 0.7061934846953144,
+ "learning_rate": 3.6031961117717144e-06,
+ "loss": 0.2810915470123291,
+ "num_input_tokens_seen": 766686368,
+ "step": 17530,
+ "train_runtime": 135665.2165,
+ "train_tokens_per_second": 5651.311
+ },
+ {
+ "epoch": 1.180649070832211,
+ "grad_norm": 0.6723872835220442,
+ "learning_rate": 3.600657495975034e-06,
+ "loss": 0.29715049266815186,
+ "num_input_tokens_seen": 766912752,
+ "step": 17535,
+ "train_runtime": 135706.1892,
+ "train_tokens_per_second": 5651.273
+ },
+ {
+ "epoch": 1.1809857258281713,
+ "grad_norm": 0.7032026671628777,
+ "learning_rate": 3.598119271500024e-06,
+ "loss": 0.30086421966552734,
+ "num_input_tokens_seen": 767131128,
+ "step": 17540,
+ "train_runtime": 135738.6404,
+ "train_tokens_per_second": 5651.531
+ },
+ {
+ "epoch": 1.1813223808241315,
+ "grad_norm": 0.7440024692945637,
+ "learning_rate": 3.5955814390564882e-06,
+ "loss": 0.30273914337158203,
+ "num_input_tokens_seen": 767352664,
+ "step": 17545,
+ "train_runtime": 135778.2276,
+ "train_tokens_per_second": 5651.515
+ },
+ {
+ "epoch": 1.1816590358200916,
+ "grad_norm": 0.7568009197320091,
+ "learning_rate": 3.5930439993541265e-06,
+ "loss": 0.3133152961730957,
+ "num_input_tokens_seen": 767564224,
+ "step": 17550,
+ "train_runtime": 135812.8257,
+ "train_tokens_per_second": 5651.633
+ },
+ {
+ "epoch": 1.1819956908160516,
+ "grad_norm": 0.8131849513084429,
+ "learning_rate": 3.590506953102522e-06,
+ "loss": 0.3256534576416016,
+ "num_input_tokens_seen": 767781400,
+ "step": 17555,
+ "train_runtime": 135853.5865,
+ "train_tokens_per_second": 5651.536
+ },
+ {
+ "epoch": 1.1823323458120119,
+ "grad_norm": 0.7021756461446376,
+ "learning_rate": 3.5879703010111554e-06,
+ "loss": 0.3131730318069458,
+ "num_input_tokens_seen": 767996072,
+ "step": 17560,
+ "train_runtime": 135897.3639,
+ "train_tokens_per_second": 5651.295
+ },
+ {
+ "epoch": 1.182669000807972,
+ "grad_norm": 0.6947691337677638,
+ "learning_rate": 3.5854340437893898e-06,
+ "loss": 0.29679055213928224,
+ "num_input_tokens_seen": 768218592,
+ "step": 17565,
+ "train_runtime": 135933.4998,
+ "train_tokens_per_second": 5651.43
+ },
+ {
+ "epoch": 1.1830056558039321,
+ "grad_norm": 0.7588414880559752,
+ "learning_rate": 3.5828981821464838e-06,
+ "loss": 0.31975440979003905,
+ "num_input_tokens_seen": 768427128,
+ "step": 17570,
+ "train_runtime": 135972.4916,
+ "train_tokens_per_second": 5651.343
+ },
+ {
+ "epoch": 1.1833423107998922,
+ "grad_norm": 0.708470213652483,
+ "learning_rate": 3.580362716791582e-06,
+ "loss": 0.31680521965026853,
+ "num_input_tokens_seen": 768660704,
+ "step": 17575,
+ "train_runtime": 136004.1836,
+ "train_tokens_per_second": 5651.743
+ },
+ {
+ "epoch": 1.1836789657958524,
+ "grad_norm": 0.8012972832226694,
+ "learning_rate": 3.5778276484337205e-06,
+ "loss": 0.3270276069641113,
+ "num_input_tokens_seen": 768871496,
+ "step": 17580,
+ "train_runtime": 136039.0476,
+ "train_tokens_per_second": 5651.844
+ },
+ {
+ "epoch": 1.1840156207918127,
+ "grad_norm": 0.6791587926903686,
+ "learning_rate": 3.575292977781821e-06,
+ "loss": 0.2917544603347778,
+ "num_input_tokens_seen": 769081176,
+ "step": 17585,
+ "train_runtime": 136078.1123,
+ "train_tokens_per_second": 5651.763
+ },
+ {
+ "epoch": 1.1843522757877727,
+ "grad_norm": 0.774600266373974,
+ "learning_rate": 3.5727587055446976e-06,
+ "loss": 0.30277276039123535,
+ "num_input_tokens_seen": 769310368,
+ "step": 17590,
+ "train_runtime": 136115.3881,
+ "train_tokens_per_second": 5651.899
+ },
+ {
+ "epoch": 1.1846889307837327,
+ "grad_norm": 0.707413300444675,
+ "learning_rate": 3.570224832431051e-06,
+ "loss": 0.3234860420227051,
+ "num_input_tokens_seen": 769534008,
+ "step": 17595,
+ "train_runtime": 136150.4904,
+ "train_tokens_per_second": 5652.084
+ },
+ {
+ "epoch": 1.185025585779693,
+ "grad_norm": 0.8625892185888415,
+ "learning_rate": 3.5676913591494715e-06,
+ "loss": 0.3072806358337402,
+ "num_input_tokens_seen": 769757384,
+ "step": 17600,
+ "train_runtime": 136188.6995,
+ "train_tokens_per_second": 5652.138
+ },
+ {
+ "epoch": 1.1853622407756532,
+ "grad_norm": 0.6997799283337246,
+ "learning_rate": 3.5651582864084345e-06,
+ "loss": 0.32632970809936523,
+ "num_input_tokens_seen": 769975064,
+ "step": 17605,
+ "train_runtime": 136230.7878,
+ "train_tokens_per_second": 5651.99
+ },
+ {
+ "epoch": 1.1856988957716132,
+ "grad_norm": 0.7412339983103909,
+ "learning_rate": 3.5626256149163096e-06,
+ "loss": 0.3052994966506958,
+ "num_input_tokens_seen": 770193168,
+ "step": 17610,
+ "train_runtime": 136266.0259,
+ "train_tokens_per_second": 5652.129
+ },
+ {
+ "epoch": 1.1860355507675733,
+ "grad_norm": 0.7442659857634908,
+ "learning_rate": 3.5600933453813464e-06,
+ "loss": 0.3075841188430786,
+ "num_input_tokens_seen": 770408384,
+ "step": 17615,
+ "train_runtime": 136300.5956,
+ "train_tokens_per_second": 5652.275
+ },
+ {
+ "epoch": 1.1863722057635335,
+ "grad_norm": 0.7199031556232233,
+ "learning_rate": 3.5575614785116886e-06,
+ "loss": 0.28684291839599607,
+ "num_input_tokens_seen": 770620728,
+ "step": 17620,
+ "train_runtime": 136345.9364,
+ "train_tokens_per_second": 5651.952
+ },
+ {
+ "epoch": 1.1867088607594938,
+ "grad_norm": 0.8819942138485812,
+ "learning_rate": 3.5550300150153618e-06,
+ "loss": 0.3115447998046875,
+ "num_input_tokens_seen": 770842752,
+ "step": 17625,
+ "train_runtime": 136383.5236,
+ "train_tokens_per_second": 5652.023
+ },
+ {
+ "epoch": 1.1870455157554538,
+ "grad_norm": 0.7246608362400021,
+ "learning_rate": 3.5524989556002863e-06,
+ "loss": 0.29783077239990235,
+ "num_input_tokens_seen": 771060848,
+ "step": 17630,
+ "train_runtime": 136428.5453,
+ "train_tokens_per_second": 5651.756
+ },
+ {
+ "epoch": 1.1873821707514138,
+ "grad_norm": 0.7193175427066145,
+ "learning_rate": 3.5499683009742604e-06,
+ "loss": 0.30118565559387206,
+ "num_input_tokens_seen": 771263296,
+ "step": 17635,
+ "train_runtime": 136462.119,
+ "train_tokens_per_second": 5651.849
+ },
+ {
+ "epoch": 1.187718825747374,
+ "grad_norm": 0.7869452467124062,
+ "learning_rate": 3.5474380518449757e-06,
+ "loss": 0.2976966381072998,
+ "num_input_tokens_seen": 771484736,
+ "step": 17640,
+ "train_runtime": 136496.9783,
+ "train_tokens_per_second": 5652.028
+ },
+ {
+ "epoch": 1.1880554807433343,
+ "grad_norm": 0.7932366417755232,
+ "learning_rate": 3.5449082089200072e-06,
+ "loss": 0.301554536819458,
+ "num_input_tokens_seen": 771704632,
+ "step": 17645,
+ "train_runtime": 136536.9073,
+ "train_tokens_per_second": 5651.986
+ },
+ {
+ "epoch": 1.1883921357392944,
+ "grad_norm": 0.6938403571190614,
+ "learning_rate": 3.5423787729068206e-06,
+ "loss": 0.32130672931671145,
+ "num_input_tokens_seen": 771927880,
+ "step": 17650,
+ "train_runtime": 136573.1716,
+ "train_tokens_per_second": 5652.119
+ },
+ {
+ "epoch": 1.1887287907352544,
+ "grad_norm": 0.82729296746894,
+ "learning_rate": 3.539849744512759e-06,
+ "loss": 0.29790964126586916,
+ "num_input_tokens_seen": 772137016,
+ "step": 17655,
+ "train_runtime": 136621.1016,
+ "train_tokens_per_second": 5651.667
+ },
+ {
+ "epoch": 1.1890654457312146,
+ "grad_norm": 0.7549562336474883,
+ "learning_rate": 3.5373211244450633e-06,
+ "loss": 0.29845280647277833,
+ "num_input_tokens_seen": 772361904,
+ "step": 17660,
+ "train_runtime": 136662.555,
+ "train_tokens_per_second": 5651.599
+ },
+ {
+ "epoch": 1.189402100727175,
+ "grad_norm": 0.7357102335902723,
+ "learning_rate": 3.5347929134108483e-06,
+ "loss": 0.31483676433563235,
+ "num_input_tokens_seen": 772575080,
+ "step": 17665,
+ "train_runtime": 136699.3985,
+ "train_tokens_per_second": 5651.635
+ },
+ {
+ "epoch": 1.189738755723135,
+ "grad_norm": 0.8439288436808993,
+ "learning_rate": 3.532265112117125e-06,
+ "loss": 0.33406968116760255,
+ "num_input_tokens_seen": 772803264,
+ "step": 17670,
+ "train_runtime": 136740.7168,
+ "train_tokens_per_second": 5651.596
+ },
+ {
+ "epoch": 1.190075410719095,
+ "grad_norm": 0.7786740028253488,
+ "learning_rate": 3.5297377212707805e-06,
+ "loss": 0.2623422861099243,
+ "num_input_tokens_seen": 773015632,
+ "step": 17675,
+ "train_runtime": 136775.9051,
+ "train_tokens_per_second": 5651.695
+ },
+ {
+ "epoch": 1.1904120657150552,
+ "grad_norm": 0.7870032609519265,
+ "learning_rate": 3.527210741578596e-06,
+ "loss": 0.3026540279388428,
+ "num_input_tokens_seen": 773239976,
+ "step": 17680,
+ "train_runtime": 136808.3986,
+ "train_tokens_per_second": 5651.992
+ },
+ {
+ "epoch": 1.1907487207110155,
+ "grad_norm": 0.7671210970920903,
+ "learning_rate": 3.5246841737472304e-06,
+ "loss": 0.3219871520996094,
+ "num_input_tokens_seen": 773448912,
+ "step": 17685,
+ "train_runtime": 136855.0232,
+ "train_tokens_per_second": 5651.593
+ },
+ {
+ "epoch": 1.1910853757069755,
+ "grad_norm": 0.7580157770145218,
+ "learning_rate": 3.522158018483231e-06,
+ "loss": 0.2920974254608154,
+ "num_input_tokens_seen": 773675944,
+ "step": 17690,
+ "train_runtime": 136891.1287,
+ "train_tokens_per_second": 5651.761
+ },
+ {
+ "epoch": 1.1914220307029355,
+ "grad_norm": 0.7290657363390212,
+ "learning_rate": 3.519632276493029e-06,
+ "loss": 0.31039018630981446,
+ "num_input_tokens_seen": 773894272,
+ "step": 17695,
+ "train_runtime": 136937.1303,
+ "train_tokens_per_second": 5651.457
+ },
+ {
+ "epoch": 1.1917586856988958,
+ "grad_norm": 0.7233052465646747,
+ "learning_rate": 3.5171069484829417e-06,
+ "loss": 0.32564287185668944,
+ "num_input_tokens_seen": 774110520,
+ "step": 17700,
+ "train_runtime": 136981.8929,
+ "train_tokens_per_second": 5651.189
+ },
+ {
+ "epoch": 1.192095340694856,
+ "grad_norm": 0.8073757654663947,
+ "learning_rate": 3.5145820351591663e-06,
+ "loss": 0.28891453742980955,
+ "num_input_tokens_seen": 774319632,
+ "step": 17705,
+ "train_runtime": 137022.015,
+ "train_tokens_per_second": 5651.06
+ },
+ {
+ "epoch": 1.192431995690816,
+ "grad_norm": 0.7652622798191697,
+ "learning_rate": 3.51205753722779e-06,
+ "loss": 0.3185939073562622,
+ "num_input_tokens_seen": 774550784,
+ "step": 17710,
+ "train_runtime": 137056.4812,
+ "train_tokens_per_second": 5651.325
+ },
+ {
+ "epoch": 1.1927686506867763,
+ "grad_norm": 0.8416381367688279,
+ "learning_rate": 3.5095334553947757e-06,
+ "loss": 0.33892183303833007,
+ "num_input_tokens_seen": 774781848,
+ "step": 17715,
+ "train_runtime": 137094.7795,
+ "train_tokens_per_second": 5651.432
+ },
+ {
+ "epoch": 1.1931053056827363,
+ "grad_norm": 0.7454120054728669,
+ "learning_rate": 3.50700979036598e-06,
+ "loss": 0.2988421440124512,
+ "num_input_tokens_seen": 774997592,
+ "step": 17720,
+ "train_runtime": 137134.1311,
+ "train_tokens_per_second": 5651.384
+ },
+ {
+ "epoch": 1.1934419606786966,
+ "grad_norm": 0.7748610171469188,
+ "learning_rate": 3.5044865428471343e-06,
+ "loss": 0.28873891830444337,
+ "num_input_tokens_seen": 775203648,
+ "step": 17725,
+ "train_runtime": 137170.5574,
+ "train_tokens_per_second": 5651.385
+ },
+ {
+ "epoch": 1.1937786156746566,
+ "grad_norm": 0.7302857347443209,
+ "learning_rate": 3.50196371354386e-06,
+ "loss": 0.30784637928009034,
+ "num_input_tokens_seen": 775418528,
+ "step": 17730,
+ "train_runtime": 137207.7458,
+ "train_tokens_per_second": 5651.419
+ },
+ {
+ "epoch": 1.1941152706706168,
+ "grad_norm": 0.8321187550816316,
+ "learning_rate": 3.499441303161655e-06,
+ "loss": 0.3373323202133179,
+ "num_input_tokens_seen": 775644000,
+ "step": 17735,
+ "train_runtime": 137243.8108,
+ "train_tokens_per_second": 5651.577
+ },
+ {
+ "epoch": 1.1944519256665769,
+ "grad_norm": 0.6939534771748743,
+ "learning_rate": 3.4969193124059042e-06,
+ "loss": 0.29674830436706545,
+ "num_input_tokens_seen": 775872464,
+ "step": 17740,
+ "train_runtime": 137280.866,
+ "train_tokens_per_second": 5651.716
+ },
+ {
+ "epoch": 1.1947885806625371,
+ "grad_norm": 0.6781828643380845,
+ "learning_rate": 3.4943977419818748e-06,
+ "loss": 0.2963244438171387,
+ "num_input_tokens_seen": 776107304,
+ "step": 17745,
+ "train_runtime": 137317.7425,
+ "train_tokens_per_second": 5651.908
+ },
+ {
+ "epoch": 1.1951252356584972,
+ "grad_norm": 0.7205121541036261,
+ "learning_rate": 3.4918765925947173e-06,
+ "loss": 0.2947703838348389,
+ "num_input_tokens_seen": 776333360,
+ "step": 17750,
+ "train_runtime": 137361.1596,
+ "train_tokens_per_second": 5651.768
+ },
+ {
+ "epoch": 1.1954618906544574,
+ "grad_norm": 0.7958380388514483,
+ "learning_rate": 3.4893558649494595e-06,
+ "loss": 0.32445640563964845,
+ "num_input_tokens_seen": 776565352,
+ "step": 17755,
+ "train_runtime": 137407.4556,
+ "train_tokens_per_second": 5651.552
+ },
+ {
+ "epoch": 1.1957985456504174,
+ "grad_norm": 0.7848842866143078,
+ "learning_rate": 3.4868355597510194e-06,
+ "loss": 0.29216394424438474,
+ "num_input_tokens_seen": 776782064,
+ "step": 17760,
+ "train_runtime": 137451.8285,
+ "train_tokens_per_second": 5651.304
+ },
+ {
+ "epoch": 1.1961352006463777,
+ "grad_norm": 0.7869373477637197,
+ "learning_rate": 3.4843156777041887e-06,
+ "loss": 0.27714879512786866,
+ "num_input_tokens_seen": 777006896,
+ "step": 17765,
+ "train_runtime": 137487.3294,
+ "train_tokens_per_second": 5651.48
+ },
+ {
+ "epoch": 1.1964718556423377,
+ "grad_norm": 0.7742008308908815,
+ "learning_rate": 3.4817962195136457e-06,
+ "loss": 0.3087809085845947,
+ "num_input_tokens_seen": 777222560,
+ "step": 17770,
+ "train_runtime": 137530.2067,
+ "train_tokens_per_second": 5651.286
+ },
+ {
+ "epoch": 1.196808510638298,
+ "grad_norm": 0.8109631152985817,
+ "learning_rate": 3.479277185883947e-06,
+ "loss": 0.3070957183837891,
+ "num_input_tokens_seen": 777436752,
+ "step": 17775,
+ "train_runtime": 137571.1937,
+ "train_tokens_per_second": 5651.159
+ },
+ {
+ "epoch": 1.197145165634258,
+ "grad_norm": 0.956312966639188,
+ "learning_rate": 3.476758577519537e-06,
+ "loss": 0.29088459014892576,
+ "num_input_tokens_seen": 777654880,
+ "step": 17780,
+ "train_runtime": 137605.082,
+ "train_tokens_per_second": 5651.353
+ },
+ {
+ "epoch": 1.1974818206302182,
+ "grad_norm": 0.7882126606326822,
+ "learning_rate": 3.4742403951247318e-06,
+ "loss": 0.3081394672393799,
+ "num_input_tokens_seen": 777876192,
+ "step": 17785,
+ "train_runtime": 137646.6579,
+ "train_tokens_per_second": 5651.254
+ },
+ {
+ "epoch": 1.1978184756261783,
+ "grad_norm": 0.7275982569821285,
+ "learning_rate": 3.471722639403735e-06,
+ "loss": 0.29077887535095215,
+ "num_input_tokens_seen": 778090376,
+ "step": 17790,
+ "train_runtime": 137688.0934,
+ "train_tokens_per_second": 5651.109
+ },
+ {
+ "epoch": 1.1981551306221385,
+ "grad_norm": 0.7256020434598892,
+ "learning_rate": 3.4692053110606277e-06,
+ "loss": 0.2993917942047119,
+ "num_input_tokens_seen": 778289376,
+ "step": 17795,
+ "train_runtime": 137728.124,
+ "train_tokens_per_second": 5650.911
+ },
+ {
+ "epoch": 1.1984917856180985,
+ "grad_norm": 0.6738742049588294,
+ "learning_rate": 3.4666884107993766e-06,
+ "loss": 0.2942534923553467,
+ "num_input_tokens_seen": 778502056,
+ "step": 17800,
+ "train_runtime": 137764.3256,
+ "train_tokens_per_second": 5650.97
+ },
+ {
+ "epoch": 1.1988284406140588,
+ "grad_norm": 0.6924946642523322,
+ "learning_rate": 3.464171939323819e-06,
+ "loss": 0.30054686069488523,
+ "num_input_tokens_seen": 778716688,
+ "step": 17805,
+ "train_runtime": 137804.6948,
+ "train_tokens_per_second": 5650.872
+ },
+ {
+ "epoch": 1.1991650956100188,
+ "grad_norm": 0.7464221349536175,
+ "learning_rate": 3.4616558973376848e-06,
+ "loss": 0.3028707981109619,
+ "num_input_tokens_seen": 778927512,
+ "step": 17810,
+ "train_runtime": 137840.357,
+ "train_tokens_per_second": 5650.939
+ },
+ {
+ "epoch": 1.199501750605979,
+ "grad_norm": 0.8392280191250697,
+ "learning_rate": 3.4591402855445697e-06,
+ "loss": 0.30633254051208497,
+ "num_input_tokens_seen": 779150112,
+ "step": 17815,
+ "train_runtime": 137873.0033,
+ "train_tokens_per_second": 5651.216
+ },
+ {
+ "epoch": 1.199838405601939,
+ "grad_norm": 0.7385160748310561,
+ "learning_rate": 3.4566251046479625e-06,
+ "loss": 0.3079712152481079,
+ "num_input_tokens_seen": 779369912,
+ "step": 17820,
+ "train_runtime": 137915.2772,
+ "train_tokens_per_second": 5651.077
+ },
+ {
+ "epoch": 1.2001750605978994,
+ "grad_norm": 0.7785026542998826,
+ "learning_rate": 3.454110355351221e-06,
+ "loss": 0.3142574310302734,
+ "num_input_tokens_seen": 779590624,
+ "step": 17825,
+ "train_runtime": 137954.41,
+ "train_tokens_per_second": 5651.074
+ },
+ {
+ "epoch": 1.2005117155938594,
+ "grad_norm": 0.7070886799527935,
+ "learning_rate": 3.451596038357593e-06,
+ "loss": 0.3034769773483276,
+ "num_input_tokens_seen": 779821712,
+ "step": 17830,
+ "train_runtime": 137993.4334,
+ "train_tokens_per_second": 5651.151
+ },
+ {
+ "epoch": 1.2008483705898196,
+ "grad_norm": 0.8286609047748951,
+ "learning_rate": 3.4490821543701924e-06,
+ "loss": 0.32932286262512206,
+ "num_input_tokens_seen": 780045176,
+ "step": 17835,
+ "train_runtime": 138029.8077,
+ "train_tokens_per_second": 5651.281
+ },
+ {
+ "epoch": 1.2011850255857797,
+ "grad_norm": 0.7276434685048679,
+ "learning_rate": 3.4465687040920227e-06,
+ "loss": 0.30574448108673097,
+ "num_input_tokens_seen": 780268064,
+ "step": 17840,
+ "train_runtime": 138072.0423,
+ "train_tokens_per_second": 5651.166
+ },
+ {
+ "epoch": 1.20152168058174,
+ "grad_norm": 0.6824095105587578,
+ "learning_rate": 3.44405568822596e-06,
+ "loss": 0.278488302230835,
+ "num_input_tokens_seen": 780486632,
+ "step": 17845,
+ "train_runtime": 138113.2062,
+ "train_tokens_per_second": 5651.064
+ },
+ {
+ "epoch": 1.2018583355777,
+ "grad_norm": 0.833968026923619,
+ "learning_rate": 3.441543107474765e-06,
+ "loss": 0.3007037162780762,
+ "num_input_tokens_seen": 780694864,
+ "step": 17850,
+ "train_runtime": 138148.5762,
+ "train_tokens_per_second": 5651.125
+ },
+ {
+ "epoch": 1.2021949905736602,
+ "grad_norm": 0.7390135533381705,
+ "learning_rate": 3.439030962541069e-06,
+ "loss": 0.293853759765625,
+ "num_input_tokens_seen": 780910776,
+ "step": 17855,
+ "train_runtime": 138187.3343,
+ "train_tokens_per_second": 5651.102
+ },
+ {
+ "epoch": 1.2025316455696202,
+ "grad_norm": 0.7449577348762936,
+ "learning_rate": 3.4365192541273885e-06,
+ "loss": 0.30453169345855713,
+ "num_input_tokens_seen": 781111576,
+ "step": 17860,
+ "train_runtime": 138223.1035,
+ "train_tokens_per_second": 5651.093
+ },
+ {
+ "epoch": 1.2028683005655805,
+ "grad_norm": 0.7209045215476029,
+ "learning_rate": 3.4340079829361108e-06,
+ "loss": 0.3177401304244995,
+ "num_input_tokens_seen": 781317608,
+ "step": 17865,
+ "train_runtime": 138256.2897,
+ "train_tokens_per_second": 5651.226
+ },
+ {
+ "epoch": 1.2032049555615405,
+ "grad_norm": 0.7033530226860483,
+ "learning_rate": 3.4314971496695078e-06,
+ "loss": 0.31711881160736083,
+ "num_input_tokens_seen": 781545368,
+ "step": 17870,
+ "train_runtime": 138290.4596,
+ "train_tokens_per_second": 5651.477
+ },
+ {
+ "epoch": 1.2035416105575008,
+ "grad_norm": 0.7085516748330253,
+ "learning_rate": 3.428986755029724e-06,
+ "loss": 0.2896330118179321,
+ "num_input_tokens_seen": 781774192,
+ "step": 17875,
+ "train_runtime": 138336.6049,
+ "train_tokens_per_second": 5651.246
+ },
+ {
+ "epoch": 1.2038782655534608,
+ "grad_norm": 0.7125099135090971,
+ "learning_rate": 3.4264767997187874e-06,
+ "loss": 0.2978813171386719,
+ "num_input_tokens_seen": 782004920,
+ "step": 17880,
+ "train_runtime": 138372.9121,
+ "train_tokens_per_second": 5651.431
+ },
+ {
+ "epoch": 1.204214920549421,
+ "grad_norm": 0.7572580628329799,
+ "learning_rate": 3.423967284438594e-06,
+ "loss": 0.3156701326370239,
+ "num_input_tokens_seen": 782217384,
+ "step": 17885,
+ "train_runtime": 138408.8011,
+ "train_tokens_per_second": 5651.5
+ },
+ {
+ "epoch": 1.204551575545381,
+ "grad_norm": 0.6615339386741577,
+ "learning_rate": 3.421458209890925e-06,
+ "loss": 0.2962948799133301,
+ "num_input_tokens_seen": 782450248,
+ "step": 17890,
+ "train_runtime": 138443.1781,
+ "train_tokens_per_second": 5651.779
+ },
+ {
+ "epoch": 1.2048882305413413,
+ "grad_norm": 0.700079916053248,
+ "learning_rate": 3.418949576777433e-06,
+ "loss": 0.28663947582244875,
+ "num_input_tokens_seen": 782681456,
+ "step": 17895,
+ "train_runtime": 138480.6285,
+ "train_tokens_per_second": 5651.92
+ },
+ {
+ "epoch": 1.2052248855373013,
+ "grad_norm": 0.775430671158068,
+ "learning_rate": 3.4164413857996513e-06,
+ "loss": 0.30020842552185056,
+ "num_input_tokens_seen": 782891192,
+ "step": 17900,
+ "train_runtime": 138522.9513,
+ "train_tokens_per_second": 5651.707
+ },
+ {
+ "epoch": 1.2055615405332616,
+ "grad_norm": 0.7147718137595105,
+ "learning_rate": 3.4139336376589853e-06,
+ "loss": 0.291434121131897,
+ "num_input_tokens_seen": 783119112,
+ "step": 17905,
+ "train_runtime": 138554.6985,
+ "train_tokens_per_second": 5652.057
+ },
+ {
+ "epoch": 1.2058981955292216,
+ "grad_norm": 0.8762127003247577,
+ "learning_rate": 3.411426333056722e-06,
+ "loss": 0.30569398403167725,
+ "num_input_tokens_seen": 783321392,
+ "step": 17910,
+ "train_runtime": 138597.0294,
+ "train_tokens_per_second": 5651.791
+ },
+ {
+ "epoch": 1.2062348505251819,
+ "grad_norm": 0.7685183853693496,
+ "learning_rate": 3.408919472694017e-06,
+ "loss": 0.321109676361084,
+ "num_input_tokens_seen": 783543848,
+ "step": 17915,
+ "train_runtime": 138637.4131,
+ "train_tokens_per_second": 5651.749
+ },
+ {
+ "epoch": 1.206571505521142,
+ "grad_norm": 0.7046016721897478,
+ "learning_rate": 3.406413057271909e-06,
+ "loss": 0.30394299030303956,
+ "num_input_tokens_seen": 783758592,
+ "step": 17920,
+ "train_runtime": 138676.5892,
+ "train_tokens_per_second": 5651.701
+ },
+ {
+ "epoch": 1.2069081605171021,
+ "grad_norm": 0.7508302781476457,
+ "learning_rate": 3.403907087491306e-06,
+ "loss": 0.303491997718811,
+ "num_input_tokens_seen": 783976568,
+ "step": 17925,
+ "train_runtime": 138712.5584,
+ "train_tokens_per_second": 5651.807
+ },
+ {
+ "epoch": 1.2072448155130622,
+ "grad_norm": 0.7413375117669119,
+ "learning_rate": 3.4014015640529996e-06,
+ "loss": 0.30534958839416504,
+ "num_input_tokens_seen": 784193960,
+ "step": 17930,
+ "train_runtime": 138748.3843,
+ "train_tokens_per_second": 5651.914
+ },
+ {
+ "epoch": 1.2075814705090224,
+ "grad_norm": 0.7576411855768873,
+ "learning_rate": 3.398896487657646e-06,
+ "loss": 0.31456284523010253,
+ "num_input_tokens_seen": 784414136,
+ "step": 17935,
+ "train_runtime": 138791.0719,
+ "train_tokens_per_second": 5651.762
+ },
+ {
+ "epoch": 1.2079181255049825,
+ "grad_norm": 0.7960278015173246,
+ "learning_rate": 3.396391859005785e-06,
+ "loss": 0.2981652975082397,
+ "num_input_tokens_seen": 784638336,
+ "step": 17940,
+ "train_runtime": 138825.2652,
+ "train_tokens_per_second": 5651.985
+ },
+ {
+ "epoch": 1.2082547805009427,
+ "grad_norm": 0.6970907266821021,
+ "learning_rate": 3.393887678797826e-06,
+ "loss": 0.3018413305282593,
+ "num_input_tokens_seen": 784844856,
+ "step": 17945,
+ "train_runtime": 138863.6497,
+ "train_tokens_per_second": 5651.91
+ },
+ {
+ "epoch": 1.2085914354969027,
+ "grad_norm": 0.7592630519307038,
+ "learning_rate": 3.3913839477340575e-06,
+ "loss": 0.30618746280670167,
+ "num_input_tokens_seen": 785058936,
+ "step": 17950,
+ "train_runtime": 138902.6182,
+ "train_tokens_per_second": 5651.866
+ },
+ {
+ "epoch": 1.208928090492863,
+ "grad_norm": 0.7766479409436013,
+ "learning_rate": 3.388880666514637e-06,
+ "loss": 0.28571882247924807,
+ "num_input_tokens_seen": 785286832,
+ "step": 17955,
+ "train_runtime": 138937.3108,
+ "train_tokens_per_second": 5652.095
+ },
+ {
+ "epoch": 1.209264745488823,
+ "grad_norm": 0.7267074307937196,
+ "learning_rate": 3.3863778358396033e-06,
+ "loss": 0.3155225276947021,
+ "num_input_tokens_seen": 785504144,
+ "step": 17960,
+ "train_runtime": 138972.8709,
+ "train_tokens_per_second": 5652.212
+ },
+ {
+ "epoch": 1.2096014004847833,
+ "grad_norm": 0.7097174073868221,
+ "learning_rate": 3.3838754564088592e-06,
+ "loss": 0.3005061149597168,
+ "num_input_tokens_seen": 785729392,
+ "step": 17965,
+ "train_runtime": 139011.4611,
+ "train_tokens_per_second": 5652.263
+ },
+ {
+ "epoch": 1.2099380554807433,
+ "grad_norm": 0.6597216885356778,
+ "learning_rate": 3.3813735289221923e-06,
+ "loss": 0.29288544654846194,
+ "num_input_tokens_seen": 785948848,
+ "step": 17970,
+ "train_runtime": 139049.8184,
+ "train_tokens_per_second": 5652.282
+ },
+ {
+ "epoch": 1.2102747104767035,
+ "grad_norm": 0.6269995153972749,
+ "learning_rate": 3.378872054079255e-06,
+ "loss": 0.29325079917907715,
+ "num_input_tokens_seen": 786168672,
+ "step": 17975,
+ "train_runtime": 139089.6764,
+ "train_tokens_per_second": 5652.243
+ },
+ {
+ "epoch": 1.2106113654726636,
+ "grad_norm": 0.7838620260591436,
+ "learning_rate": 3.3763710325795784e-06,
+ "loss": 0.3322885513305664,
+ "num_input_tokens_seen": 786371976,
+ "step": 17980,
+ "train_runtime": 139126.705,
+ "train_tokens_per_second": 5652.2
+ },
+ {
+ "epoch": 1.2109480204686238,
+ "grad_norm": 0.6771354234860218,
+ "learning_rate": 3.3738704651225673e-06,
+ "loss": 0.3040419340133667,
+ "num_input_tokens_seen": 786600776,
+ "step": 17985,
+ "train_runtime": 139162.6054,
+ "train_tokens_per_second": 5652.386
+ },
+ {
+ "epoch": 1.2112846754645838,
+ "grad_norm": 0.7878835736125291,
+ "learning_rate": 3.3713703524074936e-06,
+ "loss": 0.2779910326004028,
+ "num_input_tokens_seen": 786804904,
+ "step": 17990,
+ "train_runtime": 139199.5468,
+ "train_tokens_per_second": 5652.352
+ },
+ {
+ "epoch": 1.211621330460544,
+ "grad_norm": 0.8227585744222804,
+ "learning_rate": 3.3688706951335082e-06,
+ "loss": 0.31528596878051757,
+ "num_input_tokens_seen": 787008024,
+ "step": 17995,
+ "train_runtime": 139235.8527,
+ "train_tokens_per_second": 5652.337
+ },
+ {
+ "epoch": 1.2119579854565041,
+ "grad_norm": 0.7422262732589775,
+ "learning_rate": 3.3663714939996305e-06,
+ "loss": 0.30718555450439455,
+ "num_input_tokens_seen": 787245096,
+ "step": 18000,
+ "train_runtime": 139272.1742,
+ "train_tokens_per_second": 5652.566
+ },
+ {
+ "epoch": 1.2122946404524644,
+ "grad_norm": 0.7541583112520067,
+ "learning_rate": 3.3638727497047575e-06,
+ "loss": 0.2900678634643555,
+ "num_input_tokens_seen": 787449376,
+ "step": 18005,
+ "train_runtime": 139315.1398,
+ "train_tokens_per_second": 5652.289
+ },
+ {
+ "epoch": 1.2126312954484244,
+ "grad_norm": 0.7220759362847328,
+ "learning_rate": 3.3613744629476507e-06,
+ "loss": 0.29748926162719724,
+ "num_input_tokens_seen": 787661624,
+ "step": 18010,
+ "train_runtime": 139356.0904,
+ "train_tokens_per_second": 5652.151
+ },
+ {
+ "epoch": 1.2129679504443847,
+ "grad_norm": 0.7372103337616008,
+ "learning_rate": 3.358876634426952e-06,
+ "loss": 0.2816985845565796,
+ "num_input_tokens_seen": 787888360,
+ "step": 18015,
+ "train_runtime": 139393.2354,
+ "train_tokens_per_second": 5652.271
+ },
+ {
+ "epoch": 1.2133046054403447,
+ "grad_norm": 0.6737993858813106,
+ "learning_rate": 3.3563792648411676e-06,
+ "loss": 0.32491137981414797,
+ "num_input_tokens_seen": 788091136,
+ "step": 18020,
+ "train_runtime": 139429.8296,
+ "train_tokens_per_second": 5652.242
+ },
+ {
+ "epoch": 1.213641260436305,
+ "grad_norm": 0.7236596806287144,
+ "learning_rate": 3.353882354888681e-06,
+ "loss": 0.288041353225708,
+ "num_input_tokens_seen": 788312048,
+ "step": 18025,
+ "train_runtime": 139470.5579,
+ "train_tokens_per_second": 5652.175
+ },
+ {
+ "epoch": 1.213977915432265,
+ "grad_norm": 0.7377693831933381,
+ "learning_rate": 3.3513859052677444e-06,
+ "loss": 0.2752676486968994,
+ "num_input_tokens_seen": 788512224,
+ "step": 18030,
+ "train_runtime": 139508.9565,
+ "train_tokens_per_second": 5652.054
+ },
+ {
+ "epoch": 1.2143145704282252,
+ "grad_norm": 0.7799869819088461,
+ "learning_rate": 3.3488899166764842e-06,
+ "loss": 0.3178478479385376,
+ "num_input_tokens_seen": 788737240,
+ "step": 18035,
+ "train_runtime": 139542.6128,
+ "train_tokens_per_second": 5652.304
+ },
+ {
+ "epoch": 1.2146512254241852,
+ "grad_norm": 0.8463476604239449,
+ "learning_rate": 3.346394389812892e-06,
+ "loss": 0.3054813385009766,
+ "num_input_tokens_seen": 788959120,
+ "step": 18040,
+ "train_runtime": 139577.1377,
+ "train_tokens_per_second": 5652.495
+ },
+ {
+ "epoch": 1.2149878804201455,
+ "grad_norm": 0.7048797167467936,
+ "learning_rate": 3.3438993253748365e-06,
+ "loss": 0.2954627752304077,
+ "num_input_tokens_seen": 789168384,
+ "step": 18045,
+ "train_runtime": 139620.335,
+ "train_tokens_per_second": 5652.245
+ },
+ {
+ "epoch": 1.2153245354161055,
+ "grad_norm": 0.7007505922657411,
+ "learning_rate": 3.3414047240600526e-06,
+ "loss": 0.3028736591339111,
+ "num_input_tokens_seen": 789390496,
+ "step": 18050,
+ "train_runtime": 139660.3558,
+ "train_tokens_per_second": 5652.216
+ },
+ {
+ "epoch": 1.2156611904120658,
+ "grad_norm": 0.8100526851783998,
+ "learning_rate": 3.338910586566151e-06,
+ "loss": 0.3098297119140625,
+ "num_input_tokens_seen": 789600216,
+ "step": 18055,
+ "train_runtime": 139695.7464,
+ "train_tokens_per_second": 5652.285
+ },
+ {
+ "epoch": 1.2159978454080258,
+ "grad_norm": 0.7531507425187864,
+ "learning_rate": 3.336416913590605e-06,
+ "loss": 0.3021399974822998,
+ "num_input_tokens_seen": 789807928,
+ "step": 18060,
+ "train_runtime": 139731.0332,
+ "train_tokens_per_second": 5652.344
+ },
+ {
+ "epoch": 1.216334500403986,
+ "grad_norm": 0.8809448489320266,
+ "learning_rate": 3.333923705830766e-06,
+ "loss": 0.3216712474822998,
+ "num_input_tokens_seen": 790032400,
+ "step": 18065,
+ "train_runtime": 139770.9889,
+ "train_tokens_per_second": 5652.335
+ },
+ {
+ "epoch": 1.216671155399946,
+ "grad_norm": 0.7668922416234147,
+ "learning_rate": 3.3314309639838484e-06,
+ "loss": 0.3249736547470093,
+ "num_input_tokens_seen": 790253040,
+ "step": 18070,
+ "train_runtime": 139801.8412,
+ "train_tokens_per_second": 5652.665
+ },
+ {
+ "epoch": 1.2170078103959063,
+ "grad_norm": 0.7501530270251235,
+ "learning_rate": 3.328938688746942e-06,
+ "loss": 0.30673856735229493,
+ "num_input_tokens_seen": 790481600,
+ "step": 18075,
+ "train_runtime": 139848.0026,
+ "train_tokens_per_second": 5652.434
+ },
+ {
+ "epoch": 1.2173444653918664,
+ "grad_norm": 0.7405483975043581,
+ "learning_rate": 3.3264468808170012e-06,
+ "loss": 0.3439304828643799,
+ "num_input_tokens_seen": 790707832,
+ "step": 18080,
+ "train_runtime": 139884.7103,
+ "train_tokens_per_second": 5652.568
+ },
+ {
+ "epoch": 1.2176811203878266,
+ "grad_norm": 0.7461154895905937,
+ "learning_rate": 3.3239555408908562e-06,
+ "loss": 0.3145643711090088,
+ "num_input_tokens_seen": 790904072,
+ "step": 18085,
+ "train_runtime": 139927.9767,
+ "train_tokens_per_second": 5652.223
+ },
+ {
+ "epoch": 1.2180177753837866,
+ "grad_norm": 0.7450961063246211,
+ "learning_rate": 3.3214646696651974e-06,
+ "loss": 0.2987529277801514,
+ "num_input_tokens_seen": 791119712,
+ "step": 18090,
+ "train_runtime": 139967.4008,
+ "train_tokens_per_second": 5652.171
+ },
+ {
+ "epoch": 1.2183544303797469,
+ "grad_norm": 0.7553620001734184,
+ "learning_rate": 3.3189742678365923e-06,
+ "loss": 0.31772985458374026,
+ "num_input_tokens_seen": 791340456,
+ "step": 18095,
+ "train_runtime": 140002.9288,
+ "train_tokens_per_second": 5652.314
+ },
+ {
+ "epoch": 1.218691085375707,
+ "grad_norm": 0.6740962905213788,
+ "learning_rate": 3.316484336101472e-06,
+ "loss": 0.28296403884887694,
+ "num_input_tokens_seen": 791546784,
+ "step": 18100,
+ "train_runtime": 140038.0985,
+ "train_tokens_per_second": 5652.367
+ },
+ {
+ "epoch": 1.2190277403716672,
+ "grad_norm": 0.7648105865576331,
+ "learning_rate": 3.313994875156141e-06,
+ "loss": 0.31280903816223143,
+ "num_input_tokens_seen": 791764024,
+ "step": 18105,
+ "train_runtime": 140081.5209,
+ "train_tokens_per_second": 5652.166
+ },
+ {
+ "epoch": 1.2193643953676272,
+ "grad_norm": 0.7674425963272242,
+ "learning_rate": 3.3115058856967643e-06,
+ "loss": 0.31598572731018065,
+ "num_input_tokens_seen": 791993616,
+ "step": 18110,
+ "train_runtime": 140122.4895,
+ "train_tokens_per_second": 5652.152
+ },
+ {
+ "epoch": 1.2197010503635874,
+ "grad_norm": 0.7813158906447104,
+ "learning_rate": 3.3090173684193845e-06,
+ "loss": 0.27434990406036375,
+ "num_input_tokens_seen": 792207856,
+ "step": 18115,
+ "train_runtime": 140159.8456,
+ "train_tokens_per_second": 5652.174
+ },
+ {
+ "epoch": 1.2200377053595475,
+ "grad_norm": 0.7015793012796032,
+ "learning_rate": 3.3065293240199037e-06,
+ "loss": 0.31843407154083253,
+ "num_input_tokens_seen": 792431928,
+ "step": 18120,
+ "train_runtime": 140202.8473,
+ "train_tokens_per_second": 5652.039
+ },
+ {
+ "epoch": 1.2203743603555077,
+ "grad_norm": 0.8362321813673071,
+ "learning_rate": 3.3040417531940982e-06,
+ "loss": 0.3226888179779053,
+ "num_input_tokens_seen": 792655832,
+ "step": 18125,
+ "train_runtime": 140238.3502,
+ "train_tokens_per_second": 5652.204
+ },
+ {
+ "epoch": 1.2207110153514678,
+ "grad_norm": 0.691317135330766,
+ "learning_rate": 3.301554656637607e-06,
+ "loss": 0.2980408906936646,
+ "num_input_tokens_seen": 792868664,
+ "step": 18130,
+ "train_runtime": 140271.1305,
+ "train_tokens_per_second": 5652.401
+ },
+ {
+ "epoch": 1.221047670347428,
+ "grad_norm": 0.7033787564345391,
+ "learning_rate": 3.2990680350459427e-06,
+ "loss": 0.29863677024841306,
+ "num_input_tokens_seen": 793080312,
+ "step": 18135,
+ "train_runtime": 140311.0004,
+ "train_tokens_per_second": 5652.303
+ },
+ {
+ "epoch": 1.221384325343388,
+ "grad_norm": 0.7340650029847289,
+ "learning_rate": 3.2965818891144763e-06,
+ "loss": 0.3160408020019531,
+ "num_input_tokens_seen": 793303920,
+ "step": 18140,
+ "train_runtime": 140340.856,
+ "train_tokens_per_second": 5652.694
+ },
+ {
+ "epoch": 1.2217209803393483,
+ "grad_norm": 0.7408714024481081,
+ "learning_rate": 3.294096219538454e-06,
+ "loss": 0.322130012512207,
+ "num_input_tokens_seen": 793536920,
+ "step": 18145,
+ "train_runtime": 140387.2152,
+ "train_tokens_per_second": 5652.487
+ },
+ {
+ "epoch": 1.2220576353353083,
+ "grad_norm": 0.725033878974291,
+ "learning_rate": 3.291611027012983e-06,
+ "loss": 0.3116767406463623,
+ "num_input_tokens_seen": 793761088,
+ "step": 18150,
+ "train_runtime": 140429.0399,
+ "train_tokens_per_second": 5652.4
+ },
+ {
+ "epoch": 1.2223942903312686,
+ "grad_norm": 0.8302224573076524,
+ "learning_rate": 3.289126312233043e-06,
+ "loss": 0.29644060134887695,
+ "num_input_tokens_seen": 793993376,
+ "step": 18155,
+ "train_runtime": 140469.6701,
+ "train_tokens_per_second": 5652.419
+ },
+ {
+ "epoch": 1.2227309453272286,
+ "grad_norm": 0.6981694762570436,
+ "learning_rate": 3.2866420758934714e-06,
+ "loss": 0.3092994213104248,
+ "num_input_tokens_seen": 794216288,
+ "step": 18160,
+ "train_runtime": 140503.4652,
+ "train_tokens_per_second": 5652.646
+ },
+ {
+ "epoch": 1.2230676003231888,
+ "grad_norm": 0.6856222569708069,
+ "learning_rate": 3.2841583186889826e-06,
+ "loss": 0.30644276142120364,
+ "num_input_tokens_seen": 794442800,
+ "step": 18165,
+ "train_runtime": 140540.8201,
+ "train_tokens_per_second": 5652.755
+ },
+ {
+ "epoch": 1.2234042553191489,
+ "grad_norm": 0.76204653073967,
+ "learning_rate": 3.281675041314146e-06,
+ "loss": 0.2917419195175171,
+ "num_input_tokens_seen": 794656160,
+ "step": 18170,
+ "train_runtime": 140583.221,
+ "train_tokens_per_second": 5652.568
+ },
+ {
+ "epoch": 1.2237409103151091,
+ "grad_norm": 0.7110260250075573,
+ "learning_rate": 3.279192244463406e-06,
+ "loss": 0.29529433250427245,
+ "num_input_tokens_seen": 794869448,
+ "step": 18175,
+ "train_runtime": 140624.4232,
+ "train_tokens_per_second": 5652.428
+ },
+ {
+ "epoch": 1.2240775653110691,
+ "grad_norm": 0.7343704316676943,
+ "learning_rate": 3.2767099288310667e-06,
+ "loss": 0.29960105419158933,
+ "num_input_tokens_seen": 795093624,
+ "step": 18180,
+ "train_runtime": 140664.5318,
+ "train_tokens_per_second": 5652.41
+ },
+ {
+ "epoch": 1.2244142203070294,
+ "grad_norm": 0.639609983163744,
+ "learning_rate": 3.2742280951113025e-06,
+ "loss": 0.2962838888168335,
+ "num_input_tokens_seen": 795322792,
+ "step": 18185,
+ "train_runtime": 140702.7441,
+ "train_tokens_per_second": 5652.504
+ },
+ {
+ "epoch": 1.2247508753029894,
+ "grad_norm": 0.743438894841407,
+ "learning_rate": 3.2717467439981464e-06,
+ "loss": 0.30564584732055666,
+ "num_input_tokens_seen": 795532552,
+ "step": 18190,
+ "train_runtime": 140745.8939,
+ "train_tokens_per_second": 5652.261
+ },
+ {
+ "epoch": 1.2250875302989497,
+ "grad_norm": 0.7489547177708321,
+ "learning_rate": 3.2692658761855033e-06,
+ "loss": 0.27375977039337157,
+ "num_input_tokens_seen": 795735568,
+ "step": 18195,
+ "train_runtime": 140781.1568,
+ "train_tokens_per_second": 5652.287
+ },
+ {
+ "epoch": 1.2254241852949097,
+ "grad_norm": 0.6829987732324682,
+ "learning_rate": 3.266785492367138e-06,
+ "loss": 0.2808785915374756,
+ "num_input_tokens_seen": 795955960,
+ "step": 18200,
+ "train_runtime": 140815.8619,
+ "train_tokens_per_second": 5652.46
+ },
+ {
+ "epoch": 1.22576084029087,
+ "grad_norm": 0.7035746454375664,
+ "learning_rate": 3.2643055932366852e-06,
+ "loss": 0.28676323890686034,
+ "num_input_tokens_seen": 796165624,
+ "step": 18205,
+ "train_runtime": 140854.0679,
+ "train_tokens_per_second": 5652.415
+ },
+ {
+ "epoch": 1.22609749528683,
+ "grad_norm": 0.8214707476930566,
+ "learning_rate": 3.2618261794876365e-06,
+ "loss": 0.3010567665100098,
+ "num_input_tokens_seen": 796395392,
+ "step": 18210,
+ "train_runtime": 140889.0235,
+ "train_tokens_per_second": 5652.643
+ },
+ {
+ "epoch": 1.2264341502827902,
+ "grad_norm": 0.6985742460828861,
+ "learning_rate": 3.259347251813356e-06,
+ "loss": 0.2966732978820801,
+ "num_input_tokens_seen": 796617856,
+ "step": 18215,
+ "train_runtime": 140927.9467,
+ "train_tokens_per_second": 5652.661
+ },
+ {
+ "epoch": 1.2267708052787503,
+ "grad_norm": 0.727228038188911,
+ "learning_rate": 3.256868810907064e-06,
+ "loss": 0.3215049982070923,
+ "num_input_tokens_seen": 796852648,
+ "step": 18220,
+ "train_runtime": 140967.9622,
+ "train_tokens_per_second": 5652.722
+ },
+ {
+ "epoch": 1.2271074602747105,
+ "grad_norm": 0.7512872700129827,
+ "learning_rate": 3.2543908574618514e-06,
+ "loss": 0.3135005712509155,
+ "num_input_tokens_seen": 797070736,
+ "step": 18225,
+ "train_runtime": 141012.3331,
+ "train_tokens_per_second": 5652.49
+ },
+ {
+ "epoch": 1.2274441152706705,
+ "grad_norm": 0.6464160299656531,
+ "learning_rate": 3.251913392170668e-06,
+ "loss": 0.2944124460220337,
+ "num_input_tokens_seen": 797290688,
+ "step": 18230,
+ "train_runtime": 141057.431,
+ "train_tokens_per_second": 5652.242
+ },
+ {
+ "epoch": 1.2277807702666308,
+ "grad_norm": 0.7214923202638721,
+ "learning_rate": 3.249436415726333e-06,
+ "loss": 0.29227581024169924,
+ "num_input_tokens_seen": 797516256,
+ "step": 18235,
+ "train_runtime": 141095.1719,
+ "train_tokens_per_second": 5652.328
+ },
+ {
+ "epoch": 1.2281174252625908,
+ "grad_norm": 0.8506838334171156,
+ "learning_rate": 3.24695992882152e-06,
+ "loss": 0.3139004707336426,
+ "num_input_tokens_seen": 797725232,
+ "step": 18240,
+ "train_runtime": 141134.7796,
+ "train_tokens_per_second": 5652.223
+ },
+ {
+ "epoch": 1.228454080258551,
+ "grad_norm": 0.6887059927032259,
+ "learning_rate": 3.244483932148773e-06,
+ "loss": 0.3086080551147461,
+ "num_input_tokens_seen": 797943968,
+ "step": 18245,
+ "train_runtime": 141166.1,
+ "train_tokens_per_second": 5652.518
+ },
+ {
+ "epoch": 1.228790735254511,
+ "grad_norm": 0.7231712696156054,
+ "learning_rate": 3.2420084264004966e-06,
+ "loss": 0.30695390701293945,
+ "num_input_tokens_seen": 798161800,
+ "step": 18250,
+ "train_runtime": 141201.6515,
+ "train_tokens_per_second": 5652.638
+ },
+ {
+ "epoch": 1.2291273902504714,
+ "grad_norm": 0.7340695565349038,
+ "learning_rate": 3.2395334122689594e-06,
+ "loss": 0.30524253845214844,
+ "num_input_tokens_seen": 798379968,
+ "step": 18255,
+ "train_runtime": 141240.9452,
+ "train_tokens_per_second": 5652.61
+ },
+ {
+ "epoch": 1.2294640452464314,
+ "grad_norm": 0.8802102442010598,
+ "learning_rate": 3.2370588904462873e-06,
+ "loss": 0.3120718955993652,
+ "num_input_tokens_seen": 798609312,
+ "step": 18260,
+ "train_runtime": 141276.169,
+ "train_tokens_per_second": 5652.824
+ },
+ {
+ "epoch": 1.2298007002423916,
+ "grad_norm": 0.7640356901782597,
+ "learning_rate": 3.2345848616244775e-06,
+ "loss": 0.3038747549057007,
+ "num_input_tokens_seen": 798804840,
+ "step": 18265,
+ "train_runtime": 141315.8853,
+ "train_tokens_per_second": 5652.619
+ },
+ {
+ "epoch": 1.2301373552383517,
+ "grad_norm": 0.8040217158342741,
+ "learning_rate": 3.2321113264953797e-06,
+ "loss": 0.307041072845459,
+ "num_input_tokens_seen": 799027080,
+ "step": 18270,
+ "train_runtime": 141347.3832,
+ "train_tokens_per_second": 5652.932
+ },
+ {
+ "epoch": 1.230474010234312,
+ "grad_norm": 0.8528931267954475,
+ "learning_rate": 3.2296382857507124e-06,
+ "loss": 0.2948000431060791,
+ "num_input_tokens_seen": 799238584,
+ "step": 18275,
+ "train_runtime": 141386.4656,
+ "train_tokens_per_second": 5652.865
+ },
+ {
+ "epoch": 1.230810665230272,
+ "grad_norm": 0.7187404855989243,
+ "learning_rate": 3.227165740082053e-06,
+ "loss": 0.31221818923950195,
+ "num_input_tokens_seen": 799452544,
+ "step": 18280,
+ "train_runtime": 141428.6252,
+ "train_tokens_per_second": 5652.693
+ },
+ {
+ "epoch": 1.2311473202262322,
+ "grad_norm": 0.762382560173622,
+ "learning_rate": 3.2246936901808424e-06,
+ "loss": 0.2914170980453491,
+ "num_input_tokens_seen": 799675552,
+ "step": 18285,
+ "train_runtime": 141467.594,
+ "train_tokens_per_second": 5652.712
+ },
+ {
+ "epoch": 1.2314839752221922,
+ "grad_norm": 0.7606748879884372,
+ "learning_rate": 3.2222221367383786e-06,
+ "loss": 0.32509727478027345,
+ "num_input_tokens_seen": 799898272,
+ "step": 18290,
+ "train_runtime": 141515.3162,
+ "train_tokens_per_second": 5652.379
+ },
+ {
+ "epoch": 1.2318206302181525,
+ "grad_norm": 0.7608834756975011,
+ "learning_rate": 3.2197510804458265e-06,
+ "loss": 0.30308525562286376,
+ "num_input_tokens_seen": 800105152,
+ "step": 18295,
+ "train_runtime": 141554.2953,
+ "train_tokens_per_second": 5652.285
+ },
+ {
+ "epoch": 1.2321572852141125,
+ "grad_norm": 0.6370726326925514,
+ "learning_rate": 3.217280521994207e-06,
+ "loss": 0.2986292362213135,
+ "num_input_tokens_seen": 800331312,
+ "step": 18300,
+ "train_runtime": 141590.2519,
+ "train_tokens_per_second": 5652.446
+ },
+ {
+ "epoch": 1.2324939402100727,
+ "grad_norm": 0.7624529476398764,
+ "learning_rate": 3.214810462074407e-06,
+ "loss": 0.3016390562057495,
+ "num_input_tokens_seen": 800551688,
+ "step": 18305,
+ "train_runtime": 141623.377,
+ "train_tokens_per_second": 5652.68
+ },
+ {
+ "epoch": 1.2328305952060328,
+ "grad_norm": 0.7693052320987479,
+ "learning_rate": 3.2123409013771666e-06,
+ "loss": 0.30267770290374757,
+ "num_input_tokens_seen": 800777344,
+ "step": 18310,
+ "train_runtime": 141660.7953,
+ "train_tokens_per_second": 5652.78
+ },
+ {
+ "epoch": 1.233167250201993,
+ "grad_norm": 0.6772837658060004,
+ "learning_rate": 3.209871840593095e-06,
+ "loss": 0.31267609596252444,
+ "num_input_tokens_seen": 801009256,
+ "step": 18315,
+ "train_runtime": 141706.1002,
+ "train_tokens_per_second": 5652.61
+ },
+ {
+ "epoch": 1.233503905197953,
+ "grad_norm": 0.7125458851611337,
+ "learning_rate": 3.207403280412652e-06,
+ "loss": 0.30893242359161377,
+ "num_input_tokens_seen": 801228256,
+ "step": 18320,
+ "train_runtime": 141743.076,
+ "train_tokens_per_second": 5652.68
+ },
+ {
+ "epoch": 1.2338405601939133,
+ "grad_norm": 0.6945330021115061,
+ "learning_rate": 3.2049352215261677e-06,
+ "loss": 0.28535709381103513,
+ "num_input_tokens_seen": 801458800,
+ "step": 18325,
+ "train_runtime": 141780.6123,
+ "train_tokens_per_second": 5652.81
+ },
+ {
+ "epoch": 1.2341772151898733,
+ "grad_norm": 0.6960194289052382,
+ "learning_rate": 3.2024676646238222e-06,
+ "loss": 0.29358530044555664,
+ "num_input_tokens_seen": 801679448,
+ "step": 18330,
+ "train_runtime": 141816.9621,
+ "train_tokens_per_second": 5652.917
+ },
+ {
+ "epoch": 1.2345138701858336,
+ "grad_norm": 0.8171703953778717,
+ "learning_rate": 3.2000006103956653e-06,
+ "loss": 0.3098346948623657,
+ "num_input_tokens_seen": 801900976,
+ "step": 18335,
+ "train_runtime": 141857.1454,
+ "train_tokens_per_second": 5652.877
+ },
+ {
+ "epoch": 1.2348505251817936,
+ "grad_norm": 0.7299590696983723,
+ "learning_rate": 3.1975340595315956e-06,
+ "loss": 0.28231058120727537,
+ "num_input_tokens_seen": 802111128,
+ "step": 18340,
+ "train_runtime": 141893.2475,
+ "train_tokens_per_second": 5652.92
+ },
+ {
+ "epoch": 1.2351871801777539,
+ "grad_norm": 0.7644787058330598,
+ "learning_rate": 3.1950680127213796e-06,
+ "loss": 0.2976810455322266,
+ "num_input_tokens_seen": 802330912,
+ "step": 18345,
+ "train_runtime": 141932.7239,
+ "train_tokens_per_second": 5652.896
+ },
+ {
+ "epoch": 1.235523835173714,
+ "grad_norm": 0.7526412783611733,
+ "learning_rate": 3.192602470654638e-06,
+ "loss": 0.32010388374328613,
+ "num_input_tokens_seen": 802544448,
+ "step": 18350,
+ "train_runtime": 141974.2746,
+ "train_tokens_per_second": 5652.746
+ },
+ {
+ "epoch": 1.2358604901696741,
+ "grad_norm": 0.7865649130269652,
+ "learning_rate": 3.190137434020853e-06,
+ "loss": 0.2869425296783447,
+ "num_input_tokens_seen": 802766184,
+ "step": 18355,
+ "train_runtime": 142021.1129,
+ "train_tokens_per_second": 5652.443
+ },
+ {
+ "epoch": 1.2361971451656342,
+ "grad_norm": 0.8600627915587952,
+ "learning_rate": 3.1876729035093616e-06,
+ "loss": 0.29856894016265867,
+ "num_input_tokens_seen": 802986112,
+ "step": 18360,
+ "train_runtime": 142056.2119,
+ "train_tokens_per_second": 5652.594
+ },
+ {
+ "epoch": 1.2365338001615944,
+ "grad_norm": 0.7568409452496826,
+ "learning_rate": 3.185208879809366e-06,
+ "loss": 0.29802355766296384,
+ "num_input_tokens_seen": 803214056,
+ "step": 18365,
+ "train_runtime": 142100.6909,
+ "train_tokens_per_second": 5652.429
+ },
+ {
+ "epoch": 1.2368704551575544,
+ "grad_norm": 0.7529248541722069,
+ "learning_rate": 3.1827453636099183e-06,
+ "loss": 0.3028829574584961,
+ "num_input_tokens_seen": 803449464,
+ "step": 18370,
+ "train_runtime": 142136.647,
+ "train_tokens_per_second": 5652.655
+ },
+ {
+ "epoch": 1.2372071101535147,
+ "grad_norm": 0.9260284783600946,
+ "learning_rate": 3.1802823555999356e-06,
+ "loss": 0.2902355194091797,
+ "num_input_tokens_seen": 803675088,
+ "step": 18375,
+ "train_runtime": 142173.9871,
+ "train_tokens_per_second": 5652.758
+ },
+ {
+ "epoch": 1.2375437651494747,
+ "grad_norm": 0.7733188398886425,
+ "learning_rate": 3.1778198564681885e-06,
+ "loss": 0.32343668937683107,
+ "num_input_tokens_seen": 803903288,
+ "step": 18380,
+ "train_runtime": 142205.7209,
+ "train_tokens_per_second": 5653.101
+ },
+ {
+ "epoch": 1.237880420145435,
+ "grad_norm": 0.7192786766579897,
+ "learning_rate": 3.17535786690331e-06,
+ "loss": 0.30882773399353025,
+ "num_input_tokens_seen": 804121616,
+ "step": 18385,
+ "train_runtime": 142246.3225,
+ "train_tokens_per_second": 5653.022
+ },
+ {
+ "epoch": 1.238217075141395,
+ "grad_norm": 0.7729011461174,
+ "learning_rate": 3.172896387593784e-06,
+ "loss": 0.304299521446228,
+ "num_input_tokens_seen": 804348264,
+ "step": 18390,
+ "train_runtime": 142288.5576,
+ "train_tokens_per_second": 5652.937
+ },
+ {
+ "epoch": 1.2385537301373553,
+ "grad_norm": 0.8194549580251287,
+ "learning_rate": 3.1704354192279573e-06,
+ "loss": 0.31928482055664065,
+ "num_input_tokens_seen": 804564656,
+ "step": 18395,
+ "train_runtime": 142323.8728,
+ "train_tokens_per_second": 5653.055
+ },
+ {
+ "epoch": 1.2388903851333153,
+ "grad_norm": 0.7962633908276077,
+ "learning_rate": 3.1679749624940316e-06,
+ "loss": 0.318212103843689,
+ "num_input_tokens_seen": 804786400,
+ "step": 18400,
+ "train_runtime": 142360.3102,
+ "train_tokens_per_second": 5653.166
+ },
+ {
+ "epoch": 1.2392270401292755,
+ "grad_norm": 0.7855805493097535,
+ "learning_rate": 3.165515018080067e-06,
+ "loss": 0.3026642084121704,
+ "num_input_tokens_seen": 805005768,
+ "step": 18405,
+ "train_runtime": 142394.1464,
+ "train_tokens_per_second": 5653.363
+ },
+ {
+ "epoch": 1.2395636951252356,
+ "grad_norm": 0.938130030820677,
+ "learning_rate": 3.163055586673976e-06,
+ "loss": 0.29436650276184084,
+ "num_input_tokens_seen": 805226016,
+ "step": 18410,
+ "train_runtime": 142427.7333,
+ "train_tokens_per_second": 5653.576
+ },
+ {
+ "epoch": 1.2399003501211958,
+ "grad_norm": 0.7559775139780297,
+ "learning_rate": 3.1605966689635348e-06,
+ "loss": 0.32002964019775393,
+ "num_input_tokens_seen": 805453752,
+ "step": 18415,
+ "train_runtime": 142464.7236,
+ "train_tokens_per_second": 5653.707
+ },
+ {
+ "epoch": 1.2402370051171558,
+ "grad_norm": 0.8613984382442207,
+ "learning_rate": 3.1581382656363666e-06,
+ "loss": 0.30423684120178224,
+ "num_input_tokens_seen": 805658856,
+ "step": 18420,
+ "train_runtime": 142494.9465,
+ "train_tokens_per_second": 5653.947
+ },
+ {
+ "epoch": 1.240573660113116,
+ "grad_norm": 0.8697803980165589,
+ "learning_rate": 3.1556803773799616e-06,
+ "loss": 0.3185387372970581,
+ "num_input_tokens_seen": 805880136,
+ "step": 18425,
+ "train_runtime": 142536.9031,
+ "train_tokens_per_second": 5653.835
+ },
+ {
+ "epoch": 1.2409103151090761,
+ "grad_norm": 0.7422635875140698,
+ "learning_rate": 3.1532230048816564e-06,
+ "loss": 0.3022879123687744,
+ "num_input_tokens_seen": 806097056,
+ "step": 18430,
+ "train_runtime": 142575.9432,
+ "train_tokens_per_second": 5653.808
+ },
+ {
+ "epoch": 1.2412469701050364,
+ "grad_norm": 0.6776328969826404,
+ "learning_rate": 3.150766148828651e-06,
+ "loss": 0.3036804676055908,
+ "num_input_tokens_seen": 806315480,
+ "step": 18435,
+ "train_runtime": 142613.0736,
+ "train_tokens_per_second": 5653.868
+ },
+ {
+ "epoch": 1.2415836251009964,
+ "grad_norm": 0.7561483870082674,
+ "learning_rate": 3.1483098099079934e-06,
+ "loss": 0.2874735116958618,
+ "num_input_tokens_seen": 806538456,
+ "step": 18440,
+ "train_runtime": 142661.6297,
+ "train_tokens_per_second": 5653.507
+ },
+ {
+ "epoch": 1.2419202800969567,
+ "grad_norm": 0.7257415829423263,
+ "learning_rate": 3.1458539888065937e-06,
+ "loss": 0.3029060363769531,
+ "num_input_tokens_seen": 806750424,
+ "step": 18445,
+ "train_runtime": 142702.778,
+ "train_tokens_per_second": 5653.362
+ },
+ {
+ "epoch": 1.2422569350929167,
+ "grad_norm": 0.84249717889109,
+ "learning_rate": 3.143398686211212e-06,
+ "loss": 0.29645981788635256,
+ "num_input_tokens_seen": 806958552,
+ "step": 18450,
+ "train_runtime": 142740.2229,
+ "train_tokens_per_second": 5653.337
+ },
+ {
+ "epoch": 1.242593590088877,
+ "grad_norm": 0.7359383150490025,
+ "learning_rate": 3.1409439028084706e-06,
+ "loss": 0.3129459857940674,
+ "num_input_tokens_seen": 807178464,
+ "step": 18455,
+ "train_runtime": 142779.9615,
+ "train_tokens_per_second": 5653.304
+ },
+ {
+ "epoch": 1.2429302450848372,
+ "grad_norm": 0.7356032513359525,
+ "learning_rate": 3.1384896392848363e-06,
+ "loss": 0.28752105236053466,
+ "num_input_tokens_seen": 807397576,
+ "step": 18460,
+ "train_runtime": 142821.147,
+ "train_tokens_per_second": 5653.207
+ },
+ {
+ "epoch": 1.2432669000807972,
+ "grad_norm": 0.7330504631691293,
+ "learning_rate": 3.1360358963266404e-06,
+ "loss": 0.32807331085205077,
+ "num_input_tokens_seen": 807624936,
+ "step": 18465,
+ "train_runtime": 142852.5383,
+ "train_tokens_per_second": 5653.557
+ },
+ {
+ "epoch": 1.2436035550767572,
+ "grad_norm": 0.7686799242262236,
+ "learning_rate": 3.1335826746200594e-06,
+ "loss": 0.3099084377288818,
+ "num_input_tokens_seen": 807820280,
+ "step": 18470,
+ "train_runtime": 142890.1797,
+ "train_tokens_per_second": 5653.435
+ },
+ {
+ "epoch": 1.2439402100727175,
+ "grad_norm": 0.7644480228425453,
+ "learning_rate": 3.1311299748511335e-06,
+ "loss": 0.32409205436706545,
+ "num_input_tokens_seen": 808042832,
+ "step": 18475,
+ "train_runtime": 142931.8612,
+ "train_tokens_per_second": 5653.343
+ },
+ {
+ "epoch": 1.2442768650686777,
+ "grad_norm": 0.7485842376379487,
+ "learning_rate": 3.1286777977057504e-06,
+ "loss": 0.29155697822570803,
+ "num_input_tokens_seen": 808252592,
+ "step": 18480,
+ "train_runtime": 142970.1026,
+ "train_tokens_per_second": 5653.298
+ },
+ {
+ "epoch": 1.2446135200646378,
+ "grad_norm": 0.8186958870732818,
+ "learning_rate": 3.126226143869655e-06,
+ "loss": 0.3249779224395752,
+ "num_input_tokens_seen": 808465720,
+ "step": 18485,
+ "train_runtime": 143009.3714,
+ "train_tokens_per_second": 5653.236
+ },
+ {
+ "epoch": 1.2449501750605978,
+ "grad_norm": 0.6799895472703079,
+ "learning_rate": 3.1237750140284424e-06,
+ "loss": 0.27743706703186033,
+ "num_input_tokens_seen": 808678680,
+ "step": 18490,
+ "train_runtime": 143043.1574,
+ "train_tokens_per_second": 5653.389
+ },
+ {
+ "epoch": 1.245286830056558,
+ "grad_norm": 0.8374856400249484,
+ "learning_rate": 3.1213244088675647e-06,
+ "loss": 0.30376696586608887,
+ "num_input_tokens_seen": 808892656,
+ "step": 18495,
+ "train_runtime": 143081.3078,
+ "train_tokens_per_second": 5653.378
+ },
+ {
+ "epoch": 1.2456234850525183,
+ "grad_norm": 0.7662091689060238,
+ "learning_rate": 3.1188743290723246e-06,
+ "loss": 0.317877197265625,
+ "num_input_tokens_seen": 809111464,
+ "step": 18500,
+ "train_runtime": 143125.016,
+ "train_tokens_per_second": 5653.18
+ },
+ {
+ "epoch": 1.2459601400484783,
+ "grad_norm": 0.8647091169854892,
+ "learning_rate": 3.116424775327882e-06,
+ "loss": 0.3008237361907959,
+ "num_input_tokens_seen": 809328360,
+ "step": 18505,
+ "train_runtime": 143173.0534,
+ "train_tokens_per_second": 5652.798
+ },
+ {
+ "epoch": 1.2462967950444384,
+ "grad_norm": 0.6945551074504395,
+ "learning_rate": 3.1139757483192423e-06,
+ "loss": 0.3034308433532715,
+ "num_input_tokens_seen": 809557648,
+ "step": 18510,
+ "train_runtime": 143207.7193,
+ "train_tokens_per_second": 5653.031
+ },
+ {
+ "epoch": 1.2466334500403986,
+ "grad_norm": 0.8318013387227233,
+ "learning_rate": 3.1115272487312727e-06,
+ "loss": 0.29793879985809324,
+ "num_input_tokens_seen": 809766040,
+ "step": 18515,
+ "train_runtime": 143239.1385,
+ "train_tokens_per_second": 5653.246
+ },
+ {
+ "epoch": 1.2469701050363589,
+ "grad_norm": 0.8696915186885799,
+ "learning_rate": 3.109079277248684e-06,
+ "loss": 0.29328839778900145,
+ "num_input_tokens_seen": 809972912,
+ "step": 18520,
+ "train_runtime": 143278.4616,
+ "train_tokens_per_second": 5653.138
+ },
+ {
+ "epoch": 1.2473067600323189,
+ "grad_norm": 0.7294833612317175,
+ "learning_rate": 3.1066318345560463e-06,
+ "loss": 0.32238657474517823,
+ "num_input_tokens_seen": 810187728,
+ "step": 18525,
+ "train_runtime": 143312.263,
+ "train_tokens_per_second": 5653.304
+ },
+ {
+ "epoch": 1.247643415028279,
+ "grad_norm": 0.7391236990985328,
+ "learning_rate": 3.1041849213377782e-06,
+ "loss": 0.3177293300628662,
+ "num_input_tokens_seen": 810419880,
+ "step": 18530,
+ "train_runtime": 143347.6909,
+ "train_tokens_per_second": 5653.526
+ },
+ {
+ "epoch": 1.2479800700242392,
+ "grad_norm": 0.7271664362688767,
+ "learning_rate": 3.101738538278154e-06,
+ "loss": 0.29890563488006594,
+ "num_input_tokens_seen": 810642576,
+ "step": 18535,
+ "train_runtime": 143382.8102,
+ "train_tokens_per_second": 5653.694
+ },
+ {
+ "epoch": 1.2483167250201994,
+ "grad_norm": 0.7689863894703435,
+ "learning_rate": 3.0992926860612914e-06,
+ "loss": 0.30640869140625,
+ "num_input_tokens_seen": 810855296,
+ "step": 18540,
+ "train_runtime": 143422.92,
+ "train_tokens_per_second": 5653.596
+ },
+ {
+ "epoch": 1.2486533800161594,
+ "grad_norm": 0.7776443662476371,
+ "learning_rate": 3.0968473653711693e-06,
+ "loss": 0.2990746259689331,
+ "num_input_tokens_seen": 811094136,
+ "step": 18545,
+ "train_runtime": 143462.1432,
+ "train_tokens_per_second": 5653.715
+ },
+ {
+ "epoch": 1.2489900350121195,
+ "grad_norm": 0.7266832056756357,
+ "learning_rate": 3.094402576891613e-06,
+ "loss": 0.3187594413757324,
+ "num_input_tokens_seen": 811315008,
+ "step": 18550,
+ "train_runtime": 143497.5164,
+ "train_tokens_per_second": 5653.861
+ },
+ {
+ "epoch": 1.2493266900080797,
+ "grad_norm": 0.6721184594034306,
+ "learning_rate": 3.0919583213063004e-06,
+ "loss": 0.3195340156555176,
+ "num_input_tokens_seen": 811546128,
+ "step": 18555,
+ "train_runtime": 143531.9724,
+ "train_tokens_per_second": 5654.114
+ },
+ {
+ "epoch": 1.24966334500404,
+ "grad_norm": 0.75584993986847,
+ "learning_rate": 3.0895145992987574e-06,
+ "loss": 0.30277261734008787,
+ "num_input_tokens_seen": 811776888,
+ "step": 18560,
+ "train_runtime": 143569.4998,
+ "train_tokens_per_second": 5654.243
+ },
+ {
+ "epoch": 1.25,
+ "grad_norm": 0.8137422225809046,
+ "learning_rate": 3.087071411552366e-06,
+ "loss": 0.2947868824005127,
+ "num_input_tokens_seen": 811992848,
+ "step": 18565,
+ "train_runtime": 143605.6285,
+ "train_tokens_per_second": 5654.325
+ },
+ {
+ "epoch": 1.25033665499596,
+ "grad_norm": 0.7305460030824085,
+ "learning_rate": 3.084628758750353e-06,
+ "loss": 0.31966290473937986,
+ "num_input_tokens_seen": 812211480,
+ "step": 18570,
+ "train_runtime": 143644.3856,
+ "train_tokens_per_second": 5654.321
+ },
+ {
+ "epoch": 1.2506733099919203,
+ "grad_norm": 0.9296238525143283,
+ "learning_rate": 3.082186641575801e-06,
+ "loss": 0.3240236759185791,
+ "num_input_tokens_seen": 812427616,
+ "step": 18575,
+ "train_runtime": 143686.3373,
+ "train_tokens_per_second": 5654.174
+ },
+ {
+ "epoch": 1.2510099649878805,
+ "grad_norm": 0.7503570194002699,
+ "learning_rate": 3.0797450607116384e-06,
+ "loss": 0.30055992603302,
+ "num_input_tokens_seen": 812646144,
+ "step": 18580,
+ "train_runtime": 143727.8546,
+ "train_tokens_per_second": 5654.062
+ },
+ {
+ "epoch": 1.2513466199838406,
+ "grad_norm": 0.7124035622045924,
+ "learning_rate": 3.0773040168406475e-06,
+ "loss": 0.31271209716796877,
+ "num_input_tokens_seen": 812878472,
+ "step": 18585,
+ "train_runtime": 143761.8158,
+ "train_tokens_per_second": 5654.342
+ },
+ {
+ "epoch": 1.2516832749798006,
+ "grad_norm": 0.7943563134878205,
+ "learning_rate": 3.074863510645456e-06,
+ "loss": 0.31382598876953127,
+ "num_input_tokens_seen": 813099832,
+ "step": 18590,
+ "train_runtime": 143801.9795,
+ "train_tokens_per_second": 5654.302
+ },
+ {
+ "epoch": 1.2520199299757608,
+ "grad_norm": 0.6636865747124504,
+ "learning_rate": 3.072423542808546e-06,
+ "loss": 0.30039525032043457,
+ "num_input_tokens_seen": 813316560,
+ "step": 18595,
+ "train_runtime": 143835.7935,
+ "train_tokens_per_second": 5654.48
+ },
+ {
+ "epoch": 1.252356584971721,
+ "grad_norm": 0.7917515381442752,
+ "learning_rate": 3.0699841140122443e-06,
+ "loss": 0.3078362464904785,
+ "num_input_tokens_seen": 813538248,
+ "step": 18600,
+ "train_runtime": 143877.45,
+ "train_tokens_per_second": 5654.383
+ },
+ {
+ "epoch": 1.2526932399676811,
+ "grad_norm": 0.7542366254079307,
+ "learning_rate": 3.0675452249387327e-06,
+ "loss": 0.2891314744949341,
+ "num_input_tokens_seen": 813765928,
+ "step": 18605,
+ "train_runtime": 143916.9674,
+ "train_tokens_per_second": 5654.413
+ },
+ {
+ "epoch": 1.2530298949636411,
+ "grad_norm": 0.769680983787988,
+ "learning_rate": 3.0651068762700354e-06,
+ "loss": 0.3373257637023926,
+ "num_input_tokens_seen": 813974344,
+ "step": 18610,
+ "train_runtime": 143956.4928,
+ "train_tokens_per_second": 5654.308
+ },
+ {
+ "epoch": 1.2533665499596014,
+ "grad_norm": 0.7698186645906399,
+ "learning_rate": 3.0626690686880313e-06,
+ "loss": 0.32206950187683103,
+ "num_input_tokens_seen": 814211744,
+ "step": 18615,
+ "train_runtime": 143994.9815,
+ "train_tokens_per_second": 5654.445
+ },
+ {
+ "epoch": 1.2537032049555616,
+ "grad_norm": 0.7376028562088249,
+ "learning_rate": 3.060231802874443e-06,
+ "loss": 0.3022738218307495,
+ "num_input_tokens_seen": 814438872,
+ "step": 18620,
+ "train_runtime": 144031.2128,
+ "train_tokens_per_second": 5654.6
+ },
+ {
+ "epoch": 1.2540398599515217,
+ "grad_norm": 0.7155062346403165,
+ "learning_rate": 3.0577950795108462e-06,
+ "loss": 0.30109267234802245,
+ "num_input_tokens_seen": 814661808,
+ "step": 18625,
+ "train_runtime": 144071.0237,
+ "train_tokens_per_second": 5654.585
+ },
+ {
+ "epoch": 1.2543765149474817,
+ "grad_norm": 0.7986947159381109,
+ "learning_rate": 3.055358899278662e-06,
+ "loss": 0.2987940788269043,
+ "num_input_tokens_seen": 814871016,
+ "step": 18630,
+ "train_runtime": 144109.5084,
+ "train_tokens_per_second": 5654.526
+ },
+ {
+ "epoch": 1.254713169943442,
+ "grad_norm": 0.8209503548573042,
+ "learning_rate": 3.052923262859162e-06,
+ "loss": 0.3024166107177734,
+ "num_input_tokens_seen": 815098200,
+ "step": 18635,
+ "train_runtime": 144149.6967,
+ "train_tokens_per_second": 5654.526
+ },
+ {
+ "epoch": 1.2550498249394022,
+ "grad_norm": 0.7146717304289526,
+ "learning_rate": 3.0504881709334603e-06,
+ "loss": 0.29818005561828614,
+ "num_input_tokens_seen": 815319256,
+ "step": 18640,
+ "train_runtime": 144189.1823,
+ "train_tokens_per_second": 5654.511
+ },
+ {
+ "epoch": 1.2553864799353622,
+ "grad_norm": 0.7132265699815106,
+ "learning_rate": 3.0480536241825263e-06,
+ "loss": 0.28137338161468506,
+ "num_input_tokens_seen": 815537448,
+ "step": 18645,
+ "train_runtime": 144230.2973,
+ "train_tokens_per_second": 5654.411
+ },
+ {
+ "epoch": 1.2557231349313223,
+ "grad_norm": 0.7429234432398791,
+ "learning_rate": 3.0456196232871715e-06,
+ "loss": 0.31012353897094724,
+ "num_input_tokens_seen": 815749224,
+ "step": 18650,
+ "train_runtime": 144268.7463,
+ "train_tokens_per_second": 5654.372
+ },
+ {
+ "epoch": 1.2560597899272825,
+ "grad_norm": 0.8026609003022944,
+ "learning_rate": 3.043186168928059e-06,
+ "loss": 0.30425825119018557,
+ "num_input_tokens_seen": 815954440,
+ "step": 18655,
+ "train_runtime": 144316.2977,
+ "train_tokens_per_second": 5653.931
+ },
+ {
+ "epoch": 1.2563964449232428,
+ "grad_norm": 0.9158794619186253,
+ "learning_rate": 3.040753261785692e-06,
+ "loss": 0.2835689067840576,
+ "num_input_tokens_seen": 816164248,
+ "step": 18660,
+ "train_runtime": 144353.3412,
+ "train_tokens_per_second": 5653.934
+ },
+ {
+ "epoch": 1.2567330999192028,
+ "grad_norm": 0.7533631787720405,
+ "learning_rate": 3.038320902540429e-06,
+ "loss": 0.28765387535095216,
+ "num_input_tokens_seen": 816379936,
+ "step": 18665,
+ "train_runtime": 144394.4225,
+ "train_tokens_per_second": 5653.819
+ },
+ {
+ "epoch": 1.2570697549151628,
+ "grad_norm": 0.7215132949101682,
+ "learning_rate": 3.0358890918724706e-06,
+ "loss": 0.303347110748291,
+ "num_input_tokens_seen": 816611832,
+ "step": 18670,
+ "train_runtime": 144441.5414,
+ "train_tokens_per_second": 5653.58
+ },
+ {
+ "epoch": 1.257406409911123,
+ "grad_norm": 0.8254057548339484,
+ "learning_rate": 3.0334578304618646e-06,
+ "loss": 0.30634298324584963,
+ "num_input_tokens_seen": 816814520,
+ "step": 18675,
+ "train_runtime": 144483.9095,
+ "train_tokens_per_second": 5653.325
+ },
+ {
+ "epoch": 1.2577430649070833,
+ "grad_norm": 0.7191371330865693,
+ "learning_rate": 3.0310271189885037e-06,
+ "loss": 0.3063030481338501,
+ "num_input_tokens_seen": 817024696,
+ "step": 18680,
+ "train_runtime": 144517.3756,
+ "train_tokens_per_second": 5653.47
+ },
+ {
+ "epoch": 1.2580797199030433,
+ "grad_norm": 0.7073217630646664,
+ "learning_rate": 3.0285969581321328e-06,
+ "loss": 0.296706485748291,
+ "num_input_tokens_seen": 817244872,
+ "step": 18685,
+ "train_runtime": 144554.5892,
+ "train_tokens_per_second": 5653.538
+ },
+ {
+ "epoch": 1.2584163748990034,
+ "grad_norm": 0.7607230653856618,
+ "learning_rate": 3.0261673485723343e-06,
+ "loss": 0.3513062000274658,
+ "num_input_tokens_seen": 817456408,
+ "step": 18690,
+ "train_runtime": 144592.758,
+ "train_tokens_per_second": 5653.509
+ },
+ {
+ "epoch": 1.2587530298949636,
+ "grad_norm": 0.7009162874744629,
+ "learning_rate": 3.023738290988544e-06,
+ "loss": 0.28501205444335936,
+ "num_input_tokens_seen": 817681112,
+ "step": 18695,
+ "train_runtime": 144634.3333,
+ "train_tokens_per_second": 5653.437
+ },
+ {
+ "epoch": 1.2590896848909239,
+ "grad_norm": 0.8160907857921498,
+ "learning_rate": 3.021309786060037e-06,
+ "loss": 0.30846569538116453,
+ "num_input_tokens_seen": 817885488,
+ "step": 18700,
+ "train_runtime": 144674.0076,
+ "train_tokens_per_second": 5653.299
+ },
+ {
+ "epoch": 1.259426339886884,
+ "grad_norm": 0.7466687057656971,
+ "learning_rate": 3.018881834465941e-06,
+ "loss": 0.290647029876709,
+ "num_input_tokens_seen": 818098808,
+ "step": 18705,
+ "train_runtime": 144712.1439,
+ "train_tokens_per_second": 5653.284
+ },
+ {
+ "epoch": 1.259762994882844,
+ "grad_norm": 0.7496967934224382,
+ "learning_rate": 3.016454436885221e-06,
+ "loss": 0.3127418041229248,
+ "num_input_tokens_seen": 818317480,
+ "step": 18710,
+ "train_runtime": 144748.3739,
+ "train_tokens_per_second": 5653.379
+ },
+ {
+ "epoch": 1.2600996498788042,
+ "grad_norm": 0.7415188653692539,
+ "learning_rate": 3.0140275939966934e-06,
+ "loss": 0.3062695026397705,
+ "num_input_tokens_seen": 818539128,
+ "step": 18715,
+ "train_runtime": 144789.8872,
+ "train_tokens_per_second": 5653.289
+ },
+ {
+ "epoch": 1.2604363048747644,
+ "grad_norm": 0.7926838386125326,
+ "learning_rate": 3.0116013064790166e-06,
+ "loss": 0.3510124921798706,
+ "num_input_tokens_seen": 818770136,
+ "step": 18720,
+ "train_runtime": 144822.5073,
+ "train_tokens_per_second": 5653.611
+ },
+ {
+ "epoch": 1.2607729598707245,
+ "grad_norm": 0.838754094026344,
+ "learning_rate": 3.0091755750106937e-06,
+ "loss": 0.31281888484954834,
+ "num_input_tokens_seen": 818987192,
+ "step": 18725,
+ "train_runtime": 144857.9964,
+ "train_tokens_per_second": 5653.724
+ },
+ {
+ "epoch": 1.2611096148666845,
+ "grad_norm": 0.7065966064188791,
+ "learning_rate": 3.006750400270072e-06,
+ "loss": 0.2828061580657959,
+ "num_input_tokens_seen": 819201568,
+ "step": 18730,
+ "train_runtime": 144893.0397,
+ "train_tokens_per_second": 5653.837
+ },
+ {
+ "epoch": 1.2614462698626447,
+ "grad_norm": 0.6761766202190553,
+ "learning_rate": 3.0043257829353466e-06,
+ "loss": 0.2915013790130615,
+ "num_input_tokens_seen": 819423736,
+ "step": 18735,
+ "train_runtime": 144932.2044,
+ "train_tokens_per_second": 5653.842
+ },
+ {
+ "epoch": 1.261782924858605,
+ "grad_norm": 0.7399210906210721,
+ "learning_rate": 3.0019017236845504e-06,
+ "loss": 0.3179666042327881,
+ "num_input_tokens_seen": 819640880,
+ "step": 18740,
+ "train_runtime": 144969.2626,
+ "train_tokens_per_second": 5653.894
+ },
+ {
+ "epoch": 1.262119579854565,
+ "grad_norm": 1.0185583570035832,
+ "learning_rate": 2.9994782231955673e-06,
+ "loss": 0.3084989547729492,
+ "num_input_tokens_seen": 819858960,
+ "step": 18745,
+ "train_runtime": 145006.0911,
+ "train_tokens_per_second": 5653.962
+ },
+ {
+ "epoch": 1.262456234850525,
+ "grad_norm": 0.6872092978916245,
+ "learning_rate": 2.9970552821461185e-06,
+ "loss": 0.28854634761810305,
+ "num_input_tokens_seen": 820070184,
+ "step": 18750,
+ "train_runtime": 145045.6889,
+ "train_tokens_per_second": 5653.875
+ },
+ {
+ "epoch": 1.2627928898464853,
+ "grad_norm": 0.8684996226363714,
+ "learning_rate": 2.994632901213775e-06,
+ "loss": 0.2994524955749512,
+ "num_input_tokens_seen": 820280600,
+ "step": 18755,
+ "train_runtime": 145083.5094,
+ "train_tokens_per_second": 5653.851
+ },
+ {
+ "epoch": 1.2631295448424456,
+ "grad_norm": 0.8319193699909638,
+ "learning_rate": 2.9922110810759443e-06,
+ "loss": 0.3167592763900757,
+ "num_input_tokens_seen": 820492832,
+ "step": 18760,
+ "train_runtime": 145127.9985,
+ "train_tokens_per_second": 5653.581
+ },
+ {
+ "epoch": 1.2634661998384056,
+ "grad_norm": 0.7753044461490659,
+ "learning_rate": 2.989789822409883e-06,
+ "loss": 0.3189209461212158,
+ "num_input_tokens_seen": 820717400,
+ "step": 18765,
+ "train_runtime": 145162.0094,
+ "train_tokens_per_second": 5653.803
+ },
+ {
+ "epoch": 1.2638028548343656,
+ "grad_norm": 0.7322817419625537,
+ "learning_rate": 2.9873691258926864e-06,
+ "loss": 0.3269504547119141,
+ "num_input_tokens_seen": 820956960,
+ "step": 18770,
+ "train_runtime": 145198.2506,
+ "train_tokens_per_second": 5654.042
+ },
+ {
+ "epoch": 1.2641395098303259,
+ "grad_norm": 0.696541167423223,
+ "learning_rate": 2.9849489922012985e-06,
+ "loss": 0.3042433261871338,
+ "num_input_tokens_seen": 821165096,
+ "step": 18775,
+ "train_runtime": 145239.0804,
+ "train_tokens_per_second": 5653.885
+ },
+ {
+ "epoch": 1.264476164826286,
+ "grad_norm": 0.7452468533447384,
+ "learning_rate": 2.982529422012498e-06,
+ "loss": 0.338847017288208,
+ "num_input_tokens_seen": 821398992,
+ "step": 18780,
+ "train_runtime": 145277.8958,
+ "train_tokens_per_second": 5653.985
+ },
+ {
+ "epoch": 1.2648128198222461,
+ "grad_norm": 0.7815954898874985,
+ "learning_rate": 2.9801104160029127e-06,
+ "loss": 0.2665702819824219,
+ "num_input_tokens_seen": 821606024,
+ "step": 18785,
+ "train_runtime": 145319.4203,
+ "train_tokens_per_second": 5653.794
+ },
+ {
+ "epoch": 1.2651494748182062,
+ "grad_norm": 0.7521512397733022,
+ "learning_rate": 2.9776919748490063e-06,
+ "loss": 0.31261477470397947,
+ "num_input_tokens_seen": 821822040,
+ "step": 18790,
+ "train_runtime": 145359.9037,
+ "train_tokens_per_second": 5653.705
+ },
+ {
+ "epoch": 1.2654861298141664,
+ "grad_norm": 0.8383923854339487,
+ "learning_rate": 2.975274099227092e-06,
+ "loss": 0.32703323364257814,
+ "num_input_tokens_seen": 822027760,
+ "step": 18795,
+ "train_runtime": 145403.5301,
+ "train_tokens_per_second": 5653.424
+ },
+ {
+ "epoch": 1.2658227848101267,
+ "grad_norm": 0.8931253914400595,
+ "learning_rate": 2.9728567898133183e-06,
+ "loss": 0.31384882926940916,
+ "num_input_tokens_seen": 822249072,
+ "step": 18800,
+ "train_runtime": 145447.7238,
+ "train_tokens_per_second": 5653.227
+ },
+ {
+ "epoch": 1.2661594398060867,
+ "grad_norm": 0.7239471432554654,
+ "learning_rate": 2.9704400472836816e-06,
+ "loss": 0.3009427309036255,
+ "num_input_tokens_seen": 822468088,
+ "step": 18805,
+ "train_runtime": 145491.919,
+ "train_tokens_per_second": 5653.016
+ },
+ {
+ "epoch": 1.266496094802047,
+ "grad_norm": 0.8248452170258889,
+ "learning_rate": 2.9680238723140108e-06,
+ "loss": 0.33100879192352295,
+ "num_input_tokens_seen": 822682520,
+ "step": 18810,
+ "train_runtime": 145531.6555,
+ "train_tokens_per_second": 5652.946
+ },
+ {
+ "epoch": 1.266832749798007,
+ "grad_norm": 0.7312225706048872,
+ "learning_rate": 2.9656082655799866e-06,
+ "loss": 0.3179032325744629,
+ "num_input_tokens_seen": 822894336,
+ "step": 18815,
+ "train_runtime": 145569.5437,
+ "train_tokens_per_second": 5652.929
+ },
+ {
+ "epoch": 1.2671694047939672,
+ "grad_norm": 0.8914282863081362,
+ "learning_rate": 2.9631932277571225e-06,
+ "loss": 0.31571259498596194,
+ "num_input_tokens_seen": 823116688,
+ "step": 18820,
+ "train_runtime": 145602.2926,
+ "train_tokens_per_second": 5653.185
+ },
+ {
+ "epoch": 1.2675060597899273,
+ "grad_norm": 0.7436639756998058,
+ "learning_rate": 2.9607787595207794e-06,
+ "loss": 0.2941970109939575,
+ "num_input_tokens_seen": 823334680,
+ "step": 18825,
+ "train_runtime": 145642.6854,
+ "train_tokens_per_second": 5653.114
+ },
+ {
+ "epoch": 1.2678427147858875,
+ "grad_norm": 0.6064296816907161,
+ "learning_rate": 2.9583648615461514e-06,
+ "loss": 0.2951462507247925,
+ "num_input_tokens_seen": 823548976,
+ "step": 18830,
+ "train_runtime": 145684.7379,
+ "train_tokens_per_second": 5652.953
+ },
+ {
+ "epoch": 1.2681793697818475,
+ "grad_norm": 0.8015900434605868,
+ "learning_rate": 2.955951534508281e-06,
+ "loss": 0.32911660671234133,
+ "num_input_tokens_seen": 823759784,
+ "step": 18835,
+ "train_runtime": 145720.3213,
+ "train_tokens_per_second": 5653.019
+ },
+ {
+ "epoch": 1.2685160247778078,
+ "grad_norm": 0.6806231841224424,
+ "learning_rate": 2.9535387790820452e-06,
+ "loss": 0.3123172283172607,
+ "num_input_tokens_seen": 823989856,
+ "step": 18840,
+ "train_runtime": 145763.7657,
+ "train_tokens_per_second": 5652.913
+ },
+ {
+ "epoch": 1.2688526797737678,
+ "grad_norm": 0.827759809808618,
+ "learning_rate": 2.951126595942165e-06,
+ "loss": 0.29312863349914553,
+ "num_input_tokens_seen": 824217960,
+ "step": 18845,
+ "train_runtime": 145800.553,
+ "train_tokens_per_second": 5653.051
+ },
+ {
+ "epoch": 1.269189334769728,
+ "grad_norm": 0.7164470604004327,
+ "learning_rate": 2.948714985763197e-06,
+ "loss": 0.2757392883300781,
+ "num_input_tokens_seen": 824421568,
+ "step": 18850,
+ "train_runtime": 145841.8634,
+ "train_tokens_per_second": 5652.846
+ },
+ {
+ "epoch": 1.269525989765688,
+ "grad_norm": 0.8496090603821623,
+ "learning_rate": 2.9463039492195445e-06,
+ "loss": 0.3131323575973511,
+ "num_input_tokens_seen": 824655760,
+ "step": 18855,
+ "train_runtime": 145880.5542,
+ "train_tokens_per_second": 5652.952
+ },
+ {
+ "epoch": 1.2698626447616483,
+ "grad_norm": 0.7352726813474353,
+ "learning_rate": 2.9438934869854417e-06,
+ "loss": 0.2902811527252197,
+ "num_input_tokens_seen": 824879032,
+ "step": 18860,
+ "train_runtime": 145923.738,
+ "train_tokens_per_second": 5652.809
+ },
+ {
+ "epoch": 1.2701992997576084,
+ "grad_norm": 0.7811715951881948,
+ "learning_rate": 2.9414835997349705e-06,
+ "loss": 0.2894813060760498,
+ "num_input_tokens_seen": 825095288,
+ "step": 18865,
+ "train_runtime": 145960.032,
+ "train_tokens_per_second": 5652.885
+ },
+ {
+ "epoch": 1.2705359547535686,
+ "grad_norm": 0.6562307917959052,
+ "learning_rate": 2.939074288142045e-06,
+ "loss": 0.32269885540008547,
+ "num_input_tokens_seen": 825326376,
+ "step": 18870,
+ "train_runtime": 145998.5987,
+ "train_tokens_per_second": 5652.975
+ },
+ {
+ "epoch": 1.2708726097495286,
+ "grad_norm": 0.7677795002971816,
+ "learning_rate": 2.9366655528804257e-06,
+ "loss": 0.29350924491882324,
+ "num_input_tokens_seen": 825544816,
+ "step": 18875,
+ "train_runtime": 146042.0539,
+ "train_tokens_per_second": 5652.788
+ },
+ {
+ "epoch": 1.271209264745489,
+ "grad_norm": 0.6905930617244008,
+ "learning_rate": 2.934257394623702e-06,
+ "loss": 0.2879796504974365,
+ "num_input_tokens_seen": 825768752,
+ "step": 18880,
+ "train_runtime": 146079.3121,
+ "train_tokens_per_second": 5652.88
+ },
+ {
+ "epoch": 1.271545919741449,
+ "grad_norm": 0.7151332172339365,
+ "learning_rate": 2.9318498140453134e-06,
+ "loss": 0.32280969619750977,
+ "num_input_tokens_seen": 825987712,
+ "step": 18885,
+ "train_runtime": 146122.0243,
+ "train_tokens_per_second": 5652.726
+ },
+ {
+ "epoch": 1.2718825747374092,
+ "grad_norm": 0.8504766443077242,
+ "learning_rate": 2.929442811818527e-06,
+ "loss": 0.3143021106719971,
+ "num_input_tokens_seen": 826196048,
+ "step": 18890,
+ "train_runtime": 146161.5263,
+ "train_tokens_per_second": 5652.623
+ },
+ {
+ "epoch": 1.2722192297333692,
+ "grad_norm": 0.797394168026311,
+ "learning_rate": 2.927036388616457e-06,
+ "loss": 0.29908573627471924,
+ "num_input_tokens_seen": 826399384,
+ "step": 18895,
+ "train_runtime": 146195.9133,
+ "train_tokens_per_second": 5652.685
+ },
+ {
+ "epoch": 1.2725558847293295,
+ "grad_norm": 0.8320805975958718,
+ "learning_rate": 2.9246305451120494e-06,
+ "loss": 0.3228297233581543,
+ "num_input_tokens_seen": 826603896,
+ "step": 18900,
+ "train_runtime": 146234.1471,
+ "train_tokens_per_second": 5652.605
+ },
+ {
+ "epoch": 1.2728925397252895,
+ "grad_norm": 0.7307797880302476,
+ "learning_rate": 2.922225281978095e-06,
+ "loss": 0.3007101058959961,
+ "num_input_tokens_seen": 826821328,
+ "step": 18905,
+ "train_runtime": 146279.6181,
+ "train_tokens_per_second": 5652.334
+ },
+ {
+ "epoch": 1.2732291947212497,
+ "grad_norm": 0.7406348258598292,
+ "learning_rate": 2.919820599887212e-06,
+ "loss": 0.3024411916732788,
+ "num_input_tokens_seen": 827039792,
+ "step": 18910,
+ "train_runtime": 146321.5475,
+ "train_tokens_per_second": 5652.208
+ },
+ {
+ "epoch": 1.2735658497172098,
+ "grad_norm": 0.8447453780821402,
+ "learning_rate": 2.917416499511866e-06,
+ "loss": 0.30447540283203123,
+ "num_input_tokens_seen": 827242576,
+ "step": 18915,
+ "train_runtime": 146363.2378,
+ "train_tokens_per_second": 5651.983
+ },
+ {
+ "epoch": 1.27390250471317,
+ "grad_norm": 0.7936589269610094,
+ "learning_rate": 2.9150129815243553e-06,
+ "loss": 0.288972806930542,
+ "num_input_tokens_seen": 827457896,
+ "step": 18920,
+ "train_runtime": 146404.1905,
+ "train_tokens_per_second": 5651.873
+ },
+ {
+ "epoch": 1.27423915970913,
+ "grad_norm": 0.850907349149432,
+ "learning_rate": 2.912610046596816e-06,
+ "loss": 0.2999333620071411,
+ "num_input_tokens_seen": 827666848,
+ "step": 18925,
+ "train_runtime": 146446.094,
+ "train_tokens_per_second": 5651.683
+ },
+ {
+ "epoch": 1.2745758147050903,
+ "grad_norm": 0.6871438604531044,
+ "learning_rate": 2.91020769540122e-06,
+ "loss": 0.3152765274047852,
+ "num_input_tokens_seen": 827866000,
+ "step": 18930,
+ "train_runtime": 146485.6336,
+ "train_tokens_per_second": 5651.517
+ },
+ {
+ "epoch": 1.2749124697010503,
+ "grad_norm": 0.679666459947843,
+ "learning_rate": 2.9078059286093795e-06,
+ "loss": 0.2963433027267456,
+ "num_input_tokens_seen": 828072488,
+ "step": 18935,
+ "train_runtime": 146524.0873,
+ "train_tokens_per_second": 5651.443
+ },
+ {
+ "epoch": 1.2752491246970106,
+ "grad_norm": 0.6861610947075387,
+ "learning_rate": 2.905404746892937e-06,
+ "loss": 0.28579273223876955,
+ "num_input_tokens_seen": 828287296,
+ "step": 18940,
+ "train_runtime": 146560.9928,
+ "train_tokens_per_second": 5651.485
+ },
+ {
+ "epoch": 1.2755857796929706,
+ "grad_norm": 0.7154785961890833,
+ "learning_rate": 2.903004150923382e-06,
+ "loss": 0.3039486169815063,
+ "num_input_tokens_seen": 828510696,
+ "step": 18945,
+ "train_runtime": 146598.3451,
+ "train_tokens_per_second": 5651.569
+ },
+ {
+ "epoch": 1.2759224346889309,
+ "grad_norm": 0.7669932436785718,
+ "learning_rate": 2.9006041413720252e-06,
+ "loss": 0.3282867431640625,
+ "num_input_tokens_seen": 828734304,
+ "step": 18950,
+ "train_runtime": 146645.3601,
+ "train_tokens_per_second": 5651.282
+ },
+ {
+ "epoch": 1.2762590896848909,
+ "grad_norm": 0.6719242223720826,
+ "learning_rate": 2.8982047189100276e-06,
+ "loss": 0.30024127960205077,
+ "num_input_tokens_seen": 828967864,
+ "step": 18955,
+ "train_runtime": 146680.0954,
+ "train_tokens_per_second": 5651.536
+ },
+ {
+ "epoch": 1.2765957446808511,
+ "grad_norm": 0.7852643316945973,
+ "learning_rate": 2.895805884208378e-06,
+ "loss": 0.3082295894622803,
+ "num_input_tokens_seen": 829167224,
+ "step": 18960,
+ "train_runtime": 146717.6947,
+ "train_tokens_per_second": 5651.447
+ },
+ {
+ "epoch": 1.2769323996768112,
+ "grad_norm": 0.8516866669455334,
+ "learning_rate": 2.8934076379379016e-06,
+ "loss": 0.31248860359191893,
+ "num_input_tokens_seen": 829396808,
+ "step": 18965,
+ "train_runtime": 146755.3528,
+ "train_tokens_per_second": 5651.561
+ },
+ {
+ "epoch": 1.2772690546727714,
+ "grad_norm": 0.7752902430798366,
+ "learning_rate": 2.891009980769266e-06,
+ "loss": 0.3060382127761841,
+ "num_input_tokens_seen": 829619304,
+ "step": 18970,
+ "train_runtime": 146798.1501,
+ "train_tokens_per_second": 5651.429
+ },
+ {
+ "epoch": 1.2776057096687314,
+ "grad_norm": 0.8744465210673298,
+ "learning_rate": 2.88861291337296e-06,
+ "loss": 0.29961209297180175,
+ "num_input_tokens_seen": 829837784,
+ "step": 18975,
+ "train_runtime": 146836.9106,
+ "train_tokens_per_second": 5651.425
+ },
+ {
+ "epoch": 1.2779423646646917,
+ "grad_norm": 0.7351797146412703,
+ "learning_rate": 2.8862164364193236e-06,
+ "loss": 0.2904377460479736,
+ "num_input_tokens_seen": 830058208,
+ "step": 18980,
+ "train_runtime": 146873.1011,
+ "train_tokens_per_second": 5651.533
+ },
+ {
+ "epoch": 1.2782790196606517,
+ "grad_norm": 0.7312431717294958,
+ "learning_rate": 2.8838205505785168e-06,
+ "loss": 0.3134331703186035,
+ "num_input_tokens_seen": 830283008,
+ "step": 18985,
+ "train_runtime": 146914.2968,
+ "train_tokens_per_second": 5651.479
+ },
+ {
+ "epoch": 1.278615674656612,
+ "grad_norm": 0.663622797522999,
+ "learning_rate": 2.881425256520547e-06,
+ "loss": 0.2921187400817871,
+ "num_input_tokens_seen": 830500192,
+ "step": 18990,
+ "train_runtime": 146951.0522,
+ "train_tokens_per_second": 5651.543
+ },
+ {
+ "epoch": 1.278952329652572,
+ "grad_norm": 0.7237194461676807,
+ "learning_rate": 2.8790305549152487e-06,
+ "loss": 0.26155910491943357,
+ "num_input_tokens_seen": 830725544,
+ "step": 18995,
+ "train_runtime": 146990.9253,
+ "train_tokens_per_second": 5651.543
+ },
+ {
+ "epoch": 1.2792889846485322,
+ "grad_norm": 0.7631364393415948,
+ "learning_rate": 2.876636446432292e-06,
+ "loss": 0.3240873098373413,
+ "num_input_tokens_seen": 830948824,
+ "step": 19000,
+ "train_runtime": 147025.5963,
+ "train_tokens_per_second": 5651.729
+ },
+ {
+ "epoch": 1.2796256396444923,
+ "grad_norm": 0.7251110163689186,
+ "learning_rate": 2.8742429317411826e-06,
+ "loss": 0.30750136375427245,
+ "num_input_tokens_seen": 831173768,
+ "step": 19005,
+ "train_runtime": 147062.3862,
+ "train_tokens_per_second": 5651.845
+ },
+ {
+ "epoch": 1.2799622946404525,
+ "grad_norm": 0.9679509381325586,
+ "learning_rate": 2.8718500115112585e-06,
+ "loss": 0.30974340438842773,
+ "num_input_tokens_seen": 831380944,
+ "step": 19010,
+ "train_runtime": 147096.7442,
+ "train_tokens_per_second": 5651.933
+ },
+ {
+ "epoch": 1.2802989496364126,
+ "grad_norm": 0.7640168557000633,
+ "learning_rate": 2.869457686411691e-06,
+ "loss": 0.33095221519470214,
+ "num_input_tokens_seen": 831589040,
+ "step": 19015,
+ "train_runtime": 147137.7505,
+ "train_tokens_per_second": 5651.772
+ },
+ {
+ "epoch": 1.2806356046323728,
+ "grad_norm": 0.7459470398332059,
+ "learning_rate": 2.8670659571114912e-06,
+ "loss": 0.30541186332702636,
+ "num_input_tokens_seen": 831816448,
+ "step": 19020,
+ "train_runtime": 147176.0251,
+ "train_tokens_per_second": 5651.847
+ },
+ {
+ "epoch": 1.2809722596283328,
+ "grad_norm": 0.6423476554873053,
+ "learning_rate": 2.864674824279491e-06,
+ "loss": 0.2939921855926514,
+ "num_input_tokens_seen": 832035736,
+ "step": 19025,
+ "train_runtime": 147213.464,
+ "train_tokens_per_second": 5651.9
+ },
+ {
+ "epoch": 1.281308914624293,
+ "grad_norm": 0.7355078108437945,
+ "learning_rate": 2.862284288584371e-06,
+ "loss": 0.2885020971298218,
+ "num_input_tokens_seen": 832264232,
+ "step": 19030,
+ "train_runtime": 147251.1103,
+ "train_tokens_per_second": 5652.006
+ },
+ {
+ "epoch": 1.2816455696202531,
+ "grad_norm": 0.8373034806587798,
+ "learning_rate": 2.8598943506946297e-06,
+ "loss": 0.300329327583313,
+ "num_input_tokens_seen": 832482744,
+ "step": 19035,
+ "train_runtime": 147288.7273,
+ "train_tokens_per_second": 5652.047
+ },
+ {
+ "epoch": 1.2819822246162134,
+ "grad_norm": 0.7138141742816384,
+ "learning_rate": 2.85750501127861e-06,
+ "loss": 0.277694034576416,
+ "num_input_tokens_seen": 832697464,
+ "step": 19040,
+ "train_runtime": 147329.6861,
+ "train_tokens_per_second": 5651.933
+ },
+ {
+ "epoch": 1.2823188796121734,
+ "grad_norm": 0.6608599560904919,
+ "learning_rate": 2.8551162710044827e-06,
+ "loss": 0.30247793197631834,
+ "num_input_tokens_seen": 832912640,
+ "step": 19045,
+ "train_runtime": 147365.9859,
+ "train_tokens_per_second": 5652.001
+ },
+ {
+ "epoch": 1.2826555346081336,
+ "grad_norm": 0.7589903744520924,
+ "learning_rate": 2.8527281305402503e-06,
+ "loss": 0.32385740280151365,
+ "num_input_tokens_seen": 833133224,
+ "step": 19050,
+ "train_runtime": 147403.1332,
+ "train_tokens_per_second": 5652.073
+ },
+ {
+ "epoch": 1.2829921896040937,
+ "grad_norm": 0.8057473648914902,
+ "learning_rate": 2.8503405905537496e-06,
+ "loss": 0.29269657135009763,
+ "num_input_tokens_seen": 833346936,
+ "step": 19055,
+ "train_runtime": 147442.1312,
+ "train_tokens_per_second": 5652.027
+ },
+ {
+ "epoch": 1.283328844600054,
+ "grad_norm": 0.7212257083936001,
+ "learning_rate": 2.847953651712647e-06,
+ "loss": 0.3061253309249878,
+ "num_input_tokens_seen": 833581464,
+ "step": 19060,
+ "train_runtime": 147477.4905,
+ "train_tokens_per_second": 5652.262
+ },
+ {
+ "epoch": 1.283665499596014,
+ "grad_norm": 0.8465650818228247,
+ "learning_rate": 2.8455673146844426e-06,
+ "loss": 0.30112695693969727,
+ "num_input_tokens_seen": 833784496,
+ "step": 19065,
+ "train_runtime": 147521.4424,
+ "train_tokens_per_second": 5651.955
+ },
+ {
+ "epoch": 1.2840021545919742,
+ "grad_norm": 0.7588100951853031,
+ "learning_rate": 2.8431815801364717e-06,
+ "loss": 0.3190593719482422,
+ "num_input_tokens_seen": 834001896,
+ "step": 19070,
+ "train_runtime": 147563.6694,
+ "train_tokens_per_second": 5651.811
+ },
+ {
+ "epoch": 1.2843388095879342,
+ "grad_norm": 0.8096844572768415,
+ "learning_rate": 2.8407964487358906e-06,
+ "loss": 0.3083981513977051,
+ "num_input_tokens_seen": 834226528,
+ "step": 19075,
+ "train_runtime": 147606.1012,
+ "train_tokens_per_second": 5651.708
+ },
+ {
+ "epoch": 1.2846754645838945,
+ "grad_norm": 0.8121583620232876,
+ "learning_rate": 2.838411921149702e-06,
+ "loss": 0.30108425617218015,
+ "num_input_tokens_seen": 834428112,
+ "step": 19080,
+ "train_runtime": 147643.2603,
+ "train_tokens_per_second": 5651.651
+ },
+ {
+ "epoch": 1.2850121195798545,
+ "grad_norm": 0.7654588268965005,
+ "learning_rate": 2.836027998044722e-06,
+ "loss": 0.3017698287963867,
+ "num_input_tokens_seen": 834638200,
+ "step": 19085,
+ "train_runtime": 147687.2931,
+ "train_tokens_per_second": 5651.388
+ },
+ {
+ "epoch": 1.2853487745758148,
+ "grad_norm": 0.7769786248168525,
+ "learning_rate": 2.8336446800876143e-06,
+ "loss": 0.3067366600036621,
+ "num_input_tokens_seen": 834833144,
+ "step": 19090,
+ "train_runtime": 147722.5693,
+ "train_tokens_per_second": 5651.358
+ },
+ {
+ "epoch": 1.2856854295717748,
+ "grad_norm": 0.8520512402767025,
+ "learning_rate": 2.8312619679448637e-06,
+ "loss": 0.2865460872650146,
+ "num_input_tokens_seen": 835055128,
+ "step": 19095,
+ "train_runtime": 147762.1093,
+ "train_tokens_per_second": 5651.348
+ },
+ {
+ "epoch": 1.286022084567735,
+ "grad_norm": 0.8168300340689107,
+ "learning_rate": 2.8288798622827884e-06,
+ "loss": 0.31798462867736815,
+ "num_input_tokens_seen": 835252248,
+ "step": 19100,
+ "train_runtime": 147802.5202,
+ "train_tokens_per_second": 5651.137
+ },
+ {
+ "epoch": 1.286358739563695,
+ "grad_norm": 0.9604203872867969,
+ "learning_rate": 2.826498363767536e-06,
+ "loss": 0.316135311126709,
+ "num_input_tokens_seen": 835451680,
+ "step": 19105,
+ "train_runtime": 147840.3751,
+ "train_tokens_per_second": 5651.039
+ },
+ {
+ "epoch": 1.2866953945596553,
+ "grad_norm": 0.8232657359064803,
+ "learning_rate": 2.8241174730650855e-06,
+ "loss": 0.3065540313720703,
+ "num_input_tokens_seen": 835675080,
+ "step": 19110,
+ "train_runtime": 147880.4439,
+ "train_tokens_per_second": 5651.018
+ },
+ {
+ "epoch": 1.2870320495556153,
+ "grad_norm": 0.7796058041196822,
+ "learning_rate": 2.821737190841244e-06,
+ "loss": 0.29743661880493166,
+ "num_input_tokens_seen": 835891040,
+ "step": 19115,
+ "train_runtime": 147915.5543,
+ "train_tokens_per_second": 5651.137
+ },
+ {
+ "epoch": 1.2873687045515756,
+ "grad_norm": 0.7312714159541509,
+ "learning_rate": 2.8193575177616537e-06,
+ "loss": 0.2965965986251831,
+ "num_input_tokens_seen": 836111592,
+ "step": 19120,
+ "train_runtime": 147959.8125,
+ "train_tokens_per_second": 5650.937
+ },
+ {
+ "epoch": 1.2877053595475356,
+ "grad_norm": 0.7000449200994382,
+ "learning_rate": 2.816978454491777e-06,
+ "loss": 0.2831472396850586,
+ "num_input_tokens_seen": 836342224,
+ "step": 19125,
+ "train_runtime": 147994.2807,
+ "train_tokens_per_second": 5651.179
+ },
+ {
+ "epoch": 1.2880420145434959,
+ "grad_norm": 0.8484575372461595,
+ "learning_rate": 2.814600001696919e-06,
+ "loss": 0.31551055908203124,
+ "num_input_tokens_seen": 836555312,
+ "step": 19130,
+ "train_runtime": 148030.887,
+ "train_tokens_per_second": 5651.221
+ },
+ {
+ "epoch": 1.288378669539456,
+ "grad_norm": 0.7939086544066017,
+ "learning_rate": 2.812222160042196e-06,
+ "loss": 0.3081847906112671,
+ "num_input_tokens_seen": 836781384,
+ "step": 19135,
+ "train_runtime": 148071.8667,
+ "train_tokens_per_second": 5651.184
+ },
+ {
+ "epoch": 1.2887153245354162,
+ "grad_norm": 0.7293883099186741,
+ "learning_rate": 2.809844930192573e-06,
+ "loss": 0.31632413864135744,
+ "num_input_tokens_seen": 836996448,
+ "step": 19140,
+ "train_runtime": 148110.3971,
+ "train_tokens_per_second": 5651.166
+ },
+ {
+ "epoch": 1.2890519795313762,
+ "grad_norm": 0.7845647411236363,
+ "learning_rate": 2.807468312812831e-06,
+ "loss": 0.3241159677505493,
+ "num_input_tokens_seen": 837211800,
+ "step": 19145,
+ "train_runtime": 148145.8757,
+ "train_tokens_per_second": 5651.266
+ },
+ {
+ "epoch": 1.2893886345273364,
+ "grad_norm": 0.7305402054857416,
+ "learning_rate": 2.805092308567584e-06,
+ "loss": 0.2902689933776855,
+ "num_input_tokens_seen": 837439120,
+ "step": 19150,
+ "train_runtime": 148177.9274,
+ "train_tokens_per_second": 5651.578
+ },
+ {
+ "epoch": 1.2897252895232965,
+ "grad_norm": 0.706253740316737,
+ "learning_rate": 2.802716918121273e-06,
+ "loss": 0.2991164207458496,
+ "num_input_tokens_seen": 837665040,
+ "step": 19155,
+ "train_runtime": 148222.7975,
+ "train_tokens_per_second": 5651.391
+ },
+ {
+ "epoch": 1.2900619445192567,
+ "grad_norm": 0.7258823029517253,
+ "learning_rate": 2.8003421421381704e-06,
+ "loss": 0.29270741939544676,
+ "num_input_tokens_seen": 837893216,
+ "step": 19160,
+ "train_runtime": 148255.3261,
+ "train_tokens_per_second": 5651.69
+ },
+ {
+ "epoch": 1.2903985995152167,
+ "grad_norm": 0.7437197036098059,
+ "learning_rate": 2.797967981282371e-06,
+ "loss": 0.3194019317626953,
+ "num_input_tokens_seen": 838115864,
+ "step": 19165,
+ "train_runtime": 148297.7873,
+ "train_tokens_per_second": 5651.574
+ },
+ {
+ "epoch": 1.290735254511177,
+ "grad_norm": 0.7205908995167096,
+ "learning_rate": 2.795594436217808e-06,
+ "loss": 0.29573659896850585,
+ "num_input_tokens_seen": 838332696,
+ "step": 19170,
+ "train_runtime": 148338.6228,
+ "train_tokens_per_second": 5651.48
+ },
+ {
+ "epoch": 1.291071909507137,
+ "grad_norm": 0.6798097348635077,
+ "learning_rate": 2.7932215076082277e-06,
+ "loss": 0.29705004692077636,
+ "num_input_tokens_seen": 838541408,
+ "step": 19175,
+ "train_runtime": 148382.7784,
+ "train_tokens_per_second": 5651.204
+ },
+ {
+ "epoch": 1.2914085645030973,
+ "grad_norm": 0.7142148877160069,
+ "learning_rate": 2.790849196117219e-06,
+ "loss": 0.30868139266967776,
+ "num_input_tokens_seen": 838777616,
+ "step": 19180,
+ "train_runtime": 148425.245,
+ "train_tokens_per_second": 5651.179
+ },
+ {
+ "epoch": 1.2917452194990573,
+ "grad_norm": 0.7147263101151003,
+ "learning_rate": 2.7884775024081856e-06,
+ "loss": 0.30340185165405276,
+ "num_input_tokens_seen": 838994528,
+ "step": 19185,
+ "train_runtime": 148461.9522,
+ "train_tokens_per_second": 5651.243
+ },
+ {
+ "epoch": 1.2920818744950175,
+ "grad_norm": 0.7411183792886813,
+ "learning_rate": 2.786106427144367e-06,
+ "loss": 0.292337703704834,
+ "num_input_tokens_seen": 839211816,
+ "step": 19190,
+ "train_runtime": 148497.1246,
+ "train_tokens_per_second": 5651.367
+ },
+ {
+ "epoch": 1.2924185294909776,
+ "grad_norm": 0.784444211412864,
+ "learning_rate": 2.783735970988828e-06,
+ "loss": 0.27770419120788575,
+ "num_input_tokens_seen": 839425992,
+ "step": 19195,
+ "train_runtime": 148536.5284,
+ "train_tokens_per_second": 5651.31
+ },
+ {
+ "epoch": 1.2927551844869378,
+ "grad_norm": 0.7627571829537649,
+ "learning_rate": 2.7813661346044566e-06,
+ "loss": 0.31344926357269287,
+ "num_input_tokens_seen": 839639736,
+ "step": 19200,
+ "train_runtime": 148570.4019,
+ "train_tokens_per_second": 5651.46
+ },
+ {
+ "epoch": 1.2930918394828979,
+ "grad_norm": 0.8342081884001865,
+ "learning_rate": 2.7789969186539713e-06,
+ "loss": 0.30696806907653806,
+ "num_input_tokens_seen": 839843624,
+ "step": 19205,
+ "train_runtime": 148611.1575,
+ "train_tokens_per_second": 5651.282
+ },
+ {
+ "epoch": 1.293428494478858,
+ "grad_norm": 0.7673215752721638,
+ "learning_rate": 2.776628323799915e-06,
+ "loss": 0.32203989028930663,
+ "num_input_tokens_seen": 840061080,
+ "step": 19210,
+ "train_runtime": 148644.4524,
+ "train_tokens_per_second": 5651.48
+ },
+ {
+ "epoch": 1.2937651494748181,
+ "grad_norm": 0.8083666833563818,
+ "learning_rate": 2.774260350704657e-06,
+ "loss": 0.3181305408477783,
+ "num_input_tokens_seen": 840276920,
+ "step": 19215,
+ "train_runtime": 148679.4894,
+ "train_tokens_per_second": 5651.599
+ },
+ {
+ "epoch": 1.2941018044707784,
+ "grad_norm": 0.7678745273221902,
+ "learning_rate": 2.771893000030398e-06,
+ "loss": 0.3186459541320801,
+ "num_input_tokens_seen": 840489952,
+ "step": 19220,
+ "train_runtime": 148716.8291,
+ "train_tokens_per_second": 5651.613
+ },
+ {
+ "epoch": 1.2944384594667384,
+ "grad_norm": 0.7272017085399577,
+ "learning_rate": 2.7695262724391526e-06,
+ "loss": 0.29137458801269533,
+ "num_input_tokens_seen": 840702144,
+ "step": 19225,
+ "train_runtime": 148759.5159,
+ "train_tokens_per_second": 5651.418
+ },
+ {
+ "epoch": 1.2947751144626987,
+ "grad_norm": 0.7630834988917379,
+ "learning_rate": 2.767160168592777e-06,
+ "loss": 0.33965349197387695,
+ "num_input_tokens_seen": 840931688,
+ "step": 19230,
+ "train_runtime": 148806.4026,
+ "train_tokens_per_second": 5651.179
+ },
+ {
+ "epoch": 1.2951117694586587,
+ "grad_norm": 0.7684748875099028,
+ "learning_rate": 2.7647946891529355e-06,
+ "loss": 0.3225222110748291,
+ "num_input_tokens_seen": 841129968,
+ "step": 19235,
+ "train_runtime": 148842.7883,
+ "train_tokens_per_second": 5651.13
+ },
+ {
+ "epoch": 1.295448424454619,
+ "grad_norm": 0.728042153912699,
+ "learning_rate": 2.7624298347811342e-06,
+ "loss": 0.3056331157684326,
+ "num_input_tokens_seen": 841331904,
+ "step": 19240,
+ "train_runtime": 148884.4042,
+ "train_tokens_per_second": 5650.907
+ },
+ {
+ "epoch": 1.295785079450579,
+ "grad_norm": 0.7854514457019474,
+ "learning_rate": 2.7600656061386945e-06,
+ "loss": 0.31122641563415526,
+ "num_input_tokens_seen": 841562064,
+ "step": 19245,
+ "train_runtime": 148919.1545,
+ "train_tokens_per_second": 5651.134
+ },
+ {
+ "epoch": 1.2961217344465392,
+ "grad_norm": 0.7775133786016957,
+ "learning_rate": 2.757702003886765e-06,
+ "loss": 0.2915956974029541,
+ "num_input_tokens_seen": 841780328,
+ "step": 19250,
+ "train_runtime": 148953.299,
+ "train_tokens_per_second": 5651.304
+ },
+ {
+ "epoch": 1.2964583894424992,
+ "grad_norm": 1.058051599935181,
+ "learning_rate": 2.7553390286863205e-06,
+ "loss": 0.29277448654174804,
+ "num_input_tokens_seen": 841982320,
+ "step": 19255,
+ "train_runtime": 148990.52,
+ "train_tokens_per_second": 5651.248
+ },
+ {
+ "epoch": 1.2967950444384595,
+ "grad_norm": 0.7544624265455544,
+ "learning_rate": 2.752976681198159e-06,
+ "loss": 0.3195603609085083,
+ "num_input_tokens_seen": 842210920,
+ "step": 19260,
+ "train_runtime": 149028.3612,
+ "train_tokens_per_second": 5651.347
+ },
+ {
+ "epoch": 1.2971316994344195,
+ "grad_norm": 0.8113254179825854,
+ "learning_rate": 2.750614962082901e-06,
+ "loss": 0.28080661296844484,
+ "num_input_tokens_seen": 842433856,
+ "step": 19265,
+ "train_runtime": 149066.4868,
+ "train_tokens_per_second": 5651.397
+ },
+ {
+ "epoch": 1.2974683544303798,
+ "grad_norm": 0.7219059530126368,
+ "learning_rate": 2.748253872001e-06,
+ "loss": 0.3025611162185669,
+ "num_input_tokens_seen": 842651984,
+ "step": 19270,
+ "train_runtime": 149103.7773,
+ "train_tokens_per_second": 5651.446
+ },
+ {
+ "epoch": 1.2978050094263398,
+ "grad_norm": 0.7933392458748796,
+ "learning_rate": 2.74589341161272e-06,
+ "loss": 0.30447611808776853,
+ "num_input_tokens_seen": 842849968,
+ "step": 19275,
+ "train_runtime": 149143.5367,
+ "train_tokens_per_second": 5651.267
+ },
+ {
+ "epoch": 1.2981416644223,
+ "grad_norm": 0.8775197487075084,
+ "learning_rate": 2.743533581578163e-06,
+ "loss": 0.31205015182495116,
+ "num_input_tokens_seen": 843059320,
+ "step": 19280,
+ "train_runtime": 149180.2644,
+ "train_tokens_per_second": 5651.279
+ },
+ {
+ "epoch": 1.29847831941826,
+ "grad_norm": 0.7923357381559762,
+ "learning_rate": 2.7411743825572407e-06,
+ "loss": 0.3455479145050049,
+ "num_input_tokens_seen": 843291856,
+ "step": 19285,
+ "train_runtime": 149215.3246,
+ "train_tokens_per_second": 5651.51
+ },
+ {
+ "epoch": 1.2988149744142203,
+ "grad_norm": 0.7093696654952529,
+ "learning_rate": 2.7388158152097015e-06,
+ "loss": 0.30530009269714353,
+ "num_input_tokens_seen": 843521128,
+ "step": 19290,
+ "train_runtime": 149245.9801,
+ "train_tokens_per_second": 5651.885
+ },
+ {
+ "epoch": 1.2991516294101804,
+ "grad_norm": 0.7343455792050803,
+ "learning_rate": 2.736457880195109e-06,
+ "loss": 0.3188295841217041,
+ "num_input_tokens_seen": 843733872,
+ "step": 19295,
+ "train_runtime": 149283.513,
+ "train_tokens_per_second": 5651.889
+ },
+ {
+ "epoch": 1.2994882844061406,
+ "grad_norm": 0.7396876861043844,
+ "learning_rate": 2.7341005781728523e-06,
+ "loss": 0.3281911849975586,
+ "num_input_tokens_seen": 843964944,
+ "step": 19300,
+ "train_runtime": 149329.5165,
+ "train_tokens_per_second": 5651.695
+ },
+ {
+ "epoch": 1.2998249394021006,
+ "grad_norm": 0.752304677269571,
+ "learning_rate": 2.7317439098021444e-06,
+ "loss": 0.3072381496429443,
+ "num_input_tokens_seen": 844183152,
+ "step": 19305,
+ "train_runtime": 149373.377,
+ "train_tokens_per_second": 5651.497
+ },
+ {
+ "epoch": 1.300161594398061,
+ "grad_norm": 0.7866331506852371,
+ "learning_rate": 2.7293878757420185e-06,
+ "loss": 0.32221331596374514,
+ "num_input_tokens_seen": 844400152,
+ "step": 19310,
+ "train_runtime": 149415.5476,
+ "train_tokens_per_second": 5651.354
+ },
+ {
+ "epoch": 1.3004982493940211,
+ "grad_norm": 0.7241103171317756,
+ "learning_rate": 2.7270324766513306e-06,
+ "loss": 0.3104707717895508,
+ "num_input_tokens_seen": 844631680,
+ "step": 19315,
+ "train_runtime": 149454.6137,
+ "train_tokens_per_second": 5651.426
+ },
+ {
+ "epoch": 1.3008349043899812,
+ "grad_norm": 0.7716528370818122,
+ "learning_rate": 2.724677713188767e-06,
+ "loss": 0.3056204795837402,
+ "num_input_tokens_seen": 844836448,
+ "step": 19320,
+ "train_runtime": 149496.2843,
+ "train_tokens_per_second": 5651.22
+ },
+ {
+ "epoch": 1.3011715593859412,
+ "grad_norm": 0.7945005896939524,
+ "learning_rate": 2.722323586012821e-06,
+ "loss": 0.313628101348877,
+ "num_input_tokens_seen": 845053136,
+ "step": 19325,
+ "train_runtime": 149533.6826,
+ "train_tokens_per_second": 5651.256
+ },
+ {
+ "epoch": 1.3015082143819015,
+ "grad_norm": 0.7917610556705151,
+ "learning_rate": 2.7199700957818252e-06,
+ "loss": 0.3013652801513672,
+ "num_input_tokens_seen": 845276504,
+ "step": 19330,
+ "train_runtime": 149575.9646,
+ "train_tokens_per_second": 5651.152
+ },
+ {
+ "epoch": 1.3018448693778617,
+ "grad_norm": 0.8720139344385164,
+ "learning_rate": 2.717617243153918e-06,
+ "loss": 0.3019449710845947,
+ "num_input_tokens_seen": 845488208,
+ "step": 19335,
+ "train_runtime": 149622.3515,
+ "train_tokens_per_second": 5650.815
+ },
+ {
+ "epoch": 1.3021815243738217,
+ "grad_norm": 0.7397233088744621,
+ "learning_rate": 2.715265028787073e-06,
+ "loss": 0.31660406589508056,
+ "num_input_tokens_seen": 845695240,
+ "step": 19340,
+ "train_runtime": 149660.9196,
+ "train_tokens_per_second": 5650.742
+ },
+ {
+ "epoch": 1.3025181793697818,
+ "grad_norm": 0.8121569406562228,
+ "learning_rate": 2.7129134533390765e-06,
+ "loss": 0.33323040008544924,
+ "num_input_tokens_seen": 845925768,
+ "step": 19345,
+ "train_runtime": 149697.7788,
+ "train_tokens_per_second": 5650.891
+ },
+ {
+ "epoch": 1.302854834365742,
+ "grad_norm": 0.7696563717601685,
+ "learning_rate": 2.7105625174675404e-06,
+ "loss": 0.2952426433563232,
+ "num_input_tokens_seen": 846155368,
+ "step": 19350,
+ "train_runtime": 149736.5358,
+ "train_tokens_per_second": 5650.961
+ },
+ {
+ "epoch": 1.3031914893617023,
+ "grad_norm": 0.7453975160800163,
+ "learning_rate": 2.7082122218298957e-06,
+ "loss": 0.2969479560852051,
+ "num_input_tokens_seen": 846367080,
+ "step": 19355,
+ "train_runtime": 149773.8436,
+ "train_tokens_per_second": 5650.967
+ },
+ {
+ "epoch": 1.3035281443576623,
+ "grad_norm": 0.7913525098809976,
+ "learning_rate": 2.7058625670833947e-06,
+ "loss": 0.3046135902404785,
+ "num_input_tokens_seen": 846584096,
+ "step": 19360,
+ "train_runtime": 149810.105,
+ "train_tokens_per_second": 5651.048
+ },
+ {
+ "epoch": 1.3038647993536223,
+ "grad_norm": 0.7456322958807796,
+ "learning_rate": 2.70351355388511e-06,
+ "loss": 0.306339168548584,
+ "num_input_tokens_seen": 846800416,
+ "step": 19365,
+ "train_runtime": 149844.731,
+ "train_tokens_per_second": 5651.186
+ },
+ {
+ "epoch": 1.3042014543495826,
+ "grad_norm": 0.7705924730860086,
+ "learning_rate": 2.7011651828919407e-06,
+ "loss": 0.30881786346435547,
+ "num_input_tokens_seen": 847024880,
+ "step": 19370,
+ "train_runtime": 149886.5908,
+ "train_tokens_per_second": 5651.105
+ },
+ {
+ "epoch": 1.3045381093455428,
+ "grad_norm": 0.7634554372376511,
+ "learning_rate": 2.698817454760593e-06,
+ "loss": 0.305006217956543,
+ "num_input_tokens_seen": 847244184,
+ "step": 19375,
+ "train_runtime": 149919.4984,
+ "train_tokens_per_second": 5651.327
+ },
+ {
+ "epoch": 1.3048747643415028,
+ "grad_norm": 0.7420517594807788,
+ "learning_rate": 2.696470370147611e-06,
+ "loss": 0.3149790048599243,
+ "num_input_tokens_seen": 847461096,
+ "step": 19380,
+ "train_runtime": 149955.0762,
+ "train_tokens_per_second": 5651.433
+ },
+ {
+ "epoch": 1.3052114193374629,
+ "grad_norm": 0.7739344518736725,
+ "learning_rate": 2.69412392970934e-06,
+ "loss": 0.3325794219970703,
+ "num_input_tokens_seen": 847680648,
+ "step": 19385,
+ "train_runtime": 149993.1514,
+ "train_tokens_per_second": 5651.462
+ },
+ {
+ "epoch": 1.3055480743334231,
+ "grad_norm": 0.6530513267019178,
+ "learning_rate": 2.6917781341019615e-06,
+ "loss": 0.2991641044616699,
+ "num_input_tokens_seen": 847901040,
+ "step": 19390,
+ "train_runtime": 150031.3101,
+ "train_tokens_per_second": 5651.494
+ },
+ {
+ "epoch": 1.3058847293293834,
+ "grad_norm": 0.8035047874324559,
+ "learning_rate": 2.689432983981467e-06,
+ "loss": 0.2986927032470703,
+ "num_input_tokens_seen": 848128768,
+ "step": 19395,
+ "train_runtime": 150080.1123,
+ "train_tokens_per_second": 5651.174
+ },
+ {
+ "epoch": 1.3062213843253434,
+ "grad_norm": 0.7325503260453592,
+ "learning_rate": 2.6870884800036713e-06,
+ "loss": 0.3080791234970093,
+ "num_input_tokens_seen": 848340072,
+ "step": 19400,
+ "train_runtime": 150113.6518,
+ "train_tokens_per_second": 5651.319
+ },
+ {
+ "epoch": 1.3065580393213034,
+ "grad_norm": 0.948401336461252,
+ "learning_rate": 2.684744622824207e-06,
+ "loss": 0.32254228591918943,
+ "num_input_tokens_seen": 848548544,
+ "step": 19405,
+ "train_runtime": 150152.8491,
+ "train_tokens_per_second": 5651.232
+ },
+ {
+ "epoch": 1.3068946943172637,
+ "grad_norm": 0.7795604426950599,
+ "learning_rate": 2.6824014130985255e-06,
+ "loss": 0.31671667098999023,
+ "num_input_tokens_seen": 848765608,
+ "step": 19410,
+ "train_runtime": 150190.0674,
+ "train_tokens_per_second": 5651.277
+ },
+ {
+ "epoch": 1.307231349313224,
+ "grad_norm": 0.7275328657604889,
+ "learning_rate": 2.680058851481896e-06,
+ "loss": 0.2992823600769043,
+ "num_input_tokens_seen": 848984336,
+ "step": 19415,
+ "train_runtime": 150225.1677,
+ "train_tokens_per_second": 5651.412
+ },
+ {
+ "epoch": 1.307568004309184,
+ "grad_norm": 0.8231036715155717,
+ "learning_rate": 2.6777169386294154e-06,
+ "loss": 0.3124098777770996,
+ "num_input_tokens_seen": 849202320,
+ "step": 19420,
+ "train_runtime": 150263.4801,
+ "train_tokens_per_second": 5651.422
+ },
+ {
+ "epoch": 1.307904659305144,
+ "grad_norm": 0.784487047120117,
+ "learning_rate": 2.6753756751959846e-06,
+ "loss": 0.30565834045410156,
+ "num_input_tokens_seen": 849432016,
+ "step": 19425,
+ "train_runtime": 150304.88,
+ "train_tokens_per_second": 5651.393
+ },
+ {
+ "epoch": 1.3082413143011042,
+ "grad_norm": 0.7160689409622006,
+ "learning_rate": 2.6730350618363367e-06,
+ "loss": 0.3007196426391602,
+ "num_input_tokens_seen": 849652096,
+ "step": 19430,
+ "train_runtime": 150339.418,
+ "train_tokens_per_second": 5651.559
+ },
+ {
+ "epoch": 1.3085779692970645,
+ "grad_norm": 1.0693121673231818,
+ "learning_rate": 2.6706950992050097e-06,
+ "loss": 0.2976806640625,
+ "num_input_tokens_seen": 849873624,
+ "step": 19435,
+ "train_runtime": 150377.7065,
+ "train_tokens_per_second": 5651.593
+ },
+ {
+ "epoch": 1.3089146242930245,
+ "grad_norm": 0.7521055380512028,
+ "learning_rate": 2.6683557879563726e-06,
+ "loss": 0.3227806091308594,
+ "num_input_tokens_seen": 850094824,
+ "step": 19440,
+ "train_runtime": 150411.5882,
+ "train_tokens_per_second": 5651.791
+ },
+ {
+ "epoch": 1.3092512792889845,
+ "grad_norm": 0.7577610588393768,
+ "learning_rate": 2.6660171287446045e-06,
+ "loss": 0.33275818824768066,
+ "num_input_tokens_seen": 850318184,
+ "step": 19445,
+ "train_runtime": 150448.3342,
+ "train_tokens_per_second": 5651.895
+ },
+ {
+ "epoch": 1.3095879342849448,
+ "grad_norm": 0.7044234887744943,
+ "learning_rate": 2.6636791222237034e-06,
+ "loss": 0.3194304943084717,
+ "num_input_tokens_seen": 850537512,
+ "step": 19450,
+ "train_runtime": 150485.0966,
+ "train_tokens_per_second": 5651.972
+ },
+ {
+ "epoch": 1.309924589280905,
+ "grad_norm": 0.7351419738193509,
+ "learning_rate": 2.6613417690474864e-06,
+ "loss": 0.3084381580352783,
+ "num_input_tokens_seen": 850762520,
+ "step": 19455,
+ "train_runtime": 150516.0117,
+ "train_tokens_per_second": 5652.306
+ },
+ {
+ "epoch": 1.310261244276865,
+ "grad_norm": 0.6957251747847025,
+ "learning_rate": 2.6590050698695856e-06,
+ "loss": 0.30487637519836425,
+ "num_input_tokens_seen": 850980888,
+ "step": 19460,
+ "train_runtime": 150552.8661,
+ "train_tokens_per_second": 5652.373
+ },
+ {
+ "epoch": 1.310597899272825,
+ "grad_norm": 0.7124189387924086,
+ "learning_rate": 2.6566690253434508e-06,
+ "loss": 0.31617035865783694,
+ "num_input_tokens_seen": 851212384,
+ "step": 19465,
+ "train_runtime": 150591.0113,
+ "train_tokens_per_second": 5652.478
+ },
+ {
+ "epoch": 1.3109345542687854,
+ "grad_norm": 0.7885036457021949,
+ "learning_rate": 2.654333636122354e-06,
+ "loss": 0.29657797813415526,
+ "num_input_tokens_seen": 851438520,
+ "step": 19470,
+ "train_runtime": 150631.5054,
+ "train_tokens_per_second": 5652.46
+ },
+ {
+ "epoch": 1.3112712092647456,
+ "grad_norm": 0.7024735867575697,
+ "learning_rate": 2.6519989028593728e-06,
+ "loss": 0.3158660173416138,
+ "num_input_tokens_seen": 851666592,
+ "step": 19475,
+ "train_runtime": 150671.4222,
+ "train_tokens_per_second": 5652.476
+ },
+ {
+ "epoch": 1.3116078642607056,
+ "grad_norm": 0.8346584854469274,
+ "learning_rate": 2.649664826207414e-06,
+ "loss": 0.31950523853302004,
+ "num_input_tokens_seen": 851881728,
+ "step": 19480,
+ "train_runtime": 150710.0418,
+ "train_tokens_per_second": 5652.455
+ },
+ {
+ "epoch": 1.3119445192566657,
+ "grad_norm": 0.8221325321759024,
+ "learning_rate": 2.647331406819189e-06,
+ "loss": 0.3350534915924072,
+ "num_input_tokens_seen": 852092288,
+ "step": 19485,
+ "train_runtime": 150747.5697,
+ "train_tokens_per_second": 5652.445
+ },
+ {
+ "epoch": 1.312281174252626,
+ "grad_norm": 0.7024744149478919,
+ "learning_rate": 2.644998645347236e-06,
+ "loss": 0.29921770095825195,
+ "num_input_tokens_seen": 852306504,
+ "step": 19490,
+ "train_runtime": 150791.8993,
+ "train_tokens_per_second": 5652.204
+ },
+ {
+ "epoch": 1.3126178292485862,
+ "grad_norm": 0.7284467267787711,
+ "learning_rate": 2.6426665424439025e-06,
+ "loss": 0.316993260383606,
+ "num_input_tokens_seen": 852532944,
+ "step": 19495,
+ "train_runtime": 150836.5301,
+ "train_tokens_per_second": 5652.032
+ },
+ {
+ "epoch": 1.3129544842445462,
+ "grad_norm": 0.8271670284793109,
+ "learning_rate": 2.6403350987613526e-06,
+ "loss": 0.31356039047241213,
+ "num_input_tokens_seen": 852756160,
+ "step": 19500,
+ "train_runtime": 150879.3547,
+ "train_tokens_per_second": 5651.908
+ },
+ {
+ "epoch": 1.3132911392405062,
+ "grad_norm": 0.7670722817590104,
+ "learning_rate": 2.6380043149515676e-06,
+ "loss": 0.3076605796813965,
+ "num_input_tokens_seen": 852971624,
+ "step": 19505,
+ "train_runtime": 150913.3012,
+ "train_tokens_per_second": 5652.064
+ },
+ {
+ "epoch": 1.3136277942364665,
+ "grad_norm": 0.7538754898846585,
+ "learning_rate": 2.6356741916663453e-06,
+ "loss": 0.3078343868255615,
+ "num_input_tokens_seen": 853194624,
+ "step": 19510,
+ "train_runtime": 150952.8524,
+ "train_tokens_per_second": 5652.06
+ },
+ {
+ "epoch": 1.3139644492324267,
+ "grad_norm": 0.8299463216741857,
+ "learning_rate": 2.6333447295572933e-06,
+ "loss": 0.310660195350647,
+ "num_input_tokens_seen": 853414920,
+ "step": 19515,
+ "train_runtime": 150985.9986,
+ "train_tokens_per_second": 5652.279
+ },
+ {
+ "epoch": 1.3143011042283868,
+ "grad_norm": 0.7460970673432722,
+ "learning_rate": 2.631015929275845e-06,
+ "loss": 0.3059165239334106,
+ "num_input_tokens_seen": 853622712,
+ "step": 19520,
+ "train_runtime": 151026.6956,
+ "train_tokens_per_second": 5652.131
+ },
+ {
+ "epoch": 1.3146377592243468,
+ "grad_norm": 0.770739755897705,
+ "learning_rate": 2.628687791473234e-06,
+ "loss": 0.29071602821350095,
+ "num_input_tokens_seen": 853853296,
+ "step": 19525,
+ "train_runtime": 151057.0111,
+ "train_tokens_per_second": 5652.523
+ },
+ {
+ "epoch": 1.314974414220307,
+ "grad_norm": 0.8302523376916937,
+ "learning_rate": 2.6263603168005224e-06,
+ "loss": 0.2854055643081665,
+ "num_input_tokens_seen": 854070296,
+ "step": 19530,
+ "train_runtime": 151102.3064,
+ "train_tokens_per_second": 5652.265
+ },
+ {
+ "epoch": 1.3153110692162673,
+ "grad_norm": 0.6658939434607352,
+ "learning_rate": 2.6240335059085797e-06,
+ "loss": 0.30446481704711914,
+ "num_input_tokens_seen": 854298216,
+ "step": 19535,
+ "train_runtime": 151143.3911,
+ "train_tokens_per_second": 5652.237
+ },
+ {
+ "epoch": 1.3156477242122273,
+ "grad_norm": 0.7452286529155747,
+ "learning_rate": 2.62170735944809e-06,
+ "loss": 0.3120380163192749,
+ "num_input_tokens_seen": 854517360,
+ "step": 19540,
+ "train_runtime": 151180.803,
+ "train_tokens_per_second": 5652.287
+ },
+ {
+ "epoch": 1.3159843792081873,
+ "grad_norm": 0.7794163551487556,
+ "learning_rate": 2.6193818780695537e-06,
+ "loss": 0.298751163482666,
+ "num_input_tokens_seen": 854745560,
+ "step": 19545,
+ "train_runtime": 151214.1843,
+ "train_tokens_per_second": 5652.549
+ },
+ {
+ "epoch": 1.3163210342041476,
+ "grad_norm": 0.7606004553161916,
+ "learning_rate": 2.6170570624232837e-06,
+ "loss": 0.3143803119659424,
+ "num_input_tokens_seen": 854941592,
+ "step": 19550,
+ "train_runtime": 151257.5538,
+ "train_tokens_per_second": 5652.224
+ },
+ {
+ "epoch": 1.3166576892001078,
+ "grad_norm": 0.7540362996447706,
+ "learning_rate": 2.6147329131594078e-06,
+ "loss": 0.2849498510360718,
+ "num_input_tokens_seen": 855170576,
+ "step": 19555,
+ "train_runtime": 151298.4772,
+ "train_tokens_per_second": 5652.209
+ },
+ {
+ "epoch": 1.3169943441960679,
+ "grad_norm": 0.6379893471729663,
+ "learning_rate": 2.612409430927867e-06,
+ "loss": 0.2712059020996094,
+ "num_input_tokens_seen": 855390328,
+ "step": 19560,
+ "train_runtime": 151335.988,
+ "train_tokens_per_second": 5652.26
+ },
+ {
+ "epoch": 1.317330999192028,
+ "grad_norm": 1.705002769226515,
+ "learning_rate": 2.610086616378413e-06,
+ "loss": 0.29099502563476565,
+ "num_input_tokens_seen": 855599680,
+ "step": 19565,
+ "train_runtime": 151369.3355,
+ "train_tokens_per_second": 5652.398
+ },
+ {
+ "epoch": 1.3176676541879881,
+ "grad_norm": 0.6619066534695688,
+ "learning_rate": 2.60776447016062e-06,
+ "loss": 0.3147557497024536,
+ "num_input_tokens_seen": 855821216,
+ "step": 19570,
+ "train_runtime": 151406.8413,
+ "train_tokens_per_second": 5652.461
+ },
+ {
+ "epoch": 1.3180043091839484,
+ "grad_norm": 0.7847891473320998,
+ "learning_rate": 2.6054429929238607e-06,
+ "loss": 0.31384851932525637,
+ "num_input_tokens_seen": 856037064,
+ "step": 19575,
+ "train_runtime": 151440.0977,
+ "train_tokens_per_second": 5652.645
+ },
+ {
+ "epoch": 1.3183409641799084,
+ "grad_norm": 0.6462398801805597,
+ "learning_rate": 2.603122185317335e-06,
+ "loss": 0.29209444522857664,
+ "num_input_tokens_seen": 856240856,
+ "step": 19580,
+ "train_runtime": 151475.745,
+ "train_tokens_per_second": 5652.66
+ },
+ {
+ "epoch": 1.3186776191758685,
+ "grad_norm": 0.7747243284477985,
+ "learning_rate": 2.6008020479900465e-06,
+ "loss": 0.2708132743835449,
+ "num_input_tokens_seen": 856453280,
+ "step": 19585,
+ "train_runtime": 151515.0863,
+ "train_tokens_per_second": 5652.594
+ },
+ {
+ "epoch": 1.3190142741718287,
+ "grad_norm": 0.6709060817752821,
+ "learning_rate": 2.598482581590815e-06,
+ "loss": 0.30856051445007326,
+ "num_input_tokens_seen": 856685160,
+ "step": 19590,
+ "train_runtime": 151550.6826,
+ "train_tokens_per_second": 5652.796
+ },
+ {
+ "epoch": 1.319350929167789,
+ "grad_norm": 0.698540481486877,
+ "learning_rate": 2.5961637867682715e-06,
+ "loss": 0.280717658996582,
+ "num_input_tokens_seen": 856916440,
+ "step": 19595,
+ "train_runtime": 151591.3298,
+ "train_tokens_per_second": 5652.806
+ },
+ {
+ "epoch": 1.319687584163749,
+ "grad_norm": 0.6957573651733252,
+ "learning_rate": 2.5938456641708598e-06,
+ "loss": 0.28789021968841555,
+ "num_input_tokens_seen": 857134672,
+ "step": 19600,
+ "train_runtime": 151620.8055,
+ "train_tokens_per_second": 5653.147
+ },
+ {
+ "epoch": 1.320024239159709,
+ "grad_norm": 0.772599706020318,
+ "learning_rate": 2.591528214446836e-06,
+ "loss": 0.3085463523864746,
+ "num_input_tokens_seen": 857366576,
+ "step": 19605,
+ "train_runtime": 151656.731,
+ "train_tokens_per_second": 5653.337
+ },
+ {
+ "epoch": 1.3203608941556693,
+ "grad_norm": 0.7024918619799805,
+ "learning_rate": 2.5892114382442664e-06,
+ "loss": 0.3224550724029541,
+ "num_input_tokens_seen": 857583784,
+ "step": 19610,
+ "train_runtime": 151693.8363,
+ "train_tokens_per_second": 5653.386
+ },
+ {
+ "epoch": 1.3206975491516295,
+ "grad_norm": 0.632625020427823,
+ "learning_rate": 2.5868953362110287e-06,
+ "loss": 0.29941411018371583,
+ "num_input_tokens_seen": 857806240,
+ "step": 19615,
+ "train_runtime": 151729.7798,
+ "train_tokens_per_second": 5653.513
+ },
+ {
+ "epoch": 1.3210342041475895,
+ "grad_norm": 0.6986436139568517,
+ "learning_rate": 2.5845799089948196e-06,
+ "loss": 0.29595515727996824,
+ "num_input_tokens_seen": 858015600,
+ "step": 19620,
+ "train_runtime": 151767.663,
+ "train_tokens_per_second": 5653.481
+ },
+ {
+ "epoch": 1.3213708591435496,
+ "grad_norm": 0.7085148449641908,
+ "learning_rate": 2.5822651572431323e-06,
+ "loss": 0.30276174545288087,
+ "num_input_tokens_seen": 858241536,
+ "step": 19625,
+ "train_runtime": 151807.9703,
+ "train_tokens_per_second": 5653.468
+ },
+ {
+ "epoch": 1.3217075141395098,
+ "grad_norm": 0.7897947876767634,
+ "learning_rate": 2.5799510816032867e-06,
+ "loss": 0.3044028043746948,
+ "num_input_tokens_seen": 858431176,
+ "step": 19630,
+ "train_runtime": 151849.2654,
+ "train_tokens_per_second": 5653.18
+ },
+ {
+ "epoch": 1.32204416913547,
+ "grad_norm": 0.7421651364229699,
+ "learning_rate": 2.577637682722404e-06,
+ "loss": 0.31405105590820315,
+ "num_input_tokens_seen": 858649160,
+ "step": 19635,
+ "train_runtime": 151893.2046,
+ "train_tokens_per_second": 5652.979
+ },
+ {
+ "epoch": 1.32238082413143,
+ "grad_norm": 0.9243365365036531,
+ "learning_rate": 2.5753249612474195e-06,
+ "loss": 0.3185321807861328,
+ "num_input_tokens_seen": 858864688,
+ "step": 19640,
+ "train_runtime": 151935.4081,
+ "train_tokens_per_second": 5652.828
+ },
+ {
+ "epoch": 1.3227174791273901,
+ "grad_norm": 0.714378409765336,
+ "learning_rate": 2.5730129178250774e-06,
+ "loss": 0.3150054693222046,
+ "num_input_tokens_seen": 859090984,
+ "step": 19645,
+ "train_runtime": 151973.3637,
+ "train_tokens_per_second": 5652.905
+ },
+ {
+ "epoch": 1.3230541341233504,
+ "grad_norm": 0.7478713199114753,
+ "learning_rate": 2.570701553101933e-06,
+ "loss": 0.3087807893753052,
+ "num_input_tokens_seen": 859303816,
+ "step": 19650,
+ "train_runtime": 152016.815,
+ "train_tokens_per_second": 5652.689
+ },
+ {
+ "epoch": 1.3233907891193106,
+ "grad_norm": 0.7180167108657562,
+ "learning_rate": 2.568390867724354e-06,
+ "loss": 0.3278617858886719,
+ "num_input_tokens_seen": 859526840,
+ "step": 19655,
+ "train_runtime": 152059.4571,
+ "train_tokens_per_second": 5652.571
+ },
+ {
+ "epoch": 1.3237274441152707,
+ "grad_norm": 0.8472549748110341,
+ "learning_rate": 2.5660808623385143e-06,
+ "loss": 0.3053846836090088,
+ "num_input_tokens_seen": 859736800,
+ "step": 19660,
+ "train_runtime": 152097.7055,
+ "train_tokens_per_second": 5652.53
+ },
+ {
+ "epoch": 1.3240640991112307,
+ "grad_norm": 0.6854888856572094,
+ "learning_rate": 2.5637715375903984e-06,
+ "loss": 0.2824453830718994,
+ "num_input_tokens_seen": 859955360,
+ "step": 19665,
+ "train_runtime": 152139.8395,
+ "train_tokens_per_second": 5652.401
+ },
+ {
+ "epoch": 1.324400754107191,
+ "grad_norm": 0.8425691445759216,
+ "learning_rate": 2.5614628941258075e-06,
+ "loss": 0.3014561176300049,
+ "num_input_tokens_seen": 860163408,
+ "step": 19670,
+ "train_runtime": 152179.5969,
+ "train_tokens_per_second": 5652.291
+ },
+ {
+ "epoch": 1.3247374091031512,
+ "grad_norm": 0.785890365158059,
+ "learning_rate": 2.5591549325903388e-06,
+ "loss": 0.3199615478515625,
+ "num_input_tokens_seen": 860403016,
+ "step": 19675,
+ "train_runtime": 152220.8565,
+ "train_tokens_per_second": 5652.333
+ },
+ {
+ "epoch": 1.3250740640991112,
+ "grad_norm": 0.7049194900739367,
+ "learning_rate": 2.5568476536294105e-06,
+ "loss": 0.3173452615737915,
+ "num_input_tokens_seen": 860626712,
+ "step": 19680,
+ "train_runtime": 152253.5963,
+ "train_tokens_per_second": 5652.587
+ },
+ {
+ "epoch": 1.3254107190950712,
+ "grad_norm": 0.7458641790263951,
+ "learning_rate": 2.5545410578882466e-06,
+ "loss": 0.2950143814086914,
+ "num_input_tokens_seen": 860857936,
+ "step": 19685,
+ "train_runtime": 152289.1953,
+ "train_tokens_per_second": 5652.784
+ },
+ {
+ "epoch": 1.3257473740910315,
+ "grad_norm": 0.8429589187979019,
+ "learning_rate": 2.5522351460118776e-06,
+ "loss": 0.2903186321258545,
+ "num_input_tokens_seen": 861063800,
+ "step": 19690,
+ "train_runtime": 152319.4076,
+ "train_tokens_per_second": 5653.014
+ },
+ {
+ "epoch": 1.3260840290869917,
+ "grad_norm": 0.6773218912537953,
+ "learning_rate": 2.549929918645144e-06,
+ "loss": 0.31392974853515626,
+ "num_input_tokens_seen": 861297432,
+ "step": 19695,
+ "train_runtime": 152355.3898,
+ "train_tokens_per_second": 5653.213
+ },
+ {
+ "epoch": 1.3264206840829518,
+ "grad_norm": 0.7356765211271618,
+ "learning_rate": 2.5476253764326963e-06,
+ "loss": 0.2875689506530762,
+ "num_input_tokens_seen": 861503656,
+ "step": 19700,
+ "train_runtime": 152391.6663,
+ "train_tokens_per_second": 5653.22
+ },
+ {
+ "epoch": 1.3267573390789118,
+ "grad_norm": 0.7016223117638092,
+ "learning_rate": 2.5453215200189927e-06,
+ "loss": 0.2943393707275391,
+ "num_input_tokens_seen": 861717016,
+ "step": 19705,
+ "train_runtime": 152429.9059,
+ "train_tokens_per_second": 5653.202
+ },
+ {
+ "epoch": 1.327093994074872,
+ "grad_norm": 0.7845624304921898,
+ "learning_rate": 2.543018350048298e-06,
+ "loss": 0.29305644035339357,
+ "num_input_tokens_seen": 861937640,
+ "step": 19710,
+ "train_runtime": 152468.0419,
+ "train_tokens_per_second": 5653.235
+ },
+ {
+ "epoch": 1.3274306490708323,
+ "grad_norm": 0.7711375220547478,
+ "learning_rate": 2.540715867164685e-06,
+ "loss": 0.3149756908416748,
+ "num_input_tokens_seen": 862156848,
+ "step": 19715,
+ "train_runtime": 152506.2922,
+ "train_tokens_per_second": 5653.254
+ },
+ {
+ "epoch": 1.3277673040667923,
+ "grad_norm": 0.8404428785990262,
+ "learning_rate": 2.538414072012042e-06,
+ "loss": 0.30890326499938964,
+ "num_input_tokens_seen": 862362672,
+ "step": 19720,
+ "train_runtime": 152547.9973,
+ "train_tokens_per_second": 5653.058
+ },
+ {
+ "epoch": 1.3281039590627524,
+ "grad_norm": 0.8169832466518516,
+ "learning_rate": 2.53611296523405e-06,
+ "loss": 0.3164846420288086,
+ "num_input_tokens_seen": 862573904,
+ "step": 19725,
+ "train_runtime": 152580.4337,
+ "train_tokens_per_second": 5653.241
+ },
+ {
+ "epoch": 1.3284406140587126,
+ "grad_norm": 0.793059236401201,
+ "learning_rate": 2.5338125474742127e-06,
+ "loss": 0.2981910228729248,
+ "num_input_tokens_seen": 862787064,
+ "step": 19730,
+ "train_runtime": 152618.0709,
+ "train_tokens_per_second": 5653.243
+ },
+ {
+ "epoch": 1.3287772690546729,
+ "grad_norm": 0.6549096029319272,
+ "learning_rate": 2.531512819375832e-06,
+ "loss": 0.2908968448638916,
+ "num_input_tokens_seen": 863007040,
+ "step": 19735,
+ "train_runtime": 152662.0691,
+ "train_tokens_per_second": 5653.055
+ },
+ {
+ "epoch": 1.3291139240506329,
+ "grad_norm": 0.7643581112810446,
+ "learning_rate": 2.52921378158202e-06,
+ "loss": 0.2935811996459961,
+ "num_input_tokens_seen": 863219112,
+ "step": 19740,
+ "train_runtime": 152697.5407,
+ "train_tokens_per_second": 5653.13
+ },
+ {
+ "epoch": 1.329450579046593,
+ "grad_norm": 0.8140047984203814,
+ "learning_rate": 2.5269154347356957e-06,
+ "loss": 0.2813416957855225,
+ "num_input_tokens_seen": 863441536,
+ "step": 19745,
+ "train_runtime": 152736.5725,
+ "train_tokens_per_second": 5653.142
+ },
+ {
+ "epoch": 1.3297872340425532,
+ "grad_norm": 0.8712638461066203,
+ "learning_rate": 2.5246177794795834e-06,
+ "loss": 0.30994884967803954,
+ "num_input_tokens_seen": 863641736,
+ "step": 19750,
+ "train_runtime": 152783.7657,
+ "train_tokens_per_second": 5652.706
+ },
+ {
+ "epoch": 1.3301238890385134,
+ "grad_norm": 0.6601157103213744,
+ "learning_rate": 2.5223208164562163e-06,
+ "loss": 0.30829753875732424,
+ "num_input_tokens_seen": 863857152,
+ "step": 19755,
+ "train_runtime": 152823.124,
+ "train_tokens_per_second": 5652.66
+ },
+ {
+ "epoch": 1.3304605440344734,
+ "grad_norm": 0.7196822514488103,
+ "learning_rate": 2.5200245463079315e-06,
+ "loss": 0.31928205490112305,
+ "num_input_tokens_seen": 864070800,
+ "step": 19760,
+ "train_runtime": 152859.6114,
+ "train_tokens_per_second": 5652.708
+ },
+ {
+ "epoch": 1.3307971990304335,
+ "grad_norm": 0.7776014244495211,
+ "learning_rate": 2.5177289696768725e-06,
+ "loss": 0.32152862548828126,
+ "num_input_tokens_seen": 864291928,
+ "step": 19765,
+ "train_runtime": 152893.0449,
+ "train_tokens_per_second": 5652.919
+ },
+ {
+ "epoch": 1.3311338540263937,
+ "grad_norm": 0.7457812155053926,
+ "learning_rate": 2.515434087204996e-06,
+ "loss": 0.29552688598632815,
+ "num_input_tokens_seen": 864526144,
+ "step": 19770,
+ "train_runtime": 152933.1454,
+ "train_tokens_per_second": 5652.968
+ },
+ {
+ "epoch": 1.331470509022354,
+ "grad_norm": 0.628218974067636,
+ "learning_rate": 2.51313989953405e-06,
+ "loss": 0.3004995107650757,
+ "num_input_tokens_seen": 864756000,
+ "step": 19775,
+ "train_runtime": 152967.0366,
+ "train_tokens_per_second": 5653.218
+ },
+ {
+ "epoch": 1.331807164018314,
+ "grad_norm": 0.7135273557700192,
+ "learning_rate": 2.510846407305603e-06,
+ "loss": 0.298743748664856,
+ "num_input_tokens_seen": 864976528,
+ "step": 19780,
+ "train_runtime": 153002.9046,
+ "train_tokens_per_second": 5653.334
+ },
+ {
+ "epoch": 1.332143819014274,
+ "grad_norm": 0.678096834485325,
+ "learning_rate": 2.508553611161021e-06,
+ "loss": 0.3048709392547607,
+ "num_input_tokens_seen": 865197968,
+ "step": 19785,
+ "train_runtime": 153045.6277,
+ "train_tokens_per_second": 5653.203
+ },
+ {
+ "epoch": 1.3324804740102343,
+ "grad_norm": 0.7272471213502572,
+ "learning_rate": 2.5062615117414773e-06,
+ "loss": 0.3029000759124756,
+ "num_input_tokens_seen": 865431584,
+ "step": 19790,
+ "train_runtime": 153089.6551,
+ "train_tokens_per_second": 5653.103
+ },
+ {
+ "epoch": 1.3328171290061945,
+ "grad_norm": 0.7477356879857912,
+ "learning_rate": 2.50397010968795e-06,
+ "loss": 0.2859670400619507,
+ "num_input_tokens_seen": 865646552,
+ "step": 19795,
+ "train_runtime": 153133.4962,
+ "train_tokens_per_second": 5652.888
+ },
+ {
+ "epoch": 1.3331537840021546,
+ "grad_norm": 0.7549750199167686,
+ "learning_rate": 2.501679405641222e-06,
+ "loss": 0.32290475368499755,
+ "num_input_tokens_seen": 865868968,
+ "step": 19800,
+ "train_runtime": 153168.7301,
+ "train_tokens_per_second": 5653.04
+ },
+ {
+ "epoch": 1.3334904389981148,
+ "grad_norm": 0.7654205877160946,
+ "learning_rate": 2.499389400241881e-06,
+ "loss": 0.30891990661621094,
+ "num_input_tokens_seen": 866072568,
+ "step": 19805,
+ "train_runtime": 153205.9089,
+ "train_tokens_per_second": 5652.997
+ },
+ {
+ "epoch": 1.3338270939940748,
+ "grad_norm": 0.6968224775968341,
+ "learning_rate": 2.497100094130321e-06,
+ "loss": 0.27096853256225584,
+ "num_input_tokens_seen": 866291048,
+ "step": 19810,
+ "train_runtime": 153242.9077,
+ "train_tokens_per_second": 5653.058
+ },
+ {
+ "epoch": 1.334163748990035,
+ "grad_norm": 0.6748296546699505,
+ "learning_rate": 2.494811487946737e-06,
+ "loss": 0.2911532402038574,
+ "num_input_tokens_seen": 866515272,
+ "step": 19815,
+ "train_runtime": 153277.1916,
+ "train_tokens_per_second": 5653.256
+ },
+ {
+ "epoch": 1.3345004039859951,
+ "grad_norm": 0.7534075118050922,
+ "learning_rate": 2.492523582331135e-06,
+ "loss": 0.3089088201522827,
+ "num_input_tokens_seen": 866738320,
+ "step": 19820,
+ "train_runtime": 153318.6584,
+ "train_tokens_per_second": 5653.182
+ },
+ {
+ "epoch": 1.3348370589819554,
+ "grad_norm": 0.7981812627641985,
+ "learning_rate": 2.4902363779233134e-06,
+ "loss": 0.29610767364501955,
+ "num_input_tokens_seen": 866959032,
+ "step": 19825,
+ "train_runtime": 153358.9779,
+ "train_tokens_per_second": 5653.135
+ },
+ {
+ "epoch": 1.3351737139779154,
+ "grad_norm": 0.7457013614340272,
+ "learning_rate": 2.4879498753628885e-06,
+ "loss": 0.31332998275756835,
+ "num_input_tokens_seen": 867155816,
+ "step": 19830,
+ "train_runtime": 153400.6658,
+ "train_tokens_per_second": 5652.882
+ },
+ {
+ "epoch": 1.3355103689738757,
+ "grad_norm": 0.6543399547857148,
+ "learning_rate": 2.4856640752892702e-06,
+ "loss": 0.27421958446502687,
+ "num_input_tokens_seen": 867382880,
+ "step": 19835,
+ "train_runtime": 153438.0486,
+ "train_tokens_per_second": 5652.984
+ },
+ {
+ "epoch": 1.3358470239698357,
+ "grad_norm": 0.7540236121688988,
+ "learning_rate": 2.4833789783416757e-06,
+ "loss": 0.28227849006652833,
+ "num_input_tokens_seen": 867604504,
+ "step": 19840,
+ "train_runtime": 153478.0421,
+ "train_tokens_per_second": 5652.955
+ },
+ {
+ "epoch": 1.336183678965796,
+ "grad_norm": 0.7797596388766936,
+ "learning_rate": 2.4810945851591245e-06,
+ "loss": 0.3176878452301025,
+ "num_input_tokens_seen": 867810880,
+ "step": 19845,
+ "train_runtime": 153514.5274,
+ "train_tokens_per_second": 5652.956
+ },
+ {
+ "epoch": 1.336520333961756,
+ "grad_norm": 0.7762325041420014,
+ "learning_rate": 2.4788108963804406e-06,
+ "loss": 0.2944268941879272,
+ "num_input_tokens_seen": 868031240,
+ "step": 19850,
+ "train_runtime": 153555.0069,
+ "train_tokens_per_second": 5652.901
+ },
+ {
+ "epoch": 1.3368569889577162,
+ "grad_norm": 0.7515387143609674,
+ "learning_rate": 2.4765279126442502e-06,
+ "loss": 0.30596926212310793,
+ "num_input_tokens_seen": 868253728,
+ "step": 19855,
+ "train_runtime": 153594.4952,
+ "train_tokens_per_second": 5652.896
+ },
+ {
+ "epoch": 1.3371936439536762,
+ "grad_norm": 0.7609490688259355,
+ "learning_rate": 2.4742456345889828e-06,
+ "loss": 0.3069468975067139,
+ "num_input_tokens_seen": 868466312,
+ "step": 19860,
+ "train_runtime": 153633.742,
+ "train_tokens_per_second": 5652.836
+ },
+ {
+ "epoch": 1.3375302989496365,
+ "grad_norm": 0.7746964460091734,
+ "learning_rate": 2.471964062852867e-06,
+ "loss": 0.30750603675842286,
+ "num_input_tokens_seen": 868689096,
+ "step": 19865,
+ "train_runtime": 153665.3062,
+ "train_tokens_per_second": 5653.124
+ },
+ {
+ "epoch": 1.3378669539455965,
+ "grad_norm": 0.6717681437842503,
+ "learning_rate": 2.469683198073943e-06,
+ "loss": 0.29093983173370364,
+ "num_input_tokens_seen": 868917088,
+ "step": 19870,
+ "train_runtime": 153696.1235,
+ "train_tokens_per_second": 5653.474
+ },
+ {
+ "epoch": 1.3382036089415568,
+ "grad_norm": 0.9244971353347443,
+ "learning_rate": 2.4674030408900414e-06,
+ "loss": 0.2891664505004883,
+ "num_input_tokens_seen": 869105376,
+ "step": 19875,
+ "train_runtime": 153735.3162,
+ "train_tokens_per_second": 5653.258
+ },
+ {
+ "epoch": 1.3385402639375168,
+ "grad_norm": 0.8136223037135817,
+ "learning_rate": 2.465123591938805e-06,
+ "loss": 0.3123000144958496,
+ "num_input_tokens_seen": 869319584,
+ "step": 19880,
+ "train_runtime": 153771.8119,
+ "train_tokens_per_second": 5653.309
+ },
+ {
+ "epoch": 1.338876918933477,
+ "grad_norm": 0.73409884799906,
+ "learning_rate": 2.462844851857673e-06,
+ "loss": 0.31596615314483645,
+ "num_input_tokens_seen": 869526576,
+ "step": 19885,
+ "train_runtime": 153815.0765,
+ "train_tokens_per_second": 5653.065
+ },
+ {
+ "epoch": 1.339213573929437,
+ "grad_norm": 0.7895572970153802,
+ "learning_rate": 2.4605668212838878e-06,
+ "loss": 0.31440134048461915,
+ "num_input_tokens_seen": 869742096,
+ "step": 19890,
+ "train_runtime": 153855.5507,
+ "train_tokens_per_second": 5652.978
+ },
+ {
+ "epoch": 1.3395502289253973,
+ "grad_norm": 0.786706111069309,
+ "learning_rate": 2.458289500854493e-06,
+ "loss": 0.34726858139038086,
+ "num_input_tokens_seen": 869961536,
+ "step": 19895,
+ "train_runtime": 153890.7632,
+ "train_tokens_per_second": 5653.111
+ },
+ {
+ "epoch": 1.3398868839213574,
+ "grad_norm": 0.7370468195124786,
+ "learning_rate": 2.4560128912063353e-06,
+ "loss": 0.3165955305099487,
+ "num_input_tokens_seen": 870165224,
+ "step": 19900,
+ "train_runtime": 153932.5225,
+ "train_tokens_per_second": 5652.9
+ },
+ {
+ "epoch": 1.3402235389173176,
+ "grad_norm": 0.7650689001742789,
+ "learning_rate": 2.4537369929760596e-06,
+ "loss": 0.34606313705444336,
+ "num_input_tokens_seen": 870396672,
+ "step": 19905,
+ "train_runtime": 153969.9255,
+ "train_tokens_per_second": 5653.03
+ },
+ {
+ "epoch": 1.3405601939132776,
+ "grad_norm": 0.7725397454862113,
+ "learning_rate": 2.4514618068001144e-06,
+ "loss": 0.305745005607605,
+ "num_input_tokens_seen": 870610560,
+ "step": 19910,
+ "train_runtime": 154008.56,
+ "train_tokens_per_second": 5653.001
+ },
+ {
+ "epoch": 1.3408968489092379,
+ "grad_norm": 0.7591073107632155,
+ "learning_rate": 2.4491873333147463e-06,
+ "loss": 0.29233665466308595,
+ "num_input_tokens_seen": 870825304,
+ "step": 19915,
+ "train_runtime": 154050.5577,
+ "train_tokens_per_second": 5652.854
+ },
+ {
+ "epoch": 1.341233503905198,
+ "grad_norm": 0.8155190700964781,
+ "learning_rate": 2.4469135731560106e-06,
+ "loss": 0.2776491641998291,
+ "num_input_tokens_seen": 871033624,
+ "step": 19920,
+ "train_runtime": 154088.9114,
+ "train_tokens_per_second": 5652.799
+ },
+ {
+ "epoch": 1.3415701589011582,
+ "grad_norm": 0.7835805613524144,
+ "learning_rate": 2.4446405269597494e-06,
+ "loss": 0.31336369514465334,
+ "num_input_tokens_seen": 871261480,
+ "step": 19925,
+ "train_runtime": 154138.0941,
+ "train_tokens_per_second": 5652.473
+ },
+ {
+ "epoch": 1.3419068138971182,
+ "grad_norm": 0.7131677555675145,
+ "learning_rate": 2.442368195361618e-06,
+ "loss": 0.27323174476623535,
+ "num_input_tokens_seen": 871484784,
+ "step": 19930,
+ "train_runtime": 154175.5255,
+ "train_tokens_per_second": 5652.549
+ },
+ {
+ "epoch": 1.3422434688930784,
+ "grad_norm": 0.7614586629617044,
+ "learning_rate": 2.4400965789970647e-06,
+ "loss": 0.2858338594436646,
+ "num_input_tokens_seen": 871694432,
+ "step": 19935,
+ "train_runtime": 154214.3119,
+ "train_tokens_per_second": 5652.487
+ },
+ {
+ "epoch": 1.3425801238890385,
+ "grad_norm": 0.7849206092021981,
+ "learning_rate": 2.437825678501339e-06,
+ "loss": 0.3141868352890015,
+ "num_input_tokens_seen": 871916776,
+ "step": 19940,
+ "train_runtime": 154256.1387,
+ "train_tokens_per_second": 5652.396
+ },
+ {
+ "epoch": 1.3429167788849987,
+ "grad_norm": 0.6972967949553635,
+ "learning_rate": 2.4355554945094956e-06,
+ "loss": 0.3040180683135986,
+ "num_input_tokens_seen": 872136592,
+ "step": 19945,
+ "train_runtime": 154298.8906,
+ "train_tokens_per_second": 5652.254
+ },
+ {
+ "epoch": 1.3432534338809587,
+ "grad_norm": 0.7860171584750419,
+ "learning_rate": 2.4332860276563763e-06,
+ "loss": 0.2974998950958252,
+ "num_input_tokens_seen": 872357128,
+ "step": 19950,
+ "train_runtime": 154335.8822,
+ "train_tokens_per_second": 5652.329
+ },
+ {
+ "epoch": 1.343590088876919,
+ "grad_norm": 0.678809783107802,
+ "learning_rate": 2.4310172785766396e-06,
+ "loss": 0.30943219661712645,
+ "num_input_tokens_seen": 872573624,
+ "step": 19955,
+ "train_runtime": 154378.2017,
+ "train_tokens_per_second": 5652.182
+ },
+ {
+ "epoch": 1.343926743872879,
+ "grad_norm": 0.8435675441786462,
+ "learning_rate": 2.428749247904724e-06,
+ "loss": 0.31881489753723147,
+ "num_input_tokens_seen": 872789824,
+ "step": 19960,
+ "train_runtime": 154415.7323,
+ "train_tokens_per_second": 5652.208
+ },
+ {
+ "epoch": 1.3442633988688393,
+ "grad_norm": 0.7510529942793552,
+ "learning_rate": 2.4264819362748856e-06,
+ "loss": 0.3125936031341553,
+ "num_input_tokens_seen": 873009864,
+ "step": 19965,
+ "train_runtime": 154452.1402,
+ "train_tokens_per_second": 5652.3
+ },
+ {
+ "epoch": 1.3446000538647993,
+ "grad_norm": 0.8178579575509619,
+ "learning_rate": 2.4242153443211663e-06,
+ "loss": 0.3280599594116211,
+ "num_input_tokens_seen": 873230264,
+ "step": 19970,
+ "train_runtime": 154483.7052,
+ "train_tokens_per_second": 5652.572
+ },
+ {
+ "epoch": 1.3449367088607596,
+ "grad_norm": 0.7189596121215553,
+ "learning_rate": 2.4219494726774124e-06,
+ "loss": 0.2943082809448242,
+ "num_input_tokens_seen": 873453312,
+ "step": 19975,
+ "train_runtime": 154519.1177,
+ "train_tokens_per_second": 5652.72
+ },
+ {
+ "epoch": 1.3452733638567196,
+ "grad_norm": 0.7355041064223901,
+ "learning_rate": 2.419684321977268e-06,
+ "loss": 0.2926133632659912,
+ "num_input_tokens_seen": 873660984,
+ "step": 19980,
+ "train_runtime": 154553.7989,
+ "train_tokens_per_second": 5652.795
+ },
+ {
+ "epoch": 1.3456100188526798,
+ "grad_norm": 0.6865570643674106,
+ "learning_rate": 2.417419892854175e-06,
+ "loss": 0.29560866355896,
+ "num_input_tokens_seen": 873892712,
+ "step": 19985,
+ "train_runtime": 154598.2899,
+ "train_tokens_per_second": 5652.667
+ },
+ {
+ "epoch": 1.3459466738486399,
+ "grad_norm": 0.885274027016748,
+ "learning_rate": 2.415156185941371e-06,
+ "loss": 0.29713964462280273,
+ "num_input_tokens_seen": 874102984,
+ "step": 19990,
+ "train_runtime": 154633.4222,
+ "train_tokens_per_second": 5652.743
+ },
+ {
+ "epoch": 1.3462833288446001,
+ "grad_norm": 0.7645675693693532,
+ "learning_rate": 2.412893201871902e-06,
+ "loss": 0.30781164169311526,
+ "num_input_tokens_seen": 874317048,
+ "step": 19995,
+ "train_runtime": 154674.4314,
+ "train_tokens_per_second": 5652.628
+ },
+ {
+ "epoch": 1.3466199838405601,
+ "grad_norm": 0.7929208093171044,
+ "learning_rate": 2.410630941278595e-06,
+ "loss": 0.31746225357055663,
+ "num_input_tokens_seen": 874537536,
+ "step": 20000,
+ "train_runtime": 154717.3345,
+ "train_tokens_per_second": 5652.486
+ },
+ {
+ "epoch": 1.3469566388365204,
+ "grad_norm": 0.8665886750624643,
+ "learning_rate": 2.408369404794092e-06,
+ "loss": 0.3095715045928955,
+ "num_input_tokens_seen": 874752264,
+ "step": 20005,
+ "train_runtime": 154752.6238,
+ "train_tokens_per_second": 5652.584
+ },
+ {
+ "epoch": 1.3472932938324804,
+ "grad_norm": 0.7516875018971654,
+ "learning_rate": 2.4061085930508176e-06,
+ "loss": 0.30551714897155763,
+ "num_input_tokens_seen": 874964968,
+ "step": 20010,
+ "train_runtime": 154789.9589,
+ "train_tokens_per_second": 5652.595
+ },
+ {
+ "epoch": 1.3476299488284407,
+ "grad_norm": 0.8041861814830855,
+ "learning_rate": 2.4038485066810058e-06,
+ "loss": 0.3143286228179932,
+ "num_input_tokens_seen": 875176368,
+ "step": 20015,
+ "train_runtime": 154836.4632,
+ "train_tokens_per_second": 5652.263
+ },
+ {
+ "epoch": 1.3479666038244007,
+ "grad_norm": 0.7821727058895627,
+ "learning_rate": 2.40158914631668e-06,
+ "loss": 0.276894736289978,
+ "num_input_tokens_seen": 875385288,
+ "step": 20020,
+ "train_runtime": 154870.1836,
+ "train_tokens_per_second": 5652.381
+ },
+ {
+ "epoch": 1.348303258820361,
+ "grad_norm": 0.723635333292079,
+ "learning_rate": 2.399330512589664e-06,
+ "loss": 0.29750776290893555,
+ "num_input_tokens_seen": 875609880,
+ "step": 20025,
+ "train_runtime": 154911.8697,
+ "train_tokens_per_second": 5652.31
+ },
+ {
+ "epoch": 1.348639913816321,
+ "grad_norm": 0.7243913395939594,
+ "learning_rate": 2.3970726061315775e-06,
+ "loss": 0.30971436500549315,
+ "num_input_tokens_seen": 875831888,
+ "step": 20030,
+ "train_runtime": 154948.8908,
+ "train_tokens_per_second": 5652.392
+ },
+ {
+ "epoch": 1.3489765688122812,
+ "grad_norm": 0.7399215488091135,
+ "learning_rate": 2.394815427573836e-06,
+ "loss": 0.30228476524353026,
+ "num_input_tokens_seen": 876061376,
+ "step": 20035,
+ "train_runtime": 154992.8431,
+ "train_tokens_per_second": 5652.27
+ },
+ {
+ "epoch": 1.3493132238082413,
+ "grad_norm": 0.7012240220598052,
+ "learning_rate": 2.3925589775476516e-06,
+ "loss": 0.2862797021865845,
+ "num_input_tokens_seen": 876272016,
+ "step": 20040,
+ "train_runtime": 155032.043,
+ "train_tokens_per_second": 5652.199
+ },
+ {
+ "epoch": 1.3496498788042015,
+ "grad_norm": 0.7194801409568423,
+ "learning_rate": 2.3903032566840377e-06,
+ "loss": 0.31173529624938967,
+ "num_input_tokens_seen": 876494336,
+ "step": 20045,
+ "train_runtime": 155073.9927,
+ "train_tokens_per_second": 5652.104
+ },
+ {
+ "epoch": 1.3499865338001615,
+ "grad_norm": 0.7007455859084586,
+ "learning_rate": 2.3880482656137926e-06,
+ "loss": 0.31239540576934816,
+ "num_input_tokens_seen": 876727136,
+ "step": 20050,
+ "train_runtime": 155114.2632,
+ "train_tokens_per_second": 5652.137
+ },
+ {
+ "epoch": 1.3503231887961218,
+ "grad_norm": 0.7248173637285731,
+ "learning_rate": 2.3857940049675254e-06,
+ "loss": 0.3017667055130005,
+ "num_input_tokens_seen": 876951616,
+ "step": 20055,
+ "train_runtime": 155154.1439,
+ "train_tokens_per_second": 5652.131
+ },
+ {
+ "epoch": 1.3506598437920818,
+ "grad_norm": 0.7421571706340755,
+ "learning_rate": 2.383540475375625e-06,
+ "loss": 0.2906331062316895,
+ "num_input_tokens_seen": 877172960,
+ "step": 20060,
+ "train_runtime": 155195.6002,
+ "train_tokens_per_second": 5652.048
+ },
+ {
+ "epoch": 1.350996498788042,
+ "grad_norm": 0.7848990495009128,
+ "learning_rate": 2.3812876774682886e-06,
+ "loss": 0.30210440158843993,
+ "num_input_tokens_seen": 877405296,
+ "step": 20065,
+ "train_runtime": 155229.0295,
+ "train_tokens_per_second": 5652.327
+ },
+ {
+ "epoch": 1.351333153784002,
+ "grad_norm": 0.7518085342917976,
+ "learning_rate": 2.379035611875503e-06,
+ "loss": 0.3308712959289551,
+ "num_input_tokens_seen": 877636392,
+ "step": 20070,
+ "train_runtime": 155261.3122,
+ "train_tokens_per_second": 5652.641
+ },
+ {
+ "epoch": 1.3516698087799623,
+ "grad_norm": 0.7278323335362945,
+ "learning_rate": 2.376784279227051e-06,
+ "loss": 0.2974379062652588,
+ "num_input_tokens_seen": 877871256,
+ "step": 20075,
+ "train_runtime": 155301.2047,
+ "train_tokens_per_second": 5652.701
+ },
+ {
+ "epoch": 1.3520064637759224,
+ "grad_norm": 0.7168837215922106,
+ "learning_rate": 2.3745336801525105e-06,
+ "loss": 0.3024725437164307,
+ "num_input_tokens_seen": 878104688,
+ "step": 20080,
+ "train_runtime": 155345.4421,
+ "train_tokens_per_second": 5652.594
+ },
+ {
+ "epoch": 1.3523431187718826,
+ "grad_norm": 0.7794337072066637,
+ "learning_rate": 2.3722838152812543e-06,
+ "loss": 0.3230156898498535,
+ "num_input_tokens_seen": 878325104,
+ "step": 20085,
+ "train_runtime": 155391.9834,
+ "train_tokens_per_second": 5652.319
+ },
+ {
+ "epoch": 1.3526797737678427,
+ "grad_norm": 0.607568547988963,
+ "learning_rate": 2.3700346852424485e-06,
+ "loss": 0.2935077428817749,
+ "num_input_tokens_seen": 878541320,
+ "step": 20090,
+ "train_runtime": 155423.8708,
+ "train_tokens_per_second": 5652.551
+ },
+ {
+ "epoch": 1.353016428763803,
+ "grad_norm": 0.7909886746041502,
+ "learning_rate": 2.3677862906650594e-06,
+ "loss": 0.3263228416442871,
+ "num_input_tokens_seen": 878755824,
+ "step": 20095,
+ "train_runtime": 155457.5114,
+ "train_tokens_per_second": 5652.707
+ },
+ {
+ "epoch": 1.353353083759763,
+ "grad_norm": 0.8833083944310808,
+ "learning_rate": 2.365538632177837e-06,
+ "loss": 0.3379992961883545,
+ "num_input_tokens_seen": 878961448,
+ "step": 20100,
+ "train_runtime": 155499.7479,
+ "train_tokens_per_second": 5652.494
+ },
+ {
+ "epoch": 1.3536897387557232,
+ "grad_norm": 0.7244128821181046,
+ "learning_rate": 2.36329171040934e-06,
+ "loss": 0.3040963649749756,
+ "num_input_tokens_seen": 879189696,
+ "step": 20105,
+ "train_runtime": 155531.2684,
+ "train_tokens_per_second": 5652.816
+ },
+ {
+ "epoch": 1.3540263937516832,
+ "grad_norm": 0.7531385915612084,
+ "learning_rate": 2.3610455259879045e-06,
+ "loss": 0.2924042224884033,
+ "num_input_tokens_seen": 879399600,
+ "step": 20110,
+ "train_runtime": 155567.1802,
+ "train_tokens_per_second": 5652.861
+ },
+ {
+ "epoch": 1.3543630487476435,
+ "grad_norm": 0.8365089742949873,
+ "learning_rate": 2.358800079541674e-06,
+ "loss": 0.3243632078170776,
+ "num_input_tokens_seen": 879611920,
+ "step": 20115,
+ "train_runtime": 155607.878,
+ "train_tokens_per_second": 5652.747
+ },
+ {
+ "epoch": 1.3546997037436035,
+ "grad_norm": 0.8166182835087867,
+ "learning_rate": 2.3565553716985794e-06,
+ "loss": 0.3272987365722656,
+ "num_input_tokens_seen": 879834920,
+ "step": 20120,
+ "train_runtime": 155641.8573,
+ "train_tokens_per_second": 5652.945
+ },
+ {
+ "epoch": 1.3550363587395637,
+ "grad_norm": 0.8219130483551738,
+ "learning_rate": 2.3543114030863457e-06,
+ "loss": 0.2879811763763428,
+ "num_input_tokens_seen": 880035976,
+ "step": 20125,
+ "train_runtime": 155679.3392,
+ "train_tokens_per_second": 5652.876
+ },
+ {
+ "epoch": 1.3553730137355238,
+ "grad_norm": 0.778830195040845,
+ "learning_rate": 2.352068174332491e-06,
+ "loss": 0.2902420997619629,
+ "num_input_tokens_seen": 880240280,
+ "step": 20130,
+ "train_runtime": 155721.9307,
+ "train_tokens_per_second": 5652.642
+ },
+ {
+ "epoch": 1.355709668731484,
+ "grad_norm": 0.7246075057224634,
+ "learning_rate": 2.349825686064329e-06,
+ "loss": 0.2633873462677002,
+ "num_input_tokens_seen": 880457816,
+ "step": 20135,
+ "train_runtime": 155760.9984,
+ "train_tokens_per_second": 5652.621
+ },
+ {
+ "epoch": 1.356046323727444,
+ "grad_norm": 0.7082973857821522,
+ "learning_rate": 2.3475839389089605e-06,
+ "loss": 0.31766154766082766,
+ "num_input_tokens_seen": 880688632,
+ "step": 20140,
+ "train_runtime": 155802.16,
+ "train_tokens_per_second": 5652.609
+ },
+ {
+ "epoch": 1.3563829787234043,
+ "grad_norm": 0.7717628495827882,
+ "learning_rate": 2.345342933493289e-06,
+ "loss": 0.3041571617126465,
+ "num_input_tokens_seen": 880900704,
+ "step": 20145,
+ "train_runtime": 155842.8065,
+ "train_tokens_per_second": 5652.495
+ },
+ {
+ "epoch": 1.3567196337193643,
+ "grad_norm": 0.6914688947775744,
+ "learning_rate": 2.343102670443998e-06,
+ "loss": 0.2988900661468506,
+ "num_input_tokens_seen": 881116800,
+ "step": 20150,
+ "train_runtime": 155885.0508,
+ "train_tokens_per_second": 5652.35
+ },
+ {
+ "epoch": 1.3570562887153246,
+ "grad_norm": 0.742550661860349,
+ "learning_rate": 2.3408631503875763e-06,
+ "loss": 0.32421116828918456,
+ "num_input_tokens_seen": 881349184,
+ "step": 20155,
+ "train_runtime": 155925.098,
+ "train_tokens_per_second": 5652.388
+ },
+ {
+ "epoch": 1.3573929437112846,
+ "grad_norm": 0.878343673750393,
+ "learning_rate": 2.338624373950291e-06,
+ "loss": 0.2701118946075439,
+ "num_input_tokens_seen": 881566872,
+ "step": 20160,
+ "train_runtime": 155957.1429,
+ "train_tokens_per_second": 5652.623
+ },
+ {
+ "epoch": 1.3577295987072449,
+ "grad_norm": 0.755841373165112,
+ "learning_rate": 2.336386341758216e-06,
+ "loss": 0.30777089595794677,
+ "num_input_tokens_seen": 881802424,
+ "step": 20165,
+ "train_runtime": 155994.1426,
+ "train_tokens_per_second": 5652.792
+ },
+ {
+ "epoch": 1.3580662537032049,
+ "grad_norm": 0.8144086492913879,
+ "learning_rate": 2.3341490544372063e-06,
+ "loss": 0.3082592964172363,
+ "num_input_tokens_seen": 882020264,
+ "step": 20170,
+ "train_runtime": 156033.3282,
+ "train_tokens_per_second": 5652.768
+ },
+ {
+ "epoch": 1.3584029086991651,
+ "grad_norm": 0.7512527703728554,
+ "learning_rate": 2.3319125126129128e-06,
+ "loss": 0.3048335790634155,
+ "num_input_tokens_seen": 882240304,
+ "step": 20175,
+ "train_runtime": 156070.3454,
+ "train_tokens_per_second": 5652.838
+ },
+ {
+ "epoch": 1.3587395636951252,
+ "grad_norm": 0.7584329932938183,
+ "learning_rate": 2.329676716910777e-06,
+ "loss": 0.32310731410980226,
+ "num_input_tokens_seen": 882456760,
+ "step": 20180,
+ "train_runtime": 156107.9369,
+ "train_tokens_per_second": 5652.863
+ },
+ {
+ "epoch": 1.3590762186910854,
+ "grad_norm": 0.791075197267678,
+ "learning_rate": 2.3274416679560325e-06,
+ "loss": 0.3185749053955078,
+ "num_input_tokens_seen": 882677128,
+ "step": 20185,
+ "train_runtime": 156139.8506,
+ "train_tokens_per_second": 5653.119
+ },
+ {
+ "epoch": 1.3594128736870454,
+ "grad_norm": 0.789149126017219,
+ "learning_rate": 2.3252073663737016e-06,
+ "loss": 0.31891803741455077,
+ "num_input_tokens_seen": 882909016,
+ "step": 20190,
+ "train_runtime": 156176.5363,
+ "train_tokens_per_second": 5653.276
+ },
+ {
+ "epoch": 1.3597495286830057,
+ "grad_norm": 0.8825428696038273,
+ "learning_rate": 2.3229738127886054e-06,
+ "loss": 0.312581205368042,
+ "num_input_tokens_seen": 883114832,
+ "step": 20195,
+ "train_runtime": 156221.1973,
+ "train_tokens_per_second": 5652.977
+ },
+ {
+ "epoch": 1.3600861836789657,
+ "grad_norm": 0.666411052579731,
+ "learning_rate": 2.3207410078253416e-06,
+ "loss": 0.30325491428375245,
+ "num_input_tokens_seen": 883339080,
+ "step": 20200,
+ "train_runtime": 156268.3408,
+ "train_tokens_per_second": 5652.707
+ },
+ {
+ "epoch": 1.360422838674926,
+ "grad_norm": 0.7726612121097549,
+ "learning_rate": 2.3185089521083154e-06,
+ "loss": 0.3062901496887207,
+ "num_input_tokens_seen": 883551600,
+ "step": 20205,
+ "train_runtime": 156305.042,
+ "train_tokens_per_second": 5652.739
+ },
+ {
+ "epoch": 1.360759493670886,
+ "grad_norm": 0.7956523367599065,
+ "learning_rate": 2.3162776462617067e-06,
+ "loss": 0.3058837175369263,
+ "num_input_tokens_seen": 883769552,
+ "step": 20210,
+ "train_runtime": 156351.0682,
+ "train_tokens_per_second": 5652.469
+ },
+ {
+ "epoch": 1.3610961486668463,
+ "grad_norm": 0.8171843690979657,
+ "learning_rate": 2.314047090909498e-06,
+ "loss": 0.28980307579040526,
+ "num_input_tokens_seen": 883987024,
+ "step": 20215,
+ "train_runtime": 156388.3039,
+ "train_tokens_per_second": 5652.514
+ },
+ {
+ "epoch": 1.3614328036628063,
+ "grad_norm": 0.7366060863576702,
+ "learning_rate": 2.3118172866754557e-06,
+ "loss": 0.289400577545166,
+ "num_input_tokens_seen": 884204584,
+ "step": 20220,
+ "train_runtime": 156436.651,
+ "train_tokens_per_second": 5652.157
+ },
+ {
+ "epoch": 1.3617694586587665,
+ "grad_norm": 0.8418632778595053,
+ "learning_rate": 2.309588234183137e-06,
+ "loss": 0.29900097846984863,
+ "num_input_tokens_seen": 884430104,
+ "step": 20225,
+ "train_runtime": 156479.2021,
+ "train_tokens_per_second": 5652.062
+ },
+ {
+ "epoch": 1.3621061136547266,
+ "grad_norm": 0.7596247998423821,
+ "learning_rate": 2.3073599340558896e-06,
+ "loss": 0.3193103313446045,
+ "num_input_tokens_seen": 884638136,
+ "step": 20230,
+ "train_runtime": 156513.4901,
+ "train_tokens_per_second": 5652.153
+ },
+ {
+ "epoch": 1.3624427686506868,
+ "grad_norm": 0.794889195242685,
+ "learning_rate": 2.3051323869168506e-06,
+ "loss": 0.31695847511291503,
+ "num_input_tokens_seen": 884867312,
+ "step": 20235,
+ "train_runtime": 156552.9308,
+ "train_tokens_per_second": 5652.193
+ },
+ {
+ "epoch": 1.3627794236466468,
+ "grad_norm": 0.6894100091265923,
+ "learning_rate": 2.3029055933889443e-06,
+ "loss": 0.299372124671936,
+ "num_input_tokens_seen": 885078648,
+ "step": 20240,
+ "train_runtime": 156591.403,
+ "train_tokens_per_second": 5652.154
+ },
+ {
+ "epoch": 1.363116078642607,
+ "grad_norm": 0.679829103468669,
+ "learning_rate": 2.300679554094892e-06,
+ "loss": 0.27194240093231203,
+ "num_input_tokens_seen": 885288824,
+ "step": 20245,
+ "train_runtime": 156624.8066,
+ "train_tokens_per_second": 5652.29
+ },
+ {
+ "epoch": 1.3634527336385671,
+ "grad_norm": 0.7428557022690266,
+ "learning_rate": 2.298454269657191e-06,
+ "loss": 0.2977342367172241,
+ "num_input_tokens_seen": 885509192,
+ "step": 20250,
+ "train_runtime": 156665.816,
+ "train_tokens_per_second": 5652.217
+ },
+ {
+ "epoch": 1.3637893886345274,
+ "grad_norm": 0.7495727549443245,
+ "learning_rate": 2.2962297406981432e-06,
+ "loss": 0.30438592433929446,
+ "num_input_tokens_seen": 885728368,
+ "step": 20255,
+ "train_runtime": 156704.8388,
+ "train_tokens_per_second": 5652.208
+ },
+ {
+ "epoch": 1.3641260436304874,
+ "grad_norm": 0.7491077517728351,
+ "learning_rate": 2.2940059678398223e-06,
+ "loss": 0.2986233949661255,
+ "num_input_tokens_seen": 885946024,
+ "step": 20260,
+ "train_runtime": 156744.7021,
+ "train_tokens_per_second": 5652.159
+ },
+ {
+ "epoch": 1.3644626986264476,
+ "grad_norm": 0.7341728818382711,
+ "learning_rate": 2.291782951704106e-06,
+ "loss": 0.30042304992675783,
+ "num_input_tokens_seen": 886159456,
+ "step": 20265,
+ "train_runtime": 156781.0468,
+ "train_tokens_per_second": 5652.21
+ },
+ {
+ "epoch": 1.3647993536224077,
+ "grad_norm": 0.7856616622328413,
+ "learning_rate": 2.289560692912651e-06,
+ "loss": 0.30829734802246095,
+ "num_input_tokens_seen": 886378184,
+ "step": 20270,
+ "train_runtime": 156822.443,
+ "train_tokens_per_second": 5652.113
+ },
+ {
+ "epoch": 1.365136008618368,
+ "grad_norm": 0.722302764482047,
+ "learning_rate": 2.287339192086906e-06,
+ "loss": 0.30705411434173585,
+ "num_input_tokens_seen": 886596912,
+ "step": 20275,
+ "train_runtime": 156862.8796,
+ "train_tokens_per_second": 5652.05
+ },
+ {
+ "epoch": 1.365472663614328,
+ "grad_norm": 0.7818607969471453,
+ "learning_rate": 2.2851184498481056e-06,
+ "loss": 0.30739502906799315,
+ "num_input_tokens_seen": 886810320,
+ "step": 20280,
+ "train_runtime": 156900.5368,
+ "train_tokens_per_second": 5652.054
+ },
+ {
+ "epoch": 1.3658093186102882,
+ "grad_norm": 0.8017159372977292,
+ "learning_rate": 2.2828984668172732e-06,
+ "loss": 0.29115149974822996,
+ "num_input_tokens_seen": 887018752,
+ "step": 20285,
+ "train_runtime": 156938.5878,
+ "train_tokens_per_second": 5652.012
+ },
+ {
+ "epoch": 1.3661459736062485,
+ "grad_norm": 0.8305588542630102,
+ "learning_rate": 2.2806792436152185e-06,
+ "loss": 0.3051044464111328,
+ "num_input_tokens_seen": 887236112,
+ "step": 20290,
+ "train_runtime": 156975.4855,
+ "train_tokens_per_second": 5652.068
+ },
+ {
+ "epoch": 1.3664826286022085,
+ "grad_norm": 0.80831246019114,
+ "learning_rate": 2.2784607808625463e-06,
+ "loss": 0.30942864418029786,
+ "num_input_tokens_seen": 887448744,
+ "step": 20295,
+ "train_runtime": 157014.5677,
+ "train_tokens_per_second": 5652.015
+ },
+ {
+ "epoch": 1.3668192835981685,
+ "grad_norm": 0.7675600985733588,
+ "learning_rate": 2.276243079179634e-06,
+ "loss": 0.2999575138092041,
+ "num_input_tokens_seen": 887665496,
+ "step": 20300,
+ "train_runtime": 157053.8678,
+ "train_tokens_per_second": 5651.981
+ },
+ {
+ "epoch": 1.3671559385941288,
+ "grad_norm": 0.811512893486603,
+ "learning_rate": 2.2740261391866634e-06,
+ "loss": 0.31462917327880857,
+ "num_input_tokens_seen": 887885544,
+ "step": 20305,
+ "train_runtime": 157085.2659,
+ "train_tokens_per_second": 5652.252
+ },
+ {
+ "epoch": 1.367492593590089,
+ "grad_norm": 0.7496759995290464,
+ "learning_rate": 2.2718099615035865e-06,
+ "loss": 0.2609941720962524,
+ "num_input_tokens_seen": 888107128,
+ "step": 20310,
+ "train_runtime": 157126.9618,
+ "train_tokens_per_second": 5652.163
+ },
+ {
+ "epoch": 1.367829248586049,
+ "grad_norm": 0.6960078882906409,
+ "learning_rate": 2.2695945467501567e-06,
+ "loss": 0.3140725612640381,
+ "num_input_tokens_seen": 888330336,
+ "step": 20315,
+ "train_runtime": 157165.5698,
+ "train_tokens_per_second": 5652.194
+ },
+ {
+ "epoch": 1.368165903582009,
+ "grad_norm": 0.736910787731632,
+ "learning_rate": 2.267379895545905e-06,
+ "loss": 0.2964684009552002,
+ "num_input_tokens_seen": 888549184,
+ "step": 20320,
+ "train_runtime": 157205.25,
+ "train_tokens_per_second": 5652.16
+ },
+ {
+ "epoch": 1.3685025585779693,
+ "grad_norm": 0.6504305837119558,
+ "learning_rate": 2.2651660085101513e-06,
+ "loss": 0.3142958641052246,
+ "num_input_tokens_seen": 888773632,
+ "step": 20325,
+ "train_runtime": 157243.6633,
+ "train_tokens_per_second": 5652.206
+ },
+ {
+ "epoch": 1.3688392135739296,
+ "grad_norm": 0.7317039701175643,
+ "learning_rate": 2.262952886262003e-06,
+ "loss": 0.30502166748046877,
+ "num_input_tokens_seen": 889003680,
+ "step": 20330,
+ "train_runtime": 157285.4107,
+ "train_tokens_per_second": 5652.169
+ },
+ {
+ "epoch": 1.3691758685698896,
+ "grad_norm": 0.6676476248897819,
+ "learning_rate": 2.260740529420352e-06,
+ "loss": 0.29352798461914065,
+ "num_input_tokens_seen": 889224368,
+ "step": 20335,
+ "train_runtime": 157329.8859,
+ "train_tokens_per_second": 5651.974
+ },
+ {
+ "epoch": 1.3695125235658496,
+ "grad_norm": 0.8139919345906689,
+ "learning_rate": 2.2585289386038754e-06,
+ "loss": 0.29394705295562745,
+ "num_input_tokens_seen": 889456760,
+ "step": 20340,
+ "train_runtime": 157369.7584,
+ "train_tokens_per_second": 5652.018
+ },
+ {
+ "epoch": 1.3698491785618099,
+ "grad_norm": 0.6922241241359759,
+ "learning_rate": 2.256318114431043e-06,
+ "loss": 0.3354970932006836,
+ "num_input_tokens_seen": 889693056,
+ "step": 20345,
+ "train_runtime": 157404.9626,
+ "train_tokens_per_second": 5652.255
+ },
+ {
+ "epoch": 1.3701858335577701,
+ "grad_norm": 0.8121467144758657,
+ "learning_rate": 2.2541080575200973e-06,
+ "loss": 0.2972100734710693,
+ "num_input_tokens_seen": 889902256,
+ "step": 20350,
+ "train_runtime": 157447.764,
+ "train_tokens_per_second": 5652.048
+ },
+ {
+ "epoch": 1.3705224885537302,
+ "grad_norm": 0.7645000522207971,
+ "learning_rate": 2.251898768489081e-06,
+ "loss": 0.3170339822769165,
+ "num_input_tokens_seen": 890133696,
+ "step": 20355,
+ "train_runtime": 157481.0705,
+ "train_tokens_per_second": 5652.322
+ },
+ {
+ "epoch": 1.3708591435496902,
+ "grad_norm": 0.7817916261633198,
+ "learning_rate": 2.249690247955807e-06,
+ "loss": 0.31004133224487307,
+ "num_input_tokens_seen": 890349704,
+ "step": 20360,
+ "train_runtime": 157525.1026,
+ "train_tokens_per_second": 5652.113
+ },
+ {
+ "epoch": 1.3711957985456504,
+ "grad_norm": 1.0121499916406171,
+ "learning_rate": 2.247482496537887e-06,
+ "loss": 0.32703423500061035,
+ "num_input_tokens_seen": 890570904,
+ "step": 20365,
+ "train_runtime": 157561.604,
+ "train_tokens_per_second": 5652.208
+ },
+ {
+ "epoch": 1.3715324535416107,
+ "grad_norm": 0.7908431166811735,
+ "learning_rate": 2.24527551485271e-06,
+ "loss": 0.3082599639892578,
+ "num_input_tokens_seen": 890789232,
+ "step": 20370,
+ "train_runtime": 157598.6794,
+ "train_tokens_per_second": 5652.263
+ },
+ {
+ "epoch": 1.3718691085375707,
+ "grad_norm": 0.739831014749458,
+ "learning_rate": 2.243069303517451e-06,
+ "loss": 0.3278525352478027,
+ "num_input_tokens_seen": 891003792,
+ "step": 20375,
+ "train_runtime": 157631.5929,
+ "train_tokens_per_second": 5652.444
+ },
+ {
+ "epoch": 1.3722057635335307,
+ "grad_norm": 0.6731804830965303,
+ "learning_rate": 2.2408638631490696e-06,
+ "loss": 0.2803332328796387,
+ "num_input_tokens_seen": 891222136,
+ "step": 20380,
+ "train_runtime": 157671.5132,
+ "train_tokens_per_second": 5652.398
+ },
+ {
+ "epoch": 1.372542418529491,
+ "grad_norm": 0.7577807518738551,
+ "learning_rate": 2.2386591943643117e-06,
+ "loss": 0.2959303617477417,
+ "num_input_tokens_seen": 891447208,
+ "step": 20385,
+ "train_runtime": 157709.6607,
+ "train_tokens_per_second": 5652.458
+ },
+ {
+ "epoch": 1.3728790735254512,
+ "grad_norm": 0.740650904611458,
+ "learning_rate": 2.2364552977797023e-06,
+ "loss": 0.3103938579559326,
+ "num_input_tokens_seen": 891668296,
+ "step": 20390,
+ "train_runtime": 157751.8454,
+ "train_tokens_per_second": 5652.348
+ },
+ {
+ "epoch": 1.3732157285214113,
+ "grad_norm": 0.8801339361162619,
+ "learning_rate": 2.23425217401156e-06,
+ "loss": 0.3227031707763672,
+ "num_input_tokens_seen": 891888096,
+ "step": 20395,
+ "train_runtime": 157787.2294,
+ "train_tokens_per_second": 5652.473
+ },
+ {
+ "epoch": 1.3735523835173713,
+ "grad_norm": 0.766299417789016,
+ "learning_rate": 2.2320498236759746e-06,
+ "loss": 0.2954218864440918,
+ "num_input_tokens_seen": 892110048,
+ "step": 20400,
+ "train_runtime": 157817.9321,
+ "train_tokens_per_second": 5652.78
+ },
+ {
+ "epoch": 1.3738890385133316,
+ "grad_norm": 0.6837353560545685,
+ "learning_rate": 2.2298482473888335e-06,
+ "loss": 0.2996779441833496,
+ "num_input_tokens_seen": 892325640,
+ "step": 20405,
+ "train_runtime": 157859.5289,
+ "train_tokens_per_second": 5652.656
+ },
+ {
+ "epoch": 1.3742256935092918,
+ "grad_norm": 0.7214654235542152,
+ "learning_rate": 2.227647445765792e-06,
+ "loss": 0.3093862533569336,
+ "num_input_tokens_seen": 892551144,
+ "step": 20410,
+ "train_runtime": 157899.5496,
+ "train_tokens_per_second": 5652.652
+ },
+ {
+ "epoch": 1.3745623485052518,
+ "grad_norm": 0.9006378011875812,
+ "learning_rate": 2.225447419422305e-06,
+ "loss": 0.2831476926803589,
+ "num_input_tokens_seen": 892763912,
+ "step": 20415,
+ "train_runtime": 157937.9848,
+ "train_tokens_per_second": 5652.623
+ },
+ {
+ "epoch": 1.3748990035012119,
+ "grad_norm": 0.751847783359078,
+ "learning_rate": 2.223248168973599e-06,
+ "loss": 0.29417757987976073,
+ "num_input_tokens_seen": 892984176,
+ "step": 20420,
+ "train_runtime": 157976.464,
+ "train_tokens_per_second": 5652.641
+ },
+ {
+ "epoch": 1.375235658497172,
+ "grad_norm": 0.762461921564537,
+ "learning_rate": 2.221049695034688e-06,
+ "loss": 0.2913064479827881,
+ "num_input_tokens_seen": 893210312,
+ "step": 20425,
+ "train_runtime": 158009.5129,
+ "train_tokens_per_second": 5652.889
+ },
+ {
+ "epoch": 1.3755723134931324,
+ "grad_norm": 0.7837668271345243,
+ "learning_rate": 2.218851998220369e-06,
+ "loss": 0.3103857755661011,
+ "num_input_tokens_seen": 893433344,
+ "step": 20430,
+ "train_runtime": 158043.6487,
+ "train_tokens_per_second": 5653.08
+ },
+ {
+ "epoch": 1.3759089684890924,
+ "grad_norm": 0.7225824849007935,
+ "learning_rate": 2.21665507914522e-06,
+ "loss": 0.2799984931945801,
+ "num_input_tokens_seen": 893630528,
+ "step": 20435,
+ "train_runtime": 158082.8071,
+ "train_tokens_per_second": 5652.927
+ },
+ {
+ "epoch": 1.3762456234850524,
+ "grad_norm": 0.7975885820333312,
+ "learning_rate": 2.214458938423601e-06,
+ "loss": 0.311299204826355,
+ "num_input_tokens_seen": 893854672,
+ "step": 20440,
+ "train_runtime": 158125.2748,
+ "train_tokens_per_second": 5652.826
+ },
+ {
+ "epoch": 1.3765822784810127,
+ "grad_norm": 0.6830394522468479,
+ "learning_rate": 2.212263576669661e-06,
+ "loss": 0.2979740142822266,
+ "num_input_tokens_seen": 894084224,
+ "step": 20445,
+ "train_runtime": 158169.3177,
+ "train_tokens_per_second": 5652.703
+ },
+ {
+ "epoch": 1.376918933476973,
+ "grad_norm": 0.8072743303598326,
+ "learning_rate": 2.2100689944973182e-06,
+ "loss": 0.3165121555328369,
+ "num_input_tokens_seen": 894311056,
+ "step": 20450,
+ "train_runtime": 158215.4891,
+ "train_tokens_per_second": 5652.487
+ },
+ {
+ "epoch": 1.377255588472933,
+ "grad_norm": 0.7485842516852395,
+ "learning_rate": 2.2078751925202886e-06,
+ "loss": 0.34031267166137696,
+ "num_input_tokens_seen": 894543864,
+ "step": 20455,
+ "train_runtime": 158263.0938,
+ "train_tokens_per_second": 5652.258
+ },
+ {
+ "epoch": 1.377592243468893,
+ "grad_norm": 0.878465392483748,
+ "learning_rate": 2.205682171352054e-06,
+ "loss": 0.30671300888061526,
+ "num_input_tokens_seen": 894767288,
+ "step": 20460,
+ "train_runtime": 158302.6816,
+ "train_tokens_per_second": 5652.256
+ },
+ {
+ "epoch": 1.3779288984648532,
+ "grad_norm": 0.7614994645886192,
+ "learning_rate": 2.2034899316058926e-06,
+ "loss": 0.287479043006897,
+ "num_input_tokens_seen": 894979648,
+ "step": 20465,
+ "train_runtime": 158343.8582,
+ "train_tokens_per_second": 5652.127
+ },
+ {
+ "epoch": 1.3782655534608135,
+ "grad_norm": 0.7877267867305047,
+ "learning_rate": 2.201298473894854e-06,
+ "loss": 0.29344854354858396,
+ "num_input_tokens_seen": 895182680,
+ "step": 20470,
+ "train_runtime": 158380.2408,
+ "train_tokens_per_second": 5652.111
+ },
+ {
+ "epoch": 1.3786022084567735,
+ "grad_norm": 0.7361412279331169,
+ "learning_rate": 2.1991077988317723e-06,
+ "loss": 0.30328798294067383,
+ "num_input_tokens_seen": 895414024,
+ "step": 20475,
+ "train_runtime": 158422.7502,
+ "train_tokens_per_second": 5652.055
+ },
+ {
+ "epoch": 1.3789388634527335,
+ "grad_norm": 0.751463034805612,
+ "learning_rate": 2.1969179070292637e-06,
+ "loss": 0.2803157329559326,
+ "num_input_tokens_seen": 895624392,
+ "step": 20480,
+ "train_runtime": 158463.0376,
+ "train_tokens_per_second": 5651.945
+ },
+ {
+ "epoch": 1.3792755184486938,
+ "grad_norm": 0.7411499573407221,
+ "learning_rate": 2.1947287990997236e-06,
+ "loss": 0.3004035711288452,
+ "num_input_tokens_seen": 895840912,
+ "step": 20485,
+ "train_runtime": 158494.4467,
+ "train_tokens_per_second": 5652.191
+ },
+ {
+ "epoch": 1.379612173444654,
+ "grad_norm": 0.7715375986241239,
+ "learning_rate": 2.192540475655328e-06,
+ "loss": 0.299238920211792,
+ "num_input_tokens_seen": 896053216,
+ "step": 20490,
+ "train_runtime": 158532.5086,
+ "train_tokens_per_second": 5652.173
+ },
+ {
+ "epoch": 1.379948828440614,
+ "grad_norm": 0.8053384795009297,
+ "learning_rate": 2.1903529373080397e-06,
+ "loss": 0.3175162076950073,
+ "num_input_tokens_seen": 896270408,
+ "step": 20495,
+ "train_runtime": 158571.7359,
+ "train_tokens_per_second": 5652.145
+ },
+ {
+ "epoch": 1.380285483436574,
+ "grad_norm": 0.8121078413056436,
+ "learning_rate": 2.1881661846695895e-06,
+ "loss": 0.2798121452331543,
+ "num_input_tokens_seen": 896464640,
+ "step": 20500,
+ "train_runtime": 158604.3337,
+ "train_tokens_per_second": 5652.208
+ },
+ {
+ "epoch": 1.3806221384325343,
+ "grad_norm": 0.8616911766005805,
+ "learning_rate": 2.185980218351503e-06,
+ "loss": 0.31658909320831297,
+ "num_input_tokens_seen": 896678032,
+ "step": 20505,
+ "train_runtime": 158643.7892,
+ "train_tokens_per_second": 5652.147
+ },
+ {
+ "epoch": 1.3809587934284946,
+ "grad_norm": 0.7365179371963805,
+ "learning_rate": 2.1837950389650714e-06,
+ "loss": 0.3018453598022461,
+ "num_input_tokens_seen": 896910472,
+ "step": 20510,
+ "train_runtime": 158684.9601,
+ "train_tokens_per_second": 5652.145
+ },
+ {
+ "epoch": 1.3812954484244546,
+ "grad_norm": 0.6736956456373172,
+ "learning_rate": 2.1816106471213787e-06,
+ "loss": 0.30421650409698486,
+ "num_input_tokens_seen": 897142376,
+ "step": 20515,
+ "train_runtime": 158727.7032,
+ "train_tokens_per_second": 5652.084
+ },
+ {
+ "epoch": 1.3816321034204146,
+ "grad_norm": 0.747870924932963,
+ "learning_rate": 2.179427043431282e-06,
+ "loss": 0.32655086517333987,
+ "num_input_tokens_seen": 897368312,
+ "step": 20520,
+ "train_runtime": 158760.2315,
+ "train_tokens_per_second": 5652.349
+ },
+ {
+ "epoch": 1.381968758416375,
+ "grad_norm": 0.697584938785786,
+ "learning_rate": 2.177244228505418e-06,
+ "loss": 0.32046022415161135,
+ "num_input_tokens_seen": 897570856,
+ "step": 20525,
+ "train_runtime": 158800.7072,
+ "train_tokens_per_second": 5652.184
+ },
+ {
+ "epoch": 1.3823054134123351,
+ "grad_norm": 0.8926152387651795,
+ "learning_rate": 2.1750622029542047e-06,
+ "loss": 0.2898330926895142,
+ "num_input_tokens_seen": 897801520,
+ "step": 20530,
+ "train_runtime": 158836.9929,
+ "train_tokens_per_second": 5652.345
+ },
+ {
+ "epoch": 1.3826420684082952,
+ "grad_norm": 0.8132938665366518,
+ "learning_rate": 2.1728809673878374e-06,
+ "loss": 0.31139211654663085,
+ "num_input_tokens_seen": 898013200,
+ "step": 20535,
+ "train_runtime": 158877.4038,
+ "train_tokens_per_second": 5652.24
+ },
+ {
+ "epoch": 1.3829787234042552,
+ "grad_norm": 0.7797244695242583,
+ "learning_rate": 2.1707005224162915e-06,
+ "loss": 0.2919374227523804,
+ "num_input_tokens_seen": 898234880,
+ "step": 20540,
+ "train_runtime": 158916.8206,
+ "train_tokens_per_second": 5652.233
+ },
+ {
+ "epoch": 1.3833153784002155,
+ "grad_norm": 0.7364923903516897,
+ "learning_rate": 2.1685208686493257e-06,
+ "loss": 0.30305407047271726,
+ "num_input_tokens_seen": 898436912,
+ "step": 20545,
+ "train_runtime": 158958.896,
+ "train_tokens_per_second": 5652.008
+ },
+ {
+ "epoch": 1.3836520333961757,
+ "grad_norm": 0.735813175372172,
+ "learning_rate": 2.166342006696467e-06,
+ "loss": 0.30848112106323244,
+ "num_input_tokens_seen": 898651112,
+ "step": 20550,
+ "train_runtime": 158991.467,
+ "train_tokens_per_second": 5652.197
+ },
+ {
+ "epoch": 1.3839886883921357,
+ "grad_norm": 0.7637368092940925,
+ "learning_rate": 2.1641639371670335e-06,
+ "loss": 0.3011596202850342,
+ "num_input_tokens_seen": 898866064,
+ "step": 20555,
+ "train_runtime": 159024.145,
+ "train_tokens_per_second": 5652.387
+ },
+ {
+ "epoch": 1.3843253433880958,
+ "grad_norm": 0.9110668442903609,
+ "learning_rate": 2.161986660670109e-06,
+ "loss": 0.3262584686279297,
+ "num_input_tokens_seen": 899090152,
+ "step": 20560,
+ "train_runtime": 159055.0547,
+ "train_tokens_per_second": 5652.698
+ },
+ {
+ "epoch": 1.384661998384056,
+ "grad_norm": 0.6343085979931745,
+ "learning_rate": 2.159810177814566e-06,
+ "loss": 0.2925774574279785,
+ "num_input_tokens_seen": 899299560,
+ "step": 20565,
+ "train_runtime": 159088.6315,
+ "train_tokens_per_second": 5652.821
+ },
+ {
+ "epoch": 1.3849986533800163,
+ "grad_norm": 0.7193981263314468,
+ "learning_rate": 2.1576344892090507e-06,
+ "loss": 0.3194908618927002,
+ "num_input_tokens_seen": 899514728,
+ "step": 20570,
+ "train_runtime": 159127.1256,
+ "train_tokens_per_second": 5652.806
+ },
+ {
+ "epoch": 1.3853353083759763,
+ "grad_norm": 0.7312989213302724,
+ "learning_rate": 2.1554595954619866e-06,
+ "loss": 0.30469086170196535,
+ "num_input_tokens_seen": 899730168,
+ "step": 20575,
+ "train_runtime": 159168.3547,
+ "train_tokens_per_second": 5652.695
+ },
+ {
+ "epoch": 1.3856719633719363,
+ "grad_norm": 0.7885283712306526,
+ "learning_rate": 2.1532854971815763e-06,
+ "loss": 0.30314340591430666,
+ "num_input_tokens_seen": 899940496,
+ "step": 20580,
+ "train_runtime": 159210.4856,
+ "train_tokens_per_second": 5652.52
+ },
+ {
+ "epoch": 1.3860086183678966,
+ "grad_norm": 0.7120887950890258,
+ "learning_rate": 2.151112194975799e-06,
+ "loss": 0.2792366504669189,
+ "num_input_tokens_seen": 900163032,
+ "step": 20585,
+ "train_runtime": 159256.144,
+ "train_tokens_per_second": 5652.297
+ },
+ {
+ "epoch": 1.3863452733638568,
+ "grad_norm": 0.6730346420964233,
+ "learning_rate": 2.1489396894524106e-06,
+ "loss": 0.2858273506164551,
+ "num_input_tokens_seen": 900392200,
+ "step": 20590,
+ "train_runtime": 159303.4591,
+ "train_tokens_per_second": 5652.057
+ },
+ {
+ "epoch": 1.3866819283598169,
+ "grad_norm": 0.6233675139772793,
+ "learning_rate": 2.1467679812189497e-06,
+ "loss": 0.3015708923339844,
+ "num_input_tokens_seen": 900610928,
+ "step": 20595,
+ "train_runtime": 159340.5828,
+ "train_tokens_per_second": 5652.113
+ },
+ {
+ "epoch": 1.3870185833557769,
+ "grad_norm": 1.0595463617660366,
+ "learning_rate": 2.1445970708827213e-06,
+ "loss": 0.30586059093475343,
+ "num_input_tokens_seen": 900814704,
+ "step": 20600,
+ "train_runtime": 159380.6541,
+ "train_tokens_per_second": 5651.97
+ },
+ {
+ "epoch": 1.3873552383517371,
+ "grad_norm": 0.6602500948938497,
+ "learning_rate": 2.14242695905082e-06,
+ "loss": 0.289659309387207,
+ "num_input_tokens_seen": 901036328,
+ "step": 20605,
+ "train_runtime": 159412.9764,
+ "train_tokens_per_second": 5652.214
+ },
+ {
+ "epoch": 1.3876918933476974,
+ "grad_norm": 0.7593497019073732,
+ "learning_rate": 2.1402576463301044e-06,
+ "loss": 0.3204467535018921,
+ "num_input_tokens_seen": 901271056,
+ "step": 20610,
+ "train_runtime": 159456.2832,
+ "train_tokens_per_second": 5652.151
+ },
+ {
+ "epoch": 1.3880285483436574,
+ "grad_norm": 0.8198802661040533,
+ "learning_rate": 2.13808913332722e-06,
+ "loss": 0.28775959014892577,
+ "num_input_tokens_seen": 901485976,
+ "step": 20615,
+ "train_runtime": 159494.9253,
+ "train_tokens_per_second": 5652.13
+ },
+ {
+ "epoch": 1.3883652033396174,
+ "grad_norm": 0.7717803821263688,
+ "learning_rate": 2.1359214206485845e-06,
+ "loss": 0.32880635261535646,
+ "num_input_tokens_seen": 901703104,
+ "step": 20620,
+ "train_runtime": 159533.389,
+ "train_tokens_per_second": 5652.128
+ },
+ {
+ "epoch": 1.3887018583355777,
+ "grad_norm": 0.7172069769325149,
+ "learning_rate": 2.1337545089003905e-06,
+ "loss": 0.2626586675643921,
+ "num_input_tokens_seen": 901902224,
+ "step": 20625,
+ "train_runtime": 159566.5463,
+ "train_tokens_per_second": 5652.201
+ },
+ {
+ "epoch": 1.389038513331538,
+ "grad_norm": 0.764053491586925,
+ "learning_rate": 2.131588398688608e-06,
+ "loss": 0.31684534549713134,
+ "num_input_tokens_seen": 902125584,
+ "step": 20630,
+ "train_runtime": 159611.5339,
+ "train_tokens_per_second": 5652.007
+ },
+ {
+ "epoch": 1.389375168327498,
+ "grad_norm": 0.7374395690303461,
+ "learning_rate": 2.1294230906189833e-06,
+ "loss": 0.28316431045532225,
+ "num_input_tokens_seen": 902333832,
+ "step": 20635,
+ "train_runtime": 159653.8037,
+ "train_tokens_per_second": 5651.815
+ },
+ {
+ "epoch": 1.389711823323458,
+ "grad_norm": 0.7355709013834859,
+ "learning_rate": 2.1272585852970358e-06,
+ "loss": 0.29244768619537354,
+ "num_input_tokens_seen": 902541064,
+ "step": 20640,
+ "train_runtime": 159692.4958,
+ "train_tokens_per_second": 5651.744
+ },
+ {
+ "epoch": 1.3900484783194182,
+ "grad_norm": 0.7160331969470104,
+ "learning_rate": 2.125094883328068e-06,
+ "loss": 0.2933189392089844,
+ "num_input_tokens_seen": 902745264,
+ "step": 20645,
+ "train_runtime": 159733.1324,
+ "train_tokens_per_second": 5651.584
+ },
+ {
+ "epoch": 1.3903851333153785,
+ "grad_norm": 0.8057277953126102,
+ "learning_rate": 2.1229319853171447e-06,
+ "loss": 0.2766158103942871,
+ "num_input_tokens_seen": 902957784,
+ "step": 20650,
+ "train_runtime": 159767.4906,
+ "train_tokens_per_second": 5651.699
+ },
+ {
+ "epoch": 1.3907217883113385,
+ "grad_norm": 0.8447973197859374,
+ "learning_rate": 2.1207698918691217e-06,
+ "loss": 0.3175971508026123,
+ "num_input_tokens_seen": 903176368,
+ "step": 20655,
+ "train_runtime": 159809.1069,
+ "train_tokens_per_second": 5651.595
+ },
+ {
+ "epoch": 1.3910584433072986,
+ "grad_norm": 0.7509364554421408,
+ "learning_rate": 2.1186086035886133e-06,
+ "loss": 0.28777422904968264,
+ "num_input_tokens_seen": 903403424,
+ "step": 20660,
+ "train_runtime": 159847.5357,
+ "train_tokens_per_second": 5651.657
+ },
+ {
+ "epoch": 1.3913950983032588,
+ "grad_norm": 0.7719727435620117,
+ "learning_rate": 2.116448121080023e-06,
+ "loss": 0.309741735458374,
+ "num_input_tokens_seen": 903643016,
+ "step": 20665,
+ "train_runtime": 159888.1841,
+ "train_tokens_per_second": 5651.719
+ },
+ {
+ "epoch": 1.391731753299219,
+ "grad_norm": 0.7650108977382473,
+ "learning_rate": 2.1142884449475205e-06,
+ "loss": 0.30854334831237795,
+ "num_input_tokens_seen": 903861496,
+ "step": 20670,
+ "train_runtime": 159925.9335,
+ "train_tokens_per_second": 5651.751
+ },
+ {
+ "epoch": 1.392068408295179,
+ "grad_norm": 0.8857585605046503,
+ "learning_rate": 2.112129575795052e-06,
+ "loss": 0.2980179309844971,
+ "num_input_tokens_seen": 904082696,
+ "step": 20675,
+ "train_runtime": 159970.8097,
+ "train_tokens_per_second": 5651.548
+ },
+ {
+ "epoch": 1.3924050632911391,
+ "grad_norm": 0.6579126875016472,
+ "learning_rate": 2.1099715142263394e-06,
+ "loss": 0.2740116357803345,
+ "num_input_tokens_seen": 904295832,
+ "step": 20680,
+ "train_runtime": 160012.6441,
+ "train_tokens_per_second": 5651.402
+ },
+ {
+ "epoch": 1.3927417182870994,
+ "grad_norm": 0.7692183867198543,
+ "learning_rate": 2.1078142608448773e-06,
+ "loss": 0.28600759506225587,
+ "num_input_tokens_seen": 904516064,
+ "step": 20685,
+ "train_runtime": 160046.5372,
+ "train_tokens_per_second": 5651.582
+ },
+ {
+ "epoch": 1.3930783732830596,
+ "grad_norm": 0.8142763298016777,
+ "learning_rate": 2.1056578162539327e-06,
+ "loss": 0.32352604866027834,
+ "num_input_tokens_seen": 904737152,
+ "step": 20690,
+ "train_runtime": 160088.4432,
+ "train_tokens_per_second": 5651.483
+ },
+ {
+ "epoch": 1.3934150282790196,
+ "grad_norm": 0.871875644642827,
+ "learning_rate": 2.103502181056554e-06,
+ "loss": 0.30927181243896484,
+ "num_input_tokens_seen": 904951984,
+ "step": 20695,
+ "train_runtime": 160130.5591,
+ "train_tokens_per_second": 5651.338
+ },
+ {
+ "epoch": 1.3937516832749797,
+ "grad_norm": 0.704322984008134,
+ "learning_rate": 2.1013473558555503e-06,
+ "loss": 0.2932461738586426,
+ "num_input_tokens_seen": 905188432,
+ "step": 20700,
+ "train_runtime": 160177.8633,
+ "train_tokens_per_second": 5651.146
+ },
+ {
+ "epoch": 1.39408833827094,
+ "grad_norm": 0.7189726909358929,
+ "learning_rate": 2.099193341253519e-06,
+ "loss": 0.29349417686462403,
+ "num_input_tokens_seen": 905395280,
+ "step": 20705,
+ "train_runtime": 160214.1418,
+ "train_tokens_per_second": 5651.157
+ },
+ {
+ "epoch": 1.3944249932669002,
+ "grad_norm": 0.6639597772905609,
+ "learning_rate": 2.0970401378528167e-06,
+ "loss": 0.286403751373291,
+ "num_input_tokens_seen": 905601488,
+ "step": 20710,
+ "train_runtime": 160261.6101,
+ "train_tokens_per_second": 5650.77
+ },
+ {
+ "epoch": 1.3947616482628602,
+ "grad_norm": 0.702286400471809,
+ "learning_rate": 2.0948877462555844e-06,
+ "loss": 0.2749554872512817,
+ "num_input_tokens_seen": 905827264,
+ "step": 20715,
+ "train_runtime": 160304.5265,
+ "train_tokens_per_second": 5650.666
+ },
+ {
+ "epoch": 1.3950983032588202,
+ "grad_norm": 0.7033676607004119,
+ "learning_rate": 2.09273616706373e-06,
+ "loss": 0.32062878608703616,
+ "num_input_tokens_seen": 906057568,
+ "step": 20720,
+ "train_runtime": 160337.276,
+ "train_tokens_per_second": 5650.948
+ },
+ {
+ "epoch": 1.3954349582547805,
+ "grad_norm": 0.8756497939875741,
+ "learning_rate": 2.090585400878936e-06,
+ "loss": 0.3149780035018921,
+ "num_input_tokens_seen": 906277360,
+ "step": 20725,
+ "train_runtime": 160380.6002,
+ "train_tokens_per_second": 5650.792
+ },
+ {
+ "epoch": 1.3957716132507407,
+ "grad_norm": 0.7980623610419325,
+ "learning_rate": 2.0884354483026576e-06,
+ "loss": 0.3101172924041748,
+ "num_input_tokens_seen": 906499640,
+ "step": 20730,
+ "train_runtime": 160413.5552,
+ "train_tokens_per_second": 5651.016
+ },
+ {
+ "epoch": 1.3961082682467008,
+ "grad_norm": 0.6980186229936438,
+ "learning_rate": 2.0862863099361214e-06,
+ "loss": 0.30629117488861085,
+ "num_input_tokens_seen": 906734072,
+ "step": 20735,
+ "train_runtime": 160445.4472,
+ "train_tokens_per_second": 5651.354
+ },
+ {
+ "epoch": 1.3964449232426608,
+ "grad_norm": 0.7214735662789337,
+ "learning_rate": 2.084137986380326e-06,
+ "loss": 0.3239948272705078,
+ "num_input_tokens_seen": 906945840,
+ "step": 20740,
+ "train_runtime": 160486.1913,
+ "train_tokens_per_second": 5651.239
+ },
+ {
+ "epoch": 1.396781578238621,
+ "grad_norm": 0.7503921044839111,
+ "learning_rate": 2.081990478236048e-06,
+ "loss": 0.28044352531433103,
+ "num_input_tokens_seen": 907146456,
+ "step": 20745,
+ "train_runtime": 160522.0232,
+ "train_tokens_per_second": 5651.227
+ },
+ {
+ "epoch": 1.3971182332345813,
+ "grad_norm": 0.7170605848167944,
+ "learning_rate": 2.079843786103824e-06,
+ "loss": 0.2951197624206543,
+ "num_input_tokens_seen": 907356536,
+ "step": 20750,
+ "train_runtime": 160565.1412,
+ "train_tokens_per_second": 5651.018
+ },
+ {
+ "epoch": 1.3974548882305413,
+ "grad_norm": 0.7858210180931035,
+ "learning_rate": 2.0776979105839774e-06,
+ "loss": 0.31067094802856443,
+ "num_input_tokens_seen": 907585320,
+ "step": 20755,
+ "train_runtime": 160605.9994,
+ "train_tokens_per_second": 5651.005
+ },
+ {
+ "epoch": 1.3977915432265013,
+ "grad_norm": 0.811518268399912,
+ "learning_rate": 2.075552852276588e-06,
+ "loss": 0.33663692474365237,
+ "num_input_tokens_seen": 907812376,
+ "step": 20760,
+ "train_runtime": 160647.5068,
+ "train_tokens_per_second": 5650.958
+ },
+ {
+ "epoch": 1.3981281982224616,
+ "grad_norm": 0.843500732222522,
+ "learning_rate": 2.07340861178152e-06,
+ "loss": 0.3083709716796875,
+ "num_input_tokens_seen": 908030480,
+ "step": 20765,
+ "train_runtime": 160684.2393,
+ "train_tokens_per_second": 5651.024
+ },
+ {
+ "epoch": 1.3984648532184218,
+ "grad_norm": 0.7409888586409921,
+ "learning_rate": 2.0712651896984013e-06,
+ "loss": 0.30435686111450194,
+ "num_input_tokens_seen": 908257064,
+ "step": 20770,
+ "train_runtime": 160718.9627,
+ "train_tokens_per_second": 5651.213
+ },
+ {
+ "epoch": 1.3988015082143819,
+ "grad_norm": 0.7528366267193225,
+ "learning_rate": 2.0691225866266335e-06,
+ "loss": 0.3061100244522095,
+ "num_input_tokens_seen": 908466856,
+ "step": 20775,
+ "train_runtime": 160758.0637,
+ "train_tokens_per_second": 5651.143
+ },
+ {
+ "epoch": 1.3991381632103421,
+ "grad_norm": 0.8274013015791201,
+ "learning_rate": 2.066980803165388e-06,
+ "loss": 0.3086406707763672,
+ "num_input_tokens_seen": 908690664,
+ "step": 20780,
+ "train_runtime": 160789.4191,
+ "train_tokens_per_second": 5651.433
+ },
+ {
+ "epoch": 1.3994748182063022,
+ "grad_norm": 0.7972254073123841,
+ "learning_rate": 2.0648398399136088e-06,
+ "loss": 0.30843343734741213,
+ "num_input_tokens_seen": 908916400,
+ "step": 20785,
+ "train_runtime": 160833.2267,
+ "train_tokens_per_second": 5651.297
+ },
+ {
+ "epoch": 1.3998114732022624,
+ "grad_norm": 0.7390702899298262,
+ "learning_rate": 2.062699697470007e-06,
+ "loss": 0.295151948928833,
+ "num_input_tokens_seen": 909128960,
+ "step": 20790,
+ "train_runtime": 160877.6756,
+ "train_tokens_per_second": 5651.057
+ },
+ {
+ "epoch": 1.4001481281982224,
+ "grad_norm": 0.6966665870930083,
+ "learning_rate": 2.060560376433072e-06,
+ "loss": 0.33457579612731936,
+ "num_input_tokens_seen": 909357768,
+ "step": 20795,
+ "train_runtime": 160920.3143,
+ "train_tokens_per_second": 5650.982
+ },
+ {
+ "epoch": 1.4004847831941827,
+ "grad_norm": 0.6567105821545446,
+ "learning_rate": 2.05842187740105e-06,
+ "loss": 0.3098001956939697,
+ "num_input_tokens_seen": 909572480,
+ "step": 20800,
+ "train_runtime": 160963.4825,
+ "train_tokens_per_second": 5650.8
+ },
+ {
+ "epoch": 1.4008214381901427,
+ "grad_norm": 0.8379053674392297,
+ "learning_rate": 2.0562842009719745e-06,
+ "loss": 0.2820146560668945,
+ "num_input_tokens_seen": 909757296,
+ "step": 20805,
+ "train_runtime": 161008.9129,
+ "train_tokens_per_second": 5650.354
+ },
+ {
+ "epoch": 1.401158093186103,
+ "grad_norm": 0.8065231699303244,
+ "learning_rate": 2.0541473477436317e-06,
+ "loss": 0.2844945669174194,
+ "num_input_tokens_seen": 909975984,
+ "step": 20810,
+ "train_runtime": 161043.343,
+ "train_tokens_per_second": 5650.504
+ },
+ {
+ "epoch": 1.401494748182063,
+ "grad_norm": 0.7694176813911128,
+ "learning_rate": 2.052011318313591e-06,
+ "loss": 0.28799991607666015,
+ "num_input_tokens_seen": 910199280,
+ "step": 20815,
+ "train_runtime": 161080.8517,
+ "train_tokens_per_second": 5650.574
+ },
+ {
+ "epoch": 1.4018314031780232,
+ "grad_norm": 0.7254285849515797,
+ "learning_rate": 2.049876113279185e-06,
+ "loss": 0.32419145107269287,
+ "num_input_tokens_seen": 910419968,
+ "step": 20820,
+ "train_runtime": 161115.1641,
+ "train_tokens_per_second": 5650.74
+ },
+ {
+ "epoch": 1.4021680581739833,
+ "grad_norm": 0.663299004599166,
+ "learning_rate": 2.0477417332375165e-06,
+ "loss": 0.28672337532043457,
+ "num_input_tokens_seen": 910648216,
+ "step": 20825,
+ "train_runtime": 161146.0692,
+ "train_tokens_per_second": 5651.073
+ },
+ {
+ "epoch": 1.4025047131699435,
+ "grad_norm": 0.7218766432038832,
+ "learning_rate": 2.045608178785458e-06,
+ "loss": 0.29271349906921384,
+ "num_input_tokens_seen": 910871288,
+ "step": 20830,
+ "train_runtime": 161185.4894,
+ "train_tokens_per_second": 5651.075
+ },
+ {
+ "epoch": 1.4028413681659035,
+ "grad_norm": 0.670423396219908,
+ "learning_rate": 2.043475450519652e-06,
+ "loss": 0.29650259017944336,
+ "num_input_tokens_seen": 911087016,
+ "step": 20835,
+ "train_runtime": 161217.0941,
+ "train_tokens_per_second": 5651.305
+ },
+ {
+ "epoch": 1.4031780231618638,
+ "grad_norm": 0.7755384152878778,
+ "learning_rate": 2.0413435490365064e-06,
+ "loss": 0.31277635097503664,
+ "num_input_tokens_seen": 911295624,
+ "step": 20840,
+ "train_runtime": 161255.7945,
+ "train_tokens_per_second": 5651.243
+ },
+ {
+ "epoch": 1.4035146781578238,
+ "grad_norm": 0.7431255146543578,
+ "learning_rate": 2.0392124749322064e-06,
+ "loss": 0.31907684803009034,
+ "num_input_tokens_seen": 911522672,
+ "step": 20845,
+ "train_runtime": 161298.269,
+ "train_tokens_per_second": 5651.162
+ },
+ {
+ "epoch": 1.403851333153784,
+ "grad_norm": 0.8355426134565388,
+ "learning_rate": 2.037082228802693e-06,
+ "loss": 0.28636276721954346,
+ "num_input_tokens_seen": 911742024,
+ "step": 20850,
+ "train_runtime": 161332.2795,
+ "train_tokens_per_second": 5651.33
+ },
+ {
+ "epoch": 1.404187988149744,
+ "grad_norm": 0.7290848719366411,
+ "learning_rate": 2.03495281124369e-06,
+ "loss": 0.279879093170166,
+ "num_input_tokens_seen": 911958464,
+ "step": 20855,
+ "train_runtime": 161371.514,
+ "train_tokens_per_second": 5651.298
+ },
+ {
+ "epoch": 1.4045246431457044,
+ "grad_norm": 0.7216593366596518,
+ "learning_rate": 2.0328242228506746e-06,
+ "loss": 0.28923535346984863,
+ "num_input_tokens_seen": 912172576,
+ "step": 20860,
+ "train_runtime": 161410.452,
+ "train_tokens_per_second": 5651.261
+ },
+ {
+ "epoch": 1.4048612981416644,
+ "grad_norm": 0.7898442019168748,
+ "learning_rate": 2.0306964642189053e-06,
+ "loss": 0.3036557674407959,
+ "num_input_tokens_seen": 912389440,
+ "step": 20865,
+ "train_runtime": 161449.3775,
+ "train_tokens_per_second": 5651.242
+ },
+ {
+ "epoch": 1.4051979531376246,
+ "grad_norm": 0.7311964977804852,
+ "learning_rate": 2.028569535943402e-06,
+ "loss": 0.2902818202972412,
+ "num_input_tokens_seen": 912599400,
+ "step": 20870,
+ "train_runtime": 161487.5229,
+ "train_tokens_per_second": 5651.207
+ },
+ {
+ "epoch": 1.4055346081335847,
+ "grad_norm": 0.7771281147529382,
+ "learning_rate": 2.0264434386189524e-06,
+ "loss": 0.30182933807373047,
+ "num_input_tokens_seen": 912816512,
+ "step": 20875,
+ "train_runtime": 161522.8112,
+ "train_tokens_per_second": 5651.316
+ },
+ {
+ "epoch": 1.405871263129545,
+ "grad_norm": 0.6711489743793063,
+ "learning_rate": 2.0243181728401136e-06,
+ "loss": 0.29143590927124025,
+ "num_input_tokens_seen": 913038160,
+ "step": 20880,
+ "train_runtime": 161565.9228,
+ "train_tokens_per_second": 5651.18
+ },
+ {
+ "epoch": 1.406207918125505,
+ "grad_norm": 0.7946566060485273,
+ "learning_rate": 2.022193739201208e-06,
+ "loss": 0.31396236419677737,
+ "num_input_tokens_seen": 913238344,
+ "step": 20885,
+ "train_runtime": 161599.9418,
+ "train_tokens_per_second": 5651.229
+ },
+ {
+ "epoch": 1.4065445731214652,
+ "grad_norm": 0.7053560827598276,
+ "learning_rate": 2.0200701382963267e-06,
+ "loss": 0.2820024013519287,
+ "num_input_tokens_seen": 913456008,
+ "step": 20890,
+ "train_runtime": 161640.9986,
+ "train_tokens_per_second": 5651.141
+ },
+ {
+ "epoch": 1.4068812281174252,
+ "grad_norm": 0.8293767478590064,
+ "learning_rate": 2.0179473707193324e-06,
+ "loss": 0.2967421531677246,
+ "num_input_tokens_seen": 913682016,
+ "step": 20895,
+ "train_runtime": 161681.1903,
+ "train_tokens_per_second": 5651.134
+ },
+ {
+ "epoch": 1.4072178831133855,
+ "grad_norm": 0.681242664001568,
+ "learning_rate": 2.0158254370638435e-06,
+ "loss": 0.30292630195617676,
+ "num_input_tokens_seen": 913903464,
+ "step": 20900,
+ "train_runtime": 161723.6924,
+ "train_tokens_per_second": 5651.018
+ },
+ {
+ "epoch": 1.4075545381093455,
+ "grad_norm": 0.6662676288835188,
+ "learning_rate": 2.0137043379232585e-06,
+ "loss": 0.3011344909667969,
+ "num_input_tokens_seen": 914128144,
+ "step": 20905,
+ "train_runtime": 161765.7508,
+ "train_tokens_per_second": 5650.937
+ },
+ {
+ "epoch": 1.4078911931053057,
+ "grad_norm": 0.7275678076333139,
+ "learning_rate": 2.0115840738907293e-06,
+ "loss": 0.2961131572723389,
+ "num_input_tokens_seen": 914332560,
+ "step": 20910,
+ "train_runtime": 161803.517,
+ "train_tokens_per_second": 5650.882
+ },
+ {
+ "epoch": 1.4082278481012658,
+ "grad_norm": 0.8438338939411848,
+ "learning_rate": 2.0094646455591865e-06,
+ "loss": 0.31174521446228026,
+ "num_input_tokens_seen": 914554936,
+ "step": 20915,
+ "train_runtime": 161846.6205,
+ "train_tokens_per_second": 5650.751
+ },
+ {
+ "epoch": 1.408564503097226,
+ "grad_norm": 0.7029690243486251,
+ "learning_rate": 2.007346053521319e-06,
+ "loss": 0.3245561599731445,
+ "num_input_tokens_seen": 914776376,
+ "step": 20920,
+ "train_runtime": 161885.6534,
+ "train_tokens_per_second": 5650.756
+ },
+ {
+ "epoch": 1.408901158093186,
+ "grad_norm": 0.7673612026423596,
+ "learning_rate": 2.005228298369583e-06,
+ "loss": 0.3163960695266724,
+ "num_input_tokens_seen": 914979152,
+ "step": 20925,
+ "train_runtime": 161924.9473,
+ "train_tokens_per_second": 5650.637
+ },
+ {
+ "epoch": 1.4092378130891463,
+ "grad_norm": 0.8203680546301054,
+ "learning_rate": 2.003111380696208e-06,
+ "loss": 0.282145619392395,
+ "num_input_tokens_seen": 915204416,
+ "step": 20930,
+ "train_runtime": 161972.9634,
+ "train_tokens_per_second": 5650.353
+ },
+ {
+ "epoch": 1.4095744680851063,
+ "grad_norm": 0.826854532975397,
+ "learning_rate": 2.0009953010931747e-06,
+ "loss": 0.2879758834838867,
+ "num_input_tokens_seen": 915424616,
+ "step": 20935,
+ "train_runtime": 162018.6585,
+ "train_tokens_per_second": 5650.118
+ },
+ {
+ "epoch": 1.4099111230810666,
+ "grad_norm": 0.8165582356041448,
+ "learning_rate": 1.998880060152244e-06,
+ "loss": 0.32476277351379396,
+ "num_input_tokens_seen": 915651288,
+ "step": 20940,
+ "train_runtime": 162057.5342,
+ "train_tokens_per_second": 5650.162
+ },
+ {
+ "epoch": 1.4102477780770266,
+ "grad_norm": 0.7599467895290581,
+ "learning_rate": 1.9967656584649354e-06,
+ "loss": 0.3187078952789307,
+ "num_input_tokens_seen": 915870040,
+ "step": 20945,
+ "train_runtime": 162094.7907,
+ "train_tokens_per_second": 5650.213
+ },
+ {
+ "epoch": 1.4105844330729869,
+ "grad_norm": 0.7168782752872359,
+ "learning_rate": 1.994652096622533e-06,
+ "loss": 0.3237470149993896,
+ "num_input_tokens_seen": 916090120,
+ "step": 20950,
+ "train_runtime": 162135.1509,
+ "train_tokens_per_second": 5650.164
+ },
+ {
+ "epoch": 1.410921088068947,
+ "grad_norm": 0.8403414065661872,
+ "learning_rate": 1.992539375216089e-06,
+ "loss": 0.3227944612503052,
+ "num_input_tokens_seen": 916315064,
+ "step": 20955,
+ "train_runtime": 162172.94,
+ "train_tokens_per_second": 5650.234
+ },
+ {
+ "epoch": 1.4112577430649071,
+ "grad_norm": 0.7048786926884193,
+ "learning_rate": 1.9904274948364184e-06,
+ "loss": 0.30767481327056884,
+ "num_input_tokens_seen": 916539152,
+ "step": 20960,
+ "train_runtime": 162213.0654,
+ "train_tokens_per_second": 5650.218
+ },
+ {
+ "epoch": 1.4115943980608672,
+ "grad_norm": 0.7655887084377769,
+ "learning_rate": 1.988316456074102e-06,
+ "loss": 0.30903284549713134,
+ "num_input_tokens_seen": 916766256,
+ "step": 20965,
+ "train_runtime": 162250.1786,
+ "train_tokens_per_second": 5650.325
+ },
+ {
+ "epoch": 1.4119310530568274,
+ "grad_norm": 0.7205405993664067,
+ "learning_rate": 1.9862062595194854e-06,
+ "loss": 0.31888184547424314,
+ "num_input_tokens_seen": 916978512,
+ "step": 20970,
+ "train_runtime": 162288.8233,
+ "train_tokens_per_second": 5650.288
+ },
+ {
+ "epoch": 1.4122677080527875,
+ "grad_norm": 0.8681175003128851,
+ "learning_rate": 1.984096905762676e-06,
+ "loss": 0.2966637372970581,
+ "num_input_tokens_seen": 917190424,
+ "step": 20975,
+ "train_runtime": 162331.7616,
+ "train_tokens_per_second": 5650.098
+ },
+ {
+ "epoch": 1.4126043630487477,
+ "grad_norm": 0.7477822351336473,
+ "learning_rate": 1.9819883953935533e-06,
+ "loss": 0.3042705535888672,
+ "num_input_tokens_seen": 917420304,
+ "step": 20980,
+ "train_runtime": 162368.0639,
+ "train_tokens_per_second": 5650.251
+ },
+ {
+ "epoch": 1.4129410180447077,
+ "grad_norm": 0.8100441236586595,
+ "learning_rate": 1.9798807290017485e-06,
+ "loss": 0.2959493637084961,
+ "num_input_tokens_seen": 917628664,
+ "step": 20985,
+ "train_runtime": 162403.0968,
+ "train_tokens_per_second": 5650.315
+ },
+ {
+ "epoch": 1.413277673040668,
+ "grad_norm": 0.7188686347683017,
+ "learning_rate": 1.9777739071766693e-06,
+ "loss": 0.30003719329833983,
+ "num_input_tokens_seen": 917841064,
+ "step": 20990,
+ "train_runtime": 162445.3947,
+ "train_tokens_per_second": 5650.151
+ },
+ {
+ "epoch": 1.413614328036628,
+ "grad_norm": 0.7153354531724283,
+ "learning_rate": 1.97566793050748e-06,
+ "loss": 0.2987996578216553,
+ "num_input_tokens_seen": 918043080,
+ "step": 20995,
+ "train_runtime": 162480.8482,
+ "train_tokens_per_second": 5650.162
+ },
+ {
+ "epoch": 1.4139509830325883,
+ "grad_norm": 0.8260680916935863,
+ "learning_rate": 1.9735627995831096e-06,
+ "loss": 0.2807154178619385,
+ "num_input_tokens_seen": 918268936,
+ "step": 21000,
+ "train_runtime": 162511.5804,
+ "train_tokens_per_second": 5650.483
+ },
+ {
+ "epoch": 1.4142876380285483,
+ "grad_norm": 0.8413087670817464,
+ "learning_rate": 1.9714585149922523e-06,
+ "loss": 0.322419261932373,
+ "num_input_tokens_seen": 918495704,
+ "step": 21005,
+ "train_runtime": 162548.3519,
+ "train_tokens_per_second": 5650.6
+ },
+ {
+ "epoch": 1.4146242930245085,
+ "grad_norm": 0.7701984865141125,
+ "learning_rate": 1.9693550773233634e-06,
+ "loss": 0.2878579139709473,
+ "num_input_tokens_seen": 918721536,
+ "step": 21010,
+ "train_runtime": 162584.2427,
+ "train_tokens_per_second": 5650.742
+ },
+ {
+ "epoch": 1.4149609480204686,
+ "grad_norm": 0.7629047291809269,
+ "learning_rate": 1.967252487164663e-06,
+ "loss": 0.27617802619934084,
+ "num_input_tokens_seen": 918949576,
+ "step": 21015,
+ "train_runtime": 162625.7531,
+ "train_tokens_per_second": 5650.701
+ },
+ {
+ "epoch": 1.4152976030164288,
+ "grad_norm": 0.8987812118267688,
+ "learning_rate": 1.9651507451041344e-06,
+ "loss": 0.31108310222625735,
+ "num_input_tokens_seen": 919161544,
+ "step": 21020,
+ "train_runtime": 162664.4459,
+ "train_tokens_per_second": 5650.66
+ },
+ {
+ "epoch": 1.4156342580123888,
+ "grad_norm": 0.7347239736627051,
+ "learning_rate": 1.9630498517295204e-06,
+ "loss": 0.3056536436080933,
+ "num_input_tokens_seen": 919390784,
+ "step": 21025,
+ "train_runtime": 162697.0609,
+ "train_tokens_per_second": 5650.937
+ },
+ {
+ "epoch": 1.415970913008349,
+ "grad_norm": 0.7770976786280579,
+ "learning_rate": 1.960949807628335e-06,
+ "loss": 0.31153440475463867,
+ "num_input_tokens_seen": 919610680,
+ "step": 21030,
+ "train_runtime": 162731.7215,
+ "train_tokens_per_second": 5651.084
+ },
+ {
+ "epoch": 1.4163075680043091,
+ "grad_norm": 0.797661242102853,
+ "learning_rate": 1.9588506133878415e-06,
+ "loss": 0.33806021213531495,
+ "num_input_tokens_seen": 919830696,
+ "step": 21035,
+ "train_runtime": 162768.2447,
+ "train_tokens_per_second": 5651.168
+ },
+ {
+ "epoch": 1.4166442230002694,
+ "grad_norm": 0.7171305458599561,
+ "learning_rate": 1.956752269595078e-06,
+ "loss": 0.31885838508605957,
+ "num_input_tokens_seen": 920055416,
+ "step": 21040,
+ "train_runtime": 162811.2347,
+ "train_tokens_per_second": 5651.056
+ },
+ {
+ "epoch": 1.4169808779962294,
+ "grad_norm": 0.795769696712416,
+ "learning_rate": 1.9546547768368384e-06,
+ "loss": 0.3016183853149414,
+ "num_input_tokens_seen": 920286312,
+ "step": 21045,
+ "train_runtime": 162846.1897,
+ "train_tokens_per_second": 5651.261
+ },
+ {
+ "epoch": 1.4173175329921897,
+ "grad_norm": 0.9186674926118773,
+ "learning_rate": 1.9525581356996793e-06,
+ "loss": 0.28722732067108153,
+ "num_input_tokens_seen": 920500576,
+ "step": 21050,
+ "train_runtime": 162885.5207,
+ "train_tokens_per_second": 5651.212
+ },
+ {
+ "epoch": 1.4176541879881497,
+ "grad_norm": 0.7889483479656868,
+ "learning_rate": 1.95046234676992e-06,
+ "loss": 0.30141215324401854,
+ "num_input_tokens_seen": 920738648,
+ "step": 21055,
+ "train_runtime": 162931.4756,
+ "train_tokens_per_second": 5651.079
+ },
+ {
+ "epoch": 1.41799084298411,
+ "grad_norm": 0.8574903736733213,
+ "learning_rate": 1.9483674106336416e-06,
+ "loss": 0.2801323413848877,
+ "num_input_tokens_seen": 920949800,
+ "step": 21060,
+ "train_runtime": 162973.0326,
+ "train_tokens_per_second": 5650.934
+ },
+ {
+ "epoch": 1.41832749798007,
+ "grad_norm": 0.7178113150312732,
+ "learning_rate": 1.9462733278766853e-06,
+ "loss": 0.2972491025924683,
+ "num_input_tokens_seen": 921159152,
+ "step": 21065,
+ "train_runtime": 163013.3989,
+ "train_tokens_per_second": 5650.819
+ },
+ {
+ "epoch": 1.4186641529760302,
+ "grad_norm": 0.704039276053053,
+ "learning_rate": 1.9441800990846553e-06,
+ "loss": 0.29517397880554197,
+ "num_input_tokens_seen": 921390072,
+ "step": 21070,
+ "train_runtime": 163044.8203,
+ "train_tokens_per_second": 5651.146
+ },
+ {
+ "epoch": 1.4190008079719902,
+ "grad_norm": 0.9714518989901114,
+ "learning_rate": 1.942087724842915e-06,
+ "loss": 0.3248338460922241,
+ "num_input_tokens_seen": 921598928,
+ "step": 21075,
+ "train_runtime": 163079.8467,
+ "train_tokens_per_second": 5651.213
+ },
+ {
+ "epoch": 1.4193374629679505,
+ "grad_norm": 0.8145914643998322,
+ "learning_rate": 1.9399962057365944e-06,
+ "loss": 0.30252177715301515,
+ "num_input_tokens_seen": 921803544,
+ "step": 21080,
+ "train_runtime": 163114.8336,
+ "train_tokens_per_second": 5651.255
+ },
+ {
+ "epoch": 1.4196741179639105,
+ "grad_norm": 0.7419159789443908,
+ "learning_rate": 1.937905542350574e-06,
+ "loss": 0.32372136116027833,
+ "num_input_tokens_seen": 922027736,
+ "step": 21085,
+ "train_runtime": 163156.4557,
+ "train_tokens_per_second": 5651.188
+ },
+ {
+ "epoch": 1.4200107729598708,
+ "grad_norm": 0.7910617461936362,
+ "learning_rate": 1.935815735269506e-06,
+ "loss": 0.30223875045776366,
+ "num_input_tokens_seen": 922247592,
+ "step": 21090,
+ "train_runtime": 163190.2127,
+ "train_tokens_per_second": 5651.366
+ },
+ {
+ "epoch": 1.4203474279558308,
+ "grad_norm": 0.8193345009678259,
+ "learning_rate": 1.933726785077797e-06,
+ "loss": 0.3228907108306885,
+ "num_input_tokens_seen": 922462256,
+ "step": 21095,
+ "train_runtime": 163230.7674,
+ "train_tokens_per_second": 5651.277
+ },
+ {
+ "epoch": 1.420684082951791,
+ "grad_norm": 0.8060301943601257,
+ "learning_rate": 1.9316386923596146e-06,
+ "loss": 0.3193122386932373,
+ "num_input_tokens_seen": 922683024,
+ "step": 21100,
+ "train_runtime": 163264.9379,
+ "train_tokens_per_second": 5651.446
+ },
+ {
+ "epoch": 1.421020737947751,
+ "grad_norm": 0.9170237047964257,
+ "learning_rate": 1.929551457698888e-06,
+ "loss": 0.31147422790527346,
+ "num_input_tokens_seen": 922907808,
+ "step": 21105,
+ "train_runtime": 163294.9124,
+ "train_tokens_per_second": 5651.785
+ },
+ {
+ "epoch": 1.4213573929437113,
+ "grad_norm": 0.7054636815045637,
+ "learning_rate": 1.9274650816793044e-06,
+ "loss": 0.3035153388977051,
+ "num_input_tokens_seen": 923140672,
+ "step": 21110,
+ "train_runtime": 163327.9489,
+ "train_tokens_per_second": 5652.068
+ },
+ {
+ "epoch": 1.4216940479396714,
+ "grad_norm": 0.7667361817429569,
+ "learning_rate": 1.9253795648843136e-06,
+ "loss": 0.3144373893737793,
+ "num_input_tokens_seen": 923369128,
+ "step": 21115,
+ "train_runtime": 163362.795,
+ "train_tokens_per_second": 5652.261
+ },
+ {
+ "epoch": 1.4220307029356316,
+ "grad_norm": 0.8143437243469857,
+ "learning_rate": 1.9232949078971236e-06,
+ "loss": 0.3086548805236816,
+ "num_input_tokens_seen": 923585136,
+ "step": 21120,
+ "train_runtime": 163401.3138,
+ "train_tokens_per_second": 5652.25
+ },
+ {
+ "epoch": 1.4223673579315916,
+ "grad_norm": 0.6659717947233721,
+ "learning_rate": 1.9212111113006998e-06,
+ "loss": 0.28868756294250486,
+ "num_input_tokens_seen": 923819592,
+ "step": 21125,
+ "train_runtime": 163444.6265,
+ "train_tokens_per_second": 5652.187
+ },
+ {
+ "epoch": 1.4227040129275519,
+ "grad_norm": 0.8511283738043982,
+ "learning_rate": 1.919128175677775e-06,
+ "loss": 0.2978044986724854,
+ "num_input_tokens_seen": 924024304,
+ "step": 21130,
+ "train_runtime": 163482.4066,
+ "train_tokens_per_second": 5652.133
+ },
+ {
+ "epoch": 1.423040667923512,
+ "grad_norm": 0.7092905997999482,
+ "learning_rate": 1.917046101610827e-06,
+ "loss": 0.30063571929931643,
+ "num_input_tokens_seen": 924252928,
+ "step": 21135,
+ "train_runtime": 163517.7525,
+ "train_tokens_per_second": 5652.309
+ },
+ {
+ "epoch": 1.4233773229194722,
+ "grad_norm": 0.6893485352461994,
+ "learning_rate": 1.9149648896821087e-06,
+ "loss": 0.3125055551528931,
+ "num_input_tokens_seen": 924467024,
+ "step": 21140,
+ "train_runtime": 163562.0209,
+ "train_tokens_per_second": 5652.089
+ },
+ {
+ "epoch": 1.4237139779154322,
+ "grad_norm": 0.8321835207068441,
+ "learning_rate": 1.912884540473621e-06,
+ "loss": 0.2941263675689697,
+ "num_input_tokens_seen": 924669616,
+ "step": 21145,
+ "train_runtime": 163600.3041,
+ "train_tokens_per_second": 5652.004
+ },
+ {
+ "epoch": 1.4240506329113924,
+ "grad_norm": 0.8081464869797703,
+ "learning_rate": 1.9108050545671283e-06,
+ "loss": 0.27542452812194823,
+ "num_input_tokens_seen": 924875728,
+ "step": 21150,
+ "train_runtime": 163635.3214,
+ "train_tokens_per_second": 5652.054
+ },
+ {
+ "epoch": 1.4243872879073525,
+ "grad_norm": 0.787738441929922,
+ "learning_rate": 1.9087264325441507e-06,
+ "loss": 0.27084195613861084,
+ "num_input_tokens_seen": 925083640,
+ "step": 21155,
+ "train_runtime": 163676.6739,
+ "train_tokens_per_second": 5651.897
+ },
+ {
+ "epoch": 1.4247239429033127,
+ "grad_norm": 0.7323327553262341,
+ "learning_rate": 1.9066486749859686e-06,
+ "loss": 0.3030634880065918,
+ "num_input_tokens_seen": 925295656,
+ "step": 21160,
+ "train_runtime": 163710.5311,
+ "train_tokens_per_second": 5652.023
+ },
+ {
+ "epoch": 1.4250605978992728,
+ "grad_norm": 0.6814532189083007,
+ "learning_rate": 1.9045717824736209e-06,
+ "loss": 0.3159012794494629,
+ "num_input_tokens_seen": 925523192,
+ "step": 21165,
+ "train_runtime": 163748.7467,
+ "train_tokens_per_second": 5652.093
+ },
+ {
+ "epoch": 1.425397252895233,
+ "grad_norm": 0.7944933793290381,
+ "learning_rate": 1.9024957555879037e-06,
+ "loss": 0.30693175792694094,
+ "num_input_tokens_seen": 925733280,
+ "step": 21170,
+ "train_runtime": 163785.8609,
+ "train_tokens_per_second": 5652.095
+ },
+ {
+ "epoch": 1.425733907891193,
+ "grad_norm": 0.7669788666621243,
+ "learning_rate": 1.9004205949093695e-06,
+ "loss": 0.31174449920654296,
+ "num_input_tokens_seen": 925952408,
+ "step": 21175,
+ "train_runtime": 163819.8869,
+ "train_tokens_per_second": 5652.259
+ },
+ {
+ "epoch": 1.4260705628871533,
+ "grad_norm": 0.686815248193178,
+ "learning_rate": 1.8983463010183355e-06,
+ "loss": 0.3182375907897949,
+ "num_input_tokens_seen": 926166584,
+ "step": 21180,
+ "train_runtime": 163852.9598,
+ "train_tokens_per_second": 5652.425
+ },
+ {
+ "epoch": 1.4264072178831133,
+ "grad_norm": 0.7134163125741153,
+ "learning_rate": 1.8962728744948644e-06,
+ "loss": 0.30806000232696534,
+ "num_input_tokens_seen": 926394392,
+ "step": 21185,
+ "train_runtime": 163892.3417,
+ "train_tokens_per_second": 5652.457
+ },
+ {
+ "epoch": 1.4267438728790736,
+ "grad_norm": 0.7517845683613481,
+ "learning_rate": 1.894200315918789e-06,
+ "loss": 0.3054067850112915,
+ "num_input_tokens_seen": 926630992,
+ "step": 21190,
+ "train_runtime": 163926.2105,
+ "train_tokens_per_second": 5652.732
+ },
+ {
+ "epoch": 1.4270805278750336,
+ "grad_norm": 0.8347312289874191,
+ "learning_rate": 1.8921286258696913e-06,
+ "loss": 0.2958783388137817,
+ "num_input_tokens_seen": 926834656,
+ "step": 21195,
+ "train_runtime": 163965.1465,
+ "train_tokens_per_second": 5652.632
+ },
+ {
+ "epoch": 1.4274171828709938,
+ "grad_norm": 0.695621933573107,
+ "learning_rate": 1.890057804926913e-06,
+ "loss": 0.2869250297546387,
+ "num_input_tokens_seen": 927061800,
+ "step": 21200,
+ "train_runtime": 164003.1629,
+ "train_tokens_per_second": 5652.707
+ },
+ {
+ "epoch": 1.4277538378669539,
+ "grad_norm": 0.6966348686251759,
+ "learning_rate": 1.887987853669553e-06,
+ "loss": 0.2620058536529541,
+ "num_input_tokens_seen": 927286432,
+ "step": 21205,
+ "train_runtime": 164046.6271,
+ "train_tokens_per_second": 5652.578
+ },
+ {
+ "epoch": 1.4280904928629141,
+ "grad_norm": 0.830983696417087,
+ "learning_rate": 1.8859187726764654e-06,
+ "loss": 0.3177152633666992,
+ "num_input_tokens_seen": 927512824,
+ "step": 21210,
+ "train_runtime": 164084.865,
+ "train_tokens_per_second": 5652.641
+ },
+ {
+ "epoch": 1.4284271478588741,
+ "grad_norm": 0.7634247104794691,
+ "learning_rate": 1.8838505625262638e-06,
+ "loss": 0.29890553951263427,
+ "num_input_tokens_seen": 927741616,
+ "step": 21215,
+ "train_runtime": 164125.1634,
+ "train_tokens_per_second": 5652.647
+ },
+ {
+ "epoch": 1.4287638028548344,
+ "grad_norm": 0.8416951428048186,
+ "learning_rate": 1.8817832237973154e-06,
+ "loss": 0.2987877368927002,
+ "num_input_tokens_seen": 927959432,
+ "step": 21220,
+ "train_runtime": 164164.0626,
+ "train_tokens_per_second": 5652.634
+ },
+ {
+ "epoch": 1.4291004578507944,
+ "grad_norm": 0.7473862936478309,
+ "learning_rate": 1.8797167570677439e-06,
+ "loss": 0.28604328632354736,
+ "num_input_tokens_seen": 928181056,
+ "step": 21225,
+ "train_runtime": 164213.2671,
+ "train_tokens_per_second": 5652.29
+ },
+ {
+ "epoch": 1.4294371128467547,
+ "grad_norm": 0.7676020809790378,
+ "learning_rate": 1.8776511629154342e-06,
+ "loss": 0.3128670215606689,
+ "num_input_tokens_seen": 928412792,
+ "step": 21230,
+ "train_runtime": 164257.191,
+ "train_tokens_per_second": 5652.19
+ },
+ {
+ "epoch": 1.4297737678427147,
+ "grad_norm": 0.7060907183473086,
+ "learning_rate": 1.8755864419180176e-06,
+ "loss": 0.30298452377319335,
+ "num_input_tokens_seen": 928637368,
+ "step": 21235,
+ "train_runtime": 164304.219,
+ "train_tokens_per_second": 5651.939
+ },
+ {
+ "epoch": 1.430110422838675,
+ "grad_norm": 0.7454649225429147,
+ "learning_rate": 1.8735225946528906e-06,
+ "loss": 0.3099356174468994,
+ "num_input_tokens_seen": 928865136,
+ "step": 21240,
+ "train_runtime": 164340.2939,
+ "train_tokens_per_second": 5652.084
+ },
+ {
+ "epoch": 1.430447077834635,
+ "grad_norm": 0.8028586780145847,
+ "learning_rate": 1.8714596216972008e-06,
+ "loss": 0.3137706518173218,
+ "num_input_tokens_seen": 929082920,
+ "step": 21245,
+ "train_runtime": 164371.8405,
+ "train_tokens_per_second": 5652.324
+ },
+ {
+ "epoch": 1.4307837328305952,
+ "grad_norm": 0.7853913811139581,
+ "learning_rate": 1.8693975236278512e-06,
+ "loss": 0.29973363876342773,
+ "num_input_tokens_seen": 929295744,
+ "step": 21250,
+ "train_runtime": 164413.6617,
+ "train_tokens_per_second": 5652.181
+ },
+ {
+ "epoch": 1.4311203878265553,
+ "grad_norm": 0.8450932910736807,
+ "learning_rate": 1.8673363010215017e-06,
+ "loss": 0.32357001304626465,
+ "num_input_tokens_seen": 929509608,
+ "step": 21255,
+ "train_runtime": 164454.4421,
+ "train_tokens_per_second": 5652.08
+ },
+ {
+ "epoch": 1.4314570428225155,
+ "grad_norm": 0.9839403436936655,
+ "learning_rate": 1.8652759544545661e-06,
+ "loss": 0.3077038288116455,
+ "num_input_tokens_seen": 929730176,
+ "step": 21260,
+ "train_runtime": 164491.0098,
+ "train_tokens_per_second": 5652.164
+ },
+ {
+ "epoch": 1.4317936978184758,
+ "grad_norm": 0.7604757449953816,
+ "learning_rate": 1.863216484503214e-06,
+ "loss": 0.31038436889648435,
+ "num_input_tokens_seen": 929946784,
+ "step": 21265,
+ "train_runtime": 164534.2411,
+ "train_tokens_per_second": 5651.995
+ },
+ {
+ "epoch": 1.4321303528144358,
+ "grad_norm": 0.7883194464998438,
+ "learning_rate": 1.8611578917433698e-06,
+ "loss": 0.3216505527496338,
+ "num_input_tokens_seen": 930172032,
+ "step": 21270,
+ "train_runtime": 164567.2868,
+ "train_tokens_per_second": 5652.229
+ },
+ {
+ "epoch": 1.4324670078103958,
+ "grad_norm": 0.7899893399884377,
+ "learning_rate": 1.85910017675071e-06,
+ "loss": 0.32173566818237304,
+ "num_input_tokens_seen": 930388552,
+ "step": 21275,
+ "train_runtime": 164609.009,
+ "train_tokens_per_second": 5652.112
+ },
+ {
+ "epoch": 1.432803662806356,
+ "grad_norm": 0.7217767584914241,
+ "learning_rate": 1.8570433401006737e-06,
+ "loss": 0.2714304208755493,
+ "num_input_tokens_seen": 930614664,
+ "step": 21280,
+ "train_runtime": 164643.9676,
+ "train_tokens_per_second": 5652.285
+ },
+ {
+ "epoch": 1.4331403178023163,
+ "grad_norm": 0.6750386059644308,
+ "learning_rate": 1.8549873823684418e-06,
+ "loss": 0.2765933990478516,
+ "num_input_tokens_seen": 930827792,
+ "step": 21285,
+ "train_runtime": 164684.8033,
+ "train_tokens_per_second": 5652.178
+ },
+ {
+ "epoch": 1.4334769727982763,
+ "grad_norm": 0.6817317088910789,
+ "learning_rate": 1.8529323041289616e-06,
+ "loss": 0.32932586669921876,
+ "num_input_tokens_seen": 931052808,
+ "step": 21290,
+ "train_runtime": 164719.7957,
+ "train_tokens_per_second": 5652.343
+ },
+ {
+ "epoch": 1.4338136277942364,
+ "grad_norm": 0.7091930055386446,
+ "learning_rate": 1.850878105956927e-06,
+ "loss": 0.30021870136260986,
+ "num_input_tokens_seen": 931280240,
+ "step": 21295,
+ "train_runtime": 164748.8947,
+ "train_tokens_per_second": 5652.725
+ },
+ {
+ "epoch": 1.4341502827901966,
+ "grad_norm": 0.7271899404201645,
+ "learning_rate": 1.8488247884267885e-06,
+ "loss": 0.29010815620422364,
+ "num_input_tokens_seen": 931510352,
+ "step": 21300,
+ "train_runtime": 164786.0486,
+ "train_tokens_per_second": 5652.847
+ },
+ {
+ "epoch": 1.4344869377861569,
+ "grad_norm": 0.7528962182100777,
+ "learning_rate": 1.8467723521127495e-06,
+ "loss": 0.31166844367980956,
+ "num_input_tokens_seen": 931718464,
+ "step": 21305,
+ "train_runtime": 164828.9311,
+ "train_tokens_per_second": 5652.639
+ },
+ {
+ "epoch": 1.434823592782117,
+ "grad_norm": 0.7144128400418023,
+ "learning_rate": 1.8447207975887671e-06,
+ "loss": 0.28693480491638185,
+ "num_input_tokens_seen": 931935064,
+ "step": 21310,
+ "train_runtime": 164871.6945,
+ "train_tokens_per_second": 5652.487
+ },
+ {
+ "epoch": 1.435160247778077,
+ "grad_norm": 0.7690115630964821,
+ "learning_rate": 1.8426701254285522e-06,
+ "loss": 0.3131574630737305,
+ "num_input_tokens_seen": 932161648,
+ "step": 21315,
+ "train_runtime": 164919.0598,
+ "train_tokens_per_second": 5652.237
+ },
+ {
+ "epoch": 1.4354969027740372,
+ "grad_norm": 0.7673435561570116,
+ "learning_rate": 1.840620336205569e-06,
+ "loss": 0.3200539588928223,
+ "num_input_tokens_seen": 932386520,
+ "step": 21320,
+ "train_runtime": 164962.3028,
+ "train_tokens_per_second": 5652.119
+ },
+ {
+ "epoch": 1.4358335577699974,
+ "grad_norm": 0.7307859186284956,
+ "learning_rate": 1.8385714304930324e-06,
+ "loss": 0.2745990753173828,
+ "num_input_tokens_seen": 932614272,
+ "step": 21325,
+ "train_runtime": 164999.3023,
+ "train_tokens_per_second": 5652.232
+ },
+ {
+ "epoch": 1.4361702127659575,
+ "grad_norm": 0.6957468061230739,
+ "learning_rate": 1.8365234088639178e-06,
+ "loss": 0.2942768096923828,
+ "num_input_tokens_seen": 932835368,
+ "step": 21330,
+ "train_runtime": 165036.968,
+ "train_tokens_per_second": 5652.281
+ },
+ {
+ "epoch": 1.4365068677619175,
+ "grad_norm": 0.7649094984689576,
+ "learning_rate": 1.834476271890941e-06,
+ "loss": 0.2779007196426392,
+ "num_input_tokens_seen": 933035248,
+ "step": 21335,
+ "train_runtime": 165075.7307,
+ "train_tokens_per_second": 5652.165
+ },
+ {
+ "epoch": 1.4368435227578777,
+ "grad_norm": 0.8467030298955615,
+ "learning_rate": 1.8324300201465823e-06,
+ "loss": 0.28595268726348877,
+ "num_input_tokens_seen": 933258792,
+ "step": 21340,
+ "train_runtime": 165110.7752,
+ "train_tokens_per_second": 5652.319
+ },
+ {
+ "epoch": 1.437180177753838,
+ "grad_norm": 0.9120057814047616,
+ "learning_rate": 1.8303846542030674e-06,
+ "loss": 0.3224499225616455,
+ "num_input_tokens_seen": 933469416,
+ "step": 21345,
+ "train_runtime": 165149.1219,
+ "train_tokens_per_second": 5652.282
+ },
+ {
+ "epoch": 1.437516832749798,
+ "grad_norm": 0.6674633214236769,
+ "learning_rate": 1.828340174632377e-06,
+ "loss": 0.28315880298614504,
+ "num_input_tokens_seen": 933691024,
+ "step": 21350,
+ "train_runtime": 165182.0127,
+ "train_tokens_per_second": 5652.498
+ },
+ {
+ "epoch": 1.437853487745758,
+ "grad_norm": 0.8225297648308803,
+ "learning_rate": 1.8262965820062434e-06,
+ "loss": 0.29636235237121583,
+ "num_input_tokens_seen": 933911248,
+ "step": 21355,
+ "train_runtime": 165216.2714,
+ "train_tokens_per_second": 5652.659
+ },
+ {
+ "epoch": 1.4381901427417183,
+ "grad_norm": 0.6767595245186808,
+ "learning_rate": 1.8242538768961499e-06,
+ "loss": 0.2795313835144043,
+ "num_input_tokens_seen": 934130456,
+ "step": 21360,
+ "train_runtime": 165249.3613,
+ "train_tokens_per_second": 5652.854
+ },
+ {
+ "epoch": 1.4385267977376786,
+ "grad_norm": 0.7672480972220348,
+ "learning_rate": 1.8222120598733317e-06,
+ "loss": 0.28819353580474855,
+ "num_input_tokens_seen": 934359280,
+ "step": 21365,
+ "train_runtime": 165289.8341,
+ "train_tokens_per_second": 5652.854
+ },
+ {
+ "epoch": 1.4388634527336386,
+ "grad_norm": 0.6461035328546648,
+ "learning_rate": 1.820171131508781e-06,
+ "loss": 0.32363121509552,
+ "num_input_tokens_seen": 934570728,
+ "step": 21370,
+ "train_runtime": 165324.9001,
+ "train_tokens_per_second": 5652.934
+ },
+ {
+ "epoch": 1.4392001077295986,
+ "grad_norm": 0.7231126098153248,
+ "learning_rate": 1.8181310923732304e-06,
+ "loss": 0.2775664091110229,
+ "num_input_tokens_seen": 934787104,
+ "step": 21375,
+ "train_runtime": 165367.2208,
+ "train_tokens_per_second": 5652.796
+ },
+ {
+ "epoch": 1.4395367627255589,
+ "grad_norm": 0.7398980756791921,
+ "learning_rate": 1.8160919430371764e-06,
+ "loss": 0.3209282875061035,
+ "num_input_tokens_seen": 934994512,
+ "step": 21380,
+ "train_runtime": 165407.8952,
+ "train_tokens_per_second": 5652.66
+ },
+ {
+ "epoch": 1.439873417721519,
+ "grad_norm": 0.7357422737413674,
+ "learning_rate": 1.814053684070855e-06,
+ "loss": 0.29017982482910154,
+ "num_input_tokens_seen": 935221480,
+ "step": 21385,
+ "train_runtime": 165444.6956,
+ "train_tokens_per_second": 5652.774
+ },
+ {
+ "epoch": 1.4402100727174791,
+ "grad_norm": 0.6759617094745846,
+ "learning_rate": 1.8120163160442633e-06,
+ "loss": 0.28193049430847167,
+ "num_input_tokens_seen": 935448880,
+ "step": 21390,
+ "train_runtime": 165481.2265,
+ "train_tokens_per_second": 5652.9
+ },
+ {
+ "epoch": 1.4405467277134392,
+ "grad_norm": 0.7834824232339231,
+ "learning_rate": 1.8099798395271428e-06,
+ "loss": 0.31296658515930176,
+ "num_input_tokens_seen": 935684528,
+ "step": 21395,
+ "train_runtime": 165519.1626,
+ "train_tokens_per_second": 5653.028
+ },
+ {
+ "epoch": 1.4408833827093994,
+ "grad_norm": 0.8498195490700376,
+ "learning_rate": 1.807944255088988e-06,
+ "loss": 0.28637051582336426,
+ "num_input_tokens_seen": 935896728,
+ "step": 21400,
+ "train_runtime": 165555.3887,
+ "train_tokens_per_second": 5653.073
+ },
+ {
+ "epoch": 1.4412200377053597,
+ "grad_norm": 0.748373198382384,
+ "learning_rate": 1.8059095632990437e-06,
+ "loss": 0.29571921825408937,
+ "num_input_tokens_seen": 936114728,
+ "step": 21405,
+ "train_runtime": 165587.8193,
+ "train_tokens_per_second": 5653.283
+ },
+ {
+ "epoch": 1.4415566927013197,
+ "grad_norm": 0.7727984893995797,
+ "learning_rate": 1.8038757647263045e-06,
+ "loss": 0.312378978729248,
+ "num_input_tokens_seen": 936331896,
+ "step": 21410,
+ "train_runtime": 165625.2428,
+ "train_tokens_per_second": 5653.316
+ },
+ {
+ "epoch": 1.4418933476972797,
+ "grad_norm": 0.7518266240149807,
+ "learning_rate": 1.8018428599395143e-06,
+ "loss": 0.28726487159729003,
+ "num_input_tokens_seen": 936563512,
+ "step": 21415,
+ "train_runtime": 165667.7133,
+ "train_tokens_per_second": 5653.265
+ },
+ {
+ "epoch": 1.44223000269324,
+ "grad_norm": 0.7820091979263111,
+ "learning_rate": 1.7998108495071743e-06,
+ "loss": 0.2904759645462036,
+ "num_input_tokens_seen": 936786872,
+ "step": 21420,
+ "train_runtime": 165709.7298,
+ "train_tokens_per_second": 5653.18
+ },
+ {
+ "epoch": 1.4425666576892002,
+ "grad_norm": 0.6751633930702853,
+ "learning_rate": 1.797779733997521e-06,
+ "loss": 0.2962934970855713,
+ "num_input_tokens_seen": 937007176,
+ "step": 21425,
+ "train_runtime": 165749.8126,
+ "train_tokens_per_second": 5653.142
+ },
+ {
+ "epoch": 1.4429033126851603,
+ "grad_norm": 0.8092929943597413,
+ "learning_rate": 1.7957495139785585e-06,
+ "loss": 0.3217646837234497,
+ "num_input_tokens_seen": 937228952,
+ "step": 21430,
+ "train_runtime": 165788.6583,
+ "train_tokens_per_second": 5653.155
+ },
+ {
+ "epoch": 1.4432399676811203,
+ "grad_norm": 0.7953086319181529,
+ "learning_rate": 1.7937201900180228e-06,
+ "loss": 0.3023351192474365,
+ "num_input_tokens_seen": 937432152,
+ "step": 21435,
+ "train_runtime": 165829.5232,
+ "train_tokens_per_second": 5652.987
+ },
+ {
+ "epoch": 1.4435766226770805,
+ "grad_norm": 0.779971454215353,
+ "learning_rate": 1.7916917626834134e-06,
+ "loss": 0.339884877204895,
+ "num_input_tokens_seen": 937653160,
+ "step": 21440,
+ "train_runtime": 165865.7943,
+ "train_tokens_per_second": 5653.083
+ },
+ {
+ "epoch": 1.4439132776730408,
+ "grad_norm": 0.7683843965957411,
+ "learning_rate": 1.7896642325419722e-06,
+ "loss": 0.31574015617370604,
+ "num_input_tokens_seen": 937871808,
+ "step": 21445,
+ "train_runtime": 165902.9614,
+ "train_tokens_per_second": 5653.135
+ },
+ {
+ "epoch": 1.4442499326690008,
+ "grad_norm": 0.8231498052304078,
+ "learning_rate": 1.7876376001606915e-06,
+ "loss": 0.30422613620758054,
+ "num_input_tokens_seen": 938077432,
+ "step": 21450,
+ "train_runtime": 165937.6243,
+ "train_tokens_per_second": 5653.193
+ },
+ {
+ "epoch": 1.4445865876649608,
+ "grad_norm": 0.7326842394414514,
+ "learning_rate": 1.7856118661063128e-06,
+ "loss": 0.2780045747756958,
+ "num_input_tokens_seen": 938306192,
+ "step": 21455,
+ "train_runtime": 165979.0063,
+ "train_tokens_per_second": 5653.162
+ },
+ {
+ "epoch": 1.444923242660921,
+ "grad_norm": 0.7610326872210833,
+ "learning_rate": 1.7835870309453251e-06,
+ "loss": 0.3040392637252808,
+ "num_input_tokens_seen": 938518824,
+ "step": 21460,
+ "train_runtime": 166016.1945,
+ "train_tokens_per_second": 5653.176
+ },
+ {
+ "epoch": 1.4452598976568813,
+ "grad_norm": 0.8223404953379267,
+ "learning_rate": 1.7815630952439667e-06,
+ "loss": 0.29809937477111814,
+ "num_input_tokens_seen": 938731144,
+ "step": 21465,
+ "train_runtime": 166055.7098,
+ "train_tokens_per_second": 5653.11
+ },
+ {
+ "epoch": 1.4455965526528414,
+ "grad_norm": 0.6879100802906614,
+ "learning_rate": 1.7795400595682289e-06,
+ "loss": 0.2990201473236084,
+ "num_input_tokens_seen": 938939520,
+ "step": 21470,
+ "train_runtime": 166093.1609,
+ "train_tokens_per_second": 5653.09
+ },
+ {
+ "epoch": 1.4459332076488014,
+ "grad_norm": 0.813015218983521,
+ "learning_rate": 1.7775179244838408e-06,
+ "loss": 0.342014741897583,
+ "num_input_tokens_seen": 939148712,
+ "step": 21475,
+ "train_runtime": 166128.378,
+ "train_tokens_per_second": 5653.15
+ },
+ {
+ "epoch": 1.4462698626447617,
+ "grad_norm": 0.9260324228291221,
+ "learning_rate": 1.7754966905562925e-06,
+ "loss": 0.29918684959411623,
+ "num_input_tokens_seen": 939352016,
+ "step": 21480,
+ "train_runtime": 166162.1908,
+ "train_tokens_per_second": 5653.224
+ },
+ {
+ "epoch": 1.446606517640722,
+ "grad_norm": 0.8748856406745196,
+ "learning_rate": 1.7734763583508098e-06,
+ "loss": 0.29963269233703616,
+ "num_input_tokens_seen": 939569112,
+ "step": 21485,
+ "train_runtime": 166205.207,
+ "train_tokens_per_second": 5653.067
+ },
+ {
+ "epoch": 1.446943172636682,
+ "grad_norm": 0.7627386667181976,
+ "learning_rate": 1.7714569284323757e-06,
+ "loss": 0.31328535079956055,
+ "num_input_tokens_seen": 939800608,
+ "step": 21490,
+ "train_runtime": 166243.8986,
+ "train_tokens_per_second": 5653.143
+ },
+ {
+ "epoch": 1.447279827632642,
+ "grad_norm": 0.8015396995375884,
+ "learning_rate": 1.7694384013657173e-06,
+ "loss": 0.29224376678466796,
+ "num_input_tokens_seen": 940016152,
+ "step": 21495,
+ "train_runtime": 166286.6326,
+ "train_tokens_per_second": 5652.987
+ },
+ {
+ "epoch": 1.4476164826286022,
+ "grad_norm": 0.8776810745850877,
+ "learning_rate": 1.767420777715308e-06,
+ "loss": 0.31572024822235106,
+ "num_input_tokens_seen": 940248920,
+ "step": 21500,
+ "train_runtime": 166324.4801,
+ "train_tokens_per_second": 5653.1
+ },
+ {
+ "epoch": 1.4479531376245625,
+ "grad_norm": 0.7580449081333255,
+ "learning_rate": 1.7654040580453702e-06,
+ "loss": 0.2916224241256714,
+ "num_input_tokens_seen": 940476808,
+ "step": 21505,
+ "train_runtime": 166359.4179,
+ "train_tokens_per_second": 5653.283
+ },
+ {
+ "epoch": 1.4482897926205225,
+ "grad_norm": 0.639413137355145,
+ "learning_rate": 1.7633882429198734e-06,
+ "loss": 0.3200731039047241,
+ "num_input_tokens_seen": 940704272,
+ "step": 21510,
+ "train_runtime": 166400.2408,
+ "train_tokens_per_second": 5653.263
+ },
+ {
+ "epoch": 1.4486264476164825,
+ "grad_norm": 0.7602925689681508,
+ "learning_rate": 1.7613733329025318e-06,
+ "loss": 0.29833736419677737,
+ "num_input_tokens_seen": 940937904,
+ "step": 21515,
+ "train_runtime": 166440.5197,
+ "train_tokens_per_second": 5653.298
+ },
+ {
+ "epoch": 1.4489631026124428,
+ "grad_norm": 0.7097451959902662,
+ "learning_rate": 1.759359328556814e-06,
+ "loss": 0.3069751739501953,
+ "num_input_tokens_seen": 941164048,
+ "step": 21520,
+ "train_runtime": 166482.7242,
+ "train_tokens_per_second": 5653.223
+ },
+ {
+ "epoch": 1.449299757608403,
+ "grad_norm": 0.710759458078187,
+ "learning_rate": 1.7573462304459227e-06,
+ "loss": 0.3017618179321289,
+ "num_input_tokens_seen": 941384664,
+ "step": 21525,
+ "train_runtime": 166515.4855,
+ "train_tokens_per_second": 5653.436
+ },
+ {
+ "epoch": 1.449636412604363,
+ "grad_norm": 0.7685693129079793,
+ "learning_rate": 1.7553340391328216e-06,
+ "loss": 0.28963470458984375,
+ "num_input_tokens_seen": 941603416,
+ "step": 21530,
+ "train_runtime": 166553.5695,
+ "train_tokens_per_second": 5653.457
+ },
+ {
+ "epoch": 1.449973067600323,
+ "grad_norm": 0.738257634670664,
+ "learning_rate": 1.7533227551802068e-06,
+ "loss": 0.3050259590148926,
+ "num_input_tokens_seen": 941831232,
+ "step": 21535,
+ "train_runtime": 166590.2874,
+ "train_tokens_per_second": 5653.578
+ },
+ {
+ "epoch": 1.4503097225962833,
+ "grad_norm": 0.754114134507954,
+ "learning_rate": 1.7513123791505321e-06,
+ "loss": 0.32262425422668456,
+ "num_input_tokens_seen": 942039264,
+ "step": 21540,
+ "train_runtime": 166635.0494,
+ "train_tokens_per_second": 5653.308
+ },
+ {
+ "epoch": 1.4506463775922436,
+ "grad_norm": 0.7204342650845901,
+ "learning_rate": 1.7493029116059913e-06,
+ "loss": 0.3137768268585205,
+ "num_input_tokens_seen": 942254168,
+ "step": 21545,
+ "train_runtime": 166675.6352,
+ "train_tokens_per_second": 5653.221
+ },
+ {
+ "epoch": 1.4509830325882036,
+ "grad_norm": 0.697936428942608,
+ "learning_rate": 1.7472943531085252e-06,
+ "loss": 0.29879701137542725,
+ "num_input_tokens_seen": 942479744,
+ "step": 21550,
+ "train_runtime": 166714.327,
+ "train_tokens_per_second": 5653.262
+ },
+ {
+ "epoch": 1.4513196875841636,
+ "grad_norm": 0.7740871394950247,
+ "learning_rate": 1.7452867042198212e-06,
+ "loss": 0.28659663200378416,
+ "num_input_tokens_seen": 942691224,
+ "step": 21555,
+ "train_runtime": 166751.6931,
+ "train_tokens_per_second": 5653.263
+ },
+ {
+ "epoch": 1.4516563425801239,
+ "grad_norm": 0.6502354382689284,
+ "learning_rate": 1.7432799655013122e-06,
+ "loss": 0.2767528533935547,
+ "num_input_tokens_seen": 942915144,
+ "step": 21560,
+ "train_runtime": 166790.4888,
+ "train_tokens_per_second": 5653.291
+ },
+ {
+ "epoch": 1.4519929975760841,
+ "grad_norm": 0.779285918218484,
+ "learning_rate": 1.741274137514174e-06,
+ "loss": 0.31976196765899656,
+ "num_input_tokens_seen": 943124816,
+ "step": 21565,
+ "train_runtime": 166827.7671,
+ "train_tokens_per_second": 5653.284
+ },
+ {
+ "epoch": 1.4523296525720442,
+ "grad_norm": 0.730492770357248,
+ "learning_rate": 1.7392692208193351e-06,
+ "loss": 0.28882317543029784,
+ "num_input_tokens_seen": 943341696,
+ "step": 21570,
+ "train_runtime": 166874.6737,
+ "train_tokens_per_second": 5652.995
+ },
+ {
+ "epoch": 1.4526663075680042,
+ "grad_norm": 0.7541365898746943,
+ "learning_rate": 1.7372652159774578e-06,
+ "loss": 0.27709264755249025,
+ "num_input_tokens_seen": 943566704,
+ "step": 21575,
+ "train_runtime": 166921.1117,
+ "train_tokens_per_second": 5652.77
+ },
+ {
+ "epoch": 1.4530029625639644,
+ "grad_norm": 0.7928699130255181,
+ "learning_rate": 1.7352621235489625e-06,
+ "loss": 0.2953044414520264,
+ "num_input_tokens_seen": 943783040,
+ "step": 21580,
+ "train_runtime": 166963.3208,
+ "train_tokens_per_second": 5652.637
+ },
+ {
+ "epoch": 1.4533396175599247,
+ "grad_norm": 0.9056536373531542,
+ "learning_rate": 1.733259944094e-06,
+ "loss": 0.3123426914215088,
+ "num_input_tokens_seen": 944002144,
+ "step": 21585,
+ "train_runtime": 167004.2159,
+ "train_tokens_per_second": 5652.565
+ },
+ {
+ "epoch": 1.4536762725558847,
+ "grad_norm": 0.8189634122392393,
+ "learning_rate": 1.7312586781724789e-06,
+ "loss": 0.2972878456115723,
+ "num_input_tokens_seen": 944220536,
+ "step": 21590,
+ "train_runtime": 167035.0784,
+ "train_tokens_per_second": 5652.828
+ },
+ {
+ "epoch": 1.4540129275518447,
+ "grad_norm": 0.732986430525691,
+ "learning_rate": 1.7292583263440454e-06,
+ "loss": 0.31235198974609374,
+ "num_input_tokens_seen": 944434048,
+ "step": 21595,
+ "train_runtime": 167074.9914,
+ "train_tokens_per_second": 5652.755
+ },
+ {
+ "epoch": 1.454349582547805,
+ "grad_norm": 0.7844415288989635,
+ "learning_rate": 1.7272588891680914e-06,
+ "loss": 0.28976993560791015,
+ "num_input_tokens_seen": 944637656,
+ "step": 21600,
+ "train_runtime": 167114.3019,
+ "train_tokens_per_second": 5652.644
+ },
+ {
+ "epoch": 1.4546862375437652,
+ "grad_norm": 0.7510452926746763,
+ "learning_rate": 1.7252603672037527e-06,
+ "loss": 0.29179136753082274,
+ "num_input_tokens_seen": 944849080,
+ "step": 21605,
+ "train_runtime": 167151.8596,
+ "train_tokens_per_second": 5652.639
+ },
+ {
+ "epoch": 1.4550228925397253,
+ "grad_norm": 0.7069433117953492,
+ "learning_rate": 1.72326276100991e-06,
+ "loss": 0.2792097568511963,
+ "num_input_tokens_seen": 945079208,
+ "step": 21610,
+ "train_runtime": 167188.5022,
+ "train_tokens_per_second": 5652.776
+ },
+ {
+ "epoch": 1.4553595475356853,
+ "grad_norm": 0.7362903218179828,
+ "learning_rate": 1.721266071145186e-06,
+ "loss": 0.3040107250213623,
+ "num_input_tokens_seen": 945305216,
+ "step": 21615,
+ "train_runtime": 167227.9102,
+ "train_tokens_per_second": 5652.796
+ },
+ {
+ "epoch": 1.4556962025316456,
+ "grad_norm": 0.6951834691585765,
+ "learning_rate": 1.7192702981679531e-06,
+ "loss": 0.30785455703735354,
+ "num_input_tokens_seen": 945532656,
+ "step": 21620,
+ "train_runtime": 167262.6432,
+ "train_tokens_per_second": 5652.982
+ },
+ {
+ "epoch": 1.4560328575276058,
+ "grad_norm": 0.821045679169631,
+ "learning_rate": 1.717275442636317e-06,
+ "loss": 0.27479963302612304,
+ "num_input_tokens_seen": 945753800,
+ "step": 21625,
+ "train_runtime": 167298.9162,
+ "train_tokens_per_second": 5653.078
+ },
+ {
+ "epoch": 1.4563695125235658,
+ "grad_norm": 0.8193382447695793,
+ "learning_rate": 1.7152815051081384e-06,
+ "loss": 0.30873837471008303,
+ "num_input_tokens_seen": 945972456,
+ "step": 21630,
+ "train_runtime": 167336.3125,
+ "train_tokens_per_second": 5653.121
+ },
+ {
+ "epoch": 1.4567061675195259,
+ "grad_norm": 0.638886315430458,
+ "learning_rate": 1.7132884861410093e-06,
+ "loss": 0.3021131753921509,
+ "num_input_tokens_seen": 946181664,
+ "step": 21635,
+ "train_runtime": 167372.1705,
+ "train_tokens_per_second": 5653.16
+ },
+ {
+ "epoch": 1.4570428225154861,
+ "grad_norm": 0.718262205464768,
+ "learning_rate": 1.7112963862922766e-06,
+ "loss": 0.28493921756744384,
+ "num_input_tokens_seen": 946392928,
+ "step": 21640,
+ "train_runtime": 167406.6012,
+ "train_tokens_per_second": 5653.259
+ },
+ {
+ "epoch": 1.4573794775114464,
+ "grad_norm": 0.7426374869134673,
+ "learning_rate": 1.7093052061190224e-06,
+ "loss": 0.30887959003448484,
+ "num_input_tokens_seen": 946623584,
+ "step": 21645,
+ "train_runtime": 167442.3926,
+ "train_tokens_per_second": 5653.428
+ },
+ {
+ "epoch": 1.4577161325074064,
+ "grad_norm": 0.772847955691719,
+ "learning_rate": 1.7073149461780742e-06,
+ "loss": 0.3057863712310791,
+ "num_input_tokens_seen": 946857512,
+ "step": 21650,
+ "train_runtime": 167480.3097,
+ "train_tokens_per_second": 5653.545
+ },
+ {
+ "epoch": 1.4580527875033664,
+ "grad_norm": 0.8197339025409829,
+ "learning_rate": 1.705325607026001e-06,
+ "loss": 0.31071517467498777,
+ "num_input_tokens_seen": 947062648,
+ "step": 21655,
+ "train_runtime": 167515.2805,
+ "train_tokens_per_second": 5653.59
+ },
+ {
+ "epoch": 1.4583894424993267,
+ "grad_norm": 0.8087443650582465,
+ "learning_rate": 1.7033371892191164e-06,
+ "loss": 0.30583949089050294,
+ "num_input_tokens_seen": 947284464,
+ "step": 21660,
+ "train_runtime": 167551.6527,
+ "train_tokens_per_second": 5653.686
+ },
+ {
+ "epoch": 1.458726097495287,
+ "grad_norm": 0.8458934309863756,
+ "learning_rate": 1.7013496933134726e-06,
+ "loss": 0.3118964433670044,
+ "num_input_tokens_seen": 947494920,
+ "step": 21665,
+ "train_runtime": 167591.8547,
+ "train_tokens_per_second": 5653.586
+ },
+ {
+ "epoch": 1.459062752491247,
+ "grad_norm": 0.7530131041192967,
+ "learning_rate": 1.6993631198648724e-06,
+ "loss": 0.34239089488983154,
+ "num_input_tokens_seen": 947716632,
+ "step": 21670,
+ "train_runtime": 167625.9367,
+ "train_tokens_per_second": 5653.759
+ },
+ {
+ "epoch": 1.459399407487207,
+ "grad_norm": 0.7606267649483974,
+ "learning_rate": 1.697377469428847e-06,
+ "loss": 0.29693849086761476,
+ "num_input_tokens_seen": 947941544,
+ "step": 21675,
+ "train_runtime": 167662.1721,
+ "train_tokens_per_second": 5653.878
+ },
+ {
+ "epoch": 1.4597360624831672,
+ "grad_norm": 0.759611478078839,
+ "learning_rate": 1.6953927425606842e-06,
+ "loss": 0.32032740116119385,
+ "num_input_tokens_seen": 948159704,
+ "step": 21680,
+ "train_runtime": 167706.2866,
+ "train_tokens_per_second": 5653.692
+ },
+ {
+ "epoch": 1.4600727174791275,
+ "grad_norm": 0.6996571917020317,
+ "learning_rate": 1.6934089398154002e-06,
+ "loss": 0.2828591585159302,
+ "num_input_tokens_seen": 948375840,
+ "step": 21685,
+ "train_runtime": 167745.1778,
+ "train_tokens_per_second": 5653.67
+ },
+ {
+ "epoch": 1.4604093724750875,
+ "grad_norm": 0.6597372627621453,
+ "learning_rate": 1.6914260617477634e-06,
+ "loss": 0.284234356880188,
+ "num_input_tokens_seen": 948601288,
+ "step": 21690,
+ "train_runtime": 167778.0303,
+ "train_tokens_per_second": 5653.906
+ },
+ {
+ "epoch": 1.4607460274710475,
+ "grad_norm": 0.630497034262492,
+ "learning_rate": 1.6894441089122788e-06,
+ "loss": 0.2686470985412598,
+ "num_input_tokens_seen": 948816664,
+ "step": 21695,
+ "train_runtime": 167813.8414,
+ "train_tokens_per_second": 5653.983
+ },
+ {
+ "epoch": 1.4610826824670078,
+ "grad_norm": 0.8999505558321718,
+ "learning_rate": 1.6874630818631926e-06,
+ "loss": 0.31239607334136965,
+ "num_input_tokens_seen": 949036440,
+ "step": 21700,
+ "train_runtime": 167856.1735,
+ "train_tokens_per_second": 5653.867
+ },
+ {
+ "epoch": 1.461419337462968,
+ "grad_norm": 0.7748359789057865,
+ "learning_rate": 1.685482981154492e-06,
+ "loss": 0.31434266567230223,
+ "num_input_tokens_seen": 949257872,
+ "step": 21705,
+ "train_runtime": 167886.9016,
+ "train_tokens_per_second": 5654.151
+ },
+ {
+ "epoch": 1.461755992458928,
+ "grad_norm": 0.6957529322275174,
+ "learning_rate": 1.683503807339907e-06,
+ "loss": 0.29442682266235354,
+ "num_input_tokens_seen": 949486984,
+ "step": 21710,
+ "train_runtime": 167930.9226,
+ "train_tokens_per_second": 5654.033
+ },
+ {
+ "epoch": 1.462092647454888,
+ "grad_norm": 0.7012919327076385,
+ "learning_rate": 1.6815255609729047e-06,
+ "loss": 0.3005738496780396,
+ "num_input_tokens_seen": 949713336,
+ "step": 21715,
+ "train_runtime": 167966.4841,
+ "train_tokens_per_second": 5654.184
+ },
+ {
+ "epoch": 1.4624293024508483,
+ "grad_norm": 0.8858327435153073,
+ "learning_rate": 1.6795482426067012e-06,
+ "loss": 0.30016207695007324,
+ "num_input_tokens_seen": 949920736,
+ "step": 21720,
+ "train_runtime": 168001.7371,
+ "train_tokens_per_second": 5654.232
+ },
+ {
+ "epoch": 1.4627659574468086,
+ "grad_norm": 0.8206508131279838,
+ "learning_rate": 1.6775718527942403e-06,
+ "loss": 0.2928857564926147,
+ "num_input_tokens_seen": 950115240,
+ "step": 21725,
+ "train_runtime": 168039.1435,
+ "train_tokens_per_second": 5654.13
+ },
+ {
+ "epoch": 1.4631026124427686,
+ "grad_norm": 0.8111515519361249,
+ "learning_rate": 1.675596392088219e-06,
+ "loss": 0.3051966667175293,
+ "num_input_tokens_seen": 950343040,
+ "step": 21730,
+ "train_runtime": 168073.7856,
+ "train_tokens_per_second": 5654.32
+ },
+ {
+ "epoch": 1.4634392674387287,
+ "grad_norm": 0.68703851940888,
+ "learning_rate": 1.6736218610410633e-06,
+ "loss": 0.28461332321166993,
+ "num_input_tokens_seen": 950576064,
+ "step": 21735,
+ "train_runtime": 168112.6372,
+ "train_tokens_per_second": 5654.4
+ },
+ {
+ "epoch": 1.463775922434689,
+ "grad_norm": 0.7370612171362331,
+ "learning_rate": 1.6716482602049483e-06,
+ "loss": 0.30020885467529296,
+ "num_input_tokens_seen": 950795848,
+ "step": 21740,
+ "train_runtime": 168149.267,
+ "train_tokens_per_second": 5654.475
+ },
+ {
+ "epoch": 1.4641125774306492,
+ "grad_norm": 0.8143604362673187,
+ "learning_rate": 1.669675590131784e-06,
+ "loss": 0.3058682203292847,
+ "num_input_tokens_seen": 950999432,
+ "step": 21745,
+ "train_runtime": 168183.0542,
+ "train_tokens_per_second": 5654.55
+ },
+ {
+ "epoch": 1.4644492324266092,
+ "grad_norm": 0.8042094006923218,
+ "learning_rate": 1.6677038513732214e-06,
+ "loss": 0.29998371601104734,
+ "num_input_tokens_seen": 951224688,
+ "step": 21750,
+ "train_runtime": 168221.299,
+ "train_tokens_per_second": 5654.603
+ },
+ {
+ "epoch": 1.4647858874225694,
+ "grad_norm": 0.7585438265622613,
+ "learning_rate": 1.6657330444806508e-06,
+ "loss": 0.3094784259796143,
+ "num_input_tokens_seen": 951438456,
+ "step": 21755,
+ "train_runtime": 168256.7081,
+ "train_tokens_per_second": 5654.684
+ },
+ {
+ "epoch": 1.4651225424185295,
+ "grad_norm": 0.7062718544616226,
+ "learning_rate": 1.6637631700052021e-06,
+ "loss": 0.29812400341033934,
+ "num_input_tokens_seen": 951670352,
+ "step": 21760,
+ "train_runtime": 168294.5166,
+ "train_tokens_per_second": 5654.791
+ },
+ {
+ "epoch": 1.4654591974144897,
+ "grad_norm": 0.7939195773072991,
+ "learning_rate": 1.6617942284977428e-06,
+ "loss": 0.2911461591720581,
+ "num_input_tokens_seen": 951904024,
+ "step": 21765,
+ "train_runtime": 168329.8691,
+ "train_tokens_per_second": 5654.992
+ },
+ {
+ "epoch": 1.4657958524104497,
+ "grad_norm": 0.8580033189722117,
+ "learning_rate": 1.6598262205088862e-06,
+ "loss": 0.3206942081451416,
+ "num_input_tokens_seen": 952128864,
+ "step": 21770,
+ "train_runtime": 168367.7998,
+ "train_tokens_per_second": 5655.053
+ },
+ {
+ "epoch": 1.46613250740641,
+ "grad_norm": 0.8326314241061893,
+ "learning_rate": 1.6578591465889722e-06,
+ "loss": 0.306864070892334,
+ "num_input_tokens_seen": 952357128,
+ "step": 21775,
+ "train_runtime": 168407.1966,
+ "train_tokens_per_second": 5655.086
+ },
+ {
+ "epoch": 1.46646916240237,
+ "grad_norm": 0.788622072145791,
+ "learning_rate": 1.6558930072880942e-06,
+ "loss": 0.3149855136871338,
+ "num_input_tokens_seen": 952581712,
+ "step": 21780,
+ "train_runtime": 168443.31,
+ "train_tokens_per_second": 5655.207
+ },
+ {
+ "epoch": 1.4668058173983303,
+ "grad_norm": 0.7690789166422503,
+ "learning_rate": 1.6539278031560695e-06,
+ "loss": 0.3216851234436035,
+ "num_input_tokens_seen": 952795432,
+ "step": 21785,
+ "train_runtime": 168482.8899,
+ "train_tokens_per_second": 5655.147
+ },
+ {
+ "epoch": 1.4671424723942903,
+ "grad_norm": 0.8893746538486333,
+ "learning_rate": 1.6519635347424672e-06,
+ "loss": 0.29906625747680665,
+ "num_input_tokens_seen": 953010928,
+ "step": 21790,
+ "train_runtime": 168519.6074,
+ "train_tokens_per_second": 5655.193
+ },
+ {
+ "epoch": 1.4674791273902505,
+ "grad_norm": 0.8072532370811785,
+ "learning_rate": 1.650000202596586e-06,
+ "loss": 0.3279500246047974,
+ "num_input_tokens_seen": 953243200,
+ "step": 21795,
+ "train_runtime": 168562.3622,
+ "train_tokens_per_second": 5655.137
+ },
+ {
+ "epoch": 1.4678157823862106,
+ "grad_norm": 0.7974849910062847,
+ "learning_rate": 1.6480378072674659e-06,
+ "loss": 0.3130605220794678,
+ "num_input_tokens_seen": 953466792,
+ "step": 21800,
+ "train_runtime": 168598.3186,
+ "train_tokens_per_second": 5655.257
+ },
+ {
+ "epoch": 1.4681524373821708,
+ "grad_norm": 0.7397056696771876,
+ "learning_rate": 1.646076349303884e-06,
+ "loss": 0.29395854473114014,
+ "num_input_tokens_seen": 953688232,
+ "step": 21805,
+ "train_runtime": 168638.0597,
+ "train_tokens_per_second": 5655.237
+ },
+ {
+ "epoch": 1.4684890923781309,
+ "grad_norm": 0.8246357078420278,
+ "learning_rate": 1.6441158292543557e-06,
+ "loss": 0.29840645790100095,
+ "num_input_tokens_seen": 953899232,
+ "step": 21810,
+ "train_runtime": 168676.8656,
+ "train_tokens_per_second": 5655.187
+ },
+ {
+ "epoch": 1.468825747374091,
+ "grad_norm": 0.6910590596780838,
+ "learning_rate": 1.6421562476671339e-06,
+ "loss": 0.325870156288147,
+ "num_input_tokens_seen": 954116752,
+ "step": 21815,
+ "train_runtime": 168714.8794,
+ "train_tokens_per_second": 5655.202
+ },
+ {
+ "epoch": 1.4691624023700511,
+ "grad_norm": 0.6572349294585963,
+ "learning_rate": 1.640197605090213e-06,
+ "loss": 0.27675824165344237,
+ "num_input_tokens_seen": 954338024,
+ "step": 21820,
+ "train_runtime": 168751.1585,
+ "train_tokens_per_second": 5655.298
+ },
+ {
+ "epoch": 1.4694990573660114,
+ "grad_norm": 0.7330863432474112,
+ "learning_rate": 1.6382399020713146e-06,
+ "loss": 0.29128522872924806,
+ "num_input_tokens_seen": 954538472,
+ "step": 21825,
+ "train_runtime": 168789.4804,
+ "train_tokens_per_second": 5655.201
+ },
+ {
+ "epoch": 1.4698357123619714,
+ "grad_norm": 0.8821884108200732,
+ "learning_rate": 1.6362831391579115e-06,
+ "loss": 0.2947737455368042,
+ "num_input_tokens_seen": 954753480,
+ "step": 21830,
+ "train_runtime": 168828.8811,
+ "train_tokens_per_second": 5655.155
+ },
+ {
+ "epoch": 1.4701723673579317,
+ "grad_norm": 0.7790169887599169,
+ "learning_rate": 1.634327316897199e-06,
+ "loss": 0.3025893926620483,
+ "num_input_tokens_seen": 954963128,
+ "step": 21835,
+ "train_runtime": 168866.0318,
+ "train_tokens_per_second": 5655.152
+ },
+ {
+ "epoch": 1.4705090223538917,
+ "grad_norm": 0.7612078391890548,
+ "learning_rate": 1.6323724358361216e-06,
+ "loss": 0.2903714656829834,
+ "num_input_tokens_seen": 955171016,
+ "step": 21840,
+ "train_runtime": 168904.9902,
+ "train_tokens_per_second": 5655.079
+ },
+ {
+ "epoch": 1.470845677349852,
+ "grad_norm": 0.740158711633362,
+ "learning_rate": 1.6304184965213537e-06,
+ "loss": 0.29531302452087405,
+ "num_input_tokens_seen": 955381568,
+ "step": 21845,
+ "train_runtime": 168943.7731,
+ "train_tokens_per_second": 5655.027
+ },
+ {
+ "epoch": 1.471182332345812,
+ "grad_norm": 0.7896656186315674,
+ "learning_rate": 1.6284654994993088e-06,
+ "loss": 0.30909888744354247,
+ "num_input_tokens_seen": 955612384,
+ "step": 21850,
+ "train_runtime": 168982.4971,
+ "train_tokens_per_second": 5655.097
+ },
+ {
+ "epoch": 1.4715189873417722,
+ "grad_norm": 0.6615277768870618,
+ "learning_rate": 1.6265134453161358e-06,
+ "loss": 0.29757180213928225,
+ "num_input_tokens_seen": 955824680,
+ "step": 21855,
+ "train_runtime": 169021.4477,
+ "train_tokens_per_second": 5655.05
+ },
+ {
+ "epoch": 1.4718556423377323,
+ "grad_norm": 0.8019048450650291,
+ "learning_rate": 1.6245623345177203e-06,
+ "loss": 0.3056344985961914,
+ "num_input_tokens_seen": 956030528,
+ "step": 21860,
+ "train_runtime": 169060.2008,
+ "train_tokens_per_second": 5654.971
+ },
+ {
+ "epoch": 1.4721922973336925,
+ "grad_norm": 0.6742658933683465,
+ "learning_rate": 1.622612167649683e-06,
+ "loss": 0.28513193130493164,
+ "num_input_tokens_seen": 956238824,
+ "step": 21865,
+ "train_runtime": 169094.7819,
+ "train_tokens_per_second": 5655.046
+ },
+ {
+ "epoch": 1.4725289523296525,
+ "grad_norm": 0.6663845283359301,
+ "learning_rate": 1.6206629452573868e-06,
+ "loss": 0.2704298496246338,
+ "num_input_tokens_seen": 956461272,
+ "step": 21870,
+ "train_runtime": 169128.5661,
+ "train_tokens_per_second": 5655.232
+ },
+ {
+ "epoch": 1.4728656073256128,
+ "grad_norm": 0.719412564793699,
+ "learning_rate": 1.618714667885919e-06,
+ "loss": 0.2946579933166504,
+ "num_input_tokens_seen": 956679360,
+ "step": 21875,
+ "train_runtime": 169169.36,
+ "train_tokens_per_second": 5655.157
+ },
+ {
+ "epoch": 1.4732022623215728,
+ "grad_norm": 0.7215311614958766,
+ "learning_rate": 1.6167673360801151e-06,
+ "loss": 0.29600162506103517,
+ "num_input_tokens_seen": 956909072,
+ "step": 21880,
+ "train_runtime": 169211.1899,
+ "train_tokens_per_second": 5655.117
+ },
+ {
+ "epoch": 1.473538917317533,
+ "grad_norm": 0.8145724755385428,
+ "learning_rate": 1.6148209503845336e-06,
+ "loss": 0.33778862953186034,
+ "num_input_tokens_seen": 957127864,
+ "step": 21885,
+ "train_runtime": 169247.483,
+ "train_tokens_per_second": 5655.197
+ },
+ {
+ "epoch": 1.473875572313493,
+ "grad_norm": 0.831154654556167,
+ "learning_rate": 1.6128755113434808e-06,
+ "loss": 0.2987597703933716,
+ "num_input_tokens_seen": 957342536,
+ "step": 21890,
+ "train_runtime": 169286.3383,
+ "train_tokens_per_second": 5655.167
+ },
+ {
+ "epoch": 1.4742122273094533,
+ "grad_norm": 0.6930146071437139,
+ "learning_rate": 1.6109310195009908e-06,
+ "loss": 0.303321099281311,
+ "num_input_tokens_seen": 957573936,
+ "step": 21895,
+ "train_runtime": 169325.1989,
+ "train_tokens_per_second": 5655.236
+ },
+ {
+ "epoch": 1.4745488823054134,
+ "grad_norm": 0.8269128549821896,
+ "learning_rate": 1.6089874754008323e-06,
+ "loss": 0.29223430156707764,
+ "num_input_tokens_seen": 957779504,
+ "step": 21900,
+ "train_runtime": 169360.8254,
+ "train_tokens_per_second": 5655.26
+ },
+ {
+ "epoch": 1.4748855373013736,
+ "grad_norm": 0.9691951191701637,
+ "learning_rate": 1.6070448795865162e-06,
+ "loss": 0.3116729974746704,
+ "num_input_tokens_seen": 957995792,
+ "step": 21905,
+ "train_runtime": 169396.4164,
+ "train_tokens_per_second": 5655.349
+ },
+ {
+ "epoch": 1.4752221922973336,
+ "grad_norm": 0.7309284982091603,
+ "learning_rate": 1.605103232601277e-06,
+ "loss": 0.29828710556030275,
+ "num_input_tokens_seen": 958213080,
+ "step": 21910,
+ "train_runtime": 169439.4719,
+ "train_tokens_per_second": 5655.194
+ },
+ {
+ "epoch": 1.475558847293294,
+ "grad_norm": 0.716113695585292,
+ "learning_rate": 1.603162534988096e-06,
+ "loss": 0.2677397966384888,
+ "num_input_tokens_seen": 958431216,
+ "step": 21915,
+ "train_runtime": 169477.1448,
+ "train_tokens_per_second": 5655.224
+ },
+ {
+ "epoch": 1.475895502289254,
+ "grad_norm": 0.6939181157623007,
+ "learning_rate": 1.6012227872896791e-06,
+ "loss": 0.28536086082458495,
+ "num_input_tokens_seen": 958642264,
+ "step": 21920,
+ "train_runtime": 169510.665,
+ "train_tokens_per_second": 5655.351
+ },
+ {
+ "epoch": 1.4762321572852142,
+ "grad_norm": 0.7628495263736654,
+ "learning_rate": 1.599283990048472e-06,
+ "loss": 0.2980980396270752,
+ "num_input_tokens_seen": 958850528,
+ "step": 21925,
+ "train_runtime": 169545.9545,
+ "train_tokens_per_second": 5655.402
+ },
+ {
+ "epoch": 1.4765688122811742,
+ "grad_norm": 0.7345760703144733,
+ "learning_rate": 1.5973461438066533e-06,
+ "loss": 0.2987915754318237,
+ "num_input_tokens_seen": 959072368,
+ "step": 21930,
+ "train_runtime": 169591.6892,
+ "train_tokens_per_second": 5655.185
+ },
+ {
+ "epoch": 1.4769054672771345,
+ "grad_norm": 0.7604661278462718,
+ "learning_rate": 1.5954092491061351e-06,
+ "loss": 0.28679566383361815,
+ "num_input_tokens_seen": 959287640,
+ "step": 21935,
+ "train_runtime": 169633.6673,
+ "train_tokens_per_second": 5655.055
+ },
+ {
+ "epoch": 1.4772421222730945,
+ "grad_norm": 0.6981553171592313,
+ "learning_rate": 1.5934733064885637e-06,
+ "loss": 0.30909249782562254,
+ "num_input_tokens_seen": 959515464,
+ "step": 21940,
+ "train_runtime": 169674.7339,
+ "train_tokens_per_second": 5655.029
+ },
+ {
+ "epoch": 1.4775787772690547,
+ "grad_norm": 0.7357160098109768,
+ "learning_rate": 1.5915383164953203e-06,
+ "loss": 0.3021413803100586,
+ "num_input_tokens_seen": 959722168,
+ "step": 21945,
+ "train_runtime": 169712.7213,
+ "train_tokens_per_second": 5654.981
+ },
+ {
+ "epoch": 1.4779154322650148,
+ "grad_norm": 0.8743355744735111,
+ "learning_rate": 1.5896042796675155e-06,
+ "loss": 0.2926707983016968,
+ "num_input_tokens_seen": 959926384,
+ "step": 21950,
+ "train_runtime": 169749.0885,
+ "train_tokens_per_second": 5654.972
+ },
+ {
+ "epoch": 1.478252087260975,
+ "grad_norm": 0.7663596131904904,
+ "learning_rate": 1.5876711965460023e-06,
+ "loss": 0.3000811576843262,
+ "num_input_tokens_seen": 960147848,
+ "step": 21955,
+ "train_runtime": 169793.9752,
+ "train_tokens_per_second": 5654.782
+ },
+ {
+ "epoch": 1.478588742256935,
+ "grad_norm": 0.7506605855075004,
+ "learning_rate": 1.5857390676713546e-06,
+ "loss": 0.3030038118362427,
+ "num_input_tokens_seen": 960354248,
+ "step": 21960,
+ "train_runtime": 169826.7418,
+ "train_tokens_per_second": 5654.906
+ },
+ {
+ "epoch": 1.4789253972528953,
+ "grad_norm": 0.6893712500395903,
+ "learning_rate": 1.5838078935838913e-06,
+ "loss": 0.2610928058624268,
+ "num_input_tokens_seen": 960559904,
+ "step": 21965,
+ "train_runtime": 169865.4378,
+ "train_tokens_per_second": 5654.828
+ },
+ {
+ "epoch": 1.4792620522488553,
+ "grad_norm": 0.7970970921801804,
+ "learning_rate": 1.581877674823657e-06,
+ "loss": 0.286395788192749,
+ "num_input_tokens_seen": 960763880,
+ "step": 21970,
+ "train_runtime": 169904.7926,
+ "train_tokens_per_second": 5654.719
+ },
+ {
+ "epoch": 1.4795987072448156,
+ "grad_norm": 0.7469532730634993,
+ "learning_rate": 1.5799484119304308e-06,
+ "loss": 0.3111164093017578,
+ "num_input_tokens_seen": 960983256,
+ "step": 21975,
+ "train_runtime": 169948.1325,
+ "train_tokens_per_second": 5654.568
+ },
+ {
+ "epoch": 1.4799353622407756,
+ "grad_norm": 0.7131812605057424,
+ "learning_rate": 1.578020105443726e-06,
+ "loss": 0.2867883205413818,
+ "num_input_tokens_seen": 961204296,
+ "step": 21980,
+ "train_runtime": 169989.5356,
+ "train_tokens_per_second": 5654.491
+ },
+ {
+ "epoch": 1.4802720172367358,
+ "grad_norm": 0.8560168052021411,
+ "learning_rate": 1.576092755902787e-06,
+ "loss": 0.3115290641784668,
+ "num_input_tokens_seen": 961419104,
+ "step": 21985,
+ "train_runtime": 170025.6173,
+ "train_tokens_per_second": 5654.554
+ },
+ {
+ "epoch": 1.4806086722326959,
+ "grad_norm": 0.7910019975942408,
+ "learning_rate": 1.5741663638465903e-06,
+ "loss": 0.32115159034729,
+ "num_input_tokens_seen": 961624552,
+ "step": 21990,
+ "train_runtime": 170059.2178,
+ "train_tokens_per_second": 5654.645
+ },
+ {
+ "epoch": 1.4809453272286561,
+ "grad_norm": 0.7740362774361383,
+ "learning_rate": 1.5722409298138463e-06,
+ "loss": 0.2831075668334961,
+ "num_input_tokens_seen": 961828112,
+ "step": 21995,
+ "train_runtime": 170097.4303,
+ "train_tokens_per_second": 5654.572
+ },
+ {
+ "epoch": 1.4812819822246162,
+ "grad_norm": 0.887435473761687,
+ "learning_rate": 1.570316454342994e-06,
+ "loss": 0.3216114044189453,
+ "num_input_tokens_seen": 962043280,
+ "step": 22000,
+ "train_runtime": 170135.4044,
+ "train_tokens_per_second": 5654.574
+ },
+ {
+ "epoch": 1.4816186372205764,
+ "grad_norm": 0.7331141791279817,
+ "learning_rate": 1.568392937972212e-06,
+ "loss": 0.30264816284179685,
+ "num_input_tokens_seen": 962277624,
+ "step": 22005,
+ "train_runtime": 170177.2637,
+ "train_tokens_per_second": 5654.56
+ },
+ {
+ "epoch": 1.4819552922165364,
+ "grad_norm": 0.7730593120342526,
+ "learning_rate": 1.5664703812393988e-06,
+ "loss": 0.30075869560241697,
+ "num_input_tokens_seen": 962510352,
+ "step": 22010,
+ "train_runtime": 170215.2224,
+ "train_tokens_per_second": 5654.667
+ },
+ {
+ "epoch": 1.4822919472124967,
+ "grad_norm": 0.7777724901976949,
+ "learning_rate": 1.5645487846821967e-06,
+ "loss": 0.2980914831161499,
+ "num_input_tokens_seen": 962736264,
+ "step": 22015,
+ "train_runtime": 170256.3094,
+ "train_tokens_per_second": 5654.629
+ },
+ {
+ "epoch": 1.4826286022084567,
+ "grad_norm": 0.7843674786251581,
+ "learning_rate": 1.5626281488379724e-06,
+ "loss": 0.30644855499267576,
+ "num_input_tokens_seen": 962954856,
+ "step": 22020,
+ "train_runtime": 170290.947,
+ "train_tokens_per_second": 5654.762
+ },
+ {
+ "epoch": 1.482965257204417,
+ "grad_norm": 0.7409514511968761,
+ "learning_rate": 1.560708474243825e-06,
+ "loss": 0.28392848968505857,
+ "num_input_tokens_seen": 963172840,
+ "step": 22025,
+ "train_runtime": 170330.7319,
+ "train_tokens_per_second": 5654.721
+ },
+ {
+ "epoch": 1.483301912200377,
+ "grad_norm": 0.8527324584193815,
+ "learning_rate": 1.5587897614365854e-06,
+ "loss": 0.3137376070022583,
+ "num_input_tokens_seen": 963402704,
+ "step": 22030,
+ "train_runtime": 170363.8783,
+ "train_tokens_per_second": 5654.97
+ },
+ {
+ "epoch": 1.4836385671963372,
+ "grad_norm": 0.817954831707004,
+ "learning_rate": 1.5568720109528163e-06,
+ "loss": 0.29897422790527345,
+ "num_input_tokens_seen": 963622656,
+ "step": 22035,
+ "train_runtime": 170407.9186,
+ "train_tokens_per_second": 5654.8
+ },
+ {
+ "epoch": 1.4839752221922973,
+ "grad_norm": 0.9090048918137543,
+ "learning_rate": 1.5549552233288095e-06,
+ "loss": 0.32768850326538085,
+ "num_input_tokens_seen": 963839112,
+ "step": 22040,
+ "train_runtime": 170444.0634,
+ "train_tokens_per_second": 5654.871
+ },
+ {
+ "epoch": 1.4843118771882575,
+ "grad_norm": 0.7108214887496512,
+ "learning_rate": 1.5530393991005893e-06,
+ "loss": 0.2966951370239258,
+ "num_input_tokens_seen": 964056312,
+ "step": 22045,
+ "train_runtime": 170495.1004,
+ "train_tokens_per_second": 5654.452
+ },
+ {
+ "epoch": 1.4846485321842176,
+ "grad_norm": 0.7628438783439455,
+ "learning_rate": 1.551124538803908e-06,
+ "loss": 0.3010308265686035,
+ "num_input_tokens_seen": 964282488,
+ "step": 22050,
+ "train_runtime": 170534.3626,
+ "train_tokens_per_second": 5654.476
+ },
+ {
+ "epoch": 1.4849851871801778,
+ "grad_norm": 0.766886301451967,
+ "learning_rate": 1.5492106429742547e-06,
+ "loss": 0.30379014015197753,
+ "num_input_tokens_seen": 964505592,
+ "step": 22055,
+ "train_runtime": 170574.175,
+ "train_tokens_per_second": 5654.464
+ },
+ {
+ "epoch": 1.4853218421761378,
+ "grad_norm": 0.8077546385043058,
+ "learning_rate": 1.5472977121468386e-06,
+ "loss": 0.3085505962371826,
+ "num_input_tokens_seen": 964734320,
+ "step": 22060,
+ "train_runtime": 170605.4495,
+ "train_tokens_per_second": 5654.768
+ },
+ {
+ "epoch": 1.485658497172098,
+ "grad_norm": 0.8224102700502363,
+ "learning_rate": 1.545385746856609e-06,
+ "loss": 0.2914187669754028,
+ "num_input_tokens_seen": 964946888,
+ "step": 22065,
+ "train_runtime": 170641.0717,
+ "train_tokens_per_second": 5654.834
+ },
+ {
+ "epoch": 1.485995152168058,
+ "grad_norm": 0.73340844909282,
+ "learning_rate": 1.5434747476382388e-06,
+ "loss": 0.2726666212081909,
+ "num_input_tokens_seen": 965178856,
+ "step": 22070,
+ "train_runtime": 170676.1913,
+ "train_tokens_per_second": 5655.029
+ },
+ {
+ "epoch": 1.4863318071640184,
+ "grad_norm": 0.8499683640218707,
+ "learning_rate": 1.5415647150261337e-06,
+ "loss": 0.2993777751922607,
+ "num_input_tokens_seen": 965382264,
+ "step": 22075,
+ "train_runtime": 170714.8439,
+ "train_tokens_per_second": 5654.94
+ },
+ {
+ "epoch": 1.4866684621599784,
+ "grad_norm": 0.8088856371624081,
+ "learning_rate": 1.539655649554428e-06,
+ "loss": 0.2940176010131836,
+ "num_input_tokens_seen": 965597136,
+ "step": 22080,
+ "train_runtime": 170755.1927,
+ "train_tokens_per_second": 5654.862
+ },
+ {
+ "epoch": 1.4870051171559386,
+ "grad_norm": 0.7548642358843536,
+ "learning_rate": 1.5377475517569856e-06,
+ "loss": 0.29267857074737547,
+ "num_input_tokens_seen": 965803184,
+ "step": 22085,
+ "train_runtime": 170791.9589,
+ "train_tokens_per_second": 5654.852
+ },
+ {
+ "epoch": 1.4873417721518987,
+ "grad_norm": 0.765784318943008,
+ "learning_rate": 1.5358404221674e-06,
+ "loss": 0.30087296962738036,
+ "num_input_tokens_seen": 966009728,
+ "step": 22090,
+ "train_runtime": 170831.4433,
+ "train_tokens_per_second": 5654.754
+ },
+ {
+ "epoch": 1.487678427147859,
+ "grad_norm": 0.7556445487670026,
+ "learning_rate": 1.533934261318994e-06,
+ "loss": 0.30022130012512205,
+ "num_input_tokens_seen": 966233096,
+ "step": 22095,
+ "train_runtime": 170870.3846,
+ "train_tokens_per_second": 5654.772
+ },
+ {
+ "epoch": 1.488015082143819,
+ "grad_norm": 0.7251976781164082,
+ "learning_rate": 1.5320290697448177e-06,
+ "loss": 0.2852438449859619,
+ "num_input_tokens_seen": 966445288,
+ "step": 22100,
+ "train_runtime": 170911.1982,
+ "train_tokens_per_second": 5654.663
+ },
+ {
+ "epoch": 1.4883517371397792,
+ "grad_norm": 0.859114292106197,
+ "learning_rate": 1.530124847977657e-06,
+ "loss": 0.320376992225647,
+ "num_input_tokens_seen": 966659272,
+ "step": 22105,
+ "train_runtime": 170948.93,
+ "train_tokens_per_second": 5654.667
+ },
+ {
+ "epoch": 1.4886883921357392,
+ "grad_norm": 0.7697511007972789,
+ "learning_rate": 1.528221596550014e-06,
+ "loss": 0.2888146638870239,
+ "num_input_tokens_seen": 966874112,
+ "step": 22110,
+ "train_runtime": 170988.9473,
+ "train_tokens_per_second": 5654.6
+ },
+ {
+ "epoch": 1.4890250471316995,
+ "grad_norm": 0.7792399302642892,
+ "learning_rate": 1.5263193159941331e-06,
+ "loss": 0.2921769618988037,
+ "num_input_tokens_seen": 967092032,
+ "step": 22115,
+ "train_runtime": 171028.9685,
+ "train_tokens_per_second": 5654.551
+ },
+ {
+ "epoch": 1.4893617021276595,
+ "grad_norm": 0.7804399534999346,
+ "learning_rate": 1.5244180068419783e-06,
+ "loss": 0.28116235733032224,
+ "num_input_tokens_seen": 967308280,
+ "step": 22120,
+ "train_runtime": 171077.0397,
+ "train_tokens_per_second": 5654.226
+ },
+ {
+ "epoch": 1.4896983571236198,
+ "grad_norm": 0.8657428518226007,
+ "learning_rate": 1.5225176696252453e-06,
+ "loss": 0.32404913902282717,
+ "num_input_tokens_seen": 967517656,
+ "step": 22125,
+ "train_runtime": 171124.2522,
+ "train_tokens_per_second": 5653.89
+ },
+ {
+ "epoch": 1.4900350121195798,
+ "grad_norm": 0.7792695604281658,
+ "learning_rate": 1.5206183048753575e-06,
+ "loss": 0.30272302627563474,
+ "num_input_tokens_seen": 967731128,
+ "step": 22130,
+ "train_runtime": 171161.8669,
+ "train_tokens_per_second": 5653.894
+ },
+ {
+ "epoch": 1.49037166711554,
+ "grad_norm": 0.8874127014480637,
+ "learning_rate": 1.5187199131234653e-06,
+ "loss": 0.34148373603820803,
+ "num_input_tokens_seen": 967953912,
+ "step": 22135,
+ "train_runtime": 171197.9915,
+ "train_tokens_per_second": 5654.003
+ },
+ {
+ "epoch": 1.4907083221115,
+ "grad_norm": 0.8158011732743956,
+ "learning_rate": 1.5168224949004484e-06,
+ "loss": 0.2946127653121948,
+ "num_input_tokens_seen": 968163840,
+ "step": 22140,
+ "train_runtime": 171239.0138,
+ "train_tokens_per_second": 5653.874
+ },
+ {
+ "epoch": 1.4910449771074603,
+ "grad_norm": 0.7094644713372431,
+ "learning_rate": 1.514926050736914e-06,
+ "loss": 0.2730919361114502,
+ "num_input_tokens_seen": 968387800,
+ "step": 22145,
+ "train_runtime": 171288.0971,
+ "train_tokens_per_second": 5653.562
+ },
+ {
+ "epoch": 1.4913816321034203,
+ "grad_norm": 0.7887316321901531,
+ "learning_rate": 1.5130305811631945e-06,
+ "loss": 0.3045715570449829,
+ "num_input_tokens_seen": 968613216,
+ "step": 22150,
+ "train_runtime": 171326.4957,
+ "train_tokens_per_second": 5653.61
+ },
+ {
+ "epoch": 1.4917182870993806,
+ "grad_norm": 0.8209012844760429,
+ "learning_rate": 1.5111360867093571e-06,
+ "loss": 0.3100686550140381,
+ "num_input_tokens_seen": 968837880,
+ "step": 22155,
+ "train_runtime": 171360.4409,
+ "train_tokens_per_second": 5653.801
+ },
+ {
+ "epoch": 1.4920549420953406,
+ "grad_norm": 0.7096360424168036,
+ "learning_rate": 1.5092425679051848e-06,
+ "loss": 0.2817687511444092,
+ "num_input_tokens_seen": 969060560,
+ "step": 22160,
+ "train_runtime": 171394.8343,
+ "train_tokens_per_second": 5653.966
+ },
+ {
+ "epoch": 1.4923915970913009,
+ "grad_norm": 0.74898568213781,
+ "learning_rate": 1.5073500252801988e-06,
+ "loss": 0.31572632789611815,
+ "num_input_tokens_seen": 969287272,
+ "step": 22165,
+ "train_runtime": 171423.1343,
+ "train_tokens_per_second": 5654.355
+ },
+ {
+ "epoch": 1.492728252087261,
+ "grad_norm": 0.7194688541867551,
+ "learning_rate": 1.5054584593636412e-06,
+ "loss": 0.2992145776748657,
+ "num_input_tokens_seen": 969517872,
+ "step": 22170,
+ "train_runtime": 171459.1961,
+ "train_tokens_per_second": 5654.511
+ },
+ {
+ "epoch": 1.4930649070832211,
+ "grad_norm": 0.8380897247829227,
+ "learning_rate": 1.5035678706844814e-06,
+ "loss": 0.3096607208251953,
+ "num_input_tokens_seen": 969734072,
+ "step": 22175,
+ "train_runtime": 171496.6903,
+ "train_tokens_per_second": 5654.535
+ },
+ {
+ "epoch": 1.4934015620791812,
+ "grad_norm": 0.719052290185219,
+ "learning_rate": 1.501678259771418e-06,
+ "loss": 0.311397123336792,
+ "num_input_tokens_seen": 969956280,
+ "step": 22180,
+ "train_runtime": 171541.9334,
+ "train_tokens_per_second": 5654.339
+ },
+ {
+ "epoch": 1.4937382170751414,
+ "grad_norm": 0.8005847376324827,
+ "learning_rate": 1.499789627152874e-06,
+ "loss": 0.3107875108718872,
+ "num_input_tokens_seen": 970180640,
+ "step": 22185,
+ "train_runtime": 171584.1364,
+ "train_tokens_per_second": 5654.256
+ },
+ {
+ "epoch": 1.4940748720711015,
+ "grad_norm": 0.7488870093377558,
+ "learning_rate": 1.497901973356999e-06,
+ "loss": 0.26401643753051757,
+ "num_input_tokens_seen": 970405336,
+ "step": 22190,
+ "train_runtime": 171621.5048,
+ "train_tokens_per_second": 5654.334
+ },
+ {
+ "epoch": 1.4944115270670617,
+ "grad_norm": 0.6586409758641915,
+ "learning_rate": 1.4960152989116694e-06,
+ "loss": 0.29029550552368166,
+ "num_input_tokens_seen": 970628528,
+ "step": 22195,
+ "train_runtime": 171659.7512,
+ "train_tokens_per_second": 5654.375
+ },
+ {
+ "epoch": 1.4947481820630217,
+ "grad_norm": 0.7846315012803541,
+ "learning_rate": 1.4941296043444869e-06,
+ "loss": 0.31101787090301514,
+ "num_input_tokens_seen": 970833256,
+ "step": 22200,
+ "train_runtime": 171700.279,
+ "train_tokens_per_second": 5654.232
+ },
+ {
+ "epoch": 1.495084837058982,
+ "grad_norm": 0.6807750672867574,
+ "learning_rate": 1.4922448901827847e-06,
+ "loss": 0.2868825912475586,
+ "num_input_tokens_seen": 971056192,
+ "step": 22205,
+ "train_runtime": 171746.7294,
+ "train_tokens_per_second": 5654.001
+ },
+ {
+ "epoch": 1.495421492054942,
+ "grad_norm": 0.9017046482285462,
+ "learning_rate": 1.4903611569536103e-06,
+ "loss": 0.27285065650939944,
+ "num_input_tokens_seen": 971251664,
+ "step": 22210,
+ "train_runtime": 171784.3198,
+ "train_tokens_per_second": 5653.902
+ },
+ {
+ "epoch": 1.4957581470509023,
+ "grad_norm": 0.7546903370960244,
+ "learning_rate": 1.488478405183748e-06,
+ "loss": 0.3076265096664429,
+ "num_input_tokens_seen": 971461824,
+ "step": 22215,
+ "train_runtime": 171819.829,
+ "train_tokens_per_second": 5653.956
+ },
+ {
+ "epoch": 1.4960948020468623,
+ "grad_norm": 0.7286152762441251,
+ "learning_rate": 1.4865966353997024e-06,
+ "loss": 0.2991771221160889,
+ "num_input_tokens_seen": 971691984,
+ "step": 22220,
+ "train_runtime": 171853.9282,
+ "train_tokens_per_second": 5654.174
+ },
+ {
+ "epoch": 1.4964314570428225,
+ "grad_norm": 0.6811343761957527,
+ "learning_rate": 1.4847158481277047e-06,
+ "loss": 0.27810006141662597,
+ "num_input_tokens_seen": 971910752,
+ "step": 22225,
+ "train_runtime": 171892.8709,
+ "train_tokens_per_second": 5654.166
+ },
+ {
+ "epoch": 1.4967681120387826,
+ "grad_norm": 0.8778824220267035,
+ "learning_rate": 1.48283604389371e-06,
+ "loss": 0.274126935005188,
+ "num_input_tokens_seen": 972116168,
+ "step": 22230,
+ "train_runtime": 171929.1838,
+ "train_tokens_per_second": 5654.166
+ },
+ {
+ "epoch": 1.4971047670347428,
+ "grad_norm": 0.6703777252052292,
+ "learning_rate": 1.480957223223401e-06,
+ "loss": 0.28500995635986326,
+ "num_input_tokens_seen": 972336272,
+ "step": 22235,
+ "train_runtime": 171966.7609,
+ "train_tokens_per_second": 5654.211
+ },
+ {
+ "epoch": 1.497441422030703,
+ "grad_norm": 0.8037319519905057,
+ "learning_rate": 1.4790793866421826e-06,
+ "loss": 0.32386600971221924,
+ "num_input_tokens_seen": 972548536,
+ "step": 22240,
+ "train_runtime": 172003.2466,
+ "train_tokens_per_second": 5654.245
+ },
+ {
+ "epoch": 1.497778077026663,
+ "grad_norm": 0.8609976203519863,
+ "learning_rate": 1.4772025346751862e-06,
+ "loss": 0.30510640144348145,
+ "num_input_tokens_seen": 972767336,
+ "step": 22245,
+ "train_runtime": 172035.4661,
+ "train_tokens_per_second": 5654.458
+ },
+ {
+ "epoch": 1.4981147320226231,
+ "grad_norm": 0.6833659788420334,
+ "learning_rate": 1.475326667847266e-06,
+ "loss": 0.3076331615447998,
+ "num_input_tokens_seen": 972987240,
+ "step": 22250,
+ "train_runtime": 172073.245,
+ "train_tokens_per_second": 5654.495
+ },
+ {
+ "epoch": 1.4984513870185834,
+ "grad_norm": 0.7212227153724603,
+ "learning_rate": 1.4734517866830073e-06,
+ "loss": 0.30481748580932616,
+ "num_input_tokens_seen": 973215336,
+ "step": 22255,
+ "train_runtime": 172107.067,
+ "train_tokens_per_second": 5654.709
+ },
+ {
+ "epoch": 1.4987880420145436,
+ "grad_norm": 0.7269815306579177,
+ "learning_rate": 1.4715778917067081e-06,
+ "loss": 0.303895378112793,
+ "num_input_tokens_seen": 973425600,
+ "step": 22260,
+ "train_runtime": 172149.3633,
+ "train_tokens_per_second": 5654.541
+ },
+ {
+ "epoch": 1.4991246970105037,
+ "grad_norm": 0.6350255641888927,
+ "learning_rate": 1.4697049834424016e-06,
+ "loss": 0.29153730869293215,
+ "num_input_tokens_seen": 973643560,
+ "step": 22265,
+ "train_runtime": 172188.7254,
+ "train_tokens_per_second": 5654.514
+ },
+ {
+ "epoch": 1.4994613520064637,
+ "grad_norm": 0.9326241952838099,
+ "learning_rate": 1.4678330624138387e-06,
+ "loss": 0.285983943939209,
+ "num_input_tokens_seen": 973854824,
+ "step": 22270,
+ "train_runtime": 172221.8921,
+ "train_tokens_per_second": 5654.652
+ },
+ {
+ "epoch": 1.499798007002424,
+ "grad_norm": 0.6824859115358656,
+ "learning_rate": 1.465962129144497e-06,
+ "loss": 0.2954267978668213,
+ "num_input_tokens_seen": 974073608,
+ "step": 22275,
+ "train_runtime": 172256.0105,
+ "train_tokens_per_second": 5654.802
+ },
+ {
+ "epoch": 1.5001346619983842,
+ "grad_norm": 0.8186629815319795,
+ "learning_rate": 1.4640921841575766e-06,
+ "loss": 0.32407686710357664,
+ "num_input_tokens_seen": 974307744,
+ "step": 22280,
+ "train_runtime": 172292.4565,
+ "train_tokens_per_second": 5654.965
+ },
+ {
+ "epoch": 1.5004713169943442,
+ "grad_norm": 0.6972447701320164,
+ "learning_rate": 1.4622232279760018e-06,
+ "loss": 0.29003033638000486,
+ "num_input_tokens_seen": 974527448,
+ "step": 22285,
+ "train_runtime": 172326.5609,
+ "train_tokens_per_second": 5655.12
+ },
+ {
+ "epoch": 1.5008079719903042,
+ "grad_norm": 0.8003233393761351,
+ "learning_rate": 1.4603552611224202e-06,
+ "loss": 0.3260036468505859,
+ "num_input_tokens_seen": 974732192,
+ "step": 22290,
+ "train_runtime": 172364.9722,
+ "train_tokens_per_second": 5655.048
+ },
+ {
+ "epoch": 1.5011446269862645,
+ "grad_norm": 0.7801634179627978,
+ "learning_rate": 1.4584882841192027e-06,
+ "loss": 0.3202426195144653,
+ "num_input_tokens_seen": 974950184,
+ "step": 22295,
+ "train_runtime": 172402.4948,
+ "train_tokens_per_second": 5655.082
+ },
+ {
+ "epoch": 1.5014812819822247,
+ "grad_norm": 0.7560222588514484,
+ "learning_rate": 1.456622297488442e-06,
+ "loss": 0.2875128030776978,
+ "num_input_tokens_seen": 975160280,
+ "step": 22300,
+ "train_runtime": 172436.8178,
+ "train_tokens_per_second": 5655.174
+ },
+ {
+ "epoch": 1.5018179369781848,
+ "grad_norm": 0.8392814865686347,
+ "learning_rate": 1.4547573017519595e-06,
+ "loss": 0.29420223236083987,
+ "num_input_tokens_seen": 975368840,
+ "step": 22305,
+ "train_runtime": 172476.8287,
+ "train_tokens_per_second": 5655.072
+ },
+ {
+ "epoch": 1.5021545919741448,
+ "grad_norm": 0.7889313036819893,
+ "learning_rate": 1.4528932974312897e-06,
+ "loss": 0.28504467010498047,
+ "num_input_tokens_seen": 975600768,
+ "step": 22310,
+ "train_runtime": 172519.0021,
+ "train_tokens_per_second": 5655.034
+ },
+ {
+ "epoch": 1.502491246970105,
+ "grad_norm": 0.8405102432636552,
+ "learning_rate": 1.4510302850477005e-06,
+ "loss": 0.29707143306732176,
+ "num_input_tokens_seen": 975818888,
+ "step": 22315,
+ "train_runtime": 172557.3322,
+ "train_tokens_per_second": 5655.042
+ },
+ {
+ "epoch": 1.5028279019660653,
+ "grad_norm": 0.7839583808405233,
+ "learning_rate": 1.4491682651221746e-06,
+ "loss": 0.289170503616333,
+ "num_input_tokens_seen": 976024992,
+ "step": 22320,
+ "train_runtime": 172593.7247,
+ "train_tokens_per_second": 5655.043
+ },
+ {
+ "epoch": 1.5031645569620253,
+ "grad_norm": 0.7899317516379861,
+ "learning_rate": 1.447307238175421e-06,
+ "loss": 0.298907732963562,
+ "num_input_tokens_seen": 976231792,
+ "step": 22325,
+ "train_runtime": 172634.9012,
+ "train_tokens_per_second": 5654.892
+ },
+ {
+ "epoch": 1.5035012119579854,
+ "grad_norm": 0.7367720389125396,
+ "learning_rate": 1.4454472047278688e-06,
+ "loss": 0.29713735580444334,
+ "num_input_tokens_seen": 976442768,
+ "step": 22330,
+ "train_runtime": 172676.2447,
+ "train_tokens_per_second": 5654.76
+ },
+ {
+ "epoch": 1.5038378669539456,
+ "grad_norm": 0.7850128485368816,
+ "learning_rate": 1.443588165299672e-06,
+ "loss": 0.2721333265304565,
+ "num_input_tokens_seen": 976670000,
+ "step": 22335,
+ "train_runtime": 172721.7292,
+ "train_tokens_per_second": 5654.587
+ },
+ {
+ "epoch": 1.5041745219499059,
+ "grad_norm": 0.7958028090619667,
+ "learning_rate": 1.4417301204107037e-06,
+ "loss": 0.31305713653564454,
+ "num_input_tokens_seen": 976892192,
+ "step": 22340,
+ "train_runtime": 172757.5302,
+ "train_tokens_per_second": 5654.701
+ },
+ {
+ "epoch": 1.504511176945866,
+ "grad_norm": 0.7865067258558222,
+ "learning_rate": 1.4398730705805607e-06,
+ "loss": 0.3005058288574219,
+ "num_input_tokens_seen": 977116432,
+ "step": 22345,
+ "train_runtime": 172795.4681,
+ "train_tokens_per_second": 5654.757
+ },
+ {
+ "epoch": 1.504847831941826,
+ "grad_norm": 0.6971438552549611,
+ "learning_rate": 1.43801701632856e-06,
+ "loss": 0.3000321388244629,
+ "num_input_tokens_seen": 977343248,
+ "step": 22350,
+ "train_runtime": 172831.4992,
+ "train_tokens_per_second": 5654.891
+ },
+ {
+ "epoch": 1.5051844869377862,
+ "grad_norm": 0.7852554320922348,
+ "learning_rate": 1.436161958173745e-06,
+ "loss": 0.32420189380645753,
+ "num_input_tokens_seen": 977567152,
+ "step": 22355,
+ "train_runtime": 172868.2683,
+ "train_tokens_per_second": 5654.983
+ },
+ {
+ "epoch": 1.5055211419337464,
+ "grad_norm": 0.7387859896989003,
+ "learning_rate": 1.4343078966348718e-06,
+ "loss": 0.3067188262939453,
+ "num_input_tokens_seen": 977777560,
+ "step": 22360,
+ "train_runtime": 172910.0558,
+ "train_tokens_per_second": 5654.833
+ },
+ {
+ "epoch": 1.5058577969297064,
+ "grad_norm": 0.7529457682896904,
+ "learning_rate": 1.432454832230426e-06,
+ "loss": 0.28587291240692136,
+ "num_input_tokens_seen": 978009320,
+ "step": 22365,
+ "train_runtime": 172953.4465,
+ "train_tokens_per_second": 5654.755
+ },
+ {
+ "epoch": 1.5061944519256665,
+ "grad_norm": 0.6792382751676833,
+ "learning_rate": 1.4306027654786103e-06,
+ "loss": 0.28476083278656006,
+ "num_input_tokens_seen": 978236352,
+ "step": 22370,
+ "train_runtime": 172991.2837,
+ "train_tokens_per_second": 5654.83
+ },
+ {
+ "epoch": 1.5065311069216267,
+ "grad_norm": 0.7207520305966407,
+ "learning_rate": 1.4287516968973491e-06,
+ "loss": 0.3085874080657959,
+ "num_input_tokens_seen": 978453896,
+ "step": 22375,
+ "train_runtime": 173028.6858,
+ "train_tokens_per_second": 5654.865
+ },
+ {
+ "epoch": 1.506867761917587,
+ "grad_norm": 0.816106376696343,
+ "learning_rate": 1.4269016270042884e-06,
+ "loss": 0.3164313077926636,
+ "num_input_tokens_seen": 978673624,
+ "step": 22380,
+ "train_runtime": 173069.187,
+ "train_tokens_per_second": 5654.811
+ },
+ {
+ "epoch": 1.507204416913547,
+ "grad_norm": 0.7895982117278676,
+ "learning_rate": 1.425052556316794e-06,
+ "loss": 0.2872413873672485,
+ "num_input_tokens_seen": 978891952,
+ "step": 22385,
+ "train_runtime": 173110.0103,
+ "train_tokens_per_second": 5654.739
+ },
+ {
+ "epoch": 1.507541071909507,
+ "grad_norm": 0.9301659133996552,
+ "learning_rate": 1.4232044853519522e-06,
+ "loss": 0.3223658323287964,
+ "num_input_tokens_seen": 979097208,
+ "step": 22390,
+ "train_runtime": 173149.0058,
+ "train_tokens_per_second": 5654.651
+ },
+ {
+ "epoch": 1.5078777269054673,
+ "grad_norm": 0.6813929579263441,
+ "learning_rate": 1.4213574146265708e-06,
+ "loss": 0.28975491523742675,
+ "num_input_tokens_seen": 979304416,
+ "step": 22395,
+ "train_runtime": 173184.2372,
+ "train_tokens_per_second": 5654.697
+ },
+ {
+ "epoch": 1.5082143819014275,
+ "grad_norm": 0.7646303429188873,
+ "learning_rate": 1.4195113446571752e-06,
+ "loss": 0.32438082695007325,
+ "num_input_tokens_seen": 979527856,
+ "step": 22400,
+ "train_runtime": 173219.7965,
+ "train_tokens_per_second": 5654.826
+ },
+ {
+ "epoch": 1.5085510368973876,
+ "grad_norm": 0.7391053441959036,
+ "learning_rate": 1.417666275960018e-06,
+ "loss": 0.2848972797393799,
+ "num_input_tokens_seen": 979761880,
+ "step": 22405,
+ "train_runtime": 173250.7995,
+ "train_tokens_per_second": 5655.165
+ },
+ {
+ "epoch": 1.5088876918933476,
+ "grad_norm": 0.6903765320662747,
+ "learning_rate": 1.4158222090510605e-06,
+ "loss": 0.27116708755493163,
+ "num_input_tokens_seen": 979990616,
+ "step": 22410,
+ "train_runtime": 173289.5475,
+ "train_tokens_per_second": 5655.221
+ },
+ {
+ "epoch": 1.5092243468893078,
+ "grad_norm": 0.7236456392467262,
+ "learning_rate": 1.4139791444459945e-06,
+ "loss": 0.3101805686950684,
+ "num_input_tokens_seen": 980209928,
+ "step": 22415,
+ "train_runtime": 173323.1137,
+ "train_tokens_per_second": 5655.391
+ },
+ {
+ "epoch": 1.509561001885268,
+ "grad_norm": 0.8526682680286438,
+ "learning_rate": 1.4121370826602254e-06,
+ "loss": 0.3069303512573242,
+ "num_input_tokens_seen": 980417304,
+ "step": 22420,
+ "train_runtime": 173363.0188,
+ "train_tokens_per_second": 5655.285
+ },
+ {
+ "epoch": 1.5098976568812281,
+ "grad_norm": 0.7042288015488988,
+ "learning_rate": 1.4102960242088803e-06,
+ "loss": 0.3054825782775879,
+ "num_input_tokens_seen": 980633968,
+ "step": 22425,
+ "train_runtime": 173407.4082,
+ "train_tokens_per_second": 5655.087
+ },
+ {
+ "epoch": 1.5102343118771882,
+ "grad_norm": 0.7504743822892517,
+ "learning_rate": 1.4084559696068057e-06,
+ "loss": 0.28913288116455077,
+ "num_input_tokens_seen": 980850136,
+ "step": 22430,
+ "train_runtime": 173445.4607,
+ "train_tokens_per_second": 5655.093
+ },
+ {
+ "epoch": 1.5105709668731484,
+ "grad_norm": 0.7252926106988145,
+ "learning_rate": 1.406616919368567e-06,
+ "loss": 0.2985984325408936,
+ "num_input_tokens_seen": 981066808,
+ "step": 22435,
+ "train_runtime": 173488.0196,
+ "train_tokens_per_second": 5654.954
+ },
+ {
+ "epoch": 1.5109076218691087,
+ "grad_norm": 0.801327813808047,
+ "learning_rate": 1.4047788740084474e-06,
+ "loss": 0.29711313247680665,
+ "num_input_tokens_seen": 981290632,
+ "step": 22440,
+ "train_runtime": 173530.0228,
+ "train_tokens_per_second": 5654.875
+ },
+ {
+ "epoch": 1.5112442768650687,
+ "grad_norm": 0.6984253644671601,
+ "learning_rate": 1.402941834040452e-06,
+ "loss": 0.2857701539993286,
+ "num_input_tokens_seen": 981524520,
+ "step": 22445,
+ "train_runtime": 173571.2058,
+ "train_tokens_per_second": 5654.881
+ },
+ {
+ "epoch": 1.5115809318610287,
+ "grad_norm": 0.7193805868113159,
+ "learning_rate": 1.4011057999783006e-06,
+ "loss": 0.28029732704162597,
+ "num_input_tokens_seen": 981733352,
+ "step": 22450,
+ "train_runtime": 173613.2142,
+ "train_tokens_per_second": 5654.716
+ },
+ {
+ "epoch": 1.511917586856989,
+ "grad_norm": 0.7608894196010775,
+ "learning_rate": 1.3992707723354398e-06,
+ "loss": 0.3050365924835205,
+ "num_input_tokens_seen": 981955216,
+ "step": 22455,
+ "train_runtime": 173651.5688,
+ "train_tokens_per_second": 5654.744
+ },
+ {
+ "epoch": 1.5122542418529492,
+ "grad_norm": 0.7183358803573994,
+ "learning_rate": 1.397436751625022e-06,
+ "loss": 0.2846760988235474,
+ "num_input_tokens_seen": 982166184,
+ "step": 22460,
+ "train_runtime": 173692.0688,
+ "train_tokens_per_second": 5654.64
+ },
+ {
+ "epoch": 1.5125908968489092,
+ "grad_norm": 0.8684762367144154,
+ "learning_rate": 1.3956037383599308e-06,
+ "loss": 0.2913294792175293,
+ "num_input_tokens_seen": 982385952,
+ "step": 22465,
+ "train_runtime": 173731.1563,
+ "train_tokens_per_second": 5654.633
+ },
+ {
+ "epoch": 1.5129275518448693,
+ "grad_norm": 0.7292786314768155,
+ "learning_rate": 1.39377173305276e-06,
+ "loss": 0.27708570957183837,
+ "num_input_tokens_seen": 982592624,
+ "step": 22470,
+ "train_runtime": 173769.1128,
+ "train_tokens_per_second": 5654.587
+ },
+ {
+ "epoch": 1.5132642068408295,
+ "grad_norm": 0.7362638583219614,
+ "learning_rate": 1.3919407362158254e-06,
+ "loss": 0.3209893465042114,
+ "num_input_tokens_seen": 982825112,
+ "step": 22475,
+ "train_runtime": 173806.2328,
+ "train_tokens_per_second": 5654.717
+ },
+ {
+ "epoch": 1.5136008618367898,
+ "grad_norm": 0.8685451291862974,
+ "learning_rate": 1.3901107483611582e-06,
+ "loss": 0.2871732711791992,
+ "num_input_tokens_seen": 983031648,
+ "step": 22480,
+ "train_runtime": 173844.6509,
+ "train_tokens_per_second": 5654.656
+ },
+ {
+ "epoch": 1.5139375168327498,
+ "grad_norm": 0.7432181034545295,
+ "learning_rate": 1.3882817700005091e-06,
+ "loss": 0.2984759330749512,
+ "num_input_tokens_seen": 983247128,
+ "step": 22485,
+ "train_runtime": 173892.4519,
+ "train_tokens_per_second": 5654.34
+ },
+ {
+ "epoch": 1.5142741718287098,
+ "grad_norm": 0.7851212386445382,
+ "learning_rate": 1.3864538016453466e-06,
+ "loss": 0.3003609418869019,
+ "num_input_tokens_seen": 983445264,
+ "step": 22490,
+ "train_runtime": 173929.1801,
+ "train_tokens_per_second": 5654.286
+ },
+ {
+ "epoch": 1.51461082682467,
+ "grad_norm": 0.929685801907135,
+ "learning_rate": 1.384626843806855e-06,
+ "loss": 0.304855489730835,
+ "num_input_tokens_seen": 983646472,
+ "step": 22495,
+ "train_runtime": 173969.1302,
+ "train_tokens_per_second": 5654.144
+ },
+ {
+ "epoch": 1.5149474818206303,
+ "grad_norm": 0.7557647415601482,
+ "learning_rate": 1.3828008969959367e-06,
+ "loss": 0.31099758148193357,
+ "num_input_tokens_seen": 983875960,
+ "step": 22500,
+ "train_runtime": 174009.2793,
+ "train_tokens_per_second": 5654.158
+ },
+ {
+ "epoch": 1.5152841368165904,
+ "grad_norm": 0.850008061053931,
+ "learning_rate": 1.3809759617232166e-06,
+ "loss": 0.2885129928588867,
+ "num_input_tokens_seen": 984089952,
+ "step": 22505,
+ "train_runtime": 174050.4492,
+ "train_tokens_per_second": 5654.05
+ },
+ {
+ "epoch": 1.5156207918125504,
+ "grad_norm": 0.7124328658676645,
+ "learning_rate": 1.3791520384990247e-06,
+ "loss": 0.29050498008728026,
+ "num_input_tokens_seen": 984303984,
+ "step": 22510,
+ "train_runtime": 174085.4118,
+ "train_tokens_per_second": 5654.144
+ },
+ {
+ "epoch": 1.5159574468085106,
+ "grad_norm": 0.6783534298885123,
+ "learning_rate": 1.3773291278334216e-06,
+ "loss": 0.2980940818786621,
+ "num_input_tokens_seen": 984518832,
+ "step": 22515,
+ "train_runtime": 174122.6841,
+ "train_tokens_per_second": 5654.168
+ },
+ {
+ "epoch": 1.5162941018044709,
+ "grad_norm": 0.772415132712806,
+ "learning_rate": 1.3755072302361754e-06,
+ "loss": 0.3086568355560303,
+ "num_input_tokens_seen": 984751312,
+ "step": 22520,
+ "train_runtime": 174163.3837,
+ "train_tokens_per_second": 5654.181
+ },
+ {
+ "epoch": 1.516630756800431,
+ "grad_norm": 0.7777502241934342,
+ "learning_rate": 1.3736863462167749e-06,
+ "loss": 0.28640379905700686,
+ "num_input_tokens_seen": 984953136,
+ "step": 22525,
+ "train_runtime": 174203.7264,
+ "train_tokens_per_second": 5654.03
+ },
+ {
+ "epoch": 1.516967411796391,
+ "grad_norm": 0.7811964811897554,
+ "learning_rate": 1.3718664762844242e-06,
+ "loss": 0.28820300102233887,
+ "num_input_tokens_seen": 985162064,
+ "step": 22530,
+ "train_runtime": 174238.0664,
+ "train_tokens_per_second": 5654.115
+ },
+ {
+ "epoch": 1.5173040667923512,
+ "grad_norm": 0.7667814818121892,
+ "learning_rate": 1.370047620948044e-06,
+ "loss": 0.2942743062973022,
+ "num_input_tokens_seen": 985382480,
+ "step": 22535,
+ "train_runtime": 174279.1667,
+ "train_tokens_per_second": 5654.046
+ },
+ {
+ "epoch": 1.5176407217883114,
+ "grad_norm": 0.7745949128302583,
+ "learning_rate": 1.368229780716272e-06,
+ "loss": 0.3100618839263916,
+ "num_input_tokens_seen": 985602688,
+ "step": 22540,
+ "train_runtime": 174310.4991,
+ "train_tokens_per_second": 5654.293
+ },
+ {
+ "epoch": 1.5179773767842715,
+ "grad_norm": 0.8932217604026287,
+ "learning_rate": 1.3664129560974604e-06,
+ "loss": 0.2899731397628784,
+ "num_input_tokens_seen": 985820032,
+ "step": 22545,
+ "train_runtime": 174346.7623,
+ "train_tokens_per_second": 5654.364
+ },
+ {
+ "epoch": 1.5183140317802315,
+ "grad_norm": 0.7820206063355695,
+ "learning_rate": 1.3645971475996777e-06,
+ "loss": 0.29310288429260256,
+ "num_input_tokens_seen": 986025552,
+ "step": 22550,
+ "train_runtime": 174382.5794,
+ "train_tokens_per_second": 5654.381
+ },
+ {
+ "epoch": 1.5186506867761917,
+ "grad_norm": 0.8104637399237582,
+ "learning_rate": 1.3627823557307135e-06,
+ "loss": 0.29608545303344724,
+ "num_input_tokens_seen": 986257600,
+ "step": 22555,
+ "train_runtime": 174424.2399,
+ "train_tokens_per_second": 5654.361
+ },
+ {
+ "epoch": 1.518987341772152,
+ "grad_norm": 0.7046023346650935,
+ "learning_rate": 1.3609685809980616e-06,
+ "loss": 0.3058366298675537,
+ "num_input_tokens_seen": 986486440,
+ "step": 22560,
+ "train_runtime": 174466.7289,
+ "train_tokens_per_second": 5654.295
+ },
+ {
+ "epoch": 1.519323996768112,
+ "grad_norm": 0.7418847854253624,
+ "learning_rate": 1.3591558239089443e-06,
+ "loss": 0.2807319164276123,
+ "num_input_tokens_seen": 986707576,
+ "step": 22565,
+ "train_runtime": 174512.4598,
+ "train_tokens_per_second": 5654.081
+ },
+ {
+ "epoch": 1.519660651764072,
+ "grad_norm": 0.8450682344923751,
+ "learning_rate": 1.3573440849702902e-06,
+ "loss": 0.317004919052124,
+ "num_input_tokens_seen": 986937912,
+ "step": 22570,
+ "train_runtime": 174554.2288,
+ "train_tokens_per_second": 5654.048
+ },
+ {
+ "epoch": 1.5199973067600323,
+ "grad_norm": 0.7386948536050674,
+ "learning_rate": 1.3555333646887476e-06,
+ "loss": 0.3014618635177612,
+ "num_input_tokens_seen": 987163824,
+ "step": 22575,
+ "train_runtime": 174589.7315,
+ "train_tokens_per_second": 5654.192
+ },
+ {
+ "epoch": 1.5203339617559926,
+ "grad_norm": 0.7377369303219051,
+ "learning_rate": 1.3537236635706779e-06,
+ "loss": 0.29477739334106445,
+ "num_input_tokens_seen": 987396616,
+ "step": 22580,
+ "train_runtime": 174623.2763,
+ "train_tokens_per_second": 5654.439
+ },
+ {
+ "epoch": 1.5206706167519526,
+ "grad_norm": 0.8578989985964252,
+ "learning_rate": 1.351914982122159e-06,
+ "loss": 0.27982220649719236,
+ "num_input_tokens_seen": 987623048,
+ "step": 22585,
+ "train_runtime": 174659.6001,
+ "train_tokens_per_second": 5654.559
+ },
+ {
+ "epoch": 1.5210072717479126,
+ "grad_norm": 0.7750049777273713,
+ "learning_rate": 1.3501073208489823e-06,
+ "loss": 0.3012231349945068,
+ "num_input_tokens_seen": 987842064,
+ "step": 22590,
+ "train_runtime": 174696.8944,
+ "train_tokens_per_second": 5654.606
+ },
+ {
+ "epoch": 1.5213439267438729,
+ "grad_norm": 0.7792774723783644,
+ "learning_rate": 1.3483006802566546e-06,
+ "loss": 0.29522716999053955,
+ "num_input_tokens_seen": 988065680,
+ "step": 22595,
+ "train_runtime": 174733.3309,
+ "train_tokens_per_second": 5654.706
+ },
+ {
+ "epoch": 1.5216805817398331,
+ "grad_norm": 0.8105588967824191,
+ "learning_rate": 1.3464950608503957e-06,
+ "loss": 0.3093279838562012,
+ "num_input_tokens_seen": 988279680,
+ "step": 22600,
+ "train_runtime": 174772.1153,
+ "train_tokens_per_second": 5654.676
+ },
+ {
+ "epoch": 1.5220172367357931,
+ "grad_norm": 0.8485409938047664,
+ "learning_rate": 1.344690463135146e-06,
+ "loss": 0.30634384155273436,
+ "num_input_tokens_seen": 988495592,
+ "step": 22605,
+ "train_runtime": 174817.3086,
+ "train_tokens_per_second": 5654.449
+ },
+ {
+ "epoch": 1.5223538917317532,
+ "grad_norm": 0.7715064487874997,
+ "learning_rate": 1.3428868876155493e-06,
+ "loss": 0.3291673183441162,
+ "num_input_tokens_seen": 988698000,
+ "step": 22610,
+ "train_runtime": 174853.7041,
+ "train_tokens_per_second": 5654.43
+ },
+ {
+ "epoch": 1.5226905467277134,
+ "grad_norm": 0.6665959654227848,
+ "learning_rate": 1.3410843347959745e-06,
+ "loss": 0.2833380699157715,
+ "num_input_tokens_seen": 988914088,
+ "step": 22615,
+ "train_runtime": 174891.5132,
+ "train_tokens_per_second": 5654.443
+ },
+ {
+ "epoch": 1.5230272017236737,
+ "grad_norm": 0.7178804958210366,
+ "learning_rate": 1.3392828051804979e-06,
+ "loss": 0.3043062686920166,
+ "num_input_tokens_seen": 989141848,
+ "step": 22620,
+ "train_runtime": 174934.5815,
+ "train_tokens_per_second": 5654.353
+ },
+ {
+ "epoch": 1.5233638567196337,
+ "grad_norm": 0.7316276122750418,
+ "learning_rate": 1.3374822992729114e-06,
+ "loss": 0.3176832437515259,
+ "num_input_tokens_seen": 989359560,
+ "step": 22625,
+ "train_runtime": 174977.0909,
+ "train_tokens_per_second": 5654.223
+ },
+ {
+ "epoch": 1.5237005117155937,
+ "grad_norm": 0.8151737151768202,
+ "learning_rate": 1.3356828175767212e-06,
+ "loss": 0.30327634811401366,
+ "num_input_tokens_seen": 989566784,
+ "step": 22630,
+ "train_runtime": 175016.0102,
+ "train_tokens_per_second": 5654.15
+ },
+ {
+ "epoch": 1.524037166711554,
+ "grad_norm": 0.7458938184161533,
+ "learning_rate": 1.3338843605951462e-06,
+ "loss": 0.27550592422485354,
+ "num_input_tokens_seen": 989793688,
+ "step": 22635,
+ "train_runtime": 175058.1324,
+ "train_tokens_per_second": 5654.086
+ },
+ {
+ "epoch": 1.5243738217075142,
+ "grad_norm": 0.7254475228515918,
+ "learning_rate": 1.33208692883112e-06,
+ "loss": 0.2927609205245972,
+ "num_input_tokens_seen": 990030416,
+ "step": 22640,
+ "train_runtime": 175099.372,
+ "train_tokens_per_second": 5654.106
+ },
+ {
+ "epoch": 1.5247104767034743,
+ "grad_norm": 0.6451473463172038,
+ "learning_rate": 1.3302905227872876e-06,
+ "loss": 0.2735952377319336,
+ "num_input_tokens_seen": 990266880,
+ "step": 22645,
+ "train_runtime": 175138.0414,
+ "train_tokens_per_second": 5654.208
+ },
+ {
+ "epoch": 1.5250471316994343,
+ "grad_norm": 0.7523183374053246,
+ "learning_rate": 1.328495142966007e-06,
+ "loss": 0.3053655862808228,
+ "num_input_tokens_seen": 990479584,
+ "step": 22650,
+ "train_runtime": 175176.6743,
+ "train_tokens_per_second": 5654.175
+ },
+ {
+ "epoch": 1.5253837866953945,
+ "grad_norm": 0.7119868993623765,
+ "learning_rate": 1.3267007898693552e-06,
+ "loss": 0.29980125427246096,
+ "num_input_tokens_seen": 990712176,
+ "step": 22655,
+ "train_runtime": 175223.5228,
+ "train_tokens_per_second": 5653.991
+ },
+ {
+ "epoch": 1.5257204416913548,
+ "grad_norm": 0.7695418964162357,
+ "learning_rate": 1.324907463999111e-06,
+ "loss": 0.3055927276611328,
+ "num_input_tokens_seen": 990938448,
+ "step": 22660,
+ "train_runtime": 175261.0683,
+ "train_tokens_per_second": 5654.071
+ },
+ {
+ "epoch": 1.5260570966873148,
+ "grad_norm": 0.750866798684191,
+ "learning_rate": 1.3231151658567776e-06,
+ "loss": 0.3090615510940552,
+ "num_input_tokens_seen": 991152664,
+ "step": 22665,
+ "train_runtime": 175294.7905,
+ "train_tokens_per_second": 5654.205
+ },
+ {
+ "epoch": 1.5263937516832748,
+ "grad_norm": 0.7700423401788759,
+ "learning_rate": 1.321323895943563e-06,
+ "loss": 0.3019981861114502,
+ "num_input_tokens_seen": 991370424,
+ "step": 22670,
+ "train_runtime": 175335.3762,
+ "train_tokens_per_second": 5654.138
+ },
+ {
+ "epoch": 1.526730406679235,
+ "grad_norm": 0.6944631282349509,
+ "learning_rate": 1.3195336547603905e-06,
+ "loss": 0.31381664276123045,
+ "num_input_tokens_seen": 991588192,
+ "step": 22675,
+ "train_runtime": 175378.1626,
+ "train_tokens_per_second": 5654.0
+ },
+ {
+ "epoch": 1.5270670616751953,
+ "grad_norm": 0.711842537575573,
+ "learning_rate": 1.3177444428078951e-06,
+ "loss": 0.28606882095336916,
+ "num_input_tokens_seen": 991798856,
+ "step": 22680,
+ "train_runtime": 175413.7546,
+ "train_tokens_per_second": 5654.054
+ },
+ {
+ "epoch": 1.5274037166711554,
+ "grad_norm": 0.815411302830813,
+ "learning_rate": 1.3159562605864245e-06,
+ "loss": 0.30185737609863283,
+ "num_input_tokens_seen": 992026624,
+ "step": 22685,
+ "train_runtime": 175457.606,
+ "train_tokens_per_second": 5653.939
+ },
+ {
+ "epoch": 1.5277403716671154,
+ "grad_norm": 0.7981498875678553,
+ "learning_rate": 1.3141691085960385e-06,
+ "loss": 0.29338903427124025,
+ "num_input_tokens_seen": 992265864,
+ "step": 22690,
+ "train_runtime": 175496.9155,
+ "train_tokens_per_second": 5654.036
+ },
+ {
+ "epoch": 1.5280770266630757,
+ "grad_norm": 0.7568915202824599,
+ "learning_rate": 1.3123829873365073e-06,
+ "loss": 0.2825155258178711,
+ "num_input_tokens_seen": 992478536,
+ "step": 22695,
+ "train_runtime": 175535.4671,
+ "train_tokens_per_second": 5654.006
+ },
+ {
+ "epoch": 1.528413681659036,
+ "grad_norm": 0.7962079639131174,
+ "learning_rate": 1.3105978973073136e-06,
+ "loss": 0.2889784097671509,
+ "num_input_tokens_seen": 992692016,
+ "step": 22700,
+ "train_runtime": 175577.4777,
+ "train_tokens_per_second": 5653.869
+ },
+ {
+ "epoch": 1.528750336654996,
+ "grad_norm": 0.7986814122905842,
+ "learning_rate": 1.3088138390076565e-06,
+ "loss": 0.3144156694412231,
+ "num_input_tokens_seen": 992915000,
+ "step": 22705,
+ "train_runtime": 175610.667,
+ "train_tokens_per_second": 5654.07
+ },
+ {
+ "epoch": 1.529086991650956,
+ "grad_norm": 0.784101134617306,
+ "learning_rate": 1.3070308129364357e-06,
+ "loss": 0.28853483200073243,
+ "num_input_tokens_seen": 993131448,
+ "step": 22710,
+ "train_runtime": 175653.2502,
+ "train_tokens_per_second": 5653.932
+ },
+ {
+ "epoch": 1.5294236466469162,
+ "grad_norm": 0.8613062066807416,
+ "learning_rate": 1.3052488195922736e-06,
+ "loss": 0.28805036544799806,
+ "num_input_tokens_seen": 993337608,
+ "step": 22715,
+ "train_runtime": 175693.4521,
+ "train_tokens_per_second": 5653.811
+ },
+ {
+ "epoch": 1.5297603016428765,
+ "grad_norm": 0.7554420455835269,
+ "learning_rate": 1.3034678594734972e-06,
+ "loss": 0.2852388620376587,
+ "num_input_tokens_seen": 993560896,
+ "step": 22720,
+ "train_runtime": 175731.737,
+ "train_tokens_per_second": 5653.85
+ },
+ {
+ "epoch": 1.5300969566388365,
+ "grad_norm": 0.8715494762310279,
+ "learning_rate": 1.3016879330781468e-06,
+ "loss": 0.30547394752502444,
+ "num_input_tokens_seen": 993766904,
+ "step": 22725,
+ "train_runtime": 175768.7081,
+ "train_tokens_per_second": 5653.833
+ },
+ {
+ "epoch": 1.5304336116347965,
+ "grad_norm": 0.7476831667764691,
+ "learning_rate": 1.2999090409039728e-06,
+ "loss": 0.29537832736968994,
+ "num_input_tokens_seen": 993988440,
+ "step": 22730,
+ "train_runtime": 175806.2319,
+ "train_tokens_per_second": 5653.886
+ },
+ {
+ "epoch": 1.5307702666307568,
+ "grad_norm": 0.7728242495534967,
+ "learning_rate": 1.2981311834484367e-06,
+ "loss": 0.30858845710754396,
+ "num_input_tokens_seen": 994207440,
+ "step": 22735,
+ "train_runtime": 175851.9938,
+ "train_tokens_per_second": 5653.66
+ },
+ {
+ "epoch": 1.531106921626717,
+ "grad_norm": 0.7739386600524095,
+ "learning_rate": 1.296354361208711e-06,
+ "loss": 0.28605356216430666,
+ "num_input_tokens_seen": 994421664,
+ "step": 22740,
+ "train_runtime": 175891.4348,
+ "train_tokens_per_second": 5653.611
+ },
+ {
+ "epoch": 1.531443576622677,
+ "grad_norm": 0.8012741513918233,
+ "learning_rate": 1.294578574681678e-06,
+ "loss": 0.3216970920562744,
+ "num_input_tokens_seen": 994629248,
+ "step": 22745,
+ "train_runtime": 175928.8973,
+ "train_tokens_per_second": 5653.587
+ },
+ {
+ "epoch": 1.531780231618637,
+ "grad_norm": 0.7817355354702294,
+ "learning_rate": 1.292803824363929e-06,
+ "loss": 0.28469209671020507,
+ "num_input_tokens_seen": 994850920,
+ "step": 22750,
+ "train_runtime": 175967.4254,
+ "train_tokens_per_second": 5653.608
+ },
+ {
+ "epoch": 1.5321168866145973,
+ "grad_norm": 0.767294892049272,
+ "learning_rate": 1.2910301107517726e-06,
+ "loss": 0.3232132434844971,
+ "num_input_tokens_seen": 995074552,
+ "step": 22755,
+ "train_runtime": 176008.5181,
+ "train_tokens_per_second": 5653.559
+ },
+ {
+ "epoch": 1.5324535416105576,
+ "grad_norm": 0.8124734532106183,
+ "learning_rate": 1.2892574343412156e-06,
+ "loss": 0.319903826713562,
+ "num_input_tokens_seen": 995294736,
+ "step": 22760,
+ "train_runtime": 176043.0639,
+ "train_tokens_per_second": 5653.7
+ },
+ {
+ "epoch": 1.5327901966065176,
+ "grad_norm": 0.7450430227623678,
+ "learning_rate": 1.2874857956279858e-06,
+ "loss": 0.3038845777511597,
+ "num_input_tokens_seen": 995505008,
+ "step": 22765,
+ "train_runtime": 176081.1467,
+ "train_tokens_per_second": 5653.672
+ },
+ {
+ "epoch": 1.5331268516024776,
+ "grad_norm": 0.7872978056072006,
+ "learning_rate": 1.2857151951075147e-06,
+ "loss": 0.3022770404815674,
+ "num_input_tokens_seen": 995729520,
+ "step": 22770,
+ "train_runtime": 176120.5406,
+ "train_tokens_per_second": 5653.682
+ },
+ {
+ "epoch": 1.5334635065984379,
+ "grad_norm": 0.7298470919630958,
+ "learning_rate": 1.2839456332749456e-06,
+ "loss": 0.2998102903366089,
+ "num_input_tokens_seen": 995943808,
+ "step": 22775,
+ "train_runtime": 176158.1432,
+ "train_tokens_per_second": 5653.692
+ },
+ {
+ "epoch": 1.5338001615943981,
+ "grad_norm": 0.8031049157762769,
+ "learning_rate": 1.2821771106251308e-06,
+ "loss": 0.2987299919128418,
+ "num_input_tokens_seen": 996160472,
+ "step": 22780,
+ "train_runtime": 176194.5973,
+ "train_tokens_per_second": 5653.752
+ },
+ {
+ "epoch": 1.5341368165903582,
+ "grad_norm": 0.7596275846497683,
+ "learning_rate": 1.2804096276526312e-06,
+ "loss": 0.27775073051452637,
+ "num_input_tokens_seen": 996378168,
+ "step": 22785,
+ "train_runtime": 176231.0577,
+ "train_tokens_per_second": 5653.817
+ },
+ {
+ "epoch": 1.5344734715863182,
+ "grad_norm": 0.8239381536070142,
+ "learning_rate": 1.2786431848517184e-06,
+ "loss": 0.2855050086975098,
+ "num_input_tokens_seen": 996584112,
+ "step": 22790,
+ "train_runtime": 176267.1349,
+ "train_tokens_per_second": 5653.828
+ },
+ {
+ "epoch": 1.5348101265822784,
+ "grad_norm": 0.7589447815362402,
+ "learning_rate": 1.2768777827163725e-06,
+ "loss": 0.29274682998657225,
+ "num_input_tokens_seen": 996797912,
+ "step": 22795,
+ "train_runtime": 176305.2925,
+ "train_tokens_per_second": 5653.817
+ },
+ {
+ "epoch": 1.5351467815782387,
+ "grad_norm": 0.8140404323951491,
+ "learning_rate": 1.2751134217402811e-06,
+ "loss": 0.29671099185943606,
+ "num_input_tokens_seen": 997003488,
+ "step": 22800,
+ "train_runtime": 176347.2741,
+ "train_tokens_per_second": 5653.637
+ },
+ {
+ "epoch": 1.5354834365741987,
+ "grad_norm": 0.7258016591513569,
+ "learning_rate": 1.2733501024168465e-06,
+ "loss": 0.2848902940750122,
+ "num_input_tokens_seen": 997225312,
+ "step": 22805,
+ "train_runtime": 176381.4588,
+ "train_tokens_per_second": 5653.799
+ },
+ {
+ "epoch": 1.5358200915701588,
+ "grad_norm": 0.8298549659043899,
+ "learning_rate": 1.2715878252391695e-06,
+ "loss": 0.30264275074005126,
+ "num_input_tokens_seen": 997430736,
+ "step": 22810,
+ "train_runtime": 176421.3628,
+ "train_tokens_per_second": 5653.685
+ },
+ {
+ "epoch": 1.536156746566119,
+ "grad_norm": 0.7905575392804943,
+ "learning_rate": 1.269826590700069e-06,
+ "loss": 0.3026357412338257,
+ "num_input_tokens_seen": 997650400,
+ "step": 22815,
+ "train_runtime": 176454.177,
+ "train_tokens_per_second": 5653.878
+ },
+ {
+ "epoch": 1.5364934015620793,
+ "grad_norm": 0.7167147078330245,
+ "learning_rate": 1.2680663992920684e-06,
+ "loss": 0.2799959659576416,
+ "num_input_tokens_seen": 997872240,
+ "step": 22820,
+ "train_runtime": 176495.7097,
+ "train_tokens_per_second": 5653.805
+ },
+ {
+ "epoch": 1.5368300565580393,
+ "grad_norm": 0.8394660285508826,
+ "learning_rate": 1.2663072515073988e-06,
+ "loss": 0.30673794746398925,
+ "num_input_tokens_seen": 998084992,
+ "step": 22825,
+ "train_runtime": 176535.1094,
+ "train_tokens_per_second": 5653.748
+ },
+ {
+ "epoch": 1.5371667115539993,
+ "grad_norm": 0.819012798672818,
+ "learning_rate": 1.2645491478380002e-06,
+ "loss": 0.29927687644958495,
+ "num_input_tokens_seen": 998297904,
+ "step": 22830,
+ "train_runtime": 176577.7119,
+ "train_tokens_per_second": 5653.59
+ },
+ {
+ "epoch": 1.5375033665499596,
+ "grad_norm": 0.8389462503007052,
+ "learning_rate": 1.2627920887755212e-06,
+ "loss": 0.3054951190948486,
+ "num_input_tokens_seen": 998508592,
+ "step": 22835,
+ "train_runtime": 176613.3774,
+ "train_tokens_per_second": 5653.641
+ },
+ {
+ "epoch": 1.5378400215459198,
+ "grad_norm": 0.7157688346659142,
+ "learning_rate": 1.261036074811317e-06,
+ "loss": 0.3088068962097168,
+ "num_input_tokens_seen": 998733664,
+ "step": 22840,
+ "train_runtime": 176645.3708,
+ "train_tokens_per_second": 5653.891
+ },
+ {
+ "epoch": 1.5381766765418798,
+ "grad_norm": 0.7683443200543241,
+ "learning_rate": 1.2592811064364524e-06,
+ "loss": 0.2938983917236328,
+ "num_input_tokens_seen": 998954872,
+ "step": 22845,
+ "train_runtime": 176689.3099,
+ "train_tokens_per_second": 5653.737
+ },
+ {
+ "epoch": 1.5385133315378399,
+ "grad_norm": 1.0039919983392351,
+ "learning_rate": 1.2575271841416963e-06,
+ "loss": 0.3196579456329346,
+ "num_input_tokens_seen": 999145416,
+ "step": 22850,
+ "train_runtime": 176727.2813,
+ "train_tokens_per_second": 5653.6
+ },
+ {
+ "epoch": 1.5388499865338001,
+ "grad_norm": 0.7489841192691534,
+ "learning_rate": 1.2557743084175327e-06,
+ "loss": 0.27573988437652586,
+ "num_input_tokens_seen": 999360320,
+ "step": 22855,
+ "train_runtime": 176762.6012,
+ "train_tokens_per_second": 5653.686
+ },
+ {
+ "epoch": 1.5391866415297604,
+ "grad_norm": 0.7473129975200795,
+ "learning_rate": 1.254022479754141e-06,
+ "loss": 0.2842655420303345,
+ "num_input_tokens_seen": 999588440,
+ "step": 22860,
+ "train_runtime": 176797.2452,
+ "train_tokens_per_second": 5653.869
+ },
+ {
+ "epoch": 1.5395232965257204,
+ "grad_norm": 0.7459689277622236,
+ "learning_rate": 1.252271698641419e-06,
+ "loss": 0.28894410133361814,
+ "num_input_tokens_seen": 999804216,
+ "step": 22865,
+ "train_runtime": 176835.9913,
+ "train_tokens_per_second": 5653.85
+ },
+ {
+ "epoch": 1.5398599515216804,
+ "grad_norm": 0.7252321531655411,
+ "learning_rate": 1.2505219655689659e-06,
+ "loss": 0.310650110244751,
+ "num_input_tokens_seen": 1000033680,
+ "step": 22870,
+ "train_runtime": 176868.215,
+ "train_tokens_per_second": 5654.118
+ },
+ {
+ "epoch": 1.5401966065176407,
+ "grad_norm": 0.7797088059008812,
+ "learning_rate": 1.2487732810260877e-06,
+ "loss": 0.3046865701675415,
+ "num_input_tokens_seen": 1000253168,
+ "step": 22875,
+ "train_runtime": 176907.967,
+ "train_tokens_per_second": 5654.088
+ },
+ {
+ "epoch": 1.540533261513601,
+ "grad_norm": 0.6927104452813211,
+ "learning_rate": 1.2470256455018027e-06,
+ "loss": 0.2814844846725464,
+ "num_input_tokens_seen": 1000469584,
+ "step": 22880,
+ "train_runtime": 176948.9419,
+ "train_tokens_per_second": 5654.002
+ },
+ {
+ "epoch": 1.540869916509561,
+ "grad_norm": 0.7567591660729376,
+ "learning_rate": 1.2452790594848257e-06,
+ "loss": 0.28022444248199463,
+ "num_input_tokens_seen": 1000680152,
+ "step": 22885,
+ "train_runtime": 176994.3264,
+ "train_tokens_per_second": 5653.741
+ },
+ {
+ "epoch": 1.541206571505521,
+ "grad_norm": 0.8836138624724967,
+ "learning_rate": 1.2435335234635892e-06,
+ "loss": 0.3237956762313843,
+ "num_input_tokens_seen": 1000898768,
+ "step": 22890,
+ "train_runtime": 177031.9969,
+ "train_tokens_per_second": 5653.773
+ },
+ {
+ "epoch": 1.5415432265014812,
+ "grad_norm": 0.8098774986288102,
+ "learning_rate": 1.2417890379262221e-06,
+ "loss": 0.29699585437774656,
+ "num_input_tokens_seen": 1001118440,
+ "step": 22895,
+ "train_runtime": 177077.1356,
+ "train_tokens_per_second": 5653.573
+ },
+ {
+ "epoch": 1.5418798814974415,
+ "grad_norm": 0.7965414488198174,
+ "learning_rate": 1.2400456033605672e-06,
+ "loss": 0.26516332626342776,
+ "num_input_tokens_seen": 1001331680,
+ "step": 22900,
+ "train_runtime": 177118.7589,
+ "train_tokens_per_second": 5653.448
+ },
+ {
+ "epoch": 1.5422165364934015,
+ "grad_norm": 0.6817693862207875,
+ "learning_rate": 1.2383032202541706e-06,
+ "loss": 0.3010143756866455,
+ "num_input_tokens_seen": 1001546592,
+ "step": 22905,
+ "train_runtime": 177156.0589,
+ "train_tokens_per_second": 5653.471
+ },
+ {
+ "epoch": 1.5425531914893615,
+ "grad_norm": 0.7874018271811691,
+ "learning_rate": 1.2365618890942827e-06,
+ "loss": 0.32308430671691896,
+ "num_input_tokens_seen": 1001766568,
+ "step": 22910,
+ "train_runtime": 177191.6126,
+ "train_tokens_per_second": 5653.578
+ },
+ {
+ "epoch": 1.5428898464853218,
+ "grad_norm": 0.7366938406257898,
+ "learning_rate": 1.2348216103678618e-06,
+ "loss": 0.3221110820770264,
+ "num_input_tokens_seen": 1001986120,
+ "step": 22915,
+ "train_runtime": 177222.995,
+ "train_tokens_per_second": 5653.816
+ },
+ {
+ "epoch": 1.543226501481282,
+ "grad_norm": 0.8883683471081493,
+ "learning_rate": 1.2330823845615713e-06,
+ "loss": 0.2979999780654907,
+ "num_input_tokens_seen": 1002184920,
+ "step": 22920,
+ "train_runtime": 177260.8843,
+ "train_tokens_per_second": 5653.729
+ },
+ {
+ "epoch": 1.543563156477242,
+ "grad_norm": 0.7325915491270013,
+ "learning_rate": 1.2313442121617785e-06,
+ "loss": 0.29711506366729734,
+ "num_input_tokens_seen": 1002416768,
+ "step": 22925,
+ "train_runtime": 177301.116,
+ "train_tokens_per_second": 5653.753
+ },
+ {
+ "epoch": 1.543899811473202,
+ "grad_norm": 0.7697827606669048,
+ "learning_rate": 1.229607093654563e-06,
+ "loss": 0.2836009502410889,
+ "num_input_tokens_seen": 1002644952,
+ "step": 22930,
+ "train_runtime": 177333.8735,
+ "train_tokens_per_second": 5653.996
+ },
+ {
+ "epoch": 1.5442364664691623,
+ "grad_norm": 0.6685394671070748,
+ "learning_rate": 1.2278710295256974e-06,
+ "loss": 0.30063090324401853,
+ "num_input_tokens_seen": 1002875328,
+ "step": 22935,
+ "train_runtime": 177368.0874,
+ "train_tokens_per_second": 5654.204
+ },
+ {
+ "epoch": 1.5445731214651226,
+ "grad_norm": 0.707781339272189,
+ "learning_rate": 1.2261360202606725e-06,
+ "loss": 0.2941617965698242,
+ "num_input_tokens_seen": 1003108632,
+ "step": 22940,
+ "train_runtime": 177406.2713,
+ "train_tokens_per_second": 5654.302
+ },
+ {
+ "epoch": 1.5449097764610826,
+ "grad_norm": 0.7708196807094642,
+ "learning_rate": 1.2244020663446731e-06,
+ "loss": 0.3176748752593994,
+ "num_input_tokens_seen": 1003320896,
+ "step": 22945,
+ "train_runtime": 177449.0984,
+ "train_tokens_per_second": 5654.134
+ },
+ {
+ "epoch": 1.5452464314570427,
+ "grad_norm": 0.9256755732496396,
+ "learning_rate": 1.222669168262598e-06,
+ "loss": 0.29861812591552733,
+ "num_input_tokens_seen": 1003542056,
+ "step": 22950,
+ "train_runtime": 177486.5774,
+ "train_tokens_per_second": 5654.186
+ },
+ {
+ "epoch": 1.545583086453003,
+ "grad_norm": 0.7130093827585298,
+ "learning_rate": 1.2209373264990448e-06,
+ "loss": 0.29040777683258057,
+ "num_input_tokens_seen": 1003762224,
+ "step": 22955,
+ "train_runtime": 177519.7915,
+ "train_tokens_per_second": 5654.368
+ },
+ {
+ "epoch": 1.5459197414489632,
+ "grad_norm": 0.807014861280199,
+ "learning_rate": 1.2192065415383176e-06,
+ "loss": 0.29322617053985595,
+ "num_input_tokens_seen": 1003976720,
+ "step": 22960,
+ "train_runtime": 177559.4075,
+ "train_tokens_per_second": 5654.314
+ },
+ {
+ "epoch": 1.5462563964449232,
+ "grad_norm": 0.7816291079417402,
+ "learning_rate": 1.217476813864425e-06,
+ "loss": 0.3234560966491699,
+ "num_input_tokens_seen": 1004209864,
+ "step": 22965,
+ "train_runtime": 177604.7394,
+ "train_tokens_per_second": 5654.184
+ },
+ {
+ "epoch": 1.5465930514408834,
+ "grad_norm": 0.7404020286277221,
+ "learning_rate": 1.2157481439610801e-06,
+ "loss": 0.27009384632110595,
+ "num_input_tokens_seen": 1004428792,
+ "step": 22970,
+ "train_runtime": 177647.5783,
+ "train_tokens_per_second": 5654.053
+ },
+ {
+ "epoch": 1.5469297064368437,
+ "grad_norm": 0.8293778191938409,
+ "learning_rate": 1.2140205323116976e-06,
+ "loss": 0.2998237609863281,
+ "num_input_tokens_seen": 1004626616,
+ "step": 22975,
+ "train_runtime": 177682.0158,
+ "train_tokens_per_second": 5654.07
+ },
+ {
+ "epoch": 1.5472663614328037,
+ "grad_norm": 0.8562313154245355,
+ "learning_rate": 1.2122939793994043e-06,
+ "loss": 0.2867993116378784,
+ "num_input_tokens_seen": 1004841312,
+ "step": 22980,
+ "train_runtime": 177716.4916,
+ "train_tokens_per_second": 5654.182
+ },
+ {
+ "epoch": 1.5476030164287637,
+ "grad_norm": 0.7604165950877463,
+ "learning_rate": 1.2105684857070182e-06,
+ "loss": 0.292344331741333,
+ "num_input_tokens_seen": 1005062880,
+ "step": 22985,
+ "train_runtime": 177754.2016,
+ "train_tokens_per_second": 5654.229
+ },
+ {
+ "epoch": 1.547939671424724,
+ "grad_norm": 0.7094120408682225,
+ "learning_rate": 1.2088440517170729e-06,
+ "loss": 0.30412988662719725,
+ "num_input_tokens_seen": 1005287360,
+ "step": 22990,
+ "train_runtime": 177790.0242,
+ "train_tokens_per_second": 5654.352
+ },
+ {
+ "epoch": 1.5482763264206842,
+ "grad_norm": 0.8682158123872002,
+ "learning_rate": 1.2071206779117994e-06,
+ "loss": 0.3056650161743164,
+ "num_input_tokens_seen": 1005509112,
+ "step": 22995,
+ "train_runtime": 177827.549,
+ "train_tokens_per_second": 5654.406
+ },
+ {
+ "epoch": 1.5486129814166443,
+ "grad_norm": 0.7363938814599378,
+ "learning_rate": 1.2053983647731337e-06,
+ "loss": 0.29947643280029296,
+ "num_input_tokens_seen": 1005742432,
+ "step": 23000,
+ "train_runtime": 177862.3567,
+ "train_tokens_per_second": 5654.611
+ },
+ {
+ "epoch": 1.5489496364126043,
+ "grad_norm": 0.818386721631332,
+ "learning_rate": 1.2036771127827152e-06,
+ "loss": 0.3036951541900635,
+ "num_input_tokens_seen": 1005948408,
+ "step": 23005,
+ "train_runtime": 177896.9071,
+ "train_tokens_per_second": 5654.671
+ },
+ {
+ "epoch": 1.5492862914085646,
+ "grad_norm": 0.7210690822091025,
+ "learning_rate": 1.2019569224218869e-06,
+ "loss": 0.3014035224914551,
+ "num_input_tokens_seen": 1006184672,
+ "step": 23010,
+ "train_runtime": 177940.4278,
+ "train_tokens_per_second": 5654.615
+ },
+ {
+ "epoch": 1.5496229464045248,
+ "grad_norm": 0.7336526052679622,
+ "learning_rate": 1.2002377941716936e-06,
+ "loss": 0.27741825580596924,
+ "num_input_tokens_seen": 1006397816,
+ "step": 23015,
+ "train_runtime": 177978.0712,
+ "train_tokens_per_second": 5654.617
+ },
+ {
+ "epoch": 1.5499596014004848,
+ "grad_norm": 0.8612322343025608,
+ "learning_rate": 1.1985197285128853e-06,
+ "loss": 0.30047073364257815,
+ "num_input_tokens_seen": 1006606872,
+ "step": 23020,
+ "train_runtime": 178024.3698,
+ "train_tokens_per_second": 5654.321
+ },
+ {
+ "epoch": 1.5502962563964449,
+ "grad_norm": 0.7706187528021623,
+ "learning_rate": 1.1968027259259107e-06,
+ "loss": 0.29756712913513184,
+ "num_input_tokens_seen": 1006810960,
+ "step": 23025,
+ "train_runtime": 178061.8932,
+ "train_tokens_per_second": 5654.275
+ },
+ {
+ "epoch": 1.5506329113924051,
+ "grad_norm": 0.7669195811239113,
+ "learning_rate": 1.1950867868909293e-06,
+ "loss": 0.27481400966644287,
+ "num_input_tokens_seen": 1007021520,
+ "step": 23030,
+ "train_runtime": 178104.3,
+ "train_tokens_per_second": 5654.111
+ },
+ {
+ "epoch": 1.5509695663883654,
+ "grad_norm": 0.8825553327883765,
+ "learning_rate": 1.1933719118877923e-06,
+ "loss": 0.30696120262146,
+ "num_input_tokens_seen": 1007265144,
+ "step": 23035,
+ "train_runtime": 178134.9312,
+ "train_tokens_per_second": 5654.507
+ },
+ {
+ "epoch": 1.5513062213843254,
+ "grad_norm": 0.6530163696183422,
+ "learning_rate": 1.1916581013960632e-06,
+ "loss": 0.2742566347122192,
+ "num_input_tokens_seen": 1007502816,
+ "step": 23040,
+ "train_runtime": 178176.3835,
+ "train_tokens_per_second": 5654.525
+ },
+ {
+ "epoch": 1.5516428763802854,
+ "grad_norm": 0.8012588977520706,
+ "learning_rate": 1.189945355895002e-06,
+ "loss": 0.31151793003082273,
+ "num_input_tokens_seen": 1007718880,
+ "step": 23045,
+ "train_runtime": 178217.168,
+ "train_tokens_per_second": 5654.443
+ },
+ {
+ "epoch": 1.5519795313762457,
+ "grad_norm": 0.7331978765364463,
+ "learning_rate": 1.1882336758635727e-06,
+ "loss": 0.27026982307434083,
+ "num_input_tokens_seen": 1007951160,
+ "step": 23050,
+ "train_runtime": 178252.7434,
+ "train_tokens_per_second": 5654.618
+ },
+ {
+ "epoch": 1.552316186372206,
+ "grad_norm": 0.8000089411786985,
+ "learning_rate": 1.1865230617804412e-06,
+ "loss": 0.32370221614837646,
+ "num_input_tokens_seen": 1008186432,
+ "step": 23055,
+ "train_runtime": 178288.4347,
+ "train_tokens_per_second": 5654.806
+ },
+ {
+ "epoch": 1.552652841368166,
+ "grad_norm": 0.7285802544354917,
+ "learning_rate": 1.1848135141239753e-06,
+ "loss": 0.2949348211288452,
+ "num_input_tokens_seen": 1008403400,
+ "step": 23060,
+ "train_runtime": 178329.6018,
+ "train_tokens_per_second": 5654.717
+ },
+ {
+ "epoch": 1.552989496364126,
+ "grad_norm": 0.8133624441552181,
+ "learning_rate": 1.1831050333722438e-06,
+ "loss": 0.3110039710998535,
+ "num_input_tokens_seen": 1008619128,
+ "step": 23065,
+ "train_runtime": 178364.4191,
+ "train_tokens_per_second": 5654.822
+ },
+ {
+ "epoch": 1.5533261513600862,
+ "grad_norm": 0.7041630401525898,
+ "learning_rate": 1.1813976200030191e-06,
+ "loss": 0.29149832725524905,
+ "num_input_tokens_seen": 1008834768,
+ "step": 23070,
+ "train_runtime": 178400.3205,
+ "train_tokens_per_second": 5654.893
+ },
+ {
+ "epoch": 1.5536628063560465,
+ "grad_norm": 0.7088406276262311,
+ "learning_rate": 1.1796912744937717e-06,
+ "loss": 0.31109683513641356,
+ "num_input_tokens_seen": 1009058816,
+ "step": 23075,
+ "train_runtime": 178437.2135,
+ "train_tokens_per_second": 5654.98
+ },
+ {
+ "epoch": 1.5539994613520065,
+ "grad_norm": 0.8006442613984758,
+ "learning_rate": 1.1779859973216802e-06,
+ "loss": 0.2948103904724121,
+ "num_input_tokens_seen": 1009270824,
+ "step": 23080,
+ "train_runtime": 178479.8877,
+ "train_tokens_per_second": 5654.815
+ },
+ {
+ "epoch": 1.5543361163479665,
+ "grad_norm": 0.8721550220509111,
+ "learning_rate": 1.1762817889636142e-06,
+ "loss": 0.29932336807250975,
+ "num_input_tokens_seen": 1009470192,
+ "step": 23085,
+ "train_runtime": 178518.382,
+ "train_tokens_per_second": 5654.713
+ },
+ {
+ "epoch": 1.5546727713439268,
+ "grad_norm": 0.7952095441240394,
+ "learning_rate": 1.1745786498961542e-06,
+ "loss": 0.31789186000823977,
+ "num_input_tokens_seen": 1009708792,
+ "step": 23090,
+ "train_runtime": 178557.7116,
+ "train_tokens_per_second": 5654.804
+ },
+ {
+ "epoch": 1.555009426339887,
+ "grad_norm": 0.7311486152172996,
+ "learning_rate": 1.1728765805955766e-06,
+ "loss": 0.2769877195358276,
+ "num_input_tokens_seen": 1009922120,
+ "step": 23095,
+ "train_runtime": 178592.2137,
+ "train_tokens_per_second": 5654.906
+ },
+ {
+ "epoch": 1.555346081335847,
+ "grad_norm": 0.8415310169008027,
+ "learning_rate": 1.1711755815378595e-06,
+ "loss": 0.32232234477996824,
+ "num_input_tokens_seen": 1010130920,
+ "step": 23100,
+ "train_runtime": 178626.0283,
+ "train_tokens_per_second": 5655.004
+ },
+ {
+ "epoch": 1.555682736331807,
+ "grad_norm": 0.6963455165526707,
+ "learning_rate": 1.1694756531986818e-06,
+ "loss": 0.293909478187561,
+ "num_input_tokens_seen": 1010358016,
+ "step": 23105,
+ "train_runtime": 178670.1633,
+ "train_tokens_per_second": 5654.878
+ },
+ {
+ "epoch": 1.5560193913277673,
+ "grad_norm": 0.7864173646107584,
+ "learning_rate": 1.1677767960534232e-06,
+ "loss": 0.3079185962677002,
+ "num_input_tokens_seen": 1010578448,
+ "step": 23110,
+ "train_runtime": 178714.6961,
+ "train_tokens_per_second": 5654.703
+ },
+ {
+ "epoch": 1.5563560463237276,
+ "grad_norm": 0.7334560465799055,
+ "learning_rate": 1.166079010577163e-06,
+ "loss": 0.31560115814208983,
+ "num_input_tokens_seen": 1010787032,
+ "step": 23115,
+ "train_runtime": 178752.3631,
+ "train_tokens_per_second": 5654.678
+ },
+ {
+ "epoch": 1.5566927013196876,
+ "grad_norm": 0.8841111887561086,
+ "learning_rate": 1.1643822972446823e-06,
+ "loss": 0.3215439796447754,
+ "num_input_tokens_seen": 1011011240,
+ "step": 23120,
+ "train_runtime": 178786.2102,
+ "train_tokens_per_second": 5654.861
+ },
+ {
+ "epoch": 1.5570293563156477,
+ "grad_norm": 0.7142397763016451,
+ "learning_rate": 1.1626866565304594e-06,
+ "loss": 0.2977125883102417,
+ "num_input_tokens_seen": 1011244280,
+ "step": 23125,
+ "train_runtime": 178828.8659,
+ "train_tokens_per_second": 5654.816
+ },
+ {
+ "epoch": 1.557366011311608,
+ "grad_norm": 0.7277792379867785,
+ "learning_rate": 1.16099208890868e-06,
+ "loss": 0.2924909830093384,
+ "num_input_tokens_seen": 1011459816,
+ "step": 23130,
+ "train_runtime": 178867.1737,
+ "train_tokens_per_second": 5654.81
+ },
+ {
+ "epoch": 1.5577026663075682,
+ "grad_norm": 0.7072278904651554,
+ "learning_rate": 1.1592985948532187e-06,
+ "loss": 0.29182374477386475,
+ "num_input_tokens_seen": 1011668760,
+ "step": 23135,
+ "train_runtime": 178901.1993,
+ "train_tokens_per_second": 5654.902
+ },
+ {
+ "epoch": 1.5580393213035282,
+ "grad_norm": 0.7122762244099401,
+ "learning_rate": 1.1576061748376594e-06,
+ "loss": 0.28834197521209715,
+ "num_input_tokens_seen": 1011885552,
+ "step": 23140,
+ "train_runtime": 178941.129,
+ "train_tokens_per_second": 5654.852
+ },
+ {
+ "epoch": 1.5583759762994882,
+ "grad_norm": 0.7581459438101124,
+ "learning_rate": 1.1559148293352805e-06,
+ "loss": 0.293502950668335,
+ "num_input_tokens_seen": 1012108496,
+ "step": 23145,
+ "train_runtime": 178981.2231,
+ "train_tokens_per_second": 5654.831
+ },
+ {
+ "epoch": 1.5587126312954485,
+ "grad_norm": 0.6988252432908294,
+ "learning_rate": 1.154224558819062e-06,
+ "loss": 0.2947830677032471,
+ "num_input_tokens_seen": 1012331792,
+ "step": 23150,
+ "train_runtime": 179021.8936,
+ "train_tokens_per_second": 5654.793
+ },
+ {
+ "epoch": 1.5590492862914087,
+ "grad_norm": 0.7771507807990884,
+ "learning_rate": 1.1525353637616821e-06,
+ "loss": 0.29758148193359374,
+ "num_input_tokens_seen": 1012552976,
+ "step": 23155,
+ "train_runtime": 179058.4207,
+ "train_tokens_per_second": 5654.875
+ },
+ {
+ "epoch": 1.5593859412873687,
+ "grad_norm": 0.7491488781107557,
+ "learning_rate": 1.1508472446355196e-06,
+ "loss": 0.3053610324859619,
+ "num_input_tokens_seen": 1012785584,
+ "step": 23160,
+ "train_runtime": 179099.292,
+ "train_tokens_per_second": 5654.883
+ },
+ {
+ "epoch": 1.5597225962833288,
+ "grad_norm": 0.697759562324447,
+ "learning_rate": 1.1491602019126501e-06,
+ "loss": 0.2998149871826172,
+ "num_input_tokens_seen": 1013018136,
+ "step": 23165,
+ "train_runtime": 179129.3593,
+ "train_tokens_per_second": 5655.232
+ },
+ {
+ "epoch": 1.560059251279289,
+ "grad_norm": 0.7568078853356832,
+ "learning_rate": 1.1474742360648515e-06,
+ "loss": 0.31585729122161865,
+ "num_input_tokens_seen": 1013244520,
+ "step": 23170,
+ "train_runtime": 179174.5036,
+ "train_tokens_per_second": 5655.071
+ },
+ {
+ "epoch": 1.5603959062752493,
+ "grad_norm": 0.8520933038767174,
+ "learning_rate": 1.1457893475635968e-06,
+ "loss": 0.3203889846801758,
+ "num_input_tokens_seen": 1013457472,
+ "step": 23175,
+ "train_runtime": 179219.5043,
+ "train_tokens_per_second": 5654.839
+ },
+ {
+ "epoch": 1.5607325612712093,
+ "grad_norm": 0.6916094191053404,
+ "learning_rate": 1.144105536880063e-06,
+ "loss": 0.2820396900177002,
+ "num_input_tokens_seen": 1013672336,
+ "step": 23180,
+ "train_runtime": 179252.287,
+ "train_tokens_per_second": 5655.004
+ },
+ {
+ "epoch": 1.5610692162671693,
+ "grad_norm": 0.7609947532948533,
+ "learning_rate": 1.1424228044851176e-06,
+ "loss": 0.29037587642669677,
+ "num_input_tokens_seen": 1013859016,
+ "step": 23185,
+ "train_runtime": 179288.0236,
+ "train_tokens_per_second": 5654.918
+ },
+ {
+ "epoch": 1.5614058712631296,
+ "grad_norm": 0.6991374139098863,
+ "learning_rate": 1.1407411508493355e-06,
+ "loss": 0.3020007133483887,
+ "num_input_tokens_seen": 1014067352,
+ "step": 23190,
+ "train_runtime": 179325.7985,
+ "train_tokens_per_second": 5654.888
+ },
+ {
+ "epoch": 1.5617425262590898,
+ "grad_norm": 0.7525302404327049,
+ "learning_rate": 1.1390605764429846e-06,
+ "loss": 0.2727111101150513,
+ "num_input_tokens_seen": 1014277952,
+ "step": 23195,
+ "train_runtime": 179365.198,
+ "train_tokens_per_second": 5654.82
+ },
+ {
+ "epoch": 1.5620791812550499,
+ "grad_norm": 0.755912903107464,
+ "learning_rate": 1.1373810817360314e-06,
+ "loss": 0.29233598709106445,
+ "num_input_tokens_seen": 1014509144,
+ "step": 23200,
+ "train_runtime": 179403.2626,
+ "train_tokens_per_second": 5654.909
+ },
+ {
+ "epoch": 1.5624158362510099,
+ "grad_norm": 0.8490364171656128,
+ "learning_rate": 1.135702667198142e-06,
+ "loss": 0.3186174392700195,
+ "num_input_tokens_seen": 1014733760,
+ "step": 23205,
+ "train_runtime": 179436.0347,
+ "train_tokens_per_second": 5655.128
+ },
+ {
+ "epoch": 1.5627524912469701,
+ "grad_norm": 0.7833432896868294,
+ "learning_rate": 1.134025333298679e-06,
+ "loss": 0.2658507823944092,
+ "num_input_tokens_seen": 1014933448,
+ "step": 23210,
+ "train_runtime": 179473.8421,
+ "train_tokens_per_second": 5655.049
+ },
+ {
+ "epoch": 1.5630891462429304,
+ "grad_norm": 0.6767171989222308,
+ "learning_rate": 1.1323490805067045e-06,
+ "loss": 0.27293944358825684,
+ "num_input_tokens_seen": 1015151416,
+ "step": 23215,
+ "train_runtime": 179514.1502,
+ "train_tokens_per_second": 5654.994
+ },
+ {
+ "epoch": 1.5634258012388904,
+ "grad_norm": 0.7564520547243825,
+ "learning_rate": 1.1306739092909757e-06,
+ "loss": 0.27212421894073485,
+ "num_input_tokens_seen": 1015357104,
+ "step": 23220,
+ "train_runtime": 179558.059,
+ "train_tokens_per_second": 5654.757
+ },
+ {
+ "epoch": 1.5637624562348504,
+ "grad_norm": 0.7556836048466959,
+ "learning_rate": 1.1289998201199493e-06,
+ "loss": 0.28475046157836914,
+ "num_input_tokens_seen": 1015578216,
+ "step": 23225,
+ "train_runtime": 179598.1851,
+ "train_tokens_per_second": 5654.724
+ },
+ {
+ "epoch": 1.5640991112308107,
+ "grad_norm": 0.7771413672780467,
+ "learning_rate": 1.1273268134617816e-06,
+ "loss": 0.3049982786178589,
+ "num_input_tokens_seen": 1015803864,
+ "step": 23230,
+ "train_runtime": 179637.2745,
+ "train_tokens_per_second": 5654.75
+ },
+ {
+ "epoch": 1.564435766226771,
+ "grad_norm": 0.6387132976491532,
+ "learning_rate": 1.1256548897843189e-06,
+ "loss": 0.2998204708099365,
+ "num_input_tokens_seen": 1016035144,
+ "step": 23235,
+ "train_runtime": 179676.6807,
+ "train_tokens_per_second": 5654.797
+ },
+ {
+ "epoch": 1.564772421222731,
+ "grad_norm": 0.677517945115812,
+ "learning_rate": 1.1239840495551136e-06,
+ "loss": 0.2948010921478271,
+ "num_input_tokens_seen": 1016256464,
+ "step": 23240,
+ "train_runtime": 179716.2122,
+ "train_tokens_per_second": 5654.785
+ },
+ {
+ "epoch": 1.565109076218691,
+ "grad_norm": 0.8189763586455887,
+ "learning_rate": 1.1223142932414087e-06,
+ "loss": 0.3241511583328247,
+ "num_input_tokens_seen": 1016483368,
+ "step": 23245,
+ "train_runtime": 179758.4552,
+ "train_tokens_per_second": 5654.718
+ },
+ {
+ "epoch": 1.5654457312146512,
+ "grad_norm": 0.6822186571021763,
+ "learning_rate": 1.120645621310147e-06,
+ "loss": 0.2887975454330444,
+ "num_input_tokens_seen": 1016700976,
+ "step": 23250,
+ "train_runtime": 179800.864,
+ "train_tokens_per_second": 5654.594
+ },
+ {
+ "epoch": 1.5657823862106115,
+ "grad_norm": 0.8030985846370237,
+ "learning_rate": 1.1189780342279666e-06,
+ "loss": 0.2927762508392334,
+ "num_input_tokens_seen": 1016907680,
+ "step": 23255,
+ "train_runtime": 179835.7597,
+ "train_tokens_per_second": 5654.647
+ },
+ {
+ "epoch": 1.5661190412065715,
+ "grad_norm": 0.7041599888250881,
+ "learning_rate": 1.1173115324612033e-06,
+ "loss": 0.30506865978240966,
+ "num_input_tokens_seen": 1017110400,
+ "step": 23260,
+ "train_runtime": 179876.7454,
+ "train_tokens_per_second": 5654.485
+ },
+ {
+ "epoch": 1.5664556962025316,
+ "grad_norm": 0.8207082370707505,
+ "learning_rate": 1.1156461164758891e-06,
+ "loss": 0.3330164909362793,
+ "num_input_tokens_seen": 1017339120,
+ "step": 23265,
+ "train_runtime": 179913.8123,
+ "train_tokens_per_second": 5654.592
+ },
+ {
+ "epoch": 1.5667923511984918,
+ "grad_norm": 0.6985440775195174,
+ "learning_rate": 1.1139817867377522e-06,
+ "loss": 0.2630832433700562,
+ "num_input_tokens_seen": 1017568520,
+ "step": 23270,
+ "train_runtime": 179951.8501,
+ "train_tokens_per_second": 5654.671
+ },
+ {
+ "epoch": 1.567129006194452,
+ "grad_norm": 0.7538202300333461,
+ "learning_rate": 1.1123185437122153e-06,
+ "loss": 0.29378499984741213,
+ "num_input_tokens_seen": 1017797256,
+ "step": 23275,
+ "train_runtime": 179995.1502,
+ "train_tokens_per_second": 5654.582
+ },
+ {
+ "epoch": 1.567465661190412,
+ "grad_norm": 0.8033728394499458,
+ "learning_rate": 1.1106563878644038e-06,
+ "loss": 0.30549144744873047,
+ "num_input_tokens_seen": 1018024880,
+ "step": 23280,
+ "train_runtime": 180035.9497,
+ "train_tokens_per_second": 5654.564
+ },
+ {
+ "epoch": 1.5678023161863721,
+ "grad_norm": 0.7853141000994098,
+ "learning_rate": 1.1089953196591286e-06,
+ "loss": 0.30190200805664064,
+ "num_input_tokens_seen": 1018241712,
+ "step": 23285,
+ "train_runtime": 180077.9944,
+ "train_tokens_per_second": 5654.448
+ },
+ {
+ "epoch": 1.5681389711823324,
+ "grad_norm": 0.7859724149410223,
+ "learning_rate": 1.107335339560906e-06,
+ "loss": 0.3117091417312622,
+ "num_input_tokens_seen": 1018448032,
+ "step": 23290,
+ "train_runtime": 180115.2413,
+ "train_tokens_per_second": 5654.424
+ },
+ {
+ "epoch": 1.5684756261782926,
+ "grad_norm": 0.7863228372315338,
+ "learning_rate": 1.1056764480339427e-06,
+ "loss": 0.3014981269836426,
+ "num_input_tokens_seen": 1018674080,
+ "step": 23295,
+ "train_runtime": 180158.4273,
+ "train_tokens_per_second": 5654.324
+ },
+ {
+ "epoch": 1.5688122811742526,
+ "grad_norm": 0.7789425478674455,
+ "learning_rate": 1.1040186455421425e-06,
+ "loss": 0.33310580253601074,
+ "num_input_tokens_seen": 1018908720,
+ "step": 23300,
+ "train_runtime": 180194.8091,
+ "train_tokens_per_second": 5654.484
+ },
+ {
+ "epoch": 1.5691489361702127,
+ "grad_norm": 0.8291198693575462,
+ "learning_rate": 1.1023619325491052e-06,
+ "loss": 0.310105037689209,
+ "num_input_tokens_seen": 1019122152,
+ "step": 23305,
+ "train_runtime": 180234.0997,
+ "train_tokens_per_second": 5654.436
+ },
+ {
+ "epoch": 1.569485591166173,
+ "grad_norm": 0.6253815970834292,
+ "learning_rate": 1.1007063095181248e-06,
+ "loss": 0.27553586959838866,
+ "num_input_tokens_seen": 1019342832,
+ "step": 23310,
+ "train_runtime": 180278.8765,
+ "train_tokens_per_second": 5654.256
+ },
+ {
+ "epoch": 1.5698222461621332,
+ "grad_norm": 0.8786406413208016,
+ "learning_rate": 1.0990517769121905e-06,
+ "loss": 0.29708888530731203,
+ "num_input_tokens_seen": 1019546240,
+ "step": 23315,
+ "train_runtime": 180314.3153,
+ "train_tokens_per_second": 5654.272
+ },
+ {
+ "epoch": 1.5701589011580932,
+ "grad_norm": 0.8222072276717016,
+ "learning_rate": 1.0973983351939876e-06,
+ "loss": 0.3073767900466919,
+ "num_input_tokens_seen": 1019767384,
+ "step": 23320,
+ "train_runtime": 180352.2362,
+ "train_tokens_per_second": 5654.31
+ },
+ {
+ "epoch": 1.5704955561540532,
+ "grad_norm": 0.7054404712911847,
+ "learning_rate": 1.095745984825895e-06,
+ "loss": 0.2895050048828125,
+ "num_input_tokens_seen": 1019996880,
+ "step": 23325,
+ "train_runtime": 180386.2267,
+ "train_tokens_per_second": 5654.516
+ },
+ {
+ "epoch": 1.5708322111500135,
+ "grad_norm": 0.8092139157721354,
+ "learning_rate": 1.0940947262699902e-06,
+ "loss": 0.30101957321166994,
+ "num_input_tokens_seen": 1020200320,
+ "step": 23330,
+ "train_runtime": 180428.7048,
+ "train_tokens_per_second": 5654.313
+ },
+ {
+ "epoch": 1.5711688661459737,
+ "grad_norm": 0.7111960834790324,
+ "learning_rate": 1.092444559988038e-06,
+ "loss": 0.28864340782165526,
+ "num_input_tokens_seen": 1020417768,
+ "step": 23335,
+ "train_runtime": 180466.2384,
+ "train_tokens_per_second": 5654.342
+ },
+ {
+ "epoch": 1.5715055211419338,
+ "grad_norm": 0.660638688067613,
+ "learning_rate": 1.0907954864415048e-06,
+ "loss": 0.29140138626098633,
+ "num_input_tokens_seen": 1020645240,
+ "step": 23340,
+ "train_runtime": 180507.8532,
+ "train_tokens_per_second": 5654.298
+ },
+ {
+ "epoch": 1.5718421761378938,
+ "grad_norm": 0.8672767958383719,
+ "learning_rate": 1.089147506091549e-06,
+ "loss": 0.28663978576660154,
+ "num_input_tokens_seen": 1020864992,
+ "step": 23345,
+ "train_runtime": 180542.9652,
+ "train_tokens_per_second": 5654.416
+ },
+ {
+ "epoch": 1.572178831133854,
+ "grad_norm": 0.8719743776750728,
+ "learning_rate": 1.0875006193990217e-06,
+ "loss": 0.3154577732086182,
+ "num_input_tokens_seen": 1021076760,
+ "step": 23350,
+ "train_runtime": 180584.1055,
+ "train_tokens_per_second": 5654.3
+ },
+ {
+ "epoch": 1.5725154861298143,
+ "grad_norm": 0.7880930108374784,
+ "learning_rate": 1.0858548268244706e-06,
+ "loss": 0.2915508270263672,
+ "num_input_tokens_seen": 1021295960,
+ "step": 23355,
+ "train_runtime": 180623.241,
+ "train_tokens_per_second": 5654.289
+ },
+ {
+ "epoch": 1.5728521411257743,
+ "grad_norm": 0.8007349258213046,
+ "learning_rate": 1.0842101288281359e-06,
+ "loss": 0.293584418296814,
+ "num_input_tokens_seen": 1021516488,
+ "step": 23360,
+ "train_runtime": 180662.7604,
+ "train_tokens_per_second": 5654.273
+ },
+ {
+ "epoch": 1.5731887961217343,
+ "grad_norm": 0.7656889523789541,
+ "learning_rate": 1.082566525869952e-06,
+ "loss": 0.28592419624328613,
+ "num_input_tokens_seen": 1021745864,
+ "step": 23365,
+ "train_runtime": 180702.808,
+ "train_tokens_per_second": 5654.289
+ },
+ {
+ "epoch": 1.5735254511176946,
+ "grad_norm": 0.8722272978728876,
+ "learning_rate": 1.0809240184095476e-06,
+ "loss": 0.3062607288360596,
+ "num_input_tokens_seen": 1021959216,
+ "step": 23370,
+ "train_runtime": 180737.0204,
+ "train_tokens_per_second": 5654.399
+ },
+ {
+ "epoch": 1.5738621061136548,
+ "grad_norm": 0.6745653891846737,
+ "learning_rate": 1.0792826069062429e-06,
+ "loss": 0.26804351806640625,
+ "num_input_tokens_seen": 1022178816,
+ "step": 23375,
+ "train_runtime": 180775.1712,
+ "train_tokens_per_second": 5654.42
+ },
+ {
+ "epoch": 1.5741987611096149,
+ "grad_norm": 0.6985205733037011,
+ "learning_rate": 1.0776422918190576e-06,
+ "loss": 0.31443793773651124,
+ "num_input_tokens_seen": 1022402224,
+ "step": 23380,
+ "train_runtime": 180806.1938,
+ "train_tokens_per_second": 5654.686
+ },
+ {
+ "epoch": 1.574535416105575,
+ "grad_norm": 0.8000413255647302,
+ "learning_rate": 1.0760030736066952e-06,
+ "loss": 0.28395633697509765,
+ "num_input_tokens_seen": 1022607104,
+ "step": 23385,
+ "train_runtime": 180846.4339,
+ "train_tokens_per_second": 5654.561
+ },
+ {
+ "epoch": 1.5748720711015352,
+ "grad_norm": 0.7301802635358972,
+ "learning_rate": 1.0743649527275619e-06,
+ "loss": 0.3159285068511963,
+ "num_input_tokens_seen": 1022834416,
+ "step": 23390,
+ "train_runtime": 180888.2807,
+ "train_tokens_per_second": 5654.509
+ },
+ {
+ "epoch": 1.5752087260974954,
+ "grad_norm": 0.8194485612881878,
+ "learning_rate": 1.0727279296397514e-06,
+ "loss": 0.3103215456008911,
+ "num_input_tokens_seen": 1023064384,
+ "step": 23395,
+ "train_runtime": 180932.3993,
+ "train_tokens_per_second": 5654.401
+ },
+ {
+ "epoch": 1.5755453810934554,
+ "grad_norm": 0.8975237265847585,
+ "learning_rate": 1.0710920048010526e-06,
+ "loss": 0.3091814279556274,
+ "num_input_tokens_seen": 1023286312,
+ "step": 23400,
+ "train_runtime": 180973.834,
+ "train_tokens_per_second": 5654.333
+ },
+ {
+ "epoch": 1.5758820360894155,
+ "grad_norm": 0.7551422109604019,
+ "learning_rate": 1.0694571786689462e-06,
+ "loss": 0.29776315689086913,
+ "num_input_tokens_seen": 1023528528,
+ "step": 23405,
+ "train_runtime": 181012.8344,
+ "train_tokens_per_second": 5654.453
+ },
+ {
+ "epoch": 1.5762186910853757,
+ "grad_norm": 0.7923202761413153,
+ "learning_rate": 1.067823451700606e-06,
+ "loss": 0.29261269569396975,
+ "num_input_tokens_seen": 1023748656,
+ "step": 23410,
+ "train_runtime": 181047.5433,
+ "train_tokens_per_second": 5654.585
+ },
+ {
+ "epoch": 1.576555346081336,
+ "grad_norm": 0.7616310783411705,
+ "learning_rate": 1.066190824352899e-06,
+ "loss": 0.29994802474975585,
+ "num_input_tokens_seen": 1023974480,
+ "step": 23415,
+ "train_runtime": 181087.6631,
+ "train_tokens_per_second": 5654.579
+ },
+ {
+ "epoch": 1.576892001077296,
+ "grad_norm": 0.6523336414139479,
+ "learning_rate": 1.0645592970823837e-06,
+ "loss": 0.27131869792938235,
+ "num_input_tokens_seen": 1024177832,
+ "step": 23420,
+ "train_runtime": 181123.7413,
+ "train_tokens_per_second": 5654.575
+ },
+ {
+ "epoch": 1.577228656073256,
+ "grad_norm": 0.6772527419522956,
+ "learning_rate": 1.0629288703453105e-06,
+ "loss": 0.3091407299041748,
+ "num_input_tokens_seen": 1024403608,
+ "step": 23425,
+ "train_runtime": 181163.8389,
+ "train_tokens_per_second": 5654.57
+ },
+ {
+ "epoch": 1.5775653110692163,
+ "grad_norm": 0.8870052312783957,
+ "learning_rate": 1.061299544597627e-06,
+ "loss": 0.29852406978607177,
+ "num_input_tokens_seen": 1024608360,
+ "step": 23430,
+ "train_runtime": 181201.9913,
+ "train_tokens_per_second": 5654.509
+ },
+ {
+ "epoch": 1.5779019660651765,
+ "grad_norm": 0.7385143164519493,
+ "learning_rate": 1.0596713202949637e-06,
+ "loss": 0.3082388162612915,
+ "num_input_tokens_seen": 1024836472,
+ "step": 23435,
+ "train_runtime": 181244.3758,
+ "train_tokens_per_second": 5654.446
+ },
+ {
+ "epoch": 1.5782386210611365,
+ "grad_norm": 0.7117628849410049,
+ "learning_rate": 1.0580441978926527e-06,
+ "loss": 0.3000252485275269,
+ "num_input_tokens_seen": 1025047408,
+ "step": 23440,
+ "train_runtime": 181284.1102,
+ "train_tokens_per_second": 5654.37
+ },
+ {
+ "epoch": 1.5785752760570966,
+ "grad_norm": 0.8021413628123183,
+ "learning_rate": 1.056418177845711e-06,
+ "loss": 0.2908348083496094,
+ "num_input_tokens_seen": 1025252528,
+ "step": 23445,
+ "train_runtime": 181326.465,
+ "train_tokens_per_second": 5654.18
+ },
+ {
+ "epoch": 1.5789119310530568,
+ "grad_norm": 0.7083241483136948,
+ "learning_rate": 1.0547932606088506e-06,
+ "loss": 0.3012411594390869,
+ "num_input_tokens_seen": 1025462120,
+ "step": 23450,
+ "train_runtime": 181365.3534,
+ "train_tokens_per_second": 5654.124
+ },
+ {
+ "epoch": 1.579248586049017,
+ "grad_norm": 0.749765463777595,
+ "learning_rate": 1.053169446636475e-06,
+ "loss": 0.29930336475372316,
+ "num_input_tokens_seen": 1025676472,
+ "step": 23455,
+ "train_runtime": 181402.7027,
+ "train_tokens_per_second": 5654.141
+ },
+ {
+ "epoch": 1.579585241044977,
+ "grad_norm": 0.870290848746088,
+ "learning_rate": 1.0515467363826782e-06,
+ "loss": 0.3104248046875,
+ "num_input_tokens_seen": 1025904840,
+ "step": 23460,
+ "train_runtime": 181436.5696,
+ "train_tokens_per_second": 5654.344
+ },
+ {
+ "epoch": 1.5799218960409371,
+ "grad_norm": 0.7377471679217029,
+ "learning_rate": 1.0499251303012454e-06,
+ "loss": 0.2840702533721924,
+ "num_input_tokens_seen": 1026129016,
+ "step": 23465,
+ "train_runtime": 181472.8949,
+ "train_tokens_per_second": 5654.448
+ },
+ {
+ "epoch": 1.5802585510368974,
+ "grad_norm": 0.7923429690046462,
+ "learning_rate": 1.0483046288456544e-06,
+ "loss": 0.30331013202667234,
+ "num_input_tokens_seen": 1026336600,
+ "step": 23470,
+ "train_runtime": 181508.7471,
+ "train_tokens_per_second": 5654.475
+ },
+ {
+ "epoch": 1.5805952060328576,
+ "grad_norm": 0.7779123855214451,
+ "learning_rate": 1.0466852324690719e-06,
+ "loss": 0.30037317276000974,
+ "num_input_tokens_seen": 1026549792,
+ "step": 23475,
+ "train_runtime": 181547.6766,
+ "train_tokens_per_second": 5654.436
+ },
+ {
+ "epoch": 1.5809318610288177,
+ "grad_norm": 0.812157038252082,
+ "learning_rate": 1.0450669416243596e-06,
+ "loss": 0.2942227363586426,
+ "num_input_tokens_seen": 1026774584,
+ "step": 23480,
+ "train_runtime": 181585.7621,
+ "train_tokens_per_second": 5654.488
+ },
+ {
+ "epoch": 1.5812685160247777,
+ "grad_norm": 0.7370282264500914,
+ "learning_rate": 1.0434497567640639e-06,
+ "loss": 0.3083514213562012,
+ "num_input_tokens_seen": 1026996640,
+ "step": 23485,
+ "train_runtime": 181622.5429,
+ "train_tokens_per_second": 5654.566
+ },
+ {
+ "epoch": 1.581605171020738,
+ "grad_norm": 0.8340284951059088,
+ "learning_rate": 1.0418336783404282e-06,
+ "loss": 0.3067812204360962,
+ "num_input_tokens_seen": 1027217736,
+ "step": 23490,
+ "train_runtime": 181660.7149,
+ "train_tokens_per_second": 5654.595
+ },
+ {
+ "epoch": 1.5819418260166982,
+ "grad_norm": 0.694715006207709,
+ "learning_rate": 1.0402187068053825e-06,
+ "loss": 0.3050783395767212,
+ "num_input_tokens_seen": 1027433032,
+ "step": 23495,
+ "train_runtime": 181694.8222,
+ "train_tokens_per_second": 5654.718
+ },
+ {
+ "epoch": 1.5822784810126582,
+ "grad_norm": 0.6925827695690472,
+ "learning_rate": 1.0386048426105495e-06,
+ "loss": 0.28964962959289553,
+ "num_input_tokens_seen": 1027649992,
+ "step": 23500,
+ "train_runtime": 181732.3475,
+ "train_tokens_per_second": 5654.744
+ },
+ {
+ "epoch": 1.5826151360086183,
+ "grad_norm": 0.7154905995519729,
+ "learning_rate": 1.0369920862072396e-06,
+ "loss": 0.28251526355743406,
+ "num_input_tokens_seen": 1027858824,
+ "step": 23505,
+ "train_runtime": 181776.3615,
+ "train_tokens_per_second": 5654.524
+ },
+ {
+ "epoch": 1.5829517910045785,
+ "grad_norm": 0.7732969474103675,
+ "learning_rate": 1.0353804380464556e-06,
+ "loss": 0.28255496025085447,
+ "num_input_tokens_seen": 1028079384,
+ "step": 23510,
+ "train_runtime": 181818.5723,
+ "train_tokens_per_second": 5654.424
+ },
+ {
+ "epoch": 1.5832884460005388,
+ "grad_norm": 0.7444357762125697,
+ "learning_rate": 1.0337698985788903e-06,
+ "loss": 0.2897408246994019,
+ "num_input_tokens_seen": 1028306808,
+ "step": 23515,
+ "train_runtime": 181859.8638,
+ "train_tokens_per_second": 5654.391
+ },
+ {
+ "epoch": 1.5836251009964988,
+ "grad_norm": 0.7396832464568777,
+ "learning_rate": 1.0321604682549247e-06,
+ "loss": 0.27561049461364745,
+ "num_input_tokens_seen": 1028540048,
+ "step": 23520,
+ "train_runtime": 181903.878,
+ "train_tokens_per_second": 5654.305
+ },
+ {
+ "epoch": 1.5839617559924588,
+ "grad_norm": 0.7843713186397172,
+ "learning_rate": 1.0305521475246311e-06,
+ "loss": 0.3138261318206787,
+ "num_input_tokens_seen": 1028749872,
+ "step": 23525,
+ "train_runtime": 181940.9808,
+ "train_tokens_per_second": 5654.305
+ },
+ {
+ "epoch": 1.584298410988419,
+ "grad_norm": 0.811370185654451,
+ "learning_rate": 1.028944936837774e-06,
+ "loss": 0.31621651649475097,
+ "num_input_tokens_seen": 1028951280,
+ "step": 23530,
+ "train_runtime": 181974.7241,
+ "train_tokens_per_second": 5654.364
+ },
+ {
+ "epoch": 1.5846350659843793,
+ "grad_norm": 0.7315538832578915,
+ "learning_rate": 1.0273388366438003e-06,
+ "loss": 0.30298333168029784,
+ "num_input_tokens_seen": 1029175832,
+ "step": 23535,
+ "train_runtime": 182014.8854,
+ "train_tokens_per_second": 5654.35
+ },
+ {
+ "epoch": 1.5849717209803393,
+ "grad_norm": 0.6822526332510949,
+ "learning_rate": 1.0257338473918538e-06,
+ "loss": 0.2973572015762329,
+ "num_input_tokens_seen": 1029399728,
+ "step": 23540,
+ "train_runtime": 182052.6379,
+ "train_tokens_per_second": 5654.407
+ },
+ {
+ "epoch": 1.5853083759762994,
+ "grad_norm": 0.7092783516071793,
+ "learning_rate": 1.0241299695307644e-06,
+ "loss": 0.3015413761138916,
+ "num_input_tokens_seen": 1029616800,
+ "step": 23545,
+ "train_runtime": 182093.0118,
+ "train_tokens_per_second": 5654.345
+ },
+ {
+ "epoch": 1.5856450309722596,
+ "grad_norm": 0.6440724656411346,
+ "learning_rate": 1.0225272035090506e-06,
+ "loss": 0.282063889503479,
+ "num_input_tokens_seen": 1029848952,
+ "step": 23550,
+ "train_runtime": 182128.0658,
+ "train_tokens_per_second": 5654.532
+ },
+ {
+ "epoch": 1.5859816859682199,
+ "grad_norm": 0.7978640593248636,
+ "learning_rate": 1.0209255497749209e-06,
+ "loss": 0.30583910942077636,
+ "num_input_tokens_seen": 1030073088,
+ "step": 23555,
+ "train_runtime": 182170.9808,
+ "train_tokens_per_second": 5654.43
+ },
+ {
+ "epoch": 1.58631834096418,
+ "grad_norm": 0.7044002037356724,
+ "learning_rate": 1.0193250087762735e-06,
+ "loss": 0.28136019706726073,
+ "num_input_tokens_seen": 1030294120,
+ "step": 23560,
+ "train_runtime": 182212.5582,
+ "train_tokens_per_second": 5654.353
+ },
+ {
+ "epoch": 1.58665499596014,
+ "grad_norm": 0.626174961728994,
+ "learning_rate": 1.0177255809606934e-06,
+ "loss": 0.29512906074523926,
+ "num_input_tokens_seen": 1030518512,
+ "step": 23565,
+ "train_runtime": 182245.4993,
+ "train_tokens_per_second": 5654.562
+ },
+ {
+ "epoch": 1.5869916509561002,
+ "grad_norm": 0.7958252805077025,
+ "learning_rate": 1.0161272667754562e-06,
+ "loss": 0.2986447334289551,
+ "num_input_tokens_seen": 1030753488,
+ "step": 23570,
+ "train_runtime": 182291.1583,
+ "train_tokens_per_second": 5654.435
+ },
+ {
+ "epoch": 1.5873283059520604,
+ "grad_norm": 0.6503902978361567,
+ "learning_rate": 1.014530066667524e-06,
+ "loss": 0.2920704364776611,
+ "num_input_tokens_seen": 1030980864,
+ "step": 23575,
+ "train_runtime": 182338.3663,
+ "train_tokens_per_second": 5654.218
+ },
+ {
+ "epoch": 1.5876649609480205,
+ "grad_norm": 0.7822423572349132,
+ "learning_rate": 1.0129339810835526e-06,
+ "loss": 0.26558356285095214,
+ "num_input_tokens_seen": 1031184200,
+ "step": 23580,
+ "train_runtime": 182380.0639,
+ "train_tokens_per_second": 5654.04
+ },
+ {
+ "epoch": 1.5880016159439805,
+ "grad_norm": 0.7740865888482027,
+ "learning_rate": 1.0113390104698767e-06,
+ "loss": 0.2880537986755371,
+ "num_input_tokens_seen": 1031415416,
+ "step": 23585,
+ "train_runtime": 182415.6709,
+ "train_tokens_per_second": 5654.204
+ },
+ {
+ "epoch": 1.5883382709399407,
+ "grad_norm": 0.6444406633812003,
+ "learning_rate": 1.009745155272529e-06,
+ "loss": 0.32445645332336426,
+ "num_input_tokens_seen": 1031642064,
+ "step": 23590,
+ "train_runtime": 182452.8245,
+ "train_tokens_per_second": 5654.295
+ },
+ {
+ "epoch": 1.588674925935901,
+ "grad_norm": 0.716940757316124,
+ "learning_rate": 1.0081524159372246e-06,
+ "loss": 0.30813934803009035,
+ "num_input_tokens_seen": 1031852072,
+ "step": 23595,
+ "train_runtime": 182489.063,
+ "train_tokens_per_second": 5654.323
+ },
+ {
+ "epoch": 1.589011580931861,
+ "grad_norm": 0.7929537083603673,
+ "learning_rate": 1.0065607929093685e-06,
+ "loss": 0.2949239253997803,
+ "num_input_tokens_seen": 1032066320,
+ "step": 23600,
+ "train_runtime": 182530.2797,
+ "train_tokens_per_second": 5654.22
+ },
+ {
+ "epoch": 1.589348235927821,
+ "grad_norm": 0.7144421636045533,
+ "learning_rate": 1.0049702866340521e-06,
+ "loss": 0.2600874423980713,
+ "num_input_tokens_seen": 1032291736,
+ "step": 23605,
+ "train_runtime": 182571.9611,
+ "train_tokens_per_second": 5654.164
+ },
+ {
+ "epoch": 1.5896848909237813,
+ "grad_norm": 0.7689607902349916,
+ "learning_rate": 1.0033808975560556e-06,
+ "loss": 0.28646035194396974,
+ "num_input_tokens_seen": 1032506240,
+ "step": 23610,
+ "train_runtime": 182609.5452,
+ "train_tokens_per_second": 5654.175
+ },
+ {
+ "epoch": 1.5900215459197415,
+ "grad_norm": 0.7558437464479234,
+ "learning_rate": 1.0017926261198473e-06,
+ "loss": 0.29119462966918946,
+ "num_input_tokens_seen": 1032720936,
+ "step": 23615,
+ "train_runtime": 182649.2323,
+ "train_tokens_per_second": 5654.121
+ },
+ {
+ "epoch": 1.5903582009157016,
+ "grad_norm": 0.8842680705984803,
+ "learning_rate": 1.0002054727695814e-06,
+ "loss": 0.30643343925476074,
+ "num_input_tokens_seen": 1032929224,
+ "step": 23620,
+ "train_runtime": 182690.3573,
+ "train_tokens_per_second": 5653.989
+ },
+ {
+ "epoch": 1.5906948559116616,
+ "grad_norm": 0.6572152521341675,
+ "learning_rate": 9.986194379490993e-07,
+ "loss": 0.29436190128326417,
+ "num_input_tokens_seen": 1033140248,
+ "step": 23625,
+ "train_runtime": 182726.5478,
+ "train_tokens_per_second": 5654.024
+ },
+ {
+ "epoch": 1.5910315109076218,
+ "grad_norm": 0.7264405013312064,
+ "learning_rate": 9.970345221019345e-07,
+ "loss": 0.30518579483032227,
+ "num_input_tokens_seen": 1033354080,
+ "step": 23630,
+ "train_runtime": 182768.4987,
+ "train_tokens_per_second": 5653.896
+ },
+ {
+ "epoch": 1.591368165903582,
+ "grad_norm": 0.9173611218409357,
+ "learning_rate": 9.954507256712987e-07,
+ "loss": 0.30224905014038084,
+ "num_input_tokens_seen": 1033566992,
+ "step": 23635,
+ "train_runtime": 182805.0001,
+ "train_tokens_per_second": 5653.932
+ },
+ {
+ "epoch": 1.5917048208995421,
+ "grad_norm": 0.9443978197431155,
+ "learning_rate": 9.938680491000991e-07,
+ "loss": 0.3042686700820923,
+ "num_input_tokens_seen": 1033777624,
+ "step": 23640,
+ "train_runtime": 182847.8991,
+ "train_tokens_per_second": 5653.757
+ },
+ {
+ "epoch": 1.5920414758955022,
+ "grad_norm": 0.7432434366050173,
+ "learning_rate": 9.92286492830925e-07,
+ "loss": 0.2958357810974121,
+ "num_input_tokens_seen": 1034004184,
+ "step": 23645,
+ "train_runtime": 182887.2125,
+ "train_tokens_per_second": 5653.781
+ },
+ {
+ "epoch": 1.5923781308914624,
+ "grad_norm": 0.7076131103520957,
+ "learning_rate": 9.907060573060535e-07,
+ "loss": 0.31322293281555175,
+ "num_input_tokens_seen": 1034231736,
+ "step": 23650,
+ "train_runtime": 182921.8257,
+ "train_tokens_per_second": 5653.955
+ },
+ {
+ "epoch": 1.5927147858874227,
+ "grad_norm": 0.7877155616601498,
+ "learning_rate": 9.891267429674484e-07,
+ "loss": 0.3153421401977539,
+ "num_input_tokens_seen": 1034440472,
+ "step": 23655,
+ "train_runtime": 182960.1809,
+ "train_tokens_per_second": 5653.91
+ },
+ {
+ "epoch": 1.5930514408833827,
+ "grad_norm": 0.7335260153611561,
+ "learning_rate": 9.8754855025676e-07,
+ "loss": 0.28733248710632325,
+ "num_input_tokens_seen": 1034659216,
+ "step": 23660,
+ "train_runtime": 182996.9749,
+ "train_tokens_per_second": 5653.969
+ },
+ {
+ "epoch": 1.5933880958793427,
+ "grad_norm": 0.7659546851429526,
+ "learning_rate": 9.85971479615324e-07,
+ "loss": 0.29557249546051023,
+ "num_input_tokens_seen": 1034888536,
+ "step": 23665,
+ "train_runtime": 183037.5139,
+ "train_tokens_per_second": 5653.97
+ },
+ {
+ "epoch": 1.593724750875303,
+ "grad_norm": 0.7118609901015792,
+ "learning_rate": 9.843955314841647e-07,
+ "loss": 0.28273298740386965,
+ "num_input_tokens_seen": 1035112904,
+ "step": 23670,
+ "train_runtime": 183070.8286,
+ "train_tokens_per_second": 5654.166
+ },
+ {
+ "epoch": 1.5940614058712632,
+ "grad_norm": 0.7513624785773189,
+ "learning_rate": 9.82820706303989e-07,
+ "loss": 0.29895691871643065,
+ "num_input_tokens_seen": 1035338280,
+ "step": 23675,
+ "train_runtime": 183114.0185,
+ "train_tokens_per_second": 5654.063
+ },
+ {
+ "epoch": 1.5943980608672232,
+ "grad_norm": 0.8383529935018983,
+ "learning_rate": 9.812470045151952e-07,
+ "loss": 0.3054283618927002,
+ "num_input_tokens_seen": 1035542816,
+ "step": 23680,
+ "train_runtime": 183151.8635,
+ "train_tokens_per_second": 5654.012
+ },
+ {
+ "epoch": 1.5947347158631833,
+ "grad_norm": 0.7538875645456946,
+ "learning_rate": 9.79674426557859e-07,
+ "loss": 0.29663498401641847,
+ "num_input_tokens_seen": 1035752536,
+ "step": 23685,
+ "train_runtime": 183193.8197,
+ "train_tokens_per_second": 5653.862
+ },
+ {
+ "epoch": 1.5950713708591435,
+ "grad_norm": 0.6945464407471583,
+ "learning_rate": 9.781029728717513e-07,
+ "loss": 0.2824697732925415,
+ "num_input_tokens_seen": 1035967072,
+ "step": 23690,
+ "train_runtime": 183231.5083,
+ "train_tokens_per_second": 5653.87
+ },
+ {
+ "epoch": 1.5954080258551038,
+ "grad_norm": 0.7148723682038612,
+ "learning_rate": 9.765326438963218e-07,
+ "loss": 0.31726827621459963,
+ "num_input_tokens_seen": 1036183056,
+ "step": 23695,
+ "train_runtime": 183268.6587,
+ "train_tokens_per_second": 5653.902
+ },
+ {
+ "epoch": 1.5957446808510638,
+ "grad_norm": 0.7520572590185046,
+ "learning_rate": 9.749634400707087e-07,
+ "loss": 0.29088969230651857,
+ "num_input_tokens_seen": 1036386224,
+ "step": 23700,
+ "train_runtime": 183304.0574,
+ "train_tokens_per_second": 5653.919
+ },
+ {
+ "epoch": 1.5960813358470238,
+ "grad_norm": 0.691989732417355,
+ "learning_rate": 9.733953618337344e-07,
+ "loss": 0.3045504570007324,
+ "num_input_tokens_seen": 1036590880,
+ "step": 23705,
+ "train_runtime": 183345.8827,
+ "train_tokens_per_second": 5653.745
+ },
+ {
+ "epoch": 1.596417990842984,
+ "grad_norm": 0.7360691080511433,
+ "learning_rate": 9.718284096239077e-07,
+ "loss": 0.315799617767334,
+ "num_input_tokens_seen": 1036831896,
+ "step": 23710,
+ "train_runtime": 183384.7735,
+ "train_tokens_per_second": 5653.86
+ },
+ {
+ "epoch": 1.5967546458389443,
+ "grad_norm": 0.780467411583108,
+ "learning_rate": 9.702625838794215e-07,
+ "loss": 0.28366994857788086,
+ "num_input_tokens_seen": 1037051528,
+ "step": 23715,
+ "train_runtime": 183424.345,
+ "train_tokens_per_second": 5653.838
+ },
+ {
+ "epoch": 1.5970913008349044,
+ "grad_norm": 0.7083055249265116,
+ "learning_rate": 9.686978850381535e-07,
+ "loss": 0.28309922218322753,
+ "num_input_tokens_seen": 1037274432,
+ "step": 23720,
+ "train_runtime": 183463.0678,
+ "train_tokens_per_second": 5653.86
+ },
+ {
+ "epoch": 1.5974279558308644,
+ "grad_norm": 0.7040502210402872,
+ "learning_rate": 9.67134313537666e-07,
+ "loss": 0.28775806427001954,
+ "num_input_tokens_seen": 1037487448,
+ "step": 23725,
+ "train_runtime": 183498.8739,
+ "train_tokens_per_second": 5653.917
+ },
+ {
+ "epoch": 1.5977646108268246,
+ "grad_norm": 0.7648445160032851,
+ "learning_rate": 9.6557186981521e-07,
+ "loss": 0.2938704013824463,
+ "num_input_tokens_seen": 1037711184,
+ "step": 23730,
+ "train_runtime": 183532.5183,
+ "train_tokens_per_second": 5654.1
+ },
+ {
+ "epoch": 1.5981012658227849,
+ "grad_norm": 0.7590116723636553,
+ "learning_rate": 9.640105543077133e-07,
+ "loss": 0.29409012794494627,
+ "num_input_tokens_seen": 1037930696,
+ "step": 23735,
+ "train_runtime": 183569.9815,
+ "train_tokens_per_second": 5654.142
+ },
+ {
+ "epoch": 1.598437920818745,
+ "grad_norm": 0.6999829490635439,
+ "learning_rate": 9.624503674517972e-07,
+ "loss": 0.31141953468322753,
+ "num_input_tokens_seen": 1038151728,
+ "step": 23740,
+ "train_runtime": 183616.0849,
+ "train_tokens_per_second": 5653.926
+ },
+ {
+ "epoch": 1.598774575814705,
+ "grad_norm": 0.6931872830922609,
+ "learning_rate": 9.608913096837603e-07,
+ "loss": 0.2785954475402832,
+ "num_input_tokens_seen": 1038375320,
+ "step": 23745,
+ "train_runtime": 183658.3574,
+ "train_tokens_per_second": 5653.842
+ },
+ {
+ "epoch": 1.5991112308106652,
+ "grad_norm": 0.7826269007767029,
+ "learning_rate": 9.593333814395889e-07,
+ "loss": 0.30640616416931155,
+ "num_input_tokens_seen": 1038579600,
+ "step": 23750,
+ "train_runtime": 183690.2536,
+ "train_tokens_per_second": 5653.972
+ },
+ {
+ "epoch": 1.5994478858066254,
+ "grad_norm": 0.659801248076501,
+ "learning_rate": 9.577765831549529e-07,
+ "loss": 0.29427356719970704,
+ "num_input_tokens_seen": 1038803840,
+ "step": 23755,
+ "train_runtime": 183723.8696,
+ "train_tokens_per_second": 5654.158
+ },
+ {
+ "epoch": 1.5997845408025855,
+ "grad_norm": 0.7271704349083015,
+ "learning_rate": 9.562209152652058e-07,
+ "loss": 0.30373120307922363,
+ "num_input_tokens_seen": 1039013760,
+ "step": 23760,
+ "train_runtime": 183766.5421,
+ "train_tokens_per_second": 5653.988
+ },
+ {
+ "epoch": 1.6001211957985455,
+ "grad_norm": 0.8875347712272086,
+ "learning_rate": 9.54666378205385e-07,
+ "loss": 0.3227171182632446,
+ "num_input_tokens_seen": 1039242488,
+ "step": 23765,
+ "train_runtime": 183813.6297,
+ "train_tokens_per_second": 5653.784
+ },
+ {
+ "epoch": 1.6004578507945058,
+ "grad_norm": 0.8111840363520786,
+ "learning_rate": 9.531129724102117e-07,
+ "loss": 0.2936609983444214,
+ "num_input_tokens_seen": 1039454848,
+ "step": 23770,
+ "train_runtime": 183852.3211,
+ "train_tokens_per_second": 5653.749
+ },
+ {
+ "epoch": 1.600794505790466,
+ "grad_norm": 0.7937834563085631,
+ "learning_rate": 9.515606983140896e-07,
+ "loss": 0.2982936859130859,
+ "num_input_tokens_seen": 1039677264,
+ "step": 23775,
+ "train_runtime": 183886.8713,
+ "train_tokens_per_second": 5653.896
+ },
+ {
+ "epoch": 1.601131160786426,
+ "grad_norm": 0.7845551974583694,
+ "learning_rate": 9.500095563511119e-07,
+ "loss": 0.3003803253173828,
+ "num_input_tokens_seen": 1039906360,
+ "step": 23780,
+ "train_runtime": 183931.0004,
+ "train_tokens_per_second": 5653.785
+ },
+ {
+ "epoch": 1.601467815782386,
+ "grad_norm": 0.6656969784743623,
+ "learning_rate": 9.484595469550445e-07,
+ "loss": 0.28896503448486327,
+ "num_input_tokens_seen": 1040131192,
+ "step": 23785,
+ "train_runtime": 183970.5264,
+ "train_tokens_per_second": 5653.793
+ },
+ {
+ "epoch": 1.6018044707783463,
+ "grad_norm": 0.6946762642734486,
+ "learning_rate": 9.469106705593462e-07,
+ "loss": 0.27666046619415285,
+ "num_input_tokens_seen": 1040333032,
+ "step": 23790,
+ "train_runtime": 184009.1267,
+ "train_tokens_per_second": 5653.703
+ },
+ {
+ "epoch": 1.6021411257743066,
+ "grad_norm": 0.7736179336799578,
+ "learning_rate": 9.453629275971549e-07,
+ "loss": 0.29598097801208495,
+ "num_input_tokens_seen": 1040552408,
+ "step": 23795,
+ "train_runtime": 184047.3995,
+ "train_tokens_per_second": 5653.72
+ },
+ {
+ "epoch": 1.6024777807702666,
+ "grad_norm": 0.6664557365960093,
+ "learning_rate": 9.438163185012916e-07,
+ "loss": 0.28930160999298093,
+ "num_input_tokens_seen": 1040764544,
+ "step": 23800,
+ "train_runtime": 184091.6726,
+ "train_tokens_per_second": 5653.512
+ },
+ {
+ "epoch": 1.6028144357662266,
+ "grad_norm": 0.7303797761962147,
+ "learning_rate": 9.422708437042604e-07,
+ "loss": 0.2819626808166504,
+ "num_input_tokens_seen": 1040990752,
+ "step": 23805,
+ "train_runtime": 184135.0608,
+ "train_tokens_per_second": 5653.409
+ },
+ {
+ "epoch": 1.6031510907621869,
+ "grad_norm": 0.7712489855582633,
+ "learning_rate": 9.40726503638249e-07,
+ "loss": 0.31176731586456297,
+ "num_input_tokens_seen": 1041206528,
+ "step": 23810,
+ "train_runtime": 184167.4154,
+ "train_tokens_per_second": 5653.587
+ },
+ {
+ "epoch": 1.6034877457581471,
+ "grad_norm": 0.8231902257626785,
+ "learning_rate": 9.391832987351268e-07,
+ "loss": 0.29704809188842773,
+ "num_input_tokens_seen": 1041421728,
+ "step": 23815,
+ "train_runtime": 184202.3075,
+ "train_tokens_per_second": 5653.684
+ },
+ {
+ "epoch": 1.6038244007541071,
+ "grad_norm": 0.8241864917839168,
+ "learning_rate": 9.376412294264458e-07,
+ "loss": 0.29626688957214353,
+ "num_input_tokens_seen": 1041643056,
+ "step": 23820,
+ "train_runtime": 184245.0385,
+ "train_tokens_per_second": 5653.575
+ },
+ {
+ "epoch": 1.6041610557500672,
+ "grad_norm": 0.6892308812819989,
+ "learning_rate": 9.361002961434401e-07,
+ "loss": 0.29646663665771483,
+ "num_input_tokens_seen": 1041859688,
+ "step": 23825,
+ "train_runtime": 184286.6378,
+ "train_tokens_per_second": 5653.474
+ },
+ {
+ "epoch": 1.6044977107460274,
+ "grad_norm": 0.8147077105803181,
+ "learning_rate": 9.345604993170299e-07,
+ "loss": 0.31340832710266114,
+ "num_input_tokens_seen": 1042088840,
+ "step": 23830,
+ "train_runtime": 184324.2778,
+ "train_tokens_per_second": 5653.563
+ },
+ {
+ "epoch": 1.6048343657419877,
+ "grad_norm": 0.7853709628304697,
+ "learning_rate": 9.330218393778101e-07,
+ "loss": 0.29099485874176023,
+ "num_input_tokens_seen": 1042295560,
+ "step": 23835,
+ "train_runtime": 184359.1879,
+ "train_tokens_per_second": 5653.613
+ },
+ {
+ "epoch": 1.6051710207379477,
+ "grad_norm": 0.8167824477623031,
+ "learning_rate": 9.314843167560656e-07,
+ "loss": 0.2913378715515137,
+ "num_input_tokens_seen": 1042510744,
+ "step": 23840,
+ "train_runtime": 184393.2673,
+ "train_tokens_per_second": 5653.735
+ },
+ {
+ "epoch": 1.6055076757339077,
+ "grad_norm": 0.7205357729428846,
+ "learning_rate": 9.299479318817578e-07,
+ "loss": 0.29028098583221434,
+ "num_input_tokens_seen": 1042725176,
+ "step": 23845,
+ "train_runtime": 184433.6182,
+ "train_tokens_per_second": 5653.661
+ },
+ {
+ "epoch": 1.605844330729868,
+ "grad_norm": 0.766245897972215,
+ "learning_rate": 9.284126851845316e-07,
+ "loss": 0.2781973361968994,
+ "num_input_tokens_seen": 1042960480,
+ "step": 23850,
+ "train_runtime": 184473.0014,
+ "train_tokens_per_second": 5653.73
+ },
+ {
+ "epoch": 1.6061809857258282,
+ "grad_norm": 0.7749016859550374,
+ "learning_rate": 9.268785770937172e-07,
+ "loss": 0.3008106231689453,
+ "num_input_tokens_seen": 1043174632,
+ "step": 23855,
+ "train_runtime": 184513.6412,
+ "train_tokens_per_second": 5653.645
+ },
+ {
+ "epoch": 1.6065176407217883,
+ "grad_norm": 0.8470162466279486,
+ "learning_rate": 9.253456080383178e-07,
+ "loss": 0.3095564842224121,
+ "num_input_tokens_seen": 1043395416,
+ "step": 23860,
+ "train_runtime": 184550.3827,
+ "train_tokens_per_second": 5653.716
+ },
+ {
+ "epoch": 1.6068542957177483,
+ "grad_norm": 0.806209780755035,
+ "learning_rate": 9.238137784470286e-07,
+ "loss": 0.32625675201416016,
+ "num_input_tokens_seen": 1043602960,
+ "step": 23865,
+ "train_runtime": 184582.4698,
+ "train_tokens_per_second": 5653.857
+ },
+ {
+ "epoch": 1.6071909507137085,
+ "grad_norm": 0.7896528179568558,
+ "learning_rate": 9.222830887482153e-07,
+ "loss": 0.2661672353744507,
+ "num_input_tokens_seen": 1043826208,
+ "step": 23870,
+ "train_runtime": 184619.7546,
+ "train_tokens_per_second": 5653.925
+ },
+ {
+ "epoch": 1.6075276057096688,
+ "grad_norm": 0.8946960868592196,
+ "learning_rate": 9.207535393699351e-07,
+ "loss": 0.3128221035003662,
+ "num_input_tokens_seen": 1044034312,
+ "step": 23875,
+ "train_runtime": 184658.2466,
+ "train_tokens_per_second": 5653.873
+ },
+ {
+ "epoch": 1.6078642607056288,
+ "grad_norm": 0.8295488010306658,
+ "learning_rate": 9.192251307399197e-07,
+ "loss": 0.29769034385681153,
+ "num_input_tokens_seen": 1044252784,
+ "step": 23880,
+ "train_runtime": 184694.2905,
+ "train_tokens_per_second": 5653.953
+ },
+ {
+ "epoch": 1.6082009157015889,
+ "grad_norm": 0.7652692381284338,
+ "learning_rate": 9.176978632855842e-07,
+ "loss": 0.2840586185455322,
+ "num_input_tokens_seen": 1044471704,
+ "step": 23885,
+ "train_runtime": 184732.0494,
+ "train_tokens_per_second": 5653.982
+ },
+ {
+ "epoch": 1.608537570697549,
+ "grad_norm": 0.7068737921677537,
+ "learning_rate": 9.161717374340235e-07,
+ "loss": 0.2887522459030151,
+ "num_input_tokens_seen": 1044709256,
+ "step": 23890,
+ "train_runtime": 184770.2163,
+ "train_tokens_per_second": 5654.1
+ },
+ {
+ "epoch": 1.6088742256935094,
+ "grad_norm": 0.865885020527099,
+ "learning_rate": 9.14646753612014e-07,
+ "loss": 0.3057317495346069,
+ "num_input_tokens_seen": 1044932856,
+ "step": 23895,
+ "train_runtime": 184802.9042,
+ "train_tokens_per_second": 5654.31
+ },
+ {
+ "epoch": 1.6092108806894694,
+ "grad_norm": 0.7260701569352984,
+ "learning_rate": 9.131229122460112e-07,
+ "loss": 0.29195594787597656,
+ "num_input_tokens_seen": 1045169176,
+ "step": 23900,
+ "train_runtime": 184847.6389,
+ "train_tokens_per_second": 5654.22
+ },
+ {
+ "epoch": 1.6095475356854294,
+ "grad_norm": 0.7232801129725709,
+ "learning_rate": 9.116002137621566e-07,
+ "loss": 0.30215911865234374,
+ "num_input_tokens_seen": 1045388344,
+ "step": 23905,
+ "train_runtime": 184888.5248,
+ "train_tokens_per_second": 5654.155
+ },
+ {
+ "epoch": 1.6098841906813897,
+ "grad_norm": 0.8623009349911321,
+ "learning_rate": 9.100786585862626e-07,
+ "loss": 0.33268914222717283,
+ "num_input_tokens_seen": 1045590472,
+ "step": 23910,
+ "train_runtime": 184932.0284,
+ "train_tokens_per_second": 5653.918
+ },
+ {
+ "epoch": 1.61022084567735,
+ "grad_norm": 0.8126716994808482,
+ "learning_rate": 9.085582471438326e-07,
+ "loss": 0.3007976531982422,
+ "num_input_tokens_seen": 1045807448,
+ "step": 23915,
+ "train_runtime": 184968.2847,
+ "train_tokens_per_second": 5653.983
+ },
+ {
+ "epoch": 1.61055750067331,
+ "grad_norm": 0.9161097317548726,
+ "learning_rate": 9.070389798600394e-07,
+ "loss": 0.3141302585601807,
+ "num_input_tokens_seen": 1046022008,
+ "step": 23920,
+ "train_runtime": 185008.228,
+ "train_tokens_per_second": 5653.922
+ },
+ {
+ "epoch": 1.61089415566927,
+ "grad_norm": 0.7937859830875522,
+ "learning_rate": 9.055208571597451e-07,
+ "loss": 0.3039546489715576,
+ "num_input_tokens_seen": 1046240136,
+ "step": 23925,
+ "train_runtime": 185046.349,
+ "train_tokens_per_second": 5653.936
+ },
+ {
+ "epoch": 1.6112308106652302,
+ "grad_norm": 0.7922862286264548,
+ "learning_rate": 9.040038794674872e-07,
+ "loss": 0.30896403789520266,
+ "num_input_tokens_seen": 1046460096,
+ "step": 23930,
+ "train_runtime": 185087.8953,
+ "train_tokens_per_second": 5653.855
+ },
+ {
+ "epoch": 1.6115674656611905,
+ "grad_norm": 0.7338711700679107,
+ "learning_rate": 9.024880472074831e-07,
+ "loss": 0.28701066970825195,
+ "num_input_tokens_seen": 1046662408,
+ "step": 23935,
+ "train_runtime": 185132.2278,
+ "train_tokens_per_second": 5653.594
+ },
+ {
+ "epoch": 1.6119041206571505,
+ "grad_norm": 0.7141524995931083,
+ "learning_rate": 9.009733608036308e-07,
+ "loss": 0.30707645416259766,
+ "num_input_tokens_seen": 1046889064,
+ "step": 23940,
+ "train_runtime": 185169.7651,
+ "train_tokens_per_second": 5653.672
+ },
+ {
+ "epoch": 1.6122407756531107,
+ "grad_norm": 0.7277986509792717,
+ "learning_rate": 8.994598206795068e-07,
+ "loss": 0.29622645378112794,
+ "num_input_tokens_seen": 1047104032,
+ "step": 23945,
+ "train_runtime": 185205.0679,
+ "train_tokens_per_second": 5653.755
+ },
+ {
+ "epoch": 1.612577430649071,
+ "grad_norm": 0.735906736712434,
+ "learning_rate": 8.979474272583672e-07,
+ "loss": 0.28798418045043944,
+ "num_input_tokens_seen": 1047312440,
+ "step": 23950,
+ "train_runtime": 185247.3251,
+ "train_tokens_per_second": 5653.59
+ },
+ {
+ "epoch": 1.612914085645031,
+ "grad_norm": 0.7572638632134422,
+ "learning_rate": 8.964361809631517e-07,
+ "loss": 0.3181246280670166,
+ "num_input_tokens_seen": 1047539480,
+ "step": 23955,
+ "train_runtime": 185289.8341,
+ "train_tokens_per_second": 5653.518
+ },
+ {
+ "epoch": 1.613250740640991,
+ "grad_norm": 0.8041739758256639,
+ "learning_rate": 8.949260822164707e-07,
+ "loss": 0.31255478858947755,
+ "num_input_tokens_seen": 1047761560,
+ "step": 23960,
+ "train_runtime": 185326.4519,
+ "train_tokens_per_second": 5653.6
+ },
+ {
+ "epoch": 1.6135873956369513,
+ "grad_norm": 0.7237860013192642,
+ "learning_rate": 8.934171314406226e-07,
+ "loss": 0.2988784074783325,
+ "num_input_tokens_seen": 1047987312,
+ "step": 23965,
+ "train_runtime": 185361.8266,
+ "train_tokens_per_second": 5653.739
+ },
+ {
+ "epoch": 1.6139240506329116,
+ "grad_norm": 0.7942034276838248,
+ "learning_rate": 8.919093290575765e-07,
+ "loss": 0.30140008926391604,
+ "num_input_tokens_seen": 1048194320,
+ "step": 23970,
+ "train_runtime": 185405.0815,
+ "train_tokens_per_second": 5653.536
+ },
+ {
+ "epoch": 1.6142607056288716,
+ "grad_norm": 0.8222295997811078,
+ "learning_rate": 8.904026754889877e-07,
+ "loss": 0.30414538383483886,
+ "num_input_tokens_seen": 1048404056,
+ "step": 23975,
+ "train_runtime": 185444.5814,
+ "train_tokens_per_second": 5653.463
+ },
+ {
+ "epoch": 1.6145973606248316,
+ "grad_norm": 0.8609222631961138,
+ "learning_rate": 8.888971711561867e-07,
+ "loss": 0.32020394802093505,
+ "num_input_tokens_seen": 1048613480,
+ "step": 23980,
+ "train_runtime": 185486.5715,
+ "train_tokens_per_second": 5653.312
+ },
+ {
+ "epoch": 1.6149340156207919,
+ "grad_norm": 0.8930736881841469,
+ "learning_rate": 8.873928164801821e-07,
+ "loss": 0.32930703163146974,
+ "num_input_tokens_seen": 1048836768,
+ "step": 23985,
+ "train_runtime": 185525.6042,
+ "train_tokens_per_second": 5653.326
+ },
+ {
+ "epoch": 1.6152706706167521,
+ "grad_norm": 0.7714631943477365,
+ "learning_rate": 8.858896118816624e-07,
+ "loss": 0.2910914421081543,
+ "num_input_tokens_seen": 1049026440,
+ "step": 23990,
+ "train_runtime": 185568.2914,
+ "train_tokens_per_second": 5653.048
+ },
+ {
+ "epoch": 1.6156073256127121,
+ "grad_norm": 0.759228215290319,
+ "learning_rate": 8.84387557780994e-07,
+ "loss": 0.2792020797729492,
+ "num_input_tokens_seen": 1049235720,
+ "step": 23995,
+ "train_runtime": 185611.4217,
+ "train_tokens_per_second": 5652.862
+ },
+ {
+ "epoch": 1.6159439806086722,
+ "grad_norm": 0.7757710745934641,
+ "learning_rate": 8.828866545982196e-07,
+ "loss": 0.2844292163848877,
+ "num_input_tokens_seen": 1049455984,
+ "step": 24000,
+ "train_runtime": 185652.6171,
+ "train_tokens_per_second": 5652.794
+ },
+ {
+ "epoch": 1.6162806356046324,
+ "grad_norm": 0.7617714598274377,
+ "learning_rate": 8.813869027530664e-07,
+ "loss": 0.28838918209075926,
+ "num_input_tokens_seen": 1049680856,
+ "step": 24005,
+ "train_runtime": 185698.9704,
+ "train_tokens_per_second": 5652.594
+ },
+ {
+ "epoch": 1.6166172906005927,
+ "grad_norm": 0.7142148786302278,
+ "learning_rate": 8.798883026649302e-07,
+ "loss": 0.31560914516448973,
+ "num_input_tokens_seen": 1049914184,
+ "step": 24010,
+ "train_runtime": 185735.3739,
+ "train_tokens_per_second": 5652.742
+ },
+ {
+ "epoch": 1.6169539455965527,
+ "grad_norm": 0.6136150732667002,
+ "learning_rate": 8.783908547528947e-07,
+ "loss": 0.2625638246536255,
+ "num_input_tokens_seen": 1050126648,
+ "step": 24015,
+ "train_runtime": 185774.3978,
+ "train_tokens_per_second": 5652.698
+ },
+ {
+ "epoch": 1.6172906005925127,
+ "grad_norm": 0.7756733989123253,
+ "learning_rate": 8.768945594357109e-07,
+ "loss": 0.29282054901123045,
+ "num_input_tokens_seen": 1050339456,
+ "step": 24020,
+ "train_runtime": 185813.7393,
+ "train_tokens_per_second": 5652.647
+ },
+ {
+ "epoch": 1.617627255588473,
+ "grad_norm": 0.900598117098456,
+ "learning_rate": 8.753994171318175e-07,
+ "loss": 0.3160220146179199,
+ "num_input_tokens_seen": 1050547328,
+ "step": 24025,
+ "train_runtime": 185851.1561,
+ "train_tokens_per_second": 5652.627
+ },
+ {
+ "epoch": 1.6179639105844332,
+ "grad_norm": 0.7405411228891803,
+ "learning_rate": 8.739054282593245e-07,
+ "loss": 0.3164478063583374,
+ "num_input_tokens_seen": 1050771952,
+ "step": 24030,
+ "train_runtime": 185890.3938,
+ "train_tokens_per_second": 5652.643
+ },
+ {
+ "epoch": 1.6183005655803933,
+ "grad_norm": 0.835344576020692,
+ "learning_rate": 8.724125932360217e-07,
+ "loss": 0.3083709716796875,
+ "num_input_tokens_seen": 1050984192,
+ "step": 24035,
+ "train_runtime": 185929.4023,
+ "train_tokens_per_second": 5652.598
+ },
+ {
+ "epoch": 1.6186372205763533,
+ "grad_norm": 0.7799187017517342,
+ "learning_rate": 8.70920912479375e-07,
+ "loss": 0.3174473762512207,
+ "num_input_tokens_seen": 1051196368,
+ "step": 24040,
+ "train_runtime": 185964.2309,
+ "train_tokens_per_second": 5652.68
+ },
+ {
+ "epoch": 1.6189738755723135,
+ "grad_norm": 0.7850184374405595,
+ "learning_rate": 8.694303864065285e-07,
+ "loss": 0.29400835037231443,
+ "num_input_tokens_seen": 1051425808,
+ "step": 24045,
+ "train_runtime": 186005.0047,
+ "train_tokens_per_second": 5652.675
+ },
+ {
+ "epoch": 1.6193105305682738,
+ "grad_norm": 0.7182128587421389,
+ "learning_rate": 8.679410154343015e-07,
+ "loss": 0.29439725875854494,
+ "num_input_tokens_seen": 1051646216,
+ "step": 24050,
+ "train_runtime": 186042.8773,
+ "train_tokens_per_second": 5652.709
+ },
+ {
+ "epoch": 1.6196471855642338,
+ "grad_norm": 0.7969951778334732,
+ "learning_rate": 8.664527999791961e-07,
+ "loss": 0.3235488414764404,
+ "num_input_tokens_seen": 1051864880,
+ "step": 24055,
+ "train_runtime": 186082.635,
+ "train_tokens_per_second": 5652.676
+ },
+ {
+ "epoch": 1.6199838405601938,
+ "grad_norm": 0.7800713274850012,
+ "learning_rate": 8.649657404573819e-07,
+ "loss": 0.29723272323608396,
+ "num_input_tokens_seen": 1052093008,
+ "step": 24060,
+ "train_runtime": 186126.3919,
+ "train_tokens_per_second": 5652.573
+ },
+ {
+ "epoch": 1.620320495556154,
+ "grad_norm": 0.8274231375759122,
+ "learning_rate": 8.634798372847148e-07,
+ "loss": 0.317378568649292,
+ "num_input_tokens_seen": 1052316944,
+ "step": 24065,
+ "train_runtime": 186162.2307,
+ "train_tokens_per_second": 5652.688
+ },
+ {
+ "epoch": 1.6206571505521143,
+ "grad_norm": 0.8056126386548218,
+ "learning_rate": 8.619950908767189e-07,
+ "loss": 0.2666431665420532,
+ "num_input_tokens_seen": 1052516136,
+ "step": 24070,
+ "train_runtime": 186205.2929,
+ "train_tokens_per_second": 5652.45
+ },
+ {
+ "epoch": 1.6209938055480744,
+ "grad_norm": 0.6724696382638301,
+ "learning_rate": 8.605115016486016e-07,
+ "loss": 0.2880539894104004,
+ "num_input_tokens_seen": 1052748400,
+ "step": 24075,
+ "train_runtime": 186243.1874,
+ "train_tokens_per_second": 5652.547
+ },
+ {
+ "epoch": 1.6213304605440344,
+ "grad_norm": 0.9066394567633849,
+ "learning_rate": 8.590290700152426e-07,
+ "loss": 0.2833811044692993,
+ "num_input_tokens_seen": 1052963936,
+ "step": 24080,
+ "train_runtime": 186289.0516,
+ "train_tokens_per_second": 5652.313
+ },
+ {
+ "epoch": 1.6216671155399947,
+ "grad_norm": 0.799201189010084,
+ "learning_rate": 8.575477963912005e-07,
+ "loss": 0.29882135391235354,
+ "num_input_tokens_seen": 1053188096,
+ "step": 24085,
+ "train_runtime": 186329.9225,
+ "train_tokens_per_second": 5652.276
+ },
+ {
+ "epoch": 1.622003770535955,
+ "grad_norm": 0.8755519590986088,
+ "learning_rate": 8.560676811907071e-07,
+ "loss": 0.3093993663787842,
+ "num_input_tokens_seen": 1053409680,
+ "step": 24090,
+ "train_runtime": 186370.6512,
+ "train_tokens_per_second": 5652.229
+ },
+ {
+ "epoch": 1.622340425531915,
+ "grad_norm": 1.5513430020002343,
+ "learning_rate": 8.54588724827673e-07,
+ "loss": 0.29143998622894285,
+ "num_input_tokens_seen": 1053604304,
+ "step": 24095,
+ "train_runtime": 186412.1977,
+ "train_tokens_per_second": 5652.014
+ },
+ {
+ "epoch": 1.622677080527875,
+ "grad_norm": 0.7491596315112452,
+ "learning_rate": 8.531109277156818e-07,
+ "loss": 0.2952460050582886,
+ "num_input_tokens_seen": 1053836472,
+ "step": 24100,
+ "train_runtime": 186445.3192,
+ "train_tokens_per_second": 5652.255
+ },
+ {
+ "epoch": 1.6230137355238352,
+ "grad_norm": 0.6130646121890772,
+ "learning_rate": 8.516342902679986e-07,
+ "loss": 0.26687233448028563,
+ "num_input_tokens_seen": 1054042928,
+ "step": 24105,
+ "train_runtime": 186487.115,
+ "train_tokens_per_second": 5652.095
+ },
+ {
+ "epoch": 1.6233503905197955,
+ "grad_norm": 0.7118183132847752,
+ "learning_rate": 8.501588128975557e-07,
+ "loss": 0.274417781829834,
+ "num_input_tokens_seen": 1054266256,
+ "step": 24110,
+ "train_runtime": 186525.4577,
+ "train_tokens_per_second": 5652.131
+ },
+ {
+ "epoch": 1.6236870455157555,
+ "grad_norm": 0.7790848429598987,
+ "learning_rate": 8.486844960169704e-07,
+ "loss": 0.28249123096466067,
+ "num_input_tokens_seen": 1054484008,
+ "step": 24115,
+ "train_runtime": 186565.9108,
+ "train_tokens_per_second": 5652.072
+ },
+ {
+ "epoch": 1.6240237005117155,
+ "grad_norm": 0.7894998002519212,
+ "learning_rate": 8.472113400385257e-07,
+ "loss": 0.2758941650390625,
+ "num_input_tokens_seen": 1054711848,
+ "step": 24120,
+ "train_runtime": 186605.8621,
+ "train_tokens_per_second": 5652.083
+ },
+ {
+ "epoch": 1.6243603555076758,
+ "grad_norm": 0.7543857092013571,
+ "learning_rate": 8.457393453741886e-07,
+ "loss": 0.3019557476043701,
+ "num_input_tokens_seen": 1054923912,
+ "step": 24125,
+ "train_runtime": 186648.2364,
+ "train_tokens_per_second": 5651.936
+ },
+ {
+ "epoch": 1.624697010503636,
+ "grad_norm": 0.7442690058985666,
+ "learning_rate": 8.442685124355965e-07,
+ "loss": 0.28956966400146483,
+ "num_input_tokens_seen": 1055138184,
+ "step": 24130,
+ "train_runtime": 186689.7766,
+ "train_tokens_per_second": 5651.826
+ },
+ {
+ "epoch": 1.625033665499596,
+ "grad_norm": 0.7707121718508636,
+ "learning_rate": 8.427988416340633e-07,
+ "loss": 0.31919891834259034,
+ "num_input_tokens_seen": 1055341616,
+ "step": 24135,
+ "train_runtime": 186736.5031,
+ "train_tokens_per_second": 5651.501
+ },
+ {
+ "epoch": 1.625370320495556,
+ "grad_norm": 0.8345320866926412,
+ "learning_rate": 8.413303333805777e-07,
+ "loss": 0.2868987560272217,
+ "num_input_tokens_seen": 1055555328,
+ "step": 24140,
+ "train_runtime": 186778.9844,
+ "train_tokens_per_second": 5651.36
+ },
+ {
+ "epoch": 1.6257069754915163,
+ "grad_norm": 0.832357010463937,
+ "learning_rate": 8.398629880858034e-07,
+ "loss": 0.3065673589706421,
+ "num_input_tokens_seen": 1055788776,
+ "step": 24145,
+ "train_runtime": 186814.9017,
+ "train_tokens_per_second": 5651.523
+ },
+ {
+ "epoch": 1.6260436304874766,
+ "grad_norm": 0.7264477849943358,
+ "learning_rate": 8.383968061600778e-07,
+ "loss": 0.30824875831604004,
+ "num_input_tokens_seen": 1055998904,
+ "step": 24150,
+ "train_runtime": 186851.2697,
+ "train_tokens_per_second": 5651.548
+ },
+ {
+ "epoch": 1.6263802854834366,
+ "grad_norm": 0.6983483917587298,
+ "learning_rate": 8.36931788013417e-07,
+ "loss": 0.2875575304031372,
+ "num_input_tokens_seen": 1056224128,
+ "step": 24155,
+ "train_runtime": 186885.9871,
+ "train_tokens_per_second": 5651.703
+ },
+ {
+ "epoch": 1.6267169404793966,
+ "grad_norm": 0.8238457852516533,
+ "learning_rate": 8.354679340555044e-07,
+ "loss": 0.30007798671722413,
+ "num_input_tokens_seen": 1056436296,
+ "step": 24160,
+ "train_runtime": 186918.9055,
+ "train_tokens_per_second": 5651.843
+ },
+ {
+ "epoch": 1.6270535954753569,
+ "grad_norm": 0.7716224609530045,
+ "learning_rate": 8.34005244695707e-07,
+ "loss": 0.28886971473693845,
+ "num_input_tokens_seen": 1056648824,
+ "step": 24165,
+ "train_runtime": 186953.1199,
+ "train_tokens_per_second": 5651.945
+ },
+ {
+ "epoch": 1.6273902504713171,
+ "grad_norm": 0.6680421386267813,
+ "learning_rate": 8.325437203430558e-07,
+ "loss": 0.29457979202270507,
+ "num_input_tokens_seen": 1056867768,
+ "step": 24170,
+ "train_runtime": 186995.0671,
+ "train_tokens_per_second": 5651.848
+ },
+ {
+ "epoch": 1.6277269054672772,
+ "grad_norm": 0.7051397648416529,
+ "learning_rate": 8.310833614062652e-07,
+ "loss": 0.295581316947937,
+ "num_input_tokens_seen": 1057083272,
+ "step": 24175,
+ "train_runtime": 187032.373,
+ "train_tokens_per_second": 5651.873
+ },
+ {
+ "epoch": 1.6280635604632372,
+ "grad_norm": 0.8804185228424846,
+ "learning_rate": 8.296241682937189e-07,
+ "loss": 0.2983792543411255,
+ "num_input_tokens_seen": 1057309512,
+ "step": 24180,
+ "train_runtime": 187074.9311,
+ "train_tokens_per_second": 5651.797
+ },
+ {
+ "epoch": 1.6284002154591974,
+ "grad_norm": 0.7277060223716454,
+ "learning_rate": 8.281661414134761e-07,
+ "loss": 0.2972233295440674,
+ "num_input_tokens_seen": 1057522792,
+ "step": 24185,
+ "train_runtime": 187108.976,
+ "train_tokens_per_second": 5651.908
+ },
+ {
+ "epoch": 1.6287368704551577,
+ "grad_norm": 0.7892086033151113,
+ "learning_rate": 8.267092811732686e-07,
+ "loss": 0.31084840297698973,
+ "num_input_tokens_seen": 1057745072,
+ "step": 24190,
+ "train_runtime": 187148.6775,
+ "train_tokens_per_second": 5651.897
+ },
+ {
+ "epoch": 1.6290735254511177,
+ "grad_norm": 0.730388450752643,
+ "learning_rate": 8.252535879805029e-07,
+ "loss": 0.2759894847869873,
+ "num_input_tokens_seen": 1057972008,
+ "step": 24195,
+ "train_runtime": 187187.3763,
+ "train_tokens_per_second": 5651.941
+ },
+ {
+ "epoch": 1.6294101804470777,
+ "grad_norm": 0.863594940750271,
+ "learning_rate": 8.237990622422576e-07,
+ "loss": 0.289616322517395,
+ "num_input_tokens_seen": 1058191104,
+ "step": 24200,
+ "train_runtime": 187223.8174,
+ "train_tokens_per_second": 5652.011
+ },
+ {
+ "epoch": 1.629746835443038,
+ "grad_norm": 0.7266468720434132,
+ "learning_rate": 8.223457043652905e-07,
+ "loss": 0.3101085662841797,
+ "num_input_tokens_seen": 1058412784,
+ "step": 24205,
+ "train_runtime": 187256.143,
+ "train_tokens_per_second": 5652.219
+ },
+ {
+ "epoch": 1.6300834904389983,
+ "grad_norm": 0.9420820301682334,
+ "learning_rate": 8.208935147560227e-07,
+ "loss": 0.2965271472930908,
+ "num_input_tokens_seen": 1058609720,
+ "step": 24210,
+ "train_runtime": 187288.1131,
+ "train_tokens_per_second": 5652.306
+ },
+ {
+ "epoch": 1.6304201454349583,
+ "grad_norm": 0.7345571411845794,
+ "learning_rate": 8.194424938205597e-07,
+ "loss": 0.3259326696395874,
+ "num_input_tokens_seen": 1058845344,
+ "step": 24215,
+ "train_runtime": 187328.0703,
+ "train_tokens_per_second": 5652.358
+ },
+ {
+ "epoch": 1.6307568004309183,
+ "grad_norm": 0.7486771661318714,
+ "learning_rate": 8.179926419646705e-07,
+ "loss": 0.28400454521179197,
+ "num_input_tokens_seen": 1059062096,
+ "step": 24220,
+ "train_runtime": 187371.6715,
+ "train_tokens_per_second": 5652.2
+ },
+ {
+ "epoch": 1.6310934554268786,
+ "grad_norm": 0.8281741934458996,
+ "learning_rate": 8.165439595938047e-07,
+ "loss": 0.32569429874420164,
+ "num_input_tokens_seen": 1059262648,
+ "step": 24225,
+ "train_runtime": 187414.9864,
+ "train_tokens_per_second": 5651.963
+ },
+ {
+ "epoch": 1.6314301104228388,
+ "grad_norm": 0.8133179081823259,
+ "learning_rate": 8.150964471130801e-07,
+ "loss": 0.2735768795013428,
+ "num_input_tokens_seen": 1059491256,
+ "step": 24230,
+ "train_runtime": 187448.4434,
+ "train_tokens_per_second": 5652.174
+ },
+ {
+ "epoch": 1.6317667654187988,
+ "grad_norm": 0.7466167890530678,
+ "learning_rate": 8.136501049272899e-07,
+ "loss": 0.30480170249938965,
+ "num_input_tokens_seen": 1059706760,
+ "step": 24235,
+ "train_runtime": 187485.3647,
+ "train_tokens_per_second": 5652.211
+ },
+ {
+ "epoch": 1.6321034204147589,
+ "grad_norm": 0.7655169967068027,
+ "learning_rate": 8.122049334408983e-07,
+ "loss": 0.29867599010467527,
+ "num_input_tokens_seen": 1059920888,
+ "step": 24240,
+ "train_runtime": 187522.586,
+ "train_tokens_per_second": 5652.231
+ },
+ {
+ "epoch": 1.6324400754107191,
+ "grad_norm": 0.7048483211838898,
+ "learning_rate": 8.107609330580429e-07,
+ "loss": 0.29615049362182616,
+ "num_input_tokens_seen": 1060137784,
+ "step": 24245,
+ "train_runtime": 187563.0743,
+ "train_tokens_per_second": 5652.167
+ },
+ {
+ "epoch": 1.6327767304066794,
+ "grad_norm": 0.7788326112043563,
+ "learning_rate": 8.093181041825332e-07,
+ "loss": 0.3028998613357544,
+ "num_input_tokens_seen": 1060360840,
+ "step": 24250,
+ "train_runtime": 187603.6663,
+ "train_tokens_per_second": 5652.133
+ },
+ {
+ "epoch": 1.6331133854026394,
+ "grad_norm": 0.8246354271264723,
+ "learning_rate": 8.078764472178552e-07,
+ "loss": 0.29851455688476564,
+ "num_input_tokens_seen": 1060573408,
+ "step": 24255,
+ "train_runtime": 187643.6101,
+ "train_tokens_per_second": 5652.062
+ },
+ {
+ "epoch": 1.6334500403985994,
+ "grad_norm": 0.7136275615926145,
+ "learning_rate": 8.06435962567158e-07,
+ "loss": 0.32530434131622316,
+ "num_input_tokens_seen": 1060766968,
+ "step": 24260,
+ "train_runtime": 187685.5425,
+ "train_tokens_per_second": 5651.831
+ },
+ {
+ "epoch": 1.6337866953945597,
+ "grad_norm": 0.7353400312549586,
+ "learning_rate": 8.049966506332746e-07,
+ "loss": 0.26018400192260743,
+ "num_input_tokens_seen": 1060985712,
+ "step": 24265,
+ "train_runtime": 187718.4957,
+ "train_tokens_per_second": 5652.004
+ },
+ {
+ "epoch": 1.63412335039052,
+ "grad_norm": 0.736074674468452,
+ "learning_rate": 8.035585118186984e-07,
+ "loss": 0.2798137664794922,
+ "num_input_tokens_seen": 1061195896,
+ "step": 24270,
+ "train_runtime": 187751.2703,
+ "train_tokens_per_second": 5652.137
+ },
+ {
+ "epoch": 1.63446000538648,
+ "grad_norm": 0.7545879942996399,
+ "learning_rate": 8.021215465256038e-07,
+ "loss": 0.2788818836212158,
+ "num_input_tokens_seen": 1061420544,
+ "step": 24275,
+ "train_runtime": 187795.6161,
+ "train_tokens_per_second": 5651.999
+ },
+ {
+ "epoch": 1.63479666038244,
+ "grad_norm": 0.7515197813453762,
+ "learning_rate": 8.006857551558328e-07,
+ "loss": 0.30008883476257325,
+ "num_input_tokens_seen": 1061636496,
+ "step": 24280,
+ "train_runtime": 187842.4419,
+ "train_tokens_per_second": 5651.739
+ },
+ {
+ "epoch": 1.6351333153784002,
+ "grad_norm": 0.8049426928966876,
+ "learning_rate": 7.992511381108997e-07,
+ "loss": 0.3101008415222168,
+ "num_input_tokens_seen": 1061878752,
+ "step": 24285,
+ "train_runtime": 187875.4149,
+ "train_tokens_per_second": 5652.037
+ },
+ {
+ "epoch": 1.6354699703743605,
+ "grad_norm": 0.7928126374708939,
+ "learning_rate": 7.978176957919909e-07,
+ "loss": 0.3050835132598877,
+ "num_input_tokens_seen": 1062088984,
+ "step": 24290,
+ "train_runtime": 187921.6004,
+ "train_tokens_per_second": 5651.766
+ },
+ {
+ "epoch": 1.6358066253703205,
+ "grad_norm": 0.660553687086565,
+ "learning_rate": 7.963854285999634e-07,
+ "loss": 0.27643623352050783,
+ "num_input_tokens_seen": 1062308496,
+ "step": 24295,
+ "train_runtime": 187966.5083,
+ "train_tokens_per_second": 5651.584
+ },
+ {
+ "epoch": 1.6361432803662805,
+ "grad_norm": 0.8664436436523667,
+ "learning_rate": 7.949543369353452e-07,
+ "loss": 0.30392279624938967,
+ "num_input_tokens_seen": 1062508472,
+ "step": 24300,
+ "train_runtime": 188008.6162,
+ "train_tokens_per_second": 5651.382
+ },
+ {
+ "epoch": 1.6364799353622408,
+ "grad_norm": 0.7681401573545563,
+ "learning_rate": 7.9352442119834e-07,
+ "loss": 0.3222590446472168,
+ "num_input_tokens_seen": 1062721224,
+ "step": 24305,
+ "train_runtime": 188049.5257,
+ "train_tokens_per_second": 5651.284
+ },
+ {
+ "epoch": 1.636816590358201,
+ "grad_norm": 0.7578044675754301,
+ "learning_rate": 7.920956817888148e-07,
+ "loss": 0.3079147577285767,
+ "num_input_tokens_seen": 1062938784,
+ "step": 24310,
+ "train_runtime": 188085.9323,
+ "train_tokens_per_second": 5651.347
+ },
+ {
+ "epoch": 1.637153245354161,
+ "grad_norm": 0.8771063708461605,
+ "learning_rate": 7.906681191063165e-07,
+ "loss": 0.2923039436340332,
+ "num_input_tokens_seen": 1063160648,
+ "step": 24315,
+ "train_runtime": 188122.7957,
+ "train_tokens_per_second": 5651.418
+ },
+ {
+ "epoch": 1.637489900350121,
+ "grad_norm": 0.7603147520393745,
+ "learning_rate": 7.892417335500541e-07,
+ "loss": 0.3094671964645386,
+ "num_input_tokens_seen": 1063371504,
+ "step": 24320,
+ "train_runtime": 188158.9471,
+ "train_tokens_per_second": 5651.453
+ },
+ {
+ "epoch": 1.6378265553460813,
+ "grad_norm": 0.828515228743802,
+ "learning_rate": 7.878165255189146e-07,
+ "loss": 0.32140753269195554,
+ "num_input_tokens_seen": 1063574440,
+ "step": 24325,
+ "train_runtime": 188196.1254,
+ "train_tokens_per_second": 5651.415
+ },
+ {
+ "epoch": 1.6381632103420416,
+ "grad_norm": 0.8506425251973367,
+ "learning_rate": 7.863924954114522e-07,
+ "loss": 0.28619048595428465,
+ "num_input_tokens_seen": 1063782584,
+ "step": 24330,
+ "train_runtime": 188232.3371,
+ "train_tokens_per_second": 5651.434
+ },
+ {
+ "epoch": 1.6384998653380016,
+ "grad_norm": 0.6750210682265397,
+ "learning_rate": 7.849696436258919e-07,
+ "loss": 0.2759582042694092,
+ "num_input_tokens_seen": 1063995392,
+ "step": 24335,
+ "train_runtime": 188272.2346,
+ "train_tokens_per_second": 5651.366
+ },
+ {
+ "epoch": 1.6388365203339617,
+ "grad_norm": 0.71118679601102,
+ "learning_rate": 7.835479705601307e-07,
+ "loss": 0.30858869552612306,
+ "num_input_tokens_seen": 1064201992,
+ "step": 24340,
+ "train_runtime": 188310.5847,
+ "train_tokens_per_second": 5651.313
+ },
+ {
+ "epoch": 1.639173175329922,
+ "grad_norm": 0.7067326301637101,
+ "learning_rate": 7.821274766117337e-07,
+ "loss": 0.3018312692642212,
+ "num_input_tokens_seen": 1064418504,
+ "step": 24345,
+ "train_runtime": 188354.2732,
+ "train_tokens_per_second": 5651.151
+ },
+ {
+ "epoch": 1.6395098303258822,
+ "grad_norm": 0.7701230824468515,
+ "learning_rate": 7.807081621779367e-07,
+ "loss": 0.3180276393890381,
+ "num_input_tokens_seen": 1064640776,
+ "step": 24350,
+ "train_runtime": 188396.5704,
+ "train_tokens_per_second": 5651.062
+ },
+ {
+ "epoch": 1.6398464853218422,
+ "grad_norm": 0.7453739839923318,
+ "learning_rate": 7.792900276556509e-07,
+ "loss": 0.30577759742736815,
+ "num_input_tokens_seen": 1064866144,
+ "step": 24355,
+ "train_runtime": 188437.6782,
+ "train_tokens_per_second": 5651.026
+ },
+ {
+ "epoch": 1.6401831403178022,
+ "grad_norm": 0.7840785820426818,
+ "learning_rate": 7.778730734414469e-07,
+ "loss": 0.3030556201934814,
+ "num_input_tokens_seen": 1065082080,
+ "step": 24360,
+ "train_runtime": 188478.6851,
+ "train_tokens_per_second": 5650.942
+ },
+ {
+ "epoch": 1.6405197953137625,
+ "grad_norm": 0.7537497153937553,
+ "learning_rate": 7.764572999315772e-07,
+ "loss": 0.2959816217422485,
+ "num_input_tokens_seen": 1065308184,
+ "step": 24365,
+ "train_runtime": 188514.0839,
+ "train_tokens_per_second": 5651.08
+ },
+ {
+ "epoch": 1.6408564503097227,
+ "grad_norm": 0.7520641802122896,
+ "learning_rate": 7.750427075219536e-07,
+ "loss": 0.3010838031768799,
+ "num_input_tokens_seen": 1065519384,
+ "step": 24370,
+ "train_runtime": 188554.9706,
+ "train_tokens_per_second": 5650.975
+ },
+ {
+ "epoch": 1.6411931053056827,
+ "grad_norm": 0.8548907262270895,
+ "learning_rate": 7.736292966081655e-07,
+ "loss": 0.28882603645324706,
+ "num_input_tokens_seen": 1065712464,
+ "step": 24375,
+ "train_runtime": 188592.4974,
+ "train_tokens_per_second": 5650.874
+ },
+ {
+ "epoch": 1.6415297603016428,
+ "grad_norm": 0.6533945003956135,
+ "learning_rate": 7.722170675854673e-07,
+ "loss": 0.26800405979156494,
+ "num_input_tokens_seen": 1065939064,
+ "step": 24380,
+ "train_runtime": 188629.38,
+ "train_tokens_per_second": 5650.971
+ },
+ {
+ "epoch": 1.641866415297603,
+ "grad_norm": 0.7624504132460348,
+ "learning_rate": 7.70806020848785e-07,
+ "loss": 0.2844860076904297,
+ "num_input_tokens_seen": 1066159000,
+ "step": 24385,
+ "train_runtime": 188660.9475,
+ "train_tokens_per_second": 5651.191
+ },
+ {
+ "epoch": 1.6422030702935633,
+ "grad_norm": 0.7515316499838955,
+ "learning_rate": 7.693961567927133e-07,
+ "loss": 0.2823920726776123,
+ "num_input_tokens_seen": 1066381848,
+ "step": 24390,
+ "train_runtime": 188700.2202,
+ "train_tokens_per_second": 5651.196
+ },
+ {
+ "epoch": 1.6425397252895233,
+ "grad_norm": 0.7063503273233152,
+ "learning_rate": 7.679874758115153e-07,
+ "loss": 0.2843328475952148,
+ "num_input_tokens_seen": 1066612256,
+ "step": 24395,
+ "train_runtime": 188738.2146,
+ "train_tokens_per_second": 5651.279
+ },
+ {
+ "epoch": 1.6428763802854833,
+ "grad_norm": 0.697327098375531,
+ "learning_rate": 7.66579978299124e-07,
+ "loss": 0.30366389751434325,
+ "num_input_tokens_seen": 1066840032,
+ "step": 24400,
+ "train_runtime": 188775.3375,
+ "train_tokens_per_second": 5651.374
+ },
+ {
+ "epoch": 1.6432130352814436,
+ "grad_norm": 1.0022326064476843,
+ "learning_rate": 7.651736646491447e-07,
+ "loss": 0.32686214447021483,
+ "num_input_tokens_seen": 1067042480,
+ "step": 24405,
+ "train_runtime": 188812.9262,
+ "train_tokens_per_second": 5651.321
+ },
+ {
+ "epoch": 1.6435496902774038,
+ "grad_norm": 0.623590025551479,
+ "learning_rate": 7.63768535254844e-07,
+ "loss": 0.27104051113128663,
+ "num_input_tokens_seen": 1067263472,
+ "step": 24410,
+ "train_runtime": 188849.6348,
+ "train_tokens_per_second": 5651.393
+ },
+ {
+ "epoch": 1.6438863452733639,
+ "grad_norm": 0.830785224698193,
+ "learning_rate": 7.623645905091664e-07,
+ "loss": 0.3036444902420044,
+ "num_input_tokens_seen": 1067483856,
+ "step": 24415,
+ "train_runtime": 188892.9648,
+ "train_tokens_per_second": 5651.263
+ },
+ {
+ "epoch": 1.6442230002693239,
+ "grad_norm": 0.8051008708219101,
+ "learning_rate": 7.609618308047156e-07,
+ "loss": 0.3053473472595215,
+ "num_input_tokens_seen": 1067700112,
+ "step": 24420,
+ "train_runtime": 188925.2763,
+ "train_tokens_per_second": 5651.441
+ },
+ {
+ "epoch": 1.6445596552652841,
+ "grad_norm": 0.7189604815408457,
+ "learning_rate": 7.595602565337729e-07,
+ "loss": 0.29787912368774416,
+ "num_input_tokens_seen": 1067928768,
+ "step": 24425,
+ "train_runtime": 188973.0073,
+ "train_tokens_per_second": 5651.224
+ },
+ {
+ "epoch": 1.6448963102612444,
+ "grad_norm": 0.8282739062629259,
+ "learning_rate": 7.581598680882829e-07,
+ "loss": 0.31963634490966797,
+ "num_input_tokens_seen": 1068150624,
+ "step": 24430,
+ "train_runtime": 189012.1223,
+ "train_tokens_per_second": 5651.228
+ },
+ {
+ "epoch": 1.6452329652572044,
+ "grad_norm": 0.8570504742841146,
+ "learning_rate": 7.567606658598592e-07,
+ "loss": 0.29708251953125,
+ "num_input_tokens_seen": 1068367600,
+ "step": 24435,
+ "train_runtime": 189050.0223,
+ "train_tokens_per_second": 5651.243
+ },
+ {
+ "epoch": 1.6455696202531644,
+ "grad_norm": 0.7274203559032685,
+ "learning_rate": 7.553626502397848e-07,
+ "loss": 0.3157409906387329,
+ "num_input_tokens_seen": 1068589480,
+ "step": 24440,
+ "train_runtime": 189094.3938,
+ "train_tokens_per_second": 5651.09
+ },
+ {
+ "epoch": 1.6459062752491247,
+ "grad_norm": 0.7307127250711936,
+ "learning_rate": 7.539658216190099e-07,
+ "loss": 0.2788280248641968,
+ "num_input_tokens_seen": 1068810256,
+ "step": 24445,
+ "train_runtime": 189133.9599,
+ "train_tokens_per_second": 5651.075
+ },
+ {
+ "epoch": 1.646242930245085,
+ "grad_norm": 0.8440106593735498,
+ "learning_rate": 7.525701803881525e-07,
+ "loss": 0.305745792388916,
+ "num_input_tokens_seen": 1069020600,
+ "step": 24450,
+ "train_runtime": 189174.0585,
+ "train_tokens_per_second": 5650.989
+ },
+ {
+ "epoch": 1.646579585241045,
+ "grad_norm": 0.9140651733358113,
+ "learning_rate": 7.511757269375019e-07,
+ "loss": 0.2875369548797607,
+ "num_input_tokens_seen": 1069228640,
+ "step": 24455,
+ "train_runtime": 189220.4481,
+ "train_tokens_per_second": 5650.703
+ },
+ {
+ "epoch": 1.646916240237005,
+ "grad_norm": 0.767638448585773,
+ "learning_rate": 7.497824616570087e-07,
+ "loss": 0.31629951000213624,
+ "num_input_tokens_seen": 1069428344,
+ "step": 24460,
+ "train_runtime": 189254.1834,
+ "train_tokens_per_second": 5650.751
+ },
+ {
+ "epoch": 1.6472528952329653,
+ "grad_norm": 0.7331009881607108,
+ "learning_rate": 7.483903849362995e-07,
+ "loss": 0.3073263168334961,
+ "num_input_tokens_seen": 1069649936,
+ "step": 24465,
+ "train_runtime": 189293.0368,
+ "train_tokens_per_second": 5650.762
+ },
+ {
+ "epoch": 1.6475895502289255,
+ "grad_norm": 0.7388708195492701,
+ "learning_rate": 7.469994971646594e-07,
+ "loss": 0.3005094289779663,
+ "num_input_tokens_seen": 1069880992,
+ "step": 24470,
+ "train_runtime": 189337.0814,
+ "train_tokens_per_second": 5650.668
+ },
+ {
+ "epoch": 1.6479262052248855,
+ "grad_norm": 0.7085554942280863,
+ "learning_rate": 7.456097987310495e-07,
+ "loss": 0.3060446262359619,
+ "num_input_tokens_seen": 1070091864,
+ "step": 24475,
+ "train_runtime": 189375.3021,
+ "train_tokens_per_second": 5650.641
+ },
+ {
+ "epoch": 1.6482628602208456,
+ "grad_norm": 0.9191804958073158,
+ "learning_rate": 7.442212900240931e-07,
+ "loss": 0.3100743770599365,
+ "num_input_tokens_seen": 1070303776,
+ "step": 24480,
+ "train_runtime": 189415.09,
+ "train_tokens_per_second": 5650.573
+ },
+ {
+ "epoch": 1.6485995152168058,
+ "grad_norm": 0.7298994648086058,
+ "learning_rate": 7.42833971432082e-07,
+ "loss": 0.2700855255126953,
+ "num_input_tokens_seen": 1070514192,
+ "step": 24485,
+ "train_runtime": 189447.1005,
+ "train_tokens_per_second": 5650.729
+ },
+ {
+ "epoch": 1.648936170212766,
+ "grad_norm": 0.7999546159714445,
+ "learning_rate": 7.414478433429756e-07,
+ "loss": 0.3064833641052246,
+ "num_input_tokens_seen": 1070735968,
+ "step": 24490,
+ "train_runtime": 189484.8803,
+ "train_tokens_per_second": 5650.773
+ },
+ {
+ "epoch": 1.649272825208726,
+ "grad_norm": 0.7881065041204545,
+ "learning_rate": 7.400629061444003e-07,
+ "loss": 0.29147264957427976,
+ "num_input_tokens_seen": 1070953008,
+ "step": 24495,
+ "train_runtime": 189528.9666,
+ "train_tokens_per_second": 5650.603
+ },
+ {
+ "epoch": 1.6496094802046861,
+ "grad_norm": 0.6964365331001868,
+ "learning_rate": 7.386791602236482e-07,
+ "loss": 0.32079496383666994,
+ "num_input_tokens_seen": 1071182864,
+ "step": 24500,
+ "train_runtime": 189563.7574,
+ "train_tokens_per_second": 5650.779
+ },
+ {
+ "epoch": 1.6499461352006464,
+ "grad_norm": 0.7757976985222431,
+ "learning_rate": 7.372966059676834e-07,
+ "loss": 0.34608256816864014,
+ "num_input_tokens_seen": 1071404416,
+ "step": 24505,
+ "train_runtime": 189604.7721,
+ "train_tokens_per_second": 5650.725
+ },
+ {
+ "epoch": 1.6502827901966066,
+ "grad_norm": 0.7009325679276259,
+ "learning_rate": 7.359152437631273e-07,
+ "loss": 0.31566901206970216,
+ "num_input_tokens_seen": 1071628384,
+ "step": 24510,
+ "train_runtime": 189642.0172,
+ "train_tokens_per_second": 5650.796
+ },
+ {
+ "epoch": 1.6506194451925666,
+ "grad_norm": 0.7512626090141364,
+ "learning_rate": 7.345350739962781e-07,
+ "loss": 0.29537224769592285,
+ "num_input_tokens_seen": 1071849024,
+ "step": 24515,
+ "train_runtime": 189682.4667,
+ "train_tokens_per_second": 5650.754
+ },
+ {
+ "epoch": 1.6509561001885267,
+ "grad_norm": 0.7652445412623243,
+ "learning_rate": 7.331560970530921e-07,
+ "loss": 0.29905951023101807,
+ "num_input_tokens_seen": 1072071032,
+ "step": 24520,
+ "train_runtime": 189724.0471,
+ "train_tokens_per_second": 5650.686
+ },
+ {
+ "epoch": 1.651292755184487,
+ "grad_norm": 0.6948758340555818,
+ "learning_rate": 7.317783133191986e-07,
+ "loss": 0.30941028594970704,
+ "num_input_tokens_seen": 1072293992,
+ "step": 24525,
+ "train_runtime": 189754.0392,
+ "train_tokens_per_second": 5650.968
+ },
+ {
+ "epoch": 1.6516294101804472,
+ "grad_norm": 0.7212176612484024,
+ "learning_rate": 7.304017231798893e-07,
+ "loss": 0.2879075050354004,
+ "num_input_tokens_seen": 1072519464,
+ "step": 24530,
+ "train_runtime": 189793.2385,
+ "train_tokens_per_second": 5650.989
+ },
+ {
+ "epoch": 1.6519660651764072,
+ "grad_norm": 0.8807710558741058,
+ "learning_rate": 7.290263270201231e-07,
+ "loss": 0.29528236389160156,
+ "num_input_tokens_seen": 1072738280,
+ "step": 24535,
+ "train_runtime": 189830.8758,
+ "train_tokens_per_second": 5651.021
+ },
+ {
+ "epoch": 1.6523027201723672,
+ "grad_norm": 0.7720732199868177,
+ "learning_rate": 7.276521252245261e-07,
+ "loss": 0.28215551376342773,
+ "num_input_tokens_seen": 1072948632,
+ "step": 24540,
+ "train_runtime": 189866.7872,
+ "train_tokens_per_second": 5651.06
+ },
+ {
+ "epoch": 1.6526393751683275,
+ "grad_norm": 0.8981368488565471,
+ "learning_rate": 7.262791181773882e-07,
+ "loss": 0.29338951110839845,
+ "num_input_tokens_seen": 1073174384,
+ "step": 24545,
+ "train_runtime": 189911.0819,
+ "train_tokens_per_second": 5650.931
+ },
+ {
+ "epoch": 1.6529760301642877,
+ "grad_norm": 0.8332745819366663,
+ "learning_rate": 7.24907306262666e-07,
+ "loss": 0.30844817161560056,
+ "num_input_tokens_seen": 1073386448,
+ "step": 24550,
+ "train_runtime": 189955.2277,
+ "train_tokens_per_second": 5650.734
+ },
+ {
+ "epoch": 1.6533126851602478,
+ "grad_norm": 0.7405815929469131,
+ "learning_rate": 7.235366898639856e-07,
+ "loss": 0.3130614995956421,
+ "num_input_tokens_seen": 1073610192,
+ "step": 24555,
+ "train_runtime": 189996.8678,
+ "train_tokens_per_second": 5650.673
+ },
+ {
+ "epoch": 1.6536493401562078,
+ "grad_norm": 0.7968518861817051,
+ "learning_rate": 7.221672693646309e-07,
+ "loss": 0.3168208122253418,
+ "num_input_tokens_seen": 1073834968,
+ "step": 24560,
+ "train_runtime": 190036.3942,
+ "train_tokens_per_second": 5650.681
+ },
+ {
+ "epoch": 1.653985995152168,
+ "grad_norm": 0.7509915696488056,
+ "learning_rate": 7.207990451475594e-07,
+ "loss": 0.30944645404815674,
+ "num_input_tokens_seen": 1074068032,
+ "step": 24565,
+ "train_runtime": 190069.3229,
+ "train_tokens_per_second": 5650.928
+ },
+ {
+ "epoch": 1.6543226501481283,
+ "grad_norm": 0.8218450305806543,
+ "learning_rate": 7.194320175953901e-07,
+ "loss": 0.3064124584197998,
+ "num_input_tokens_seen": 1074280488,
+ "step": 24570,
+ "train_runtime": 190114.2338,
+ "train_tokens_per_second": 5650.71
+ },
+ {
+ "epoch": 1.6546593051440883,
+ "grad_norm": 0.7421711436175688,
+ "learning_rate": 7.180661870904066e-07,
+ "loss": 0.285313081741333,
+ "num_input_tokens_seen": 1074502024,
+ "step": 24575,
+ "train_runtime": 190155.2205,
+ "train_tokens_per_second": 5650.658
+ },
+ {
+ "epoch": 1.6549959601400483,
+ "grad_norm": 0.7026987080451035,
+ "learning_rate": 7.167015540145599e-07,
+ "loss": 0.2826228618621826,
+ "num_input_tokens_seen": 1074710072,
+ "step": 24580,
+ "train_runtime": 190194.1052,
+ "train_tokens_per_second": 5650.596
+ },
+ {
+ "epoch": 1.6553326151360086,
+ "grad_norm": 0.7787998439821507,
+ "learning_rate": 7.153381187494657e-07,
+ "loss": 0.3077234745025635,
+ "num_input_tokens_seen": 1074926392,
+ "step": 24585,
+ "train_runtime": 190228.7504,
+ "train_tokens_per_second": 5650.704
+ },
+ {
+ "epoch": 1.6556692701319689,
+ "grad_norm": 0.7756508425427905,
+ "learning_rate": 7.139758816764036e-07,
+ "loss": 0.28848824501037595,
+ "num_input_tokens_seen": 1075131776,
+ "step": 24590,
+ "train_runtime": 190266.2138,
+ "train_tokens_per_second": 5650.671
+ },
+ {
+ "epoch": 1.6560059251279289,
+ "grad_norm": 0.6980517503857505,
+ "learning_rate": 7.126148431763191e-07,
+ "loss": 0.28489990234375,
+ "num_input_tokens_seen": 1075362760,
+ "step": 24595,
+ "train_runtime": 190301.6247,
+ "train_tokens_per_second": 5650.833
+ },
+ {
+ "epoch": 1.656342580123889,
+ "grad_norm": 0.7359183935227057,
+ "learning_rate": 7.11255003629821e-07,
+ "loss": 0.30712342262268066,
+ "num_input_tokens_seen": 1075591552,
+ "step": 24600,
+ "train_runtime": 190339.4831,
+ "train_tokens_per_second": 5650.911
+ },
+ {
+ "epoch": 1.6566792351198492,
+ "grad_norm": 0.7209263296219092,
+ "learning_rate": 7.098963634171874e-07,
+ "loss": 0.3017916679382324,
+ "num_input_tokens_seen": 1075828544,
+ "step": 24605,
+ "train_runtime": 190376.2279,
+ "train_tokens_per_second": 5651.066
+ },
+ {
+ "epoch": 1.6570158901158094,
+ "grad_norm": 0.8863531557142312,
+ "learning_rate": 7.085389229183537e-07,
+ "loss": 0.3059419631958008,
+ "num_input_tokens_seen": 1076024056,
+ "step": 24610,
+ "train_runtime": 190418.8793,
+ "train_tokens_per_second": 5650.827
+ },
+ {
+ "epoch": 1.6573525451117694,
+ "grad_norm": 0.812027087160403,
+ "learning_rate": 7.071826825129264e-07,
+ "loss": 0.28927106857299806,
+ "num_input_tokens_seen": 1076235696,
+ "step": 24615,
+ "train_runtime": 190461.2181,
+ "train_tokens_per_second": 5650.682
+ },
+ {
+ "epoch": 1.6576892001077295,
+ "grad_norm": 0.7711258281303948,
+ "learning_rate": 7.058276425801735e-07,
+ "loss": 0.30096821784973143,
+ "num_input_tokens_seen": 1076456872,
+ "step": 24620,
+ "train_runtime": 190502.0918,
+ "train_tokens_per_second": 5650.63
+ },
+ {
+ "epoch": 1.6580258551036897,
+ "grad_norm": 0.7870890841283721,
+ "learning_rate": 7.044738034990273e-07,
+ "loss": 0.29817709922790525,
+ "num_input_tokens_seen": 1076662016,
+ "step": 24625,
+ "train_runtime": 190533.526,
+ "train_tokens_per_second": 5650.775
+ },
+ {
+ "epoch": 1.65836251009965,
+ "grad_norm": 0.7826364707728534,
+ "learning_rate": 7.031211656480841e-07,
+ "loss": 0.3162134885787964,
+ "num_input_tokens_seen": 1076875184,
+ "step": 24630,
+ "train_runtime": 190573.3113,
+ "train_tokens_per_second": 5650.714
+ },
+ {
+ "epoch": 1.65869916509561,
+ "grad_norm": 0.7671703072883405,
+ "learning_rate": 7.017697294056058e-07,
+ "loss": 0.27853074073791506,
+ "num_input_tokens_seen": 1077089776,
+ "step": 24635,
+ "train_runtime": 190615.0578,
+ "train_tokens_per_second": 5650.602
+ },
+ {
+ "epoch": 1.65903582009157,
+ "grad_norm": 0.8067540848795106,
+ "learning_rate": 7.004194951495163e-07,
+ "loss": 0.3217106819152832,
+ "num_input_tokens_seen": 1077318536,
+ "step": 24640,
+ "train_runtime": 190649.8372,
+ "train_tokens_per_second": 5650.771
+ },
+ {
+ "epoch": 1.6593724750875303,
+ "grad_norm": 0.7719725402304747,
+ "learning_rate": 6.990704632574046e-07,
+ "loss": 0.29686970710754396,
+ "num_input_tokens_seen": 1077547440,
+ "step": 24645,
+ "train_runtime": 190685.7981,
+ "train_tokens_per_second": 5650.906
+ },
+ {
+ "epoch": 1.6597091300834905,
+ "grad_norm": 0.8088510895933726,
+ "learning_rate": 6.977226341065219e-07,
+ "loss": 0.299147891998291,
+ "num_input_tokens_seen": 1077764672,
+ "step": 24650,
+ "train_runtime": 190723.2912,
+ "train_tokens_per_second": 5650.934
+ },
+ {
+ "epoch": 1.6600457850794506,
+ "grad_norm": 0.7515417585035541,
+ "learning_rate": 6.96376008073788e-07,
+ "loss": 0.3067599296569824,
+ "num_input_tokens_seen": 1077969096,
+ "step": 24655,
+ "train_runtime": 190758.1707,
+ "train_tokens_per_second": 5650.972
+ },
+ {
+ "epoch": 1.6603824400754106,
+ "grad_norm": 0.7301895968368473,
+ "learning_rate": 6.950305855357781e-07,
+ "loss": 0.27429049015045165,
+ "num_input_tokens_seen": 1078181392,
+ "step": 24660,
+ "train_runtime": 190792.7332,
+ "train_tokens_per_second": 5651.061
+ },
+ {
+ "epoch": 1.6607190950713708,
+ "grad_norm": 0.7577040604392636,
+ "learning_rate": 6.936863668687383e-07,
+ "loss": 0.30942230224609374,
+ "num_input_tokens_seen": 1078397824,
+ "step": 24665,
+ "train_runtime": 190830.5609,
+ "train_tokens_per_second": 5651.075
+ },
+ {
+ "epoch": 1.661055750067331,
+ "grad_norm": 0.7766234601269146,
+ "learning_rate": 6.923433524485745e-07,
+ "loss": 0.323834753036499,
+ "num_input_tokens_seen": 1078622016,
+ "step": 24670,
+ "train_runtime": 190872.8273,
+ "train_tokens_per_second": 5650.998
+ },
+ {
+ "epoch": 1.6613924050632911,
+ "grad_norm": 0.7623394697980529,
+ "learning_rate": 6.910015426508554e-07,
+ "loss": 0.296146297454834,
+ "num_input_tokens_seen": 1078847744,
+ "step": 24675,
+ "train_runtime": 190912.4664,
+ "train_tokens_per_second": 5651.007
+ },
+ {
+ "epoch": 1.6617290600592511,
+ "grad_norm": 0.7664054805943663,
+ "learning_rate": 6.896609378508152e-07,
+ "loss": 0.309325647354126,
+ "num_input_tokens_seen": 1079050512,
+ "step": 24680,
+ "train_runtime": 190950.4474,
+ "train_tokens_per_second": 5650.945
+ },
+ {
+ "epoch": 1.6620657150552114,
+ "grad_norm": 0.7252601816921024,
+ "learning_rate": 6.883215384233488e-07,
+ "loss": 0.2978062629699707,
+ "num_input_tokens_seen": 1079282736,
+ "step": 24685,
+ "train_runtime": 190991.7978,
+ "train_tokens_per_second": 5650.938
+ },
+ {
+ "epoch": 1.6624023700511716,
+ "grad_norm": 0.8104244979601277,
+ "learning_rate": 6.869833447430152e-07,
+ "loss": 0.28819808959960935,
+ "num_input_tokens_seen": 1079488008,
+ "step": 24690,
+ "train_runtime": 191037.7882,
+ "train_tokens_per_second": 5650.652
+ },
+ {
+ "epoch": 1.6627390250471317,
+ "grad_norm": 0.7648578275299457,
+ "learning_rate": 6.856463571840361e-07,
+ "loss": 0.2954545021057129,
+ "num_input_tokens_seen": 1079693528,
+ "step": 24695,
+ "train_runtime": 191074.9276,
+ "train_tokens_per_second": 5650.629
+ },
+ {
+ "epoch": 1.6630756800430917,
+ "grad_norm": 0.7776160702348354,
+ "learning_rate": 6.843105761202945e-07,
+ "loss": 0.3160905599594116,
+ "num_input_tokens_seen": 1079922112,
+ "step": 24700,
+ "train_runtime": 191116.2384,
+ "train_tokens_per_second": 5650.604
+ },
+ {
+ "epoch": 1.663412335039052,
+ "grad_norm": 0.795996876323385,
+ "learning_rate": 6.829760019253412e-07,
+ "loss": 0.3114809513092041,
+ "num_input_tokens_seen": 1080138072,
+ "step": 24705,
+ "train_runtime": 191153.6294,
+ "train_tokens_per_second": 5650.628
+ },
+ {
+ "epoch": 1.6637489900350122,
+ "grad_norm": 0.7322911069723348,
+ "learning_rate": 6.816426349723804e-07,
+ "loss": 0.30275850296020507,
+ "num_input_tokens_seen": 1080349800,
+ "step": 24710,
+ "train_runtime": 191191.3873,
+ "train_tokens_per_second": 5650.62
+ },
+ {
+ "epoch": 1.6640856450309722,
+ "grad_norm": 0.6721252860586514,
+ "learning_rate": 6.803104756342877e-07,
+ "loss": 0.2700210094451904,
+ "num_input_tokens_seen": 1080576104,
+ "step": 24715,
+ "train_runtime": 191227.8925,
+ "train_tokens_per_second": 5650.724
+ },
+ {
+ "epoch": 1.6644223000269323,
+ "grad_norm": 0.7184313716768772,
+ "learning_rate": 6.789795242835962e-07,
+ "loss": 0.26948862075805663,
+ "num_input_tokens_seen": 1080793368,
+ "step": 24720,
+ "train_runtime": 191269.297,
+ "train_tokens_per_second": 5650.637
+ },
+ {
+ "epoch": 1.6647589550228925,
+ "grad_norm": 0.7596106189238437,
+ "learning_rate": 6.776497812925015e-07,
+ "loss": 0.3047426700592041,
+ "num_input_tokens_seen": 1081001160,
+ "step": 24725,
+ "train_runtime": 191307.1705,
+ "train_tokens_per_second": 5650.605
+ },
+ {
+ "epoch": 1.6650956100188528,
+ "grad_norm": 0.8078999876649536,
+ "learning_rate": 6.763212470328628e-07,
+ "loss": 0.3334406614303589,
+ "num_input_tokens_seen": 1081221944,
+ "step": 24730,
+ "train_runtime": 191348.8888,
+ "train_tokens_per_second": 5650.526
+ },
+ {
+ "epoch": 1.6654322650148128,
+ "grad_norm": 0.7785010647558146,
+ "learning_rate": 6.749939218762003e-07,
+ "loss": 0.2895099878311157,
+ "num_input_tokens_seen": 1081439048,
+ "step": 24735,
+ "train_runtime": 191390.3574,
+ "train_tokens_per_second": 5650.436
+ },
+ {
+ "epoch": 1.6657689200107728,
+ "grad_norm": 0.7473396001618127,
+ "learning_rate": 6.736678061936958e-07,
+ "loss": 0.3205715656280518,
+ "num_input_tokens_seen": 1081670648,
+ "step": 24740,
+ "train_runtime": 191432.0544,
+ "train_tokens_per_second": 5650.416
+ },
+ {
+ "epoch": 1.666105575006733,
+ "grad_norm": 0.8395952302334243,
+ "learning_rate": 6.723429003561937e-07,
+ "loss": 0.325761604309082,
+ "num_input_tokens_seen": 1081907840,
+ "step": 24745,
+ "train_runtime": 191466.0452,
+ "train_tokens_per_second": 5650.651
+ },
+ {
+ "epoch": 1.6664422300026933,
+ "grad_norm": 0.7553522775373878,
+ "learning_rate": 6.710192047341984e-07,
+ "loss": 0.2971907615661621,
+ "num_input_tokens_seen": 1082129744,
+ "step": 24750,
+ "train_runtime": 191505.5101,
+ "train_tokens_per_second": 5650.645
+ },
+ {
+ "epoch": 1.6667788849986533,
+ "grad_norm": 0.7670068129458077,
+ "learning_rate": 6.69696719697881e-07,
+ "loss": 0.29376702308654784,
+ "num_input_tokens_seen": 1082352688,
+ "step": 24755,
+ "train_runtime": 191544.0035,
+ "train_tokens_per_second": 5650.674
+ },
+ {
+ "epoch": 1.6671155399946134,
+ "grad_norm": 0.7937246542094001,
+ "learning_rate": 6.683754456170655e-07,
+ "loss": 0.27995355129241944,
+ "num_input_tokens_seen": 1082568000,
+ "step": 24760,
+ "train_runtime": 191576.7889,
+ "train_tokens_per_second": 5650.831
+ },
+ {
+ "epoch": 1.6674521949905736,
+ "grad_norm": 0.7863716640239266,
+ "learning_rate": 6.670553828612448e-07,
+ "loss": 0.28558523654937745,
+ "num_input_tokens_seen": 1082762664,
+ "step": 24765,
+ "train_runtime": 191615.9134,
+ "train_tokens_per_second": 5650.693
+ },
+ {
+ "epoch": 1.6677888499865339,
+ "grad_norm": 0.6755324668287188,
+ "learning_rate": 6.657365317995696e-07,
+ "loss": 0.2899418592453003,
+ "num_input_tokens_seen": 1082980808,
+ "step": 24770,
+ "train_runtime": 191656.4559,
+ "train_tokens_per_second": 5650.636
+ },
+ {
+ "epoch": 1.668125504982494,
+ "grad_norm": 0.7089937331683779,
+ "learning_rate": 6.644188928008521e-07,
+ "loss": 0.31790828704833984,
+ "num_input_tokens_seen": 1083191992,
+ "step": 24775,
+ "train_runtime": 191696.5978,
+ "train_tokens_per_second": 5650.554
+ },
+ {
+ "epoch": 1.668462159978454,
+ "grad_norm": 0.8389878605630636,
+ "learning_rate": 6.631024662335667e-07,
+ "loss": 0.31216111183166506,
+ "num_input_tokens_seen": 1083409256,
+ "step": 24780,
+ "train_runtime": 191745.9321,
+ "train_tokens_per_second": 5650.233
+ },
+ {
+ "epoch": 1.6687988149744142,
+ "grad_norm": 0.7518345608186,
+ "learning_rate": 6.617872524658469e-07,
+ "loss": 0.29103267192840576,
+ "num_input_tokens_seen": 1083638560,
+ "step": 24785,
+ "train_runtime": 191781.1772,
+ "train_tokens_per_second": 5650.391
+ },
+ {
+ "epoch": 1.6691354699703744,
+ "grad_norm": 0.8193494154451724,
+ "learning_rate": 6.604732518654882e-07,
+ "loss": 0.3304905414581299,
+ "num_input_tokens_seen": 1083853344,
+ "step": 24790,
+ "train_runtime": 191823.5191,
+ "train_tokens_per_second": 5650.263
+ },
+ {
+ "epoch": 1.6694721249663345,
+ "grad_norm": 0.8968284580615762,
+ "learning_rate": 6.591604647999473e-07,
+ "loss": 0.32892765998840334,
+ "num_input_tokens_seen": 1084083632,
+ "step": 24795,
+ "train_runtime": 191860.6602,
+ "train_tokens_per_second": 5650.37
+ },
+ {
+ "epoch": 1.6698087799622945,
+ "grad_norm": 0.6910607916103046,
+ "learning_rate": 6.57848891636339e-07,
+ "loss": 0.2843611478805542,
+ "num_input_tokens_seen": 1084309072,
+ "step": 24800,
+ "train_runtime": 191903.5578,
+ "train_tokens_per_second": 5650.281
+ },
+ {
+ "epoch": 1.6701454349582547,
+ "grad_norm": 0.7582706445392589,
+ "learning_rate": 6.565385327414441e-07,
+ "loss": 0.30639631748199464,
+ "num_input_tokens_seen": 1084535760,
+ "step": 24805,
+ "train_runtime": 191937.737,
+ "train_tokens_per_second": 5650.456
+ },
+ {
+ "epoch": 1.670482089954215,
+ "grad_norm": 0.8230500682984503,
+ "learning_rate": 6.552293884816968e-07,
+ "loss": 0.2900113582611084,
+ "num_input_tokens_seen": 1084761472,
+ "step": 24810,
+ "train_runtime": 191977.452,
+ "train_tokens_per_second": 5650.463
+ },
+ {
+ "epoch": 1.670818744950175,
+ "grad_norm": 0.6579326078890422,
+ "learning_rate": 6.539214592231968e-07,
+ "loss": 0.29627652168273927,
+ "num_input_tokens_seen": 1084987552,
+ "step": 24815,
+ "train_runtime": 192018.4764,
+ "train_tokens_per_second": 5650.433
+ },
+ {
+ "epoch": 1.671155399946135,
+ "grad_norm": 0.7194038569880106,
+ "learning_rate": 6.526147453317028e-07,
+ "loss": 0.26366333961486815,
+ "num_input_tokens_seen": 1085201944,
+ "step": 24820,
+ "train_runtime": 192058.3704,
+ "train_tokens_per_second": 5650.376
+ },
+ {
+ "epoch": 1.6714920549420953,
+ "grad_norm": 0.7597781908310334,
+ "learning_rate": 6.513092471726307e-07,
+ "loss": 0.299773645401001,
+ "num_input_tokens_seen": 1085434936,
+ "step": 24825,
+ "train_runtime": 192099.0636,
+ "train_tokens_per_second": 5650.392
+ },
+ {
+ "epoch": 1.6718287099380555,
+ "grad_norm": 0.7278296909083918,
+ "learning_rate": 6.500049651110635e-07,
+ "loss": 0.2928581714630127,
+ "num_input_tokens_seen": 1085656760,
+ "step": 24830,
+ "train_runtime": 192139.5855,
+ "train_tokens_per_second": 5650.354
+ },
+ {
+ "epoch": 1.6721653649340156,
+ "grad_norm": 0.7587076551124298,
+ "learning_rate": 6.487018995117339e-07,
+ "loss": 0.2924539089202881,
+ "num_input_tokens_seen": 1085894200,
+ "step": 24835,
+ "train_runtime": 192174.6364,
+ "train_tokens_per_second": 5650.559
+ },
+ {
+ "epoch": 1.6725020199299756,
+ "grad_norm": 0.7823318054145896,
+ "learning_rate": 6.474000507390455e-07,
+ "loss": 0.2911822319030762,
+ "num_input_tokens_seen": 1086112280,
+ "step": 24840,
+ "train_runtime": 192208.7768,
+ "train_tokens_per_second": 5650.69
+ },
+ {
+ "epoch": 1.6728386749259359,
+ "grad_norm": 0.7008110558611228,
+ "learning_rate": 6.460994191570513e-07,
+ "loss": 0.3034616470336914,
+ "num_input_tokens_seen": 1086334184,
+ "step": 24845,
+ "train_runtime": 192246.1123,
+ "train_tokens_per_second": 5650.747
+ },
+ {
+ "epoch": 1.673175329921896,
+ "grad_norm": 0.7292792811187331,
+ "learning_rate": 6.448000051294717e-07,
+ "loss": 0.2885445594787598,
+ "num_input_tokens_seen": 1086553800,
+ "step": 24850,
+ "train_runtime": 192284.9452,
+ "train_tokens_per_second": 5650.748
+ },
+ {
+ "epoch": 1.6735119849178561,
+ "grad_norm": 0.6923122611597387,
+ "learning_rate": 6.435018090196832e-07,
+ "loss": 0.30068202018737794,
+ "num_input_tokens_seen": 1086783696,
+ "step": 24855,
+ "train_runtime": 192321.9072,
+ "train_tokens_per_second": 5650.858
+ },
+ {
+ "epoch": 1.6738486399138162,
+ "grad_norm": 1.2221971616080631,
+ "learning_rate": 6.42204831190722e-07,
+ "loss": 0.2921877861022949,
+ "num_input_tokens_seen": 1087006792,
+ "step": 24860,
+ "train_runtime": 192356.0704,
+ "train_tokens_per_second": 5651.014
+ },
+ {
+ "epoch": 1.6741852949097764,
+ "grad_norm": 0.7759288823452887,
+ "learning_rate": 6.409090720052835e-07,
+ "loss": 0.30269834995269773,
+ "num_input_tokens_seen": 1087221944,
+ "step": 24865,
+ "train_runtime": 192396.3099,
+ "train_tokens_per_second": 5650.95
+ },
+ {
+ "epoch": 1.6745219499057367,
+ "grad_norm": 0.7489484593312903,
+ "learning_rate": 6.396145318257229e-07,
+ "loss": 0.27388882637023926,
+ "num_input_tokens_seen": 1087443824,
+ "step": 24870,
+ "train_runtime": 192433.614,
+ "train_tokens_per_second": 5651.008
+ },
+ {
+ "epoch": 1.6748586049016967,
+ "grad_norm": 0.7471747560717638,
+ "learning_rate": 6.383212110140535e-07,
+ "loss": 0.31777362823486327,
+ "num_input_tokens_seen": 1087659568,
+ "step": 24875,
+ "train_runtime": 192469.464,
+ "train_tokens_per_second": 5651.076
+ },
+ {
+ "epoch": 1.6751952598976567,
+ "grad_norm": 0.9939208015357779,
+ "learning_rate": 6.370291099319509e-07,
+ "loss": 0.3115004301071167,
+ "num_input_tokens_seen": 1087852096,
+ "step": 24880,
+ "train_runtime": 192505.3832,
+ "train_tokens_per_second": 5651.022
+ },
+ {
+ "epoch": 1.675531914893617,
+ "grad_norm": 0.7589414568505637,
+ "learning_rate": 6.357382289407433e-07,
+ "loss": 0.29903695583343504,
+ "num_input_tokens_seen": 1088082304,
+ "step": 24885,
+ "train_runtime": 192540.2873,
+ "train_tokens_per_second": 5651.193
+ },
+ {
+ "epoch": 1.6758685698895772,
+ "grad_norm": 0.7522829988493125,
+ "learning_rate": 6.344485684014251e-07,
+ "loss": 0.3122910499572754,
+ "num_input_tokens_seen": 1088296296,
+ "step": 24890,
+ "train_runtime": 192578.9655,
+ "train_tokens_per_second": 5651.169
+ },
+ {
+ "epoch": 1.6762052248855372,
+ "grad_norm": 0.7718700735790412,
+ "learning_rate": 6.33160128674642e-07,
+ "loss": 0.31540684700012206,
+ "num_input_tokens_seen": 1088501864,
+ "step": 24895,
+ "train_runtime": 192615.6028,
+ "train_tokens_per_second": 5651.161
+ },
+ {
+ "epoch": 1.6765418798814973,
+ "grad_norm": 0.7172203437810806,
+ "learning_rate": 6.318729101207055e-07,
+ "loss": 0.2925147533416748,
+ "num_input_tokens_seen": 1088727288,
+ "step": 24900,
+ "train_runtime": 192657.5288,
+ "train_tokens_per_second": 5651.102
+ },
+ {
+ "epoch": 1.6768785348774575,
+ "grad_norm": 0.6793461071435637,
+ "learning_rate": 6.305869130995807e-07,
+ "loss": 0.30594768524169924,
+ "num_input_tokens_seen": 1088947848,
+ "step": 24905,
+ "train_runtime": 192692.2858,
+ "train_tokens_per_second": 5651.227
+ },
+ {
+ "epoch": 1.6772151898734178,
+ "grad_norm": 0.8065807442493926,
+ "learning_rate": 6.293021379708924e-07,
+ "loss": 0.30313916206359864,
+ "num_input_tokens_seen": 1089171240,
+ "step": 24910,
+ "train_runtime": 192732.2584,
+ "train_tokens_per_second": 5651.214
+ },
+ {
+ "epoch": 1.6775518448693778,
+ "grad_norm": 0.812051527961556,
+ "learning_rate": 6.280185850939241e-07,
+ "loss": 0.31810295581817627,
+ "num_input_tokens_seen": 1089387216,
+ "step": 24915,
+ "train_runtime": 192774.0539,
+ "train_tokens_per_second": 5651.109
+ },
+ {
+ "epoch": 1.677888499865338,
+ "grad_norm": 0.8171798449848718,
+ "learning_rate": 6.267362548276173e-07,
+ "loss": 0.30693349838256834,
+ "num_input_tokens_seen": 1089597552,
+ "step": 24920,
+ "train_runtime": 192811.4332,
+ "train_tokens_per_second": 5651.104
+ },
+ {
+ "epoch": 1.6782251548612983,
+ "grad_norm": 0.9137416038709986,
+ "learning_rate": 6.254551475305703e-07,
+ "loss": 0.30268676280975343,
+ "num_input_tokens_seen": 1089790616,
+ "step": 24925,
+ "train_runtime": 192849.8209,
+ "train_tokens_per_second": 5650.981
+ },
+ {
+ "epoch": 1.6785618098572583,
+ "grad_norm": 0.80355688108705,
+ "learning_rate": 6.241752635610443e-07,
+ "loss": 0.31978979110717776,
+ "num_input_tokens_seen": 1090007856,
+ "step": 24930,
+ "train_runtime": 192886.1002,
+ "train_tokens_per_second": 5651.044
+ },
+ {
+ "epoch": 1.6788984648532184,
+ "grad_norm": 0.9368900433932702,
+ "learning_rate": 6.228966032769496e-07,
+ "loss": 0.31137659549713137,
+ "num_input_tokens_seen": 1090222152,
+ "step": 24935,
+ "train_runtime": 192922.3211,
+ "train_tokens_per_second": 5651.094
+ },
+ {
+ "epoch": 1.6792351198491786,
+ "grad_norm": 0.8015786582216125,
+ "learning_rate": 6.216191670358651e-07,
+ "loss": 0.29123806953430176,
+ "num_input_tokens_seen": 1090437952,
+ "step": 24940,
+ "train_runtime": 192965.473,
+ "train_tokens_per_second": 5650.949
+ },
+ {
+ "epoch": 1.6795717748451389,
+ "grad_norm": 0.7556404800034242,
+ "learning_rate": 6.203429551950163e-07,
+ "loss": 0.32757062911987306,
+ "num_input_tokens_seen": 1090662752,
+ "step": 24945,
+ "train_runtime": 193011.4944,
+ "train_tokens_per_second": 5650.766
+ },
+ {
+ "epoch": 1.679908429841099,
+ "grad_norm": 0.8290622091556459,
+ "learning_rate": 6.19067968111296e-07,
+ "loss": 0.28361237049102783,
+ "num_input_tokens_seen": 1090869792,
+ "step": 24950,
+ "train_runtime": 193048.7244,
+ "train_tokens_per_second": 5650.749
+ },
+ {
+ "epoch": 1.680245084837059,
+ "grad_norm": 0.7116384964210178,
+ "learning_rate": 6.177942061412478e-07,
+ "loss": 0.2768688678741455,
+ "num_input_tokens_seen": 1091083944,
+ "step": 24955,
+ "train_runtime": 193088.2417,
+ "train_tokens_per_second": 5650.701
+ },
+ {
+ "epoch": 1.6805817398330192,
+ "grad_norm": 0.7009278311455667,
+ "learning_rate": 6.16521669641077e-07,
+ "loss": 0.2864533424377441,
+ "num_input_tokens_seen": 1091313248,
+ "step": 24960,
+ "train_runtime": 193120.829,
+ "train_tokens_per_second": 5650.935
+ },
+ {
+ "epoch": 1.6809183948289794,
+ "grad_norm": 0.7261926667581532,
+ "learning_rate": 6.152503589666426e-07,
+ "loss": 0.30624988079071047,
+ "num_input_tokens_seen": 1091544272,
+ "step": 24965,
+ "train_runtime": 193157.6847,
+ "train_tokens_per_second": 5651.053
+ },
+ {
+ "epoch": 1.6812550498249395,
+ "grad_norm": 0.7142592236914898,
+ "learning_rate": 6.139802744734636e-07,
+ "loss": 0.26999435424804685,
+ "num_input_tokens_seen": 1091761648,
+ "step": 24970,
+ "train_runtime": 193193.4372,
+ "train_tokens_per_second": 5651.132
+ },
+ {
+ "epoch": 1.6815917048208995,
+ "grad_norm": 0.8017434808157823,
+ "learning_rate": 6.127114165167142e-07,
+ "loss": 0.26949467658996584,
+ "num_input_tokens_seen": 1091969864,
+ "step": 24975,
+ "train_runtime": 193228.9139,
+ "train_tokens_per_second": 5651.172
+ },
+ {
+ "epoch": 1.6819283598168597,
+ "grad_norm": 0.8597502504635929,
+ "learning_rate": 6.11443785451229e-07,
+ "loss": 0.30141243934631345,
+ "num_input_tokens_seen": 1092190952,
+ "step": 24980,
+ "train_runtime": 193269.634,
+ "train_tokens_per_second": 5651.125
+ },
+ {
+ "epoch": 1.68226501481282,
+ "grad_norm": 0.7558509297387581,
+ "learning_rate": 6.101773816314926e-07,
+ "loss": 0.29298968315124513,
+ "num_input_tokens_seen": 1092417104,
+ "step": 24985,
+ "train_runtime": 193306.8721,
+ "train_tokens_per_second": 5651.207
+ },
+ {
+ "epoch": 1.68260166980878,
+ "grad_norm": 0.7852414890479306,
+ "learning_rate": 6.089122054116558e-07,
+ "loss": 0.3086401462554932,
+ "num_input_tokens_seen": 1092645416,
+ "step": 24990,
+ "train_runtime": 193342.0721,
+ "train_tokens_per_second": 5651.359
+ },
+ {
+ "epoch": 1.68293832480474,
+ "grad_norm": 0.8486422869311524,
+ "learning_rate": 6.076482571455161e-07,
+ "loss": 0.2916858196258545,
+ "num_input_tokens_seen": 1092868304,
+ "step": 24995,
+ "train_runtime": 193383.0382,
+ "train_tokens_per_second": 5651.314
+ },
+ {
+ "epoch": 1.6832749798007003,
+ "grad_norm": 0.8736609060726518,
+ "learning_rate": 6.063855371865369e-07,
+ "loss": 0.30973148345947266,
+ "num_input_tokens_seen": 1093073112,
+ "step": 25000,
+ "train_runtime": 193420.2455,
+ "train_tokens_per_second": 5651.286
+ },
+ {
+ "epoch": 1.6836116347966605,
+ "grad_norm": 0.8270587506153307,
+ "learning_rate": 6.051240458878316e-07,
+ "loss": 0.3318582057952881,
+ "num_input_tokens_seen": 1093292280,
+ "step": 25005,
+ "train_runtime": 193460.4002,
+ "train_tokens_per_second": 5651.246
+ },
+ {
+ "epoch": 1.6839482897926206,
+ "grad_norm": 0.7861426638271823,
+ "learning_rate": 6.038637836021721e-07,
+ "loss": 0.30891571044921873,
+ "num_input_tokens_seen": 1093512488,
+ "step": 25010,
+ "train_runtime": 193500.033,
+ "train_tokens_per_second": 5651.226
+ },
+ {
+ "epoch": 1.6842849447885806,
+ "grad_norm": 0.7570592424235427,
+ "learning_rate": 6.026047506819882e-07,
+ "loss": 0.29696998596191404,
+ "num_input_tokens_seen": 1093724280,
+ "step": 25015,
+ "train_runtime": 193535.9844,
+ "train_tokens_per_second": 5651.271
+ },
+ {
+ "epoch": 1.6846215997845408,
+ "grad_norm": 0.8023108005761309,
+ "learning_rate": 6.013469474793632e-07,
+ "loss": 0.2949728012084961,
+ "num_input_tokens_seen": 1093945480,
+ "step": 25020,
+ "train_runtime": 193575.5488,
+ "train_tokens_per_second": 5651.259
+ },
+ {
+ "epoch": 1.684958254780501,
+ "grad_norm": 0.792969947232272,
+ "learning_rate": 6.000903743460368e-07,
+ "loss": 0.2946459770202637,
+ "num_input_tokens_seen": 1094164496,
+ "step": 25025,
+ "train_runtime": 193619.1001,
+ "train_tokens_per_second": 5651.119
+ },
+ {
+ "epoch": 1.6852949097764611,
+ "grad_norm": 0.7249873667459386,
+ "learning_rate": 5.98835031633409e-07,
+ "loss": 0.296980881690979,
+ "num_input_tokens_seen": 1094387696,
+ "step": 25030,
+ "train_runtime": 193658.0253,
+ "train_tokens_per_second": 5651.135
+ },
+ {
+ "epoch": 1.6856315647724212,
+ "grad_norm": 0.7187877526122118,
+ "learning_rate": 5.97580919692528e-07,
+ "loss": 0.30903480052947996,
+ "num_input_tokens_seen": 1094621216,
+ "step": 25035,
+ "train_runtime": 193690.3611,
+ "train_tokens_per_second": 5651.397
+ },
+ {
+ "epoch": 1.6859682197683814,
+ "grad_norm": 0.7738587622533499,
+ "learning_rate": 5.963280388741072e-07,
+ "loss": 0.2778623104095459,
+ "num_input_tokens_seen": 1094843312,
+ "step": 25040,
+ "train_runtime": 193726.0711,
+ "train_tokens_per_second": 5651.502
+ },
+ {
+ "epoch": 1.6863048747643417,
+ "grad_norm": 0.8921293625641102,
+ "learning_rate": 5.950763895285055e-07,
+ "loss": 0.3040936946868896,
+ "num_input_tokens_seen": 1095048760,
+ "step": 25045,
+ "train_runtime": 193766.9375,
+ "train_tokens_per_second": 5651.371
+ },
+ {
+ "epoch": 1.6866415297603017,
+ "grad_norm": 0.6626481867816244,
+ "learning_rate": 5.938259720057465e-07,
+ "loss": 0.29845240116119387,
+ "num_input_tokens_seen": 1095275144,
+ "step": 25050,
+ "train_runtime": 193803.6462,
+ "train_tokens_per_second": 5651.468
+ },
+ {
+ "epoch": 1.6869781847562617,
+ "grad_norm": 0.7608946538453196,
+ "learning_rate": 5.925767866555043e-07,
+ "loss": 0.28672802448272705,
+ "num_input_tokens_seen": 1095488208,
+ "step": 25055,
+ "train_runtime": 193846.2174,
+ "train_tokens_per_second": 5651.326
+ },
+ {
+ "epoch": 1.687314839752222,
+ "grad_norm": 0.8656369475130261,
+ "learning_rate": 5.913288338271095e-07,
+ "loss": 0.31831324100494385,
+ "num_input_tokens_seen": 1095711504,
+ "step": 25060,
+ "train_runtime": 193884.7583,
+ "train_tokens_per_second": 5651.355
+ },
+ {
+ "epoch": 1.6876514947481822,
+ "grad_norm": 0.7023811573666321,
+ "learning_rate": 5.900821138695478e-07,
+ "loss": 0.29593048095703123,
+ "num_input_tokens_seen": 1095931728,
+ "step": 25065,
+ "train_runtime": 193924.6678,
+ "train_tokens_per_second": 5651.327
+ },
+ {
+ "epoch": 1.6879881497441422,
+ "grad_norm": 0.760116486923957,
+ "learning_rate": 5.888366271314605e-07,
+ "loss": 0.2997978448867798,
+ "num_input_tokens_seen": 1096158920,
+ "step": 25070,
+ "train_runtime": 193970.5562,
+ "train_tokens_per_second": 5651.161
+ },
+ {
+ "epoch": 1.6883248047401023,
+ "grad_norm": 0.6921190819902673,
+ "learning_rate": 5.875923739611439e-07,
+ "loss": 0.3102496385574341,
+ "num_input_tokens_seen": 1096386400,
+ "step": 25075,
+ "train_runtime": 194005.6003,
+ "train_tokens_per_second": 5651.313
+ },
+ {
+ "epoch": 1.6886614597360625,
+ "grad_norm": 0.708393747684828,
+ "learning_rate": 5.863493547065513e-07,
+ "loss": 0.288178825378418,
+ "num_input_tokens_seen": 1096598208,
+ "step": 25080,
+ "train_runtime": 194046.4638,
+ "train_tokens_per_second": 5651.215
+ },
+ {
+ "epoch": 1.6889981147320228,
+ "grad_norm": 0.7325173487118091,
+ "learning_rate": 5.851075697152858e-07,
+ "loss": 0.2890732765197754,
+ "num_input_tokens_seen": 1096830840,
+ "step": 25085,
+ "train_runtime": 194089.2226,
+ "train_tokens_per_second": 5651.168
+ },
+ {
+ "epoch": 1.6893347697279828,
+ "grad_norm": 1.084295866928116,
+ "learning_rate": 5.838670193346118e-07,
+ "loss": 0.3075998067855835,
+ "num_input_tokens_seen": 1097042768,
+ "step": 25090,
+ "train_runtime": 194131.2034,
+ "train_tokens_per_second": 5651.038
+ },
+ {
+ "epoch": 1.6896714247239428,
+ "grad_norm": 0.708754132607845,
+ "learning_rate": 5.826277039114426e-07,
+ "loss": 0.2823533773422241,
+ "num_input_tokens_seen": 1097258936,
+ "step": 25095,
+ "train_runtime": 194173.4531,
+ "train_tokens_per_second": 5650.921
+ },
+ {
+ "epoch": 1.690008079719903,
+ "grad_norm": 0.6971867426568854,
+ "learning_rate": 5.813896237923505e-07,
+ "loss": 0.33135075569152833,
+ "num_input_tokens_seen": 1097479600,
+ "step": 25100,
+ "train_runtime": 194218.3152,
+ "train_tokens_per_second": 5650.752
+ },
+ {
+ "epoch": 1.6903447347158633,
+ "grad_norm": 0.7270149386031489,
+ "learning_rate": 5.801527793235606e-07,
+ "loss": 0.2957238912582397,
+ "num_input_tokens_seen": 1097708984,
+ "step": 25105,
+ "train_runtime": 194258.1829,
+ "train_tokens_per_second": 5650.773
+ },
+ {
+ "epoch": 1.6906813897118234,
+ "grad_norm": 0.882744897146511,
+ "learning_rate": 5.789171708509516e-07,
+ "loss": 0.28771932125091554,
+ "num_input_tokens_seen": 1097906256,
+ "step": 25110,
+ "train_runtime": 194299.2332,
+ "train_tokens_per_second": 5650.595
+ },
+ {
+ "epoch": 1.6910180447077834,
+ "grad_norm": 0.7410230789420439,
+ "learning_rate": 5.776827987200584e-07,
+ "loss": 0.30194315910339353,
+ "num_input_tokens_seen": 1098128136,
+ "step": 25115,
+ "train_runtime": 194336.6611,
+ "train_tokens_per_second": 5650.648
+ },
+ {
+ "epoch": 1.6913546997037436,
+ "grad_norm": 0.7896654861282312,
+ "learning_rate": 5.764496632760691e-07,
+ "loss": 0.2923448085784912,
+ "num_input_tokens_seen": 1098334664,
+ "step": 25120,
+ "train_runtime": 194373.4203,
+ "train_tokens_per_second": 5650.642
+ },
+ {
+ "epoch": 1.6916913546997039,
+ "grad_norm": 0.7314826062008781,
+ "learning_rate": 5.752177648638241e-07,
+ "loss": 0.27889845371246336,
+ "num_input_tokens_seen": 1098559728,
+ "step": 25125,
+ "train_runtime": 194408.4815,
+ "train_tokens_per_second": 5650.781
+ },
+ {
+ "epoch": 1.692028009695664,
+ "grad_norm": 0.8906343966024878,
+ "learning_rate": 5.739871038278239e-07,
+ "loss": 0.3101338863372803,
+ "num_input_tokens_seen": 1098782712,
+ "step": 25130,
+ "train_runtime": 194440.3466,
+ "train_tokens_per_second": 5651.002
+ },
+ {
+ "epoch": 1.692364664691624,
+ "grad_norm": 0.7281493386147246,
+ "learning_rate": 5.727576805122142e-07,
+ "loss": 0.28267052173614504,
+ "num_input_tokens_seen": 1098991872,
+ "step": 25135,
+ "train_runtime": 194480.4862,
+ "train_tokens_per_second": 5650.911
+ },
+ {
+ "epoch": 1.6927013196875842,
+ "grad_norm": 0.6893488772809538,
+ "learning_rate": 5.715294952608041e-07,
+ "loss": 0.30869436264038086,
+ "num_input_tokens_seen": 1099224528,
+ "step": 25140,
+ "train_runtime": 194512.1799,
+ "train_tokens_per_second": 5651.186
+ },
+ {
+ "epoch": 1.6930379746835444,
+ "grad_norm": 0.8413021899719699,
+ "learning_rate": 5.703025484170471e-07,
+ "loss": 0.3042963981628418,
+ "num_input_tokens_seen": 1099455640,
+ "step": 25145,
+ "train_runtime": 194548.5571,
+ "train_tokens_per_second": 5651.317
+ },
+ {
+ "epoch": 1.6933746296795045,
+ "grad_norm": 0.760681189560044,
+ "learning_rate": 5.690768403240587e-07,
+ "loss": 0.3153714656829834,
+ "num_input_tokens_seen": 1099676720,
+ "step": 25150,
+ "train_runtime": 194585.4115,
+ "train_tokens_per_second": 5651.383
+ },
+ {
+ "epoch": 1.6937112846754645,
+ "grad_norm": 0.7173964449350725,
+ "learning_rate": 5.67852371324602e-07,
+ "loss": 0.30834920406341554,
+ "num_input_tokens_seen": 1099907704,
+ "step": 25155,
+ "train_runtime": 194620.3712,
+ "train_tokens_per_second": 5651.555
+ },
+ {
+ "epoch": 1.6940479396714248,
+ "grad_norm": 0.7329549717993644,
+ "learning_rate": 5.66629141761097e-07,
+ "loss": 0.2930551528930664,
+ "num_input_tokens_seen": 1100142256,
+ "step": 25160,
+ "train_runtime": 194657.3311,
+ "train_tokens_per_second": 5651.687
+ },
+ {
+ "epoch": 1.694384594667385,
+ "grad_norm": 0.7342138353588805,
+ "learning_rate": 5.654071519756154e-07,
+ "loss": 0.30519957542419435,
+ "num_input_tokens_seen": 1100366192,
+ "step": 25165,
+ "train_runtime": 194704.0482,
+ "train_tokens_per_second": 5651.481
+ },
+ {
+ "epoch": 1.694721249663345,
+ "grad_norm": 0.7791350168211661,
+ "learning_rate": 5.641864023098831e-07,
+ "loss": 0.33337717056274413,
+ "num_input_tokens_seen": 1100584800,
+ "step": 25170,
+ "train_runtime": 194745.9485,
+ "train_tokens_per_second": 5651.387
+ },
+ {
+ "epoch": 1.695057904659305,
+ "grad_norm": 0.7179815820915966,
+ "learning_rate": 5.629668931052773e-07,
+ "loss": 0.29319195747375487,
+ "num_input_tokens_seen": 1100817816,
+ "step": 25175,
+ "train_runtime": 194785.1895,
+ "train_tokens_per_second": 5651.445
+ },
+ {
+ "epoch": 1.6953945596552653,
+ "grad_norm": 0.73809182276721,
+ "learning_rate": 5.61748624702833e-07,
+ "loss": 0.31841015815734863,
+ "num_input_tokens_seen": 1101043616,
+ "step": 25180,
+ "train_runtime": 194820.4275,
+ "train_tokens_per_second": 5651.582
+ },
+ {
+ "epoch": 1.6957312146512256,
+ "grad_norm": 0.7597610918234663,
+ "learning_rate": 5.605315974432313e-07,
+ "loss": 0.29940404891967776,
+ "num_input_tokens_seen": 1101262632,
+ "step": 25185,
+ "train_runtime": 194863.2161,
+ "train_tokens_per_second": 5651.465
+ },
+ {
+ "epoch": 1.6960678696471856,
+ "grad_norm": 0.7151694523792368,
+ "learning_rate": 5.593158116668146e-07,
+ "loss": 0.2952594757080078,
+ "num_input_tokens_seen": 1101502480,
+ "step": 25190,
+ "train_runtime": 194904.3942,
+ "train_tokens_per_second": 5651.502
+ },
+ {
+ "epoch": 1.6964045246431456,
+ "grad_norm": 0.8157887300040912,
+ "learning_rate": 5.581012677135683e-07,
+ "loss": 0.2999080181121826,
+ "num_input_tokens_seen": 1101706720,
+ "step": 25195,
+ "train_runtime": 194939.3381,
+ "train_tokens_per_second": 5651.536
+ },
+ {
+ "epoch": 1.6967411796391059,
+ "grad_norm": 0.7188880993897309,
+ "learning_rate": 5.56887965923139e-07,
+ "loss": 0.30147387981414797,
+ "num_input_tokens_seen": 1101936320,
+ "step": 25200,
+ "train_runtime": 194980.9761,
+ "train_tokens_per_second": 5651.507
+ },
+ {
+ "epoch": 1.6970778346350661,
+ "grad_norm": 0.8551105269669304,
+ "learning_rate": 5.556759066348227e-07,
+ "loss": 0.29673826694488525,
+ "num_input_tokens_seen": 1102155368,
+ "step": 25205,
+ "train_runtime": 195018.2983,
+ "train_tokens_per_second": 5651.548
+ },
+ {
+ "epoch": 1.6974144896310261,
+ "grad_norm": 0.7898708049199309,
+ "learning_rate": 5.54465090187567e-07,
+ "loss": 0.3003216743469238,
+ "num_input_tokens_seen": 1102367352,
+ "step": 25210,
+ "train_runtime": 195053.3847,
+ "train_tokens_per_second": 5651.619
+ },
+ {
+ "epoch": 1.6977511446269862,
+ "grad_norm": 0.7892657487617118,
+ "learning_rate": 5.532555169199727e-07,
+ "loss": 0.29103655815124513,
+ "num_input_tokens_seen": 1102583040,
+ "step": 25215,
+ "train_runtime": 195087.2838,
+ "train_tokens_per_second": 5651.742
+ },
+ {
+ "epoch": 1.6980877996229464,
+ "grad_norm": 0.729451521498232,
+ "learning_rate": 5.520471871702937e-07,
+ "loss": 0.28641371726989745,
+ "num_input_tokens_seen": 1102810576,
+ "step": 25220,
+ "train_runtime": 195127.6822,
+ "train_tokens_per_second": 5651.738
+ },
+ {
+ "epoch": 1.6984244546189067,
+ "grad_norm": 0.8574198765928966,
+ "learning_rate": 5.508401012764341e-07,
+ "loss": 0.2999880313873291,
+ "num_input_tokens_seen": 1103028192,
+ "step": 25225,
+ "train_runtime": 195157.8665,
+ "train_tokens_per_second": 5651.979
+ },
+ {
+ "epoch": 1.6987611096148667,
+ "grad_norm": 0.7201969737459759,
+ "learning_rate": 5.496342595759552e-07,
+ "loss": 0.28559165000915526,
+ "num_input_tokens_seen": 1103256944,
+ "step": 25230,
+ "train_runtime": 195204.2497,
+ "train_tokens_per_second": 5651.808
+ },
+ {
+ "epoch": 1.6990977646108267,
+ "grad_norm": 0.719407135820255,
+ "learning_rate": 5.484296624060626e-07,
+ "loss": 0.2883117198944092,
+ "num_input_tokens_seen": 1103482616,
+ "step": 25235,
+ "train_runtime": 195249.5833,
+ "train_tokens_per_second": 5651.652
+ },
+ {
+ "epoch": 1.699434419606787,
+ "grad_norm": 0.8370197571942842,
+ "learning_rate": 5.472263101036212e-07,
+ "loss": 0.29885265827178953,
+ "num_input_tokens_seen": 1103706952,
+ "step": 25240,
+ "train_runtime": 195285.64,
+ "train_tokens_per_second": 5651.757
+ },
+ {
+ "epoch": 1.6997710746027472,
+ "grad_norm": 0.839267857665248,
+ "learning_rate": 5.460242030051421e-07,
+ "loss": 0.3131263732910156,
+ "num_input_tokens_seen": 1103918168,
+ "step": 25245,
+ "train_runtime": 195322.8829,
+ "train_tokens_per_second": 5651.761
+ },
+ {
+ "epoch": 1.7001077295987073,
+ "grad_norm": 0.7353115375828058,
+ "learning_rate": 5.448233414467924e-07,
+ "loss": 0.3075102806091309,
+ "num_input_tokens_seen": 1104153064,
+ "step": 25250,
+ "train_runtime": 195358.9751,
+ "train_tokens_per_second": 5651.919
+ },
+ {
+ "epoch": 1.7004443845946673,
+ "grad_norm": 0.73339929516101,
+ "learning_rate": 5.43623725764389e-07,
+ "loss": 0.3154341697692871,
+ "num_input_tokens_seen": 1104367200,
+ "step": 25255,
+ "train_runtime": 195396.1799,
+ "train_tokens_per_second": 5651.939
+ },
+ {
+ "epoch": 1.7007810395906275,
+ "grad_norm": 0.7555499440673459,
+ "learning_rate": 5.424253562934001e-07,
+ "loss": 0.29883978366851804,
+ "num_input_tokens_seen": 1104597088,
+ "step": 25260,
+ "train_runtime": 195441.1857,
+ "train_tokens_per_second": 5651.813
+ },
+ {
+ "epoch": 1.7011176945865878,
+ "grad_norm": 0.7445236468484644,
+ "learning_rate": 5.412282333689461e-07,
+ "loss": 0.29061620235443114,
+ "num_input_tokens_seen": 1104824056,
+ "step": 25265,
+ "train_runtime": 195480.4688,
+ "train_tokens_per_second": 5651.839
+ },
+ {
+ "epoch": 1.7014543495825478,
+ "grad_norm": 0.7030179795843015,
+ "learning_rate": 5.400323573257987e-07,
+ "loss": 0.29516527652740476,
+ "num_input_tokens_seen": 1105049880,
+ "step": 25270,
+ "train_runtime": 195517.0208,
+ "train_tokens_per_second": 5651.937
+ },
+ {
+ "epoch": 1.7017910045785078,
+ "grad_norm": 0.7158916904045084,
+ "learning_rate": 5.388377284983803e-07,
+ "loss": 0.287148642539978,
+ "num_input_tokens_seen": 1105277608,
+ "step": 25275,
+ "train_runtime": 195553.5443,
+ "train_tokens_per_second": 5652.046
+ },
+ {
+ "epoch": 1.702127659574468,
+ "grad_norm": 0.7855552714949071,
+ "learning_rate": 5.37644347220767e-07,
+ "loss": 0.2909886360168457,
+ "num_input_tokens_seen": 1105492920,
+ "step": 25280,
+ "train_runtime": 195593.7926,
+ "train_tokens_per_second": 5651.984
+ },
+ {
+ "epoch": 1.7024643145704283,
+ "grad_norm": 0.7992003312183278,
+ "learning_rate": 5.364522138266809e-07,
+ "loss": 0.30830817222595214,
+ "num_input_tokens_seen": 1105725216,
+ "step": 25285,
+ "train_runtime": 195629.7512,
+ "train_tokens_per_second": 5652.132
+ },
+ {
+ "epoch": 1.7028009695663884,
+ "grad_norm": 0.7548214170551567,
+ "learning_rate": 5.352613286495018e-07,
+ "loss": 0.2986649990081787,
+ "num_input_tokens_seen": 1105936272,
+ "step": 25290,
+ "train_runtime": 195672.9754,
+ "train_tokens_per_second": 5651.962
+ },
+ {
+ "epoch": 1.7031376245623484,
+ "grad_norm": 0.6650218012220956,
+ "learning_rate": 5.340716920222527e-07,
+ "loss": 0.28066387176513674,
+ "num_input_tokens_seen": 1106155736,
+ "step": 25295,
+ "train_runtime": 195711.2334,
+ "train_tokens_per_second": 5651.979
+ },
+ {
+ "epoch": 1.7034742795583087,
+ "grad_norm": 0.8192022199015255,
+ "learning_rate": 5.32883304277616e-07,
+ "loss": 0.3220468521118164,
+ "num_input_tokens_seen": 1106367880,
+ "step": 25300,
+ "train_runtime": 195754.2351,
+ "train_tokens_per_second": 5651.821
+ },
+ {
+ "epoch": 1.703810934554269,
+ "grad_norm": 0.8220156836674571,
+ "learning_rate": 5.31696165747918e-07,
+ "loss": 0.2808229446411133,
+ "num_input_tokens_seen": 1106591160,
+ "step": 25305,
+ "train_runtime": 195793.8848,
+ "train_tokens_per_second": 5651.817
+ },
+ {
+ "epoch": 1.704147589550229,
+ "grad_norm": 0.7943490070647735,
+ "learning_rate": 5.305102767651393e-07,
+ "loss": 0.28664450645446776,
+ "num_input_tokens_seen": 1106820392,
+ "step": 25310,
+ "train_runtime": 195840.7495,
+ "train_tokens_per_second": 5651.635
+ },
+ {
+ "epoch": 1.704484244546189,
+ "grad_norm": 0.7913194850774943,
+ "learning_rate": 5.293256376609091e-07,
+ "loss": 0.30523157119750977,
+ "num_input_tokens_seen": 1107043152,
+ "step": 25315,
+ "train_runtime": 195882.9879,
+ "train_tokens_per_second": 5651.553
+ },
+ {
+ "epoch": 1.7048208995421492,
+ "grad_norm": 0.7081852025829084,
+ "learning_rate": 5.281422487665089e-07,
+ "loss": 0.27099974155426027,
+ "num_input_tokens_seen": 1107273600,
+ "step": 25320,
+ "train_runtime": 195922.0686,
+ "train_tokens_per_second": 5651.602
+ },
+ {
+ "epoch": 1.7051575545381095,
+ "grad_norm": 0.7673688868553783,
+ "learning_rate": 5.269601104128674e-07,
+ "loss": 0.2887356042861938,
+ "num_input_tokens_seen": 1107500568,
+ "step": 25325,
+ "train_runtime": 195968.4871,
+ "train_tokens_per_second": 5651.422
+ },
+ {
+ "epoch": 1.7054942095340695,
+ "grad_norm": 0.7390671103279993,
+ "learning_rate": 5.257792229305702e-07,
+ "loss": 0.3096314430236816,
+ "num_input_tokens_seen": 1107721408,
+ "step": 25330,
+ "train_runtime": 196009.11,
+ "train_tokens_per_second": 5651.377
+ },
+ {
+ "epoch": 1.7058308645300295,
+ "grad_norm": 0.9154523312754923,
+ "learning_rate": 5.245995866498438e-07,
+ "loss": 0.27949187755584715,
+ "num_input_tokens_seen": 1107927904,
+ "step": 25335,
+ "train_runtime": 196050.2257,
+ "train_tokens_per_second": 5651.245
+ },
+ {
+ "epoch": 1.7061675195259898,
+ "grad_norm": 0.779691321745416,
+ "learning_rate": 5.234212019005741e-07,
+ "loss": 0.3069715738296509,
+ "num_input_tokens_seen": 1108137040,
+ "step": 25340,
+ "train_runtime": 196092.6734,
+ "train_tokens_per_second": 5651.088
+ },
+ {
+ "epoch": 1.70650417452195,
+ "grad_norm": 0.7743911753132502,
+ "learning_rate": 5.222440690122882e-07,
+ "loss": 0.3008026838302612,
+ "num_input_tokens_seen": 1108352000,
+ "step": 25345,
+ "train_runtime": 196133.018,
+ "train_tokens_per_second": 5651.022
+ },
+ {
+ "epoch": 1.70684082951791,
+ "grad_norm": 0.771105877517414,
+ "learning_rate": 5.210681883141716e-07,
+ "loss": 0.2749685764312744,
+ "num_input_tokens_seen": 1108571744,
+ "step": 25350,
+ "train_runtime": 196169.374,
+ "train_tokens_per_second": 5651.095
+ },
+ {
+ "epoch": 1.70717748451387,
+ "grad_norm": 0.7792881082810094,
+ "learning_rate": 5.19893560135053e-07,
+ "loss": 0.3137876749038696,
+ "num_input_tokens_seen": 1108803560,
+ "step": 25355,
+ "train_runtime": 196203.8165,
+ "train_tokens_per_second": 5651.284
+ },
+ {
+ "epoch": 1.7075141395098303,
+ "grad_norm": 0.7995549624581778,
+ "learning_rate": 5.187201848034146e-07,
+ "loss": 0.3042750835418701,
+ "num_input_tokens_seen": 1109023488,
+ "step": 25360,
+ "train_runtime": 196245.4211,
+ "train_tokens_per_second": 5651.207
+ },
+ {
+ "epoch": 1.7078507945057906,
+ "grad_norm": 0.7706690517207955,
+ "learning_rate": 5.175480626473873e-07,
+ "loss": 0.290797758102417,
+ "num_input_tokens_seen": 1109234112,
+ "step": 25365,
+ "train_runtime": 196284.3763,
+ "train_tokens_per_second": 5651.158
+ },
+ {
+ "epoch": 1.7081874495017506,
+ "grad_norm": 0.888485346761317,
+ "learning_rate": 5.163771939947504e-07,
+ "loss": 0.30904016494750974,
+ "num_input_tokens_seen": 1109454048,
+ "step": 25370,
+ "train_runtime": 196324.1237,
+ "train_tokens_per_second": 5651.135
+ },
+ {
+ "epoch": 1.7085241044977106,
+ "grad_norm": 0.6868307931042189,
+ "learning_rate": 5.152075791729327e-07,
+ "loss": 0.3244988203048706,
+ "num_input_tokens_seen": 1109687648,
+ "step": 25375,
+ "train_runtime": 196364.0378,
+ "train_tokens_per_second": 5651.176
+ },
+ {
+ "epoch": 1.7088607594936709,
+ "grad_norm": 0.7101255126564905,
+ "learning_rate": 5.140392185090165e-07,
+ "loss": 0.2983409881591797,
+ "num_input_tokens_seen": 1109893064,
+ "step": 25380,
+ "train_runtime": 196405.8629,
+ "train_tokens_per_second": 5651.018
+ },
+ {
+ "epoch": 1.7091974144896311,
+ "grad_norm": 0.8655143566241545,
+ "learning_rate": 5.128721123297265e-07,
+ "loss": 0.3025853157043457,
+ "num_input_tokens_seen": 1110127392,
+ "step": 25385,
+ "train_runtime": 196448.4648,
+ "train_tokens_per_second": 5650.985
+ },
+ {
+ "epoch": 1.7095340694855912,
+ "grad_norm": 0.7765483897971529,
+ "learning_rate": 5.117062609614438e-07,
+ "loss": 0.2883389949798584,
+ "num_input_tokens_seen": 1110349016,
+ "step": 25390,
+ "train_runtime": 196484.3558,
+ "train_tokens_per_second": 5651.081
+ },
+ {
+ "epoch": 1.7098707244815512,
+ "grad_norm": 0.8068466453475462,
+ "learning_rate": 5.105416647301909e-07,
+ "loss": 0.2935046195983887,
+ "num_input_tokens_seen": 1110572696,
+ "step": 25395,
+ "train_runtime": 196527.8258,
+ "train_tokens_per_second": 5650.969
+ },
+ {
+ "epoch": 1.7102073794775114,
+ "grad_norm": 0.8272038940843907,
+ "learning_rate": 5.093783239616468e-07,
+ "loss": 0.26650362014770507,
+ "num_input_tokens_seen": 1110779632,
+ "step": 25400,
+ "train_runtime": 196565.2985,
+ "train_tokens_per_second": 5650.945
+ },
+ {
+ "epoch": 1.7105440344734717,
+ "grad_norm": 0.8912009620998278,
+ "learning_rate": 5.082162389811352e-07,
+ "loss": 0.28749451637268064,
+ "num_input_tokens_seen": 1110997128,
+ "step": 25405,
+ "train_runtime": 196600.4482,
+ "train_tokens_per_second": 5651.041
+ },
+ {
+ "epoch": 1.7108806894694317,
+ "grad_norm": 0.8227134244564941,
+ "learning_rate": 5.070554101136288e-07,
+ "loss": 0.2878044605255127,
+ "num_input_tokens_seen": 1111200552,
+ "step": 25410,
+ "train_runtime": 196637.1184,
+ "train_tokens_per_second": 5651.021
+ },
+ {
+ "epoch": 1.7112173444653918,
+ "grad_norm": 0.6957933161324612,
+ "learning_rate": 5.058958376837497e-07,
+ "loss": 0.2986098289489746,
+ "num_input_tokens_seen": 1111427560,
+ "step": 25415,
+ "train_runtime": 196678.1193,
+ "train_tokens_per_second": 5650.997
+ },
+ {
+ "epoch": 1.711553999461352,
+ "grad_norm": 0.7375797617626252,
+ "learning_rate": 5.047375220157691e-07,
+ "loss": 0.3050894021987915,
+ "num_input_tokens_seen": 1111652600,
+ "step": 25420,
+ "train_runtime": 196718.6261,
+ "train_tokens_per_second": 5650.978
+ },
+ {
+ "epoch": 1.7118906544573123,
+ "grad_norm": 0.7728802385790842,
+ "learning_rate": 5.035804634336056e-07,
+ "loss": 0.3212940216064453,
+ "num_input_tokens_seen": 1111871248,
+ "step": 25425,
+ "train_runtime": 196760.4219,
+ "train_tokens_per_second": 5650.889
+ },
+ {
+ "epoch": 1.7122273094532723,
+ "grad_norm": 0.7147851806060915,
+ "learning_rate": 5.024246622608292e-07,
+ "loss": 0.293403172492981,
+ "num_input_tokens_seen": 1112108600,
+ "step": 25430,
+ "train_runtime": 196799.535,
+ "train_tokens_per_second": 5650.972
+ },
+ {
+ "epoch": 1.7125639644492323,
+ "grad_norm": 0.8753202500856931,
+ "learning_rate": 5.01270118820652e-07,
+ "loss": 0.3186427354812622,
+ "num_input_tokens_seen": 1112335328,
+ "step": 25435,
+ "train_runtime": 196834.1465,
+ "train_tokens_per_second": 5651.13
+ },
+ {
+ "epoch": 1.7129006194451926,
+ "grad_norm": 0.8169092701177705,
+ "learning_rate": 5.001168334359429e-07,
+ "loss": 0.3163825511932373,
+ "num_input_tokens_seen": 1112566808,
+ "step": 25440,
+ "train_runtime": 196867.059,
+ "train_tokens_per_second": 5651.361
+ },
+ {
+ "epoch": 1.7132372744411528,
+ "grad_norm": 0.7130198658873356,
+ "learning_rate": 4.989648064292102e-07,
+ "loss": 0.3128488540649414,
+ "num_input_tokens_seen": 1112795792,
+ "step": 25445,
+ "train_runtime": 196905.6927,
+ "train_tokens_per_second": 5651.415
+ },
+ {
+ "epoch": 1.7135739294371128,
+ "grad_norm": 0.8272498786011496,
+ "learning_rate": 4.978140381226177e-07,
+ "loss": 0.31338427066802976,
+ "num_input_tokens_seen": 1113015816,
+ "step": 25450,
+ "train_runtime": 196950.2891,
+ "train_tokens_per_second": 5651.253
+ },
+ {
+ "epoch": 1.7139105844330729,
+ "grad_norm": 0.6817930969478092,
+ "learning_rate": 4.966645288379729e-07,
+ "loss": 0.2886411428451538,
+ "num_input_tokens_seen": 1113242208,
+ "step": 25455,
+ "train_runtime": 196986.8049,
+ "train_tokens_per_second": 5651.354
+ },
+ {
+ "epoch": 1.7142472394290331,
+ "grad_norm": 0.8193216088135635,
+ "learning_rate": 4.955162788967327e-07,
+ "loss": 0.3108631134033203,
+ "num_input_tokens_seen": 1113433336,
+ "step": 25460,
+ "train_runtime": 197028.9568,
+ "train_tokens_per_second": 5651.115
+ },
+ {
+ "epoch": 1.7145838944249934,
+ "grad_norm": 0.7410886402030697,
+ "learning_rate": 4.943692886200013e-07,
+ "loss": 0.30652599334716796,
+ "num_input_tokens_seen": 1113646408,
+ "step": 25465,
+ "train_runtime": 197065.6766,
+ "train_tokens_per_second": 5651.143
+ },
+ {
+ "epoch": 1.7149205494209534,
+ "grad_norm": 0.7490199419500515,
+ "learning_rate": 4.932235583285305e-07,
+ "loss": 0.31150734424591064,
+ "num_input_tokens_seen": 1113881008,
+ "step": 25470,
+ "train_runtime": 197104.2655,
+ "train_tokens_per_second": 5651.227
+ },
+ {
+ "epoch": 1.7152572044169134,
+ "grad_norm": 0.8168471127164367,
+ "learning_rate": 4.920790883427201e-07,
+ "loss": 0.2943288326263428,
+ "num_input_tokens_seen": 1114079888,
+ "step": 25475,
+ "train_runtime": 197148.0427,
+ "train_tokens_per_second": 5650.981
+ },
+ {
+ "epoch": 1.7155938594128737,
+ "grad_norm": 0.8217572900775035,
+ "learning_rate": 4.909358789826196e-07,
+ "loss": 0.3120744228363037,
+ "num_input_tokens_seen": 1114268728,
+ "step": 25480,
+ "train_runtime": 197186.0762,
+ "train_tokens_per_second": 5650.849
+ },
+ {
+ "epoch": 1.715930514408834,
+ "grad_norm": 0.842144785636648,
+ "learning_rate": 4.897939305679206e-07,
+ "loss": 0.30831499099731446,
+ "num_input_tokens_seen": 1114498760,
+ "step": 25485,
+ "train_runtime": 197217.8311,
+ "train_tokens_per_second": 5651.105
+ },
+ {
+ "epoch": 1.716267169404794,
+ "grad_norm": 0.7113758542252159,
+ "learning_rate": 4.886532434179686e-07,
+ "loss": 0.2864993572235107,
+ "num_input_tokens_seen": 1114714920,
+ "step": 25490,
+ "train_runtime": 197261.5876,
+ "train_tokens_per_second": 5650.948
+ },
+ {
+ "epoch": 1.716603824400754,
+ "grad_norm": 0.7070024392396191,
+ "learning_rate": 4.875138178517497e-07,
+ "loss": 0.2663728713989258,
+ "num_input_tokens_seen": 1114931832,
+ "step": 25495,
+ "train_runtime": 197300.3097,
+ "train_tokens_per_second": 5650.938
+ },
+ {
+ "epoch": 1.7169404793967142,
+ "grad_norm": 0.8884464092351633,
+ "learning_rate": 4.863756541879028e-07,
+ "loss": 0.31972756385803225,
+ "num_input_tokens_seen": 1115156536,
+ "step": 25500,
+ "train_runtime": 197338.9855,
+ "train_tokens_per_second": 5650.969
+ },
+ {
+ "epoch": 1.7172771343926745,
+ "grad_norm": 0.693402281770043,
+ "learning_rate": 4.852387527447117e-07,
+ "loss": 0.2761407375335693,
+ "num_input_tokens_seen": 1115391624,
+ "step": 25505,
+ "train_runtime": 197375.3791,
+ "train_tokens_per_second": 5651.118
+ },
+ {
+ "epoch": 1.7176137893886345,
+ "grad_norm": 0.7175083758215528,
+ "learning_rate": 4.84103113840107e-07,
+ "loss": 0.2971747398376465,
+ "num_input_tokens_seen": 1115621104,
+ "step": 25510,
+ "train_runtime": 197410.3395,
+ "train_tokens_per_second": 5651.28
+ },
+ {
+ "epoch": 1.7179504443845945,
+ "grad_norm": 0.7429321287522253,
+ "learning_rate": 4.829687377916664e-07,
+ "loss": 0.2947702407836914,
+ "num_input_tokens_seen": 1115833104,
+ "step": 25515,
+ "train_runtime": 197441.5837,
+ "train_tokens_per_second": 5651.459
+ },
+ {
+ "epoch": 1.7182870993805548,
+ "grad_norm": 0.8107818792796975,
+ "learning_rate": 4.81835624916614e-07,
+ "loss": 0.2821854591369629,
+ "num_input_tokens_seen": 1116048240,
+ "step": 25520,
+ "train_runtime": 197479.0855,
+ "train_tokens_per_second": 5651.476
+ },
+ {
+ "epoch": 1.718623754376515,
+ "grad_norm": 0.802522255622259,
+ "learning_rate": 4.807037755318211e-07,
+ "loss": 0.28911077976226807,
+ "num_input_tokens_seen": 1116280624,
+ "step": 25525,
+ "train_runtime": 197512.9072,
+ "train_tokens_per_second": 5651.684
+ },
+ {
+ "epoch": 1.718960409372475,
+ "grad_norm": 0.6592568584202634,
+ "learning_rate": 4.795731899538086e-07,
+ "loss": 0.3094839096069336,
+ "num_input_tokens_seen": 1116492320,
+ "step": 25530,
+ "train_runtime": 197560.5492,
+ "train_tokens_per_second": 5651.393
+ },
+ {
+ "epoch": 1.719297064368435,
+ "grad_norm": 0.7237886999876593,
+ "learning_rate": 4.78443868498737e-07,
+ "loss": 0.2828222274780273,
+ "num_input_tokens_seen": 1116713608,
+ "step": 25535,
+ "train_runtime": 197596.8732,
+ "train_tokens_per_second": 5651.474
+ },
+ {
+ "epoch": 1.7196337193643954,
+ "grad_norm": 0.7185713199478717,
+ "learning_rate": 4.773158114824206e-07,
+ "loss": 0.28137338161468506,
+ "num_input_tokens_seen": 1116938536,
+ "step": 25540,
+ "train_runtime": 197632.4385,
+ "train_tokens_per_second": 5651.595
+ },
+ {
+ "epoch": 1.7199703743603556,
+ "grad_norm": 1.1879567747502189,
+ "learning_rate": 4.7618901922031445e-07,
+ "loss": 0.2812469959259033,
+ "num_input_tokens_seen": 1117161488,
+ "step": 25545,
+ "train_runtime": 197670.8796,
+ "train_tokens_per_second": 5651.624
+ },
+ {
+ "epoch": 1.7203070293563156,
+ "grad_norm": 0.8094448913417437,
+ "learning_rate": 4.750634920275243e-07,
+ "loss": 0.3022423267364502,
+ "num_input_tokens_seen": 1117374592,
+ "step": 25550,
+ "train_runtime": 197712.1996,
+ "train_tokens_per_second": 5651.521
+ },
+ {
+ "epoch": 1.7206436843522757,
+ "grad_norm": 0.752864591432035,
+ "learning_rate": 4.739392302188001e-07,
+ "loss": 0.319273042678833,
+ "num_input_tokens_seen": 1117603384,
+ "step": 25555,
+ "train_runtime": 197750.9243,
+ "train_tokens_per_second": 5651.571
+ },
+ {
+ "epoch": 1.720980339348236,
+ "grad_norm": 0.8294658057367533,
+ "learning_rate": 4.728162341085368e-07,
+ "loss": 0.31023268699645995,
+ "num_input_tokens_seen": 1117820784,
+ "step": 25560,
+ "train_runtime": 197793.1161,
+ "train_tokens_per_second": 5651.465
+ },
+ {
+ "epoch": 1.7213169943441962,
+ "grad_norm": 0.7053663821231337,
+ "learning_rate": 4.7169450401077843e-07,
+ "loss": 0.312652063369751,
+ "num_input_tokens_seen": 1118048160,
+ "step": 25565,
+ "train_runtime": 197836.8692,
+ "train_tokens_per_second": 5651.364
+ },
+ {
+ "epoch": 1.7216536493401562,
+ "grad_norm": 0.7700362624042942,
+ "learning_rate": 4.7057404023921174e-07,
+ "loss": 0.3073885440826416,
+ "num_input_tokens_seen": 1118274672,
+ "step": 25570,
+ "train_runtime": 197872.3987,
+ "train_tokens_per_second": 5651.494
+ },
+ {
+ "epoch": 1.7219903043361162,
+ "grad_norm": 0.8048237014417249,
+ "learning_rate": 4.694548431071705e-07,
+ "loss": 0.31366183757781985,
+ "num_input_tokens_seen": 1118492208,
+ "step": 25575,
+ "train_runtime": 197919.5839,
+ "train_tokens_per_second": 5651.246
+ },
+ {
+ "epoch": 1.7223269593320765,
+ "grad_norm": 0.7167405150431256,
+ "learning_rate": 4.683369129276377e-07,
+ "loss": 0.3049338340759277,
+ "num_input_tokens_seen": 1118716936,
+ "step": 25580,
+ "train_runtime": 197953.2698,
+ "train_tokens_per_second": 5651.419
+ },
+ {
+ "epoch": 1.7226636143280367,
+ "grad_norm": 0.663480470579991,
+ "learning_rate": 4.6722025001323444e-07,
+ "loss": 0.2772920846939087,
+ "num_input_tokens_seen": 1118917560,
+ "step": 25585,
+ "train_runtime": 197998.4722,
+ "train_tokens_per_second": 5651.142
+ },
+ {
+ "epoch": 1.7230002693239967,
+ "grad_norm": 0.8472714336560921,
+ "learning_rate": 4.6610485467623643e-07,
+ "loss": 0.3002648830413818,
+ "num_input_tokens_seen": 1119122208,
+ "step": 25590,
+ "train_runtime": 198037.4639,
+ "train_tokens_per_second": 5651.063
+ },
+ {
+ "epoch": 1.7233369243199568,
+ "grad_norm": 0.73662714701174,
+ "learning_rate": 4.649907272285564e-07,
+ "loss": 0.29230647087097167,
+ "num_input_tokens_seen": 1119341064,
+ "step": 25595,
+ "train_runtime": 198077.5592,
+ "train_tokens_per_second": 5651.024
+ },
+ {
+ "epoch": 1.723673579315917,
+ "grad_norm": 0.7420030033003574,
+ "learning_rate": 4.638778679817596e-07,
+ "loss": 0.28714590072631835,
+ "num_input_tokens_seen": 1119570192,
+ "step": 25600,
+ "train_runtime": 198120.1877,
+ "train_tokens_per_second": 5650.965
+ },
+ {
+ "epoch": 1.7240102343118773,
+ "grad_norm": 0.824030176081278,
+ "learning_rate": 4.6276627724705204e-07,
+ "loss": 0.3071947813034058,
+ "num_input_tokens_seen": 1119786040,
+ "step": 25605,
+ "train_runtime": 198161.3927,
+ "train_tokens_per_second": 5650.879
+ },
+ {
+ "epoch": 1.7243468893078373,
+ "grad_norm": 0.7154654354680215,
+ "learning_rate": 4.616559553352873e-07,
+ "loss": 0.2772822380065918,
+ "num_input_tokens_seen": 1120003688,
+ "step": 25610,
+ "train_runtime": 198197.9173,
+ "train_tokens_per_second": 5650.936
+ },
+ {
+ "epoch": 1.7246835443037973,
+ "grad_norm": 0.7332839778498565,
+ "learning_rate": 4.6054690255696256e-07,
+ "loss": 0.3121823310852051,
+ "num_input_tokens_seen": 1120225000,
+ "step": 25615,
+ "train_runtime": 198232.3778,
+ "train_tokens_per_second": 5651.07
+ },
+ {
+ "epoch": 1.7250201992997576,
+ "grad_norm": 0.8304618732220158,
+ "learning_rate": 4.594391192222214e-07,
+ "loss": 0.299558162689209,
+ "num_input_tokens_seen": 1120436736,
+ "step": 25620,
+ "train_runtime": 198275.4196,
+ "train_tokens_per_second": 5650.911
+ },
+ {
+ "epoch": 1.7253568542957178,
+ "grad_norm": 0.7210943628722162,
+ "learning_rate": 4.583326056408505e-07,
+ "loss": 0.2828629732131958,
+ "num_input_tokens_seen": 1120656400,
+ "step": 25625,
+ "train_runtime": 198318.2954,
+ "train_tokens_per_second": 5650.797
+ },
+ {
+ "epoch": 1.7256935092916779,
+ "grad_norm": 0.8701484329791821,
+ "learning_rate": 4.5722736212228613e-07,
+ "loss": 0.3062956094741821,
+ "num_input_tokens_seen": 1120853112,
+ "step": 25630,
+ "train_runtime": 198360.6419,
+ "train_tokens_per_second": 5650.582
+ },
+ {
+ "epoch": 1.726030164287638,
+ "grad_norm": 0.7789016257709495,
+ "learning_rate": 4.5612338897560227e-07,
+ "loss": 0.3126507759094238,
+ "num_input_tokens_seen": 1121085288,
+ "step": 25635,
+ "train_runtime": 198403.9135,
+ "train_tokens_per_second": 5650.52
+ },
+ {
+ "epoch": 1.7263668192835981,
+ "grad_norm": 0.8101928313460407,
+ "learning_rate": 4.5502068650952527e-07,
+ "loss": 0.31250529289245604,
+ "num_input_tokens_seen": 1121299992,
+ "step": 25640,
+ "train_runtime": 198443.451,
+ "train_tokens_per_second": 5650.476
+ },
+ {
+ "epoch": 1.7267034742795584,
+ "grad_norm": 0.8289277705819492,
+ "learning_rate": 4.5391925503241894e-07,
+ "loss": 0.2995558023452759,
+ "num_input_tokens_seen": 1121526960,
+ "step": 25645,
+ "train_runtime": 198482.6522,
+ "train_tokens_per_second": 5650.504
+ },
+ {
+ "epoch": 1.7270401292755184,
+ "grad_norm": 0.7957658163232814,
+ "learning_rate": 4.5281909485229704e-07,
+ "loss": 0.29502243995666505,
+ "num_input_tokens_seen": 1121740648,
+ "step": 25650,
+ "train_runtime": 198523.6274,
+ "train_tokens_per_second": 5650.414
+ },
+ {
+ "epoch": 1.7273767842714784,
+ "grad_norm": 0.8283561928306208,
+ "learning_rate": 4.51720206276815e-07,
+ "loss": 0.27261307239532473,
+ "num_input_tokens_seen": 1121954360,
+ "step": 25655,
+ "train_runtime": 198559.855,
+ "train_tokens_per_second": 5650.459
+ },
+ {
+ "epoch": 1.7277134392674387,
+ "grad_norm": 0.7838264977883859,
+ "learning_rate": 4.506225896132743e-07,
+ "loss": 0.2796020984649658,
+ "num_input_tokens_seen": 1122168368,
+ "step": 25660,
+ "train_runtime": 198597.1547,
+ "train_tokens_per_second": 5650.476
+ },
+ {
+ "epoch": 1.728050094263399,
+ "grad_norm": 0.7911207040134044,
+ "learning_rate": 4.4952624516861944e-07,
+ "loss": 0.2953200101852417,
+ "num_input_tokens_seen": 1122372424,
+ "step": 25665,
+ "train_runtime": 198632.1048,
+ "train_tokens_per_second": 5650.509
+ },
+ {
+ "epoch": 1.728386749259359,
+ "grad_norm": 0.7188871555974835,
+ "learning_rate": 4.4843117324944005e-07,
+ "loss": 0.2733217716217041,
+ "num_input_tokens_seen": 1122608584,
+ "step": 25670,
+ "train_runtime": 198671.3821,
+ "train_tokens_per_second": 5650.58
+ },
+ {
+ "epoch": 1.728723404255319,
+ "grad_norm": 0.8236529240166243,
+ "learning_rate": 4.4733737416196745e-07,
+ "loss": 0.2921212911605835,
+ "num_input_tokens_seen": 1122786752,
+ "step": 25675,
+ "train_runtime": 198705.9189,
+ "train_tokens_per_second": 5650.495
+ },
+ {
+ "epoch": 1.7290600592512793,
+ "grad_norm": 0.6891675836585306,
+ "learning_rate": 4.4624484821208346e-07,
+ "loss": 0.29300737380981445,
+ "num_input_tokens_seen": 1123000352,
+ "step": 25680,
+ "train_runtime": 198743.7451,
+ "train_tokens_per_second": 5650.494
+ },
+ {
+ "epoch": 1.7293967142472395,
+ "grad_norm": 0.7600535970631421,
+ "learning_rate": 4.451535957053044e-07,
+ "loss": 0.26680479049682615,
+ "num_input_tokens_seen": 1123218520,
+ "step": 25685,
+ "train_runtime": 198786.9925,
+ "train_tokens_per_second": 5650.362
+ },
+ {
+ "epoch": 1.7297333692431995,
+ "grad_norm": 0.7452499443166702,
+ "learning_rate": 4.440636169467999e-07,
+ "loss": 0.28789448738098145,
+ "num_input_tokens_seen": 1123429840,
+ "step": 25690,
+ "train_runtime": 198828.3828,
+ "train_tokens_per_second": 5650.249
+ },
+ {
+ "epoch": 1.7300700242391596,
+ "grad_norm": 0.801697711689962,
+ "learning_rate": 4.429749122413757e-07,
+ "loss": 0.2902831077575684,
+ "num_input_tokens_seen": 1123644168,
+ "step": 25695,
+ "train_runtime": 198871.0647,
+ "train_tokens_per_second": 5650.114
+ },
+ {
+ "epoch": 1.7304066792351198,
+ "grad_norm": 0.7857956805272411,
+ "learning_rate": 4.418874818934865e-07,
+ "loss": 0.29162254333496096,
+ "num_input_tokens_seen": 1123867240,
+ "step": 25700,
+ "train_runtime": 198907.5425,
+ "train_tokens_per_second": 5650.199
+ },
+ {
+ "epoch": 1.73074333423108,
+ "grad_norm": 0.829177173381663,
+ "learning_rate": 4.4080132620722806e-07,
+ "loss": 0.30682687759399413,
+ "num_input_tokens_seen": 1124064016,
+ "step": 25705,
+ "train_runtime": 198944.8498,
+ "train_tokens_per_second": 5650.129
+ },
+ {
+ "epoch": 1.73107998922704,
+ "grad_norm": 0.7300356398237123,
+ "learning_rate": 4.3971644548634116e-07,
+ "loss": 0.30285561084747314,
+ "num_input_tokens_seen": 1124276384,
+ "step": 25710,
+ "train_runtime": 198983.4747,
+ "train_tokens_per_second": 5650.099
+ },
+ {
+ "epoch": 1.7314166442230001,
+ "grad_norm": 1.020129891409066,
+ "learning_rate": 4.3863284003420826e-07,
+ "loss": 0.3177463531494141,
+ "num_input_tokens_seen": 1124494424,
+ "step": 25715,
+ "train_runtime": 199023.7261,
+ "train_tokens_per_second": 5650.052
+ },
+ {
+ "epoch": 1.7317532992189604,
+ "grad_norm": 0.7523401828963102,
+ "learning_rate": 4.375505101538563e-07,
+ "loss": 0.2902905702590942,
+ "num_input_tokens_seen": 1124712920,
+ "step": 25720,
+ "train_runtime": 199061.3982,
+ "train_tokens_per_second": 5650.08
+ },
+ {
+ "epoch": 1.7320899542149206,
+ "grad_norm": 0.8005644638652999,
+ "learning_rate": 4.3646945614795487e-07,
+ "loss": 0.3163978099822998,
+ "num_input_tokens_seen": 1124936520,
+ "step": 25725,
+ "train_runtime": 199095.1882,
+ "train_tokens_per_second": 5650.245
+ },
+ {
+ "epoch": 1.7324266092108807,
+ "grad_norm": 0.7714782744078659,
+ "learning_rate": 4.3538967831881995e-07,
+ "loss": 0.31358091831207274,
+ "num_input_tokens_seen": 1125167472,
+ "step": 25730,
+ "train_runtime": 199132.5807,
+ "train_tokens_per_second": 5650.343
+ },
+ {
+ "epoch": 1.7327632642068407,
+ "grad_norm": 0.7624866372005362,
+ "learning_rate": 4.343111769684039e-07,
+ "loss": 0.29058811664581297,
+ "num_input_tokens_seen": 1125373784,
+ "step": 25735,
+ "train_runtime": 199172.4308,
+ "train_tokens_per_second": 5650.249
+ },
+ {
+ "epoch": 1.733099919202801,
+ "grad_norm": 0.6799023332475078,
+ "learning_rate": 4.332339523983103e-07,
+ "loss": 0.27836108207702637,
+ "num_input_tokens_seen": 1125601280,
+ "step": 25740,
+ "train_runtime": 199212.9839,
+ "train_tokens_per_second": 5650.241
+ },
+ {
+ "epoch": 1.7334365741987612,
+ "grad_norm": 0.7561206882286643,
+ "learning_rate": 4.321580049097773e-07,
+ "loss": 0.28482909202575685,
+ "num_input_tokens_seen": 1125825176,
+ "step": 25745,
+ "train_runtime": 199246.8335,
+ "train_tokens_per_second": 5650.404
+ },
+ {
+ "epoch": 1.7337732291947212,
+ "grad_norm": 0.8118236999029766,
+ "learning_rate": 4.310833348036941e-07,
+ "loss": 0.27461097240447996,
+ "num_input_tokens_seen": 1126041128,
+ "step": 25750,
+ "train_runtime": 199290.7549,
+ "train_tokens_per_second": 5650.243
+ },
+ {
+ "epoch": 1.7341098841906812,
+ "grad_norm": 0.7118638487915145,
+ "learning_rate": 4.300099423805865e-07,
+ "loss": 0.2958423614501953,
+ "num_input_tokens_seen": 1126244560,
+ "step": 25755,
+ "train_runtime": 199333.5162,
+ "train_tokens_per_second": 5650.051
+ },
+ {
+ "epoch": 1.7344465391866415,
+ "grad_norm": 0.7506783418387937,
+ "learning_rate": 4.289378279406264e-07,
+ "loss": 0.2995717525482178,
+ "num_input_tokens_seen": 1126468312,
+ "step": 25760,
+ "train_runtime": 199366.8538,
+ "train_tokens_per_second": 5650.229
+ },
+ {
+ "epoch": 1.7347831941826017,
+ "grad_norm": 0.8246621666903383,
+ "learning_rate": 4.2786699178362647e-07,
+ "loss": 0.30245399475097656,
+ "num_input_tokens_seen": 1126689568,
+ "step": 25765,
+ "train_runtime": 199407.3048,
+ "train_tokens_per_second": 5650.192
+ },
+ {
+ "epoch": 1.7351198491785618,
+ "grad_norm": 0.8135304755844003,
+ "learning_rate": 4.2679743420904254e-07,
+ "loss": 0.3159316539764404,
+ "num_input_tokens_seen": 1126917656,
+ "step": 25770,
+ "train_runtime": 199441.5657,
+ "train_tokens_per_second": 5650.365
+ },
+ {
+ "epoch": 1.7354565041745218,
+ "grad_norm": 0.8164341987242794,
+ "learning_rate": 4.257291555159726e-07,
+ "loss": 0.26204819679260255,
+ "num_input_tokens_seen": 1127132592,
+ "step": 25775,
+ "train_runtime": 199474.9349,
+ "train_tokens_per_second": 5650.497
+ },
+ {
+ "epoch": 1.735793159170482,
+ "grad_norm": 0.7634016130973207,
+ "learning_rate": 4.246621560031594e-07,
+ "loss": 0.2924259424209595,
+ "num_input_tokens_seen": 1127339712,
+ "step": 25780,
+ "train_runtime": 199509.5428,
+ "train_tokens_per_second": 5650.555
+ },
+ {
+ "epoch": 1.7361298141664423,
+ "grad_norm": 0.7910333110781415,
+ "learning_rate": 4.235964359689837e-07,
+ "loss": 0.27674455642700196,
+ "num_input_tokens_seen": 1127566704,
+ "step": 25785,
+ "train_runtime": 199548.866,
+ "train_tokens_per_second": 5650.579
+ },
+ {
+ "epoch": 1.7364664691624023,
+ "grad_norm": 0.8840197427923319,
+ "learning_rate": 4.225319957114726e-07,
+ "loss": 0.3013903617858887,
+ "num_input_tokens_seen": 1127772128,
+ "step": 25790,
+ "train_runtime": 199588.6729,
+ "train_tokens_per_second": 5650.482
+ },
+ {
+ "epoch": 1.7368031241583624,
+ "grad_norm": 0.7325596847908312,
+ "learning_rate": 4.214688355282909e-07,
+ "loss": 0.28219919204711913,
+ "num_input_tokens_seen": 1127988664,
+ "step": 25795,
+ "train_runtime": 199631.9302,
+ "train_tokens_per_second": 5650.342
+ },
+ {
+ "epoch": 1.7371397791543226,
+ "grad_norm": 0.8173650460918741,
+ "learning_rate": 4.204069557167506e-07,
+ "loss": 0.2839889764785767,
+ "num_input_tokens_seen": 1128202648,
+ "step": 25800,
+ "train_runtime": 199667.9001,
+ "train_tokens_per_second": 5650.396
+ },
+ {
+ "epoch": 1.7374764341502829,
+ "grad_norm": 0.876625098531604,
+ "learning_rate": 4.1934635657380153e-07,
+ "loss": 0.304730224609375,
+ "num_input_tokens_seen": 1128423984,
+ "step": 25805,
+ "train_runtime": 199702.6037,
+ "train_tokens_per_second": 5650.522
+ },
+ {
+ "epoch": 1.7378130891462429,
+ "grad_norm": 0.7480060231681754,
+ "learning_rate": 4.1828703839603637e-07,
+ "loss": 0.2796186447143555,
+ "num_input_tokens_seen": 1128651136,
+ "step": 25810,
+ "train_runtime": 199740.7074,
+ "train_tokens_per_second": 5650.581
+ },
+ {
+ "epoch": 1.738149744142203,
+ "grad_norm": 0.8165869116083727,
+ "learning_rate": 4.172290014796926e-07,
+ "loss": 0.32165372371673584,
+ "num_input_tokens_seen": 1128867360,
+ "step": 25815,
+ "train_runtime": 199775.1172,
+ "train_tokens_per_second": 5650.691
+ },
+ {
+ "epoch": 1.7384863991381632,
+ "grad_norm": 0.8122569762079913,
+ "learning_rate": 4.1617224612064353e-07,
+ "loss": 0.29702818393707275,
+ "num_input_tokens_seen": 1129077456,
+ "step": 25820,
+ "train_runtime": 199815.7373,
+ "train_tokens_per_second": 5650.593
+ },
+ {
+ "epoch": 1.7388230541341234,
+ "grad_norm": 0.7151797283997641,
+ "learning_rate": 4.151167726144101e-07,
+ "loss": 0.2909691333770752,
+ "num_input_tokens_seen": 1129312768,
+ "step": 25825,
+ "train_runtime": 199850.2397,
+ "train_tokens_per_second": 5650.795
+ },
+ {
+ "epoch": 1.7391597091300834,
+ "grad_norm": 0.7870049200470505,
+ "learning_rate": 4.1406258125615106e-07,
+ "loss": 0.3070362567901611,
+ "num_input_tokens_seen": 1129537904,
+ "step": 25830,
+ "train_runtime": 199880.6637,
+ "train_tokens_per_second": 5651.061
+ },
+ {
+ "epoch": 1.7394963641260435,
+ "grad_norm": 0.6937152886038869,
+ "learning_rate": 4.130096723406674e-07,
+ "loss": 0.29037325382232665,
+ "num_input_tokens_seen": 1129757208,
+ "step": 25835,
+ "train_runtime": 199915.4723,
+ "train_tokens_per_second": 5651.174
+ },
+ {
+ "epoch": 1.7398330191220037,
+ "grad_norm": 0.8109854430003198,
+ "learning_rate": 4.119580461624023e-07,
+ "loss": 0.3026682615280151,
+ "num_input_tokens_seen": 1129966384,
+ "step": 25840,
+ "train_runtime": 199951.9393,
+ "train_tokens_per_second": 5651.19
+ },
+ {
+ "epoch": 1.740169674117964,
+ "grad_norm": 0.77895028602625,
+ "learning_rate": 4.109077030154396e-07,
+ "loss": 0.30320086479187014,
+ "num_input_tokens_seen": 1130186152,
+ "step": 25845,
+ "train_runtime": 199986.3251,
+ "train_tokens_per_second": 5651.317
+ },
+ {
+ "epoch": 1.740506329113924,
+ "grad_norm": 0.7121386019107564,
+ "learning_rate": 4.0985864319350366e-07,
+ "loss": 0.3050692081451416,
+ "num_input_tokens_seen": 1130406368,
+ "step": 25850,
+ "train_runtime": 200030.0945,
+ "train_tokens_per_second": 5651.181
+ },
+ {
+ "epoch": 1.740842984109884,
+ "grad_norm": 0.8089630935847927,
+ "learning_rate": 4.0881086698996154e-07,
+ "loss": 0.2959964513778687,
+ "num_input_tokens_seen": 1130629712,
+ "step": 25855,
+ "train_runtime": 200072.2089,
+ "train_tokens_per_second": 5651.108
+ },
+ {
+ "epoch": 1.7411796391058443,
+ "grad_norm": 0.6596855615199905,
+ "learning_rate": 4.077643746978194e-07,
+ "loss": 0.2970879316329956,
+ "num_input_tokens_seen": 1130854672,
+ "step": 25860,
+ "train_runtime": 200107.7191,
+ "train_tokens_per_second": 5651.23
+ },
+ {
+ "epoch": 1.7415162941018045,
+ "grad_norm": 0.7016863080677025,
+ "learning_rate": 4.067191666097281e-07,
+ "loss": 0.29431335926055907,
+ "num_input_tokens_seen": 1131071712,
+ "step": 25865,
+ "train_runtime": 200143.6703,
+ "train_tokens_per_second": 5651.299
+ },
+ {
+ "epoch": 1.7418529490977646,
+ "grad_norm": 0.721345869315641,
+ "learning_rate": 4.056752430179728e-07,
+ "loss": 0.28345699310302735,
+ "num_input_tokens_seen": 1131294176,
+ "step": 25870,
+ "train_runtime": 200187.8959,
+ "train_tokens_per_second": 5651.162
+ },
+ {
+ "epoch": 1.7421896040937246,
+ "grad_norm": 0.7756187950368285,
+ "learning_rate": 4.046326042144866e-07,
+ "loss": 0.2875986576080322,
+ "num_input_tokens_seen": 1131517072,
+ "step": 25875,
+ "train_runtime": 200228.1281,
+ "train_tokens_per_second": 5651.139
+ },
+ {
+ "epoch": 1.7425262590896848,
+ "grad_norm": 0.6831782860497421,
+ "learning_rate": 4.0359125049083956e-07,
+ "loss": 0.2836933135986328,
+ "num_input_tokens_seen": 1131734008,
+ "step": 25880,
+ "train_runtime": 200260.4735,
+ "train_tokens_per_second": 5651.31
+ },
+ {
+ "epoch": 1.742862914085645,
+ "grad_norm": 0.7787460128342357,
+ "learning_rate": 4.025511821382422e-07,
+ "loss": 0.30028643608093264,
+ "num_input_tokens_seen": 1131948944,
+ "step": 25885,
+ "train_runtime": 200306.9879,
+ "train_tokens_per_second": 5651.071
+ },
+ {
+ "epoch": 1.7431995690816051,
+ "grad_norm": 0.7541542349539324,
+ "learning_rate": 4.015123994475462e-07,
+ "loss": 0.28558173179626467,
+ "num_input_tokens_seen": 1132155112,
+ "step": 25890,
+ "train_runtime": 200346.4536,
+ "train_tokens_per_second": 5650.987
+ },
+ {
+ "epoch": 1.7435362240775654,
+ "grad_norm": 0.7540947221379003,
+ "learning_rate": 4.0047490270924417e-07,
+ "loss": 0.29801325798034667,
+ "num_input_tokens_seen": 1132391904,
+ "step": 25895,
+ "train_runtime": 200377.6526,
+ "train_tokens_per_second": 5651.288
+ },
+ {
+ "epoch": 1.7438728790735256,
+ "grad_norm": 0.7097292901302847,
+ "learning_rate": 3.994386922134691e-07,
+ "loss": 0.2968616247177124,
+ "num_input_tokens_seen": 1132613664,
+ "step": 25900,
+ "train_runtime": 200410.4186,
+ "train_tokens_per_second": 5651.471
+ },
+ {
+ "epoch": 1.7442095340694856,
+ "grad_norm": 0.6871287862369735,
+ "learning_rate": 3.984037682499936e-07,
+ "loss": 0.2899972438812256,
+ "num_input_tokens_seen": 1132836560,
+ "step": 25905,
+ "train_runtime": 200448.2196,
+ "train_tokens_per_second": 5651.517
+ },
+ {
+ "epoch": 1.7445461890654457,
+ "grad_norm": 0.7897783179799106,
+ "learning_rate": 3.973701311082301e-07,
+ "loss": 0.31725754737854006,
+ "num_input_tokens_seen": 1133050712,
+ "step": 25910,
+ "train_runtime": 200489.2159,
+ "train_tokens_per_second": 5651.43
+ },
+ {
+ "epoch": 1.744882844061406,
+ "grad_norm": 0.8896464404276014,
+ "learning_rate": 3.9633778107723454e-07,
+ "loss": 0.2900426387786865,
+ "num_input_tokens_seen": 1133254224,
+ "step": 25915,
+ "train_runtime": 200526.2526,
+ "train_tokens_per_second": 5651.401
+ },
+ {
+ "epoch": 1.7452194990573662,
+ "grad_norm": 0.7236782781112409,
+ "learning_rate": 3.953067184456966e-07,
+ "loss": 0.3090181350708008,
+ "num_input_tokens_seen": 1133470824,
+ "step": 25920,
+ "train_runtime": 200564.9033,
+ "train_tokens_per_second": 5651.392
+ },
+ {
+ "epoch": 1.7455561540533262,
+ "grad_norm": 0.8647141709585514,
+ "learning_rate": 3.9427694350195224e-07,
+ "loss": 0.31003541946411134,
+ "num_input_tokens_seen": 1133700592,
+ "step": 25925,
+ "train_runtime": 200601.8067,
+ "train_tokens_per_second": 5651.497
+ },
+ {
+ "epoch": 1.7458928090492862,
+ "grad_norm": 0.7588158649781692,
+ "learning_rate": 3.9324845653397446e-07,
+ "loss": 0.3075569152832031,
+ "num_input_tokens_seen": 1133925704,
+ "step": 25930,
+ "train_runtime": 200639.6339,
+ "train_tokens_per_second": 5651.554
+ },
+ {
+ "epoch": 1.7462294640452465,
+ "grad_norm": 0.7340500241842518,
+ "learning_rate": 3.922212578293755e-07,
+ "loss": 0.2702242374420166,
+ "num_input_tokens_seen": 1134140200,
+ "step": 25935,
+ "train_runtime": 200682.8182,
+ "train_tokens_per_second": 5651.407
+ },
+ {
+ "epoch": 1.7465661190412067,
+ "grad_norm": 0.82037212018234,
+ "learning_rate": 3.911953476754088e-07,
+ "loss": 0.2941112518310547,
+ "num_input_tokens_seen": 1134363752,
+ "step": 25940,
+ "train_runtime": 200715.6542,
+ "train_tokens_per_second": 5651.596
+ },
+ {
+ "epoch": 1.7469027740371668,
+ "grad_norm": 0.8812494234102816,
+ "learning_rate": 3.9017072635896716e-07,
+ "loss": 0.31668860912323,
+ "num_input_tokens_seen": 1134590888,
+ "step": 25945,
+ "train_runtime": 200756.942,
+ "train_tokens_per_second": 5651.565
+ },
+ {
+ "epoch": 1.7472394290331268,
+ "grad_norm": 0.76480797109403,
+ "learning_rate": 3.891473941665819e-07,
+ "loss": 0.3079123020172119,
+ "num_input_tokens_seen": 1134821944,
+ "step": 25950,
+ "train_runtime": 200795.9315,
+ "train_tokens_per_second": 5651.618
+ },
+ {
+ "epoch": 1.747576084029087,
+ "grad_norm": 0.8623020733146947,
+ "learning_rate": 3.881253513844246e-07,
+ "loss": 0.30869007110595703,
+ "num_input_tokens_seen": 1135026112,
+ "step": 25955,
+ "train_runtime": 200836.4857,
+ "train_tokens_per_second": 5651.494
+ },
+ {
+ "epoch": 1.7479127390250473,
+ "grad_norm": 0.7551876844288087,
+ "learning_rate": 3.871045982983063e-07,
+ "loss": 0.29468765258789065,
+ "num_input_tokens_seen": 1135250784,
+ "step": 25960,
+ "train_runtime": 200877.5576,
+ "train_tokens_per_second": 5651.457
+ },
+ {
+ "epoch": 1.7482493940210073,
+ "grad_norm": 0.7546664443489978,
+ "learning_rate": 3.8608513519367897e-07,
+ "loss": 0.2700496673583984,
+ "num_input_tokens_seen": 1135481416,
+ "step": 25965,
+ "train_runtime": 200922.6162,
+ "train_tokens_per_second": 5651.337
+ },
+ {
+ "epoch": 1.7485860490169673,
+ "grad_norm": 0.9863347976919302,
+ "learning_rate": 3.8506696235562915e-07,
+ "loss": 0.2968393087387085,
+ "num_input_tokens_seen": 1135693800,
+ "step": 25970,
+ "train_runtime": 200960.332,
+ "train_tokens_per_second": 5651.333
+ },
+ {
+ "epoch": 1.7489227040129276,
+ "grad_norm": 0.7481085355968223,
+ "learning_rate": 3.840500800688884e-07,
+ "loss": 0.2980025768280029,
+ "num_input_tokens_seen": 1135925744,
+ "step": 25975,
+ "train_runtime": 201006.8964,
+ "train_tokens_per_second": 5651.178
+ },
+ {
+ "epoch": 1.7492593590088878,
+ "grad_norm": 0.8214495132803978,
+ "learning_rate": 3.8303448861782333e-07,
+ "loss": 0.2952871322631836,
+ "num_input_tokens_seen": 1136130424,
+ "step": 25980,
+ "train_runtime": 201040.8482,
+ "train_tokens_per_second": 5651.242
+ },
+ {
+ "epoch": 1.7495960140048479,
+ "grad_norm": 0.8019257337743148,
+ "learning_rate": 3.8202018828644105e-07,
+ "loss": 0.27895636558532716,
+ "num_input_tokens_seen": 1136347560,
+ "step": 25985,
+ "train_runtime": 201079.8679,
+ "train_tokens_per_second": 5651.225
+ },
+ {
+ "epoch": 1.749932669000808,
+ "grad_norm": 0.9034770504035424,
+ "learning_rate": 3.8100717935838804e-07,
+ "loss": 0.32080807685852053,
+ "num_input_tokens_seen": 1136562648,
+ "step": 25990,
+ "train_runtime": 201121.1998,
+ "train_tokens_per_second": 5651.133
+ },
+ {
+ "epoch": 1.7502693239967682,
+ "grad_norm": 0.7869916616338762,
+ "learning_rate": 3.799954621169477e-07,
+ "loss": 0.27611637115478516,
+ "num_input_tokens_seen": 1136781352,
+ "step": 25995,
+ "train_runtime": 201160.3173,
+ "train_tokens_per_second": 5651.121
+ },
+ {
+ "epoch": 1.7506059789927284,
+ "grad_norm": 0.7949557024166627,
+ "learning_rate": 3.7898503684504483e-07,
+ "loss": 0.29705018997192384,
+ "num_input_tokens_seen": 1136994720,
+ "step": 26000,
+ "train_runtime": 201200.2712,
+ "train_tokens_per_second": 5651.06
+ },
+ {
+ "epoch": 1.7509426339886884,
+ "grad_norm": 0.7027696556054698,
+ "learning_rate": 3.7797590382524066e-07,
+ "loss": 0.28871927261352537,
+ "num_input_tokens_seen": 1137197696,
+ "step": 26005,
+ "train_runtime": 201242.4798,
+ "train_tokens_per_second": 5650.883
+ },
+ {
+ "epoch": 1.7512792889846485,
+ "grad_norm": 0.7864204996333114,
+ "learning_rate": 3.76968063339736e-07,
+ "loss": 0.3024580955505371,
+ "num_input_tokens_seen": 1137430488,
+ "step": 26010,
+ "train_runtime": 201274.5947,
+ "train_tokens_per_second": 5651.138
+ },
+ {
+ "epoch": 1.7516159439806087,
+ "grad_norm": 0.8310573898266609,
+ "learning_rate": 3.759615156703722e-07,
+ "loss": 0.28331038951873777,
+ "num_input_tokens_seen": 1137644120,
+ "step": 26015,
+ "train_runtime": 201312.2778,
+ "train_tokens_per_second": 5651.141
+ },
+ {
+ "epoch": 1.751952598976569,
+ "grad_norm": 0.7194041093377634,
+ "learning_rate": 3.749562610986235e-07,
+ "loss": 0.3102069854736328,
+ "num_input_tokens_seen": 1137861320,
+ "step": 26020,
+ "train_runtime": 201348.9701,
+ "train_tokens_per_second": 5651.19
+ },
+ {
+ "epoch": 1.752289253972529,
+ "grad_norm": 0.7426038956707869,
+ "learning_rate": 3.739522999056094e-07,
+ "loss": 0.28222415447235105,
+ "num_input_tokens_seen": 1138066152,
+ "step": 26025,
+ "train_runtime": 201385.6107,
+ "train_tokens_per_second": 5651.179
+ },
+ {
+ "epoch": 1.752625908968489,
+ "grad_norm": 0.7686985240205324,
+ "learning_rate": 3.7294963237208326e-07,
+ "loss": 0.3015716552734375,
+ "num_input_tokens_seen": 1138288896,
+ "step": 26030,
+ "train_runtime": 201430.5418,
+ "train_tokens_per_second": 5651.024
+ },
+ {
+ "epoch": 1.7529625639644493,
+ "grad_norm": 0.7939046805197044,
+ "learning_rate": 3.7194825877843786e-07,
+ "loss": 0.2880394458770752,
+ "num_input_tokens_seen": 1138513200,
+ "step": 26035,
+ "train_runtime": 201479.5121,
+ "train_tokens_per_second": 5650.764
+ },
+ {
+ "epoch": 1.7532992189604095,
+ "grad_norm": 0.7428618998986647,
+ "learning_rate": 3.7094817940470375e-07,
+ "loss": 0.30432782173156736,
+ "num_input_tokens_seen": 1138721632,
+ "step": 26040,
+ "train_runtime": 201520.1099,
+ "train_tokens_per_second": 5650.66
+ },
+ {
+ "epoch": 1.7536358739563696,
+ "grad_norm": 0.738527693119671,
+ "learning_rate": 3.6994939453055044e-07,
+ "loss": 0.2994675159454346,
+ "num_input_tokens_seen": 1138951136,
+ "step": 26045,
+ "train_runtime": 201561.9244,
+ "train_tokens_per_second": 5650.626
+ },
+ {
+ "epoch": 1.7539725289523296,
+ "grad_norm": 0.9322929614127222,
+ "learning_rate": 3.6895190443528515e-07,
+ "loss": 0.2943981170654297,
+ "num_input_tokens_seen": 1139175408,
+ "step": 26050,
+ "train_runtime": 201601.723,
+ "train_tokens_per_second": 5650.623
+ },
+ {
+ "epoch": 1.7543091839482898,
+ "grad_norm": 0.7070168137348763,
+ "learning_rate": 3.679557093978525e-07,
+ "loss": 0.312309193611145,
+ "num_input_tokens_seen": 1139396600,
+ "step": 26055,
+ "train_runtime": 201638.5926,
+ "train_tokens_per_second": 5650.687
+ },
+ {
+ "epoch": 1.75464583894425,
+ "grad_norm": 0.7986110214626218,
+ "learning_rate": 3.669608096968341e-07,
+ "loss": 0.29059772491455077,
+ "num_input_tokens_seen": 1139603752,
+ "step": 26060,
+ "train_runtime": 201681.4232,
+ "train_tokens_per_second": 5650.514
+ },
+ {
+ "epoch": 1.75498249394021,
+ "grad_norm": 0.8095452658769865,
+ "learning_rate": 3.6596720561045295e-07,
+ "loss": 0.2955317974090576,
+ "num_input_tokens_seen": 1139829360,
+ "step": 26065,
+ "train_runtime": 201718.99,
+ "train_tokens_per_second": 5650.58
+ },
+ {
+ "epoch": 1.7553191489361701,
+ "grad_norm": 0.791001274129923,
+ "learning_rate": 3.6497489741656455e-07,
+ "loss": 0.2989501476287842,
+ "num_input_tokens_seen": 1140053168,
+ "step": 26070,
+ "train_runtime": 201754.5113,
+ "train_tokens_per_second": 5650.695
+ },
+ {
+ "epoch": 1.7556558039321304,
+ "grad_norm": 0.6815029838821709,
+ "learning_rate": 3.639838853926669e-07,
+ "loss": 0.29015383720397947,
+ "num_input_tokens_seen": 1140284016,
+ "step": 26075,
+ "train_runtime": 201793.6507,
+ "train_tokens_per_second": 5650.743
+ },
+ {
+ "epoch": 1.7559924589280906,
+ "grad_norm": 0.7697158876317151,
+ "learning_rate": 3.629941698158923e-07,
+ "loss": 0.3135251998901367,
+ "num_input_tokens_seen": 1140503592,
+ "step": 26080,
+ "train_runtime": 201837.3895,
+ "train_tokens_per_second": 5650.606
+ },
+ {
+ "epoch": 1.7563291139240507,
+ "grad_norm": 0.6797973936849687,
+ "learning_rate": 3.6200575096301207e-07,
+ "loss": 0.2774399757385254,
+ "num_input_tokens_seen": 1140721216,
+ "step": 26085,
+ "train_runtime": 201873.4172,
+ "train_tokens_per_second": 5650.676
+ },
+ {
+ "epoch": 1.7566657689200107,
+ "grad_norm": 0.8393124553061458,
+ "learning_rate": 3.61018629110434e-07,
+ "loss": 0.32251529693603515,
+ "num_input_tokens_seen": 1140946680,
+ "step": 26090,
+ "train_runtime": 201907.6529,
+ "train_tokens_per_second": 5650.834
+ },
+ {
+ "epoch": 1.757002423915971,
+ "grad_norm": 0.7056858714134263,
+ "learning_rate": 3.6003280453420386e-07,
+ "loss": 0.2859219551086426,
+ "num_input_tokens_seen": 1141168176,
+ "step": 26095,
+ "train_runtime": 201940.7733,
+ "train_tokens_per_second": 5651.004
+ },
+ {
+ "epoch": 1.7573390789119312,
+ "grad_norm": 0.7740995437255472,
+ "learning_rate": 3.59048277510004e-07,
+ "loss": 0.28550260066986083,
+ "num_input_tokens_seen": 1141389904,
+ "step": 26100,
+ "train_runtime": 201981.2605,
+ "train_tokens_per_second": 5650.969
+ },
+ {
+ "epoch": 1.7576757339078912,
+ "grad_norm": 0.7658978202078977,
+ "learning_rate": 3.5806504831315423e-07,
+ "loss": 0.28779993057250974,
+ "num_input_tokens_seen": 1141608056,
+ "step": 26105,
+ "train_runtime": 202020.9745,
+ "train_tokens_per_second": 5650.938
+ },
+ {
+ "epoch": 1.7580123889038513,
+ "grad_norm": 0.7409514848212948,
+ "learning_rate": 3.570831172186112e-07,
+ "loss": 0.3048838138580322,
+ "num_input_tokens_seen": 1141832104,
+ "step": 26110,
+ "train_runtime": 202059.9812,
+ "train_tokens_per_second": 5650.956
+ },
+ {
+ "epoch": 1.7583490438998115,
+ "grad_norm": 0.7982423185812808,
+ "learning_rate": 3.5610248450097084e-07,
+ "loss": 0.30502030849456785,
+ "num_input_tokens_seen": 1142049928,
+ "step": 26115,
+ "train_runtime": 202100.022,
+ "train_tokens_per_second": 5650.914
+ },
+ {
+ "epoch": 1.7586856988957718,
+ "grad_norm": 0.7105646801738283,
+ "learning_rate": 3.551231504344604e-07,
+ "loss": 0.2915775775909424,
+ "num_input_tokens_seen": 1142270008,
+ "step": 26120,
+ "train_runtime": 202138.5884,
+ "train_tokens_per_second": 5650.925
+ },
+ {
+ "epoch": 1.7590223538917318,
+ "grad_norm": 0.6784969862052538,
+ "learning_rate": 3.5414511529295036e-07,
+ "loss": 0.30109832286834715,
+ "num_input_tokens_seen": 1142484296,
+ "step": 26125,
+ "train_runtime": 202171.0989,
+ "train_tokens_per_second": 5651.076
+ },
+ {
+ "epoch": 1.7593590088876918,
+ "grad_norm": 0.8313599079016897,
+ "learning_rate": 3.5316837934994465e-07,
+ "loss": 0.3126358509063721,
+ "num_input_tokens_seen": 1142703184,
+ "step": 26130,
+ "train_runtime": 202211.0287,
+ "train_tokens_per_second": 5651.043
+ },
+ {
+ "epoch": 1.759695663883652,
+ "grad_norm": 0.7462437378743789,
+ "learning_rate": 3.521929428785836e-07,
+ "loss": 0.3012665271759033,
+ "num_input_tokens_seen": 1142932000,
+ "step": 26135,
+ "train_runtime": 202249.327,
+ "train_tokens_per_second": 5651.104
+ },
+ {
+ "epoch": 1.7600323188796123,
+ "grad_norm": 0.7707749450724266,
+ "learning_rate": 3.5121880615164584e-07,
+ "loss": 0.3177218198776245,
+ "num_input_tokens_seen": 1143147936,
+ "step": 26140,
+ "train_runtime": 202288.9562,
+ "train_tokens_per_second": 5651.064
+ },
+ {
+ "epoch": 1.7603689738755723,
+ "grad_norm": 0.7201463000961772,
+ "learning_rate": 3.5024596944154445e-07,
+ "loss": 0.30140066146850586,
+ "num_input_tokens_seen": 1143365592,
+ "step": 26145,
+ "train_runtime": 202327.958,
+ "train_tokens_per_second": 5651.051
+ },
+ {
+ "epoch": 1.7607056288715324,
+ "grad_norm": 0.7940935130035773,
+ "learning_rate": 3.4927443302033127e-07,
+ "loss": 0.2727382659912109,
+ "num_input_tokens_seen": 1143591488,
+ "step": 26150,
+ "train_runtime": 202375.4812,
+ "train_tokens_per_second": 5650.84
+ },
+ {
+ "epoch": 1.7610422838674926,
+ "grad_norm": 0.7992454946312759,
+ "learning_rate": 3.4830419715969233e-07,
+ "loss": 0.2696976900100708,
+ "num_input_tokens_seen": 1143804104,
+ "step": 26155,
+ "train_runtime": 202410.9519,
+ "train_tokens_per_second": 5650.9
+ },
+ {
+ "epoch": 1.7613789388634529,
+ "grad_norm": 0.8375508909384706,
+ "learning_rate": 3.4733526213095114e-07,
+ "loss": 0.30320181846618655,
+ "num_input_tokens_seen": 1144024536,
+ "step": 26160,
+ "train_runtime": 202453.0452,
+ "train_tokens_per_second": 5650.814
+ },
+ {
+ "epoch": 1.761715593859413,
+ "grad_norm": 0.8560064029859269,
+ "learning_rate": 3.4636762820506876e-07,
+ "loss": 0.3035079002380371,
+ "num_input_tokens_seen": 1144244512,
+ "step": 26165,
+ "train_runtime": 202492.5015,
+ "train_tokens_per_second": 5650.799
+ },
+ {
+ "epoch": 1.762052248855373,
+ "grad_norm": 0.8794486904268961,
+ "learning_rate": 3.4540129565263867e-07,
+ "loss": 0.30885040760040283,
+ "num_input_tokens_seen": 1144457424,
+ "step": 26170,
+ "train_runtime": 202528.6337,
+ "train_tokens_per_second": 5650.843
+ },
+ {
+ "epoch": 1.7623889038513332,
+ "grad_norm": 0.7747667521193755,
+ "learning_rate": 3.444362647438942e-07,
+ "loss": 0.30580987930297854,
+ "num_input_tokens_seen": 1144685136,
+ "step": 26175,
+ "train_runtime": 202566.2912,
+ "train_tokens_per_second": 5650.916
+ },
+ {
+ "epoch": 1.7627255588472934,
+ "grad_norm": 0.749987526461497,
+ "learning_rate": 3.434725357487029e-07,
+ "loss": 0.29648699760437014,
+ "num_input_tokens_seen": 1144912488,
+ "step": 26180,
+ "train_runtime": 202608.8203,
+ "train_tokens_per_second": 5650.852
+ },
+ {
+ "epoch": 1.7630622138432535,
+ "grad_norm": 0.7973029748008016,
+ "learning_rate": 3.4251010893656844e-07,
+ "loss": 0.3241448402404785,
+ "num_input_tokens_seen": 1145132240,
+ "step": 26185,
+ "train_runtime": 202646.9809,
+ "train_tokens_per_second": 5650.872
+ },
+ {
+ "epoch": 1.7633988688392135,
+ "grad_norm": 0.8079881889044294,
+ "learning_rate": 3.4154898457663065e-07,
+ "loss": 0.2964663505554199,
+ "num_input_tokens_seen": 1145339072,
+ "step": 26190,
+ "train_runtime": 202684.9934,
+ "train_tokens_per_second": 5650.833
+ },
+ {
+ "epoch": 1.7637355238351737,
+ "grad_norm": 0.74358995342413,
+ "learning_rate": 3.40589162937664e-07,
+ "loss": 0.2934283256530762,
+ "num_input_tokens_seen": 1145537280,
+ "step": 26195,
+ "train_runtime": 202723.2034,
+ "train_tokens_per_second": 5650.746
+ },
+ {
+ "epoch": 1.764072178831134,
+ "grad_norm": 0.7784930364611832,
+ "learning_rate": 3.3963064428808034e-07,
+ "loss": 0.2671190738677979,
+ "num_input_tokens_seen": 1145744432,
+ "step": 26200,
+ "train_runtime": 202765.8103,
+ "train_tokens_per_second": 5650.58
+ },
+ {
+ "epoch": 1.764408833827094,
+ "grad_norm": 0.6964223743604306,
+ "learning_rate": 3.386734288959265e-07,
+ "loss": 0.29965527057647706,
+ "num_input_tokens_seen": 1145941000,
+ "step": 26205,
+ "train_runtime": 202807.8719,
+ "train_tokens_per_second": 5650.377
+ },
+ {
+ "epoch": 1.764745488823054,
+ "grad_norm": 0.7356574014692101,
+ "learning_rate": 3.377175170288838e-07,
+ "loss": 0.3009481430053711,
+ "num_input_tokens_seen": 1146164576,
+ "step": 26210,
+ "train_runtime": 202841.3847,
+ "train_tokens_per_second": 5650.546
+ },
+ {
+ "epoch": 1.7650821438190143,
+ "grad_norm": 0.7941269061669325,
+ "learning_rate": 3.367629089542712e-07,
+ "loss": 0.2912009239196777,
+ "num_input_tokens_seen": 1146397144,
+ "step": 26215,
+ "train_runtime": 202889.1986,
+ "train_tokens_per_second": 5650.361
+ },
+ {
+ "epoch": 1.7654187988149745,
+ "grad_norm": 0.7636350648934562,
+ "learning_rate": 3.3580960493904024e-07,
+ "loss": 0.3059426546096802,
+ "num_input_tokens_seen": 1146616808,
+ "step": 26220,
+ "train_runtime": 202930.3093,
+ "train_tokens_per_second": 5650.298
+ },
+ {
+ "epoch": 1.7657554538109346,
+ "grad_norm": 0.7999400483263438,
+ "learning_rate": 3.3485760524978093e-07,
+ "loss": 0.31472842693328856,
+ "num_input_tokens_seen": 1146822992,
+ "step": 26225,
+ "train_runtime": 202962.5617,
+ "train_tokens_per_second": 5650.416
+ },
+ {
+ "epoch": 1.7660921088068946,
+ "grad_norm": 0.826503171384712,
+ "learning_rate": 3.3390691015271646e-07,
+ "loss": 0.29501938819885254,
+ "num_input_tokens_seen": 1147041408,
+ "step": 26230,
+ "train_runtime": 202998.4752,
+ "train_tokens_per_second": 5650.493
+ },
+ {
+ "epoch": 1.7664287638028549,
+ "grad_norm": 0.7752263216759548,
+ "learning_rate": 3.329575199137053e-07,
+ "loss": 0.2826894521713257,
+ "num_input_tokens_seen": 1147263888,
+ "step": 26235,
+ "train_runtime": 203037.2378,
+ "train_tokens_per_second": 5650.51
+ },
+ {
+ "epoch": 1.766765418798815,
+ "grad_norm": 0.8585013195673774,
+ "learning_rate": 3.3200943479824113e-07,
+ "loss": 0.327055025100708,
+ "num_input_tokens_seen": 1147480536,
+ "step": 26240,
+ "train_runtime": 203079.6569,
+ "train_tokens_per_second": 5650.396
+ },
+ {
+ "epoch": 1.7671020737947751,
+ "grad_norm": 0.6517474645162997,
+ "learning_rate": 3.3106265507145406e-07,
+ "loss": 0.2585421562194824,
+ "num_input_tokens_seen": 1147695160,
+ "step": 26245,
+ "train_runtime": 203115.9008,
+ "train_tokens_per_second": 5650.445
+ },
+ {
+ "epoch": 1.7674387287907352,
+ "grad_norm": 0.7376728837137707,
+ "learning_rate": 3.3011718099810687e-07,
+ "loss": 0.28537108898162844,
+ "num_input_tokens_seen": 1147904984,
+ "step": 26250,
+ "train_runtime": 203155.1468,
+ "train_tokens_per_second": 5650.386
+ },
+ {
+ "epoch": 1.7677753837866954,
+ "grad_norm": 0.7607253054674292,
+ "learning_rate": 3.2917301284259904e-07,
+ "loss": 0.3015407085418701,
+ "num_input_tokens_seen": 1148133136,
+ "step": 26255,
+ "train_runtime": 203195.8171,
+ "train_tokens_per_second": 5650.378
+ },
+ {
+ "epoch": 1.7681120387826557,
+ "grad_norm": 0.7368433932513038,
+ "learning_rate": 3.2823015086896335e-07,
+ "loss": 0.31117079257965086,
+ "num_input_tokens_seen": 1148351808,
+ "step": 26260,
+ "train_runtime": 203231.7124,
+ "train_tokens_per_second": 5650.456
+ },
+ {
+ "epoch": 1.7684486937786157,
+ "grad_norm": 0.7455974069716451,
+ "learning_rate": 3.2728859534087e-07,
+ "loss": 0.28980340957641604,
+ "num_input_tokens_seen": 1148578048,
+ "step": 26265,
+ "train_runtime": 203270.0794,
+ "train_tokens_per_second": 5650.502
+ },
+ {
+ "epoch": 1.7687853487745757,
+ "grad_norm": 0.726415832314535,
+ "learning_rate": 3.263483465216194e-07,
+ "loss": 0.30661208629608155,
+ "num_input_tokens_seen": 1148808400,
+ "step": 26270,
+ "train_runtime": 203304.8486,
+ "train_tokens_per_second": 5650.669
+ },
+ {
+ "epoch": 1.769122003770536,
+ "grad_norm": 0.7198274225186322,
+ "learning_rate": 3.2540940467415126e-07,
+ "loss": 0.29364948272705077,
+ "num_input_tokens_seen": 1149026320,
+ "step": 26275,
+ "train_runtime": 203344.2573,
+ "train_tokens_per_second": 5650.646
+ },
+ {
+ "epoch": 1.7694586587664962,
+ "grad_norm": 0.675379666160097,
+ "learning_rate": 3.2447177006103726e-07,
+ "loss": 0.29972095489501954,
+ "num_input_tokens_seen": 1149241504,
+ "step": 26280,
+ "train_runtime": 203385.3582,
+ "train_tokens_per_second": 5650.562
+ },
+ {
+ "epoch": 1.7697953137624562,
+ "grad_norm": 0.7180951826553738,
+ "learning_rate": 3.235354429444831e-07,
+ "loss": 0.29863905906677246,
+ "num_input_tokens_seen": 1149442960,
+ "step": 26285,
+ "train_runtime": 203420.2428,
+ "train_tokens_per_second": 5650.583
+ },
+ {
+ "epoch": 1.7701319687584163,
+ "grad_norm": 0.7550877294810541,
+ "learning_rate": 3.226004235863306e-07,
+ "loss": 0.3004556655883789,
+ "num_input_tokens_seen": 1149664768,
+ "step": 26290,
+ "train_runtime": 203464.2499,
+ "train_tokens_per_second": 5650.451
+ },
+ {
+ "epoch": 1.7704686237543765,
+ "grad_norm": 0.7735355193352721,
+ "learning_rate": 3.2166671224805437e-07,
+ "loss": 0.2912909984588623,
+ "num_input_tokens_seen": 1149894072,
+ "step": 26295,
+ "train_runtime": 203504.8957,
+ "train_tokens_per_second": 5650.449
+ },
+ {
+ "epoch": 1.7708052787503368,
+ "grad_norm": 0.7860723514589903,
+ "learning_rate": 3.207343091907633e-07,
+ "loss": 0.28990769386291504,
+ "num_input_tokens_seen": 1150100408,
+ "step": 26300,
+ "train_runtime": 203547.7921,
+ "train_tokens_per_second": 5650.272
+ },
+ {
+ "epoch": 1.7711419337462968,
+ "grad_norm": 0.8144219098196478,
+ "learning_rate": 3.1980321467520226e-07,
+ "loss": 0.30245256423950195,
+ "num_input_tokens_seen": 1150300712,
+ "step": 26305,
+ "train_runtime": 203587.3111,
+ "train_tokens_per_second": 5650.159
+ },
+ {
+ "epoch": 1.7714785887422568,
+ "grad_norm": 0.8081994567096525,
+ "learning_rate": 3.1887342896174735e-07,
+ "loss": 0.31411175727844237,
+ "num_input_tokens_seen": 1150501592,
+ "step": 26310,
+ "train_runtime": 203630.5459,
+ "train_tokens_per_second": 5649.946
+ },
+ {
+ "epoch": 1.771815243738217,
+ "grad_norm": 0.6657629683425634,
+ "learning_rate": 3.1794495231041276e-07,
+ "loss": 0.28406760692596433,
+ "num_input_tokens_seen": 1150727232,
+ "step": 26315,
+ "train_runtime": 203671.5283,
+ "train_tokens_per_second": 5649.917
+ },
+ {
+ "epoch": 1.7721518987341773,
+ "grad_norm": 0.7568239836133204,
+ "learning_rate": 3.1701778498084035e-07,
+ "loss": 0.3084291458129883,
+ "num_input_tokens_seen": 1150932320,
+ "step": 26320,
+ "train_runtime": 203712.8008,
+ "train_tokens_per_second": 5649.779
+ },
+ {
+ "epoch": 1.7724885537301374,
+ "grad_norm": 0.7723629219802286,
+ "learning_rate": 3.160919272323121e-07,
+ "loss": 0.3075510263442993,
+ "num_input_tokens_seen": 1151162984,
+ "step": 26325,
+ "train_runtime": 203750.2121,
+ "train_tokens_per_second": 5649.874
+ },
+ {
+ "epoch": 1.7728252087260974,
+ "grad_norm": 0.7180286793509041,
+ "learning_rate": 3.1516737932374087e-07,
+ "loss": 0.30257492065429686,
+ "num_input_tokens_seen": 1151371272,
+ "step": 26330,
+ "train_runtime": 203784.7687,
+ "train_tokens_per_second": 5649.938
+ },
+ {
+ "epoch": 1.7731618637220576,
+ "grad_norm": 0.7988698082761704,
+ "learning_rate": 3.142441415136727e-07,
+ "loss": 0.29022831916809083,
+ "num_input_tokens_seen": 1151601408,
+ "step": 26335,
+ "train_runtime": 203817.7911,
+ "train_tokens_per_second": 5650.152
+ },
+ {
+ "epoch": 1.773498518718018,
+ "grad_norm": 0.7403905711500696,
+ "learning_rate": 3.133222140602893e-07,
+ "loss": 0.2794740438461304,
+ "num_input_tokens_seen": 1151807024,
+ "step": 26340,
+ "train_runtime": 203851.848,
+ "train_tokens_per_second": 5650.216
+ },
+ {
+ "epoch": 1.773835173713978,
+ "grad_norm": 0.7358244297636568,
+ "learning_rate": 3.124015972214034e-07,
+ "loss": 0.30714993476867675,
+ "num_input_tokens_seen": 1152030904,
+ "step": 26345,
+ "train_runtime": 203880.8052,
+ "train_tokens_per_second": 5650.512
+ },
+ {
+ "epoch": 1.774171828709938,
+ "grad_norm": 0.7987243174003847,
+ "learning_rate": 3.114822912544635e-07,
+ "loss": 0.31725311279296875,
+ "num_input_tokens_seen": 1152240824,
+ "step": 26350,
+ "train_runtime": 203921.8118,
+ "train_tokens_per_second": 5650.405
+ },
+ {
+ "epoch": 1.7745084837058982,
+ "grad_norm": 0.7884531597674449,
+ "learning_rate": 3.1056429641655007e-07,
+ "loss": 0.2949615716934204,
+ "num_input_tokens_seen": 1152447328,
+ "step": 26355,
+ "train_runtime": 203963.8804,
+ "train_tokens_per_second": 5650.252
+ },
+ {
+ "epoch": 1.7748451387018584,
+ "grad_norm": 0.8631624259760383,
+ "learning_rate": 3.0964761296437665e-07,
+ "loss": 0.2921284198760986,
+ "num_input_tokens_seen": 1152671024,
+ "step": 26360,
+ "train_runtime": 204005.1311,
+ "train_tokens_per_second": 5650.206
+ },
+ {
+ "epoch": 1.7751817936978185,
+ "grad_norm": 0.7806230963793127,
+ "learning_rate": 3.087322411542937e-07,
+ "loss": 0.3069296836853027,
+ "num_input_tokens_seen": 1152898216,
+ "step": 26365,
+ "train_runtime": 204043.7868,
+ "train_tokens_per_second": 5650.249
+ },
+ {
+ "epoch": 1.7755184486937785,
+ "grad_norm": 0.8045496935454175,
+ "learning_rate": 3.0781818124227867e-07,
+ "loss": 0.27917664051055907,
+ "num_input_tokens_seen": 1153109064,
+ "step": 26370,
+ "train_runtime": 204083.2918,
+ "train_tokens_per_second": 5650.188
+ },
+ {
+ "epoch": 1.7758551036897388,
+ "grad_norm": 0.7628962650971699,
+ "learning_rate": 3.0690543348394776e-07,
+ "loss": 0.2973165988922119,
+ "num_input_tokens_seen": 1153332104,
+ "step": 26375,
+ "train_runtime": 204124.6151,
+ "train_tokens_per_second": 5650.137
+ },
+ {
+ "epoch": 1.776191758685699,
+ "grad_norm": 0.735769438564152,
+ "learning_rate": 3.059939981345467e-07,
+ "loss": 0.2835052490234375,
+ "num_input_tokens_seen": 1153547680,
+ "step": 26380,
+ "train_runtime": 204153.8046,
+ "train_tokens_per_second": 5650.385
+ },
+ {
+ "epoch": 1.776528413681659,
+ "grad_norm": 0.7910223938595813,
+ "learning_rate": 3.050838754489566e-07,
+ "loss": 0.28968124389648436,
+ "num_input_tokens_seen": 1153764144,
+ "step": 26385,
+ "train_runtime": 204197.8486,
+ "train_tokens_per_second": 5650.227
+ },
+ {
+ "epoch": 1.776865068677619,
+ "grad_norm": 0.7189107694256088,
+ "learning_rate": 3.041750656816894e-07,
+ "loss": 0.29620354175567626,
+ "num_input_tokens_seen": 1153982600,
+ "step": 26390,
+ "train_runtime": 204235.7042,
+ "train_tokens_per_second": 5650.249
+ },
+ {
+ "epoch": 1.7772017236735793,
+ "grad_norm": 0.8335040707060406,
+ "learning_rate": 3.0326756908689135e-07,
+ "loss": 0.3176565647125244,
+ "num_input_tokens_seen": 1154204320,
+ "step": 26395,
+ "train_runtime": 204274.2047,
+ "train_tokens_per_second": 5650.27
+ },
+ {
+ "epoch": 1.7775383786695396,
+ "grad_norm": 0.687809411831832,
+ "learning_rate": 3.0236138591833995e-07,
+ "loss": 0.2815841197967529,
+ "num_input_tokens_seen": 1154426800,
+ "step": 26400,
+ "train_runtime": 204312.8632,
+ "train_tokens_per_second": 5650.289
+ },
+ {
+ "epoch": 1.7778750336654996,
+ "grad_norm": 0.7809149749160714,
+ "learning_rate": 3.0145651642944863e-07,
+ "loss": 0.31723904609680176,
+ "num_input_tokens_seen": 1154656600,
+ "step": 26405,
+ "train_runtime": 204357.2069,
+ "train_tokens_per_second": 5650.188
+ },
+ {
+ "epoch": 1.7782116886614596,
+ "grad_norm": 0.739025964030064,
+ "learning_rate": 3.005529608732588e-07,
+ "loss": 0.26880898475646975,
+ "num_input_tokens_seen": 1154869552,
+ "step": 26410,
+ "train_runtime": 204394.2963,
+ "train_tokens_per_second": 5650.204
+ },
+ {
+ "epoch": 1.7785483436574199,
+ "grad_norm": 0.7646452369868884,
+ "learning_rate": 2.996507195024495e-07,
+ "loss": 0.322553014755249,
+ "num_input_tokens_seen": 1155101520,
+ "step": 26415,
+ "train_runtime": 204435.1594,
+ "train_tokens_per_second": 5650.21
+ },
+ {
+ "epoch": 1.7788849986533801,
+ "grad_norm": 0.8474473108460225,
+ "learning_rate": 2.9874979256932614e-07,
+ "loss": 0.2975711107254028,
+ "num_input_tokens_seen": 1155322552,
+ "step": 26420,
+ "train_runtime": 204473.0585,
+ "train_tokens_per_second": 5650.243
+ },
+ {
+ "epoch": 1.7792216536493402,
+ "grad_norm": 0.8062221329371094,
+ "learning_rate": 2.9785018032583325e-07,
+ "loss": 0.3099265813827515,
+ "num_input_tokens_seen": 1155544168,
+ "step": 26425,
+ "train_runtime": 204516.1566,
+ "train_tokens_per_second": 5650.136
+ },
+ {
+ "epoch": 1.7795583086453002,
+ "grad_norm": 0.8324573418937758,
+ "learning_rate": 2.969518830235435e-07,
+ "loss": 0.298630952835083,
+ "num_input_tokens_seen": 1155762856,
+ "step": 26430,
+ "train_runtime": 204549.2337,
+ "train_tokens_per_second": 5650.292
+ },
+ {
+ "epoch": 1.7798949636412604,
+ "grad_norm": 0.7975157170801572,
+ "learning_rate": 2.960549009136626e-07,
+ "loss": 0.30402772426605223,
+ "num_input_tokens_seen": 1155983360,
+ "step": 26435,
+ "train_runtime": 204590.4666,
+ "train_tokens_per_second": 5650.231
+ },
+ {
+ "epoch": 1.7802316186372207,
+ "grad_norm": 0.8344419171870143,
+ "learning_rate": 2.951592342470289e-07,
+ "loss": 0.33509268760681155,
+ "num_input_tokens_seen": 1156211056,
+ "step": 26440,
+ "train_runtime": 204625.1719,
+ "train_tokens_per_second": 5650.385
+ },
+ {
+ "epoch": 1.7805682736331807,
+ "grad_norm": 0.7399401765355089,
+ "learning_rate": 2.942648832741124e-07,
+ "loss": 0.28286046981811525,
+ "num_input_tokens_seen": 1156416184,
+ "step": 26445,
+ "train_runtime": 204663.2394,
+ "train_tokens_per_second": 5650.337
+ },
+ {
+ "epoch": 1.7809049286291407,
+ "grad_norm": 0.6734864094384185,
+ "learning_rate": 2.933718482450154e-07,
+ "loss": 0.2894281387329102,
+ "num_input_tokens_seen": 1156635048,
+ "step": 26450,
+ "train_runtime": 204698.0899,
+ "train_tokens_per_second": 5650.444
+ },
+ {
+ "epoch": 1.781241583625101,
+ "grad_norm": 0.8706400424083963,
+ "learning_rate": 2.9248012940947423e-07,
+ "loss": 0.3163520097732544,
+ "num_input_tokens_seen": 1156864984,
+ "step": 26455,
+ "train_runtime": 204729.4569,
+ "train_tokens_per_second": 5650.701
+ },
+ {
+ "epoch": 1.7815782386210612,
+ "grad_norm": 0.7638164398502649,
+ "learning_rate": 2.915897270168522e-07,
+ "loss": 0.2976935625076294,
+ "num_input_tokens_seen": 1157071688,
+ "step": 26460,
+ "train_runtime": 204767.7005,
+ "train_tokens_per_second": 5650.655
+ },
+ {
+ "epoch": 1.7819148936170213,
+ "grad_norm": 0.8606236851862529,
+ "learning_rate": 2.9070064131615004e-07,
+ "loss": 0.30093798637390134,
+ "num_input_tokens_seen": 1157281584,
+ "step": 26465,
+ "train_runtime": 204809.9681,
+ "train_tokens_per_second": 5650.514
+ },
+ {
+ "epoch": 1.7822515486129813,
+ "grad_norm": 0.8072069252667485,
+ "learning_rate": 2.8981287255599567e-07,
+ "loss": 0.291579532623291,
+ "num_input_tokens_seen": 1157474720,
+ "step": 26470,
+ "train_runtime": 204843.9736,
+ "train_tokens_per_second": 5650.519
+ },
+ {
+ "epoch": 1.7825882036089415,
+ "grad_norm": 0.7787592014629802,
+ "learning_rate": 2.8892642098465205e-07,
+ "loss": 0.2964937210083008,
+ "num_input_tokens_seen": 1157694912,
+ "step": 26475,
+ "train_runtime": 204887.0752,
+ "train_tokens_per_second": 5650.405
+ },
+ {
+ "epoch": 1.7829248586049018,
+ "grad_norm": 0.741114575623905,
+ "learning_rate": 2.8804128685001263e-07,
+ "loss": 0.29356560707092283,
+ "num_input_tokens_seen": 1157923424,
+ "step": 26480,
+ "train_runtime": 204930.4827,
+ "train_tokens_per_second": 5650.323
+ },
+ {
+ "epoch": 1.7832615136008618,
+ "grad_norm": 0.8320550507276643,
+ "learning_rate": 2.8715747039960216e-07,
+ "loss": 0.2721066474914551,
+ "num_input_tokens_seen": 1158134336,
+ "step": 26485,
+ "train_runtime": 204962.95,
+ "train_tokens_per_second": 5650.457
+ },
+ {
+ "epoch": 1.7835981685968219,
+ "grad_norm": 0.8218511157843951,
+ "learning_rate": 2.8627497188057684e-07,
+ "loss": 0.30223267078399657,
+ "num_input_tokens_seen": 1158358632,
+ "step": 26490,
+ "train_runtime": 205002.9913,
+ "train_tokens_per_second": 5650.447
+ },
+ {
+ "epoch": 1.783934823592782,
+ "grad_norm": 0.7711493227331807,
+ "learning_rate": 2.853937915397248e-07,
+ "loss": 0.32183716297149656,
+ "num_input_tokens_seen": 1158575032,
+ "step": 26495,
+ "train_runtime": 205041.1723,
+ "train_tokens_per_second": 5650.451
+ },
+ {
+ "epoch": 1.7842714785887424,
+ "grad_norm": 0.7845009644246281,
+ "learning_rate": 2.845139296234639e-07,
+ "loss": 0.28379087448120116,
+ "num_input_tokens_seen": 1158803496,
+ "step": 26500,
+ "train_runtime": 205077.698,
+ "train_tokens_per_second": 5650.558
+ },
+ {
+ "epoch": 1.7846081335847024,
+ "grad_norm": 0.6932426732560467,
+ "learning_rate": 2.836353863778479e-07,
+ "loss": 0.29067995548248293,
+ "num_input_tokens_seen": 1159023072,
+ "step": 26505,
+ "train_runtime": 205114.9773,
+ "train_tokens_per_second": 5650.602
+ },
+ {
+ "epoch": 1.7849447885806624,
+ "grad_norm": 0.66733187111737,
+ "learning_rate": 2.8275816204855534e-07,
+ "loss": 0.3040937900543213,
+ "num_input_tokens_seen": 1159239184,
+ "step": 26510,
+ "train_runtime": 205163.8168,
+ "train_tokens_per_second": 5650.31
+ },
+ {
+ "epoch": 1.7852814435766227,
+ "grad_norm": 0.8094709155559127,
+ "learning_rate": 2.818822568809015e-07,
+ "loss": 0.31920318603515624,
+ "num_input_tokens_seen": 1159445280,
+ "step": 26515,
+ "train_runtime": 205199.8163,
+ "train_tokens_per_second": 5650.323
+ },
+ {
+ "epoch": 1.785618098572583,
+ "grad_norm": 0.7397186499607165,
+ "learning_rate": 2.810076711198284e-07,
+ "loss": 0.2857333183288574,
+ "num_input_tokens_seen": 1159658816,
+ "step": 26520,
+ "train_runtime": 205238.5515,
+ "train_tokens_per_second": 5650.297
+ },
+ {
+ "epoch": 1.785954753568543,
+ "grad_norm": 0.7573212272405271,
+ "learning_rate": 2.801344050099131e-07,
+ "loss": 0.3130965709686279,
+ "num_input_tokens_seen": 1159896520,
+ "step": 26525,
+ "train_runtime": 205275.2217,
+ "train_tokens_per_second": 5650.446
+ },
+ {
+ "epoch": 1.786291408564503,
+ "grad_norm": 0.7444021852294227,
+ "learning_rate": 2.7926245879536076e-07,
+ "loss": 0.3018080711364746,
+ "num_input_tokens_seen": 1160116400,
+ "step": 26530,
+ "train_runtime": 205318.3954,
+ "train_tokens_per_second": 5650.329
+ },
+ {
+ "epoch": 1.7866280635604632,
+ "grad_norm": 0.7159851006520612,
+ "learning_rate": 2.7839183272000857e-07,
+ "loss": 0.28679165840148924,
+ "num_input_tokens_seen": 1160350120,
+ "step": 26535,
+ "train_runtime": 205353.8407,
+ "train_tokens_per_second": 5650.491
+ },
+ {
+ "epoch": 1.7869647185564235,
+ "grad_norm": 0.9695258135131855,
+ "learning_rate": 2.775225270273241e-07,
+ "loss": 0.30443391799926756,
+ "num_input_tokens_seen": 1160577328,
+ "step": 26540,
+ "train_runtime": 205388.8223,
+ "train_tokens_per_second": 5650.635
+ },
+ {
+ "epoch": 1.7873013735523835,
+ "grad_norm": 0.6888517858579661,
+ "learning_rate": 2.7665454196040665e-07,
+ "loss": 0.2736891508102417,
+ "num_input_tokens_seen": 1160803320,
+ "step": 26545,
+ "train_runtime": 205426.5857,
+ "train_tokens_per_second": 5650.697
+ },
+ {
+ "epoch": 1.7876380285483435,
+ "grad_norm": 0.8050232282526216,
+ "learning_rate": 2.7578787776198325e-07,
+ "loss": 0.31282496452331543,
+ "num_input_tokens_seen": 1161013768,
+ "step": 26550,
+ "train_runtime": 205464.4875,
+ "train_tokens_per_second": 5650.679
+ },
+ {
+ "epoch": 1.7879746835443038,
+ "grad_norm": 0.7051417852744916,
+ "learning_rate": 2.7492253467441775e-07,
+ "loss": 0.2743107318878174,
+ "num_input_tokens_seen": 1161234120,
+ "step": 26555,
+ "train_runtime": 205500.8557,
+ "train_tokens_per_second": 5650.751
+ },
+ {
+ "epoch": 1.788311338540264,
+ "grad_norm": 0.7282394673289367,
+ "learning_rate": 2.740585129396961e-07,
+ "loss": 0.30170412063598634,
+ "num_input_tokens_seen": 1161458560,
+ "step": 26560,
+ "train_runtime": 205545.0743,
+ "train_tokens_per_second": 5650.627
+ },
+ {
+ "epoch": 1.788647993536224,
+ "grad_norm": 0.7602917600626389,
+ "learning_rate": 2.731958127994433e-07,
+ "loss": 0.2769881248474121,
+ "num_input_tokens_seen": 1161674768,
+ "step": 26565,
+ "train_runtime": 205586.7314,
+ "train_tokens_per_second": 5650.534
+ },
+ {
+ "epoch": 1.788984648532184,
+ "grad_norm": 0.8213091294594603,
+ "learning_rate": 2.7233443449490695e-07,
+ "loss": 0.307672905921936,
+ "num_input_tokens_seen": 1161890360,
+ "step": 26570,
+ "train_runtime": 205630.1097,
+ "train_tokens_per_second": 5650.39
+ },
+ {
+ "epoch": 1.7893213035281443,
+ "grad_norm": 0.7931628273831316,
+ "learning_rate": 2.7147437826697153e-07,
+ "loss": 0.29656333923339845,
+ "num_input_tokens_seen": 1162107656,
+ "step": 26575,
+ "train_runtime": 205671.7811,
+ "train_tokens_per_second": 5650.302
+ },
+ {
+ "epoch": 1.7896579585241046,
+ "grad_norm": 1.1103372341248579,
+ "learning_rate": 2.706156443561481e-07,
+ "loss": 0.3016365528106689,
+ "num_input_tokens_seen": 1162338136,
+ "step": 26580,
+ "train_runtime": 205709.1075,
+ "train_tokens_per_second": 5650.397
+ },
+ {
+ "epoch": 1.7899946135200646,
+ "grad_norm": 0.7165987576366394,
+ "learning_rate": 2.6975823300257884e-07,
+ "loss": 0.29532361030578613,
+ "num_input_tokens_seen": 1162551688,
+ "step": 26585,
+ "train_runtime": 205742.7144,
+ "train_tokens_per_second": 5650.512
+ },
+ {
+ "epoch": 1.7903312685160246,
+ "grad_norm": 0.6849565677358419,
+ "learning_rate": 2.6890214444603604e-07,
+ "loss": 0.29342079162597656,
+ "num_input_tokens_seen": 1162773064,
+ "step": 26590,
+ "train_runtime": 205775.5364,
+ "train_tokens_per_second": 5650.687
+ },
+ {
+ "epoch": 1.790667923511985,
+ "grad_norm": 0.8242687417078883,
+ "learning_rate": 2.6804737892592193e-07,
+ "loss": 0.30418787002563474,
+ "num_input_tokens_seen": 1162987848,
+ "step": 26595,
+ "train_runtime": 205824.7685,
+ "train_tokens_per_second": 5650.379
+ },
+ {
+ "epoch": 1.7910045785079451,
+ "grad_norm": 0.8268380789860988,
+ "learning_rate": 2.671939366812687e-07,
+ "loss": 0.30772976875305175,
+ "num_input_tokens_seen": 1163202312,
+ "step": 26600,
+ "train_runtime": 205862.7926,
+ "train_tokens_per_second": 5650.377
+ },
+ {
+ "epoch": 1.7913412335039052,
+ "grad_norm": 0.767877157549208,
+ "learning_rate": 2.66341817950741e-07,
+ "loss": 0.29974985122680664,
+ "num_input_tokens_seen": 1163438080,
+ "step": 26605,
+ "train_runtime": 205898.0882,
+ "train_tokens_per_second": 5650.553
+ },
+ {
+ "epoch": 1.7916778884998652,
+ "grad_norm": 0.7009232602836326,
+ "learning_rate": 2.6549102297262764e-07,
+ "loss": 0.29571521282196045,
+ "num_input_tokens_seen": 1163660200,
+ "step": 26610,
+ "train_runtime": 205939.8237,
+ "train_tokens_per_second": 5650.487
+ },
+ {
+ "epoch": 1.7920145434958255,
+ "grad_norm": 0.871322486063614,
+ "learning_rate": 2.6464155198485387e-07,
+ "loss": 0.30173745155334475,
+ "num_input_tokens_seen": 1163887248,
+ "step": 26615,
+ "train_runtime": 205979.0591,
+ "train_tokens_per_second": 5650.512
+ },
+ {
+ "epoch": 1.7923511984917857,
+ "grad_norm": 0.803169204691401,
+ "learning_rate": 2.6379340522496853e-07,
+ "loss": 0.29706802368164065,
+ "num_input_tokens_seen": 1164110704,
+ "step": 26620,
+ "train_runtime": 206013.7386,
+ "train_tokens_per_second": 5650.646
+ },
+ {
+ "epoch": 1.7926878534877457,
+ "grad_norm": 0.7782352100729737,
+ "learning_rate": 2.6294658293015584e-07,
+ "loss": 0.3205864429473877,
+ "num_input_tokens_seen": 1164337200,
+ "step": 26625,
+ "train_runtime": 206050.8157,
+ "train_tokens_per_second": 5650.728
+ },
+ {
+ "epoch": 1.7930245084837058,
+ "grad_norm": 0.8588354428150048,
+ "learning_rate": 2.621010853372258e-07,
+ "loss": 0.31758651733398435,
+ "num_input_tokens_seen": 1164544968,
+ "step": 26630,
+ "train_runtime": 206091.8464,
+ "train_tokens_per_second": 5650.612
+ },
+ {
+ "epoch": 1.793361163479666,
+ "grad_norm": 0.8321453808487478,
+ "learning_rate": 2.612569126826187e-07,
+ "loss": 0.3288443088531494,
+ "num_input_tokens_seen": 1164763392,
+ "step": 26635,
+ "train_runtime": 206129.4651,
+ "train_tokens_per_second": 5650.64
+ },
+ {
+ "epoch": 1.7936978184756263,
+ "grad_norm": 0.7646490512058014,
+ "learning_rate": 2.6041406520240563e-07,
+ "loss": 0.3208475589752197,
+ "num_input_tokens_seen": 1164996760,
+ "step": 26640,
+ "train_runtime": 206162.845,
+ "train_tokens_per_second": 5650.857
+ },
+ {
+ "epoch": 1.7940344734715863,
+ "grad_norm": 0.9618962883016744,
+ "learning_rate": 2.5957254313228595e-07,
+ "loss": 0.3083490371704102,
+ "num_input_tokens_seen": 1165209184,
+ "step": 26645,
+ "train_runtime": 206203.1869,
+ "train_tokens_per_second": 5650.782
+ },
+ {
+ "epoch": 1.7943711284675463,
+ "grad_norm": 0.6865797037478947,
+ "learning_rate": 2.5873234670758753e-07,
+ "loss": 0.2793558597564697,
+ "num_input_tokens_seen": 1165437296,
+ "step": 26650,
+ "train_runtime": 206245.195,
+ "train_tokens_per_second": 5650.737
+ },
+ {
+ "epoch": 1.7947077834635066,
+ "grad_norm": 0.743249619367278,
+ "learning_rate": 2.578934761632712e-07,
+ "loss": 0.29536702632904055,
+ "num_input_tokens_seen": 1165668480,
+ "step": 26655,
+ "train_runtime": 206276.3477,
+ "train_tokens_per_second": 5651.004
+ },
+ {
+ "epoch": 1.7950444384594668,
+ "grad_norm": 0.775356540520281,
+ "learning_rate": 2.570559317339211e-07,
+ "loss": 0.30143041610717775,
+ "num_input_tokens_seen": 1165893536,
+ "step": 26660,
+ "train_runtime": 206318.3881,
+ "train_tokens_per_second": 5650.943
+ },
+ {
+ "epoch": 1.7953810934554268,
+ "grad_norm": 0.6994994001366991,
+ "learning_rate": 2.56219713653757e-07,
+ "loss": 0.2995075941085815,
+ "num_input_tokens_seen": 1166122240,
+ "step": 26665,
+ "train_runtime": 206352.7461,
+ "train_tokens_per_second": 5651.111
+ },
+ {
+ "epoch": 1.7957177484513869,
+ "grad_norm": 0.8569243475539816,
+ "learning_rate": 2.5538482215662187e-07,
+ "loss": 0.3236057758331299,
+ "num_input_tokens_seen": 1166338248,
+ "step": 26670,
+ "train_runtime": 206386.7169,
+ "train_tokens_per_second": 5651.227
+ },
+ {
+ "epoch": 1.7960544034473471,
+ "grad_norm": 0.766892273093451,
+ "learning_rate": 2.545512574759923e-07,
+ "loss": 0.2701665163040161,
+ "num_input_tokens_seen": 1166555592,
+ "step": 26675,
+ "train_runtime": 206433.7742,
+ "train_tokens_per_second": 5650.992
+ },
+ {
+ "epoch": 1.7963910584433074,
+ "grad_norm": 0.7756837374057447,
+ "learning_rate": 2.537190198449713e-07,
+ "loss": 0.31007416248321534,
+ "num_input_tokens_seen": 1166772216,
+ "step": 26680,
+ "train_runtime": 206467.3628,
+ "train_tokens_per_second": 5651.122
+ },
+ {
+ "epoch": 1.7967277134392674,
+ "grad_norm": 0.7601571953732823,
+ "learning_rate": 2.5288810949629207e-07,
+ "loss": 0.3006875991821289,
+ "num_input_tokens_seen": 1166996856,
+ "step": 26685,
+ "train_runtime": 206509.338,
+ "train_tokens_per_second": 5651.061
+ },
+ {
+ "epoch": 1.7970643684352274,
+ "grad_norm": 0.7713983757495535,
+ "learning_rate": 2.5205852666231544e-07,
+ "loss": 0.2778863668441772,
+ "num_input_tokens_seen": 1167208832,
+ "step": 26690,
+ "train_runtime": 206552.5087,
+ "train_tokens_per_second": 5650.906
+ },
+ {
+ "epoch": 1.7974010234311877,
+ "grad_norm": 0.8187900042183005,
+ "learning_rate": 2.5123027157503144e-07,
+ "loss": 0.28261337280273435,
+ "num_input_tokens_seen": 1167408464,
+ "step": 26695,
+ "train_runtime": 206590.339,
+ "train_tokens_per_second": 5650.838
+ },
+ {
+ "epoch": 1.797737678427148,
+ "grad_norm": 0.7565722539390317,
+ "learning_rate": 2.504033444660592e-07,
+ "loss": 0.3025941848754883,
+ "num_input_tokens_seen": 1167643608,
+ "step": 26700,
+ "train_runtime": 206627.1922,
+ "train_tokens_per_second": 5650.968
+ },
+ {
+ "epoch": 1.798074333423108,
+ "grad_norm": 0.6833026230882182,
+ "learning_rate": 2.4957774556664706e-07,
+ "loss": 0.28271756172180174,
+ "num_input_tokens_seen": 1167868528,
+ "step": 26705,
+ "train_runtime": 206670.4456,
+ "train_tokens_per_second": 5650.873
+ },
+ {
+ "epoch": 1.798410988419068,
+ "grad_norm": 0.8308124821063052,
+ "learning_rate": 2.487534751076692e-07,
+ "loss": 0.29103024005889894,
+ "num_input_tokens_seen": 1168090128,
+ "step": 26710,
+ "train_runtime": 206715.3981,
+ "train_tokens_per_second": 5650.717
+ },
+ {
+ "epoch": 1.7987476434150282,
+ "grad_norm": 0.7543688224419066,
+ "learning_rate": 2.47930533319633e-07,
+ "loss": 0.2806957721710205,
+ "num_input_tokens_seen": 1168302824,
+ "step": 26715,
+ "train_runtime": 206757.8975,
+ "train_tokens_per_second": 5650.584
+ },
+ {
+ "epoch": 1.7990842984109885,
+ "grad_norm": 0.8079691544469572,
+ "learning_rate": 2.4710892043266823e-07,
+ "loss": 0.3256166219711304,
+ "num_input_tokens_seen": 1168526496,
+ "step": 26720,
+ "train_runtime": 206793.1384,
+ "train_tokens_per_second": 5650.702
+ },
+ {
+ "epoch": 1.7994209534069485,
+ "grad_norm": 0.7709686680304173,
+ "learning_rate": 2.462886366765388e-07,
+ "loss": 0.2893906354904175,
+ "num_input_tokens_seen": 1168747720,
+ "step": 26725,
+ "train_runtime": 206834.9728,
+ "train_tokens_per_second": 5650.629
+ },
+ {
+ "epoch": 1.7997576084029085,
+ "grad_norm": 0.861388639740031,
+ "learning_rate": 2.454696822806341e-07,
+ "loss": 0.3119854211807251,
+ "num_input_tokens_seen": 1168955928,
+ "step": 26730,
+ "train_runtime": 206876.7703,
+ "train_tokens_per_second": 5650.494
+ },
+ {
+ "epoch": 1.8000942633988688,
+ "grad_norm": 0.7333451652936307,
+ "learning_rate": 2.446520574739708e-07,
+ "loss": 0.2833410739898682,
+ "num_input_tokens_seen": 1169181640,
+ "step": 26735,
+ "train_runtime": 206910.3727,
+ "train_tokens_per_second": 5650.667
+ },
+ {
+ "epoch": 1.800430918394829,
+ "grad_norm": 0.8244240692612769,
+ "learning_rate": 2.438357624851967e-07,
+ "loss": 0.265745735168457,
+ "num_input_tokens_seen": 1169399504,
+ "step": 26740,
+ "train_runtime": 206944.8765,
+ "train_tokens_per_second": 5650.778
+ },
+ {
+ "epoch": 1.800767573390789,
+ "grad_norm": 0.7646663892198701,
+ "learning_rate": 2.430207975425847e-07,
+ "loss": 0.30634517669677735,
+ "num_input_tokens_seen": 1169627072,
+ "step": 26745,
+ "train_runtime": 206986.9735,
+ "train_tokens_per_second": 5650.728
+ },
+ {
+ "epoch": 1.801104228386749,
+ "grad_norm": 0.7977131343420332,
+ "learning_rate": 2.4220716287403744e-07,
+ "loss": 0.306353759765625,
+ "num_input_tokens_seen": 1169859232,
+ "step": 26750,
+ "train_runtime": 207024.3013,
+ "train_tokens_per_second": 5650.83
+ },
+ {
+ "epoch": 1.8014408833827094,
+ "grad_norm": 0.7225010612739647,
+ "learning_rate": 2.413948587070869e-07,
+ "loss": 0.2988481044769287,
+ "num_input_tokens_seen": 1170041408,
+ "step": 26755,
+ "train_runtime": 207063.8036,
+ "train_tokens_per_second": 5650.632
+ },
+ {
+ "epoch": 1.8017775383786696,
+ "grad_norm": 0.775593581305895,
+ "learning_rate": 2.405838852688891e-07,
+ "loss": 0.3097994327545166,
+ "num_input_tokens_seen": 1170267480,
+ "step": 26760,
+ "train_runtime": 207095.0651,
+ "train_tokens_per_second": 5650.871
+ },
+ {
+ "epoch": 1.8021141933746296,
+ "grad_norm": 0.8152686402481596,
+ "learning_rate": 2.397742427862315e-07,
+ "loss": 0.2666957378387451,
+ "num_input_tokens_seen": 1170495840,
+ "step": 26765,
+ "train_runtime": 207129.3856,
+ "train_tokens_per_second": 5651.037
+ },
+ {
+ "epoch": 1.8024508483705897,
+ "grad_norm": 0.7668475732368676,
+ "learning_rate": 2.3896593148552695e-07,
+ "loss": 0.3030317544937134,
+ "num_input_tokens_seen": 1170707968,
+ "step": 26770,
+ "train_runtime": 207164.983,
+ "train_tokens_per_second": 5651.09
+ },
+ {
+ "epoch": 1.80278750336655,
+ "grad_norm": 0.7115287772275755,
+ "learning_rate": 2.3815895159281844e-07,
+ "loss": 0.28127307891845704,
+ "num_input_tokens_seen": 1170932136,
+ "step": 26775,
+ "train_runtime": 207201.4624,
+ "train_tokens_per_second": 5651.177
+ },
+ {
+ "epoch": 1.8031241583625102,
+ "grad_norm": 0.7814253275211129,
+ "learning_rate": 2.3735330333377493e-07,
+ "loss": 0.2792118310928345,
+ "num_input_tokens_seen": 1171162408,
+ "step": 26780,
+ "train_runtime": 207238.2441,
+ "train_tokens_per_second": 5651.285
+ },
+ {
+ "epoch": 1.8034608133584702,
+ "grad_norm": 0.8200002096055589,
+ "learning_rate": 2.3654898693369232e-07,
+ "loss": 0.29738240242004393,
+ "num_input_tokens_seen": 1171377408,
+ "step": 26785,
+ "train_runtime": 207281.7346,
+ "train_tokens_per_second": 5651.137
+ },
+ {
+ "epoch": 1.8037974683544302,
+ "grad_norm": 0.7884729475059873,
+ "learning_rate": 2.3574600261749735e-07,
+ "loss": 0.30119688510894777,
+ "num_input_tokens_seen": 1171597216,
+ "step": 26790,
+ "train_runtime": 207321.2969,
+ "train_tokens_per_second": 5651.118
+ },
+ {
+ "epoch": 1.8041341233503905,
+ "grad_norm": 0.8577997537906676,
+ "learning_rate": 2.3494435060973875e-07,
+ "loss": 0.3168250322341919,
+ "num_input_tokens_seen": 1171822008,
+ "step": 26795,
+ "train_runtime": 207355.9199,
+ "train_tokens_per_second": 5651.259
+ },
+ {
+ "epoch": 1.8044707783463507,
+ "grad_norm": 0.9036192566903659,
+ "learning_rate": 2.3414403113459937e-07,
+ "loss": 0.30097527503967286,
+ "num_input_tokens_seen": 1172040272,
+ "step": 26800,
+ "train_runtime": 207392.3163,
+ "train_tokens_per_second": 5651.32
+ },
+ {
+ "epoch": 1.8048074333423108,
+ "grad_norm": 0.7489283622158472,
+ "learning_rate": 2.3334504441588413e-07,
+ "loss": 0.2781896352767944,
+ "num_input_tokens_seen": 1172260704,
+ "step": 26805,
+ "train_runtime": 207432.2609,
+ "train_tokens_per_second": 5651.294
+ },
+ {
+ "epoch": 1.8051440883382708,
+ "grad_norm": 0.7236248733853186,
+ "learning_rate": 2.3254739067702704e-07,
+ "loss": 0.2797673225402832,
+ "num_input_tokens_seen": 1172485680,
+ "step": 26810,
+ "train_runtime": 207469.9941,
+ "train_tokens_per_second": 5651.351
+ },
+ {
+ "epoch": 1.805480743334231,
+ "grad_norm": 0.8892645318062625,
+ "learning_rate": 2.3175107014109077e-07,
+ "loss": 0.2838028907775879,
+ "num_input_tokens_seen": 1172707720,
+ "step": 26815,
+ "train_runtime": 207507.3876,
+ "train_tokens_per_second": 5651.402
+ },
+ {
+ "epoch": 1.8058173983301913,
+ "grad_norm": 0.7651063250427357,
+ "learning_rate": 2.3095608303076223e-07,
+ "loss": 0.311909818649292,
+ "num_input_tokens_seen": 1172928712,
+ "step": 26820,
+ "train_runtime": 207551.5584,
+ "train_tokens_per_second": 5651.264
+ },
+ {
+ "epoch": 1.8061540533261513,
+ "grad_norm": 0.734890627438004,
+ "learning_rate": 2.3016242956835798e-07,
+ "loss": 0.31490211486816405,
+ "num_input_tokens_seen": 1173154336,
+ "step": 26825,
+ "train_runtime": 207594.3613,
+ "train_tokens_per_second": 5651.186
+ },
+ {
+ "epoch": 1.8064907083221113,
+ "grad_norm": 0.7733770590852225,
+ "learning_rate": 2.2937010997581999e-07,
+ "loss": 0.284251070022583,
+ "num_input_tokens_seen": 1173384008,
+ "step": 26830,
+ "train_runtime": 207634.0828,
+ "train_tokens_per_second": 5651.211
+ },
+ {
+ "epoch": 1.8068273633180716,
+ "grad_norm": 0.9236592739480588,
+ "learning_rate": 2.2857912447471818e-07,
+ "loss": 0.2854511260986328,
+ "num_input_tokens_seen": 1173591424,
+ "step": 26835,
+ "train_runtime": 207670.999,
+ "train_tokens_per_second": 5651.205
+ },
+ {
+ "epoch": 1.8071640183140318,
+ "grad_norm": 0.709148295682169,
+ "learning_rate": 2.2778947328625012e-07,
+ "loss": 0.2829423427581787,
+ "num_input_tokens_seen": 1173795056,
+ "step": 26840,
+ "train_runtime": 207707.7029,
+ "train_tokens_per_second": 5651.187
+ },
+ {
+ "epoch": 1.8075006733099919,
+ "grad_norm": 0.9439136755079099,
+ "learning_rate": 2.2700115663123745e-07,
+ "loss": 0.3229234218597412,
+ "num_input_tokens_seen": 1174010232,
+ "step": 26845,
+ "train_runtime": 207744.5589,
+ "train_tokens_per_second": 5651.22
+ },
+ {
+ "epoch": 1.807837328305952,
+ "grad_norm": 0.9241155338367218,
+ "learning_rate": 2.2621417473013164e-07,
+ "loss": 0.28131952285766604,
+ "num_input_tokens_seen": 1174230336,
+ "step": 26850,
+ "train_runtime": 207785.5145,
+ "train_tokens_per_second": 5651.166
+ },
+ {
+ "epoch": 1.8081739833019121,
+ "grad_norm": 0.8397274812785074,
+ "learning_rate": 2.254285278030094e-07,
+ "loss": 0.30530600547790526,
+ "num_input_tokens_seen": 1174459216,
+ "step": 26855,
+ "train_runtime": 207819.3658,
+ "train_tokens_per_second": 5651.346
+ },
+ {
+ "epoch": 1.8085106382978724,
+ "grad_norm": 0.8065249949734656,
+ "learning_rate": 2.24644216069575e-07,
+ "loss": 0.30002517700195314,
+ "num_input_tokens_seen": 1174667136,
+ "step": 26860,
+ "train_runtime": 207866.2773,
+ "train_tokens_per_second": 5651.071
+ },
+ {
+ "epoch": 1.8088472932938324,
+ "grad_norm": 0.6707793753515161,
+ "learning_rate": 2.238612397491574e-07,
+ "loss": 0.2729137420654297,
+ "num_input_tokens_seen": 1174893104,
+ "step": 26865,
+ "train_runtime": 207905.4724,
+ "train_tokens_per_second": 5651.093
+ },
+ {
+ "epoch": 1.8091839482897925,
+ "grad_norm": 0.7616275632457961,
+ "learning_rate": 2.230795990607143e-07,
+ "loss": 0.29819502830505373,
+ "num_input_tokens_seen": 1175120424,
+ "step": 26870,
+ "train_runtime": 207946.3976,
+ "train_tokens_per_second": 5651.074
+ },
+ {
+ "epoch": 1.809520603285753,
+ "grad_norm": 0.7786936080708119,
+ "learning_rate": 2.2229929422282915e-07,
+ "loss": 0.27527265548706054,
+ "num_input_tokens_seen": 1175334208,
+ "step": 26875,
+ "train_runtime": 207982.2651,
+ "train_tokens_per_second": 5651.127
+ },
+ {
+ "epoch": 1.809857258281713,
+ "grad_norm": 0.7904035393585753,
+ "learning_rate": 2.215203254537107e-07,
+ "loss": 0.3183107614517212,
+ "num_input_tokens_seen": 1175555296,
+ "step": 26880,
+ "train_runtime": 208022.7623,
+ "train_tokens_per_second": 5651.09
+ },
+ {
+ "epoch": 1.810193913277673,
+ "grad_norm": 0.7604104589550896,
+ "learning_rate": 2.2074269297119588e-07,
+ "loss": 0.3029907464981079,
+ "num_input_tokens_seen": 1175780944,
+ "step": 26885,
+ "train_runtime": 208056.1021,
+ "train_tokens_per_second": 5651.269
+ },
+ {
+ "epoch": 1.8105305682736332,
+ "grad_norm": 0.8292516099336918,
+ "learning_rate": 2.199663969927479e-07,
+ "loss": 0.29438438415527346,
+ "num_input_tokens_seen": 1175996056,
+ "step": 26890,
+ "train_runtime": 208096.3568,
+ "train_tokens_per_second": 5651.209
+ },
+ {
+ "epoch": 1.8108672232695935,
+ "grad_norm": 0.8271865279223709,
+ "learning_rate": 2.1919143773545316e-07,
+ "loss": 0.29373531341552733,
+ "num_input_tokens_seen": 1176219504,
+ "step": 26895,
+ "train_runtime": 208134.3303,
+ "train_tokens_per_second": 5651.252
+ },
+ {
+ "epoch": 1.8112038782655535,
+ "grad_norm": 0.9028884805515789,
+ "learning_rate": 2.184178154160288e-07,
+ "loss": 0.29395351409912107,
+ "num_input_tokens_seen": 1176436808,
+ "step": 26900,
+ "train_runtime": 208170.4974,
+ "train_tokens_per_second": 5651.314
+ },
+ {
+ "epoch": 1.8115405332615135,
+ "grad_norm": 0.7863867287916279,
+ "learning_rate": 2.1764553025081514e-07,
+ "loss": 0.2931424379348755,
+ "num_input_tokens_seen": 1176656768,
+ "step": 26905,
+ "train_runtime": 208207.3522,
+ "train_tokens_per_second": 5651.37
+ },
+ {
+ "epoch": 1.8118771882574738,
+ "grad_norm": 0.8798746009059921,
+ "learning_rate": 2.168745824557794e-07,
+ "loss": 0.28325557708740234,
+ "num_input_tokens_seen": 1176871952,
+ "step": 26910,
+ "train_runtime": 208250.7917,
+ "train_tokens_per_second": 5651.224
+ },
+ {
+ "epoch": 1.812213843253434,
+ "grad_norm": 0.8072441204009388,
+ "learning_rate": 2.161049722465136e-07,
+ "loss": 0.28316502571105956,
+ "num_input_tokens_seen": 1177088104,
+ "step": 26915,
+ "train_runtime": 208293.9071,
+ "train_tokens_per_second": 5651.092
+ },
+ {
+ "epoch": 1.812550498249394,
+ "grad_norm": 0.7553112409241435,
+ "learning_rate": 2.1533669983823835e-07,
+ "loss": 0.29533112049102783,
+ "num_input_tokens_seen": 1177320360,
+ "step": 26920,
+ "train_runtime": 208336.7997,
+ "train_tokens_per_second": 5651.044
+ },
+ {
+ "epoch": 1.812887153245354,
+ "grad_norm": 0.6386036144723565,
+ "learning_rate": 2.1456976544579732e-07,
+ "loss": 0.2751606464385986,
+ "num_input_tokens_seen": 1177560024,
+ "step": 26925,
+ "train_runtime": 208382.272,
+ "train_tokens_per_second": 5650.961
+ },
+ {
+ "epoch": 1.8132238082413143,
+ "grad_norm": 0.7537084466037821,
+ "learning_rate": 2.138041692836623e-07,
+ "loss": 0.29282474517822266,
+ "num_input_tokens_seen": 1177776744,
+ "step": 26930,
+ "train_runtime": 208414.4067,
+ "train_tokens_per_second": 5651.129
+ },
+ {
+ "epoch": 1.8135604632372746,
+ "grad_norm": 0.766246409563199,
+ "learning_rate": 2.1303991156592818e-07,
+ "loss": 0.29790616035461426,
+ "num_input_tokens_seen": 1177986848,
+ "step": 26935,
+ "train_runtime": 208452.677,
+ "train_tokens_per_second": 5651.1
+ },
+ {
+ "epoch": 1.8138971182332346,
+ "grad_norm": 0.7584006159785801,
+ "learning_rate": 2.122769925063195e-07,
+ "loss": 0.2797182559967041,
+ "num_input_tokens_seen": 1178200832,
+ "step": 26940,
+ "train_runtime": 208493.3034,
+ "train_tokens_per_second": 5651.025
+ },
+ {
+ "epoch": 1.8142337732291947,
+ "grad_norm": 0.8751221362570611,
+ "learning_rate": 2.1151541231818174e-07,
+ "loss": 0.2855216503143311,
+ "num_input_tokens_seen": 1178415944,
+ "step": 26945,
+ "train_runtime": 208536.0031,
+ "train_tokens_per_second": 5650.899
+ },
+ {
+ "epoch": 1.814570428225155,
+ "grad_norm": 0.8086208091486383,
+ "learning_rate": 2.1075517121449063e-07,
+ "loss": 0.31327362060546876,
+ "num_input_tokens_seen": 1178637904,
+ "step": 26950,
+ "train_runtime": 208579.4059,
+ "train_tokens_per_second": 5650.788
+ },
+ {
+ "epoch": 1.8149070832211152,
+ "grad_norm": 0.8301036812650167,
+ "learning_rate": 2.0999626940784334e-07,
+ "loss": 0.27339811325073243,
+ "num_input_tokens_seen": 1178843008,
+ "step": 26955,
+ "train_runtime": 208615.5879,
+ "train_tokens_per_second": 5650.791
+ },
+ {
+ "epoch": 1.8152437382170752,
+ "grad_norm": 0.8710877367090695,
+ "learning_rate": 2.0923870711046568e-07,
+ "loss": 0.2946287155151367,
+ "num_input_tokens_seen": 1179067688,
+ "step": 26960,
+ "train_runtime": 208653.4478,
+ "train_tokens_per_second": 5650.842
+ },
+ {
+ "epoch": 1.8155803932130352,
+ "grad_norm": 0.6554427095056514,
+ "learning_rate": 2.0848248453420704e-07,
+ "loss": 0.2847181797027588,
+ "num_input_tokens_seen": 1179293168,
+ "step": 26965,
+ "train_runtime": 208690.2441,
+ "train_tokens_per_second": 5650.926
+ },
+ {
+ "epoch": 1.8159170482089955,
+ "grad_norm": 0.8746736454560675,
+ "learning_rate": 2.077276018905422e-07,
+ "loss": 0.2915675163269043,
+ "num_input_tokens_seen": 1179511568,
+ "step": 26970,
+ "train_runtime": 208729.2372,
+ "train_tokens_per_second": 5650.917
+ },
+ {
+ "epoch": 1.8162537032049557,
+ "grad_norm": 0.7805453963758747,
+ "learning_rate": 2.0697405939057225e-07,
+ "loss": 0.3022061824798584,
+ "num_input_tokens_seen": 1179708384,
+ "step": 26975,
+ "train_runtime": 208763.205,
+ "train_tokens_per_second": 5650.94
+ },
+ {
+ "epoch": 1.8165903582009157,
+ "grad_norm": 0.6389930066356448,
+ "learning_rate": 2.0622185724502253e-07,
+ "loss": 0.27937405109405516,
+ "num_input_tokens_seen": 1179925616,
+ "step": 26980,
+ "train_runtime": 208797.998,
+ "train_tokens_per_second": 5651.039
+ },
+ {
+ "epoch": 1.8169270131968758,
+ "grad_norm": 0.8421462718448113,
+ "learning_rate": 2.0547099566424367e-07,
+ "loss": 0.2782083511352539,
+ "num_input_tokens_seen": 1180142704,
+ "step": 26985,
+ "train_runtime": 208842.2398,
+ "train_tokens_per_second": 5650.881
+ },
+ {
+ "epoch": 1.817263668192836,
+ "grad_norm": 0.8420718246560043,
+ "learning_rate": 2.0472147485821325e-07,
+ "loss": 0.27927365303039553,
+ "num_input_tokens_seen": 1180348248,
+ "step": 26990,
+ "train_runtime": 208879.8268,
+ "train_tokens_per_second": 5650.848
+ },
+ {
+ "epoch": 1.8176003231887963,
+ "grad_norm": 0.7460943379350691,
+ "learning_rate": 2.0397329503653084e-07,
+ "loss": 0.2773773670196533,
+ "num_input_tokens_seen": 1180563544,
+ "step": 26995,
+ "train_runtime": 208916.8282,
+ "train_tokens_per_second": 5650.878
+ },
+ {
+ "epoch": 1.8179369781847563,
+ "grad_norm": 0.8475284436859102,
+ "learning_rate": 2.03226456408423e-07,
+ "loss": 0.3106630802154541,
+ "num_input_tokens_seen": 1180762848,
+ "step": 27000,
+ "train_runtime": 208960.3449,
+ "train_tokens_per_second": 5650.655
+ },
+ {
+ "epoch": 1.8182736331807163,
+ "grad_norm": 0.7183644291988805,
+ "learning_rate": 2.024809591827409e-07,
+ "loss": 0.30523037910461426,
+ "num_input_tokens_seen": 1180988840,
+ "step": 27005,
+ "train_runtime": 209002.5667,
+ "train_tokens_per_second": 5650.595
+ },
+ {
+ "epoch": 1.8186102881766766,
+ "grad_norm": 0.801765388255765,
+ "learning_rate": 2.0173680356796066e-07,
+ "loss": 0.26985788345336914,
+ "num_input_tokens_seen": 1181189432,
+ "step": 27010,
+ "train_runtime": 209041.262,
+ "train_tokens_per_second": 5650.509
+ },
+ {
+ "epoch": 1.8189469431726368,
+ "grad_norm": 0.8150765259723309,
+ "learning_rate": 2.00993989772183e-07,
+ "loss": 0.2943392515182495,
+ "num_input_tokens_seen": 1181407992,
+ "step": 27015,
+ "train_runtime": 209073.5583,
+ "train_tokens_per_second": 5650.681
+ },
+ {
+ "epoch": 1.8192835981685969,
+ "grad_norm": 0.7914385589041727,
+ "learning_rate": 2.0025251800313284e-07,
+ "loss": 0.3112603187561035,
+ "num_input_tokens_seen": 1181622192,
+ "step": 27020,
+ "train_runtime": 209120.2047,
+ "train_tokens_per_second": 5650.445
+ },
+ {
+ "epoch": 1.8196202531645569,
+ "grad_norm": 0.7561624538382505,
+ "learning_rate": 1.9951238846816045e-07,
+ "loss": 0.3185489892959595,
+ "num_input_tokens_seen": 1181854088,
+ "step": 27025,
+ "train_runtime": 209152.1413,
+ "train_tokens_per_second": 5650.691
+ },
+ {
+ "epoch": 1.8199569081605171,
+ "grad_norm": 0.7921686074062922,
+ "learning_rate": 1.987736013742414e-07,
+ "loss": 0.3113109111785889,
+ "num_input_tokens_seen": 1182065984,
+ "step": 27030,
+ "train_runtime": 209191.0211,
+ "train_tokens_per_second": 5650.654
+ },
+ {
+ "epoch": 1.8202935631564774,
+ "grad_norm": 0.7653273331531749,
+ "learning_rate": 1.9803615692797375e-07,
+ "loss": 0.27271738052368166,
+ "num_input_tokens_seen": 1182287536,
+ "step": 27035,
+ "train_runtime": 209228.5228,
+ "train_tokens_per_second": 5650.7
+ },
+ {
+ "epoch": 1.8206302181524374,
+ "grad_norm": 0.7068384362133963,
+ "learning_rate": 1.9730005533558417e-07,
+ "loss": 0.29187893867492676,
+ "num_input_tokens_seen": 1182520232,
+ "step": 27040,
+ "train_runtime": 209272.9747,
+ "train_tokens_per_second": 5650.611
+ },
+ {
+ "epoch": 1.8209668731483974,
+ "grad_norm": 0.7399705032741323,
+ "learning_rate": 1.9656529680291802e-07,
+ "loss": 0.27612121105194093,
+ "num_input_tokens_seen": 1182742904,
+ "step": 27045,
+ "train_runtime": 209312.4766,
+ "train_tokens_per_second": 5650.609
+ },
+ {
+ "epoch": 1.8213035281443577,
+ "grad_norm": 0.7369131913718883,
+ "learning_rate": 1.9583188153544986e-07,
+ "loss": 0.28553037643432616,
+ "num_input_tokens_seen": 1182945672,
+ "step": 27050,
+ "train_runtime": 209347.8795,
+ "train_tokens_per_second": 5650.622
+ },
+ {
+ "epoch": 1.821640183140318,
+ "grad_norm": 1.1539568423391955,
+ "learning_rate": 1.9509980973827724e-07,
+ "loss": 0.2979405641555786,
+ "num_input_tokens_seen": 1183165608,
+ "step": 27055,
+ "train_runtime": 209386.946,
+ "train_tokens_per_second": 5650.618
+ },
+ {
+ "epoch": 1.821976838136278,
+ "grad_norm": 0.7640565156115381,
+ "learning_rate": 1.943690816161209e-07,
+ "loss": 0.3077999114990234,
+ "num_input_tokens_seen": 1183395432,
+ "step": 27060,
+ "train_runtime": 209424.1043,
+ "train_tokens_per_second": 5650.713
+ },
+ {
+ "epoch": 1.822313493132238,
+ "grad_norm": 0.7149421204366606,
+ "learning_rate": 1.9363969737332677e-07,
+ "loss": 0.2753284454345703,
+ "num_input_tokens_seen": 1183617608,
+ "step": 27065,
+ "train_runtime": 209460.0325,
+ "train_tokens_per_second": 5650.804
+ },
+ {
+ "epoch": 1.8226501481281983,
+ "grad_norm": 0.652062441800595,
+ "learning_rate": 1.9291165721386507e-07,
+ "loss": 0.27479655742645265,
+ "num_input_tokens_seen": 1183826768,
+ "step": 27070,
+ "train_runtime": 209495.4576,
+ "train_tokens_per_second": 5650.847
+ },
+ {
+ "epoch": 1.8229868031241585,
+ "grad_norm": 0.9271372389885489,
+ "learning_rate": 1.9218496134132957e-07,
+ "loss": 0.3049477100372314,
+ "num_input_tokens_seen": 1184052800,
+ "step": 27075,
+ "train_runtime": 209531.6276,
+ "train_tokens_per_second": 5650.95
+ },
+ {
+ "epoch": 1.8233234581201185,
+ "grad_norm": 0.6911714020764224,
+ "learning_rate": 1.9145960995893831e-07,
+ "loss": 0.282161283493042,
+ "num_input_tokens_seen": 1184270392,
+ "step": 27080,
+ "train_runtime": 209567.2517,
+ "train_tokens_per_second": 5651.028
+ },
+ {
+ "epoch": 1.8236601131160786,
+ "grad_norm": 0.7018056304999722,
+ "learning_rate": 1.9073560326953344e-07,
+ "loss": 0.3012186050415039,
+ "num_input_tokens_seen": 1184494848,
+ "step": 27085,
+ "train_runtime": 209601.3891,
+ "train_tokens_per_second": 5651.178
+ },
+ {
+ "epoch": 1.8239967681120388,
+ "grad_norm": 0.7756023060864273,
+ "learning_rate": 1.9001294147558303e-07,
+ "loss": 0.3109905242919922,
+ "num_input_tokens_seen": 1184709808,
+ "step": 27090,
+ "train_runtime": 209641.9441,
+ "train_tokens_per_second": 5651.111
+ },
+ {
+ "epoch": 1.824333423107999,
+ "grad_norm": 0.8354480678136353,
+ "learning_rate": 1.892916247791743e-07,
+ "loss": 0.3057056188583374,
+ "num_input_tokens_seen": 1184920824,
+ "step": 27095,
+ "train_runtime": 209678.6955,
+ "train_tokens_per_second": 5651.126
+ },
+ {
+ "epoch": 1.824670078103959,
+ "grad_norm": 0.7450481139042144,
+ "learning_rate": 1.885716533820231e-07,
+ "loss": 0.2988601207733154,
+ "num_input_tokens_seen": 1185141448,
+ "step": 27100,
+ "train_runtime": 209721.9756,
+ "train_tokens_per_second": 5651.012
+ },
+ {
+ "epoch": 1.8250067330999191,
+ "grad_norm": 0.7070713398300131,
+ "learning_rate": 1.878530274854662e-07,
+ "loss": 0.2867612838745117,
+ "num_input_tokens_seen": 1185358376,
+ "step": 27105,
+ "train_runtime": 209757.3918,
+ "train_tokens_per_second": 5651.092
+ },
+ {
+ "epoch": 1.8253433880958794,
+ "grad_norm": 0.7229583190371239,
+ "learning_rate": 1.8713574729046557e-07,
+ "loss": 0.3091785669326782,
+ "num_input_tokens_seen": 1185587344,
+ "step": 27110,
+ "train_runtime": 209793.505,
+ "train_tokens_per_second": 5651.211
+ },
+ {
+ "epoch": 1.8256800430918396,
+ "grad_norm": 0.7520860741774413,
+ "learning_rate": 1.8641981299760582e-07,
+ "loss": 0.29340548515319825,
+ "num_input_tokens_seen": 1185795904,
+ "step": 27115,
+ "train_runtime": 209826.9024,
+ "train_tokens_per_second": 5651.305
+ },
+ {
+ "epoch": 1.8260166980877997,
+ "grad_norm": 0.7995613128971399,
+ "learning_rate": 1.857052248070962e-07,
+ "loss": 0.30235800743103025,
+ "num_input_tokens_seen": 1186003992,
+ "step": 27120,
+ "train_runtime": 209868.1712,
+ "train_tokens_per_second": 5651.186
+ },
+ {
+ "epoch": 1.8263533530837597,
+ "grad_norm": 0.8063565780315961,
+ "learning_rate": 1.8499198291876907e-07,
+ "loss": 0.2838723659515381,
+ "num_input_tokens_seen": 1186206928,
+ "step": 27125,
+ "train_runtime": 209908.8621,
+ "train_tokens_per_second": 5651.057
+ },
+ {
+ "epoch": 1.82669000807972,
+ "grad_norm": 0.659822757313705,
+ "learning_rate": 1.8428008753207937e-07,
+ "loss": 0.3044004440307617,
+ "num_input_tokens_seen": 1186430512,
+ "step": 27130,
+ "train_runtime": 209952.433,
+ "train_tokens_per_second": 5650.949
+ },
+ {
+ "epoch": 1.8270266630756802,
+ "grad_norm": 0.7534355877635904,
+ "learning_rate": 1.8356953884610673e-07,
+ "loss": 0.3086956262588501,
+ "num_input_tokens_seen": 1186652392,
+ "step": 27135,
+ "train_runtime": 209996.3146,
+ "train_tokens_per_second": 5650.825
+ },
+ {
+ "epoch": 1.8273633180716402,
+ "grad_norm": 0.7536622469259906,
+ "learning_rate": 1.8286033705955554e-07,
+ "loss": 0.2920703411102295,
+ "num_input_tokens_seen": 1186878432,
+ "step": 27140,
+ "train_runtime": 210028.5754,
+ "train_tokens_per_second": 5651.033
+ },
+ {
+ "epoch": 1.8276999730676002,
+ "grad_norm": 0.8189703081756119,
+ "learning_rate": 1.8215248237074879e-07,
+ "loss": 0.2958489656448364,
+ "num_input_tokens_seen": 1187100664,
+ "step": 27145,
+ "train_runtime": 210067.8385,
+ "train_tokens_per_second": 5651.035
+ },
+ {
+ "epoch": 1.8280366280635605,
+ "grad_norm": 0.6908418891259124,
+ "learning_rate": 1.8144597497763816e-07,
+ "loss": 0.2915064334869385,
+ "num_input_tokens_seen": 1187314688,
+ "step": 27150,
+ "train_runtime": 210105.6707,
+ "train_tokens_per_second": 5651.036
+ },
+ {
+ "epoch": 1.8283732830595207,
+ "grad_norm": 0.7771306934508,
+ "learning_rate": 1.8074081507779506e-07,
+ "loss": 0.2950667142868042,
+ "num_input_tokens_seen": 1187543272,
+ "step": 27155,
+ "train_runtime": 210145.1035,
+ "train_tokens_per_second": 5651.063
+ },
+ {
+ "epoch": 1.8287099380554808,
+ "grad_norm": 0.7951975641517514,
+ "learning_rate": 1.8003700286841563e-07,
+ "loss": 0.3046703815460205,
+ "num_input_tokens_seen": 1187764808,
+ "step": 27160,
+ "train_runtime": 210185.6937,
+ "train_tokens_per_second": 5651.026
+ },
+ {
+ "epoch": 1.8290465930514408,
+ "grad_norm": 0.8247212903362624,
+ "learning_rate": 1.793345385463191e-07,
+ "loss": 0.2969427824020386,
+ "num_input_tokens_seen": 1187985984,
+ "step": 27165,
+ "train_runtime": 210225.4532,
+ "train_tokens_per_second": 5651.009
+ },
+ {
+ "epoch": 1.829383248047401,
+ "grad_norm": 0.8389229794347001,
+ "learning_rate": 1.7863342230794667e-07,
+ "loss": 0.306946849822998,
+ "num_input_tokens_seen": 1188205016,
+ "step": 27170,
+ "train_runtime": 210256.0102,
+ "train_tokens_per_second": 5651.23
+ },
+ {
+ "epoch": 1.8297199030433613,
+ "grad_norm": 0.6995002304359613,
+ "learning_rate": 1.7793365434936316e-07,
+ "loss": 0.2987263917922974,
+ "num_input_tokens_seen": 1188436064,
+ "step": 27175,
+ "train_runtime": 210296.1751,
+ "train_tokens_per_second": 5651.249
+ },
+ {
+ "epoch": 1.8300565580393213,
+ "grad_norm": 0.8195984422353864,
+ "learning_rate": 1.7723523486625704e-07,
+ "loss": 0.29759798049926756,
+ "num_input_tokens_seen": 1188648440,
+ "step": 27180,
+ "train_runtime": 210337.9552,
+ "train_tokens_per_second": 5651.136
+ },
+ {
+ "epoch": 1.8303932130352814,
+ "grad_norm": 0.7388330413443034,
+ "learning_rate": 1.7653816405393766e-07,
+ "loss": 0.3033345937728882,
+ "num_input_tokens_seen": 1188864712,
+ "step": 27185,
+ "train_runtime": 210377.2381,
+ "train_tokens_per_second": 5651.109
+ },
+ {
+ "epoch": 1.8307298680312416,
+ "grad_norm": 0.8567770406651678,
+ "learning_rate": 1.758424421073418e-07,
+ "loss": 0.308400821685791,
+ "num_input_tokens_seen": 1189093376,
+ "step": 27190,
+ "train_runtime": 210414.6866,
+ "train_tokens_per_second": 5651.19
+ },
+ {
+ "epoch": 1.8310665230272019,
+ "grad_norm": 0.7475393949595193,
+ "learning_rate": 1.7514806922102222e-07,
+ "loss": 0.2787260293960571,
+ "num_input_tokens_seen": 1189290992,
+ "step": 27195,
+ "train_runtime": 210452.2542,
+ "train_tokens_per_second": 5651.12
+ },
+ {
+ "epoch": 1.8314031780231619,
+ "grad_norm": 0.7779392879208484,
+ "learning_rate": 1.7445504558916027e-07,
+ "loss": 0.2868196964263916,
+ "num_input_tokens_seen": 1189504064,
+ "step": 27200,
+ "train_runtime": 210482.3919,
+ "train_tokens_per_second": 5651.323
+ },
+ {
+ "epoch": 1.831739833019122,
+ "grad_norm": 0.7793252529447385,
+ "learning_rate": 1.7376337140555754e-07,
+ "loss": 0.283552360534668,
+ "num_input_tokens_seen": 1189712800,
+ "step": 27205,
+ "train_runtime": 210518.9894,
+ "train_tokens_per_second": 5651.332
+ },
+ {
+ "epoch": 1.8320764880150822,
+ "grad_norm": 0.7136774949097324,
+ "learning_rate": 1.7307304686363825e-07,
+ "loss": 0.30210127830505373,
+ "num_input_tokens_seen": 1189932032,
+ "step": 27210,
+ "train_runtime": 210557.2729,
+ "train_tokens_per_second": 5651.346
+ },
+ {
+ "epoch": 1.8324131430110424,
+ "grad_norm": 0.6733587332513272,
+ "learning_rate": 1.723840721564496e-07,
+ "loss": 0.2896436214447021,
+ "num_input_tokens_seen": 1190162824,
+ "step": 27215,
+ "train_runtime": 210596.5269,
+ "train_tokens_per_second": 5651.389
+ },
+ {
+ "epoch": 1.8327497980070024,
+ "grad_norm": 0.7975602822015874,
+ "learning_rate": 1.7169644747666081e-07,
+ "loss": 0.29418466091156004,
+ "num_input_tokens_seen": 1190392976,
+ "step": 27220,
+ "train_runtime": 210635.9714,
+ "train_tokens_per_second": 5651.423
+ },
+ {
+ "epoch": 1.8330864530029625,
+ "grad_norm": 0.6803205298907224,
+ "learning_rate": 1.710101730165642e-07,
+ "loss": 0.30214219093322753,
+ "num_input_tokens_seen": 1190618688,
+ "step": 27225,
+ "train_runtime": 210673.7247,
+ "train_tokens_per_second": 5651.482
+ },
+ {
+ "epoch": 1.8334231079989227,
+ "grad_norm": 0.7468300724636248,
+ "learning_rate": 1.7032524896807402e-07,
+ "loss": 0.27968459129333495,
+ "num_input_tokens_seen": 1190825912,
+ "step": 27230,
+ "train_runtime": 210712.9044,
+ "train_tokens_per_second": 5651.414
+ },
+ {
+ "epoch": 1.833759762994883,
+ "grad_norm": 0.75132763867745,
+ "learning_rate": 1.6964167552272702e-07,
+ "loss": 0.3041102647781372,
+ "num_input_tokens_seen": 1191031648,
+ "step": 27235,
+ "train_runtime": 210750.2864,
+ "train_tokens_per_second": 5651.388
+ },
+ {
+ "epoch": 1.834096417990843,
+ "grad_norm": 0.8236438444986357,
+ "learning_rate": 1.6895945287168314e-07,
+ "loss": 0.30144453048706055,
+ "num_input_tokens_seen": 1191243504,
+ "step": 27240,
+ "train_runtime": 210787.0749,
+ "train_tokens_per_second": 5651.407
+ },
+ {
+ "epoch": 1.834433072986803,
+ "grad_norm": 0.8116073242293598,
+ "learning_rate": 1.682785812057225e-07,
+ "loss": 0.31502351760864256,
+ "num_input_tokens_seen": 1191444928,
+ "step": 27245,
+ "train_runtime": 210818.024,
+ "train_tokens_per_second": 5651.533
+ },
+ {
+ "epoch": 1.8347697279827633,
+ "grad_norm": 0.7705506997654222,
+ "learning_rate": 1.6759906071525e-07,
+ "loss": 0.2920320987701416,
+ "num_input_tokens_seen": 1191662640,
+ "step": 27250,
+ "train_runtime": 210862.0161,
+ "train_tokens_per_second": 5651.386
+ },
+ {
+ "epoch": 1.8351063829787235,
+ "grad_norm": 0.7712074370148209,
+ "learning_rate": 1.669208915902909e-07,
+ "loss": 0.33261871337890625,
+ "num_input_tokens_seen": 1191900592,
+ "step": 27255,
+ "train_runtime": 210897.8718,
+ "train_tokens_per_second": 5651.553
+ },
+ {
+ "epoch": 1.8354430379746836,
+ "grad_norm": 0.7737113976334139,
+ "learning_rate": 1.6624407402049237e-07,
+ "loss": 0.2747644901275635,
+ "num_input_tokens_seen": 1192112584,
+ "step": 27260,
+ "train_runtime": 210931.9902,
+ "train_tokens_per_second": 5651.644
+ },
+ {
+ "epoch": 1.8357796929706436,
+ "grad_norm": 0.7765514097289985,
+ "learning_rate": 1.655686081951252e-07,
+ "loss": 0.28942217826843264,
+ "num_input_tokens_seen": 1192328048,
+ "step": 27265,
+ "train_runtime": 210962.6642,
+ "train_tokens_per_second": 5651.844
+ },
+ {
+ "epoch": 1.8361163479666038,
+ "grad_norm": 0.7986187558729846,
+ "learning_rate": 1.6489449430308058e-07,
+ "loss": 0.32457714080810546,
+ "num_input_tokens_seen": 1192555872,
+ "step": 27270,
+ "train_runtime": 211006.3008,
+ "train_tokens_per_second": 5651.755
+ },
+ {
+ "epoch": 1.836453002962564,
+ "grad_norm": 0.7814336439979509,
+ "learning_rate": 1.6422173253287323e-07,
+ "loss": 0.31779980659484863,
+ "num_input_tokens_seen": 1192784792,
+ "step": 27275,
+ "train_runtime": 211044.4069,
+ "train_tokens_per_second": 5651.819
+ },
+ {
+ "epoch": 1.8367896579585241,
+ "grad_norm": 0.8231817333366194,
+ "learning_rate": 1.6355032307263764e-07,
+ "loss": 0.2909559726715088,
+ "num_input_tokens_seen": 1193008024,
+ "step": 27280,
+ "train_runtime": 211087.1712,
+ "train_tokens_per_second": 5651.732
+ },
+ {
+ "epoch": 1.8371263129544841,
+ "grad_norm": 0.7297603518728784,
+ "learning_rate": 1.628802661101314e-07,
+ "loss": 0.27639720439910886,
+ "num_input_tokens_seen": 1193222040,
+ "step": 27285,
+ "train_runtime": 211126.1642,
+ "train_tokens_per_second": 5651.701
+ },
+ {
+ "epoch": 1.8374629679504444,
+ "grad_norm": 0.7393851625213684,
+ "learning_rate": 1.6221156183273524e-07,
+ "loss": 0.2929367542266846,
+ "num_input_tokens_seen": 1193445752,
+ "step": 27290,
+ "train_runtime": 211166.2819,
+ "train_tokens_per_second": 5651.687
+ },
+ {
+ "epoch": 1.8377996229464046,
+ "grad_norm": 0.8626095075985437,
+ "learning_rate": 1.615442104274484e-07,
+ "loss": 0.2990311622619629,
+ "num_input_tokens_seen": 1193676024,
+ "step": 27295,
+ "train_runtime": 211203.0891,
+ "train_tokens_per_second": 5651.792
+ },
+ {
+ "epoch": 1.8381362779423647,
+ "grad_norm": 0.7989960018689986,
+ "learning_rate": 1.6087821208089492e-07,
+ "loss": 0.2992110252380371,
+ "num_input_tokens_seen": 1193898896,
+ "step": 27300,
+ "train_runtime": 211239.9537,
+ "train_tokens_per_second": 5651.861
+ },
+ {
+ "epoch": 1.8384729329383247,
+ "grad_norm": 0.675694068212015,
+ "learning_rate": 1.6021356697931866e-07,
+ "loss": 0.2737618684768677,
+ "num_input_tokens_seen": 1194124640,
+ "step": 27305,
+ "train_runtime": 211274.8199,
+ "train_tokens_per_second": 5651.997
+ },
+ {
+ "epoch": 1.838809587934285,
+ "grad_norm": 0.7414825082624487,
+ "learning_rate": 1.5955027530858537e-07,
+ "loss": 0.27455065250396726,
+ "num_input_tokens_seen": 1194345288,
+ "step": 27310,
+ "train_runtime": 211311.0878,
+ "train_tokens_per_second": 5652.071
+ },
+ {
+ "epoch": 1.8391462429302452,
+ "grad_norm": 0.6826873569949812,
+ "learning_rate": 1.5888833725418275e-07,
+ "loss": 0.2740659236907959,
+ "num_input_tokens_seen": 1194562672,
+ "step": 27315,
+ "train_runtime": 211347.4146,
+ "train_tokens_per_second": 5652.128
+ },
+ {
+ "epoch": 1.8394828979262052,
+ "grad_norm": 0.7227868616604537,
+ "learning_rate": 1.582277530012194e-07,
+ "loss": 0.26207771301269533,
+ "num_input_tokens_seen": 1194797416,
+ "step": 27320,
+ "train_runtime": 211380.4831,
+ "train_tokens_per_second": 5652.354
+ },
+ {
+ "epoch": 1.8398195529221653,
+ "grad_norm": 0.8448912125013232,
+ "learning_rate": 1.5756852273442536e-07,
+ "loss": 0.29030356407165525,
+ "num_input_tokens_seen": 1195005200,
+ "step": 27325,
+ "train_runtime": 211422.3023,
+ "train_tokens_per_second": 5652.219
+ },
+ {
+ "epoch": 1.8401562079181255,
+ "grad_norm": 0.697440308697788,
+ "learning_rate": 1.569106466381526e-07,
+ "loss": 0.3182931900024414,
+ "num_input_tokens_seen": 1195227200,
+ "step": 27330,
+ "train_runtime": 211462.6591,
+ "train_tokens_per_second": 5652.19
+ },
+ {
+ "epoch": 1.8404928629140858,
+ "grad_norm": 0.7422044369619762,
+ "learning_rate": 1.5625412489637337e-07,
+ "loss": 0.27002391815185545,
+ "num_input_tokens_seen": 1195454536,
+ "step": 27335,
+ "train_runtime": 211507.1847,
+ "train_tokens_per_second": 5652.075
+ },
+ {
+ "epoch": 1.8408295179100458,
+ "grad_norm": 0.7443593742932311,
+ "learning_rate": 1.555989576926842e-07,
+ "loss": 0.2888377904891968,
+ "num_input_tokens_seen": 1195681944,
+ "step": 27340,
+ "train_runtime": 211549.0676,
+ "train_tokens_per_second": 5652.031
+ },
+ {
+ "epoch": 1.8411661729060058,
+ "grad_norm": 0.801773457702004,
+ "learning_rate": 1.5494514521029736e-07,
+ "loss": 0.29750452041625974,
+ "num_input_tokens_seen": 1195905376,
+ "step": 27345,
+ "train_runtime": 211588.7755,
+ "train_tokens_per_second": 5652.027
+ },
+ {
+ "epoch": 1.841502827901966,
+ "grad_norm": 0.7523395284692173,
+ "learning_rate": 1.5429268763205108e-07,
+ "loss": 0.2998412370681763,
+ "num_input_tokens_seen": 1196127744,
+ "step": 27350,
+ "train_runtime": 211635.9655,
+ "train_tokens_per_second": 5651.817
+ },
+ {
+ "epoch": 1.8418394828979263,
+ "grad_norm": 0.8119149425227684,
+ "learning_rate": 1.5364158514040328e-07,
+ "loss": 0.3214261531829834,
+ "num_input_tokens_seen": 1196344424,
+ "step": 27355,
+ "train_runtime": 211676.2418,
+ "train_tokens_per_second": 5651.765
+ },
+ {
+ "epoch": 1.8421761378938863,
+ "grad_norm": 0.9267135845555328,
+ "learning_rate": 1.5299183791743223e-07,
+ "loss": 0.30311391353607176,
+ "num_input_tokens_seen": 1196551984,
+ "step": 27360,
+ "train_runtime": 211712.3977,
+ "train_tokens_per_second": 5651.78
+ },
+ {
+ "epoch": 1.8425127928898464,
+ "grad_norm": 0.79776476761183,
+ "learning_rate": 1.523434461448381e-07,
+ "loss": 0.31781580448150637,
+ "num_input_tokens_seen": 1196784808,
+ "step": 27365,
+ "train_runtime": 211750.7126,
+ "train_tokens_per_second": 5651.857
+ },
+ {
+ "epoch": 1.8428494478858066,
+ "grad_norm": 0.7201171985633329,
+ "learning_rate": 1.5169641000394086e-07,
+ "loss": 0.29927232265472414,
+ "num_input_tokens_seen": 1197018080,
+ "step": 27370,
+ "train_runtime": 211789.2549,
+ "train_tokens_per_second": 5651.93
+ },
+ {
+ "epoch": 1.8431861028817669,
+ "grad_norm": 0.730174194835056,
+ "learning_rate": 1.5105072967568246e-07,
+ "loss": 0.2902745246887207,
+ "num_input_tokens_seen": 1197229392,
+ "step": 27375,
+ "train_runtime": 211827.3609,
+ "train_tokens_per_second": 5651.911
+ },
+ {
+ "epoch": 1.843522757877727,
+ "grad_norm": 0.6910545329116042,
+ "learning_rate": 1.504064053406251e-07,
+ "loss": 0.29667439460754397,
+ "num_input_tokens_seen": 1197443992,
+ "step": 27380,
+ "train_runtime": 211869.9495,
+ "train_tokens_per_second": 5651.788
+ },
+ {
+ "epoch": 1.843859412873687,
+ "grad_norm": 0.7540793648170933,
+ "learning_rate": 1.4976343717895192e-07,
+ "loss": 0.3001372814178467,
+ "num_input_tokens_seen": 1197673640,
+ "step": 27385,
+ "train_runtime": 211902.7264,
+ "train_tokens_per_second": 5651.997
+ },
+ {
+ "epoch": 1.8441960678696472,
+ "grad_norm": 0.7814817716047526,
+ "learning_rate": 1.491218253704685e-07,
+ "loss": 0.30607056617736816,
+ "num_input_tokens_seen": 1197898760,
+ "step": 27390,
+ "train_runtime": 211943.3379,
+ "train_tokens_per_second": 5651.976
+ },
+ {
+ "epoch": 1.8445327228656074,
+ "grad_norm": 0.7515118407147103,
+ "learning_rate": 1.484815700945974e-07,
+ "loss": 0.30979294776916505,
+ "num_input_tokens_seen": 1198115856,
+ "step": 27395,
+ "train_runtime": 211982.4794,
+ "train_tokens_per_second": 5651.957
+ },
+ {
+ "epoch": 1.8448693778615675,
+ "grad_norm": 0.7192859496811161,
+ "learning_rate": 1.478426715303849e-07,
+ "loss": 0.2884451150894165,
+ "num_input_tokens_seen": 1198320608,
+ "step": 27400,
+ "train_runtime": 212021.1992,
+ "train_tokens_per_second": 5651.891
+ },
+ {
+ "epoch": 1.8452060328575275,
+ "grad_norm": 0.7754550279674132,
+ "learning_rate": 1.4720512985649694e-07,
+ "loss": 0.29123849868774415,
+ "num_input_tokens_seen": 1198532416,
+ "step": 27405,
+ "train_runtime": 212063.5064,
+ "train_tokens_per_second": 5651.762
+ },
+ {
+ "epoch": 1.8455426878534877,
+ "grad_norm": 0.9501576369289315,
+ "learning_rate": 1.4656894525122034e-07,
+ "loss": 0.30306599140167234,
+ "num_input_tokens_seen": 1198734552,
+ "step": 27410,
+ "train_runtime": 212105.602,
+ "train_tokens_per_second": 5651.593
+ },
+ {
+ "epoch": 1.845879342849448,
+ "grad_norm": 0.7974261229774522,
+ "learning_rate": 1.459341178924617e-07,
+ "loss": 0.30881667137145996,
+ "num_input_tokens_seen": 1198955512,
+ "step": 27415,
+ "train_runtime": 212144.96,
+ "train_tokens_per_second": 5651.586
+ },
+ {
+ "epoch": 1.846215997845408,
+ "grad_norm": 0.7988079730512095,
+ "learning_rate": 1.453006479577479e-07,
+ "loss": 0.28842964172363283,
+ "num_input_tokens_seen": 1199167896,
+ "step": 27420,
+ "train_runtime": 212180.9912,
+ "train_tokens_per_second": 5651.627
+ },
+ {
+ "epoch": 1.846552652841368,
+ "grad_norm": 0.7202060611505903,
+ "learning_rate": 1.4466853562422778e-07,
+ "loss": 0.3094759464263916,
+ "num_input_tokens_seen": 1199366136,
+ "step": 27425,
+ "train_runtime": 212214.506,
+ "train_tokens_per_second": 5651.669
+ },
+ {
+ "epoch": 1.8468893078373283,
+ "grad_norm": 0.8214802158441638,
+ "learning_rate": 1.440377810686694e-07,
+ "loss": 0.294615650177002,
+ "num_input_tokens_seen": 1199589824,
+ "step": 27430,
+ "train_runtime": 212247.8835,
+ "train_tokens_per_second": 5651.834
+ },
+ {
+ "epoch": 1.8472259628332885,
+ "grad_norm": 0.7935797505217701,
+ "learning_rate": 1.4340838446746052e-07,
+ "loss": 0.2963638067245483,
+ "num_input_tokens_seen": 1199816200,
+ "step": 27435,
+ "train_runtime": 212284.2378,
+ "train_tokens_per_second": 5651.933
+ },
+ {
+ "epoch": 1.8475626178292486,
+ "grad_norm": 0.7255235631748863,
+ "learning_rate": 1.4278034599661094e-07,
+ "loss": 0.2769758701324463,
+ "num_input_tokens_seen": 1200031328,
+ "step": 27440,
+ "train_runtime": 212314.5115,
+ "train_tokens_per_second": 5652.14
+ },
+ {
+ "epoch": 1.8478992728252086,
+ "grad_norm": 0.78374561254918,
+ "learning_rate": 1.421536658317485e-07,
+ "loss": 0.2689784288406372,
+ "num_input_tokens_seen": 1200227096,
+ "step": 27445,
+ "train_runtime": 212349.6526,
+ "train_tokens_per_second": 5652.126
+ },
+ {
+ "epoch": 1.8482359278211689,
+ "grad_norm": 0.7784687447417672,
+ "learning_rate": 1.41528344148123e-07,
+ "loss": 0.28493242263793944,
+ "num_input_tokens_seen": 1200462264,
+ "step": 27450,
+ "train_runtime": 212386.6742,
+ "train_tokens_per_second": 5652.249
+ },
+ {
+ "epoch": 1.848572582817129,
+ "grad_norm": 0.8647273624224125,
+ "learning_rate": 1.40904381120604e-07,
+ "loss": 0.2554482936859131,
+ "num_input_tokens_seen": 1200679968,
+ "step": 27455,
+ "train_runtime": 212417.8431,
+ "train_tokens_per_second": 5652.444
+ },
+ {
+ "epoch": 1.8489092378130891,
+ "grad_norm": 0.7867893330661792,
+ "learning_rate": 1.4028177692367973e-07,
+ "loss": 0.2979857921600342,
+ "num_input_tokens_seen": 1200888896,
+ "step": 27460,
+ "train_runtime": 212455.3142,
+ "train_tokens_per_second": 5652.43
+ },
+ {
+ "epoch": 1.8492458928090492,
+ "grad_norm": 0.9388552543962256,
+ "learning_rate": 1.3966053173146032e-07,
+ "loss": 0.29952003955841067,
+ "num_input_tokens_seen": 1201115544,
+ "step": 27465,
+ "train_runtime": 212489.4425,
+ "train_tokens_per_second": 5652.589
+ },
+ {
+ "epoch": 1.8495825478050094,
+ "grad_norm": 0.6831577223182406,
+ "learning_rate": 1.3904064571767462e-07,
+ "loss": 0.29733192920684814,
+ "num_input_tokens_seen": 1201343992,
+ "step": 27470,
+ "train_runtime": 212529.6948,
+ "train_tokens_per_second": 5652.594
+ },
+ {
+ "epoch": 1.8499192028009697,
+ "grad_norm": 0.7057151389200952,
+ "learning_rate": 1.3842211905567227e-07,
+ "loss": 0.2762616157531738,
+ "num_input_tokens_seen": 1201569224,
+ "step": 27475,
+ "train_runtime": 212561.6614,
+ "train_tokens_per_second": 5652.803
+ },
+ {
+ "epoch": 1.8502558577969297,
+ "grad_norm": 0.8452116887810223,
+ "learning_rate": 1.378049519184216e-07,
+ "loss": 0.2911623239517212,
+ "num_input_tokens_seen": 1201779840,
+ "step": 27480,
+ "train_runtime": 212599.5548,
+ "train_tokens_per_second": 5652.786
+ },
+ {
+ "epoch": 1.8505925127928897,
+ "grad_norm": 0.7069439995002406,
+ "learning_rate": 1.3718914447851116e-07,
+ "loss": 0.2653970718383789,
+ "num_input_tokens_seen": 1202007472,
+ "step": 27485,
+ "train_runtime": 212638.9418,
+ "train_tokens_per_second": 5652.81
+ },
+ {
+ "epoch": 1.85092916778885,
+ "grad_norm": 0.767853195090442,
+ "learning_rate": 1.3657469690815106e-07,
+ "loss": 0.28401851654052734,
+ "num_input_tokens_seen": 1202246288,
+ "step": 27490,
+ "train_runtime": 212671.529,
+ "train_tokens_per_second": 5653.066
+ },
+ {
+ "epoch": 1.8512658227848102,
+ "grad_norm": 0.7219513957308681,
+ "learning_rate": 1.359616093791677e-07,
+ "loss": 0.3170768737792969,
+ "num_input_tokens_seen": 1202478016,
+ "step": 27495,
+ "train_runtime": 212715.4107,
+ "train_tokens_per_second": 5652.99
+ },
+ {
+ "epoch": 1.8516024777807703,
+ "grad_norm": 0.7333849781257291,
+ "learning_rate": 1.353498820630106e-07,
+ "loss": 0.29583718776702883,
+ "num_input_tokens_seen": 1202698624,
+ "step": 27500,
+ "train_runtime": 212754.0646,
+ "train_tokens_per_second": 5653.0
+ },
+ {
+ "epoch": 1.8519391327767303,
+ "grad_norm": 0.7743102457437921,
+ "learning_rate": 1.3473951513074625e-07,
+ "loss": 0.31031112670898436,
+ "num_input_tokens_seen": 1202918504,
+ "step": 27505,
+ "train_runtime": 212795.3126,
+ "train_tokens_per_second": 5652.937
+ },
+ {
+ "epoch": 1.8522757877726905,
+ "grad_norm": 0.9893994757413681,
+ "learning_rate": 1.341305087530631e-07,
+ "loss": 0.29344775676727297,
+ "num_input_tokens_seen": 1203135384,
+ "step": 27510,
+ "train_runtime": 212826.4122,
+ "train_tokens_per_second": 5653.13
+ },
+ {
+ "epoch": 1.8526124427686508,
+ "grad_norm": 0.8824830489862262,
+ "learning_rate": 1.3352286310026663e-07,
+ "loss": 0.3105172634124756,
+ "num_input_tokens_seen": 1203348136,
+ "step": 27515,
+ "train_runtime": 212864.5316,
+ "train_tokens_per_second": 5653.117
+ },
+ {
+ "epoch": 1.8529490977646108,
+ "grad_norm": 0.8177611033215039,
+ "learning_rate": 1.329165783422831e-07,
+ "loss": 0.31962890625,
+ "num_input_tokens_seen": 1203560776,
+ "step": 27520,
+ "train_runtime": 212898.9135,
+ "train_tokens_per_second": 5653.203
+ },
+ {
+ "epoch": 1.8532857527605708,
+ "grad_norm": 0.8203874085361618,
+ "learning_rate": 1.3231165464865914e-07,
+ "loss": 0.29946393966674806,
+ "num_input_tokens_seen": 1203778736,
+ "step": 27525,
+ "train_runtime": 212929.6442,
+ "train_tokens_per_second": 5653.411
+ },
+ {
+ "epoch": 1.853622407756531,
+ "grad_norm": 0.6941728137863973,
+ "learning_rate": 1.3170809218855884e-07,
+ "loss": 0.29173548221588136,
+ "num_input_tokens_seen": 1204009952,
+ "step": 27530,
+ "train_runtime": 212966.9944,
+ "train_tokens_per_second": 5653.505
+ },
+ {
+ "epoch": 1.8539590627524913,
+ "grad_norm": 0.7342203887402693,
+ "learning_rate": 1.3110589113076554e-07,
+ "loss": 0.2878772258758545,
+ "num_input_tokens_seen": 1204230200,
+ "step": 27535,
+ "train_runtime": 213008.2192,
+ "train_tokens_per_second": 5653.445
+ },
+ {
+ "epoch": 1.8542957177484514,
+ "grad_norm": 0.7315091280926438,
+ "learning_rate": 1.3050505164368566e-07,
+ "loss": 0.29323248863220214,
+ "num_input_tokens_seen": 1204451192,
+ "step": 27540,
+ "train_runtime": 213039.3969,
+ "train_tokens_per_second": 5653.655
+ },
+ {
+ "epoch": 1.8546323727444114,
+ "grad_norm": 0.8014357486751242,
+ "learning_rate": 1.299055738953392e-07,
+ "loss": 0.2807599067687988,
+ "num_input_tokens_seen": 1204650768,
+ "step": 27545,
+ "train_runtime": 213080.939,
+ "train_tokens_per_second": 5653.489
+ },
+ {
+ "epoch": 1.8549690277403716,
+ "grad_norm": 0.7816529327087509,
+ "learning_rate": 1.293074580533704e-07,
+ "loss": 0.28776376247406005,
+ "num_input_tokens_seen": 1204864152,
+ "step": 27550,
+ "train_runtime": 213127.338,
+ "train_tokens_per_second": 5653.26
+ },
+ {
+ "epoch": 1.855305682736332,
+ "grad_norm": 0.7431632363177251,
+ "learning_rate": 1.2871070428503885e-07,
+ "loss": 0.3120842456817627,
+ "num_input_tokens_seen": 1205072184,
+ "step": 27555,
+ "train_runtime": 213166.7662,
+ "train_tokens_per_second": 5653.19
+ },
+ {
+ "epoch": 1.855642337732292,
+ "grad_norm": 0.6740193343425465,
+ "learning_rate": 1.2811531275722543e-07,
+ "loss": 0.3159989595413208,
+ "num_input_tokens_seen": 1205302840,
+ "step": 27560,
+ "train_runtime": 213202.2706,
+ "train_tokens_per_second": 5653.33
+ },
+ {
+ "epoch": 1.855978992728252,
+ "grad_norm": 0.850999270296055,
+ "learning_rate": 1.275212836364298e-07,
+ "loss": 0.30814902782440184,
+ "num_input_tokens_seen": 1205521608,
+ "step": 27565,
+ "train_runtime": 213246.1453,
+ "train_tokens_per_second": 5653.193
+ },
+ {
+ "epoch": 1.8563156477242122,
+ "grad_norm": 0.7919555393902858,
+ "learning_rate": 1.2692861708877014e-07,
+ "loss": 0.28333234786987305,
+ "num_input_tokens_seen": 1205730248,
+ "step": 27570,
+ "train_runtime": 213289.0034,
+ "train_tokens_per_second": 5653.035
+ },
+ {
+ "epoch": 1.8566523027201725,
+ "grad_norm": 0.7321483909457048,
+ "learning_rate": 1.2633731327998332e-07,
+ "loss": 0.2803944110870361,
+ "num_input_tokens_seen": 1205950784,
+ "step": 27575,
+ "train_runtime": 213322.9311,
+ "train_tokens_per_second": 5653.17
+ },
+ {
+ "epoch": 1.8569889577161325,
+ "grad_norm": 0.9024845578541033,
+ "learning_rate": 1.25747372375426e-07,
+ "loss": 0.2885521411895752,
+ "num_input_tokens_seen": 1206159224,
+ "step": 27580,
+ "train_runtime": 213368.2449,
+ "train_tokens_per_second": 5652.946
+ },
+ {
+ "epoch": 1.8573256127120925,
+ "grad_norm": 0.8389667628017906,
+ "learning_rate": 1.251587945400723e-07,
+ "loss": 0.3000046253204346,
+ "num_input_tokens_seen": 1206379280,
+ "step": 27585,
+ "train_runtime": 213408.9939,
+ "train_tokens_per_second": 5652.898
+ },
+ {
+ "epoch": 1.8576622677080528,
+ "grad_norm": 0.7189413123955991,
+ "learning_rate": 1.2457157993851832e-07,
+ "loss": 0.2905135154724121,
+ "num_input_tokens_seen": 1206604064,
+ "step": 27590,
+ "train_runtime": 213446.507,
+ "train_tokens_per_second": 5652.958
+ },
+ {
+ "epoch": 1.857998922704013,
+ "grad_norm": 0.756958619394132,
+ "learning_rate": 1.2398572873497438e-07,
+ "loss": 0.2975008487701416,
+ "num_input_tokens_seen": 1206821040,
+ "step": 27595,
+ "train_runtime": 213487.2557,
+ "train_tokens_per_second": 5652.895
+ },
+ {
+ "epoch": 1.858335577699973,
+ "grad_norm": 0.7936162080166628,
+ "learning_rate": 1.2340124109327334e-07,
+ "loss": 0.2905877590179443,
+ "num_input_tokens_seen": 1207029960,
+ "step": 27600,
+ "train_runtime": 213526.3845,
+ "train_tokens_per_second": 5652.838
+ },
+ {
+ "epoch": 1.858672232695933,
+ "grad_norm": 0.7617351149236978,
+ "learning_rate": 1.22818117176865e-07,
+ "loss": 0.3129683494567871,
+ "num_input_tokens_seen": 1207251408,
+ "step": 27605,
+ "train_runtime": 213571.4503,
+ "train_tokens_per_second": 5652.682
+ },
+ {
+ "epoch": 1.8590088876918933,
+ "grad_norm": 0.6863825753058197,
+ "learning_rate": 1.2223635714881777e-07,
+ "loss": 0.30173707008361816,
+ "num_input_tokens_seen": 1207460752,
+ "step": 27610,
+ "train_runtime": 213606.2956,
+ "train_tokens_per_second": 5652.74
+ },
+ {
+ "epoch": 1.8593455426878536,
+ "grad_norm": 0.6602683521069512,
+ "learning_rate": 1.2165596117181931e-07,
+ "loss": 0.28788735866546633,
+ "num_input_tokens_seen": 1207688200,
+ "step": 27615,
+ "train_runtime": 213640.7332,
+ "train_tokens_per_second": 5652.893
+ },
+ {
+ "epoch": 1.8596821976838136,
+ "grad_norm": 0.7244916772239959,
+ "learning_rate": 1.2107692940817594e-07,
+ "loss": 0.30665135383605957,
+ "num_input_tokens_seen": 1207904320,
+ "step": 27620,
+ "train_runtime": 213682.5989,
+ "train_tokens_per_second": 5652.797
+ },
+ {
+ "epoch": 1.8600188526797736,
+ "grad_norm": 0.8628325975987942,
+ "learning_rate": 1.2049926201981088e-07,
+ "loss": 0.2934282302856445,
+ "num_input_tokens_seen": 1208124344,
+ "step": 27625,
+ "train_runtime": 213723.4491,
+ "train_tokens_per_second": 5652.746
+ },
+ {
+ "epoch": 1.8603555076757339,
+ "grad_norm": 0.719201813819039,
+ "learning_rate": 1.199229591682677e-07,
+ "loss": 0.3092188835144043,
+ "num_input_tokens_seen": 1208348632,
+ "step": 27630,
+ "train_runtime": 213757.6689,
+ "train_tokens_per_second": 5652.89
+ },
+ {
+ "epoch": 1.8606921626716941,
+ "grad_norm": 0.8213286351201897,
+ "learning_rate": 1.193480210147069e-07,
+ "loss": 0.3101773262023926,
+ "num_input_tokens_seen": 1208576560,
+ "step": 27635,
+ "train_runtime": 213793.1342,
+ "train_tokens_per_second": 5653.019
+ },
+ {
+ "epoch": 1.8610288176676542,
+ "grad_norm": 0.7546904535243726,
+ "learning_rate": 1.1877444771991042e-07,
+ "loss": 0.2830918788909912,
+ "num_input_tokens_seen": 1208805448,
+ "step": 27640,
+ "train_runtime": 213830.9044,
+ "train_tokens_per_second": 5653.09
+ },
+ {
+ "epoch": 1.8613654726636142,
+ "grad_norm": 0.7654042884672665,
+ "learning_rate": 1.1820223944427279e-07,
+ "loss": 0.28035907745361327,
+ "num_input_tokens_seen": 1209015368,
+ "step": 27645,
+ "train_runtime": 213873.433,
+ "train_tokens_per_second": 5652.948
+ },
+ {
+ "epoch": 1.8617021276595744,
+ "grad_norm": 0.7531392542074884,
+ "learning_rate": 1.1763139634781262e-07,
+ "loss": 0.29651951789855957,
+ "num_input_tokens_seen": 1209240952,
+ "step": 27650,
+ "train_runtime": 213911.0698,
+ "train_tokens_per_second": 5653.008
+ },
+ {
+ "epoch": 1.8620387826555347,
+ "grad_norm": 0.7502735387914459,
+ "learning_rate": 1.1706191859016392e-07,
+ "loss": 0.3263824224472046,
+ "num_input_tokens_seen": 1209461064,
+ "step": 27655,
+ "train_runtime": 213953.9938,
+ "train_tokens_per_second": 5652.902
+ },
+ {
+ "epoch": 1.8623754376514947,
+ "grad_norm": 0.8331137280204224,
+ "learning_rate": 1.1649380633057927e-07,
+ "loss": 0.3037287950515747,
+ "num_input_tokens_seen": 1209677248,
+ "step": 27660,
+ "train_runtime": 213997.5656,
+ "train_tokens_per_second": 5652.762
+ },
+ {
+ "epoch": 1.8627120926474547,
+ "grad_norm": 0.7575852346311494,
+ "learning_rate": 1.1592705972792939e-07,
+ "loss": 0.27420315742492674,
+ "num_input_tokens_seen": 1209892576,
+ "step": 27665,
+ "train_runtime": 214039.3874,
+ "train_tokens_per_second": 5652.663
+ },
+ {
+ "epoch": 1.863048747643415,
+ "grad_norm": 0.7320879410064924,
+ "learning_rate": 1.1536167894070249e-07,
+ "loss": 0.3063500165939331,
+ "num_input_tokens_seen": 1210118104,
+ "step": 27670,
+ "train_runtime": 214082.5531,
+ "train_tokens_per_second": 5652.577
+ },
+ {
+ "epoch": 1.8633854026393752,
+ "grad_norm": 0.7458641244137307,
+ "learning_rate": 1.1479766412700655e-07,
+ "loss": 0.33033363819122313,
+ "num_input_tokens_seen": 1210345472,
+ "step": 27675,
+ "train_runtime": 214121.2922,
+ "train_tokens_per_second": 5652.616
+ },
+ {
+ "epoch": 1.8637220576353353,
+ "grad_norm": 0.7595248981178473,
+ "learning_rate": 1.1423501544456595e-07,
+ "loss": 0.31739757061004636,
+ "num_input_tokens_seen": 1210558512,
+ "step": 27680,
+ "train_runtime": 214156.2124,
+ "train_tokens_per_second": 5652.689
+ },
+ {
+ "epoch": 1.8640587126312953,
+ "grad_norm": 0.7625771638849843,
+ "learning_rate": 1.1367373305072315e-07,
+ "loss": 0.2853879451751709,
+ "num_input_tokens_seen": 1210779944,
+ "step": 27685,
+ "train_runtime": 214192.52,
+ "train_tokens_per_second": 5652.765
+ },
+ {
+ "epoch": 1.8643953676272556,
+ "grad_norm": 0.7246559600061148,
+ "learning_rate": 1.1311381710244096e-07,
+ "loss": 0.2670130729675293,
+ "num_input_tokens_seen": 1211003328,
+ "step": 27690,
+ "train_runtime": 214233.2878,
+ "train_tokens_per_second": 5652.732
+ },
+ {
+ "epoch": 1.8647320226232158,
+ "grad_norm": 0.8080414702851416,
+ "learning_rate": 1.1255526775629521e-07,
+ "loss": 0.27195014953613283,
+ "num_input_tokens_seen": 1211215040,
+ "step": 27695,
+ "train_runtime": 214278.5375,
+ "train_tokens_per_second": 5652.526
+ },
+ {
+ "epoch": 1.8650686776191758,
+ "grad_norm": 0.7987137339242992,
+ "learning_rate": 1.1199808516848487e-07,
+ "loss": 0.30287680625915525,
+ "num_input_tokens_seen": 1211429152,
+ "step": 27700,
+ "train_runtime": 214324.0073,
+ "train_tokens_per_second": 5652.326
+ },
+ {
+ "epoch": 1.8654053326151359,
+ "grad_norm": 0.7407863475801888,
+ "learning_rate": 1.1144226949482306e-07,
+ "loss": 0.28324642181396487,
+ "num_input_tokens_seen": 1211638376,
+ "step": 27705,
+ "train_runtime": 214360.0889,
+ "train_tokens_per_second": 5652.351
+ },
+ {
+ "epoch": 1.865741987611096,
+ "grad_norm": 0.6793586684952472,
+ "learning_rate": 1.1088782089074157e-07,
+ "loss": 0.27295451164245604,
+ "num_input_tokens_seen": 1211854440,
+ "step": 27710,
+ "train_runtime": 214400.1883,
+ "train_tokens_per_second": 5652.301
+ },
+ {
+ "epoch": 1.8660786426070564,
+ "grad_norm": 0.8291474455329964,
+ "learning_rate": 1.1033473951129137e-07,
+ "loss": 0.3024445056915283,
+ "num_input_tokens_seen": 1212077896,
+ "step": 27715,
+ "train_runtime": 214434.9545,
+ "train_tokens_per_second": 5652.427
+ },
+ {
+ "epoch": 1.8664152976030164,
+ "grad_norm": 0.7458532162025366,
+ "learning_rate": 1.0978302551113929e-07,
+ "loss": 0.3134200096130371,
+ "num_input_tokens_seen": 1212305392,
+ "step": 27720,
+ "train_runtime": 214476.0312,
+ "train_tokens_per_second": 5652.405
+ },
+ {
+ "epoch": 1.8667519525989764,
+ "grad_norm": 0.7681313743079988,
+ "learning_rate": 1.0923267904457025e-07,
+ "loss": 0.30133209228515623,
+ "num_input_tokens_seen": 1212533032,
+ "step": 27725,
+ "train_runtime": 214516.8657,
+ "train_tokens_per_second": 5652.39
+ },
+ {
+ "epoch": 1.8670886075949367,
+ "grad_norm": 0.7916528062724383,
+ "learning_rate": 1.086837002654867e-07,
+ "loss": 0.30214452743530273,
+ "num_input_tokens_seen": 1212743440,
+ "step": 27730,
+ "train_runtime": 214558.688,
+ "train_tokens_per_second": 5652.269
+ },
+ {
+ "epoch": 1.867425262590897,
+ "grad_norm": 0.7468188583517545,
+ "learning_rate": 1.0813608932740915e-07,
+ "loss": 0.3034868001937866,
+ "num_input_tokens_seen": 1212977168,
+ "step": 27735,
+ "train_runtime": 214594.7214,
+ "train_tokens_per_second": 5652.409
+ },
+ {
+ "epoch": 1.867761917586857,
+ "grad_norm": 0.7831956841711861,
+ "learning_rate": 1.0758984638347569e-07,
+ "loss": 0.2956989765167236,
+ "num_input_tokens_seen": 1213207904,
+ "step": 27740,
+ "train_runtime": 214629.3326,
+ "train_tokens_per_second": 5652.573
+ },
+ {
+ "epoch": 1.868098572582817,
+ "grad_norm": 0.7955988160557622,
+ "learning_rate": 1.0704497158644022e-07,
+ "loss": 0.32259302139282225,
+ "num_input_tokens_seen": 1213425128,
+ "step": 27745,
+ "train_runtime": 214673.9293,
+ "train_tokens_per_second": 5652.41
+ },
+ {
+ "epoch": 1.8684352275787772,
+ "grad_norm": 0.6842277034050441,
+ "learning_rate": 1.0650146508867642e-07,
+ "loss": 0.2915182590484619,
+ "num_input_tokens_seen": 1213645624,
+ "step": 27750,
+ "train_runtime": 214712.349,
+ "train_tokens_per_second": 5652.426
+ },
+ {
+ "epoch": 1.8687718825747375,
+ "grad_norm": 0.7772804158541299,
+ "learning_rate": 1.0595932704217327e-07,
+ "loss": 0.2860418796539307,
+ "num_input_tokens_seen": 1213868720,
+ "step": 27755,
+ "train_runtime": 214747.403,
+ "train_tokens_per_second": 5652.542
+ },
+ {
+ "epoch": 1.8691085375706975,
+ "grad_norm": 0.7387951063630381,
+ "learning_rate": 1.0541855759853837e-07,
+ "loss": 0.2798731327056885,
+ "num_input_tokens_seen": 1214099608,
+ "step": 27760,
+ "train_runtime": 214776.3703,
+ "train_tokens_per_second": 5652.855
+ },
+ {
+ "epoch": 1.8694451925666575,
+ "grad_norm": 0.6909728784946527,
+ "learning_rate": 1.0487915690899685e-07,
+ "loss": 0.29963858127593995,
+ "num_input_tokens_seen": 1214319216,
+ "step": 27765,
+ "train_runtime": 214814.2285,
+ "train_tokens_per_second": 5652.881
+ },
+ {
+ "epoch": 1.8697818475626178,
+ "grad_norm": 0.7184706301388402,
+ "learning_rate": 1.0434112512438921e-07,
+ "loss": 0.29176318645477295,
+ "num_input_tokens_seen": 1214533368,
+ "step": 27770,
+ "train_runtime": 214852.4866,
+ "train_tokens_per_second": 5652.871
+ },
+ {
+ "epoch": 1.870118502558578,
+ "grad_norm": 0.8307447666812611,
+ "learning_rate": 1.0380446239517506e-07,
+ "loss": 0.2727836608886719,
+ "num_input_tokens_seen": 1214730928,
+ "step": 27775,
+ "train_runtime": 214893.9185,
+ "train_tokens_per_second": 5652.7
+ },
+ {
+ "epoch": 1.870455157554538,
+ "grad_norm": 0.8916967588803947,
+ "learning_rate": 1.032691688714299e-07,
+ "loss": 0.2895357608795166,
+ "num_input_tokens_seen": 1214918680,
+ "step": 27780,
+ "train_runtime": 214939.4108,
+ "train_tokens_per_second": 5652.377
+ },
+ {
+ "epoch": 1.870791812550498,
+ "grad_norm": 0.7913603727738847,
+ "learning_rate": 1.0273524470284734e-07,
+ "loss": 0.3172266483306885,
+ "num_input_tokens_seen": 1215146136,
+ "step": 27785,
+ "train_runtime": 214976.005,
+ "train_tokens_per_second": 5652.473
+ },
+ {
+ "epoch": 1.8711284675464583,
+ "grad_norm": 0.9142096797553524,
+ "learning_rate": 1.0220269003873739e-07,
+ "loss": 0.29100470542907714,
+ "num_input_tokens_seen": 1215365024,
+ "step": 27790,
+ "train_runtime": 215011.3451,
+ "train_tokens_per_second": 5652.562
+ },
+ {
+ "epoch": 1.8714651225424186,
+ "grad_norm": 0.8041418730532507,
+ "learning_rate": 1.0167150502802703e-07,
+ "loss": 0.2920061111450195,
+ "num_input_tokens_seen": 1215587488,
+ "step": 27795,
+ "train_runtime": 215049.5211,
+ "train_tokens_per_second": 5652.593
+ },
+ {
+ "epoch": 1.8718017775383786,
+ "grad_norm": 0.7464357375845391,
+ "learning_rate": 1.0114168981926075e-07,
+ "loss": 0.31276350021362304,
+ "num_input_tokens_seen": 1215810000,
+ "step": 27800,
+ "train_runtime": 215091.2811,
+ "train_tokens_per_second": 5652.53
+ },
+ {
+ "epoch": 1.8721384325343386,
+ "grad_norm": 0.7898446735642499,
+ "learning_rate": 1.006132445605995e-07,
+ "loss": 0.3080112457275391,
+ "num_input_tokens_seen": 1216014256,
+ "step": 27805,
+ "train_runtime": 215124.1859,
+ "train_tokens_per_second": 5652.615
+ },
+ {
+ "epoch": 1.872475087530299,
+ "grad_norm": 0.7329730105281138,
+ "learning_rate": 1.0008616939982118e-07,
+ "loss": 0.28853521347045896,
+ "num_input_tokens_seen": 1216231488,
+ "step": 27810,
+ "train_runtime": 215162.2144,
+ "train_tokens_per_second": 5652.626
+ },
+ {
+ "epoch": 1.8728117425262591,
+ "grad_norm": 0.7289363854898081,
+ "learning_rate": 9.956046448432177e-08,
+ "loss": 0.283016300201416,
+ "num_input_tokens_seen": 1216451408,
+ "step": 27815,
+ "train_runtime": 215198.9756,
+ "train_tokens_per_second": 5652.682
+ },
+ {
+ "epoch": 1.8731483975222192,
+ "grad_norm": 0.7391182483442761,
+ "learning_rate": 9.90361299611109e-08,
+ "loss": 0.30707416534423826,
+ "num_input_tokens_seen": 1216680872,
+ "step": 27820,
+ "train_runtime": 215231.7273,
+ "train_tokens_per_second": 5652.888
+ },
+ {
+ "epoch": 1.8734850525181792,
+ "grad_norm": 0.8057353688260167,
+ "learning_rate": 9.851316597681959e-08,
+ "loss": 0.3057239055633545,
+ "num_input_tokens_seen": 1216892024,
+ "step": 27825,
+ "train_runtime": 215268.9874,
+ "train_tokens_per_second": 5652.891
+ },
+ {
+ "epoch": 1.8738217075141395,
+ "grad_norm": 0.8107611348319113,
+ "learning_rate": 9.799157267769088e-08,
+ "loss": 0.3025531768798828,
+ "num_input_tokens_seen": 1217114336,
+ "step": 27830,
+ "train_runtime": 215300.5052,
+ "train_tokens_per_second": 5653.096
+ },
+ {
+ "epoch": 1.8741583625100997,
+ "grad_norm": 0.6867341450134844,
+ "learning_rate": 9.747135020958753e-08,
+ "loss": 0.2927044153213501,
+ "num_input_tokens_seen": 1217335976,
+ "step": 27835,
+ "train_runtime": 215340.9059,
+ "train_tokens_per_second": 5653.064
+ },
+ {
+ "epoch": 1.8744950175060597,
+ "grad_norm": 0.7604489307015202,
+ "learning_rate": 9.695249871798818e-08,
+ "loss": 0.2803203582763672,
+ "num_input_tokens_seen": 1217552464,
+ "step": 27840,
+ "train_runtime": 215388.9005,
+ "train_tokens_per_second": 5652.81
+ },
+ {
+ "epoch": 1.8748316725020198,
+ "grad_norm": 0.7584373162773101,
+ "learning_rate": 9.643501834798786e-08,
+ "loss": 0.30305957794189453,
+ "num_input_tokens_seen": 1217765392,
+ "step": 27845,
+ "train_runtime": 215429.0423,
+ "train_tokens_per_second": 5652.745
+ },
+ {
+ "epoch": 1.8751683274979802,
+ "grad_norm": 0.7728951165449691,
+ "learning_rate": 9.591890924429858e-08,
+ "loss": 0.2971980094909668,
+ "num_input_tokens_seen": 1218000984,
+ "step": 27850,
+ "train_runtime": 215470.184,
+ "train_tokens_per_second": 5652.759
+ },
+ {
+ "epoch": 1.8755049824939403,
+ "grad_norm": 0.7960632667781874,
+ "learning_rate": 9.540417155124826e-08,
+ "loss": 0.278977370262146,
+ "num_input_tokens_seen": 1218198848,
+ "step": 27855,
+ "train_runtime": 215509.0395,
+ "train_tokens_per_second": 5652.658
+ },
+ {
+ "epoch": 1.8758416374899003,
+ "grad_norm": 0.73841359058749,
+ "learning_rate": 9.489080541278117e-08,
+ "loss": 0.2986736536026001,
+ "num_input_tokens_seen": 1218407032,
+ "step": 27860,
+ "train_runtime": 215554.3706,
+ "train_tokens_per_second": 5652.435
+ },
+ {
+ "epoch": 1.8761782924858605,
+ "grad_norm": 0.7691921305890641,
+ "learning_rate": 9.437881097246026e-08,
+ "loss": 0.3069639205932617,
+ "num_input_tokens_seen": 1218632248,
+ "step": 27865,
+ "train_runtime": 215593.7195,
+ "train_tokens_per_second": 5652.448
+ },
+ {
+ "epoch": 1.8765149474818208,
+ "grad_norm": 0.7084528384556897,
+ "learning_rate": 9.3868188373461e-08,
+ "loss": 0.2972809076309204,
+ "num_input_tokens_seen": 1218846656,
+ "step": 27870,
+ "train_runtime": 215628.931,
+ "train_tokens_per_second": 5652.519
+ },
+ {
+ "epoch": 1.8768516024777808,
+ "grad_norm": 0.8347410289645919,
+ "learning_rate": 9.335893775857918e-08,
+ "loss": 0.3106103420257568,
+ "num_input_tokens_seen": 1219056688,
+ "step": 27875,
+ "train_runtime": 215663.948,
+ "train_tokens_per_second": 5652.575
+ },
+ {
+ "epoch": 1.8771882574737409,
+ "grad_norm": 0.7646297115981436,
+ "learning_rate": 9.285105927022309e-08,
+ "loss": 0.28227670192718507,
+ "num_input_tokens_seen": 1219278728,
+ "step": 27880,
+ "train_runtime": 215702.0527,
+ "train_tokens_per_second": 5652.606
+ },
+ {
+ "epoch": 1.877524912469701,
+ "grad_norm": 0.8949871913293714,
+ "learning_rate": 9.234455305042133e-08,
+ "loss": 0.2822525978088379,
+ "num_input_tokens_seen": 1219485440,
+ "step": 27885,
+ "train_runtime": 215750.0835,
+ "train_tokens_per_second": 5652.306
+ },
+ {
+ "epoch": 1.8778615674656614,
+ "grad_norm": 0.8356384310040774,
+ "learning_rate": 9.183941924081563e-08,
+ "loss": 0.2904726266860962,
+ "num_input_tokens_seen": 1219702416,
+ "step": 27890,
+ "train_runtime": 215787.36,
+ "train_tokens_per_second": 5652.335
+ },
+ {
+ "epoch": 1.8781982224616214,
+ "grad_norm": 0.7934445345932645,
+ "learning_rate": 9.133565798266574e-08,
+ "loss": 0.2883297443389893,
+ "num_input_tokens_seen": 1219921880,
+ "step": 27895,
+ "train_runtime": 215820.3228,
+ "train_tokens_per_second": 5652.488
+ },
+ {
+ "epoch": 1.8785348774575814,
+ "grad_norm": 0.783421786892747,
+ "learning_rate": 9.083326941684622e-08,
+ "loss": 0.3204617500305176,
+ "num_input_tokens_seen": 1220148280,
+ "step": 27900,
+ "train_runtime": 215855.8328,
+ "train_tokens_per_second": 5652.607
+ },
+ {
+ "epoch": 1.8788715324535417,
+ "grad_norm": 0.7556141302042848,
+ "learning_rate": 9.033225368384912e-08,
+ "loss": 0.3104442358016968,
+ "num_input_tokens_seen": 1220358200,
+ "step": 27905,
+ "train_runtime": 215897.0384,
+ "train_tokens_per_second": 5652.501
+ },
+ {
+ "epoch": 1.879208187449502,
+ "grad_norm": 0.6646972540097822,
+ "learning_rate": 8.98326109237807e-08,
+ "loss": 0.27222867012023927,
+ "num_input_tokens_seen": 1220576488,
+ "step": 27910,
+ "train_runtime": 215939.5159,
+ "train_tokens_per_second": 5652.4
+ },
+ {
+ "epoch": 1.879544842445462,
+ "grad_norm": 0.73208942270682,
+ "learning_rate": 8.933434127636586e-08,
+ "loss": 0.31929314136505127,
+ "num_input_tokens_seen": 1220801344,
+ "step": 27915,
+ "train_runtime": 215983.142,
+ "train_tokens_per_second": 5652.299
+ },
+ {
+ "epoch": 1.879881497441422,
+ "grad_norm": 0.7379086364893702,
+ "learning_rate": 8.883744488094314e-08,
+ "loss": 0.3021604061126709,
+ "num_input_tokens_seen": 1221012952,
+ "step": 27920,
+ "train_runtime": 216020.1906,
+ "train_tokens_per_second": 5652.309
+ },
+ {
+ "epoch": 1.8802181524373822,
+ "grad_norm": 0.7315736019754095,
+ "learning_rate": 8.834192187646917e-08,
+ "loss": 0.30646262168884275,
+ "num_input_tokens_seen": 1221243752,
+ "step": 27925,
+ "train_runtime": 216054.4565,
+ "train_tokens_per_second": 5652.481
+ },
+ {
+ "epoch": 1.8805548074333425,
+ "grad_norm": 0.8178242002248947,
+ "learning_rate": 8.784777240151365e-08,
+ "loss": 0.30849490165710447,
+ "num_input_tokens_seen": 1221469512,
+ "step": 27930,
+ "train_runtime": 216085.3674,
+ "train_tokens_per_second": 5652.717
+ },
+ {
+ "epoch": 1.8808914624293025,
+ "grad_norm": 0.745298782226071,
+ "learning_rate": 8.73549965942655e-08,
+ "loss": 0.3049221754074097,
+ "num_input_tokens_seen": 1221695328,
+ "step": 27935,
+ "train_runtime": 216123.8064,
+ "train_tokens_per_second": 5652.757
+ },
+ {
+ "epoch": 1.8812281174252625,
+ "grad_norm": 0.7296969049019549,
+ "learning_rate": 8.686359459252725e-08,
+ "loss": 0.29515538215637205,
+ "num_input_tokens_seen": 1221906744,
+ "step": 27940,
+ "train_runtime": 216162.3363,
+ "train_tokens_per_second": 5652.727
+ },
+ {
+ "epoch": 1.8815647724212228,
+ "grad_norm": 0.6723805170647249,
+ "learning_rate": 8.637356653371787e-08,
+ "loss": 0.28774356842041016,
+ "num_input_tokens_seen": 1222134512,
+ "step": 27945,
+ "train_runtime": 216197.844,
+ "train_tokens_per_second": 5652.852
+ },
+ {
+ "epoch": 1.881901427417183,
+ "grad_norm": 0.8473708457816271,
+ "learning_rate": 8.588491255487163e-08,
+ "loss": 0.2720442056655884,
+ "num_input_tokens_seen": 1222344728,
+ "step": 27950,
+ "train_runtime": 216239.4622,
+ "train_tokens_per_second": 5652.737
+ },
+ {
+ "epoch": 1.882238082413143,
+ "grad_norm": 0.7624946480602264,
+ "learning_rate": 8.539763279264035e-08,
+ "loss": 0.2836963176727295,
+ "num_input_tokens_seen": 1222560856,
+ "step": 27955,
+ "train_runtime": 216284.8082,
+ "train_tokens_per_second": 5652.551
+ },
+ {
+ "epoch": 1.882574737409103,
+ "grad_norm": 0.7053326289175388,
+ "learning_rate": 8.49117273832889e-08,
+ "loss": 0.3163050651550293,
+ "num_input_tokens_seen": 1222789864,
+ "step": 27960,
+ "train_runtime": 216318.1103,
+ "train_tokens_per_second": 5652.739
+ },
+ {
+ "epoch": 1.8829113924050633,
+ "grad_norm": 0.6499749519322318,
+ "learning_rate": 8.442719646270082e-08,
+ "loss": 0.28265206813812255,
+ "num_input_tokens_seen": 1223005096,
+ "step": 27965,
+ "train_runtime": 216357.5685,
+ "train_tokens_per_second": 5652.703
+ },
+ {
+ "epoch": 1.8832480474010236,
+ "grad_norm": 0.7141334939696597,
+ "learning_rate": 8.394404016637159e-08,
+ "loss": 0.28957691192626955,
+ "num_input_tokens_seen": 1223223864,
+ "step": 27970,
+ "train_runtime": 216399.1348,
+ "train_tokens_per_second": 5652.628
+ },
+ {
+ "epoch": 1.8835847023969836,
+ "grad_norm": 0.7972265729262933,
+ "learning_rate": 8.346225862941592e-08,
+ "loss": 0.28935163021087645,
+ "num_input_tokens_seen": 1223430136,
+ "step": 27975,
+ "train_runtime": 216434.0905,
+ "train_tokens_per_second": 5652.668
+ },
+ {
+ "epoch": 1.8839213573929436,
+ "grad_norm": 0.8292085959022567,
+ "learning_rate": 8.298185198656161e-08,
+ "loss": 0.31040134429931643,
+ "num_input_tokens_seen": 1223661888,
+ "step": 27980,
+ "train_runtime": 216475.8962,
+ "train_tokens_per_second": 5652.647
+ },
+ {
+ "epoch": 1.884258012388904,
+ "grad_norm": 0.7497952405897995,
+ "learning_rate": 8.25028203721534e-08,
+ "loss": 0.2839097023010254,
+ "num_input_tokens_seen": 1223872392,
+ "step": 27985,
+ "train_runtime": 216516.2523,
+ "train_tokens_per_second": 5652.566
+ },
+ {
+ "epoch": 1.8845946673848641,
+ "grad_norm": 0.8204275222628281,
+ "learning_rate": 8.202516392015081e-08,
+ "loss": 0.30760276317596436,
+ "num_input_tokens_seen": 1224095496,
+ "step": 27990,
+ "train_runtime": 216553.2529,
+ "train_tokens_per_second": 5652.63
+ },
+ {
+ "epoch": 1.8849313223808242,
+ "grad_norm": 0.9261868943922493,
+ "learning_rate": 8.154888276412865e-08,
+ "loss": 0.31410627365112304,
+ "num_input_tokens_seen": 1224301616,
+ "step": 27995,
+ "train_runtime": 216587.9714,
+ "train_tokens_per_second": 5652.676
+ },
+ {
+ "epoch": 1.8852679773767842,
+ "grad_norm": 0.7105028791923415,
+ "learning_rate": 8.10739770372776e-08,
+ "loss": 0.2736332416534424,
+ "num_input_tokens_seen": 1224525064,
+ "step": 28000,
+ "train_runtime": 216624.6668,
+ "train_tokens_per_second": 5652.75
+ },
+ {
+ "epoch": 1.8856046323727444,
+ "grad_norm": 0.6544014559458133,
+ "learning_rate": 8.060044687240365e-08,
+ "loss": 0.2702982425689697,
+ "num_input_tokens_seen": 1224764712,
+ "step": 28005,
+ "train_runtime": 216655.2066,
+ "train_tokens_per_second": 5653.059
+ },
+ {
+ "epoch": 1.8859412873687047,
+ "grad_norm": 0.7293503705566087,
+ "learning_rate": 8.012829240192754e-08,
+ "loss": 0.31306607723236085,
+ "num_input_tokens_seen": 1224979848,
+ "step": 28010,
+ "train_runtime": 216695.9204,
+ "train_tokens_per_second": 5652.99
+ },
+ {
+ "epoch": 1.8862779423646647,
+ "grad_norm": 0.7181373637032492,
+ "learning_rate": 7.9657513757887e-08,
+ "loss": 0.2802762985229492,
+ "num_input_tokens_seen": 1225200744,
+ "step": 28015,
+ "train_runtime": 216735.8221,
+ "train_tokens_per_second": 5652.968
+ },
+ {
+ "epoch": 1.8866145973606248,
+ "grad_norm": 0.753661015701672,
+ "learning_rate": 7.918811107193225e-08,
+ "loss": 0.28357934951782227,
+ "num_input_tokens_seen": 1225430328,
+ "step": 28020,
+ "train_runtime": 216771.4604,
+ "train_tokens_per_second": 5653.098
+ },
+ {
+ "epoch": 1.886951252356585,
+ "grad_norm": 0.8998847487777263,
+ "learning_rate": 7.872008447533109e-08,
+ "loss": 0.3160097122192383,
+ "num_input_tokens_seen": 1225648792,
+ "step": 28025,
+ "train_runtime": 216808.3453,
+ "train_tokens_per_second": 5653.144
+ },
+ {
+ "epoch": 1.8872879073525453,
+ "grad_norm": 0.7934211783590108,
+ "learning_rate": 7.825343409896547e-08,
+ "loss": 0.3094397306442261,
+ "num_input_tokens_seen": 1225864272,
+ "step": 28030,
+ "train_runtime": 216844.6438,
+ "train_tokens_per_second": 5653.191
+ },
+ {
+ "epoch": 1.8876245623485053,
+ "grad_norm": 0.7748967589073342,
+ "learning_rate": 7.778816007333268e-08,
+ "loss": 0.3081014156341553,
+ "num_input_tokens_seen": 1226087184,
+ "step": 28035,
+ "train_runtime": 216887.2541,
+ "train_tokens_per_second": 5653.109
+ },
+ {
+ "epoch": 1.8879612173444653,
+ "grad_norm": 0.8144894017280363,
+ "learning_rate": 7.73242625285453e-08,
+ "loss": 0.29062952995300295,
+ "num_input_tokens_seen": 1226306464,
+ "step": 28040,
+ "train_runtime": 216925.8357,
+ "train_tokens_per_second": 5653.114
+ },
+ {
+ "epoch": 1.8882978723404256,
+ "grad_norm": 0.7574566262911345,
+ "learning_rate": 7.686174159433069e-08,
+ "loss": 0.3064660787582397,
+ "num_input_tokens_seen": 1226520296,
+ "step": 28045,
+ "train_runtime": 216966.9693,
+ "train_tokens_per_second": 5653.028
+ },
+ {
+ "epoch": 1.8886345273363858,
+ "grad_norm": 0.627537289336894,
+ "learning_rate": 7.640059740003091e-08,
+ "loss": 0.2722774982452393,
+ "num_input_tokens_seen": 1226746648,
+ "step": 28050,
+ "train_runtime": 217006.2516,
+ "train_tokens_per_second": 5653.047
+ },
+ {
+ "epoch": 1.8889711823323458,
+ "grad_norm": 0.7705422901820207,
+ "learning_rate": 7.594083007460396e-08,
+ "loss": 0.29419803619384766,
+ "num_input_tokens_seen": 1226953312,
+ "step": 28055,
+ "train_runtime": 217048.1468,
+ "train_tokens_per_second": 5652.908
+ },
+ {
+ "epoch": 1.8893078373283059,
+ "grad_norm": 0.7044024327241764,
+ "learning_rate": 7.54824397466225e-08,
+ "loss": 0.2896998167037964,
+ "num_input_tokens_seen": 1227175648,
+ "step": 28060,
+ "train_runtime": 217084.5331,
+ "train_tokens_per_second": 5652.985
+ },
+ {
+ "epoch": 1.8896444923242661,
+ "grad_norm": 0.715442564314742,
+ "learning_rate": 7.502542654427347e-08,
+ "loss": 0.28265087604522704,
+ "num_input_tokens_seen": 1227383304,
+ "step": 28065,
+ "train_runtime": 217116.2593,
+ "train_tokens_per_second": 5653.116
+ },
+ {
+ "epoch": 1.8899811473202264,
+ "grad_norm": 0.7174922222370509,
+ "learning_rate": 7.456979059535907e-08,
+ "loss": 0.29991984367370605,
+ "num_input_tokens_seen": 1227603976,
+ "step": 28070,
+ "train_runtime": 217161.5657,
+ "train_tokens_per_second": 5652.952
+ },
+ {
+ "epoch": 1.8903178023161864,
+ "grad_norm": 0.8821755978176467,
+ "learning_rate": 7.411553202729682e-08,
+ "loss": 0.2972327470779419,
+ "num_input_tokens_seen": 1227795992,
+ "step": 28075,
+ "train_runtime": 217202.7149,
+ "train_tokens_per_second": 5652.765
+ },
+ {
+ "epoch": 1.8906544573121464,
+ "grad_norm": 0.7547240032942905,
+ "learning_rate": 7.3662650967119e-08,
+ "loss": 0.2765605926513672,
+ "num_input_tokens_seen": 1228009920,
+ "step": 28080,
+ "train_runtime": 217242.1532,
+ "train_tokens_per_second": 5652.724
+ },
+ {
+ "epoch": 1.8909911123081067,
+ "grad_norm": 0.7381759970020896,
+ "learning_rate": 7.321114754147151e-08,
+ "loss": 0.28570716381072997,
+ "num_input_tokens_seen": 1228232816,
+ "step": 28085,
+ "train_runtime": 217280.844,
+ "train_tokens_per_second": 5652.743
+ },
+ {
+ "epoch": 1.891327767304067,
+ "grad_norm": 0.9586290374485588,
+ "learning_rate": 7.276102187661615e-08,
+ "loss": 0.31825072765350343,
+ "num_input_tokens_seen": 1228448704,
+ "step": 28090,
+ "train_runtime": 217319.8127,
+ "train_tokens_per_second": 5652.723
+ },
+ {
+ "epoch": 1.891664422300027,
+ "grad_norm": 0.7106822110807806,
+ "learning_rate": 7.231227409842944e-08,
+ "loss": 0.30093338489532473,
+ "num_input_tokens_seen": 1228675704,
+ "step": 28095,
+ "train_runtime": 217355.8762,
+ "train_tokens_per_second": 5652.829
+ },
+ {
+ "epoch": 1.892001077295987,
+ "grad_norm": 0.7479457461633977,
+ "learning_rate": 7.186490433240156e-08,
+ "loss": 0.30986011028289795,
+ "num_input_tokens_seen": 1228877808,
+ "step": 28100,
+ "train_runtime": 217390.778,
+ "train_tokens_per_second": 5652.852
+ },
+ {
+ "epoch": 1.8923377322919472,
+ "grad_norm": 0.7862198031676131,
+ "learning_rate": 7.14189127036391e-08,
+ "loss": 0.2832916259765625,
+ "num_input_tokens_seen": 1229095280,
+ "step": 28105,
+ "train_runtime": 217425.4702,
+ "train_tokens_per_second": 5652.95
+ },
+ {
+ "epoch": 1.8926743872879075,
+ "grad_norm": 0.7610693134796771,
+ "learning_rate": 7.097429933686118e-08,
+ "loss": 0.3074744701385498,
+ "num_input_tokens_seen": 1229309216,
+ "step": 28110,
+ "train_runtime": 217466.3238,
+ "train_tokens_per_second": 5652.872
+ },
+ {
+ "epoch": 1.8930110422838675,
+ "grad_norm": 0.9062932954935005,
+ "learning_rate": 7.053106435640333e-08,
+ "loss": 0.31830539703369143,
+ "num_input_tokens_seen": 1229541912,
+ "step": 28115,
+ "train_runtime": 217504.1355,
+ "train_tokens_per_second": 5652.959
+ },
+ {
+ "epoch": 1.8933476972798275,
+ "grad_norm": 0.7426318413849856,
+ "learning_rate": 7.008920788621366e-08,
+ "loss": 0.299678635597229,
+ "num_input_tokens_seen": 1229758840,
+ "step": 28120,
+ "train_runtime": 217545.4661,
+ "train_tokens_per_second": 5652.882
+ },
+ {
+ "epoch": 1.8936843522757878,
+ "grad_norm": 0.7413896799320939,
+ "learning_rate": 6.964873004985717e-08,
+ "loss": 0.3041519641876221,
+ "num_input_tokens_seen": 1229975280,
+ "step": 28125,
+ "train_runtime": 217582.095,
+ "train_tokens_per_second": 5652.925
+ },
+ {
+ "epoch": 1.894021007271748,
+ "grad_norm": 0.8140877400473608,
+ "learning_rate": 6.920963097051148e-08,
+ "loss": 0.3081005096435547,
+ "num_input_tokens_seen": 1230195248,
+ "step": 28130,
+ "train_runtime": 217622.0274,
+ "train_tokens_per_second": 5652.899
+ },
+ {
+ "epoch": 1.894357662267708,
+ "grad_norm": 0.8171774682734737,
+ "learning_rate": 6.877191077096945e-08,
+ "loss": 0.28580417633056643,
+ "num_input_tokens_seen": 1230403720,
+ "step": 28135,
+ "train_runtime": 217663.6137,
+ "train_tokens_per_second": 5652.776
+ },
+ {
+ "epoch": 1.894694317263668,
+ "grad_norm": 0.7796944173521188,
+ "learning_rate": 6.833556957363818e-08,
+ "loss": 0.31174564361572266,
+ "num_input_tokens_seen": 1230619400,
+ "step": 28140,
+ "train_runtime": 217701.7663,
+ "train_tokens_per_second": 5652.776
+ },
+ {
+ "epoch": 1.8950309722596284,
+ "grad_norm": 0.7394530047484759,
+ "learning_rate": 6.790060750053895e-08,
+ "loss": 0.2922555446624756,
+ "num_input_tokens_seen": 1230833368,
+ "step": 28145,
+ "train_runtime": 217738.7294,
+ "train_tokens_per_second": 5652.799
+ },
+ {
+ "epoch": 1.8953676272555886,
+ "grad_norm": 0.6213839509137753,
+ "learning_rate": 6.746702467330723e-08,
+ "loss": 0.2662560701370239,
+ "num_input_tokens_seen": 1231064032,
+ "step": 28150,
+ "train_runtime": 217778.1192,
+ "train_tokens_per_second": 5652.836
+ },
+ {
+ "epoch": 1.8957042822515486,
+ "grad_norm": 0.7209769414897946,
+ "learning_rate": 6.703482121319383e-08,
+ "loss": 0.2952676296234131,
+ "num_input_tokens_seen": 1231282944,
+ "step": 28155,
+ "train_runtime": 217815.4664,
+ "train_tokens_per_second": 5652.872
+ },
+ {
+ "epoch": 1.8960409372475087,
+ "grad_norm": 0.7719176273534697,
+ "learning_rate": 6.660399724106204e-08,
+ "loss": 0.31438963413238524,
+ "num_input_tokens_seen": 1231513400,
+ "step": 28160,
+ "train_runtime": 217855.297,
+ "train_tokens_per_second": 5652.896
+ },
+ {
+ "epoch": 1.896377592243469,
+ "grad_norm": 0.7554115178407177,
+ "learning_rate": 6.617455287739161e-08,
+ "loss": 0.28926453590393064,
+ "num_input_tokens_seen": 1231738296,
+ "step": 28165,
+ "train_runtime": 217889.6272,
+ "train_tokens_per_second": 5653.038
+ },
+ {
+ "epoch": 1.8967142472394292,
+ "grad_norm": 0.7869091031125452,
+ "learning_rate": 6.574648824227369e-08,
+ "loss": 0.29925999641418455,
+ "num_input_tokens_seen": 1231961800,
+ "step": 28170,
+ "train_runtime": 217924.705,
+ "train_tokens_per_second": 5653.153
+ },
+ {
+ "epoch": 1.8970509022353892,
+ "grad_norm": 0.7908040273802075,
+ "learning_rate": 6.5319803455417e-08,
+ "loss": 0.3044144153594971,
+ "num_input_tokens_seen": 1232184856,
+ "step": 28175,
+ "train_runtime": 217965.4264,
+ "train_tokens_per_second": 5653.121
+ },
+ {
+ "epoch": 1.8973875572313492,
+ "grad_norm": 0.7737887474560637,
+ "learning_rate": 6.489449863614106e-08,
+ "loss": 0.3226012706756592,
+ "num_input_tokens_seen": 1232404696,
+ "step": 28180,
+ "train_runtime": 218003.5275,
+ "train_tokens_per_second": 5653.141
+ },
+ {
+ "epoch": 1.8977242122273095,
+ "grad_norm": 0.7245308841502345,
+ "learning_rate": 6.44705739033824e-08,
+ "loss": 0.29819202423095703,
+ "num_input_tokens_seen": 1232620728,
+ "step": 28185,
+ "train_runtime": 218043.0047,
+ "train_tokens_per_second": 5653.108
+ },
+ {
+ "epoch": 1.8980608672232697,
+ "grad_norm": 0.883569648340652,
+ "learning_rate": 6.404802937568899e-08,
+ "loss": 0.283476448059082,
+ "num_input_tokens_seen": 1232822576,
+ "step": 28190,
+ "train_runtime": 218076.1787,
+ "train_tokens_per_second": 5653.174
+ },
+ {
+ "epoch": 1.8983975222192297,
+ "grad_norm": 0.6097934961985845,
+ "learning_rate": 6.362686517122463e-08,
+ "loss": 0.276761531829834,
+ "num_input_tokens_seen": 1233038648,
+ "step": 28195,
+ "train_runtime": 218115.3719,
+ "train_tokens_per_second": 5653.149
+ },
+ {
+ "epoch": 1.8987341772151898,
+ "grad_norm": 0.7436026418824351,
+ "learning_rate": 6.320708140776676e-08,
+ "loss": 0.28359341621398926,
+ "num_input_tokens_seen": 1233258352,
+ "step": 28200,
+ "train_runtime": 218158.6419,
+ "train_tokens_per_second": 5653.035
+ },
+ {
+ "epoch": 1.89907083221115,
+ "grad_norm": 0.7681913781823284,
+ "learning_rate": 6.278867820270596e-08,
+ "loss": 0.299153208732605,
+ "num_input_tokens_seen": 1233480288,
+ "step": 28205,
+ "train_runtime": 218195.6989,
+ "train_tokens_per_second": 5653.092
+ },
+ {
+ "epoch": 1.8994074872071103,
+ "grad_norm": 0.7830138695146227,
+ "learning_rate": 6.237165567304749e-08,
+ "loss": 0.29818384647369384,
+ "num_input_tokens_seen": 1233691368,
+ "step": 28210,
+ "train_runtime": 218237.5859,
+ "train_tokens_per_second": 5652.974
+ },
+ {
+ "epoch": 1.8997441422030703,
+ "grad_norm": 0.7003448488149383,
+ "learning_rate": 6.19560139354114e-08,
+ "loss": 0.30460848808288576,
+ "num_input_tokens_seen": 1233930016,
+ "step": 28215,
+ "train_runtime": 218270.3105,
+ "train_tokens_per_second": 5653.22
+ },
+ {
+ "epoch": 1.9000807971990303,
+ "grad_norm": 0.7370403299032923,
+ "learning_rate": 6.154175310602916e-08,
+ "loss": 0.30403151512146,
+ "num_input_tokens_seen": 1234154432,
+ "step": 28220,
+ "train_runtime": 218302.9228,
+ "train_tokens_per_second": 5653.403
+ },
+ {
+ "epoch": 1.9004174521949906,
+ "grad_norm": 0.7562586977964543,
+ "learning_rate": 6.112887330074812e-08,
+ "loss": 0.2886775255203247,
+ "num_input_tokens_seen": 1234363888,
+ "step": 28225,
+ "train_runtime": 218340.6417,
+ "train_tokens_per_second": 5653.386
+ },
+ {
+ "epoch": 1.9007541071909508,
+ "grad_norm": 0.8023435948804223,
+ "learning_rate": 6.07173746350287e-08,
+ "loss": 0.3039447069168091,
+ "num_input_tokens_seen": 1234586704,
+ "step": 28230,
+ "train_runtime": 218377.1498,
+ "train_tokens_per_second": 5653.461
+ },
+ {
+ "epoch": 1.9010907621869109,
+ "grad_norm": 0.7102448231458076,
+ "learning_rate": 6.030725722394548e-08,
+ "loss": 0.30956282615661623,
+ "num_input_tokens_seen": 1234811984,
+ "step": 28235,
+ "train_runtime": 218419.5266,
+ "train_tokens_per_second": 5653.396
+ },
+ {
+ "epoch": 1.901427417182871,
+ "grad_norm": 0.8917953406376619,
+ "learning_rate": 5.98985211821862e-08,
+ "loss": 0.31682801246643066,
+ "num_input_tokens_seen": 1235026864,
+ "step": 28240,
+ "train_runtime": 218463.2255,
+ "train_tokens_per_second": 5653.248
+ },
+ {
+ "epoch": 1.9017640721788311,
+ "grad_norm": 3.1213239281278913,
+ "learning_rate": 5.9491166624052185e-08,
+ "loss": 0.29059295654296874,
+ "num_input_tokens_seen": 1235257424,
+ "step": 28245,
+ "train_runtime": 218502.1071,
+ "train_tokens_per_second": 5653.298
+ },
+ {
+ "epoch": 1.9021007271747914,
+ "grad_norm": 0.800370397152095,
+ "learning_rate": 5.908519366345955e-08,
+ "loss": 0.2929649829864502,
+ "num_input_tokens_seen": 1235465968,
+ "step": 28250,
+ "train_runtime": 218531.1775,
+ "train_tokens_per_second": 5653.5
+ },
+ {
+ "epoch": 1.9024373821707514,
+ "grad_norm": 0.7899542839676547,
+ "learning_rate": 5.868060241393747e-08,
+ "loss": 0.32121825218200684,
+ "num_input_tokens_seen": 1235701752,
+ "step": 28255,
+ "train_runtime": 218574.1157,
+ "train_tokens_per_second": 5653.468
+ },
+ {
+ "epoch": 1.9027740371667115,
+ "grad_norm": 0.8945208045237926,
+ "learning_rate": 5.8277392988627665e-08,
+ "loss": 0.30273237228393557,
+ "num_input_tokens_seen": 1235918048,
+ "step": 28260,
+ "train_runtime": 218613.0158,
+ "train_tokens_per_second": 5653.451
+ },
+ {
+ "epoch": 1.9031106921626717,
+ "grad_norm": 1.0956824749607716,
+ "learning_rate": 5.787556550028772e-08,
+ "loss": 0.2830909967422485,
+ "num_input_tokens_seen": 1236133248,
+ "step": 28265,
+ "train_runtime": 218655.8574,
+ "train_tokens_per_second": 5653.328
+ },
+ {
+ "epoch": 1.903447347158632,
+ "grad_norm": 0.7232891426044542,
+ "learning_rate": 5.747512006128553e-08,
+ "loss": 0.3094156265258789,
+ "num_input_tokens_seen": 1236365056,
+ "step": 28270,
+ "train_runtime": 218690.6793,
+ "train_tokens_per_second": 5653.488
+ },
+ {
+ "epoch": 1.903784002154592,
+ "grad_norm": 0.825207510049438,
+ "learning_rate": 5.707605678360595e-08,
+ "loss": 0.30304322242736814,
+ "num_input_tokens_seen": 1236586568,
+ "step": 28275,
+ "train_runtime": 218729.618,
+ "train_tokens_per_second": 5653.494
+ },
+ {
+ "epoch": 1.904120657150552,
+ "grad_norm": 0.6746156299713596,
+ "learning_rate": 5.6678375778845276e-08,
+ "loss": 0.2991356611251831,
+ "num_input_tokens_seen": 1236816112,
+ "step": 28280,
+ "train_runtime": 218761.9014,
+ "train_tokens_per_second": 5653.709
+ },
+ {
+ "epoch": 1.9044573121465123,
+ "grad_norm": 0.7102199653388521,
+ "learning_rate": 5.6282077158213435e-08,
+ "loss": 0.29000029563903806,
+ "num_input_tokens_seen": 1237040088,
+ "step": 28285,
+ "train_runtime": 218799.9174,
+ "train_tokens_per_second": 5653.75
+ },
+ {
+ "epoch": 1.9047939671424725,
+ "grad_norm": 0.8451189106225591,
+ "learning_rate": 5.588716103253511e-08,
+ "loss": 0.29954440593719484,
+ "num_input_tokens_seen": 1237269408,
+ "step": 28290,
+ "train_runtime": 218845.6994,
+ "train_tokens_per_second": 5653.615
+ },
+ {
+ "epoch": 1.9051306221384325,
+ "grad_norm": 0.8190739439948348,
+ "learning_rate": 5.549362751224585e-08,
+ "loss": 0.31754138469696047,
+ "num_input_tokens_seen": 1237476144,
+ "step": 28295,
+ "train_runtime": 218883.6501,
+ "train_tokens_per_second": 5653.58
+ },
+ {
+ "epoch": 1.9054672771343926,
+ "grad_norm": 0.7766937595152473,
+ "learning_rate": 5.5101476707397075e-08,
+ "loss": 0.2879055976867676,
+ "num_input_tokens_seen": 1237699056,
+ "step": 28300,
+ "train_runtime": 218922.8288,
+ "train_tokens_per_second": 5653.586
+ },
+ {
+ "epoch": 1.9058039321303528,
+ "grad_norm": 0.6581310767139287,
+ "learning_rate": 5.471070872765216e-08,
+ "loss": 0.31143577098846437,
+ "num_input_tokens_seen": 1237922344,
+ "step": 28305,
+ "train_runtime": 218962.3366,
+ "train_tokens_per_second": 5653.586
+ },
+ {
+ "epoch": 1.906140587126313,
+ "grad_norm": 0.8595902577669746,
+ "learning_rate": 5.432132368228815e-08,
+ "loss": 0.3075572490692139,
+ "num_input_tokens_seen": 1238144360,
+ "step": 28310,
+ "train_runtime": 218996.6769,
+ "train_tokens_per_second": 5653.713
+ },
+ {
+ "epoch": 1.906477242122273,
+ "grad_norm": 0.7260821752374543,
+ "learning_rate": 5.393332168019572e-08,
+ "loss": 0.29381508827209474,
+ "num_input_tokens_seen": 1238376024,
+ "step": 28315,
+ "train_runtime": 219038.2854,
+ "train_tokens_per_second": 5653.697
+ },
+ {
+ "epoch": 1.9068138971182331,
+ "grad_norm": 0.7695214541702009,
+ "learning_rate": 5.354670282987695e-08,
+ "loss": 0.31864123344421386,
+ "num_input_tokens_seen": 1238615352,
+ "step": 28320,
+ "train_runtime": 219071.5756,
+ "train_tokens_per_second": 5653.93
+ },
+ {
+ "epoch": 1.9071505521141934,
+ "grad_norm": 0.7626491001109459,
+ "learning_rate": 5.316146723945037e-08,
+ "loss": 0.3147014379501343,
+ "num_input_tokens_seen": 1238839600,
+ "step": 28325,
+ "train_runtime": 219109.8816,
+ "train_tokens_per_second": 5653.965
+ },
+ {
+ "epoch": 1.9074872071101536,
+ "grad_norm": 0.7542385697702055,
+ "learning_rate": 5.277761501664535e-08,
+ "loss": 0.28926112651824953,
+ "num_input_tokens_seen": 1239055736,
+ "step": 28330,
+ "train_runtime": 219151.3861,
+ "train_tokens_per_second": 5653.88
+ },
+ {
+ "epoch": 1.9078238621061137,
+ "grad_norm": 0.8539701413858525,
+ "learning_rate": 5.2395146268804934e-08,
+ "loss": 0.3053768157958984,
+ "num_input_tokens_seen": 1239282360,
+ "step": 28335,
+ "train_runtime": 219183.9605,
+ "train_tokens_per_second": 5654.074
+ },
+ {
+ "epoch": 1.9081605171020737,
+ "grad_norm": 0.6987538717491121,
+ "learning_rate": 5.201406110288465e-08,
+ "loss": 0.2954925537109375,
+ "num_input_tokens_seen": 1239505344,
+ "step": 28340,
+ "train_runtime": 219222.4201,
+ "train_tokens_per_second": 5654.099
+ },
+ {
+ "epoch": 1.908497172098034,
+ "grad_norm": 0.7258797958708753,
+ "learning_rate": 5.1634359625454266e-08,
+ "loss": 0.26157379150390625,
+ "num_input_tokens_seen": 1239718552,
+ "step": 28345,
+ "train_runtime": 219261.4852,
+ "train_tokens_per_second": 5654.064
+ },
+ {
+ "epoch": 1.9088338270939942,
+ "grad_norm": 0.833392494175178,
+ "learning_rate": 5.125604194269606e-08,
+ "loss": 0.32343695163726804,
+ "num_input_tokens_seen": 1239933464,
+ "step": 28350,
+ "train_runtime": 219303.7788,
+ "train_tokens_per_second": 5653.954
+ },
+ {
+ "epoch": 1.9091704820899542,
+ "grad_norm": 0.8588011467833099,
+ "learning_rate": 5.087910816040542e-08,
+ "loss": 0.32165529727935793,
+ "num_input_tokens_seen": 1240153424,
+ "step": 28355,
+ "train_runtime": 219348.9617,
+ "train_tokens_per_second": 5653.792
+ },
+ {
+ "epoch": 1.9095071370859142,
+ "grad_norm": 0.7170515438107391,
+ "learning_rate": 5.0503558383990235e-08,
+ "loss": 0.29596068859100344,
+ "num_input_tokens_seen": 1240382984,
+ "step": 28360,
+ "train_runtime": 219391.3692,
+ "train_tokens_per_second": 5653.746
+ },
+ {
+ "epoch": 1.9098437920818745,
+ "grad_norm": 0.9056425664871629,
+ "learning_rate": 5.0129392718472614e-08,
+ "loss": 0.29495375156402587,
+ "num_input_tokens_seen": 1240602456,
+ "step": 28365,
+ "train_runtime": 219428.0549,
+ "train_tokens_per_second": 5653.801
+ },
+ {
+ "epoch": 1.9101804470778347,
+ "grad_norm": 0.7683712573518254,
+ "learning_rate": 4.9756611268486077e-08,
+ "loss": 0.3231898307800293,
+ "num_input_tokens_seen": 1240834056,
+ "step": 28370,
+ "train_runtime": 219470.1885,
+ "train_tokens_per_second": 5653.77
+ },
+ {
+ "epoch": 1.9105171020737948,
+ "grad_norm": 0.8473839083939165,
+ "learning_rate": 4.9385214138277795e-08,
+ "loss": 0.3085433006286621,
+ "num_input_tokens_seen": 1241053560,
+ "step": 28375,
+ "train_runtime": 219515.4349,
+ "train_tokens_per_second": 5653.605
+ },
+ {
+ "epoch": 1.9108537570697548,
+ "grad_norm": 0.7530840258849333,
+ "learning_rate": 4.901520143170857e-08,
+ "loss": 0.31458406448364257,
+ "num_input_tokens_seen": 1241278752,
+ "step": 28380,
+ "train_runtime": 219551.5142,
+ "train_tokens_per_second": 5653.702
+ },
+ {
+ "epoch": 1.911190412065715,
+ "grad_norm": 0.8440231962915516,
+ "learning_rate": 4.864657325225064e-08,
+ "loss": 0.2959612846374512,
+ "num_input_tokens_seen": 1241501272,
+ "step": 28385,
+ "train_runtime": 219592.0284,
+ "train_tokens_per_second": 5653.672
+ },
+ {
+ "epoch": 1.9115270670616753,
+ "grad_norm": 0.7788969789388622,
+ "learning_rate": 4.827932970298932e-08,
+ "loss": 0.28577983379364014,
+ "num_input_tokens_seen": 1241706992,
+ "step": 28390,
+ "train_runtime": 219631.7102,
+ "train_tokens_per_second": 5653.587
+ },
+ {
+ "epoch": 1.9118637220576353,
+ "grad_norm": 0.7570563184078223,
+ "learning_rate": 4.7913470886624125e-08,
+ "loss": 0.30133938789367676,
+ "num_input_tokens_seen": 1241929008,
+ "step": 28395,
+ "train_runtime": 219667.8047,
+ "train_tokens_per_second": 5653.669
+ },
+ {
+ "epoch": 1.9122003770535954,
+ "grad_norm": 0.7674867971456765,
+ "learning_rate": 4.754899690546544e-08,
+ "loss": 0.2801696538925171,
+ "num_input_tokens_seen": 1242147848,
+ "step": 28400,
+ "train_runtime": 219701.4788,
+ "train_tokens_per_second": 5653.798
+ },
+ {
+ "epoch": 1.9125370320495556,
+ "grad_norm": 0.7214457077784336,
+ "learning_rate": 4.718590786143729e-08,
+ "loss": 0.3026761531829834,
+ "num_input_tokens_seen": 1242368696,
+ "step": 28405,
+ "train_runtime": 219734.8895,
+ "train_tokens_per_second": 5653.944
+ },
+ {
+ "epoch": 1.9128736870455159,
+ "grad_norm": 0.7424922888900161,
+ "learning_rate": 4.6824203856076776e-08,
+ "loss": 0.3060077905654907,
+ "num_input_tokens_seen": 1242605672,
+ "step": 28410,
+ "train_runtime": 219770.358,
+ "train_tokens_per_second": 5654.11
+ },
+ {
+ "epoch": 1.9132103420414759,
+ "grad_norm": 0.7488946195290975,
+ "learning_rate": 4.646388499053356e-08,
+ "loss": 0.2775508642196655,
+ "num_input_tokens_seen": 1242816568,
+ "step": 28415,
+ "train_runtime": 219806.033,
+ "train_tokens_per_second": 5654.151
+ },
+ {
+ "epoch": 1.913546997037436,
+ "grad_norm": 0.7402940285995784,
+ "learning_rate": 4.6104951365568696e-08,
+ "loss": 0.2732607126235962,
+ "num_input_tokens_seen": 1243037664,
+ "step": 28420,
+ "train_runtime": 219842.7305,
+ "train_tokens_per_second": 5654.213
+ },
+ {
+ "epoch": 1.9138836520333962,
+ "grad_norm": 0.8182194298111709,
+ "learning_rate": 4.574740308155801e-08,
+ "loss": 0.2992802858352661,
+ "num_input_tokens_seen": 1243248672,
+ "step": 28425,
+ "train_runtime": 219884.9,
+ "train_tokens_per_second": 5654.088
+ },
+ {
+ "epoch": 1.9142203070293564,
+ "grad_norm": 0.7646860170343411,
+ "learning_rate": 4.539124023848762e-08,
+ "loss": 0.2957939624786377,
+ "num_input_tokens_seen": 1243481672,
+ "step": 28430,
+ "train_runtime": 219924.6589,
+ "train_tokens_per_second": 5654.126
+ },
+ {
+ "epoch": 1.9145569620253164,
+ "grad_norm": 0.7725331026025192,
+ "learning_rate": 4.503646293595787e-08,
+ "loss": 0.32006833553314207,
+ "num_input_tokens_seen": 1243698024,
+ "step": 28435,
+ "train_runtime": 219958.6796,
+ "train_tokens_per_second": 5654.235
+ },
+ {
+ "epoch": 1.9148936170212765,
+ "grad_norm": 0.7967463881139475,
+ "learning_rate": 4.468307127318105e-08,
+ "loss": 0.2883774280548096,
+ "num_input_tokens_seen": 1243923240,
+ "step": 28440,
+ "train_runtime": 219994.0164,
+ "train_tokens_per_second": 5654.35
+ },
+ {
+ "epoch": 1.9152302720172367,
+ "grad_norm": 0.8721239693546312,
+ "learning_rate": 4.4331065348982014e-08,
+ "loss": 0.2830281496047974,
+ "num_input_tokens_seen": 1244140864,
+ "step": 28445,
+ "train_runtime": 220025.1373,
+ "train_tokens_per_second": 5654.54
+ },
+ {
+ "epoch": 1.915566927013197,
+ "grad_norm": 0.7395255680690924,
+ "learning_rate": 4.3980445261798125e-08,
+ "loss": 0.3066460371017456,
+ "num_input_tokens_seen": 1244363064,
+ "step": 28450,
+ "train_runtime": 220071.6297,
+ "train_tokens_per_second": 5654.355
+ },
+ {
+ "epoch": 1.915903582009157,
+ "grad_norm": 0.8552277238547642,
+ "learning_rate": 4.363121110967927e-08,
+ "loss": 0.2869099140167236,
+ "num_input_tokens_seen": 1244567184,
+ "step": 28455,
+ "train_runtime": 220111.9824,
+ "train_tokens_per_second": 5654.245
+ },
+ {
+ "epoch": 1.916240237005117,
+ "grad_norm": 0.7565359558538893,
+ "learning_rate": 4.328336299028735e-08,
+ "loss": 0.27414395809173586,
+ "num_input_tokens_seen": 1244775824,
+ "step": 28460,
+ "train_runtime": 220156.5249,
+ "train_tokens_per_second": 5654.049
+ },
+ {
+ "epoch": 1.9165768920010773,
+ "grad_norm": 0.8537030065422263,
+ "learning_rate": 4.293690100089731e-08,
+ "loss": 0.30220975875854494,
+ "num_input_tokens_seen": 1244988464,
+ "step": 28465,
+ "train_runtime": 220195.8129,
+ "train_tokens_per_second": 5654.006
+ },
+ {
+ "epoch": 1.9169135469970375,
+ "grad_norm": 0.8148234793335632,
+ "learning_rate": 4.259182523839556e-08,
+ "loss": 0.29154458045959475,
+ "num_input_tokens_seen": 1245214008,
+ "step": 28470,
+ "train_runtime": 220234.437,
+ "train_tokens_per_second": 5654.039
+ },
+ {
+ "epoch": 1.9172502019929976,
+ "grad_norm": 0.7939908369257882,
+ "learning_rate": 4.224813579928211e-08,
+ "loss": 0.3210744857788086,
+ "num_input_tokens_seen": 1245434864,
+ "step": 28475,
+ "train_runtime": 220278.6573,
+ "train_tokens_per_second": 5653.906
+ },
+ {
+ "epoch": 1.9175868569889576,
+ "grad_norm": 0.7166123345087314,
+ "learning_rate": 4.190583277966842e-08,
+ "loss": 0.29584832191467286,
+ "num_input_tokens_seen": 1245653840,
+ "step": 28480,
+ "train_runtime": 220319.9507,
+ "train_tokens_per_second": 5653.84
+ },
+ {
+ "epoch": 1.9179235119849178,
+ "grad_norm": 0.7696523687780278,
+ "learning_rate": 4.1564916275278477e-08,
+ "loss": 0.2821512699127197,
+ "num_input_tokens_seen": 1245862296,
+ "step": 28485,
+ "train_runtime": 220358.5767,
+ "train_tokens_per_second": 5653.795
+ },
+ {
+ "epoch": 1.918260166980878,
+ "grad_norm": 0.7811203217746909,
+ "learning_rate": 4.122538638144824e-08,
+ "loss": 0.2951045989990234,
+ "num_input_tokens_seen": 1246082824,
+ "step": 28490,
+ "train_runtime": 220395.6576,
+ "train_tokens_per_second": 5653.845
+ },
+ {
+ "epoch": 1.9185968219768381,
+ "grad_norm": 0.7044342809522225,
+ "learning_rate": 4.08872431931262e-08,
+ "loss": 0.2757328271865845,
+ "num_input_tokens_seen": 1246295392,
+ "step": 28495,
+ "train_runtime": 220435.9021,
+ "train_tokens_per_second": 5653.777
+ },
+ {
+ "epoch": 1.9189334769727981,
+ "grad_norm": 0.8062793001735447,
+ "learning_rate": 4.055048680487339e-08,
+ "loss": 0.29179792404174804,
+ "num_input_tokens_seen": 1246522464,
+ "step": 28500,
+ "train_runtime": 220473.702,
+ "train_tokens_per_second": 5653.837
+ },
+ {
+ "epoch": 1.9192701319687584,
+ "grad_norm": 0.7448550831538739,
+ "learning_rate": 4.0215117310862806e-08,
+ "loss": 0.3171137809753418,
+ "num_input_tokens_seen": 1246721184,
+ "step": 28505,
+ "train_runtime": 220514.8027,
+ "train_tokens_per_second": 5653.685
+ },
+ {
+ "epoch": 1.9196067869647186,
+ "grad_norm": 0.8454561648217672,
+ "learning_rate": 3.9881134804878316e-08,
+ "loss": 0.3017882347106934,
+ "num_input_tokens_seen": 1246923736,
+ "step": 28510,
+ "train_runtime": 220555.7088,
+ "train_tokens_per_second": 5653.555
+ },
+ {
+ "epoch": 1.9199434419606787,
+ "grad_norm": 0.818261031372424,
+ "learning_rate": 3.9548539380318e-08,
+ "loss": 0.3179955005645752,
+ "num_input_tokens_seen": 1247152480,
+ "step": 28515,
+ "train_runtime": 220588.4295,
+ "train_tokens_per_second": 5653.753
+ },
+ {
+ "epoch": 1.9202800969566387,
+ "grad_norm": 0.7797530065110128,
+ "learning_rate": 3.921733113019077e-08,
+ "loss": 0.3016833305358887,
+ "num_input_tokens_seen": 1247354448,
+ "step": 28520,
+ "train_runtime": 220628.9625,
+ "train_tokens_per_second": 5653.63
+ },
+ {
+ "epoch": 1.920616751952599,
+ "grad_norm": 0.7210084261051174,
+ "learning_rate": 3.888751014711867e-08,
+ "loss": 0.3128508567810059,
+ "num_input_tokens_seen": 1247557312,
+ "step": 28525,
+ "train_runtime": 220662.2317,
+ "train_tokens_per_second": 5653.697
+ },
+ {
+ "epoch": 1.9209534069485592,
+ "grad_norm": 0.6727498501097389,
+ "learning_rate": 3.855907652333402e-08,
+ "loss": 0.28232321739196775,
+ "num_input_tokens_seen": 1247776920,
+ "step": 28530,
+ "train_runtime": 220702.1687,
+ "train_tokens_per_second": 5653.669
+ },
+ {
+ "epoch": 1.9212900619445192,
+ "grad_norm": 0.7326711637458728,
+ "learning_rate": 3.8232030350683344e-08,
+ "loss": 0.2884651184082031,
+ "num_input_tokens_seen": 1248006808,
+ "step": 28535,
+ "train_runtime": 220743.7071,
+ "train_tokens_per_second": 5653.646
+ },
+ {
+ "epoch": 1.9216267169404793,
+ "grad_norm": 0.749333938315018,
+ "learning_rate": 3.7906371720622946e-08,
+ "loss": 0.2990164041519165,
+ "num_input_tokens_seen": 1248234992,
+ "step": 28540,
+ "train_runtime": 220782.3848,
+ "train_tokens_per_second": 5653.689
+ },
+ {
+ "epoch": 1.9219633719364395,
+ "grad_norm": 0.7532923064142389,
+ "learning_rate": 3.758210072422275e-08,
+ "loss": 0.28374738693237306,
+ "num_input_tokens_seen": 1248453088,
+ "step": 28545,
+ "train_runtime": 220818.7593,
+ "train_tokens_per_second": 5653.746
+ },
+ {
+ "epoch": 1.9223000269323998,
+ "grad_norm": 0.8243046568432454,
+ "learning_rate": 3.72592174521641e-08,
+ "loss": 0.30987684726715087,
+ "num_input_tokens_seen": 1248674912,
+ "step": 28550,
+ "train_runtime": 220861.5131,
+ "train_tokens_per_second": 5653.656
+ },
+ {
+ "epoch": 1.9226366819283598,
+ "grad_norm": 0.7456873109655421,
+ "learning_rate": 3.693772199474033e-08,
+ "loss": 0.30152506828308107,
+ "num_input_tokens_seen": 1248896320,
+ "step": 28555,
+ "train_runtime": 220899.3543,
+ "train_tokens_per_second": 5653.689
+ },
+ {
+ "epoch": 1.9229733369243198,
+ "grad_norm": 0.8268089260806681,
+ "learning_rate": 3.661761444185674e-08,
+ "loss": 0.30233216285705566,
+ "num_input_tokens_seen": 1249110256,
+ "step": 28560,
+ "train_runtime": 220941.1918,
+ "train_tokens_per_second": 5653.587
+ },
+ {
+ "epoch": 1.92330999192028,
+ "grad_norm": 0.7596988638251234,
+ "learning_rate": 3.6298894883030066e-08,
+ "loss": 0.3038165807723999,
+ "num_input_tokens_seen": 1249337408,
+ "step": 28565,
+ "train_runtime": 220975.0974,
+ "train_tokens_per_second": 5653.748
+ },
+ {
+ "epoch": 1.9236466469162403,
+ "grad_norm": 0.8230343694103864,
+ "learning_rate": 3.5981563407389007e-08,
+ "loss": 0.3213151454925537,
+ "num_input_tokens_seen": 1249559176,
+ "step": 28570,
+ "train_runtime": 221018.9022,
+ "train_tokens_per_second": 5653.63
+ },
+ {
+ "epoch": 1.9239833019122003,
+ "grad_norm": 0.8032414208109102,
+ "learning_rate": 3.566562010367536e-08,
+ "loss": 0.2826304197311401,
+ "num_input_tokens_seen": 1249766312,
+ "step": 28575,
+ "train_runtime": 221051.4492,
+ "train_tokens_per_second": 5653.735
+ },
+ {
+ "epoch": 1.9243199569081604,
+ "grad_norm": 0.6984009548103529,
+ "learning_rate": 3.535106506024011e-08,
+ "loss": 0.2978407621383667,
+ "num_input_tokens_seen": 1249983696,
+ "step": 28580,
+ "train_runtime": 221088.9385,
+ "train_tokens_per_second": 5653.76
+ },
+ {
+ "epoch": 1.9246566119041206,
+ "grad_norm": 0.7453143461192999,
+ "learning_rate": 3.5037898365049005e-08,
+ "loss": 0.2644341230392456,
+ "num_input_tokens_seen": 1250204168,
+ "step": 28585,
+ "train_runtime": 221131.5643,
+ "train_tokens_per_second": 5653.667
+ },
+ {
+ "epoch": 1.9249932669000809,
+ "grad_norm": 0.8872235907895941,
+ "learning_rate": 3.472612010567755e-08,
+ "loss": 0.29390702247619627,
+ "num_input_tokens_seen": 1250412232,
+ "step": 28590,
+ "train_runtime": 221167.7612,
+ "train_tokens_per_second": 5653.682
+ },
+ {
+ "epoch": 1.925329921896041,
+ "grad_norm": 0.7629679742085492,
+ "learning_rate": 3.441573036931323e-08,
+ "loss": 0.3093965291976929,
+ "num_input_tokens_seen": 1250638880,
+ "step": 28595,
+ "train_runtime": 221205.5961,
+ "train_tokens_per_second": 5653.74
+ },
+ {
+ "epoch": 1.925666576892001,
+ "grad_norm": 0.9764963685352374,
+ "learning_rate": 3.410672924275604e-08,
+ "loss": 0.273404598236084,
+ "num_input_tokens_seen": 1250870096,
+ "step": 28600,
+ "train_runtime": 221249.308,
+ "train_tokens_per_second": 5653.668
+ },
+ {
+ "epoch": 1.9260032318879612,
+ "grad_norm": 0.8128740784863975,
+ "learning_rate": 3.3799116812416876e-08,
+ "loss": 0.3114511489868164,
+ "num_input_tokens_seen": 1251082376,
+ "step": 28605,
+ "train_runtime": 221285.4439,
+ "train_tokens_per_second": 5653.704
+ },
+ {
+ "epoch": 1.9263398868839214,
+ "grad_norm": 0.8683827269674369,
+ "learning_rate": 3.349289316431803e-08,
+ "loss": 0.31067667007446287,
+ "num_input_tokens_seen": 1251307384,
+ "step": 28610,
+ "train_runtime": 221322.0142,
+ "train_tokens_per_second": 5653.786
+ },
+ {
+ "epoch": 1.9266765418798815,
+ "grad_norm": 0.7478140934985664,
+ "learning_rate": 3.3188058384095446e-08,
+ "loss": 0.29280407428741456,
+ "num_input_tokens_seen": 1251525400,
+ "step": 28615,
+ "train_runtime": 221360.0032,
+ "train_tokens_per_second": 5653.801
+ },
+ {
+ "epoch": 1.9270131968758415,
+ "grad_norm": 0.8025594638672747,
+ "learning_rate": 3.2884612556993714e-08,
+ "loss": 0.32468104362487793,
+ "num_input_tokens_seen": 1251744936,
+ "step": 28620,
+ "train_runtime": 221403.9081,
+ "train_tokens_per_second": 5653.671
+ },
+ {
+ "epoch": 1.9273498518718017,
+ "grad_norm": 0.8491439802692531,
+ "learning_rate": 3.258255576787161e-08,
+ "loss": 0.3124075174331665,
+ "num_input_tokens_seen": 1251958328,
+ "step": 28625,
+ "train_runtime": 221443.8187,
+ "train_tokens_per_second": 5653.616
+ },
+ {
+ "epoch": 1.927686506867762,
+ "grad_norm": 0.8546654392090276,
+ "learning_rate": 3.22818881011977e-08,
+ "loss": 0.31823225021362306,
+ "num_input_tokens_seen": 1252192752,
+ "step": 28630,
+ "train_runtime": 221481.8825,
+ "train_tokens_per_second": 5653.703
+ },
+ {
+ "epoch": 1.928023161863722,
+ "grad_norm": 0.7973579543913553,
+ "learning_rate": 3.198260964105249e-08,
+ "loss": 0.27967541217803954,
+ "num_input_tokens_seen": 1252407792,
+ "step": 28635,
+ "train_runtime": 221519.1612,
+ "train_tokens_per_second": 5653.722
+ },
+ {
+ "epoch": 1.928359816859682,
+ "grad_norm": 0.7954284380902052,
+ "learning_rate": 3.168472047112903e-08,
+ "loss": 0.2996687412261963,
+ "num_input_tokens_seen": 1252623280,
+ "step": 28640,
+ "train_runtime": 221566.1063,
+ "train_tokens_per_second": 5653.497
+ },
+ {
+ "epoch": 1.9286964718556423,
+ "grad_norm": 0.7984207095399551,
+ "learning_rate": 3.138822067473068e-08,
+ "loss": 0.28585450649261473,
+ "num_input_tokens_seen": 1252835408,
+ "step": 28645,
+ "train_runtime": 221604.288,
+ "train_tokens_per_second": 5653.48
+ },
+ {
+ "epoch": 1.9290331268516026,
+ "grad_norm": 0.7748173909126452,
+ "learning_rate": 3.10931103347728e-08,
+ "loss": 0.30099103450775144,
+ "num_input_tokens_seen": 1253057512,
+ "step": 28650,
+ "train_runtime": 221646.7214,
+ "train_tokens_per_second": 5653.4
+ },
+ {
+ "epoch": 1.9293697818475626,
+ "grad_norm": 0.7586349259346754,
+ "learning_rate": 3.0799389533781587e-08,
+ "loss": 0.3233338356018066,
+ "num_input_tokens_seen": 1253285920,
+ "step": 28655,
+ "train_runtime": 221683.9501,
+ "train_tokens_per_second": 5653.481
+ },
+ {
+ "epoch": 1.9297064368435226,
+ "grad_norm": 0.7596388594233524,
+ "learning_rate": 3.050705835389578e-08,
+ "loss": 0.28039982318878176,
+ "num_input_tokens_seen": 1253502504,
+ "step": 28660,
+ "train_runtime": 221722.1476,
+ "train_tokens_per_second": 5653.484
+ },
+ {
+ "epoch": 1.9300430918394829,
+ "grad_norm": 0.8761598764006204,
+ "learning_rate": 3.021611687686443e-08,
+ "loss": 0.2949859380722046,
+ "num_input_tokens_seen": 1253714448,
+ "step": 28665,
+ "train_runtime": 221759.6137,
+ "train_tokens_per_second": 5653.484
+ },
+ {
+ "epoch": 1.9303797468354431,
+ "grad_norm": 0.8219444240328218,
+ "learning_rate": 2.992656518404857e-08,
+ "loss": 0.3135673999786377,
+ "num_input_tokens_seen": 1253921528,
+ "step": 28670,
+ "train_runtime": 221798.9051,
+ "train_tokens_per_second": 5653.416
+ },
+ {
+ "epoch": 1.9307164018314031,
+ "grad_norm": 0.661270135482443,
+ "learning_rate": 2.96384033564201e-08,
+ "loss": 0.28868966102600097,
+ "num_input_tokens_seen": 1254137480,
+ "step": 28675,
+ "train_runtime": 221835.7558,
+ "train_tokens_per_second": 5653.451
+ },
+ {
+ "epoch": 1.9310530568273632,
+ "grad_norm": 0.767415898484409,
+ "learning_rate": 2.935163147456288e-08,
+ "loss": 0.27773294448852537,
+ "num_input_tokens_seen": 1254360912,
+ "step": 28680,
+ "train_runtime": 221876.4785,
+ "train_tokens_per_second": 5653.42
+ },
+ {
+ "epoch": 1.9313897118233234,
+ "grad_norm": 0.8423049211038874,
+ "learning_rate": 2.906624961867166e-08,
+ "loss": 0.3049351215362549,
+ "num_input_tokens_seen": 1254581976,
+ "step": 28685,
+ "train_runtime": 221913.3786,
+ "train_tokens_per_second": 5653.476
+ },
+ {
+ "epoch": 1.9317263668192837,
+ "grad_norm": 0.8039083333559773,
+ "learning_rate": 2.8782257868553154e-08,
+ "loss": 0.2789355754852295,
+ "num_input_tokens_seen": 1254798040,
+ "step": 28690,
+ "train_runtime": 221950.8121,
+ "train_tokens_per_second": 5653.496
+ },
+ {
+ "epoch": 1.9320630218152437,
+ "grad_norm": 0.7939203667614896,
+ "learning_rate": 2.849965630362328e-08,
+ "loss": 0.2943142890930176,
+ "num_input_tokens_seen": 1255008552,
+ "step": 28695,
+ "train_runtime": 221991.6931,
+ "train_tokens_per_second": 5653.403
+ },
+ {
+ "epoch": 1.9323996768112037,
+ "grad_norm": 1.040825902674123,
+ "learning_rate": 2.8218445002912153e-08,
+ "loss": 0.2914332389831543,
+ "num_input_tokens_seen": 1255234552,
+ "step": 28700,
+ "train_runtime": 222032.4922,
+ "train_tokens_per_second": 5653.382
+ },
+ {
+ "epoch": 1.932736331807164,
+ "grad_norm": 0.7879730035255785,
+ "learning_rate": 2.7938624045059094e-08,
+ "loss": 0.2954190731048584,
+ "num_input_tokens_seen": 1255459816,
+ "step": 28705,
+ "train_runtime": 222067.1659,
+ "train_tokens_per_second": 5653.514
+ },
+ {
+ "epoch": 1.9330729868031242,
+ "grad_norm": 0.8446326753294154,
+ "learning_rate": 2.766019350831428e-08,
+ "loss": 0.3093540668487549,
+ "num_input_tokens_seen": 1255699664,
+ "step": 28710,
+ "train_runtime": 222106.7303,
+ "train_tokens_per_second": 5653.587
+ },
+ {
+ "epoch": 1.9334096417990843,
+ "grad_norm": 0.7880511342051334,
+ "learning_rate": 2.7383153470541546e-08,
+ "loss": 0.3093677282333374,
+ "num_input_tokens_seen": 1255930096,
+ "step": 28715,
+ "train_runtime": 222139.5127,
+ "train_tokens_per_second": 5653.79
+ },
+ {
+ "epoch": 1.9337462967950443,
+ "grad_norm": 0.7912632326756309,
+ "learning_rate": 2.710750400921336e-08,
+ "loss": 0.2810804843902588,
+ "num_input_tokens_seen": 1256129960,
+ "step": 28720,
+ "train_runtime": 222179.3382,
+ "train_tokens_per_second": 5653.676
+ },
+ {
+ "epoch": 1.9340829517910045,
+ "grad_norm": 0.7097484635650013,
+ "learning_rate": 2.683324520141417e-08,
+ "loss": 0.29244356155395507,
+ "num_input_tokens_seen": 1256355072,
+ "step": 28725,
+ "train_runtime": 222220.9674,
+ "train_tokens_per_second": 5653.63
+ },
+ {
+ "epoch": 1.9344196067869648,
+ "grad_norm": 0.8119430664301859,
+ "learning_rate": 2.65603771238393e-08,
+ "loss": 0.3020967483520508,
+ "num_input_tokens_seen": 1256563176,
+ "step": 28730,
+ "train_runtime": 222266.5769,
+ "train_tokens_per_second": 5653.406
+ },
+ {
+ "epoch": 1.9347562617829248,
+ "grad_norm": 0.7459006681748622,
+ "learning_rate": 2.6288899852796036e-08,
+ "loss": 0.2978071689605713,
+ "num_input_tokens_seen": 1256797928,
+ "step": 28735,
+ "train_runtime": 222306.2836,
+ "train_tokens_per_second": 5653.452
+ },
+ {
+ "epoch": 1.9350929167788848,
+ "grad_norm": 0.8087885741576003,
+ "learning_rate": 2.6018813464202543e-08,
+ "loss": 0.302975058555603,
+ "num_input_tokens_seen": 1257009296,
+ "step": 28740,
+ "train_runtime": 222341.5433,
+ "train_tokens_per_second": 5653.506
+ },
+ {
+ "epoch": 1.935429571774845,
+ "grad_norm": 0.727456579862242,
+ "learning_rate": 2.575011803358618e-08,
+ "loss": 0.29305691719055177,
+ "num_input_tokens_seen": 1257233712,
+ "step": 28745,
+ "train_runtime": 222380.0158,
+ "train_tokens_per_second": 5653.537
+ },
+ {
+ "epoch": 1.9357662267708053,
+ "grad_norm": 0.6572939384489086,
+ "learning_rate": 2.5482813636087955e-08,
+ "loss": 0.27554895877838137,
+ "num_input_tokens_seen": 1257442168,
+ "step": 28750,
+ "train_runtime": 222423.441,
+ "train_tokens_per_second": 5653.371
+ },
+ {
+ "epoch": 1.9361028817667654,
+ "grad_norm": 0.7993380569308158,
+ "learning_rate": 2.5216900346458073e-08,
+ "loss": 0.2871532678604126,
+ "num_input_tokens_seen": 1257660464,
+ "step": 28755,
+ "train_runtime": 222466.1845,
+ "train_tokens_per_second": 5653.266
+ },
+ {
+ "epoch": 1.9364395367627254,
+ "grad_norm": 0.766816269617841,
+ "learning_rate": 2.4952378239058716e-08,
+ "loss": 0.29356064796447756,
+ "num_input_tokens_seen": 1257889496,
+ "step": 28760,
+ "train_runtime": 222499.466,
+ "train_tokens_per_second": 5653.45
+ },
+ {
+ "epoch": 1.9367761917586857,
+ "grad_norm": 0.7954103166328962,
+ "learning_rate": 2.4689247387862934e-08,
+ "loss": 0.2963290214538574,
+ "num_input_tokens_seen": 1258103416,
+ "step": 28765,
+ "train_runtime": 222532.7425,
+ "train_tokens_per_second": 5653.565
+ },
+ {
+ "epoch": 1.937112846754646,
+ "grad_norm": 0.7997832806288906,
+ "learning_rate": 2.4427507866453537e-08,
+ "loss": 0.28399271965026857,
+ "num_input_tokens_seen": 1258294944,
+ "step": 28770,
+ "train_runtime": 222573.7553,
+ "train_tokens_per_second": 5653.384
+ },
+ {
+ "epoch": 1.937449501750606,
+ "grad_norm": 0.885443863513371,
+ "learning_rate": 2.416715974802586e-08,
+ "loss": 0.2864212989807129,
+ "num_input_tokens_seen": 1258505472,
+ "step": 28775,
+ "train_runtime": 222608.5067,
+ "train_tokens_per_second": 5653.447
+ },
+ {
+ "epoch": 1.937786156746566,
+ "grad_norm": 0.7164017486860768,
+ "learning_rate": 2.3908203105384997e-08,
+ "loss": 0.30834760665893557,
+ "num_input_tokens_seen": 1258727136,
+ "step": 28780,
+ "train_runtime": 222642.3336,
+ "train_tokens_per_second": 5653.584
+ },
+ {
+ "epoch": 1.9381228117425262,
+ "grad_norm": 0.7587198432537924,
+ "learning_rate": 2.365063801094747e-08,
+ "loss": 0.2865851402282715,
+ "num_input_tokens_seen": 1258942528,
+ "step": 28785,
+ "train_runtime": 222676.6074,
+ "train_tokens_per_second": 5653.681
+ },
+ {
+ "epoch": 1.9384594667384865,
+ "grad_norm": 0.706256825592826,
+ "learning_rate": 2.3394464536740658e-08,
+ "loss": 0.30954954624176023,
+ "num_input_tokens_seen": 1259175632,
+ "step": 28790,
+ "train_runtime": 222718.1525,
+ "train_tokens_per_second": 5653.673
+ },
+ {
+ "epoch": 1.9387961217344465,
+ "grad_norm": 0.8885250264575596,
+ "learning_rate": 2.3139682754402816e-08,
+ "loss": 0.3149877548217773,
+ "num_input_tokens_seen": 1259405216,
+ "step": 28795,
+ "train_runtime": 222755.681,
+ "train_tokens_per_second": 5653.751
+ },
+ {
+ "epoch": 1.9391327767304065,
+ "grad_norm": 0.7626354653879915,
+ "learning_rate": 2.2886292735182502e-08,
+ "loss": 0.2907993316650391,
+ "num_input_tokens_seen": 1259637856,
+ "step": 28800,
+ "train_runtime": 222792.0279,
+ "train_tokens_per_second": 5653.873
+ },
+ {
+ "epoch": 1.9394694317263668,
+ "grad_norm": 0.7549952496575574,
+ "learning_rate": 2.2634294549939705e-08,
+ "loss": 0.2934614896774292,
+ "num_input_tokens_seen": 1259866088,
+ "step": 28805,
+ "train_runtime": 222833.3252,
+ "train_tokens_per_second": 5653.85
+ },
+ {
+ "epoch": 1.939806086722327,
+ "grad_norm": 0.7716438013359953,
+ "learning_rate": 2.2383688269144723e-08,
+ "loss": 0.3583078145980835,
+ "num_input_tokens_seen": 1260088088,
+ "step": 28810,
+ "train_runtime": 222873.4346,
+ "train_tokens_per_second": 5653.828
+ },
+ {
+ "epoch": 1.940142741718287,
+ "grad_norm": 0.7760006769888061,
+ "learning_rate": 2.2134473962878712e-08,
+ "loss": 0.2929667949676514,
+ "num_input_tokens_seen": 1260302240,
+ "step": 28815,
+ "train_runtime": 222909.5666,
+ "train_tokens_per_second": 5653.872
+ },
+ {
+ "epoch": 1.940479396714247,
+ "grad_norm": 0.9788787208283655,
+ "learning_rate": 2.1886651700833705e-08,
+ "loss": 0.28723697662353515,
+ "num_input_tokens_seen": 1260489696,
+ "step": 28820,
+ "train_runtime": 222944.9358,
+ "train_tokens_per_second": 5653.816
+ },
+ {
+ "epoch": 1.9408160517102075,
+ "grad_norm": 0.6725253717418324,
+ "learning_rate": 2.1640221552312603e-08,
+ "loss": 0.3041691780090332,
+ "num_input_tokens_seen": 1260707328,
+ "step": 28825,
+ "train_runtime": 222983.6272,
+ "train_tokens_per_second": 5653.811
+ },
+ {
+ "epoch": 1.9411527067061676,
+ "grad_norm": 1.0073697163500734,
+ "learning_rate": 2.1395183586229163e-08,
+ "loss": 0.30638680458068845,
+ "num_input_tokens_seen": 1260942368,
+ "step": 28830,
+ "train_runtime": 223022.3981,
+ "train_tokens_per_second": 5653.882
+ },
+ {
+ "epoch": 1.9414893617021276,
+ "grad_norm": 0.7599893500344853,
+ "learning_rate": 2.1151537871106353e-08,
+ "loss": 0.28965296745300295,
+ "num_input_tokens_seen": 1261152728,
+ "step": 28835,
+ "train_runtime": 223058.9354,
+ "train_tokens_per_second": 5653.899
+ },
+ {
+ "epoch": 1.9418260166980879,
+ "grad_norm": 0.7939541327760055,
+ "learning_rate": 2.0909284475080226e-08,
+ "loss": 0.285279655456543,
+ "num_input_tokens_seen": 1261373808,
+ "step": 28840,
+ "train_runtime": 223107.3758,
+ "train_tokens_per_second": 5653.663
+ },
+ {
+ "epoch": 1.942162671694048,
+ "grad_norm": 0.8217446318942616,
+ "learning_rate": 2.0668423465894928e-08,
+ "loss": 0.2754975318908691,
+ "num_input_tokens_seen": 1261592168,
+ "step": 28845,
+ "train_runtime": 223148.9614,
+ "train_tokens_per_second": 5653.587
+ },
+ {
+ "epoch": 1.9424993266900081,
+ "grad_norm": 0.7874416991261182,
+ "learning_rate": 2.0428954910907684e-08,
+ "loss": 0.28652522563934324,
+ "num_input_tokens_seen": 1261787104,
+ "step": 28850,
+ "train_runtime": 223186.7342,
+ "train_tokens_per_second": 5653.504
+ },
+ {
+ "epoch": 1.9428359816859682,
+ "grad_norm": 0.8264865579012179,
+ "learning_rate": 2.019087887708382e-08,
+ "loss": 0.3006443977355957,
+ "num_input_tokens_seen": 1262007728,
+ "step": 28855,
+ "train_runtime": 223227.7335,
+ "train_tokens_per_second": 5653.454
+ },
+ {
+ "epoch": 1.9431726366819284,
+ "grad_norm": 0.7815316131606884,
+ "learning_rate": 1.9954195431001744e-08,
+ "loss": 0.29778175354003905,
+ "num_input_tokens_seen": 1262233872,
+ "step": 28860,
+ "train_runtime": 223262.3109,
+ "train_tokens_per_second": 5653.591
+ },
+ {
+ "epoch": 1.9435092916778887,
+ "grad_norm": 0.7625140991861158,
+ "learning_rate": 1.9718904638848513e-08,
+ "loss": 0.2813864231109619,
+ "num_input_tokens_seen": 1262445256,
+ "step": 28865,
+ "train_runtime": 223305.0134,
+ "train_tokens_per_second": 5653.457
+ },
+ {
+ "epoch": 1.9438459466738487,
+ "grad_norm": 0.819180819579518,
+ "learning_rate": 1.9485006566422602e-08,
+ "loss": 0.2955453395843506,
+ "num_input_tokens_seen": 1262680216,
+ "step": 28870,
+ "train_runtime": 223347.1343,
+ "train_tokens_per_second": 5653.443
+ },
+ {
+ "epoch": 1.9441826016698087,
+ "grad_norm": 0.7141341923837556,
+ "learning_rate": 1.9252501279132806e-08,
+ "loss": 0.3031654357910156,
+ "num_input_tokens_seen": 1262895920,
+ "step": 28875,
+ "train_runtime": 223391.4022,
+ "train_tokens_per_second": 5653.288
+ },
+ {
+ "epoch": 1.944519256665769,
+ "grad_norm": 0.7599819750131713,
+ "learning_rate": 1.902138884199878e-08,
+ "loss": 0.2994793176651001,
+ "num_input_tokens_seen": 1263122360,
+ "step": 28880,
+ "train_runtime": 223427.1868,
+ "train_tokens_per_second": 5653.396
+ },
+ {
+ "epoch": 1.9448559116617292,
+ "grad_norm": 0.7754812262214777,
+ "learning_rate": 1.879166931964993e-08,
+ "loss": 0.30690946578979494,
+ "num_input_tokens_seen": 1263337688,
+ "step": 28885,
+ "train_runtime": 223467.6117,
+ "train_tokens_per_second": 5653.337
+ },
+ {
+ "epoch": 1.9451925666576892,
+ "grad_norm": 0.7766593010208253,
+ "learning_rate": 1.856334277632765e-08,
+ "loss": 0.29476094245910645,
+ "num_input_tokens_seen": 1263574904,
+ "step": 28890,
+ "train_runtime": 223506.8009,
+ "train_tokens_per_second": 5653.407
+ },
+ {
+ "epoch": 1.9455292216536493,
+ "grad_norm": 0.6831350388874682,
+ "learning_rate": 1.8336409275880872e-08,
+ "loss": 0.3002762794494629,
+ "num_input_tokens_seen": 1263807760,
+ "step": 28895,
+ "train_runtime": 223552.6379,
+ "train_tokens_per_second": 5653.289
+ },
+ {
+ "epoch": 1.9458658766496095,
+ "grad_norm": 0.7320234419132584,
+ "learning_rate": 1.8110868881772713e-08,
+ "loss": 0.2916655302047729,
+ "num_input_tokens_seen": 1264016568,
+ "step": 28900,
+ "train_runtime": 223592.718,
+ "train_tokens_per_second": 5653.21
+ },
+ {
+ "epoch": 1.9462025316455698,
+ "grad_norm": 0.7177337064879638,
+ "learning_rate": 1.788672165707328e-08,
+ "loss": 0.2689183712005615,
+ "num_input_tokens_seen": 1264236144,
+ "step": 28905,
+ "train_runtime": 223637.4105,
+ "train_tokens_per_second": 5653.062
+ },
+ {
+ "epoch": 1.9465391866415298,
+ "grad_norm": 0.7335996504510445,
+ "learning_rate": 1.7663967664465763e-08,
+ "loss": 0.2718139886856079,
+ "num_input_tokens_seen": 1264443160,
+ "step": 28910,
+ "train_runtime": 223677.4438,
+ "train_tokens_per_second": 5652.976
+ },
+ {
+ "epoch": 1.9468758416374898,
+ "grad_norm": 0.7751823359014647,
+ "learning_rate": 1.7442606966242005e-08,
+ "loss": 0.3149592399597168,
+ "num_input_tokens_seen": 1264681216,
+ "step": 28915,
+ "train_runtime": 223720.453,
+ "train_tokens_per_second": 5652.953
+ },
+ {
+ "epoch": 1.94721249663345,
+ "grad_norm": 0.7519570365655589,
+ "learning_rate": 1.722263962430526e-08,
+ "loss": 0.3035134315490723,
+ "num_input_tokens_seen": 1264904128,
+ "step": 28920,
+ "train_runtime": 223755.8381,
+ "train_tokens_per_second": 5653.055
+ },
+ {
+ "epoch": 1.9475491516294103,
+ "grad_norm": 0.828398864844539,
+ "learning_rate": 1.7004065700168547e-08,
+ "loss": 0.28637185096740725,
+ "num_input_tokens_seen": 1265126560,
+ "step": 28925,
+ "train_runtime": 223788.0098,
+ "train_tokens_per_second": 5653.237
+ },
+ {
+ "epoch": 1.9478858066253704,
+ "grad_norm": 0.700009166031884,
+ "learning_rate": 1.6786885254954644e-08,
+ "loss": 0.3050400733947754,
+ "num_input_tokens_seen": 1265361040,
+ "step": 28930,
+ "train_runtime": 223824.3291,
+ "train_tokens_per_second": 5653.367
+ },
+ {
+ "epoch": 1.9482224616213304,
+ "grad_norm": 0.8369964157697908,
+ "learning_rate": 1.6571098349398296e-08,
+ "loss": 0.2956183910369873,
+ "num_input_tokens_seen": 1265565976,
+ "step": 28935,
+ "train_runtime": 223863.4274,
+ "train_tokens_per_second": 5653.295
+ },
+ {
+ "epoch": 1.9485591166172906,
+ "grad_norm": 0.8754073900671024,
+ "learning_rate": 1.635670504384346e-08,
+ "loss": 0.31615753173828126,
+ "num_input_tokens_seen": 1265765816,
+ "step": 28940,
+ "train_runtime": 223908.9438,
+ "train_tokens_per_second": 5653.038
+ },
+ {
+ "epoch": 1.948895771613251,
+ "grad_norm": 0.941511353921259,
+ "learning_rate": 1.6143705398243837e-08,
+ "loss": 0.29006495475769045,
+ "num_input_tokens_seen": 1265969312,
+ "step": 28945,
+ "train_runtime": 223952.0727,
+ "train_tokens_per_second": 5652.858
+ },
+ {
+ "epoch": 1.949232426609211,
+ "grad_norm": 0.7913567769706255,
+ "learning_rate": 1.5932099472165674e-08,
+ "loss": 0.29108805656433107,
+ "num_input_tokens_seen": 1266179488,
+ "step": 28950,
+ "train_runtime": 223990.7097,
+ "train_tokens_per_second": 5652.821
+ },
+ {
+ "epoch": 1.949569081605171,
+ "grad_norm": 0.7310227670184213,
+ "learning_rate": 1.572188732478164e-08,
+ "loss": 0.2916900157928467,
+ "num_input_tokens_seen": 1266409776,
+ "step": 28955,
+ "train_runtime": 224030.8188,
+ "train_tokens_per_second": 5652.837
+ },
+ {
+ "epoch": 1.9499057366011312,
+ "grad_norm": 0.7938589935789288,
+ "learning_rate": 1.551306901487859e-08,
+ "loss": 0.3035112857818604,
+ "num_input_tokens_seen": 1266624136,
+ "step": 28960,
+ "train_runtime": 224074.3669,
+ "train_tokens_per_second": 5652.695
+ },
+ {
+ "epoch": 1.9502423915970915,
+ "grad_norm": 0.6920718267717534,
+ "learning_rate": 1.5305644600852043e-08,
+ "loss": 0.2758214473724365,
+ "num_input_tokens_seen": 1266855056,
+ "step": 28965,
+ "train_runtime": 224114.5981,
+ "train_tokens_per_second": 5652.711
+ },
+ {
+ "epoch": 1.9505790465930515,
+ "grad_norm": 0.7554921758498927,
+ "learning_rate": 1.5099614140706154e-08,
+ "loss": 0.2854607105255127,
+ "num_input_tokens_seen": 1267072344,
+ "step": 28970,
+ "train_runtime": 224158.7838,
+ "train_tokens_per_second": 5652.566
+ },
+ {
+ "epoch": 1.9509157015890115,
+ "grad_norm": 0.8000972769551457,
+ "learning_rate": 1.489497769205761e-08,
+ "loss": 0.31021819114685056,
+ "num_input_tokens_seen": 1267277496,
+ "step": 28975,
+ "train_runtime": 224199.1913,
+ "train_tokens_per_second": 5652.462
+ },
+ {
+ "epoch": 1.9512523565849718,
+ "grad_norm": 0.7228114152004868,
+ "learning_rate": 1.4691735312132304e-08,
+ "loss": 0.2921774864196777,
+ "num_input_tokens_seen": 1267506744,
+ "step": 28980,
+ "train_runtime": 224237.6676,
+ "train_tokens_per_second": 5652.515
+ },
+ {
+ "epoch": 1.951589011580932,
+ "grad_norm": 0.7982672860216988,
+ "learning_rate": 1.4489887057766439e-08,
+ "loss": 0.30820393562316895,
+ "num_input_tokens_seen": 1267724224,
+ "step": 28985,
+ "train_runtime": 224272.862,
+ "train_tokens_per_second": 5652.598
+ },
+ {
+ "epoch": 1.951925666576892,
+ "grad_norm": 0.7019172612562962,
+ "learning_rate": 1.428943298540597e-08,
+ "loss": 0.29095611572265623,
+ "num_input_tokens_seen": 1267950120,
+ "step": 28990,
+ "train_runtime": 224309.6136,
+ "train_tokens_per_second": 5652.678
+ },
+ {
+ "epoch": 1.952262321572852,
+ "grad_norm": 0.6605820071712591,
+ "learning_rate": 1.4090373151107173e-08,
+ "loss": 0.3069890022277832,
+ "num_input_tokens_seen": 1268176376,
+ "step": 28995,
+ "train_runtime": 224342.3845,
+ "train_tokens_per_second": 5652.861
+ },
+ {
+ "epoch": 1.9525989765688123,
+ "grad_norm": 0.7607077702464043,
+ "learning_rate": 1.3892707610536627e-08,
+ "loss": 0.2565622329711914,
+ "num_input_tokens_seen": 1268384472,
+ "step": 29000,
+ "train_runtime": 224386.5442,
+ "train_tokens_per_second": 5652.676
+ },
+ {
+ "epoch": 1.9529356315647726,
+ "grad_norm": 0.806764475508148,
+ "learning_rate": 1.3696436418970672e-08,
+ "loss": 0.2830059051513672,
+ "num_input_tokens_seen": 1268601536,
+ "step": 29005,
+ "train_runtime": 224428.0534,
+ "train_tokens_per_second": 5652.598
+ },
+ {
+ "epoch": 1.9532722865607326,
+ "grad_norm": 0.8503139141015114,
+ "learning_rate": 1.3501559631296512e-08,
+ "loss": 0.31444830894470216,
+ "num_input_tokens_seen": 1268805592,
+ "step": 29010,
+ "train_runtime": 224471.139,
+ "train_tokens_per_second": 5652.422
+ },
+ {
+ "epoch": 1.9536089415566926,
+ "grad_norm": 0.7710744016084916,
+ "learning_rate": 1.3308077302010557e-08,
+ "loss": 0.31547060012817385,
+ "num_input_tokens_seen": 1269039096,
+ "step": 29015,
+ "train_runtime": 224509.2303,
+ "train_tokens_per_second": 5652.503
+ },
+ {
+ "epoch": 1.9539455965526529,
+ "grad_norm": 0.8026179441993274,
+ "learning_rate": 1.3115989485218972e-08,
+ "loss": 0.2989495754241943,
+ "num_input_tokens_seen": 1269246368,
+ "step": 29020,
+ "train_runtime": 224543.5644,
+ "train_tokens_per_second": 5652.562
+ },
+ {
+ "epoch": 1.9542822515486131,
+ "grad_norm": 0.805001102582023,
+ "learning_rate": 1.2925296234638784e-08,
+ "loss": 0.2807016849517822,
+ "num_input_tokens_seen": 1269464104,
+ "step": 29025,
+ "train_runtime": 224584.5623,
+ "train_tokens_per_second": 5652.499
+ },
+ {
+ "epoch": 1.9546189065445732,
+ "grad_norm": 0.6954871573935902,
+ "learning_rate": 1.2735997603597339e-08,
+ "loss": 0.29611780643463137,
+ "num_input_tokens_seen": 1269699728,
+ "step": 29030,
+ "train_runtime": 224616.5365,
+ "train_tokens_per_second": 5652.744
+ },
+ {
+ "epoch": 1.9549555615405332,
+ "grad_norm": 0.857067719947332,
+ "learning_rate": 1.2548093645030623e-08,
+ "loss": 0.2928828001022339,
+ "num_input_tokens_seen": 1269928480,
+ "step": 29035,
+ "train_runtime": 224656.3805,
+ "train_tokens_per_second": 5652.759
+ },
+ {
+ "epoch": 1.9552922165364934,
+ "grad_norm": 0.7481342627141493,
+ "learning_rate": 1.236158441148605e-08,
+ "loss": 0.28606395721435546,
+ "num_input_tokens_seen": 1270155816,
+ "step": 29040,
+ "train_runtime": 224692.812,
+ "train_tokens_per_second": 5652.855
+ },
+ {
+ "epoch": 1.9556288715324537,
+ "grad_norm": 0.8112388374034099,
+ "learning_rate": 1.2176469955119119e-08,
+ "loss": 0.32508511543273927,
+ "num_input_tokens_seen": 1270385632,
+ "step": 29045,
+ "train_runtime": 224733.026,
+ "train_tokens_per_second": 5652.866
+ },
+ {
+ "epoch": 1.9559655265284137,
+ "grad_norm": 0.8038650607300725,
+ "learning_rate": 1.199275032769842e-08,
+ "loss": 0.3148582696914673,
+ "num_input_tokens_seen": 1270610312,
+ "step": 29050,
+ "train_runtime": 224776.4013,
+ "train_tokens_per_second": 5652.775
+ },
+ {
+ "epoch": 1.9563021815243737,
+ "grad_norm": 0.8038054887828693,
+ "learning_rate": 1.1810425580598971e-08,
+ "loss": 0.32245888710021975,
+ "num_input_tokens_seen": 1270844152,
+ "step": 29055,
+ "train_runtime": 224817.2922,
+ "train_tokens_per_second": 5652.786
+ },
+ {
+ "epoch": 1.956638836520334,
+ "grad_norm": 0.8285772488267888,
+ "learning_rate": 1.1629495764807763e-08,
+ "loss": 0.29970650672912597,
+ "num_input_tokens_seen": 1271079472,
+ "step": 29060,
+ "train_runtime": 224861.1966,
+ "train_tokens_per_second": 5652.729
+ },
+ {
+ "epoch": 1.9569754915162942,
+ "grad_norm": 0.7500252998485927,
+ "learning_rate": 1.1449960930921544e-08,
+ "loss": 0.29777934551239016,
+ "num_input_tokens_seen": 1271292624,
+ "step": 29065,
+ "train_runtime": 224898.4968,
+ "train_tokens_per_second": 5652.74
+ },
+ {
+ "epoch": 1.9573121465122543,
+ "grad_norm": 0.7492263033388418,
+ "learning_rate": 1.1271821129146266e-08,
+ "loss": 0.28283419609069826,
+ "num_input_tokens_seen": 1271514960,
+ "step": 29070,
+ "train_runtime": 224932.5444,
+ "train_tokens_per_second": 5652.872
+ },
+ {
+ "epoch": 1.9576488015082143,
+ "grad_norm": 0.794721710026827,
+ "learning_rate": 1.1095076409298189e-08,
+ "loss": 0.28227386474609373,
+ "num_input_tokens_seen": 1271727712,
+ "step": 29075,
+ "train_runtime": 224970.8077,
+ "train_tokens_per_second": 5652.857
+ },
+ {
+ "epoch": 1.9579854565041745,
+ "grad_norm": 0.8361082033343284,
+ "learning_rate": 1.0919726820803889e-08,
+ "loss": 0.2864542484283447,
+ "num_input_tokens_seen": 1271937480,
+ "step": 29080,
+ "train_runtime": 225007.7296,
+ "train_tokens_per_second": 5652.861
+ },
+ {
+ "epoch": 1.9583221115001348,
+ "grad_norm": 0.7529640945800179,
+ "learning_rate": 1.0745772412698585e-08,
+ "loss": 0.2920374870300293,
+ "num_input_tokens_seen": 1272150888,
+ "step": 29085,
+ "train_runtime": 225047.2154,
+ "train_tokens_per_second": 5652.818
+ },
+ {
+ "epoch": 1.9586587664960948,
+ "grad_norm": 0.7725658212667705,
+ "learning_rate": 1.057321323362892e-08,
+ "loss": 0.28984389305114744,
+ "num_input_tokens_seen": 1272385000,
+ "step": 29090,
+ "train_runtime": 225088.6801,
+ "train_tokens_per_second": 5652.816
+ },
+ {
+ "epoch": 1.9589954214920549,
+ "grad_norm": 0.7321569817209439,
+ "learning_rate": 1.0402049331849629e-08,
+ "loss": 0.2945473432540894,
+ "num_input_tokens_seen": 1272608520,
+ "step": 29095,
+ "train_runtime": 225120.0817,
+ "train_tokens_per_second": 5653.021
+ },
+ {
+ "epoch": 1.959332076488015,
+ "grad_norm": 0.7684463547894764,
+ "learning_rate": 1.0232280755226865e-08,
+ "loss": 0.30609171390533446,
+ "num_input_tokens_seen": 1272824152,
+ "step": 29100,
+ "train_runtime": 225164.1788,
+ "train_tokens_per_second": 5652.871
+ },
+ {
+ "epoch": 1.9596687314839754,
+ "grad_norm": 0.6970891800486236,
+ "learning_rate": 1.006390755123543e-08,
+ "loss": 0.2889437675476074,
+ "num_input_tokens_seen": 1273039232,
+ "step": 29105,
+ "train_runtime": 225204.746,
+ "train_tokens_per_second": 5652.808
+ },
+ {
+ "epoch": 1.9600053864799354,
+ "grad_norm": 0.8951268014858879,
+ "learning_rate": 9.89692976696044e-09,
+ "loss": 0.3118439674377441,
+ "num_input_tokens_seen": 1273253968,
+ "step": 29110,
+ "train_runtime": 225250.2013,
+ "train_tokens_per_second": 5652.621
+ },
+ {
+ "epoch": 1.9603420414758954,
+ "grad_norm": 0.7116868073327594,
+ "learning_rate": 9.731347449097316e-09,
+ "loss": 0.31283159255981446,
+ "num_input_tokens_seen": 1273474224,
+ "step": 29115,
+ "train_runtime": 225292.9511,
+ "train_tokens_per_second": 5652.526
+ },
+ {
+ "epoch": 1.9606786964718557,
+ "grad_norm": 0.7624467705283283,
+ "learning_rate": 9.56716064394958e-09,
+ "loss": 0.30971055030822753,
+ "num_input_tokens_seen": 1273701776,
+ "step": 29120,
+ "train_runtime": 225324.6547,
+ "train_tokens_per_second": 5652.74
+ },
+ {
+ "epoch": 1.961015351467816,
+ "grad_norm": 0.8807522971861178,
+ "learning_rate": 9.404369397432166e-09,
+ "loss": 0.287065315246582,
+ "num_input_tokens_seen": 1273920400,
+ "step": 29125,
+ "train_runtime": 225360.8369,
+ "train_tokens_per_second": 5652.803
+ },
+ {
+ "epoch": 1.961352006463776,
+ "grad_norm": 0.766183012473875,
+ "learning_rate": 9.242973755068663e-09,
+ "loss": 0.2945993900299072,
+ "num_input_tokens_seen": 1274143544,
+ "step": 29130,
+ "train_runtime": 225396.198,
+ "train_tokens_per_second": 5652.906
+ },
+ {
+ "epoch": 1.961688661459736,
+ "grad_norm": 0.8619210725642108,
+ "learning_rate": 9.082973761993518e-09,
+ "loss": 0.2969123363494873,
+ "num_input_tokens_seen": 1274356376,
+ "step": 29135,
+ "train_runtime": 225437.914,
+ "train_tokens_per_second": 5652.804
+ },
+ {
+ "epoch": 1.9620253164556962,
+ "grad_norm": 0.878826623744484,
+ "learning_rate": 8.924369462949834e-09,
+ "loss": 0.3293659448623657,
+ "num_input_tokens_seen": 1274559280,
+ "step": 29140,
+ "train_runtime": 225478.4148,
+ "train_tokens_per_second": 5652.689
+ },
+ {
+ "epoch": 1.9623619714516565,
+ "grad_norm": 0.795997521175958,
+ "learning_rate": 8.767160902291016e-09,
+ "loss": 0.2915841817855835,
+ "num_input_tokens_seen": 1274776792,
+ "step": 29145,
+ "train_runtime": 225514.841,
+ "train_tokens_per_second": 5652.74
+ },
+ {
+ "epoch": 1.9626986264476165,
+ "grad_norm": 0.7676235515949815,
+ "learning_rate": 8.61134812397968e-09,
+ "loss": 0.30088181495666505,
+ "num_input_tokens_seen": 1274996960,
+ "step": 29150,
+ "train_runtime": 225551.2935,
+ "train_tokens_per_second": 5652.803
+ },
+ {
+ "epoch": 1.9630352814435765,
+ "grad_norm": 0.7939420108805821,
+ "learning_rate": 8.45693117158819e-09,
+ "loss": 0.2771022796630859,
+ "num_input_tokens_seen": 1275203528,
+ "step": 29155,
+ "train_runtime": 225587.6966,
+ "train_tokens_per_second": 5652.806
+ },
+ {
+ "epoch": 1.9633719364395368,
+ "grad_norm": 0.8267377385515322,
+ "learning_rate": 8.303910088299228e-09,
+ "loss": 0.2890202045440674,
+ "num_input_tokens_seen": 1275420104,
+ "step": 29160,
+ "train_runtime": 225627.9039,
+ "train_tokens_per_second": 5652.759
+ },
+ {
+ "epoch": 1.963708591435497,
+ "grad_norm": 0.7872969059306067,
+ "learning_rate": 8.152284916904674e-09,
+ "loss": 0.297850513458252,
+ "num_input_tokens_seen": 1275648864,
+ "step": 29165,
+ "train_runtime": 225665.3758,
+ "train_tokens_per_second": 5652.834
+ },
+ {
+ "epoch": 1.964045246431457,
+ "grad_norm": 0.7526544096353636,
+ "learning_rate": 8.002055699805612e-09,
+ "loss": 0.2845508098602295,
+ "num_input_tokens_seen": 1275859328,
+ "step": 29170,
+ "train_runtime": 225711.406,
+ "train_tokens_per_second": 5652.613
+ },
+ {
+ "epoch": 1.964381901427417,
+ "grad_norm": 0.7307972307788215,
+ "learning_rate": 7.853222479013434e-09,
+ "loss": 0.28490164279937746,
+ "num_input_tokens_seen": 1276098304,
+ "step": 29175,
+ "train_runtime": 225747.3603,
+ "train_tokens_per_second": 5652.772
+ },
+ {
+ "epoch": 1.9647185564233773,
+ "grad_norm": 0.7126777979052846,
+ "learning_rate": 7.705785296148737e-09,
+ "loss": 0.29223477840423584,
+ "num_input_tokens_seen": 1276313944,
+ "step": 29180,
+ "train_runtime": 225782.9902,
+ "train_tokens_per_second": 5652.835
+ },
+ {
+ "epoch": 1.9650552114193376,
+ "grad_norm": 0.7780760910871221,
+ "learning_rate": 7.55974419244132e-09,
+ "loss": 0.3048643350601196,
+ "num_input_tokens_seen": 1276534976,
+ "step": 29185,
+ "train_runtime": 225815.2561,
+ "train_tokens_per_second": 5653.006
+ },
+ {
+ "epoch": 1.9653918664152976,
+ "grad_norm": 0.7124335281108779,
+ "learning_rate": 7.415099208732402e-09,
+ "loss": 0.2795270919799805,
+ "num_input_tokens_seen": 1276756464,
+ "step": 29190,
+ "train_runtime": 225850.916,
+ "train_tokens_per_second": 5653.094
+ },
+ {
+ "epoch": 1.9657285214112576,
+ "grad_norm": 0.7090900998979741,
+ "learning_rate": 7.2718503854707355e-09,
+ "loss": 0.3065217971801758,
+ "num_input_tokens_seen": 1276983008,
+ "step": 29195,
+ "train_runtime": 225891.8245,
+ "train_tokens_per_second": 5653.073
+ },
+ {
+ "epoch": 1.966065176407218,
+ "grad_norm": 0.8593525853790884,
+ "learning_rate": 7.129997762715391e-09,
+ "loss": 0.26297545433044434,
+ "num_input_tokens_seen": 1277205144,
+ "step": 29200,
+ "train_runtime": 225929.616,
+ "train_tokens_per_second": 5653.111
+ },
+ {
+ "epoch": 1.9664018314031781,
+ "grad_norm": 0.8188907230159181,
+ "learning_rate": 6.98954138013519e-09,
+ "loss": 0.2864915132522583,
+ "num_input_tokens_seen": 1277414784,
+ "step": 29205,
+ "train_runtime": 225967.4368,
+ "train_tokens_per_second": 5653.092
+ },
+ {
+ "epoch": 1.9667384863991382,
+ "grad_norm": 0.7346378968927514,
+ "learning_rate": 6.850481277008159e-09,
+ "loss": 0.3049713373184204,
+ "num_input_tokens_seen": 1277643008,
+ "step": 29210,
+ "train_runtime": 226014.3298,
+ "train_tokens_per_second": 5652.929
+ },
+ {
+ "epoch": 1.9670751413950982,
+ "grad_norm": 0.7674532896602834,
+ "learning_rate": 6.712817492222079e-09,
+ "loss": 0.32462234497070314,
+ "num_input_tokens_seen": 1277862488,
+ "step": 29215,
+ "train_runtime": 226054.1024,
+ "train_tokens_per_second": 5652.906
+ },
+ {
+ "epoch": 1.9674117963910585,
+ "grad_norm": 0.6823823423034494,
+ "learning_rate": 6.576550064273934e-09,
+ "loss": 0.29960381984710693,
+ "num_input_tokens_seen": 1278083144,
+ "step": 29220,
+ "train_runtime": 226087.3475,
+ "train_tokens_per_second": 5653.05
+ },
+ {
+ "epoch": 1.9677484513870187,
+ "grad_norm": 0.6768114788982357,
+ "learning_rate": 6.44167903127102e-09,
+ "loss": 0.30608859062194826,
+ "num_input_tokens_seen": 1278306920,
+ "step": 29225,
+ "train_runtime": 226128.6262,
+ "train_tokens_per_second": 5653.008
+ },
+ {
+ "epoch": 1.9680851063829787,
+ "grad_norm": 0.7867457093292909,
+ "learning_rate": 6.3082044309287215e-09,
+ "loss": 0.2798271656036377,
+ "num_input_tokens_seen": 1278530936,
+ "step": 29230,
+ "train_runtime": 226163.2382,
+ "train_tokens_per_second": 5653.133
+ },
+ {
+ "epoch": 1.9684217613789388,
+ "grad_norm": 0.7363218232718188,
+ "learning_rate": 6.176126300573848e-09,
+ "loss": 0.29613399505615234,
+ "num_input_tokens_seen": 1278747232,
+ "step": 29235,
+ "train_runtime": 226204.0451,
+ "train_tokens_per_second": 5653.07
+ },
+ {
+ "epoch": 1.968758416374899,
+ "grad_norm": 0.772729635299263,
+ "learning_rate": 6.045444677140744e-09,
+ "loss": 0.28275036811828613,
+ "num_input_tokens_seen": 1278973944,
+ "step": 29240,
+ "train_runtime": 226241.1384,
+ "train_tokens_per_second": 5653.145
+ },
+ {
+ "epoch": 1.9690950713708593,
+ "grad_norm": 0.7892151646300906,
+ "learning_rate": 5.91615959717462e-09,
+ "loss": 0.2926478385925293,
+ "num_input_tokens_seen": 1279195664,
+ "step": 29245,
+ "train_runtime": 226280.1718,
+ "train_tokens_per_second": 5653.15
+ },
+ {
+ "epoch": 1.9694317263668193,
+ "grad_norm": 0.6905193077369461,
+ "learning_rate": 5.78827109682989e-09,
+ "loss": 0.27014522552490233,
+ "num_input_tokens_seen": 1279416864,
+ "step": 29250,
+ "train_runtime": 226314.3102,
+ "train_tokens_per_second": 5653.274
+ },
+ {
+ "epoch": 1.9697683813627793,
+ "grad_norm": 0.7987578510711066,
+ "learning_rate": 5.661779211869056e-09,
+ "loss": 0.2848729848861694,
+ "num_input_tokens_seen": 1279631848,
+ "step": 29255,
+ "train_runtime": 226352.8714,
+ "train_tokens_per_second": 5653.261
+ },
+ {
+ "epoch": 1.9701050363587396,
+ "grad_norm": 0.7803335673063395,
+ "learning_rate": 5.536683977666601e-09,
+ "loss": 0.3105973243713379,
+ "num_input_tokens_seen": 1279865120,
+ "step": 29260,
+ "train_runtime": 226385.8852,
+ "train_tokens_per_second": 5653.467
+ },
+ {
+ "epoch": 1.9704416913546998,
+ "grad_norm": 0.6890074934726101,
+ "learning_rate": 5.412985429203988e-09,
+ "loss": 0.28957743644714357,
+ "num_input_tokens_seen": 1280078376,
+ "step": 29265,
+ "train_runtime": 226423.2957,
+ "train_tokens_per_second": 5653.475
+ },
+ {
+ "epoch": 1.9707783463506598,
+ "grad_norm": 0.7625841866074922,
+ "learning_rate": 5.290683601073543e-09,
+ "loss": 0.2784770965576172,
+ "num_input_tokens_seen": 1280304192,
+ "step": 29270,
+ "train_runtime": 226459.6491,
+ "train_tokens_per_second": 5653.564
+ },
+ {
+ "epoch": 1.9711150013466199,
+ "grad_norm": 0.7840668749162247,
+ "learning_rate": 5.169778527476243e-09,
+ "loss": 0.299863862991333,
+ "num_input_tokens_seen": 1280532272,
+ "step": 29275,
+ "train_runtime": 226498.8885,
+ "train_tokens_per_second": 5653.592
+ },
+ {
+ "epoch": 1.9714516563425801,
+ "grad_norm": 0.8729160388891067,
+ "learning_rate": 5.050270242222821e-09,
+ "loss": 0.31504406929016116,
+ "num_input_tokens_seen": 1280731456,
+ "step": 29280,
+ "train_runtime": 226536.9074,
+ "train_tokens_per_second": 5653.522
+ },
+ {
+ "epoch": 1.9717883113385404,
+ "grad_norm": 0.8134946456239271,
+ "learning_rate": 4.932158778733765e-09,
+ "loss": 0.2948625087738037,
+ "num_input_tokens_seen": 1280954184,
+ "step": 29285,
+ "train_runtime": 226580.571,
+ "train_tokens_per_second": 5653.416
+ },
+ {
+ "epoch": 1.9721249663345004,
+ "grad_norm": 0.9015410062046074,
+ "learning_rate": 4.8154441700387636e-09,
+ "loss": 0.3150008201599121,
+ "num_input_tokens_seen": 1281153920,
+ "step": 29290,
+ "train_runtime": 226613.4169,
+ "train_tokens_per_second": 5653.478
+ },
+ {
+ "epoch": 1.9724616213304604,
+ "grad_norm": 0.7835120290516451,
+ "learning_rate": 4.7001264487761545e-09,
+ "loss": 0.31816887855529785,
+ "num_input_tokens_seen": 1281377960,
+ "step": 29295,
+ "train_runtime": 226653.002,
+ "train_tokens_per_second": 5653.479
+ },
+ {
+ "epoch": 1.9727982763264207,
+ "grad_norm": 0.7177281483739487,
+ "learning_rate": 4.586205647194031e-09,
+ "loss": 0.3077888250350952,
+ "num_input_tokens_seen": 1281598760,
+ "step": 29300,
+ "train_runtime": 226691.7054,
+ "train_tokens_per_second": 5653.488
+ },
+ {
+ "epoch": 1.973134931322381,
+ "grad_norm": 0.8554091191816099,
+ "learning_rate": 4.4736817971507974e-09,
+ "loss": 0.3019340991973877,
+ "num_input_tokens_seen": 1281813592,
+ "step": 29305,
+ "train_runtime": 226737.7122,
+ "train_tokens_per_second": 5653.288
+ },
+ {
+ "epoch": 1.973471586318341,
+ "grad_norm": 0.7693611386783255,
+ "learning_rate": 4.362554930112395e-09,
+ "loss": 0.30186340808868406,
+ "num_input_tokens_seen": 1282029480,
+ "step": 29310,
+ "train_runtime": 226775.6925,
+ "train_tokens_per_second": 5653.293
+ },
+ {
+ "epoch": 1.973808241314301,
+ "grad_norm": 0.7405776360693889,
+ "learning_rate": 4.252825077155631e-09,
+ "loss": 0.2891343593597412,
+ "num_input_tokens_seen": 1282250600,
+ "step": 29315,
+ "train_runtime": 226814.7389,
+ "train_tokens_per_second": 5653.295
+ },
+ {
+ "epoch": 1.9741448963102612,
+ "grad_norm": 0.7831694807451932,
+ "learning_rate": 4.1444922689665156e-09,
+ "loss": 0.29642043113708494,
+ "num_input_tokens_seen": 1282463216,
+ "step": 29320,
+ "train_runtime": 226855.4258,
+ "train_tokens_per_second": 5653.218
+ },
+ {
+ "epoch": 1.9744815513062215,
+ "grad_norm": 0.8733709826420818,
+ "learning_rate": 4.037556535839149e-09,
+ "loss": 0.2692773103713989,
+ "num_input_tokens_seen": 1282685880,
+ "step": 29325,
+ "train_runtime": 226887.8468,
+ "train_tokens_per_second": 5653.392
+ },
+ {
+ "epoch": 1.9748182063021815,
+ "grad_norm": 0.7090978050641338,
+ "learning_rate": 3.932017907677943e-09,
+ "loss": 0.28195557594299314,
+ "num_input_tokens_seen": 1282915496,
+ "step": 29330,
+ "train_runtime": 226931.5302,
+ "train_tokens_per_second": 5653.315
+ },
+ {
+ "epoch": 1.9751548612981416,
+ "grad_norm": 0.7330650828079108,
+ "learning_rate": 3.827876413997067e-09,
+ "loss": 0.310672664642334,
+ "num_input_tokens_seen": 1283112440,
+ "step": 29335,
+ "train_runtime": 226962.5074,
+ "train_tokens_per_second": 5653.411
+ },
+ {
+ "epoch": 1.9754915162941018,
+ "grad_norm": 0.8645271417610324,
+ "learning_rate": 3.725132083918781e-09,
+ "loss": 0.2937079668045044,
+ "num_input_tokens_seen": 1283319208,
+ "step": 29340,
+ "train_runtime": 227004.0095,
+ "train_tokens_per_second": 5653.289
+ },
+ {
+ "epoch": 1.975828171290062,
+ "grad_norm": 0.7553494839592546,
+ "learning_rate": 3.623784946175102e-09,
+ "loss": 0.3103309631347656,
+ "num_input_tokens_seen": 1283519312,
+ "step": 29345,
+ "train_runtime": 227044.1149,
+ "train_tokens_per_second": 5653.171
+ },
+ {
+ "epoch": 1.976164826286022,
+ "grad_norm": 0.7711706358606312,
+ "learning_rate": 3.5238350291083577e-09,
+ "loss": 0.28435301780700684,
+ "num_input_tokens_seen": 1283726744,
+ "step": 29350,
+ "train_runtime": 227085.1855,
+ "train_tokens_per_second": 5653.062
+ },
+ {
+ "epoch": 1.976501481281982,
+ "grad_norm": 0.7233862194158842,
+ "learning_rate": 3.4252823606678588e-09,
+ "loss": 0.31540963649749754,
+ "num_input_tokens_seen": 1283952928,
+ "step": 29355,
+ "train_runtime": 227118.6155,
+ "train_tokens_per_second": 5653.226
+ },
+ {
+ "epoch": 1.9768381362779424,
+ "grad_norm": 0.7946045825578855,
+ "learning_rate": 3.328126968414336e-09,
+ "loss": 0.3130497455596924,
+ "num_input_tokens_seen": 1284147936,
+ "step": 29360,
+ "train_runtime": 227154.485,
+ "train_tokens_per_second": 5653.192
+ },
+ {
+ "epoch": 1.9771747912739026,
+ "grad_norm": 0.7860291343433327,
+ "learning_rate": 3.232368879517167e-09,
+ "loss": 0.31641981601715086,
+ "num_input_tokens_seen": 1284366488,
+ "step": 29365,
+ "train_runtime": 227200.0074,
+ "train_tokens_per_second": 5653.021
+ },
+ {
+ "epoch": 1.9775114462698626,
+ "grad_norm": 0.7901349246218392,
+ "learning_rate": 3.1380081207543766e-09,
+ "loss": 0.27493464946746826,
+ "num_input_tokens_seen": 1284580288,
+ "step": 29370,
+ "train_runtime": 227236.8347,
+ "train_tokens_per_second": 5653.046
+ },
+ {
+ "epoch": 1.9778481012658227,
+ "grad_norm": 0.7901988601043518,
+ "learning_rate": 3.0450447185137454e-09,
+ "loss": 0.300215482711792,
+ "num_input_tokens_seen": 1284790176,
+ "step": 29375,
+ "train_runtime": 227276.5524,
+ "train_tokens_per_second": 5652.982
+ },
+ {
+ "epoch": 1.978184756261783,
+ "grad_norm": 0.7905431031827841,
+ "learning_rate": 2.953478698792256e-09,
+ "loss": 0.29309883117675783,
+ "num_input_tokens_seen": 1285019536,
+ "step": 29380,
+ "train_runtime": 227311.7967,
+ "train_tokens_per_second": 5653.114
+ },
+ {
+ "epoch": 1.9785214112577432,
+ "grad_norm": 0.8196057690436338,
+ "learning_rate": 2.8633100871960917e-09,
+ "loss": 0.28349852561950684,
+ "num_input_tokens_seen": 1285225968,
+ "step": 29385,
+ "train_runtime": 227354.1237,
+ "train_tokens_per_second": 5652.97
+ },
+ {
+ "epoch": 1.9788580662537032,
+ "grad_norm": 0.8115546785952105,
+ "learning_rate": 2.774538908940638e-09,
+ "loss": 0.3020082950592041,
+ "num_input_tokens_seen": 1285433856,
+ "step": 29390,
+ "train_runtime": 227391.991,
+ "train_tokens_per_second": 5652.943
+ },
+ {
+ "epoch": 1.9791947212496632,
+ "grad_norm": 0.8333323595696321,
+ "learning_rate": 2.687165188849927e-09,
+ "loss": 0.2933856248855591,
+ "num_input_tokens_seen": 1285662136,
+ "step": 29395,
+ "train_runtime": 227423.7759,
+ "train_tokens_per_second": 5653.156
+ },
+ {
+ "epoch": 1.9795313762456235,
+ "grad_norm": 0.7761984521860247,
+ "learning_rate": 2.6011889513588574e-09,
+ "loss": 0.29178383350372317,
+ "num_input_tokens_seen": 1285879472,
+ "step": 29400,
+ "train_runtime": 227461.1494,
+ "train_tokens_per_second": 5653.183
+ },
+ {
+ "epoch": 1.9798680312415837,
+ "grad_norm": 0.7709375069582487,
+ "learning_rate": 2.5166102205093077e-09,
+ "loss": 0.31009225845336913,
+ "num_input_tokens_seen": 1286104184,
+ "step": 29405,
+ "train_runtime": 227495.7175,
+ "train_tokens_per_second": 5653.312
+ },
+ {
+ "epoch": 1.9802046862375438,
+ "grad_norm": 0.7529670231526651,
+ "learning_rate": 2.4334290199540254e-09,
+ "loss": 0.2966364860534668,
+ "num_input_tokens_seen": 1286326480,
+ "step": 29410,
+ "train_runtime": 227533.9984,
+ "train_tokens_per_second": 5653.337
+ },
+ {
+ "epoch": 1.9805413412335038,
+ "grad_norm": 0.7423799860118283,
+ "learning_rate": 2.351645372953848e-09,
+ "loss": 0.27869043350219724,
+ "num_input_tokens_seen": 1286539544,
+ "step": 29415,
+ "train_runtime": 227574.621,
+ "train_tokens_per_second": 5653.265
+ },
+ {
+ "epoch": 1.980877996229464,
+ "grad_norm": 0.830934035934882,
+ "learning_rate": 2.2712593023804795e-09,
+ "loss": 0.292054557800293,
+ "num_input_tokens_seen": 1286749296,
+ "step": 29420,
+ "train_runtime": 227603.0505,
+ "train_tokens_per_second": 5653.48
+ },
+ {
+ "epoch": 1.9812146512254243,
+ "grad_norm": 0.776340332148109,
+ "learning_rate": 2.1922708307120512e-09,
+ "loss": 0.3023597240447998,
+ "num_input_tokens_seen": 1286966040,
+ "step": 29425,
+ "train_runtime": 227646.316,
+ "train_tokens_per_second": 5653.358
+ },
+ {
+ "epoch": 1.9815513062213843,
+ "grad_norm": 0.7541577609267867,
+ "learning_rate": 2.1146799800386698e-09,
+ "loss": 0.3239266872406006,
+ "num_input_tokens_seen": 1287186672,
+ "step": 29430,
+ "train_runtime": 227678.08,
+ "train_tokens_per_second": 5653.538
+ },
+ {
+ "epoch": 1.9818879612173443,
+ "grad_norm": 0.8192940550401582,
+ "learning_rate": 2.0384867720579794e-09,
+ "loss": 0.278280234336853,
+ "num_input_tokens_seen": 1287391672,
+ "step": 29435,
+ "train_runtime": 227717.0952,
+ "train_tokens_per_second": 5653.47
+ },
+ {
+ "epoch": 1.9822246162133046,
+ "grad_norm": 0.7527540217103025,
+ "learning_rate": 1.96369122807738e-09,
+ "loss": 0.30411410331726074,
+ "num_input_tokens_seen": 1287620424,
+ "step": 29440,
+ "train_runtime": 227752.8072,
+ "train_tokens_per_second": 5653.588
+ },
+ {
+ "epoch": 1.9825612712092648,
+ "grad_norm": 0.8768248029758434,
+ "learning_rate": 1.8902933690123637e-09,
+ "loss": 0.2814857721328735,
+ "num_input_tokens_seen": 1287839296,
+ "step": 29445,
+ "train_runtime": 227783.5419,
+ "train_tokens_per_second": 5653.786
+ },
+ {
+ "epoch": 1.9828979262052249,
+ "grad_norm": 0.7931820825672209,
+ "learning_rate": 1.8182932153892884e-09,
+ "loss": 0.3011634826660156,
+ "num_input_tokens_seen": 1288057400,
+ "step": 29450,
+ "train_runtime": 227821.3297,
+ "train_tokens_per_second": 5653.805
+ },
+ {
+ "epoch": 1.983234581201185,
+ "grad_norm": 0.8278654551866558,
+ "learning_rate": 1.7476907873426041e-09,
+ "loss": 0.29094853401184084,
+ "num_input_tokens_seen": 1288283336,
+ "step": 29455,
+ "train_runtime": 227858.4256,
+ "train_tokens_per_second": 5653.876
+ },
+ {
+ "epoch": 1.9835712361971451,
+ "grad_norm": 0.7009932914918141,
+ "learning_rate": 1.6784861046159617e-09,
+ "loss": 0.29619736671447755,
+ "num_input_tokens_seen": 1288507000,
+ "step": 29460,
+ "train_runtime": 227904.983,
+ "train_tokens_per_second": 5653.703
+ },
+ {
+ "epoch": 1.9839078911931054,
+ "grad_norm": 0.696357225779694,
+ "learning_rate": 1.610679186561659e-09,
+ "loss": 0.3063995599746704,
+ "num_input_tokens_seen": 1288727264,
+ "step": 29465,
+ "train_runtime": 227944.9622,
+ "train_tokens_per_second": 5653.677
+ },
+ {
+ "epoch": 1.9842445461890654,
+ "grad_norm": 0.6754486614043711,
+ "learning_rate": 1.5442700521428599e-09,
+ "loss": 0.2689430952072144,
+ "num_input_tokens_seen": 1288937368,
+ "step": 29470,
+ "train_runtime": 227983.4782,
+ "train_tokens_per_second": 5653.644
+ },
+ {
+ "epoch": 1.9845812011850255,
+ "grad_norm": 0.7807038119601082,
+ "learning_rate": 1.4792587199297103e-09,
+ "loss": 0.29396581649780273,
+ "num_input_tokens_seen": 1289161136,
+ "step": 29475,
+ "train_runtime": 228025.749,
+ "train_tokens_per_second": 5653.577
+ },
+ {
+ "epoch": 1.9849178561809857,
+ "grad_norm": 0.8625253120569166,
+ "learning_rate": 1.4156452081026672e-09,
+ "loss": 0.310637903213501,
+ "num_input_tokens_seen": 1289388008,
+ "step": 29480,
+ "train_runtime": 228066.1418,
+ "train_tokens_per_second": 5653.57
+ },
+ {
+ "epoch": 1.985254511176946,
+ "grad_norm": 0.8386731246211869,
+ "learning_rate": 1.3534295344513893e-09,
+ "loss": 0.29576449394226073,
+ "num_input_tokens_seen": 1289614784,
+ "step": 29485,
+ "train_runtime": 228110.4974,
+ "train_tokens_per_second": 5653.465
+ },
+ {
+ "epoch": 1.985591166172906,
+ "grad_norm": 0.8448203849362906,
+ "learning_rate": 1.2926117163741813e-09,
+ "loss": 0.2955544710159302,
+ "num_input_tokens_seen": 1289834936,
+ "step": 29490,
+ "train_runtime": 228149.3127,
+ "train_tokens_per_second": 5653.468
+ },
+ {
+ "epoch": 1.985927821168866,
+ "grad_norm": 0.7551993663361088,
+ "learning_rate": 1.233191770877995e-09,
+ "loss": 0.3009219169616699,
+ "num_input_tokens_seen": 1290054672,
+ "step": 29495,
+ "train_runtime": 228188.1429,
+ "train_tokens_per_second": 5653.469
+ },
+ {
+ "epoch": 1.9862644761648263,
+ "grad_norm": 0.7385225987005458,
+ "learning_rate": 1.1751697145800934e-09,
+ "loss": 0.2817996978759766,
+ "num_input_tokens_seen": 1290281064,
+ "step": 29500,
+ "train_runtime": 228224.0221,
+ "train_tokens_per_second": 5653.573
+ },
+ {
+ "epoch": 1.9866011311607865,
+ "grad_norm": 0.8551674733846094,
+ "learning_rate": 1.1185455637052756e-09,
+ "loss": 0.3367082357406616,
+ "num_input_tokens_seen": 1290509624,
+ "step": 29505,
+ "train_runtime": 228265.6636,
+ "train_tokens_per_second": 5653.542
+ },
+ {
+ "epoch": 1.9869377861567465,
+ "grad_norm": 0.8607951663242355,
+ "learning_rate": 1.063319334089763e-09,
+ "loss": 0.3116638660430908,
+ "num_input_tokens_seen": 1290724632,
+ "step": 29510,
+ "train_runtime": 228308.5459,
+ "train_tokens_per_second": 5653.422
+ },
+ {
+ "epoch": 1.9872744411527066,
+ "grad_norm": 0.7586696985053731,
+ "learning_rate": 1.0094910411762027e-09,
+ "loss": 0.2789448261260986,
+ "num_input_tokens_seen": 1290947232,
+ "step": 29515,
+ "train_runtime": 228353.3345,
+ "train_tokens_per_second": 5653.288
+ },
+ {
+ "epoch": 1.9876110961486668,
+ "grad_norm": 0.8786362198012819,
+ "learning_rate": 9.570607000175536e-10,
+ "loss": 0.31690254211425783,
+ "num_input_tokens_seen": 1291165048,
+ "step": 29520,
+ "train_runtime": 228396.0938,
+ "train_tokens_per_second": 5653.184
+ },
+ {
+ "epoch": 1.987947751144627,
+ "grad_norm": 0.867401433919585,
+ "learning_rate": 9.060283252765311e-10,
+ "loss": 0.29549479484558105,
+ "num_input_tokens_seen": 1291376632,
+ "step": 29525,
+ "train_runtime": 228437.0857,
+ "train_tokens_per_second": 5653.095
+ },
+ {
+ "epoch": 1.988284406140587,
+ "grad_norm": 0.7283183527191147,
+ "learning_rate": 8.563939312233871e-10,
+ "loss": 0.2834463357925415,
+ "num_input_tokens_seen": 1291592160,
+ "step": 29530,
+ "train_runtime": 228475.8857,
+ "train_tokens_per_second": 5653.079
+ },
+ {
+ "epoch": 1.9886210611365471,
+ "grad_norm": 0.780933918989674,
+ "learning_rate": 8.081575317386847e-10,
+ "loss": 0.29933602809906007,
+ "num_input_tokens_seen": 1291812360,
+ "step": 29535,
+ "train_runtime": 228512.8712,
+ "train_tokens_per_second": 5653.127
+ },
+ {
+ "epoch": 1.9889577161325074,
+ "grad_norm": 0.7275658389704259,
+ "learning_rate": 7.613191403116338e-10,
+ "loss": 0.30489234924316405,
+ "num_input_tokens_seen": 1292039232,
+ "step": 29540,
+ "train_runtime": 228545.6355,
+ "train_tokens_per_second": 5653.31
+ },
+ {
+ "epoch": 1.9892943711284676,
+ "grad_norm": 0.7546810071907879,
+ "learning_rate": 7.158787700400905e-10,
+ "loss": 0.3050061225891113,
+ "num_input_tokens_seen": 1292268968,
+ "step": 29545,
+ "train_runtime": 228588.7605,
+ "train_tokens_per_second": 5653.248
+ },
+ {
+ "epoch": 1.9896310261244277,
+ "grad_norm": 0.7400073186881632,
+ "learning_rate": 6.718364336316674e-10,
+ "loss": 0.2916685104370117,
+ "num_input_tokens_seen": 1292491832,
+ "step": 29550,
+ "train_runtime": 228627.3323,
+ "train_tokens_per_second": 5653.269
+ },
+ {
+ "epoch": 1.9899676811203877,
+ "grad_norm": 0.8064188223143873,
+ "learning_rate": 6.291921434026238e-10,
+ "loss": 0.29037137031555177,
+ "num_input_tokens_seen": 1292713384,
+ "step": 29555,
+ "train_runtime": 228664.2829,
+ "train_tokens_per_second": 5653.324
+ },
+ {
+ "epoch": 1.990304336116348,
+ "grad_norm": 0.6299629051575274,
+ "learning_rate": 5.879459112778651e-10,
+ "loss": 0.26992483139038087,
+ "num_input_tokens_seen": 1292916816,
+ "step": 29560,
+ "train_runtime": 228707.5074,
+ "train_tokens_per_second": 5653.145
+ },
+ {
+ "epoch": 1.9906409911123082,
+ "grad_norm": 0.7057723847440374,
+ "learning_rate": 5.480977487920536e-10,
+ "loss": 0.3113460302352905,
+ "num_input_tokens_seen": 1293152760,
+ "step": 29565,
+ "train_runtime": 228744.4402,
+ "train_tokens_per_second": 5653.264
+ },
+ {
+ "epoch": 1.9909776461082682,
+ "grad_norm": 0.6985242431575329,
+ "learning_rate": 5.096476670890526e-10,
+ "loss": 0.29065916538238523,
+ "num_input_tokens_seen": 1293370744,
+ "step": 29570,
+ "train_runtime": 228786.0909,
+ "train_tokens_per_second": 5653.188
+ },
+ {
+ "epoch": 1.9913143011042282,
+ "grad_norm": 0.8726876861778484,
+ "learning_rate": 4.725956769202622e-10,
+ "loss": 0.3043833255767822,
+ "num_input_tokens_seen": 1293580568,
+ "step": 29575,
+ "train_runtime": 228821.7096,
+ "train_tokens_per_second": 5653.225
+ },
+ {
+ "epoch": 1.9916509561001885,
+ "grad_norm": 0.7789301695690733,
+ "learning_rate": 4.369417886485039e-10,
+ "loss": 0.3031582832336426,
+ "num_input_tokens_seen": 1293788160,
+ "step": 29580,
+ "train_runtime": 228859.1238,
+ "train_tokens_per_second": 5653.208
+ },
+ {
+ "epoch": 1.9919876110961487,
+ "grad_norm": 0.7445218059693559,
+ "learning_rate": 4.026860122430254e-10,
+ "loss": 0.293507981300354,
+ "num_input_tokens_seen": 1294011512,
+ "step": 29585,
+ "train_runtime": 228891.545,
+ "train_tokens_per_second": 5653.383
+ },
+ {
+ "epoch": 1.9923242660921088,
+ "grad_norm": 0.776822247685104,
+ "learning_rate": 3.698283572839412e-10,
+ "loss": 0.29284181594848635,
+ "num_input_tokens_seen": 1294233288,
+ "step": 29590,
+ "train_runtime": 228930.2438,
+ "train_tokens_per_second": 5653.396
+ },
+ {
+ "epoch": 1.9926609210880688,
+ "grad_norm": 0.8637476385715652,
+ "learning_rate": 3.38368832959457e-10,
+ "loss": 0.2951065540313721,
+ "num_input_tokens_seen": 1294439056,
+ "step": 29595,
+ "train_runtime": 228966.3788,
+ "train_tokens_per_second": 5653.402
+ },
+ {
+ "epoch": 1.992997576084029,
+ "grad_norm": 0.8185926706066555,
+ "learning_rate": 3.0830744806753523e-10,
+ "loss": 0.30010027885437013,
+ "num_input_tokens_seen": 1294656608,
+ "step": 29600,
+ "train_runtime": 229002.6257,
+ "train_tokens_per_second": 5653.457
+ },
+ {
+ "epoch": 1.9933342310799893,
+ "grad_norm": 0.7968229979488,
+ "learning_rate": 2.7964421101478457e-10,
+ "loss": 0.29233975410461427,
+ "num_input_tokens_seen": 1294866112,
+ "step": 29605,
+ "train_runtime": 229033.3425,
+ "train_tokens_per_second": 5653.614
+ },
+ {
+ "epoch": 1.9936708860759493,
+ "grad_norm": 0.6760975642489492,
+ "learning_rate": 2.5237912981646016e-10,
+ "loss": 0.3044607162475586,
+ "num_input_tokens_seen": 1295082752,
+ "step": 29610,
+ "train_runtime": 229068.0198,
+ "train_tokens_per_second": 5653.704
+ },
+ {
+ "epoch": 1.9940075410719094,
+ "grad_norm": 0.8833991167375352,
+ "learning_rate": 2.265122120975738e-10,
+ "loss": 0.2928770542144775,
+ "num_input_tokens_seen": 1295294464,
+ "step": 29615,
+ "train_runtime": 229108.5875,
+ "train_tokens_per_second": 5653.627
+ },
+ {
+ "epoch": 1.9943441960678696,
+ "grad_norm": 0.7026921457755877,
+ "learning_rate": 2.020434650906733e-10,
+ "loss": 0.2951038360595703,
+ "num_input_tokens_seen": 1295518440,
+ "step": 29620,
+ "train_runtime": 229144.7225,
+ "train_tokens_per_second": 5653.713
+ },
+ {
+ "epoch": 1.9946808510638299,
+ "grad_norm": 0.735480223574441,
+ "learning_rate": 1.7897289563972854e-10,
+ "loss": 0.28854942321777344,
+ "num_input_tokens_seen": 1295752200,
+ "step": 29625,
+ "train_runtime": 229181.9465,
+ "train_tokens_per_second": 5653.814
+ },
+ {
+ "epoch": 1.99501750605979,
+ "grad_norm": 0.7785852910887519,
+ "learning_rate": 1.5730051019569037e-10,
+ "loss": 0.28860130310058596,
+ "num_input_tokens_seen": 1295966672,
+ "step": 29630,
+ "train_runtime": 229217.582,
+ "train_tokens_per_second": 5653.871
+ },
+ {
+ "epoch": 1.99535416105575,
+ "grad_norm": 0.7483503409930179,
+ "learning_rate": 1.3702631481926633e-10,
+ "loss": 0.29309055805206297,
+ "num_input_tokens_seen": 1296186448,
+ "step": 29635,
+ "train_runtime": 229253.3432,
+ "train_tokens_per_second": 5653.948
+ },
+ {
+ "epoch": 1.9956908160517102,
+ "grad_norm": 0.8075255441882612,
+ "learning_rate": 1.1815031517981023e-10,
+ "loss": 0.28788046836853026,
+ "num_input_tokens_seen": 1296387216,
+ "step": 29640,
+ "train_runtime": 229284.8545,
+ "train_tokens_per_second": 5654.046
+ },
+ {
+ "epoch": 1.9960274710476704,
+ "grad_norm": 0.7216194041021678,
+ "learning_rate": 1.0067251655643262e-10,
+ "loss": 0.28290114402770994,
+ "num_input_tokens_seen": 1296614056,
+ "step": 29645,
+ "train_runtime": 229320.1711,
+ "train_tokens_per_second": 5654.165
+ },
+ {
+ "epoch": 1.9963641260436304,
+ "grad_norm": 0.8745531907030398,
+ "learning_rate": 8.459292383633522e-11,
+ "loss": 0.2950653076171875,
+ "num_input_tokens_seen": 1296815776,
+ "step": 29650,
+ "train_runtime": 229363.6026,
+ "train_tokens_per_second": 5653.974
+ },
+ {
+ "epoch": 1.9967007810395905,
+ "grad_norm": 0.8156583509555599,
+ "learning_rate": 6.991154151592128e-11,
+ "loss": 0.2833098888397217,
+ "num_input_tokens_seen": 1297021184,
+ "step": 29655,
+ "train_runtime": 229403.9752,
+ "train_tokens_per_second": 5653.874
+ },
+ {
+ "epoch": 1.9970374360355507,
+ "grad_norm": 0.734921335121816,
+ "learning_rate": 5.662837370135066e-11,
+ "loss": 0.3091259956359863,
+ "num_input_tokens_seen": 1297247832,
+ "step": 29660,
+ "train_runtime": 229441.4684,
+ "train_tokens_per_second": 5653.938
+ },
+ {
+ "epoch": 1.997374091031511,
+ "grad_norm": 0.8453937499810635,
+ "learning_rate": 4.47434241074296e-11,
+ "loss": 0.3012515068054199,
+ "num_input_tokens_seen": 1297458968,
+ "step": 29665,
+ "train_runtime": 229478.4365,
+ "train_tokens_per_second": 5653.947
+ },
+ {
+ "epoch": 1.997710746027471,
+ "grad_norm": 0.8579563343326696,
+ "learning_rate": 3.4256696057055615e-11,
+ "loss": 0.3033904552459717,
+ "num_input_tokens_seen": 1297673200,
+ "step": 29670,
+ "train_runtime": 229522.3122,
+ "train_tokens_per_second": 5653.8
+ },
+ {
+ "epoch": 1.998047401023431,
+ "grad_norm": 0.804937600894727,
+ "learning_rate": 2.5168192482882824e-11,
+ "loss": 0.30843775272369384,
+ "num_input_tokens_seen": 1297896616,
+ "step": 29675,
+ "train_runtime": 229556.9513,
+ "train_tokens_per_second": 5653.92
+ },
+ {
+ "epoch": 1.9983840560193913,
+ "grad_norm": 0.8014462539211099,
+ "learning_rate": 1.747791592676684e-11,
+ "loss": 0.27484722137451173,
+ "num_input_tokens_seen": 1298137256,
+ "step": 29680,
+ "train_runtime": 229589.775,
+ "train_tokens_per_second": 5654.16
+ },
+ {
+ "epoch": 1.9987207110153515,
+ "grad_norm": 0.7667553211393875,
+ "learning_rate": 1.1185868539209665e-11,
+ "loss": 0.3029062032699585,
+ "num_input_tokens_seen": 1298347688,
+ "step": 29685,
+ "train_runtime": 229623.1373,
+ "train_tokens_per_second": 5654.255
+ },
+ {
+ "epoch": 1.9990573660113116,
+ "grad_norm": 0.7997172260231348,
+ "learning_rate": 6.29205207991479e-12,
+ "loss": 0.30252552032470703,
+ "num_input_tokens_seen": 1298551880,
+ "step": 29690,
+ "train_runtime": 229659.802,
+ "train_tokens_per_second": 5654.241
+ },
+ {
+ "epoch": 1.9993940210072716,
+ "grad_norm": 0.8058607018219679,
+ "learning_rate": 2.7964679166769815e-12,
+ "loss": 0.2877742528915405,
+ "num_input_tokens_seen": 1298772440,
+ "step": 29695,
+ "train_runtime": 229699.7546,
+ "train_tokens_per_second": 5654.218
+ },
+ {
+ "epoch": 1.9997306760032318,
+ "grad_norm": 0.897980038082396,
+ "learning_rate": 6.991170281578364e-13,
+ "loss": 0.26116390228271485,
+ "num_input_tokens_seen": 1298981184,
+ "step": 29700,
+ "train_runtime": 229737.0922,
+ "train_tokens_per_second": 5654.207
+ },
+ {
+ "epoch": 2.0,
+ "num_input_tokens_seen": 1299153424,
+ "step": 29704,
+ "total_flos": 2402581590032384.0,
+ "train_loss": 0.34685841939968143,
+ "train_runtime": 229818.4762,
+ "train_samples_per_second": 0.517,
+ "train_steps_per_second": 0.129
+ }
+ ],
+ "logging_steps": 5,
+ "max_steps": 29704,
+ "num_input_tokens_seen": 1299153424,
+ "num_train_epochs": 2,
+ "save_steps": 500,
+ "stateful_callbacks": {
+ "TrainerControl": {
+ "args": {
+ "should_epoch_stop": false,
+ "should_evaluate": false,
+ "should_log": false,
+ "should_save": true,
+ "should_training_stop": true
+ },
+ "attributes": {}
+ }
+ },
+ "total_flos": 2402581590032384.0,
+ "train_batch_size": 1,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/training_args.bin b/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..9f35ce99940d0d7f4628bd149b2a2e946f1e8ec6
--- /dev/null
+++ b/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c414b624070ab6682af9c82937e863fa17942c0203f8862756f767014d4fbaf7
+size 7761
diff --git a/training_loss.png b/training_loss.png
new file mode 100644
index 0000000000000000000000000000000000000000..f9d979085a11d4e04c0f0d03cf0a236364850d0a
Binary files /dev/null and b/training_loss.png differ